Aminoacyl tRNA synthetases and their connection to autoimmune rheumatic disease

Abstract

Idiopathic inflammatory myopathies (IIM, collectively called myositis) are rare autoimmune rheumatic diseases characterized by chronic inflammation and associated with high mortality and morbidity (1, 2). The disease nomenclature tells us that IIM has traditionally been described as a group of diseases mainly affecting the skeletal muscle, with muscle weakness and weakened muscle endurance as the dominating clinical symptoms (3). The incidence of IIM in Sweden has been reported to be 11 per 1 million persons per year, with the highest incidence rate between 50 and 79 years of age (4). Gender demographics of IIM are similar to many other autoimmune diseases, with about two-thirds of the patients being female (5). Because of their low incidence, the mechanisms behind these diseases are largely unknown.

Autoantibodies have been identified in the majority of IIM patients (2). The autoantibodies present in IIM are divided into two groups; myositis-specific autoantibodies (MSAs) and myositis-associated autoantibodies (MAAs), with both groups of antibodies targeting both cytoplasmic and nuclear components (6). Autoantibodies targeting different aminoacyl transfer RNA synthetases (aaRS) are the hallmark of anti-synthetase syndrome (ASSD), one of the subgroups of IIM. There are seven anti-aaRS autoantibodies described in the literature associated with IIM and ASSD. Anti-Jo1 is the most common anti-aaRS antibody, targeting histidyl tRNA synthetase (HisRS), and is present in 20-30% of patients with IIM (7). In human cells, there are nineteen cytoplasmic aaRS, one for each amino acid. By only looking at the function and structure of the autoantigenic aaRS proteins, it is difficult to understand why these become autoantigens and not the other aaRS proteins. Research on aaRS has increased during the last decade due to their connection to disease and the novel findings regarding their noncanonical functions.

This thesis describes translational research combining clinical knowledge and patient material with advanced technological methods. It aimed at providing more profound knowledge about the connection between aminoacyl tRNA synthetase family members and rheumatic disease. More specifically, to support the scientific community with validated research tools and to study aaRS, and autoantibodies targeting these, in autoimmune rheumatic diseases, particularly IIM.

Paper I: Aminoacyl tRNA synthetases (aaRS) have been associated with non-canonical functions and several diseases. Understanding and determining their roles has been a challenge due to the lack of research tools to enable their study. In this study, well-validated research tools were produced and generated in the form of recombinant aaRS proteins and recombinant antibody fragments that target these. A pipeline was established to produce high-quality recombinant proteins and use them in phage display selections for scFv generation and validation. The recombinant proteins were produced in Escherichia coli (E. coli) with an Avi-tag to enable site-specific biotinylation and used as antigens in phage display selections to generate single-chain fragment variables (scFv). The scFv were validated in several immunoassays such as ELISA, multiplex bead array assay, and HTRF, as well as in cell-based assays such as immunoprecipitation followed by massspectrometry and immunofluorescence. The tools generated in this study are open-access and can be used to study human cytoplasmic aaRS and their connection to disease.

Paper II: In this study, we investigated anti-Jo1 autoantibodies in more detail by addressing the reactivity and affinity of anti-Jo1 against its antigen, HisRS, in sera and bronchoalveolar lavage fluid (BALF) in patients with IIM/ASSD. The associations between the reactivity profile and clinical data over time were investigated. Anti-Jo1 IgG reactivity against full-length HisRS (HisRS-FL), three HisRS domains (WHEP, catalytic domain (CD), and anticodon binding domain (ABD)), as well as the HisRS splice variant (SV), was evaluated in ELISA and western blot. Additionally, anti-Jo1 IgG and IgA reactivity were assessed in matched serum and BALF samples in ELISA, and the affinity of anti-Jo1 IgG from sera against HisRS-FL was measured using surface plasmon resonance. Finally, correlations at diagnosis and longitudinally between autoantibody reactivity and clinical data were evaluated. Anti-Jo1 IgG from serum and BALF bound HisRS-FL, SV, and WHEP with high reactivity at diagnosis. In addition, anti-Jo1 IgG displays a high affinity against HisRS-FL close to the time of diagnosis. The autoantibody levels against HisRS-FL were highest at the time of diagnosis in patients ever-developing interstitial lung disease (ILD) and arthritis, and anti-WHEP IgG reactivity in BALF correlated with poor pulmonary function.

Paper III: In this study, we hypothesized that there is a possibility of finding autoantibodies targeting other aaRS than the eight previously described. Samples and clinical data from 217 IIM patients and 156 age and gender-matched nonrheumatic population controls were included and tested against 118 antigens representing 33 proteins, including all nineteen cytoplasmic aaRS and other myositis-related proteins. In 72 IIM individuals (33%), we could identify reactivity against any of the included aaRS. In total, 16/19 aaRS were found with autoantigenic properties. Twelve IIM patients displayed reactivity against nine novel aaRS. Notably, some of these individuals were previously known as seronegative for MSAs. Besides the novel reactivities, we could identify anti-Zo, -KS, and - HA, positive individuals that were not previously tested in clinical settings, and we could confirm the reactivity to four of the most common aaRS (Jo1, PL12, PL7, and EJ). A low frequency of anti-aaRS autoantibodies was also detected in controls.

Paper IV: Here, we started by investigating the presence of HisRS in BALF and examining potential interaction partners. We did not find any new possible interaction partners but instead revealed that anti-Jo1 autoantibodies were present in different isotypes. Therefore, we continued to investigate the presence of different autoantibody isotypes in IIM. Analysis of a small cohort (n=5) revealed that the isotype distribution differed between sera and BALF taken at the same time point. To further investigate, sera from a larger cohort of IIM individuals (n=366) were screened for the presence of autoantibodies of isotypes IgG, IgA, and IgM, against Jo1 (HisRS), PL7 (ThrRS), PL12 (AlaRS), EJ (GlyRS), OJ (IleRS), and Ro52 (TRIM21) using a multiplex bead array assay. We found that for individuals with anti-Jo1 autoantibodies targeting HisRS-FL, IgG and IgM were the most common, followed by only having IgG, and about 11% displayed reactivity with all three isotypes. Most individuals with anti-PL12, -PL7, -EJ, and -OJ autoantibodies had only one isotype of autoantibodies present against the respective aaRS. However, the specific isotype differed between individuals, and many positive individuals did not harbor anti-aaRS IgG autoantibodies. Most anti-Ro52 positive individuals only had IgG anti-Ro52, in contrast to what was found for anti-aaRS autoantibodies.

To conclude, this thesis helped empower future research by generating new open-access research tools. Furthermore, we studied autoantibodies targeting aaRS and found that high levels and high-affinity anti-Jo1 autoantibodies towards HisRS can be found already early in the disease. In addition, we found that serum anti-Jo1 autoantibody levels correlated with ILD, and anti-Jo1 autoantibodies in BALF correlated with poor pulmonary function. These findings support the previously raised hypothesis that the lung might play a role in the pathogenic mechanism in anti-Jo1+ patients. We further suggest that most cytoplasmic aaRS may become autoantigenic. Autoantibodies against new aaRS may be found in the circulation of IIM individuals previously classified as seronegative. Adding these individuals to the seropositive group may have high clinical relevance. In addition, the frequency of the different autoantibody isotypes suggests that for anti-aaRS autoantibodies, it could be essential to investigate additional autoantibody isotypes, as some patients only harbor autoantibodies of IgM or IgA isotypes but not IgG. The clinical relevance of the different antibody isotypes still needs to be determined. The discoveries obtained in this thesis have helped improve the knowledge of aminoacyl tRNA synthetase autoantibodies and their connection to disease and open up for more studies in larger IIM cohorts and population controls.