Article Text

Mass spectrometry-based identification of new anti-Ly and known antisynthetase autoantibodies
  1. Jean-Baptiste Vulsteke1,2,
  2. Rita Derua3,4,
  3. Sylvain Dubucquoi5,
  4. Frédéric Coutant6,7,
  5. Sebastien Sanges8,9,
  6. David Goncalves7,
  7. Greet Wuyts10,
  8. Petra De Haes11,12,
  9. Daniel Blockmans13,14,
  10. Wim A Wuyts15,16,
  11. Kristl G Claeys17,18,
  12. Ellen De Langhe19,20,
  13. Nicole Fabien7,
  14. Xavier Bossuyt10,21
  1. 1 Development and Regeneration, Skeletal Biology Engineering and Research Center, KU Leuven, Leuven, Belgium
  2. 2 Rheumatology, KU Leuven University Hospitals Leuven, Leuven, Belgium
  3. 3 Molecular and Cellular Medicine: Laboratory of Protein Phosphorylation and Proteomics, KU Leuven, Leuven, Belgium
  4. 4 SyBioMa, KU Leuven, Leuven, Belgium
  5. 5 Inserm, CHU Lille, U1286 - INFINITE - Institute for Translational Research In Inflammation, University of Lille, Lille, France
  6. 6 Eduard Herriot Hospital, Immunogenomics and Inflammation Research Team, University of Lyon, Lyon, France
  7. 7 Immunology, Lyon-Sud Hospital, Hospices Civils de Lyon, Pierre-Bénite, France
  8. 8 Inserm, CHU Lille, U1286 - INFINITE - Institute for Translational Research in Inflammation, University of Lille, Lille, France
  9. 9 Service de Médecine Interne et Immunologie Clinique, Centre de référence des maladies autoimmunes systémiques rares du Nord et Nord-Ouest de France (CeRAINO), CHU Lille, Lille, France
  10. 10 Microbiology, Immunology and Transplantation, Clinical and Diagnostic Immunology, KU Leuven, Leuven, Belgium
  11. 11 Microbiology, Immunology and Transplantation, KU Leuven University Hospitals Leuven, Leuven, Belgium
  12. 12 Dermatology, KU Leuven University Hospitals Leuven, Leuven, Belgium
  13. 13 Microbiology, Immunology and Transplantation, Laboratory for Clinical Infectious and Inflammatory Disorders, KU Leuven, Leuven, Belgium
  14. 14 General Internal Medicine, KU Leuven University Hospitals Leuven, Leuven, Belgium
  15. 15 Chronic Diseases and Metabolism, Laboratory of Respiratory Diseases and Thoracic Surgery, KU Leuven, Leuven, Belgium
  16. 16 Respiratory Diseases, KU Leuven University Hospitals Leuven, Leuven, Belgium
  17. 17 Neurosciences, Laboratory for Muscle Diseases and Neuropathies, KU Leuven, Leuven, Belgium
  18. 18 Neurology, European Reference Network on Rare Neuromuscular Diseases (ERN EURO-NMD), KU Leuven University Hospitals Leuven, Leuven, Belgium
  19. 19 Development and Regeneration, Skeletal Biology and Engineering Research Center, KU Leuven, Leuven, Belgium
  20. 20 Rheumatology, European Reference Network on Rare and Complex Connective Tissue and Musculoskeletal Diseases (ReCONNET), KU Leuven University Hospitals Leuven, Leuven, Belgium
  21. 21 Laboratory Medicine, KU Leuven University Hospitals Leuven, Leuven, Belgium
  1. Correspondence to Xavier Bossuyt, Microbiology, Immunology and Transplantation, Clinical and Diagnostic Immunology, KU Leuven, 3000 Leuven, Belgium; xavier.bossuyt{at}uzleuven.be

Abstract

Objectives To discover new and detect known antisynthetase autoantibodies (ASAs) through protein immunoprecipitation combined with gel-free liquid chromatography-tandem mass spectrometry (IP-MS).

Methods IP-MS was performed using sera of individuals showing features of antisynthetase syndrome (ASyS) without (n=5) and with (n=12) previously detected ASAs, and healthy controls (n=4). New candidate aminoacyl-tRNA-synthetase (ARS) autoantigens identified through unbiased IP-MS were confirmed by IP-western blot. A targeted IP-MS assay for various ASA specificities was developed and validated with sera of patients with known ASAs (n=16), disease controls (n=20) and healthy controls (n=25). The targeted IP-MS assay was applied in an additional cohort of patients with multiple ASyS features or isolated myositis without previously detected ASAs (n=26).

Results Autoantibodies to cytoplasmic cysteinyl-tRNA-synthetase (CARS1) were identified by IP-MS and confirmed by western blot as a new ASA specificity, named anti-Ly, in the serum of a patient with ASyS features. Rare ASAs, such as anti-OJ, anti-Zo and anti-KS, and common ASAs could also be identified by IP-MS. A targeted IP-MS approach for ASA detection was developed and validated. Application of this method in an additional cohort identified an additional patient with anti-OJ autoantibodies that were missed by line and dot immunoassays.

Discussion CARS1 is the dominant cognate ARS autoantigen of the newly discovered anti-Ly ASA specificity. Rare and common ASA specificities could be detected by both unbiased and targeted IP-MS. Unbiased and targeted IP-MS are promising methods for discovery and detection of autoantibodies, especially autoantibodies that target complex autoantigens.

  • Autoantibodies
  • Polymyositis
  • Autoimmune Diseases
  • Autoimmunity

Data availability statement

Data are available on reasonable request.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

WHAT IS ALREADY KNOWN ON THIS TOPIC

  • Detection of antisynthetase autoantibodies (ASAs) is crucial for diagnosis of the antisynthetase syndrome (ASyS), but there are issues with current detection methods and in some patients with a compatible phenotype no ASA can be found.

WHAT THIS STUDY ADDS

  • Through unbiased immunoprecipitation-mass spectrometry (IP-MS) the new anti-Ly autoantibody specificity, directed against cysteinyl-tRNA synthetase, was identified, in addition to known ASAs and other relevant autoantibodies.

  • Targeted IP-MS is a feasible and reliable detection method for ASAs, including the difficult-to-detect anti-OJ autoantibodies.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

  • The novel IP-MS approach could improve diagnosis of ASyS as it can confidently identify both new and known ASAs.

  • Unbiased and targeted IP-MS represent new discovery and detection methods that are especially suitable for complex autoantigens.

Introduction

Antisynthetase syndrome (ASyS) is a rare systemic autoimmune disease characterised by the presence of autoantibodies against cytoplasmic aminoacyl-tRNA-synthetases (ARSs) and a clinical syndrome of myositis, interstitial lung disease, arthritis, Raynaud’s phenomenon and specific cutaneous manifestations, with not all clinical features present in each person.1 Anti-histidyl-tRNA-synthetase autoantibodies, more widely known as anti-Jo-1 autoantibodies, are the most common antisynthetase autoantibodies (ASAs), but autoantibodies against most, but not all, other ARSs, including the multisynthetase complex (MSC) that contains nine ARS activities, have been described.2–4 Detection of ASAs is included in criteria for idiopathic inflammatory myopathy, interstitial pneumonia with autoimmune features and ASyS as a distinct entity, and is highly relevant for clinical management.5–8

Radiolabeled protein and RNA immunoprecipitation are considered to be the reference detection method, but currently solid-phase assays (SPAs) are more widely used.9 While SPAs perform well for most ASA specificities, for some ASAs sensitivity of SPA is low (eg, anti-OJ autoantibodies) and for the rarest ASAs SPAs are not rigorously validated or not available.4 In addition, there are patients with clinical features compatible with ASyS but in whom no known ASA can be identified. Immunoprecipitation combined with gel-free liquid chromatography-tandem mass spectrometry (IP-MS) could be a complementary discovery and detection method for ASAs. As technical proof of this concept, we here describe an unbiased and targeted IP-MS approach for identification of new and known ASAs.

Methods

Patients and patient material

The study outline is illustrated in figure 1. For unbiased IP-MS and development of targeted IP-MS, a serum aliquot of 5 individuals with multiple clinical features of ASyS (as included in the Connor’s criteria7) and a cytoplasmic pattern on the HEp-2 indirect immunofluorescence assay (HEp-2 IIF) without known ASAs, as defined by a negative result on a myositis dot immunoassay (Myositis 12 IgG, D-tek, Belgium), from the University Hospitals Leuven (Leuven, Belgium), Hospices Civils de Lyon, (Lyon, France) and University Hospitals of Lille (Lille, France) was retrieved. As controls, sera of 12 patients with a positive result on a myositis dot immunoassay (online supplemental table 1) for anti-Jo-1, anti-PL-7, anti-PL12 and anti-EJ autoantibodies were retrieved from the department of Laboratory Medicine at the University Hospitals Leuven (Leuven, Belgium), in addition to sera from four healthy controls.

Supplemental material

Figure 1

Outline study. Immunoprecipitation is performed by incubation of serum with magnetic protein A/G beads, subsequent cross-linking of the bound immunoglobulins with BS3, and incubation with HeLa cytoplasmic extract. In the first step (arrow 1), the eluted precipitated proteins are processed and then analysed by liquid chromatography-tandem mass spectrometry (MS) in data-acquisition mode. New and known aminoacyl-tRNA-synthetases (ARS) are identified by high total spectral counts, a semiquantitative measure for protein abundance. In the second step (arrow 2), proteotypic peptides are selected for each ARS and a targeted MS/MS assay is developed and evaluated with the samples from step 1. In the third step (arrow 3), a validation cohort (n=61) and an additional cohort (n=26) were evaluated. Created with Biorender.

For validation of the targeted IP-MS assay, sera of patients fulfilling the Connor’s criteria for ASyS with known anti-Jo-1, anti-PL-7, anti-PL-12 and anti-EJ autoantibodies, as detected on line immunoassay (Euroimmun Autoimmune Inflammatory Myopathies 16 Ag, Germany, n=16 in total) and controls (systemic lupus erythematosus (SLE), n=10, autoantibody-positive dermatomyositis (DM), n=10, healthy controls, n=25) were evaluated.

In addition, a new cohort of individuals with multiple features of ASyS (≥ 2 clinical features from Connors’ criteria or 1 clinical feature in combination with a cytoplasmic HEp-2 IIF pattern≥1/80, n=17) or biopsy-proven myositis without other ASyS features (n=9, all HEp-2 IIF negative) and without known ASAs or other relevant autoantibodies was evaluated. The clinical and serological features of these patients are summarised in online supplemental table 2.

Immunoprecipitation

Sera (1/20 in 5% Tris-buffered saline (TBS)) were cross-linked with 5 mM BS3 (Thermo Fisher Scientific, USA) to 50 µg protein A/G magnetic beads (Thermo Fisher Scientific, USA) in a 96-well plate and subsequently incubated with 200 µg HeLa cytoplasmic extract in 200 µL TBS overnight at 4°C. For mass spectrometry analysis, samples were eluted with 50 µL 0,2% formic acid after extensive washing and thereafter prepared for downstream analysis by in-solution digestion with trypsin and desalting (online supplemental methods). For western blot analysis, samples were eluted with 20 µL NuPAGE or Bolt sample buffer (Thermo Fisher Scientific, USA).

Unbiased mass spectrometry

Peptides were resuspended in 15 µL 0,1% formic acid/5% acetonitrile and subjected to high performance liquid chromatography (online supplemental methods) coupled to a Q Exactive Orbitrap mass spectrometer (Thermo Fisher Scientific, USA) operated in data-dependent acquisition mode. Proteins were identified by Mascot (Matrix Science, USA) using the UniProt Homo sapiens database (#entries: 194619). Peptide and protein identifications were validated with Scaffold V.4 (Proteome Software, USA). Relative quantification of proteins between samples was assessed by total spectral counting (the number of tandem MS spectra that can be matched to the protein) and ranking of proteins according to their abundance within a complex sample was estimated by the normalised exponentially modified protein abundance index (emPAI).10

Targeted mass spectrometry

For each target protein, two proteotypic peptides (peptides that uniquely identify a specific protein) from the unbiased IP-MS analysis, verified by a PeptideAtlas and Protein BLAST search, displaying minimal retention time overlap across all finally included peptides, were selected for targeted MS analysis (online supplemental methods). A parallel reaction monitoring assay based on the theoretical m/z ratios and retention times (as determined from the unbiased IP-MS experiment) was developed in Skyline (MacCoss Lab Software, USA) software and performed on the Q Exactive Orbitrap Mass Spectrometer. The data were analysed by Skyline software and matched to a spectral library built from the unbiased IP-MS experiment exported from Scaffold V.4. A minimum dotp value (a measure of matching with fragmentation spectra of the spectral library) of 0.7 was used as a threshold for correct peptide identification. Relative quantification of peptides between samples was assessed by summing the five most intense fragment peak areas (ie, sum peak area), normalised to the maximal peptide intensity observed across all samples. Data were visualised with the ggplot2 package in Rstudio and Microsoft Excel (Microsoft, USA).11 12

Results

Identification of new anti-Ly autoantibodies

Cytoplasmic cysteinyl-tRNA synthetase (CARS1) and valyl-tRNA synthetase (VARS1) were precipitated by serum #1, with a total spectral count of 51 and 10 respectively, but not by the other evaluated sera (table 1). We named this new ARS reactivity anti-Ly (as the corresponding serum was identified in CHU Lyon, France). CARS1 was much more abundant than VARS1 in the immunoprecipitate, as indicated by a more than tenfold higher emPAI value (5.68 vs 0.31). The strong precipitation of CARS1 by anti-Ly was also observed by western blot of the immunoprecipitate incubated with rabbit polyclonal antibodies, in contrast to the non-specific band for VARS1 (figure 2A). The Ly serum also showed strong reactivity with recombinant CARS1 on western blot, whereas reactivity with recombinant VARS1 on western blot was not specific (figure 2B). Finally, CARS1 and VARS1 did not clearly coimmunoprecipitate through IP with rabbit polyclonal anti-CARS1 or anti-VARS1 (figure 2C). Based on these results, we consider CARS1 the dominant cognate ARS autoantigen of anti-Ly autoantibodies. Besides CARS1 and VARS1, acetyl-coenzyme A carboxylase (ACACA), a key enzyme of de novo fatty acid biosynthesis, was also precipitated by the Ly serum (table 1).

Figure 2

anti-Ly autoantibodies. (A) Immunoprecipitation-western blot for CARS1 and VARS1, LC loading control (25 µg HeLa cytoplasmic extract), healthy control (HC). B) Western blot of 100 ng recombinant CARS1 and VARS1 protein (RP) and 25 µg HeLa cytoplasmic extract (CE) incubated with human serum (1/100 5% bovine serum albumin/Tris-buffered saline). (C) Immunoprecipitation (IP) with rabbit polyclonal anti-CARS1 and anti-VARS antibody (Proteintech, USA) as bait (1 µg/50 µg Pierce A/G magnetic beads); the same rabbit polyclonal antibodies were used for western blot (1/1000 and 1/500 5% non-fat milk/TBST, respectively). LC loading control (25 µg HeLa cytoplasmic extract), NC negative control, WB western blot. (D) HEp-2 indirect immunofluorescence of serum #1 at 1/1280 dilution, ×40 (Kallstad HEp-2, Bio-Rad Laboratories, USA). E–F) Left: reference spectra from spectral library for one proteotypic peptide for CARS1 (E) and VARS1 (F), right: chromatogram of 5 most intense product ions in targeted IP-MS for serum #1 for each peptide. IP-MS, immunoprecipitation-mass spectrometry.

Table 1

Total spectral count in unbiased IP-MS

The patient with anti-Ly autoantibodies had multiple ASyS features, including arthritis, interstitial lung disease, myalgia and mechanic’s hands (online supplemental table 3). The discovery of the new anti-Ly autoantibodies thus allowed a definite diagnosis as ASyS. HEp-2 IIF showed a cytoplasmic speckled pattern with nuclear dots (figure 2D), with the titre and cytoplasmic speckle density varying across two different HEp-2 assays (online supplemental figure 1).

Unbiased IP-MS of known ASAs

In the four other individuals with multiple ASyS features, previously undetected known ASAs were identified through unbiased IP-MS, thereby shifting their diagnosis to definite ASyS (online supplemental table 3). Serum #2 and #3 jointly precipitated all known components of the MSC, the target of anti-OJ autoantibodies. However, the total spectral counts of some individual MSC components varied strongly between the two sera. This difference was most marked for leucine-tRNA synthetase (LARS1), which was strongly immunoprecipitated by serum #2 but absent in the precipitate of serum #3, suggesting heterogeneity in the reactivity of anti-OJ autoantibodies. Serum #4 precipitated both the alpha and beta subunits (FARSA/FARSB) of the phenylalanine tRNA-synthetase complex/Zo autoantigen. Serum #5 precipitated asparagine-tRNA synthetase (NARS1), the KS autoantigen, but also a complex consisting of ATPase family protein 2 homolog (SPATA5), C1orf109 and spermatogenesis-associated protein 5-like protein 1 (SPATA5L1). This set of interacting proteins, which are involved in ribosome maturation, have hitherto not been described as autoantigen.13

The cognate ARS autoantigen of anti-Jo-1 (histidyl-tRNA synthetase, HARS1), anti-PL-7 (threonine-tRNA synthetase, TARS1) and anti-PL-12 (alanine-tRNA synthetase, AARS1) autoantibodies was detected in the immunoprecipitate of all sera with known ASA status and technically valid IP-MS results. While glycine-tRNA synthetase (GARS1) was precipitated in two out of three putative anti-EJ sera (as detected by dot immunoassay), one serum surprisingly precipitated the RuvBL1/2 dimer instead. Anti-RuvBL1/2 autoantibody positivity was confirmed through IP-western blot (online supplemental figure 2).14

Targeted IP-MS assay development

For each protein, the sum peak area of minimally one proteotypic peptide showed adequate discrimination between the different ASA specificities (figure 3, online supplemental table 4 for absolute values). Moreover, the sum peak area values were proportional to the total spectral count of the unbiased IP-MS analysis for corresponding sera. There was a high peak intensity and sum peak area for the selected CARS1 peptide in the patient with anti-Ly autoantibodies, not present in the other evaluated sera (figure 2E and figure 3). There was also a high peak intensity for the selected VARS1 peptide and a more than fivefold higher value for the VARS1 peptide in the anti-Ly serum than in the other evaluated samples (figure 2F and figure 3). The differences in normalised sum peak area values of MSC proteins between the two anti-OJ sera also corresponded to the differences between the two sera observed by unbiased IP-MS. This confirms the different relative abundances of the constituents of the precipitated complex.

Supplemental material

Figure 3

Targeted IP-MS evaluation of unbiased IP-MS cohort. relative quantification of proteotypic peptides between samples, assessed by the sum peak area of the five most intense ions, normalised to its maximal value across all samples (values ranging between 0 and 1, with 1 corresponding to the maximal peptide-specific value obtained in this experiment). U17 and U19 were excluded due to insufficient chromatogram quality. U16 was anti-EJ positive on dot immunoassay, but was found to be anti-RuvBL1/2-positive by IP-MS (online supplemental figure 2). IP-MS, immunoprecipitation-mass spectrometry.

Validation cohort for targeted IP-MS

In the validation group, there was full concordance of targeted IP-MS for 14/16 sera with previously detected ASAs by line immunoassay (figure 4A). All sera with anti-Jo-1, anti-PL-12 and anti-EJ were concordant. One serum with a positive line immunoassay result for anti-PL-7 and anti-OJ was only concordant for anti-PL-7. Two out of three other sera with anti-PL-7 autoantibodies by line immunoassay (Euroimmun, Germany) were concordant. The serum with discordant results for anti-PL-7 autoantibodies (V7) was also negative on a separate dot immunoassay (D-tek, Belgium). The range of results observed by IP-MS was larger than the range of the results obtained by line immunoassay (figure 4A). In a dilution series analysis of the sera with the highest targeted IP-MS values for each ASA, a saturation effect was visible for the dot immunoassay (figure 4B). These findings suggest that the wide range of the sum peak area observed in targeted IP-MS better reflects differences in autoantibody levels than results reported by line or dot immunoassay at the standard dilution.

Figure 4

Targeted IP-MS in validation cohort. (A) Relative quantification of proteotypic peptides of samples with previously detected anti-Jo-1. (n=4), anti-PL-7 (n=4), anti-PL-12 (n=4) and anti-EJ (n=4) autoantibodies from the Lyon cohort. (V1–V16). The highest normalised sum peak area values obtained in the disease controls. (n=20, max DC) and healthy controls (n=25, max HC) are shown as well. sum peak area values were normalised to the maximal peptide-specific value across all samples from this experiment (values ranging between 0 and 1, with 1 corresponding to the maximal peptide-specific value obtained). Asterisks indicate that no spectra were identified with dotp values ≥ 0.7. Arrowheads indicate sera that were selected for dilution series with dot immunoassay. Dia dot immunoassay, LIA line immunoassay. Both dia and LIA results were expressed in test-specific arbitrary units. (B) Dilution series of sera indicated with an arrowhead from panel a, as evaluated by targeted IP-MS (bars) and dot immunoassay (line) (myositis 12 IgG, D-tek, Belgium). Colours correspond to A. SD: standard dilution according to manufacturer. IP-MS, immunoprecipitation-mass spectrometry.

In 25 healthy controls no or very low levels of the target peptides were found by targeted IP-MS for all included ASAs (figure 4A, online supplemental figure 3). If the highest level of target peptide in healthy controls is considered as a threshold for positivity, none of the 10 controls with SLE was positive. One serum of an individual with myositis and anti-MDA5 autoantibodies (serum V54, figure 5) showed high values for QARS1 (which belongs to MSC subcomplex I) and DARS1 (which belongs to MSC subcomplex II) but not for IARS1 or KARS1. In none of the other control sera ASAs were identified (figure 4A, online supplemental figure 3).

Figure 5

Heterogeneity in IP-MS results of anti-OJ autoantibodies. relative quantification of proteotypic peptides for multisynthetase complex components for anti-OJ positive sera (D15, V54, U2 and U3). for each peptide, the highest normalised sum peak area value obtained in healthy controls (n=25) is shown as well. For each peptide, sum peak area values were normalised to its maximal value across all samples from this experiment (values ranging between 0 and 1, with 1 corresponding to the maximal peptide-specific value for each individual MSC component). Asterisks indicate that no spectra were identified with dotp values ≥ 0.7. Proteotypic peptide sequences are outlined in the inclusion list available in online supplemental methods. IP-MS, immunoprecipitation combined with liquid chromatography-tandem mass spectrometry; MSC, multisynthetase complex.

Additional cohort analysed by targeted IP-MS

In the additional cohort, anti-OJ autoantibodies were identified in 1 out of 17 patients with multiple ASyS features (serum D15, figure 5). These antibodies were not detected by conventional line and dot immunoassay. Serum D15 precipitated mainly QARS1. IARS1 was weakly precipitated and KARS1 not at all. No other ASAs, including anti-Ly, anti-Zo and anti-KS, were identified in the other patients with ASyS features. In the nine patients with isolated myositis and HEp-2 IIF negativity, no additional patients with ASAs were identified.

Discussion

In this report, we identified anti-Ly autoantibodies as a new ASA reactivity in a person with ASyS features through a new unbiased IP-MS approach. Anti-Ly autoantibodies precipitated CARS1 and, to a much lesser extent, VARS1 in an unbiased IP-MS analysis. CARS1 was confirmed as cognate ARS autoantigen on IP-WB, western blot with recombinant protein and targeted IP-MS whereas VARS1 (which has no significant sequence similarity to CARS1) could only be confirmed through targeted IP-MS. Based on these experiments, CARS1 is presumably the dominant cognate ARS autoantigen of anti-Ly autoantibodies. The consistent finding of VARS1 by sensitive unbiased and targeted IP-MS with this serum, but not with any other evaluated serum, could reflect low-grade coimmunoprecipitation of VARS1, as has been noted in some protein interaction databases.15 However, we could not identify strong coimmunoprecipitation through an IP-WB approach using commercial rabbit polyclonal antibodies. Relevantly, few instances of individuals with two distinct ASA specificities have been documented.16 Notably, during the review process of this paper, Muro et al reported antibodies against CARS1 and VARS1 in two different patients out of a cohort mainly consisting of DM, by screening with a custom ELISA.17 The patient with anti-CARS reactivity was diagnosed with DM but had multiple features compatible with ASyS. Our study, together with the study of Muro et al, thus establish CARS1 as definite and relevant ARS autoantigen in ASyS.

The additional precipitation of ACACA by the anti-Ly serum is another intriguing finding as there are no known direct interactions of ACACA with CARS1 or VARS1, though other ARS proteins are known to interact with non-ARS proteins outside the context of aminoacylation.18 Both CARS1 and ACACA, however, are essential in ferroptosis, a process which might play a role in intracellular autoantigen exposure.19–21 Similar to other ARSs, CARS1 itself can also be secreted, thereby exerting an immunomodulatory effect.22 23

Rare and common ARS autoantigens were also readily identified by unbiased IP-MS. As illustrated by two sera with anti-OJ autoantibodies, unbiased IP-MS can highlight heterogeneity in composition of complex autoantigens, like the MSC. For instance, the patient with anti-OJ autoantibodies identified in the additional cohort did not clearly precipitate IARS1 or KARS1, which are used as recombinant proteins in current anti-OJ assays.24 Information on heterogeneity in IP patterns might thus in turn inform development of high-throughput assays. Besides ARSs, the identification of RuvBL1/2 as autoantigen in a patient with presumed anti-EJ autoantibodies shows the added value of an unbiased approach and illustrates the specificity issues of dot immunoassays. Importantly, this added value of unbiased IP-MS is contingent on matching the right sera with the right protein source; in this case, a cytoplasmic extract for sera with a cytoplasmic HEp-2 IIF pattern.

While unbiased IP-MS is ideally suited for discovery purposes, implementation in routine clinical practice is difficult. A targeted IP-MS protocol, however, could have its place in reference centres, especially as an increasing number of clinical laboratories already possess a mass spectrometer for other targeted MS/MS assays. As proof of concept, our study shows that multi-targeted IP-MS is a viable approach to detect ASAs, including the new anti-Ly autoantibodies and difficult-to-detect anti-OJ autoantibodies. In the limited number of evaluated samples, already three false-positive and two false-negative ASAs, as detected by dot or line immunoassays, were identified. The false-positive and false-negative rate of dot/line immunoassays thus remains a point of concern. Further optimisation and validation of solid-phase assays, including comparison with IP-based methods, should be a priority. In the targeted IP-MS assay a wide range of values was noted for some ASAs, which most likely reflect real differences in autoantibody levels. It has been shown that autoantibody levels of anti-Jo-1 autoantibodies correlate with disease activity, but it is unclear if there is a ceiling effect to this correlation and whether this also applies to other ASAs.25

In an evaluation of an additional cohort, we could identify anti-OJ autoantibodies in one patient. With a mean of 2.2 clinical Connor’s criteria and presence of a cytoplasmic HEp-2 IIF pattern in 7/17 patients in the group with multiple ASyS features, and only isolated myositis as feature in the second group, the pretest probability for ASyS might have been too low for identification of additional cases. The identified patient with anti-OJ autoantibodies only had interstitial lung disease and a HEp-2 IIF cytoplasmic pattern, illustrating that ASyS, including anti-OJ-positive ASyS, can present with only a single clinical manifestation.4 A cytoplasmic HEp-2 IIF pattern in patients with any manifestation compatible with ASyS should thus always be further explored. Notwithstanding these considerations, the targeted IP-MS approach did not reveal other anti-Zo, anti-KS and anti-Ly autoantibodies in the evaluated sera. Overall, anti-Ly autoantibodies most likely represent an ultra-rare autoantibody specificity, similar to anti-Zo and anti-KS, whereas anti-OJ autoantibodies are likely underdetected.

Targeted IP-MS represents a new and promising method for autoantibody detection, but some technical hurdles still need to be addressed to translate this technique to the clinical laboratory. In the current proof-of-concept study, peptides were not absolutely quantified as this requires inclusion of isotopically labelled standards. To determine a threshold of positivity, an absolute quantification is desirable as there is intrinsic variability in IP-MS due to biological (eg, differences in individual protein concentrations between batches of cell extracts) and technical reasons. Furthermore, as mentioned above, the clinical relevance of the wide range of the IP-MS results merits further evaluation.

To conclude, through unbiased IP-MS CARS1 was identified as dominant cognate ARS of the newly identified anti-Ly autoantibodies in an individual with ASyS, thereby further expanding the spectrum of antisynthetase autoimmunity. Unbiased IP-MS also allowed reliable identification and characterisation of both rare and common known ASA specificities. A targeted IP-MS approach for ASAs is feasible and merits further evaluation as it has unique benefits as compared with the currently used detection methods. A high index of suspicion for both known or hitherto unidentified ASAs is warranted in individuals with ASyS features, especially in the presence of a cytoplasmic HEp-2 IIF pattern. Lastly, unbiased and targeted IP-MS could represent valuable discovery and detection methods for autoantibodies in other systemic autoimmune diseases, especially for those directed against a complex autoantigen.

Data availability statement

Data are available on reasonable request.

Ethics statements

Patient consent for publication

Ethics approval

This study involves human participants and was approved by University Hospitals Leuven (S60347/B32220071204) (Leuven) Hospices civils de Lyon (CT 69HCL21_0501) (Lyon)Comité de Protection des Personnes Sud-Est II (CPP #2019-87, RCB and EUDRACT # 2019-A01083-54) (Lille). An informed consent waiver was authorised by the ethical committees of the University Hospitals Leuven and Hospices civils de Lyon as this was a retrospective study with left-over material (secondary use) of samples submitted to the clinical laboratory. Patients were notified of the content and aim of the study. Informed consent was obtained for patients from Lille.

Acknowledgments

We thank Sebastien Carpentier and Kusay Arat from SyBioMa, the Proteomics Core Facility of the Biomedical Science Group, KU Leuven. We thank Doreen Dillaerts and Tom Dehaemers for technical assistance. We thank the clinicians from the Respiratory Medicine department, (Hôpital Louis Pradel, Bron, France) and the Biobank CRBSUD Hospices Civils de Lyon (no BRIF: BB-0033–00046).

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • EDL, NF and XB are joint senior authors.

  • Handling editor Josef S Smolen

  • Twitter @JBVulsteke

  • Contributors J-BV, RD and XB designed the study, analysed the data and drafted the manuscript. J-BV, RD and GW performed the experiments. SD, FC, SS, DG, PDH, DB, WAW, EDL and NF are involved in clinical patient management or clinical laboratory evaluation of autoantibodies. All authors critically revised the manuscript. J-BV and XB act as guarantor for the overall content.

  • Funding J-BV holds a Research Foundation – Flanders (FWO) SB Fellowship (1S62419N). This study was supported by the Research Fund KU Leuven (C3/20/042).

  • Competing interests SS received consulting fees from Novartis and Biotest, and support for attending meetings from Novartis, Sobi, Shire Takeda and Sanofi Genzyme. WAW has received research grants and/or consultancy fee’s from Roche, Boehringer-Ingelheim and Galapagos. XB was part of a scientific advisory committee for Werfen and Thermo Fisher Scientific, and received speaker’s fees from Werfen and Thermo Fisher Scientific. All abovementioned interests were outside the scope of the current work. J-BV, RD, SD, FC, DG, GW, PDH, DB, EDL and NF do not declare any competing interests.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.