Detection of anti-PL-12 autoantibodies by ELISA using a recombinant antigen; study of the immunoreactive region

Abstract

Autoantibodies to aminoacyl-tRNA synthetases are highly associated with myositis and detection is important in clinical diagnosis; however, current methods of screening limit its clinical utility. In the present study, alanyl-tRNA synthetase (PL-12) recombinant protein was obtained by immunological screening of a HeLa expression library and used in an ELISA with 22 anti-PL-12 sera, 200 autoimmune sera negative for PL-12 and 100 healthy individual sera. Sensitivity of the method was 95% (21/22) and specificity 100%. Mapping of the immunoreactive region was carried out using three anti-PL-12 sera and different recombinant protein-derived peptides. Results show that the same conformational epitope located within amino acids 730–951 of the PL-12 antigen outside the catalytic region was recognized by the three anti-PL-12 sera tested. We conclude that ELISA using recombinant protein is an effective and useful method for routine screening for anti-PL-12 autoantibodies.

INTRODUCTION

Autoantibodies to aminoacyl-tRNA synthetases (anti-synthetases) are found in 30% of sera from patients with myositis, being highly specific for this disorder [1] and strongly associated with complicating interstitial lung disease [2]. So far, autoantibodies recognizing five different synthetases have been described: anti-Jo-1 (histidine), anti-PL-12 (alanine), anti-PL-7 (threonine), anti-EJ (glycine) and anti-OJ (isoleucine) [3,4]. Autoantibody to Jo-1 antigen is the most common of these myositis-associated specificities, and although non-Jo-1 anti-synthetase autoantibodies are very uncommon, together they account for 5–8% of myositis patients [3].

Immunoprecipitation of nucleic acids and proteins is currently the most sensitive and specific method for detection of non-Jo-1 anti-synthetase autoantibodies, but these tests are unfortunately intricate and time-consuming, diminishing their clinical utility [4]. Immunoblotting is not reliable, since many sera appear to react exclusively with conformational epitopes. Anti-PL-7 can be identified by immunodiffusion, but anti-PL-12, anti-OJ and anti-EJ are not reliably detected by this method [5]. ELISA is highly sensitive for detection of anti-Jo-1 [6]. Since detection of the different anti-synthetase autoantibodies needs the use of multiple techniques and diverse antigen sources, an ELISA method using recombinant antigens is expected to be the most useful way of clinical detection of these antibodies. In this way, Jo-1 and EJ recombinant antigens reacted similarly to the natural protein [7,8].

The structural localization of epitopes recognized by the anti-synthetase antibodies may lead to a better understanding of the aetiopathogenesis of myositis. It has been suggested that the presence of these anti-synthetase antibodies in sera from patients suffering from myositis reflects a previous specific viral infection. This speculation is based on the interaction between some synthetases and the RNA of certain picornaviruses [9] which have been associated with this disorder in both epidemiological and animal model studies [10,11]. At present, only the epitopes recognized for the anti-Jo-1 and anti-EJ autoantibodies are known [7,12].

In the present study, we cloned the cDNA of PL-12 antigen, which was used to develop an ELISA capable of detecting anti-PL-12 autoantibodies and to study the immunoreactive region of PL-12 antigen.

PATIENTS AND METHODS

Human sera

RCD prototype serum was from the serum bank of the Hospital Universitario Virgen del Rocío, Sevilla. Barc-1 and Barc-2 were from the laboratory of Hospital de Sant Pau, Barcelona. These three sera showed anti-PL-12 antibodies by immunoprecipitation for nucleic acids and proteins. Sera from 20 other patients found to have anti-PL-12 activity by immunoprecipitation for nucleic acid and proteins (performed as in [13] and including some sera reported there) were taken from the serum bank of the Oklahoma Medical Research Foundation, Oklahoma City. Sera from 100 patients with systemic lupus erythematosus (SLE), 100 with other autoimmune diseases (27 scleroderma, 21 rheumatoid arthritis, 10 CREST, 16 mixed connective tissue diseases, 12 primary Sjögren's syndrome, and 14 polymyositis with anti-Jo-1), and 100 from healthy controls were collected from the serum bank in Sevilla.

cDNA library screening and sequence determination

A HeLa (D98/AH-2, HPRT-subclone) cDNA library cloned in λZAP II expression vector was obtained from Stratagene Inc. (La Jolla, CA). Clones were initially selected by immunological screening as described by Young & Davis [14]. The test was performed on duplicate filters and those bacteriophages reacting with antiserum RCD were subsequently purified to homogeneity. Before screening, serum was extensively adsorbed against bacterial proteins and wild-type λZAP II phage to eliminate background. Bound antibodies were revealed by using a biotin/streptavidin detection system as described previously [15].

Affinity purification of autoantibodies

Affinity-purified antibodies from λZAP II clones were obtained from confluent plates induced to produce recombinant protein with isopropyl-β-d-thiogalactopyranoside (IPTG)-impregnated nitrocellulose filters. After incubating the plates overnight, the filters were blocked and probed with primary antibody and washed as described previously [15]. Bound antibody was eluted from the filters with glycine buffer pH 2.5–2.8 and neutralized with Tris–HCl pH 9.5. Antibodies were concentrated with Centricon 30 microconcentrators (Amicon Corp., Danvers, MA) and used for immunofluorescence and immunoblotting analysis.

DNA subcloning and sequence determination

Purified λZAP II clones were subcloned in vivo into pBluescript SK(–) plasmids by using R408 helper phage as recommended by the manufacturer (Stratagene). DNA sequencing was carried out on an ABI PRISM 310 Genetic Analyser (Perkin-Elmer Corp., Foster City, CA). Computer analysis of nucleic acid and protein sequences was done using the University of Wisconsin Genetics Group Sequence Analysis Software Package [16].

Purification of PL-12 recombinant antigen

Whole cDNA of clone RCD-6 was ligated into pGEX-4T-3 expression vector (Pharmacia, Piscataway, NJ). DNA sequence analysis confirmed that the constructed cDNA clones had the predicted structure. Glutathione-S-transferase (GST) fusion proteins were purified directly from bacterial lysates using the Bulk GST Purification Modules according to the manufacturer's instructions (Pharmacia). Successful purification of GST fusion protein was confirmed by immunoblot analysis using goat anti-GST antiserum (Pharmacia).

ELISA

The optimum concentration of purified antigen and sera dilutions was determined as described previously [17]. Polystyrene microtitre plates (Nunc, Copenhagen, Denmark) were coated overnight at 4°C with 100 μl PL-12 recombinant antigen or GST protein (concentration 2.5 μg/ml) in TBS buffer (10 mm Tris–HCl pH 7.5, 150 mm NaCl). After washing with TBS–T (TBS + 0.05% Tween 20), the plates were post-coated with 3% non-fat dry milk in TBS–T (200 μl) for 1 h at 37°C and washed thoroughly with TBS–T. Sera were diluted in 3% non-fat dry milk in TBS–T at 1:400 dilution in duplicate and 100 μl of the dilution to be tested were added to the plates for 1 h at room temperature with agitation. This was followed, after similar washing, by the addition of 100 μl alkaline phosphatase-conjugated rabbit anti-human IgG diluted 1:1000 in TBS–T as recommended by the manufacturer (Dakopatts a/s, Glostrup, Denmark) for 1 h at room temperature in agitation. After five washes the alkaline-phosphatase activity was developed with 100 μl of p-nitrophenylphosphate (Sigma, St Louis, MO) at 1 mg/ml in 1 m diethanolamine buffer pH 9.8 with 1 mm MgCl₂ for 1 h at room temperature. The mean optical density (OD) for each duplicate was measured on a micro-ELISA reader (Bio-Tek, Winooski, VT) at 405 nm. Results were expressed in ELISA activity units based on a standard curve of serum RCD dilution versus OD, with 1 unit being the activity of RCD at 1:400 000 dilution. Inhibition studies of the ELISA reactivity were performed by incubating PL-12 recombinant antigen or GST protein overnight at 4°C with diluted sera. Further steps of the assay were as described above.

Mapping of the immunoreactive region

Fragments of clone RCD-6, prepared by treatment with restriction enzymes or with exonuclease III using the Erase-a-Base kit (Promega Biotech, Madison, WI), were ligated into pGEX-4T-3 expression vector (Pharmacia). DNA sequence analysis of the mutants confirmed that the constructed cDNA clones had the predicted structure. Fusion proteins were purified and ELISA analysis was performed as described above.

Other methods

The method of Lerner & Steitz [18] for immunoprecipitation for nucleic acids and proteins was used. Immunoblotting techniques were carried out by using an extract derived from HeLa and HEp-2 cell lines as described previously [15]. Protein concentrations were determined by the BioRad protein assay (BioRad Labs, München, Germany). HeLa and HEp-2 cells (ATCC CCL 2.2 and 23; American Type Culture Collection, Rockville, MD) were maintained in RPMI 1640 (Biochrom, Berlin, Germany) supplemented with 10% (v/v) fetal calf serum (FCS; Sera-Lab, Crawley Down, UK), 2 mml-glutamine, and gentamicin sulphate (5 μg/ml) in a 5% CO₂/95% air incubator.

RESULTS

Cloning and purification of PL-12 recombinant antigen

RCD serum was used for screening a HeLa λZAP II expression library. After screening 500 000 plaques, four positive signals (RCD-1, RCD-6, RCD-9 and RCD-10) were obtained, which retained specificity after a secondary and tertiary screening to 100% purity. An insert of 3200 bp was observed in clones RCD-6 and RCD-9, of 3000 bp in clone RCD-10 and of 1700 bp in clone RCD-1, showing identical sequences in the 3′ end after digestion with restriction enzymes. cDNA from clone RCD-6 and RCD-9 was sequenced, encoding a 968 amino acid polypeptide, 100% homologous to the cDNA encoding human alanyl-tRNA synthetase (PL-12 antigen) [19]. RCD-10 and RCD-1 were shown to be fragments of that protein.

Antibodies from the RCD serum were purified on nitrocellulose filters containing 50 000 phages expressing each recombinant protein. Once eluted and adjusted to a similar concentration, the affinity-purified antibodies were tested by indirect immunofluorescence (IIF), immunoblot analysis and immunoprecipitation. All antibodies eluted from these clones showed a diffuse cytoplasmic pattern similar to those described for anti-PL-12 by IIF [13]. Protein immunoprecipitation revealed a band of 110 kD. Immunoblot with cell extracts and immunoprecipitation of tRNA were negative.

cDNA of clone RCD-6 was cloned into pGEX expression vector and the protein product was purified by affinity chromatography. Immunoblot analysis with goat anti-GST antiserum showed a single band of 140 kD corresponding to GST-PL-12 fusion protein (110 kD of PL-12 antigen + 29 kD of GST protein) (data not shown).

ELISA for detection of anti-PL-12 antibodies

RCD serum was used to establish a standard curve to define arbitrary ELISA units, with 1 unit being the activity of RCD at 1:400 000 dilution. The normal (negative) range of this assay was determined with a series of 100 sera from healthy individuals, 100 from SLE and 100 from other autoimmune diseases (14 of them with anti-Jo-1). All samples were tested in two independent assays and the average coefficients of variation were < 5%. The mean and the highest activity units in each group are shown in Table 1. The upper limit of the negative range was established at 7000 U.

Table 1.

Anti-PL-12 ELISA with 300 anti-PL-12⁻ sera. The mean and the highest values were expressed in ELISA units

To study the sensitivity of this assay, sera from 22 patients with anti-PL-12 antibodies were tested (RCD prototype serum was not included). Only one serum had < 7000 units and was considered negative. In a previous study [13], this serum was found to have weaker immunoprecipitation activity than the nine other anti-PL-12 sera and did not inhibit alanyl-tRNA synthetase activity. Sixteen sera had > 100 000 units, three had between 50 000 and 100 000 units, and two had between 10 000 and 50 000 units (Fig. 1). All anti-PL-12 sera showed values in the same range as controls by ELISA with GST protein. In addition, ELISA activity from all anti-PL-12 sera was inhibited by preincubation with PL-12 antigen but not with GST protein.

Fig. 1.

Graph comparing anti-PL-12-positive and -negative sera in the anti-PL-12 ELISA. Results were converted to ELISA activity units to compare between plates. The horizontal line represents the 7000-U level. NL, Healthly individuals; SLE, systemic lupus erythematosus; other, scleroderma (n = 27), rheumatoid arthritis (n = 21), CREST (n = 10), mixed connective tissue disease (MCTD) (n = 16), Sjögren's syndrome (n = 12), and anti-Jo-1 sera (n = 14).

Mapping of the immunoreactive region

Sera from three Spanish patients (RCD, Barc-1 and Barc-2) were used to determine the immunoreactive region of PL-12 antigen. By immunoblotting assay using HeLa and HEp-2 cell extracts, the characteristic 110-kD band described for immunoprecipitation of PL-12 antigen did not appear in any case. Likewise, using PL-12 recombinant antigen these sera did not recognize any band by immunoblot.

Autoantibodies from RCD, Barc-1 and Barc-2 sera reacted by ELISA with a fragment containing amino acids 435–968 (OD > 1.4) but did not react with amino acids 1–435 (OD < 0.2). The subfragment containing amino acids 730–951 was equally reactive (OD > 1.4) but not the sequence of amino acids 435–730 (OD < 0.2). None of the three sera exhibited anti-GST protein reactivity when tested (Table 2). After denaturation by heat during 5 min at 95°C, PL-12 sera did not recognize the full recombinant protein or its fragments, but anti-GST serum retained its reactivity.

Table 2.

Reactivity of RCD, Barc-1 and Barc-2 sera (diluted 1:400) in ELISA with GST-PL-12 recombinant full-length protein (1–968 aa), different fragments and GST protein

DISCUSSION

In the present study, we have cloned the cDNA of PL-12 antigen by immunological screening of a HeLa λZAP II expression library with an anti-PL-12 serum. PL-12 recombinant antigen was used to develop an ELISA to detect and quantify anti-PL-12 autoantibodies and to study the immunological region recognized by three anti-PL-12 sera.

The specificity study for the ELISA was performed by testing sera from 100 healthy individuals and 200 autoimmune sera containing a great diversity of autoantibodies (SSA, SSB, DNA, RNP, Sm, Jo-1, etc.). SLE patients were selected because of their high titres and diversity of autoantibodies, and anti-Jo-1 sera in order to study the possible cross-reaction with other anti-synthetase autoantibodies.

For convenience, the cut-off point was established at 7000 U in order to exclude all 300 PL-12⁻ sera (assay specificity 100%). Using this cut-off, the sensitivity was 95% (21/22 anti-PL-12 sera) and the reactivity was very strong, having > 50 000 U in 90% [19/21] of anti-PL-12⁺ sera. The specific activity of PL-12 recombinant antigen was demonstrated by the lack of reactivity of anti-PL-12 sera against the GST protein and specific inhibition of anti-PL-12 ELISA activity by preincubation with fusion protein but not GST protein.

We studied the fine specificity of three anti-PL-12 sera. The results showed that all three sera recognized a conformational epitope. This assertion is based on two observations: first, all three sera were negative by immunoblot with cell extracts and with recombinant antigen (i.e. the band of 110 kD described for anti-PL-12 sera does not appear in any case); and second, reactivity of anti-PL-12 serum with the recombinant antigen in ELISA disappeared when the antigen was heated. These observations are similar to those obtained with anti-Jo-1 antisera: although most of these reacted well by immunoblot and with the recombinant protein, the major epitope of human Jo-1 is presumably conformationally determined, because overlapping synthetic hexapeptides of Jo-1 antigen did not react with anti-Jo-1 sera [20]. Additionally, there is indirect evidence of a major conformational epitope of the EJ antigen located in the C-terminal region, because no antigen subfragment reacted as strongly as the whole antigen [12]. Our results support the idea that the major epitopes recognized by anti-synthetase autoantibodies are conformational.

Anti-PL-12 sera contain separate populations of autoantibodies reacting with the synthetase and autoantibodies reacting with tRNA^ala [4]. Anti-tRNA^ala antibodies react with a selected set of tRNA^ala, all of them having the inosine-guanine-cytosine anticodon, but are negative against tRNA^ala having other alanine anticodons. The epitope recognized by anti-tRNA^ala antibodies has been located at a nine-base region that includes the anticodon [21]. In this study, fragments of PL-12 recombinant antigen were used to localize the position of the immunoreactive region(s) of the synthetase. The result of the epitope mapping demonstrates that the immunoreactive region is located within amino acids 730–951, outside the catalytic region (Fig. 2). The human catalytic domain of PL-12 extends from the N-terminus to amino acid 499 [19], a domain highly conserved between Esherichia coli and human and which is not recognized by the anti-PL-12 sera tested in this study.

Fig. 2.

Schematic drawing of alanyl-tRNA synthetase. Regions needed for adenylate formation, aminoacylation, and oligomerization and the class-defining motifs 1, 2, and 3 are shown [14]. The epitope is found in the C-terminal region, outside the catalytic region.

In conclusion, we verified the usefulness of ELISA using PL-12 recombinant protein as an effective method for screening for anti-PL-12 antibodies, as an alternative to immunoprecipitation, and we have demonstrated that the same immunoreactive region located within amino acids 730–951, outside the catalytic region, is recognized by three anti-PL-12 sera.