Routine immunofluorescence detection of Ro/SS-A autoantibody using HEp-2 cells transfected with human 60 kDa Ro/SS-A

BACKGROUND: Ro/SS-A autoantibodies associated with systemic lupus erythematosus (SLE) and Sjögren syndrome may be missed during routine screening for antinuclear autoantibodies (ANA) by immunofluorescence using HEp-2 cells. AIMS: To investigate the use of HEp-2 cells transfected with human 60 kDa Ro/SS-A for routine detection of these antibodies. METHODS: 10,500 sera were screened at a dilution of 1:200 for Ro/SS-A antibodies, identified by intense immunofluorescence staining in 10-15% of hyperexpressing cells of either the nucleus and nucleolus combined or the nucleus alone. RESULTS: Ro/SS-A antibodies were identified in 160/2100 ANA positive sera (8%), of which seven were ANA negative (titre < 200) and 33 had weak ANA titres (200) in 85-90% of non-hyperexpressing "background" cells. Enzyme linked immunosorbent assay (ELISA) confirmed the presence of Ro/SS-A antibodies in 110 newly diagnosed Ro/SS-A positive sera. Of these, 50 reacted with Ro/SS-A, 51 with Ro/SS-A and La/SS-B, and nine with Ro/SS-A and other extractable nuclear antigen (ENA) specificities. Fifteen sera which did not show Ro/SS-A antibodies by immunofluorescence tested positive for Ro/SS-A by immunodiffusion, counter-immunoelectrophesis, or ELISA; of these, 14 had ANA titres > 200. Clinical data from 95 Ro/SS-A positive patients showed that 52% had SLE, 24% Sjögren syndrome, 8% rheumatoid arthritis, and 16% other diseases. CONCLUSIONS: (1) HEp-2 cells transfected with human 60 kDa Ro/SSA are useful for routine immunofluorescence detection for Ro/SS-A antibodies with a positive predictive value of 100%; (2) sera positive for Ro/SS-A antibodies by immunofluorescence should be tested for ENA by other methods because > 50% of these sera will have another ENA reactivity in addition to Ro/SS-A; (3) detection of Ro/SS-A by immunofluorescence may be missed in the presence of high titre ANAs; (4) with a detection sensitivity of 91%, a negative immunofluorescence results for Ro/SS-A does not exclude the presence of this autoantibody.     

A simple method for the biochemical purification of Ro/SS-A antigen

In the present study, Ro/SS-A antigen has been isolated from human spleen by a two-step procedure. In the first step most of the non-antigenic material was removed by means of ammonium sulphate precipitation and ion exchange chromatography. The final purification was obtained by passing the Ro/SS-A-containing fractions twice through a Mono Q ion exchange fast protein liquid chromatography (FPLC) column. The purified antigen showed identical immunoreactivity with crude material on CIE and was composed of two polypeptides with a molecular weight of approximately 60,000 and 55,000 respectively on SDS-PAGE, both reacting on Western blotting with a panel of anti-Ro/SS-A antisera. This system permits milligrams of highly purified antigen to be obtained from grams of human spleen.     

Analysis of B-cell epitopes of the Ro/SS-A autoantigen.

Autoantibodies to the Ro and La ribonucleoprotein antigens are found in several rheumatoid diseases. The important epitopes have been studied using synthetic peptides and recombinant antigens to understand how they arise and their implication in disease. Here, we analyse the results of epitope mapping studies of Ro60 and Ro52 autoantigens and focus on the major findings to date.     

A major autoepitope is present on the amino terminus of a human SS-A/Ro polypeptide

A human SS-A/Ro antigen is present on the polypeptide component of a particle composed of hyRNA and a 60 kD protein. We have now purified the Wil-2 cell 60 kilo dalton (kD) SS-A/Ro protein and determined its amino-terminal amino acid sequence. A synthetic peptide corresponding to residues 7 to 24 of this sequence (RoSP7-24) exhibited enzyme-linked immunosorbent assay (ELISA) binding activity with immunodiffusion-defined, monospecific anti-SS-A/Ro sera. In addition, ELISA binding of monospecific anti-SS-A/Ro sera to native SS-A/Ro antigen was partially inhibited (35%) by KLH-RoSP7-24. Sera from patients known to frequently produce precipitating anti-SS-A/Ro antibody (subacute cutaneous lupus erythematosus [SCLE], 56 patients; Sj?gren's syndrome [SS], 41 patients; mothers of infants with neonatal LE [NLE], 10 individuals; infants with congenital heart block [CHB], 5 patients) were tested for reactivity to RoSP7-24 in ELISA. Overall, 38% of SCLE sera, 36% of SS sera, 50% of maternal NLE sera and 20% of CHB infant sera had anti-RoSP7-24 binding levels greater than 2 standard deviations above the mean of that of normal individuals. Of the sera which had anti-SS-A/Ro detected by double immunodiffusion and/or counterimmunoelectrophoresis, 68% of SCLE patients, 71% of SS patients, 55% of NLE mothers and 20% of CHB infants had significantly elevated RoSP7-24 ELISA binding levels. These findings strongly suggest that a major autoepitope of native human SS-A/Ro resides on the amino terminal portion of the Wil-2 SS-A/Ro 60 kD polypeptide.(ABSTRACT TRUNCATED AT 250 WORDS)     

Antinuclear, anticytoplasmic, and anti-Sjogren's syndrome antigen A (SS-A/Ro) antibodies in female blood donors

A study of 2500 sera from female blood donors between the ages of 20 and 50 years was undertaken to determine the frequency of antinuclear (ANA), anticytoplasmic (ACA), and antimitochondrial (AMA) antibodies. When sera were tested by immunofluorescence (IF) on HEp-2 cells, 15.9 and 1.1% had ANA titers greater than 1/20 and 1/80, respectively. Analysis of these sera for autoantibody specificity showed: 1.5% antinucleolar, 1.0% anti-nuclear matrix, 0.2% anti-mitotic spindle apparatus, and 0.2% anti-primary biliary cirrhosis nuclear antigen. AMA titers of greater than 1/80 were seen in 2.5% and AMA titers greater than 1/160 were seen in 1.0%. None of the sera had anti-double stranded DNA. Testing of an additional 2500 sera for anti-Sjogren's Syndrome antigen A (anti-SS-A/Ro) revealed a frequency of 22/5000 (0.44%) with the highest frequency (0.72%) being in the 45-50 age group and a relatively high frequency (0.58%) in the 20-24 age group.     

Autoantibodies to the basal cells of squamous epithelium react with thymic epithelial cells

Sera from carriers of hepatitis B surface antigen (17 out of 21), which reacted in immunofluorescence with the basal cell layer (BCL) of squamous epithelium, were also shown to react with a thymic stellate epithelial cell (SEC) characterized by long, dendritic-like cytoplasmic processes. Absorption of the autoantibodies against BCL of squamous epithelium (BCL-Ab) with a thymic homogenate abolished the reactivity with BCL and SEC, demonstrating that the same antigenic determinant was recognized in both cells. In the human thymus, SEC were present both in the cortex and in the medulla. In the outer cortex SEC delineated the septal spaces. SEC were also stained by anti-HLA-DR (Ia) but not by antiactin monoclonal antibodies. The morphology and distribution of SEC were similar to those of the previously described thymic epithelial cells containing α-1 thymosin (K. Hirokawa, J. E. McClure, and A. L. Goldstein, Thymus4, 19, 1982). BCL-Ab were also found to react with five human epithelial thymomas. BCL-Ab seemed to be useful for further characterization of the thymic epithelial cells and for the immunodiagnosis of thymoma.     

A radioimmunoassay for antibodies to the Ro/SSA particle

A radioimmunoassay (RIA) for antibodies to the Ro/SSA particle is described using Iodogen to radiolabel the antigenic protein moiety of Ro/SSA with 125I. RIA methods utilizing either 33% saturated ammonium sulfate (NH4)2SO4 or 3.5% polyethylene glycol to separate bound from free antigen are comparable. Either method is similar to ELISA in sensitivity and specificity. This RIA provides a fluid phase assay for the Ro/SSA-anti-Ro/SSA system not previously available.     

Autoantibodies to human endogenous retrovirus-K are frequently detected in health and disease and react with multiple epitopes

A number of studies have found increased levels of antibodies to human endogenous retroviruses (HERVs) in autoimmune rheumatic diseases. It is not clear whether this immune response is driven by the HERV itself or by cross-reactions with an exogenous virus or an autoantigen. To address this question, we examined the antibody response to the Env protein of two closely related members of the HERV-K family, HERV-K10 and IDDMK1,222. By immunoblotting of recombinant proteins, antibodies were found in 32-47% of 84 sera from patients with autoimmune rheumatic disease, and 29% of 35 normal controls. Epitope mapping with overlapping 15mers identified multiple reactive peptides on both antigens, with one (GKTCPKEIPKGSKNT) containing immunodominant epitope(s). By ELISA, the median titre of antibody to this peptide was significantly increased in 39 patients with SLE compared to 39 healthy controls and 86 patients with other rheumatic diseases (P < 0.003). We have shown that there is a high frequency of IgG antibodies to HERV-K env sequences in human sera, both in health and autoimmune rheumatic disease, and that the response is to multiple epitopes. This supports the hypothesis that the autoimmune response to HERV-K is antigen-driven and may be an early stage in the chain of events that leads to tolerance breakdown to other autoantigens.     

Autoantibodies in SLE but not in scleroderma react with protein-stripped nucleosomes

Autoantibodies against nucleosomes (ANuA) are known to be sensitive markers for systemic lupus erythematosus (SLE), but their clinical relevance seemed to be limited because sera from patients with progressive systemic sclerosis (PSS) also showed positive reactions with conventional ANuA ELISA test systems (anti-Nu1 ELISA). It was generally assumed thatANuA were associated with both diseases. Using discontinuous sucrose gradient centrifugation to generate pure nucleosomes, we discovered by chance that at the 30-50% sucrose interface an antigen (Nu2) banded which was demonstrably free of non-histone components and histone H1. The two different nucleosome preparations, Nu1 and Nu2, were used in parallel as antigenic substrates in standardised ELISA tests to analyse sera from SLE (295 patients), PSS (119) and patients with other rheumatic diseases (101). With Nu1, 62% of the SLE and 52% of the PSS sera showed positive reactions. Two sera from patients suffering from Sj?gren's syndrome (SS) and one from polymyositis were also positive. Using the Nu2 preparation, 58% of the SLE but none of the PSS sera showed a positive reaction. One serum from a patient with SS was also positive. It could be shown that it was the PSS-specific autoantigen Scl-70 in the nucleosome preparation (Nu1) which contributed to the positive reactions of the PSS sera in conventional ANuA test systems, whereas in the Nu2 preparation no remaining Scl-70 was detectable. The present study definitely proved that ANuA are highly and specifically associated with SLE but not with PSS.     

Characterization of murine monoclonal antibodies against 60-kD Ro/SS-A and La/SS-B autoantigens

Small cytoplasmic ribonucleoproteins (scRNPs) are important autoantigens in patients with systemic lupus erythematosus and Sjögren's syndrome. MoAbs against these proteins were made by immunization of BALB/c mice with purified human recombinant 60-kD Ro/SS-A or 50-kD La/SS-B proteins. Five stable hybridoma cell lines were obtained, of which four secreted anti-Ro/SS-A antibodies (clones 1D8, 1D11, 2G10 and 6G8) and one produced anti-La/SS-B antibodies (clone 7F6). The MoAbs were further characterized using four different immunoassays: immunofluorescence, immunoblotting, RNA precipitation combined with Northern blotting, and recombinant protein precipitation. All lour MoAbs against Ro/SS-A recognized the native protein and one of them (2G10) recognized also intact scRNP particles. Interestingly, hY3-RNA was reproducibly not efficiently precipitated by MoAb 2G10. Epitope mapping using deletion mutants of the 60-kD Ro/SS-A antigen showed that MoAb ID8 recognized the C-terminal part of this protein, while 1D11 and 2G10 recognized distinct epitopes in the region between the RNP motif and the putative zinc finger domain. The epitopes recognized by these MoAbs lire highly conserved among species, and the epitope recognized by MoAb 2G10 may be identical to an autoepitope recognized by sera of patients. This is the first report describing the isolation and characterization of MoAbs of the IgG class against the 60-kD Ro/SS-A and La/SS-B autoantigens obtained by immunization with purified human recombinant proteins. These MoAbs can be of great use in studying the cellular processes in which scRNPs are involved, and may help to determine why these scRNPs become autoantigenic in autoimmune diseases.     

Heritable factors shape natural human IgM reactivity to Ro60/SS-A and may predispose for SLE-associated IgG anti-Ro and anti-La autoantibody production

Systemic lupus erythematosus (SLE) is characterized by various IgG autoreactivities, among which anti-Ro/SS-A is particularly pathology-associated and early detectable. SLE also shows significant familial aggregation, but genetic factors are not well understood and remain controversial for disease-associated IgG. Here we report that IgM anti-Ro showed a uniquely high degree of heritability in a study of SLE-affected families. Unlike IgM anti-La or anti-dsDNA, IgM anti-Ro was also significantly correlated to IgG anti-Ro among SLE patients, as well as to IgG anti-La and anti-dsDNA. We conclude that largely genetically determined, thus natural IgM anti-Ro-bearing precursor B-cells, may be an important factor for class switching and determinant spreading in early phases of SLE pathogenesis. Furthermore, we found unexpected sex differences in isotype/specificity correlations among SLE-unaffected relatives and control subjects, which could help understand the strong gender bias associated with SLE. We propose that the study of such correlation structures may reveal characteristic spreading pathways relevant for human SLE.     

Sera from patients with rheumatic diseases recognize different epitope regions on the 52-kD Ro/SS-A protein.

Patients suffering from systemic lupus erythematosus (SLE) or Sjögren's syndrome (SS) often contain autoantibodies directed to the Ro(SS-A) complex. In this study the antigenic determinants on two of the components of the Ro complex, i.e. the Ro60 and the Ro52 polypeptides, were investigated. Anti-Ro+ sera were selected by counter-immunoelectrophoresis. Depending on the detection method, 59-68% of the SLE patients produced anti-Ro but not anti-La antibody, while 72-81% of the SS patients produced both anti-Ro and anti-La antibody. Immunoprecipitation of recombinant Ro-proteins showed that 61 sera (87%) were reactive with both Ro proteins, seven sera with Ro60 only, one serum with Ro52 only, and one serum did not precipitate the proteins at all. The anti-Ro60 reactivity of human sera is strongly associated with the native form of Ro60, suggesting that conformational autoepitopes are an important feature of Ro60. In the case of Ro52, frequently the residues located between amino acids 216 and 292 were essential for reactivity with the antibodies. With 70% of the lupus sera tested this appeared to be the only region important for reactivity. The antibodies of SS patients generally recognized multiple B cell epitopes located between amino acids 55 and 292. The results of this study indicate that the antigenic determinants on Ro52 are different for autoantibodies produced by lupus patients compared with those of SS patients.     

Autoantibodies in human anti-Ro sera specifically recognize deproteinized hY5 Ro RNA.

We report the existence of a novel autoantibody specificity linked to anti-Ro antibodies. Sera from two patients with anti-Ro ribonucleoprotein (RNP) antibodies also contained antibodies that immunoprecipitated specifically either the deproteinized RNA component of the RohY5 RNP particle, or intact in vitro transcribed hY5 RNA. No serum recognized specifically the other hY RNAs. A mutant hY5 RNA with additional nucleotides (nt) at both extremities was not immunoprecipitated, possibly because of altered secondary structure. Following digestion of hY5 RNA with ribonuclease T1, the smallest immunoprecipitable RNA fragments were 27 and 31 nt long, and respectively mapped to the 5' and 3' ends of hY5 RNA, excluding the La-binding region. Base pairing between the 27 and 31 nt long fragments was required for recognition by antibodies. Our data indicate that the epitope bound by anti-hY5 RNA antibodies is conformational. We have previously reported that most anti-Ro sera contain a population of antibodies specific for the RohY5 RNP. Since antibodies to the deproteinized hY RNAs within anti-Ro sera are also restricted to anti-hY5 RNA, a direct role for the human-specific RohY5 particles in the immunization process leading to the production of anti-Ro antibodies is suggested.     

Autoantibodies to the Ro/SSA autoantigen are conformation dependent. II: Antibodies to the denatured form of 52 kD Ro/SSA are a cross reacting subset of antibodies to the native 60 kD Ro/SSA molecule.

The immune response to the Ro/SSA particles is conformation dependent. In sera with only anti-Ro/SSA precipitins, the autoantibodies to the 60 kD Ro/SSA are largely to the native 60 kD Ro/SSA while autoantibodies to the 52 kD Ro/SSA particle when present are exclusively to the denatured 52 kD Ro/SSA particle. Antibodies eluted from a recombinant 52 kD Ro/SSA fusion protein reacted in a sandwich ELISA which only measures antibody to native 60 kD Ro/SSA antigens and this reaction is largely inhibited by native homogeneous 60 kD Ro/SSA. In addition, antibody binding to the 52 kD Ro/SSA antigen in Western blot is also strongly inhibited by native 60 kD Ro/SSA. These experiment strongly suggest that reactivity of denatured 52 kD Ro/SSA antigen represents a cross reaction with autoantibodies directed to the native 60 kD Ro/SSA antigen. As a corollary of these experiments, data are presented that suggest the hY-RNAs are not associated with the 52 kD Ro/SSA protein but only with the 60 kD Ro/SSA protein. These data are consistent with the hypothesis that the autoanti-Ro/SSA response is driven by native 60 kD Ro/SSA and the immune response to denatured 52 kD Ro/SSA is largely a cross-reactive subset of the immune response to native 60 kD Ro/SSA.     

Human Anti-Ro Autoantibodies Bind Peptides Accessible to the Surface of the Native Ro Autoantigen

The relationship between fine specificity of linear epitopes and conformational determinants has been explored in a naturally arising human autoimmune response. In particular, the hypothesis tested is that the linear epitopes of the human Ro autoantigen are components of its conformational epitopes. Twenty groups among the 531 overlapping octapeptides 60kDa Ro are variably bound by anti-Ro precipitin positive lupus sera whose reactivity was easily distinguished from sera of normal controls and of anti-Ro precipitin negative lupus patients. The specific activities of anti-peptide antibodies and of anti-native Ro autoantibodies are similarly increased after affinity enrichment using native human Ro as ligand. Moreover, affinity-enriched anti-native Ro autoantibodies bind virtually the same 20 groups of epitopes recognized by whole anti-Ro positive sera. Two peptides (residues 274–290 and 480–494) from the defined 60 kDa Ro octapeptide epitopes have been prepared and used as ligands for affinity purification of peptide specific autoantibodies. The binding of both whole IgG and affinity-enriched peptide specific autoantibodies is inhibited by native Ro autoantigen. Thus, none of the available data can be construed to support the existence of cryptic linear epitopes in this system. Indeed, the data are only consistent with the conclusion that all of the anti-Ro octapeptide autoantibodies are part of the population of anti-native Ro autoantibodies in this naturally arising autoimmune response.     

Significance of the Ro antigen system

Summary Knowledge about antibodies to the Ro/SSA and La/SSB antigens has expanded greatly. Recognition of these antibodies was probably achieved 25 years ago but their macromolecular structure, clinical associations, and genetic relationships have come to light only in the past 7 years. It seems clear that these antibodies have a special place in the nosology of SLE and SS and that in certain instances (e.g., neonatal LE) the antibodies play a direct pathogenic role, while in other circumstances (e.g., vasculitis, nephritis, SS) tissue damage might result from immune complex deposition on vascular structures. Certainly, the latter problems will be active areas of investigation in the coming years. If the pace of recent progress continues, many of the questions raised in this review should soon have clear answers.     

False-positive sera do not react with human immunodeficiency virus (HIV) gag-encoded recombinant antigen

Ten sera from healthy blood donors positive by enzyme-linked immunoadsorbent assay (ELISA) were studied by immunoblot assay using natural and recombinant proteins. They interacted only with p17 or p24 proteins but were nonreactive with a recombinant protein (RP 50), which carries antigenic determinants to p17 and p24. Reactions were not blocked by preincubation of sera with genetically engineered p17 and p24 or purified viral p24, indicating that some new epitopes were formed during the Western blot procedure. Recombinant gag-encoded protein is required for confirmation of human immunodeficiency virus (HIV) seropositivity.