RHEUMATOID FACTOR ASSOCIATED WITH A SECRETORY COMPONENT IN RHEUMATOID ARTHRITIS

Transcription

1 British Journal of Rheumatology 1995;34: RHEUMATOID FACTOR ASSOCIATED WITH A SECRETORY COMPONENT IN RHEUMATOID ARTHRITIS C. JORGENSEN,*t M. MOYNIER,t C. BOLOGNA,*t P. YOUINOUJ and J. SANY*t * Department of Immuno- Rheumatology, Centre Gui-de-Chauliac, Montpellier Cedex 5, VNSERM U291, 99 rue Puech-Villa, Zolad, Montpellier Cedex and %Laboratory of Immunology, Brest University Medical School Hospital, F-2969 Brest Cedex, France SUMMARY The authors' objective was to study the serum secretory immunoglobulin A (S-IgA) concentration and the presence of rheumatoid factor (RF) complexed with a secretory component (SQ in rheumatoid arthritis (RA). Sixty-three RA patients were studied. There were 49 healthy subjects in the control group. The S-IgA concentration and the presence of IgA isotype RF were determined by ELISA in the serum. The presence of SC complexed to RF (SC-RF) was studied by a sandwich-type enzyme-linked immunosorbent assay with an antibody against the SC used to capture S-immunoglobulin, and associated anti-globulin activity was revealed with a peroxidase-conjugated human IgG Fc fragment. We observed a significant increase in S-IgA in RA (mean 76.8*gml± S.D.), as compared to controls (mean 13.6igml± 11.9 S.D.) (P <.1). Forty-one per cent of RA patients presented a S-IgA concentration above the upper threshold, but we did not observe any association with disease activity. S-IgA concentration was correlated with the presence of IgA-RF. Twenty-seven RA patients had a positive SC-RF versus one in the control group (P <.1). The presence of SC-RF was associated with an increased S-IgA concentration (P <.1), and the presence of RF-IgA (P <.2). However, no association with disease activity was noted. Our study showed that serum S-IgA was increased in RA, and that part of the RF were complexed with SC. These results suggest contribution of mucosal lymphocytes in the pathogenesis of RA. KEY WORDS: Secretory immunoglobulin, Rheumatoid factor, Rheumatoid arthritis, Mucosa-associated lymphoid tissue. RHEUMATOID factor immunoglobulin A (RF-IgA) is present in 71 % of rheumatoid arthritis (RA) patients [1]. In early RA, the presence of RF-IgA has been associated with active disease and erosive arthritis with poor prognosis [2]. Moreover, RF-IgA is present in 86% of RA with extra-articular features [3]. A high serum IgA concentration has been reported in 23% of patients with RA [1]. IgA may be associated with an a-antitrypsin and this complex is correlated with the development of erosions in early RA [4]. We found RA to be associated with a serum IgA concentration above 5 g1, characterized by an increase of distal interphalangeal joints, sacro-iliitis and recurrent haematuria, but disease activity was not correlated with IgA concentration [5]. The origin of increased IgA in RA is not known, but a persistently high serum IgA concentration is considered to be a sign of chronic activation of mucosal-associated lymphoid tissue (MALT) [6]. However, it was suggested that in ankylosing spondylitis (AS), most serum IgA is derived from the bone marrow and belongs to the IgA, subclass [7]. To study the origin of serum IgA and RF in RA, we determined the presence of secretory IgA and RF bound to secretory component (SC). Secretory immunoglobulin A (S-IgA) is the predominant immunoglobulin in external secretions and plays a role both in defence against pathogens and in Submitted 12 September 1994; revised version accepted 23 November Correspondence to: C. Jorgensen, Department of Immune- Rheumatology, Centre Gui-de-Chauliac, Montpellier Cedex 5, France. preventing contact with environmental allergens. IgA is produced locally in a polymeric form by plasma cells in the musosal tissue, associated with the J chain. Epithelial cells in mucosal tissue synthesize the secretory components, a 7 kd protein homologous to the variable domain of the L chain in immunoglobulins [8]. SC is covalently bound to the Fc portion of IgA while internalized within epithelial cells [9]. The complex is transported to the apical pole of the cell and released in the lumen where SC stabilizes the quaternary structure of the IgA molecule and enhances its resistance to proteolysis. The serum concentration of S-IgA was found to be increased in inflammatory bowel diseases, and in AS, suggesting persistent stimulation of MALT in these two diseases [1,11]. In this study we determined the serum concentration of S-IgA, the presence of RF-IgA and describe an ELISA to detect the presence of immunoglobulin complexed with the secretory component having RF activity in RA. We found an association between RF-IgA, SC-RF and the serum concentration of S-IgA. The origin of SC-RF production is further discussed. MATERIALS AND METHODS Patients We studied 63 out-patients with RA according to the American College of Rheumatology 1987 revised criteria [12]. There were 49 women and 14 men, from 26 to 78 yr of age (mean 55 yr), with a mean disease duration of 8 yr (range 1-2 yr). The clinical and radiological features were assessed in each patient. The Downloaded from at Pennsylvania State University on March 6, British Society for Rheumatology

2 JORGENSEN ET AL.\ SECRETORY COMPONENT COMPLEXED WITH RF IN RA 237 treatments were: methotrexate (14 cases), hydroxychloroquine (eight cases), D penicillamine (five cases) and gold salts (12 cases). Thirty-nine were taking non-steroidal anti-inflammatory drugs (NSAIDs) at the time of the study. The disease was considered active if three of the following criteria were present: morning stiffness for longer than 45 min, more than six painful joints, more than four swollen joints and an erythrocyte sedimentation rate (ESR) above 28 mm in the first hour. Thirty-seven patients had active disease according to these criteria. Synovial fluids were available in 1 patients with active disease. The control group included 49 healthy subjects (they were 31 women, with a mean age of 52 yr) without evidence of gastrointestinal, liver or rheumatic diseases. Methods The Westergren ESR was determined in the first hour. Serum concentrations of IgA ( g1), IgG (8-18 g1), IgM ( g1) were measured by nephelometry (Behring). Determination of the IgA subclass RF. RF concentration was determined by a double-sandwich enzymelinked immunosorbent assay (ELISA). Wells were coated by incubating the plates overnight with affinity purified goat F(ab')2 against human IgA or IgM (Immunotech, Marseilles, France). Unbound plastic sites were saturated with 1% bovine serum albumin. After six washings with 25 il PBS tween.5%, wells were incubated with sera of patients and controls at two different dilutions (11 and 11) for 1 h. Plates were than incubated for 1 h with peroxidaseconjugated human IgG Fc fragments (Jackson, West Grove, PA, USA) and peroxydase activity revealed with OPD. Absorbances were read on a multiscan plate reader at 495 nm. Tests were considered positive if the upper-limit was equal to the mean (OD) of the controls plus two standard deviations. Determination of the S-IgA concentration. Wells were coated by incubation overnight with affinity purified goat monoclonal antibody specific to human secretory components (Sigma, St Quentin Fallavier, France). Unbound plastic sites were saturated with 1% bovine serujn-albumin. After six washings with 25il PBS tween.5%, wells were incubated with sera of patients and controls at two different dilutions (11 and 11). The bound S-IgA was detected after 1 h incubation with peroxydase-conjugated goat anti-human IgA antibodies (Immunotech, Marseilles, France) as previously described [1]. Absorbances were read on a multiscan plate reader at 495 nm. Concentrations were calculated from a standardized curve with a known human colostral IgA concentration (Sigma, St Quentin Fallavier, France). The serum of a patient with IgA, myeloma of known concentration was used as negative control. To exclude any non-specific binding the same experiments were conducted using a goat polyclonal anti-human serum albumin (Sigma, St Quentin Fallavier, France). Determination of secretory immunoglobulin with RF activity. Wells were coated overnight with affinity purified goat monoclonal antibody specific to human secretory components at 11 (Sigma, St Quentin Fallavier, France). Unbound plastic sites were saturated with 2 I of 1 % bovine serum albumin. After six washings with 25il PBS tween.5%, wells were incubated for 1 h with sera of patients and controls diluted at 11 and 11. The wells were washed six times before 1 h incubation with peroxydaseconjugated human IgG Fc fragment at 11 dilution (Jackson, West Grove, PA, USA). Absorbances were read on a multiscan plate reader at 495 nm. The test was considered positive if the upper limit was equal to the mean OD of the controls plus two standard deviations. To exclude any non-specific binding of the RF present in the serum of the patients with the antiserum against human SC the same experiments were conducted using a goat polyclonal antihuman serum albumin (Sigma, St Quentin Fallavier, France). Statistical analysis. Differences between RA patients with active disease and inactive disease, concerning mean concentrations of secretory IgA, SC-RF and IgA-RF expressed as OD, were compared using Student's (-test to determine the statistical significance. The number of RA patients with a SC-RF above the upper threshold (mean concentration of the controls plus two standard deviations) were compared to controls using the chi-squared test. The correlation coefficients between the concentration of S-IgA, the concentrations of RF-IgA or SC-RF expressed as OD were analysed using simple regression analysis. A "-value less than.5 was considered significant. RESULTS S-IgA serum levels in patients with RA Forty-one per cent (26 of 63) of RA patients had an increased serum S-IgA concentration. Figure 1 shows the increased serum S-IgA in RA patients (76.83 igml, s.d , range tgml) compared to controls (13.6igml, s.d. 11.9, range 2-63igml) (P<.1). The specificity of our ELISA to detect secretory IgA was determined by the fact that no monoclonal IgA from myeloma was captured. The mean S-IgA concentration in 37 RA patients with active disease was 94.4igml (s.d. 197, range zgml), but was not significantly different from the remaining 26 patients with inactive disease (55.6igml, s.d. 68.1). We found a significant association between the serum S-IgA concentration in RA and the total serum IgA concentration (P <.3). We did not detect any significant concentration of S-IgA in synovial fluid obtained from 1 patients with active disease. No difference was observed between patients taking NSAIDs (data not shown). IgA class RF Twenty-nine of the 63 RA patients (49%) has positive RF-IgA. We did not find any association between the presence of RF-IgA and the disease activity, the disease duration or the patient's age. Downloaded from at Pennsylvania State University on March 6, 216

3 238 BRITISH JOURNAL OF RHEUMATOLOGY VOL. 34 NO. 3 S-lgA jigml 3-, 25 2a Active RA RA in remission controls FIG. 1. Comparison of serum secretory immunoglobulin A (S-IgA) distribution in rheumatoid arthritis (RA) patients (n = 63), and controls (n - 49) by ELISA (upper threshold = mean controls +2 s.d. = 37.2 igml) (P <.1). RA with active disease, n - 37, RA in remission, n = 26. The serum concentrations of secretory IgA were associated with the presence of RF-IgA in RA patients (R 2 =.36, P <.1, Fig. 2). RF with a secretory component Twenty-seven RA patients (48%) had a significant concentration of SC-RF, as compared to one patient in the control group (P <.1). The concentration of S-RF expressed in OD correlated with that of RF-IgA, as shown in Fig. 3 (R 2 =.158, P<.2) and also with RF-IgM (P <.1). We observed a significant association between the concentration of S-RF expressed as OD with the S-IgA concentration (R 2 =.332, P <.1) as shown in Fig. 4. However, the serum concentration of SC-RF was not associated with disease activity or the CRP or the ESR in RA patients. ngml T DISCUSSION IgA is the main isotype produced locally by mucous tissue. Mucosal antibodies are functionally important to protect against viruses and bacteria in the respiratory and intestinal tracts. Moreover, IgA prevents the introduction of foreign antigens into the circulation, and excrete immune complexes present in the lamina propria [13]. IgA enters mucosal secretions via epithelial transcytosis mediated by the membrane secretory component. At the apical surface, proteolytic cleavage of the SC results in the release of S-IgA into the lumen, S-IgA is a complex of IgA and the soluble extracellular domain of the SC [14]. An increased serum IgA concentration is considered to result from increased MALT activity [6]. Moreover, increased S-IgA in the serum may occur in cases of hepatobiliary disorders or in gut inflammation [1, 15]. An Downloaded from at Pennsylvania State University on March 6, 216,1,2,3,4,5,6,7,8,9 1 FIG. 2. Simple regression analysis comparing the concentration of RF-IgA expressed as optical density obtained by ELISA with the S-IgA concentration (in figjml) from 63 RA patients (* 2 =.36, P <.2).

4 JORGENSEN ET AL.: SECRETORY COMPONENT COMPLEXED WITH RF IN RA 239 1,,9,8 j,6,5 u.,3,2,1 A I d Oo Ko >,2,4 ^<< ; v J^--*^ ^ " O,6,8 1 1,2 1,4 1,6 1,8 Secretory RF Fio. 3. Simple regression analysis comparing the concentration of RF-IgA with SC-RF expressed as optical density obtained by ELISA from 63 RA patients (R^ =.158, P <.2). association between disease activity in patients with AS and S-IgA serum concentrations indicates the role of mucosal stimulation in the pathogenesis of AS [11]. Mucosal tissue has been implicated in the pathogenesis of RA [16, 17]. Our results show an increased serum S-IgA concentration in RA patients, without any association with disease activity, and this might suggest an activation of mucous lymphocytes. The presence of RF-IgA early in the course of RA is reported to be predictive of early erosions and to be of poor prognosis [2]. Moreover, extra-articular features are more frequently observed in RA patients with RF-IgA and associated with increased mucosal symptoms [3, 18]. However, in our study, the presence of RF-IgA was not associated with disease activity, but < itory Ig 25, A u ( J ),2 FRIgA associated with the serum S-IgA concentration. To determine whether the IgA isotype RF could be S-IgA we developed a method to capture immunoglobulin complexed with a secretory component having activity against the immunoglobulin Fc fragment. Forty-eight per cent of our RA patients presented with FR complexed with SC. The specificity of the method was demonstrated by the negative results obtained with goat anti-serum against human bovine serum albumin in the same experimental conditions. Despite an association between SC-RF with RF-IgA, we were not able to demonstrate that SC-RF is an IgA isotype. SC binding to RF suggests that B cells producing RF are surrounded by epithelial cells which synthesize SC. secretory IgA o s» o *,4,6,8 1 1,2 1,4 Secretory RF Downloaded from at Pennsylvania State University on March 6, 216 Fio. 4. Simple regression analysis comparing the concentration of SC-RF expressed as optical density obtained by ELISA with the secretory IgA concentration (in igml) from 63 RA patients (R 2 =.332, P <.1).

5 24 BRITISH JOURNAL OF RHEUMATOLOGY VOL. 34 NO. 3 Since we were unable to demonstrate the presence of SC-IgA in the synovial fluid of RA patients, it seems unlikely that SC bound to RF is produced locally within the joints. However, we cannot exclude the possibility that the S-IgA has been destroyed by proteolytic enzymes within the SF. Non-specific binding of soluble SC to RF in the serum seems unlikely, since immunoglobulin binding to SC requires active transcytosis through the epithelial cells before enzymatic cleavage occurs at the apical pole of the cell. The detection of RF complexed with SC thus suggests local production of RF by mucosal lymphocytes; RF being complexed to SC produced by mucosal epithelial cells. Increased secretory IgA in the serum of patients with RA and the presence of S-RF might suggest a stimulation of mucous-associated lymphocytes by an environmental antigen. Further studies on the mucosal tissue and mucosal immune response are clearly needed to establish the role of mucosal lymphocytes in RA. REFERENCES 1. Pillemer SR, Reynolds WJ, Yoon SJ, Pcrera M, Ncwkirk M, Klein M. IgA related disorders in rheumatoid arthritis. Rheumatol 1987;14: Teitsson I, Withrington RH, Seifert MH, VaJdimarsson H. Prospective study of early rheumatoid arthritis. Ann Rheum Dis 1984;43: Gioud Paquet M, Auvinet M, Raffin T, Girard P. IgM-RF, IgA-RF, IgE-RF detected by ELISA in RA. Ann Rheum Dis 1987;46: Stanworth DR. Modulation of auto-immunity and disease. In: Maini RN, Berry H, eds. New York: Praeger Scientific, 1981: Jorgensen C, Anaya JM, Cognot C, Sany J. Rheumatoid arthritis associated with high levels of immunoglobulin A: clinical and biological characteristics. Clin Exp Rheumatol 1992;1: Peeters AJ, Daha MR, Smeets TJM, Breedveld C. Bone marrow IgA and IgA subclass synthesis in ankylosing spondylitis. J Rheumatol 1992; 19: Hocini H, Iscaki S, Benlahrache C, Vitalis L, Chevalier X, Larget-Piet B, Bonnet JP. Increased levels of serum IgA as IgA,, monomers in ankylosing spondylitis. Ann Rheum Dis 1992;51: Koopmann WJ, Griffin JA. B lymphocytes. In: Harris ED, ed. Textbook of rheumatology, Philadelphia: WB Saunders, Mestecky J, McGhee JR. Immunoglobulin A: molecular and cellular interactions involved in IgA biosynthesis and immune response. Ado Immunol 1987;4: Delacroix D, Vaerman JP. Reassessment of levels secretory IgA in pathological sera using a quantitative radio-immunoassay. Clin Exp Immunol 1981;43: Collado A, Sanmarti R, Serra C, Gallart T, Canete JD, Gratacos J, Vives J, Munoz-Gomez J. Serum levels of secretory IgA in ankylosing spondylitis. Scand J Rheumatol 1991;2: Amett FC, Edworthy SM, Bloch DA. The American Rheumatoid Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31: Mazanec MB, Nechud JG, Kaetzel CS, Lamm ME. A three tiered view of the role of IgA in mucosal defense. Immunol Today 1993;14: Solari R, Kraehenbuhl JP. The biosynthesis of secretory component and its role in the transepithelial transport of IgA dimer. Immunol Today 1985;6: Furuda Y, Imoto M, Hayakwa T. Serum levels of secretory IgA in liver disease. Ann J Gastroenterol 1985;8: Marcolongo R, Bayeli PF, Montagnani M. Gastrointestinal involvement in RA: a biopsy study. J Rheumatol 1979;6: Segal AW, Isenberg DA, Hajirrousou V, Tolfree S, Clare J, Smaith ML. Preliminary evidence for gut involvement in the pathogenesis of rheumatoid arthritis. Br J Rheumatol 1986;25: Ludviksson BK, Jonsson T, Erlendsson K,.Sigfusson A. Disease manifestations in patients with isolated elevation of IgA RF. Scand J Rheumatol 1992;21:l Zaphitopoulus GC. Rheumatoid arthritis and the gut. Br J Rheumatol 1986;25: Downloaded from at Pennsylvania State University on March 6, 216