Downloaded from http://ard.bmj.com/ on June 4, 2017 - Published by group.bmj.com Annals of the Rheumatic Diseases 1994; 53: 535-539 535 Yersinia-specific antibodies in serum and synovial fluid in patients with yersinia triggered reactive arthritis Outi Maiki-Ikola, Riitta Lahesmaa, Jurgen Heesemann, Riitta Merilahti-Palo, Riitta Saario, Auli Toivanen, Kaisa Granfors National Public Health Institute, Department in Turku, Finland 0 Maki-Ikola* K Granfors* Turku University, Turku, Finland Department of Medical Abstract Objectives-To further evaluate the role of bacterial antigens in triggering inflammation in the joint in patients with reactive arthritis by studying local antibody synthesis in the joint. Methods-Yersinia-specific antibodies in paired serum and synovial fluid samples from 29 patients with yersinia triggered reactive arthritis were studied using an enzyme linked immunosorbent assay (ELISA), an inhibition ELISA with six monoclonal antibodies against lipopolysaccharide or released proteins of yersinia and immunoblotting. Antibodies of IgM, IgG and IgA classes, as well as antibodies of IgA subclasses and those containing secretory component were measured against the lipopolysaccharide and the sodium dodecyl sulphate extract of whole Yersinia enterocolitica 0:3 bacteria. Results-It was shown that yersiniaspecific antibodies, as well as antibodies against other microbial antigens (rubella, measles, Bordetella pertussis, tetanus toxoid and Candida albicans) in synovial fluid mirror those in serum by concentration, by specificity and by distribution in classes and subclasses. Conclusion-These results do not suggest any strong local antibody production, but indicate that the majority of yersinia antibodies in the synovial fluid are derived from the circulation. (Anni Rheum Dis 1994; 53: 535-539) Patients and methods Microbiology* R Lahesmaa Department of Medicine R Merilahti-Palo R Saario A Toivanen Insitute of Hygiene and Microbiology, University of Wurzburg, Germany J Heesemann Correspondence to: Dr Kaisa Granfors, National Public Health Institute, Department in Turku, SF-20520 Turku, Finland. Accepted for publication 29 April 1994 IgA has been shown to be higher in synovial fluid than in serum in patients with Reiter's disease and ReA to indicate local antibody production.6 Further, in two studies with small groups of patients with salmonella and Chlamydia trachomnatis triggered ReA, evidence for intra-articular production of specific antibodies was obtained.7 However, mechanisms leading to inflammation in synovium and the possible role for microbial antigens and antibodies in propagating these events is not understood. In the present work, to further evaluate the role of bacterial antigens in triggering inflammation in the joint, we studied local antibody synthesis by measuring yersiniaspecific antibody levels in paired samples of serum and synovial fluid from patients with yersinia triggered ReA using an enzyme linked immunosorbent assay (ELISA). Antibodies of IgM, IgG and IgA classes, as well as antibodies of IgA subclasses and those containing secretory component (sIgA) were measured against LPS and/or sodium dodecyl sulphate (SDS) extract of whole Yersinia enterocolitica 0:3 bacteria. Antibody levels were also measured by inhibition-ELISA against two of the plasmid encoded released proteins (RPs), which are proteins released into calcium deficient media by human enteropathogenetic strains of yersinia.9 The pattern of antigen specificity was further evaluated by inhibitionELISA with four monoclonal antibodies (MoAbs) directed against different parts of LPS, as well as by immunoblotting. PATIENTS AND SAMPLES Certain infections of the gastrointestinal and urogenital tract, such as those caused by yersiniae and salmonellae, are sometimes followed by development of reactive arthritis (ReA), especially in patients carrying HLAB27. In the development of ReA host-microbe interaction is important, but the exact pathogenic mechanisms of arthritis are largely unknown.' 2 Despite negative attempts to culture bacteria from synovial fluids, there is strong direct evidence for the presence of microbial components, especially lipopolysaccharide (LPS), in the affected joints in patients with ReA.'-5 In addition, the concentration of total polymeric Paired serum and synovial fluid samples from 29 patients infected with Y enterocolitica 0:3 were chosen from our large panel of samples collected from reactive arthritis patients for the present studies. Diagnosis of Yersinia infection was based in all patients on the typical clinical picture (diarrhoea, abdominal pain, vomiting and/or arthritis) and clearly increased levels of anti-yersinia antibodies detected by ELISA,'0 in 16 cases, additionally, the pathogen was isolated from the stools. Age of the patients ranged from 13 to 72 years (mean 35), and the female to male ratio was 9:20. All patients developed ReA as a post infectious complication within three weeks after the onset of Downloaded from http://ard.bmj.com/ on June 4, 2017 - Published by group.bmj.com Mdki-Ikola, Lahesmaa, Heesemann, et al 536 infection. Of the 27 patients tested, 23 were positive for the HLA B27 antigen. Serum and synovial fluid samples were collected simultaneously at 13 days to two months after the onset of infection; from one patient samples were taken at 8&5 months after the onset of infection. Synovial fluid aspirated from the knee was mixed with heparin (50 IU/ml), centrifugated at 200 g for 10 minutes and the supernatant was used for analysis. Both synovial fluid and serum samples were stored at -20C and studied simultaneously. Serum and synovial fluid samples from nine patients with ReA triggered by microbes other than yersiniae (5, salmonellae; 2, Chlamydia trachomatis; 1, Y pseudotuberculosis and 1, unknown aetiology), from four patients with rheumatoid arthritis and from nine patients with swollen joints for other reasons (6 with swollen knee with unknown reason; 4 with chronic synovial inflammation with unknown reason, and 1 with colitis ulcerosa) served as controls. was included on each plate. Antibody concentrations in the samples were expressed as relative units (enzyme immunoassay units; EIU): 1 U is 1/100 of the corresponding antibody concentration in the reference serum. Samples were tested as duplicates, and the results are expressed as the mean values. MONOCLONAL ANTIBODIES (MOAbs) The MoAb 9-200 recognises the 46 kD/57 kD RP (YopM) and the MoAb PU-174 the 26 kD RP (YopE) of Y enterocolitica 0:3." '3 The MoAbs 2C1, 6B6 and A6 react with the O-polysaccharide part of the LPS of Y enterocolitica 0:3.3 '0 MoAb 2B5 reacts with the outer core component of the Y enterocolitica 0:3 LPS.'0 All these MoAbs have been described earlier. INHIBITION- ELISAS WITH MOAbs The procedure was similar to that described earlier.'0 " Polystyrene microtitre plates were coated with RPs (Yops) or LPS of Y ANTIGENS enterocolitica 0:3. A 75 ,ul portion of serum As the source of RPs, plasmid-positive strain of samples diluted 1:10 in RP-ELISA and 1:10 Y enterocolitica 0:3 was grown and RPs and 1:50 in LPS ELISA were incubated on the prepared as previously described.9 11 The LPS plates for 2 hours at 37'C. After washings, a was extracted from Y enterocolitica 0:3 using 60 pl of the MoAb (at dilutions optimal for the phenol-water method.'0 The SDS extract each MoAb; that is, 9-200 at 1:50, PU-174 at of whole Y enterocolitica 0:3 bacteria was also 1:2000, 2C 1 at 1:2000, 6B6 at 1: 15 000, 2B5 at 1:800 and A6 at 1:800 dilutions) was added used as antigen.8 to incubate overnight at room temperature. After washings, 60 pl of goat alkaline ELISA FOR ANTI-YERSINIA ANTIBODIES phosphatase-conjugated antiserum (absorbed The ELISA procedures for IgM, IgG, IgA, with human serum) specific for mouse IgG and IgAl and IgA2 class yersinia antibodies and for IgM (TAGO, Burlingame, California) was yersinia antibodies containing secretory com- added at a dilution of 1:4000 and incubated at ponent (sIgA) have been reported earlier.'0 12 3 hours at 37°C. Quantitation by enzyme Patient serum or synovial fluid samples at substrate was carried out as described earlier. 1:250 dilution were allowed to react with SDS- Results were compared with values obtained by extract of whole yersinia bacteria or LPS the MoAb alone (without serum sample) to antigen attached to polystyrene microtitre calculate the inhibition percentage as described plates. In IgM, IgG and IgA class assays alka- earlier. 10 l line phosphatase-conjugated swine antibody to human IgM, IgG and IgA (Orion Diagnostica, Espoo, Finland) were used to detect bound IMMUNOBLOTTING ANALYSIS OF IgG AND IgA antibodies. In IgA subclass analysis MoAbs CLASS ANTI-YERSINIA ANTIBODIES against IgAl or IgA2 (Becton Dickinson, The SDS-polyacrylamide gel electrophoresis of Mountain View, California) were used to whole Y enterocolitica 0:3 bacteria and detect bound antibodies, after which rabbit immunoblotting were carried out as previously anti-mouse immunoglobulin (Miles-Yeda, described.'4 1' Kiryat Weizmann, Rehovot, Israel) absorbed with human IgG and swine alkaline phosphatase conjugated anti-rabbit IgG (Orion CONTROL ELISAS Diagnostica) were used. The sIgA antibodies To compare serum and synovial fluid were detected using rabbit immunoglobulins to distribution of antibodies against other than human secretory component (DAKO Patts the infecting microbe, the ELISAs for A/S, Glostrup, Denmark) and alkaline antibodies against Bordetella pertussis, tetanus phosphatase conjugated swine anti-rabbit IgG toxoid, Candida albicans, rubella and measles (Orion Diagnostica). Fresh p-nitrophenyl were performed as described elsewhere.'6-18 phosphate in diethanolamine-MgCl2-buffer solution (Orion Diagnostica) was used as a substrate and the reaction was stopped with STATISTICS 1 M sodium hydroxide. The optical density Wilcoxon test was used for comparison of was measured with a Titertek Multiscan antibody levels in serum and synovial fluid. Photometer (Labsystems, Helsinki, Finland) at Spearman's correlation test was used to assess a wavelength of 405 nm. Positive reference the degree of correlation between serum and serum with high levels of yersinia antibodies synovial fluid antibody concentrations. Downloaded from http://ard.bmj.com/ on June 4, 2017 - Published by group.bmj.com Yersinia-specific antibodies in serum and synovial fluid in patients with versinia triggered reactive arthritis Table 1 Antibody concentrations (in EIU) against sodium dodecyl sulphate extract of whole Yersinia enterocolitica 0:3 bacteria in paired serum and synovialfluid samples of patients with yersinia triggered reactive arthritis Synovialfluid Seruz IgM IgG IgA IgAl IgA2 sIgA 122-8 85 9 176 8 64-6 42-4 76 1 (67-6) (32-2) (99 4) (33 6) (44-5) (59 8) 97-9 (76-1) 75.6 (36-4) 151-6 (110-5) 55-2 (36 4) 44-2 (59-4) 51 2 (48 6) Spearmann 's correlation P* 0-0002 0-03 0 04 NS NS 0 0001 Numlber of sample coefficientt/P pairs 0 88/0 0000 0 74/0 00003 0-87/0-0000 0 65/0 0006 0-59/0-0022 0 90/0 0000 25 25 25 23 23 23 Mean (SD) values are given. EIU, enzyme immunoassay unit. NS, not significant. sIgA, IgA antibody containing secretory component. *Antibody concentrations between serum and synovial fluid samples are compared (Wilcoxon test). tBetween antibody concentrations in serum and synovial fluid samples. Table 2 Inhibition of the binding of Yersinia enterocolitica 0:3 lipopolysaccharide (LPS)-specific monoclonal antibodies (MoAbs) to Yersinia enterocolitica 0:3 LPS by paired serum and synovialfluid samples from patients with yersinia triggered reactive arthritis MoAb Specficity 2B5 A6 6B6 2C1 Outer core of LPS O-polysaccharide of LPS O-polysaccharide of LPS O-polysaccharide of LPS Samtiple Inhibition % in dilution serumt/synovialfluid 1:10 1:50 1:10 1:50 1:10 1:50 1:10 1:50 57-6 37 9 53 3 31 9 71 7 56 0 73-1 58-6 (29 1)/60 2 (28-4)/35-1 (33-2)/52-1 (31-0)/29-1 (28 6)/69 8 (35 0)/52-7 (30 7)/69-9 (36-5)/54-0 (28-7) (24 9) (31-2) (27 2) (35 5) (35-1) (37 4) (36-3) P* Spearmann's Numiber of correlation samiple pairs coefficientt/P NS NS NS NS NS NS NS NS 77/0 0000 95/0 0000 82/0-0000 96/0 0000 70/0 0002 93/0 0000 81/0-0000 0-95/0-0000 0 0 0 0 0 0 0 23 24 22 24 23 24 23 24 Mean (SD) values are given. *Antibody concentrations between serum and synovial fluid samples are compared (Wilcoxon test). tBetween antibody concentrations in serum and synovial fluid samples. 537 [inhibition-% > mean + 2 SD in controls]." The corresponding figures were eight for serum samples and 12 for synovial fluid samples when inhibition of MoAb 9-200 to 46 kD RP was studied. When the mean (SD) inhibition-% of all the serum or synovial fluid samples were studied, the serum samples were able to inhibit the binding of 26 kD RP-specific MoAb 38-4% (34 8), compared with 20-3% (29 9) inhibition with synovial fluid samples. Thus there were no statistically significant differences in the concentrations of antibodies to the MoAb PU- 174-defined epitope of 26 kD RP between the sera and synovial fluids. Accordingly, there were no differences in the ability to inhibit the binding of 46 kD RPspecific MoAb to 46 kD RP between serum [mean (SD) 58-8% (28-7)] and synovial fluid [65-9% (31-8)] samples. Similarly, when paired serum and synovial fluid samples were studied, there were no significant differences in the antibody concentrations against the MoAb 2B5-, A6-, 6B6and 2C1-defined epitopes of Y enterocolitica 0:3 LPS between sera and synovial fluids (table 2). There was a good correlation between the antibody concentrations in sera and synovial fluids in all assays; Spearman's correlation coefficient was from 0 70 to 0-96 (p from 0-0002 to 0-0000). IMMUNOBLOTr-TING Results DIRECT ELISA FOR ANTI-YERSINIA ANTIBODIES Paired samples of serum and synovial fluid from the patients with Y enterocolitica 0:3 infection were studied for the presence of antiyersinia antibodies (table 1). In all assays the correlation between the antibody concentrations in serum and synovial fluid was good: Spearman's correlation coefficient was from 0 59 to 090 (p from 00022 to 0 0001). The antibody levels against the SDS extract of whole bacteria as well as the LPS were significantly higher in the sera compared with the synovial fluid samples in IgM, IgG and IgA classes, as well as in sIgA (p from < 0 04 to 0 0001); especially clearly the difference was seen in the IgM and sIgA classes. In the IgAl and IgA2 subclass analysis there were no differences in the antibody levels between the sera and synovial fluid samples. No significant differences were found between the absorbance values in the SDS extract ELISA and the LPS ELISA indicating that the majority of the antibodies are directed against the LPS part of the bacteria. To study whether we can identify qualitative differences in the antibody reactivity to various yersinia antigens, paired serum and synovial fluid samples were analysed by immunblotting with whole Y enterocolitica 0:3 bacteria as antigen in the nitrocellulose sheet. Based on this extensive analysis, identical panel of Y enterocolitica antigens was recognised by IgG and IgA of the serum and synovial fluid samples (figure). CONTROL ELISAS The antibody level in serum samples was higher than in synovial fluid samples when IgG class antibodies against tetanus toxoid, measles and rubella were studied. Also the level of IgM, IgG and IgA class antibodies against Candida albicans was higher in serum samples compared with synovial fluid samples (table 3). CONTROL SAMPLES Paired serum and synovial fluid samples from 22 control patients with other rheumatic diseases than yersinia triggered ReA were studied for the presence of yersinia-specific antibodies, as well as antibodies against the INHIBITION ELISAS FOR YERSINIA ANTIBODIES above mentioned control microbes. The Paired samples of serum and synovial fluid yersinia antibody levels were always low. There from 17 patients with Y enterocolitica 0:3 were no differences in the antibody levels infection were used to inhibit the binding of between sera and synovial fluids in IgM, IgG, MoAbs to 26 kD and 46 kD RPs. Nine of the IgA and sIgA class ELISAs using LPS as serum samples and four of the synovial fluid antigen, or in IgG and sIgA classes using SDS samples are able to inhibit the binding of extract of whole bacteria as antigen. In IgM MoAb PU- 174 to 26 kD RP significantly more and IgA class with SDS extract antigen the than sera taken from healthy controls antibody concentration was higher in serum Downloaded from http://ard.bmj.com/ on June 4, 2017 - Published by group.bmj.com 538 Mdki-Ikola, Lahesmaa, Heesemann, et al Discussion In the present study we demonstrate that in patients with yersinia triggered ReA yersiniaspecific antibodies in the synovial fluid mirror those in the serum as judged by concentration, by specificity or by distribution in antibody classes or subclasses (p always <0 003 for correlation). When there was a statistically significant difference in the antibody concentrations between serum and synovial fluid samples, the antibody level was always higher in serum compared to synovial fluid. These results do not speak for any strong local antibody production, but indicate that the majority of yersinia antibodies in the synovial fluid are derived from the circulation. The results from assays with microbes unrelated to yersinia infections as antigens support this concept. Accordingly, no significant differences were found in the antibody levels against Chlamydia trachomatis, Yersinia enterocolitica or Borrelia burgdorferi between serum and synovial fluid samples from patients with undifferentiated oligoarthritis.'9 Further, the difference in antibody concentrations between serum and synovial fluid samples was biggest in IgM and sIgA (the largest immunoglobulin molecules in size) assays, which is in line with the above suggestion of filtration of the antibodies from circulation into the joint. Controversially, we have recently found evidence for intra-articular antibody production of IgA2 subclass against salmonella LPS in patients with salmonella triggered ReA.8 Most of the circulating IgA belongs to the subclass IgAl, whereas in locations close to mucosal surfaces, especially the distal gut where salmonella invades the mucosa, the number of IgA2 producing cells is increased.20 21 Thus antibodies of IgA2 subclass in synovial fluid were thought to Immunoblotting analysis of serum and synovialfluid IgG indicate mucosal origin. It was speculated that response to Yersinia enterocolitica 0:3 in tw o patients the high level of IgA2 class salmonella specific with reactive arthritis showing that identical p)anel of Y antibodies in the joint was a result of selective enterocolitica antigens was recognised by the serum and synovialfluid samples. migration of IgA2-secreting lymphocytes from the distal gut mucosa to synovium,22 where the compared with synovial fluid (p < 005). The existing salmonella LPS may have stimulated antibody level was higher in serum compared local IgA2 class antibody production.4 Howwith synovial fluid also when antibo dies against ever, yersinia penetrates the mucosa in the rubella (p < 00004), measles (p < 0.0003), proximal part of the gut, where no increased tetanus toxoid (p < 0-0004), Bordet ella pertussis number of IgA2 producing cells are seen. Thus (p < 0002) and Candida albicans (Ip < 0 0003) if the intra-articular antibody production for IgM, p < 0 0003 for IgA and p < 0 0001 for suggested in salmonella triggered ReA is subclass specific, the difference in the location IgG) were measured. of invasion between salmonella and yersinia may explain why IgA2 class anti-yersinia antibodies may not have been produced locally Table 3 Antibody concentrations against control microbes in paired serum a? in the joints in patients with yersinia triggered fluid samples ofpatients with yersinia triggered reactive arthritis ReA. It is, however, important to note that Number Serum Synovialfluid P* higher antibody concentrations are generally of sample detected in the serum, as shown here and also, pairs for example, in patients with rheumatoid 19 < 0-01 7-3 (8-0) Tetanus toxoid, IgG* 10-1 (9-2) arthritis.23 Therefore even if the concentration 25 <0-05 37-1 (33 5) 42-5 (30 4) Measles, IgG* 25 < 0-01 of antibodies in synovial fluid was lower than 39-2 (32-5) 49-0 (29-7) Rubella, IgG* Candida albicans that observed in serum, the synovial fluid may 22 < 0 0005 0 559 (0-226) 0-370 (0 205) gMIt have contained locally produced antibodies. 22 < 0-01 1-047 (0 444) 1-186 (0-416) IgGt 22 Further, it is also possible that some of the <0 005 0-421 (0 249) 0-538 (0-241) IgAt synovial fluid antibodies are lost in the assays Mean (SD) values are given. *EIU, enzyme immunoassay unit. due to the formation of immune complexes tAbsorbance value. and subsequent trapping in the cartilage. (Wilcoxon test). compared samples are fluid serum and synovial tAntibody concentrations between Downloaded from http://ard.bmj.com/ on June 4, 2017 - Published by group.bmj.com Yersinia-specific antibodies in serumn and synovial fluid in patients with versinia triggered reactive arthritis Likewise, it is possible that the finding of similar levels in serum and synovial fluid yersinia-specific IgAl and IgA2 subclasses, as well as of IgA and IgG classes would be of some importance, since the differences between antibody levels in serum and synovial fluid in other immunoglobulin classes and in controls against other organisms are much more striking than in these mentioned tests. The role of RPs in the pathogenesis of ReA has not been well studied yet. There are some indications for RPs to play some role in the pathogenesis of ReA: 1) patients with yersinia triggered ReA have more often antibodies against the 36 kD RP and higher concentrations of antibodies against the 26 kD RP in the beginning of the disease than the patients who recover from the infection without arthritis;" '4 2) total amount of IgA class antibodies against RPs is significantly higher in arthritic patients in the beginning of the disease;'5 3) yersinia bacilli was demonstrated in the intestinal biopsy specimens of patients with yersinia-triggered ReA by immunofluorescence using monospecific rabbit antiserum to 46 kD RP24 and 4) strong T-cell reactivity to RPs was detected in synovial fluid samples taken from patients with reactive arthritis.25 26 In the present study we could find no evidence for intra-articular antibody production against the two MoAb-defined epitopes of 26 kD and 46 kD RPs. The intrasynovial antibodies also against RPs may have therefore filtrated to the joint from the circulation rather than been produced locally. Thus the involvement of RPs in the pathogenesis of ReA remains open to speculation. In ReA, microbial antigens enter the body via mucosal surfaces. Subsequently, microbial LPS is transported to the joints. -5 Microbial antibodies are produced and may filtrate into the synovium from circulation, or as in case of salmonella triggered ReA, they may be produced locally in the joints.8 Antibodies may also be transported to the joint as part of immune complexes, as immune complexes consisting of yersinia antigen and specific antibody are found in the circulation as well as in synovial fluid in yersinia triggered ReA.27 The specific antibodies together with microbial antigens in the joint may participate in inflammation and tissue injury through complement activation. LPS may play an important role in these events, as a major part of yersinia-specific antibodies in serum and synovial fluid in patients with ReA were directed against LPS. Our previous studies also suggest a crucial role for LPS in the pathogenesis of ReA. 1-4 8 10 28 We thank Professor David T Y Yu for 2C1 and 6B6 monoclonal antibodies, Dr Rauli Leino for some of the patient samples and records, Raija Vainionpaa for analysing the viral antibodies and Mrs Soile Niittoaho and Mrs Maija Salonen for technical assistance. Our study was supported by the grants from the Rheumatism Research Foundation, the Turku University Foundation, the Yrjo Jahnsson Foundation, the Sigrid Juselius Foundation and the Finnish Academy, and by contract with the Finnish Life and Pension Insurance Companies. 539 1 Granfors K. Do bacterial antigens cause reactive arthritis? Rheuiii Dis Clilii North Aim, 1992; 18: 37-48. 2 Maki-Ikola 0, Granfors K. Salmonella-triggered reactive arthritis. Lanicet 1992; 339: 1096-8. 3 Granfors K, Jalkanen S, von Essen R, et al. Yersinia antigens in synovial-fluid cells from patients with reactive arthritis. NE?igl _Med 1989; 320: 216-21. 4 Granfors K, Jalkanen S, Lindberg A A, et al. Salmonella lipopolysaccharide in synovial cells from patients with reactive arthritis. Lauzcet 1990; 335: 685-8. 5 Hammer M, Zeidler H, Klimsa S, Heesemann J. Yersinia enterocolitica in the synovial membrane of patients with Yersinia-induced arthritis. Arthritis Rhetumi 1990; 33: 1795-800. 6 Inman R D, Johnston M E A, Klein M H. Analysis of serum and synovial fluid IgA in Reiter's syndrome and reactive arthritis. Clii Ihnniityiol Ih)i;nioiqpathlol 1987; 43: 195-203. 7 Hughes R A, Treharne J D, Keat A C. Comparison of serum and synovial fluid chlamydial antibody titres. Arthritis RhellM 1989; 32: sl 13 (c129). 8 Maki-Ikola 0, Yli-Kerttula U, Saario R, Toivanen P, Granfors K. Salmonella-specific antibodies in serum and svnovial fluid in reactive arthritis. Br _7 Rheniapatol 1992; 31: 25-9. 9 Heesemann J, Gross U, Schmidt N, Laufs R. Immunochemical analysis of plasmid-encoded proteins released by enteropathogenic yersinia sp. grown in calcium-deficient media. Inifect Inoniiini 1986; 54: 561-7. 10 Granfors K, Ogasawara M, Hill J L, Lahesmaa-Rantala R, Toivanen A, Yu D T Y. Analysis of IgA antibodies to lipopolysaccharide in yersinia-triggered reactive arthritis. _JlIfecrDis 1989; 159: 1142-7. 11 Maki-Ikola 0, Pulz M, Heesemann J, et al. Antibody response against 26 kD and 46 kilodalton released proteins of yersinia in yersinia-triggered reactive arthritis. Amiii Rheioni Dis 1992; 51: 1247-9. 12 Granfors K, Toivanen A. IgA-anti-yersinia antibodies in yersinia triggered reactive arthritis. Ainn Rheiini Dis 1986; 45: 561-5. 13 Heesemann J, Kalthoff H, Koch F. Monoclonal antibodies directed against plasmid-encoded released proteins of enteropathogenic Yersinia. FEMS Microbiol Lett 1986; 36: 15-9. 14 Stahlberg T H, Heesemann J, Granfors K, Toivanen A. Immunoblot analysis of IgM, IgG and IgA responses to plasmid encoded released proteins of yersinia enterocolitica in patients with or without yersinia triggered reactive arthritis. Annii Rheiton Dis 1989; 48: 577-81. 15 Lahesmaa-Rantala R, Heesemann J, Lehtonen O-P, Granfors K, Toivanen A. Avidity of antibodies against released proteins of yersinia spp: comparison of patients with or without reactive arthritis. Annii Rheumii Dis 1989; 48: 1003-6. 16 Ruuskanen 0, Viljanen M K, Salmi T T, Lehtonen O-P, Kouvalainen K, Peltonen T. DTP and DPT-inactivated polio vaccines: comparison of adverse reactions and IgG, IgM and IgA antibody responses to DTP. Acta Paediatr Scanid 1980; 69: 177-82. 17 Salonen J, Vainionpaa R, Halonen P. Assav of measles virus IgM and IgG class antibodies by use of peroxidaselabelled viral antigens. Arch Virol 1986; 91: 93-106. 18 Kaukonen K, Savolainen J, Viander M, Terho E 0. Characterization of Aspergillus umbrosus carbohydrate antigens by biotinylated lectins and IgG response to mannan/mannoprotein antigens in patients with farmer's lung. Cliii ExpAllergy 1993; 23: 21-7. 19 Sieper J, Braun J, Brandt J, et al. Pathogenetic role of Chlamydia, Yersinia and Borrelia in undifferentiated oligoarthritis. JRheumiiatol 1992; 19: 1236-42. 20 Crago S S, Kutteh W H, Moro I, et al. Distribution of IgA 1-, IgA2-, and J chain-containing cells in human tissues. Jmlnniunol 1984; 132: 16-8. 21 Kett K, Brandtzaeg P, Fausa 0. J-chain expression is more prominent in immunoglobulin A2 than in immunoglobulin Al colonic immunocvtes and is decreased in both subclasses associated with inflammatory bowel disease. Gastroewterol 1988; 94: 1419-25. 22 Salmi M, Granfors K, Leirisalo-Repo M, et al. Selective endothelial binding of interleukin-2-dependent human T-cell lines derived from different tissues. Proc Natl Acad Sci 1992; 89: 11436-40. 23 Musiani M, Zerbini M, Ferri S, Plazzi M, Gentilomi G, La Placa M. Comparison of the immune response to EpsteinBarr virus and cytomegalovirus in sera and synovial fluids of patients with rheumatoid arthritis. Annii Rheitni Dis 1987;46: 837-42. 24 Hoogkamp-Korstanje J A A, de Koning J, Heesemann J. Persistence of yersinia enterocolitica in man. Inifection 1988; 16: 81-5. 25 Lahesmaa R, Yssel H, Batsford S, et al. Yersinia enterocolitica activates a T helper type 1-like T cell subset in reactive arthritis. Innnunol 1992; 148: 3079-85. 26 Lahesmaa R, Leirisalo-Repo M, Soderberg C, et al. Preferential TCR VP usage by Yersinia antigen-reactive T cells isolated from a patient with reactive arthritis. A rthntis Rheuwm 199 3; 36: S 151. 27 Lahesmaa-Rantala R, Granfors K, Isomaki H, Toivanen A. Yersinia specific immune complexes in the synovial fluid of patients with yersinia triggered reactive arthritis. Apin Rheu,n Dis 1987; 46: 510-4. 28 Maki-Ikola 0, Leirisalo-Repo M, Kantele A, Toivanen P, Granfors K. Salmonella-specific antibodies in reactive arthritis. l Infect Dis 1991; 164: 1141-8. Downloaded from http://ard.bmj.com/ on June 4, 2017 - Published by group.bmj.com Yersinia-specific antibodies in serum and synovial fluid in patients with Yersinia triggered reactive arthritis. O Mäki-Ikola, R Lahesmaa, J Heesemann, R Merilahti-Palo, R Saario, A Toivanen and K Granfors Ann Rheum Dis 1994 53: 535-539 doi: 10.1136/ard.53.8.535 Updated information and services can be found at: http://ard.bmj.com/content/53/8/535 These include: Email alerting service Receive free email alerts when new articles cite this article. Sign up in the box at the top right corner of the online article. Notes To request permissions go to: http://group.bmj.com/group/rights-licensing/permissions To order reprints go to: http://journals.bmj.com/cgi/reprintform To subscribe to BMJ go to: http://group.bmj.com/subscribe/