Influence of HLADR genes on the production of rheumatoid arthritisspecific autoantibodies to citrullinated fibrinogen.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 52, No. 11, November 2005, pp 3424–3432 DOI 10.1002/art.21391 © 2005, American College of Rheumatology Influence of HLA–DR Genes on the Production of Rheumatoid Arthritis–Specific Autoantibodies to Citrullinated Fibrinogen Isabelle Auger,1 Mireille Sebbag,2 Christian Vincent,2 Nathalie Balandraud,1 Sandrine Guis,1 Leonor Nogueira,2 Björn Svensson,3 Alain Cantagrel,4 Guy Serre,2 and Jean Roudier1 Objective. Antibodies directed against citrullinated fibrinogen are highly specific for rheumatoid arthritis (RA). This study was undertaken to test whether RA-associated HLA–DR alleles are associated with anti–citrullinated fibrinogen in RA patient sera and whether replacement of arginyl by citrullyl residues on fibrinogen peptides modifies their binding to HLA–DR molecules and their recognition by T cells. Methods. Antikeratin, antifilaggrin, and anti– citrullinated fibrinogen antibodies were assayed in RA patients who had undergone HLA–DR typing. Direct assays were performed to investigate binding of citrullinated or native fibrinogen peptides (encompassing the entire ␣- and ␤-chains of fibrinogen) to purified HLA–DR molecules. T cell proliferative responses to citrullinated or native fibrinogen peptides were measured in RA patients and controls. Results. HLA–DRB1*0404 was associated with anti–citrullinated fibrinogen in RA sera (P ⴝ 0.002). For the RA-associated alleles HLA–DRB1*0401 and HLA–DR1, there was a nonsignificant trend toward association (P ⴝ 0.07). Multiple peptides from the ␣and ␤-chains of fibrinogen bound many HLA–DR alleles; DRB1*0404 was the best fibrinogen peptide binder. Citrullination did not influence fibrinogen pep- tide binding to HLA–DR or fibrinogen peptide recognition by T cells. Peripheral blood T cells that recognized native or citrullinated fibrinogen peptides were common in RA patients but not in healthy controls. Conclusion. The RA-associated HLA–DRB1*0404 allele is also associated with production of antibodies to citrullinated fibrinogen. DRB1*0401 and DRB1*01 tend to be associated with anti–citrullinated fibrinogen, but this is not statistically significant. Citrullination of fibrinogen peptide does not influence peptide–DR–T cell interaction. Finally, T cell proliferation in response to citrullinated or uncitrullinated fibrinogen peptides is frequent in RA patients and very infrequent in controls. Rheumatoid arthritis (RA) is a chronic inflammatory joint disease with a prevalence of 0.5–1% worldwide. The mechanisms leading to RA are unknown. RA is considered a polygenic autoimmune disease. HLA–DR genes are the strongest genetic component. DR alleles whose B1-chain contains the “shared epitope” (SE), a conserved 5–amino acid motif, are associated with RA susceptibility (1). Susceptibility alleles include HLA–DR4 subtypes HLA–DRB1*0401, DRB1*0404, DRB1*0405, and DRB1*0408, and HLA– DR1. HLA–DR genotypes containing 2 RA susceptibility alleles (“double-dose” genotypes) confer a higher risk than genotypes containing only 1 susceptibility allele (“single-dose” genotypes), which confer a higher risk than DR genotypes containing no susceptibility allele (2). RA patient sera contain autoantibodies to “citrullinated” (deiminated) proteins (ACPAs) which are specific for RA and recognize peptidic epitopes centered by a citrulline (a posttranslationally modified form of arginine, produced by deimination) on various proteins, such as filaggrin, vimentin, and fibrinogen (3–6). Citrullinated fibrinogen is found in RA synovium, where ACPAs are produced (7). 1 Isabelle Auger, PhD, Nathalie Balandraud, MD, PhD, Sandrine Guis, MD, PhD, Jean Roudier, MD, PhD: INSERM UMR 639, La Conception Hospital, Marseille, France; 2Mireille Sebbag, PhD, Christian Vincent, MD, PhD, Leonor Nogueira, MD, PhD, Guy Serre, MD, PhD: CNRS-Toulouse III University, Purpan Hospital, Toulouse, France; 3Björn Svensson, MD: Helsingborg Lasarett AB, Helsingborg, Sweden; 4Alain Cantagrel, MD: Toulouse III University, Rangueil Hospital, Toulouse, France. Address correspondence and reprint requests to Jean Roudier, MD, INSERM UMR 639, Faculté de Médecine, 27 BD Jean Moulin, 13005 Marseille, France. E-mail: jean.roudier@ medecine.univ-mrs.fr. Submitted for publication December 16, 2004; accepted in revised form July 27, 2005. 3424 HLA–DR AND ANTI–CITRULLINATED FIBRINOGEN IN RA 3425 Figure 1. Position of peptides on the ␣- and ␤-chains of fibrinogen. Empty boxes represent each of 53 15-mer peptides encompassing the whole ␣-chain or 43 15-mer peptides encompassing the whole ␤-chain. For each 15-mer peptide with an arginine residue (R) (40 peptides on the ␣-chain and 31 on the ␤-chain), a citrullinated (CIT) variant was also synthesized. Extra boxes represent extra, overlapping peptides synthesized in areas where the position of the R residue might have compromised its interaction with the P4 pocket. In this study, we examined the role of RAassociated HLA–DR alleles in the production of antibodies to citrullinated fibrinogen in RA patient sera. We also investigated the binding of native and citrullinated fibrinogen peptides to different HLA–DR alleles and measured T cell proliferative responses to native and citrullinated fibrinogen peptides in RA patients and controls. We found that expression of RA-associated HLA–DR alleles was not mandatory for anti– citrullinated fibrinogen production. The RA-associated HLA–DRB1*0404 allele was strongly associated with the presence of anti–citrullinated fibrinogen in RA sera, whereas HLA–DRB1*01 and HLA–DRB1*0401 showed only a trend toward association. Both RAassociated and non–RA-associated HLA–DR alleles bound multiple peptides from the ␣- and ␤-chains of fibrinogen, in both their native and citrullinated forms. Finally, peripheral blood mononuclear cells (PBMCs) from most RA patients, but not healthy controls, were able to proliferate in response to fibrinogen peptides, in both their native and citrullinated forms. PATIENTS AND METHODS RA patients and controls. We studied 388 RA patients from the rheumatology units at La Conception Hospital in Marseille (122 patients; mean age 58 years, mean disease duration 10 years, 78% female), Rangueil Hospital in Toulouse (157 patients; mean age 54 years, mean disease duration 5 years, 79% female), and Helsingborg Lazarett AB in Helsingborg (109 patients; mean age 53 years, mean disease duration 8 years, 73% female). All patients fulfilled the American College of Rheumatology (formerly, the American Rheumatism Association) 1987 criteria for RA (8). Laboratory staff volunteers served as normal controls. All patients and controls underwent HLA–DR oligotyping. The distribution of 3426 HLA alleles was similar among the 3 groups of patients (Marseille 30% SE negative, 45% single-dose SE, 24% doubledose SE; Toulouse 23% SE negative, 58% single-dose SE, 19% double-dose SE; Helsingborg 17% SE negative, 69 % singledose SE, 14% double-dose SE). Detection of ACPAs. For each serum sample, we assayed 3 ACPA antibody systems: the classic antikeratin and antifilaggrin antibodies, as well as anti–citrullinated fibrinogen antibodies. Antikeratin antibodies were detected by immunofluorescence on rat esophagus epithelium as described previously (9). Antifilaggrin antibodies were assayed by immunoblotting on human epidermis filaggrin as described previously (10). Anti–citrullinated fibrinogen was detected with a recently developed in-house enzyme-linked immunosorbent assay (ELISA), using in vitro deiminated human fibrinogen as immunosorbent (11). Briefly, microtiter plates were coated overnight with human deiminated fibrinogen (5 g/ml) diluted in phosphate buffered saline (PBS; pH 7.4). The plates were blocked with PBS containing 2% bovine serum albumin and 100 l sera, diluted to 1:50 in 2M NaCl/PBS, and incubated for 1 hour. After washing, horseradish peroxidase–labeled, Fcspecific goat antibody to human IgG was added, followed by incubation for 1 hour and further washing. All incubations and washing steps were performed at 4°C. Bound antibody was detected with orthophenylenediamine dihydrochloride (Sigma, St. Quentin, France). Plates were read using a Multiskan plate reader (Thermo Labsystem, Cergy-Pontoise, France). For each of the 3 above-described assays, a serum was considered positive when its titer reached a previously established diagnostic threshold for 98.5% specificity (9–11). Classification of HLA–DR alleles and genotype groups. HLA–DR alleles were classified into 4 groups: group A (DRB1*0404/0405/0408), group B (DRB1*0401), group C (DRB1*0101/*0102), and group D (all others). Based on these allelic groups, RA patients were divided in genotypic groups. One hundred eighty-four “single-dose” patients were classified as A/D (n ⫽ 47), B/D (n ⫽ 68), or C/D (n ⫽ 69). The D/D genotypic group consisted of 117 patients expressing no RAassociated HLA–DRB1 allele. Eighty-seven “double-dose” patients were classified as A/B (n ⫽ 28), B/B (n ⫽ 13), B/C (n ⫽ 20), C/C (n ⫽ 9), A/C (n ⫽ 12), or A/A (n ⫽ 5). The chi-square test (12) was used to compare antibody status (positive or negative) between genotypic groups. Synthetic peptides. Peptides (Neosystem, Strasbourg, France) were synthesized using the solid-phase system and purified (⬎60%). We synthesized 93 15-mer peptides encompassing the entire sequence of the ␣-chain of human fibrinogen (locus NP_000499). These were 40 arginine-containing peptides, their 40 citrulline-substituted variants, and 13 peptides without arginine. We also synthesized 74 15-mer peptides encompassing the entire sequence of the ␤-chain of human fibrinogen (locus NP_005132). These were 31 argininecontaining peptides, their 31 citrullinated variants, and 12 peptides without arginine. The positions of peptides on the ␣- and ␤-chains of fibrinogen are indicated in Figure 1.Whenever we thought the position of the R residue compromised interaction with the P4 pocket, we synthesized extra, overlapping peptides, which appear as extra boxes in Figure 1. The amino acid sequence of each peptide is indicated in Figure 2. AUGER ET AL Figure 2. Amino acid sequences of ␣- and ␤-chain fibrinogen peptides. r ⫽ arginine; r* ⫽ citrulline. HLA–DR AND ANTI–CITRULLINATED FIBRINOGEN IN RA 3427 T cell proliferation assay. Mononuclear cells from RA patients and controls were isolated from 30 ml of heparinized blood by centrifugation through Ficoll-Histopaque (Sigma). Cells were cultured at a density of 106 cells/ml in RPMI 1640 with 10% self serum in the presence of 2 g/ml stimulatory peptide. After 7 days of culture at 37°C, proliferative response to the peptide was evaluated using the colorimetric immunoassay bromodeoxyuridine kit (Roche Diagnostics, Meylan, France). Positive responses were defined as optical density (OD) higher than twice the OD for cells cultured without peptide (13). Lymphoblastoid cell lines. The HLA-homozygous lymphoblastoid cell lines JESTHOM (HLA–DRB1*0101), SAVC (HLA–DRB1*0401), YAR (HLA–DRB1*0402), PEYSSON (HLA–DRB1*0404), and MOU (HLA–DRB1*0701) were cultured in RPMI 1640 with 10% fetal calf serum. Antibodies. The anti–HLA–DR monoclonal antibodies LB3.1 (a gift from M. F. Del Guercio, Cytel, San Diego, CA) and biotinylated B8122 (Immunotech, Marseille, France) were used. Peroxidase-conjugated avidin was supplied by Sigma. Purification of HLA–DR molecules. DRB1-homozygous cells (2 ⫻ 109) were lysed in 10 mM Tris (pH 8), 10 mM NaCl, 10 mM MgCl2, 1% Triton X-100, 0.05 mg/ml DNase, and protease inhibitors. Total protein extracts were immunoprecipitated with anti–HLA–DR antibody LB3.1 covalently coupled to cyanogen bromide–activated Sepharose 4B (Sigma). After washing, HLA–DR molecules were eluted in PBS (pH 2) with 0.5% n-octylglucoside, neutralized in 1M Tris, and quantified. Peptide binding assay. The peptide binding assay used in these studies is a variant of the protein/HLA–DR binding assay we used previously (14). ELISA plates were coated with 10 g of peptide per well and blocked with 1% bovine serum albumin. One microgram of purified HLA–DR molecule was added to each well. After washing, bound HLA–DR was detected by administration of biotinylated anti–HLA–DR antibody followed by peroxidase-conjugated avidin. OD was read at 405 nm. The binding of each of the 5 purified HLA–DR alleles was assayed on a set of 6 96-well ELISA plates coated with 167 peptides from the ␣- or ␤-chain of fibrinogen (each in duplicate wells) and, as controls, 2 empty wells (not coated with any peptide) and 2 wells coated with a positive binder, influenza hemagglutinin (HA) peptide (PKYVKQNTLKLAT) (15). The total number of ELISA plates needed to assay the binding of the 167 fibrinogen peptides to the 5 different HLA–DR alleles was 30 (5 ⫻ 6). Tests of the binding of the citrullinated variant of a given peptide were always done on the same plate as the native peptide. Positive binding was defined as an OD value more than twice the OD for the empty well (HLA–DR without peptide). In each of the 30 plates, this cutoff value was very close to the OD obtained with HA peptide. To ensure that our data were consistent with binding data obtained by conventional methods, we prepared an extra plate coated with 6 specific peptides: 3 from the third hypervariable regions of HLA–DRB1*0401 (KDLLEQKRAAVDTYC), *0402 (KDILEDERAAVDTYC), and *0301 (KDLLEQKRGRVDNYC) (which had been found to be negative by O’Sullivan et al ), the HA peptide (positive control), fibrinogen peptide 167 (which we had classified as a very good binder), and fibrinogen peptide 115 (which we had classified as a nonbinder). For the 5 HLA–DR alleles tested, Table 1. Percentage of rheumatoid arthritis (RA) patients whose sera were positive for anti–citrullinated protein, by number of shared epitopes* No. of shared epitopes (n) Two (87) A/B (28) B/B (13) B/C (20) C/C (9) A/C (12) A/A (5) One (184) A/D (47) B/D (68) C/D (69) Zero (117) D/D (117) Anti–citrullinated protein Keratin Filaggrin Fibrinogen 46 38 50 44 17 40 57 38 45 44 42 40 96† 85 80 78 75 60 40 31 32 45 34 46 83‡ 71 71 24 26 55 * A ⫽ HLA–DRB1*0404/0405/0408; B ⫽ HLA–DRB1*0401; C ⫽ HLA–DRB1*0101*/*0102; D ⫽ non-RA HLA–DR allele. † P ⬍ 0.001 versus D/D group. ‡ P ⫽ 0.002 versus D/D group. the data from our direct binding assay were consistent with data obtained by the conventional technique (15). RESULTS RA-associated HLA–DR alleles are not mandatory for production of autoantibodies to citrullinated proteins. The sera of 388 RA patients were tested for the presence of antikeratin, antifilaggrin, and anti– citrullinated fibrinogen antibodies (9–11). To evaluate the influence of individual RA-associated HLA–DR alleles on anti–citrullinated protein antibody production, we compared the frequency of antikeratin-, antifilaggrin-, and anti–citrullinated fibrinogen–positive sera among the groups of patients expressing either 1 (single-dose genotypes), 2 (double-dose genotypes) or no RA-associated HLA–DR alleles. Fifty-five percent of the SE-negative patients (D/D genotype) had antibodies to citrullinated fibrinogen. Among single-dose SE positive patients, only the A/D group (DRB1*0404/0405/0408) had a significantly higher frequency of anti–citrullinated fibrinogen antibodies than the D/D group (83% anti–citrullinated fibrinogen positive; P ⫽ 0.002 by chi-square analysis) (Table 1). In the B/D and C/D groups, there was only a trend toward an increased frequency of anti– citrullinated fibrinogen positivity compared with the D/D group (both 71% anti–citrullinated fibrinogen positive; P ⫽ 0.07 by chi-square analysis) (Table 1). Among the double-dose genotypes, only the A/B group (com- 3428 AUGER ET AL Figure 3. Binding of peptides from the ␣-chain (A) and the ␤-chain (B) of fibrinogen to HLA–DR alleles. Enzyme-linked immunosorbent assay plates were coated with 10 g of fibrinogen peptide per well. One microgram of purified HLA–DR molecule was then added to the plates. After washing, bound HLA–DR was detected by administration of biotinylated anti–HLA–DR antibody followed by peroxidase-conjugated avidin. Optical density (OD) was read at 405 nm. Positive binding was defined as an OD value higher than twice the OD for HLA–DR without peptide. pound DRB1*0401/0404-05-08 heterozygotes) had a frequency of anti–citrullinated fibrinogen positivity (96%) higher than that in the D/D group (P ⬍ 0.001). The analysis of antikeratin and antifilaggrin antibody production in the various genotypic groups yielded similar data, highlighting the particular association of anti–citrullinated protein antibodies with HLA– DRB1*0404/0405/0408 single-dose genotypes and with the compound heterozygous HLA–DRB1*0404-05-08/ 0401 double-dose genotype. Promiscuous binding of fibrinogen peptides to HLA–DR alleles, and lack of influence of citrullination on peptide binding to HLA–DR. The association (or, for HLA–DRB1*0401 and HLA–DR1, trend toward association) between RA-associated HLA–DR alleles and production of anti–citrullinated fibrinogen suggested HLA–DR AND ANTI–CITRULLINATED FIBRINOGEN IN RA 3429 Figure 4. Binding of ␣-chain fibrinogen peptides to HLA–DR alleles, and proliferative responses of peripheral blood mononuclear cells from rheumatoid arthritis (RA) patients and controls to fibrinogen peptides. that these alleles might bind peptides from citrullinated fibrinogen. To test this hypothesis, we used a direct binding assay. The 53 15-mer peptides encompassing the entire fibrinogen ␣-chain and each of their 40 citrullinesubstituted variants, as well as the 43 15-mer peptides encompassing the entire ␤-chain of fibrinogen and each of their 31 citrulline-substituted variants, were tested for binding to purified HLA–DRB1*0401, *0404, and *0101 (susceptibility alleles) and DRB1*07 and *0402 (non– susceptibility alleles) (Figures 3A and B). For each HLA–DR allele, positively binding peptides were defined as those yielding OD values more than twice the OD for HLA–DR without peptide. We found that every HLA–DR allele tested was able to bind peptides from the ␣- and ␤-chains of fibrinogen. The HLA–DR allele that bound the highest number of fibrinogen peptides was HLA–DRB1*0404, which bound 18 of 93 peptides from the ␣-chain and 28 of 74 from the ␤-chain. Peptide citrullination did not appear to influence binding to HLA–DR alleles (Figures 3A and B). T cell proliferative responses to fibrinogen peptides are common in RA patients and uncommon in controls, and are independent of citrullination. Thirtyseven peptides from the ␣-chain of fibrinogen and 43 from the ␤-chain of fibrinogen that, in native or citrullinated form, bound at least 1 of the 5 tested HLA–DR alleles were used in proliferation studies in the presence of PBMCs from 12 RA patients and 10 healthy controls. In general, proliferative responses to fibrinogen peptides were quite common in patients and extremely uncommon in controls (Figures 4 and 5). Eighteen peptides from the ␣-chain of fibrinogen and 27 from the ␤-chain triggered in vitro proliferative responses in RA PBMCs, whereas only 2 peptides from the ␣-chain and 5 peptides from the ␤-chain triggered such responses in controls (Figure 6). Proliferative responses to both citrullinated and native fibrinogen peptides were observed. There was no indication that citrullinated peptides could stimulate proliferative response more efficiently than native peptides (Figure 6). When we observed a positive proliferative re- 3430 AUGER ET AL Figure 5. Binding of ␤-chain fibrinogen peptides to HLA–DR alleles, and proliferative responses of peripheral blood mononuclear cells from rheumatoid arthritis (RA) patients and controls to fibrinogen peptides. sponse to a fibrinogen peptide in an individual patient, we performed an ELISA for antipeptide antibodies, to eliminate cell proliferation induced by immune complexes. We did not observe the presence of both an antibody and a cellular response to the same (citrullinated or native) fibrinogen peptide in any patient, despite the use of control antigens (Hsp70, Epstein-Barr virus gp110) in our antifibrin peptide ELISA. DISCUSSION The sera of patients with rheumatoid arthritis contain a family of highly disease-specific autoantibodies (anti–citrullinated protein autoantibodies) which recognize citrulline-centered peptidic epitopes on various proteins, such as filaggrin, vimentin, and fibrinogen. Citrullyls are arginyl residues that have been converted from their native basic form into a neutral variant by a posttranslational modification called deimination, mediated by peptidylarginine deiminases. The origin of immunization against peptidic epitopes containing citrullyl residues is unknown. However, ACPAs are produced in the synovium of RA patients, and their most abundant target in the synovium is deiminated fibrinogen (6). Therefore, it is likely that deiminated fibrinogen plays a role in ACPA production. In the present study we analyzed the molecular basis for production of autoantibodies to deiminated (or citrullinated) fibrinogen in patients with RA. We found that expression of RA-associated HLA–DR alleles is not mandatory in order for RA patients to develop anti–citrullinated fibrinogen. Indeed, 55% of sera from SE-negative RA patients (D/D group) were positive for anti–citrullinated fibrinogen. Eighty-three percent of the RA patients in group A/D (single-dose, HLA–DRB1*0404) had serum anti– citrullinated fibrinogen, a significantly higher proportion than in the D/D group. For other RA-associated HLA–DR alleles there was only a trend toward an association. Our fibrinogen peptide binding results are con- HLA–DR AND ANTI–CITRULLINATED FIBRINOGEN IN RA 3431 Figure 6. Proliferative responses of peripheral blood mononuclear cells (PBMCs) from rheumatoid arthritis (RA) patients and controls to peptides from the ␣- and ␤-chains of fibrinogen. PBMCs from 12 RA patients and 10 healthy controls were cultured at a density of 106 cells with 2 g/ml of stimulatory peptide. Thirty-seven peptides from the ␣-chain and 43 peptides from the ␤-chain of fibrinogen, which, in native or citrullinesubstituted form, bound at least 1 of the 5 tested HLA–DR alleles, were used for these proliferation studies. Positive responses was defined as an optical density (OD) value higher than twice the OD for cells cultured without peptide. sistent with the fact that RA-associated HLA–DR alleles are not mandatory for production of anti–citrullinated fibrinogen and that only HLA–DRB1*0404 is associated with production of these antibodies. Every HLA–DR allele, either RA-associated or non–RA-associated, bound multiple peptides from fibrinogen. HLA– DRB1*0404 was associated with anti–citrullinated fibrinogen production and was also the allele that bound the most fibrinogen peptides. Thus, it can be argued that because it binds more fibrinogen peptides, HLA– DRB1*0404 is more likely to be involved in the production of anti–citrullinated fibrinogen. Thus, our very extensive peptide binding data do not support the idea that citrullination is necessary for fibrinogen peptides to bind HLA–DR–associated alleles. Indeed, we found that, as a general rule, peptide citrullination did not help the binding of fibrinogen peptides to HLA–DR. It has been proposed that RA-associated HLA–DR alleles, whose peptide binding groove has a very basic P4 pocket (16), could bind citrullinated (neutral) fibrinogen peptides better than native (basic) fibrinogen peptides (17). This hypothesis was supported by the demonstration that arginine-to-citrulline substitution in a single vimentin-derived peptide allowed high-affinity peptide interaction (17). However, data from the present study of a large number of peptides in both native and citrullinated form indicate that citrullination is not a prerequisite for binding of fibrinogen peptides to HLA–DR. In this respect, immunization against citrullinated fibrinogen in RA is different from immunization against gliadin in celiac disease. In the latter case, posttranslational deamidation of gliadin peptides (turning glutamine into glutamic acid) allows specific binding of deamidated gliadin peptide to the celiac disease–predisposing HLA–DQ2 and HLA–DQ8 alleles (18). In the former case, citrullination does not influence peptide–DR interaction. Finally, we analyzed proliferative responses of PBMCs from RA patients and controls to every fibrinogen peptide that bound at least 1 HLA–DR allele. We found that proliferative responses to fibrinogen peptides were common in RA patients and rare in controls. However, they were not specific to RA; responses were also observed in 3 patients with early undifferentiated arthritis and 1 patient with psoriatic arthritis, none of whom were positive for anti–citrullinated fibrinogen (data not shown). Citrullinated peptides were not more efficient than native peptides at triggering proliferation. Our data suggest a tentative model for major histocompatibility complex (MHC) and T cell contribution to the production of IgG anti–citrullinated fibrinogen. HLA–DR alleles bind fibrinogen peptides in their 3432 AUGER ET AL native or citrullinated form. T cells proliferate in response to both native and citrullinated fibrinogen peptides in patients with inflammatory arthritis. In RA, T cells deliver help for production of disease-specific IgG anti–citrullinated fibrinogen, and this phenomenon has to be specific for RA, since anti–citrullinated fibrinogen is specific for this disease. HLA–DRB1*0404 is most likely to be associated with anti–citrullinated fibrinogen production because it binds the most fibrinogen peptides. Finally, citrulline residues are not involved in fibrinogen peptide binding to the MHC or in fibrinogen peptide recognition by T cells. This is consistent with the known importance of these residues in defining B cell epitopes. REFERENCES 1. Gregersen PK, Silver J, Winchester RJ. The shared epitope hypothesis: an approach to understanding the molecular genetics of susceptibility to rheumatoid arthritis. Arthritis Rheum 1987;30: 1205–13. 2. Ollier W, Thomson W. Population genetics of rheumatoid arthritis. Rheum Dis Clin North Am 1992;18:741–59. 3. Simon M, Girbal E, Sebbag M, Gomes-Daudrix V, Vincent C, Salama G, et al. The cytokeratin filament aggregating protein filaggrin is the target of the so called “antikeratin antibodies,” autoantibodies specific for rheumatoid arthritis. J Clin Invest 1993;92:1387–93. 4. Girbal-Neuhauser E, Durieux JJ, Arnaud M, Dalbon P, Sebbag M, Vincent C, et al. The epitopes targeted by the rheumatoid arthritis-associated antifilaggrin autoantibodies are posttranslationally generated on various sites of (pro)filaggrin by deimination of arginine residues. J Immunol 1999;162:585–94. 5. Vossenaar E, Despres N, Lapointe E, van der Heiden A, Lora M, Senshu T, et al. Rheumatoid arthritis-specific anti-Sa antibodies target citrullinated vimentin. Arthritis Res Ther 2004;6:R142-50. 6. Masson-Bessiere C, Sebbag M, Girbal-Neuhauser E, Nogueira L, Vincent C, Senshu T, et al. The major synovial target of the rheumatoid arthritis-specific antifilaggrin autoantibodies are deiminated forms of the ␣ and ␤ chains of fibrin. J Immunol 2001;166:4177–84. 7. Masson-Bessiere C, Sebbag M, Durieux JJ, Nogueira L, Vincent 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. C, Girbal-Neuhauser E, et al. In the rheumatoid pannus, antifilaggrin autoantibodies are produced by local plasma cells and constitute a higher proportion of IgG than in synovial fluid and serum. Clin Exp Immunol 2000;119:544–52. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31:315–24. Vincent C, Serre G, Lapeyre F, Fournie B, Ayrolles C, Fournie A, et al. High diagnostic value in rheumatoid arthritis of antibodies to the stratum corneum of rat oesophagus epithelium so called “antikeratin” antibodies. Ann Rheum Dis 1989;48:712–22. Vincent C, Simon M, Sebbag M, Girbal-Neuhauser E, Durieux JJ, Cantagrel A, et al. Immunoblotting detection of autoantibodies to human epidermis filaggrin: a new diagnostic test for rheumatoid arthritis. J Rheumatol 1988;25:838–46. Chapuy-Regaud S, Nogueira L, Clavel C, Sebbag M, Vincent C, Serre G. IgG subclass distribution of the rheumatoid arthritis specific autoantibodies to citrullinated fibrin. Clin Exp Immunol 2005;139:542–50. Glantz S. How to analyze rates and proportions. In: Glantz S, editor. Primer of biostatistics. New York: McGraw-Hill; 1992. p. 110–54. Porstmann T, Ternynck T, Avrameas S. Quantitation of 5-bromo2-deoxyuridine incorporation into DNA: an enzyme immunoassay for the assessment of the lymphoid cell proliferative response. J Immunol Methods 1985;82:169–79. Auger I, Lepecuchel L, Roudier J. Interaction between heat-shock protein 73 and HLA–DRB1 alleles associated or not with rheumatoid arthritis. Arthritis Rheum 2002;46:929–33. O’Sullivan D, Sidney J, del Guercio MF, Colon S, Sette A. Truncation analysis of several DR binding epitopes. J Immunol 1991;146:1240–6. Reviron D, Perdriger A, Toussirot E, Wendling D, Balandraud N, Guis S, et al. Influence of shared epitope–negative HLA–DRB1 alleles on genetic susceptibility to rheumatoid arthritis. Arthritis Rheum 2001;44:535–40. Hill JA, Southwood S, Sette A, Jevnikar AM, Bell DA, Cairns E. The conversion of arginine to citrulline allows for a high-affinity peptide interaction with the rheumatoid arthritis-associated HLADRB1*0401 MHC class II molecule. J Immunol 2003;171:538–41. Molberg O, Mcadam SN, Korner R, Quarsten H, Kristiansen C, Madsen L, et al. Tissue transglutaminase selectively modifies gliadin peptides that are recognized by gut derived T cells in celiac disease [published erratum appears in Nat Med 1998;4:974]. Nat Med 1998;4:713–7.