Rheumatoid arthritis (RA) is considered to be a prototypical autoimmune disease. However, the autoantigens that play an important role in the development of RA remain unclear. The aim of this study was to investigate whether T cells specific for citrullinated epitopes from self proteins are present in patients with RA.
Peripheral blood mononuclear cells (PBMCs) from 28 RA patients and 18 healthy controls were stimulated with citrullinated or noncitrullinated aggrecan peptide Agg84–103, and proliferative and cytokine responses were assessed using 3H-thymidine incorporation assay, enzyme-linked immunosorbent assay, and intracellular cytokine analysis.
A proliferative response to the citrullinated aggrecan peptide was detected in >60% of RA patients but not in healthy controls. Furthermore, citrullinated aggrecan peptide–stimulated PBMCs from RA patients produced high levels of the proinflammatory cytokine interleukin-17 (IL-17), accompanied by an induction of IL-17+CD4+ T cells. In contrast, PBMCs from RA patients and healthy controls exhibited no response to stimulation with the noncitrullinated aggrecan peptide.
Proinflammatory T cell responses to stimulation with a citrullinated arthritogenic aggrecan peptide were detected in RA patients but not in healthy individuals, suggesting a role for these autoantigen-specific T cells in the pathogenesis of RA. Our results suggest that the lack of response to the noncitrullinated analog peptide not only implicates the citrulline residue in T cell recognition but also highlights the potential value of citrullinated aggrecan peptide–specific responses as biomarkers of RA. To our knowledge, this is the first study to demonstrate the presence of citrullinated antigen–specific T cells in human RA.
Rheumatoid arthritis (RA) is considered a prototypical autoimmune disease. However, the autoantigens that play a key role in the development of RA remain undetermined. There is an increasing body of evidence suggesting that posttranslational modifications of proteins by processes such as citrullination can generate novel epitopes and that these new epitopes may in turn trigger autoimmunity (1). Indeed, recent studies have shown that anti–cyclic citrullinated peptide (anti-CCP) antibodies are not only highly specific for RA (2) but may also be pathogenic (3). Furthermore, B cell depletion resulting from the use of biologic therapies effectively suppresses synovial inflammation (4, 5). These observations implicate citrullinated autoantigen–specific B cells in mediating joint inflammation in RA. However, the underlying immune processes that drive the development of these autoantigen-specific B cells are poorly understood.
Immunogenetic studies have shown that >90% of RA patients have a similar amino acid sequence, commonly known as the RA “shared epitope,” at positions 70–74 of the HLA–DR β-chain (6). Since HLA–DR is a class II major histocompatibility complex (MHC) molecule, the key function of which is antigen presentation to CD4+ T cells, the strong association between the shared epitope and the development of RA suggests that CD4+ T cells play an important role in mediating the disease process. More recently, the results of genome-wide association studies have implicated other T cell–associated genes in the development of RA, reinforcing the importance of T cells in the pathogenesis of the disease (7). However, the antigens that drive T cell activation in RA remain unknown, and the mechanisms by which T cells may mediate the pathologic process are not fully understood. Since anti-CCP antibodies are strongly correlated with shared epitope alleles (2) and since anti-CCP antibodies of the IgG subclass (implying that class switching has taken place) are present in RA patients (8), one possible role of CD4+ T cells in RA pathogenesis is to provide T cell help for citrullinated antigen–specific B cells. Alternatively, CD4+ T cells may be directly involved in the pathologic process through the production of proinflammatory cytokines. It is noteworthy, therefore, that citrullination may increase the affinity of peptide binding to class II MHC molecules expressing the shared epitope (9), thus enhancing the efficiency of antigen presentation.
These observations strongly suggest that citrullinated autoantigen–specific T cells play a role in the pathogenesis of RA. Indeed, data from a recent mouse model of inflammatory arthritis, in which naive mice were immunized with citrullinated fibrinogen, revealed a strong citrullinated fibrinogen–specific T cell response in the arthritic mice (10). However, citrullinated autoantigen–specific T cell responses have not been reported in human RA to date.
The findings of studies using mouse models and human RA patients have implicated aggrecan as a candidate autoantigen (11, 12). Furthermore, the immunodominant, arthritogenic epitopes of aggrecan have been identified (13). In the current study, we showed that T cells specific for a citrullinated arthritogenic aggrecan peptide are present in RA patients but not in healthy controls, providing the first evidence that citrullinated autoantigen–specific T cells may play a key role in the pathogenesis of RA in humans.
PATIENTS AND METHODS
All cases in the study (n = 28) were patients at the rheumatology clinic of a regional teaching hospital, and all were diagnosed according to the American College of Rheumatology (formerly, the American Rheumatism Association) revised classification criteria for RA (14). Anti-CCP antibody levels were assessed at a clinical immunology laboratory (Royal Victoria Infirmary, Newcastle upon Tyne, UK) using the Diastat assay (Axis-Shield, Dundee, UK). Control samples were obtained from healthy volunteers at Newcastle University (n = 15) or from buffy coats stored at the Newcastle Blood Transfusion Centre (n = 3). The study was approved by the Newcastle and North Tyneside Local Research Ethics Committee, and all participants provided written informed consent before samples were collected. Data on clinical characteristics of the patients are available from the author upon request.
Culture medium and reagents.
RPMI 1640 medium supplemented with 3 mML-glutamine, 50 μM 2-mercaptoethanol, 30 μg/ml gentamicin, and 5% human antibody serum was used (all from Sigma-Aldrich, Gillingham, UK). All cells were incubated at 37°C in a humidified incubator containing 95% air and 5% CO2.
Citrullinated and noncitrullinated peptides were synthesized according to the peptide sequence of the human aggrecan Agg84-103, using a commercially available synthesis service (GenScript, Piscataway, NJ). (The arginine/citrulline substitution is underlined in the following: VVLLVATEGR/CitVRVNSAYQDK.) Heat-inactivated Candida hyphae (1.1 μg/ml protein equivalent of 1 × 105 colony- forming units/ml), a gift from Dr. Desa Lilic (Newcastle University), was used as a positive control antigen for stimulating interleukin-17 (IL-17) and interferon-γ (IFNγ) production.
Peripheral blood mononuclear cells (PBMCs) were isolated from whole blood samples using density-gradient centrifugation. PBMCs (2 × 105/well, in triplicate) were incubated in 96-well plates with 10 μg/ml of citrullinated or noncitrullinated aggrecan peptide for 9 days. During the final 18 hours of culture, supernatant to be used in measuring cytokine production was collected and stored at −40°C. To the remaining culture, medium containing 3H-thymidine was added (final concentration 14.8 kBq/well). Cells were harvested on glass fiber membranes after 12–16 hours, and radioactivity was quantified using a Microbeta TriLux scintillation counter (PerkinElmer, Emeryville, CA). The stimulation index was calculated as counts per minute with antigen/counts per minute without antigen. Response to phytohemagglutinin (5 μg/ml) (Sigma-Aldrich), measured on day 3, was used as a positive control. A stimulation index of >2 was considered a positive response.
Intracellular cytokine analysis.
PBMCs were stimulated either with R/Cit93-substituted Agg84-103 peptides or with Candida hyphae. Then, cells were treated with phorbol myristate acetate (50 ng/ml) and ionomycin (1 μg/ml) for 5 hours, the last 4 of which were in the presence of 5 μg/ml brefeldin A (all from Sigma-Aldrich). Nonspecific Fcγ receptor–mediated binding was blocked by incubation with human IgG. Cells were incubated with peridinin chlorophyll A protein–Cy5.5–conjugated anti-human CD4 (BD Biosciences, Oxford, UK). After permeabilization and blocking of nonspecific binding with mouse IgG and rat serum, cells were incubated with Alexa Fluor 647–conjugated anti-human IL-17 and Pacific Blue–conjugated anti-human IFNγ according to the instructions of the manufacturer (eBioscience, San Diego, CA). Cells were analyzed using an LSR II 18-color flow cytometer (BD Biosciences), and 10,000 events in the CD4+ T cell gate were collected for each sample. Analysis was performed using FACSDiva software, version 5 (BD Biosciences).
Measurement of cytokine production in culture supernatant.
Concentrations of IL-17A and IL-22 were determined using a Ready Set Go kit (eBiosciences) and enzyme-linked immunosorbent assay (Emelca Bioscience, Breda, The Netherlands), respectively, according to the manufacturers' instructions. Cytokine concentrations in individual samples were assessed by the 4-parameter Boltzmann sigmoidal curve-fitting method using GraphPad Prism, version 5.01 (GraphPad Software, San Diego, CA). Th1/Th2 cytokines (IFNγ, IL-2, IL-4, IL-5, IL-10, IL-12, and IL-13) were quantified simultaneously using electrochemiluminescence multiplex technology according to the instructions of the manufacturer (Meso Scale Discovery, Gaithersburg, MD). Signal was detected using a Sector Imager 2400 (Meso Scale Discovery).
Statistical analysis (descriptive statistics, linear regression, and correlation analysis) was performed using GraphPad Prism, and paired and unpaired 2-tailed t-tests were performed using Excel 2002, Service Pack 3 (Microsoft, Redmond, WA). P values less than or equal to 0.01 were considered significant.
A proliferative response to a citrullinated arthritogenic aggrecan peptide was detected in the majority of RA patients but not in healthy controls. In order to determine the optimal culture duration for the detection of proliferative responses to citrullinated and noncitrullinated aggrecan peptides, we measured the proliferation of PBMCs from 10 RA patients in response to citrullinated and noncitrullinated aggrecan peptide Agg84-103 over a period of 6–12 days. Throughout the culture period, none of the patient samples exhibited a response to the noncitrullinated aggrecan peptide. Strikingly, however, 7 of 10 samples exhibited a response to the citrullinated aggrecan peptide, with maximal response observed on day 9 (data available from the author upon request). We therefore extended our investigation to include a total of 28 RA patients and 18 healthy controls. Consistently, PBMCs from 17 of the 28 RA patients (60.7%) proliferated in response to the citrullinated aggrecan peptide. In contrast, none of the RA samples proliferated in response to the noncitrullinated aggrecan peptide. Furthermore, PBMCs from healthy controls did not proliferate in response to either the citrullinated or the noncitrullinated aggrecan peptide (Figure 1). The responses to phytohemagglutinin, a polyclonal T cell mitogen, were similar in RA patients and in healthy controls (data not shown).
Responses to the citrullinated aggrecan peptide are biased toward proinflammatory cytokines.
Using PBMCs from RA patients, we also examined cytokine production in response to the aggrecan peptide. The level of IL-17 in the culture supernatant of PBMCs that had proliferated in response to the citrullinated aggrecan peptide was significantly increased compared with levels in unstimulated cultures (Figure 1B). This was accompanied by an induction of IL-17+CD4+ cells, as detected using flow cytometry (Figures 2A and C). In contrast, in the culture supernatant of PBMCs that had not proliferated in response to the citrullinated aggrecan peptide, there was only a modest increase in IL-17 levels and no increase in IFNγ levels, compared with levels in unstimulated cultures (Figure 1), and there was no induction of either IL-17+CD4+ cells or IFNγ+CD4+ cells (Figure 2). Furthermore, there was no significant increase in the production of IL-17 or IFNγ in PBMCs that did or those that did not respond to stimulation with the noncitrullinated aggrecan peptide. Production of IL-17 and IFNγ in response to Candida, a control antigen, was similar in PBMCs that did and those that did not respond to the citrullinated aggrecan peptide (data available from the author upon request). Furthermore, the production of IL-22 was increased in PBMCs that responded to the citrullinated aggrecan peptide (Figure 3A). The levels of other cytokines (IL-2, IL-4, IL-5, IL-10, IL-12, and IL-13) in PBMC cultures did not change significantly after stimulation with either the citrullinated or the noncitrullinated aggrecan peptide (data not shown).
Correlation of citrullinated aggrecan peptide–specific proliferative responses with IL-17 production and anti-CCP titer.
Since our data showed that the production of IL-17 in PBMCs from RA patients was more prominent in those that exhibited a response to the citrullinated aggrecan peptide, we performed linear regression and correlation analysis to clarify the relationship between the proliferative response and the IL-17 response. We found that proliferation in response to the citrullinated aggrecan peptide strongly correlated with production of IL-17 (r = 0.724, P = 0.001) (Figure 3B). Taken together, our data suggest that Th17 cells were the cellular subset in PBMCs that were responsible for the citrullinated aggrecan peptide–specific responses in RA patients.
We also investigated whether proliferation was correlated with other clinical parameters, including anti-CCP titer, C-reactive protein level, erythrocyte sedimentation rate, disease duration, and age. Only anti-CCP titer was found to be moderately correlated with the proliferative response (r = 0.512, P = 0.009) (data not shown).
In the current study, we have demonstrated that PBMCs from ∼60% of RA patients proliferated in response to the citrullinated analog of a major arthritogenic epitope of aggrecan, which was accompanied by increased production of IL-17 in the culture supernatant and by the induction of IL-17+CD4+ T cells. The lack of response to the corresponding noncitrullinated aggrecan peptide indicated that the citrulline residue is integral to the antigen specificity of the observed immune response. Although purified T cells were not used in our experiments, the use of a 20-mer peptide in these assays, together with our observation that CD4+IL-17+ T cells were induced following stimulation with the citrullinated aggrecan peptide, strongly suggests that the citrullinated aggrecan peptide–specific responses were driven by T cells. To our knowledge, this is the first study demonstrating the presence of citrullinated autoantigen–specific T cells in patients with RA.
Aggrecan is a proteoglycan which forms a major structural component of articular cartilage. Data from animal models have implicated aggrecan in the pathogenesis of RA, since immunization with aggrecan elicited aggrecan-specific T cell and B cell responses and induced inflammatory arthritis in susceptible mouse strains (11). However, although immune responses directed against aggrecan peptides have been reported in RA patients (12), similar responses were also observed in patients with other forms of arthritis, such as ankylosing spondylitis and osteoarthritis (15). Indeed, in one study, the immune response to aggrecan peptides was more readily detectable in healthy controls than in patients with RA (16). In contrast, in our study, a citrullinated aggrecan peptide–specific response was detected exclusively in patients with RA. Therefore, we have identified not only a novel candidate, RA-specific autoantigen that could play an important role in the pathogenesis of RA, but also a potentially useful biomarker of RA.
Although it is tempting to speculate on a pathogenic role for citrullinated aggrecan peptide–specific T cells, further investigations are needed to define their role and clinical significance in RA. In the current study, we found that citrullinated aggrecan peptide–specific T cells produced proinflammatory cytokines IL-17 and IL-22, raising the possibility that these cells may be directly involved in joint and cartilage inflammation and damage. Indeed, synovial fluid mononuclear cells from 2 of 3 RA patients also proliferated in response to stimulation with the citrullinated aggrecan peptide (von Delwig A, Ng W-F: unpublished observations). Another possible role for the citrullinated aggrecan peptide–specific T cells could be to provide help for citrullinated aggrecan–specific B cells. In this regard, we found that anti-CCP titer was modestly correlated with the magnitude of citrullinated aggrecan peptide–specific proliferative responses, although citrullinated aggrecan peptide–specific responses were observed in some anti-CCP–negative patients and were not observed in some anti-CCP–positive patients. Additionally, it is unclear whether the commercially available assay that we used to measure anti-CCP antibodies was capable of detecting anti–citrullinated aggrecan antibodies. Direct measurement of anti–citrullinated aggrecan antibodies in the sera of these patients would be more informative for addressing whether citrullinated aggrecan peptide–specific T cells provide help for citrullinated aggrecan peptide–specific B cell responses.
The reason for the lack of observable responses to the citrullinated aggrecan peptide in ∼40% of RA patients and in all the healthy controls is unclear, but there are several nonmutually exclusive possibilities. First, the affinity of peptide binding to different class II MHC molecules differs; therefore, it is possible that the class II MHC molecules in patients who did not respond to the citrullinated aggrecan peptide failed to efficiently present the citrullinated aggrecan peptide, in favor of another aggrecan peptide or different autoantigens. Therefore, future studies will need to determine whether responses to the citrullinated aggrecan peptide are restricted to particular class II MHC alleles and to investigate the responses to other candidate citrullinated peptides. Second, the frequency of the citrullinated aggrecan peptide–specific T cells might be too low to be detected using the methods we employed in the current study. Third, response to the citrullinated aggrecan peptide may be suppressed (e.g., by Treg cells), which may be particularly relevant in healthy controls.
In conclusion, we have shown that citrullinated aggrecan peptide–specific Th17 cells are present in the peripheral blood of RA patients. Further investigation of the clinical significance of these autoimmune T cells and their role in mediating the inflammatory process in RA may advance our understanding of the pathogenesis of RA, inform the development of novel antigen–specific therapeutic strategies, and identify useful biomarkers of RA.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Ng had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Von Delwig, Robinson, Ng.
Acquisition of data. Von Delwig, Locke.
Analysis and interpretation of data. Von Delwig, Robinson, Ng.
We thank all the patients and volunteers for donating blood samples, and all the medical and nursing staff in the musculoskeletal department for assisting with sample collection.