Abstract
Objective. To analyze γδT cell subsets in peripheral blood (PB) and synovial fluid (SF) of patients with juvenile idiopathic arthritis (JIA), and to correlate γδT cell subsets with clinical characteristics.
Methods. γδT cell subsets as percentages of CD3+ T cells in samples of PB (n = 25) and SF (n = 93) were analyzed by flow cytometry in 93 JIA patients. The percentage of Vγ9+ γδT cells after 10 days of in vitro expansion with either interleukin 2 (IL-2) or isopentenyl pyrophosphate (IPP) plus IL-2 was determined.
Results. Both Vδ1+ and Vγ9+ γδT cell subsets were detected in SF of all patients, but only the percentage of Vδ1+ cells was higher in SF compared to PB (p < 0.01). The distribution of γδT cell subsets was similar in different JIA subgroups, whereas antinuclear antibody (ANA)-positive patients had a higher percentage of SF Vδ1+ T cells than ANA-negative patients (p < 0.01). The percentage of SF Vδ1+ T cells was inversely associated with age at onset, recurrence of synovitis, and erythrocyte sedimentation rate; and that of SF Vγ9+ T cells was inversely correlated with age at onset and was higher in patients who recovered from disease (n = 15). IPP-induced expansion of SF Vγ9+ T cells correlated with disease remission, whereas the expansion of SF Vγ9+ T cells in media with IL-2 alone was significantly greater in patients with uveitis.
Conclusion. The percentage of Vδ1+ and Vγ9+ γδT cells among the SF T cells and their ability to respond to IPP or IL-2 correlated with specific outcomes of JIA, suggesting their role in the immunopathogenesis of this disease.
- GAMMA DELTA T CELLS
- JUVENILE IDIOPATHIC ARTHRITIS
- SYNOVIAL FLUID
- PEDIATRIC RHEUMATOLOGY
- INTRAARTICULAR CORTICOSTEROID INJECTION
Juvenile idiopathic arthritis (JIA), a clinically heterogeneous group of different disorders characterized by chronic inflammatory arthritis in children, is the most common chronic rheumatic disease of childhood and an important cause of short- and longterm disability1. According to the current International League of Associations for Rheumatology (ILAR) classification, different subtypes of JIA can be distinguished on the basis of clinical manifestations (oligoarthritis, persistent and extended; rheumatoid factor (RF)-positive and RF-negative polyarthritis; systemic, psoriatic, enthesitis-related (ERA), and other undifferentiated arthritis)2. Each subtype is characterized by distinct clinical and laboratory manifestations and prognosis. Although the precise etiology is still unknown, JIA is an autoimmune disease, and T cells are thought to be key players in this process3. Indeed, quantitative and qualitative differences among functionally distinct T cell subsets may be linked to various subtypes of JIA4,5.
In this report, we examine the role of γδT cells, a subset with both proinflammatory and immunoregulatory potential, in JIA. These lymphocytes differ from conventional γδT cells in several ways: they usually do not express CD4 or CD8, and their T cell receptor (TCR) is encoded by γ and δ TCR genes6. There are 2 major γδT cell subsets in humans. The first, variable region gene (V)δ1+, is mostly localized among epithelial cells lining mucosal surfaces. These cells usually express Vγ genes other than Vγ9, and become activated by the stress-induced cell-surface major histocompatibility class I-related proteins MICA and MICB on neighboring epithelial cells7. The second subset expresses a TCR composed for the most part of the Vγ9 and Vδ2 genes and is predominantly found in the peripheral blood (PB) and the lymphatic system. Vγ9+Vδ2+ (here referred to as Vγ9+) cells are activated, in a TCR-dependent manner, by low molecular weight pyrophosphorylated intermediates [e.g., isopentenyl pyrophosphate (IPP)] in the classical mevalonate pathway of isoprenoid synthesis within eukaryotic and prokaryotic cells8. In addition, they respond with much higher affinity to (E)-4-hydroxy-3-methyl-but-2-enyl pyrophosphate (HMBPP), an intermediate in the nonmevalonate pathway of isoprenoid synthesis that is used by prokaryotes exclusively9.
The broad functional repertoire of γδT cells includes in vitro non major histocompatibility complex (MHC) restricted cytotoxic effects against activated tumoral and non-tumoral cells (e.g., activated CD4+ T cells, macrophages, and cultured synoviocytes), activation of dendritic cells, augmentation of autoantibody production, activation of polymorphonuclear cells, cytokine secretion, and even antigen presentation to TCR αß+ CD4+ T cells10. These properties suggest a central role of γδT cells in the mechanisms driving synovitis in JIA. We describe studies in which we prospectively evaluated the involvement of these unique T cells in different forms of JIA, and associated their percentage among the total T cell population with disease manifestations.
MATERIALS AND METHODS
Subjects included 93 consecutive patients with JIA attending the rheumatology clinic of Safra Children’s Hospital, Sheba Medical Center, Tel Hashomer. All met the ILAR criteria for JIA2 and had active disease, as PB and SF were taken during intraarticular corticosteroid injection (IACSI) of triamcinolone acetonide. SF was obtained from a knee in 84 and from other joints in 9 patients. In 25 patients both SF and peripheral venous blood samples were obtained concomitantly.
The ethical committee of our institute approved the study protocol, and signed informed consent was obtained from parents of all patients.
Sex, the age at examination and at JIA onset, presence of psoriasis, location of first joint involved, antinuclear antibody (ANA) and rheumatoid factor (RF) status, history of uveitis and of recently active uveitis, recent medications, and time elapsed since previous IACSI were recorded. Data collected prospectively included time to relapse in the injected joint and any other joint, months of followup after the IACSI, number of IACSI and number of joints injected, and the addition of new disease modifying antirheumatic drugs (DMARD). All patients attended our clinic at least every 4 months during the entire followup period. Clinical remission of medication was defined as absence of arthritis without medications for 12 months11.
Isolation of mononuclear cells (MC)
PBMC were isolated by Ficoll Hypaque (Sigma, St. Louis, MO, USA) density centrifugation as described12,13. MC were collected from the interface after centrifugation, washed 3 times with RPMI-1640 medium (Gibco, Invitrogen), and resuspended in RPMI-1640 supplemented with 10% heat-inactivated fetal bovine serum (FBS; Gibco BRL, UK) with 2 mM glutamine and penicillin-streptomycin solution (100 μg/ml) (Ta’assiot Biologiot, Beit Haemek, Israel).
Characterization of γδT cell subsets by flow cytometry
The immunostaining protocol and subsequent flow cytometric analysis of human T cells was carried out as described12. Mouse fluorochrome-conjugated monoclonal antibodies to human CD3-FITC were from BD Pharmingen, to human Vγ9, Vδ2 from Immunotech, and to human Vδ1 from Endogen (Pierce). We had found a 97.3% correlation of Vγ9 with Vδ2, the 2 genes predominantly coexpressed together in the γδT cell subset, in 45 samples of SF and PB from the patients in preliminary experiments; thus we have reported here only results for Vγ9 expression. The cells were analyzed on a FACSCalibur™ instrument using the Cellquest™ software (BD Biosciences). Viable lymphocytes were defined by their forward scatter/side scatter characteristics. Results are expressed as a percentage of the total CD3+ T cell count. Importantly, the number of the total T cells in samples of cultured cells, as measured in representative samples, increased by a factor of 2–3 fold routinely, showing that percentage of evaluated Vγ9 T cells was due to proliferation rather than death of non-Vγ9 T cells in the cultures (data not shown).
In vitro stimulation
SFMC or PBMC (1 × 106 cells/ml) were cultured in 96-well round-bottom tissue culture plates (Costar) in 200 μl RPMI-1640 medium supplemented with 10% FBS, 2 mM L-glutamine, 100 μg/ml penicillin-streptomycin solution, and 100 IU/ml recombinant human interleukin 2 (rhIL-2; Boehringer-Mannheim), and stimulated by addition of 0.5–4 μg/ml IPP (Sigma Aldrich). Medium was replaced on days 3 and 7 of culture. All cultures were maintained for 10 days, then cells were collected and analyzed by flow cytometry. Results reported here were at the dose of 2 μg/ml IPP, which we determined to yield the maximal proliferative response (data not shown).
Statistical analysis
Analysis of data was carried out using SPSS 11.0 statistical software (SPSS Inc., Chicago, IL, USA). For continuous variables, such as age at time of IACSI and laboratory data, normality of distribution was assessed using the Kolmogorov-Smirnov test (cutoff at p = 0.01). Descriptive statistics were calculated and described as appropriate for distribution. Categorical variables such as sex and presence of uveitis were described using frequency distributions and are presented as frequency (%). Associations between laboratory variables and continuous clinical data were described by calculating the Pearson or Spearman rho correlation coefficients as appropriate. Laboratory data were compared by categorical clinical data using the t-test for independent samples or the Mann-Whitney U-test for dichotomous clinical variables, and one-way analysis of variance (ANOVA) followed by the Bonferroni test or the Kruskal-Wallis test followed by Mann-Whitney U-test for clinical data with more than 2 categories. The chi-square test (exact as needed) was used to assess associations between categorical variables. All tests are 2-sided and considered significant at p < 0.05.
RESULTS
Patient characteristics
The main clinical characteristics of the study cohort are shown in Table 1. Disease was oligoarticular in 57 patients (47 persistent; 10 extended), 12 had RF-negative polyarthritis, 11 the systemic subtype, 5 the psoriatic, 3 ERA, and 5 had the undifferentiated JIA subtype. In addition to the ILAR classification, we performed an independent analysis based on number of joints involved. This revealed that 36 patients had polyarticular, 51 pauciarticular, and 6 monoarticular disease. At the time of T cell studies, 22 patients (23.7%) were treated with methotrexate (MTX), 11 (11.8%) with low-dose systemic corticosteroids, and 11 (11.8%) with anti-tumor necrosis factor agents (combined with MTX in 8).
γδT cell subsets in SF and PB and clinical characteristics
γδT cell subset analysis in SF revealed that levels of Vδ1+ cells were significantly higher in females than males (5.60% ± 4.5% vs 3.36% ± 3.00%, respectively; p < 0.032). The percentage of each of the major subsets of γδT cells was independently inversely correlated with the patient’s age at time of SF aspiration (Vδ1+, r = −0.333, p = 0.002; and Vγ9+, r = −0.241, p = 0.025) and also with the age at onset of JIA (Vδ1+, r = −0.275, p = 0.01; and Vγ9+, r = −0.215, p = 0.048). In the 25 patients in whom SF and PB were obtained simultaneously, Vδ1+ cells were significantly enriched among SF T cells relative to PB (5.06% ± 4.30% vs 1.78% ± 1.34%, respectively; p < 0.01)) compared with the Vγ9+ subset, which was represented equally in SF (5.39% ± 3.78%) and PB (4.66% ± 2.68%) samples (Figure 1).
Analyses in relation to JIA subtypes revealed, nevertheless, that both major subsets of γδT cells were represented in the SF of all JIA subtypes, with no significant differences (Table 2 and Figure 2). The proportions of the 2 major γδ subsets were also not associated with the number of joints involved (i.e., mono, oligo, and polyarthritis; data not shown). In addition, we found that the Vγ9+ T cells could be maintained in vitro for 10 days in the presence of IL-2 and expanded in response to IPP in SFMC cultures derived from patients of all clinical subsets (Figure 2 and Table 2). Interestingly, however, the percentage of Vγ9+ T cells stimulated with IL-2 for 10 days was increased in polyarticular JIA SFMC cultures compared to cultures from systemic JIA patients (6.99% ± 3.89% vs 2.44% ± 1.98%, respectively; p < 0.05). We also found that the percentage of Vγ9+ T cells in SFMC cultures after in vitro IL-2 stimulation was significantly higher in patients who had a history of uveitis (7.02% ± 3.14%), and particularly higher when the uveitis was active during the aspiration of sample (9.10% ± 2.24%) compared to patients without uveitis (4.62% ± 2.98% and 4.72% ± 2.91%; p < 0.008 and p < 0.002, respectively; Figure 3). In addition, the percentage of Vδ1+ T cells was significantly higher in the SF of ANA-positive patients compared to ANA-negative patients (6.44% ± 5.35% vs 3.98% ± 2.87%; p = 0.033). The percentage of Vγ9+ T cells was higher in ANA-positive patients as well, but this difference did not achieve statistical significance (Figure 4).
Relation of γδT cells to disease severity and prognosis
Associations between indices of disease severity, disease outcome, and SF γδT cells in JIA were assessed. The percentage of Vδ1+ T cells was significantly lower in patients with recurrence in any joint following the recent IACSI (4.40% ± 4.10% vs 7.00% ± 4.34%, respectively; p = 0.015), in patients with recurrence in the same joint (4.33% ± 4.22% vs 6.75% ± 4.04%; p = 0.017), and in patients starting DMARD therapy during the followup (3.51% ± 2.54% vs 5.50% ± 4.53%; p = 0.043). Inverse correlations between Vδ1+ cells and the number of IACSI during the followup (r = −0.25, p = 0.022), the number of joints injected (r = −0.291, p = 0.007), and the ESR (r = −0.54, p = 0.025) were also observed. Greater percentage of Vγ9+ T cells in SF was also associated with a better prognosis, since it correlated inversely with the number of joints injected with corticosteroids during followup (r = −0.231, p = 0.032). Most importantly, a higher frequency of SF Vγ9+ T cells, both at start of culture and after expansion on exposure to IPP in vitro, was found in patients who subsequently recovered compared to patients with active disease at the end of the followup (7.48% ± 4.14% vs 4.96% ± 3.61%, p = 0.022; and 43.1% ± 26.6% vs 26.2% ± 23.4%, p = 0.035, respectively). IPP-driven expansion of Vγ9+ T cells was also inversely associated with levels of C-reactive protein (r = −0.609, p = 0.047). We found no association between medication taken by the patient during the IACSI and the measures of γδT cells.
DISCUSSION
The results presented here are the first to demonstrate significant associations between subsets of γδT cells in the SF and clinical characteristics of patients with JIA. The major findings are as follows: (1) both Vδ1+ and Vγ9+Vδ2+ subsets are represented in SF of all subsets of JIA patients; (2) Vδ1+ γδT cells are significantly enriched among SF T cells compared to PB; (3) ANA-positive patients have significantly higher levels of SF Vδ1+ cells than ANA-negative patients; (4) elevated levels of both Vδ1+ and Vγ9+ subsets appear to be associated with specific features of favorable disease outcomes; and (5) the relative expansion of the Vγ9+ subset in vitro in response to IL-2 stimulation is positively associated with the occurrence of clinical uveitis.
Previous studies addressed the issue of γδT cells in JIA14,15,16,17. Although healthy children display a gradual increase of Vγ9+ T cells with age (contrasting with our finding that in SF of JIA patients, the percentage of these cells decreased with age), no differences in the proportion of total γδT cells in the PB of patients and controls were found by some investigators14,18. However, Massa, et al reported that the proportion of γδT cells decreased upon treatment with MTX17. In contrast, Black, et al reported a higher proportion of γδT cells in SF than in PB in a cohort of 39 patients19. Moreover, the highest numbers were found in patients with strong responses to GroEL, Yersinia, and human heat shock protein and, similar to our findings, correlated with lower systemic acute-phase reactions. Doherty, et al studied Jγ rearrangements in SF samples, and concluded that the γδT cells in synovia might be oligoclonal and identical in different joints of the same patient, implying antigenic or superantigenic selection20. These authors also reported a lower expression of the joining region JγP (Jγ1.2), which is required for IPP reactivity, in the SF than in the PB of the patients. Kjeldsen-Kragh, et al found that both Vδ1+ and Vγ9+ cells in the SF displayed increased activation markers; the former were CD45RA+CD69+, whereas the latter expressed the CD45RO isoform, suggesting different levels of activation of the respective subsets15. Although these studies indicated that both subsets of γδT cells are localized in SF in an activated state and differ from PB γδT cells, none systematically analyzed these lymphocytes in different subtypes of JIA.
Our first major finding of comparable percentages of the Vγ9+ and Vδ1+ T cell subsets in the SF in situ suggests that both subsets play a role in the immunopathogenesis of synovitis in all subtypes of JIA (Figures 1 and 2). Nevertheless, the analysis of variance revealed a significantly lower percentage of IL-2-responsive Vγ9+ T cells in systemic JIA, compared to RF-negative polyarticular JIA (Table 2). This suggests that in systemic JIA Vγ9+ T cells are defective in their ability to proliferate in response to IL-2, which may be related to the unique mechanisms underlying systemic JIA, currently perceived as autoinflammatory rather than autoimmune. Indeed, our findings complement recent observations of a decrease of γδT cells in the PB of patients with systemic but not other forms of JIA during disease flares, and suggest that this JIA subtype is associated with a unique Vγ9+ γδT cell response21.
Although the γδT cell subsets were equally distributed among JIA subtypes, ANA-positive patients had a significantly higher percentage of SF Vδ1+ T cells than ANA-negative patients (Figure 4). Vδ1+ T cell activation is, at least in part, driven by interactions with MICA/MICB molecules induced on stressed cells, and synoviocyte MICA has been shown to be functional and induce cytokine secretion from CD4+CD28– SF T cells in patients with rheumatoid arthritis7. This suggests that these molecules could similarly affect SF Vδ1+ T cells in JIA. Further, certain MICA alleles are more highly expressed in JIA patients than in controls, and an association between MICA-5 and systemic lupus erythematosus, the prototypic ANA-positive disease, has also been shown22,23,24. Thus, it is possible that the higher level of Vδ1+ T cells in the synovium of ANA-positive patients may reflect different levels and/or subtypes of MICA molecules in the synovium of ANA-positive versus ANA-negative JIA patients, a topic which will be tested in future studies. In Lyme-induced arthritis, SF Vδ1+ T cells induce the apoptosis of the Lyme-reactive SF CD4+ T cells, while enhancing maturation of dendritic cells, suggesting that Vδ1+ T cells may perform both antiinflammatory and proinflammatory roles in JIA as well25. In our study, however, elevated levels of Vδ1+ T cells were associated with less recurrences of arthritis. Together with the better longterm prognosis of ANA-positive JIA patients, who have higher levels of Vδ1+ SF T cells, this suggests that an increase in levels of these cells may play a regulatory rather than a proinflammatory role in JIA synovium. Of note, we also found higher levels of Vδ1+ cells in female patients; this finding is probably related to the higher prevalence of ANA-positive females versus males. Anterior uveitis is a serious extraarticular manifestation of JIA, occurring in 10%–20% of the patients, mainly in oligoarticular and RF-negative polyarticular subtypes and ANA-positive patients26,27. Patients with uveitis had a higher percentage of IL-2-responsive SF Vγ9+ T cells (Figure 3). Patients with polyarticular JIA, in whom uveitis is more severe, also exhibited an elevated percentage of SF IL-2-responsive Vγ9+ T cells, whereas in systemic JIA, in which uveitis is rare, a low response was observed (Table 2). These findings, suggesting a role of IL-2-responsive Vγ9+ T cells in uveitis in JIA, are consistent with other reports showing that γδT cells activated IL-17-producing uveitogenic CD4+ T cells in an animal model, and that Vγ9+ T cells are found in intraocular fluid of Behçet disease patients with uveitis28.
Finally, recovery from disease was found to be positively correlated with the percentage of SF Vγ9+ T cells and the degree of expansion of the Vγ9+ T cells in response to stimulation with IPP. Although these 2 variables may be related, it should be noted that only a subset of Vγ9+ T cells, i.e., those expressing the Vγ9J1.2 gene rearrangement, can respond to IPP and expand29. Thus, our data suggest that patients whose SF Vγ9+ T cell populations are enriched for Vγ9Jγ1.2 cells have a greater chance of recovery20. A potential beneficial role for IPP-responsive Vγ9+ T cells in the synovium could be attributable to immunoregulatory effects and/or to the ability of IPP-activated Vγ9+ T cells to mediate cytotoxicity against proliferating synovial fibroblasts12. Further, a potentially protective role for Vγ9+ T cells in JIA is consistent with our previous report of an association between a depleted Vγ9+ subset in PB and development of a more severe arthritis in adult patients with Felty syndrome30. Studies to elucidate the mechanisms underlying the association of Vγ9+ T cells with better outcomes of arthritis are clearly warranted. Measurements of Vγ9+ T cells and their response to IPP in vitro may serve as novel prognostic markers in JIA, suggesting that larger-scale studies are indicated to validate these hypotheses.
Our results show for the first time a significant correlation between relative numbers of γδT cell subsets in the SF and their ability to expand in response to IPP or IL-2 and distinct clinical forms of JIA and the prognosis. Studies of these cells might yield important new insights into pathogenetic mechanisms of these diseases as well as new prognostic markers and therapeutic tools.
- Accepted for publication January 13, 2011.
REFERENCES
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
- 10.
- 11.
- 12.
- 13.
- 14.
- 15.
- 16.
- 17.
- 18.
- 19.
- 20.
- 21.
- 22.
- 23.
- 24.
- 25.
- 26.
- 27.
- 28.
- 29.
- 30.