To investigate the phenotype and function of CD4+ T cells in synovial fluid (SF) from the affected joints of children with oligoarticular-onset juvenile idiopathic arthritis (JIA), and to establish a possible link with disease activity.
To investigate the phenotype and function of CD4+ T cells in synovial fluid (SF) from the affected joints of children with oligoarticular-onset juvenile idiopathic arthritis (JIA), and to establish a possible link with disease activity.
CD4+ T cells were obtained from the peripheral blood (PB) and SF of 23 children with oligoarticular-onset JIA, as well as from the PB of 15 healthy children. The cells were analyzed for the expression of CXCR3, CCR6, and CD161 and for the production of interferon-γ and interleukin-17A (IL-17A). Spectratyping and clonotype analyses were performed to assess different T cell subsets.
The numbers of CD4+CD161+ cells showing either the Th1 or the Th17/Th1 phenotype were higher in the SF than in the PB of children with JIA. The few Th17 cells from JIA SF underwent a spontaneous shift to the Th1 phenotype in vitro, whereas Th17 cells from the PB of healthy children shifted only in the presence of JIA SF; this effect was neutralized by antibody blockade of IL-12 activity. Spectratyping and clonotype analyses showed a similar skewing of the T cell receptor Vβ repertoire in both CD161+ Th17 cells and CD161+ Th1 cells derived from the SF of the same JIA patient. The frequencies of CD4+CD161+ cells, particularly the Th17/Th1 cells, in the JIA SF positively correlated with the erythrocyte sedimentation rate and levels of C-reactive protein.
These findings suggest that a shifting of CD4+CD161+ T cells from Th17 to the Th17/Th1 or Th1 phenotype can occur in the SF of children with oligoarticular-onset JIA, and indicate that the accumulation of these cells is correlated with parameters of inflammation. Thus, the results support the hypothesis that these cells may play a role in JIA disease activity.
A recently described subset of CD4+ T helper cells, referred to as Th17 cells, has been suggested to be pathogenic in several murine models of chronic inflammatory disorders, such as experimental autoimmune encephalomyelitis (1), collagen-induced arthritis (2), and inflammatory bowel disorders (3–8). Th17 cells have also been described in humans, and these cells have been found to be at least partially different from murine Th17 cells (9, 10).
Of particular note, human Th17 cells express molecules that are distinct from those expressed by Th1 cells, such as interleukin-23 receptor (IL-23R) and retinoic acid receptor–related orphan receptor C (RORC). However, there are also other molecules that are shared with Th1 cells, such as IL-12Rβ2 and T-box protein 21 (10). Moreover, in IL-17A–producing human cells, at least part of the cells were found to also produce interferon-γ (IFNγ) (referred to as Th17/Th1 cells), and both Th17 and Th17/Th1 cells exhibit plasticity toward Th1 cells after culturing in the presence of IL-12 (10). Furthermore, human Th17 cells originate in response to stimulation from IL-1β and IL-23, without any critical need for transforming growth factor β (11). More recently, we and other investigators (12, 13) have shown that virtually all human memory Th17 cells are contained within the CD161+ cell fraction of circulating and tissue-infiltrating CD4+ T cells, and that they originate from CD161+ precursors present in umbilical cord blood and newborn thymus (12, 14).
Juvenile idiopathic arthritis (JIA), the most common form of persistent arthritis in children, is a broad term describing a clinically heterogeneous group of arthritides of unknown cause, the onset of which typically begins before age 16 years (15). Several immunologic abnormalities have been characterized both in patients with rheumatoid arthritis (RA) (16) and in children with JIA (17, 18). Many of these abnormalities are common to both diseases, whereas others are related to a specific disease subtype. The synovial membrane shows pronounced infiltration of mononuclear cells, including T cells, B cells, macrophages, dendritic cells, and plasma cells (19, 20). T cell infiltrates predominantly consist of Th1-skewed cells, which were thought to have a central role in disease pathogenesis (21). After the association between Th17 cells and several human autoimmune diseases, including RA, was observed, some studies found increased levels of IL-17A, as well as the presence of Th17 cells and the expression of their transcription factor, RORC, in the synovial fluid (SF) of children with JIA (22–26).
In this study, we assessed the phenotypic and functional features of SF CD4+ T cells from the affected joints of children with oligoarticular-onset JIA, and compared these features to those of circulating CD4+ T cells from the same patients and to those of peripheral blood mononuclear cells (PBMCs) from healthy children. We found an accumulation of CD4+CD161+ T cells producing both IL-17A and IFNγ, which probably originated from the shift of Th17 cells to the production of IFNγ, in the SF of children with JIA, and this finding correlated with parameters of disease activity, suggesting that these cells play a major role in the inflammatory processes of JIA.
Samples of PB and SF were obtained from 23 children with oligoarticular-onset JIA, whose diagnosis was made in accordance with the International League of Associations for Rheumatology classification criteria for JIA (27). Patients with systemic-onset disease, polyarticular JIA, or enthesitis-related arthritis were excluded, in order to assess a homogeneous patient group. The patients' demographic and clinical characteristics are described in Table 1. As a control, PB samples from 15 healthy age- and sex-matched donors were also assessed. Additional PB samples were obtained from 9 adult healthy volunteers. The erythrocyte sedimentation rate (ESR) (normal values up to 30 mm/hour; Test1) and the C-reactive protein (CRP) serum levels (normal values up to 0.5 mg/dl; assessed using the particle-enhanced turbidimetric immunoassay method) were evaluated by staff in the service laboratory of the Anna Meyer Pediatric Hospital in Florence, Italy. The procedures followed in the study were in accordance with the ethics standards of the Regional Committee on Human Experimentation.
|Patient/sex/age||Therapy||ESR, mm/hour||CRP, mg/dl|
The medium used for cultures was RPMI 1640 (Seromed), supplemented with 2 mML-glutamine, 1% nonessential amino acids, 1% pyruvate, 2 × 10−5M 2-mercaptoethanol (all from Gibco), and 10% fetal calf serum. Monoclonal antibodies (mAb) to recombinant IL-12, CXCR3, and CCR6 and an anti–IL-12 neutralizing mAb (MAB219, IgG1, clone 24910; the antibody does not neutralize the bioactivity of IL-23) were from R&D Systems.
PBMCs from healthy donors and PBMCs and SF mononuclear cells (SFMCs) from patients with JIA were analyzed by flow cytometry, using a Becton Dickinson BDLSR II flow cytometer, to assess the production of intracellular cytokines and expression of surface molecules, as previously described (10, 12). In brief, the PBMCs from 3 healthy donors and the PBMCs and SFMCs from 3 patients with JIA were stimulated with phorbol myristate acetate (PMA) plus ionomycin (Sigma), and stained with the following conjugated mAb: anti-CD3–pacific blue, CD4–phycoerythrin (PE)/Cy7, CD8–allophycocyanin (APC)/Cy7, CD161-PE, IFNγ–fluorescein isothiocyanate (FITC) (all from Becton Dickinson), and IL-17–APC (eBioscience). The cells were then analyzed and sorted using a FACSAria, into subsets of CD3+ CD4+CD8−CD161+IL-17+IFNγ−, CD3+CD4+CD8− CD161+IL-17−IFNγ+, and CD3+CD4+CD8−CD161−IL-17−IFNγ+ cells. The sorted cells were used for determination of RORC messenger RNA (mRNA) by quantitative reverse transcription–polymerase chain reaction (RT-PCR) (10, 12).
For assessments of cytokine secretion, the PBMCs from 8 healthy donors and the PBMCs and SFMCs from 6 patients with JIA were stimulated with PMA plus ionomycin. Stimulated cells were recovered, washed, and stained with catch reagents for IFNγ and IL-17 (Miltenyi Biotec), in accordance with the manufacturer's instructions. Following an additional 45 minutes of incubation (at 37°C in 5% CO2), the cells were stained with the following conjugated mAb: anti-CD3–pacific blue, CD4-PE/Cy7, CD8-APC/Cy7, CD161-PE, IL-17–APC, and IFNγ-FITC. The cells were then analyzed and sorted using a FACSAria, into subsets of CD3+CD4+CD8−CD161+IL-17+IFNγ− and CD3+CD4+CD8−CD161+IL-17−IFNγ+ cells. The recovered cells were cultured for 1 week and then analyzed for their ability to produce IL-17 and IFNγ, after stimulation with PMA plus ionomycin.
CD3+CD4+CD8−CD161+IL-17+IFNγ− and CD3+CD4+CD8−CD161+IL-17−IFNγ+ cells derived from the PB of 2 healthy donors and from the SF of 2 patients with JIA were cloned under limiting dilution (0.3 cell/well), as previously reported (10, 12). In addition, CD3+CD4+ CD8−CD161+IL-17+IFNγ+ cells from 1 of the 2 healthy subjects and from the 2 patients with JIA were cultured under limiting dilution. The cloning efficiency for all cell populations ranged between 20% and 28%. CD3+CD4+CD8− CD161+IL-17+IFNγ− cells derived from the PBMCs of healthy donors were cultured in vitro, in the presence or absence of IL-12 (2 ng/ml), JIA SF (2% volume/volume; obtained by pooling the SF samples from 21 randomly selected patients with JIA), and JIA SF plus anti–IL-12 mAb (10 μg/ml). On day 7, the cells were recovered and assessed for their ability to produce IL-17 and IFNγ, after stimulation with PMA plus ionomycin.
TaqMan RT-PCR was performed as described elsewhere (10, 12). From the fixed cell subsets of CD3+CD4+CD8−CD161+IL-17+IFNγ−, CD3+CD4+CD8−CD161+IL-17−IFNγ+, and CD3+ CD4+CD8−CD161−IL-17−IFNγ+ cells, RNA was extracted using an RNeasy formalin-fixed, paraffin-embedded tissue kit (Qiagen). Primers and probes used in the RT-PCR were from Applied Biosystems.
Analysis of the TCR repertoire was done by evaluating the length of the third complementarity-determining region (CDR3), a well-established PCR-based technique (herein referred to as spectratyping or immunoscope analysis) (28–31). Assessment of the clonotype composition of the peaks expanded in spectratyping analysis was done using clonotype primers derived from the CDR3 sequences of T cell clones. The expanded peaks were resolved, by spectratyping, into a single peak of the same size and same TCR Vβ (TCRBV) family of the peak to be analyzed (sequences of the T cell clones have been submitted to GenBank), as described previously (32–34).
Cytokine levels in the sera and SF of JIA patients were assessed using commercially available ELISA kits. These kits were as follows: for IL-7, a Quantikine HS kit from R&D Systems; for IL-18, an ELISA kit from MBL; for IL-12p40, an ELISA kit from Invitrogen; and for IL-1α, IL-1β, IL-1 receptor antagonist, IL-2, IL-4, IL-6, IL-8, IL-10, IL-15, IL-17A, IL-17F, IL-21, IL-22, IL-23, IFNγ, IFNα, and tumor necrosis factor α (TNFα), a Custom LEGENDplex ELISA kit from BioLegend.
A standard 2-tailed paired t-test was used for statistical analysis. P values less than or equal to 0.05 were considered significant. Pearson's correlation coefficients were used to calculate the correlations.
CD4+ T cells present in the PB and SF of 18 patients with JIA, as well as in the PB of 15 healthy children, were analyzed for the production of IFNγ and IL-17A after polyclonal stimulation. There were no differences in the proportions of IFNγ- or IL-17A–producing cells in the circulating CD4+ T cells between healthy children and children with JIA. Similarly, there were no differences in the proportions of CD4+ T cells able to produce IL-17A alone (Th17 cells) between the PB and the SF of children with JIA. In contrast, the proportions of IFNγ-producing CD4+ T cells (Th1 cells), as well as the proportions of CD4+ T cells producing both IFNγ and IL-17A (Th17/Th1 cells), were significantly increased in the SF of children with JIA, as compared with the PB of the same patients (Figure 1A).
Since human Th1 cells have been found to express CXCR3, whereas Th17 cells mainly express CCR6, and Th17/Th1 cells usually express both CXCR3 and CCR6 (9, 10), the proportions of CD4+ T cells expressing CXCR3 or CCR6 were also assessed. Among CD4+ T cells, both CCR6−CXCR3+ and CCR6+CXCR3+ cells, but not CCR6+CXCR3− cells, were expressed at significantly higher levels in the SF than in the PB of children with JIA, whereas there were no appreciable differences in the expression of these chemokines in the PB between healthy subjects and patients with JIA (Figure 1B).
We then investigated the proportions of CD4+CD161+ T cells in the SF of children with JIA, in comparison with that observed in the PB of the same patients or the PB of healthy children. There was significant enrichment for these cells in the SF when compared with the PB of children with JIA (Figure 1C). CD4+ T cells that were able to produce IL-17A were found almost exclusively in the CD161+ fraction, and the levels of IL-17A–producing CD4+ T cells appeared to be significantly higher in the SF than in the PB of children with JIA. No differences in the levels of IL-17A–producing CD4+ T cells were observed when the PB was compared between children with JIA and healthy children (Figure 1D).
In contrast, IFNγ-producing cells could be detected within either the CD161+ or the CD161− fraction of the PB and SF from healthy children and from children with JIA, with the proportions of IFNγ-producing cells being significantly higher in the JIA SF than in the JIA PB (Figure 1D). CD161− Th1 cells were usually more prevalent than CD161+ Th1 cells in all 3 sources examined (Figure 1D).
In addition, we analyzed the CD4+CD161+ and CD4+CD161− T cell populations for the expression of the early activation marker CD69, as well as for the expression of FoxP3 and CD25, which are two molecules expressed by Treg cells. CD69 was present at very low levels (<1%) in both CD4+CD161+ and CD4+ CD161− cells from the PB or SF of patients with JIA. In contrast, the proportions of CD25+FoxP3+ cells were significantly higher in CD4+CD161− cells than in CD4+CD161+ cells from both the PB (mean ± SEM 4.2 ± 0.3% versus 1.6 ± 0.2% [n = 12]; P < 0.005) and SF (8.5 ± 0.4% versus 1.5 ± 0.2% [n = 12]; P < 0.005) of patients with JIA. Moreover, the proportions of CD161−CD25+FoxP3+ cells, but not those of CD161+CD25+FoxP3+ cells, present in the SF were significantly higher (P < 0.001) when compared to the proportions present in the PB of the same patients.
In previous studies (10, 14, 35), we found that human Th17 and Th17/Th1 cells, as well as CD161+ Th1 cells, express the transcription factor RORC, whereas this factor is absent in CD161− Th1 cells. RORC mRNA was therefore assessed by real-time quantitative RT-PCR in fluorescence-activated cell–sorted CD161+IL-17+IFNγ− (Th17), CD161+IL-17−IFNγ+ (CD161+ Th1), and CD161−IL-17−IFNγ+ (CD161− Th1) cell fractions of circulating CD4+ T cells from healthy children and children with JIA, as well as in the SF of the children with JIA. Levels of RORC mRNA were significantly higher in the CD161+ cell fraction, irrespective of whether these cells produced IL-17A or IFNγ alone, in the circulating CD4+ T cells from both healthy children and children with JIA, and were also significantly higher in the CD161+ cell fraction from JIA SF in comparison with CD161− Th1 cells from all 3 sources (Figure 1E).
In the SF of patients with JIA, the presence of high numbers of RORC+CD161+IFNγ+CD4+ T cells (herein referred to as nonclassic Th1) as well as Th17/Th1 cells, but low numbers of Th17 cells, led us to hypothesize that nonclassic Th1 cells could be derived from an in vivo shifting of Th17 cells to Th1. To verify this hypothesis, CD161+IL-17+IFNγ− (Th17) and CD161+IL-17−IFNγ+ (nonclassic Th1) cell fractions were sorted from the PB of healthy children, as well as from both the PB and the SF of children with JIA. The 2 cell subsets from the 3 sources were cultured for 7 days in IL-2–containing medium and then stimulated with PMA plus ionomycin.
Virtually all IFNγ-producing CD161+ cells from all 3 sources maintained their Th1 phenotype and did not acquire the ability to produce IL-17A in culture. In contrast, although the majority of IL-17A–producing circulating CD161+ cells from both healthy children and children with JIA maintained their Th17 phenotype, and only a fraction of them acquired the capacity to produce IFNγ (Th17/Th1 cells), the majority of IL-17A–producing CD4+ T cells from the SF of children with JIA spontaneously shifted in culture toward the nonclassic Th1 phenotype (Figure 2A).
This observation was confirmed by a T cell cloning experimental approach. Some T cell clones that were derived from Th17 cells from the PB of a healthy child maintained the Th17 phenotype (18%), but the majority of them shifted to the Th17/Th1 phenotype (68%), and a few of them shifted to the Th1 phenotype (10%). In contrast, virtually all Th17 clones derived from the SF of a child with JIA shifted to IFNγ production, acquiring either the Th17/Th1 phenotype (41%) or the Th1 phenotype (53%), whereas shifting of Th1 clones to the production of IL-17A, under conditions favoring IL-17 production in JIA SF (i.e., culturing in the presence of IL-1β plus IL-23), was never observed (Figure 2B and results not shown).
To further support the observation that Th17 cells differentiate in culture toward the nonclassic Th1 phenotype, with a first step consisting of the coproduction of IL-17 and IFNγ (Th17/Th1), we cultured, under limiting dilution, not only CD161+IL-17+IFNγ− (Th17) cells and CD161+IL-17−IFNγ+ (nonclassic Th1) cells, but also CD161+IL-17+IFNγ+ (Th17/Th1) cells derived from the PB of 1 healthy subject or the SF of 1 patient with JIA. As shown in Figure 2C, whereas the majority of Th17 cells from both the PB and SF shifted to the Th17/Th1 phenotype (67% and 66.7%, respectively) and a proportion of them shifted to the nonclassic Th1 phenotype (16% and 29%, respectively), the majority of Th17/Th1 cells from both sources acquired the nonclassic Th1 phenotype (79% and 70%, respectively). Moreover, in this case, the shifting of Th1 clones to the production of IL-17A, in the presence of IL-1β plus IL-23, was never observed (Figure 2C).
In order to identify the possible mechanisms responsible for the potential shifting of Th17 cells to Th1 in the SF of children with JIA, we measured the levels of different cytokines in both the SF and sera of patients with JIA. The majority of these cytokines (IL-1β, IL-2, IL-4, IL-7, IL-15, IL-17A, IL-17F, IL-21, IFNγ, and TNFα) in both the sera and SF had a concentration lower than 10 pg/ml (results not shown). In contrast, levels of IL-1α, IL-6, IL-8, and IL-12 were significantly higher in the SF than in the sera from the same patients (Figure 3A).
Since, among these cytokines, IL-12 is considered to be the most powerful agent in the induction of Th1 polarization, Th17 cells were purified from the PB of healthy subjects and cultured in the presence or absence of IL-12, pooled SF from children with JIA, or the same SF plus an anti–IL-12 neutralizing mAb. Culturing in the presence of either IL-12 or the pooled SF from children with JIA resulted in reduced proportions of IL-17+IFNγ− (Th17) cells and increased proportions of both IL-17+IFNγ+ (Th17/Th1) and IL-17−IFNγ+ (nonclassic Th1) cells. The addition of an anti–IL-12 mAb to the cultures containing JIA SF completely reversed these modulatory effects (Figure 3B), suggesting that the effects were at least partially due to the Th1-polarizing activity of IL-12 (Figure 3C).
The demonstration that Th17 cells from the SF of children with JIA could spontaneously shift in vitro to Th1 cells allowed us to hypothesize that a similar Th17-to-Th1 shifting could also occur in vivo in the SF of children with JIA, which would thus explain the high frequency of CD161+ Th1 cells. To provide support to this hypothesis, we decided to compare the TCRBV repertoire of highly purified CD161+IL-17−IFNγ+ cells with that of CD161+IL-17+IFNγ− cells derived from the SF of a patient with JIA. Spectratyping analysis showed a Gaussian distribution of the TCRBV profiles in PBMCs, whereas a skewing of the TCRBV repertoire was observed in SF T cells as well as in SF-derived CD161+IL-17−IFNγ+ and CD161+IL-17+IFNγ− cultured T cells from the same JIA patient (Figure 4A). The skewing of the repertoire was similar in CD161+IL-17−IFNγ+ cells and CD161+IL-17+IFNγ− cells (BV3, BV11, BV15, BV17, BV21, and BV22). Moreover, in 3 cases, the common repertoire was also detected in SF cells (BV11, BV17, and BV22) (Figure 4A).
These findings were further confirmed by analyzing the composition of the clonotypes of 3 TCRBV families (BV3, BV15, and BV22) that had been found to be expanded (Figure 4B). Clonotype analysis was performed by using primers derived from the CDR3 sequences of 3 T cell clones (clones 192, 206, and 222 in Figure 4B) obtained from the above-described T cell cultures. These clones were chosen because, in the spectratyping analysis, they expressed a single peak of the same size in the same TCRBV family found to be expanded. The clonotyping PCR showed that the sequences of each clone were present at high frequency, independent of the profile of cytokine production (Figure 4B), which thus demonstrates their common cellular origin.
The possibility that the presence of CD4+CD161+ T cells in the SF correlated with disease activity was then investigated. No correlation was found between any type of SF-infiltrating CD4+ T cells and the numbers of joints with disease involvement (results not shown). However, the proportions of CD4+CD161+ cells in the SF directly correlated with the ESR and serum levels of CRP in children with JIA (Figure 5A). Of note, a positive correlation between the proportions of CD161+ T cells producing both IL-17A and IFNγ and the ESR and levels of CRP was also found (Figure 5B). Moreover, no correlation was observed between the numbers of CD161− Th1 cells, CD161+ nonclassic Th1 cells, or CD161+ Th17 cells and the ESR and levels of CRP in children with JIA (n = 18) (in CD161− Th1 cells, R2 = 0.013 and R2 = 0.0022, respectively; in CD161+ nonclassic Th1 cells, R2 = 0.012 and R2 = 0.0025, respectively; in CD161+ Th17 cells, R2 = 0.06 and R2 = 0.07, respectively).
The results of this study demonstrate an accumulation of CD4+ T cells that have the capacity to produce IFNγ (Th1 cells) or both IFNγ and IL-17A (Th17/Th1 cells), but not IL-17A alone (Th17 cells), in the SF of children with oligoarticular-onset JIA, when compared to cells from the PB. These findings are partially in agreement with the results reported in other investigations (22–24), in which increased levels of IL-17A, IL-17A–producing cells, and RORC have been demonstrated in the SF of children with JIA. In addition, we also observed an enrichment for CD4+CD161+ T cells in the JIA SF, the majority of which exhibited either the Th17/Th1 phenotype or the nonclassic Th1 (RORC+IFNγ+IL-17−) phenotype.
In order to explain the minimal presence of Th17 cells in the SF of children with JIA, as well as the prevalence of Th17/Th1 cells and, even more, of Th1 cells (part of which expressed CD161), we hypothesized that a shifting of Th17 cells to the production of IFNγ could occur in the SF microenvironment. Accordingly, while the majority of purified Th17 cells from the PB of both healthy children and children with JIA maintained their cytokine profile after in vitro culture, virtually all of the cells present in the JIA SF spontaneously shifted, under the same experimental conditions, to the Th1 profile.
Of note, we found significantly increased levels of IL-12 in the SF of JIA patients when compared to the plasma levels of IL-12 in the same patients. Moreover, Th17 cells from the PB of healthy children were induced to shift to Th1 cells when cultured in vitro in the presence of JIA SF, to an extent similar to that induced by culturing of the cells in the presence of recombinant IL-12. More importantly, this effect was completely reversed by a neutralizing anti–IL-12 mAb, strongly suggesting that the shift was related mainly to the activity of IL-12 present in the SF. Accordingly, a trend toward a positive correlation, although not statistically significant (P < 0.1), between SF levels of IL-12 and the proportions of CD161+ Th17/Th1 cells or CD161+ Th1 cells was observed. In contrast, culturing CD161+ Th1 cells in the presence of IL-1β plus IL-23, both of which are cytokines involved in the differentiation of Th17 cells (12), never induced the cells to shift to the production of IL-17A (results not shown). These findings are in agreement with our previous results obtained by assessing the shifting ability of human Th17 and Th1 clones that had been generated from the intestinal mucosa of patients with Crohn's disease (10).
Finally, and most importantly, spectratyping analysis demonstrated that both Th17 cells and nonclassic Th1 cells in the SF exhibited a similar TCRBV repertoire, despite their different cytokine profiles. This finding was strengthened by our assessment, at the clonotype level, of 3 TCRBV families that were found to be expanded. These findings, together with the demonstration that CD4+CD161+ T cells that produce both IL-17A and IFNγ can only originate from Th17 cells, but not from Th1 cells, strongly suggest that CD161+ Th17/Th1 cells, as well as CD161+ Th1 cells, present in the SF of children with JIA are derived from CD161+ Th17 cells. The late plasticity of Th17 cells to Th1 cells was recently observed in a mouse study, in which the effect was found to be related to the activity of IL-12, or the prolonged exposure to IL-23, on Th17 cells (36). Furthermore, similar results were recently reported by Nistala et al (37), who showed Th17 plasticity to Th1 driven by the inflammatory environment in human autoimmune arthritis. Very recently, the instability of the Th17 phenotype has been definitively demonstrated, at a genetic level, in mice (38).
Another interesting finding emerging from this study was the demonstration that the proportions of total CD4+CD161+ T cells, as well as that of CD4+CD161+ T cells producing both IL-17A and IFNγ (Th17/Th1 cells), present in the SF of affected joints correlated with the ESR and serum levels of CRP in the same patients. This suggests that CD4+CD161+ T cells and CD4+CD161+ Th17/Th1 cells may play a role in the activity of the disease. Moreover, the demonstration that CD4+CD25+FoxP3+ T cells are significantly enriched in the CD161− cell fraction, in comparison with that in the CD161+ cell fraction, is consistent with this interpretation.
At present, the reason that CD4+CD161+ T cells show increased expression in the SF of JIA patients is unclear. One possibility is that naive CD4+CD161+CCR6+ T cells that behave as precursors of Th17 cells (12) are recruited in the joints at early stages of the disease, wherein they differentiate into Th17, Th17/Th1, and, finally, Th1 cells, in response to environmental cytokines. Another possibility is that the memory-type subpopulation of CD4+CD161+CCR6+ Th17 cells are recruited to the inflamed joints, wherein they differentiate into Th17/Th1 cells and then into Th1 cells, in response to IL-12. Consistent with both of these hypotheses, one study showed that CCL20, the natural ligand of CCR6, is produced by SF monocytic cells and accumulates in the SF of JIA patients (39).
The respective role of Th17, Th17/Th1, and Th1 cells in the pathogenesis of chronic inflammatory disorders is still debated. It has indeed been shown that Th17 cells can promote pancreatic inflammation, but they only induce type 1 diabetes mellitus (DM) efficiently in lymphopenic mice after their conversion into Th1 cells (40). Accordingly, highly purified Th17 cells from BDC2.5NOD mice shift into Th1-like cells in NOD/SCID recipient mice. More importantly, the development of type 1 DM was prevented by treatment with anti-IFNγ–specific antibodies, but not with anti–IL-17A–specific antibodies (41). Moreover, in the murine model of the autoimmune disorder known as proteoglycan-induced arthritis, Th1 cells, but not Th17 cells, are pathogenic (42).
The results of this study strongly support the concept that among CD4+ T cells, the subset of CD161+ T cells are the most important in maintaining the inflammatory process, and that within this cell subset, a shifting may occur from Th17 to Th17/Th1 cells, and even to nonclassic Th1 cells, which is probably related to the activity of IL-12, even if other, still-unidentified cytokines present in the inflammatory microenvironment might be involved. Thus, it is reasonable to suggest that this family of CD161+ T cells is different from classic CD161− Th1 cells, probably because of its ability to produce other proinflammatory components that still need to be identified.
The debate on the respective role of Th17, Th17/Th1, and Th1 cells in the pathogenesis of JIA is particularly important for the development of possible biologic therapeutic strategies. In this view, it has recently been suggested that the neutralization of IL-17 could be a potential novel goal for the treatment of immune-mediated arthritides, by showing that a humanized anti–IL-17 mAb added to oral disease-modifying antirheumatic drugs improved the signs and symptoms of RA in patients (43). In contrast, ustekinumab, a humanized mAb that inhibits receptor-binding of IL-12 and IL-23, reduced the signs and symptoms of psoriatic arthritis (44), suggesting that both Th1 and Th17 cells cooperate in disease pathogenesis. Our results support this latter observation and identify CD161 as a link between Th17, Th17/Th1, and part of Th1 cells. This finding provides the opportunity to interfere, at the same time, with all of the cell populations, whose presence in the SF from affected joints seems to be related to biologic parameters of disease activity, thus prompting us to hypothesize that targeting of CD161 may represent a possible approach for the treatment of JIA.
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. Romagnani 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. Cosmi, Cimaz, L. Maggi, Santarlasci, Liotta, E. Maggi, Romagnani, Annunziato.
Acquisition of data. L. Maggi, Santarlasci, Capone, Borriello, Frosali, Querci, Simonini, Barra, Piccinni.
Analysis and interpretation of data. Cosmi, Liotta, De Palma, E. Maggi, Romagnani, Annunziato.