Dr. Grom has received consulting fees, speaking fees, and/or honoraria from Novartis and NovImmune (less than $10,000 each).
The limited role of interferon-γ in systemic juvenile idiopathic arthritis cannot be explained by cellular hyporesponsiveness
Article first published online: 27 OCT 2012
Copyright © 2012 by the American College of Rheumatology
Arthritis & Rheumatism
Volume 64, Issue 11, pages 3799–3808, November 2012
How to Cite
Sikora, K. A., Fall, N., Thornton, S. and Grom, A. A. (2012), The limited role of interferon-γ in systemic juvenile idiopathic arthritis cannot be explained by cellular hyporesponsiveness. Arthritis & Rheumatism, 64: 3799–3808. doi: 10.1002/art.34604
- Issue published online: 27 OCT 2012
- Article first published online: 27 OCT 2012
- Accepted manuscript online: 27 JUN 2012 11:28AM EST
- Manuscript Accepted: 21 JUN 2012
- Manuscript Received: 14 FEB 2012
- NIH. Grant Numbers: P01-AR-048929, R01-AR-059049
Systemic juvenile idiopathic arthritis (JIA) is an autoinflammatory syndrome in which the myelomonocytic lineage appears to play a pivotal role. Inflammatory macrophages are driven by interferon-γ (IFNγ), but studies have failed to demonstrate an IFN- induced gene signature in active systemic JIA. This study sought to characterize the status of an IFN-induced signature within affected tissue and to gauge the integrity of IFN signaling pathways within peripheral monocytes from patients with systemic JIA.
Synovial tissue from 12 patients with active systemic JIA and 9 with active extended oligoarticular JIA was assessed by real-time polymerase chain reaction to quantify IFN-induced chemokine gene expression. Peripheral monocytes from 3 patients with inactive systemic JIA receiving anti–interleukin-1β (anti–IL-1β) therapy, 5 patients with active systemic JIA, and 8 healthy controls were incubated with or without IFNγ to gauge changes in gene expression and to measure phosphorylated STAT-1 (pSTAT-1) levels.
IFN-induced chemokine gene expression in synovium was constrained in active systemic JIA compared to the known IFN-mediated extended oligoarticular subtype. In unstimulated peripheral monocytes, IFN-induced gene expression was similar between the groups, except that lower levels of STAT1, MIG, and PIAS were observed in patients with active disease, while higher levels of PIAS1 were observed in patients with inactive disease. Basal pSTAT-1 levels in monocytes tended to be higher in systemic JIA patients compared to healthy controls, with the highest levels seen in those with inactive disease. Upon stimulation of monocytes, the fold increase in gene expression was roughly equal between groups, except for a greater increase in STAT1 in patients with inactive systemic JIA compared to controls, and a greater increase in IRF1 in those with active compared to inactive disease. Upon stimulation, the fold increase in pSTAT-1 was highest in monocytes from patients with inactive systemic JIA.
Monocytes in patients with active systemic JIA retain the ability to respond to IFNγ, suggesting that the lack of an IFN-induced gene signature in patients with active disease reflects a limited in vivo exposure to IFNγ. In patients with inactive systemic JIA who received treatment with anti–IL-1β, hyperresponsiveness to IFNγ was observed.
Systemic juvenile idiopathic arthritis (JIA) is an autoinflammatory disease characterized by spiking fevers, arthritis, polyserositis, evanescent rash, and lymphadenopathy (1). Evidence suggests that one of the effector cells in the pathogenesis of this disease is derived from the monocyte/macrophage lineage. Patients with active systemic JIA have not only an increased proportion of CD14+ monocytes (2) but also increased numbers of myelomonocytic precursors within the peripheral blood (3, 4).
In addition, one of the feared complications of systemic JIA is development of macrophage activation syndrome (MAS), in which affected patients experience widespread activation and expansion of T lymphocytes and cells of the monocyte/macrophage lineage, leading to unremitting fever, hyperferritinemia, hypertriglyceridemia, hepatosplenomegaly, a consumptive coagulopathy, and cytopenias. One of its prominent characteristics is an expansion of macrophages actively engaged in the hemophagocytosis of multiple cellular elements within the bone marrow and other lymphoid tissue (5, 6). Yet, many of these features can also be seen in severe systemic JIA, in an intermediate phenotype called subclinical, or occult, MAS, thus suggesting that MAS and systemic JIA may be different extremes of the same clinical spectrum (7, 8).
The activated macrophage phenotype can be divided into 2 general groups, M1 and M2, depending on the stimuli being received and the subsequent differentiation pathway (9–11). In this model, M1 macrophage differentiation is mainly driven by interferon-γ (IFNγ), while M2 macrophages result from mostly antiinflammatory signals, such as interleukin-4 (IL-4), IL-10, macrophage colony-stimulating factor, transforming growth factor β, or glucocorticoids. M1 macrophages excel at intracellular pathogen eradication, producing high amounts of IL-12 and tumor necrosis factor α (TNFα). Due to their highly phagocytic properties, M2 macrophages are mostly scavengers that are heavily involved in wound healing and tissue remodeling. In contrast to M1 macrophages, M2 macrophages are a prominent source of IL-10.
Currently, the activated macrophage in systemic JIA and MAS has yet to be characterized in the M1/M2 polar model. Systemic JIA and MAS are both inflammatory conditions, and it is therefore likely that the macrophages therein are of the M1 phenotype. However, the data supporting a role for IFNγ, which is the key facilitator of M1 macrophage differentiation, as a mediator in active systemic JIA is scant. Gattorno et al demonstrated that the serum levels of both IFNγ and interferon-inducible protein 10 (IP-10) were elevated in patients with systemic JIA, regardless of the extent of disease activity (12). In contrast, Lasiglie et al showed that, upon overnight stimulation of peripheral blood mononuclear cells (PBMCs), there was no difference in the absolute number of cells producing IFNγ between patients with active systemic JIA and healthy controls (13). Congruent with this finding, 3 independent gene expression studies have failed to detect an IFN-induced genetic signature within the PBMCs of patients with active systemic JIA (3, 14, 15).
In contrast to these findings in systemic JIA, IFNγ is believed to have a prominent role in the pathogenesis of hemophagocytic lymphohistiocytosis (HLH), which is both a hereditary and an acquired hemophagocytic syndrome that is clinically similar to MAS. IFNγ expression has been found to be highly elevated in the serum of patients with HLH (16–18), and blockade of IFNγ has been suggested as a potential treatment for both the inherited and acquired forms of active HLH (19). To this end, the goals of this study were to characterize the status of an IFN-induced gene signature within the affected tissue and to gauge the integrity of IFN signaling pathways within the peripheral monocytes of patients with systemic JIA.
IFNγ is a pleotropic cytokine that canonically signals via the JAK/STAT signal transducer pathway, a signaling mechanism that is common to the many different cytokines (20). Specifically, IFNγ will lead to the recruitment of JAK-1 and JAK-2 to the intracellular portion of the IFNγ receptor, which then becomes phosphorylated on a tyrosine residue. This allows recruitment, docking, and subsequent phosphorylation of STAT-1. In its active form, STAT-1 will then form a homodimer and translocate into the nucleus in order to bind to an IFNγ-activation site in the promoter region, resulting in the initiation of specific IFN-responsive genes.
Given the inflammatory nature of active systemic JIA, it seems paradoxical that an IFN-induced genetic signature cannot be detected within the PBMCs of patients with active systemic JIA. In this study, we therefore explored the role of IFNγ in the pathogenesis of systemic JIA. Our results confirm that patients with active systemic JIA have a restricted IFN-induced genetic signature, not only in the PBMCs but also in isolated peripheral monocytes. Furthermore, we demonstrate a markedly constrained expression of IFN-responsive genes within the peripherally affected synovial tissue of patients with systemic JIA, when compared to patients with extended oligoarticular JIA, a subtype of JIA that is known to express an IFN-induced genetic signature within the synovial cells (21).
In our explorations of the ability of systemic JIA peripheral monocytes to even respond to IFNγ, we did not find the overall reactivity of the monocytes to be significantly different from that in healthy control peripheral monocytes, which suggests that the absence of an IFN-induced genetic signature is not due to monocyte hyporesponsiveness to IFNγ, but rather may be attributed to limited exposure of the cells to IFNγ. Finally, our findings also demonstrate that the basal IFN signal was increased and the reactivity to IFNγ was augmented in peripheral monocytes from patients with inactive systemic JIA who were receiving an IL-1β–blocking agent, a finding that may have an impact on the ability of these patients to respond appropriately to viral illnesses.
PATIENTS AND METHODS
Peripheral monocytes were obtained from 8 age-matched healthy controls, 6 patients with active systemic JIA (none of whom fulfilled the diagnostic criteria for MAS ), and 3 patients with inactive systemic JIA receiving an anti–IL-1β therapeutic agent. The diagnosis of systemic JIA was determined using the International League of Associations for Rheumatology (ILAR) criteria (23). IFNγ-stimulated peripheral monocytes were used for real-time reverse transcription–polymerase chain reaction (RT-PCR) analyses of gene expression and STAT-1 phosphorylation assays (a complete summary of the demographic and clinical characteristics of the patients in these analyses is shown in Table 1).
|Patient/age, years||Assay||Disease duration, months since diagnosis||Disease status||Active arthritis||Systemic features†||Cytopenias||ESR, mm/hour||CRP, mg/dl||Ferritin, ng/ml||Current medication|
|9/3||FACS||21||Active||Yes||Yes||No||ND||ND||>26,000||Prednisone, anakinra, cyclosporin A|
For real-time RT-PCR analyses of the affected synovial tissue, peripheral biopsy samples from 12 patients with systemic JIA and 9 with extended oligoarticular JIA were procured from the Cincinnati Pediatric Rheumatology Tissue Repository. Affected synovium was identified on the basis of the presence of a significant inflammatory infiltration of the tissue, as described previously (24). Patients with extended oligoarticular arthritis were chosen as a comparison group because IFNγ-producing Th1 lymphocytes are found in abundance within the synovial compartment of patients with this JIA subtype. Furthermore, the increased production of IFNγ within the synovium is associated with a prominent IFN-induced gene signature (21).
For all patients in this study, except for those who contributed synovial biopsy tissue (for whom clinical data were lacking), active systemic JIA was defined as the presence of at least 1 of the following findings (after other potential causes, such as minor infections, had been clinically excluded): an increased erythrocyte sedimentation rate and/or C-reactive protein level, evidence of arthritis, presence of rash consistent with a diagnosis of systemic JIA, and evidence of fever, serositis, or lymphadenopathy.
Written consent was collected for all participants. The study was approved by the Institutional Review Board of Cincinnati Children's Hospital Medical Center.
Real-time RT-PCR studies of peripheral blood monocytes.
PBMCs were isolated from the peripheral blood of all subjects and collected within 1 hour of venipuncture in acid citrate dextrose tubes (BD Biosciences) using the Ficoll gradient method (GE Healthcare). Whole PBMC populations were then divided into 2 subpopulations, CD14+ monocytes (positive selection) and CD14− non-monocytes (negative selection), via an AutoMACS Pro automated magnetic cell sorting system with anti-human CD14 MicroBeads (Miltenyi Biotec). According to the findings from fluorescence-activated cell sorting (FACS) analyses done in our laboratory, we were able to achieve >94% CD14+ cell purity in the CD14+ fraction, and <5% CD14+ cells in the CD14− fraction.
Approximately 1–3 × 106 CD14+ cells (or 3 × 106 CD14− cells), along with 3 ml of RPMI 1640 medium–10% fetal calf serum (Invitrogen), were placed into 2 wells of a 6-well plate (TPP) and then either stimulated with 100 units/ml of human recombinant IFNγ (R&D Systems) or left unstimulated for 3 hours at 37°C. Based on the findings from preliminary dose-response studies, we determined that 100 units/ml of IFNγ offered a substantial response, with the least amount of cytokine possible, for cell stimulation in both the gene expression analyses and STAT-1 phosphorylation assays. After stimulation, the cells were lysed with TRIzol (Invitrogen), and isolation of RNA was carried out in accordance with the manufacturer's instructions.
The amount and purity of the RNA were then analyzed using a NanoDrop ND-1000 spectrophotometer with NanoDrop 1000 software (version 3.7.1; Thermo Fisher Scientific). In total, 1,000 ng of RNA was used to synthesize complementary DNA (cDNA) using an iScript cDNA synthesis kit (Bio-Rad), performed with a Bio-Rad iCycler (version 4.006). Real-time PCRs were performed in 20 -μl volumes, using an iCycler (Bio-Rad) with gene-specific primers and iQ SYBR Green Supermix (Bio-Rad). Messenger RNA copy numbers were normalized against the copy number of the housekeeping gene, GAPDH, of the same sample. Control calculations, conducted in our laboratory, determined that GAPDH gene expression did not change in healthy control or systemic JIA peripheral monocytes upon stimulation with IFNγ.
The following primer pairs were used: for GAPDH, 5′-CAGCCCCAGCGTCAAAGG and 5′-GCTCTCCAGAACATCATCC; for IP-10, 5′-AGTGGCATTCAAGGAGTACC and 5′-ATCCTTGGAAGCACTGCATC; for IFN-inducible T cell α chemoattractant (I-TAC), 5′-GCTATAGCCTTGGCTGTGATAT and 5′-CAGGGCCTATGCAAAGACA; for monokine induced by IFNγ (MIG), 5′-GAGAAAGGGTCGCTGTTCCT and 5′-TTTGGCTGACCTGTTTCTCC; for STAT1, 5′-CTTTGGTTGAATCCCCAGGC and 5′-TGCTCCCAGTCTTGCTTTTCTAAC; for IFN regulatory factor 1 (IRF1), 5′-ATGAGACCCTGGCTAGAG and 5′-AAGCATCCGGTACACTCG; for IL-6, 5′-TTCACCAGGCAAGTCTCC and 5′-ATACTCGACGGCATCTCAG; for IL-1β, 5′-TCCCCAGCCCTTTTGTTGA and 5′-TTAGAACCAAATGTGGCCGTG; for TNFα, 5′-AACTACAGACCCCCCCTGAAAAC and 5′-AAGAGGCTGAGGAACAAGCACC; for protein inhibitor of activated STAT1 (PIAS1), 5′-AAGCACGGACGCAAACACGAAC and 5′-GTGAGTTGTGGAATGGTAGATGGAG; for suppressor of cytokine signaling 3 (SOCS3), 5′-CACTCTTCAGCATCTCTGTCGGAAG and 5′-CATAGGAGTCCAGGTGGCCGTT; and for SOCS1, 5′-TCCCCTTCCAGATTTGACCG and 5′-AAGAGGTAGGAGGTGCGAGTTCAG. The specificities of the primers were determined by measuring the PCR-product length on gel electrophoresis and via the examination of the melting point of the PCR product. Relative gene expression was calculated using the comparative threshold cycle method (25).
STAT-1 phosphorylation studies via intracellular FACS analysis.
CD14+ peripheral blood monocytes from healthy controls, patients with inactive systemic JIA, and patients with active systemic JIA were collected and isolated in the same manner as detailed above. Monocytes were then either left unstimulated or stimulated with 100 units/ml of recombinant human IFNγ (R&D Systems) for 30 minutes at 37°C. After incubation, 2% paraformaldehyde was added for 10 minutes at room temperature to stop the signaling cascade. The cells were permeabilized in 90% methanol overnight, and then stained with anti-CD14 (BD Biosciences) and the following markers of phosphorylation: phycoerythrin-conjugated pY701 (for pSTAT-1), and Alex Fluor 488–conjugated pY641 (for pSTAT-6; used as a negative control) (both from BD Biosciences). The data were acquired on a FACSCanto (BD Biosciences) and analyzed using FlowJo (Tree Star).
Comparisons between the healthy control and systemic JIA patient groups were performed using Student's unpaired t-tests. P values less than 0.05 were considered significant.
Expression of IFN-induced chemokines within systemic JIA and extended oligoarticular JIA synovium.
To explore whether an IFN-induced gene expression signature exists within peripherally affected tissue in systemic JIA, we sought to explore this signature within the inflamed synovial tissue of patients with active systemic JIA. IP-10 (CXCL10) (26), MIG (CXCL9) (27), and I-TAC (CXCL11) (28) are CXC-subfamily chemokines that primarily serve as lymphocyte chemoattractants during inflammation. Each of these proteins will display a highly dramatic fold increase in expression upon stimulation with IFNγ (29, 30), and thus all 3 serve well as markers to gauge the presence of locally produced IFNγ within an inflammatory microenvironment.
Figure 1A depicts the relative gene expression levels of IP-10, MIG, and I-TAC in synovial tissue from patients with active systemic JIA compared with synovial tissue from patients with extended oligoarticular JIA, a disease in which the synovium is characterized by an IFN-induced gene signature. Although we did find an IFN-induced chemokine gene signature within systemic JIA synovial tissue, dramatically lower expression levels of IP-10 (4.5-fold less), MIG (15.8-fold less), and I-TAC (3.6-fold less) were detected in the affected synovium of patients with systemic JIA compared to that of patients with extended oligoarticular JIA (each P < 0.01).
Expression of a restricted IFN-induced genetic signature in peripheral blood CD14+ monocytes.
To test for an IFN-induced gene signature within peripheral monocytes, the levels of IFN-induced gene expression were compared between patients with active systemic JIA, patients with inactive systemic JIA, and healthy controls. In addition to pSTAT-1 directly inducing the expression of I-TAC, MIG, IP-10, and even STAT1 itself, it also induced expression of a secondary transcription factor, IRF1, a factor that serves to further propagate the IFNγ signal by inducing the expression of a second wave of IFN-responsive genes (31). In comparing the unstimulated levels of STAT1, IRF1, I-TAC, MIG, and IP-10 gene expression in peripheral blood monocytes between all groups (Figure 1B), we found that the levels of IP-10 (2.3-fold less; P = 0.22), I-TAC (1.5-fold less; P = 0.47), and IRF1 (2.4-fold less; P = 0.11) were all lower in patients with active systemic JIA compared to healthy controls, and only the levels of MIG (7.3-fold less) and STAT1 (2.3-fold less) were significantly lower in patients with active systemic JIA compared to healthy controls (both P < 0.05). The difference in expression of these 5 genes was only minimal in unstimulated peripheral monocytes from patients with inactive systemic JIA compared to healthy controls.
Within the active systemic JIA group, no appreciable differences in gene expression were seen in those with predominantly active arthritis when compared to those with mainly systemic features; in fact, this study was not sufficiently powered to detect such a difference. Although the chemokines studied herein are mainly expressed within the myelomonocytic lineage, to ensure that we were not missing an IFN-induced signature by failing to look at non-monocytes, we studied gene expression in the CD14− monocyte fraction (a mixture of immature granulocytes, lymphocytes, and natural killer cells). In all 3 groups tested, IFN-induced chemokine gene expression within the CD14− peripheral monocyte fraction was found to be ∼2.5-fold lower than that measured within the CD14+ peripheral monocyte fraction.
Unstimulated STAT-1 phosphorylation within peripheral blood CD14+ monocytes.
Since most of the effects of IFNγ are mediated through the phosphorylation of STAT-1 (20), we wanted to assess whether the isolated cells had been previously exposed to circulating levels of IFNγ in vivo. To achieve this, we measured the unstimulated levels of pSTAT-1 within freshly isolated peripheral blood CD14+ monocytes from patients with active systemic JIA, patients with inactive systemic JIA receiving IL-1β–blocking therapy, and healthy controls. The phosphorylation of STAT-1 was detected by intracellular FACS analysis (Figure 2A). As shown in Figure 2B, unstimulated monocytes from patients with systemic JIA, in both the active disease group and the inactive disease group, had higher amounts of pSTAT-1 compared to healthy controls, as measured by the mean fluorescence intensity (MFI) of pSTAT-1 (mean ± SD MFI 226 ± 137 in active systemic JIA and 351 ± 133 in inactive systemic JIA versus 189 ± 127 in healthy controls; P = 0.69 and P = 0.14, respectively), although these differences were not statistically significant. The amounts of pSTAT-6, used as a negative control, in unstimulated cells were not statistically significantly different between each group (results not shown).
Transcriptional effects of IFNγ on peripheral blood CD14+ monocytes.
After exploring the IFN-induced genetic signature within both peripheral monocytes and CD14− cells in unstimulated conditions, we next studied the effects of IFNγ stimulation on both of these cell populations. As shown in Figure 3, upon stimulation of monocytes with IFNγ, the fold change in gene expression was higher in patients with active systemic JIA compared to healthy controls, in experiments examining expression of STAT1 (mean ± SD fold increase 34.8 ± 33.7 versus 12.5 ± 5.8; P = 0.21), IRF1 (fold increase 131 ± 51.2 versus 83.0 ± 58.5; P = 0.18), IP-10 (fold increase 49,994 ± 38,969 versus 15,488 ± 11,293; P = 0.12), MIG (fold increase 27,301 ± 44,486 versus 2,973 ± 1,486; P = 0.29), and I-TAC (fold increase 2,970 ± 1,572 versus 2,610 ± 1,428; P = 0.70), although the differences were not significant. With the exception of STAT1 induction, the fold change in IRF1, IP-10, MIG, and I-TAC gene expression was similar between patients with inactive systemic JIA and healthy controls. Surprisingly, induction of STAT1 gene expression by IFNγ in monocytes was significantly higher in the inactive disease group compared to healthy controls (mean ± SD fold increase 42.8 ± 3.5 versus 12.5 ± 5.8; P < 0.01).
Among the patients with active systemic JIA, this study was not sufficiently powered to detect a difference in IFN responsiveness between those whose predominant disease manifestation was arthritis compared to those whose disease was primarily characterized by systemic features. Upon stimulation of the CD14− cell fraction by IFNγ in each group, no significant differences in induced gene expression, the magnitude of which was several fold lower than that in stimulated CD14+ cells, were detected (results not shown).
Effects of IFNγ stimulation on STAT-1 phosphorylation within peripheral blood monocytes.
We next examined the effects of IFNγ stimulation of peripheral monocytes on STAT-1 phosphorylation within both the active and inactive systemic JIA patient groups and healthy control group. A representative example of the fold increase in pSTAT-1 levels upon IFNγ stimulation is shown in Figure 2C. Perhaps most interestingly, it was the patients with inactive systemic JIA receiving an anti–IL-1β–blocking therapy who experienced the greatest fold change in pSTAT-1 levels upon IFNγ stimulation of the cells (mean ± SD fold increase 9.0 ± 1.8) when compared to patients with active systemic JIA (fold increase 5.4 ± 1.7; P < 0.05) and healthy controls (fold increase 4.2 ± 0.9; P < 0.01) (Figure 2D). As a negative control, no appreciable increase in pSTAT-6 levels was detected in any of the groups upon stimulation of the cells with IFNγ (results not shown).
Gene expression of negative regulators of activated STAT1.
In an attempt to limit the potentially deleterious effects of an overly exuberant inflammatory response, our immune system employs safeguards to limit IFN signaling. The proteins SOCS1 (32), SOCS3 (33), and PIAS-1 (34) serve as important negative regulators of STAT1 signaling. As such, we next studied whether an increased expression of STAT1 negative regulators could be demonstrated in conjunction with the restricted IFN-induced gene signature and, if not, whether these negative regulators could potentially be induced by the presence of IFNγ. As seen in Figures 4A and B, although SOCS1 gene expression in unstimulated cells was lower in both of the patient groups (active and inactive systemic JIA) when compared to healthy controls, this finding was not statistically significant. Unstimulated SOCS3 expression was not found to be significantly different in either of the systemic JIA patient groups compared to controls. Unstimulated PIAS1 gene expression was significantly lower in patients with active systemic JIA compared to healthy controls (mean ± SD relative gene expression 5,270 ± 1,090 versus 12,260 ± 7,770; P < 0.05), and was significantly higher in patients with inactive systemic JIA (mean ± SD relative gene expression 15,990 ± 1,660) when compared to patients with active systemic JIA (P < 0.01).
Of note, no differences in PIAS1 or SOCS3 gene induction were detected upon IFNγ stimulation of the cells in each group, as shown in Figure 4C. Similarly, although the differences were not statistically significant, the gene induction of SOCS1 was depressed in both the active systemic JIA group (mean ± SD fold increase 98.9 ± 63.8) and inactive systemic JIA group (fold increase 61.0 ± 44.4) when compared to SOCS1 induction in healthy controls (fold increase 168 ± 196; P = 0.38 and P = 0.18, respectively).
Although systemic JIA is an inflammatory disorder in which the monocyte/macrophage lineage appears to be expanded and activated, it is unclear whether this phenomenon is driven by IFNγ. In this study, we provide evidence to suggest that IFNγ plays a limited role in influencing the monocyte/macrophage phenotype in systemic JIA. We demonstrate that the expression of IFN-induced chemokines is markedly limited within both circulating monocytes and the inflamed synovial tissue of patients with systemic JIA when compared to patients with extended oligoarticular JIA, a subtype of JIA in which IFNγ-producing Th1 lymphocytes are abundant within the synovial compartment (21). Furthermore, we demonstrate that this absence of an IFN-induced signature is not due to monocyte hyporesponsiveness toward IFNγ, since systemic JIA peripheral monocytes were just as reactive to IFNγ as were healthy control monocytes.
Taken together, these findings suggest that the lack of an IFN-induced signature within systemic JIA monocytes could be attributed to limited in vivo exposure to IFNγ. Although neither the serum nor synovial levels of IFNγ were directly measured in this study, previous studies have demonstrated that serum and synovial levels of IFNγ were lower in systemic JIA compared to either the oligoarticular or polyarticular JIA subtype (35), which is consistent with a lack of an IFN-induced PBMC gene expression signature in systemic JIA (3, 14, 15).
In contrast to its role in active systemic JIA, there is mounting evidence to indicate that IFNγ may be implicated as a key pathogenic mediator in hemophagocytic syndromes, including MAS secondary to systemic JIA (16–18). Strikingly high levels of IFNγ have been reported in familial HLH (16, 17), and active hemophagocytosis within an animal model of primary HLH, involving the perforin-knockout mouse, requires the presence of IFNγ (36). Although data on serum levels of IFNγ in systemic JIA/MAS are less abundant, a study by Billiau et al demonstrated an expansion of IFNγ-producing lymphocytes in liver biopsy tissue from a patient with systemic JIA/MAS (37). Furthermore, the expression of neopterin, a molecule produced specifically by IFNγ-stimulated macrophages (38), was found to be highly elevated in patients with MAS, in contrast to only mildly increased levels in patients with active systemic JIA (39). In fact, a highly increased level of serum neopterin has been proposed as a potential diagnostic marker for hemophagocytic syndromes in general (40). Based on these observations, the differing role of IFNγ in the pathogenesis of systemic JIA compared to MAS represents a divergence, thus suggesting that these are 2 distinct pathophysiologic phenomena, rather than being opposite ends of the same clinical spectrum, as has been proposed in the literature (41).
Based on the M1/M2 macrophage polarization paradigm, IFNγ is the key driver of M1 differentiation (9–11). Systemic JIA peripheral monocytes do not appear to be exposed to significant amounts of IFNγ in vivo, and thus do not display the typical characteristics of the M1 macrophage. However, their preserved ability to respond to IFNγ suggests that they could acquire this phenotype, given the appropriate microenvironment, such as during an intercurrent viral illness.
Last, and perhaps most interestingly, we demonstrate a paradoxically increased basal IFN signal and increased IFNγ responsiveness within the inactive systemic JIA patient group being treated with an anti–IL-1β agent. In support of this finding, Quartier et al revealed an up-regulation of IFN-related gene expression in patients with systemic JIA treated with anakinra, regardless of whether the patient was considered a treatment responder (42). These observations suggest the existence of a cross-regulation between IL-1β and IFNγ-induced signaling.
Consistent with this idea, IFNγ has, in some experimental systems, been shown to harbor distinct antiinflammatory and immunomodulatory properties (43, 44). For example, in mice with collagen-induced arthritis induced by immunization with type II collagen plus Freund's complete adjuvant (containing heat-inactivated mycobacteria), IFNγ seemed to suppress the excessive expansion of myeloid cells (45). Similarly, blockade of IFNγ in the experimental autoimmune encephalitis model produced a more severe phenotype (46).
In light of the apparent pathogenic role of IL-1β in systemic JIA (15), IFNγ has been reported to decrease the production of IL-1β–induced matrix metalloproteinases in fibroblast-like synoviocytes from human donors with rheumatoid arthritis, and also to reduce IL-1β expression within the synovium in an antigen-induced arthritis model (47). In fact, IFNγ was shown to cause a near global dampening of human macrophage IL-1β–mediated inflammatory cytokine production and tissue destruction, a phenomenon that was believed to be due to a STAT1-dependent down-regulation of IL-1 receptor type I (48). The same group of investigators also found that IFNγ serves as a STAT1-dependent negative regulator of monocyte chemoattractant protein 1 (CCL2)–mediated human monocyte chemotaxis (49). Whether these observations are relevant to systemic JIA is not presently clear, but our finding of increased monocyte reactivity to IFNγ in patients being treated with an anti–IL-1β therapeutic agent deserves further investigation, since it may have an impact on the ability of these patients to appropriately respond to viral illnesses. Since MAS is often triggered by viral infections in patients with systemic JIA, such monocyte hyperresponsiveness to IFNγ may also contribute to the development of this complication.
In summary, we provide evidence to support the concept of a limited role of IFNγ in active systemic JIA, a finding that cannot be explained by peripheral monocyte hyporesponsiveness to IFNγ. Interestingly, based on our findings regarding the phosphorylation of STAT-1, we also show that peripheral monocytes from patients with inactive systemic JIA receiving an IL-1β–blocking agent displayed an increased responsiveness to IFNγ stimulation. Whether this increased reactivity to IFNγ in patients being treated with an anti–IL-1β agent may have an impact on the patients' ability to respond to infections, particularly viral illnesses, or alters their risk for developing MAS needs to be further investigated.
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. Sikora 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. Sikora, Fall, Thornton, Grom.
Acquisition of data. Sikora, Fall.
Analysis and interpretation of data. Sikora, Fall, Thornton, Grom.
- 3Gene expression profiling of peripheral blood from patients with untreated new-onset systemic juvenile idiopathic arthritis reveals molecular heterogeneity that may predict macrophage activation syndrome. Arthritis Rheum 2007; 56: 3793–804., , , , , , et al.
- 7The diagnostic significance of soluble CD163 and soluble interleukin-2 receptor α-chain in macrophage activation syndrome and untreated new-onset systemic juvenile idiopathic arthritis. Arthritis Rheum 2007; 56: 965–71., , , , , , et al.