In Vivo Expansion of Activated Foxp3+ Regulatory T Cells and Establishment of a Type 2 Immune Response upon IL-33 Treatment Protect against Experimental Arthritis.

IL-33 is strongly involved in several inflammatory and autoimmune disorders with both pro- and anti-inflammatory properties. However, its contribution to chronic autoimmune inflammation, such as rheumatoid arthritis, is ill defined and probably requires tight regulation. In this study, we aimed at deciphering the complex role of IL-33 in a model of rheumatoid arthritis, namely, collagen-induced arthritis (CIA). We report that repeated injections of IL-33 during induction (early) and during development (late) of CIA strongly suppressed clinical and histological signs of arthritis. In contrast, a late IL-33 injection had no effect. The cellular mechanism involved in protection was related to an enhanced type 2 immune response, including the expansion of eosinophils, Th2 cells, and type 2 innate lymphoid cells, associated with an increase in type 2 cytokine levels in the serum of IL-33-treated mice. Moreover, our work strongly highlights the interplay between IL-33 and regulatory T cells (Tregs), demonstrated by the dramatic in vivo increase in Treg frequencies after IL-33 treatment of CIA. More importantly, Tregs from IL-33-treated mice displayed enhanced capacities to suppress IFN-γ production by effector T cells, suggesting that IL-33 not only favors Treg proliferation but also enhances their immunosuppressive properties. In concordance with these observations, we found that IL-33 induced the emergence of a CD39(high) Treg population in a ST2L-dependent manner. Our findings reveal a powerful anti-inflammatory mechanism by which IL-33 administration inhibits arthritis development.

Aortic VCAM-1: an early marker of vascular inflammation in collagen-induced arthritis.

Cardiovascular disease (CVD) is a major cause of morbidity and mortality in rheumatoid arthritis (RA). There are limited experimental data on vascular involvement in arthritis models. To study the link between CVD and inflammation in RA, we developed a model of vascular dysfunction and articular inflammation by collagen-induced arthritis (CIA) in C57Bl/6 (B6) mice. We studied the expression of vascular inflammatory markers in CIA with and without concomitant hyperlipidic diet (HD). Collagen-induced arthritis was induced with intradermal injection of chicken type-II collagen followed by a boost 21 days later. Mice with and without CIA were fed a standard diet or an HD for 12 weeks starting from the day of the boost. Arthritis severity was evaluated with a validated clinical score. Aortic mRNA levels of vascular cell adhesion molecule-1 (VCAM-1), inducible nitric oxide synthase (iNOS) and interleukin-17 were analysed by quantitative RT-PCR. Vascular cell adhesion molecule-1 localization in the aortic sinus was determined by immunohistochemistry. Atherosclerotic plaque presence was assessed in aortas. Collagen-induced arthritis was associated with increased expression of VCAM-1, independent of diet. VCAM-1 overexpression was detectable as early as 4 weeks after collagen immunization and persisted after 15 weeks. The HD induced atheroma plaque formation and aortic iNOS expression regardless of CIA. Concomitant CIA and HD had no additive effect on atheroma or VCAM-1 or iNOS expression. CIA and an HD diet induced a distinct and independent expression of large-vessel inflammation markers in B6 mice. This model may be relevant for the study of CVD in RA.

Interplay between TNF and regulatory T cells in a TNF-driven murine model of arthritis.

CD4(+)CD25(+)Foxp3(+) regulatory T cells (Treg) are involved in several autoimmune diseases, including rheumatoid arthritis. TNF-α blockers induce therapeutic benefits in rheumatoid arthritis via a variety of mechanisms. We aimed to characterize the impact on Treg of TNF-α overexpression in vivo and of TNF-α inhibiting treatments. We used human TNF-α transgenic mice as a model of strictly TNF-α-dependent arthritis. Our study showed that initial Treg frequency was lower in TNF-α transgenic mice than in wild-type mice. However, the course of arthritis was marked by elevation of Treg frequency and a dramatic increase in expression of TNFR2. Antagonizing TNF-α with either the anti-human TNF-α Ab (infliximab) or active immunotherapy (TNF-kinoid) increased the Treg frequency and upregulated CTLA-4, leading to enhancement of suppressor activity. Moreover, both anti-TNF-α strategies promoted the differentiation of a CD62L(-) Treg population. In conclusion, in an in vivo model of TNF-α-driven arthritis, Treg frequency increased with inflammation but failed to control the inflammatory process. Both passive and active TNF-α-inhibiting strategies restored the suppressor activity of Treg and induced the differentiation of a CD62L(-) Treg population.

Deoxyribonuclease 1-Mediated Clearance of Circulating Chromatin Prevents From Immune Cell Activation and Pro-inflammatory Cytokine Production, a Phenomenon Amplified by Low Trap1 Activity: Consequences for Systemic Lupus Erythematosus.

Increased concentrations of circulating chromatin, especially oligo-nucleosomes, are observed in sepsis, cancer and some inflammatory autoimmune diseases like systemic lupus erythematosus (SLE). In SLE, circulating nucleosomes mainly result from increased apoptosis and decreased clearance of apoptotic cells. Once released, nucleosomes behave both as an autoantigen and as a damage-associated molecular pattern (DAMP) by activating several immune cells, especially pro-inflammatory cells. Deoxyribonuclease 1 (DNase1) is a major serum nuclease whose activity is decreased in mouse and human lupus. Likewise, the mitochondrial chaperone tumor necrosis factor (TNF) receptor-associated protein-1 (Trap1) protects against oxidative stress, which is increased in SLE. Here, using wild type, DNase1-deficient and DNase1/Trap1-deficient mice, we demonstrate that DNase1 is a major serum nuclease involved in chromatin degradation, especially when the plasminogen system is activated. In vitro degradation assays show that chromatin digestion is strongly impaired in serum from DNase1/Trap1-deficient mice as compared to wild type mice. In vivo, after injection of purified chromatin, clearance of circulating chromatin is delayed in DNase1/Trap1-deficient mice in comparison to wild type mice. Since defective chromatin clearance may lead to chromatin deposition in tissues and subsequent immune cell activation, spleen cells were stimulated in vitro with chromatin. Splenocytes were activated by chromatin, as shown by interleukin (IL)-12 secretion and CD69 up-regulation. Moreover, cell activation was exacerbated when Trap1 is deficient. Importantly, we also show that cytokines involved in lupus pathogenesis down-regulate Trap1 expression in splenocytes. Therefore, combined low activities of both DNase1 and Trap1 lead to an impaired degradation of chromatin in vitro, delayed chromatin clearance in vivo and enhanced activation of immune cells. This situation may be encountered especially, but not exclusively, in SLE by the negative action of cytokines on Trap1 expression.

Intra-articular electrotransfer of mouse soluble tumour necrosis factor receptor in a murine model of rheumatoid arthritis.

BACKGROUND: Rheumatoid arthritis (RA) is a chronic autoimmune disease that causes inflammation and destruction of the joints. In the collagen-induced arthritis mouse model of RA, we developed a nonviral gene therapy method designed to block in situ the main cytokine tumour necrosis factor (TNF)-alpha METHODS: Electrotransfer was used to deliver a plasmid encoding extracellular domain of mouse soluble TNF-alpha receptor type I fused to the Fc fragment of mouse immunoglobulin (Ig)G1 (pTNFR-Is) corresponding to a dimeric TNF-alpha soluble receptor fusion protein (mTNFR-Is/Ig). RESULTS: Delivery of the plasmid into the knees at symptom onset improved the histological inflammation and destruction not only at the knees, but also at the ankles, indicating a local and a regional therapeutic effect. The plasmid was detected in synovial membrane and meniscus specimens from the injected joints. In the synovial membrane, 15 days post-injection, interleukin (IL)-17 and TNF-alpha mRNAs expression were increased, whereas IL-10 mRNA was unchanged. However, the empty plasmid exerted a pro-inflammatory effect 30 days post-injection. CONCLUSIONS: These data indicate that local nonviral gene therapy against TNF-alpha is effective, although further work is needed to decrease plasmid induced inflammation.

Involvement of Tumor Necrosis Factor Receptor Type II in FoxP3 Stability and as a Marker of Treg Cells Specifically Expanded by Anti-Tumor Necrosis Factor Treatments in Rheumatoid Arthritis.

OBJECTIVE: To study the involvement of Treg cells expressing tumor necrosis factor receptor type II (TNFRII) in exerting control of inflammation in experimental models and in the response to anti-TNF treatments in patients with rheumatoid arthritis (RA) or spondyloarthritis (SpA). METHODS: The role of TNFRII in Treg cells was explored using a multilevel translational approach. Treg cell stability was evaluated by analyzing the methylation status of the Foxp3 locus using bisulfite sequencing. Two models of inflammation (imiquimod-induced skin inflammation and delayed-type hypersensitivity arthritis [DTHA]) were induced in TNFRII-/- mice, with or without transfer of purified CD4+CD25+ cells from wild-type (WT) mice. In patients with RA and those with SpA, the evolution of the TNFRII+ Treg cell population before and after targeted treatment was monitored. RESULTS: Foxp3 gene methylation in Treg cells was greater in TNFRII-/- mice than in WT mice (50% versus 36.7%). In cultured Treg cells, TNF enhanced the expression, maintenance, and proliferation of Foxp3 through TNFRII signaling. Imiquimod-induced skin inflammation and DTHA were aggravated in TNFRII-/- mice (P < 0.05 for mice with skin inflammation and P < 0.0001 for mice with ankle swelling during DTHA compared to WT mice). Adoptive transfer of WT mouse Treg cells into TNFRII-/- mice prevented aggravation of arthritis. In patients with RA receiving anti-TNF treatments, but not those receiving tocilizumab, the frequency of TNFRII+ Treg cells was increased at 3 months of treatment compared to baseline (mean ± SEM 65.2 ± 3.1% versus 49.1 ± 5.5%; P < 0.01). In contrast, in anti-TNF-treated patients with SpA, the frequency of TNFRII+ Treg cells was not modified. CONCLUSION: TNFRII expression identifies a subset of Treg cells that are characterized by stable expression of Foxp3 via gene hypomethylation, and adoptive transfer of TNFRII-expressing Treg cells ameliorates inflammation in experimental models. Expansion and activation of TNFRII+ Treg cells may be one of the mechanisms by which anti-TNF agents control inflammation in RA, but not in SpA.

Relationship between angiogenesis and inflammation in experimental arthritis.

Background. Angiogenesis is involved in rheumatoid arthritis (RA) leading to leucocyte recruitment and inflammation in the synovium. Furthermore, synovial inflammation itself further potentiates endothelial proliferation and angiogenesis. In this study, we aimed at evaluating the reciprocical relationship between synovial inflammation and angiogenesis in a RA model, namely collagen-induced arthritis (CIA). Methods. CIA was induced by immunization of DBA/1 mice with collagen type II in adjuvant. Endothelial cells were detected using a GSL-1 lectin-specific immunohistochemical staining on knee joint sections. Angiogenesis, clinical scores and histological signs of arthritis were evaluated from the induction of CIA until the end of the experiment. Angiogenesis was quantified by counting both the isolated endothelial cells and vessels stained on each section. To evaluate the effect of increased angiogenesis on CIA, VEGF gene transfer was performed using an adeno-associated virus encoding VEGF (AAV-VEGF), by intra-muscular or intra-articular injection in mice with CIA. Results. We showed an increase in synovial angiogenesis from day 6 to day 55 after CIA induction, and, moreover, joint vascularization and clinical scores of arthritis were correlated (p < 0.0001, r = 0.61). Vascularization and histological scores were also correlated (p = 0.0006, r = 0.51). Systemic VEGF overexpression in mice with CIA was followed by an aggravation of arthritis as compared to AAV-lacZ control group (p < 0.0001). In contrast, there was no difference in clinical scores between control mice and mice injected within the knee with AAV-VEGF, even if joint vascularization was higher in this group than in all other groups (p = 0,05 versus non-injected group). Intra-articular AAV-VEGF injections induced more severe signs of histological inflammation and bone destruction than AAV-Lac Z or no injection. Conclusion. Angiogenesis and joint inflammation evolve in parallel during collagen-induced arthritis. Furthermore, this work shows that exogenous VEGF can aggravate CIA. It is direct evidence that the increase in joint vascularization leads to an exacerbation of arthritis. Taken together, these results emphasize the role of angiogenesis in inflammatory arthritis. It also suggests an early involvement of angiogenesis in joint inflammation.

Collagen-induced arthritis and imiquimod-induced psoriasis develop independently of interleukin-33.

BACKGROUND: Interleukin (IL)-33 is a dual cytokine with both an alarmin role and a T helper 2 cell (Th2)-like inducing effect. It is involved in the pathogenesis of rheumatoid arthritis (RA) and its models; we recently demonstrated that exogenous IL-33 could inhibit collagen-induced arthritis (CIA) in C57BL/6 mice. However, its pathophysiological role in RA is unclear. Indeed, mice deficient in the IL-33 receptor ST2 show reduced susceptibility to arthritis, and the disease is not modified in IL-33-deficient mice. We examined the immune response in wild-type (WT) and IL-33-deficient mice with CIA. To further understand the role of endogenous IL-33 in inflammatory diseases, we studied its role in a skin psoriasis model. Mice on a C57BL/6 background were deficient in IL-33 but expressed lacZ under the IL-33 promoter. Therefore, IL-33 promotor activity could be analyzed by lacZ detection and IL-33 gene expression was analyzed by X-Gal staining in various mice compartments. Frequencies of CD4(+)FoxP3(+) regulatory T cells (Tregs) and Th1 and Th17 cells were evaluated by flow cytometry in WT and IL-33(-/-) mice. Bone resorption was studied by evaluating osteoclast activity on a synthetic mineral matrix. Psoriasis-like dermatitis was induced by application of imiquimod to the skin of mice. RESULTS: Severity of CIA was similar in IL-33(-/-) and WT littermates. Joints of IL-33(-/-) mice with CIA showed IL-33 promotor activity. In mice with CIA, frequencies of Tregs, Th1 and Th17 in the spleen or lymph nodes did not differ between the genotypes; osteoclast activity was higher but not significantly in IL-33(-/-) than WT mice. Psoriasis development did not differ between the genotypes. CONCLUSIONS: Despite its expression in the synovium of arthritic mice and normal keratinocytes, IL-33 is not required for CIA development in arthritis or psoriasis. Its absence does not induce a T cell shift toward Th1, Th17 or Treg subpopulations. Altogether, these data and our previous ones, showing that exogenous IL-33 can almost completely inhibit CIA development, suggest that this cytokine is not crucial for development of chronic inflammation. Studies of RA patients are needed to determine whether treatment targeting the IL-33/ST2 axis would be effective.

Engraftment of cutaneous fibroblasts within synovial membrane in a nonhuman primate: short-term results.

OBJECTIVES: Gene therapy using cells as vectors to achieve secretion of therapeutic proteins may hold promise in the treatment of chronic diseases. Cell-based gene therapy with xenogeneic cells secreting antiinflammatory cytokines (IL-4, IL-13, or IL-1 receptor type II) has been found effective in mice with collagen-induced arthritis (CIA), a model for human rheumatoid arthritis. Autologous cells engineered to produce antiinflammatory cytokines were also effective in the mouse CIA model. In all these experiments, the cells were grafted into the subcutaneous tissue of the back, resulting in systemic treatment. To evaluate the feasibility of cell-based gene therapy confined to the joints, we performed intraarticular injections of autologous cells in a rhesus monkey with CIA, a model more similar to human RA. METHODS: We prepared ex vivo cultures of skin fibroblasts from the animal then transfected the cells with a plasmid carrying the lacZ gene. We injected these marker cells into metacarpophalangeal, metatarsophalangeal, and interphalangeal joints. RESULTS: Kinetic evaluation of synovial tissue X-gal labeling, which reflected reported gene expression by skin fibroblasts present within the synovium, showed significant labeling by transfected cells up to 6 days after intraarticular injection. Xenogeneic fibroblasts (Chinese hamster ovary cells) injected intraarticularly were also detected within synovial specimens; however, labeling intensity was less marked than with autologous cells. Our findings establish the feasibility of skin fibroblast grafting into the synovium. CONCLUSION: This preliminary study opens the door to studies of heterotopic autologous transfected cells for the treatment of CIA in monkeys by direct gene transfer within joints.

Angiogenesis markers (VEGF, soluble receptor of VEGF and angiopoietin-1) in very early arthritis and their association with inflammation and joint destruction.

OBJECTIVE: To investigate the involvement of angiogenesis markers in very early arthritis patients and their relevance to predict further joint destruction. METHODS: Levels of Vascular Endothelial Growth Factor (VEGF), angiopoietin-1 (Ang-1), and soluble Fms-like tyrosine kinase-1 (sFlt-1) were measured by ELISA in serum samples from 310 patients having polyarthritis, evolving for less than 6 months (VErA cohort). Each angiogenesis marker was measured at baseline and one year later. X-rays of hands and feet were carried out at inclusion and after 1 year and read using the van der Heidje-modified Sharp method. RESULTS: At baseline and after 1 year, VEGF levels were correlated with clinical and biological parameters of inflammation. We also observed a positive correlation between sFlt-1 levels and biological inflammation (Erythrocyte Sedimentation Rate (ESR): r=0.17, p=0.006; C Reactive Protein: r=0.14, p=0.02). Angiopoietin-1 levels were correlated with ESR (r=0.12, p=0.04). Interestingly, only VEGF levels measured at baseline were correlated with Disease Activity Score measured 1 year later. Relationship between angiogenesis markers and radiographic progression was also evaluated. VEGF and Ang-1 levels measured at inclusion were related with Sharp score after one year (VEGF: r=0.21, p<0.001; Ang-1: r=0.24, p<0.001; Spearman's test). Moreover, VEGF levels were higher in patients with radiographic progression (p=0.002). CONCLUSION: Serum concentrations of VEGF, sFlt-1 and angiopoietin-1 were correlated to parameters of inflammation and to bone destruction in early arthritis. These results contribute to demonstrate that angiogenesis reflects disease severity and angiogenesis markers might become a new useful tool to evaluate disease activity and to estimate outcome for patients with inflammatory arthritis.

Syngeneic fibroblasts transfected with a plasmid encoding interleukin-4 as non-viral vectors for anti-inflammatory gene therapy in collagen-induced arthritis.

BACKGROUND: No effective long-term treatment is available for rheumatoid arthritis. Recent advances in gene therapy and cell therapy have demonstrated efficiency in collagen-induced arthritis (CIA). Interleukin-4 (IL-4) is already known to be efficient in CIA in systemic injection or administered by gene therapy. This study was designed to evaluate the effect of a non-viral gene therapy of CIA, involving injection of syngeneic fibroblasts transfected with a plasmid encoding for IL-4. METHODS: Immortalised fibroblasts from DBA/1 mice (DBA/1/0 cells) were transfected with a plasmid expressing IL-4 cDNA (DBA/1/IL-4 cells). Xenogeneic fibroblasts from Chinese hamster ovary (CHO) transfected with a plasmid expressing IL-4 cDNA (CHO/IL-4) were studied also. The cells were engrafted in mice developing CIA by subcutaneous injection of 3 x 10(6) DBA/1/0 or DBA/1/IL-4 or CHO/IL-4 cells. RESULTS: Injection of DBA/1/IL-4 cells, on days 10 and 25 after immunisation, was associated with a significant and lasting improvement in the clinical and histological evidence of joint inflammation and destruction as compared with DBA/1/0 and CHO/IL-4 cells. DBA/1/IL-4 cell treatment decreased also the production of IgG2a antibody to CII and the proliferation of CIIB-specific nodal T cells. Later treatments (engraftments on days 23 and 35 after immunisation) exerted also an anti-inflammatory effect, as evaluated on clinical and histological signs of CIA. CONCLUSIONS: Taken together, these findings indicate that systemic administration of syngeneic cells transfected with an anti-inflammatory cytokine gene, namely IL-4, with a non-viral method is effective in CIA and may attenuate the cytokine imbalance seen in this disease.

Interleukin-4 cellular gene therapy and osteoprotegerin decrease inflammation-associated bone resorption in collagen-induced arthritis.

To evaluate the respective action of IL-4, an anti-inflammatory cytokine, and OPG, an inhibitor of bone resorption, on the inflammatory process and the associated bone resorption in collagen-induced arthritis (CIA). After CIA induction, DBA/1 mice were treated with OPG or with IL-4 DBA/1 transfected fibroblasts or both OPG + IL-4. CIA significantly improved in IL-4 groups. OPG had no effect on arthritis clinical scores but histologic scores were reduced in OPG, IL-4, and OPG + IL-4 groups vs. nontreated CIA mice. OPG increased significantly BMD and decreased by 45% D-pyridinolin levels. Moreover association of IL-4 and OPG exerted an additive effect of BMD and resorption marker (-68%). Production of IFN-gamma in the supernatants of spleen cells was reduced in IL-4 treated mice. OPG had a moderate effect on IFN-gamma, but potentiated the inhibitory effect of IL-4. OPG and IL-4 prevent bone loss in CIA-mice model and could have additive effects on IFN-gamma secretion.