The role of the innate immune system in the pathogenesis of systemic juvenile idiopathic arthritis: advances in understanding
Systemic juvenile idiopathic arthritis (sJIA) is a severe autoinflammatory childhood-onset immune disorder, classified as one of the JIA subtypes. Patients clinically present with fever, skin rash, enlarged lymph nodes (LNs), spleen and/or liver, and develop arthritis. Haematologically, neutrophilia (including immature neutrophils), leukocytosis, anaemia, and thrombocytosis is noticed. Patients have high plasma cytokine levels of interleukin-6 (IL-6), IL-18, C-reactive proteins, and S100 proteins (S100A8/A9 and S100A12). Among all JIA subtypes, sJIA has the highest mortality rate since the disease is associated with a severe life-threatening complication macrophage activation syndrome (MAS). MAS is a hyperinflammatory syndrome with excessive activation of macrophages and T cells resulting in a cytokine storm and is associated with hemophagocytosis. More recently sJIA was also associated with pulmonary complications that include interstitial lung disease (ILD), and pulmonary alveolar proteinosis (PAP).
In this doctoral thesis, we aimed to understand the role of the innate immune system in the pathogenesis of the sJIA. More specifically we aimed to gain insights into the successful use of IL-1 antagonists in patients considering that IL-1β blood levels are barely detectable. Next, we aimed to understand the effects of IL-1 on both the neutrophils and the γδ T cells. Considering the high numbers of neutrophils in blood of patients and the splenomegaly as a characteristic of sJIA, we studied extramedullary granulopoiesis and the role of neutrophils in the disease pathogenesis. Finally, we aimed to answer why sJIA patients show lung disorders, and more specifically investigated whether this is rather an epiphenomenon, or whether lung inflammation participates in the disease initiation.
To address these aims, we took advantage of a described mouse model of sJIA that is based on the injection of complete Freund’s adjuvant (CFA) in interferon-gamma (IFN-γ) knock-out (KO) mice. These mice develop all clinical and haematological features reminiscent of sJIA, whereas CFA-immunisation in wild-type (WT) animals only results in a partial, self-resolving immune response. We demonstrated increased inflammasome activation in immunised IFN-γ KO mice, as evident by increased levels of active caspase-1 and IL-1β in the lymphoid organs and the blood. This could in part explain their increased susceptibility to develop sJIA. Using neutralising anti-IL-1β antibodies, the sJIA-like features could partially be prohibited and treated mice showed reduced weight loss, rash, granulocytosis, and thrombocytosis, and had reduced plasma IL-6 levels. Interestingly, the number of IL-17- and IL-22-producing CD27 negative γδ T cells and CD4+ T cells were decreased upon treatment. Beneficial effects of IL-17-depletion were reported in the initial description of the mouse model. In this doctoral thesis, we explored the role of IL-22 in the pathogenesis of sJIA. Increased IL-22 plasma levels were measured in the CFA-immunised IFN-γ KO mice. Using IL-22 neutralising antibodies we showed a partial role of IL-22 in the mouse model since neutralisation was associated with reduced weight loss, tail damage, neutrophilia, and thrombocytosis. In sJIA patients with active disease increased plasma levels of active IL-22 were measured. Using healthy donor γδ T cells or CD4+ cells, we could show a synergistic effect of IL-22 and IL-1β in the production of IL-17.
To understand the role of neutrophils in the sJIA pathophysiology, we made use of anti-Ly6G (1A8) depleting antibodies. Anti-Ly6G can be used to study the in vivo role of neutrophils in murine models and became the gold standard considering its neutrophil-specific depletion. In the CFA-challenged BALB/C model, we demonstrated an incomplete anti-Ly6G-mediated neutrophil depletion. Furthermore, using BrdU incorporation, we demonstrated an increased proliferation in the bone marrow (BM) upon depletion, and a massive neutrophilic proliferation in the spleen, showing that a temporary depletion is followed by a fast rebound of new immature neutrophils that cannot be depleted. Therefore, we used an alternative approach in which we targeted granulocyte-colony stimulating factor (G-CSF) signalling. G-CSF is the main growth factor involved in the proliferation, maturation, release, and functionality of neutrophils. G-CSF release can be stimulated by IL-17. During excessive myelopoiesis, it can drive the formation of myeloid-derived suppressor cells (MDSCs), having an immune-tempering capacity. Increased plasma levels of G-CSF were measure in the CFA-immunised IFN-γ KO mice. Also, in patients with active disease increased plasma G-CSF levels were reported. Interestingly, these levels were correlating with some inflammatory mediators including the number of neutrophils and plasma levels of IL-6 or IL-1β. The role of G-CSF was studied in our mouse model using G-CSF receptor (G-CSFR)-blocking antibodies. We could demonstrate that G-CSF is responsible for the observed neutrophilia and extramedullary myelopoiesis. Next, by blocking the G-CSFR, we could inhibit the development of arthritis. However, treated mice showed a mild aggravation of some of the disease symptoms including weight loss, tail damage, number of immature RBCs, and had slightly increased plasma levels of IL-6, IL-17, and TNF-α. Single-cell sequencing revealed that neutrophil development in the spleen follows a similar developmental path as described in the BM and also includes stages of pre-neutrophils, immature neutrophils, and mature neutrophils. When we compared CFA-immunised WT and IFN-γ KO mice, the latter were characterised by the development of MDSCs, having a mature phenotype with high levels of CEBP/β. Administration of the G-CSFR-antagonising antibody could block the maturation and differentiation of neutrophils in CFA-immunised mice. The reduced numbers of MDSCs could in part explain their aggravated phenotype.
In the final part of this doctoral thesis, lung abnormalities in our sJIA mouse model were the subject of research considering the recent lung complications that were described in patients. Using microcomputer tomography, a moderate enlargement of the lungs was measured upon CFA-immunisation as well as a decrease in aerated lung density. In the IFN-γ KO mice only, an increased non-aerated lung density was found. Next, functional alterations were noticed in the CFA-immunised IFN-γ KO mice, with male mice showing an increased tissue hysteresivity and tissue damping, female mice presented with increased airway hyperactivity. Histopathology revealed subpleural and parenchymal cellular infiltrates associated with the formation of granulomas. Using flow cytometry and quantitative PCR, we demonstrated an excessive innate immune activation with infiltration of mainly immature and mature neutrophils or monocytes, increased γδ T cells, activated macrophages, and increased expression of both IL-1β, IL-6, TNF-α, IL-17, and IL-22. Associated with increased neutrophilic infiltrates, high levels of G-CSF and neutrophil-attracting chemokines were measured in the CFA-injected IFN-γ KO mice. Although the pulmonary manifestations in mice were distinct to those described in sJIA patients, the work demonstrates an important innate immune activation in our model. It underlines that lung inflammation should not be neglected when the pathophysiology of sJIA is studied.
Expanding our knowledge on neutrophils in autoinflammatory diseases, we studied neutrophils in both pyrin-associated inflammation with neutrophilic dermatosis (PAAND) and familial Mediterranean fever (FMF). PAAND and FMF are two IL-1β-driven autoinflammatory syndromes caused by distinct mutations in the same MEFV gene. Considering their high numbers of neutrophils in the blood and affected tissues, the transcriptional, functional, and phenotypical characteristics of neutrophils were extensively analysed. At the transcriptional level patients showed a partially overlapping activated phenotype with increased expression of S100 proteins, IL-4R, CD48, F5, MMP9, and NF-κB. Increased levels of MMP9 and S100A8/A9 were validated in plasma. Flow cytometric analysis demonstrated an immature character, an increased expression of TLR4 and TLR9, whereas a decreased expression of CXCR2, C5aR, and BLTR1 was observed on neutrophils from FMF patients only. PAAND patients showed an enhanced granularity and a functional increased random, but reduced CXCL8-induced migration. Similarly, in FMF patients a tendency for enhanced random migration and reduced chemoattractant-induced migration was noticed. Neutrophils from both FMF and PAAND patients showed an increased basal and TNF-α-induced ROS production and had an elevated release of the azurophilic granule protein myeloperoxidase (MPO). Analysing the phagocytic capacity, we showed enhanced activity in PAAND patients but reduced activity in FMF patients. Together our data describe the neutrophil abnormalities in two pyrin-associated diseases and underlines that FMF and PAAND are two distinct diseases.