Axial spondyloarthritis (axSpA) and psoriatic arthritis (PsA) belong to a group of immune-mediated inflammatory diseases with primarily axial but also peripheral arthritis, with a prevalence of approximately 0.54% in Europe (1, 2). These diseases additionally present with extra-articular manifestations like uveitis, inflammatory bowel disease and psoriasis. As the onset of disease predominates in the third decade, socioeconomic and personal effects are high and long-lasting, demanding efficient therapy. The common denominators of axSpA and PsA are inflammation of the entheses and osteoproliferation along the tendons and capsules. Presumably, inflammation precedes new bone formation and early halting of inflammation by therapies like non-steroidal anti-inflammatory drugs (NSAIDs), anti-TNF, anti-IL-17A/F and Janus kinase inhibition (JAKi) can avoid osteoproliferation and ultimately ankylosis. However, the precise mechanisms and players of this osteoproliferation have not been satisfactorily elucidated yet.

The efficacy of anti-IL-17A/F treatment in clinical studies together with genetic and functional evidence underlines the role of the IL-23/IL-17A pathway in the pathogenesis of SpA (3–5). Main players of the IL-23/IL-17A pathway in SpA are “pathogenic” Th17 cells, producing IL-17A, IL-17F, IL-22, GM-CSF and IFNγ (6). Pathogenic Th17 cells arise from Th17 cells, which are responsible for the maintenance of tissue homeostasis at mucosal barriers and host defense against fungi and extracellular bacteria, in the presence of IL-23 (7–9). IL26, a member of the IL-10 family, is also produced by pathogenic Th17 cells (10, 11). It uniquely combines proinflammatory with antimicrobial, DNA-binding and DNA-shuttling properties (12, 13). Elevated systemic levels and locally increased expression of IL26 have been reported in chronic inflammatory diseases like Crohn’s disease, rheumatoid arthritis, SpA, psoriasis and ANCA-associated vasculitis (11, 13–16). IL26 signaling is mediated by multiple pathways, including a classical cytokine receptor (IL-10R2 and IL-20R1) leading to activation of JAK1 and TYK2, and direct cell membrane penetration. IL26 expression has also been reported in CD8⁺ T cells, mucosal-associated invariant T (MAIT) cells, NKp44⁺ NK cells, γδT cells, innate lymphoid cells type 3 (ILC3), CD68⁺ macrophage-like and synoviolin⁺ fibroblast-like synoviocytes and synovial myofibroblasts (15–21). Effects have been observed on human intestinal and bronchial epithelial cells, endothelial cells, fibroblasts, monocytes, NK and B cells, plasmacytoid dendritic cells and neutrophils (12, 15, 16, 22–28).

IL26 also increases mineralization in human osteoblasts, and inhibits receptor activator of nuclear factor κB ligand-induced osteoclastogenesis and bone-resorbing activity of mature osteoclasts, making it an interesting therapeutic target for preventing osteoproliferation (16, 24). Additionally, the capacity of IL26 to bind DNA released during inflammation-induced tissue damage renders IL26 a key driver of chronic inflammation. This provides another interesting feature of anti-IL26 therapy for axSpA and PsA because, with current therapeutic options (NSAIDs, anti-TNF, anti-IL-17A/F, and JAKi), approximately 40% of patients will not enter sustained remission. Therefore, the need for new therapies remains high in these patients.

In this study we comparatively investigate the expression and cellular sources of IL26 in the peripheral blood and peripheral synovial tissue of anti-TNF therapy-naive patients with axSpA, PsA, and RA. We demonstrate significantly increased local expression of IL26 in axSpA, potentially implicating a significant role of IL26 in osteoproliferation and a new therapeutic target.

This is so far the largest comparative study of IL26 in inflammatory arthritis with a focus on axSpA and PsA, clearly showing the increased expression of IL26 in peripheral joints.

Our data uniquely add to previous findings of increased IL26 expression in peripheral joints of axSpA by demonstrating IL26 expression in axSpA as well as in PsA peripheral synovial tissue by immunohistochemistry and immunofluorescence (16). In our hands, IL26 co-localizes with CD68⁺ macrophage-like synoviocytes in the sublining of the synovium. This is of special interest as efficacy of 12 weeks of secukinumab (ACR20 response rate of 65%) in peripheral SpA was accompanied by a significant reduction of synovial sublining CD68⁺ macrophages and CD15⁺ neutrophils (34). Compared with RA, state-of-the-art analysis of synovial tissue-resident macrophages (STMs) is not available from SpA yet. In human RA joint tissue, two populations of STMs are present: MerTK⁻CD206⁻ STMs that induce proinflammatory responses in synovial fibroblasts and MerTK⁺CD206⁺ STMs that induce resolution of inflammation and repair mechanisms (35). These MerTK⁺CD206⁺ STMs are predominant in RA in remission, and a proportion of less than 47.5% is an independent predictor for disease flares after tapering or discontinuation of treatment. It remains to be elucidated if IL26-positive CD68⁺ synovial macrophages stem from circulating monocytes or from synovial tissue-resident macrophages, which both have been demonstrated in the K/BxN serum-transfer arthritis model (36). An ex vivo study on synovial fluid from PsA showed an expansion of intermediate monocytes and macrophages (37). These cells demonstrated expression of MRC1, the gene encoding CD206, which was shown to be an important marker for STMs in RA. At the moment, we cannot exclude that the IL26⁺ synovial myeloid cells are not producing IL26 themselves, but rather have shuttled extracellular DNA from cell death due to excessive inflammation in complex with IL26 into their cytosol (13). Possible local factors inducing IL26 production in resident CD68⁺ macrophage-like synoviocytes are not yet clear and need to be investigated. Studies with human alveolar macrophages have revealed bacterial products such as endotoxins and also environmental toxins (e.g., water-soluble tobacco smoke components) as inducers of IL26 (22, 38). Potential inducers of IL26 in CD68⁺ macrophage-like synoviocytes could be damage-associated molecular pattern (DAMP) molecules, such as HMGB1—either released after cell death or actively secreted—activating TLR4 and subsequently NF-κB (39). The latter was shown to be responsible for IL-1β-induced IL26 production in human memory CD4⁺ T cells (32). Due to the lack of synovial fluid samples from patients with PsA and axSpA, we were not able to further investigate the sources of IL26 in this compartment, which might differ from the actual synovial tissue. CD3⁺ T cells and fibroblast-like synoviocytes identified by cadherin-11 and CD90 are not involved in the increased expression of IL26 in the synovium of axSpA and PsA in our study, confirming the findings from SpA synovial fluid by Heftdal et al. (16). This contrasts with the findings from RA joint tissue, where IL26 co-localizes with CD3, RORγt, and synoviolin (15). However, IL26 supports Th17 immunity by upregulating IL-17A secretion from Th17 cells and inducing plasticity of memory non-Th17 cells toward Th17 cells through upregulation of IL-1β secretion from macrophages (15). In context of the DNA-binding capacity of IL26, it remains to be elucidated if IL26 in the peripheral joints of SpA patients also represents extracellular traps and, therefore, a further mechanism by which IL26 perpetuates chronic inflammation. IL26 has been shown to enhance NET-induced secretion of IL-6 from human monocytes and of CXCL8 from human neutrophils (13). As only a small fraction of IL26⁺ synovial cells are CD68⁺ in axSpA and PsA, further detailed investigations of local producers of IL26 in axSpA and PsA are warranted. Candidates include mainly CD3-negative cell types like ILC3, fibroblast-like synoviocytes and neutrophils. One caveat of our study is the fact that most of the joint samples analyzed were acquired during joint replacement surgery and, therefore, most likely represent a chronic state of inflammation with low activity in axSpA and PsA. In addition, these peripheral samples might not represent the mechanisms of inflammation and bone remodeling in the axial joints of SpA patients. This, however, is a common problem of SpA research, and the previously examined facet joints from SpA patients were also from established, chronic diseases. Nevertheless, IL26 is also present in the bone marrow of these joints (16).

Despite significantly elevated expression of IL26 in peripheral joint tissue in axSpA and PsA, we did not observe a relevant increase of IL26 in the serum of axSpA patients. This is congruent with the results of a previous study on plasma levels of IL26 in SpA patients (16). However, IL26 serum levels were significantly increased in PsA and RA patients compared with HCs. This could be explained by the fact that PsA and RA represent inflammatory arthritis with peripheral arthritis, whereas axSpA is predominately an axial disease. In our cohort of axSpA patients, only 27.6% had peripheral disease compared with 95% of PsA patients. The significantly elevated percentages of IL-1R1⁺ Th17 cells in PsA patients identified by chemokine receptor profile might also explain the significantly higher levels of serum IL26 in this group compared with HCs. Considering the significant enhancement of IL26 in the peripheral joint tissue of axSpA patients, this points to a distinct role of IL26 at the site of inflammation in axSpA compared with PsA and RA. Additional local factors involved in this distinct role of IL26 in axSpA joints need to be identified in the future. Differences between the serum levels of cytokines and local expression are not uncommon and have been described for example for IL-20 in patients with rheumatoid arthritis (40). This discrepancy has also been observed in patients with IBD, where increased protein and mRNA expression levels of IL26 in the gut were evident compared with HCs, but IL26 serum levels and IL26 expression in PBMCs were lower (41). This would fit with our observation of local IL26 attributable to tissue-resident CD68⁺ macrophage-like synoviocytes in SpA and not extravasated CD3⁺ T cells. A recent study in HCs demonstrated a large proportion of IL26⁺IL-17A⁻ cells within circulating Th17 cells (identified through chemokine receptor staining) via PrimeFlow (42). This population seems to be a Th17 intermediate as it differentiates into IL-17A⁺ Th17 cells upon TGF-β1 exposure. Such intermediates and the described differentiation were also shown in psoriatic skin lesions. Therefore, the lack of IL26⁺CD3⁺ cells in our peripheral tissue samples could be attributed to a loss of IL26 expression in chronic or long-standing disease. Additionally, the only slightly elevated percentages of circulating IL26⁺ Th17 cells in axSpA and PsA patients approximately corresponded with the results from the serum measurements of IL26, further strengthening the local role of IL26 in SpA. However, the reduced sample number in the PrimeFlow assay of IL26 (approx. 15 vs. 30 samples) might be responsible for not showing significantly elevated percentages of IL26⁺ Th17 cells. Moreover, the slightly increased percentages of IL26⁺CD8⁺ T cells in axSpA and PsA could be an alternative source of peripheral IL26. This could reflect the fact that, like IL-17A, IL26 can be produced by different cell types as part of a pathologic type 3 immunity. Nevertheless, we found memory CD4⁺ T cells from axSpA patients to be more prone to IL26 expression upon encountering typical Th17-favoring cytokine combinations like IL-2, IL-1β, and IL-23 in vitro. In addition, circulating CD4⁺ T cells from axSpA and PsA more often expressed IL26 ex vivo compared with HCs, being in congruence with previously reported data showing IL26 expression exclusively in peripheral CD4⁺ T cells (12).

IL26 could play a major part in the chronicity of inflammation in axSpA and PsA as IL26 can increase the secretion of IL-1β, IL-6, TNFα, and CCL20 (Th17 cell recruitment) from CD14⁺ myeloid cells as shown from RA synovial fluid (15). In this context, therapies targeting IL26 could be of great relevance in patients failing current therapies. Due to the employment of JAK1 and TYK2 downstream of the IL26 receptor, current JAK inhibitors like upadacitinib should reduce IL26 responses to some extent. However, specifically targeting IL26 could still have superior effects to current JAK1 or future TYK2 inhibition as several other signaling options have been described for IL26. For example, IL26/DNA complexes can enter plasmacytoid dendritic cells and activate TLR9 or activate monocytes by inflammasome-mediated activation of the STING (stimulator of interferon genes) pathway (12, 13). It also remains possible that IL26 alone due to its chemical properties could penetrate the cell membrane and illicit proinflammatory responses (13, 43). This aspect also remains interesting as it is not clear through which pathway IL26 induces bone mineralization in human osteoblasts (16). So far, the specific receptor chain for IL26, IL-20R1, has been detected in human amniotic fluid-derived stem cells, which can be differentiated into osteocytes, and in the mouse osteoblast precursor cell line MC3T3-E1 (44, 45), but not in primary human osteoblasts or osteoblast precursors. So, the exact role and mechanisms of IL26 in SpA-associated osteoproliferation remain unclear and warrant further investigation. Inhibiting RORγt, one of the key transcription factors of Th17 cells, significantly downregulates the expression of IL26 in axSpA Th17 cells (46), but unfortunately, none of the candidate compounds has been successfully trialed so far (47). As IL-17A increases IL26 secretion from RA synovial fibroblasts, IL-17A blockers like sekukinumab could also exert some effects on synovial pathophysiology in axSpA and PsA (15). Monoclonal antibodies targeting IL26 significantly suppressed imiquimod-induced skin inflammation in human IL26 transgenic mice without apparent effects on antimicrobial activity (48). In this study, the concentrations of anti-IL26 mAb used were much lower than the concentrations reported necessary for the antimicrobial effects of IL26. This allows for further evaluation of anti-IL26 mAb as a therapeutic target in inflammatory diseases like psoriasis, PsA and axSpA. As IL26 secretion from human memory CD4⁺ T cells induced by IL-1β/IL-1R1 was nearly abrogated by pharmacological inhibition of NF-κB activation, this provides a further target for blocking IL26-induced pathology assuming a role of the aforementioned IL26⁺IL-17⁻ Th17 intermediates in the active or early stages of the disease (32).

In summary, we have shown that different cells are responsible for local and systemic production of IL26 in axSpA and PsA providing evidence for the first time that CD68⁺ macrophage-like synoviocytes could be a local source of IL26 in SpA peripheral joints. Regarding its role in chronicity of inflammation and osteoproliferation, IL26 provides an interesting therapeutic approach for axSpA and PsA by simultaneously targeting the two key modules of its pathophysiology.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving humans were approved by 100/18 S, Ethical Committee of the Faculty of Medicine, Technical University Munich and 06/Q1606/139 Oxford University Hospitals NHS Foundation Trust. The study has been registered within the German Clinical Trials Register (DRKS00014672). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

AH: Conceptualization, Formal analysis, Funding acquisition, Investigation, Methodology, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing. AO: Formal analysis, Investigation, Visualization, Writing – original draft. QB: Writing – review & editing, Resources. HM-F: Writing – review & editing, Resources. GL: Writing – review & editing, Resources. MW: Writing – review & editing, Resources. FW: Writing – review & editing, Resources. HM: Writing – review & editing, Resources. DS: Writing – review & editing, Resources. PB: Supervision, Writing – review & editing, Resources. UH: Resources, Writing – review & editing. MA: Writing – review & editing, Resources. PM: Resources, Supervision, Writing – review & editing. CS: Resources, Supervision, Writing – review & editing.