IL-40: A New B Cell-Associated Cytokine Up-Regulated in Rheumatoid Arthritis Decreases Following the Rituximab Therapy and Correlates With Disease Activity, Autoantibodies, and NETosis

Background Interleukin 40 (IL-40) is a newly identified B cell-associated cytokine implicated in humoral immune responses and B cell homeostasis. As B cells play a pivotal role in autoimmunity, we investigated the function of IL-40 in rheumatoid arthritis (RA). Methods IL-40 expression was determined in the synovial tissue from RA and osteoarthritis (OA) patients. IL-40 was analysed in the serum/synovial fluid of patients with RA (n=50), systemic lupus erythematosus (SLE, n=69), OA (n=44), and healthy controls (HC, n=50). We assessed the changes of IL-40 levels in RA patients following the B cell depletion by rituximab (n=29) or after the TNF inhibition by adalimumab (n=25). We examined the relationship between IL-40, disease activity, autoantibodies, cytokines, and NETosis markers. Effect of IL-40 on synovial fibroblasts was determined. Results IL-40 was overexpressed in RA synovial tissue, particularly by synovial lining and infiltrating immune cells. The levels of IL-40 were up-regulated in the synovial fluid of RA versus OA patients (p<0.0001). Similarly, IL-40 was increased in the serum of RA patients compared to HC, OA, or SLE (p<0.0001 for all) and decreased after 16 and 24 weeks (p<0.01 and p<0.01) following rituximab treatment. No significant effect of adalimumab on IL-40 was observed. IL-40 levels in RA patients correlated with rheumatoid factor-IgM and anti-cyclic citrullinated peptides (anti-CCP) in the serum (p<0.0001 and p<0.01), as well as in the synovial fluid (p<0.0001 and p<0.001). Synovial fluid IL-40 was also associated with disease activity score DAS28 (p<0.05), synovial fluid leukocyte count (p<0.01), neutrophil attractants IL-8 (p<0.01), MIP-1α (p<0.01), and markers of neutrophil extracellular traps externalization (NETosis) such as proteinase 3 (p<0.0001) and neutrophil elastase (p<0.0001). Synovial fibroblasts exposed to IL-40 increased the secretion of IL-8 (p<0.01), MCP-1 (p<0.05), and MMP-13 (p<0.01) compared to the unstimulated cells. Conclusions We show the up-regulation of IL-40 in RA and its decrease following B cell depleting therapy. The association of IL-40 with autoantibodies, chemokines, and markers of NETosis may imply its potential involvement in RA development. Moreover, IL-40 up-regulates the secretion of chemokines and MMP-13 in synovial fibroblasts, indicating its role in the regulation of inflammation and tissue destruction in RA.


INTRODUCTION
Rheumatoid arthritis (RA) is a common chronic autoimmune disease characterized by persistent synovitis with various extraarticular manifestations (1). The pathogenic process of RA is localised particularly in the synovial joint infiltrated by immune cells, which together with synovial fibroblasts release inflammatory mediators and matrix-degrading enzymes, contributing to bone erosion and cartilage destruction (2). For decades, RA has been widely accepted as a T cell-driven autoimmune disease. Emerging evidence emphasized the significance of B cells in the pathogenesis of RA beyond antibody production (3). As documented in the experimental models of arthritis, cytokines secreted by B cells may be involved in the induction and promotion of arthritis (4,5). Recently, B cell depleting agents have proved highly effective in RA treatment (6). Nevertheless, elimination of B cells often results in a decline of autoantibody levels in RA (7), but does not always correlate with the clinical response to therapy (7,8). It is apparent that autoantibody-independent mechanisms of B cells are implicated in the progression of RA (9). Therefore, identification of new B cell-associated biomarkers embedded in the pathogenesis of RA would be beneficial.
Interleukin 40 (IL-40) is recently identified as a B cellassociated cytokine, which is related to immune response mechanisms and B cell homeostasis (10). This cytokine was originally described by Catalan-Dibene et al. in 2017 as a small (27 kDa) secreted protein encoded by the C17orf99 gene (chromosome 17 open reading frame 99) (10). Based on its unique structural properties, IL-40 fell into the category of the few so-called "orphan" cytokines (11), which do not share homologies with any of the established cytokine families. To date, there have been only few studies regarding the C17orf99 gene or its product IL-40 protein. It is well established that C17orf99 or IL-40 are expressed only in mammals and are particularly enhanced in fetal liver, bone marrow, and activated B cells (10)(11)(12)(13). As demonstrated in IL-40 knockout mice, IL-40 is associated with lactation and affects the production of IgA and, thereby, the composition of the intestinal microbiome in mice (10). Moreover, IL-40 knockout mice exhibited abnormalities in B cell populations, indicating the role of IL-40 in B cell development (10). Under in vitro conditions, IL-40 is expressed by human B cells upon activation by anti-CD40 mAb, anti-IgM, and IL-4, and its expression is further potentiated by transforming growth factor (TGF)-b1 (10). Furthermore, IL-40 was detected in several cell lines of human diffuse large B cell lymphoma (10), and its differential expression among the lymphoma subtypes was reported (14,15). Very recently, C17orf99 was found down-regulated in the co-culture of human respiratory epithelial cells with macrophages upon treatment with anti-inflammatory cytokine IL-38 (16), suggesting the role of IL-40 in inflammation. The only evidence linking IL-40 to autoimmune inflammation was reported back in 2012 by Zingaretti et al., who identified C17orf99 as one of the four autoantigens discriminating autoimmune hepatitis from healthy individuals (17).
Altogether, the implication of IL-40 in the B cell homeostasis and in the regulation of immune mechanisms makes it a suitable candidate player in the pathogenesis of autoimmune diseases. Thus, we aimed to analyse the expression of IL-40 in patients with RA, its association with disease-specific parameters, and its immunomodulatory capacity in vitro.

Patients
This study included three cohorts of patients with RA. Cohort 1 involved patients with active RA with knee effusion (total of 50

Laboratory Measurements
In cohort 1, paired blood and synovial fluid samples were collected at the time of clinically indicated knee arthrocentesis in 50 RA and 44 knee OA patients. In cohorts 2 and 3, only peripheral blood samples were obtained from 29 RA patients who received rituximab therapy (cohort 2) at baseline and at weeks 16 and 24 and from 25 patients receiving adalimumab (cohort 3) at baseline and at weeks 12 and 52. All samples were immediately processed, aliquoted, and stored at -80°C until use. Prior to analysis, the synovial fluid samples were treated with hyaluronidase (Hylase Dessau; Riemser Arzneimittel, Greifswald, Germany) for 30 minutes at 37°C. The disease severity was assessed by the disease activity score in 28 joints using the erythrocyte sedimentation rate (DAS28-ESR). Levels of C-reactive protein (CRP) were analysed by turbidimetry using an Olympus Biochemical Analyzer (Olympus CO Ltd., Tokyo, Japan), and anti-CCP and RF-IgM levels were determined using standard enzyme-linked immune sorbent assay (ELISA) kits (Test Line s.r.o., Czech Republic). CD19+ B cells were assessed by flow cytometry as previously described (20).

Measurement of Cytokines
Levels of IL-40 in the serum and synovial fluid were analysed by commercially available ELISA kits, according to the

Immunohistochemistry
Samples of synovial tissue were obtained from five patients with RA (5 females, with a mean (SD) age of 67 (7) years) from one knee joint, two elbow joints, and two small hand joints. The control group comprised of synovial tissue from four patients with OA (2 females and 2 males, with a mean (SD) age of 63 (8) years) from three knee joints and one hip joint. All synovial tissue samples were obtained during joint surgery, embedded in paraffin, and cut into 5-mm-thick sections. Subsequently, the sections were deparaffinised and rehydrated. Endogenous peroxidase activity was inhibited by adding DAKO Dual Endogenous Enzyme Block (Agilent, Santa Clara, CA, USA) for 10 min, and non-specific hydrophobic binding activity was prevented by adding 2% bovine serum albumin (BSA, Thermo Fisher Scientific, Waltham, Massachusetts, USA) diluted in PBS.
In order to detect IL-40 in the synovial tissue, the slides were immunoprobed with primary rabbit polyclonal C17orf99 antibody (Sinobiological, Eschborn, Germany) diluted 1:500 in 2% BSA in PBS, and were incubated overnight at 4°C. Isotype IgG Universal Negative Control for IS-Series Rabbit Primary Antibodies (Agilent, Santa Clara, CA, USA) was used as a negative control. After rinsing with PBS buffer, HRP conjugated polyclonal goat anti-rabbit secondary antibody (Agilent, Santa Clara, CA, USA) diluted 1:200 in 2% BSA in PBS was added for 1h at room temperature (RT

Cell Cultures and In Vitro Experiments
RA synovial fibroblasts were obtained from biopsies from RA patients (n=9) as previously described (21)

Statistical Analysis
The data are presented as median and interquartile range (IQR) if not stated otherwise. Normality of data across analysed groups was tested by Kolmogorov-Smirnov test with 5% level of statistical significance. For analysis of differences between groups, Mann-Whitney U-test for non-paired data was performed. Wilcoxon matched-pairs signed rank test was applied for the in vitro data. Bivariate relationships between variables under study were determined by Spearman's correlation coefficient. Multivariate regression analysis was used to predict levels of IL-40 in the serum and synovial fluid by a set of predictors. The predictors for both dependent variables (serum IL-40 and synovial fluid IL-40) were selected based on significant bivariate associations. Highly collinear predictors were excluded from the analysis, and only the best predictor (with the highest correlation with dependent variable) was retained in the regression model. Longitudinal observations of IL40 were examined using two-way (cohorts 2 and 3 by time) repeated-measures analysis of variance (RM-ANOVA) to assess the inter-group differences, followed by one-way RM-ANOVA with least significant difference (LSD) post-hoc tests conducted within each group. P values less than 0.05 were considered statistically significant (*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001). GraphPad Prism 6 (GraphPad Software, La Jolla, CA, USA) and IBM SPSS 25.0 ® (Chicago, IL, USA) were used to perform the analysis.

RESULTS
Increased Expression of IL-40 in the RA Synovial Tissue IL-40 was detected in both RA and OA synovial tissue (Figure 1). The IL-40 expression was significantly enhanced in RA compared to OA synovial tissue, particularly within the inflammatory infiltrate ( Figure 1).  Table 3).

Synovial Fluid IL-40 Correlates With the Count and Activation of Neutrophils and With Markers of NETosis (Cohort 1)
In patients with RA, synovial fluid IL-40 significantly correlates with the number of synovial fluid neutrophils (r=0.375, p=0.017) and with the synovial fluid levels of chemokines IL-8 (r=0.539, p=0.004) and MIP-1a (r=0.537, p=0.002), (Figures 3C, D). Given the association of IL-40 with the neutrophil attractant and activator IL-8, we sought to analyse the relation between IL-40 and markers of NETosis. We found that synovial fluid IL-40 strongly correlates with the levels of proteinase 3 (PR3) (r=0.661, p<0.0001) and neutrophil elastase (NE) (r=0.652, p<0.0001) (Figures 3E, F). No relation was found between serum IL-40 and any of the studied cytokines, chemokines or NETosis markers (data now shown).

DISCUSSION
We are the first to demonstrate the implication of the new cytokine IL-40 in the pathogenesis of RA. We show local and Data were analysed using Spearman correlation and are presented as correlation coefficient r. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. anti-CCP, anti-cyclic citrullinated peptide antibody; CRP, C-reactive protein; DAS, disease activity score; ESR, erythrocyte sedimentation rate; IgM, immunoglobulin M; RF, rheumatoid factor; SF, synovial fluid; SJC, swollen joint count; TJC, tender joint count.  systemic up-regulation of IL-40 in patients with RA and its association with disease activity, levels of autoantibodies, chemokines, and markers of NETosis. We also provide evidence that extracellular IL-40 enhances the secretion of chemokines and matrix-degrading enzyme MMP-13 in synovial fibroblasts. IL-40 is a recently discovered B cell-associated cytokine related to immune response and B cell homeostasis (10,17). Here we demonstrated that IL-40 is accumulated in the synovial tissue of RA patients. We observed intensive IL-40 staining within the hyperplastic synovial lining layer, indicating its expression by activated synovial fibroblasts. Also, immune cells infiltrating the RA synovial tissue tested positive for IL-40. According to Catalan-Dibene et al., IL-40 expression is associated with B cells (10); however, our data on synovial tissue indicate that besides B cells, other immune cells such as macrophages, T cells and neutrophils are involved in IL-40 expression in RA. This possibility is supported by our analysis of serum IL-40 in RA patients treated with B cell depleting agent rituximab. We showed that high levels of IL-40 decreased following rituximab therapy in 70% of patients, whereas the remaining 30% showed unchanged or even increased levels of serum IL-40. Indeed, no significant association of IL-40 levels with the number of B cell at baseline or following the therapy was found. Since the B cell depletion was accompanied by a decline in disease activity, dysregulation of other immune cells might have occurred and contributed to the changes in the IL-40 levels following the treatment. Altogether, these data suggest that IL-40 is produced by multiple cell types in RA. Of note, serum levels of IL-40 do not seem to be significantly affected by the treatment with TNF inhibitor adalimumab. Targeting specific upstream events, such a B cells activation, rather than TNF itself, may be responsible for reducing IL-40, which indicates that IL-40 may be related to specific mechanisms in the pathogenesis of RA linked to B cells and their regulatory effects. Furthermore, IL-40 was elevated in the paired samples of serum and synovial fluid of RA patients compared to OA. In addition to that, we assessed IL-40 in the serum of patients with SLE, a systemic autoimmune rheumatic disease characterised by abnormalities and hyperactivity of B cells (22). Surprisingly, the serum levels of IL-40 in SLE were comparable to the ones in OA and healthy controls, suggesting no specific association of IL-40 with neither B cells nor the systemic nature of the disease. These findings imply that the enhanced systemic levels of IL-40 in RA may be related to the local inflammation and originate from the inflamed joints rather than from systemic inflammation. In support of this claim, levels of IL-40 in the RA synovial fluid are about six times higher than those in the serum and are strongly associated with each other. In addition, IL-40 correlates with several indicators of local inflammation such as synovial fluid leukocyte count, but also with the disease activity score, or the number of swollen joints, but not with CRP. In light of these findings, it should also be stressed that only a negligible amount of IL-40 was detected in the OA synovial fluid, and the levels of IL-40 in the OA serum were comparable to the ones found in healthy individuals. This observation further reinforces the hypothesis that IL-40 diffuses from the RA joints to the blood circulation and reflects the intensity of the local inflammation.
We also demonstrated that IL-40 is significantly associated with the levels of RF-IgM and anti-CCP autoantibodies and thereby with the activation of B cells, which is in agreement with previously reported data (10). Of particular interest is the correlation of IL-40 with anti-CCP, a finding which outlines a possible implication of IL-40 in the process of citrullination and autoimmune reaction in RA. It is well established that the anti-CCP production in RA is fuelled by citrullinated antigens released during the formation of neutrophil extracellular traps -NETosis (23) and that anti-CCP antibodies play an important role in the early phases of RA development (24). In this regard, we showed that synovial fluid IL-40 correlates with the number of neutrophils in the synovial fluid and with the synovial fluid levels of IL-8, a neutrophil chemoattractant and potent inducer of NETosis (25). Consistently, IL-40 co-localized with MPO, one of common neutrophil markers, in the RA synovial tissue. In addition, the strong correlation of synovial fluid IL-40 with the levels of proteinase 3 and neutrophil elastase further highlight the link of IL-40 to neutrophils and NETosis. Moreover, our unpublished preliminary data indicate that IL-40 is released by RA neutrophils undergoing NETosis. A question which therefore arises is whether the high concentration of IL-40 in the synovial fluid may be a result of the crosstalk between the activated B cells and neutrophils in RA joint. In addition to all that, in vitro data show that RA synovial fibroblasts enhanced the IL-8 secretion upon IL-40 stimulation. Given the abundance of IL-40 in the RA synovial fluid, we can speculate that IL-40 stimulates fibroblasts in the synovial lining layer to produce IL-8, which in turn, can activate synovial fluid neutrophils and thereby fuel the autoimmune reaction in RA joints. Moreover, IL-40 upregulates the secretion of pro-inflammatory chemokine MCP-1 by synovial fibroblasts. MCP-1 has recently been demonstrated to act on synovial fibroblasts to promote their aggressive phenotype (26). Thus, it is tempting to assume that IL-40 may be implicated in the development of synovial hyperplasia. Noteworthy is the association of synovial fluid IL-40 with the levels of MIP-1a, which not only functions as multifunctional chemokine and neutrophil attractant (27) but, according to Jordan et al., also represents a potent regulator of bone resorption in arthritis (28). In the context of bone remodelling, the correlation of IL-40 with anti-CCP and IL-8 should be emphasized, as IL-8 is involved in the anti-CCP-driven osteoclast activation and bone loss (29,30). Last but not least, we revealed that IL-40 acts on the RA synovial fibroblasts to upregulate the secretion of MMP-13, a key molecule in the cartilage degradation network (31). These data imply that IL-40 may play a particular role in the RA joint damage, a subject that requires further investigation. Some limitations of the study should be taken into account. The relationship between the IL-40 expression in the synovial tissue and fluid could not be assessed since the tissue samples were obtained from a group of patients different from those who underwent synovial fluid aspiration. Furthermore, the correlation between IL-40 and increased NETosis observed in the synovial fluid could not be further explored in patients following biological therapies, as there are no synovial fluid samples available after the treatment.
Notwithstanding these limitations, this study shows for the first time an implication of IL-40 in the pathogenesis of RA. In fact, these novel findings may pave the way for placing IL-40 into the portfolio of molecules involved in the immune reaction and tissue remodelling control in RA. Also, this work may inspire further studies on IL-40 in other autoimmune diseases.

CONCLUSION
Taken together, our results show that IL-40 is elevated in RA and decreases following the B cell depleting therapy. Moreover, IL-40 correlates with disease activity, autoantibodies, chemokines, and markers of NETosis, indicating its potential implication in RA development. In addition, IL-40 up-regulates chemokines and MMP-13 in synovial fibroblasts, which implies its potential role in inflammation and tissue destruction in RA.

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

ETHICS STATEMENT
The studies involving human participants were reviewed and approved by Institutional Ethics Committee of Institute of Rheumatology, Prague, Czech Republic. The patients/ participants provided their written informed consent to participate in this study.