A pathogenic IFNα, BLyS and IL-17 axis in Systemic Lupus Erythematosus patients

This study aims to analyze in depth the role of IFNα in the upregulation of BLyS in different leukocyte populations and the possible relationship of these molecules with IL-17 and other pathogenic cytokines in SLE. Thus, IFNAR1 and membrane BLyS (mBLyS) expression was upregulated on various blood cell types from patients and closely correlated in all individuals. Moreover, BLyS serum levels associated positively with IFNα and IL-17A amounts, as well as with mBLyS on B cells and neutrophils. Interestingly, mBLyS on neutrophils was also correlated with IL-17A levels. Additionally, intracellular IL-17A expression was increased in both CD4+ lymphocytes and neutrophils from patients, and IL-17+CD4+ T cell frequency was associated with serum IFNα and IFNRA1 expression on B cells. Finally, in vitro assays support an IFNα role in the activation of Th17 cells in SLE. In conclusion, these data suggest that IFNα, BLyS and IL-17 could form a pathological axis in SLE, involving T and B lymphocytes, monocytes, DCs and neutrophils, which act in a vicious circle that encourage the preexisting inflammation and propagate the disease process.

Scientific RepoRts | 6:20651 | DOI: 10.1038/srep20651 characteristics. Therefore, additional research into the roles played by IFNα and BLyS is required for the identification of SLE patients appropriated for these treatments.
Similarly, IL-17A pathway inhibitors have been recently proposed as a therapeutic option for SLE patients 25 , since increased circulating levels of IL-17 correlated with disease activity and a Th17/Th1 imbalance have been reported in SLE [26][27][28] . Interestingly, it has been described that IL-17, alone or in synergy with BLyS may stimulate B cell survival and differentiation [29][30][31] , thus leading to the production of autoantibodies, and consequently IFNα secretion by pDC activated by the resulting immune-complexes 32,33 . Indeed, type I IFN has been described to exert a detrimental role in Th17 drive-autoimmune diseases 34 .
Considering previous evidence, the present study aims to evaluate the role of type I IFN signalling in the upregulation of BLyS in SLE patients by analysing the expression of BLyS and IFNRA1 (the α -chain of the common receptor for type-I IFNs) on the membrane of different leukocyte populations, as well as their possible association with the IL-17 production and the serum levels of other relevant pathogenic cytokines.

IFNAR1 overexpression on blood leukocytes is associated with membrane BLyS levels.
To extend the knowledge about the previously reported upregulatory in vitro effect of IFNα on BLyS induction and release 19 , we wanted to examine the possible relationship between BLyS and IFNAR1 expression on the surface of different circulating fresh cell types. Thus, membrane BLyS (mBLyS) and IFNAR1 levels were quantified by flow cytometry on B cells, neutrophils, monocytes, pDCs and mDCs from SLE patients and healthy controls (HC) (Fig. 1A). As expected, mBLyS expression was upregulated in all studied populations from patients, but especially in neutrophils, monocytes, mDCs and B lymphocytes, the same populations that exhibited higher levels of IFNAR1 compared with HC. Moreover, mBLyS and IFNAR1 levels were closely correlated in all individuals, but more strongly in SLE patients (Fig. 1B).
Multivariate linear regression analysis showed that mBLyS and IFNAR1 upregulation was independent of the treatment, but significant associations were detected with specific clinical manifestations. Thus, mBLyS and IFNAR1 levels on monocytes were related to the frequency of nephritis (β = 0.275, p = 0.032 and β = 0.270, p = 0.027, respectively) and serositis (β = 0.316, p = 0.012 and β = 0.283, p = 0.028), whereas expression on B cells were associated to the frequency of serositis (β = 0.257, p = 0.025 and β = 0.209, p = 0.021) and neurologic disease (β = 0.336, p = 0.005 and β = 0.289, p = 0.017). Therefore, these data indicate an increase in the expression of IFNAR1 in various cell types linked to the pathological widespread overexpression of mBLyS in SLE patients and associated to specific clinical manifestations.
Circulating BLyS correlated with IFNα and IL-17A serum levels in SLE patients. Next, we analyzed serum levels of BLyS (sBLyS) and IFNα with the aim of determining their possible association with the mBLyS and IFNAR1 overexpression on the different cell types. Interestingly, circulating amounts of sBLyS in SLE patients correlated positively with mBLyS expression on B cells (ρ = 0.273, p = 0.025) and neutrophils (ρ = 0.329, p = 0.007) and the same tendency was observed with IFNAR1 (B cells: ρ = 0.227, p = 0.064; neutrophils: ρ = 0.273, p = 0.026). However, no significant associations were detected between IFNα levels and mBLyS or IFNAR1 expression. Additionally, serum levels of several SLE related cytokines were analyzed in patients and controls (Table 1), showing significant associations of sBLyS with other soluble mediators upregulated in SLE, specifically IFNα , IL-17A, IL-12p70 and MIP-1α (Table 2). Interestingly, in a multiple linear regression model including circulating levels of cytokines, age, gender and disease activity (SLEDAI and anti-dsDNA titer), serum levels of IL-17A remained as an independent predictor of sBLyS (β = 0.252, p = 0.005). Moreover, IL-17A levels correlated with mBLyS expression on neutrophils (ρ = 0.246, p = 0.045), a cell type that is activated by this cytokine in addition to being able to produce it. Therefore, these data suggest a relationship between IFNα , BLyS and IL-17 expression in SLE patients.
IL-17 producing cells are increased and associated with IFNα levels in SLE patients. Therefore, in view of the observed results and since IL-17 can be produced by several cell subsets, including activated Th17 cells and neutrophils 35 , we analyzed the intracellular IL-17 and IFNγ expression in fresh peripheral blood cells from SLE patients and HC ( Fig. 2A). The proportion of IL-17 + cells was increased among both neutrophils and CD4 + lymphocytes from patients compared to controls (Fig. 2B), in accordance with the elevated levels of this cytokine detected in the serum. Remarkably, IL-17 + CD4 + lymphocytes exhibited a positive correlation with IFNα serum levels as well as with IFNRA1 B cell expression in SLE patients, but not in HC (Fig. 2C). These associations were not displayed by IFNγ + cells, thus suggesting that IFNα plays a role in the activation of Th17 cells from SLE patients.
To assess this hypothesis, we performed in vitro cultures with PBMCs isolated from SLE patients and HC to analyze the effect of IFNα treatment on IL-17A and BLyS secretion. As was previously reported 19 , IFNα stimulation increased BLyS release in both HC and SLE cultures compared to untreated cells. However, IL-17A secretion was increased in the supernatants of PBMCs cultures from patients but not in those with HC cells (Fig. 2D).

Discussion
In recent years, much evidence has demonstrated the crucial role of IFNα and BLyS in the etiopathogenesis of SLE 2,36 . Additionally, the upregulation and cellular mobilization of BLyS after IFNα stimulation has been reported [19][20][21] . Since clinical trials in SLE are conducted with agents that counteract BLyS 23 , its induction by IFNα in patients becomes an important topic. Consequently, the present study extends the proposed relationship between IFNα signaling and BLyS expression in SLE, and reveals a possible pathogenic axis involving both molecules and IL-17 in these patients. Although prior studies have pointed to a contribution of IFNAR signalling in autoantibody production and renal disease in murine models [9][10][11]37,38 , no previous works have analysed the expression of IFNAR in SLE patients. The findings of this study revealed an increased expression of IFNAR1 on various SLE leukocyte populations, which were closely related to the enhanced mBLyS levels. Moreover, both molecules were also correlated in healthy individuals, thus suggesting an intrinsic mechanistic connection between these two pathways.
In agreement with the simultaneous overexpression of IFNAR1 and mBLyS, our cohort of patients showed a positive correlation, not previously reported, between serum levels of IFNα and BLyS. Therefore, it becomes clear that IFNα signalling, via the IFNAR1, could be involved in SLE pathogenesis through the upregulation of BLyS in various cell types, including B cells, monocytes, mDCs and neutrophils. However, only the expression on B cells and neutrophils was positively correlated with the circulating amounts of BLyS. Another interesting finding of this work was the relationship of IL-17 with both BLyS and IFNα in SLE patients, allowing us to propose a new cytokine pathogenic axis in this disease. IL-17 exerts a proinflammatory effect mediated by the recruitment and activation of monocytes and neutrophils, in addition to the induction of cytokines and chemokines [39][40][41] . Increased frequency of Th17 cells and elevated amounts of IL-17 have been implicated in the pathogenesis of SLE and other autoimmune diseases [26][27][28]42,43 , although the main cellular source of this cytokine in such patients is uncertain. Accordingly, our ex vivo assays of blood leukocytes indicated an increased proportion of cells expressing IL-17, not only CD4 + lymphocytes but also neutrophils, a population activated in SLE and able to produce this molecule in pathological conditions 44,45 . Interestingly, IL-17 serum levels in SLE patients correlated with circulating BLyS as well as with mBLyS expression on neutrophils. In line with this, it has been proposed that BLyS could promote the expansion of Th17 cells 46 , whereas IL-17, alone or in synergy with BLyS may stimulate B cell survival and differentiation [29][30][31] .
Likewise, IFNα serum levels and IFNRA1 expression on B cells were related to the proportion of CD4 + T cells producing IL-17 in patients, suggesting that high IFNα levels, maybe indirectly, could induce IL-17 secretion in SLE. In fact, our in vitro experiments support this hypothesis. Whereas our results agree with the observed blocking effect of type I IFNs on the IL-17 production by human PBMCs 47-50 , IFNα treatment of such cells from SLE patients induced IL-17 secretion, which was positively correlated with BLyS release. In agreement with this, IL-17 has been positively correlated with IFNα and MxA expression, a sensitive marker of type I IFNs activity 51,52 , in cutaneous lesions of lupus erythematosus 53 . This effect of IFNα on SLE cells could be due to an a non-canonical IFNAR signalling that can activate the transcription factor STAT3 or the induction of IL-6, both of them critical factors for the Th17 differentiation [54][55][56] . Moreover, higher percentages of T-helper cells producing IL-17 have been described in SLE patients expressing type I IFN inducible genes, supporting the hypothesis that type I IFN co-acts with Th17 cytokines in SLE pathogenesis 57 .
In summary, the present study has prompted us to propose the existence of a pathological axis in SLE involving IFNα , BLyS and IL-17, all of which are cytokines with well-known deleterious roles in these patients (Fig. 3). First, the high IFNα levels usually present in SLE patients could be responsible for widespread leukocyte activation and induction of BLyS, as was suggested by the coordinated overexpression of IFNAR1 and BLyS on the membrane of B cells, monocytes, mDCs and neutrophils, and by the positive correlation detected between serum

. Proposed IFNα, BLyS and IL-17 pathogenic axis in SLE. The high IFNα levels present in most SLE patients could promote the activation and induction of BLyS in B cells, monocytes, mDCs and neutrophils,
including the low-density granulocyte (LDG) subset. Thus, the co-stimulation provided by these activated antigen-presenting cells -such as B cells, monocytes or mDCs-, to the Th17 subset enhanced in SLE patients can induce the release of IL-17. Moreover, IL-23 secretion by IFNα -activated mDCs may amplify Th17 differentiation. Also, IFNα -activated neutrophils under the inflammatory conditions presented in SLE are able to secrete IL-17. Then, this vicious circle is closed by the combined effect of IL-17 and BLyS. First, they promoted B cell survival and differentiation, which lead to autoantibody production and further generation of the immune complexes able to induce IFNα secretion by pDCs. In addition, both cytokines can activate neutrophils, thus sustaining the preexisting SLE inflammation. Finally, IFNα secretion by neutrophils, and especially by the LDG subset, propagates the disease process. levels of both cytokines. Furthermore, IFNα could promote IL-17 secretion by Th17 cells, increased in SLE patients, probably through the co-stimulation provided by activated antigen-presenting cells, such as B cells, monocytes or mDCs, thus explaining the parallel release of BLyS and IL-17 in SLE cultures after IFNα treatment. Moreover, IFNα -stimulated mDCs secrete IL-23 58 , a cytokine that amplifies Th17 differentiation. Likewise, the altered neutrophils present in SLE patients, over-activated with IFNα , could be also a relevant source of IL-17 59 . Finally, the vicious circle is closed by the effect of IL-17 and BLyS promoting B cell survival and differentiation, which lead to autoantibody production and IFNα secretion by pDCs 32,33 . Also, IL-17 and BLyS can activate neutrophils, thus encouraging the preexisting inflammation. Furthermore, SLE neutrophils, and especially the recently described low-density granulocyte subset 60,61 , could secrete increased levels of IFNα and form neutrophil extracellular traps (NETs), which in turn would lead to more IFNα synthesis by pDCs leading to a self-perpetuating cycle. Therefore, the concomitant participation of IFNα , BLyS and IL-17 in a pathogenic axis could be a key factor underlying the incomplete response to therapies based on the blockade of these cytokines in a subset of SLE patients. Hence, given the heterogeneity of the disease, determination of the individual cytokine profile could be useful for the identification of SLE patients candidates for biological therapies blocking such targets, alone or in combination.

Methods
Patients and controls. Patients included in the study were not hospitalized individuals fulfilling at least four American College of Rheumatology (ACR) revised criteria for the SLE classification 62 , and that were sequentially recruited from the outpatient clinic of the Autoimmune Disease Unit (Hospital Universitario Central de Asturias, HUCA). Information on clinical features during the disease course was obtained after a retrospective review of clinical histories. At the time of sampling, patients were asked precise questions regarding the treatment received over the previous 3 months and anti-dsDNA titer and SLE disease activity index (SLEDAI) were determined. Ex vivo cytometric analysis were performed in sixty-seven patients (Table 3) and twenty-nine healthy controls (mean age ± SD: 47.73 ± 6.56 years; female/male: 21/8) recruited from the same population. For cytokine quantification, additional serum samples from 199 SLE patients (Supplementary Table S1) and 90 sex and age-matched   To determine IL-17 and IFNγ positive leukocytes, blood cells were fixed, permeabilized and intracellularly stained with monoclonal antibodies against these cytokines following the manufacturer's instructions (Fixation/permeabilization buffer set; eBiosciences). Acquisition of 200,000 events and 10,000 CD4 + cells/tube was performed using a FACSCanto II flow cytometer (BD Biosciences). The analysis was based on cells located in an area of plots termed "the living region" which was defined using forward and side scatter. Then, different leukocyte subpopulations were identified according to the expression of specific surface markers, as previously described 19 . Thus, pDCs were identified as double positive cells for the expression of BDCA-2 and CD123, mDCs were defined as CD19 − BDCA-1 + , while monocytes, CD4 + lymphocytes and B cells and were identified by expression of CD14, CD4 or CD19, respectively. Neutrophils were identified according to their distinctive forward and side-scatter signal. Supplementary Table S2 shows the number of events counted in the different cellular subpopulations. Samples were subsequently analyzed using FlowJo software (Scripps Research Institute, San Diego, CA). Results were expressed as the percentage of positive cells or mean fluorescence intensity (MFI) of gated populations after subtracting the fluorescence of the background of the respective isotype control from the total fluorescence.

In vitro cultures. Peripheral blood mononuclear cells (PBMC) from healthy donors and patients were
obtained by centrifugation over Ficoll-Hypaque gradients (Lymphoprep, Nycomed). PBMCs, at a density of 2 × 10 6 /ml, were cultured in complete RPMI medium (RPMI 1640 containing 2 mM L-glutamine and 25 mM Hepes, supplemented with 10% heat-inactivated fetal calf serum and the antibiotic streptomycin and ampicillin at 100 μ g/ml) at 37 °C and 5% carbon dioxide in presence or absence of human IFNα 2b (Intron-A, 1000 U/ml). At different times of culture (2, 4 and 6 hours), cell-free supernatants from these cultures were collected for IL-17A and BLyS quantification.

Statistical analysis.
The Kolmogorov-Smirnov test was used to assess the normal distribution of the data and nonparametric testing was used to determine differences between patient and control groups (Mann-Whitney U-test), while correlations were examined by Spearman's rank correlation test. Multivariate linear regression analyses including treatments, demographic and clinical parameters were performed to determine the influence on the mBLyS and IFNRA1 expression. Variables were log-transformed to achieve normal distribution and standardized linear regression coefficients (beta) were used as an estimate of the association. Data were expressed as the median (interquartile range). A p-value < 0.05 was considered statistically significant. Data were analysed using GraphPad Prism 5 software (GraphPad Software, USA) and SPSS 22 statistical software package (SPSS Inc.).