Endogenous stimulation is responsible for the high frequency of IL-17A-producing neutrophils in patients with rheumatoid arthritis

Background Neutrophils play an important role in the pathogenesis of rheumatoid arthritis (RA). It has recently been reported that in addition to T helper (Th) 17 cells, other cells, including neutrophils, produce IL-17A, an important inflammatory cytokine involved in the pathogenesis of RA. The purpose of this study was to examine the presence of interleukin 17A-producing neutrophils in patients with RA. Methods We performed a cross-sectional study including 106 patients with RA and 56 healthy individuals. Whole peripheral blood cells were analyzed by flow cytometry to identify CD66b+ CD177+ IL-17A+ neutrophils and CD3+ CD4+ IL-17A+ T cells. Serum levels of IL-17A and IL-6 were measured by means of cytometry bead array (CBA). In purified neutrophils, mRNA levels of IL-17 and RORγ were measured by RT-PCR. In addition, purified neutrophils from patients and healthy controls were stimulated with the cytokines IL-6 and IL-23 to evaluate differences in their capacity to produce IL-17A. Results Neutrophils from RA patients expressed IL-17 and RORγ mRNA. Consequently, these cells also expressed IL-17A. Serum IL-17A levels but not Th17 cell numbers were increased in RA patients. Neutrophils positive for cytoplasmic IL-17A were more abundant in patients with RA (mean 1.2 ± 3.18%) than in healthy individuals (mean 0.07 ± 0.1%) (p < 0.0001). Although increased IL-17A+ neutrophil numbers were present in RA patients regardless of disease activity (mean 6.5 ± 5.14%), they were more frequent in patients with a more recent diagnosis (mean time after disease onset 3.5 ± 4.24 years). IL-6 and IL-23 induced the expression of RORγ but failed to induce IL-17A expression by neutrophils from RA patients and healthy individuals after a 3 h stimulation. Conclusion IL-17A-producing neutrophils are increased in some RA patients, which are not related to disease activity but have an increased frequency in patients with recent-onset disease. This finding suggests that IL-17A-producing neutrophils play an early role in the development of RA. Electronic supplementary material The online version of this article (10.1186/s13223-019-0359-9) contains supplementary material, which is available to authorized users.

Background RA is a systemic and chronic inflammatory autoimmune that affects 1-1.5% of the population worldwide [1]. The inflammatory process behind RA is associated with an increased production of inflammatory cytokines and chemokines. Among others, tumor necrosis factor (TNF) and interleukin (IL)-6 appear to play a major role in the pathogenesis of tissue damage in RA, as evidenced by a decrease of disease activity after blocking of these cytokines [2]. Other cytokines involved in the pathogenesis of RA include interferon (IFN)-γ and IL-17, which are the major effector cytokines of the Th1 and Th17 subsets of CD4+ T lymphocytes, respectively [3,4]. Of these, IL-17 (mostly IL-17A) has been the subject of interest over the past two decades. Thus, IL-17A levels are increased in the synovial fluid and serum [5][6][7], and Th17 cells are enriched in the synovial membrane of RA patients [6]. Moreover, the K/BxN mouse model of arthritis supports the role of IL-17 in the pathogenesis of joint inflammation [8]. Although preliminary clinical trials targeting IL-17A in RA have shown some clinical benefits, these positive results have yet to be confirmed in large-scale clinical trials [4,9,10]. In addition to Th17 cells, mast cells and mononuclear cells are present in the synovium of RA patients. Additionally, human and mouse neutrophils are capable of producing IL-17A [11][12][13][14][15][16][17][18][19][20][21], but no evidence for these cells has been linked to RA. In the case of IL-17A-producing neutrophils, they are increased in the peripheral blood of patients with asthma who are allergic to fungus [22], as well as in the plaques of psoriasis patients [23], and in the joints of patients with ankylosing spondylitis but not in those of patients with osteoarthritis [24]. In RA patients, neutrophils are known to accumulate in the synovial fluid and, to a lesser extent, in the synovial tissue [25]. In mouse models of RA, IL-17A+ neutrophils appear to be important effectors of disease pathogenesis and tissue damage [8]. In addition, peripheral blood neutrophils from patients with RA are functionally different from those obtained from healthy people, with remarkable differences in gene and protein expression, such as that of TNF and myeloblastin. Moreover, RA neutrophils are primed for ROS production [1]. We hypothesized that IL-17A-producing neutrophils would be present in the peripheral blood of patients with RA but not in that of healthy counterparts. To evaluate this hypothesis, we analyzed a cohort of 106 patients with RA and 56 healthy individuals. We found a high frequency of IL-17A-producing neutrophils in patients with RA but not in healthy controls. This result suggests that neutrophils are an important source of IL-17A during RA pathogenesis.

Patients and control subjects
We enrolled 106 RA patients at the Rheumatology Department outpatient clinics of the Hospital Juárez de México (HJM) who fulfilled the ACR/EULAR 2010 criteria for the classification of RA [26,27]. Briefly, these criteria were a confirmed presence of synovitis in at least 1 joint, an absence of an alternative explanation for synovitis, and a total score of 6 of a possible 10 from the individual scores in (A) number and site of involved joints, (B) serologic abnormality (positive anti-CCP), (C) elevated acute-phase reactants, and (D) duration of symptoms. The enrolled RA patients had a mean disease duration of 9.4 ± 0.81 years (0.5-42 years) since their initial clinical symptoms; thus, all RA patients were receiving drug treatment. Demographic and clinical characteristics were obtained by an interview on the day of sample collection for laboratory studies. The data collected included age, sex, and age of disease onset. Physical examination included number of tender joint, number of swollen joint count, and the disease activity score calculated for 28 joints using C reactive protein (DAS28-CRP).
CRP was determined by means of CardioPhase hsCRP, (Siemens Healthcare Diagnostic, Marburg, Germany). Immunologic tests included rheumatoid factor (RF) determined by means of Tina-quant FR II (Roche Diagnostic, Indianapolis, IN, USA) and anticitrullinated protein antibody levels (ACPAs) determined by ARCHITECT Anti-CCP (Abbott Laboratories, Wiesbaden, Germany). The healthy controls (HC) were 56 age-and sex-matched individuals.
The study followed the guidelines of the Declaration of Helsinki and was approved by the Ethics Committee of the HJM. Patients and controls signed informed consent documentation to provide peripheral blood by a single needle puncture, and none of them had any sign of infection.

Serum collection
Peripheral blood was collected in a tube without anticoagulant and left undisturbed for approximately 30 min. The clot was removed by centrifugation, and sera was aliquoted and kept frozen until cytokine determination.

Neutrophil purification and RT-PCR
Erythrocytes were removed from HWB following dextran sedimentation, and neutrophils were isolated by a Ficoll-Hypaque gradient from the leukocyte (nonmononuclear) fraction. Neutrophil purity was 96%, as defined by staining with an anti-CD66b antibody; cell viability was 98% immediately after purification and 90% after 16

Neutrophil stimulation and IL-17A quantification
Neutrophils were isolated from healthy controls and RA patients as described above, and 2 × 10 6 cells were stimulated for 3 h with IL-6 (20 μg/ml) and IL-23 (2 μg/ ml) as described [18]; during the last 2 h of stimulation, cells were cultured in the presence of brefeldin A (Sigma Aldrich, Saint Louis, Missouri). After stimulation, cells were collected and stained for flow cytometry as described before. Supernatants were collected after 16 h of stimulation, and IL-17A was quantified by BD cytometry bead array (CBA) following the manufacturer's instructions.

Statistics
Statistical analysis was performed using GraphPad Prism Software v5 (GraphPad Prism Software Inc.). All data were analyzed for the D' Agostino & Pearson omnibus normality test. The definition of statistical significance was set at p < 0.05 by means of the nonparametric Mann-Whitney test and by the nonparametric ANOVA Kruskal-Wallis test when more than two groups were examined. Correlation analysis was performed using Spearman's nonparametric method.

Patients and control characteristics
Demographic characteristics and ACPA, RF, CRP and ESR data for the 106 patients and 56 healthy controls are shown in Table 1. Eighty-eight percent of patients were women, and twelve percent were men. The mean DAS28-CRP was 3.2 ± 1.27. Among all patients, 13.8% had at least one first-degree relative with autoimmune disease. The percentage of RF-seropositive patients was 86.1% (> 60 U/ml). The extra-articular manifestations secondary to RA were observed, as follows: 79.4% of the patients had kerato-conjunctivitis sicca, 16% of the patients

IL-17A serum levels but not Th17 cells are increased in RA patients
Increased levels of IL-17A in the serum of patients with RA have been described [5][6][7]. We measured serum levels of IL-17A in the cohort of RA patients and healthy controls included in this study. We found slightly but statistically significantly increased levels of IL-17A in RA patients compared with those in control subjects (8. Fig. 1a). Because Th17 cells are considered the main producers of IL-17A, we measured the frequency of already polarized Th17 cells in the peripheral blood. This result means that any polyclonal stimuli, such as ionomycin/PMA or anti-CD3/ CD28 antibodies, were used to activate T cells in vitro. Therefore, we observed a low frequency of Th17 cells, but there were higher numbers of Th17 cells in RA patients than in healthy controls (0.06% ± 0.14% [0.0-0.92%] in RA patients vs 0.02% ± 0.04% [0.0-0.22%] in controls, p = 0.08); however, this difference was not statistically significant (Fig. 1b).

The frequency of IL-17A+ neutrophils is elevated in patients with RA
To evaluate the frequency of IL-17A+ neutrophils, we first determined the expression of IL-17A and RORγ mRNA in purified neutrophils from RA patients by means of RT-PCR, both of which were present in neutrophils from RA patients. There was no statistically significant correlation between mRNA levels and the frequency of IL-17A-producing neutrophils (Fig. 2a, b) nor with the mean fluorescence intensity (MFI) of intracellular IL-17A (Fig. 2c). However, a positive correlation (p ≤ 0.0001, r = 0.84) was observed between IL-17A and RORγ mRNA levels (Fig. 2d). Afterward, we studied the frequency of IL-17A+ circulating neutrophils in RA patients and healthy controls. Neutrophil identification was achieved by double staining using anti-CD177 (FITC) and anti-CD66b (PE) antibodies (Additional file 2: Figure S1). IL-17A mRNA expression also correlated with the frequency of IL-17A+ neutrophils in these patients (p = 0.008, r = 0.74) (Fig. 2b).
Based on disease activity, as defined by DAS28-CRP, there was no significant correlation with the frequency of IL-17A+ neutrophils (Fig. 3d, p = 0.45, r = 0.07). Moreover, there was no correlation between serum IL-17A and the frequency of IL-17A+ neutrophils.
Neutrophils are innate immune cells that are important in the early stages of the inflammatory response; we observed a positive correlation between the frequency of IL-17A+ neutrophils and the time with clinical diagnosis of RA (Fig. 3f ) (p = 0.007 r = 0.26). Patients with an increased frequency of IL-17A+ neutrophils were those with a relatively recent diagnosis of RA, with a mean time of 3.5 ± 4.24 years from disease onset. In contrast, RA patients negative for IL-17A-producing neutrophils had a mean of 8.2 ± 8.1 years since disease onset.
As IL-6 is necessary for the induction of IL-17A production by neutrophils [28,29], we measured the serum levels of IL-16, which were increased in RA patients (7.99 ± 16.7 pg/ml in RA vs 0.67 ± 1.18 pg/ml in HC). Additionally, there was a positive correlation between IL-6 serum levels and DAS28-CRP (p < 0.0001, r = 0.39), as others have observed [30][31][32], but IL-6 levels did not correlate with an increased numbers of IL-17A+ neutrophils (data not shown).

IL-6 and IL-23 induce expression of RORγ but not of IL-17A in neutrophils
As it was reported that IL-6 and IL-23 induce the production of IL-17A and the expression of its master transcription factor RORγ by CD4+ T cells [19,28,29], it was of interest to examine whether these cytokines had similar effects on neutrophils. Therefore, purified neutrophils from healthy donors were stimulated with the same concentration of IL-6 and IL-23 as described by Li et al. [18]. The frequency of IL-17A+ neutrophils increased, but it was not statistically significant when compared with that of nonstimulated neutrophils (US = 0.4% ± 0.14%, IL-6 + IL-23 = 2.5% ± 2.1%, p = 0.15, Fig. 4a). To analyze whether this effect was similar in 72.9% ± 17.7%, p = 0.02) (Fig. 4b). Importantly, although there was a trend toward an increased frequency of IL-17A-producing neutrophils in the RA patient group compared with that in the healthy control group, this difference was not statistically significant (HC 2% ± 2, RA 15.7% ± 13%, p = 0.08) (Fig. 4c). We wondered if IL-17A could be detected at a later time point of stimulation; for this reason, we measured the cytokine that was being released in the supernatant after 16 h of stimulation. We did not observe IL-17A production by neutrophils from healthy controls (US 36.3 pg/ml ± 42.8 pg/ml; IL-6+ IL-23 46.9 pg/ml ± 54.8 pg/ml, p = 0.7) (Fig. 4d).

Discussion
It has been established that IL-17A is an important cytokine in the pathogenesis of RA. Increased levels of IL-17A have been shown in the synovial fluid of RA patients, and a high concentration of this cytokine is a marker of disease severity. In this study, we observed an increase in the circulating levels of this cytokine. However, a low frequency of Th17 cells ex vivo was observed, with no significant difference from the healthy control group. Nevertheless, Leipe et al. showed that despite the low frequency of Th17 cells observed ex vivo, there was significant difference with healthy controls. This difference from our results might be due to the kind of patients studied; they analyzed patients with very early activity disease (mean disease duration < 3 months), and these patients were treatment-naive patients [33]. In contrast, all RA patients analyzed in the present study were receiving drug treatment (methotrexate, leflunomide or other DMARDs, Additional file 1: Table S1), and the most recently diagnosed patients had a mean of 3.5 years. However, we studied a wide range of patients regarding the mean disease duration since their initial clinical symptoms had occurred 9.4 ± 0.81 years (0-42 years) prior. These criteria were used to improve the understanding of T cell and neutrophil biology in the presence of the inflammatory milieu and immunosuppressive medication. In this context, it has Consequently, this result suggests that IL-17A might be produced by an alternative source rather than by Th17 cells.
We and other groups have reported neutrophils as a source of IL-17A in asthma patients where an increased frequency of IL-17A+ neutrophils was observed, especially in patients with asthma allergy to fungus [22]. The pathogenic role of IL-17 neutrophils during inflammatory processes in patients with spondyloarthritis has been suggested by observations of this subpopulation of cells found in the facet joints of patients with advanced ankylosing spondylitis [24].
In the present study, we found that peripheral blood IL-17A-producing neutrophil levels were elevated in RA patients, regardless of the level of disease activity. It is likely that these circulating neutrophils might contribute to systemic consequences of RA [34]; by this manner; they might also be involved in the inability to achieve a complete absence of symptoms. However, some of these cells will migrate to the synovial fluid where they might . The Mann-Whitney U-test was utilized and resulted in a p value < 0.0001. d Correlation plot between IL-17A + neutrophils and DAS28-CRP. e Correlation graph between IL-17A-producing neutrophils and time after onset of the disease. A Spearman correlation test was performed for these samples. Error bars indicate SD  15:44 contribute to the local inflammatory response, because, it has been demonstrated that neutrophils CD177+, which co-expressed on their plasma membrane proteinase-3 appear to be selectively recruited for transmigration [35]. IL-17 and RORγ mRNA was detected in neutrophils purified from the peripheral blood of RA patients. The correlation between the levels of both mRNAs suggests that RORγ is a key transcription factor for the development of IL-17A-producing neutrophils. However, the absence of a correlation between IL-17A mRNA levels and IL-17A protein suggests the participation of a regulatory mechanism for both molecules within neutrophils. IL-17A was detected by intracellular staining. Thus, these findings reflect that the production of IL-17A is de novo synthesis and not due to a protein storage model, as has been suggested for neutrophils  It has been demonstrated that neutrophils from asthmatic patients or from healthy individuals produce increased amounts of IL-17A when stimulated with IL-6, IL-23 and IL-21 [29]. It has also been reported that IL-6 and IL-23 induce the expression of IL-17A mRNA and protein in mouse neutrophils and human neutrophils from healthy individuals [28]. We found high levels of IL-6 in patients with RA, although there was no correlation with the frequency of IL-17A+ neutrophils. We observed, as has been described for Th17 cells, that IL-6 and IL-23 induce the expression of the transcription factor RORγt but only a slight increase in the production of IL-17A. Thus, in addition to cytokines, neutrophils need a second signal that triggers the production and secretion of IL-17A. However, it has been recently published that purified neutrophils from healthy donors did not produce IL-17A when they were stimulated with a broad array of stimuli, including LPS, curdlan or TNFα [37]. We hypothesize, as it was posited before, that perhaps some DAMPS produced during the disease might be involved in the production of this cytokine.
To our knowledge, this is the first time that IL-17A-producing neutrophils have been detected in the peripheral blood of RA patients. However, a prospective study should be conducted to identify whether the increased population of circulating neutrophils expressing IL-17A has a predictive value for relapses or worsening of the disease. Additionally, it will be important to define whether the population of neutrophils are stimulated and expanded in bone marrow and then released into circulation or if they are polarized to produce IL-17 in the periphery when they find the stimuli. Furthermore, the frequency of IL-17A+ neutrophils is higher in patients with recent onset of the disease (less than 3.5 years), whereas patients with more than 8 years of disease had lower frequencies, suggesting that these cells might play an important role during the early phases of RA but are still active during even the very chronic stage of the disease. Therefore, we suggest that peripheral blood IL-17A-producing neutrophils might be involved in the innate immune response that contributes to sustaining the inflamed joints observed in RA patients. In conclusion, RA patients have IL-17A-producing neutrophils that appear to result from endogenous stimulation and that could be an important source of this cytokine at the site of inflammation. Thus, it is important to consider neutrophils as potential therapeutic targets in RA.

Conclusions
IL-17A-producing neutrophils are increased in RA patients, particularly in recent onset disease patients. This finding suggests that IL-17A-producing neutrophils play an early role in the development of RA.

Additional files
Additional file 1: Table S1. DMARD therapy administrated to patients.
Additional file 2. Figure S1. Strategy used for neutrophils identification. A) Neutrophils were identify according to size (forward scatter) and complexity (side scatter) and B) to the expression of CD66b and CD177, an isotype control antibody were used to identify positive to IL-17A.