Background
Spondyloarthropathies (SpA) are a group of inflammatory disorders characterized by involvement of the spine and peripheral joints. In addition, entheseal inflammation is central to its pathogenesis. Adult axial SpA is seen in adults who present with inflammatory back pain (IBP) and may or may not have radiological evidence of sacroiliitis (
1). Ankylosing spondylitis (AS) is the prototype of axial SpA that has a very strong association with human leukocyte antigen (HLA)-B27 with more than 95% patients being positive for it (
2,
3). HLA-B27 along with environmental factors like mechanical stress and gut microbes are implicated in the pathogenesis of SpA.
Subclinical gut inflammation has been reported in 60% of patients with AS (
4). Gastrointestinal symptoms and antibodies associated with gastrointestinal diseases, and IgA levels were associated with high disease activity in patients with SpA (
5). Gut dysbiosis leads to leakage and translocation of bacterial products into the systemic circulation (
6). These can then act via toll like receptors (TLR) s especially 2 and 4 (
7) and activate monocytes/macrophages leading to the production of pro-inflammatory cytokines as well as IL-23 which skews the immune response towards IL-17 production by T cells and innate immune cells. Patients with AS and ERA show high expression of TLR 2 and 4 on monocytes as well as produce higher pro-inflammatory cytokines compared to healthy controls (HC) (
8‐
11). In addition, we have recently shown that endogenous ligands like myeloid related protein (MRP)8/14 and Tenascin-C (TNC) are elevated in sera from AS patients compared to HC and their levels correlate with activity of the disease (
12,
13).
Enthesitis related arthritis (ERA) category of juvenile idiopathic arthritis (JIA) is characterized by lower limb arthritis and enthesitis and is usually seen in boys older than 6 years (
14). HLA-B27 is present in 60–80% of these children (
15). IBP, the hallmark of AS develops in a proportion of patients after 5–10 years and about one-third progress to AS in adulthood (
16). ERA patients also show gut dysbiosis and presence of autoantibodies associated with inflammation of gastrointestinal mucosa like antibodies to tissue transglutaminase or anti-
Saccharomyces cerevisiae antibodies (
17). They have elevated fecal calprotectin, high pro-inflammatory cytokines (TNF and IL-6) in synovial fluid (SF) and TLR4 endogenous ligands (TNC and MRP8/14) in serum (
18‐
22). In addition, they also have increased TLR2 and 4 expression on monocytes in blood and SF (
8).
Though ERA and AS show similarities in their clinical features, HLA-B27 association and some of the immune abnormalities, there are differences like higher prevalence of enthesitis and arthritis in juvenile onset SpA and a higher prevalence of IBP and uveitis in AS (
23,
24) Thus, we aimed to see if the monocyte response seen in these two diseases is similar or different. To study this, we analyzed the frequency of cytokine producing monocytes in peripheral blood (PB) and synovial fluid mononuclear cells (SFMC) at baseline as well as on stimulation with both exogenous and endogenous TLR ligands. In addition, we also measured the production of cytokines as well as endogenous TLR ligands at baseline as well as on stimulation. Finally, we assessed the mRNA expression levels post-stimulation with the endogenous and exogenous TLR ligands.
Patients and methods
Patients
Since SpA is a male-dominant disease, and also to avoid gender difference as a potential confounder, we enrolled only male subjects. Adult axial SpA patients fulfilling the Assessment of SpondyloArthritis international Society [ASAS] classification criteria for axial SpA (
25) and ERA patients fulfilling the JIA-International League of Associations for Rheumatology (ILAR) classification criteria (
26) were included as study subjects. Young adult males were included as HC. None of the patient’s with ERA had psoriasis or inflammatory bowel disease.
Clinical assessments included Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) (
27) for adult axial SpA patients and Juvenile Spondyloarthropathy Disease Activity Score (JSpADA) (
28) for ERA patients. These were assessed by the treating rheumatologist at the time of sample collection.
The PB was collected from all subjects whereas SF was collected from only those patients who required intra-articular steroid injection as a part of their treatment. All experiments were performed on whole blood (WB) as it is difficult to get a higher quantity of blood from children.
Assessment of frequency of pro-inflammatory cytokine producing monocytes
500 μl PB was diluted (1:1) with RPMI (Sigma Aldrich, MO, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, MA, USA) and 1% antibiotic (Antibiotic-antimycotic, Gibco, MA, USA) (complete culture medium) and cultured in 24-well plates for 4 h in a 5% CO2 incubator at 37 °C.
Stimulation of cells with TLR ligands
Cells were stimulated with following ligands for 4 h: LPS (100 ng/ml; a TLR4 ligand, Sigma Aldrich, MO, USA), Peptidoglycan (PG, 5 μg/ml; a TLR2 ligand, Sigma Aldrich, MO, USA), MRP8 (5 μg/ml; Abcam, Cambridge, UK) or TNC (10 μg/ml; Merck Millipore, MA, USA). Unstimulated cells served as baseline. 10 μg/ml Brefeldin A (Sigma Aldrich, MO, USA) was added as secretion inhibitor in all cultures.
Staining of cells and flow cytometry
Cells were surface-stained with fluorescent monoclonal antibody to CD14 (BD Biosciences, CA, USA). After RBC lysis (BD FACS lysis solution, BD Biosciences, CA, USA), cells were fixed and permeabilized (with BD Cytofix/Cytoperm solution, BD Biosciences, CA, USA). The cells were later stained for intracellular IL-6 and TNF, using fluorescent monoclonal antibodies (BD Biosciences, CA, USA). 10
5 cells were acquired in the flow cytometer (Beckman Coulter, CA USA). The monocytes were gated in the side scatter (SSC) vs CD14 plot. In the CD14
+ gate, cells were analysed and the frequency of CD14
+ IL-6
+ and CD14
+ TNF
+ cells was calculated using Navios software (Beckman Coulter, CA, USA). Gating strategy is shown in Additional file
1.
In vitro production of pro-inflammatory cytokines and MMP3
500 μl PB was diluted (1:1) with complete culture medium and dispensed in 24-well plates. Cells were stimulated with following ligands (in a 5% CO2 incubator at 37 °C) for 24 h: LPS (2.5 μg/ml), PG (5 μg/ml), MRP8 (5 μg/ml), TNC (10 μg/ml). Unstimulated cells served as the baseline. In the culture supernatants, levels of TNF (BD OptEIA Kit, CA, USA), IL-6 (BD OptEIA Kit, CA, USA) and MMP3 (R&D systems, MN, USA) were measured by enzyme linked immunosorbent assay (ELISA), as per the manufacturer’s instructions. The minimum detection limit was 7.8 pg/ml for TNF, 4.7 pg/ml for IL-6 and 31.3 pg/ml for MMP3.
Validation of pro-inflammatory cytokine production by quantitative PCR
In a subset of patients and HC (n = 5), 500 μl PB was diluted (1:1) with complete culture medium and dispensed in 24-well plates. Cells were stimulated with following ligands (in a 5% CO2 incubator at 37 °C) for 4 h: LPS (2.5 μg/ml), MRP8 (5 μg/ml) and TNC (10 μg/ml). Unstimulated cells served as baseline. The cells were later stored in TRIzol at − 80 °C (Thermo Fischer Scientific, MA, USA) till RNA isolation.
After thawing the cells and vortexing to facilitate lysis, the sample was centrifuged at 10,000×g for 10 mins at 4 °C. The supernatant was carefully aspirated and transferred to fresh microcentrifuge tubes. Following this, 200 μl of chloroform (Sigma Aldrich, MO, USA) was added to the suspension, mixed manually for 15 mins and kept at room temperature for 10 mins. It was then centrifuged at 12,000×g for 15 min at 4 °C. The aqueous layer was carefully aspirated and transferred to fresh microcentrifuge tubes. 500 μl of isopropyl-alcohol (Sigma Aldrich, MO, USA) was added to the separated aqueous layer and mixed. The suspension was centrifuged at 12,000×g for 10 mins at 4 °C. The RNA pellet was then washed twice with 1 ml of 75% ethanol (Merck Millipore, MA, USA) and centrifuged at 7500×g. The pellet was then dried at RT. Finally, the RNA pellet was dissolved in 10 μl of RNase free water. The absorbance was taken at 260 and 280 nm in Nanodrop spectrophotometer (Thermo Fisher, MA, USA).
cDNA was prepared using High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, CA, USA) as done in previous studies [
6]. Subsequently, real-time PCR was performed using TaqMan Fast Advanced Master mix (Thermo Fischer, MA, USA). Taqman gene expression assay kits were purchased from Applied Biosystems (CA, USA), the IDs being Hs00174128_m1 (TNF), Hs00174131_m1 (IL-6) and HS02786624_g1 (GAPDH). Each 20 μl reaction mixture comprised of 10 μl TaqMan Fast Advanced Master mix, 1 μl TaqMan assay probe, 7 μl RNase free water and 2 μl cDNA. The reaction conditions in the real-time PCR amplification and detection instrument (LightCycler 480 Instrument II, Roche Molecular Systems Inc., CA, USA) were 50 °C for 2 min, 95 °C for 2 min and finally 40 PCR cycles of 95 °C for 3 s and 60 °C for 30 s.
GAPDH was used as a housekeeping gene. Relative fold change was determined by the ΔΔCt method (Ct = cycle threshold), where fold change =2-ΔΔCt and ΔΔCt = [Ct(TNF/IL-6)-CtGAPDH] for stimulated (LPS/TNC/MRP8) sample - [Ct(TNF/IL-6)-CtGAPDH] for unstimulated sample. More than a 2-fold increase in expression was considered significant.
In vitro production of TNC and MRP8/14 in response to LPS
500 μl PB was diluted (1:1) with complete culture medium and dispensed in 24 well plates. The cells were stimulated (24 h in a 5% CO2 incubator at 37 °C) with either LPS (2.5 μg/ml) or left unstimulated (baseline). Following this, levels of MRP8/14 (LEGEND MAX Human MRP8/14 Calprotectin ELISA kit, BioLegend, CA, USA) and TNC [Tenascin C large (FNIIIC), IBL International; Hamburg, Germany] in culture supernatants were measured by ELISA as per the manufacturer’s instructions. The minimum detection limit of MRP8/14 was 3.13 ng/ml and that of TNC was 0.38 ng/ml.
Studies on synovial fluid cells
The SFMC’s were isolated by density gradient centrifugation using Histopaque 1077 (Sigma, MO, USA). Inflammatory response of SFMC was studied using the protocol described above for PB, except that 106 SFMC/ml were used. Cells were stimulated with LPS, PG, TNC, MRP8 or left unstimulated (baseline); following which the TNF and IL-6 producing monocyte frequencies were measured as per the above-mentioned protocol for flow cytometry.
Statistical analysis
All results are represented as median [Interquartile range (IQR)]. Intergroup comparison was done using non-parametric tests. Correlation with disease activity was assessed using Spearman’s correlation and values are expressed as r (95% confidence intervals). p-value < 0.05 was taken as significant. Exact p values are given unless it exceeded p < 0.0001. Graph pad prism 7 (trial version) was used for all statistical analysis.
Discussion
Both ERA and adult axial SpA patients had similar monocyte responses with minor differences. Both had higher frequencies of pro inflammatory cytokine producing monocytes as well as TNF, IL-6, MMP3, TNC and MRP8/14 production at baseline and post-stimulation with TLR exogenous and endogenous ligands as compared to HC. The cytokine producing monocyte frequency at baseline also correlated with the disease activity. Stimulation with TLR ligands led to similar increase in frequency of TNF producing monocytes in SFMC’s. However, the IL-6 producing monocyte frequency was higher in ERA patients than adult axial SpA post stimulation with exogenous ligands of TLR4.
Adult axial SpA and ERA patients had higher frequency of cytokine producing monocyte frequency at baseline as compared to HC. This suggests towards the presence of pre-activated monocytes in both patient subsets. Previously in patients with adult axial SpA, elevated frequencies of pro-inflammatory cytokine producing monocytes was observed. However, this was observed only in patients receiving conventional disease modifying anti-rheumatic drugs (DMARDs) and not in those receiving anti-TNF therapy (
29). Methotrexate (MTX) has not been observed to inhibit cytokine production by monocytes (
30) while prednisolone has been shown to reduce TNF production from monocytes (
31). None of the patients included in the present study were receiving anti-TNF therapy and a small proportion of them were on MTX and prednisolone.
Positive correlation of the baseline pro-inflammatory cytokine producing monocyte frequency with disease activity was observed for both adult axial SpA and ERA patients. This suggests that cytokines produced by monocytes contribute to inflammation. An association between the serum levels of inflammatory mediators and disease activity in AS has been reported (
32,
33) Moreover, therapeutic success of anti-TNF agents in both AS and ERA further supports the role of pro-inflammatory cytokines in SpA (
34,
35).
The elevated production of IL-6 as well as MMP3 from ERA patients at baseline as compared to adult axial SpA may indicate that the monocytes from ERA patients are more activated and this could be related to higher disease activity as reflected by presence of more active peripheral arthritis and enthesitis in ERA patients as compared to adult axial SpA. The real reason behind the difference in IL-6 production compared to IL-6 producing monocyte frequency in the patient subsets is difficult to explain. However, some other blood cells along with monocytes may have contributed to IL-6 production in WB cultures whereas in flow cytometry, we have assessed only the IL-6 producing monocyte frequency.
The present study also provides the data that endogenous ligands like MRP8/14 and TNC also activate monocytes in patients with adult axial SpA and ERA. Here again, monocytes from ERA patients showed a slightly higher response as compared to adult axial SpA. MRP8/14 has been observed to be expressed in monocytes and infiltrating neutrophils in the inflamed joints of JIA patients (
36). Stimulation of PBMCs from healthy subjects with MRP8/14 has already been shown to cause production of pro-inflammatory cytokines (
37). TNC promotes the recruitment of monocytes/macrophages to the site of inflammation (
38). It also binds to TLR4 on monocyte/macrophages and causes production of pro-inflammatory cytokines in a dose dependent manner (
39). These cytokines can induce production of TNC through an ‘auto amplification’ loop (
40).
Endogenous ligands were less potent then LPS in stimulating monocytes. TNC has also been shown to elicit less potent cytokine response than LPS in macrophages (
39). It is known that endogenous ligands like TNC enhance expression of extracellular matrix proteins which work towards tissue repair (
41), whereas LPS activates several downstream signalling pathways, causing production of pro-inflammatory cytokines (
42).
Increased production of MMP3 on stimulation with endogenous ligands suggests that MMP3 production could be another mechanism by which the activated monocytes cause joint damage. MMP3 is a potent protease which causes degradation of matrix as well as cartilage, leading to joint damage. Children suffering from ERA have increased levels of MMP3 in the serum (
43).
We have shown that endogenous ligands induce pro-inflammatory response from monocytes in patients and it has been reported that both MRP8/14 and TNC levels are elevated in patients with ERA and AS (
12,
13,
18,
21). We studied the production of these ligands on stimulation of monocytes and found that indeed, patients cells produced higher levels of TNC & MRP8/14 as compared to HC. MRP8/14 and TNC are produced by monocytes upon stimulation by microbes or stress (
44,
45).
SFMC’s from both adult axial SpA and ERA patients had similar frequency of TNF producing monocytes after stimulation. However, the presence of higher IL-6 producing monocyte frequency from ERA patients than adult axial SpA post stimulation with endogenous ligands suggests that at the local site, the monocyte/macrophage lineage cells are more activated in ERA patients. This could be due to selective homing of activated cells in the synovium or their higher activation at local site due to excessive production of endogenous ligands. Indeed, levels of MRP8/14 in SF are higher than plasma in children with ERA (
21).
The strength of this study includes a good sample size, validation of data using three methods, i.e., frequency of cytokine producing monocytes, cytokine levels in culture supernatants and mRNA quantification, and inclusion of SF samples. The limitations of this study are lack of healthy children control, significant disease duration at time of inclusion, use of glucocorticoids and MTX by a proportion of patients at the time of analysis. However, the abnormalities seen are similar in juvenile and adult SpA, suggesting that they are not influenced by the age of patients but are specific to the disease. Though glucocorticoids and MTX can affect cytokine production however, we did not find any difference in the two groups (data not shown). It would be ideal to study cytokine production by monocytes separated from WB. However, this requires a minimum of 15–20 ml blood to get 1.5–2 million monocytes which is difficult in children. Thus, monocyte response is mostly similar in ERA and adult axial SpA.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.