Introduction
Rheumatoid arthritis (RA) is a chronic, inflammatory, autoimmune disease characterized by persistent synovitis and hyperplasia of the joint synovium, development of pannus, and invasion of leukocytes into the joint followed by destruction of local articular components such as cartilage and bone [
1,
2]. In the RA synovium a variety of cell types can be found, specifically T cells, B cells, macrophages and dendritic cells (DCs) [
3,
4]. DCs derive from two sources: stem cells in the bone marrow, and precursor cells found in the circulation. In humans there are four major groups of DCs so far characterized: myeloid DCs (mDCs), plasmacytoid DCs (pDCs), migratory DCs such as Langerhans cells and dermal DCs, and monocyte-derived DCs (mo-DC) [
5]. Although DCs represent a relatively small subset of immune cells, they are widely distributed throughout lymphoid and nonlymphoid tissues [
6].
DCs have a crucial role in the initiation of primary immune responses. Individuals with autoimmune disease show a high number of aberrantly activated DCs either in circulation or in the autoimmune lesions, secreting large amounts of proinflammatory cytokines that mediate inflammation and differentiation of pathogenic T-helper type 1 and T-helper type 17 cells [
7]. Rheumatoid synovial DCs have been described as having a more mature, differentiated phenotype, expressing high levels of HLA-DR, CD86 and nuclear RelB, and have been observed to associate with T cells in perivascular mononuclear cell aggregates surrounding the postcapillary venules, and in germinal center-like structures [
8]. In addition, the RA synovium contains abundant immature mDCs and pDCs that express cytokines (interleukin (IL)-12, IL-15, IL-18, and IL-23), HLA class II molecules, and costimulatory molecules that are necessary for T-cell activation and antigen presentation [
9]. In the synovial fluid (SF), DCs exhibit a semi-mature phenotype showing low levels of CD80 and CD83 expression [
9]. An important sequel of continued antigenic stimulation via DCs is the formation of lymphoid structures at the site of inflammation. By coordinating the recruitment and/or activation of other immune cells, DCs can drive the generation of ectopic lymphoid tissues, as in the case of inflamed synovia in RA and systemic lupus erythematosus [
10].
FMS-related tyrosine kinase 3 ligand (Flt3L) is crucial for steady-state pDC and mDC development. Mice lacking Flt3L have reduced numbers of DCs [
11], as do mice that are deficient in signal transducer and activator of transcription 3 [
12], which is an important molecule in the Flt3L signaling cascade. Conversely, administration of Flt3L to mice or humans leads to a dramatic increase in DC numbers both in lymphoid and nonlymphoid organs [
13]. Flt3L is abundantly expressed in most human tissues, as a membrane-bound form and/or as a secreted form. Flt3L is initially synthesized as a membrane-bound protein, which must be cleaved to become a soluble growth factor. The extracellular domain alone has been shown to be sufficient for bioactivity [
14]. Ectodomain shedding of Flt3L is metalloproteinase dependent and is mediated by tumor necrosis factor-converting enzyme (TACE) [
15], a type 1 membrane protein belonging to a large family of transmembrane metalloproteases (a disintegrin and metalloprotease domain gene family) that was originally identified as the enzyme responsible for the cleavage of pro-tumor necrosis factor (TNF) alpha [
16], but also has numerous additional substrates and functions, including a critical role in activating the ligands of the epidermal growth factor receptor, and in the modulation of immune reactions [
17]. The receptor for Flt3L, CD135 is a transmembrane receptor tyrosine kinase expressed in bone marrow cells during the early stages of hematopoiesis [
18], where it is involved in the control of maintenance, expansion, mobilization and differentiation of progenitor cells [
19]. CD135 is required for DC homeostasis, and inhibition of CD135-mediated signals results in fewer DCs [
20]. The effects of CD135 deficiency are most evident in the periphery, where this receptor is essential for the homeostatic expansion of DC progenitor populations in lymphoid organs [
21].
Flt3L has been shown to accumulate in RA SF and induces arthritis when injected into healthy mouse knee joints. In addition, administration of Flt3L worsens experimental arthritis, while tyrosine kinase inhibitors that target CD135 alleviate experimental arthritis in mice models [
22,
23]. Given the relevance of Flt3L/CD135 in early hematopoiesis and DC generation, and possible involvement in RA, we characterized in detail the expression of both receptor and ligand in RA patients in comparison with healthy individuals (HI) and non-RA disease controls.
Methods
Patients and controls
Patients with RA diagnosed according to the 2010 criteria defined by the European League Against Rheumatism [
24] were included in the study. Gout patients and HI were used as controls.
Peripheral blood mononuclear cells (PBMC) and synovial fluid mononuclear cells were isolated by gradient centrifugation with Lymphoprep (Axis-Shield PoPAS, Dieren, the Netherlands). Cells were frozen in fetal calf serum (Invitrogen, Breda, the Netherlands) containing 10% dimethyl sulfoxide (Sigma Aldrich, Zwijndrecht, the Netherlands) until further experimentation. SF samples were obtained by arthrocentesis of inflamed knee joints. Cell-free SF samples were stored at -80°C. Synovial tissue (ST) specimens were obtained during arthroscopy (2.7 mm arthroscope; Storz, Tuttlingen, Germany) under local anesthesia [
25]. The samples were snap frozen
en bloc in Tissue-Tek OCT (Miles Diagnostics, Elkhart, IN, USA). The frozen blocks were stored in liquid nitrogen. Cryostat sections (5 μm) were mounted on glass slides (Star Frost adhesive slides; Knittelgläser, Braunschweig, Germany). The glass slides were sealed and stored at -80°C until immunohistological analysis. Demographic and clinical data of the patients are presented in Table
1. All patients gave written informed consent before inclusion in the study, and the study was approved by the Local Ethics Committee of the Academic Medical Center, University of Amsterdam.
Table 1
Demographic and clinical characteristics of rheumatoid arthritis and gout synovial tissue immunohistochemistry
Age (years) | 56 ± 13 | 74 ± 17 |
Sex, female/male | 13/3 | 2/10 |
Disease duration (years) | 10.3 ± 14.4 | 2.2 ± 3.5 |
Swollen joint count | 7 ± 5 | 2 ± 3 |
C-reactive protein (mg/l) | 26.6 ± 21.2 | 18.2 ± 17.75 |
Disease Activity Score in 28 joints | 5.0 ± 1.2 | nd |
Erythrocyte sedimentation rate (mm/hour) | 40.2 ± 30.3 | 27.7 ± 19.2 |
IgM-RF (kU/l) | 493 ± 1406 | nd |
Anti-CCP (kAU/ml) | 2472 ± 3600 | nd |
Number taking NSAIDS (positive/negative) | 9/15 | 5/12 |
Number taking corticosteroids (positive/negative) | 0/15 | 0/12 |
Number taking DMARDs (positive/negative) | 0/15 | 0/12 |
Number taking anti-tumor necrosis factor (positive/negative) | 0/15 | 0/12 |
Enzyme-linked immunosorbent assay
Serum and SF levels of Flt3L were determined by enzyme-linked immunosorbent assay (R&D Systems, Abingdon, UK) following the manufacturer’s instructions. Demographic and clinical data of the patients used in each experiment are presented in Table
2. To determine the relationship between serum Flt3L levels and clinical response in RA patients, Flt3L serum levels were also measured in patients who started treatment with either a different regimen of glucocorticoids or adalimumab.
Table 2
Demographic and clinical characteristics of rheumatoid arthritis paired serum and synovial fluid (enzyme-linked immunosorbent assay)
Age (years) | 58 ± 14 |
Sex, female/male | 5/2 |
Disease duration (years) | 15 ± 6 |
Swollen joint count | 4 ± 4 |
C-reactive protein (mg/l) | 20 ± 25 |
Disease Activity Score in 28 joints | 4.3 ± 2.5 |
Erythrocyte sedimentation rate (mm/hour) | 31.1 ± 23.1 |
IgM-RF (kU/l) | 117.25 ± 273.1 |
Anti-CCP (kAU/ml) | 290.7 ± 447.9 |
Number taking NSAIDS (positive/negative) | 1/9 |
Number taking corticosteroids (positive/negative) | 6/9 |
Number taking DMARDs (positive/negative) | 0/9 |
Number taking anti-tumor necrosis factor (positive/negative) | 2/9 |
Patients treated with high-dose glucocorticoids
Nine patients from the active arm of a previously conducted, double-blind, randomized, placebo-controlled trial were treated with 60 mg oral prednisolone daily for 1 week followed by 40 mg prednisolone daily during the second week [
26]. Flt3L serum levels were measured at baseline and after 2 weeks. In this study, response was defined as a decrease in Disease Activity Score in 28 joints ≥1.2 after 2 weeks of glucocorticoid (prednisolone) treatment.
Patients treated with adalimumab
Baseline demographic and clinical features of patients from the larger open-label, prospective, single-center adalimumab clinical trial have been described previously [
27]. Forty-eight patients were included for the present analysis. All patients received 40 mg adalimumab subcutaneously every other week, in combination with a stable methotrexate dose for at least 16 weeks. Use of oral glucocorticoids (prednisone ≤10 mg/day) was allowed. Clinical response at 16 weeks was determined according to the European League Against Rheumatism response criteria [
24].
Flow cytometry
The expression of specific markers was investigated in PBMC/synovial fluid mononuclear cells by fluorescence-activated cell sorting analysis after surface or intracellular staining with specific antibodies that were conjugated to different fluorescent dyes. For the extracellular staining, cells were washed in phosphate-buffered saline containing 1% bovine serum albumin and 0.02% sodium azide and were incubated with specific antibodies for 30 minutes at 4°C. Intracellular staining for Flt3L was performed after staining for surface markers was completed (T cell, B cell, natural killer (NK) cell and DC markers) using paraformaldehyde 4% as a fixation method followed by saponin permeabilization and Flt3L staining for 30 minutes at 4°C. For further details see Additional file
1. Flow cytometry was performed using a FACS CANTO (Becton Dickinson, Breda, the Netherlands) and analyzed with Flowjo analysis software (Tree Star, Ashland, OR, USA).
Immunohistochemistry
Briefly, endogenous peroxidase activity was inhibited in the acetone-fixed sections by 0.1% sodium azide and 0.3% hydrogen peroxide in phosphate-buffered saline. Sections were stained using mouse monoclonal antibodies against CD68 (clone EBM-11; Dako, Heverlee, the Netherlands), CD163 (clone 5cFAT; BMA Biomedicals, Augst, Switzerland) or CD135 (clone BV10A4H2; eBiosciences, Vienna, Austria). Sections were sequentially incubated with a secondary horseradish peroxidase-labeled antibody, followed by horseradish peroxidase detection using the AEC kit (Brunschwig, Amsterdam, the Netherlands), and hematoxylin (Klinipath, Duiven, the Netherlands) as the counterstain. Parallel sections were incubated with isotype-matched and concentration-matched monoclonal antibodies as negative controls. After immunohistochemical staining, coded sections stained for CD135, CD68 or CD163 were analyzed in a random order by computer-assisted image analysis [
28]. For all markers, 18 high-power fields were analyzed. Images were analyzed with the Qwin analysis system (Leica, Cambridge, UK).
Immunofluorescence
Frozen ST sections were fixed in acetone and blocked with 10% human serum (Dako, Glostrup, Denmark), followed by incubation with mouse purified monoclonal antibody against TACE (R&D Systems ) for 1 hour. After washing with phosphate-buffered saline/bovine serum albumin 1%, sections were incubated with a secondary horseradish peroxidase-labeled antibody for 30 minutes. For detection of the first primary antibody, a biotin-conjugated tyramide signal amplification (PerkinElmer Life Sciences, Boston, MA, USA) was used followed by streptavidin Alexa 594 antibody (Invitrogen, Bleiswijk, the Netherlands). After blocking with 10% mouse serum (Dako), the sections were incubated with fluorescein isothiocyanate-labeled mouse monoclonal antibodies against CD68 (BioLegend, London, UK), CD163 (BioLegend), CD19 (eBiosciences), CD55 (Becton Dickinson), CD3 (eBiosciences), CD31 (eBiosciences), or purified rabbit polyclonal von Willebrand factor (Dako) followed by a secondary antibody labeled with Alexa 488. The slides were mounted with Vectashield containing diamidino-2-phenylindole (Vector Laboratories, Burlingame, CA, USA) and were analyzed on a fluorescent imaging microscope (Leica DMRA, Wetzlar, Germany) coupled to a charge-coupled device camera.
Monocyte purification, macrophage and dendritic cell differentiation
PBMCs were isolated from volunteer donor blood buffy coats (Sanquin, Amsterdam, the Netherlands) by gradient centrifugation with Lymphoprep (Axis-Shield PoPAS), and monocytes were further isolated by Percoll gradient separation (GE Healthcare, Zeist, the Netherlands). Differentiation of monocytes into macrophages was performed in IMDM/10% fetal calf serum supplemented with 100 μg/ml gentamycin (Invitrogen), in the presence of granulocyte–macrophage colony-stimulating factor (5 ng/ml), macrophage colony-stimulating factor (25 ng/ml), interferon-gamma (IFNγ, 10 ng/ml) or IL-10 (10 ng/ml) (all from R&D Systems) for 7 days. mo-DCs were differentiated in IMDM/5% fetal calf serum supplemented with 100 μg/ml gentamycin (Invitrogen), in the presence of granulocyte–macrophage colony-stimulating factor (50 ng/ml) and IL-4 (100 ng/ml) for 6 days.
Quantitative measurement of mRNA expression
Gene expression (mRNA) in synovial biopsies from RA and gout patients, polarized macrophages and mo-DC was assessed by quantitative polymerase chain reaction (qPCR) as described in detail in Table
3. Demographic and clinical data of the patients used in each experiment are presented in Table
3.
Table 3
Demographic and clinical characteristics of rheumatoid arthritis and gout synovial tissue (quantitative polymerase chain reaction)
Age (years) | 60 ± 8 | 62 ± 14 |
Sex, female/male | 13/9 | 2/10 |
Disease duration (years) | 13 ± 14 | 2.3 ± 1.7 |
Swollen joint count | 7 ± 5 | 2 ± 3 |
C-reactive protein (mg/l) | 20 ± 26 | 36 ± 31 |
Disease Activity Score in 28 joints | 5.0 ± 1.2 | nd |
Erythrocyte sedimentation rate (mm/hour) | 38 ± 31 | 35 ± 20 |
IgM-RF (kU/l) | 464 ± 1366 | nd |
Anti-CCP (kAU/ml) | 2197 ± 3475 | nd |
Number taking NSAIDS (positive/negative) | 12/22 | 4/12 |
Number taking corticosteroids (positive/negative) | 9/22 | 1/12 |
Number taking DMARDs (positive/negative) | 12/22 | 0/12 |
Number taking anti-tumor necrosis factor (positive/negative) | 6/22 | 0/12 |
Statistical evaluation
All analyses were performed using Prism software (GraphPad, La Jolla, CA, USA). Flt3L levels in paired serum and SF from RA patients and controls were compared using the Wilcoxon matched-pairs test and the Mann–Whitney U test, respectively. Results from qPCR and immunohistochemistry were analyzed using the Mann–Whitney U test. Results from fluorescence-activated cell sorting analysis and qPCR for polarized macrophages and mo-DCs were analyzed using the Kruskal–Wallis test. P <0.05 was considered statistically significant.
Discussion
In the present study we show in RA patients the expression of Flt3L and its receptor in three different compartments: blood, SF and ST. Flt3L was significantly elevated in RA SF compared with paired serum, confirming previous observations [
23]. In addition, we reported for the first time that RA serum contains significantly elevated levels of Flt3L compared with HI serum. Moreover, we observed a higher expression of Flt3L (mRNA) in STs of RA patients compared with gout patients. RA and gout samples were matched in terms of macrophage numbers and IL-6 and IL-8 expression. This might indicate that Flt3L levels might reflect disease specificity more than just inflammation.
A detailed analysis of cellular components in paired blood and SF revealed that monocytes are the major cell population that expresses extracellular Flt3L. Monocytes are bone marrow-derived cells that mediate essential regulatory and effector functions in innate and adaptive immunity [
32]. Circulating PB monocytes migrate into tissues, where they differentiate into different effector cells such as macrophages, DC and osteoclasts [
33] that are of importance in RA pathology. The percentage of circulating CD14
+ Flt3L
+ monocytes is increased in RA patients, compared with HI. Moreover, in RA SF the percentage of Flt3L-expressing monocytes is significantly higher compared with paired PB. Importantly, RA SF monocytes have a superior capacity to express Flt3L on a per cell basis compared with PB (circulating) monocytes. Mobilization of preformed Flt3L from intracellular stores rather than
de novo synthesis might be responsible for increased Flt3L levels at the site of inflammation. Flt3L is a crucial growth factor for DC differentiation that leads to increased numbers of these cells both in mice and humans [
13]. Interestingly, Flt3L can substitute for macrophage colony-stimulating factor in support of osteoclast differentiation and function [
34], raising the possibility of a direct role for Flt3L in bone damage in RA.
The expression of the receptor for Flt3L, CD135, is elevated in RA PB monocytes compared with HI PB monocytes. In addition, we show that RA SF monocytes express higher levels of CD135 compared with paired PB monocytes. Coexpression of CD135 and its ligand in the same cell (in this case by RA monocytes) suggests a possible autocrine stimulatory mechanism, as already reported for primary acute myeloid leukemia (AML) [
35].
TACE is the major sheddase for Flt3L [
15]. Flt3L is primarily produced as a membrane-bound protein [
36] and soluble Flt3L is generated through ectodomain shedding [
15]. TACE has been implicated in RA due to its role in processing membrane-bound TNF to its soluble form [
37]. Previous studies have demonstrated a central role for TNF in RA, and early preclinical studies indicated that inhibition of TACE was beneficial for patients with arthritis [
38]. We observed that in RA ST the main sources of TACE were macrophages (both CD68
+ and CD163
+ populations), CD55
+ fibroblast-like synoviocytes, activated endothelial cells or infiltrating monocytes (CD31
+) and B cells (CD19
+). As soluble Flt3L levels are highly dependent on TACE activity, the above-mentioned cells might contribute to local levels of soluble Flt3L. Since the TACE expression level (mRNA) in RA and gout STs was the same and Flt3L levels were increased in RA ST, it is tempting to speculate that TACE biological activity in RA might be elevated compared with gout patients. This observation cannot be attributed to a differential degree of ST inflammation between RA and gout STs since the expression of the inflammatory cytokines IL-6 and IL-8 and macrophages numbers (CD68
+ and CD163
+ cells) was similar. Macrophages are critically involved in the pathogenesis of RA [
39]. Not only do they produce a variety of proinflammatory cytokines and chemokines, but macrophages also contribute to cartilage and bone destruction [
40]. In addition, it has been reported that the number of macrophages in RA ST correlates with bone damage and that increased numbers of macrophages are an early hallmark of active disease [
3]. One of the main features of macrophages is their high plasticity during development. The nomenclature of general macrophage polarization has been proposed in the last decade, in which M1 (classical, inflammatory) and M2 (alternative, anti-inflammatory) refer to the two extremes of a spectrum of possible macrophage activation status [
41]. M1 macrophages are mainly present in RA and are characterized by a proinflammatory phenotype, producing high levels of TNF, IL-1, IL-6, IL-12, reactive oxygen species, and low levels of IL-10 [
42]. Here we show for the first time that, in addition to the above-mentioned inflammatory mediators, Flt3L might be considered a specific marker for IFNγ-differentiated macrophages. Polarizing cytokines such as IFNγ might contribute to the high levels of Flt3L found in RA synovium by shifting the macrophage polarization into a M1-like phenotype. Overall, these data suggest that in RA, in addition to circulating monocytes, IFNγ-differentiated macrophages might be an important source of Flt3L.
Sublining macrophages are a reliable biomarker for response to therapy in RA [
43]. We have previously shown that oral prednisolone, an effective therapy in RA, was associated with a reduction in macrophage infiltration in ST [
26]. In this study we observed a marked reduction of Flt3L serum levels in RA patients after prednisolone treatment, and a significant correlation between the Disease Activity Score in 28 joints and Flt3L serum levels. In addition, in the responder group of RA patients treated with adalimumab we observed a trend toward reduced serum levels of Flt3L. The reduction of Flt3L serum levels observed after effective treatment might reflect the reduction in the numbers of the main source(s) of Flt3L: circulating monocytes and/or ST macrophages. Flt3L has recently been outlined within a panel of preclinical biomarkers of predictive value for the development of RA [
44]. Flt3L levels might therefore be valuable as a potential biomarker of inflammation or response to treatment.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
MIR, PPT and MCL were responsible for study conception and design. MIR, SGP, SA, BH and PB were responsible for acquisition of data. MIR, SGP, DMG, KAR and MCL were responsible for analysis and interpretation of data. MIR and MCL drafted the manuscript. All authors revised the manuscript critically for important intellectual content and approved the final version.