Reduced expression of monocyte CD200R is associated with enhanced proinflammatory cytokine production in sarcoidosis

In sarcoidosis, the proinflammatory cytokines interferon gamma, tumour necrosis factor and interleukin-6 are released by monocyte-derived macrophages and lymphocytes in the lungs and other affected tissues. Regulatory receptors expressed on monocytes and macrophages act to suppress cytokine production, and reduced expression of regulatory receptors may thus promote tissue inflammation. The aim of this study was to characterise the role of regulatory receptors on blood monocytes in patients with sarcoidosis. Cytokine release in response to stimulation of whole blood was measured in healthy controls and Caucasian non-smoking patients with sarcoidosis who were not taking disease modifying therapy. Expression of the regulatory molecules IL-10R, SIRP-α/β, CD47, CD200R, and CD200L was measured by flow cytometry, and functional activity was assessed using blocking antibodies. Stimulated whole blood and monocytes from patients with sarcoidosis produced more TNF and IL-6 compared with healthy controls. 52.9% of sarcoidosis patients had monocytes characterised by low expression of CD200R, compared with 11.7% of controls (p < 0.0001). Patients with low monocyte CD200R expression produced higher levels of proinflammatory cytokines. In functional studies, blocking the CD200 axis increased production of TNF and IL-6. Reduced expression of CD200R on monocytes may be a mechanism contributing to monocyte and macrophage hyper-activation in sarcoidosis.

Sarcoidosis monocytes produce more TNF and IL-6. A whole blood assay was used for ex vivo stimulation studies. Significantly higher concentrations of secreted TNF and IL-6 were found in stimulated whole blood from patients with sarcoidosis compared with healthy controls (Fig. 1). IFNγ and IL-10 were not significantly different between sarcoidosis patients and healthy controls (Fig. 1). When the kinetics of cytokine production in PHA-stimulated whole blood were measured, IFNγ and IL-10 were produced with kinetics commensurate with T lymphocyte activation, whereas TNF production was rapid, peaking at 3-6 hours and declining thereafter ( Supplementary Fig. S2). Similar kinetics have been observed by others for monocyte-derived TNF 20,21 . PHA is a T cell mitogen 22 and it also stimulates monocytes by cross-linking Toll-like receptors 23 . To further explore the relative contribution of T lymphocytes and monocytes to the enhanced cytokine release observed in sarcoidosis, whole blood was stimulated with SEA, a more selective T cell mitogen. IL-6 production in response to SEA was significantly lower in blood from sarcoidosis patients (Fig. 2), consistent with the T lymphocytopenia.
To confirm that monocytes were responsible for the enhanced TNF and IL-6 production observed in sarcoidosis patients following PHA stimulation, intracellular accumulation of TNF was quantified by flow cytometry in Brefeldin A-treated PBMCs. PHA led to a substantial accumulation of TNF in monocytes, but not lymphocytes (Fig. 3). In sarcoidosis monocytes, stimulated intracellular TNF accumulation was significantly greater than in healthy controls (Fig. 3). More IL-6 accumulation was also seen in sarcoidosis monocytes ( Supplementary Fig. S3).

Sarcoidosis patients have reduced CD200R expression on monocytes.
Monocyte-and macrophage-derived TNF can be regulated by a variety of cytokines and cell surface receptor-ligand interactions 15 . Therefore, IL-10R, SIRP-α, and CD200R on monocytes were examined to determine whether expression of these receptors was altered in patients with sarcoidosis. Expression profiles of IL-10R and SIRP-α/β were similar in sarcoidosis patients and healthy controls (Fig. 5). Similarly, expression of CD47, the ligand for SIRP-α/β, was the same in sarcoidosis and controls (data not shown).

Category
Healthy Sarcoidosis In contrast, CD200R expression on blood monocytes from patients with sarcoidosis was reduced, with patients clustering as high or low expressers (p < 0.05) (Fig. 6). Using a boundary defining low expression below the standard deviation of the healthy population (geometric mean < 7.28 fluorescence units), 52.9% of sarcoidosis patients were classified as CD200R low compared with 11.7% of healthy donors (P < 0.0001) (Fig. 6). Monocyte CD200R expression was not related to serum ACE activity, C-reactive protein, or plasma viscosity (data not shown). CD200R expression on T lymphocytes in sarcoidosis also polarised as either CD200R low or CD200R high in a bimodal distribution, compared with a unimodal distribution in healthy controls (Fig. S4).
Expression of the ligand CD200L was absent on monocytes, and the percentage of CD200L + T lymphocytes was similar in sarcoidosis patients and healthy subjects (data not shown). However, because of the T cell lymphocytopenia there were fewer T cells expressing CD200L in sarcoidosis blood (0.16 ± 0.03 × 10 6 compared with 0.39 ± 0.03 × 10 6 cells/ml, p < 0.0001).
CD200L/CD200R interactions regulate TNF and IL-6 production. To investigate how CD200L/ CD200R interactions regulate proinflammatory cytokine release, monocyte CD200R expression was correlated with stimulated cytokine release in whole blood assays and PBMC cultures. In sarcoidosis, IL-6 responses were significantly higher in subjects with low monocyte CD200R expression when compared with high CD200R expressers (Fig. 6), and a similar trend was observed for TNF. When data from sarcoidosis and healthy subjects were pooled there was a trend towards an inverse correlation between IL-6 release and monocyte CD200R expression (r = 0.39, p = 0.21) (Fig. 6).

Discussion
This study is the first to characterise the expression of the regulatory receptor CD200R and its ligand CD200L on blood mononuclear cells in sarcoidosis. We have demonstrated that the distribution of monocyte CD200R expression in a cohort of patients with sarcoidosis is distinctly different from healthy control subjects, with a significantly higher proportion of sarcoidosis patients exhibiting a CD200R low monocyte phenotype. Low monocyte CD200R expression was associated with heightened TNF and IL-6 production in response to PHA stimulation, and blockade of CD200R or CD200L in healthy controls led to increased TNF and IL-6 secretion by monocytes, recapitulating the hyper-activated monocyte state seen in sarcoidosis.
Monocyte CD200R expression was not associated with serum ACE activity, suggesting that CD200R is not simply a marker of total granuloma burden 25 . Other studies have reported reduced CD200R expression on monocytes in inflammatory arthritis 16 . However, CD200R expression in sarcoidosis did not correlate with serum C-reactive protein or plasma viscosity, signifying that a CD200R low monocyte phenotype is not a manifestation of systemic inflammation. Thus monocyte CD200R appears to reflect a unique aspect of disease activity in sarcoidosis. Further studies are required to assess its value in predicting disease progression and response to anti-inflammatory therapy. Whether a similar hyper-responsive phenotype is exhibited by local tissue macrophages in sarcoidosis has yet to be determined; responses in peripheral blood monocytes may not translate directly to the macrophages within granulomata. Murine models suggest that in lung inflammation, monocytes actively contribute to resident macrophage populations 4 . Lung macrophages are potent producers of TNF and IL-6 5,6 which contribute to the formation of sarcoid granulomata 8 .
Despite the evidence for active cellular immunity within tissue granulomata, previous research on peripheral blood in sarcoidosis has been subject to contradictory reports suggesting immune responses may be enhanced 26,27 , diminished 28,29 or similar to healthy populations 30 . Reports of suppressed peripheral blood T lymphocyte responses and skin test anergy to specific antigens and non-specific mitogens may be explained by reduced CD4+ T helper cells or increased regulatory T cells. In the present study, patients with sarcoidosis displayed a reduced peripheral blood immune response when T lymphocytes are selectively stimulated with SEA, consistent with the observed T lymphocytopenia. The increased responsiveness to PHA in sarcoidosis patients was observed particularly at higher concentrations of stimulant, which would be consistent with a dose-dependent model where regulation over a specific threshold is reduced.
The present study implicates blood monocytes as potentially important regulatory cells in sarcoidosis. Cranshaw et al. reported that sarcoidosis monocytes had reduced IL-10 production 11 . In the whole blood assay used in the present study, we did not demonstrate that the IL-10 regulatory axis was dysfunctional in sarcoidosis, and the expression profiles of regulatory SIRP-α/CD47 were also similar between sarcoidosis and healthy subjects. The choice of methods for cell isolation and stimulation could explain the differences in the findings, since Cranshaw et al. studied monocytes that had been isolated by magnetic bead negative selection and stimulated with lipopolysaccharide 11 . Depending on the nature of the stimulant or other factors, deficiencies in more than one monocyte regulatory pathway may contribute to immune hyper-activation in sarcoidosis. We also demonstrated that in patients with sarcoidosis the intermediate CD16 + CD14 ++ blood monocyte population was expanded. Since this monocyte subset has been shown to express the highest levels of the TNF receptor TNFR1 31 , monocytes in sarcoidosis may be more responsive to TNF.
Although the whole blood assay used in the present study has an advantage in minimally perturbing cells such as monocytes that are highly pleiotropic and respond rapidly to changing environments on isolation and in tissue culture, the assay does have some limitations. For example, due to T cell lymphocytopenia, fewer CD200L + lymphocytes will be present in sarcoidosis blood for potential ligation of monocyte CD200R.
Anti-inflammatory drug treatment in sarcoidosis is generally considered to be clinically indicated if, following an initial observation period, there is evidence of progressive disease. Corticosteroid treatment reduces macrophage TNF release and leads to improvement in disease markers, but the effect of steroids on long term Results are presented as mean ± SEM geometric mean fluorescence intensity; *p < 0.05, **p < 0.01 using twoway ANOVA with Bonferroni's post hoc test.
outcomes is uncertain and the benefits have to be balanced against the potential for serious side effects 32,33 . TNF-blocking biological therapies have been tested, with mixed results. Treatment with the TNF monoclonal antibody infliximab reduced tissue inflammation as measured by 18 F-fluorodeoxyglucose uptake on positron emission tomography imaging in an open label study 34 , and led to a small (2.5%) improvement in FVC at 24 weeks compared with placebo in patients with chronic pulmonary sarcoidosis in a randomised controlled trial 35 . Conversely, studies with TNF-blocking protein etanercept showed high treatment failure rates 36 , and anti-TNF therapy for rheumatic diseases has been linked in many reports to de novo development of sarcoidosis-like granulomata 37 . These findings indicate that therapeutic TNF blockade in sarcoidosis is not a panacea. Alternatively, CD200R is potentially targetable as an anti-inflammatory therapeutic strategy. CD200 recombinant proteins and agonistic antibodies have been reported to reduce macrophage activation and promote immunological tolerance in human cells in vitro and in mice in vivo [38][39][40][41] .
The present study provides evidence that the CD200R/CD200L axis is important in modulating proinflammatory cytokine release from monocytes in sarcoidosis, and raises the possibility that disease progression could be propagated by reduced monocyte CD200R expression. There is a need for improved treatment options for patients with sarcoidosis, and the CD200L/CD200R axis is potentially tractable for future therapy 42 .

Methods
Subjects. Ethical approval for the use of blood samples from healthy volunteers was obtained from the Ethics Committee, Hull York Medical School. Regional ethical approval was obtained to obtain blood and tissue samples (c) Percentages of CD14 ++ CD16 + intermediate monocytes, CD14 ++ CD16 − classical monocytes, and CD14 + CD16 ++ non-classical monocytes in patients with sarcoidosis (n = 14) and controls (n = 18). Results are presented as mean ± SEM; *p < 0.05, **p < 0.01 using two-way ANOVA with Sidak's Post hoc test.
Scientific RepoRts | 6:38689 | DOI: 10.1038/srep38689 from sarcoidosis patients from the Hull and East Riding local research ethics committee (LREC 08/H1304/54). Written informed consent was obtained from all subjects, and all methods were performed in accordance with the relevant guidelines and regulations of Good Clinical Practice. Patients with sarcoidosis were diagnosed by multi-disciplinary assessment and displayed non-caseating granulomata on histopathological analysis. Patients with other granulomatous diseases were excluded. Exclusion criteria for sarcoidosis patients and healthy volunteers were concomitant inflammatory or immunological diseases, treatment with corticosteroids or other immune-modulatory therapies, and current tobacco smoking. Blood was collected in sodium heparin tubes (BD) and processed within 1 hour. The results of each set of experiments informed subsequent experimental protocols so not all patients and controls contributed to all of the experiments. To avoid bias, subjects included in each set of experiments were not selected in any way. The numbers of subjects contributing to each experiment are shown in the figure legends. Supplementary Tables S1 and S2.

Reagents and Antibodies. See
Peripheral blood mononuclear cell (PBMC) isolation. Heparinised whole blood was diluted 1:1 with sterile phosphate-buffered saline (PBS) and layered over Histopaque 1077 in a 1:1 ratio. The gradient was centrifuged at 600 × g for 30 minutes without braking. The PBMC layer was removed with a sterile pipette and washed twice before suspending in RPMI medium. Whole blood or isolated PBMCs were incubated at 37 °C in a 5% CO 2 atmosphere.
Whole blood cytokine release assay. Phytohaemagglutinin (PHA) or Staphylococcus aureus enterotoxin A (SEA) were added to 1 ml aliquots of fresh whole blood in microfuge tubes and incubated at 37 °C for 16 hrs (1-72 hrs for time course experiments). Following centrifugation at 3000 × g for 8 minutes, plasma was collected and stored at −20 °C. Isolated PBMCs were resuspended in the same volume of RPMI media as the original blood volume and incubated under the same conditions as whole blood. ELISA kits (Biolegend, San Diego, USA) were used to measure IFN-γ, IL-10, TNF, IL-6, IL-4 and IL-12(p70) according to the manufacturer's instructions. Plates were analysed using a Multiskan FC Microplate Photometer (Thermo Scientific, Waltham, USA). Results below the minimum detectable limits with 1:2 dilutions of plasma were recorded as zero.   (a) Change in IL-6 (pg/ml) in PBMCs and whole blood after incubation with CD200R (n = 5) or CD200L (n = 6) antibodies compared with isotype control IgG. Cells were stimulated with 100 μg/ml PHA for 16 hours. (b) Change in TNF (pg/ml) in PBMCs and whole blood after incubation with CD200R or CD200L antibodies compared with isotype control IgG. Cells were stimulated with 100 μg/ml PHA for 16 hours. Results are presented as differences from isotype control IgG; *p < 0.05 using paired t-tests.