GM-CSF Primes Proinflammatory Monocyte Responses in Ankylosing Spondylitis

Abstract

Objectives: GM-CSF is a pro-inflammatory cytokine with multiple actions predominantly on myeloid cells. Enhanced GM-CSF expression by lymphocytes from patients with Ankylosing Spondylitis (AS) has recently been described, however, its potential pathogenic role(s) in AS are unknown. Methods: The effects of GM-CSF on TNF, IL-23, and CCL17 production by blood, PBMCs and isolated CD14+ monocytes from AS patients and healthy controls (HCs) were studied using ELISA. Serum CCL17 and GM-CSF and T cell GM-CSF production were studied in AS patients including pre-and on TNFi therapy. Results: GM-CSF markedly increased TNF production by LPS-stimulated whole blood, peripheral blood mononuclear cells (PBMC) and purified monocytes from AS patients, with 2 h GM-CSF exposure sufficient for monocyte "priming." Blocking of GM-CSF significantly reduced the production of TNF by whole blood from AS patients but not HCs. GM-CSF priming increased IL-23 production from LPS-stimulated AS and HC whole blood 5-fold, with baseline and stimulated IL-23 levels being significantly higher in AS whole blood. GM-CSF also stimulated CCL17 production from AS and HC blood and CCL17 levels were elevated in AS plasma. GM-CSF could be detected in plasma from 14/46 (30%) AS patients compared to 3/18 (17%) HC. Conclusion: We provide evidence that GM-CSF primes TNF and IL-23 responses in myeloid cells from AS patients and HC. We also show CCL17 levels, downstream of GM-CSF, were elevated in plasma samples of AS patients. Taken together these observations are supportive of GM-CSF neutralization as a potential novel therapeutic approach for the treatment of AS.

Keywords: CCL17; IL-23; TNF-alpha; ankylosing spondylitis; monocyte GM-CSF.

Figure 1

GM-CSF potentiates TNF and IL-23 production by LPS-stimulated whole blood from AS patients and healthy controls. Fresh whole blood from AS patients or healthy controls was treated with 10 ng/ml recombinant GM-CSF (A,B), anti-GM-CSF or IgG control for 2 h (C,D), then stimulated with 10 ng/ml LPS overnight. TNF (A,C) or IL-23 (B,D) measured in supernatants using ELISA. Data are represented as mean and SEM of independent donors in independent experiments. p-value was calculated using paired t-test. *p < 0.05, **p < 0.01, ***p < 0.001, ns: not significant. (A) AS n = 6, HC n = 5. TNF and IL-23 concentrations in the absence of LPS were less than 90 and 40 pg/ml respectively, AS n = 8, HC n = 6, (C) AS n = 3, HC n = 4, (D) AS n = 4, HC n = 4.

Figure 2

GM-CSF-potentiation of TNF production by LPS-stimulated PBMCs is contributed primarily by CD14+ monocytes. GM-CSF potentiates the response of monocytes to agonists for multiple TLRs. (A) GM-CSF receptorα(CD116) expression on CD14+ monocytes. CyTOF analyzed by viSNE clustering analysis of all live CD116 (GM-CSF Receptor α) events from PBMCs from a HC. Representative of N = 1. (B,C) CD14+monocyte (n = 5 AS patients and HCs) or CD14-depleted PBMCs (n = 3 AS patients) were treated with GM-CSF for 2 h followed by LPS (10 ng/ml) stimulation overnight. (D) CD14+ monocytes isolated from 5 HCs were treated with GM-CSF then with 15 ng/ml Pam3Csk4 or 1 μg/mlRX848 (resiquimod) or 10 μM CpGODN overnight. TNF levels in supernatant were determined by ELISA. Data are represented as mean and SEM of independent experiments, p-value was calculated using paired t-test, *p < 0.05, **p < 0.01, ns: not significant.

Figure 3

Short term (2 h) GM-CSF exposure enhances cytokine production by stimulated monocytes and is associated with altered expression of multiple cytokine regulators. (A) CD14+ monocytes were isolated from 5 AS patients or (B) 7 HCs. 1 million cells/ml CD14 monocytes were treated with GM-CSF for 2 h (and then washed) or overnight. Cells were then treated with LPS stimulation overnight. TNF level in supernatant was determined by ELISA. (C,D) Expression levels of IRF4 and IRF5 in GM-CSF-treated and untreated monocytes were measured using taqman qPCR. Red lines indicate AS patients, green lines indicate HCs. (E) CD14+ monocytes isolated from 4 AS and 4 HCs were treated with 10 ng/ml GM-CSF for 2 h. The effect of GM-CSF on genes in JAK/STAT signaling pathway was measured using the ProfilerRT2 PCR assay. The statistics for Profiler RT2 PCR array was carried out using QIAGEN data analysis website, P < 0.05 and fold change >1.5 was considered significant. Red dots indicate up-regulated genes, and blue dots indicate down-regulated genes. For (A,B), data are represented as mean and SEM of independent donors, P-value was calculated using one way ANOVA and Dunnett's multiple comparison test. For (C,D), p-value was calculated using paired t-test. *p < 0.05, **p < 0.01, ****p < 0.0001.

Figure 4

GM-CSF increases the production of CCL17 by whole blood from AS and HC. CCL17 is elevated in AS plasma and is reduced after TNFi treatment. Whole blood from AS patients or healthy controls was treated with 10 ng/ml recombinant GM-CSF for 2 h (A), then cultured overnight. Supernatant was measured for CCL17 using ELISA. Data are represented as mean and SEM of independent donors, p-value was calculated using paired t-test. *p < 0.05. (A) AS, HC n = 5. (B) Plasma CCL17 levels from 36 AS, 18 HC and 18 RA. For AS, 18 of them were biologics naive and 18 of them were on biologics. P-value was calculated using one-way ANOVA on the log-transformed data. (C) 8 paired pre- and on TNFi treatment AS plasma samples were collected, CCL17 in plasma was determined by ELISA. Red dots show TNFi non-responders, and green dots show TNFi responders. (D) Whole blood, PBMCs, or CD14+ monocytes (n = 3) from healthy controls was treated with 10 ng/ml recombinant GM-CSF then cultured overnight. Supernatant was collected and measured for CCL17 production using ELISA. Data was normalized as pg/ml/0.1 million CD14+ monocytes. Data are represented as mean and SEM of independent donors, p-value was calculated using paired t-test. *p < 0.05, **p < 0.01, ***p < 0.001, ns: not significant.

Figure 5

Cartoon model suggesting possible role of GM-CSF and monocytes in AS pathogenesis. Note this inflammation amplification may occur in gut-associated lymphoid tissue, blood or within the joint.