CD83 expressed by macrophages is an important immune checkpoint molecule for the resolution of inflammation

Abstract

Excessive macrophage (Mφ) activation results in chronic inflammatory responses or autoimmune diseases. Therefore, identification of novel immune checkpoints on Mφ, which contribute to resolution of inflammation, is crucial for the development of new therapeutic agents. Herein, we identify CD83 as a marker for IL-4 stimulated pro-resolving alternatively activated Mφ (AAM). Using a conditional KO mouse (cKO), we show that CD83 is important for the phenotype and function of pro-resolving Mφ. CD83-deletion in IL-4 stimulated Mφ results in decreased levels of inhibitory receptors, such as CD200R and MSR-1, which correlates with a reduced phagocytic capacity. In addition, CD83-deficient Mφ upon IL-4 stimulation, show an altered STAT-6 phosphorylation pattern, which is characterized by reduced pSTAT-6 levels and expression of the target gene Gata3. Concomitantly, functional studies in IL-4 stimulated CD83 KO Mφ reveal an increased production of pro-inflammatory mediators, such as TNF-α, IL-6, CXCL1 and G-CSF. Furthermore, we show that CD83-deficient Mφ have enhanced capacities to stimulate the proliferation of allo-reactive T cells, which was accompanied by reduced frequencies of Tregs. In addition, we show that CD83 expressed by Mφ is important to limit the inflammatory phase using a full-thickness excision wound healing model, since inflammatory transcripts (e.g. Cxcl1, Il6) were increased, whilst resolving transcripts (e.g. Ym1, Cd200r, Msr-1) were decreased in wounds at day 3 after wound infliction, which reflects the CD83 resolving function on Mφ also in vivo. Consequently, this enhanced inflammatory milieu led to an altered tissue reconstitution after wound infliction. Thus, our data provide evidence that CD83 acts as a gatekeeper for the phenotype and function of pro-resolving Mφ.

Keywords: CD83; STAT-6; checkpoint molecule; macrophages; resolution of inflammation; wound healing.

Figure 1

Analyses of CD83 surface expression by murine bone-marrow derived Mφ using different stimuli. Murine bone-marrow derived macrophages were generated and CD83 expression levels were analyzed after inflammatory or alternative activation. (A) Flow cytometric analyses show no CD83 expression on murine Mφ upon stimulation with IFN-ɣ (300 U/ml), LPS (100ng/ml), TNF-α (300 U/ml) for 16h. In contrast, stimulation with IL-4 (40 ng/ml), IL-13 (40 ng/ml) or IL4+IL-13 resulted in high CD83 expression on the surface of Mφ at the 16h time point (n ≥ 4) (left bar graph); Representative FACS histograms are presented (right graph). (B) qPCR analyses of Cd83 mRNA expression after different stimulations in murine Mφ. (C) Time-dependent regulation of CD83 expression on BMDM after inflammatory activation with IFN-ɣ (300 U/ml), LPS (100ng/ml), TNF-α (1000 U/ml) or Zymosan 10 µg/ml (n ≥ 4), analyzed by flow cytometry. (D) Time-dependent regulation of CD83 expression on BMDM upon stimulation with IL-4 (40 ng/ml), IL-13 (40 ng/ml) or IL-10 (10 ng/ml) (n ≥ 4), analyzed by flow cytometry. Gating strategy for the Mφ population is depicted in Supplementary Figure 2 . Data are represented as mean ± SEM. Statistical analysis was performed using a Two-way ANOVA or the appropriate corresponding non-parametric test. Experiments were performed at least three times. n.s., not significant, which indicates there is no statistical signficance; * p<0.05; ** < 0.01; *** p < 0.001; **** p < 0.0001.

Figure 2

Analyses of CD83 deficient murine Mφ. Mφ were generated from CD83wt or CD83 cKO mice and subsequently stimulated with IFN-γ or IL-4 for 16h or left untreated. (A) Cd83 expression levels were determined by qPCR and normalized to CD83wt BMDMs (n = 10). (B) Assessment of CD83 expression levels by flow cytometry (n = 20). (C) Assessment of knock-out efficiency in whole cell lysates from mock-, IFN-γ or IL-4 stimulated Mφ by Western blotting. β-actin served as a loading control. See full uncut gels in Supplemental Material (S1) (D) Cell viability assessment using flow cytometry (n = 24). (E) Differentiation efficacy assessing the percentage of F4/80⁺CD11b⁺ cells by flow cytometry, representing the Mφ population (n ≥ 24). (F) Expression levels of F4/80 and CD11b within the Mφ population (n ≥ 40). The gating strategy for the Mφ population is depicted in Supplementary Figure 2 . Statistical analyses were performed by One-way ANOVA or the appropriate corresponding non-parametric test. Data are represented as mean ± SEM. Experiments were performed at least three times. ***p<0.001; **** p< 0.0001. The absence of asterisks indicates that there is no statistical significance.

Figure 3

CD83 deficiency modulates the pro-resolving phenotype of IL-4 stimulated Mφ. BMDM were generated and differentiated into inflammatory CAM or AAM, using IFN-γ or IL-4 respectively, for 16h. (A) Assessment of surface MHC-II (n ≥ 47) and CD86 (n ≥ 47) expression levels by flow cytometry on stimulated murine wt and CD83-deficient Mφ. (B) qPCR analyses of Marchf1 and Marchf8 in Mφ derived from CD83wt or CD83 cKO mice (n = 5). (C) Analyses of Dectin-1 surface expression levels (left bar graph) and mRNA expression levels (right bar graph, via flow cytometry and qPCR, respectively). Significantly increased Dectin-1 expression levels on IL-4 stimulated CD83-deficient Mφ (n = 7). (D) Analyses of CD200R surface expression (left bar graph, n ≥ 29-31) and mRNA expression (right bar graph, n ≥ 8) revealed significantly decreased levels on IL-4 stimulated CD83-deficient Mφ (E) Analyses of MSR-1 surface expression (left bar graph, n ≥ 31) and Msr-1 mRNA expression (right bar graph, n = 6) revealed significantly decreased levels on IL-4 stimulated CD83-deficient Mφ. (F) Assessment of phagocytic activity via gentamicin protection assays revealed significantly decreased capacity to engulf E.coli. Statistical analyses were performed by One-way ANOVA or the appropriate corresponding non-parametric test. Data are represented as mean ± SEM. Experiments were performed at least three times. *p< 0.05; **p<0.01; ***p<0.001; **** p< 0.0001. The absence of asterisks indicates that there is no statistical significance.

Figure 4

STAT6 phosphorylation is altered in CD83-deficient macrophages upon IL-4 stimulation. Bone-marrow derived Mφ were generated from CD83wt or CD83 cKO mice and stimulated with IL-4 for 15 or 30 min. Unstimulated Mφ served as control. Subsequently, whole cell lysates were prepared and analyzed by Western blot. (A) Representative Western blot showing pSTAT-6, STAT6 and β Actin levels in whole cell lysates derived from CD83wt and CD83ΔMφ animals (B) Quantification of the ratio of pSTAT-6 and STAT6 normalized to β-actin (n ≥ 9). (C) qPCR analyses of Gata3 mRNA levels in Mφ generated from cKO or wt mice, stimulated with IFNg or IL-4. Significantly reduced Gata3 mRNA levels are observed in IL-4 stimulated CD83-deficient Mφ. Statistical analyses were performed by One-way ANOVA or the appropriate corresponding non-parametric test. Data are represented as mean ± SEM. Experiments were performed at least three times. *p< 0.05; **p<0.01. The absence of asterisks indicates that there is no statistical significance.

Figure 5

CD83-deficient IL-4- as well as IFN-γ-stimulated Mφ show a pro-inflammatory profile. Bone-marrow derived Mφ were generated and differentiated either into CAM or AAM, via IFN-γ or IL-4 respectively, or were left untreated for 16h. Afterwards, the supernatants were analyzed by CBA and cells via qPCR. (A) IL-4 stimulated CD83-deficient Mφ show increased secretion levels of IL-6, TNF-α, CXCL1 and G-CSF (upper bar graphs). qPCR analyses showed significantly increased mRNA levels of Il-6, Tnfa, Cxcl1 and Csf3 in IL-4-stimulated CD83 KO Mφ (lower bar graphs). (B) CCL5/RANTES and MCP-1 expression levels are increased in supernatants of IFN-γ-stimulated CD83-deficient Mφ. Statistical analyses were performed by One-way ANOVA or the appropriate corresponding non-parametric test. Data are represented as mean ± SEM. Experiments were performed at least three times. *p< 0.05; **p<0.01; **** p< 0.0001. The absence of asterisks indicates that there is no statistical significance.

Figure 6

CD83-deficient Mφ show enhanced capacity to stimulate allo-reactive T cells. Mφ were generated from CD83wt and CD83 cKO mice and differentiated using IFNγ or IL-4. Afterwards, the medium was discarded and splenocytes derived from BALB/c mice (4x10⁵ cells/well) were co-cultured with differentiated Mφ in 96-well plates, at different Mφ:splenocyte ratios, as indicated for 48h. T cell proliferation was assessed using tritium (A–C). Co-cultures of unstimulated, IFN-γ- and IL-4 stimulated Mφ, derived from CD83-deficient Mφ, show enhanced proliferative responses, when compared to co-cultures with CD83wt derived Mφ (left bar graphs, A–C). This observation is reflected by decreased T cell clusters shown in representative microscopic images (A–C, right side). (D) Flow cytometric analyses revealed a significantly decreased frequencies of Tregs (CD4⁺Foxp3⁺ cells) in co-cultures of CD83-deficient Mφ with allo-reactive splenocytes. Statistical analyses were performed by using an Unpaired t-test (A–C) or Two-way ANOVA (D) or the appropriate corresponding non-parametric test (n ≥ 4). Data are represented as mean ± SEM. Experiments were performed at least three times. *p< 0.05; **p<0.01. The absence of asterisks indicates that there is no statistical significance.

Figure 7

CD83-deficient Mφ accelerate the inflammatory phase of wound healing and promote upregulation of fibrosis associated transcripts (A) Experimental set-up for the full-thickness excisional wound healing model. Biopsy punches (6mm) were placed into the dorsal skin of CD83wt as well as CD83 cKO mice. (B) Wound closure was calculated relative to the initial d0 wound dimension. 8mm silicone rings (Thermo scientific) were mounted around the wound area, using vetbond (3M). Imaging was performed on day 0, 3, and 6 and wound diameters were determined by ImageJ. (C–E) qPCR analyses were performed using skin biopsies from day 0, 3 and 6 (n = 5 per group). (F) Representative H&E slides of day 6 wound biopsies from CD83wt as well as CD83 cKO mice. Statistical analyses were performed by using a Two-way ANOVA or the appropriate corresponding non-parametric test. Data are represented as mean ± SEM. *p< 0.05; **p<0.01; ***p<0.001; **** p< 0.0001. The absence of asterisks indicates that there is no statistical significance.

Figure 8

CD83 expressed by Mφ is an important immune checkpoint molecule that contributes to resolution of inflammation. CD83 is an early marker for IL-4 stimulated AAMs and its deletion in Mφ results in striking phenotypic and functional changes. CD83-deficient, IL-4 stimulated Mφ are characterized by a decreased STAT-6 phosphorylation status when compared to CD83wt Mφ. This goes along with reduced expression levels of AAM-associated marker molecules such as CD200R and Msr-1. Reduction in MSR-1 expression correlates with a reduced phagocytic activity of E.coli in CD83-deficient IL-4 stimulated Mφ. In contrast, CAM associated Dectin-1 expression is upregulated. Furthermore, CD83-deficient, IL-4 stimulated Mφ express increased levels of pro-inflammatory modulators, such as IL-6, TNF-α, CXCL1 and G-CSF. Functionally, Mφ generated from CD83 cKO mice show enhanced allogeneic T cell proliferative capacities and reduced frequencies of Tregs in Mφ - cell co-cultures. Finally, IFNγ-stimulated Mφ generated from cKO mice show an increased production of RANTES and MCP-1, indicating that CD83 also modifies the production of these pro-inflammatory chemokines.