Neutrophils induce macrophage anti-inflammatory reprogramming by suppressing NF-κB activation

Abstract

Apoptotic cells modulate the function of macrophages to control and resolve inflammation. Here, we show that neutrophils induce a rapid and sustained suppression of NF-κB signalling in the macrophage through a unique regulatory relationship which is independent of apoptosis. The reduction of macrophage NF-κB activation occurs through a blockade in transforming growth factor β-activated kinase 1 (TAK1) and IKKβ activation. As a consequence, NF-κB (p65) phosphorylation is reduced, its translocation to the nucleus is inhibited and NF-κB-mediated inflammatory cytokine transcription is suppressed. Gene Set Enrichment Analysis reveals that this suppression of NF-κB activation is not restricted to post-translational modifications of the canonical NF-κB pathway, but is also imprinted at the transcriptional level. Thus neutrophils exert a sustained anti-inflammatory phenotypic reprogramming of the macrophage, which is reflected by the sustained reduction in the release of pro- but not anti- inflammatory cytokines from the macrophage. Together, our findings identify a novel apoptosis-independent mechanism by which neutrophils regulate the mediator profile and reprogramming of monocytes/macrophages, representing an important nodal point for inflammatory control.

Fig. 1. Neutrophil-mediated changes to the macrophage inflammatory cytokine profile.

a LPS-induced (1 ng/ml; 6 h) TNF release from MDM co-cultured with viable and apoptotic BL-2 cells, Jurkat cells (both n = 6) and neutrophils (n = 10). b LPS-induced (1 ng/ml; 6 h) TNF release from co-cultures of MDM with apoptotic (closed bars) and viable (open bars) neutrophils at the ratio of neutrophil to MDM as indicated (n = 4). c LPS-induced (1 ng/ml; 6 h) TNF release from monocytes (n = 12), INFγ (20 ng/ml) primed MDM (n = 6) and alveolar macrophages (n = 11) co-cultured with apoptotic or viable neutrophils. d LPS-induced (1 ng/ml; 6 h) TNF release from MDM and monocytes co-cultured with neutrophils isolated from BAL (n = 4): cytocentrifugation preparation of BAL neutrophils stained with Diff-Quik shows the multi-lobed morphology of viable neutrophils, cells stained pink are eosinophils. Flow cytometry of BAL neutrophils stained with annexin V and propidium iodide where the upper left quadrant represents cell debris, the upper right quadrant represents necrotic cells, the lower right quadrant represents apoptotic cells and the lower left represents viable cells. e LPS-induced (1 ng/ml; 6 h) cytokine profiles from MDM co-cultured with apoptotic (i) and viable (ii) neutrophils (n = 6-12). f Basal and LPS-induced (1 ng/ml; 6 h) TNF release from MDM co-cultured with apoptotic neutrophils purified by cell sorting annexin V⁺/propidium iodide^- cells (n = 7). Neutrophils are negative for bound annexin V after sorting due to reduction of the extracellular Ca²⁺ concentration and thus annexin V does not affect the subsequent incubation with MDM. g Percentage of viable neutrophils that become apoptotic, as measured by either (i) annexin V⁺ or (ii) activation of caspase 3/7, over the period of a 6 h co-culture with MDM and LPS (1 ng/ml; n = 6). All n-numbers represent data derived from separate healthy donors with data plotted as mean ± s.e.m. Neutrophil to MDM or monocyte ratio in co-cultures were 3:1, unless otherwise stated. Image scale bars = 10 μm. ***P > 0.001, **P > 0.01, *P > 0.05 using paired t-test (a–d) or ANOVA with Tukey post-hoc test (f). Ct Control, MDM monocyte derived macrophage, A-N apoptotic neutrophil, V-N viable neutrophil, Mono monocyte

Fig. 2. Neutrophil-mediated suppression of macrophage inflammatory cytokine release in response to different inflammatory signalling.

The release of TNF (a), CXCL-8 (b, c) from MDM co-cultured with neutrophils and stimulated with LPS (1 ng/ml), TNF (20 ng/ml) and IL-1β (10 ng/ml), respectively, for 6 h. Data derived from 8 separate healthy donors and plotted as mean ± s.e.m. ***P > 0.001, **P > 0.01, *P > 0.05 using ANOVA with Tukey post-hoc test. Ct control, A-N apoptotic neutrophil, V-N viable neutrophil

Fig. 3. The impact of neutrophils on NF-κB activity in MDM.

a LPS-induced (1 ng/ml; 20 min) p65 phosphorylation measured by (i-iv) flow cytometry (serine 529; n = 4) and (v) immunoblotting (serine 536, 529 and 276; n = 3) in MDM co-cultured with apoptotic or viable neutrophils (total time point: 30 min). b LPS-induced (1 ng/ml; 20 min) p65 nuclear translocation assessed by (i) quantitative high-content imaging and confocal microscopy (scale bars 100 μm and 7.5 μm, respectively), (ii) quantification of p65 nuclear translocation from high-content imaging (n = 8) and (iii) immunoblotting in MDM co-cultured with apoptotic and viable neutrophils (n = 3). Full images of individual stains and merged staining panels are available in supplemental data (supplemental data Fig. 1). c LPS-induced (1 ng/ml; 1 h) p65 binding to the TNF promotor in MDM co-cultured with apoptotic or viable neutrophils (n = 3). d LPS-induced (1 ng/ml; 20 min) p65 phosphorylation in alveolar macrophages (n = 3). e Phagocytosis of apoptotic neutrophils labelled with CellTracker Green and pHrodo Red by MDM at 40 min of co-culture in the presence of LPS (1 ng/ml; n = 14). f LPS-induced (1 ng/ml; 6 h) TNF release from MDM co-incubated with either apoptotic or viable neutrophils (cells), the supernatants from apoptotic or viable neutrophils cultured alone for 6 h (Sup) or the supernatants from apoptotic or viable neutrophils co-cultured with MDM for 6 h before (CoC-Sup) or after (CoC-Sup (UC) ultracentrifugation at 100,000 g for 1 h (n = 4). All n-numbers represent data derived from separate healthy donors with data plotted as mean ± s.e.m. ***P > 0.001, *P > 0.05 using paired t-test. A-N apoptotic neutrophil, V-N viable neutrophil, ser serine, Ct control, Nuc nuclear, Cyt. Ext. Ct cytoplasmic extract control, AM alveolar macrophage

Fig. 4. Sustained influence of apoptotic neutrophils on the macrophage cytokine profile.

a Work flow for MDM and AM culture and stimulation post co-culture (pMDM and pAM, respectively). b LPS-induced (1 ng/ml, 6 h) TNF release from pMDM at 0, 24, 48 and 72 h (n = 6). c, d Release of CXCL-8 (c) and TNF (d) from pMDM stimulated with TNF (20 ng/ml) and IL-1β (10 ng/ml), respectively, for 6 h at 0 or 24 h (n = 4). e high-content imaging (i; scale bar 100 μm) and quantification (ii, n = 6) of p65 translocation in pMDM stimulated with LPS (1 ng/ml, 1 h) at 0 h and 24 h, (iii) LPS-induced (1 ng/ml, 30 min) p65 phosphorylation (serine 536) in pMDM at 0 h or 24 h (n = 3). Full images of individual stains and merged staining panels from high content imaging are available in supplemental data (supplemental data Fig. 2). f TNF release (i) and p65 phosphorylation (serine 536) (ii) from pAM at 0 and 24 h (n = 11 and 3, respectively). g Table (i) of significantly different NF-κB-related pathways (with a frequency distribution ratio of less than 0.2) identified by gene set enrichment analysis (GSEA) of an affymetrix 2.1 ST array and (ii) Hierarchical clustering of the combined genes (326) from these significant NF-κB pathways from LPS-stimulated MDM (1 ng/ml; 9 h) cultured with or without apoptotic neutrophils (n = 4). All n-numbers represent data derived from separate healthy donors with data plotted as mean ± s.e.m. ***P > 0.001, **P > 0.01 using paired t-test. A-N apoptotic neutrophil, pMDM MDM post co-culture with apoptotic neutrophils, pAM AM post co-culture with apoptotic neutrophils, Ct control, h hour

Fig. 5. Neutrophil-mediated regulation of TAK1 and IKKβ activation in macrophages.

a Inhibition of LPS-induced (1 ng/ml) TNF release (n = 6) (i), IKKβ and p65 (serine 536) phosphorylation (n = 3) (ii) and p65 translocation (n = 7; scale bar 100 μm) (iii) in MDM using the IKKβ selective inhibitor TPCA-1. Full images of individual stains and merged staining panels from high content imaging are available in supplemental data. b LPS (1 ng/ml; 20 min) induced IKKβ phosphorylation in MDM co-cultured with apoptotic and live neutrophils (i) and in pMDM (ii) and pAM (iii) at 0 h or 24 h (n = 3). c LPS (1 ng/ml; 20 min) induced TAK1 phosphorylation in MDM co-cultured with apoptotic and live neutrophils (i) and in pMDM (ii) and pAM (iii) at 0 h or 24 h (n = 3). d Summary flow diagram of proposed inhibition of NF-κB activity in macrophages by neutrophils. All n-numbers represent data derived from separate healthy donors with data plotted as mean ± s.e.m. ***P > 0.001 using paired t-test. Conc. Concentration, A-N apoptotic neutrophil, V-N viable neutrophil, Ct Control, h hour

Fig. 6. Canonical activation of NF-κB through the TAK1-IKKβ-NF-κB axis.

Adaptor proteins are recruited to the cell surface receptors upon its activation. These include the E3 ubiquitin ligase TRAF6 (in the case of TRL4 and IL-1R) and TRAF2/RIP1 (in the case of TNFR1) resulting the formation of non-degradative ubiquitin chains on which TAK1/TAB and IKK complexes are recruited. TAK1 phosphorylates IKKβ which then induces IκB degradation and phosphorylates p65 at serine 536 resulting in NF-κB transactivation and inflammatory gene transcription

Fig. 7. Neutrophil-mediated modulation of monocyte and macrophage mediator profile influences inflammatory control and resolution.

Recruitment of neutrophils to the site of injury during an innate inflammatory response has the potential to influence the inflammatory function of recruited monocyte and macrophage populations. These interactions will control the intensity of the inflammatory response. Neutrophils undergoing apoptosis will further regulate responses of monocyte and macrophage populations and drive sustained anti-inflammatory reprogramming, which is pivotal for inflammation resolution and tissue repair