Glucocorticoids limit acute lung inflammation in concert with inflammatory stimuli by induction of SphK1

Abstract

Acute lung injury (ALI) is a severe inflammatory disease for which no specific treatment exists. As glucocorticoids have potent immunosuppressive effects, their application in ALI is currently being tested in clinical trials. However, the benefits of this type of regimen remain unclear. Here we identify a mechanism of glucocorticoid action that challenges the long-standing dogma of cytokine repression by the glucocorticoid receptor. Contrarily, synergistic gene induction of sphingosine kinase 1 (SphK1) by glucocorticoids and pro-inflammatory stimuli via the glucocorticoid receptor in macrophages increases circulating sphingosine 1-phosphate levels, which proves essential for the inhibition of inflammation. Chemical or genetic inhibition of SphK1 abrogates the therapeutic effects of glucocorticoids. Inflammatory p38 MAPK- and mitogen- and stress-activated protein kinase 1 (MSK1)-dependent pathways cooperate with glucocorticoids to upregulate SphK1 expression. Our findings support a critical role for SphK1 induction in the suppression of lung inflammation by glucocorticoids, and therefore provide rationales for effective anti-inflammatory therapies.

Figures

Figure 1. GCs inhibit lung inflammation by the GR in myeloid cells requiring an intact GR dimerization interface.

(a) GRflox and GRLysMCre mice were injected with vehicle (Co), LPS and OA (hereafter, the LPS and OA co-treatment will be designated as ‘LPS') or LPS and Dex (LPS+Dex). After 18 h, EB was injected i.v., and the mice were killed 30 min later. EB accumulation in the lung was determined after perfusion with PBS+5 mM EDTA. (b) Representative haematoxylin/eosin (H&E) staining of lung sections (scale bars, 0.1 mm), and (c) histological scores of GRflox and GRLysMCre mice treated as described in a. (d) GRflox and GRLysMCre mice were treated as described in a. Cell numbers were determined in the BAL 18 h after the indicated treatments. (e) GRdim and wild-type (WT) littermate mice were injected with vehicle (Co), LPS or LPS+Dex. After 24 h, EB was injected i.v. and 30 min later the mice were killed. EB accumulation in the lung was quantified as described in a. (f) Representative H&E staining of lung sections (scale bars, 0.1 mm) and (g) histological scores of lungs of littermate WT and GRdim mice from the experiment described in e. (h) GRdim and WT mice were treated as described in e. Cell numbers were determined in the BAL 24 h after the indicated treatments. Results are presented as mean±s.e.m. Number of biological replicates: (a) (6–10), (c) (18–20), (d) (7–13), (e) (4–8), (g) (12–16) and (h) (7–11). In a, data are from three independent biological experiments; in c, data are from four independent biological experiments; in d, data are from two independent biological experiments; in e, data are from two independent biological experiments; in g, data are from three independent biological experiments out of four; and in h, data are from two independent biological experiments. Statistical analysis was performed by one-way analysis of variance. *P<0.05, **P<0.01, ***P<0.001; ND, not detectable; NS, not significant.

Figure 2. S1P signalling increased by Dex or S1PR1 agonist reduces ALI in GRLysMCre mice.

(a) GRdim and wild-type (WT) littermate control mice were treated as described in Fig. 1e. The S1P level in plasma (pmol ml−1) was determined by LC/MS/MS after 24 h. (b) GRflox and GRLysMCre mice were treated as described in Fig. 1a. The S1P level in plasma (pmol ml−1) was determined by LC/MS/MS after 18 h. (c) GRflox and GRLysMCre mice were treated as described in Fig. 1a, and an additional group of LPS-treated mice received SEW2781 (LPS+SEW). EB was injected i.v. 30 min before killing. Representative lungs are shown to demonstrate EB accumulation. (d) Quantification of EB accumulation of the experiment described in c determined as described in Fig. 1a. Results are presented as mean±s.e.m. Number of biological replicates: (a) (9), (b) (15–20) and (d) (9 or 10). Data in a are from two independent biological experiments, and data from b and d are from three independent biological experiments. Statistical analysis was performed by one-way analysis of variance. *P<0.05, **P<0.01 and ***P<0.001; NS, not significant.

Figure 3. SphK1 expression in myeloid cells is essential for the anti-inflammatory effects of GCs in ALI.

(a) SphK1flox and SphK1LysMCre mice were treated as described in Fig. 1a. The S1P level in plasma (pmol ml−1) was determined 24 h later by LC/MS/MS. (b) SphK1flox and SphK1LysMCre mice were treated as described in Fig. 1a, and lung specimens with accumulated EB are shown. (c) EB accumulation of isolated lungs was quantified as described in Fig. 1a. (d) Representative haematoxylin/eosin staining of lung sections from SphK1flox and SphK1LysMCre mice treated as described in Fig. 1a (scale bars, 0.1 mm). (e) Cell numbers were determined in the BAL 24 h after the indicated treatments. (f) IL-6 plasma levels (ng ml−1) in SphK1flox and SphK1LysMCre mice were determined by ELISA 5 h after LPS or LPS+Dex treatment. Results are presented as mean±s.e.m. Number of biological replicates: (a) (8–15), (c) (5–12), (e) (5–12) and (f) (5–12). In a, c and e, the data are from two independent biological experiments; and in f, the data are from one biological experiment. Statistical analysis was performed by one-way analysis of variance. *P<0.05, **P<0.01 and ***P<0.001; ND, not detectable; NS, not significant.

Figure 4. Dex and inflammatory stimuli synergistically elevate SphK1 mRNA and activity dependent on the GR in macrophages.

(a) BMDMs derived from wild-type (WT) and GRdim mice were treated with vehicle (Co), LPS, Dex or LPS+Dex for 4 h. SphK1 mRNA expression was analysed by quantitative RT–PCR. (b) BMDMs derived from WT and GRdim mice were treated as described in a for 4 h. SphK1 activity was determined by the conversion of C17- and C18-sphingosine to S1P by LC/MS/MS. (c–e) BMDMs derived from WT mice were treated for 2 h with vehicle (Co), (c) Pam3CSK4 (TLR1/2 agonist), (d) polyIC (TLR3 agonist), (e) TNFα, Dex, or co-treatment with (c) Pam3CSK4 and Dex (Pam+Dex), (d) polyIC and Dex (polyIC+Dex), (e) TNFα and Dex (TNF+Dex). SphK1 mRNA expression was analysed by quantitative RT–PCR. Results are presented as mean±s.e.m. Number of biological replicates: (a) (5 or 6), (b) (4–8), (c) (5–7), (d) (3–5) and (e) (6 or 7). Statistical analysis was performed by one-way analysis of variance. *P<0.05, **P<0.01 and ***P<0.001.

Figure 5. Dex and LPS treatment synergistically elevate SphK1 mRNA dependent on the p38 MAPK–MSK1 pathway.

(a) ChIP of wild-type (WT) and GRdim BMDMs was conducted 4 h after the indicated treatment (Dex, LPS+Dex). Shown are the quantitative RT–PCR results with SphK1 primer. (b) Quantitative analysis of SphK1 mRNA expression in BMDMs treated for 2 h with vehicle (Co), LPS, Dex or LPS+Dex in the presence or absence of 10 μM SB203580 (p38 MAPK inhibitor). (c) BMDMs from WT and p38αMx1Cre mice were treated for 2 h with vehicle (Co), LPS, Dex or LPS+Dex, then SphK1 mRNA expression was analysed. (d) Quantitative analysis of SphK1 mRNA expression in BMDMs treated for 2 h with vehicle (Co), LPS, Dex or LPS+Dex in the presence or absence of 10 μM SB747651A (MSK1 inhibitor). (e) The direct (physical) and indirect (functional) association among the downregulated genes after MSK1 inhibition is achieved using the STRING 10.1 database, and finally the network is visualized using Cytoscape 3.1.0. Results are presented as mean±s.e.m. Number of technical replicates: (a) (3); number of biological replicates: (b) (3), (c) (3 or 4) and (d) (6–8). Statistical analysis was performed by one-way analysis of variance. **P<0.01 and ***P<0.001; NS, not significant.

Figure 6. GCs and LPS regulate SphK1 and S1P dependent on GR dimerization and the p38–MSK1 pathway in macrophages to limit ALI.

During inflammation, SphK1 is moderately expressed. This results in a low activation of S1PR1, an increase in lung endothelial barrier permeability and exacerbation of ALI (left). During GC treatment of ALI, SphK1 expression in macrophages is synergistically induced by the GR- and LPS-triggered signalling via activation of TLR4 p38 MAPK–MSK1. This leads to an elevation of S1P, which acts on S1PR1 to decrease lung endothelial barrier permeability. This enhanced S1P level is a prerequisite to limit ALI (right).