Roflumilast N-oxide prevents cytokine secretion induced by cigarette smoke combined with LPS through JAK/STAT and ERK1/2 inhibition in airway epithelial cells

Abstract

Cigarette smoke is a major cause of chronic obstructive pulmonary disease (COPD). Airway epithelial cells and macrophages are the first defense cells against cigarette smoke and these cells are an important source of pro-inflammatory cytokines. These cytokines play a role in progressive airflow limitation and chronic airways inflammation. Furthermore, the chronic colonization of airways by Gram-negative bacteria, contributes to the persistent airways inflammation and progression of COPD. The current study addressed the effects of cigarette smoke along with lipolysaccharide (LPS) in airway epithelial cells as a representative in vitro model of COPD exacerbations. Furthermore, we evaluated the effects of PDE4 inhibitor, the roflumilast N-oxide (RNO), in this experimental model. A549 cells were stimulated with cigarette smoke extract (CSE) alone (0.4% to 10%) or in combination with a low concentration of LPS (0.1 µg/ml) for 2 h or 24 h for measurement of chemokine protein and mRNAs and 5-120 min for protein phosphorylation. Cells were also pre-incubated with MAP kinases inhibitors and Prostaglandin E2 alone or combined with RNO, before the addition of CSE+LPS. Production of cytokines was determined by ELISA and protein phosphorylation by western blotting and phospho-kinase array. CSE did not induce production of IL-8/CXCL8 and Gro-α/CXCL1 from A549 cells, but increase production of CCL2/MCP-1. However the combination of LPS 0.1 µg/ml with CSE 2% or 4% induced an important production of these chemokines, that appears to be dependent of ERK1/2 and JAK/STAT pathways but did not require JNK and p38 pathways. Moreover, RNO associated with PGE2 reduced CSE+LPS-induced cytokine release, which can happen by occur through of ERK1/2 and JAK/STAT pathways. We report here an in vitro model that can reflect what happen in airway epithelial cells in COPD exacerbation. We also showed a new pathway where CSE+LPS can induce cytokine release from A549 cells, which is reduced by RNO.

Figure 1. Effects of CSE (0.1%–10%) and LPS (10 ng/ml–10 µg/ml) on chemokine release and cell viability.

Serum-starved A549 cells were incubated with medium alone (control) or with different concentration of CSE (0.1%–10%) and LPS (10 ng/ml–10 µg/ml) for 24 h. The culture supernatants were collected and the concentrations of IL-8/CXCL8, Gro-α/CXCL1 and MCP-1/CCL2 were measured by ELISA (A, B, C and D) and viability was determined by MTT test (E). The data represent the mean ± SEM of 3 experiments. * p<0.05 compared to control.

Figure 2. CSE associated with LPS induced chemokine release from A549 cells and mRNA expression.

Serum-starved A549 cells were incubated with medium alone (control), CSE 2% or 4% (blank bars), LPS 0.1 µg/ml alone or in combination with 2 or 4% of CSE (hatched bars) for 24 h. The culture supernatants were collected and the concentrations of chemokines were measured by ELISA (A, B and C). The data represent the mean ± SEM of 7 experiments. A549 cells were stimulated for 2 h. mRNA expression was then determined by real-time quantitative RT-PCR (D, E and F). The results are normalized to the gene expression of GAPDH. The data represent the mean ± SEM of 4 experiments. * p<0.05 compared to control; α p<0.05 compared to LPS 0.1 µg/ml.

Figure 3. Effects of CSE and LPS alone or in combination in ERK1/2, p38 and JNK activation.

Serum-starved A549 cells were incubated with serum free medium alone (control, Ctrl), LPS 0.1 µg/ml, CSE 4% or CSE 4% associated with LPS 0.1 µg/ml for 5, 15, 30, 60 or 120 min. Total cell lysates were immunoblotted with antibodies specific for phospho-p38 kinase and total p38 kinase (A), phospho-JNK and total JNK (B) or phospho ERK1/2 and total ERK1/2 (C). A549 cells stimulated with IL-1β (7 ng/ml) were used as positive controls. The data represent the mean ± SEM of 3 experiments. * p<0.05 compared to control; α p<0.05 compared to LPS 0.1 µg/ml.

Figure 4. Effect of ERK 1/2, p38 and JNK inhibitor on chemokine release induce by CSE+LPS.

Serum-starved A549 cells were preincubated with serum-free medium alone (control: blank bars), vehicle (0.1% DMSO), the p38 MAPKinhibitor SB203580 (10 µM and 20 µM), the MEK1/2 inhibitor U0126 (3 µM and 5 µM) or the JNK inhibitor SP600125 (2 µM and 4 µM) for 2 h and then stimulated with CSE (4%) together with LPS (0.1 µg/ml) for 24 h (black bars). At the end of the incubation period culture supernatants were collected for IL-8/CXCL8, MCP-1/CCL2 and Gro-α/CXCL1 quantification by ELISA. Data are expressed as means ± SEM of 3 independent experiments. * p<0.05 compared to control; β p<0.05 compared to vehicle.

Figure 5. Effects of CSE and LPS alone or in combination on selected phosphoproteins in A549 cells.

Cells were incubated for 40-free medium alone (control), LPS 0.1 µg/ml (A), CSE 4% (B), or CSE 4% and LPS 0.1 µg/ml (C). Then cell extracts were probed on human phosphoprotein arrays. Results representing the signal intensity (AU) were expressed as % of the unstimulated control. Data are expressed as means ± SEM of 3 independent experiments. * p<0.05 compared to control.

Figure 6. Effects of CSE and LPS alone or in combination on STAT3 phosphorylation in A549 cells.

Serum-starved A549 cells were stimulated with serum free medium alone (control), LPS 0.1 µg/ml, CSE 4% or CSE 4% and LPS 0.1 µg/ml for 5, 15, 30 or 60 min. Total cell lysates were immunoblotted with antibodies specific for phospho STAT3 and total STAT3. Data are expressed as means ± SEM of 3 independent experiments.* p<0.05 compared to control; α p<0.05 compared to LPS 0.1 µg/ml.

Figure 7. Roflumilast N-oxide associated with PGE₂ partly inhibits chemokines release from A549 cells stimulated with CSE+LPS.

Cells were preincubated with vehicle (blank bars) or 10 nM PGE₂ with roflumilast N-oxide at 1 nM (A, B and C) or 1 µM (D, E and F) (hatched bars) for 2 h and then stimulated or not with CSE at 2% or 4% in combination with LPS at 0.1 µg/ml. After 24 hours cell culture supernatants were collected and chemokines were quantified by ELISA. Results are expressed as means ± SEM of 3 independent experiments. * p<0.05 compared to control (0.01% DMSO); # p<0.05 compared to vehicle.

Figure 8. Effects of roflumilast N-oxide on selected phosphoproteins in A549 cells.

Cells were preincubated for 2(blank bars) or PGE₂ (10 nM) and roflumilast N-oxide (1 nM) (hatched bars) then co-stimulated with CSE (4%) and LPS (0.1 µg/ml). Cell extracts were probed on human phosphoprotein arrays. Results representing the signal intensity (AU) were expressed as % of the unstimulated control. Data are expressed as means ± SEM of 3 independent experiments. * p<0.05 compared to control; # p<0.05 compared to respective vehicle.