Pseudomonas aeruginosa induces p38MAP kinase-dependent IL-6 and CXCL8 release from bronchial epithelial cells via a Syk kinase pathway

Abstract

Pseudomonas aeruginosa (Pa) infection is a major cause of airway inflammation in immunocompromised and cystic fibrosis (CF) patients. Mitogen-activated protein (MAP) and tyrosine kinases are integral to inflammatory responses and are therefore potential targets for novel anti-inflammatory therapies. We have determined the involvement of specific kinases in Pa-induced inflammation. The effects of kinase inhibitors against p38MAPK, MEK 1/2, JNK 1/2, Syk or c-Src, a combination of a p38MAPK with Syk inhibitor, or a novel narrow spectrum kinase inhibitor (NSKI), were evaluated against the release of the proinflammatory cytokine/chemokine, IL-6 and CXCL8 from BEAS-2B and CFBE41o- epithelial cells by Pa. Effects of a Syk inhibitor against phosphorylation of the MAPKs were also evaluated. IL-6 and CXCL8 release by Pa were significantly inhibited by p38MAPK and Syk inhibitors (p<0.05). Phosphorylation of HSP27, but not ERK or JNK, was significantly inhibited by Syk kinase inhibition. A combination of p38MAPK and Syk inhibitors showed synergy against IL-6 and CXCL8 induction and an NSKI completely inhibited IL-6 and CXCL8 at low concentrations. Pa-induced inflammation is dependent on p38MAPK primarily, and Syk partially, which is upstream of p38MAPK. The NSKI suggests that inhibiting specific combinations of kinases is a potent potential therapy for Pa-induced inflammation.

Fig 1. Stimulation of IL-6 and CXCL8 from BEAS-2B cells.

BEAS-2B cells were stimulated with (A) TNFα at 10 ng/ml or LPS at 10, 100 or 100 ng/ml for four hours. Cell free supernatant was collected, and IL-6 measured by sandwich ELISA (n = 3, mean ± SD). (B) BEAS-2B cells were infected with Pa at varying concentrations for one hour, followed by addition of gentamicin at 100 μg/ml and a further 24 hours incubation. Cell viability was measured by MTT assay (n = 3, mean ± SD). (C) Cell free supernatant from cells infected with Pa for 24 hours was analysed for IL-6 by sandwich ELISA (n = 5, median with IQR). BEAS-2B cells were infected with Pa at 2.5 x 10⁷ CFU/ml for one hour, followed by addition of gentamicin at 100 μg/ml. Supernatant was collected at 2, 4, 8, 12 or 24 hours later. (D) IL-6 and (E) CXCL8 were measured by sandwich ELISA (n = 4, median with IQR). (A-C) Stimulants were compared to NT using Friedman test with Dunn’s correction. (D-E) At each time point, Pa treatments were compared with NT using the Mann-Whitney U test (* = p<0.05** = p<0.01, *** = p<0.001).

Fig 2. Head to head comparison of kinase inhibitors against Pa-induced IL-6 and CXCL8 release, and cell viability.

BEAS-2B cells underwent a two-hour pre-treatment with kinase inhibitors at a single concentration known to completely inhibit the target kinase enzyme activity. The cells were stimulated with Pa at 2.5 x 10⁷ CFU/ml for one hour followed by addition of gentamicin at 100 μg/ml and a further four-hour incubation. The concentration of (A) IL-6 and (B) CXCL8 in the cell-free supernatant was measured by sandwich ELISA. (C) BEAS-2B cell viability was measured after compound treatment and Pa infection, using an MTT assay; cell viability was compared to cells treated with DMSO and Pa. Friedman test with Dunn’s multiple comparisons was used to compare each kinase inhibitor with vehicle treatment. n = 6, showing median with IQR (* = p<0.05, ** = p<0.01 and *** = p<0.001). Compound concentrations: SB203580–3.8 μg/ml; BIRB796–0.5 μg/ml; SP600125–2.2 μg/ml; PD98059–8.0 μg/ml; BAY 61–3606–0.5 μg/ml and Dasatinib– 0.1 μg/ml.

Fig 3. Concentration dependent inhibition of Pa-induced IL-6 and CXCL8 release by selected kinase inhibitors.

BEAS-2B cells underwent a two-hour pre-treatment with 10-fold dilutions of (A and D) SB203580, (B and E) BIRB796 and (C and F) BAY 61–3606. The cells were stimulated with Pa at 2.5 x 10⁷ CFU/ml for one hour followed by addition of gentamicin at 100 μg/ml and a further four-hour incubation. The concentration of (A, B and C) IL-6 and (D, E and F) CXCL8 in the cell free supernatant was measured by sandwich ELISA. Friedman test with Dunn’s multiple comparisons was used to compare each kinase inhibitor with vehicle treatment. n = 6, showing median with IQR (* = p<0.05, ** = p<0.01 and *** = p<0.001).

Fig 4. Inhibition of Pa-induced phosphorylation of HSP27 by p38MAPK inhibitors.

BEAS-2B cells in six well plates were pre-incubated with either DMSO (NT), SB203580 or BIRB796 at 1 μg/ml for two hours at 37°C, 5% CO₂. The cells were then stimulated with Pa at 2.5 x 10⁷ CFU/ml for two hours at 37°C, 5% CO₂, after which whole cell protein was collected and levels of phosphorylated- and total-HSP27 were measured by Western blot. (A) Representative image of the four experiments. (B) Band intensity was determined using Syngene Gene Tools software and the levels of phosphorylated-HSP27 were corrected to the levels of total-HSP27. A Friedman test with Dunn’s multiple comparison was used to compare Pa treatment alone with all other conditions. n = 4 showing median with IQR (** = p≤0.01).

Fig 5. Inhibition of MAP kinase activity by BAY 61–3606.

BEAS-2B cells in six well plates were pre-incubated with either DMSO, or BAY 61–3606 at 1 μg/ml for two hours at 37°C, 5% CO₂. The cells were then incubated with Pa at 2.5 x 10⁷ CFU/ml for two hours at 37°C, after which whole cell protein was collected. Western blots were carried out to measure the levels of total and phosphorylated (A) HSP27, (C) ERK and (E) JNK. (B, D and F) Band intensities were calculated using Syngene Gene Tools and phosphorylated protein levels were corrected to total protein. BAY 61–3606 treatment was compared to DMSO using a one-tailed Wilcoxon sign ranked test. (B) and (F), n = 5, median with IQR, (D) n = 3, mean ± SD. Images are representative of the repeats conducted (* = p<0.05).

Fig 6. Synergy of p38MAPK and Syk inhibition against Pa-induced IL-6 and CXCL8 release.

BEAS-2B cells underwent a two-hour pre-treatment with four-fold dilutions of either BIRB796 or BAY 61–3606, or a combination of the two. The cells were stimulated with Pa at 2.5 x 10⁷ CFU/ml for one hour followed by addition of gentamicin at 100 μg/ml and a further four-hour incubation. The concentration of (A) IL-6 and (C) CXCL8 in the cell free supernatant were measured by sandwich ELISA, n = 3, showing mean ± SD. Synergy was calculated using the Chou-Talalay method using CompuSyn software [64]. Isobolograms of the ED₉₀ concentrations for the compound concentrations against (B) IL-6 and (D) CXCL8 were drawn using Microsoft Excel. The ED₉₀ of each compound applied alone is plotted on the x and y axes and joined with a line. The ED₉₀ values of each compound when used in combination are plotted as an X. An X on the line would indicate an additive effect; X on the left of the line would indicate synergy (as in these experiments), and to the right would indicate compound antagonism.

Fig 7. Inhibition of IL-6 and CXCL8 by a novel narrow spectrum kinase inhibitor.

BEAS-2B cells underwent a two-hour pre-treatment with 10-fold dilutions of RV1088. The cells were stimulated with Pa at 2.5 x 10⁷ CFU/ml for one hour followed by addition of gentamicin at 100 μg/ml and a further four-hour incubation. The concentration of (A) IL-6 and (B) CXCL8 in the cell free supernatant was measured by sandwich ELISA. Friedman test with Dunn’s multiple comparisons was used to compare each kinase inhibitor with vehicle treatment. n = 6, showing median with IQR (* = p<0.05, ** = p<0.01 and *** = p<0.001).

Fig 8. Inhibition of Pa-induced IL-6 from bronchial epithelial cells expressing Phe508del- and WT-CFTR.

CFBE41o- cells expressing either Phe508del (A-C) or WTCFTR (D-F) underwent a two-hour pre-treatment with 10-fold dilutions of BIRB796 (A and D), BAY 61–3606 (B and E) and RV1088 (C and E). The cells were stimulated with Pa at 2.5 x 10⁷ CFU/ml for one hour followed by addition of gentamicin at 100 μg/ml and a further four-hour incubation. The concentration of IL-6 in the cell free supernatant was measured by sandwich ELISA. Friedman test with Dunn’s multiple comparisons was used to compare each kinase inhibitor with vehicle treatment. n = 6 for BIRB796 and RV1088, showing median with IQR, n = 3 for BAY 61–3606, showing mean and SD (* = p<0.05, ** = p<0.01 and *** = p<0.001, **** = p<0.0001).