Suppression of NF-kappaB-mediated beta-defensin gene expression in the mammalian airway by the Bordetella type III secretion system

Abstract

Expression of innate immune genes such as beta-defensins is induced in airway epithelium by bacterial components via activation of NF-kappaB. We show here that live Gram-negative bacteria can similarly stimulate this pathway, resulting in upregulation of the beta-defensin tracheal antimicrobial peptide (TAP) in primary cultures of bovine tracheal epithelial cells (TECs), by a Toll-like receptor 4 (TLR4)-mediated pathway. The Gram-negative airway pathogen Bordetella bronchiseptica possesses a type III secretion system previously suggested to inhibit the nuclear translocation of NF-kappaB in a cell line by immunohistochemistry. We therefore hypothesized that this pathogen might interfere in the innate immune response of the epithelium. Exposure of TECs to wild-type B. bronchiseptica suppressed the activation of NF-kappaB and the subsequent induction of TAP mRNA levels, whereas a type III secretion-defective strain did not. These results suggest a mechanism for bacterial evasion of the innate immune response in the airway, which could allow for the observed persistent colonization of this pathogen.

Fig. 1. Stimulation of TAP expression by LPS and bacteria via TLR4 in bovine TEC

A. TECs were pretreated with 10 μg ml⁻¹ RsDPLA for 1 h prior to stimulation with 100 ng ml⁻¹ of P. aeruginosa LPS or live B. bronchiseptica for 18 h. Dose–response studies revealed that 10 μg ml⁻¹ of RsDPLA was the lowest concentration necessary to observe the inhibition of TAP expression (not shown). Total RNA was subjected to semi-quantitative RT-PCR for 15 cycles, followed by Southern blot hybridization, as described in Experimental procedures. B. Suppression of TAP upregulation by the type III secretion system of B. bronchiseptica. TEC were challenged with live B. bronchiseptica for 6 h. Graphical representation of the data (n = 3) indicates the mean of the fold increase over the untreated control samples ± standard error of the mean. Significance (*) of the difference between RB50 and WD3-stimulated cells was determined by t-test analysis, when P < 0.05.

Fig. 2

Detection of TAP peptide by mass spectrometry. Primary cultures of bovine TEC were incubated in the absence (A) or the presence (B) of 100 ng ml⁻¹ Pseudomonas aeruginosa LPS for 18 h. Cells were lysed, and cytoplasmic extracts were partially purified by C-18 chromatography, and subjected to high resolution mass spectrometry as described in Experimental procedures.

Fig. 3. Role of NF-κB activation in TAP upregulation

A. TECs were pretreated with 20 μM of SN50 for 1 h prior to stimulation with 100 ng ml⁻¹ of P. aeruginosa LPS, 100 ng ml⁻¹ or of rhTNFα for 18 h. Semi-quantitative RT-PCR was used to examine TAP and GAPDH expression. The products were visualized by gel electrophoresis and ethidium bromide staining. The blot shown is representative of three experiments. B. Graphical representation of the data (n = 3) is presented as a percentage of the control treated with LPS or rhTNFα, which is set as 100%, ±standard error of the mean. Significance of the difference between cells treated with and without SN50 stimulated with LPS (*), or rhTNFα (**) was determined by t- test analysis, when P < 0.05. The optimal, sublethal concentration of SN50 was determined by dose–response studies (not shown).

Fig. 4

Inhibition of NF-κB binding by B. bronchiseptica. TECs were stimulated with viable RB50 or WD3 strains of B. bronchiseptica at an moi of 1000 : 1 for 1 h. Nuclear extracts were obtained, and 40 μg were incubated with [γ-³²P]-ATP labelled, double-stranded oligonucleotide (ds-TAP/NF32) containing the NF-κB consensus sequence from the 5′ flanking region of the TAP gene (upper gel). Competition experiments were performed to determine specificity of binding by using unlabelled specific (ds-TAP/NF32) and non-specific (ds-TAP/NFmut32) double-stranded oligonucleotides. The blot shown is representative of two separate experiments. The same extracts were used in an EMSA with labelled oligonucleotide containing the NF-IL6 consensus sequence from the 5′ flanking region of the TAP gene (lower gel).

Fig. 5. Activation of the ERK pathway by LPS and live bacteria

A. Inhibition of ERK phosphorylation of LPS-stimulated cells by U0126. TECs were stimulated with 100 ng ml⁻¹ P. aeruginosa LPS for 30, 60 or 90 min, and ERK phosphorylation was determined by Western blot analysis. Phosphorylated ERK is shown in the upper panel, and total ERK is shown in the lower panel. B. TAP upregulation is not dependent on the ERK pathway. TECs were pretreated with 10 μM of U0126 or DMSO as the vehicle control for 2 h prior to stimulation with 100 ng ml⁻¹ of P. aeruginosa LPS, or 100 ng ml⁻¹ of rhTNFα for 18 h. Semi-quantitative RT-PCR was used to examine TAP and GAPDH expression. The products were visualized by gel electrophoresis and ethidium bromide staining (upper panel). The blot shown is representative of three experiments. Graphical representation of the data (lower panel, n = 3) is presented as a percentage of the control treated with LPS, or rhTNFα alone, which is set as 100%, ±standard error of the mean. C. The effect of B. bronchiseptica on ERK 1/2 phosphorylation. Total protein from TEC stimulated with 100 ng ml⁻¹ P. aeruginosa LPS or the RB50 or WD3 strains of B. bronchiseptica (moi = 1000:1) for 30 or 60 min were subjected to Western blot analysis, as above.