Cystic fibrosis (CF) is a genetic disease characterized by severe neutrophil-dominated airway inflammation. An important cause of inflammation in CF is Pseudomonas aeruginosa infection. We have evaluated the importance of a number of P. aeruginosa components, namely lipopeptides, LPS, and unmethylated CpG DNA, as proinflammatory stimuli in CF by characterizing the expression and functional activity of their cognate receptors, TLR2/6 or TLR2/1, TLR4, and TLR9, respectively, in a human tracheal epithelial line, CFTE29o−, which is homozygous for the ΔF508 CF transmembrane conductance regulator mutation. We also characterized TLR expression and function in a non-CF airway epithelial cell line 16HBE14o−. Using RT-PCR, we demonstrated TLR mRNA expression. TLR cell surface expression was assessed by fluorescence microscopy. Lipopeptides, LPS, and unmethylated CpG DNA induced IL-8 and IL-6 protein production in a time- and dose-dependent manner. The CF and non-CF cell lines were largely similar in their TLR expression and relative TLR responses. ICAM-1 expression was also up-regulated in CFTE29o− cells following stimulation with each agonist. CF bronchoalveolar lavage fluid, which contains LPS, bacterial DNA, and neutrophil elastase (a neutrophil-derived protease that can activate TLR4), up-regulated an NF-κB-linked reporter gene and increased IL-8 protein production in CFTE29o− cells. This effect was abrogated by expression of dominant-negative versions of MyD88 or Mal, key signal transducers for TLRs, thereby implicating them as potential anti-inflammatory agents for CF.
Cystic fibrosis (CF)3 is an autosomal recessive inherited disorder characterized by mutations in the gene encoding the CF transmembrane conductance regulator (CFTR) protein. It is one of the most common lethal hereditary disorders among Caucasians of European descent. The most frequent mutation in CF is a deletion of phenylalanine at position 508 of the CFTR protein, the ΔF508 mutation. This abrogates CFTR channel function and causes the protein to fold aberrantly and accumulate within the endoplasmic reticulum (1, 2, 3, 4, 5, 6).
CF is characterized in the lungs by neutrophil-dominated inflammation. IL-8, a member of the CXC chemokine family, is a potent activator and chemoattractant of neutrophils and has been shown to be expressed by epithelial cells in response to a variety of stimuli, including the neutrophil-derived protease, neutrophil elastase (NE) (7). In CF, IL-8-induced recruitment of additional neutrophils to the airways results in further release of NE and additional induction of IL-8 gene expression by bronchial epithelial cells, thereby perpetuating a chronic cycle of respiratory inflammation. A major goal in CF research is the development of improved therapies to treat pulmonary inflammation associated with this condition.
A significant factor by which pulmonary inflammation is mediated in CF is Pseudomonas aeruginosa infection. Pseudomonas Ags can exacerbate pulmonary inflammation in CF by exaggerating proinflammatory gene expression via TLR activation. TLRs belong to a family of proteins that can recognize and discriminate a diverse array of microbial Ags (8, 9, 10). Following their activation by specific factors, TLRs transduce intracellular signals to regulate proinflammatory gene expression. The first TLR to be identified was initially termed human or hToll, on the basis of its homology to the Drosophila Toll protein, which has a role in embryonic development and immunity in Drosophila melanogaster (11). This was later renamed TLR4, and is now recognized to be the principal receptor responsible for transducing the LPS signal intracellularly. The TLR family currently includes at least 10 human TLRs for which multiple agonists exist. Briefly, TLR2 is activated by microbial diacylated and triacylated lipopeptides by heterodimerizing with TLRs 6 and 1, respectively, TLR3 by dsRNA, TLR4 by LPS (and NE among others), TLR5 by flagellin, and TLR9 by unmethylated CpG (uCpG) dinucelotides (12, 13, 14, 15, 16, 17, 18, 19, 20, 21). TLR7 and 8 are activated by ssRNA (22, 23), and specific activators of TLR10 have yet to be identified. An important and interesting feature of TLR signal transduction is that a highly conserved pathway is activated by the different TLRs (24, 25). These receptors signal via a number of kinases and adaptor proteins, MyD88/Mal or TIR domain-containing adaptor inducing IFN-β (TRIF)/TIR domain-containing adaptor molecule-1, IL-1R-associated kinases, TNFR-associated factor 6, TGF-β-activated kinase 1, and IκB kinases, to activate NF-κB and induce expression of NF-κB-regulated genes. These intracellular signaling molecules represent key inhibitory targets for therapeutic drug design.
Patterns of TLR expression have been studied in many different tissues and cell types; however, until recently, TLR expression and function in airway epithelial cells (26, 27), particularly CF airway epithelial cells, remained largely unexplored (28). Furthermore, the contribution of so-called nonimmune epithelial cells to the inflammatory response in the CF lung deserves more detailed investigation. Therefore, given that there are a number of potential TLR agonists in the CF lung (e.g., NE, Pseudomonas lipopeptides, LPS, and DNA), we decided to evaluate TLR expression in CF airway epithelial cells, to use these cells as a model to determine the contribution made by CF epithelium to inflammation in the CF lung, and evaluate the potential of TLR-specific inhibitors to ameliorate the inflammatory response of CF airway epithelial cells.
Materials and Methods
Cell culture and treatments
CFTE29o− and CFBE41o− cells are ΔF508 homozygous tracheal and bronchial epithelial cell lines, respectively. The 16HBE14o− is a non-CF human bronchial epithelial cell line. These were obtained as a gift from D. Gruenert (University of Vermont, Burlington, VT) (29, 30, 31). The cells were cultured on coated plates (fibronectin) (1 mg/ml; Sigma-Aldrich), collagen (Vitrogen 100, 2.9 mg/ml; Cohesion Technologies), and BSA (1 mg/ml; Sigma-Aldrich) in Eagle’s MEM (Invitrogen Life Technologies) supplemented with 10% FCS, 1% l-glutamine, and 1% penicillin/streptomycin (Invitrogen Life Technologies). Twenty-four hours before agonist treatment, cells were washed with serum-free Eagle’s MEM and placed under serum-free conditions or in serum containing 1% FCS for LPS or macrophage-activating lipopeptide-2 (MALP-2) stimulations. Human myelomonocytic U937 cells (European Collection of Cell Culture) were cultured in RPMI 1640 containing 10% FCS, 1% l-glutamine, and 1% penicillin/streptomycin (Invitrogen Life Technologies), and were maintained at 37°C in a humidified atmosphere of 5% CO2. Agonist treatments were performed in serum-free or 1% FCS conditions, as appropriate.
Diacylated lipopeptide from Mycoplasma fermentans (MALP-2 (S-(2,3-bis acyloxypropyl)-cysteine-GNNDESNISFKEK)) (Alexis Biochemicals), triacylated lipopeptide (palmitoyl-Cys((RS)-2,3-di((palmitoyloxy)-propyl)-Ala-Gly-OH) (Pam3; Bachem), P. aeruginosa LPS (Sigma-Aldrich), uCpG (5′-TCGTCGTTTTGTCGTTTTGTCGTT-3′) or control GpG (5′-CTGGTCTTTCTGGTTTTTTTCTGG-3′) phosphodiester or phosphorothioate-modified oligonucleotides (MWG Biotec), PMA (Sigma-Aldrich), human rTNF-α (R&D Systems), polymixin B, and actinomycin D (Sigma-Aldrich) were used, as indicated. P. aeruginosa strain 01 (PAO1) was a gift from R. Hancock (University of British Columbia, Vancouver, British Columbia, Canada). PAO1-conditioned medium (PCM) was prepared by filter sterilizing culture supernatants from 72-h PAO1 trypticase soy broth cultures.
TLR mRNA analysis
Total RNA was isolated from 1 × 106 cells using TRI reagent (Sigma-Aldrich). Contaminating DNA was removed using DNase 1 treatment. For RT-PCR, 1 μg of total RNA was reverse transcribed into cDNA with an oligo(dT)15 primer using first strand cDNA synthesis kit (Roche). The integrity of RNA extraction and cDNA synthesis was verified by PCR by measuring the amounts of GAPDH cDNA in each sample using GAPDH-specific primers to generate a 211-bp product. PCR mixtures contained 2.5 mM MgCl2, 1.25 U of Taq polymerase, 0.2 mM dNTPs (Promega), and 50 pmol each of TLR2, TLR4, TLR6, and TLR9 gene-specific primers (Table I). Thermocycling conditions were 95°C for 5 min, followed by 35 cycles of 95°C for 30 s, 55°C for 30 s, and 72°C for 30 s. A final extension step of 72°C for 10 min was performed. Control PCR using an RNA template failed to generate any products. Multiplex PCR were also performed to detect GAPDH, TLRs 1–5, and CD14 using the Cytoxpress Multiplex PCR Human Signaling Gene Set 1 kit (BioSource International; QHM0172), according to the manufacturer’s instructions. Products were resolved on 1.5% Tris-borate-EDTA (TBE)-agarose gels containing 0.5 μg/ml ethidium bromide (Sigma-Aldrich), and images were captured using the GeneGenius Gel Documentation and Analysis System (Syngene).
|Gene (Accession no.) .||Primers (5′→3′) .||Bases .||Product Size (bp) .|
|Gene (Accession no.) .||Primers (5′→3′) .||Bases .||Product Size (bp) .|
Cells (2 × 104) were grown in eight-well chamber slides (Nunc), washed with PBS-0.5% BSA, Fc blocked for 15 min at room temperature with 2% BSA (Sigma-Aldrich), then labeled with anti-TLR primary Abs (goat anti-human TLR1, TLR3, TLR5, and TLR6 (Santa Cruz Biotechnology); mouse IgG2A anti-human TLR2.1 (eBioscience); mouse IgG2A anti-human TLR4 (Serotec); mouse IgG1 anti-human TLR6 (Alexis Biochemicals); or mouse IgG1 anti-human TLR9 (Imgenex)) for 30 min at 4°C. Following three washes, cells were incubated with 10 μg/ml FITC-labeled secondary Ab (anti-goat IgG or anti-mouse F(ab′)2 (DakoCytomation)). Cells were counterstained with propidium iodide (PI) (Molecular Probes), and laser-scanning cytometry (Compucyte) was used to quantify cell surface TLR expression, as previously described (17). FITC and PI cellular fluorescence of at least 3 × 103 cells were measured. TLR expression was quantified using CompuCyte software on the basis of integrated green fluorescence. CD14 expression was quantified using a PE-conjugated mouse anti-human CD14 IgG2A Ab (DakoCytomation), and PE cellular fluorescence was measured in all cells. ICAM-1 expression was quantified using an FITC-conjugated anti-human ICAM-1 Ab (R&D Systems). Appropriate goat, mouse IgG2A, or mouse IgG1 isotype controls (R&D Systems) were prepared for all samples.
Cytokine protein production
Cells (1 × 105) were left untreated or stimulated with lipopeptide (MALP-2 or Pam3), LPS, uCpG, control DNA, PMA, TNF-α, CF bronchoalveolar lavage fluid (BALF), PCM, or vehicle controls, as indicated. In some experiments, cells were pretreated with actinomycin D (10 μg/ml) or TLR2.1 (eBioscience) (5 μg/ml) for 1 h before agonist treatment. IL-8 and IL-6 protein concentrations in the cell supernatants were determined by sandwich ELISA (R&D Systems). All assays were performed in duplicate or triplicate a minimum of three times.
BALF was collected from individuals with CF (n = 7) following informed consent using a protocol approved by Beaumont Hospital Ethics Committee (32). Samples were filtered through gauze and centrifuged at 1000 × g for 10 min, and cell-free supernatants were aliquoted and stored at −80°C. Antigenic NE levels and NE activity were calculated in pooled BALF, as previously described (33, 34). LPS levels were quantified using the QCL 1000 Limulus amebocyte lysate assay (Cambrex). Pseudomonas DNA was detected in individual BALF samples by PCR using PAO1-specific gene primers for fliC (forward, 5′-GAACTGACCCGTATCTCCGA-3′ and reverse, 5′-GTCAATGGTCTCGTATGGCGT-3′, product 111 bp) and lasB (forward, 5′-CGAACCCGGTCTAAAGTTCA-3′ and reverse, 5′-CAAGTGATTGACGCAGGCTA-3′, product 103 bp). Thermocycling conditions were 95°C for 5 min, followed by 35 cycles of 95°C for 30 s, 55°C for 30 s, and 72°C for 30 s. A final extension step of 72°C for 10 min was performed. Products were resolved on 1.5% TBE agarose gels containing 0.5 μg/ml ethidium bromide (Sigma-Aldrich), and images were captured using the GeneGenius Gel Documentation and Analysis System (Syngene).
Transfection and reporter gene studies
Transfections were performed into cells (1 × 105) with TransFast Reagent (Promega) in a 1:1 ratio, according to the manufacturer’s instructions, using 100 ng of NF-κB5-linked luciferase reporter plasmid and 200 ng of either pCDNA3.1 (Invitrogen Life Technologies) or a ΔMyD88 expression plasmid (35, 36) or pDC304 or a Mal P/H expression vector (a gift from A. Bowie, Dublin, Ireland) (37). ΔMyD88 contains only a functional TIR domain and lacks the death domain required for downstream signaling, while Mal P/H is a dominant-negative version of Mal with a Pro125His point mutation in box 2 of the TIR domain. Uniform transfection efficiencies were achieved by initially optimizing transfection conditions using a constitutive luciferase expression vector, pGL3-control (Promega). Cells were then supplemented with additional growth medium for 24 h at 37°C before being left untreated or stimulated, as indicated. Cells were lysed with reporter lysis buffer (Promega), and reporter gene activity was quantified by luminometry (PerkinElmer Wallac Victor2, 1420 multilabel counter) using the Promega luciferase assay system, according to the manufacturer’s instructions.
Data were analyzed with GraphPad Prism 3.0 software package (GraphPad). Results are expressed as mean ± SE, and were compared by Mann-Whitney U test. Differences were considered significant when the p value was ≤0.05.
TLR expression by CF and non-CF airway epithelial cells
CFTE29o− and 16HBE14o− cells were assessed for GAPDH, CD14, TLR1–6, and TLR9 RNA expression by RT-PCR (Fig. 1). Similar TLR and CD14 RNA expression profiles were also seen in CFBE41o− cell lines (data not shown).
TLR1–6, TLR9, and CD14 cell surface protein expression was assessed in CFTE29o− and 16HBE14o− cells using laser-scanning microscopy. Cell surface TLRs1–5 and TLR9, but not TLR6, were detected by microscopy on both the CFTE29o− and 16HBE14o− cell lines (Fig. 1). Both airway epithelial cell lines also expressed only low levels of CD14 on the surface. Western blotting of membrane fractions also detected TLR2, TLR4, TLR5, and TLR9 expression in both cell lines (data not shown).
Diacylated and triacylated lipopeptide, LPS, and uCpG dinucleotides induce IL-8 expression from CFTE29o− cells
The effect of TLR2/6, TLR2/1, TLR4, and TLR9 agonists on IL-8 protein production in CFTE29o− cells was assessed. Diacylated lipopeptide (MALP-2) dose dependently induced IL-8 expression from CFTE29o− cells after 24 h at 4, 40, and 400 ng/ml (p ≤ 0.05). Triacylated lipopeptide (Pam3) and LPS also induced IL-8 expression from CFTE29o− cells after 24 h at doses of 1, 10, and 50 μg/ml (p ≤ 0.05) (Fig. 2,A). uCpG DNA induced IL-8 expression from CFTE29o− cells only after 48 h at 10 and 100 μg/ml (p ≤ 0.05). However, uCpG did significantly increase IL-8 expression from myelomonocytic U937 cells at both 24 and 48 h (83 ± 2 vs 120 ± 19 and 106 ± 4 pg/ml IL-8 for control vs uCpG (30 μg/ml) at 24 and 48 h, p ≤ 0.05). Both phosphodiester- and phosphorothioate-modified uCpG DNA induced similar effects; however, phosphodiester-containing oligonuceotides were used because high concentrations of phosphorothioate-containing oligonucleotides can induce CpG-independent effects (38). The lipopeptide effect was inhibited by pretreatment with a TLR2-neutralizing Ab (Fig. 2,B); the LPS effect was inhibited by polymixin B (Fig. 2,C); a control oligonucleotide that did not contain uCpG dinucelotides failed to increase IL-8 expression from CFTE29o− (Fig. 2,D); and all effects were blocked by pretreatment with actimomycin D (Fig. 2 E) (p ≤ 0.05). PCM could also dose dependently increase IL-8 expression from CFTE29o− cells at 24 h at doses of 1, 5, or 10% (v/v) (132 ± 13 vs 201 ± 20, 342 ± 75, and 452 ± 44 pg/ml IL-8 for control vs 1, 5, or 10%, respectively, p ≤ 0.05).
Similar to their effects in CFTE29o− cells, triacylated lipopeptide, LPS, and uCpG also increased IL-8 protein production from 16HBE14o− cells (p ≤ 0.05) (Fig. 2 F). Diacylated lipopeptide failed to increase IL-8 (or IL-6) expression from 16HBE14o− cells after 24 h (data not shown). The relative amounts of IL-8 induced by 16HBE14o− and CFTE29o− cells were similar, although the CFTE29o− cells expressed higher levels of IL-8 basally and following agonist treatment than the16HBE14o− cells.
IL-6 and ICAM-1 expression are increased by lipopeptides, LPS, and uCpG DNA in CFTE29o− cells
Diacylated and triacylated lipopeptides, LPS, and uCpG DNA, but not a control oligonucleotide, induced IL-6 expression from CFTE29o− cells similar to TNF-α after 48 h (data not shown, p ≤ 0.05) (Fig. 3,A). As before, these effects were blocked by polymixin B or actimomycin D, as appropriate (p ≤ 0.05, Fig. 3 B). PCM (1% v/v) also increased IL-6 production from CFTE29o− cells compared with untreated cells after 24 h (259 ± 34 vs 386 ± 12 pg/ml IL-6 for control vs PCM, p ≤ 0.05).
Expression of the adhesion molecule ICAM-1 on CFTE29o− cells was quantified in response to 48 h of stimulation with lipopeptide, LPS, uCpG, or TNF-α (Fig. 3 C). All agonists significantly up-regulated ICAM-1 compared with unstimulated control cells (p ≤ 0.05).
ΔMyD88 inhibits lipopeptide-, LPS-, and CF BALF-induced NF-κB reporter gene expression in CFTE29o− cells
BALF isolated from individuals with CF (CF BALF) contains proinflammatory factors present within the CF lung, including LPS (39), bacterial DNA, and NE, among others. The seven pooled CF BALF samples contained 677 endotoxin U/ml, equivalent to 1.4 μg/ml LPS, and 743 μg/ml NE, which had 14.3% activity and is equivalent to 3.7 μM NE. Given that the process of CF BALF collection involves a 25- to 50-fold dilution of epithelial lining fluid, these values correspond to ∼35–70 μg/ml LPS and 92–185 μM NE on the respiratory surface in vivo. PAO1-specific gene primers were used in PCR to detect the Pseudomonas flagellin subunit gene, fliC, and the Pseudomonas elastase gene, lasB, in each CF BALF sample (Fig. 4,A). fliC and lasB were detected in four BALF samples (lanes 1, 2, 4, and 7). We assessed the ability of CF BALF (10 μl) to induce IL-8 protein production from CFTE29o− cells and compared its effects with those of NE and LPS alone and in combination (Fig. 4 B). Stimulation with both NE (10 nM) and LPS (10 μg/ml) together induced higher levels of IL-8 production compared with each agonist used alone, and their combined effects were less than those induced by 10 μl of CF BALF, which contains 37 nM NE and 14 ng of LPS, and many other proinflammatory factors.
NF-κB reporter gene expression in CFTE29o− cells was measured in response to stimulation with pooled CF BALF or lipopeptide, LPS, or PMA for 18 h. Each of these agonists significantly increased NF-κB reporter gene expression (p < 0.05) compared with untreated cells (Fig. 5). uCpG DNA did not increase reporter gene expression at 18 h (data not shown). Cotransfection with a plasmid expressing a signaling-deficient mutant of MyD88, termed ΔMyD88, inhibited the stimulatory effects of lipopeptide, LPS, and CF BALF by 40 ± 15, 43 ± 4, and 33 ± 1%, respectively (p < 0.05). PMA-induced NF-κB reporter gene activation was not inhibited by ΔMyD88.
ΔMyD88 and ΔMal inhibit lipopeptide-, LPS-, and CF BALF-induced IL-8 protein expression in CFTE29o− cells
IL-8 protein production by CFTE29o− cells was also measured in response to stimulation with lipopeptide, LPS, CF BALF, or PMA for 18 h. Each of these agonists significantly increased IL-8 protein production (p < 0.05) compared with untreated cells (Fig. 6). Overexpression of ΔMyD88 or an inactive mutant of the MyD88 adaptor protein Mal, Mal P/H, inhibited the lipopeptide, LPS, and CF BALF (p < 0.05), but not the PMA-induced effects.
Inflammation in the CF lung is a neutrophil-dominated event; however, the epithelium plays an important role by regulating neutrophil recruitment, and therefore contributes significantly to the overall inflammatory response within the lung. In light of the current intensive research in the area of TLRs and general widespread interest in innate immunity, we thought it was timely to investigate in more detail innate immune responses of CF airway epithelial cells, in particular those activated by specific TLRs.
TLRs represent a major arm of the innate immune system. Their principal role is to recognize and discriminate microbial components and mount a rapid protective inflammatory response (8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25). TLR function has been most widely studied in immune cells to date; however, given their broad tissue distribution, it is likely that TLRs fulfill their known roles in other cell types as well. In this study, we have evaluated the roles of specific TLRs in CF airway epithelial cells, in particular those relevant to key proinflammatory agents present in the CF lung. Our studies found that TLRs1–5 and TLR9 are expressed on the surface of CF and non-CF tracheal and bronchial epithelial cell lines. Others have reported similar findings in 16HBE14o− cells and primary CF nasal polyp tissue in culture (28); however, it has been reported that TLR4 is not surface expressed on BEAS-2B and A549 airway epithelial cells (27), suggesting heterogeneity of TLR expression exists between different airway epithelial cell types. Although mRNA for TLR6 was present in all of the epithelial cell types examined, we failed to detect TLR6 protein expression; nonetheless, the CFTE29o− cells were responsive to the TLR2/6 agonist MALP-2. Interestingly, MALP-2 failed to induce a response in the non-CF cell line, suggesting that MALP-2 responsiveness may be enhanced in cells with impaired CFTR and raises the question of whether TLR2 may be a modifier gene for CF. Many studies have proposed the linkage between mutations in CFTR and other genetic loci (40), and it is possible that mutations in CFTR and TLR genes could be linked given the important role of TLR agonists in CF lung disease. The enhanced MALP-2 responsiveness of the CF cell line compared with non-CF cells could also be due to the less tight junctions of CF airway epithelial cells, thus enabling MALP-2 to access additional TLR2 proteins on the basolateral surface of the CFTE29o− cells (30). Alternatively, the differential responses observed may be due to tissue specificity given that CF cell line tested was derived from trachea, whereas the non-CF cell line was bronchial derived. Differential responses in these two cell types have been documented in the past (41). Our finding that TLR9 is expressed on the surface of epithelial cells is in contrast to studies performed using macrophages in which TLR9 was found to be localized to endosomes and shown to become activated following internalization of uCpG DNA (42). Similar to a previous report, we found that expression of CD14 by CF and non-CF airway epithelial cells was low (27).
Activation of TLRs 2/6, 2/1, 4, and 9 by their cognate agonists induced proinflammatory responses in the CF airway epithelial cells. Expression of IL-6, IL-8, and ICAM-1, all NF-κB-regulated genes, was increased both time and dose dependently following stimulation with lipopeptides, LPS, or uCpG DNA. MALP-2 is active at very low concentrations and induced functional responses at doses of 4–400 ng/ml. In contrast, the levels of triacylated lipopeptide and LPS required to induce similar responses were significantly higher, 1–50 μg/ml; however, these amounts are physiologically relevant in CF in vivo. In the context of CF, IL-8 is particularly important, as it is a potent neutrophil chemotactic factor (43). It attracts neutrophils to the inflammatory site, where their transepithelial passage is facilitated by ICAM-1 (44). uCpG was a less potent stimulus than lipopeptides or LPS, inducing its effects at 48–72 h rather than 24 h poststimulation. This is a cell-specific phenomenon, as we found that uCpG could strongly induce IL-8 expression in myelomonocytic U937 cells at 24 h.
A primary aim of this work was to examine whether so-called nonimmune airway epithelial cells contribute to the inflammatory response in the CF lung. Given that airway epithelial cells do induce functional TLR responses, it appears that they have an important role in pulmonary inflammation in CF given their immense surface area. In addition to lipopeptides, LPS, and uCpG DNA, we also evaluated the proinflammatory properties of PCM and CF BALF on IL-6 and IL-8 protein production by CFTE29o− cells. PCM contains factors secreted by stationary phase culture of PAO1, and is highly proinflammatory (45). As well as lipopeptides, LPS, and uCpG DNA, other immunostimulatory factors present in PCM are likely to include flagellin, pyocyanins, Pseudomonas elastases, and exotoxins. Flagellin, the major structural component of flagella, is encoded by the fliC gene. Purified rFliC from Salmonella enteritidis is a potent inducer of cytokine synthesis in macrophages (46), and Salmonella typhimurium flagellin has been shown to induce IL-8 production by intestinal epithelial cells (47). Like other members of the TLR family, TLR5 signals via the same intracellular signaling molecules. In this study, we did not directly evaluate the proinflammatory effects of PAO1 flagellin on TLR5-mediated responses because in the CF lung Pseudomonas exists as a biofilm. Flagellin subunit genes are negatively regulated during biofilm formation (48, 49) and by azithromycin (50), an antibiotic commonly prescribed to individuals with CF, and, as such, flagellin may not be a major proinflammatory stimulus in the CF lung with an established Pseudomonas biofilm.
CF BALF contains a representative sampling of host and pathogen factors present on the epithelial surface of the lung. In addition to LPS (39), we also detected PAO1 DNA in our CF BALF samples. It has previously been shown that uCpG DNA isolated from CF sputum induces lower respiratory tract inflammation in an animal model (51); however, the epithelium may not play a major role in these events given that TLR9-induced responses were relatively weak in the CFTE29o− and 16HBE14o− cells. Viruses and ssRNA and dsRNA may also be present in CF BALF. These potentially signal via TLRs 3, 7, 8, and 9 (13, 22, 23, 52) (and TLR4, in the case of respiratory syncitial virus (53)) and are likely to be important in modulating the immune response in the CF lung during viral infection. Another abundant factor present in the CF BALF was NE, the neutrophil-derived protease that can induce IL-8 gene expression in bronchial epithelial cells via TLR4 (7, 17). Given the heterogeneous composition of CF BALF, it is unlikely that any one single factor is entirely responsible for the proinflammatory properties of CF BALF. Our studies measuring the costimulatory effects of different TLR agonists suggest that different components within CF BALF may be additive and combine to induce its potent proinflammatory effects.
We have previously demonstrated that NE-induced IL-8 expression can be inhibited by ΔMyD88 (54). Building on these studies, we show in this study that ΔMyD88 also inhibits NF-κB-linked reporter gene expression in CF airway epithelial cells in response to lipopeptide, LPS, and CF BALF. Furthermore, both ΔMyD88 and a dominant-negative mutant of its adaptor Mal can abrogate IL-8 protein production by each of these stimuli, providing direct evidence of a potential role for these inhibitors as CF therapeutics. Pulmonary inflammation in CF is characterized by production of thick inspissated mucus, bacterial colonization, and severe inflammation, and because of this, many of the efforts directed toward curing CF have targeted the lungs. The results to date with gene therapy have been disappointing, and now many CF researchers are considering a multifaceted approach to curing CF, recognizing that therapies that ameliorate abnormal mucus production, bacterial colonization, and inflammation will almost certainly lead to improved survival.
In addition to MyD88 and Mal, other TIR domain adaptor proteins exist that function in a MyD88-independent fashion. The TRIF protein is responsible for eliciting TLR3 response to viruses and some TLR4 responses to LPS (55, 56). Both TLRs using TRIF to induce IFN-regulatory factor (IRF)-3 or -7 regulated expression of genes encoding IFN-α and -β, IFN-γ-inducible protein-10, and RANTES. Two additional TIR domain-containing adaptors have been identified in humans, termed TRIF-related adaptor molecule and sterile α and HEAT-Armadillo motifs (57). TRIF-related adaptor molecule participates in the TLR4 MyD88-independent pathway, leading to activation of NF-κB and IRF-3 and -7, whereas sterile α and HEAT-Armadillo motifs have yet to be characterized fully. However, it remains to be shown whether targeting the MyD88-independent signaling pathway represents a therapeutic target for preventing inflammation in CF, as there is no information regarding activation of IRF-3/7, IFN-α or -β, or IFN-γ-inducible protein-10 in CF; however, aberrantly low levels of RANTES are secreted by CF airway epithelial cells, suggesting that this pathway may be impaired in CF cells (58, 59). Nonetheless, it is clear from this and other studies in our laboratory that inhibitors based on the TIR domain of MyD88 or its orthologues could potentially ameliorate pulmonary inflammation induced by TLR and IL-1R agonists. An important challenge for the future will be to develop suitable delivery methods for these inhibitors and determine their compatibility with current conventional CF therapies.
We are grateful to A. Bowie (Trinity College) for providing pDC304 and Mal P/H vectors; D. Gruenert (University of Vermont) for the CFTE29o−, CFBE41o−, and 16HBE14o− cell lines; and R. Hancock (University of British Columbia) for P. aeruginosa strain 01.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was funded by research grants from Enterprise Ireland (SC/2001/104), Programme from Research in Third Level Institutes administered by Higher Education Authority, Cystic Fibrosis Association of Ireland, Alpha One Foundation, and Royal College of Surgeons in Ireland.
Abbreviations used in this paper: CF, cystic fibrosis; BALF, bronchoalveolar lavage fluid; CFTR, CF transmembrane conductance regulator; IRF, IFN-regulatory factor; MALP-2, macrophage-activating lipopeptide-2; NE, neutrophil elastase; Pam3, triacylated lipopeptide; PAO1, P. aeruginosa strain O1; PCM, PAO1-conditioned medium; PI, propidium iodide; TBE, Tris-borate-EDTA; TIR, TLR- and IL-1R-related; TRIF, TIR domain-containing adaptor inducing IFN-β; uCpG, unmethylated CpG.