|Home | About | Journals | Submit | Contact Us | Français|
Hydrogen peroxide production by the NADPH oxidase Duox1 occurs during activation of respiratory epithelial cells stimulated by purified bacterial ligands, such as lipopolysaccharide. Here, we characterize Duox activation using intact bacterial cells of several airway pathogens. We found that only Pseudomonas aeruginosa, not Burkholderia cepacia or Staphylococcus aureus, triggers H2O2 production in bronchial epithelial cells in a calcium-dependent but predominantly ATP-independent manner. Moreover, by comparing mutant Pseudomonas strains, we identify several virulence factors that participate in Duox activation, including the type-three secretion system. These data provide insight on Duox activation by mechanisms unique to Pseudomonas aeruginosa.
The dual oxidases (Duox1, 2) are two members of the Nox/Duox NADPH oxidase family, transmembrane proteins that transport electrons from cytosolic NADPH to reduce molecular oxygen into superoxide (Nox1-5) or hydrogen peroxide (Duox). In the respiratory system, Duox proteins are expressed in differentiated airway epithelial cells and localize to the apical plasma membrane, where they release hydrogen peroxide (H2O2) into the airway surface liquid (ASL) through elevations of intracellular calcium concentrations .
Sterility of the lower respiratory tract is maintained by several antimicrobial systems, including lysozyme, defensins and lactoperoxidase (LPO). Lactoperoxidase, a heme-peroxidase abundant in the ASL oxidizes thiocyanate (SCN−) into antimicrobial hypothiocyanite (OSCN−) using H2O2 [2,3]. Air-liquid interface (ALI) cultures of primary tracheobronchial epithelial cells kill several bacterial pathogens in a Duox-, LPO- and SCN− -dependent manner [2,4,5]. Moreover, Duox was suggested to participate in airway epithelial wound repair through activation by extracellular ATP derived from injured cells . Duox-derived reactive oxygen species have also been implicated in mucin production by bronchial epithelial cells . Roles for Duox in innate immune processes is further suggested by its induction by inflammatory cytokines or bacterial products; the Th1 cytokine IFN-γ induces Duox2, whereas Th2 cytokines, IL-4 and IL-13 induce Duox1 [4,8]. Bacterial flagellin up-regulates Duox2 in human tracheobronchial epithelial cells (HTBE) .
Airway epithelial cells are continuously exposed to inhaled bacteria. Bacterial attachment activates epithelial signaling pathways that enhance bacterial removal and recruit other immune cells through inflammatory cytokine release. Recently, Duox was implicated during bacterial activation of epithelial cells . LPS or an activating antibody against asialo-GM1 receptor were shown to trigger ATP release by immortalized bronchial epithelial cells (HBE1), which activates Duox1 by binding to P2YR purinergic receptors . Duox-derived H2O2 activates EGFR-dependent IL-8 release amplified by ADAM17-mediated TGFa release . Whether Duox activation by bacterial products is a general mechanism applying to many bacterial species and whether other bacterial virulence factors are involved in Duox activation are unknown.
In this study, we explored airway epithelial Duox activation using intact bacterial cells of several airway pathogens, instead of purified virulence factors. Our data indicate that only Pseudomonas aeruginosa, not Burkholderia cepacia or Staphylococcus aureus, triggers Duox1-mediated H2O2 production by bronchial epithelial cells by mechanisms that are calcium-dependent and predominantly ATP-independent. Flagellum- and LPS-deficient Pseudomonas mutants elicit lower Duox activation than wild-type strains. Moreover, we identify the type-three secretion system of Pseudomonas as a novel Duox activator.
The human bronchial mucoepidermoid cell line NCI-H292 (ATCC; cat# CRL-1848) was grown in RPMI-1640 medium (Invitrogen) containing 10% FBS, 1% penicillin-streptomycin, 1% L-glutamine, 1% sodium-pyruvate and 1% HEPES. Following overnight serum starvation, IL-4 (10 ng/mL) or IL-13 (10 ng/mL) (R&D Systems) was added in serum-free RPMI-1640 medium and cells were assayed two days later.
The Cdk4/hTERT immortalized human bronchial epithelial cells were a kind gift of Dr. John Minna (University of Texas, Southwestern Medical Center) . The cells were cultured for two weeks prior used for experiments on 6-well collagen-coated semipermeable transwell supports on air-liquid interface in Rheinwal-Green/Wu medium + ascorbic acid (1 part Ham's F12, 3 parts DMEM, 5% Cosmic Calf serum, 0.1% penstrep+ 400 ng/mL hydrocortisone, 5 ng/mL EGF, 5 μg/mL apo-transferrin, 5 μg/mL insulin, 10 ng/mL cholera toxin, 13 ng/mL triiodothyronine, 49 μg/mL adenine, 50 μg/mL ascorbic acid.
Pseudomonas aeruginosa strains used were: P. aeruginosa ATCC 10145 (PA10145; ATCC); wild-type PAO1 and isogenic single gene knock-out mutants (Pseudomonas Mutant Library, University of Washington, Seattle) deficient in the following virulence factors: A-band LPS (oxidoreductase Rmd-deficient, ID#: 2918), flagellum (fliC-def., ID#: 245), pilus (pilA-def., ID#: 33348), exoenzyme S (ID#: 52760), exoenzyme T (ID#: 9353) and exoenzyme Y (ID#: 21412), alkaline protease (ID#: 13536) and alkaline metalloprotease (ID#: 16254). Wild-type PA14 and its pyocyanin-deficient mutant (PhzM) were gifts from Dr. Frederick Ausubel (Harvard Medical School, Boston). Burkholderia cepacia and Staphylococcus aureus were gifts from Dr. Steven Holland (NIAID, NIH, USA). Bacteria were grown in Luria-Bertani broth (KD Medical) for up to 3 days (shaking, 37 C). Heat killing was done at 95 C for 5 min.
Extracellular H2O2 release was measured by Luminol/HRP chemiluminescence assay . Trypsinized, IL-4- or IL-13-induced NCI-H292 cells were preincubated with or without 10μM diphenylene iodonium (DPI) (37 C, 10min) in HBSS (cell density 5×106/mL) and were stimulated by addition of an equal volume of HBSS containing 2×108/mL bacteria, 1 mM Luminol and 20 U/mL horse-radish peroxidase. Most experiments used bacteria grown for 6-8 hrs, washed twice and resuspended in HBSS, although one experiment explored cultures grown for 3-48 hours (Fig 4B). In experiments studying effects of bacterial supernatants, 1.5-day culture supernatants were diluted 1:20 in HBSS and mixed with cytokine-induced NCI-H292 cells. To inhibit effects of extracellular ATP, apyrase (10 U/mL, Sigma) or suramin (10 and 100 μM, Sigma) was used. In some experiments calcium was added to calcium-free HBSS in the range of 1-1000 μM, and a zero calcium concentration was achieved using 100 μM EGTA added to calcium-free HBSS. All other experiments used 1 mM calcium concentrations. Luminescence was recorded in a 96-well plate reader (Luminoskan Ascent, Thermo) for 1-3 hours. H2O2 release was shown either as kinetics of relative luminescence units (RLU) over time or as barograms representing integrated luminescence values for the entire measurement period (int. RLU).
NCI-H292 cell lysates were processed for Western blotting as described . Primary antibodies used were: anti-Duox (rabbit, 1:2000, gift from C. Dupuy), anti-α-tubulin (mouse, 1:2000; Santa Cruz); anti-TLR4 (rabbit, 1:2000; Novus); anti-TLR5 (rabbit, 1:2000; Santa Cruz), anti-CFTR (mouse, 1:100, R&D Systems). Secondary antibodies used: HRP-linked anti-rabbit IgG from donkey (1:1000, GE Healthcare) and HRP-linked anti-mouse IgG from sheep (1:1000, GE Healthcare).
Total RNA was isolated from NCI-H292 cells by Trizol/chloroform extraction followed by isopropanol precipitation and washes in 70% ethanol. cDNA was synthesized with Thermoscript cDNA synthesis kit (Invitrogen) using 2μg total RNA, oligo dT primers and RNaseH treatment. To detect asialo GM1 and TLR2 expression, the following gene-specific primers were used: asialo GM1 (F: 5′-gctggagaaacagcagaagg-3′; R: 5′-aggtcagacacgaactgcttc-3′), TLR2 (F: 5′-cgttctctcaggtgactgctc-3′; R: 5′- cctttggatcctgcttgc -3′). PCR reactions were carried out in a Biometra PCR thermocycler. β-actin was used as reference (F: 5′-ccaaccgcgagaagatga-3′; R: 5′-ccagaggcgtacagggatag-3′). PCR products were resolved on 10% Acrylamide/TBE gels.
To silence gene expression by RNA interference, NCI-H292 cells were transfected with 100 nM siRNAs, using Lipofectamine-2000 by transfection methods previously described . For Duox1-targeted interference, a pool of 3 targeting siRNAs was used (Duox1#1, #2, and #4) at concentrations of 33 nM each.
The following siRNAs were used (sequence of the sense strand): Duox1 #1: 5′-GGACUUAUCCUGGCUAGAG-3′; Duox1 #2: 5′-GGAUAUGAUCUGUCCCUCU-3′; Duox1 #4: 5′-GCUAUGCAGAUGGCGUGUA-3′; β-actin: 5′-GAUGAGAUUGGCAUGGCU-3′; negative control: 5′-CCGUAUCGUAAGCAGUACU-3′
Data are presented as mean by ANOVA and post-hoc Dunnett's test when results of trends were compared. Significant changes were marked as * when p<0.05, ** when p<0.01 and *** when p<0.001.
To examine Duox1 activation by exposure to airway pathogens, we used mucoepidermoid bronchial epithelial NCI-H292 cells, widely used in bronchial epithelial cell studies, induced with IL-4 or IL-13. Both cytokines were shown to up-regulate Duox1 in HTBE cells  and NCI-H292 cells  (Fig. 1A). We also detected NCI-H292 cell expression of TLR2, TLR4, TLR5 and asialoGM1, receptors that bind bacterial flagellum, pilus or LPS (Fig.1A, B) . We tested Staphylococcus aureus, Pseudomonas aeruginosa, and Burkholderia cepacia, species frequently found in patients suffering from bacterial pneumonia and cystic fibrosis (CF)  and included Escherichia coli as a reference species not typically infecting the respiratory tract. As shown in Fig. 1C, only Pseudomonas aeruginosa (PA10145) was found to trigger H2O2 release by IL-4- and IL-13-induced epithelial cells. None of the bacterial supernatants from these pathogens elicited H2O2 production (Fig.1D). We tested two other widely used Pseudomonas aeruginosa wild-type strains (PAO1 and PA14) and found that both elicited release of H2O2 by IL-4-induced epithelial cells, which was greatly reduced by the flavoenzyme inhibitor DPI (Fig. 2A). Bacterial H2O2 release was not significant (Fig. 2A, B). No significant H2O2 release by NCI-H292 cells was observed in response to P. aeruginosa without prior induction of Duox1 by the cytokines (data not shown). To show that Pseudomonas-triggered H2O2 production is not unique to cytokine-treated NCI-H292 cells, we tested Cdk4/hTERT human immortalized bronchial epithelial cells (hBEC) cultured on an air-liquid interface. These cells produce H2O2 in response to ionomycin and express Duox1 as the major Duox isoform (data not shown). We found that Cdk4/hTERT hBECs respond with H2O2 release to Pseudomonas aeruginosa (Fig. 2C), but not to Staphylococcus aureus, Burkholderia cepacia and Escherichia coli (data not shown). Next we showed that heat-killed Pseudomonas or their supernatants failed to elicit H2O2 production in IL-4-treated NCI-H292 cells (Fig. 2D), suggesting an active mechanism requiring live bacteria or heat-labile, surface-attached virulence factors. When omitting luminol or HRP (or both) from the assay, the Pseudomonas-triggered luminescent signal disappeared, verifying that detection of H2O2 does not involve bacterial or epithelial peroxidases (Fig. 2E). Our results show for the first time that bronchial epithelial cells respond with H2O2 release following exposure to intact Pseudomonas aeruginosa.
To confirm that Duox1 is the source of H2O2 produced in Pseudomonas-exposed epithelial cells, we used pooled Duox1-targeted siRNAs to suppress Duox1 expression in IL-4-induced NCI-H292 cells. Fig. 3A shows that Duox1 siRNA-treated epithelial cells have greatly diminished H2O2 release, whereas cells treated with negative control or β actin-targeted siRNAs respond with similar H2O2 production to PAO1 exposure as untreated cells. Non-targeted and actin siRNAs did not alter Duox protein levels, whereas pooled Duox1-targeted siRNAs greatly reduced Duox protein levels (Fig. 3B). This suggests Duox1 is the major source of H2O2 in epithelial cells responding to Pseudomonas.
Calcium ions required for Duox activation can originate from the extracellular space or from internal stores. By gradually decreasing extracellular Ca2+ concentrations (1000, 300, 100, 10, 0 μM), Pseudomonas-stimulated Duox activation steadily decreased in IL-4-induced NCI-H292 cells (Fig. 4A). Thus, activation of Duox by Pseudomonas aeruginosa is dependent on extracellular calcium.
Growth of Pseudomonas aeruginosa in vitro mimics its natural life cycle. The early logarithmic phase corresponds to planktonic, mobile forms of the bacterium, whereas stationary phase or saturated cultures exhibit more features of established biofilms as seen in long-term chronic infection in CF patients . Each phase has its characteristic virulence factors, some appearing only in young cultures or in old cultures or throughout the entire cycle . To gain further insight into the mechanism of Duox activation by Pseudomonas, we compared PAO1 bacteria cultured for different times (3-48 hrs) for their ability to activate Duox. We found young cultures (6-9 hrs) representing planktonic forms most effectively trigger Duox activation, which gradually decreases with older cultures (24, 48 hrs) (Fig. 4B). This may reflect the disappearance of Duox-activating factors typical in the early phase, the appearance of Duox-inhibiting factors typical for the late growth phase  or a combination of both processes.
We examined involvement of ATP release in Duox activation by Pseudomonas, since ATP release was proposed to trigger LPS- and asialoGM1-mediated activation of Duox1 in HBE1 cells . Apyrase (an extracellular ADP/ATPase) and suramin (a P2YR blocker) were shown to inhibit Duox1 activation . We tested the same concentrations of inhibitors in our system using whole bacteria and found apyrase had only minor inhibitory effects on Pseudomonas-stimulated Duox activation, whereas it completely inhibited Duox activation by exogenous ATP (Fig. 4B). We also observed only marginal inhibition by suramin (Fig. 4C). These data suggest bacteria stimulate Duox mainly through ATP-independent mechanisms.
To screen potential Pseudomonas virulence factors responsible for Duox activation, we exposed IL-4-induced NCI-H292 cells to PAO1 single-gene mutants deficient in A-band LPS, flagellum, pilus, exoS, exoY, exoT, alkaline protease and alkaline metalloproteinase, along with the parental PAO1 strain, and compared their abilities to trigger H2O2 production. Purified LPS, flagellum and pilus have already been suggested as triggers of Duox1 activation in bronchial epithelial cells , but the type-three secretion system (T3SS) and extracellular proteases have never been explored. We found that mutants deficient in flagellum, LPS and the T3SS triggered significantly less H2O2 than the wild-type strain (Fig. 5A). The difference between the parental PAO1 strain and the protease- and pilus-mutants were not significant (Fig.5A). Flagellum, pilus and T3SS are virulence factors all expressed in planktonic forms of Pseudomonas, consistent with the findings on young cultures (Fig. 4B). We also tested if pyocyanin influenced the luminescence signal by comparing wild-type PA14 and pyocyanin-deficient (PhzM) strains, but did not detect any significant differences in stimulated H2O2 production (Fig. 5B). Young cultures of Pseudomonas aeruginosa used here to elicit Duox1-mediated H2O2 production are not expected to produce significant pyocyanin typical of late-phase cultures .
Our results show for the first time Duox1 activation in bronchial epithelial cells exposed to Pseudomonas aeruginosa. Two other common airway pathogens Staphylococcus aureus and Burkholderia cepacia failed to elicit Duox activation. S. aureus is a Gram + organism whereas Pseudomonas is gram -, although the reasons why the more closely-related B. cepacia does not stimulate Duox may reflect the absence of specific activating factors or the presence of other Duox inhibitory or ROS scavenging mechanisms. Duox1 stimulation by Pseudomonas requires living bacteria and extracellular calcium, it is only marginally ATP-dependent, and is a multi-factorial event involving several bacterial surface virulence factors (flagellum, LPS) as well as the T3SS. The surface factors may collaborate to enable more efficient microbial-epithelial cell interactions, thereby promoting a more effective T3SS-dependent activation of Duox. The T3SS of Pseudomonas is an early-phase factor essential for full virulence in animal models . CF patients are colonized early by Pseudomonas aeruginosa expressing the T3SS [16,17]. Since ExoY is an adenylyl cyclase and Duox1 activity is enhanced by cAMP-dependent kinase, ExoY could directly affect Duox1 activity through phosphorylation [18,19]. ExoS and ExoT have ADP-ribosyltransferase activities, which could activate Gs protein and adenylyl cyclase and also lead to Duox1 activation by enhancing intracellular cAMP levels. Our data suggest activation of Duox1 in tracheobronchial epithelial cells exposed to Pseudomonas aeruginosa contributes to early detection and elimination of this microorganism. Duox1-derived hydrogen peroxide can support killing of Pseudomonas by the LPO/SCN− system and may contribute to epithelial IL-8 production to alert neutrophils [1,2,5,10,20]. Thus, cooperativity in activation of Duox by multiple bacterial virulence factors can be a key element for successful clearance of Pseudomonas aeruginosa from healthy airways and may explain the specificity of epithelial cell responsiveness to this airway pathogen.
We are grateful for research materials from the following sources: Anti-Duox antibodies (Dr. Corinne Dupuy, Institut Gustave-Roussy, FRE2939 CNRS), PAO1 strains (The Pseudomonas Mutant Library, UW, Seattle); PA14 strains (Dr. Frederick Ausubel, Massachusetts General Hospital, Boston); B. cepacia and S. aureus (Dr. Steven Holland, NIAID, NIH). This work was supported by the Intramural Research Program of the NIH, National Institute of Allergy and Infectious Diseases.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.