INFECTION AND IMMUNITY, Sept. 2010, p. 3822–3831 Vol. 78, No. 90019-9567/10/$12.00 doi:10.1128/IAI.00433-10Copyright © 2010, American Society for Microbiology. All Rights Reserved.

Naturally Produced Outer Membrane Vesicles from Pseudomonas aeruginosaElicit a Potent Innate Immune Response via Combined Sensing of Both

Lipopolysaccharide and Protein Components�

Terri N. Ellis, Sara A. Leiman, and Meta J. Kuehn*Department of Biochemistry, Duke University Medical Center, Durham, North Carolina 27710

Received 27 April 2010/Returned for modification 19 May 2010/Accepted 18 June 2010

Pseudomonas aeruginosa is a prevalent opportunistic human pathogen that, like other Gram-negative patho-gens, secretes outer membrane vesicles. Vesicles are complex entities composed of a subset of envelope lipidand protein components that have been observed to interact with and be internalized by host cells. This studycharacterized the inflammatory responses to naturally produced P. aeruginosa vesicles and determined thecontribution of vesicle Toll-like receptor (TLR) ligands and vesicle proteins to that response. Analysis ofmacrophage responses to purified vesicles by real-time PCR and enzyme-linked immunosorbent assay iden-tified proinflammatory cytokines upregulated by vesicles. Intact vesicles were shown to elicit a profoundlygreater inflammatory response than the response to purified lipopolysaccharide (LPS). Both TLR ligands LPSand flagellin contributed to specific vesicle cytokine responses, whereas the CpG DNA content of vesicles didnot. Neutralization of LPS sensing demonstrated that macrophage responses to the protein composition ofvesicles required the adjuvantlike activity of LPS to elicit strain specific responses. Protease treatment toremove proteins from the vesicle surface resulted in decreased interleukin-6 and tumor necrosis factor alphaproduction, indicating that the production of these specific cytokines may be linked to macrophage recognitionof vesicle proteins. Confocal microscopy of vesicle uptake by macrophages revealed that vesicle LPS allows forbinding to macrophage surfaces, whereas vesicle protein content is required for internalization. These datademonstrate that macrophage sensing of both LPS and protein components of outer membrane vesiclescombine to produce a bacterial strain-specific response that is distinct from those triggered by individual,purified vesicle components.

The innate immune response to Gram-negative bacteria isdominated by recognition of lipopolysaccharide (LPS). LPS issensed by the Toll-like receptor 4 (TLR4) complex, and nu-merous studies have focused on both how LPS sensitivitydrives effective inflammatory responses, leading to the clear-ance of infection, as well as how uncontrolled TLR4 signalingcan lead to LPS toxicity and play a role in septic shock (20, 39,48). Much of this research has been performed using purifiedLPS in the experimental treatments, and yet it is unlikely thatpure LPS is shed from bacteria in the context of an infection.Instead, LPS has been found to be shed from the bacteria inthe form of outer membrane vesicles.

Outer membrane vesicles are spherical, selective portions ofouter membrane and periplasm that are naturally secreted byall Gram-negative bacteria (7, 31, 49). Vesicles are produced atall stages of bacterial growth and have been detected in in-fected human tissues (7, 23, 28). Vesicle production has beenidentified as an independent bacterial stress response pathwaythat is activated when bacteria are exposed to environmentalstress, such as might be experienced during colonization ofhost tissues (35).

Compared to preparations of pure LPS, natural outer mem-brane vesicles are heterogeneous proteoliposomes of a larger

dimension (50 to 250 nm in diameter) than liposomes com-posed solely of LPS (4, 7). Natural outer membrane vesiclesare heterogeneous complexes of pathogen-associated molecu-lar patterns (PAMPs), such as LPS, flagellin, and CpG DNA,as well as other outer membrane proteins, virulence factors,and envelope lipids (7, 31). The molecular composition ofvesicles varies with the bacterial strain of origin, while LPSstructure remains relatively constant within a bacterial species.The combination of PAMPs, virulence factors, and other outermembrane components results in vesicles that are particularlyladen with molecules that can be recognized by the immunesystem. In the present study we have focused on characterizingmacrophage innate immune responses to the combined signalspresented by the heterogeneous PAMP ligands presented inthe native context of outer membrane vesicles.

Studies of host immune responses to outer membrane ves-icles have mainly addressed the generation of antibodies tovesicle components. Notably, a protective antibody response iselicited by outer membrane vesicles generated from Neisseriameningitidis and Vibrio cholerae (17, 18, 38, 40, 47). In contrast,there are relatively few studies characterizing how vesicles trig-ger the innate inflammatory response. Outer membrane vesi-cles from Helicobacter pylori and Pseudomonas aeruginosa havebeen shown to elicit interleukin-8 (IL-8) production by epithe-lial cells (3, 24), and Salmonella enterica serovar Typhimuriumvesicles have been shown to activate dendritic cells to secreteIL-12 and tumor necrosis factor alpha (TNF-�) (1).

This research has focused on outer membrane vesicles fromthe opportunistic pathogen P. aeruginosa. P. aeruginosa infects

* Corresponding author. Mailing address: Duke University MedicalCenter, Department of Biochemistry, Box 3711, Durham, NC 27710.Phone: (919) 684-2545. Fax: (919) 684-8885. E-mail: meta.kuehn@duke.edu.

� Published ahead of print on 6 July 2010.

3822

on June 8, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

the respiratory tract of patients with nosocomial pneumonia,cystic fibrosis, or acute respiratory distress syndrome (14).Host responses to acute pulmonary infections with P. aerugi-nosa are characterized by an intense inflammatory response.Macrophage recognition of bacterial components initiates acascade of proinflammatory cytokine secretion, resulting inlarge numbers of neutrophils infiltrating the lung to clear thebacteria (44).

We hypothesize that naturally produced vesicles may play animportant role in activating innate immune responses. Giventheir small size and high proportion of PAMPs, it is reasonableto expect that vesicles could easily infiltrate into infected tis-sues and stimulate widespread inflammation. In addition toTLR ligands, vesicles also contain a variety of protein virulencefactors that may specialize vesicles for specific functions ortypes of host cell damage. P. aeruginosa vesicles have beenshown to contain the adhesins OprF, OprG, and OprH, as wellas virulence factors that include �-lactamase, hemolysin, phos-pholipase C, antimicrobial quinolones, and quorum-sensingmolecules (3, 12, 25, 26, 31, 34). P. aeruginosa vesicle-associ-ated proteins have also been shown to directly affect host cells.Vesicle-packaged CiF protein downregulates lung epithelialexpression of the cystic fibrosis transmembrane conductanceregulator (CFTR) protein (10, 33), while an aminopeptidaseenriched in vesicles increases association of vesicles with cul-tured epithelial cells (2).

The present study describes how vesicles, with their hetero-geneous mixture of PAMPs, contribute to the innate immuneresponse during P. aeruginosa infection. We have focused onvesicle-macrophage interactions, since macrophages are senti-nel cells that initiate inflammatory defenses against colonizingbacteria. Our data demonstrate that macrophages are moresensitive to the potent stimulus of outer membrane vesiclesfrom P. aeruginosa compared to equivalent levels of pure LPS.Neutralizing the LPS reactivity of the vesicles or altering theprotein composition of the vesicles significantly impacted boththe macrophage-vesicle interactions and the cytokine re-sponses to the vesicles. Based on our findings, we conclude thata potent and distinct inflammatory response to vesicles is thecumulative effect of sensing vesicle-associated heterogeneousprotein and LPS ligands by macrophages.

MATERIALS AND METHODS

Bacterial strains. The P. aeruginosa strains used were the laboratory strainsPAO1 and PAO1�fliC and the clinical isolate S470. PAO1�fliC was provided byDan Wozniak (Ohio State University). Strain S470 was provided by J. R. Wright(Duke University) and is a minimally passaged, nonmucoid cystic fibrosis clinicalisolate.

Vesicle preparation. Vesicles were purified from cell free supernatants by themethod described by Bauman and Kuehn (3). Briefly, P. aeruginosa was grown inLB broth to early stationary phase. Cells were pelleted and supernatants wereconcentrated via a 100-kDa tangential filtration concentration unit (Pall-Gell-man) to �500 ml. The retentate was centrifuged and filtered through a 0.45-�m-pore-size Durapore polyvinylidene difluoride filter (Millipore) to remove theremaining bacteria. Vesicles were precipitated from the cell-free supernatantwith ammonium sulfate, pelleted, dialyzed overnight with HEPES (50 mM, pH7), concentrated (50-kDa MWCO Centriplus; Millipore), and adjusted to 45%Opti-Prep/HEPES-NaCl. Opti-Prep gradients were layered over the 2-ml crudevesicle samples at concentrations of 40, 35, 30, 25, and 20%. Gradients wereultracentrifuged (100,000 � g, 16 h), and 1-ml fractions were removed from thetop. Fractions were analyzed for vesicle protein content by SDS–15% PAGE andRuby Red staining. Pure vesicles were recovered from pooled peak fractions ofthe Opti-Prep gradients, reconstituted into phosphate-buffered saline (PBS; 130

mM NaCl, 8 mM Na2HPO4, 2 mM KCl, 1 mM KH2PO4 [pH 7.4]). The proteinconcentration of the vesicle preparations was determined by Bradford assay.

To calculate the quantity of LPS in each sample, serial dilutions of LPSstandards and purified vesicles were separated by SDS–15% PAGE and stainedfor LPS with a PRO-Q Emerald LPS staining kit (Molecular Probes). Thedensity of the LPS B bands was determined by using ImageJ software, and theLPS content in the vesicles was calculated by using the standard curve generatedfrom the pure LPS samples.

LPS purification. LPS from P. aeruginosa was purified by a modified version ofthe method of Zaidi et al. (55). Briefly, cells were lysed in 10% sarcosine, debriswas pelleted, and LPS was precipitated overnight with 4 volumes of 95% ethanol.LPS was pelleted by ultracentrifugation, resuspended in PBS, and digested withDNase, RNase, and pronase. The digested material was ultracentrifuged topellet the LPS, and the LPS was lyophilized.

Fluorescent labeling. Purified vesicles were fluorescently labeled by incubationwith Alexa Fluor 555-succinimidyl ester (AF555 [Invitrogen], in 0.1 M Na2CO3

[pH 9]) according to the manufacturer’s instructions for 1 h at 25°C with mixing.Free AF555 was removed from labeled vesicles by three washes in PBS(150,000 � g, 30 min).

Cell culture. RAW264.7 or MH-S mouse macrophage cells were grown inKaighn’s F-12K media with 10% fetal bovine sera and penicillin-streptomycin-amphotericin B (Fungizone). Confluent cells were harvested and seeded intowells of a 96-well plates at a concentration of 5 � 104 cells/well. Vesicles, purifiedLPS, or sham PBS treatment (10-�l doses of each) were added to each well,followed by incubation until time of assay.

Polymyxin B treatment. Purified vesicles or LPS were preincubated with 0.1mg of polymyxin B/ml for 30 min at room temperature prior to addition tomacrophages. Polymyxin B (50 �g/ml) was also added to the macrophage cellculture media for experiments using the polymyxin B-pretreated vesicles or LPS.

Proteinase K treatment. Purified vesicles were treated with 0.1 mg of protein-ase K/ml overnight at 37°C with agitation. Postincubation, soluble proteins andexcess proteinase K were removed from vesicles by three washes in PBS(150,000 � g, 30 min).

Inhibition of TLR9. RAW264.7 cells were pretreated for 1 h with 3 �g ofnuclease-resistant phosphorothioate oligodeoxynucleotide (iCpG DNA; ODN2088 5�-TCCTGGCGGGGAAGT-3�)/ml to inhibit TLR9 signaling (50).

Real-time PCR. Total RNA was harvested from cells by using a QiagenRNeasy kit and DNase treated to remove contaminating DNA. Harvested RNAwas quantitated, and 1 �g of RNA was reverse transcribed to cDNA usingoligo(dT) primers and Superscript II reverse transcriptase (Invitrogen). ThiscDNA was purified from reaction components by using a Qiagen DNA clean-upkit and reconstituted in a final volume of 50 �l. Then, 1 �l of purified cDNA wasused as the template for each real-time PCR. Real-time PCR was performed ona Bio-Rad iCycler using the iQ SYBR green Supermix. Primers for target genesare presented in Table 1. Gene expression was normalized to actin production ineach sample, and the fold induction was determined by using the ��CT method(32).

ELISA. Cytokine protein production (macrophage inflammatory protein 2[MIP-2], IL-1�, IL-6, and TNF-�) was quantified by enzyme-linked immunosor-bent assay (ELISA; R&D Systems/BD) according to the manufacturer’s instruc-tions.

Confocal microscopy. All fluorescence microscopy reagents were purchasedfrom Molecular Probes/Invitrogen unless otherwise stated. MH-S mouse alveolarmacrophages were seeded overnight onto coverslips in 24-well tissue cultureplates (105 cells per well), followed by incubation with fluorescently labeledvesicles for 15 min at 37°C. All subsequent steps were carried out on ice usingice-cold Dulbecco phosphate-buffered saline (PBS) for washes. After incubationwith vesicles, cells were washed twice to remove unbound vesicles or LPS. Cellexteriors were labeled with the FluoReporter cell surface biotinylation kit,washed twice, and incubated with AF633-conjugated streptavidin. Cells werethen washed, fixed in neutral buffered formalin, mounted with ProLong Anti-Fade reagent, and visualized on a Leica SP5 confocal microscope.

Statistics. All experiments were performed with n � 4. Statistical comparisonsbetween groups were performed using single-factor analysis of variance, followedby Tukey’s post hoc test. The statistical significance was set at P � 0.05.

RESULTS

Macrophages are more sensitive to outer membrane vesiclesthan to pure LPS. Within the lung, macrophages are critical tosensing and initiation of the inflammatory host response to P.aeruginosa (22). In order to analyze how sentinel macrophages

VOL. 78, 2010 INFLAMMATORY RESPONSES TO OUTER MEMBRANE VESICLES 3823

on June 8, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

trigger inflammation in response to outer membrane vesiclessecreted by P. aeruginosa, we developed an in vitro assay tomeasure proinflammatory cytokine gene expression by macro-phages. Outer membrane vesicles were collected from theovernight culture supernatant of clinical P. aeruginosa strainS470 and purified by density gradient. Proinflammatory genes(MIP-2, KC, TNF-�, IL-1�, IL-6, and iNOS) were selected tobe assayed based on their previously described upregulationduring active P. aeruginosa infection (5, 11, 21, 22, 41, 42, 45,51, 52). RAW 264.7 macrophages were coincubated with ves-icles for 3 h, and gene expression was assayed by real-time PCRand normalized to actin production in each sample. The foldinduction was determined by using the ��CT method (32). Todetermine the minimal dose of vesicles required to trigger aresponse, we used a 10-fold dilution series of vesicle doseswhich were standardized to LPS content. A 3-h exposure elic-ited optimal upregulation of gene expression (data not shown),and therefore this condition was used in subsequent experi-ments.

Outer membrane vesicles contain significant quantities ofLPS; however, the membrane context of vesicle LPS mightactivate a different cytokine response than extracted, pure LPS.To compare both the amount and the type of response, cyto-kine levels were measured for macrophages exposed to equiv-alent concentrations of LPS either purified or as part of a purevesicle preparation. As seen in Fig. 1A, both vesicles and pureLPS stimulated significant transcription of MIP-2, TNF-�, IL-1�, and IL-6. Other genes assayed, such as KC and iNOS, didnot exhibit significant upregulation in response to either vesi-cles or LPS (data not shown). Proinflammatory gene expres-sion was clearly dose dependent. At the highest dose (1 �g ofLPS content), both treatments triggered similar expression lev-els by all cytokines assayed, indicating that this may be a sat-uration level response (data not shown). However, sequentialdilution of doses demonstrated an increasing divergence in theresponse with vesicles triggering a more profound cytokineresponse than pure LPS as the dose decreased. Since secretedvesicles comprise a small fraction (ca. 1 to 2%) of the totalouter membrane protein of a growing culture (3), these lowerdoses represent a physiologically relevant point of comparison.

The four cytokines that demonstrated significant transcrip-tional upregulation were further analyzed by ELISA for pro-tein secretion into culture supernatant over the course of a 3-hexposure to either vesicles or purified LPS. Figure 1B showsthat similar levels of MIP-2 were secreted over this time periodin response to both LPS and vesicles, indicating that recogni-tion of LPS within vesicles may be the dominant stimulus

needed to generate this cytokine. Neither treatment stimulatedsignificant production of IL-1�, even through significantamounts of IL-1� RNA transcripts were present, as assayed byreal-time PCR. In contrast, both TNF-� and IL-6 were se-creted at significantly higher levels in response to outer mem-brane vesicles than LPS. Thus, P. aeruginosa vesicles are potentstimulators of macrophages that lead to the upregulated se-cretion of MIP-2, TNF-�, and IL-6.

CpG DNA does not significantly contribute to the inflam-matory response to vesicles. Outer membrane vesicles fromPseudomonas have been observed to contain bacterial DNA,which is a ligand for endosomal TLR9 (43). To investigate thecontribution of CpG DNA in the response to vesicles, TLR9signaling was inhibited by pretreatment of RAW 264.7 cellswith an inhibitory oligonucleotide (iCpG), which is a compet-itive inhibitor of CpG binding to TLR9. As seen in Fig. 2,inhibition of TLR9 signaling with iCpG had no effect on cyto-kine responses to S470 vesicles. No significant change in thecytokine transcriptional profiles was seen at any of the vesicledoses tested, indicating that CpG DNA is not a significanttrigger of inflammatory responses to vesicles.

Flagellin stimulates MIP-2 and IL-6 secretion in responseto outer membrane vesicles. Since flagellin is present in theprotein profiles of P. aeruginosa outer membrane vesicles andis a potent innate immune stimulator via both TLR5 and theNOD-like receptor IPAF (3), we investigated the contributionof flagellin to the cytokine response to vesicles using vesiclesfrom strain PAO1 and its isogenic flagellin mutant (�fliC). Wequantified the flagellin bands from SDS-PAGE profiles of bothPAO1 and S470 derived vesicles using densitometric analysisand demonstrated that these preparations contained equiva-lent levels of flagellin (data not shown). We could not useRAW 264.7 cells for these experiments because these cells donot express TLR5 (53). Therefore, we used the MH-S alveolarmacrophage cell line, which is known to express TLR5. Wenoted first that the MH-S cell line exhibited much lower sen-sitivity to vesicles than did RAW 264.7 cells: the lowest vesicledose (0.001 �g of LPS content) elicited only modest transcrip-tional increases for any of the genes assayed, these increasesdid not result in measurable cytokine levels by ELISA and,unlike RAW 264.7 cells, the MH-S cells did not secrete TNF-�in response to any of the vesicle doses (Fig. 3 and data notshown). However, ELISA analysis indicated that vesicles didinduce both MIP-2 and IL-6 secretion and that these weresignificantly lower in response to �fliC vesicles across multipledoses. Thus, the flagellin component of vesicles contributes

TABLE 1. RT-PCR primers used in these studies

Gene assayedPrimer sequence (5�–3�)

Forward Reverse

MIP-2 TCCATGAAAGCCATCCGACTGCAT ACATCCCACCCACACAGTGAAAGATNF-� ACCACTCTCCCTTTGCAGAACTCA TCTCATGCACCACCATCAAGGACTIL-1� TGCTTGTGAGGTGCTGATGTACCA TGGAGAGTGTGGATCCCAAGCAATIL-6 AACGCACTAGGTTTGCCGAGTAGA TGGCTAAGGACCAAGACCATCCAAKC AGTGTTGTCAGAAGCCAGCGTTCA ACCCAAACCGAAGTCATAGCCACAiNOS ATGTCATGAGCAAAGGCGCAGAAC CTGCTGGTGGTGACAAGCACATTTActin AGAGGGAAATCGTGCGTGAC CAATAGTGATGACCTGGCCGT

3824 ELLIS ET AL. INFECT. IMMUN.

on June 8, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

significantly to the stimulation of MIP-2 and IL-6 secretion bymacrophages.

Strain-specific proteins impact the inflammatory responsesto vesicles. We previously demonstrated that the protein com-position of vesicles varied with bacterial strain and can affecttheir association with host cells (2, 3). Therefore, we investi-gated whether the macrophage responses would also differ forvesicles from different strains of P. aeruginosa. A proteomiccomparison of vesicles from the clinical isolate S470 and labstrain PAO1 demonstrated some significant differences in theirprotein composition (3). The ratios of LPS to protein contentfor these two vesicle preparations were highly similar (data notshown), so any observed differences in the responses wouldlikely be the result of protein composition rather than LPSconcentration.

Cytokine upregulation and secretion were assayed in re-sponse to vesicles from S470 and PAO1. As shown in Fig. 4Asignificant differences in gene expression by real-time PCRanalysis were detected for the lowest dose (0.001 �g of LPScontent). At this dose, vesicles from strain S470 upregulatedsignificantly higher levels of all three cytokines assayed.

Greater differences were detected when cytokine levels weremeasured by ELISA (Fig. 4B). MIP-2 secretion was signifi-cantly higher in response to PAO1 vesicles, whereas the secre-tion of both TNF-� and IL-6 was greater in response to S470vesicles. TNF-� secretion above the baseline of untreated cellswas not detected at the lowest dose for either vesicle treat-ment, nor were significant IL-6 levels detected in response tothe lowest dose of PAO1 vesicles. These data indicate thatstrain-specific vesicle components affect both the nature andthe sensitivity of the macrophage proinflammatory cytokineresponse.

LPS and vesicle proteins act synergistically to generate adistinct inflammatory response. Two different biochemicaltreatments were used to investigate the contribution of partic-ular vesicle components in more detail. First, vesicles werepretreated with polymyxin B to neutralize the contribution ofLPS. Polymyxin B binds to the lipid A portion of LPS, inhib-iting sensing by the TLR4 complex (27). Second, vesicles wereincubated overnight with proteinase K to degrade vesicle pro-teins. Figure 5A shows the effect of proteinase K treatment onthe vesicle proteins. The protein profile of the vesicles was

FIG. 1. The inflammatory response of macrophages to outer membrane vesicles is more potent than to pure LPS. RAW 264.7 mousemacrophages were incubated with vesicles or purified P. aeruginosa LPS for 3 h. S470 vesicles and LPS doses were created from 10-fold serialdilutions of stocks containing 1 �g of LPS. (A) Gene expression was assayed by real-time PCR. (B) Protein secretion was assayed from culturesupernatants by ELISA. The bars represent the standard error of the mean (SEM) (n � 4). *, P 0.05; **, P 0.01.

VOL. 78, 2010 INFLAMMATORY RESPONSES TO OUTER MEMBRANE VESICLES 3825

on June 8, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

significantly altered, with many high-molecular-weight bandsdisappearing after protease treatment. All proteins were notfully degraded, as they were likely to be protected from pro-tease treatment by the vesicular membrane.

Macrophages were incubated with either of these treatedvesicle preparations, and cytokine expression was assayed byreal-time PCR and ELISA. As seen in Fig. 5B and C, poly-myxin B treatment of vesicles dramatically depressed bothcytokine transcriptional and protein responses of all genesassayed. The significant inhibition of cytokine transcription bypolymyxin B treatment was surprising given the higher potencyof whole vesicles compared to LPS (Fig. 1B). Polymyxin B-treated LPS, used as a control for these experiments, gener-ated minimal transcriptional responses by the macrophages(data not shown). These results indicate that LPS acts as anadjuvant in the reaction of macrophages toward vesicles,greatly enhancing the response to the non-LPS components ofvesicles.

In comparison to polymyxin B, proteinase K treatment ofvesicles resulted in a more selective dampening of cytokineresponses. Proteinase K treatment depressed IL-6 productionmost significantly for all tested doses of vesicles, as measuredby both reverse transcription-PCR and ELISA (Fig. 5B and C),indicating the likelihood that the production of this cytokine istriggered in response to outer membrane protein recognition.However, for several cytokines, induction levels increased withprotease treatment of the vesicles. Secreted MIP-2 levels werehigher than that seen in response to untreated vesicles (Fig.5C). TNF-� secretion was most dramatically impacted by pro-teinase K treatment at the lowest vesicle dose (0.001 �g of LPScontent), whereas significantly higher TNF-� secretion wasseen compared to the response to either intact vesicles or pureLPS. Together, these data suggest that vesicles interact withmacrophages in a manner that triggers a distinct, strain-specificresponse that reflects recognition of a combination of vesicleconstituents.

Sensing of both LPS and vesicle proteins is required forinternalization by macrophages. Based on their activation ofdistinct cytokine responses, we predicted that LPS, vesicles,and treated vesicles might interact differently with macro-phages. We developed a confocal microscopy assay to comparehow our preparations of vesicles associated with macrophages.MH-S alveolar macrophages were treated with untreated, pro-teinase K-treated, or polymyxin B-treated, Alexa Fluor 555-labeled S470 vesicles for 15 min. Vesicles were fluorescentlylabeled after proteinase K treatment, whereas polymyxin Btreatment was performed on fluorescently labeled vesicles. Af-ter vesicle coincubation, the cells were washed to remove un-bound material, surface biotinylated, and incubated with AlexaFluor 633-streptavidin.

The different LPS-containing reagents yielded substantiallydistinct localization profiles (Fig. 6), giving insight into themolecular basis for the distinct pattern of activation of themacrophages. Untreated S470 vesicles appeared as punctateparticles that were not labeled by streptavidin and thus werelikely to have been internalized by the macrophages. PolymyxinB treatment resulted in few vesicles interacting with macro-phages. The few vesicles that were observed were not labeledby streptavidin and thus were internalized by the macrophages.Proteinase K-treated vesicles adhered to the surfaces of mac-rophages and were labeled by streptavidin and thus not inter-nalized. These data indicate that vesicle LPS is required forinitial surface binding of vesicles to the macrophage surface,whereas recognition of vesicle protein triggers internalization.

DISCUSSION

The goal of the present study was to characterize both theinflammatory response to P. aeruginosa outer membrane ves-icles and the molecular components key to this response.

Outer membrane vesicles triggered a distinct, strain-specificpattern of cytokine production in macrophages compared to

FIG. 2. CpG DNA is not a major contributor to the inflammatory response to outer membrane vesicles. The contribution of CpG DNA wasdetermined by preincubating RAW 264.7 cells with CpG inhibitory oligodeoxynucleotides prior to the addition of vesicles. Gene expression wasassayed by real-time PCR normalized to actin production in each sample. The bars represent the SEM (n � 4). Statistical analysis revealed nosignificant differences between the pairs of treatments for each gene product.

3826 ELLIS ET AL. INFECT. IMMUN.

on June 8, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

that of pure LPS. LPS and flagellin both played roles in trig-gering cytokine production, whereas CpG DNA did not. Pro-teinase K digested vesicles bound to the surface of macro-phages without internalization and exhibited reduced levels ofIL-6. Further, the response to surface proteins depended onthe presence of LPS, since a lack of LPS sensing abrogated themajority of the response.

It was anticipated that the macrophage response to vesicleswould closely resemble that to pure LPS; however, this wasclearly not the case. The higher potency of vesicles to triggercytokine production was particularly evident at low vesicledoses. A titration of vesicle dose was similarly used by Alanizet al. (1) to study the ability of vesicles from Salmonella toactivate dendritic cells. Our study supports their conclusionthat vesicles are potent stimulators of proinflammatory cyto-kines. However, since Alaniz et al. standardized their vesicleand LPS doses to total dry weight rather than LPS content,these responses cannot be compared directly in terms of therelative contribution of LPS to this response. Therefore, whileboth studies clearly demonstrate the potency of vesicles to

trigger inflammatory responses, our study defines the contri-bution of LPS to the macrophage response and demonstratesthat the context of heterogeneous components present in theouter membrane is critical to the generation of a robust re-sponse.

This study supports the concept that the three-dimensionalconfiguration in which vesicle ligands (including LPS) are pre-sented to macrophages impacts both the type and sensitivity oftheir response. Beyond acting as stimuli, vesicle proteins ex-hibit the ability to block or hinder vesicle interactions withsurface receptors such as TLR4 or MD2. The pattern ofTNF-� secretion observed in our studies illustrate this abil-ity, since TNF-� secretion was enhanced by protease treat-ment (Fig. 5B and C), whereas protein-intact vesicles re-quired higher doses to trigger similar levels of TNF-�production (Fig. 1B).

Several other studies have further suggested that the spatialorientation of LPS impacts its interactions with host cells.Analysis of CD14-dependent internalization of LPS has shownthat aggregate size of LPS complexes may directly affect the

FIG. 3. Flagellin contributes to the MIP-2 and IL-6 responses to outer membrane vesicles. The contribution of flagellin was determined usingMH-S mouse macrophage cells incubated with vesicles from the PAO1 or PAO1�fliC strains. (A) Fold changes in gene expression were assessedby real-time PCR normalized to actin production. (B) Protein secretion was assayed from culture supernatants by ELISA. The bars represent theSEM (n � 4). *, P 0.05; **, P 0.01.

VOL. 78, 2010 INFLAMMATORY RESPONSES TO OUTER MEMBRANE VESICLES 3827

on June 8, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

rate of recognition and uptake (30). Further, Elsbach con-cluded that the binding of and responses to LPS by inflamma-tory cells critically depend on the context and presence of otherbacterial compounds (such as proteins) in which the LPS waspresented (16).

The dramatic effect proteinase K treatment had on the re-sponse to vesicles demonstrates that the protein content ofvesicles include key ligands of significant cytokine responses.Flagellin monomers, ligands for both TLR5 and NOD-likereceptors, have been detected in purified P. aeruginosa vesicles(3, 36). Transcriptional upregulation of IL-1� was detected inresponse to intact vesicles; however, no mature protein wassecreted, indicating that modification by caspase-1 did not oc-cur. Since previous studies demonstrated that LPS increasestranscription of the IL-1� gene without active secretion of theprotein (54), we propose a similar effect occurs for vesicles.

We found that vesicle proteins other than flagellin contrib-ute to the IL-6 response via recognition within the macro-phage. The expression profiles of IL-6 secretion to protease-treated vesicles and flagellin-deficient vesicles differedsignificantly and cannot only be explained by protease degra-dation of surface-associated flagellin. In addition, the re-

sponses to S470 and PAO1 vesicles differed, whereas the re-sponses to purified LPS from both strains did not (data notshown). Macrophage-bound vesicles, inhibited from enteringby cytochalasin D, trigger MIP-2 and TNF-� production butnot IL-6, confirming the role of vesicle proteins and internal-ization in triggering an IL-6 response (data not shown). MIP-2and TNF-� can be stimulated by binding of outer surfacereceptors such as TLR4 and TLR5.

Each cytokine analyzed exhibited a distinct pattern of re-sponses to vesicles. MIP-2, a neutrophil chemoattractant, wassecreted in a pattern that most closely matched TLR triggeredresponses to purified LPS. Diminished production of MIP-2 bydeletion of flagellin further indicates that surface TLR recep-tors are engaged to stimulate production (6). Protease treat-ment actually enhanced MIP-2 secretion, indicating that vesi-cle proteins may physically block TLR receptor engagement ofLPS. In contrast, IL-6 production appears to be predominantlya response to vesicle proteins, since it dramatically decreasedfollowing either the deletion of flagellin or general proteasetreatment. The observed TNF-� response to vesicles was in-termediate to those of MIP-2 and IL-6.

Analysis of the cytokine promoter regions indicates that the

FIG. 4. The cytokine responses to outer membrane vesicles are bacterial strain dependent. RAW 264.7 mouse macrophage cells were incubatedwith vesicles from clinical isolate S470 or lab strain PAO1 for 3 h. Gene expression was assayed by real-time PCR (A), and protein secretion wasassayed from culture supernatants by ELISA (B). The bars represent the SEM (n � 4). *, P 0.05; **, P 0.01.

3828 ELLIS ET AL. INFECT. IMMUN.

on June 8, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

MIP-2 response to vesicles may be triggered primarily by NF-B-dependent pathways, since the MIP-2 promoter only con-tains binding sites for AP-1 and NF-B (29). In contrast, theTNF-� and IL-6 promoters contain a larger variety of regula-tory binding sites. The presence of sites responsive to prosta-glandins or cyclic AMP, in addition to AP-1 and NF-B sites(13, 46), results in sensitivity to other ligands such as vesicleproteins.

The data presented here indicate that vesicle LPS acts as anadjuvant for other components. This effect was most dramat-ically demonstrated by the lack of response to vesicle proteinsin polymyxin B treatment experiments. Vesicle localizationfurther support the key contributions of vesicle proteins tointernalization and the generation of a robust cytokine re-sponse. Thus, vesicle proteins are particularly effective stimuliof cytokine production when recognized in the context of otherPAMPs such as LPS.

Recent mechanistic studies have provided support for theadjuvanticity of TLR ligands. Sensing of TLR ligands directlyimpacts phagosomal maturation, selection of antigens forMHC presentation, and maturation of dendritic cells (8, 9).

Further, the presence of LPS during an initial antigen exposurechanged the mechanism and magnitude of T-cell traffickingupon subsequent LPS-free antigen challenge, as well as ampli-fied the phagocytic inflammatory response (37). These studiessuggest that the response to the LPS-protein mixture in vesi-cles may translate to protective downstream T-cell and mem-ory responses. Further, LPS or other TLR ligands greatly en-hanced protective responses to N. meningitis Vesicle-basedvaccines (15, 19). Our demonstration that low doses of LPSwithin a vesicle context augment protein specific responsescould be used to refine efforts to engineer outer membranevesicles into vaccines.

In summary, the data presented here demonstrate that natu-rally secreted outer membrane vesicles of P. aeruginosa are potentstimulators of strain-specific proinflammatory responses. Macro-phage responses to a complex of heterogeneous PAMPs are dis-tinct from the responses to the individual components, requirethe presence of LPS for maximal potency, and trigger intenseresponses via combined sensing of multiple pathogen-specificstimuli. Future studies will focus on determining which outermembrane proteins are critical to the observed responses.

FIG. 5. LPS and vesicle proteins act synergistically to generate a distinct inflammatory response. The ability of the proteins present on vesiclesto trigger a host response was determined by treating vesicles with proteinase K or polymyxin B. (A) SDS-PAGE of S470 outer membrane vesiclesprotein profiles before and after proteinase K digestion. Macrophages were exposed to treated vesicles for 3 h and responses were assayed byreal-time PCR (B) or ELISA (C). The bars represent the SEM (n � 4). *, P 0.05; **, P 0.01.

VOL. 78, 2010 INFLAMMATORY RESPONSES TO OUTER MEMBRANE VESICLES 3829

on June 8, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

ACKNOWLEDGMENTS

This research was supported by a Burroughs Wellcome Fund Inves-tigator in Pathogenesis of Infectious Disease Award, an NIEHS-sup-ported Duke Center for Comparative Biology of Vulnerable Popula-tions Pilot Project, and an American Lung Association CareerInvestigator Award (to M.J.K.) and a postdoctoral fellowship from TheHartwell Foundation (to T.N.E.).

We thank Susanne Bauman for the pure LPS preparations.

REFERENCES

1. Alaniz, R. C., B. L. Deatherage, J. C. Lara, and B. T. Cookson. 2007.Membrane vesicles are immunogenic facsimiles of Salmonella typhimuriumthat potently activate dendritic cells, prime B- and T-cell responses, andstimulate protective immunity in vivo. J. Immunol. 179:7692–7701.

2. Bauman, S. J., and M. J. Kuehn. 2009. Pseudomonas aeruginosa vesiclesassociate with and are internalized by human lung epithelial cells. BMCMicrobiol. 9:26.

3. Bauman, S. J., and M. J. Kuehn. 2006. Purification of outer membranevesicles from Pseudomonas aeruginosa and their activation of an IL-8 re-sponse. Microbes Infect. 8:2400–2408.

4. Bennett-Guerrero, E., T. J. McIntosh, G. R. Barclay, D. S. Snyder, R. J.Gibbs, M. G. Mythen, and I. R. Poxton. 2000. Preparation and preclinicalevaluation of a novel liposomal complete-core lipopolysaccharide vaccine.Infect. Immun. 68:6202–6208.

5. Berclaz, P. Y., B. Carey, M. D. Fillipi, K. Wernke-Dollries, N. Geraci, S.Cush, T. Richardson, J. Kitzmiller, M. O’Connor, C. Hermoyian, T. Korf-hagen, J. A. Whitsett, and B. C. Trapnell. 2007. GM-CSF regulates a PU.1-dependent transcriptional program determining the pulmonary response toLPS. Am. J. Respir. Cell Mol. Biol. 36:114–121.

6. Beutler, B., K. Hoebe, X. Du, and R. J. Ulevitch. 2003. How we detectmicrobes and respond to them: the Toll-like receptors and their transducers.J. Leukoc. Biol. 74:479–485.

7. Beveridge, T. J. 1999. Structures of gram-negative cell walls and their derivedmembrane vesicles. J. Bacteriol. 181:4725–4733.

8. Blander, J. M., and R. Medzhitov. 2006. On regulation of phagosome mat-uration and antigen presentation. Nat. Immunol. 7:1029–1035.

9. Blander, J. M., and R. Medzhitov. 2006. Toll-dependent selection of micro-bial antigens for presentation by dendritic cells. Nature 440:808–812.

10. Bomberger, J. M., D. P. Maceachran, B. A. Coutermarsh, S. Ye, G. A.O’Toole, and B. A. Stanton. 2009. Long-distance delivery of bacterial viru-lence factors by Pseudomonas aeruginosa outer membrane vesicles. PLoSPathog. 5:e1000382.

11. Borgatti, M., V. Bezzerri, I. Mancini, E. Nicolis, M. C. Dechecchi, I. Lam-pronti, P. Rizzotti, G. Cabrini, and R. Gambari. 2007. Induction of IL-6 geneexpression in a CF bronchial epithelial cell line by Pseudomonas aeruginosais dependent on transcription factors belonging to the Sp1 superfamily.Biochem. Biophys. Res. Commun. 357:977–983.

12. Ciofu, O., T. J. Beveridge, J. Kadurugamuwa, J. Walther-Rasmussen, and N.Hoiby. 2000. Chromosomal beta-lactamase is packaged into membranevesicles and secreted from Pseudomonas aeruginosa. J. Antimicrob. Che-mother. 45:9–13.

13. Dendorfer, U., P. Oettgen, and T. A. Libermann. 1994. Multiple regulatoryelements in the interleukin-6 gene mediate induction by prostaglandins,cyclic AMP, and lipopolysaccharide. Mol. Cell. Biol. 14:4443–4454.

14. Driscoll, J. A., S. L. Brody, and M. H. Kollef. 2007. The epidemiology,pathogenesis, and treatment of Pseudomonas aeruginosa infections. Drugs67:351–368.

15. Durand, V., J. Mackenzie, J. de Leon, C. Mesa, V. Quesniaux, M. Montoya,A. Le Bon, and S. Y. Wong. 2009. Role of lipopolysaccharide in the inductionof type I interferon-dependent cross-priming and IL-10 production in miceby meningococcal outer membrane vesicles. Vaccine 27:1912–1922.

16. Elsbach, P. 2000. Mechanisms of disposal of bacterial lipopolysaccharides byanimal hosts. Microbes Infect. 2:1171–1180.

17. Feiring, B., J. Fuglesang, P. Oster, L. M. Naess, O. S. Helland, S. Tilman, E.Rosenqvist, M. A. Bergsaker, H. Nokleby, and I. S. Aaberge. 2006. Persistingimmune responses indicating long-term protection after booster dose withmeningococcal group B outer membrane vesicle vaccine. Clin. Vaccine Im-munol. 13:790–796.

18. Findlow, J., S. Taylor, A. Aase, R. Horton, R. Heyderman, J. Southern, N.Andrews, R. Barchha, E. Harrison, A. Lowe, E. Boxer, C. Heaton, P. Balmer,E. Kaczmarski, P. Oster, A. Gorringe, R. Borrow, and E. Miller. 2006.

FIG. 6. Macrophage recognition of outer membrane vesicle proteins is required for internalization by macrophages. MH-S alveolar macro-phages were exposed to native, proteinase K-treated, or polymyxin B-treated Alexa Fluor 555-labeled vesicles for 15 min. At the indicated timepoints, the unbound label was washed off, and the cells were surface labeled with biotin and Alexa Fluor 633-streptavidin (blue) and visualized byconfocal microscopy.

3830 ELLIS ET AL. INFECT. IMMUN.

on June 8, 2020 by guesthttp://iai.asm

.org/D

ownloaded from

Comparison and correlation of neisseria meningitidis serogroup B immuno-logic assay results and human antibody responses following three doses ofthe Norwegian meningococcal outer membrane vesicle vaccine MenBvac.Infect. Immun. 74:4557–4565.

19. Fransen, F., C. J. Boog, J. P. van Putten, and P. van der Ley. 2007. Agonistsof Toll-like receptors 3, 4, 7, and 9 are candidates for use as adjuvants in anouter membrane vaccine against Neisseria meningitidis serogroup B. Infect.Immun. 75:5939–5946.

20. Freudenberg, M. A., S. Tchaptchet, S. Keck, G. Fejer, M. Huber, N. Schutze,B. Beutler, and C. Galanos. 2008. Lipopolysaccharide sensing an importantfactor in the innate immune response to Gram-negative bacterial infections:benefits and hazards of LPS hypersensitivity. Immunobiology 213:193–203.

21. Gosselin, D., J. DeSanctis, M. Boule, E. Skamene, C. Matouk, and D.Radzioch. 1995. Role of tumor necrosis factor alpha in innate resistance tomouse pulmonary infection with Pseudomonas aeruginosa. Infect. Immun.63:3272–3278.

22. Hajjar, A. M., H. Harowicz, H. D. Liggitt, P. J. Fink, C. B. Wilson, and S. J.Skerrett. 2005. An essential role for non-bone marrow-derived cells in con-trol of Pseudomonas aeruginosa pneumonia. Am. J. Respir. Cell Mol. Biol.33:470–475.

23. Heczko, U., V. C. Smith, R. Mark Meloche, A. M. Buchan, and B. B. Finlay.2000. Characteristics of Helicobacter pylori attachment to human primaryantral epithelial cells. Microbes Infect. 2:1669–1676.

24. Ismail, S., M. B. Hampton, and J. I. Keenan. 2003. Helicobacter pylori outermembrane vesicles modulate proliferation and interleukin-8 production bygastric epithelial cells. Infect. Immun. 71:5670–5675.

25. Kadurugamuwa, J. L., and T. J. Beveridge. 1997. Natural release of virulencefactors in membrane vesicles by Pseudomonas aeruginosa and the effect ofaminoglycoside antibiotics on their release. J. Antimicrob. Chemother. 40:615–621.

26. Kadurugamuwa, J. L., and T. J. Beveridge. 1995. Virulence factors arereleased from Pseudomonas aeruginosa in association with membrane vesi-cles during normal growth and exposure to gentamicin: a novel mechanismof enzyme secretion. J. Bacteriol. 177:3998–4008.

27. Katz, M., H. Tsubery, S. Kolusheva, A. Shames, M. Fridkin, and R. Jelinek.2003. Lipid binding and membrane penetration of polymyxin B derivativesstudied in a biomimetic vesicle system. Biochem. J. 375:405–413.

28. Keenan, J., T. Day, S. Neal, B. Cook, G. Perez-Perez, R. Allardyce, and P.Bagshaw. 2000. A role for the bacterial outer membrane in the pathogenesisof Helicobacter pylori infection. FEMS Microbiol. Lett. 182:259–264.

29. Kim, D. S., J. H. Han, and H. J. Kwon. 2003. NF-B and c-Jun-dependentregulation of macrophage inflammatory protein-2 gene expression in re-sponse to lipopolysaccharide in RAW 264.7 cells. Mol. Immunol. 40:633–643.

30. Kitchens, R. L., and R. S. Munford. 1998. CD14-dependent internalizationof bacterial lipopolysaccharide (LPS) is strongly influenced by LPS aggrega-tion but not by cellular responses to LPS. J. Immunol. 160:1920–1928.

31. Kuehn, M. J., and N. C. Kesty. 2005. Bacterial outer membrane vesicles andthe host-pathogen interaction. Genes Dev. 19:2645–2655.

32. Livak, K. J., and T. D. Schmittgen. 2001. Analysis of relative gene expressiondata using real-time quantitative PCR and the 2���CT method. Methods25:402–408.

33. Maceachran, D. P., S. Ye, J. M. Bomberger, D. A. Hogan, A. Swiatecka-Urban, B. A. Stanton, and A. O’Toole, G. 2007. The Pseudomonas aerugi-nosa secreted protein PA2934 decreases apical membrane expression ofthe cystic fibrosis transmembrane conductance regulator. Infect. Immun.75:3902–3912.

34. Mashburn, L. M., and M. Whiteley. 2005. Membrane vesicles traffic signalsand facilitate group activities in a prokaryote. Nature 437:422–425.

35. McBroom, A. J., and M. J. Kuehn. 2007. Release of outer membrane vesiclesby Gram-negative bacteria is a novel envelope stress response. Mol. Micro-biol. 63:545–558.

36. Miao, E. A., E. Andersen-Nissen, S. E. Warren, and A. Aderem. 2007. TLR5

and Ipaf: dual sensors of bacterial flagellin in the innate immune system.Semin. Immunopathol. 29:275–288.

37. Mikhak, Z., A. Farsidjani, and A. D. Luster. 2009. Endotoxin augmentedantigen-induced Th1 cell trafficking amplifies airway neutrophilic inflamma-tion. J. Immunol. 182:7946–7956.

38. Norheim, G., E. Arne Hoiby, D. A. Caugant, E. Namork, T. Tangen, E.Fritzsonn, and E. Rosenqvist. 2004. Immunogenicity and bactericidal activityin mice of an outer membrane protein vesicle vaccine against Neisseriameningitidis serogroup A disease. Vaccine 22:2171–2180.

39. Opal, S. M. 2007. The host response to endotoxin, antilipopolysaccharidestrategies, and the management of severe sepsis. Int. J. Med. Microbiol.297:365–377.

40. Perrett, K. P., and A. J. Pollard. 2005. Towards an improved serogroup BNeisseria meningitidis vaccine. Expert Opin. Biol. Ther. 5:1611–1625.

41. Power, M. R., B. Li, M. Yamamoto, S. Akira, and T. J. Lin. 2007. A role ofToll-IL-1 receptor domain-containing adaptor-inducing IFN-beta in the hostresponse to Pseudomonas aeruginosa lung infection in mice. J. Immunol.178:3170–3176.

42. Reiniger, N., M. M. Lee, F. T. Coleman, C. Ray, D. E. Golan, and G. B. Pier.2007. Resistance to Pseudomonas aeruginosa chronic lung infection requirescystic fibrosis transmembrane conductance regulator-modulated interleu-kin-1 (IL-1) release and signaling through the IL-1 receptor. Infect. Immun.75:1598–1608.

43. Renelli, M., V. Matias, R. Y. Lo, and T. J. Beveridge. 2004. DNA-containingmembrane vesicles of Pseudomonas aeruginosa PAO1 and their genetictransformation potential. Microbiology 150:2161–2169.

44. Sadikot, R. T., T. S. Blackwell, J. W. Christman, and A. S. Prince. 2005.Pathogen-host interactions in Pseudomonas aeruginosa pneumonia. Am. J.Respir. Crit. Care Med. 171:1209–1223.

45. Sadikot, R. T., H. Zeng, M. Joo, M. B. Everhart, T. P. Sherrill, B. Li, D. S.Cheng, F. E. Yull, J. W. Christman, and T. S. Blackwell. 2006. Targetedimmunomodulation of the NF-B pathway in airway epithelium impacts hostdefense against Pseudomonas aeruginosa. J. Immunol. 176:4923–4930.

46. Sandoval, J., J. Pereda, J. L. Rodriguez, J. Escobar, J. Hidalgo, L. A.Joosten, L. Franco, J. Sastre, and G. Lopez-Rodas. 2010. Ordered transcrip-tional factor recruitment and epigenetic regulation of TNF-alpha in necro-tizing acute pancreatitis. Cell Mol. Life Sci. 67:1687–1697.

47. Schild, S., E. J. Nelson, A. L. Bishop, and A. Camilli. 2009. Characterizationof Vibrio cholerae outer membrane vesicles as a candidate vaccine for chol-era. Infect. Immun. 77:472–484.

48. Schnare, M., M. Rollinghoff, and S. Qureshi. 2006. Toll-like receptors:sentinels of host defence against bacterial infection. Int. Arch. Allergy Im-munol. 139:75–85.

49. Schooling, S. R., and T. J. Beveridge. 2006. Membrane vesicles: an over-looked component of the matrices of biofilms. J. Bacteriol. 188:5945–5957.

50. Talati, A. J., H. J. Kim, Y. I. Kim, A. K. Yi, and B. K. English. 2008. Role ofbacterial DNA in macrophage activation by group B streptococci. MicrobesInfect. 10:1106–1113.

51. Tsai, W. C., R. M. Strieter, B. Mehrad, M. W. Newstead, X. Zeng, and T. J.Standiford. 2000. CXC chemokine receptor CXCR2 is essential for protec-tive innate host response in murine Pseudomonas aeruginosa pneumonia.Infect. Immun. 68:4289–4296.

52. Tseng, J., J. Do, J. H. Widdicombe, and T. E. Machen. 2006. Innate immuneresponses of human tracheal epithelium to Pseudomonas aeruginosa flagellin,TNF-�, and IL-1�. Am. J. Physiol. Cell Physiol. 290:C678–C690.

53. West, A. P., B. A. Dancho, and S. B. Mizel. 2005. Gangliosides inhibitflagellin signaling in the absence of an effect on flagellin binding to Toll-likereceptor 5. J. Biol. Chem. 280:9482–9488.

54. Wewers, M. D. 2004. IL-1�: an endosomal exit. Proc. Natl. Acad. Sci. U. S. A.101:10241–10242.

55. Zaidi, T. S., M. J. Preston, and G. B. Pier. 1997. Inhibition of bacterialadherence to host tissue does not markedly affect disease in the murinemodel of Pseudomonas aeruginosa corneal infection. Infect. Immun. 65:1370–1376.

Editor: B. A. McCormick

VOL. 78, 2010 INFLAMMATORY RESPONSES TO OUTER MEMBRANE VESICLES 3831

on June 8, 2020 by guesthttp://iai.asm

.org/D

ownloaded from