01 J Immunol Bouwman2014 Campylobacter

of September 29, 2014.This information is current as

jejuni CampylobacterInflammasome Activation by

PuttenBleumink-Pluym, Richard A. Flavell and Jos P. M. van Lieneke I. Bouwman, Marcel R. de Zoete, Nancy M. C.

ol.1400648http://www.jimmunol.org/content/early/2014/09/28/jimmun

published online 29 September 2014J Immunol

MaterialSupplementary

DCSupplemental.htmlhttp://www.jimmunol.org/content/suppl/2014/09/28/content.1400648.

Subscriptionshttp://jimmunol.org/subscriptions

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/ji/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/cgi/alerts/etocReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists, Inc. All rights reserved.Copyright 2014 by The American Association of9650 Rockville Pike, Bethesda, MD 20814-3994.The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

at University of Colorado H

ealth Sciences Library on September 29, 2014

http://ww

w.jimmunol.org/

Dow

nloaded from



http://ww

w.jimmunol.org/

Dow

nloaded from

The Journal of Immunology

Inflammasome Activation by Campylobacter jejuni

Lieneke I. Bouwman,* Marcel R. de Zoete,, Nancy M. C. Bleumink-Pluym,*

Richard A. Flavell,, and Jos P. M. van Putten*

The Gram-negative pathogen Campylobacter jejuni is the most common cause of bacterial foodborne disease worldwide. The

mechanisms that lead to bacterial invasion of eukaryotic cells and massive intestinal inflammation are still unknown. In this study,

we report that C. jejuni infection of mouse macrophages induces upregulation of proIL-1b transcript and secretion of IL-1b

without eliciting cell death. Immunoblotting indicated cleavage of caspase-1 and IL-1b in infected cells. In bone marrowderived

macrophages from different knockout mice, IL-1b secretion was found to require NLRP3, ASC, and caspase-1/11 but not NLRC4.

In contrast to NLRP3 activation by ATP, C. jejuni activation did not require priming of these macrophages. C. jejuni also activated

the NLRP3 inflammasome in human macrophages as indicated by the presence of ASC foci and caspase-1positive cells. Analysis

of a vast array of C. jejuni mutants with defects in capsule formation, LPS biosynthesis, chemotaxis, flagella synthesis and flagellin

(-like) secretion, type 6 secretion system needle protein, or cytolethal distending toxin revealed a direct correlation between the

number of intracellular bacteria and NLRP3 inflammasome activation. The C. jejuni invasionrelated activation of the NLRP3

inflammasome without cytotoxicity and even in nonprimed cells extends the known repertoire of bacterial inflammasome acti-

vation and likely contributes to C. jejuniinduced intestinal inflammation. The Journal of Immunology, 2014, 193: 000000.

Inflammasomes are multiprotein complexes that form in thecytosol following sensing of intracellular threats like in-truding bacteria and viruses or cell damage. Inflammasome

complexes generally consist of sensor proteins (members of theNod-like receptor [NLR] or Pyrin and HIN200 domain [PYHIN]protein family) and effector procaspases (mainly caspase-1),which are bridged by the adaptor protein apoptosis-associatedspeck-like protein (ASC). After assembly, inflammasomes in-duce the activation of caspase-1 through autocleavage, whichsubsequently activates cytokines IL-1b and IL-18, and inducea form of cell death referred to as pyroptosis (15).The best studied inflammasomes in relation to bacterial infection

are the NLR family, CARD domaincontaining 4 (NLRC4) and theNLR family, pyrin domaincontaining 3 (NLRP3) inflammasomes(6, 7). The NLRC4 inflammasome is formed after sensing cyto-solic bacterial flagellin or components of the bacterial type IIIsecretion system (T3SS) by distinct neuronal apoptosis inhibitoryprotein (NAIP) receptors in the cell (810). NLRP3 inflamma-some activation requires a two signal process. The first (priming)

signal leads to the expression of NLRP3 and proIL-1b throughactivation of NF-kB via stimulation of TLRs, other pattern rec-

ognition receptors, or endogenous cytokines. Then, a second

stimulus (e.g., pore-forming toxins, bacterial invasion, or uptake

of large particulates) induces the formation of the NLRP3

inflammasome (4). Although NLRP3 ligands are highly diverse,

they all seem to converse in the efflux of K+ from the cell, which is

proposed to be the common trigger (11, 12).The bacterial pathogen Campylobacter jejuni is the most

common cause of bacterial foodborne disease worldwide. Symp-

tomatic infection typically involves intestinal inflammation with

abdominal pain, fever, and (bloody) diarrhea. In 1% of the cases,serious complications may develop such as the acute autoimmune

paralyzing neuropathy GuillainBarre syndrome (13, 14). In

contrast to most enteropathogens including Salmonella, C. jejuni

lacks traditional virulence factors like T3SSs. The molecular

cause of C. jejuni intestinal inflammation is still largely unknown.

After ingestion, the bacteria travel deep down into the intestinal

crypts of the colon where they colonize and replicate. At some

point the epithelial barrier is breached, resulting in acute inflam-

mation accompanied by strong neutrophil recruitment and acti-

vation of T- and B-cell responses (15, 16). Bacterial motility and

chemotaxis are crucial for causing disease (1720). Other pro-

posed virulence traits include the polysaccharide capsule, secreted

proteins (Cia proteins, HtrA protease), type 6 secretion (T6SS)

effector molecules, apoptosis-inducing proteins (cytolethal dis-

tending toxin, FspA2), and bacterial adhesion and invasion pro-

moting factors (FlaC, PEB1, JlpA, CapA, and CadF) (for review,

see Refs. 2123). The role of these factors in the development of

human infection, however, remains to be demonstrated.The induction of acute intestinal inflammation in response to

C. jejuni infection suggests the activation of innate pattern rec-

ognition receptors (24, 25). Although C. jejuni flagellin and DNA

escape TLR recognition, the bacterial LPS and lipoproteins po-

tently activate the TLR4 and TLR2 pathways (25). In addition,

C. jejuni is internalized by monocytes and macrophages and

activates NOD1 (24, 2629). Cellular infection is accompanied by

the secretion of several proinflammatory cytokines such as IL-6,

*Department of Infectious Diseases and Immunology, Utrecht University, 3584 CLUtrecht, the Netherlands; Department of Immunobiology, Yale University School ofMedicine, New Haven, CT 06520; and Howard Hughes Medical Institute, YaleUniversity, New Haven, CT 06520

Received for publication March 11, 2014. Accepted for publication August 27, 2014.

This work was supported by a Rubicon fellowship from the Netherlands and theOrganization of Scientific Research (to M.R.d.Z.).

Address correspondence and reprint requests to Dr. Jos P.M. van Putten, Departmentof Infectious Diseases and Immunology, Utrecht University, Yalelaan 1, 3584 CLUtrecht, the Netherlands. E-mail address: [email protected]

The online version of this article contains supplemental material.

Abbreviations used in this article: ASC, apoptosis-associated speck-like protein;BMM, bone marrow macrophage; cat, chloramphenicol; CDT, cytolethal distendingtoxin; F.I., fluorescence intensity; FLICA, fluorescent labeled inhibitor of caspases;HI, heart infusion; kana, kanamycin; LB, LuriaBertani; LDH, lactate dehydroge-nase; LOS, lipooligosaccharide; m.o.i., multiplicity of infection; NAIP, neuronalapoptosis inhibitory protein; NLR, Nod-like receptor; NLRC4, NLR family, CARDdomaincontaining 4; NLRP3, NLR family, pyrin domaincontaining 3; PI, propi-dium iodide; T3SS, type 3 secretion system; T4SS, type 4 secretion system; T6SS,type 6 secretion system.

Copyright 2014 by The American Association of Immunologists, Inc. 0022-1767/14/$16.00

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1400648

Published September 29, 2014, doi:10.4049/jimmunol.1400648 at U

niversity of Colorado Health Sciences Library on Septem

ber 29, 2014http://w

ww

.jimmunol.org/D

ownloaded from

IL-8, TNF-a, and IL-1b (3032). Considering the potential im-portant role of IL-1b in the clinical manifestation of C. jejuniinfection (33), we investigated in the current study the ability ofC. jejuni to activate the inflammasome. Our results reveal thatC. jejuni can induce inflammasome activation without cytotoxic-ity, which has thus far not been observed for other pathogens.

Materials and MethodsCell culture and reagents

J774.A1 cells (ATCC TIB-67) were routinely cultured in DMEM plus 10%FCS at 37C and 10% CO2. THP-1 null and THP-1 defNLRP3 (InvivoGen;thp-null, thp-dnlp) were grown according to the manufacturers protocol inRPMI 1640 medium plus 10% FCS in the presence of 200 mg/mlHygrogold (every other passage) at 37C and 5% CO2. L929 cells werecultured in RPMI 1640 medium plus 10% FCS at 37C and 5% CO2.

The following reagents were used: gentamicin, kanamycin (kana),chloramphenicol (cat), Triton X-100, Tween 20, Tris, paraformaldehyde,PMA, TCA, and goat anti-rabbit IgG-HRP (A4914) (Sigma-Aldrich);FCS and Dulbeccos PBS (PAA); DMEM, RPMI 1640 medium, Opti-MEM, penicillin, and streptomycin (Life Technologies); BCA ProteinAssay Kit, Concentrators, 9K MWCO, and SuperSignal West FemtoChemiluminescent Substrate (Pierce); cOmplete ULTRA Tablets EDTA-Free, lactate dehydrogenase (LDH), and Cytotoxicity Detection KitPLUS

(LDH) (Roche); WGA-Alexa Fluor 633, goat anti-rabbit-Alexa Fluor 488,primers, Pfx DNA polymerase (Life Technologies); ATP, DNase,29-deoxynucleoside 59-triphosphates, BamHI, KpnI, SacI, SacII, PhusionDNA Polymerase, GeneJET Gel Extraction Kit, CloneJET PCR CloningKit, and Rapid DNA ligation kit (Thermo); Mouse IL-1b ELISA Ready-SET-Go and Human IL-1b ELISA Ready-SET-Go (eBioscience); BrilliantIII Ultra-Fast Sybr Green qRT-PCR kit (Agilent); reporter lysis buffer andLuciferase Assay Agent, pGEM-T easy (Promega); Saponin agar plates,Mueller Hinton plates, heart infusion (HI) plates, LuriaBertani (LB)plates, LB broth, and HI broth (Biotrading); Campylobacter selectivesupplement and charcoal cefoperazone desoxycholate agar (SR0155)(Oxoid); RNA Bee (Bio-connect); Fluorsave (Calbiochem); Qiaex II gelextraction kit (Qiagen); FAM fluorescent labeled inhibitor of caspases(FLICA) Caspase-1 Assay Kit (Immunochemistry); and rabbit anticaspase-1 (ab17820), rabbit anti-TMS1 (ASC) (ab64808), and rabbit antiIL-1b (ab9722) (Abcam).

Cultivation of primary mouse macrophages

Bone marrow cells were isolated as described previously (34). Uponthawing, cells were collected (5 min, 4853 g, 20C), resuspended in 10 mlbone marrowderived macrophage (BMM) medium (RPMI 1640 mediumplus 10% heat-inactivated FCS and 30% L929 conditioned medium) withpenicillin (100 IU) and streptomycin (100 mg/ml) and allowed to differ-entiate for 6 d into BMM at 37C and 5% CO2. After 3 d, an additional10 ml BMM medium was added to the cells. After differentiation cellswere collected, counted, and seeded into a 96-well (1 3 105 cells/well) or24-well (2.5 3 105 cells/well) plate in BMM medium, and used the nextday. L929 conditioned medium was collected from L929 cells grown in40 ml medium for 10 d in a T75 flask, filter sterilized (0.22-mm pore size),and stored at 220C until use.

Bacterial culture

All C. jejuni strains (Supplemental Table I) were routinely grown undermicroaerophilic conditions at 37C on saponin agar plates containing 4%lysed horse blood or in 5 ml HI broth at 160 rpm for 16 h. Kana (50 mg/ml)or cat (20 mg/ml) was added to the medium when appropriate. All C. jejunistrains had similar growth rates. Escherichia coli DH5a (Netherlands Cul-ture Collection of Bacteria) was grown on LB agar plates or in 5 ml LBbroth at 37C in air. Salmonella was grown on LB agar with ampicillin(100 mg/ml), or in HI medium for 2 h at 37C in air. C. jejuni lysate wasprepared as described previously (25). Shortly, bacteria were grown for 16 hunder standard conditions, collected by centrifugation (10 min, 3000 3 g),and resuspended in DPBS to a final OD550 of 1. The suspended bacteria wereheat killed at 65C for 30 min and subsequently sonicated on ice (15-s pulse,30-s pause, six times), aliquoted, and stored at 220C until further use.

Generation of C. jejuni mutants

To generate C. jejuni mutants, the target genes and their flanking regionswere PCR amplified from chromosomal DNA from strains 108 or 81116using the primers gene name fwd and gene name rev (Table I) andPhusion polymerase or PFX polymerase using the manufacturers protocol

in a Bio-Rad iCycler. The PCR products were gel purified using the QiaexII gel extraction kit or the GeneJET gel extraction kit, ligated into pGEM-T easy (cheY, cetA, flaC) or pJet1.2 (cdt) using the Rapid DNA ligation kitor CloneJET PCR cloning kit, and transformed into E. coli DH5a. Plas-mids were isolated from the transformants using the GeneJET plasmidminiprep kit and used to inactivate the cdt, cheY, and flaC genes via anoutward PCR with primers gene name BamHI/KpnI fwd and genename BamHI/KpnI rev. BamHI or KpnI sites were introduced to enableinsertion of the cat cassette from pAV35. This yielded the vectors pcdt::cat,pcheY::cat, and pflaC::cat. The cetA gene already contains a BamHI siteenabling direct insertion of the cat cassette after BamHI digestion, yieldingpcetA::cat. Gene inactivation constructs were verified by sequencing. Pri-mers T7 and Sp6 were used for the pGEM-Teasy inserts. The pJet1.2forward and reverse primers were used to sequence the pJet1.2 inserts(Baseclear). The vectors were introduced individually via electroporation(pcdt::cat, pcheY::cat, and pcetA::cat) or natural transformation (pflaC::cat)into C. jejuni using cat (20 mg/ml) selection. The 81116flaAB::kana/flaC::cat mutant was constructed via natural transformation of 81116flaC::catwith 81116flaAB::kana chromosomal DNA. All gene disruptions wereconfirmed via PCR. For cheY complementation, the cheY gene was am-plified from the chromosomal DNA from strain 108 using the primerscheY-sacI-fwd and cheY-SacII-rev and introduced into pWM1007 via SacIand SacII resulting in pcheY. pcheY was introduced into 108cheY::cat viaelectroporation resulting in 108cheY::cat+pcheY. Sequences were analyzedand aligned using Clone Manager 9 software (Sci-Ed).

Construction of C. jejuni luciferase reporter strains andfluorescent strains

The pMA5-metK-luc plasmid was introduced into several C. jejuni strains(81116, 108cheY::cat, 108kpsM::cat, and 108cetA::cat) via conjugation (35).Strain 108cheY::cat became either GFP or mCherry positive by introducingplasmid pMA1 containing GFP or mCherry via conjugation. In short, a 16-hculture of E. coli S17.1 containing the pMA5-metK-luc, pMA1-GFP, orpMA1-mCherry plasmid was diluted to an OD550 of 0.05 in 5 ml LBmedium. Analogous 16-h cultures of the C. jejuni strains were diluted to anOD550 of 0.5 in 5 ml HI broth. When the E. coli culture reached an OD550 of0.4, 1 ml C. jejuni culture was collected by centrifugation (10 min, 50003 g)and suspended in 1 ml E. coli culture. The C. jejuni and E. coli mix wasreleased on a Mueller Hinton plate and incubated for 5 h (37C) undermicroaerophilic conditions. Then, bacteria were collected in 1 ml HI, pel-leted (10 min, 5000 3 g), suspended in 100 ml HI, and plated on saponinagar plates containing 4% lysed horse blood, charcoal cefoperazone des-oxycholate agar, Campylobacter selective supplement, 50 mg/ml kana, andwhen required, 20 mg/ml cat. Single antibiotic resistant colonies werecollected after 48 h of incubation. The pMA1-mCherry plasmid was in-troduced in E. coli via chemical transformation.

Infection assay

J774.A1macrophages were seeded into 24- or 96-well plates in DMEMplus10% FCS. BMMs were seeded as described above. The next day, cells wereprimed by the addition of C. jejuni 108 lysate (equivalent of multiplicity ofinfection [m.o.i.] 20) for 16 h when appropriate. THP-1 monocytes (2 3105 cells/well in a 96-well plate) were differentiated with 100 nM PMA for48 h in RPMI 1640 medium plus 10% FCS. Prior to inoculation, the THP-1cells were rinsed, and RPMI 1640 medium plus 10% FCS without PMAwas added. The macrophages were inoculated with the indicated amountsof C. jejuni, Salmonella, or E. coli or stimulated with 2.5 or 5 mM ATP.After 20 min, the ATP was removed, and fresh medium was given. Ex-tracellular Salmonella and E. coli were removed after 2 h by replacing themedium with fresh medium containing 50 mg/ml gentamicin. The exper-iment was stopped, and samples were collected after 12 h of incubationunless indicated otherwise.

Real time RT-PCR

J774.A1 cells were seeded into a 24-well plate and stimulated the next dayby the addition of 50 ng/ml lipooligosaccharides (LOS) of N. meningitidis,lysate of C. jejuni strain 108 (equivalent of m.o.i 20) or inoculated withviable C. jejuni 108 (m.o.i of 20). After 4 h of stimulation, RNA wasisolated using RNA Bee, according to the manufacturers protocol. RNAwas treated with 1 mg DNAse/mg RNA for 30 min at 37C. The DNase wasinactivated by heating for 10 min at 65C in the presence of 2.5 mMEDTA. mRNA levels were determined in the LightCycler 480 Real-TimePCR System using the Brilliant III Ultra-Fast Sybr-Green qRT-PCR kit,according to the manufacturers protocol with the primers listed in Table I.Per reaction 50 ng DNase-treated RNA was used as template. Real-timecycle conditions: 3 min at 95C, 40 cycles 5 s 95C and 10 s 60C, 3 min

2 INFLAMMASOME ACTIVATION BY C. JEJUNI



http://ww

w.jimmunol.org/

Dow

nloaded from

95C. mRNA levels were calculated by subtracting the corresponding Ctvalues obtained for samples before (1) and after (2) treatment using thefollowing formula: 1) DCt control = Ct target gene control Ct Actin control and 2)DCt target gene treat Ct Actin treated. The fold change in mRNA was deter-mined by fold change = 2(DCt[treated]-(DCt[control]) (36). Presented results arefrom three individual assays performed in duplicate.

ELISA

Cell culture supernatants were collected in a fresh plate, centrifuged(10 min, 485 3 g) to remove dead cells, bacteria, and other debris, andstored at 280C until further analysis. IL-b levels were determined byELISA using the manufacturers protocol. Samples were diluted in assaydiluent to stay within the range of the assay (8 - 1000 pg/ml for mice and4 - 500 pg/ml for human). Plates were measured at 450 and 570 nm forwavelength correction on the FLUOstar Omega (BMG Labtech). Presentedresults are from three individual assays performed in duplicate.

Detection of caspase-1 and IL-1b cleavage

For caspase-1 cleavage J774.A1 macrophages were seeded into a 6-wellin DMEM+10% FCS. The next day, the medium was replaced with Opti-MEM and primed by the addition of C. jejuni 108 lysate (equivalent ofm.o.i. 20) for 16 h when appropriate. The macrophages were inoculatedwith C. jejuni or E. coli. After 2 h of infection, 250 mg/ml gentamicin wasadded to the E. coliinoculated wells. After 6 h of infection, the super-natant was collected and frozen at 280C for a minimum of 1 h witha protease inhibitor. For the detection of cleaved IL-1b in the supernatant,J774.A1 macrophages or THP-1 monocytes were seeded in a T25 flask andinfected as described for the infection assay with one minor change. Beforeinfection, the medium was replaced with Opti-MEM. Postinfection (12 h)the supernatant was collected and frozen at 280C for a minimum of 1 hwith a protease inhibitor. The thawed supernatant was concentrated viaa concentrator (9K MWCO) (IL-1b detection) or TCA precipitation(caspase-1). TCA (40%) was added to the supernatant in a 1:1 volume andincubated for 30 min at 4C. The precipitate was collected by centri-fugation (10 min, 21,100 3 g), washed twice with ice cold acetone(10 min, 21,100 3 g), dried (10 min at 50C), and taken up in radio-immunoprecipitation assay buffer. Protein concentration was determinedvia BCA protein concentration kit, according to the manufacturers pro-tocol. Protein (10 mg for caspase-1 detection and 50 mg for IL-1b detec-tion) was loaded and run on a 12% SDS-Page gel. After transfer of theproteins via blotting, the polyvinylidene difluoride membrane was blockedin TBST plus 5% milk (1 h) and incubated (16 h) with rabbit anticaspase-1 (ab17820) (1:1000) or rabbit antiIL-1b (ab9722) (1:2500) in TBST plus5% milk at 4C. The next day, the membrane was washed three times withTBST plus 5% milk (10 min per wash), incubated (1 h) with goat anti-rabbit IgG-HRP (A4914) (1:10,000), and washed (10 min per wash) withTBST plus 5% milk, TBST, and TBS. HRP signal was detected usingSuperSignal West Femto Chemiluminescent Substrate on the ChemiDocMP System (Bio-Rad). Data were analyzed using Image Lab software(Bio-Rad). The images have been cropped.

Luciferase reporter assay

Bacteria were cultured for 16 h under standard conditions with a start OD550of 0.01 from an 8-h preculture. Infection assays were performed withC. jejuni strains containing pMA5-metK-luc as described above. After 6 h,the cells were washed twice, lysed with 13 reporter lysis buffer supple-mented with 1% Triton-X100, and placed at 280C for at least 30 min.The cell lysate was analyzed for luciferase activity in a luminometer(TD20/20; Turner Designs) immediately after adding 50 ml Promega Lu-ciferase Assay Agent to the sample as described previously (20). Tomeasure intracellular survival, the medium was replaced after 6 h withmedium containing gentamicin (250 mg/ml). After an additional incuba-tion (2 h), the gentamicin concentration (50 mg/ml) was reduced andremained present throughout the assay. Presented results are from threeindividual assays performed in duplicate.

Cytotoxicity assay

Primed J774.A1 macrophages were infected in 96-well plate as describedabove with some minor changes. The assay was performed in DMEMwithout FCS. After 12 h of infection, the total and secreted amount of LDHwas determined using the Cytotoxicity Detection KitPLUS (LDH), accordingto the manufacturers protocol. Plates were measured at 492 and 690 nmfor wavelength correction on the FLUOstar Omega (BMG Labtech).C. jejuni by itself had no effect on LDH and did not influence the assay.The percentage of cytotoxicity was calculated as the percentage of LDHrelease compared with the total LDH concentration (percentage of cyto-

toxicity = 100 3 [LDH released/total LDH]). Presented results are fromthree individual assays performed in triplicate.

Propidium iodide uptake

Primed J774.A1 macrophages were infected in a 96-blackwell plate witha transparent bottom (see infection assay) with some minor changes. Theassay was performed in Opti-MEM without phenol red. After 11 h of in-fection, propidium iodide (PI) (3 mM) was added to the wells, and incu-bated for 1 h after which the plate was measured. Plates were excited at492 nm and measured at 610 nm using a fixed gain, with bottom optics(orbital averaging, 35 flashes) on the FLUOstar Omega (BMG Labtech).Fluorescence intensity (F.I.) was corrected for background fluorescencefrom an unstained non stimulated well. Presented results are from threeindividual assays performed in triplicate.

Confocal microscopy

J774.A1 cells or BMMs were grown on 12 mm circular glass slides, primedfor 16 h with the indicated stimulus, and inoculated with fluorescentC. jejuni. THP-1 monocytes were PMA differentiated, as mentioned pre-viously, on 12 mm circular glass slides and inoculated with fluorescentC. jejuni. For active caspase-1 detection cells were incubated (1 h, 37C)with FAM-FLICA (0.53) and washed twice (10 min, 37C) before fixa-tion. After incubation, cells were washed twice with DPBS and fixed with4% of paraformaldehyde in 100 mM phosphate buffer (pH 7.4) (1 h, roomtemperature). The fixed cells were washed with DPBS and stained withWGA Alexa Fluor 633 (1:500 for 1 h in DPBS). When needed cells werepermeabilized with 1% Triton X-100 plus 1% BSA in DPBS (30 min,20C). ASC was stained by incubation (16 h, 4C) of the cells with anti-TMS1 plus 0.01% Triton X-100 plus 2% BSA in DPBS followed with thegoat anti-rabbit-Alexa Fluor 488 (1:100) secondary Ab in DPBS plus 2%BSA (1 h, 20C). After staining the slides were washed three times withDPBS, once with MilliQ, embedded in Fluorsave, and viewed in the Bio-Rad radiance2000 system or Leica SPE-II system. The slides used forC. jejuni uptake were viewed using fixed settings for section thickness(2 mm) and magnification. Per slide 4 random images were captured. Datawere analyzed with ImageJ software.

Statistical analysis

Results were analyzed using GraphPad Prism 5 software.Where appropriatesignificance was calculated using a paired Student t test.

ResultsC. jejuni infection primes the macrophages for inflammasomeactivation

As activation of the inflammasome may involve a two-step process(priming and activation), we first determined whether C. jejuni wascapable of priming the cells for inflammasome activation (Fig. 1).The transcription levels of proIL-1b in J774.A1 macrophageswere determined using real-time RT-PCR with the primers listedin Table I. Incubation of macrophages with C. jejuni strain 108 ledto a 125-fold induction of proIL-1b mRNA at 4 h of infection(Fig. 1A). Lysed C. jejuni yielded an even stronger effect probablydue to increased availability of TLR ligands after bacterial dis-integration (25). The transcriptional levels of ASC, NLRP3,caspase-1, or caspase-11 were similar upon stimulation withC. jejuni lysate or LPS (4 h) (Fig. 1BE).To functionally verify the effect of priming, J774.A1 macro-

phages were incubated (for 16 h) with C. jejuni lysate and sub-sequently infected (12 h) with E. coli or stimulated with ATP(known inflammasome activators). The amount of secreted IL-1bin the culture supernatant was determined using ELISA. IL-1bsecretion was only observed in primed cells with additional E. colior ATP stimulation (Fig. 1F). Taken together, these result showthat C. jejuni lysate primes the cells for inflammasome activation.

C. jejuni infection activates the inflammasome

ProIL-1b is processed by activated inflammasomes and subse-quently secreted from the cell. To assess the activation of theinflammasome by C. jejuni, the amount of IL-1b secreted fromprimed macrophages was determined using ELISA. Infection of

The Journal of Immunology 3



http://ww

w.jimmunol.org/

Dow

nloaded from

primed macrophages with live C. jejuni resulted in a dose-dependent secretion of IL-1b (Fig. 2A) and was observed forseveral C. jejuni strains (Fig. 2B). Western blot analysis of the

supernatant confirmed the presence of the active IL-1b (17 kDa)form upon infection with C. jejuni strain 108 (Fig. 2C). TheC. jejuni lysate alone did not cause IL-1b secretion, suggesting therequirement for intact bacteria to activate inflammasomes.To ascertain that the release of IL-1b from the macrophages

was caused by activation of the inflammasome, we determinedwhether caspase-1 was cleaved and secreted into the supernatantupon infection (6 h) with E. coli or C. jejuni. The presence of thep20 cleavage fragment (containing the active domain) of caspase-1in the supernatant was determined using Western blotting. BothE. coli and C. jejuni strain 108 induced caspase-1 cleavage, asevident from the increased appearance of the p20 protein band(Fig. 2D). For unknown reasons, the cleavage pattern forE. coli and C. jejuniinfected macrophages were different. Asexpected, cleavage was most evident for primed macrophages.Inflammasome activation was also confirmed by the formation ofthe ASC speck (a large multiprotein complex is being formedcontaining ASC) upon infection with C. jejuni or E. coli (2 h) asdetermined by confocal microscopy (Fig. 2E). Clearly, theseresults demonstrate that C. jejuni is capable to activate theinflammasome.

Cellular infectiondependent inflammasome activation byC. jejuni

To learn more about the mechanism(s) of C. jejuniinducedinflammasome activation, a series of C. jejuni mutants with defectsin putative virulence determinants was tested for their ability toinduce IL-1b secretion (Fig. 3AC). In these experiments an in-termediate dose of C. jejuni (m.o.i. of 20) was used to infect thecells to avoid IL-1b secretion as result of bacterial depletion ofmedia components or possible C. jejuni induced cell toxicity. Ge-netic manipulation resulted in successful inactivation of a vastnumber of putative virulence genes, albeit in different C. jejunistrains. Infection assays demonstrated that genetically definedmutants with defects in bacterial capsule assembly (kpsM::cat),LOS assembly (waaF::cat), bacterial motility (motAB::cat), pro-duction of flagellin and flagellin-like proteins (flaAB::cat; flaAB::cat + flaC::kana), cytolethal distending toxin (CDT) production(cdt::cat), or the type 6 secretion apparatus (hcp::kana + kpsM::cat)induced similar levels of IL-1b secretion as their correspondingparental strain (Fig. 3AC). In contrast, mutants lacking thechemotaxis protein CheY or the energy taxis protein CetA eliciteda strongly increased (cheY::cat) or reduced (cetA::cat) IL-1b secretioncompared with the parent strain (Fig. 3A). Western blotting con-firmed the presence of more cleaved IL-1b in the supernatant of cellsinfected with strain 108cheY::cat compared with the parent strain(108) (Fig, 2C). This suggests that bacterial taxis strongly influencesthe level of inflammasome activation.The CheY- and CetA-defective bacterial phenotypes show, re-

spectively, hyperinvasive and hypoinvasive behavior toward epi-thelial cells (18, 19, 37, 38). Therefore, we examined a possiblecorrelation between the number of bacteria that infected themacrophages and IL-1b secretion for both taxis mutants and theparental strain 108 using a luciferase reporter assay (20). Mac-rophages were infected with different strains of C. jejunipro-ducing luciferase. The amount of luciferase produced correlates tothe number of viable intracellular bacteria. After 6 h of incubation,the supernatant of the cells was removed, and the cells were lysedto determine bacterial luciferase activity. In addition, after 12 h ofinfection, the IL-1b secretion in the supernatant was determined.This demonstrated that the cheY::cat mutant yielded highest lu-ciferase activity, whereas the cetA::cat mutant yielded lower levelsthan the parent strain C. jejuni 108 (Fig. 3D), thus followinga similar pattern as observed for the IL-1b secretion. Although

FIGURE 1. C. jejuni lysate primes macrophages for inflammasome ac-

tivation. (AE) mRNA expression levels of IL-1b, NLRP3, ASC, caspase-1,

and caspase-11 in J774.A1 macrophages postinfection (4 h) with viable

C. jejuni 108 (m.o.i. 20) or stimulation with LPS (50 ng/ml) or C. jejuni

lysate (m.o.i. 20). (A) LPS, viable, and lysate C. jejuni showed significantupregulation of IL-1b mRNA (p , 0.01). (B and C) No differences inmRNA levels were observed for NLRP3 or ASC after stimulation with

C. jejuni (viable or lysate) compared with LPS. (D) Caspase-1 mRNA

expression levels were upregulated after stimulation with LPS (p , 0.001),C. jejuni (p , 0.01), and lysate (p , 0.001). (E) Caspase-11 mRNAtranscript showed a minimal increase (p , 0.05) after stimulation with LPSor C. jejuni (viable or lysate). (F) Macrophage IL-1b secretion after stim-

ulation (12 h) of nonprimed (N) or lysate primed (n) cells with ATP (5 mM)

or E. coli (m.o.i. 20). In primed macrophages, ATP and E. coli significantly

increased IL-1b secretion (p , 0.01). Values are presented as the mean 6SEM of three independent experiments performed in duplicate.




http://ww

w.jimmunol.org/

Dow

nloaded from

several strain of C. jejuni induced IL-1b secretion (Fig. 2B), therewas some variation among the strains. This variation again cor-responded with the amount of intracellular bacteria as determinedwith the luciferase reporter assay (Fig. 3E). Taken together, theseresults highlight a correlation between the amount of intracellularbacteria and inflammasome activation.

C. jejuni does not cause cell death

Besides IL-1b secretion, a common downstream effect of in-flammasome activation is cell death via pyroptosis. Indeed, in-cubation of primed J774.A1 macrophages with ATP, Salmonella,or E. coli (12 h) induced cytotoxicity as estimated from the releaseof LDH in the culture supernatants (Fig. 4A). In contrast, incu-bation of the cells with C. jejuni did not result in the release ofdetectable amounts of LDH despite activation of the inflamma-some (Fig. 3AC). Even the hyperinvasive cheY::cat mutant didnot cause any LDH release. Measurement of PI uptake in J774.A1macrophages infected with C. jejuni (strain 108 and cheY::cat,12 h) revealed a minor yet not significant increase (Figure 4B). Incontrast, infection with Salmonella did increase PI uptake in thesecells, whereas E. coli had a minimal effect under the conditionsused. Taken together, these results suggest that C. jejuni activatesthe inflammasome without causing cell death.To assess whether the lack of cytotoxicity may be related to rapid

killing of the intracellular C. jejuni, we measured the intracellularsurvival in primed J774.A1 macrophages using the bacterial lu-ciferase reporter assay (Fig. 4C). After 12 h of infection, very fewviable intracellular bacteria were detected and none after 24 h.There was also no major difference in survival between strain 108and the hyperinvasive strain 108cheY::cat.

C. jejuni activates the NLRP3 inflammasome in primary mousemacrophages

To determine which type of inflammasome is activated byC. jejuni, we used BMMs derived from either C57BL/6 wild-typeor knockout mice deficient in distinct inflammasome components.The wild-type BMMs showed the expected secretion of IL-1bafter stimulation with ATP and the enhanced IL-1b secretionpostinfection with Salmonella (Fig. 5A). Both ATP and Salmo-nella required priming of the primary macrophages to be effective.Infection of the wild-type macrophages with C. jejuni strain108cheY::cat also elicited the release of IL-1b but without theneed to previous prime the macrophages. In fact, in primed cells,C. jejuni did not induce IL-1b secretion (Fig. 5A).The type of inflammasome that was activated by the various

stimuli was determined using BMMs from caspase-12/2/112/2,ASC2/2, NLRP32/2, and NLRC42/2mice (Fig. 5B, 5C). As expected,ATP induced IL-1b secretion in wild-type and NLRC42/2 BMMbut not in BMMs deficient in NLRP3, ASC, and caspase1/11, whichall are components of the NLRP3 inflammasome (Fig. 5B). Infectionwith Salmonella induced IL-1b secretion in both wild-type andNLRP32/2 BMMs, whereas secretion was severely reduced inNLRC42/2 and ASC2/2 macrophages, consistent with previ-ous reports (Fig. 5B) (39). C. jejuni strain 108 and the cheY::catmutant induced equal levels of IL-1b secretion in wild-typeBMM and BMMs isolated from NLRC42/2 mice (Fig. 5C).C. jejuniinduced IL-b secretion was not detected postinfectionof caspase-12/2/112/2 and ASC2/2 BMMs and severely reducedin NLRP32/2 BMMs. This suggests that C. jejuni activates theNLRP3 inflammasome and that caspase-1/11 and ASC are criticalfor the activation.

Table I. Primers used in this study

Primer Name Sequence

Real-time RT-PCRmActin RT fwd 59-TCCTGTGGCATCCACGAAACT-39mActin RT rev 59-GGAGCAATGATCCTGATCTTC-39mIL-1b RT fwd 59-CCCAAGCAATACCCAAAGAAGAAG-39mIL-1b RT rev 59-TGTCCTGACCACTGTTGTTTCC-39mNLRP3 RT fwd 59-CGAGACCTCTGGGAAAAAGCT-39mNLRP3 RT rev 59-GCATACCATAGAGGAATGTGATGTACA-39mASC RT fwd 59-AAAAGTTCAAGATGAAGCTGCTG-39mASC RT rev 59-CTCCTGTAAGCCCATGTCTCTAA-39mCaspase-1 RT fwd 59-TTTCAGTAGCTCTGCGGTGT-39mCaspase-1 RT rev 59-TTTCTTCCTGATTCAGCACTCTC-39mCaspase-11 RT fwd 59-GCCACTTGCCAGGTCTACGAG-39mCaspase-11 RT rev 59-AGGCCTGCACAATGATGACTTT-39

PCRcdt fwd 59-CTACACCCAAGGCCAAAG-39cdt rev 59-GCCTCGATAATATGGCGTCC-39cdt BamHI fwd 59-CCGGATCCAATTCGCCAAATGAACG-39cdt BamHI rev 59-CCGGATCCCTTTAACAGCTGCTACCC-39cheY fwd 59-AACTACACCACTCATTGATTT-39cheY rev 59-GCTGAGGCAGTGCAACTTGT-39cheY BamHI fwd 59-CGGGATCCTTCTGGCATATTCCAATCTG-39cheY BamHI rev 59-CGGGATCCGCCTATCATCATGGTTACAA-39cetA fwd 59-TCCCGCCATAAAGCCTTGTG-39cetA rev 59-TAGAGCCGCAAGCGTACTTC-39cheY sacI fwd 59-TCCGAGCTCTAAAAAACTTTGAAAGGACGAAAT-39cheY sacII rev 59-TCCCCGCGGTTAAAAATCAGCCTTTACTCAG-39FlaC fwd 59-AATCATTTTACCGCAGAACC-39FlaC rev 59-ATCAATCCCAAAGCCTTAGA-39FlaC KpnI fwd 59-GGGGTACCATAGTTGCATCAGAGATCAT-39FlaC KpnI rev 59-GGGGTACCAAATAGGCTCAGGTATCAAT-39T7 59-TATTTAGGTGACACTATAG-39SP6 59-TAATACGACTCACTATAGGG-39pJet1.2 forward 59-CGACTCACTATAGGGAGAGCGGC-39pJet1.2 reverse 59-AAGAACATCGATTTTCCATGGCAG-39




http://ww

w.jimmunol.org/

Dow

nloaded from

To ensure that the lack of inflammasome activation in C. jejuniinfected NLRP32/2 BMMs was not caused by poor infection ofthese cells, we determined the number of intracellular bacteria.Wild-type and NLRP32/2 cells were incubated (1 h) withmCherry-producing C. jejuni strains 108 and 108cheY::cat, andbacteria were visualized by confocal microscopy (Fig. 5E, 5F).Both cell types contained equal numbers of C. jejuni strains 108and 108cheY::cat. The cheY::cat mutant was more infective(4-fold) than the parent strain, as was already noted for the J774.A1 macrophages (Fig. 3D). The increased infectivity of the cheY::cat mutant compared with the parent strain likely explains thehigher production of IL-1b (Fig. 5C). Complementation of thecheY::cat mutant with the plasmid pcheY restored IL-1b secretionto parent levels (Fig. 5D).

Activation of the human NLRP3 inflammasome by C. jejuni

Although C. jejuni efficiently infects both mouse and humanmacrophages, mice normally do not establish infection afterC. jejuni exposure in contrast to humans. In addition, speciesdifferences between the human and mouse inflammasomes have

been reported (40). To ascertain that C. jejuni also activates thehuman inflammasome, PMA-differentiated human THP-1 mono-cytes and THP-1 cells deficient in NLRP3 (THP-1 defNLRP3)were infected with Salmonella, E. coli, or C. jejuni. This experi-mental setup abolished the need for additional priming of the cellsbecause the PMA treatment (100 nM PMA for 48 h) alreadyactivates NF-kB. Infection of the cells with Salmonella and E. coliinduced IL-1b secretion in the THP-1 cells but not in THP-1defNLRP3 cells (Fig. 6A), indicating that these bacteria activatethe human NLRP3 inflammasome. All C. jejuni strains tested alsoinduced IL-1b secretion in a NLRP3-dependent fashion (Fig. 6B).Furthermore, the C. jejuni 108cheY::cat mutant caused more andthe C. jejuni 108cetA::cat less secretion than the parent strain, aswas found for the mouse macrophages (Figs. 3A, 5C). Inflam-masome activation by C. jejuni in PMA-differentiated THP-1 cells(2 h infection) was confirmed by the visualization of an ASCspeck (Fig. 6C) and the presence of active caspase-1 (FLICA-positive cells) (Fig. 6D) as determined by confocal microscopyand the secretion of cleaved IL-1b (17 kDa) into the culture su-pernatant as shown via Western blotting (12 h infection) (Fig. 6E).

FIGURE 2. C. jejuni activates the inflammasome. (A and

B) IL-1b secretion by primed J774.A1 macrophages infected

(12 h) with different numbers of C. jejuni, C. jejuni lysate

(m.o.i. 20) (A) or with different C. jejuni strains (m.o.i.200) (B). Infection with an m.o.i. higher than 20 (p , 0.05)or with different strains (p , 0.01) increased IL-1b secre-tion. Values are presented as the mean 6 SEM of three in-dependent experiments performed in duplicate. (C) Western

blot probed for active IL-1b (17 kDa) (arrowhead) in the

supernatant of primed J774.A1 macrophages postinfection

(12 h) with C. jejuni strain 108 (m.o.i. 20 or 200), 108cheY::

cat (m.o.i. 20), E. coli (m.o.i 20), or without bacteria (non-

stimulated). Lanes were loaded with 50 mg protein. (D)

Western blot probed for the cleaved caspase-1 p20 fragment

in the supernatant of nonprimed or primed J774.A1 macro-

phages postinfection (6 h) with C. jejuni strain 108 (m.o.i.

200), E. coli (m.o.i. 20), or without bacteria (nonstimulated).

Lanes were loaded with 10 mg protein. (E) Confocal mi-

croscopy showing ASC speck formation (,) in primed J774.A1 macrophages postinfection (2 h) with mCherry fluores-

cent C. jejuni strain 108 (m.o.i. 40) or E. coli (m.o.i. 20)

(red). ASC foci were stained with anti-ASC Ab in combi-

nation with goat-anti-rabbit-Alexa Fluor 488 (green). Cell

surface was stained with WGA-Alexa Fluor 633 (blue). Scale

bar, 10 mm.




http://ww

w.jimmunol.org/

Dow

nloaded from

Luciferase assays on the infected THP-1 cells demonstrated higherbacterial values for the 108cheY::cat than for the parent strain(Fig. 6F), suggesting that also in human cells inflammasome ac-tivation varies with cellular infection levels. THP-1 defNLRP3macrophages showed similar levels of luciferase activity asmeasured for the infected THP-1 cells excluding that the differ-ence in IL-1b secretion in the deficient cells was caused by lowinfection rates. Finally, intracellular survival of C. jejuni strain108 (Fig. 6G) and 108cheY::cat (Fig. 6H) was followed by theluciferase reporter assay in differentiated THP-1 and THP-1defNLRP3 cells. Intracellular levels of both strains severely re-duced overtime and were almost absent after 24 h of infection. Nodifference in bacterial survival was observed between the THP-1

and the THP-1 defNLRP3 cells, indicating that the poor intra-cellular survival of C. jejuni occurs independent of inflammasomeactivation. Overall, these results show that C. jejuni also activatesthe human NLRP3 inflammasome and that activation varies withthe amount of infection.

DiscussionIn the current study, we provide evidence that the principal bacterialfood-borne pathogen C. jejuni induces the secretion of IL-1b viaactivation of the NLRP3 but not the NLRC4 inflammasome. Theeffect required the inflammasome components NLRP3, ASC andcaspase-1/11 and was observed upon infection of both mouse andhuman macrophages. Inflammasome activation required viable

FIGURE 3. Cellular infection-induced acti-

vation of the inflammasome. (AC) Induction of

IL-1b secretion in primed J774.A1 macro-

phages (12 h) exposed to different C. jejuni

mutants and their respective parent strains 108,

81176, and 81116 (m.o.i. 20). (D) Effect of

C. jejuni infection of primed J774.A1 macro-

phages (m.o.i. 20) on IL-1b secretion (N) (after

12 h) and bacterial viability (n) (after 6 h) as

measured via the luciferase reporter assay (rel-

ative light unit [RLU]). The p values (AD) for

all mutant strains compared with the parent

strain were not significantly different, except for

the cheY::cat and cetA::cat (p , 0.05). (E) In-tracellular bacterial viability of several C. jejuni

strains (m.o.i. 20) in primed J774.A1 macro-

phages (6 h) as measured via the luciferase re-

porter assay. Strain 81176 was significantly

(p, 0.01) more present intracellular than strain108. Values are the mean 6 SEM of three in-dependent experiments performed in duplicate.

FIGURE 4. Cell viability and intracellular survival after

C. jejuni induced inflammasome activation. (AC) Cyto-

toxicity and bacterial survival in infected primed J774.A1

macrophages. (A) LDH release from primed cells at 12 h of

incubation with ATP (5 mM), Salmonella (m.o.i. 20),

E. coli (m.o.i. 20), or several C. jejuni (m.o.i. 20) or

mutants of strain 108. ATP (p , 0.001), Salmonella (p ,0.01), and E. coli (p , 0.01) caused significant cytotox-icity. None of the C. jejuni strains induced cytotoxicity. (B)

PI uptake by primed macrophages incubated (12 h) with

the indicated strains. Salmonella (p , 0.001) causeda significant increase in F.I. The increase F.I. induced by

C. jejuni cheY mutant was not statistically significant. (C)

Intracellular survival of C. jejuni strain 108 and its cheY

derivative (m.o.i. 20) in primed J774.A1 macrophages as

measured via the luciferase reporter assay. Luciferase ac-

tivity (relative light unit [RLU]) after 6 h infection was set

as 100% value. Values are the mean 6 SEM of three in-dependent experiments performed in triplicate.




http://ww

w.jimmunol.org/

Dow

nloaded from

bacteria and varied with the severity of the cellular infection.Strikingly, inflammasome activation by C. jejuni did not lead tocell death and occurred in primary mouse macrophages withoutthe need of a priming signal.The secretion of IL-1b and the structurally related IL-18 are

important in the innate immunity and systemic response againstbacterial infections (41). Many flagellated bacterial pathogens(Legionella pneumophila, Pseudomonas aeruginosa, SalmonellaTyphimurium, Shigella flexneri, and enteropathogenic E. coli) acti-vate the NLRC4 inflammasome (42, 43). Crucial for this activation isthe translocation of flagellin or components of the T3SS into thecytosol, which are sensed by members of the NAIP family (810, 4446). Our results with mouse NLRP32/2 and NLRC42/2 knockoutcells and human NLRP32/2 macrophages demonstrate that C. jejunifails to activate the NLRC4 inflammasome. This is consistent withthe absence of a T3SS or T4SS in this pathogen. The apparent ab-sence of translocation of flagellin or other NAIP ligands into thecytosol makes C. jejuni invisible for the NLRC4 inflammasome.In this study, a large body of evidence indicates that C. jejuni

activates the NLRP3 inflammasome. The C. jejuniinduced se-cretion of mature IL-1b in primed J774.A1 macrophages, theformation of an ASC speck, the generation of caspase-1 cleavagefragments, and the absence and severe reduction of IL-1b secre-tion in human and mouse NLRP3-negative cells respectively, all

indicate NLRP3-dependent IL-1b secretion. Interestingly, some lowresidual IL-1b secretion remained in the NLRP3-deficient mousemacrophages, whereas this was not observed for the caspase-1/11 andASC-deficient macrophages. This suggests a possible role for anadditional inflammasome contributing to the response.The NLRP3 inflammasome can be formed in response to a di-

verse array of agents (4, 7, 43) but, to our knowledge, for none ofthem binding of a specific ligand to NLRP3 has been demon-strated. The mechanism driving the activation of the NLRP3 byC. jejuni was investigated using different C. jejuni strains andseries of genetically defined mutants. All of the tested strains in-duced IL-1b secretion, suggesting that inflammasome activation isa stable trait of the pathogen. This trait was preserved in mutantswith defects in capsule formation, LOS biosynthesis, flagellasynthesis and flagellin(-like) secretion, T6SS needle protein, CDT,and several assumed bacterial adhesion/invasion promoting fac-tors. Inflammasome activation was affected after disruption of thecheY gene and, to a lesser extent, the cetA gene. The strong in-crease in IL-1b production observed for the CheY mutant wasaccompanied by increased cellular infection. The hyperinvasiveCheY phenotype was evident from microscopy and luciferasebacterial gene reporter assays and was observed for both mouseand human macrophages. The exact signal(s) that drive(s)C. jejuniinduced NLRP3 formation remain to be defined but

FIGURE 5. C. jejuni activation of the NLRP3 inflam-

masome in primary mouse macrophages. (A) Secretion of

IL-1b by non-primed (N) and primed (n) BMMs incubated

(12 h) with Salmonella (m.o.i. 2), C. jejuni (m.o.i. 20), or

ATP (2.5 mM). ATP (p , 0.05) and Salmonella (p ,0.001) significantly increased IL-1b secretion; C. jejuni

significantly increased IL-1b in nonprimed cells only (p ,0.001). (B) Similar assay but with primed BMMs isolated

from the indicated knockout mice and incubated with

Salmonella (N) (m.o.i. 2) or ATP (2.5 mM) (n). IL-1b se-

cretion was significantly lower in caspase-12/2/112/2 and

NLRC42/2 BMMs upon infection with Salmonella (p ,0.05). Upon ATP stimulation, IL-1b secretion was signifi-

cantly reduced in caspase-12/2/112/2, ASC2/2, or

NLRP32/2 BMMs (p , 0.05). (C) IL-1b secretion bynonprimed BMMs from several knockout mice infected

(12 h) with C. jejuni strain 108 (N) or 108cheY::cat (n)

(m.o.i. 20). Significant lower secretion was observed for

caspase-12/2/112/2 (p , 0.001), ASC2/2 (p , 0.001),and NLRP32/2 (p , 0.01) BMMs upon stimulation withboth C. jejuni strains. Strain cheY::cat induced more IL-1b

secretion (p, 0.05) than the parent strain. (D) Secretion ofIL-1b in nonprimed BMMs infected (12 h) with C. jejuni

strain 108, 108cheY::cat, or 108cheY::cat+pcheY (m.o.i.

20). Complementation of the defective cheY restored the

high IL-1b secretion induced by strain 108cheY::cat (p ,0.01) to parental levels (p . 0.0.05). (E) Confocal mi-croscopy on BMMs (wt and NLRP32/2) infected (1 h)

with mCherry fluorescent (red) C. jejuni 108 and 108cheY::

cat (m.o.i 200); cells were counterstained with WGA-

Alexa Fluor 633 (green). (F) Quantification of the number

of intracellular bacteria in the wild-type (wt) (N) and

NLRP32/2 (n) BMMs showed no significant difference in

the number of intracellular bacteria between the wt and

NLRP32/2 BMMs. Strain 108cheY::cat was more invasive

(p , 0.001) than the parent strain in the BMMs. Values arethe mean 6 SEM of three independent experiments per-formed in duplicate.




http://ww

w.jimmunol.org/

Dow

nloaded from

likely cause a transmembrane ion flux (e.g., K+-efflux), whichseems the common denominator of NLRP3 activation (12). At thistime, it is tempting to speculate that C. jejuniinduced cell dam-age caused by the cellular infection contributes to the NLRP3activation, although this was not evident from increased release ofLDH from the cells (Fig. 4).Our results indicate that C. jejuni activates the NLRP3 in-

flammasome in murine J774.A1 macrophage cells, primary mouseBMMs, and human-differentiated THP-1 cells. A striking differ-ence among these cell types however, was the apparent lack ofneed to prime the primary cells to establish a strong IL-1b pro-duction upon C. jejuni infection. ATP only activated the inflamma-some in primed BMMs, indicating that the cells were not alreadyprimed. Viable C. jejuni were most effective in inflammasome ac-tivation in nonprimed primary cells. In primed cells, IL-1b pro-duction was minimal probably because of the more efficient bacterialuptake and killing in activated macrophages.Another unexpected finding was that C. jejuni activates the

inflammasome without apparent cytotoxicity as evidenced by theabsence of LDH release or an increased PI uptake by the infectedcells. LDH release was noted postinfection with E. coli or Sal-monella, indicating that the pyroptosis pathway is intact in thesecells. Inflammasome activation by other bacterial species alwaysseems to be followed by pyroptosis (2, 4, 6, 7, 42, 47). This hasbeen linked with the presence of LPS in the cytosol (48, 49). It canbe imagined that different bacterial metabolic demands, a lowlevel of bacterial LPS in the cytosol, and/or the relative poor in-tracellular survival of C. jejuni prevent C. jejuniinduced cyto-toxicity. Activation of the inflammasome was not required to killthe intracellular C. jejuni. Alternatively, the pathogen may haveevolved a strategy to prevent bacteria-induced pyroptosis.Our results that C. jejuni induces inflammasome activation in

both murine and human cells without apparent cytotoxicity and inprimary cells without a need of priming extends the known rep-ertoire of inflammasome activation by bacterial pathogens. Thedata provide a molecular basis for the observed IL-1b secretionduring C. jejuni infection and for key features in C. jejuni path-ogenesis (22, 23, 31, 33, 5052).

FIGURE 6. C. jejuni activation of the NLRP3 inflammasome in human

macrophages. (A and B) Secretion of IL-1b by PMA differentiated THP-1

(N) and THP-1 defNLRP3 (n) cells infected with Salmonella (m.o.i. 2),

E. coli (m.o.i. 20), or different C. jejuni strains (m.o.i. 20) (12 h). Sal-

monella (p , 0.01), E. coli (p , 0.01), and all C. jejuni strains (p , 0.05)induced IL-1b secretion upon infection. Strain cheY::cat induced more

IL-1b secretion (p , 0.01) than the parent strain. (C and D) Confocalmicroscopy on PMA-differentiated THP-1 cells infected (2 h) with

mCherry positive C. jejuni strain 108 (m.o.i. 40) and E. coli (m.o.i. 20)

(red). Cell surface was stained with WGA-Alexa Fluor 633 (blue). ASC

foci [, and . in (C)] were stained with an anti-ASC Ab in combinationwith goat-anti-rabbit-Alexa Fluor 488 (green). Active caspase-1 (D) was

detected with FLICA (green) at 1 h of infection. Scale bars, 10 mm. (E)

Western blot showing the presence of active (cleaved) IL-1b (17 kDa) in

the supernatant of noninfected and C. jejuni strain 108 (m.o.i. 20)infected

(12 h) PMA-differentiated THP-1 cells. Lanes were loaded with 50 mg

protein. (F) Intracellular viability of C. jejuni mutants and parent strain in

6 h infected PMA-differentiated THP-1 (N) and THP-1 defNLRP3 (n) as

measured with the bacterial luciferase reporter assay. No significant dif-

ferences in RLU were measured between the THP-1 and THP-1

defNLRP3. Strain 108cheY::cat was more invasive (p , 0.001) than theparent strain. (G and H) Intracellular survival (24 h) of C. jejuni strains 108

(G) or 108cheY::cat (H) in THP-1 (N) or THP-1 defNLRP3 (n) cells. Lu-

ciferase activity at 6 h of infection was set as 100%. The decrease in in-

tracellular C. jejuni (108 and 108cheY::cat) at 12 h of infection was

statistically significant (p , 0.05). Values are the mean 6 SEM of threeindependent experiments performed in duplicate.




http://ww

w.jimmunol.org/

Dow

nloaded from

AcknowledgmentsWe thank Dr. Dietmar Zaiss (Utrecht University) for providing the L929

cells.

DisclosuresThe authors have no financial conflicts of interest.

References1. Thornberry, N. A., H. G. Bull, J. R. Calaycay, K. T. Chapman, A. D. Howard,

M. J. Kostura, D. K. Miller, S. M. Molineaux, J. R. Weidner, J. Aunins, et al.1992. A novel heterodimeric cysteine protease is required for interleukin-1 betaprocessing in monocytes. Nature 356: 768774.

2. Fernandes-Alnemri, T., J. Wu, J. W. Yu, P. Datta, B. Miller, W. Jankowski,S. Rosenberg, J. Zhang, and E. S. Alnemri. 2007. The pyroptosome: a supra-molecular assembly of ASC dimers mediating inflammatory cell death viacaspase-1 activation. Cell Death Differ. 14: 15901604.

3. Martinon, F., K. Burns, and J. Tschopp. 2002. The inflammasome: a molecularplatform triggering activation of inflammatory caspases and processing of proIL-b. Mol. Cell 10: 417426.

4. Latz, E., T. S. Xiao, and A. Stutz. 2013. Activation and regulation of theinflammasomes. Nat. Rev. Immunol. 13: 397411.

5. Strowig, T., J. Henao-Mejia, E. Elinav, and R. Flavell. 2012. Inflammasomes inhealth and disease. Nature 481: 278286.

6. Franchi, L., R. Munoz-Planillo, and G. Nunez. 2012. Sensing and reacting tomicrobes through the inflammasomes. Nat. Immunol. 13: 325332.

7. Broz, P., and D. M. Monack. 2011. Molecular mechanisms of inflammasomeactivation during microbial infections. Immunol. Rev. 243: 174190.

8. Franchi, L., A. Amer, M. Body-Malapel, T. D. Kanneganti, N. Ozoren,R. Jagirdar, N. Inohara, P. Vandenabeele, J. Bertin, A. Coyle, et al. 2006. Cy-tosolic flagellin requires Ipaf for activation of caspase-1 and interleukin 1b inSalmonella-infected macrophages. Nat. Immunol. 7: 576582.

9. Miao, E. A., C. M. Alpuche-Aranda, M. Dors, A. E. Clark, M. W. Bader,S. I. Miller, and A. Aderem. 2006. Cytoplasmic flagellin activates caspase-1 andsecretion of interleukin 1b via Ipaf. Nat. Immunol. 7: 569575.

10. Miao, E. A., D. P. Mao, N. Yudkovsky, R. Bonneau, C. G. Lorang, S. E. Warren,I. A. Leaf, and A. Aderem. 2010. Innate immune detection of the type III se-cretion apparatus through the NLRC4 inflammasome. Proc. Natl. Acad. Sci. USA107: 30763080.

11. Compan, V., A. Baroja-Mazo, G. Lopez-Castejon, A. I. Gomez, C. M. Martnez,D. Angosto, M. T. Montero, A. S. Herranz, E. Bazan, D. Reimers, et al. 2012.Cell volume regulation modulates NLRP3 inflammasome activation. Immunity37: 487500.

12. Munoz-Planillo, R., P. Kuffa, G. Martnez-Colon, B. L. Smith, T. M. Rajendiran,and G. Nunez. 2013. K+ efflux is the common trigger of NLRP3 inflammasomeactivation by bacterial toxins and particulate matter. Immunity 38: 11421153.

13. Yuki, N., K. Susuki, M. Koga, Y. Nishimoto, M. Odaka, K. Hirata, K. Taguchi,T. Miyatake, K. Furukawa, T. Kobata, and M. Yamada. 2004. Carbohydratemimicry between human ganglioside GM1 and Campylobacter jejuni lip-ooligosaccharide causes Guillain-Barre syndrome. Proc. Natl. Acad. Sci. USA101: 1140411409.

14. Allos, B. M. 2001. Campylobacter jejuni infections: update on emerging issuesand trends. Clin. Infect. Dis. 32: 12011206.

15. Black, R. E., M. M. Levine, M. L. Clements, T. P. Hughes, and M. J. Blaser.1988. Experimental Campylobacter jejuni infection in humans. J. Infect. Dis.157: 472479.

16. Rathinam, V. A., K. A. Hoag, and L. S. Mansfield. 2008. Dendritic cells fromC57BL/6 mice undergo activation and induce Th1-effector cell responses againstCampylobacter jejuni. Microbes Infect. 10: 13161324.

17. van Alphen, L. B., N. M. Bleumink-Pluym, K. D. Rochat, B. W. van Balkom,M. M. Wosten, and J. P. van Putten. 2008. Active migration into the subcellularspace precedes Campylobacter jejuni invasion of epithelial cells. Cell. Micro-biol. 10: 5366.

18. Yao, R., D. H. Burr, and P. Guerry. 1997. CheY-mediated modulation of Cam-pylobacter jejuni virulence. Mol. Microbiol. 23: 10211031.

19. Golden, N. J., and D. W. Acheson. 2002. Identification of motility and autoag-glutination Campylobacter jejuni mutants by random transposon mutagenesis.Infect. Immun. 70: 17611771.

20. Bouwman, L. I., P. Niewold, and J. P. van Putten. 2013. Basolateral invasion andtrafficking of Campylobacter jejuni in polarized epithelial cells. PLoS One 8:e54759.

21. van Putten, J. P., L. B. van Alphen, M. M. Wosten, and M. R. de Zoete. 2009.Molecular mechanisms of Campylobacter infection. In Current Topics in Mi-crobiology and Immunology, Vol. 337. Sasakawa, C., ed. Springer Verlag, Berlin,p. 197229.

22. Bouwman, L. I., and J. P. M. van Putten. 2012. Biology of Campylobacter In-fection. In Foodborne and Waterborne Bacterial Pathogens. S. M. Faruque, ed.Caister Academic Press, Norfolk, UK, p. 231250.

23. Young, K. T., L. M. Davis, and V. J. Dirita. 2007. Campylobacter jejuni: mo-lecular biology and pathogenesis. Nat. Rev. Microbiol. 5: 665679.

24. Zilbauer, M., N. Dorrell, A. Elmi, K. J. Lindley, S. Schuller, H. E. Jones,N. J. Klein, G. Nunez, B. W. Wren, and M. Bajaj-Elliott. 2007. A major role for

intestinal epithelial nucleotide oligomerization domain 1 (NOD1) in elicitinghost bactericidal immune responses to Campylobacter jejuni. Cell. Microbiol. 9:24042416.

25. de Zoete, M. R., A. M. Keestra, P. Roszczenko, and J. P. van Putten. 2010.Activation of human and chicken toll-like receptors by Campylobacter spp.Infect. Immun. 78: 12291238.

26. Wassenaar, T. M., M. Engelskirchen, S. Park, and A. Lastovica. 1997. Differ-ential uptake and killing potential of Campylobacter jejuni by human peripheralmonocytes/macrophages. Med. Microbiol. Immunol. (Berl.) 186: 139144.

27. Hickey, T. E., G. Majam, and P. Guerry. 2005. Intracellular survival of Cam-pylobacter jejuni in human monocytic cells and induction of apoptotic death bycytholethal distending toxin. Infect. Immun. 73: 51945197.

28. Kiehlbauch, J. A., R. A. Albach, L. L. Baum, and K. P. Chang. 1985. Phago-cytosis of Campylobacter jejuni and its intracellular survival in mononuclearphagocytes. Infect. Immun. 48: 446451.

29. Watson, R. O., and J. E. Galan. 2008. Campylobacter jejuni survives withinepithelial cells by avoiding delivery to lysosomes. PLoS Pathog. 4: e14.

30. Jones, M. A., S. Totemeyer, D. J. Maskell, C. E. Bryant, and P. A. Barrow. 2003.Induction of proinflammatory responses in the human monocytic cell line THP-1by Campylobacter jejuni. Infect. Immun. 71: 26262633.

31. Siegesmund, A. M., M. E. Konkel, J. D. Klena, and P. F. Mixter. 2004. Cam-pylobacter jejuni infection of differentiated THP-1 macrophages results in in-terleukin 1 beta release and caspase-1-independent apoptosis. Microbiology 150:561569.

32. Sun, X., D. Threadgill, and C. Jobin. 2012. Campylobacter jejuni induces colitisthrough activation of mammalian target of rapamycin signaling. Gastroenterol-ogy 142: 8695.e5.

33. Islam, D., N. Ruamsap, A. Aksomboon, P. Khantapura, A. Srijan, andC. J. Mason. Immune responses to Campylobacter (C. jejuni or C. coli)infections: a two-year study of US forces deployed to Thailand. APMIS. In press.

34. Weischenfeldt, J., and B. Porse. 2008. Bone Marrow-Derived Macrophages(BMM): Isolation and Applications. CSH Protoc. 2008: pdb.prot5080.

35. van Mourik, A., N. M. Bleumink-Pluym, L. van Dijk, J. P. van Putten, andM. M. Wosten. 2008. Functional analysis of a Campylobacter jejuni alkalinephosphatase secreted via the Tat export machinery. Microbiology 154: 584592.

36. Keestra, A. M., and J. P. van Putten. 2008. Unique properties of the chickenTLR4/MD-2 complex: selective lipopolysaccharide activation of the MyD88-dependent pathway. J. Immunol. 181: 43544362.

37. Hickey, T. E., S. Baqar, A. L. Bourgeois, C. P. Ewing, and P. Guerry. 1999.Campylobacter jejuni-stimulated secretion of interleukin-8 by INT407 cells.Infect. Immun. 67: 8893.

38. Elliott, K. T., I. B. Zhulin, J. A. Stuckey, and V. J. DiRita. 2009. Conservedresidues in the HAMP domain define a new family of proposed bipartite energytaxis receptors. J. Bacteriol. 191: 375387.

39. Broz, P., K. Newton, M. Lamkanfi, S. Mariathasan, V. M. Dixit, andD. M. Monack. 2010. Redundant roles for inflammasome receptors NLRP3 andNLRC4 in host defense against Salmonella. J. Exp. Med. 207: 17451755.

40. Atianand, M. K., E. B. Duffy, A. Shah, S. Kar, M. Malik, and J. A. Harton. 2011.Francisella tularensis reveals a disparity between human and mouse NLRP3inflammasome activation. J. Biol. Chem. 286: 3903339042.

41. Dinarello, C. A. 2009. Immunological and inflammatory functions of theinterleukin-1 family. Annu. Rev. Immunol. 27: 519550.

42. von Moltke, J., J. S. Ayres, E. M. Kofoed, J. Chavarra-Smith, and R. E. Vance.2013. Recognition of bacteria by inflammasomes. Annu. Rev. Immunol. 31: 73106.

43. de Zoete, M. R., and R. A. Flavell. 2013. Interactions between Nod-likereceptors and intestinal bacteria. Front. Immunol. 4: 462.

44. Yang, J., Y. Zhao, J. Shi, and F. Shao. 2013. Human NAIP and mouse NAIP1recognize bacterial type III secretion needle protein for inflammasome activa-tion. Proc. Natl. Acad. Sci. USA 110: 1440814413.

45. Zhao, Y., J. Yang, J. Shi, Y. N. Gong, Q. Lu, H. Xu, L. Liu, and F. Shao. 2011.The NLRC4 inflammasome receptors for bacterial flagellin and type III secretionapparatus. Nature 477: 596600.

46. Rayamajhi, M., D. E. Zak, J. Chavarria-Smith, R. E. Vance, and E. A. Miao.2013. Cutting edge: mouse NAIP1 detects the type III secretion system needleprotein. J. Immunol. 191: 39863989.

47. Cunha, L. D., and D. S. Zamboni. 2013. Subversion of inflammasome activationand pyroptosis by pathogenic bacteria. Front. Cell. Infect. Microbiol. 3: 76.

48. Hagar, J. A., D. A. Powell, Y. Aachoui, R. K. Ernst, and E. A. Miao. 2013.Cytoplasmic LPS activates caspase-11: implications in TLR4-independent en-dotoxic shock. Science 341: 12501253.

49. Kayagaki, N., M. T. Wong, I. B. Stowe, S. R. Ramani, L. C. Gonzalez,S. Akashi-Takamura, K. Miyake, J. Zhang, W. P. Lee, A. Muszynski, et al. 2013.Noncanonical inflammasome activation by intracellular LPS independent ofTLR4. Science 341: 12461249.

50. Edwards, L. A., K. Nistala, D. C. Mills, H. N. Stephenson, M. Zilbauer,B. W. Wren, N. Dorrell, K. J. Lindley, L. R. Wedderburn, and M. Bajaj-Elliott.2010. Delineation of the innate and adaptive T-cell immune outcome in thehuman host in response to Campylobacter jejuni infection. PLoS One 5: e15398.

51. Sims, J. E., and D. E. Smith. 2010. The IL-1 family: regulators of immunity. Nat.Rev. Immunol. 10: 89102.

52. Al-Banna, N., R. Raghupathy, and M. J. Albert. 2008. Correlation of proin-flammatory and anti-inflammatory cytokine levels with histopathologicalchanges in an adult mouse lung model of Campylobacter jejuni infection. Clin.Vaccine Immunol. 15: 17801787.




http://ww

w.jimmunol.org/

Dow

nloaded from

Documents

01 J Immunol Bouwman2014 Campylobacter