Tumor Biology and Immunology

Inflammasome Adaptor ASC SuppressesApoptosis of Gastric Cancer Cells by an IL18-Mediated Inflammation-Independent MechanismVirginie Deswaerte1,2, Paul Nguyen3,4, Alison West1,2, Alison F. Browning1,2,Liang Yu1,2, Saleela M. Ruwanpura1,2, Jesse Balic1,2, Thaleia Livis1,2, Charlotte Girard5,6,Adele Preaudet3,4, Hiroko Oshima7, Ka Yee Fung3,4, Hazel Tye1,2, Meri Najdovska1,2,Matthias Ernst8, Masanobu Oshima7, Cem Gabay5,6, Tracy Putoczki3,4, andBrendan J. Jenkins1,2

Abstract

Inflammasomes are key regulators of innate immunity inchronic inflammatory disorders and autoimmune diseases, buttheir role in inflammation-associated tumorigenesis remains ill-defined. Here we reveal a protumorigenic role in gastric cancer forthe key inflammasome adaptor apoptosis-related speck-like pro-tein containing a CARD (ASC) and its effector cytokine IL18.Genetic ablation of ASC in the gp130F/F spontaneous mousemodel of intestinal-type gastric cancer suppressed tumorigenesisby augmenting caspase-8-like apoptosis in the gastric epithelium,independently from effects onmyeloid cells andmucosal inflam-mation. This phenotype was characterized by reduced activationof caspase-1 and NF-kB activation and reduced expression ofmature IL18, but not IL1b, in gastric tumors. Genetic ablation ofIL18 in the same model also suppressed gastric tumorigenesis,whereas blockade of IL1b and IL1a activity upon genetic ablationof the IL1 receptor had no effect. The specific protumorigenic role

for IL18 was associated with high IL18 gene expression in thegastric tumor epithelium compared with IL1b, which was pref-erentially expressed in immune cells. Supporting an epithelial-specific role for IL18, we found it to be highly secreted fromhuman gastric cancer cell lines. Moreover, IL18 blockade eitherby a neutralizing anti-IL18 antibody or by CRISPR/Cas9-drivendeletion of ASC augmented apoptosis in human gastric cancercells. In clinical specimens of human gastric cancer tumors, weobserved a significant positive correlation between elevatedmature IL18 protein and ASC mRNA levels. Collectively, ourfindings reveal the ASC/IL18 signaling axis as a candidate ther-apeutic target in gastric cancer.

Significance: Inflammasome activation that elevates IL18helps drive gastric cancer by protecting cancer cells against apo-ptosis, with potential implications for new therapeutic strategiesin this setting. Cancer Res; 78(5); 1293–307. �2017 AACR.

IntroductionGastric cancer is the third most lethal cancer worldwide, and is

among a growing number of cancers associated with precursorychronic inflammatory responses (1–3). The predominant histo-

logic subtype of gastric cancer is intestinal-type, and despite thecausal correlation between chronic gastric inflammation triggeredby pathogenic microbes (i.e., Helicobacter pylori) and intestinal-type gastric cancer (2, 4), the identity of innate immune regulatorswithin the host gastric mucosa that promote gastric cancerremains unclear. Consistent with an altered host immuneresponse predisposing to gastric cancer, gene polymorphisms forthe proinflammatory cytokine IL1b, which are associated withaugmented gene expression, increase the risk of human gastriccancer (5, 6), and transgenic overexpression of IL1b in micetriggers gastric inflammation and tumors (7). In addition, clinicalstudies on the related IL1 cytokine family member IL18, whichcan display opposing anti- or protumorigenic effects dependentupon the tissue and cellular context in various cancers (8),demonstrate that IL18 levels are increased in gastric cancerpatients and serve as a poor prognostic marker (9–11). Althoughexperimental data from human gastric cancer cell lines alsosuggest that IL18 may contribute to the malignant progressionof tumors (9, 12, 13), a definitive role for IL18 in gastric cancerremains unproven.

Inflammasomes have recently been identified as multi-protein complexes that are essential for the production ofmature and bioactive IL1b and IL18 proteins. Accordingly, theyhave attracted much attention as key factors of the immunesystem with the potential to influence susceptibility to many

1Centre for Innate Immunity and Infectious Diseases, Hudson Institute of MedicalResearch, Clayton, Victoria, Australia. 2Department of Molecular TranslationalScience, School of Clinical Sciences, Monash University, Clayton, Victoria,Australia. 3Inflammation Division, Walter and Eliza Hall Institute of MedicalResearch, Parkville, Victoria, Australia. 4Department of Medical Biology, Uni-versity of Melbourne, Parkville, Victoria, Australia. 5Division of Rheumatology,University Hospital of Geneva, Geneva, Switzerland. 6Department of Pathologyand Immunology, University of Geneva School ofMedicine, Geneva, Switzerland.7Division of Genetics, Cancer Research Institute, Kanazawa University, Kana-zawa, Japan. 8Olivia Newton-John Cancer Research Institute, La Trobe Univer-sity School of Cancer Medicine, Heidelberg, Victoria, Australia.

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

V. Deswaerte and P. Nguyen contributed equally to this article.

Corresponding Author: Brendan J. Jenkins, Hudson Institute of MedicalResearch, 27-31 Wright Street, Clayton, Victoria 3168, Australia. Phone: 613-8572-2740; E-mail: brendan.jenkins@hudson.org.au

doi: 10.1158/0008-5472.CAN-17-1887

�2017 American Association for Cancer Research.

CancerResearch

www.aacrjournals.org 1293

autoimmune and inflammatory diseases, as well as cancers,where IL1b and IL18 are implicated (14–17). These cytokinesare initially produced as inactive proIL1b and proIL18 forms,with proIL1b expression upregulated following ligand-mediat-ed activation of pattern recognition receptors (PRR) such asToll-like receptor (TLR) or nucleotide-binding oligomerizationdomain-containing (NOD) family members, whereas proIL18is constitutively expressed. Subsequently, their inflammasome-mediated secretion as bioactive cytokines is controlled bymembers of the nucleotide-binding domain and leucine-richrepeat containing protein (NLR) family, namely NLR pyrindomain containing 1 (NLRP1), NLRP3, NLRP6, NLRP12, NLRCARD domain containing 4 (NLRC4) and NLR apoptosisinhibitory protein (NAIP), as well as the cytosolic DNA sensorabsent in melanoma 2 (AIM2; refs. 15, 18). Specifically, each ofthese NLRs and AIM2 form the core of distinct inflammasomecomplexes, whereby upon sensing host- and/or microbial-derived ligands, they associate with the key adaptor proteinapoptosis-related speck-like protein containing a CARD (ASC)to facilitate activation of caspase-1, which catalyzes the matu-ration of proIL1b and proIL18 precursors into bioactive secret-ed cytokines (15).

Although ASC is critical for inflammasome-mediated patho-logies involving IL1b and/or IL18, investigations into thedefinitive role of ASC in tumorigenesis are still in their infancyand have been largely restricted to intestinal and skin carcino-genesis, with contrasting findings (19). Here, we reveal thatelevated PYCARD (hereafter referred to as ASC) mRNA andmature IL18, but not IL1b, protein levels are a coincidentfeature of gastric tumors from both gastric cancer patients andthe gp130F/F intestinal-type gastric cancer mouse model (20).Furthermore, genetic ablation of either Asc or Il18 in gp130F/F

mice suppressed gastric tumor growth, independent of inflam-mation, which was associated with augmented neoplasticepithelial cell death. Interestingly, at the molecular level sup-pressed gastric tumorigenesis in both gp130F/F:Asc�/� andgp130F/F:Il18�/� mice was characterized by reduced activationlevels of NF-kB. In further support of these findings, we alsodemonstrate in vitro that CRISPR/Cas9-mediated genetic abla-tion of ASC in human gastric cancer cells suppressed theircolony-forming potential (i.e., growth), which was associatedwith reduced secretion of mature IL18, increased cellular apo-ptosis, and reduced activation levels of NF-kB. Collectively,these findings support the existence of a novel protumorigenicASC inflammasome/IL18 axis in gastric cancer.

Materials and MethodsMice

The gp130F/F mice (20), along with mice homozygousnull for Asc (Pycard) (21), Il1r (22), or Il18 (23) were used togenerate gp130F/F:Asc�/�, gp130F/F:Il1r�/�, and gp130F/F:Il18�/�

mice, respectively, on a mixed 129Sv � C57BL/6 background.Experiments comparing different mouse strains includedgenetically- and age-matched littermates, including whereappropriate, wild-type (gp130þ/þ) control mice. Mice werehoused under specific pathogen-free conditions on a 12-hourlight/dark cycle, and all animal studies were approved by theMonash University Monash Medical Centre "A" Animal EthicsCommittee and the Walter and Eliza Hall Institute AnimalEthics Committee.

Human biopsiesGastric biopsies were collected at the XinHuaHospital (Shang-

hai, China) frompatients, upon formalwritten informed consent,undergoing upper gastrointestinal endoscopy or surgical resec-tion. Clinicopathologic features and demographics of gastriccancer patient cohorts are described in Supplementary TableS1. Biopsies were snap-frozen in liquid nitrogen. Studies wereapproved by the Xin Hua Hospital Human Research EthicsCommittee and undertaken in accordance with the appropriateethics guidelines. Patient studies were conducted in accordancewith the World Medical Association Declaration of Helsinkistatement on the ethical principles for medical research involvinghuman subjects.

Laser microdissectionTumor epithelial and stroma samples from gp130F/F mice were

collected fromOCT-embedded frozen sections stained with tolu-idine blue using laser microdissection (Leica). Total RNA wasextracted from microdissected samples using the miRNeasymicrokit (Qiagen), and then reverse transcribed with the Prime-Script RT Reagent Kit (Takara). Quantitative RT-PCR (qPCR) wasperformed as described later.

RNA isolation and gene expression analysesTotal RNA was isolated from snap-frozen mouse and human

stomach tissues using TRI Reagent Solution (Sigma) followed byon-column RNeasy Mini Kit RNA clean-up and DNase treatment(Qiagen). qPCR was performed on cDNA with Taqman GeneExpression Assays (mouse Il1b, Il1r, Il18, Il18r1, Pycard; Thermo-Fisher Scientific) or SYBR Green chemistry (Life Technologies)using the Applied Biosystems 7300, 7900HT Fast, and Viia7 Real-Time PCR Systems (ThermoFisher Scientific). Data acquisitionand analyses were undertaken using the Sequence DetectionSystem Version 2.4 software (Applied Biosystems). Forward andreverse primer sequences for mouse 18S rRNA, Il1b, Tnfa, Cxcl1,Cxcl2, Ccnd1, Ccnd2, and c-myc have been previously published(24). Sequences for othermouse andhumanprimerswill be givenupon request.

The Cancer Genome AtlasGene expression data and clinical information from The

Cancer Genome Atlas (TCGA) gastric cancer patients wereobtained from the open access TCGA data portal (https://portal.gdc.cancer.gov/projects/TCGA-STAD). The alignment ofsample identifiers yielded 18 primary gastric cancer cases forwhich there was available tumor and matched nontumor data.We used reads per kilobase of exon model per million mappedreads (RPKM) to quantify ASC expression levels from RNAsequencing (RNA-Seq) data generated from each gastric cancerpatient within TCGA.

ELISA and immunoblottingTotal protein lysates from snap-frozen tissues were prepared

at room temperature with two incubations of 20 minutes at37�C before and after homogenization (25). ELISA for humantotal IL1b (R&D Systems), and mouse and human total IL18(MBL International Corporation) were performed according tothe manufacturer's instructions. Mature and free human IL18was detected with a recently-developed sandwich ELISA (26),and mature mouse IL18 was detected by ELISA using equivalentmethodology for the mature human IL18 ELISA. All ELISAs

Deswaerte et al.

Cancer Res; 78(5) March 1, 2018 Cancer Research1294

were performed using 50 mg of protein lysate per well of 96-wellplates. Immunoblotting was performed with antibodies againstmouse (AdipoGen) and human (Cell Signaling Technologies)caspase-1, IL18 (BioVision), phospho(p)NF-kB p65 (Ser536;Cell Signaling Technologies) and a-tubulin (Abcam), andprotein bands were visualized using either the Odyssey InfraredImaging System (LI-COR) for IL18 (p24/p18) and a-Tubulin,or enhanced chemiluminescence for caspase-1 (p45/p20). Thebands were quantified using ImageJ software (NIH).

Histology, IHC, and immunofluorescenceFormalin-fixed and paraffin-embedded tissue sections were

stained with hematoxylin and eosin (H&E) for histologic evalu-ation. The terminal deoxynucleotidyl transferase (tdT)–mediateddUDP nick-end labeling (TUNEL) assay (Millipore), as well asimmunohistochemistry to detect proliferating cell nuclear anti-gen (PCNA) and cleaved caspase-8 (Cell Signaling Technology),pNF-kB p65 (Ser536; Santa Cruz Biotechnology), and CD45,B220, CD3, and CD68 (BD BioSciences), were performed asbefore (24, 27). Positive-stained cells were counted manually[n ¼ 20 high-power (�40) fields] with a random offset, orpercentage area of positive staining was acquired using ImageJsoftware. Dual immunofluorescence was performed on paraffin-embedded stomach tissues using fluorescent-conjugated primaryantibodies, and Alexa Fluor secondary antibody. Antigens weredetected with antibodies against E-cadherin and active caspase-3(Cell Signaling Technology), pNF-kB p65 (Ser536; Santa CruzBiotechnology), and CD45 (BD Biosciences), as previouslydescribed (27). Slides were examined by confocal microscopy(Nikon) and analyzed for red and green fluorescence. Whereappropriate, stereologic techniques were applied to enumeratedual-stained cells (27).

Flow cytometric sorting and analyses of gastric single-cellsuspensions

Single-cell suspensions of dissected mouse stomach tumor-bearing tissue were prepared by collagenase digestion asdescribed previously (24). For cell sorting, single cells werecollected and stained with fluorescence-conjugated antibodiesagainst CD45 (Biolegend) and EpCAM (eBioscience), as well aspropidium iodide, and then sorted using a FACSAria instrument(BD Biosciences). For analyses, single-cell populations werestained with fluorescence-conjugated antibodies against CD3,CD69, CD19, CD4, CD11c, B220, CD86, CD45 CD11b, CD8,CD11b (eBioscience), and Gr-1 and F4/80 (BD BioSciences).Apoptosis of human AGS cells was assessed with an Annexin V:FITC Apoptosis Detection Kit (BD Pharmingen). Stained cellswere acquired on a FACSCanto instrument (BD Biosciences)and analyzed using FlowJo software (Tree Star) as describedpreviously (24).

Human cell lines and colony-forming assaysHuman gastric cancer cell lines AGS (ATCC) and MKN1

(Japanese Collection of Research Bioresources Cell Bank) weremaintained in RPMI supplemented with 10% heat-inactivatedFCS, 1% penicillin—streptomycin, and 1% L-glutamine (GIBCO).Cell lines were authenticated via short tandem repeat profiling(PowerPlex HS16 System Kit; Promega) in our laboratory afterreceipt in 2013. For experiments, cell lines were passaged forunder 3 months at a time between freeze–thaw cycles. Cell lineswere routinely tested throughout the time of experiments for

mycoplasma contamination (MycoAlert PLUS MycoplasmaDetection Kit; Lonza). For clonogenic assays, AGS and MKN1cells were seeded in six-well plates (1 � 104 cells/well) withRPMI/10% FCS media with or without either anti-hIL18-IgAantibody (1 mg/mL; InvivoGen) or anakinra (1 mg/mL; SwedishOrphan Biovitrum AB), the latter a specific IL1 receptor (IL1R)antagonist that blocks the activity of IL1b and IL1a. After 10 daysof culture, colonies (containing � 50 cells) were stained andfixed with a solution of 0.005% crystal violet (Sigma-Aldrich) in10% methanol, and colonies counted.

CRISPR-driven ASC and caspase-1 gene knockoutUsing published protocols (28), self-complementary oligonu-

cleotides (Sigma-Aldrich) for human ASC (hASC-sgRNA1-O1-caccgCGAGGGTCACAAACGTTGAG; hASC-sgRNA1-O2-aaacCT-CAACGTTTGTGACCCTCGc; hASC-sgRNA2-O1- caccgCATGTC-GCGCAGCACGTTAG; hASC-sgRNA2-O2-aaacCTAACGTGCTG-CGCGACATGc) and human caspase-1 (hCASP1-sgRNA1-O1-caccgAAAGCTGTTTATCCGTTCCA; hCASP1-sgRNA1-O2- aaacT-GGAACGGATAAACAGCTTTc; hCASP1-sgRNA2-O1- caccgGCT-CCCTAGAAGAAGCTCAA; hCASP1-sgRNA2-O2- aaacTTGAGCT-TCTTCTAGGGAGCc) were ligated into the LentiCRISPRv2 con-struct (Addgene). The single-guided (sg) RNA sequences aredesigned and constructed for human ASC to allow for sgRNAtargeting of constitutive exonic coding regions (Exon 1 andExon 3). Lentivirus was produced by transfecting vectors intoLenti-X/H293T cells with LentiCRISPR:psPAX2:pMD2.G at a ratioof 4:3:1. Virus was harvested 48 hours after transfection, filteredand used to infect AGS cell cultures containing 5 mg/mL poly-brene. Infected cells were selected with puromycin, and cellsinfected with nontarget control sgRNA vector were used as neg-ative controls.

Bone marrow chimeric miceThe gp130F/F and gp130F/F:Asc�/� mice were lethally-irradiated

(single 9.5 Gy dose) and reconstituted with 5 � 106 unfractio-nated donor bone marrow cells from the indicated genotypes asdescribed previously (24).

Statistical analysesStatistical analyses were performed using GraphPad PrismV6.0

software. Data normality were assessed using the D'Agostino andPearson omnibus K2 normality test, and the appropriate tests toidentify statistical significance (P < 0.05) between the means oftwo or multiple groups are presented in the relevant figurelegends. Data are expressed as the mean � SEM.

ResultsASC expression is elevated in human gastric cancer, and geneticablation of ASC in gp130F/F mice suppresses gastric tumorgrowth

In human gastric cancer, gene expression for the key inflam-masome adaptor, ASC, was significantly elevated in tumorsfrom two independent gastric cancer patient cohorts (Fig. 1Aand B; Supplementary Table S1), thus implicating ASC indisease pathogenesis. To interrogate the role of ASC in gastriccancer, we generated gp130F/F mice deficient for ASC (gp130F/F:Asc�/�). At 10 to 12 weeks of age, which is 4 to 6 weeks afterthe onset of antral gastric hyperplasia and tumor formation ingp130F/F mice, the stomach size, tumor mass, and incidence

ASC Inflammasomes Promote Gastric Cancer via IL18

www.aacrjournals.org Cancer Res; 78(5) March 1, 2018 1295

Figure 1.

Suppressed gastric tumorigenesis in gp130F/F:Asc�/� mice. A, qPCR expression of ASC (relative to 18S rRNA) in gastric tumor (T) and matched, adjacentnontumor (NT) tissue from 10 Chinese gastric cancer patients. ��� , P < 0.001; unpaired t test. B, ASC gene expression in gastric tumor and nontumor tissues(left), and in each gastric tumor tissue relative to matched nontumor tissue (right), from 18 TCGA gastric cancer patients. � , P < 0.05; unpaired/pairedt tests. C, Representative 20-to 24-week-old gp130F/F (F/F) and gp130F/F:Asc�/� (F/F:Asc�/�) mouse stomachs. Arrows, macroscopically visibletumors. Fundic (f), body (b), and antral (a) stomach regions are depicted. D, Scatter plots depicting total mass (g) of stomachs and gastric tumors,and incidence of tumors in total and by size, from 20-to 24-week-old mice. Data are expressed as the mean � SEM. � , P < 0.05; �� , P < 0.01; ��� , P < 0.001;���� , P < 0.0001; unpaired t test. E, Representative photomicrographs showing H&E-stained whole stomach cross-sections from 20- to 24-week-oldmice. Tumors are depicted by dotted squares. Scale bars, 1 mm. F, Left, representative H&E-stained tumor cross-sections from 20- to 24-week-old F/F andF/F:Asc�/� mice. Arrows, inflammatory cell accumulates. Scale bars, 100 mm. Right, magnified submucosa areas demonstrating the presence of plasma (P)and lymphocyte (L) inflammatory cells. Scale bars, 50 mm. Graph depicts inflammatory scores (0–3; none, mild, moderate, severe) from six mice/genotype.Shown in C and E is one from 15 to 25 representative stomach images/genotype, and in F, one from six representative images/genotype.

Deswaerte et al.

Cancer Res; 78(5) March 1, 2018 Cancer Research1296

were comparable between gp130F/F and gp130F/F:Asc�/� mice(Supplementary Fig. S1A–S1E). However, compared to age-and sex-matched gp130F/F littermates, 20- to 24-week-old and30- to 34-week-old gp130F/F:Asc�/� mouse stomachs display-ed a markedly reduced hyperplastic response and werevisibly smaller and significantly reduced in mass by approxi-mately 30% and 40%, respectively (Fig. 1C and D; Supple-mentary Fig. S1F and S1G), with no observable gender bias.Furthermore, although histologic assessment of gp130F/F andgp130F/F:Asc�/� mouse stomachs revealed similar gastric ade-nomatous hyperplastic lesions with no evidence of low-gradedysplasia nor carcinoma in situ, the gastric antral tumor massin gp130F/F:Asc�/� mice was significantly reduced by approxi-mately 40% at 20–24 weeks and approximately 60% at 30–34weeks compared with age-matched gp130F/F mice (Fig. 1D;Supplementary Fig. S1H). In addition, the total incidenceof gastric lesions in 20- to 24-week-old and 30- to 34-week-old gp130F/F: Asc�/� mice was also significantly reduced com-pared to gp130F/F mice at the corresponding ages, which corre-sponded with smaller hyperplastic lesions in gp130F/F:Asc�/�

mice (Fig. 1D and E; Supplementary Fig. S1I and S1J).Protein levels of activated caspase-1 p20 subunit, the down-

stream effector of ASC inflammasomes, were significantlyincreased 1.7- to 2-fold in stomachs of 20- to 24-week-oldgp130F/F compared with gp130þ/þ wild-type mice (Supplemen-tary Fig. S2A), and were significantly reduced by 7.1-fold(compared to 2.6-fold for procaspase-1) in gp130F/F:Asc�/�

versus gp130F/F tumors (Supplementary Fig. S2B). These datatherefore confirm that the elevated inflammasome activityduring gastric tumorigenesis in gp130F/F mice is reduced in theabsence of ASC.

ASC-driven gastric tumorigenesis in gp130F/F mice isindependent of inflammation

Because ASC inflammasome activation can instigate potentinflammatory responses, we investigated whether suppressedtumorigenesis in gp130F/F:Asc�/� mice was associated withreduced gastric inflammation. Histologic assessment ofH&E-stained gastric tissue sections from 20- to 24-week-oldmice revealed that the presence of chronic inflammatoryinfiltrates comprising primarily plasma and lymphoid cellswas comparable in the submucosa and mucosa regions oftumors from both genotypes (Fig. 1F). Immunohistochemicalanalyses also indicated comparable numbers of CD45þ leu-kocyte infiltrates, and CD68þ macrophage, B220þ B cell andCD3þ T cell immune subsets, in the gastric mucosa of gp130F/F

and gp130F/F:Asc�/� tumors (Fig. 2A and B; SupplementaryFig. S3A–S3D). Consistent with these observations, flow cyto-metry on tumor-bearing stomachs of 20- to 24-week-oldgp130F/F:Asc�/� and gp130F/F mice confirmed similar frequen-cies of gastric immune cell subsets, namely CD11bþGr-1þ

myeloid-derived suppressor cells (MDSC), CD11cþCD11b�

dendritic cells, CD11bþF4/80þ monocytes/macrophages, andB- and T-cell populations (Fig. 2C). Also, the activation statusof B (B220þCD86þ) and T (CD4þCD69þ and CD8þCD69þ)cells, which are the predominant infiltrating immune cells ingp130F/F and gp130F/F:Asc�/� gastric tumors, was comparable(Fig. 2C). The similar frequency and activation status ofdistinct immune cell subsets was also confirmed in perigastriclymph nodes of gp130F/F and gp130F/F:Asc�/� mice (Supple-mentary Fig. S3E).

The unaltered inflammation was also coincident withsimilar mRNA levels of numerous inflammatory genes ingastric tumors from 20- to 24-week-old gp130F/F andgp130F/F:Asc�/� mice (Fig. 3A). We also assessed whether thesuppressed gastric tumorigenesis was associated with reducedexpression of various genes encoding angiogenic factors impli-cated in gastric cancer, namely Cxcl1, Cxcl2, Vegf, Mmp2, andMmp9. However, their mRNA levels were not significantlyreduced in gp130F/F:Asc�/� tumors (Supplementary Fig. S3F),suggesting that ASC does not promote tumor angiogenesis ingp130F/F mice.

Because ASC is expressed in both the gp130F/F gastric tumorepithelium and immune cell-containing stroma, albeit signifi-cantly higher in the tumor epithelium (Fig. 3B), we assessedwhether ASC-expressing myeloid cells contributed to gastrictumorigenesis by generating reciprocal bone marrow chimerasbetween irradiated 8-week-old gp130F/F and gp130F/F:Asc�/�micethat were subsequently aged. The size of hyperplastic stomachsand tumor mass from 20- to 24-week-old gp130F/F recipientsreconstituted with gp130F/F:Asc�/� donor bone marrow was com-parable to control gp130F/F mice reconstituted with autologousgp130F/F bone marrow (Fig. 3C and D). Similarly, the reciprocalreconstitution of gp130F/F:Asc�/� recipients with gp130F/F bonemarrow had no effect on gastric tumor growth (Fig. 3C and D).Collectively, these data indicate that ASC-expressingmyeloid cellsdo not promote gastric tumorigenesis.

Suppressed gastric tumorigenesis in gp130F/F:Asc�/� mice ischaracterized by augmented tumor cell apoptosis

We have previously demonstrated that gastric tumorigenesisin gp130F/F mice is associated with a high PCNA proliferativeindex compared to wild-type mice (24). However, immunos-taining of gastric tumor sections from 20- to 24-week-oldgp130F/F and gp130F/F:Asc�/� mice revealed comparable PCNAþ

cell numbers, indicating that suppressed tumorigenesis ingp130F/F:Asc�/� mice was not characterized by reduced tumorcell proliferation (Fig. 4A). Similarly, the expression of cell-cycle regulatory genes Ccnb1, Ccnd1, Ccnd2, c-myc, and Cdc42was comparable in gp130F/F:Asc�/� and gp130F/F tumors (Sup-plementary Fig. S3G).

We next investigated whether ASC contributed to increasedsurvival of neoplastic gastric epithelial cells, a cellular processthat is associatedwith gastric tumorigenesis in gp130F/Fmice (20).Indeed, TUNELþ and cleaved caspase-8þ apoptotic cell numberswere significantly increased (�3-fold) in gastric tumors of 20- to24-week-old gp130F/F:Asc�/� compared with gp130F/F mice(Fig. 4B and C). Dual immunofluorescence staining for cleavedcaspase-3 and the epithelial marker E-cadherin further confirmeda significant increase in the number of apoptotic epithelial cells ingastric tumors of gp130F/F:Asc�/� versus gp130F/F mice (Fig. 4D;Supplementary Fig. S4A).

Because activation of the protumorigenic transcription factorNF-kB within the intestinal epithelium has been linked withsuppressing epithelial cell apoptosis in colitis-associated cancer(29), we investigated whether a similar function for NF-kBcould be assigned to the gastric epithelium during tumorigen-esis. Indeed, in gastric tumors of 20- to 24-week-old gp130F/F

mice, immunofluorescence indicated both nuclear and cyto-plasmic staining for the phosphorylated (Ser536) p65 subunitof NF-kB primarily in the mucosal glandular epithelium, withlittle to no pNF-kB p65 staining in the immune/inflammatory

ASC Inflammasomes Promote Gastric Cancer via IL18

www.aacrjournals.org Cancer Res; 78(5) March 1, 2018 1297

cell-rich submucosal or lamina propria regions (Supple-mentary Fig. S4B). Furthermore, in the smaller gp130F/F:Asc�/�

tumors displaying increased apoptosis, immunoblottingrevealed significantly reduced pNF-kB p65 levels (Supplemen-tary Fig. S4C). In addition, immunohistochemistry confirmedthat the reduced phosphorylation (i.e., activation) of NF-kBassociated with a significantly lower number of pNF-kB p65-positive cells in the mucosal epithelium, compared to gp130F/F

tumors (Fig. 4E). By contrast, immune/inflammatory cellaggregates found within the matching submucosal areas ofgp130F/F and gp130F/F:Asc�/� gastric tumors contained compa-rable low numbers of pNF-kB p65-positive cells (Supplemen-tary Fig. S4D). Collectively, these data suggest that ASC defi-ciency in the gp130F/F gastric tumor epithelium augmentscaspase-8-mediated cell death, which correlates with reducedNF-kB activation.

Increased production of IL18, but not IL1b, is associatedwith ASC-mediated gastric tumorigenesis in gp130F/F mice

To elucidate a role for the ASC inflammasome effector cyto-kines IL1b and/or IL18 in gastric tumorigenesis, we initiallymeasured IL1b and IL18 expression levels in gastric tumors of10- to 12-week-old and 20- to 24-week-old gp130F/F mice.Although Il1b mRNA levels were significantly increased by up toapproximately 50-fold in tumor and approximately 6-fold inadjacent tumor-free gastric antrum tissues of gp130F/F mice com-pared to normal gastric antrum tissue of gp130þ/þmice (Fig. 5A),immunoblotting indicated that protein levels of pro (31 kD) andmature (17 kD) forms of IL1b were only increased (albeit signif-icantly) by up to approximately 3-fold in gp130F/F tumor and/ortumor-free lysates comparedwith gp130þ/þ gastric antrum lysates,with the highest increase in 20- to 24-week-old gp130F/F tumors(Fig. 5B and C). By contrast, despite a <3-fold increase in Il18

Figure 2.

Genetic disruption of ASC does not suppress gastric inflammation in gp130F/F mice. A, Representative CD45-stained gastric antral tumor cross-sectionsfrom 20- to 24-week-old gp130F/F (F/F) and gp130F/F:Asc�/� (F/F:Asc�/�) mice (one of 8 representative images/genotype). Scale bars, 100 mm.B, Quantitative enumeration (mean � SEM) of CD45-positive cells/high-power field (HPF) in gastric tumor mucosa of eight mice/genotype. C, Frequenciesof cell populations, presented as the mean � SEM, in F/F and F/F:Asc�/� 20- to 24-week-old mouse stomachs (6 mice/genotype) as determined byflow cytometry.

Deswaerte et al.

Cancer Res; 78(5) March 1, 2018 Cancer Research1298

Figure 3.

Suppressed gastric tumorigenesis in gp130F/F:Asc�/�mice is independent of inflammation and hematopoietic-derivedmyeloid cells.A, qPCR expression analyses ofinflammatory genes in gastric tumors of 20- to 24-week-old gp130F/F (F/F) and gp130F/F:Asc�/� (F/F:Asc�/�) mice (eight mice/genotype). Expression dataare normalized for 18S rRNA and are presented from experimental triplicates as the mean � SEM. B, qPCR expression analyses of the indicated genes in capturedlaser microdissected gastric tumor epithelial (Epi) and stroma (Strom) tissue from 20- to 24-week-old F/F mice. Expression data from five samples/genotypeare shown following normalization for 18S rRNA and are presented from technical triplicates as the mean � SEM. � , P < 0.05; unpaired t test. C, Scatter plot,presented as the mean � SEM, depicting the total mass (g) of mouse gastric tumors. � , P < 0.05; �� , P < 0.01; one-way ANOVA followed by Tukey multiplecomparisons test. D, Representative stomachs from 20- to 24-week-old recipient F/F mice reconstituted with autologous F/F (F/FF/F) or heterologous gp130F/F:Asc�/� (F/FF/F:Asc) mouse bone marrow and recipient F/F:Asc mice reconstituted with autologous F/F:Asc (F/F:AscF/F:Asc) or heterologous F/F (F/F:AscF/F)bone marrow (shown is one of 6 representative stomach images/group). Fundic (f), body (b) and antral (a) stomach regions are depicted.

ASC Inflammasomes Promote Gastric Cancer via IL18

www.aacrjournals.org Cancer Res; 78(5) March 1, 2018 1299

mRNA levels in gp130F/F gastric tumor and/or nontumor tissuescompared with gp130þ/þ antrum tissue (Fig. 5A), immunoblotsrevealed that the mature 18 kDa form of IL18, but notthe immature 24 kDa proIL18, was specifically upregulated by

up to 27-fold in both 10- to 12-week-old and 20-to 24-week-oldgp130F/F tumor and tumor-free tissues (Fig. 5B and C).

In light of these findings suggesting increased processing ofmature IL18 protein, we next assessed whether IL18 specifically

Figure 4.

ASC promotes gastric tumor cellsurvival in gp130F/F mice. A,Representative photomicrographsshowing PCNA-stained gastric antraltumor cross-sections from 20- to 24-week-old gp130F/F (F/F) and gp130F/F:Asc�/� (F/F:Asc�/�) mice. Scale bars,50 mm. Graph depicts quantitativeenumeration of PCNA-positive cells/high-power field (HPF) in gastrictumor mucosa from mice. B,Representative photomicrographs ofTUNEL-stained antral gastric tumorcross-sections, along with graphdepicting quantitative enumeration ofTUNEL-positive cells/20HPF in gastrictumor mucosa, of mice. Arrows,TUNEL-positive cells. Scale bars, 50mm. C, Representativephotomicrographs of active caspase-8-immunostained antral gastric tumorcross-sections, along with graphdepicting quantitative enumeration ofactive caspase-8-positive cells/20HPF in gastric tumor mucosa, of mice.Arrows, active caspase-8–positivecells. Scale bars, 20 mm. D,Representative confocalimmunofluorescencephotomicrographs of cells stained forcleaved caspase-3 (green), E-cadherin(red, epithelial cell marker), and thenuclear marker 40 ,6-diamidino-2-phenylindole (DAPI; blue) in cross-sections of 20- to 24-week-old F/Fand F/F:Asc�/� gastric tumors. Whitecircles, dual-labeled E-cadherinand cleaved caspase-3 cells.Scale bars, 40 mm. Graph depictsstereological quantification of dual-labeled caspase-3/E-cadherin-positive cells in gastric tumors of20- to 24-week-old mice. E,Representative pNF-kB p65-stainedcross-sections through the antraltumor mucosal region of mousestomachs. Scale bars, 50 mm. Graphdepicting the percentage of pNF-kBp65-positive cells/area in the gastrictumor mucosa from mice of theindicated genotypes. In A–E, shown isone of 8 representative stomachimages/genotype, and data in graphsare presented as the mean � SEMfromeightmice/genotype. �� ,P<0.01;��� , P < 0.001; ����, P < 0.0001;unpaired t test.

Deswaerte et al.

Cancer Res; 78(5) March 1, 2018 Cancer Research1300

Figure 5.

Upregulated IL18 production and caspase-1 activation in gp130F/F mouse gastric tumors. A, In antral gastric tissue from 10- to 12-week-old and 20- to 24-week-old gp130þ/þ (þ/þ) mice, and tumor (T) and nontumor (NT) tissue from age-matched gp130F/F (F/F) mice, shown are Il1b and Il18 mRNA levels byqPCR. Gene expression data are normalized to 18S rRNA and are presented from technical triplicates as the mean � SEM. n ¼ 5 samples/group. � , P < 0.05;�� , P < 0.01; unpaired t test. B and C, Immunoblots of 10- to 12-week-old (B) and 20- to 24-week-old (C) þ/þ, F/FNT, and F/FT gastric antral tissuelysates with anti-IL18 and anti-IL1b antibodies detecting both pro/mature forms (24/18 kD for IL18, 31/17 kD for IL1b). Each lane represents an individual mouse.Protein loading was assessed using a-tubulin antibody. Graphs depict densitometric quantification of immunoblots from individual gastric tumor tissue lysates(6 mice/genotype) showing pro and mature IL18 and IL1b levels relative to a-tubulin. �, P < 0.05; �� , P < 0.01; ��� , P < 0.001; unpaired t test. D, Immunoblots of20- to 24-week-old F/F and F/F:Asc�/� gastric tumor tissue lysates with antibodies against pro/mature IL18 and IL1b, as well as a-tubulin. Each lanerepresents an individual mouse. Graph depicts densitometric quantification of immunoblots from individual gastric tumor tissue lysates (6 mice/genotype)showing pro and mature IL18 and IL1b levels relative to a-tubulin. � , P < 0.05; unpaired t test.

ASC Inflammasomes Promote Gastric Cancer via IL18

www.aacrjournals.org Cancer Res; 78(5) March 1, 2018 1301

acted downstream of ASC to promote gastric tumorigenesis.Indeed, immunoblotting revealed that mature IL18 protein levelswere significantly reduced in tumor lysates from 20- to 24-week-old gp130F/F:Asc�/� versus gp130F/F mice, which was not observedfor mature IL1b (Fig. 5D). The reduced processing of IL18 ingp130F/F:Asc�/� gastic tumor lysates was also confirmed by anELISA specific for mature IL18 (Supplementary Fig. S5A). More-over, gastric hyperplasia and tumor burden were significantlyreduced in 20- to 24-week-old, but not 10- to 12-week-old,gp130F/F:Il18�/� mice lacking IL18 compared with gp130F/F mice(Fig. 6A–D; Supplementary Fig. S5B–S5E), which mimickedthe suppressed gastric tumor phenotype of gp130F/F:Asc�/� mice.In contrast, in gp130F/F mice, the genetic ablation of IL1R,leading to blockade of IL1b (and IL1a) activity, had no signi-

ficant impact on tumor burden, as evidenced by comparabletumor mass (gp130F/F, 0.130 � 0.017 g vs. gp130F/F:Il1r�/�,0.121 � 0.064 g), tumor incidence and overall stomach size in20- to 24-week-old gp130F/F and gp130F/F:Il1r�/� mice (Supple-mentary Fig. S5F–S5K). Because the long-term, daily administra-tion of tumor-bearing gp130F/F mice with the IL1R antagonistanakinra (high dose, 100 mg/kg) also had no effect on tumori-genesis, these observations suggest that IL1b does not play amajor role contributing to tumorigenesis in this model. Inaddition, treatment of human AGS and MKN1 gastric cancercells with anakinra had no effect on anchorage-dependentand -independent colony formation (Supplementary Fig. S5L).

Similar to gp130F/F:Asc�/� mice, ameliorated gastric tumori-genesis in gp130F/F:Il18�/� mice was associated with significantly

Figure 6.

IL18 promotes ASC-mediated gastrictumorigenesis in gp130F/F mice. A,Representative stomachs from 20- to24-week-old (F/F) and gp130F/F:Il18�/�

(F/F:Il18�/�) mice. Arrows,macroscopically visible tumors. Fundic(f), body (b), and antral (a) regions aredepicted. B and C, Scatter plotsdepicting total mass (g) of gastrictumors and incidence of tumors in total(B) and by size (C) from 20- to 24-week-old mice. Data are expressed as themean � SEM. � , P < 0.05;�� ,P<0.01; ��� ,P <0.001; unpaired t test.D, Representative photomicrographsshowing H&E-stained whole stomachcross-sections from 20- to 24-week-oldmice. Dotted squares, tumors. Scalebars, 1 mm. In A and D, shown is one of7 representative stomach images/genotype. E and F, Representativephotomicrographs of TUNEL-stained (E)and caspase-8-stained (F) antral gastrictumor cross-sections of 20- to 24-week-old F/F and F/F:Il18�/� mice (shown isone of 6 representative images/genotype). Arrows, TUNEL-positiveand caspase-8-positive cells. Scale bars,50 mm (E) and 20 mm (F). Graphsdepict quantitative enumeration,presented as the mean � SEM ofTUNEL-positive and caspase-8-positivecells/20 high-power fields (HPF) ingastric tumor mucosa of mice(6/genotype). �� , P < 0.01; unpairedt test.

Deswaerte et al.

Cancer Res; 78(5) March 1, 2018 Cancer Research1302

elevated numbers of apoptotic TUNELþ (Fig. 6E) and caspase-8þ

(Fig. 6F) cells, and reduced numbers of cells expressing acti-vated NF-kB (Supplementary Fig. S6A), within the mucosalepithelium. In contrast, we observed no changes in the level ofproliferation, infiltration of inflammatory cells, or expression ofcell-cycle, angiogenic, and inflammatory genes (SupplementaryFigs. S6B–S6D and S7A–S7C), in the tumor epithelium. Collec-tively, these findings support a protumorigenic role for ASC ingastric cancer that is mediated largely by IL18.

Elevated gastric epithelial IL18 expression augments gastriccancer cell growth

Consistent with elevated Asc gene expression in the gastricepithelium of gp130F/F mice (Fig. 3B), mRNA levels for Il18 orits Il18r1 receptor gene were significantly increased in EpCamþ

epithelial compared to CD45þ immune (or stroma) cells ingp130F/F gastric nontumor or tumor tissues (Fig. 7A; Supple-mentary Fig. S7D). By contrast, although Il1r expression levelswere generally unchanged in the gastric epithelial and immunecompartments, Il1b was lowly-expressed in epithelial versusimmune cells or stroma in gp130F/F gastric tumors (Fig. 7A;Supplementary Fig. S7D). In support of these findings, ELISAconfirmed that total and mature IL18 protein levels weresignificantly increased in tumor compared to matched non-tumor tissue lysates from gastric cancer patients, with total IL18levels higher than those for total IL1b (Fig. 7B). Importantly,the elevated levels of IL18, but not IL1b, positively correlatedwith increased ASC mRNA levels in gastric cancer patienttumors (Fig. 7C). Furthermore, secreted (i.e., mature) IL18protein was detected by ELISA in human gastric cancer cellline supernatants, whereas as expected, secreted IL1b proteinwas undetectable due to the absence of steady-state cellularproIL1b levels (Fig. 7D).

The therapeutic utility of targeting IL18 in human gastric cancerwas supported by the observation that treatment of AGS andMKN1 human gastric cancer cells with an IL18 neutralizingantibody suppressed anchorage-dependent and -independentcolony formation, and augmented apoptosis (Fig. 7E–G). Fur-thermore, CRISPR/Cas9-mediated ASC ablation in AGS cellssuppressed colony formation, which coincided with reducedsecretion of mature IL18 (but not IL1b) and activation of cas-pase-1 andNF-kB, and conversely elevated apoptosis (Fig. 7H–K).Notably, these findings mimic those observed upon CRISPR/Cas9-mediated knockout of caspase-1 in AGS cells (Supplemen-tary Fig. S8A–S8C), and therefore support the notion that elevatedIL18 production (downstream of ASC and caspase-1) in humangastric cancer cells promotes cell-autonomous growth, which canbe readily targeted therapeutically.

DiscussionOver recent years, studies largely restricted to melanoma,

skin and colon carcinogenesis have identified complex andcontrasting tumor-promoter and tumor-suppressor roles forthe inflammasome adaptor ASC, which reveal a functionaldependency by ASC-associated inflammasomes on cell type(i.e., myeloid vs. epithelial), tissue specificity, and disease stage(19). For instance, in experimentally-induced skin carcinogen-esis, myeloid-derived ASC expression favors tumorigenesis,whereas ASC expression in keratinocytes negates tumor forma-tion (30, 31). In addition, modulating ASC expression in

primary and metastatic human melanoma demonstratedopposing anti- and protumorigenic functions, respectively, forASC, indicating that the role of ASC in cancer can depend ondisease stage (32). With respect to the gut, ASC-deficientmice are more susceptible to azoxymethane (AOM)/dextransodium sulfate (DSS)-induced colitis-associated carcinogene-sis (CAC), thus supporting an antitumorigenic role for ASC(33). By contrast, a protumorigenic role for ASC in colorectalcancer is suggested in AhR�/� and ApcMin/þ spontaneous colo-rectal cancer models, whereby ASC deficiency amelioratestumor formation (34).

Here, in gastric cancer, we reveal a protumorigenic role forASC, whose elevated expression was observed in tumors ofapproximately 75% of intestinal-type gastric cancer patients,that is independent of gastric inflammation. Specifically, wedemonstrate that suppressed gastric tumorigenesis in gp130F/F:Asc�/� mice is associated with reduced caspase-1 activation andelevated numbers of caspase-3- and caspase-8-expressingcells within the tumor epithelium, and thus uncover a hithertounknown function for ASC inflammasomes in eliciting anantiapoptotic response in neoplastic gastric epithelial cellsinvolving caspases-3 and -8. Furthermore, our discovery thatASC ablation in gastric epithelial (cancer) cells suppresses theirgrowth potential, which correlates with reduced NF-kB activa-tion (and IL18 processing/production), along with elevatedapoptosis, suggests a role for NF-kB as a signaling facilitatorof the antiapoptotic/prosurvival function for ASC in thegastric (tumor) epithelium. Indeed, this notion is consistentwith the emergence that NF-kB promotes the initiation and/orprogression of numerous epithelial cancers, as evidenced bythe observation that during colorectal cancer, in which ASChas been assigned a protumorigenic role (34), NF-kB activationin the intestinal epithelium promotes tumorigenesis by sup-pressing apoptosis (29). Therefore, in the context of gastriccancer, our findings invoke the existence of potential signalingcross-talk between activation of the canonical ASC inflamma-some (via NF-kB) and caspase-8 apoptotic machinery to damp-en the latter, thus augmenting tumor growth. Furthermore, ourobservations revealing that the predominate NF-kB signal islocalized to epithelial and not immune cells within gastrictumors is consistent with our proposed tumor (epithelial)cell-autonomous antiapoptotic role for the ASC inflam-masome/IL18 axis, and thus provide an explanation (at leastin part) for why inflammation is not affected in the gp130F/F:Asc�/� tumors.

Our current findings also expand upon the traditional role ofASC that has been linked to cell death facilitated by canonicalinflammasome-mediated pyroptosis via caspase-1, or morerecently the formation of a noncanonical, apoptosis-inducingASC apoptosome (with AIM2 and NLRP3) via interaction withcaspase-8, independent of inflammatory-related caspase-1 acti-vation (35–37). Regarding the former, it has recently emerged thatASC inflammasome/caspase-1-dependent cleavage and activa-tion of the pore-forming effector protein, gasderminD, is a criticalevent in pyroptosis (38). Although the pathophysiological role ofgasdermin D-mediated pyroptosis, including in the context ofcancer, is ill-defined, our data presented here suggest that thenovel antiapoptotic arm of the ASC inflammasome/caspase-1axis in gastric epithelial cells, which protects against caspase-8–mediated cell death, occurs independent of gasdermin D proces-sing by the ASC inflammasome. Considering that themechanistic

ASC Inflammasomes Promote Gastric Cancer via IL18

www.aacrjournals.org Cancer Res; 78(5) March 1, 2018 1303

Deswaerte et al.

Cancer Res; 78(5) March 1, 2018 Cancer Research1304

basis governing this dual mode of ASC-mediated cell death (i.e.,apoptosis and pyroptosis) remains to be fully elucidated,our notion regarding the independent role of gasdermin D ingastric cancer (and potentially other cancers) warrants furtherinvestigation.

We also note that the above-mentioned pro-apoptotic andpro-pyroptotic functions previously assigned to ASC mostlikely explain its tumor suppressor actions in certain cancers,which have also been linked to methylation-induced silencingof ASC (39). In this respect, a recent study indicated that ASCis methylated in approximately one-third of gastric cancercases, although the histologic subtype (i.e., intestinal or dif-fuse) involved and the effect on ASC expression in the tumorsanalyzed were not reported (40).

Another key finding of our study was the identification ofIL18, but not IL1b, as a major downstream inflammasomeeffector cytokine, which propagates the growth potential ofgastric cancer cells. Previous studies have shown that stomach-specific overexpression of human IL1b in transgenic miceinduces gastric inflammation-associated carcinogenesis via therecruitment of MDSCs (7), and in human gastric cancer ele-vated gastric IL1b production is associated with inflammation(41). Notably, our demonstration that IL18, like ASC, pro-motes tumor growth independent of inflammation in gp130F/F

mice contrasts these findings for IL1b. Here, in gp130F/F gastrictumors we show that gene expression for IL18 and its IL18R1receptor, as well as for ASC, is elevated in epithelial comparedto inflammatory/immune cells, which contrasts the low andhigh gene expression levels for IL1b in epithelial and inflam-matory cells, respectively. Together with the high basal pro-duction of IL18 in supernatants of human gastric cancer celllines, along with the growth inhibitory effect of a neutralizingIL18 monoclonal antibody on human gastric cancer cells, ourfindings support a direct, cell-autonomous effect of IL18(downstream of ASC) on gastric cancer cells. We thereforepropose that the functional dichotomy between IL18 and IL1bin gastric cancer can be explained by, at least in part, thepreferential requirement of epithelial (tumor) cells for IL18,which directly acts to promote growth of the tumor epi-thelium, whereas immune cells predominantly utilize IL1bto provide a proinflammatory microenvironment that cansupport tumorigenesis.

This notion could also account for our previous observationin gp130F/F mice that genetic ablation of the MyD88 adaptorprotein, which is crucial for IL18R signaling, also suppressesgrowth of the gastric tumor epithelium independent of inflam-mation (42). Furthermore, in another gastric cancer mouse

model, Gan, in which inflammation drives gastric tumorigen-esis, the ablation of MyD88 in bone marrow–derived immunecells (thus disrupting IL1b/IL1R signaling, which also requiresMyD88) suppressed the tumor-promoting inflammatorymicroenvironment (43). It is therefore perhaps not surprisingthat this latter observation contrasts our current bone marrowchimera data in gp130F/F mice whereby myeloid cells expressinglow levels of IL18 (and ASC) do not promote gastric tumor-igenesis, suggesting that at least this cellular component of thetumor microenvironment does not influence apoptosis andtumor growth in the gp130F/F gastric cancer model.

Although we reveal here a key role for IL18 as a down-stream effector of ASC inflammasomes in gastric cancer, ourstudy also raises the intriguing question concerning the identityof the specific PRR(s), as well as the source and nature of itsagonist(s) (e.g., microbial- and/or host-derived), that com-prises the protumorigenic ASC inflammasome? In the currentabsence of definitive in vivo evidence to answer this question,which for instance requires genetic ablation of specific ASCinflammasome-associated PRRs (e.g., AIM2, NLCR4, NLRP1,NLRP3) in our gp130F/F gastric cancer model, we refer to ourin vitro data demonstrating reduced levels of secreted (mature)IL18 and activated caspase-1 proteins upon CRISPR/Cas9-mediated knock-out of ASC in cultured human gastric cancercells (Fig. 7J). These observations are suggestive of constitutiveASC inflammasome activation driven by the cell intrinsicproduction and release of inflammasome-activating agonistsduring culture. Interestingly, cultured human cancer cell linesconstitutively release host-derived damage-associated molecu-lar patterns (DAMP), for example DNA and HMGB1 (44, 45),which can activate the AIM2 and NLRP3 inflammasomes,respectively (46, 47). Therefore, it is tempting to speculate thatcultured human gastric cancer cell lines, and by analogy thegastric tumor epithelium, also have the potential to constitu-tively release such host-derived DAMPs, which promote cellulargrowth via their specific PRR-associated inflammasome.

It is also worth considering our findings with previousinvestigations into the role of IL18 via inflammasome activa-tion in gastrointestinal cancers, which have been restrictedto the AOM/DSS-induced CAC model in concert with eitherIl18�/� mice or the administration of IL18 to inflammasome-deficient (Casp1�/�) mice. These studies indicate a protectiverole for IL18 in the initiating stages of CAC, attributed to highIL18 expression levels in the intestinal epithelium, whichmaintain barrier integrity through epithelial cell proliferationand survival (48, 49). Thus, although this growth-potentiatingeffect of IL18 on the intestinal epithelium protects against

Figure 7.Augmented epithelial cell IL18 expression promotes gastric cancer cell growth. A, qPCR gene expression analysis (mean � SEM) in sorted CD45-positive(immune) or EpCam-positive (epithelial) cells isolated from at least eight gp130F/F gastric tumor (T) and nontumor (NT) tissues. � , P < 0.05; �� , P < 0.01;unpaired t test. B, ELISAs for total and mature IL18 and total IL1b proteins in gastric tumor and matched nontumor tissue lysates from 10 to 15gastric cancer patients are presented as the mean � SEM. � , P < 0.05; unpaired t test. C, Linear regression of ASC mRNA and mature IL18 and IL1bprotein levels in gastric cancer patient tumors. R ¼ Pearson correlation coefficient. D, ELISA for secreted IL18 protein in human gastric cancer cellline supernatants (24-hour culture). E and I, Flow cytometry of apoptotic Annexin-V-positive human AGS cells treated with/without anti-hIL18 mAb (E)or transduced with nontargeted control sgRNA (Ctl) and ASC sgRNA (KO; I). �, P < 0.05; �� , P < 0.01; unpaired t test. F–H, Representative images (1 of 6/group) showing colony formation of AGS (F) and MKN1 (G) cells treated with anti-hIL18 mAb, and AGS Ctl and ASC KO (H) cells. Graphs depictcolony number/well (6 wells/group) expressed as the mean � SEM. �� , P < 0.01; ���� , P < 0.0001; unpaired t test. In E–I, graphs show data (n ¼ 6experiments/group) that are presented as the mean � SEM. J and K, Immunoblots with the indicated antibodies on cell culture supernatants (J) and celllysates (J and K) from AGS Ctl and ASC KO cells cultured for 24 hours. Shown are two independent ASC KO clones.

ASC Inflammasomes Promote Gastric Cancer via IL18

www.aacrjournals.org Cancer Res; 78(5) March 1, 2018 1305

colonic tumorigenesis, our current findings indicate that, con-versely, in the gastric epithelium IL18 suppresses apoptosisto promote gastric tumorigenesis. In this regard, our currentstudy reveals tissue-specific activities of IL18, thus supportingthe pleiotropic actions previously ascribed to IL18 on themigration and proliferation of human gastric cancer cell lines(9, 12, 13).

In summary, our current study suggests the existence ofa novel protumorigenic ASC/IL18/NF-kB signaling axis thataugments gastric epithelial cell survival in gastric cancer.Despite the recent development of numerous small-moleculeinhibitors against individual ASC inflammasome components,namely NLRP3, NLRP1, NLRC4, and AIM2 (50), their speci-ficity and efficacy as anticancer agents in vivo will requirerigorous evaluation in numerous cancer models that take intoaccount the tissue-specific and cell-type contexts that define themultifaceted activities of ASC in cancer. Furthermore, in gastriccancer, the potential application of such inhibitors will beinfluenced by the future identification of specific PRRs, whichcomprise the disease-associated ASC inflammasome. In light ofour findings, and given the current lack of preclinical-validatedinhibitors that can block the actions of ASC, therapies thatneutralize the biological activity of IL18 may serve as effectiveweapons against gastric cancer characterized by dysregulatedASC inflammasome-driven IL18 production.

Disclosure of Potential Conflicts of InterestC. Girard reports receiving a commercial research grant from salary sup-

ported by an unrestricted grant from AB2 Bio S.A., Lausanne, Switzerland. C.Gabay reports receiving other commercial research support from AB2 Bio and isa consultant/advisory board member of AB2 Bio. No potential conflicts ofinterest were disclosed by the other authors.

Authors' ContributionsConception and design: V. Deswaerte, T. Putoczki, B.J. JenkinsDevelopment of methodology: V. Deswaerte, A. West, T. Putoczki, B.J. JenkinsAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): V. Deswaerte, P. Nguyen, A. West, A.F. Browning,L. Yu, J. Balic, C. Girard, H. Oshima, K.Y. Fung, H. Tye, M. Najdovska,M. Oshima, C. Gabay, T. Putoczki, B.J. JenkinsAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): V. Deswaerte, P. Nguyen, A. West, A.F. Browning,L. Yu, T. Livis, H. Tye, M. Najdovska, C. Gabay, T. Putoczki, B.J. JenkinsWriting, review, and/or revision of themanuscript: P. Nguyen, A.F. Browning,C. Girard, M. Ernst, C. Gabay, T. Putoczki, B.J. JenkinsAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): T. Livis, A. Preaudet, M. Najdovska, M. Ernst,T. Putoczki, B.J. JenkinsStudy supervision: T. Putoczki, B.J. JenkinsOther (performing experiments and analyzing data): S.M. Ruwanpura

AcknowledgmentsWe thank R. Smith and P. Bouillet for comments, K. Fitzgerald (University

of Massachusetts Medical School, Worcester, MA) for providing Asc�/� mice,and C. Nold (Hudson Institute of Medical Research, Melbourne, Australia)for providing anakinra. This work was funded by the National Health andMedical Research Council (NHMRC) of Australia (to B. Jenkins), and theOperational Infrastructure Support Program by the Victorian Government ofAustralia. P. Nguyen, A. Browning, and J. Balic were supported by AustralianPostgraduate Awards from the Australian Government, and A. West wassupported by an NHMRC Early Career Fellowship. T. Putoczki was supportedby a Victorian Cancer Agency Mid-Career Fellowship. B. Jenkins was sup-ported by an NHMRC Senior Medical Research Fellowship.

The costs of publication of this articlewere defrayed inpart by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received June 26, 2017; revised October 26, 2017; accepted December 19,2017; published OnlineFirst December 27, 2017.

References1. Mantovani A, Allavena P, Sica A, Balkwill F. Cancer-related inflammation.

Nat Med 2008;454:436–44.2. Fox J, Wang TC. Inflammation, atrophy, and gastric cancer. J Clin Invest

2007;117:60–9.3. Ferlay J, Soerjomataram I, Dikshit R, Eser S, Mathers C, Rebelo M,

et al. Cancer incidence and mortality worldwide: sources, methodsand major patterns in GLOBOCAN 2012. Int J Cancer 2015;136:359–86.

4. Uemura N, Okamoto S, Yamamoto S, Matsumura N, Yamaguchi S,Yamakido M, et al. Helicobacter pylori infection and the development ofgastric cancer. N Engl J Med 2001;345:784–9.

5. El-Omar E, Carrington M, Chow WH, McColl KE, Bream JH, Young HA,et al. Interleukin-1 polymorphisms associated with increased risk ofgastric cancer. Nature 2000;404:398–402.

6. El-Omar E, Rabkin CS, Gammon MD, Vaughan TL, Risch HA, SchoenbergJB, et al. Increased risk of noncardia gastric cancer associated with proin-flammatory cytokine gene polymorphisms. Gastroenterology 2003;124:1193–201.

7. Tu S, Bhagat G, Cui G, Takaishi S, Kurt-Jones EA, Rickman B, et al. Over-expression of interleukin-1beta induces gastric inflammation and cancerand mobilizes myeloid-derived suppressor cells in mice. Cancer Cell2009;14:408–19.

8. Fabbi M, Carbotti G, Ferrini S. Context-dependent role of IL-18 incancer biology and counter-regulation by IL18BP. J Leukoc Biol 2015;97:665–75.

9. Kang JS, Bae SY, Kim HR, Kim YS, Kim DJ, Cho BJ, et al. Interleukin-18increases metastasis and immune escape of stomach cancer via the down-regulation of CD70 and maintenance of CD44. Carcinogenesis 2009;30:1987–96.

10. Haghshenas MR, Hosseini SV, Mahmoudi M, Saberi-Firozi M, Farjadian S,Ghaderi A. IL-18 serum level and IL-18 promoter gene polymorphism inIranian patients with gastrointestinal cancers. J Gastroenterol Hepatol2009;24:1119–22.

11. Thong-Ngam D, Tangkijvanich P, Lerknimitr R, Mahachai V, Theamboon-lers A, Poovorawan Y. Diagnostic role of serum interleukin-18 in gastriccancer patients. World J Gastroenterol 2006;12:4473–7.

12. KimKE, SongH, KimTS, YoonD, KimCW,Bang SI, et al. Interleukin-18 is acritical factor for vascular endothelial growth factor-enhancedmigration inhuman gastric cancer cell lines. Oncogene 2007;26:1468–76.

13. Majima T, Ichikura T, Chochi K, Kawabata T, Tsujimoto H, Sugasawa H,et al. Exploitation of interleukin-18 by gastric cancers for their growth andevasion of host immunity. Int J Cancer 2006;118:388–95.

14. Drexler SK, Yazdi AS. Complex roles of inflammasomes in carcinogenesis.Cancer J 2013;19:468–72.

15. Latz E, Xiao TS, Stutz A. Activation and regulation of the inflammasomes.Nat Rev Immunol 2013;13:397–411.

16. Stutz A, Golenbock DT, Latz E. Inflammasomes: too big to miss. J ClinInvest 2009;119:3502–11.

17. Broderick L, De Nardo D, Franklin BS, Hoffman HM, Latz E. The inflam-masomes and autoinflammatory syndromes. Annu Rev Pathol 2015;10:395–424.

18. Palomo J, Dietrich D, Martin P, Palmer G, Gabay C. The interleukin (IL)-1cytokine family-balance between agonists and antagonists in inflamma-tory diseases. Cytokine 2015;76:25–37.

19. Kolb R, Liu GH, Janowski AM, Sutterwala FS, ZhangW. Inflammasomes incancer: a double-edged sword. Protein Cell 2014;5:12–20.

20. Jenkins BJ, Grail D, Nheu T, Najdovska M, Wang B, Waring P, et al.Hyperactivation of Stat3 in gp130 mutant mice promotes gastric

Deswaerte et al.

Cancer Res; 78(5) March 1, 2018 Cancer Research1306

hyperproliferation and desensitizes TGF-beta signaling. Nat Med 2005;11:845–52.

21. Ozoren N, Masumoto J, Franchi L, Kanneganti TD, Body-Malapel M,Ert€urk I, et al. Distinct roles of TLR2 and the adaptor ASC in IL-1beta/IL-18 secretion in response to Listeria monocytogenes. J Immunol 2006;176:4337–42.

22. Labow M, Shuster D, Zetterstrom M, Nunes P, Terry R, Cullinan EB, et al.Absence of IL-1 signaling and reduced inflammatory response in IL-1 type Ireceptor-deficient mice. J Immunol 1997;159:2452–61.

23. Zwijnenburg PJ, van der Poll T, Florquin S, Akira S, Takeda K, Roord JJ, et al.Interleukin-18 gene-deficient mice show enhanced defense and reducedinflammation during pneumococcal meningitis. J Neuroimmunol 2003;138:31–7.

24. Tye H, Kennedy CL, Najdovska M, McLeod L, McCormack W, Hughes N,et al. STAT3-driven upregulation of TLR2 promotes gastric tumorigenesisindependent of tumor inflammation. Cancer Cell 2012;22:466–78.

25. Kikkawa S, Matsumoto M, Shida K, Fukumori Y, Toyoshima K, Seya T.Human macrophages produce dimeric forms of IL-18 which can bedetected with monoclonal antibodies specific for inactive IL-18. BiochemBiophys Res Commun 2001;281:461–7.

26. Girard C, Rech J, Brown M, Allali D, Roux-Lombard P, Spertini F, et al.Elevated serum levels of free interleukin-18 in adult-onset Still's disease.Rheumatology 2016;55:2237–47.

27. Ruwanpura SM, McLeod L, Dousha LF, Seow HJ, Alhayyani S, Tate MD,et al. Therapeutic targeting of the IL-6 trans-signalling/mTORC1 axisin pulmonary emphysema. Am J Respir Crit Care Med 2016;194:1494–505.

28. Ran FA, Hsu PD, Wright J, Agarwala V, Scott DA, Zhang F. Genomeengineering using the CRISPR-Cas9 system. Nat Protoc 2013;8:2281–308.

29. Greten FR, Eckmann L, Greten TF, Park JM, Li ZW, Egan LJ, et al. IKKbetalinks inflammation and tumorigenesis in a mouse model of colitis-asso-ciated cancer. Cell 2004;118:285–96.

30. Drexler SK, Bonsignore L, Masin M, Tardivel A, Jackstadt R, Hermeking H,et al. Tissue-specific opposing functions of the inflammasome adaptor ASCin the regulation of epithelial skin carcinogenesis. Proc Natl Acad Sci U S A2012;109:18384–9.

31. Chow MT, Sceneay J, Paget C, Tardivel A, Jackstadt R, Hermeking H, et al.NLRP3 suppresses NK cell-mediated responses to carcinogen-inducedtumors and metastases. Cancer Res 2012;72:5721–32.

32. Liu W, Luo Y, Dunn JH, Norris DA, Dinarello CA, Fujita M. Dual roleof apoptosis-associated speck-like protein containing a CARD (ASC)in tumorigenesis of human melanoma. J Invest Dermatol 2013;133:518–27.

33. Allen I, TeKippe EM, Woodford RM, Uronis JM, Holl EK, Rogers AB, et al.The NLRP3 inflammasome functions as a negative regulator of tumori-genesis during colitis-associated cancer. J Exp Med 2010;207:1045–56.

34. Ikuta T, Kobayashi Y, Kitazawa M, Shiizaki K, Itano N, Noda T, et al. ASC-associated inflammation promotes cecal tumorigenesis in arylh hydrocar-bon receptor-deficient mice. Carcinogenesis 2013;34:1620–7.

35. Chung H, Vilaysane A, Lau A, Stahl M, Morampudi V, Bondzi-Simpson A,et al. NLRP3 regulates a non-canonical platform for caspase-8 activationduring epithelial cell apoptosis. Cell Death Differ 2016;23:1331–46.

36. Sagulenko V, Thygesen SJ, Sester DP, Idris A, Cridland JA, Vajjhala PR, et al.AIM2 and NLRP3 inflammasomes activate both apoptotic and pyroptoticdeath pathways via ASC. Cell Death Differ 2013;20:1149–60.

37. HasegawaM, Kawase K, Inohara N, Imamura R, YehWC, Kinoshita T, et al.Mechanism of ASC-mediated apoptosis: bid-dependent apoptosis in typeII cells. Oncogene 2007;26:1748–56.

38. Man SM, Karki R, Kanneganti TD.Molecular mechanisms and functions ofpyroptosis, inflammatory caspases and inflammasomes in infectious dis-eases. Immunol Rev 2017;277:61–75.

39. Guan X, Sagara J, Yokoyama T, Koganehira Y, Oguchi M, Saida T, et al.ASC/TMS1, a caspase-1 activating adaptor, is downregulated by aberrantmethylation in human melanoma. Int J Cancer 2003;107:202–8.

40. WuL, ZhangC,WangX,DingX,Deng J, LiangH.MethylationofASC/TMS1promoter is associated with poor prognosis of patients with gastric cancer.Clin Transl Oncol 2016;18:296–303.

41. Deans DA, Wigmore SJ, Gilmour S, Paterson-Brown S, Ross JA, Fearon KC.Elevated tumor interleukin-1beta is associated with systemic inflamma-tion: amarker of reduced survival in gastro-oesophageal cancer. Br J Cancer2006;95:1568–75.

42. Kennedy CL, Najdovska M, Tye H, McLeod L, Yu L, Jarnicki A, et al.Differential role of MyD88 and Mal/TIRAP in TLR2-mediated gastrictumourigenesis. Oncogene 2014;33:2540–46.

43. Maeda Y, Echizen K, Oshima H, Yu L, Sakulsak N, Hirose O, et al. Myeloiddifferentiation factor 88 signaling in bone marrow-derived cells promotesgastric tumorigenesis by generation of inflammatory microenvironment.Cancer Prev Res (Phila) 2016;9:253–63.

44. Bronkhorst AJ, Wentzel JF, Aucamp J, van Dyk E, du Plessis L, Pretorius PJ.Characterization of the cell-free DNA released by cultured cancer cells.Biochim Biophys Acta 2016;1863:157–65.

45. Jube S, Rivera ZS, BianchiME, Powers A,Wang E, Pagano I, et al. Cancer cellsecretion of the DAMP proteinHMGB1 supports progression inmalignantmesothelioma. Cancer Res 2012;72:3290–301.

46. Hornung V, Ablasser A, Charrel-Dennis M, Bauernfeind F, Horvath G,Caffrey DR, et al. AIM2 recognizes cytosolic dsDNA and forms a caspase-1-activating inflammasome with ASC. Nature 2009;458:514–8.

47. Chi W, Chen H, Li F, Zhu Y, Yin W, Zhuo Y. HMGB1 promotes theactivation of NLRP3 and caspase-8 inflammasomes via NF-kB pathwayin acute glaucoma. J Neuroinflammation 2015;12:137.

48. Salcedo R, Worschech A, Cardone M, Jones Y, Gyulai Z, Dai RM, et al.MyD88-mediated signaling prevents development of adenocarcinomas ofthe colon: role of interleukin 18. J Exp Med 2010;207:1625–36.

49. Williams TM, LeethRA, RothschildDE,Coutermarsh-Ott SL,McDanielDK,Simmons AE, et al. The NLRP1 inflammasome attenuates colitis andcolitis-associated tumorigenesis. J Immunol 2015;194:3369–80.

50. GuoH,Callaway JB, Ting JP. Inflammasomes:mechanismof action, role indisease, and therapeutics. Nat Med 2015;21:677–87.

www.aacrjournals.org Cancer Res; 78(5) March 1, 2018 1307

ASC Inflammasomes Promote Gastric Cancer via IL18