Inflammation-Dependent IL18 Signaling Restricts Hepatocellular Carcinoma Growth by Enhancing the Accumulation and Activity of Tumor-Infiltrating Lymphocytes

Published OnlineFirst February 18, 2016; DOI: 10.1158/0008-5472.CAN-15-1548 Cancer Tumor and Stem Cell Biology Research

Inﬂammation-Dependent IL18 Signaling Restricts Hepatocellular Carcinoma Growth by Enhancing the Accumulation and Activity of Tumor- Inﬁltrating Lymphocytes Geoffrey J. Markowitz1, Pengyuan Yang1,2,3, Jing Fu3, Gregory A. Michelotti4, Rui Chen1, Jianhua Sui5, Bin Yang2, Wen-Hao Qin3, Zheng Zhang6, Fu-Sheng Wang6, Anna Mae Diehl4, Qi-Jing Li7, Hongyang Wang3, and Xiao-Fan Wang1

Abstract

Chronic inflammation in liver tissue is an underlying cause of IL18R1 deletion increased tumor burden. Mechanistically, we hepatocellular carcinoma. High levels of inflammatory cytokine foundthatIL18exertedinflammation-dependent tumor-sup- IL18 in the circulation of patients with hepatocellular carcinoma pressive effects largely by promoting the differentiation, activ- correlates with poor prognosis. However, conflicting results have ity, and survival of tumor-infiltrating T cells. Finally, differences been reported for IL18 in hepatocellular carcinoma development in the expression of IL18 in tumor tissue versus nontumor and progression. In this study, we used tissue specimens from tissueweremorepredictiveofpatientoutcomethanoverall hepatocellular carcinoma patients and clinically relevant mouse tissue expression. Taken together, our findings resolve a long- models of hepatocellular carcinoma to evaluate IL18 expression standing contradiction regarding a tumor-suppressive role for and function. In a mouse model of liver fibrosis that recapitulates IL18 in established hepatocellular carcinoma and provide a a tumor-promoting microenvironment, global deletion of the mechanistic explanation for the complex relationship between IL18 receptor IL18R1 enhanced tumor growth and burden. Sim- its expression pattern and hepatocellular carcinoma prognosis. ilarly, in a carcinogen-induced model of liver tumorigenesis, Cancer Res; 76(8); 1–12. 2016 AACR.

Introduction Because of the well-known intimate relationship of hepatocellular carcinoma with inflammation, previous studies have dem- Liver cancer is the third most deadly cancer worldwide, and onstrated both pro and antitumorigenic roles of a variety of unlike many other cancers, the incidence of this disease and its immune cell types and mediators. Important effector cells, such most common type, hepatocellular carcinoma, is rising despite as macrophages, natural killer (NK) cells, and T cells, have been advances in surveillance and therapeutics (1, 2). Hepatocellular shown to play varying and diverse roles in regulating the multistep carcinoma derives from multiple etiologic factors, including tumorigenic process, often in a stage- and even etiology-depen- infection by hepatitis viruses, alcohol abuse, and metabolic dent fashion (6–14). Cytokines, including TNFa, IL6, IL1a, IL1b, syndrome (1). These etiologic factors induce a state of chronic TGFb, IFNg, and CCL22, have been reported to display robust and inflammation, fibrosis, and cirrhosis, which in turn leads to disparate functionalities in modulating hepatocellular carcinoma malignant transformation of hepatocytes, hepatocellular carcino- initiation and progression (7, 13, 15–20). Examination of hepa- ma development, and progression (1, 3–5). tocellular carcinoma patient samples has also indicated a tight correlation among inflammation, disease progression, and prog-

1 nosis, although these correlations may vary based on the etiology Department of Pharmacology and Cancer Biology, Duke University – Medical Center, Durham, North Carolina. 2Key Laboratory of Infection (4, 14, 21 24). and Immunity, Institute of Biophysics, Chinese Academy of Sciences, IL18 is a member of the IL1 family of cytokines (25). Originally Beijing, China. 3National Center for Liver Cancer, Second Military described as IFNg-inducing factor, its canonical function is to Medical University, Shanghai, China. 4Department of Gastroenterolo- 5 promote the production of IFNg from a variety of immune cells, gy, Duke University Medical Center, Durham, North Carolina. Biolo- þ gics Research Center, National Institute of Biological Sciences, Beijing, primarily Th1 CD4 T cells and NK cells, frequently in conjunc- China. 6Research Center for Biological Therapy, Beijing 302 Hospital, tion with IL12 (25). IL18 is produced by a variety of cells, 7 Beijing, China. Department of Immunology, Duke University Medical including but not limited to dendritic cells, macrophages, and Center, Durham, North Carolina. epithelial cells (25). Its functional production is regulated both Note: Supplementary data for this article are available at Cancer Research transcriptionally and via proteolytic processing of its precursor Online (http://cancerres.aacrjournals.org/). peptide by the inﬂammasome (25). Clinical data have shown that Corresponding Author: Xiao-Fan Wang, Duke University Medical Center, Box IL18 levels in serum are elevated in patients with chronic hepatitis 3813, Durham, NC 27710. Phone: 919-681-4861; Fax: 919-681-7152; E-mail: B, hepatitis C, and hepatocellular carcinoma, and higher levels of [email protected] circulating IL18 are correlated with worse prognosis of hepato- doi: 10.1158/0008-5472.CAN-15-1548 cellular carcinoma (26–28). However, conﬂicting experimental 2016 American Association for Cancer Research. results have been reported showing either pro or antitumorigenic

www.aacrjournals.org OF1

Markowitz et al.

functions of IL18. Multiple carcinogen-induced and spontaneous Fibrosis induction tumorigenesis studies have demonstrated a protumorigenic role Carbon tetrachloride–induced liver fibrosis. Six- to 7-week-old of IL18, whereas investigations utilizing the implantation of male WT or IL18R1 / mice were treated with either olive oil or tumor cells in combination with methods to enhance IL18 carbon tetrachloride (CCl4) diluted 1:20 in olive oil at 600 mL/kg. expression revealed potential IL-18–induced tumor suppression Mice received biweekly intraperitoneal injections for 6 weeks and (10, 29, 30). Thus, despite the established connection between its were sacrificed 2 days following the 12th injection. presence in the circulation of patients and correlated poorer prognosis, a pathologic feature frequently associated with tumor Bile duct ligation–induced liver fibrosis. Twelve-week-old male WT promoters, the role of IL18 in hepatocellular carcinoma devel- or IL18R1 / mice were anesthetized and the common bile duct opment and progression is still contentious. To address this isolated and transected between two ligations. Mice were sacri- critical question, we conducted comprehensive analyses of two ficed 14 days postsurgery. sets of patient samples in conjunction with clinically relevant mouse models to examine the role of IL-18 in hepatocellular Orthotopic implantation carcinoma. We reveal a tumor-suppressive role for IL18 that is Implantation occurred either 2 days after the 8th injection of consistent with the complex expression patterns of this cytokine in CCl4 or olive oil or immediately following bile duct ligation. Mice our patient cohorts. Differing from previous reports, we demon- were anesthetized, the abdomen opened, and the liver exposed. strate a powerful effect of endogenous IL18 signaling modulating 6 þ Hepa1-6-GFP-Luc cells (3 10 ) were suspended in 30 mL growth the accumulation and activity of both CD8 and multiple subsets – þ factor reduced Matrigel and injected into the left lobe of the liver. of CD4 T cells to effect antitumor activity. We further demon- The peritoneum and skin were subsequently sutured shut. The 9th strate that IL18 signaling serves as a key mediator of complemen- dose of CCl or olive oil occurred 2 days after surgery, and the 12th þ – 4 tary CD8 T-cell- and NK cell dependent tumor suppression. dose occurred 12 days after surgery. Mice were monitored for Finally, we show that the differences in the level of IL18 expression health and sacrificed 14 days postimplantation for both fibrotic between matched tumor and nontumor tissue provided more allograft models. reliable prognostic value than the overall levels of IL18. Taken together, our findings establish that, contrary to an expected DEN-induced carcinogenesis protumorigenic activity due to its elevated levels in circulation, Post-natal-day-15 WT or IL18R1 / mice received a single IL18 actually exerts a stage-dependent tumor-suppressive effect intraperitoneal injection of N-nitrosodiethylamine (DEN) at 25 on hepatocellular carcinoma progression, primarily through the mg/kg. Mice were sacrificed 6 or 12 months postadministration. antitumor activities of tumor-infiltrating lymphocytes. Alternatively, starting one week post-DEN treatment, mice also received weekly doses of 500 mL/kg CCl4 via intraperitoneal fi Materials and Methods injection. These mice received 22 doses of CCl4 and were sacri ced Patient samples 2 days after the last injection. Patient samples were obtained following informed consent according to established protocols approved by the Ethics Com- Sample harvest and preparation Livers and spleens were harvested and either postfixed in 10% mittees of the Eastern Hepatobiliary Surgery Hospital (Shanghai, China) and the 302 Military Hospital (Beijing, China). Details on phosphate-buffered formalin at 4 C on an orbital shaker for 1 day fi patient collection, characterization, and survival categorization, before washing with 70% ethanol and embedding in paraf n fl as well as usage of the different specimens for IL18 staining, serum blocks or processed for ow cytometry/FACS. Detailed tissue fl quantification, and mRNA expression, are detailed in the Sup- processing for ow cytometry, sorting, and staining protocols are plementary Material. in the Supplementary Material.

Mice Liver and spleen as percentage of body weight All experimental procedures described were approved by the Mice were euthanized, weighed, and their livers and spleens Duke University Animal Care and Use Committee. C57BL/6 [wild dissected and weighed. Percent of body weight was calculated as type (WT)] breeding mice were purchased from The Jackson 100% (weight of liver or spleen)/(weight of the entire mouse). Laboratory. IL18R1 / mice backcrossed to C57BL/6 were provided by Dr. Yiping Yang (Duke University Medical Center, Histology Durham, NC). Tissue evaluated by histology was fixed in 10% phosphate- buffered formalin and embedded in paraffin blocks. Five microm- Reagents eter–thick sections were stained for hematoxylin and eosin and For a list of antibodies, stains, and cytokines used, clone and/or picrosirius red using standard protocols. catalog numbers, and companies from which they were purchased, see Supplementary Table S1. For qPCR primers and CBCs and serum evaluation shRNA constructs, see Supplementary Table S2. Hepa1-6 cells Blood was collected from the inferior vena cava from anesthe- [Hepa1-6][ATCC CRL-1830] were purchased directly from ATCC, tized mice at sacrifice. For complete blood counts (CBC), blood which were validated by short tandem repeat profiling and was stored in EDTA-coated tubes and examined on an Abbot passaged for fewer than 6 months after receipt, with frozen stocks Cell-Dyn 3700. For serum evaluation, blood was incubated at of both na€ve and GFP-labeled luciferase reporter–expressing room temperature for 10 minutes and then centrifuged at 3,000 Hepa1-6 cells (Hepa1-6-GFP-Luc) being thawed prior to use. rpm for 10 minutes, following which serum was aspirated, trans- Recombinant IL18BP was generated as described in ref. 31. ferred to a new tube, and stored at 80C until assay. Alanine

OF2 Cancer Res; 76(8) April 15, 2016 Cancer Research

Inﬂammatory IL18 Signaling Restricts HCC Progression

aminotransferase (ALT) and aspartate aminotransferase (AST) microenvironment resembles the diseased tissue from which were quantitatively evaluated by colorimetry using kits pur- a tumor develops in most hepatocellular carcinoma patients. chased from Biotron Diagnostics according to the manufac- To do this, we first established a mouse model combining turer's instructions. Circulating mature IL18 was detected in the CCl4-induced liver fibrosis with orthotopic implantation of serum of mice using a Mouse IL-18 ELISA Kit (Medical & tumor cells (Fig. 2A; ref. 34). This model induced robust fibrosis Biological Laboratories Co.). in WT mice, as demonstrated by increases in circulating amino- transferases, hepatic stellate cell activation, and collagen depo- qPCR analysis of T cells sition (Supplementary Fig. S2). Induction of inflammatory liver At sacrifice, tissue was harvested and samples processed as fibrosis enhanced tumor growth, with increased infiltration and þ þ described before. Samples were stained, and TCRb CD4 CD25 , altered immune composition in nontumor liver and tumor þ þ þ þ þ TCRb CD4 CD25 , and TCRb CD8 populations were collect- compared with tumor engraftment without fibrosis induction ed. RNA was extracted from isolated cells, reverse transcribed (34). IL18 is produced by many constituents of the tumor using iScript, and samples probed for targets of interest. microenvironment (25). We therefore performed immunohis- tochemical staining in several hepatocellular carcinoma models Cell differentiation and stimulation assays and found that multiple cell populations expressed IL18 (Sup- – Na€ve T cells were harvested from lymph nodes of WT or plementary Fig. S3A S3D). We further isolated populations IL18R1 / mice, sorted by commercially available untouched from the tumor and obtained similar results as shown in other isolation methods and subjected to differentiation and stimula- pathologic systems. Endothelial, myeloid, and tumor cells produced IL18, with myeloid populations dominating production, tion assays using established protocols. Information detailing þ þ these protocols can be found in the Supplementary Material. whereas lymphoid cells, such as CD4 T cells, CD8 Tcells,and NK cells, expressed higher levels of IL18R1 (Supplementary Fig. Statistical analysis S3E). Thus, to examine the overall effects of IL18 signaling in the tumor microenvironment, we employed a mouse strain with Data are presented as mean SEM. Statistics were calculated / using GraphPad Prism; tests are described in each figure legend. constitutive whole-body IL18R1 deletion (IL18R1 ;ref.35). Significance was set at P < 0.05. Consistent with a tumor-suppressive role for IL18 signaling, orthotopically implanted tumors grew to greater sizes in / Results IL18R1 mice than in WT mice (Fig. 2B). Importantly, this enhancement in tumor burden was observed only in mice A complex relationship between prognosis of hepatocellular treated with CCl4, not in those treated with the vehicle control, carcinoma patients and expression of IL18 signaling olive oil, demonstrating a reliance on the inflammatory context components for IL18 signaling to robustly control tumor growth (Fig. 2B). Previous studies indicated that IL18 is frequently elevated in the Interestingly, IL18R1 / mice displayed a slightly less severe circulation and liver tissue of patients with chronic liver diseases fibrotic phenotype compared with WT animals (Supplementary compared with healthy individuals, and a higher level is corre- Fig. S2), indicating the tumor-suppressive nature of IL18 sig- lated with poor prognosis of hepatocellular carcinoma (26–28, naling may not be shared at other disease stages. 31–33). To further explore this topic, we first examined the To validate the antitumor effect of IL18 signaling across mul- prevalence of IL18 in the serum of patients with liver disease and tiple models, we examined liver tumorigenesis induced by a found elevated levels compared with those in healthy individuals commonly used chemical carcinogen, DEN (12). Importantly, (Fig. 1A). We then determined the relationship between prognosis this model has been shown to provoke local and systemic and expression of IL18 in tissue specimens from a cohort of 138 immune responses, particularly in later stages of tumorigenesis hepatocellular carcinoma patients (described in Supplementary (12). Tumor burden at late stages in tumorigenesis was enhanced Table S3). Consistent with earlier reports, higher IL18 expression, in IL18R1 / mice compared with WT animals (Fig. 2C). Taken either intratumoral or peritumoral, was associated with poor together, these data suggest that IL18 signaling plays an important prognosis (Fig. 1B and C; stratification based on staining density inflammation-dependent tumor-suppressive role. shown in Supplementary Fig. S1), although this point was less clear with patients displaying intermediate levels of IL18. How- IL18 signaling regulates composition of tumor-infiltrating ever, when mRNA samples from a separate set of 26 chronic T-cell populations þ hepatitis B virus–positive (HBV ) patients with metastatic hepa- IL18 affects multiple cell types, which may affect tumor growth tocellular carcinoma were examined, we found significantly lower (25). To identify which populations carried out IL18 signaling– expression of both IL18 (Fig. 1D) and its receptor IL18R1 (Fig. 1E) mediated tumor suppression, we first examined circulating in primary and metastatic tumor tissue compared with matched immune cells. Na€ve WT and IL18R1 / mice did not show nontumor tissue samples. These data validate that elevated IL18 significant differences in circulating immune population propor- levels in liver tissue of hepatocellular carcinoma patients could be tions (Supplementary Fig. S4). In contrast, upon CCl4 treatment detrimental for survival. However, in comparison with the and orthotopic tumor cell implantation, we found increased inflamed peritumoral tissue, the expression level of IL18 and its neutrophil and decreased lymphocyte proportions in IL18R1 / receptor was significantly lower within the tumor itself. mice compared with WT animals (Supplementary Fig. S4B). This change associates with enhanced tumor burden in IL18R1 / IL18 signaling exhibits suppressive effects on tumor growth mice, and increased circulating neutrophil-to-lymphocyte ratios These seemingly contradictory results prompted us to address have been reported to be indicative of worse patient prognosis the functional relevance of IL18 in hepatocellular carcinoma, (36). The decreased lymphocyte counts also present a plausible utilizing a more clinically relevant model in which the liver major mechanism for impaired antitumor function.

www.aacrjournals.org Cancer Res; 76(8) April 15, 2016 OF3

Markowitz et al.

ADE4,000 3,500 NS NS 3,000 4 5 2,500 2,000 2 1,200 0 0 800

IL18 (pg/mL) –2 400 –5

0 –4

–6 –10 Relative IL18 expression ( n = 4) ( n = 3) ( n = 9) ( n = 5) ( n = 8) Relative IL18R1 expression + + + + + HBV Healthy ( n = 6) Nontumor Non-HBV ( n = 6) Hepatitis ( n = 1) Alcoholic ( n = 3) Alcoholic ( n = 5) Nontumor + Primary tumor Autoimmune ( n = 3) Autoimmune ( n = 5) Chronic HCV Chronic HBV Chronic HBV Chronic HCV Primary tumor Metastatic tumor Metastatic tumor Hepatitis Liver HCC

Acute HBV cirrhisis

B High IL18 High IL18 intratumoral intratumoral 100 100 Middle IL18 Middle IL18 intratumoral intratumoral 80 Low IL18 80 Low IL18 intratumoral intratumoral 60 60

40 40

20 20 Overall survival (%) P = 0.0340 P = 0.0399 Disease-free survival (%) 0 0 0 20 40 60 80 100 120 0 20 40 60 80 100 120 C High IL18 High IL18 peritumoral peritumoral 100 100 Middle IL18 Middle IL18 peritumoral peritumoral 80 Low IL18 80 Low IL18 peritumoral peritumoral 60 60

40 40

20 20 Overall survival (%) P = 0.0151 P = 0.0078 0 Disease-free survival (%) 0 0 20 40 60 80 100 120 0 20 40 60 80 100 120 Months Months

Figure 1. Elevated IL18 correlates with poor prognosis, but expression of IL18 and IL18R1 is often decreased within the tumor from the same patient. A, circulating plasma IL18 measured by ELISA. B and C, Kaplan–Meier curves for overall (left) and disease-free survival (right) of the cohort of 138 patients stratiﬁed into the highest, middle, or lowest third of patients based on IL18 staining density either in tumor (B) or in peritumoral (C) tissue as shown in Supplementary Fig. S1. D and E, mRNA expression of IL18 (D) and IL18R1 (E) from matched non-, primary, and metastatic tumor tissue. Samples were normalized relative to average expression levels in all nontumor tissue samples and displayed as log2 (fold change to average expression in nontumor tissue). Statistics: one-way ANOVA (P < 0.0001), followed by unpaired t tests comparing each group with healthy controls, n indicated under each column (A); log-rank Mantel–Cox tests, n ¼ 46 per group (B and C); one-way ANOVA with Tukey multiple comparisons test, n ¼ 26 matched samples per group (D and E). NS, nonsigniﬁcant. , P < 0.05; , P < 0.01; , P < 0.001; , P < 0.0001. HCC, hepatocellular carcinoma; HCV, hepatitis C virus.

We then examined the immune composition of the fibrotic S8A). Composition in the tumor demonstrated an accumula- þ þ liver and tumor allograft. We did not find significant differences tion of CD8 T cells and consequent decrease in the CD4 T-cell þ in numbers of NK cells, NKT cells, CD11b monocytes, Gr-1– proportion in WT mice compared with na€ve liver tissue, which expressing monocyte subsets, or bulk T cells in either liver or was not observed in IL18R1 / mice (Fig. 3A and C and tumor between genotypes (Supplementary Figs. S5–S7). How- Supplementary Figs. S7B and S8B). Importantly, although there þ þ ever, we found that T-cell composition was different, specifically was a marginal offset in the CD4 /CD8 T-cell ratio in the na€ve in the tumor. In the diseased nontumor liver, there was a spleens of WT and IL18R1 / mice, these differences were þ nonsignificant decrease in CD8 T cells, and a small yet signif- nonsignificant in the spleens of tumor-bearing mice (Fig. 3D þ icant increase in CD4 T cells in the IL18R1 / mice compared and E and Supplementary Fig. S8C). We also confirmed this with the WT animals, which was not observed in livers from T-cell skewing in two additional models: both bile duct ligation na€ve mice (Fig. 3A and B and Supplementary Figs. S7A and with orthotopic tumor cell implantation and tumorigenesis

OF4 Cancer Res; 76(8) April 15, 2016 Cancer Research

Inﬂammatory IL18 Signaling Restricts HCC Progression

A CCI4 Administration: D26, D23, D19 D16, D12, D9 D5, D2 D2 D5 D9 D12

D0 D14 Orthotopic Sacrifice

NS B Olive oil CCl4 WT IL18R1–/– WT IL18R1–/– 25

10 of body weight 5 Liver as percentage 0 Cells – –+ + ––+ +

CCI4 –– ++ –– ++ IL18R1 +/+ +/+ +/+ +/+ –/– –/– –/– –/–

C NS 20

15 WT

of body weight 5 Liver as percentage 0 IL18R1–/– Months 6 6 12 12 IL18R1 +/+ –/– +/+ –/–

Figure 2. IL18 signaling exerts a tumor-suppressive role on the growth of liver tumor cells orthotopically implanted in mice with a fibrotic liver and at late stages in tumorigenesis. A, experimental schematic illustrating timing of the 12 doses of CCl4 administered, orthotopic implantation, and tissue harvest. B, tumor growth (liver as a percentage of body weight); a sham Matrigel plug or Hepa1-6 cells expressing a GFP-luciferase vector (Hepa1-6-GFP-Luc) was implanted intoWTor IL18R1-deficient C57Bl/6 mice treated with olive oil or CCl4; representative images on the right. C, tumor burden (liver as percentage of body weight) in WT or IL18R1-deficient mice 6 and 12 months after DEN administration; livers harvested 12 months post-DEN treatment on right. Statistics: one-way ANOVA with Tukey multiple comparisons test, n ¼ 3–8 for olive oil-treated mice, n ¼ 33 for CCl4-treated mice pooled over 9 experiments (B); one-way ANOVA with Sidak multiple comparisons test, n ¼ 3–8 per group (C). NS, nonsignificant. , P < 0.05; , P < 0.01.

combining administration of DEN with CCl4 administration IL18 signaling regulates T-cell differentiation, activation, þ yielded comparatively increased tumor-infiltrating CD4 T-cell proliferation, and survival þ proportions and decreased tumor-infiltrating CD8 T-cell pro- As the most relevant difference between the WT and IL18R1 / þ þ portions in IL18R1 / mice (Supplementary Fig. S9). Together, tumor immune composition was the CD4 /CD8 T-cell ratio, these data indicate IL18R1 deletion leads significantly reduced and T cells are well-known mediators of tumor suppression, we þ tumor-infiltrating CD8 T cells and proportionally increased proceeded to determine the mechanism through which those þ tumor-infiltrating CD4 T cells. cells affect allograft tumor growth. We found minor differences

www.aacrjournals.org Cancer Res; 76(8) April 15, 2016 OF5

Markowitz et al.

CCl nontumor liver CCl tumor A Naïve liver BC4 4 NS NS NS NS NS NS 100 100 100

80 80 80

60 60 60

40 40 40

Percent of gate Percent 20 of gate Percent 20 of gate Percent 20

0 0 0 + + + + + + + + + + + + + + + + + +

TCRb TCRb TCRb –/– –/– –/– + in TCRb+ in TCRb+ in TCRb+ in TCRb + in TCRb in TCRb+ in TCRb in TCRb + in TCRb + in TCRb WT TCRb WT TCRb + + WT TCRb + in TCRb + in TCRb CD8 CD4 CD8 CD4 CD8 CD4 IL18R1 –/– –/– IL18R1 –/– –/– IL18R1 –/– –/– WT CD8 WT CD4 WT CD8 WT CD4 WT CD8 WT CD4

IL18R1 IL18R1 IL18R1 IL18R1 IL18R1 IL18R1 CCl spleen DENaïve spleen 4 NS NS NS NS 100 100

80 80

60 60

40 40 Percent of gate Percent 20 of gate Percent 20

0 0 + + + + + + + + + + + +

TCRb TCRb –/– –/– in TCRb in TCRb + in TCRb in TCRb+ in TCRb in TCRb WT TCRb + + in TCRb+ + in TCRb WT TCRb + + CD8 CD4 CD8 CD4 IL18R1 –/– –/– IL18R1 –/– –/– WT CD8 WT CD4 WT CD8 WT CD4

IL18R1 IL18R1 IL18R1 IL18R1

Figure 3. þ þ þ þ þ T-cell populations that infiltrated the tumor are skewed by loss of IL18R1. A–E, TCRb ,TCRb CD8 ,andTCRb CD4 cells in na€ve WT and IL18R1-deficient mice or mice treated with CCl4 and tumor implantation were quantified in na€ve liver tissue (A), nontumor liver tissue from tumor-bearing mice (B), tumor tissue from tumor- bearing mice (C), na€ve spleen (D), or spleen from tumor-bearing mice (E). Statistics: one-way ANOVA with Sidak multiple comparisons test, n ¼ 9–12 per group fornontumorliversamplespooledfromthreeexperiments,n ¼ 12–15 per group for tumor samples pooled from four experiments, n ¼ 6–8 per group for spleen samples pooled from two experiments, n ¼ 4–5perna€ve group pooled over two experiments (A–E). NS, nonsignificant. , P < 0.05; , P < 0.01; , P < 0.001; , P < 0.0001.

þ in T-cell activation and checkpoint suppression and no difference and D). We also observed a trend for reduced TNFa IL2 (P ¼ þ þ þ in immunosuppressive T regulatory cell accumulation between 0.1707) and TNFa IL2 (P ¼ 0.0896) cells in IL18R1 / CD4 T WT and IL18R1 / mice (Supplementary Figs. S10–S12). On the cells compared with WT and detected no differences in granzyme, basis of current literature, these differences in T-cell activation and T-bet, or Eomes production (Supplementary Fig. S13). þ PD-1 expression would have resulted in enhanced tumor sup- We also examined the functionality of IL18 signaling in CD8 T pression in the IL18R1 / mice, contrary to the phenotype cells. Similarly, we determined the expression of relevant cyto- observed here. We then examined the functionality of tumor- kines, enzymes, and transcription factors in tumor-infiltrating þ þ þ infiltrating conventional CD4 T cells. CD4 CD25 T cells were CD8 T cells. Expressions of Prf1, GzmA, and GzmB were mod- þ isolated from the tumor and subjected to qPCR analysis for erately increased in IL18R1 / CD8 T cells compared with WT, cytokine, enzyme, and relevant transcription factor expression. whereas IFNg and Eomes were mildly decreased, suggesting that þ þ CD4 T cells from IL18R1 / mice expressed significantly less IL2, on a per-cell basis, IL18R1 / CD8 T cells may be slightly more TNFa, and IL17a compared with WT, suggesting diminished potent killers than WT (Fig. 4B). Closer examination via flow functional potential (Fig. 4A), although GzmB expression was cytometry also demonstrated significant differences in IFNg pro- þ þ enhanced. IL2 stimulates CD8 T-cell growth and differentiation duction in tumor-infiltrating IL18R1 / CD8 T cells compared into memory cells, whereas TNFa mediates antitumor activity with WT, but no differences in granzyme production (Fig. 4C–E (37, 38). Importantly, both Th1 and Th17 cells have been shown and Supplementary Fig. S13). We also observed a significant þ to possess antitumor functionality (39, 40). We validated the reduction in T-bet Eomes and a trend for a reduction in T- þ þ þ expression profiles of these factors in tumor-infiltrating T cells via bet Eomes (P ¼ 0.0510) in IL18R1 / CD8 T cells compared flow cytometry (Fig. 4C–E and Supplementary Fig. S13). with WT (Fig. 4F–H). þ Although a decrease in IFNg IL17a cells among IL18R1 / To further test these IL18 signaling–related T cell–intrinsic þ þ CD4 T cells was observed, we surprisingly found increased defects, we first examined CD4 T cells. We differentiated na€ve þ þ proportions of IFNg IL17a cells compared with WT (Fig. 4C CD4 T cells from WT and IL18R1 / mice in vitro under Th1- and

OF6 Cancer Res; 76(8) April 15, 2016 Cancer Research

Inﬂammatory IL18 Signaling Restricts HCC Progression

A B 3 3 + WT CD4 WT CD8+ −/− + IL18R1 CD4 IL18R1−/− CD8+ 2 2

Fold of WT of Fold 1 Fold of WT of Fold 1

0 0

IL2 IL4 IL6 γ IL10 T-betIFNγ Prf1 TGFβ IFN T-bet IL17a TNFα TGFβIL17aTNFαIP-10 GZMAGZMB RORC IL18R1 GZMAGZMB Eomes MCP-1IL18R1

C DE NS NS NS − + + 100 20 5 IL17a IL17a IL17a + −

80 + 4 15 60 3 10 40 2 5 20 1

0 0 0 Percent stained: IFN γ stained: Percent Percent stained: IFN γ stained: Percent + + + + + + + + IFN γ stained: Percent + + + +

CD8 CD4 CD8 CD4 CD4 −/− −/− −/− −/− −/− CD8 −/− WT CD8 WT CD4 WT CD8 WT CD4 WT CD8 WT CD4

IL18R1 IL18R1 IL18R1 IL18R1 IL18R1 IL18R1

FGH − + NS NS NS + NS NS 40 40 10 Eomes Eomes Eomes + 8 − 30 30 + 6 20 20 4 10 10 2

0 0 0 + + + + Percent stained: T-bet stained: Percent + + + + + + + + Percent stained: T-bet stained: Percent Percent stained: T-bet stained: Percent

CD8 CD4 CD8 CD4 CD8 −/− −/− −/− −/− −/− −/− CD4 WT CD8 WT CD4 WT CD8 WT CD4 WT CD8 WT CD4

IL18R1 IL18R1 IL18R1 IL18R1 IL18R1 IL18R1

Figure 4. Functionality of tumor-infiltrating T cells is altered by loss of IL18R1. A and B, mice were treated with CCl4 and orthotopic tumor implantation, and þ þ þ þ TCRb CD4 CD25 cells (A) and TCRb CD8 cells (B) were sorted from harvested tumors, mRNA extracted and reverse transcribed, and probed for expression of targets. Fold changes are calculated by normalization to the average expression level in WT CD4þ (A) and CD8þ T cells (B) for the þ þ given experiment. Data pooled from two experiments, total n ¼ 7forWTandn ¼ 6forIL18R1-deficient mice. C–H, tumor-infiltrating TCRb CD4 and þ þ TCRb CD8 cells in WT and IL18R1-deficient mice were stained for IFNg and IL17a (C–E) or T-bet and Eomes (F–H). Single- and double-stained þ þ þ þ populations were quantified as percentages of TCRb CD4 and TCRb CD8 cells. Statistics: multiple t tests, n ¼ 6–7pergrouppooledovertwo experiments (A and B); one-way ANOVA (C–H) with uncorrected Fisher LSD test, n ¼ 11–13 per group pooled over three experiments for C–E, n ¼ 8–9 per group pooled over two experiments for F–H. NS, nonsignificant. , P < 0.05; , P < 0.01.

Th17-inducing conditions and then examined their IFNg and efficiently and failed to enhance IFNg production upon exposure þ IL17a production in the presence or absence of IL18. In WT to IL18 (Fig. 5A). IL18R1 / CD4 T cells also did not differen- þ CD4 T cells, we found efficient differentiation to both Th1 and tiate as efficiently to the Th17 lineage (Fig. 5C). We further Th17 lineages and enhanced IFNg production in Th1-differenti- assessed the proliferation and death of WT Th1 and Th17 cells ated cells upon IL18 exposure (Fig. 5A and C and Supplementary in the presence of IL18BP, an antagonistic secreted decoy receptor þ Fig. S14). Th17-differentiated WT CD4 T cells did not respond to for IL18 (31), in the absence of exogenously added IL18. We IL18 exposure with enhanced IL17a production (Fig. 5C). In found that more Th1-differentiated cells died, and Th17-differ- þ contrast, IL18R1 / CD4 T cells did not differentiate to Th1 as entiated cells proliferated more slowly when treated with

www.aacrjournals.org Cancer Res; 76(8) April 15, 2016 OF7

Markowitz et al.

NS NS NS NS NS NS A NS C NS NS NS 15 25 WT WT −/− −/− IL18R1 20 IL18R1 10 15

10 % IFN γ +

5 % IL17a + 5

0 0

0 ng/mL IL18 30 ng/mL IL18 0 ng/mL IL18 30 ng/mL IL18 100 ng/mL IL18 100 ng/mL IL18

20 0 ng/mL IL18BP NS 0 ng/mL IL18BP NS 10 ng/mL IL18BP 1.4 15 10 ng/mL IL18BP 30 ng/mL IL18BP 30 ng/mL IL18BP 100 ng/mL IL18BP 100 ng/mL IL18BP 10 300 ng/mL IL18BP 1.2 300 ng/mL IL18BP

Percent living Percent 5 Fold of CTV MFI Fold without treatment 1.0 0

0 ng/mL IL18BP 0 ng/mL IL18BP 30 ng/mL IL18BP 10 ng/mL 30IL18BP ng/mL IL18BP 10 ng/mL IL18BP 100 ng/mL300 IL18BP ng/mL IL18BP 100 ng/mL300 IL18BP ng/mL IL18BP

Figure 5. þ Differentiation and cytokine production of CD4 T cells is altered by loss of IL18R1. A, quantification of IFNg production from WT and IL18R1-deficient Th1- differentiated CD4þ T cells stimulated with IL18. B, WT CD4þ T cells were differentiated to Th1 and treated with IL18BP. Five days after the start of differentiation, Th1- þ differentiated (IFNg ) cell survival was evaluated using a fixable Live/Dead stain. C, quantification of IL17a production from WT and IL18R1-deficient Th17- þ þ differentiated CD4 T cells stimulated with IL18. D, WT CD4 T cells were differentiated to Th17, dyed with CellTrace Violet (CTV), and treated with IL18BP. Five days after the start of differentiation, proliferation was evaluated by CellTrace Violet dilution. Statistics: two-way ANOVA with uncorrected Fisher LSD test, n ¼ 3 per group, representative of twice-repeated experiments (A and C); one-way ANOVA with uncorrected Fisher LSD test, n ¼ 12 per group pooled from three experiments (B and D). NS, nonsignificant. , P < 0.05; , P < 0.01; , P < 0.001; , P < 0.0001. MFI, mean fluorescence intensity.

þ increasing doses of IL18BP (Fig. 5B and D). We conclude that in IL18R1 / , or treated IL18R1 / CD8 T cells (Fig. 6A). Treated þ Th1 cells, IL18 serves as a prosurvival factor, whereas in Th17 cells, WT CD8 T cells exposed to TCM also had a greater proliferation IL18 modulates proliferation. Combined with the diminished rate than both their untreated counterparts as well as IL18R1 / þ production of TNFa and IL2, these results argue for multifaceted CD8 T cells (Fig. 6B). Together, these data demonstrate IL18 modulation of inflammatory Th cell capacities. that deficiency in IL18 signaling modulates both the accumula- þ þ þ IL18 signaling also had a robust effect on CD8 T cells: both tion and functionality of both CD4 and CD8 tumor-infiltrating their number and functionality were significantly reduced in T cells. IL18R1 / tumors. Therefore, to evaluate the net effect of IL18 signaling, we performed in vitro restimulation experiments (uti- IL18 signaling regulates lymphocyte-mediated tumor lizing weaker T-cell antigen receptor complex stimulation than cytotoxicity in vivo þ that used during initial activation) to assess effects on CD8 T-cell IL18 signaling profoundly impacted T-cell differentiation, cyto- þ proliferation and survival. Upon restimulation, untreated WT and kine production, proliferation, and survival, so we tested if CD8 þ IL18R1 / CD8 T cells exposed to tumor conditional media T cells mediated the differential antitumor responses between WT þ (TCM) proliferate, and the majority subsequently dies (Supple- and IL18R1 / mice. We ablated CD8 T cells in both WT and þ mentary Fig. S15). However, WT CD8 T cells treated with IL18 IL18R1 / mice in our fibrotic allograft model and examined had a higher survival rate than either untreated WT, untreated tumor burden. Surprisingly, this did not equalize, nor did it

OF8 Cancer Res; 76(8) April 15, 2016 Cancer Research

Inﬂammatory IL18 Signaling Restricts HCC Progression

A B

40 3,000 No TCM TCM 30 2,000

20 MFI 1,000

Percent of gate Percent 10

0 0

WT 0 ng/mL KO 0 ng/mL WT 0 ng/mL KO 0 ng/mL WT 30 ng/mL KO 30 ng/mL WT 30 ng/mL KO 30 ng/mL WT 100 ng/mL KO 100 ng/mL WT 100 ng/mL KO 100 ng/mL

NS CDNS NS NS NS NS 20 100

80 15 60 10 40 Percent of gate Percent of body weight 5 20 Liver as percentage Liver

0 0 + + + + + + + + + + + +

+ IgG −/− NK1.1 NK1.1 + TCRb + TCRb + TCRb + TCRb TCRb TCRb −/− −/− −/− −/− WT + IgG IgG IgG + Anti-CD8 WT NK1.1WT NK1.1 WT TCRbWT TCRb −/− Anti-CD8IgG IgG Anti-CD8IgG Anti-CD8 NK1.1 Anti-CD8 WT + Anti-CD8 IL18R1 −/− IgG NK1.1 IL18R1 IL18R1 Anti-CD8 −/− IL18R1 IL18R1 WT NK1.1WT NK1.1 Anti-CD8 IL18R1−/− + Anti-NK, anti-CD8 WT + Anti-NK, anti-CD8 IL18R1 IL18R1

IL18R1

Figure 6. Accumulation of CD8þ T cells is altered by loss of IL18R1; CD8þ T cells and NK cells modulate tumor burden. A, surviving WT and IL18R1-deficient CD8þ T cells (defined þ as 7AADloCellTrace Violetlo) upon reactivation and stimulation with different doses of IL18 with or without TCM. B, proliferation of CD8 T cells [lower mean fluorescence intensity (MFI)]. C, tumor growth measured by liver weight as a percentage of body weight. Tumor-bearing mice were treated with IgG, anti-CD8a, or þ þ þ þ anti-NK1.1 and anti-CD8a. D, quantification of intratumoral NK1.1 TCRb , NK1.1 TCRb , and TCRb NK1.1 cells from WT or IL18R1-deficient C57Bl/6 mice treated with CCl4 and either IgG or anti-CD8a. Data are representative of experiments repeated three times. Statistics: two-way ANOVA with Sidak multiple comparisons test, n ¼ 4 per group pooled over two experiments (A); two-way ANOVA with Sidak multiple comparisons test, n ¼ 2 per group, representative of experiments repeated two times (B); unpaired t tests (C–D), left to right: n ¼ 6, n ¼ 6, n ¼ 6, n ¼ 10, n ¼ 7, n ¼ 3(C),n ¼ 3, n ¼ 5, n ¼ 5, n ¼ 5 (D). NS, nonsignificant. , P < 0.05; , P < 0.01; , P < 0.001; , P < 0.0001. KO, knockout.

significantly alter, tumor burden between WT and IL18R1 / therefore performed ablation experiments targeting NK cells. mice (Fig. 6C). Upon closer examination, we found that when Although single ablation of NK cells did not affect tumor burden þ þ CD8 T cells were ablated in WT mice, NK cells comprised a (data not shown), depleting both NK cells and CD8 T cells significantly increased portion of tumor-infiltrating cells com- equalized tumor burden between WT and IL18R1 / mice, dem- þ pared with IgG-treated controls (Fig. 6D). Our earlier experiments onstrating complementary CD8 T-cell– and NK cell–dependent had not shown a genotype-specific difference in intratumoral NK IL18-mediated tumor suppression (Fig. 6C). cell accumulation; however, IL18 markedly regulates NK cell activity (25). We then examined the functionality and maturation Comparisons between IL18 expression in tumor and nontumor of tumor-infiltrating NK cells in WT and IL18R1 / mice and tissue provide better survival prediction found differences in IFNg, T-bet, and Eomes production (Sup- These results suggest that, rather than promoting tumor growth plementary Fig. S16), suggesting that their antitumor efficacy may at late stages, IL18 acts as a tumor suppressor and should display be different, particularly in the absence of other cytotoxic cells. We reduced intratumoral expression. To probe this critical point, we

www.aacrjournals.org Cancer Res; 76(8) April 15, 2016 OF9

Markowitz et al.

ABFigure 7. Intratumoral Peritumoral Highest 100 difference Differences in IL18 expression between Middle tumor and nontumor tissue strongly difference 80 Lowest correlate with prognosis. A, difference representative intratumoral and 60 Sample 1 peritumoral staining (magniﬁcation, 40 400) for IL18 from three patients. Sample 1, middle intratumoral density, 20 P < 0.0001 low peritumoral density, low difference Overall survival, percent Overall 0 in staining density; sample 2, low 0 20 40 60 80 100 120 intratumoral density, low peritumoral Months density, middle difference in staining Sample 2 Highest density; sample 3, high intratumoral 100 difference Middle density, high peritumoral density, high difference difference in staining density. B, 80 Lowest difference Kaplan–Meier curves for overall (top) 60 and disease-free survival (bottom) of the 138 patients stratiﬁed into the 40 highest, middle, or lowest third of Sample 3 20 patients based on differences in IL18 P < 0.0001 staining density (Supplementary Table 0 Disease-free survival, percent 0 20 40 60 80 100 120 S4; Supplementary Fig. S18). Statistics: – Months log-rank Mantel Cox tests, n ¼ 46 per group (B).

reexamined the tissue specimens described earlier in Fig. 1B and C finding provides a novel molecular mechanism, as previous to determine the relationship between the expression profiles of alternative models had only demonstrated effects of IL18 on IL18 and prognosis of those patients. In the majority of tissue NK-mediated cytotoxic activity in gene therapy studies in immu- samples, similar to Fig. 1D, IL18 staining was stronger in the nodeficient mice (10). IL18 is a well-known regulator of Th1 peritumoral tissue than within the tumor itself (Fig. 7A and functionality, affecting differentiation, proliferation, and cytokine Supplementary Fig. S17; quantification: Supplementary Table production (25, 41–44); its role in Th17 cells is more contentious S4; Supplementary Figs. S1 and S18). Importantly, when we (45, 46). Th1 cells are critical modulators of tumor suppression quantified differences in immunostaining density between each in multiple cancers, including hepatocellular carcinoma (39), patient's intratumoral and peritumoral tissue, we found that this whereas the function of Th17 cells in hepatocellular carcinoma difference, particularly for those with the highest divergence, is unresolved. Increasing levels of IL17 in the tumor have been correlated more significantly to both overall and disease-free correlated with worse prognosis, and some reports have shown survival in patients than staining in either region alone (Fig. significant effects of IL17 on tumor cells (47, 48). However, 7B; compared with P values in Fig. 1B and C). By this new studies in other systems have shown that Th17 cells affect anti- þ measure, the ambiguity in patients with intermediate levels of tumor cytotoxicity, in part by regulating CD8 T-cell functionality IL18 staining shown in Fig. 1B and C was largely diminished. (40). Our data demonstrate IL18 signaling–mediated effects on Thus, the differential expression of this cytokine between tumor Th17 development and functionality in vitro and in vivo in hepa- and nontumor tissue could be a better biomarker for predicting tocellular carcinoma. Other systems have shown that IL18 over- þ patient outcomes. expression leads to enhanced CD8 T-cell cytotoxicity in vitro and partial efficacy in reducing tumor burden in vivo, although not at the level of individual cells; our data mirror this and add more Discussion clarity to its effects. Cumulatively, we show multifaceted IL18 Liver cancer is a devastating disease, the progression and mediation of inflammatory Th cell capacities, modulation of their þ þ prognosis of which are intricately intertwined with the chronically interactions with CD8 T cells, and effects on the CD8 T cells inflamed tissue in which it develops. This diseased microenvi- themselves to control tumor growth. Importantly, these results ronment is drastically different from the healthy liver, including were generated in the absence of IL18 overexpression, demon- different cytokines, growth factors, activated stromal cells, extra- strating a robust tumor-suppressive role for endogenous IL18. cellular matrix, and infiltrating immune cells. This altered micro- We found that discerning the effects of endogenous IL-18 on environment has been well known to be integral to liver carci- tumor growth was dependent upon an inflammatory context. nogenesis; however, studies examining the interplay between this These data suggest that important factors modulating tumor altered environment, the tumor, and effects on the progression of growth may not be easily evaluated in the absence of models the disease posttransformation have been less abundant (34). mimicking clinical situation: this disease almost always presents Numerous molecules have been shown to modulate this disease, with concurrent chronic inflammation and fibrosis. Many previ- but many more remain to be discovered and evaluated using more ous studies have utilized either immunodeficient mouse models rigorous and relevant models with intact immune systems and or models without this inflamed tissue background. Although the functional microenvironments necessary for these experiments. results generated from those experiments are still valuable, future In this study, we demonstrated that although IL18 is elevated in studies should be designed accounting for this continually evolv- serum and nontumor tissue of patients with hepatocellular car- ing inflammatory context (7, 19, 34, 49,50). cinoma, its overall function in tumor progression is suppressive, We posit that elevated IL18 in tissue and serum is indicative of modulating the accumulation and function of lymphocytes. This inflammation and ongoing tissue injury and a marker for disease

OF10 Cancer Res; 76(8) April 15, 2016 Cancer Research

Inﬂammatory IL18 Signaling Restricts HCC Progression

progression; however, its in-tumor functionality is suppressive. reduced IL18 signaling in the tumor compared with its surround- IL18 expression is induced in chronic liver disease arising from a ing nontumor tissue, indicative of less intratumoral IL18-induced variety of etiologies, including chronic HBV, chronic HCV, and tumor suppression; and destructive inflammatory IL18 signaling NAFLD, elevated in tissue and serum, and correlates with liver in the diseased nontumor liver tissue, indicative of overall tumor- injury and fibrosis (26–28, 32, 33, 51). Indeed, hepatotropic independent liver function. viruses induce expression of IL18 in vitro, and transgenic IL18 – expression induces hepatic injury (52 54). Similar to other IL1 Disclosure of Potential Conflicts of Interest family members, IL18 is integral to the hepatocarcinogenic pro- No potential conflicts of interest were disclosed. cess (15, 18, 20, 29,30). We observed decreased fibrosis yet increased tumor burden in IL18R1 / mice at later stages, as well Authors' Contributions as reduced IL18 in tumor compared with matched nontumor Conception and design: G.J. Markowitz, P. Yang, A.M. Diehl, Q.-J. Li, tissue from patients, suggesting both late-stage tumor suppression X.-F. Wang by IL18 and a disruption in the frequently observed correlation Development of methodology: G.J. Markowitz, P. Yang, G.A. Michelotti between fibrosis and tumor burden. Cumulatively, this describes Acquisition of data (provided animals, acquired and managed patients, a stage- and tissue-specific functionality for IL18: inflammation provided facilities, etc.): G.J. Markowitz, P. Yang, J. Fu, G.A. Michelotti, enables carcinogenesis; however, the same ongoing inflammation R. Chen, B. Yang, W.-H. Qin, Z. Zhang, A.M. Diehl, H. Wang in established tumors is suppressive. Indeed, tumor cells produce Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): G.J. Markowitz, P. Yang, J. Fu, A.M. Diehl less IL18 than their untransformed counterparts in distal tissue. Writing, review, and/or revision of the manuscript: G.J. Markowitz, P. Yang, Multiple mechanisms underlying this reduction in IL18 produc- Q.-J. Li, X.-F. Wang tion are possible, including clonal expansion, promoter methyl- Administrative, technical, or material support (i.e., reporting or organizing ation, transcriptional repression, and deregulation of inflamma- data, constructing databases): P. Yang, G.A. Michelotti, J. Sui, B. Yang, some components, which warrant further investigation. These F.-S. Wang data emphasize that patterns of cytokine expression may differ Study supervision: P. Yang, F.-S. Wang, A.M. Diehl, H. Wang, X.-F. Wang between tumor, nontumor tissue, and serum. Caution must be observed in choosing microenvironment-based therapeutic tar- Acknowledgments gets for development, taking into consideration clinical data from The authors thank members of the Duke Cancer Institute Flow Cytometry all relevant tissues. Core for aid in sorting cells and members of the Wang and Li labs, particularly Jing Hu, Regina Lin, and Siqi Liu, for helpful discussions and technical advice. Recent studies have shown that gene expression in the diseased liver correlates strongly with patient survival, and treatment targeting the nontumor liver dramatically affects tumor progres- Grant Support sion (22, 24, 55). Our analyses demonstrated a stronger correla- This work was supported by grants CA154151 and CA164791, NCI tion between survival and IL18 staining in the nontumor liver (X.-F. Wang); grant 2012ZX10002-009, National Science and Technology Major Project of China, grant 30921006, National Natural Science Founda- than in the tumor. However, comparing IL18 expression between tion for Creative Research Groups of China (H. Wang); grant 81472768, nontumor and tumor tissues had a stronger correlation to prog- National Natural Science Foundation of China (J. Fu); grant 81222024, nosis than either tissue alone, with the staining in the nontumor National Science Fund for Outstanding Young Scholars, grant liver determining the magnitude of this difference. Interestingly, 2012CB519005, National Key Basic Research Program of China (Z. Zhang); levels of IL18R1 in either tissue alone or by comparison did not and grant 2015CB553700-006, National Key Basic Research Program of have predictive power towards patient survival (Supplementary China (P. Yang). The costs of publication of this article were defrayed in part by the payment Fig. S19), suggesting that differential expression of ligand dictates of page charges. This article must therefore be hereby marked advertisement the effects of IL18 signaling on hepatocellular carcinoma prog- in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. nosis. These findings suggest that the prognosis of hepatocellular carcinoma patients is determined by the combinational effects of Received June 4, 2015; revised February 4, 2016; accepted February 10, 2016; two key factors intimately associated with IL18 expression: published OnlineFirst February 18, 2016.

References 1. Yang JD, Roberts LR. Hepatocellular carcinoma: a global view. Nat Rev 8. Maeda S, Kamata H, Luo J-L, Leffert H, Karin M. IKKb couples hepatocyte Gastroenterol Hepatol 2010;7:448–58. death to cytokine-driven compensatory proliferation that promotes chem- 2. Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer ical hepatocarcinogenesis. Cell 2005;121:977–90. statistics. CA Cancer J Clin 2011;61:69–90. 9. Fu J, Xu D, Liu Z, Shi M, Zhao P, Fu B, et al. Increased regulatory 3. Farazi PA, DePinho RA. Hepatocellular carcinoma pathogenesis: from T cells correlate with CD8 T-cell impairment and poor survival genes to environment. Nat Rev Cancer 2006;6:674–87. in hepatocellular carcinoma patients. Gastroenterology 2007;132: 4. Kuraishy A, Karin M, Grivennikov SI. Tumor promotion via injury- and 2328–39. death-induced inflammation. Immunity 2011;35:467–77. 10. Tsuchiyama T, Nakamoto Y, Sakai Y, Marukawa Y, Kitahara M, Mukaida N, 5. Thorgeirsson SS, Grisham JW. Molecular pathogenesis of human hepato- et al. Prolonged, NK cell-mediated antitumor effects of suicide gene therapy cellular carcinoma. Nat Genet 2002;31:339–46. combined with monocyte chemoattractant protein-1 against hepatocellu- 6. Grimm CF, Ortmann D, Mohr L, Michalak S, Krohne TU, Meckel S, et al. lar carcinoma. J Immunol 2007;178:574–83. Mouse a-fetoprotein–specific DNA-based immunotherapy of hepatocel- 11. Qin L-X. Inflammatory immune responses in tumor microenvironment lular carcinoma leads to tumor regression in mice. Gastroenterology and metastasis of hepatocellular carcinoma. Cancer Microenviron 2012;5: 2000;119:1104–12. 203–9. 7. Pikarsky E, Porat RM, Stein I, Abramovitch R, Amit S, Kasem S, et al. NF-kB 12. Schneider C, Teufel A, Yevsa T, Staib F, Hohmeyer A, Walenda G, et al. functions as a tumour promoter in inflammation-associated cancer. Nature Adaptive immunity suppresses formation and progression of diethylni- 2004;431:461–6. trosamine-induced liver cancer. Gut 2012;61:1733–43.

www.aacrjournals.org Cancer Res; 76(8) April 15, 2016 OF11

Markowitz et al.

13. Yang P, Li Q-J, Feng Y, Zhang Y, Markowitz GJ, Ning S, et al. TGF-beta-miR- 35. Hoshino K, Tsutsui H, Kawai T, Takeda K, Nakanishi K, Takeda Y, et al. 34a-CCL22 signaling-induced treg cell recruitment promotes venous Cutting edge: generation of IL-18 receptor-deficient mice: evidence for IL-1 metastases of HBV-positive hepatocellular carcinoma. Cancer Cell receptor-related protein as an essential IL-18 binding receptor. J Immunol 2012;22:291–303. 1999;162:5041–4. 14. Yang P, Markowitz GJ, Wang XF. The hepatitis B virus-associated tumor 36. Kinoshita A, Onoda H, Imai N, Iwaku A, Oishi M, Fushiya N, et al. microenvironment in hepatocellular carcinoma. Natl Sci Rev 2014;1: Comparison of the prognostic value of inflammation-based prognostic 396–412. scores in patients with hepatocellular carcinoma. Br J Cancer 2012;107: 15. Sakurai T, He G, Matsuzawa A, Yu G-Y, Maeda S, Hardiman G, et al. 988–93. Hepatocyte necrosis induced by oxidative stress and IL-1a release mediate 37. Calzascia T, Pellegrini M, Hall H, Sabbagh L, Ono N, Elford AR, et al. TNF-a carcinogen-induced compensatory proliferation and liver tumorigenesis. is critical for antitumor but not antiviral T cell immunity in mice. J Clin Cancer Cell 2008;14:156–65. Invest 2007;117:3833–45. 16. Tang Y, Kitisin K, Jogunoori W, Li C, Deng C-X, Mueller SC, et al. Progen- 38. Rosenberg SA. IL-2: the first effective immunotherapy for human cancer. itor/stem cells give rise to liver cancer due to aberrant TGF-beta and IL-6 J Immunol 2014;192:5451–8. signaling. Proc Natl Acad Sci U S A 2008;105:2445–50. 39. Budhu A, Forgues M, Ye Q-H, Jia H-L, He P, Zanetti KA, et al. Prediction of 17. Park EJ, Lee JH, Yu G-Y, He G, Ali SR, Holzer RG, et al. Dietary and genetic venous metastases, recurrence, and prognosis in hepatocellular carcinoma obesity promote liver inflammation and tumorigenesis by enhancing IL-6 based on a unique immune response signature of the liver microenviron- and TNF expression. Cell 2010;140:197–208. ment. Cancer Cell 2006;10:99–111. 18. Yan W, Chang Y, Liang X, Cardinal JS, Huang H, Thorne SH, et al. High- 40. Martin-Orozco N, Muranski P, Chung Y, Yang XO, Yamazaki T, Lu S, et al. T mobility group box 1 activates caspase-1 and promotes hepatocellular helper 17 cells promote cytotoxic T cell activation in tumor immunity. carcinoma invasiveness and metastases. Hepatology 2012;55:1863–75. Immunity 2009;31:787–98. 19. He G, Dhar D, Nakagawa H, Font-Burgada J, Ogata H, Jiang Y, et al. 41. Greene CM, Meachery G, Taggart CC, Rooney CP, Coakley R, O'Neill SJ, Identification of liver cancer progenitors whose malignant progression et al. Role of IL-18 in CD4þ T lymphocyte activation in sarcoidosis. depends on autocrine IL-6 signaling. Cell 2013;155:384–96. J Immunol 2000;165:4718–24. 20. Yoshimoto S, Loo TM, Atarashi K, Kanda H, Sato S, Oyadomari S, et al. 42. Kohno K, Kataoka J, Ohtsuki T, Suemoto Y, Okamoto I, Usui M, et al. IFN- Obesity-induced gut microbial metabolite promotes liver cancer through gamma-inducing factor (IGIF) is a costimulatory factor on the activation of senescence secretome. Nature 2014;498:97–101. Th1 but not Th2 cells and exerts its effect independently of IL-12. 21. Ye Q-H, Qin L-X, Forgues M, He P, Kim JW, Peng AC, et al. Predicting J Immunol 1997;158:1541–50. hepatitis B virus–positive metastatic hepatocellular carcinomas using gene 43. Micallef MJ, Ohtsuki T, Kohno K, Tanabe F, Ushio S, Namba M, et al. expression profiling and supervised machine learning. Nat Med 2003;9: Interferon-gamma-inducing factor enhances T helper 1 cytokine produc- 416–23. tion by stimulated human T cells: synergism with interleukin-12 for 22. Hoshida Y, Villanueva A, Kobayashi M, Peix J, Chiang DY, Camargo A, et al. interferon-gamma production. Eur J Immunol 1996;26:1647–51. Gene expression in fixed tissues and outcome in hepatocellular carcinoma. 44. Puren AJ, Fantuzzi G, Gu Y, Su MS, Dinarello CA. Interleukin-18 (IFN- N Engl J Med 2008;359:1995–2004. gamma-inducing factor) induces IL-8 and IL-1beta via TNFalpha produc- 23. Michelotti GA, Machado MV, Diehl AM. NAFLD, NASH and liver cancer. tion from non-CD14þ human blood mononuclear cells. J Clin Invest Nat Rev Gastroenterol Hepatol 2013;10:656–65. 1998;101:711–21. 24. Fuchs BC, Hoshida Y, Fujii T, Wei L, Yamada S, Lauwers GY, et al. Epidermal 45. Lalor SJ, Dungan LS, Sutton CE, Basdeo SA, Fletcher JM, Mills KHG. growth factor receptor inhibition attenuates liver fibrosis and development Caspase-1-processed cytokines IL-1 and IL-18 promote IL-17 production of hepatocellular carcinoma. Hepatology 2014;59:1577–90. by and CD4 T cells that mediate autoimmunity. J Immunol 2011;186: 25. Garlanda C, Dinarello CA, Mantovani A. The Interleukin-1 family: back to 5738–48. the future. Immunity 2013;39:1003–18. 46. Nakae S, Iwakura Y, Suto H, Galli SJ. Phenotypic differences between Th1 26. Cheng K-S, Tang H-L, Chou F-T, Chou J-W, Hsu C-H, Yu C-J, et al. Cytokine and Th17 cells and negative regulation of Th1 cell differentiation by IL-17. evaluation in liver cirrhosis and hepatocellular carcinoma. Hepatogas- J Leukoc Biol 2007;81:1258–68. troenterology 2009;56:1105–10. 47. Gu F-M, Li Q-L, Gao Q, Jiang J-H, Zhu K, Huang X-Y, et al. IL-17 induces 27. Perrella O, Cuomo O, Sbreglia C, Monaco A, Gnarini MR, Gentile B, et al. AKT-dependent IL-6/JAK2/STAT3 activation and tumor progression in IL-18 and interferon-gamma in HCV-related hepatocellular carcinoma: a hepatocellular carcinoma. Mol Cancer 2011;10:150. model of interplay between immune status and cancer. J Biol Regul 48. Liao R, Sun J, Wu H, Yi Y, Wang J-X, He H-W, et al. High expression of IL-17 Homeost Agents 2009;23:251–8. and IL-17RE associate with poor prognosis of hepatocellular carcinoma. 28. Tangkijvanich P, Thong-Ngam D, Mahachai V, Theamboonlers A, Poovor- J Exp Clin Cancer Res 2013;32:3. awan Y. Role of serum interleukin-18 as a prognostic factor in patients with 49. Dapito DH, Mencin A, Gwak G-Y, Pradere J-P, Jang M-K, Mederacke I, et al. hepatocellular carcinoma. World J Gastroenterol 2007;13:4345–9. Promotion of hepatocellular carcinoma by the intestinal microbiota and 29. Li S, Sun R, Chen Y, Wei H, Tian Z. TLR2 limits development of hepato- TLR4. Cancer Cell 2012;21:504–16. cellular carcinoma by reducing IL18-mediated immunosuppression. Can- 50. Yang M-C, Chang C-P, Lei H-Y. Induction of liver fibrosis in a murine cer Res 2015;75:986–95. hepatoma model by thioacetamide is associated with enhanced tumor 30. Terakura D, Shimizu M, Iwasa J, Baba A, Kochi T, Ohno T, et al. Preventive growth and suppressed antitumor immunity. Lab Invest 2010;90:1782–93. effects of branched-chain amino acid supplementation on the spontane- 51. Negash AA, Gale M. Hepatitis regulation by the inflammasome signaling ous development of hepatic preneoplastic lesions in C57BL/KsJ-db/db pathway. Immunol Rev 2015;265:143–55. obese mice. Carcinogenesis 2012;33:2499–506. 52. Finotto S, Siebler J, Hausding M, Schipp M, Wirtz S, Klein S, et al. Severe 31. Novick D, Schwartsburd B, Pinkus R, Suissa D, Belzer I, Sthoeger Z, et al. A hepatic injury in interleukin 18 (IL-18) transgenic mice: a key role for IL-18 novel IL-18BP ELISA shows elevated serum IL-18BP in sepsis and extensive in regulating hepatocyte apoptosis in vivo. Gut 2004;53:392–400. decrease of free IL-18. Cytokine 2001;14:334–42. 53. Lee M-O, Choi Y-H, Shin E-C, Kang H-J, Kim Y-M, Jeong S-Y, et al. Hepatitis 32. Chia CS, Ban K, Ithnin H, Singh H, Krishnan R, Mokhtar S, et al. Expression B virus X protein induced expression of interleukin 18 (IL-18): a potential of interleukin-18, interferon-gamma and interleukin-10 in hepatocellular mechanism for liver injury caused by hepatitis B virus (HBV) infection. carcinoma. Immunol Lett 2002;84:163–72. J Hepatol 2002;37:380–6. 33. Zhang Y, Li Y, Ma Y, Liu S, She Y, Zhao P, et al. Dual effects of interleukin-18: 54. Shrivastava S, Mukherjee A, Ray R, Ray RB. Hepatitis C virus induces inhibiting hepatitis B virus replication in HepG2.2.15 cells and promoting interleukin-1beta (IL-1beta)/IL-18 in circulatory and resident liver macro- hepatoma cells metastasis. Am J Physiol Gastrointest Liver Physiol phages. J Virol 2013;87:12284–90. 2011;301:G565–73. 55. Philips GM, Chan IS, Swiderska M, Schroder VT, Guy C, Karaca GF, et al. 34. Markowitz GJ, Michelotti GA, Diehl AM, Wang X-F. Inflammatory models Hedgehog signaling antagonist promotes regression of both liver fibrosis drastically alter tumor growth and the immune microenvironment in and hepatocellular carcinoma in a murine model of primary liver cancer. hepatocellular carcinoma. Sci Bull 2015;60:762–72. PLoS One 2011;6:e23943.

OF12 Cancer Res; 76(8) April 15, 2016 Cancer Research

Inflammation-Dependent IL18 Signaling Restricts Hepatocellular Carcinoma Growth by Enhancing the Accumulation and Activity of Tumor-Infiltrating Lymphocytes

Geoffrey J. Markowitz, Pengyuan Yang, Jing Fu, et al.

Cancer Res Published OnlineFirst February 18, 2016.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-15-1548

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2016/02/18/0008-5472.CAN-15-1548.DC1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/early/2016/04/05/0008-5472.CAN-15-1548. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.