Expression of Interleukin-4 Receptor Alpha in Human Pleural Mesothelioma Is Associated with Poor Survival and Promotion of Tumor Inﬂammation

Published OnlineFirst January 18, 2012; DOI: 10.1158/1078-0432.CCR-11-1808

Clinical Cancer Human Cancer Biology Research

Bryan M. Burt1, Andrew Bader1, Daniel Winter1, Scott J. Rodig2, Raphael Bueno1, and David J. Sugarbaker1

Abstract Purpose: The origin and pathogenesis of malignant pleural mesothelioma (MPM) are closely aligned with inflammation. MPM tumors express interleukin-4 receptor a (IL-4Ra), the principal subunit of the IL-4 receptor. We set out to determine the biologic function and clinical relevance of IL-4Ra in human MPM. Experimental Design: Expression of IL-4Ra by human MPM tumors was determined by quantitative real-time PCR (n ¼ 37) and immunohistochemistry (n ¼ 52). Intracellular cytokine analysis of T-cell– derived IL-4 was carried out on matched tumor and blood samples from eight patients with MPM. Four human MPM cell lines were used to determine the direct effects of IL-4 on MPM tumor cells. Results: High tumor mRNA expression of IL-4Ra was an independent predictor of poor survival in patients with epithelial MPM [HR, 3.13, 95% confidence interval (CI), 1.68–7.15; P ¼ <0.0001]. Ninety- seven percent of epithelial MPM tumors and 95% of nonepithelial MPM tumors expressed IL-4Ra protein by immunohistochemistry, and strong IL-4Ra staining correlated with worse survival in patients with epithelial histology (P ¼ 0.04). A greater percentage of tumor-infiltrating T cells produced IL-4 compared with matched blood T cells (21% 7% vs. 4% 2%, P ¼ 0.0002). In response to IL-4, human MPM cells showed increased STAT-6 phosphorylation and increased production of IL-6, IL-8, and VEGF, without effect on proliferation or apoptosis. Conclusions: Tumor expression of IL-4Ra is inversely correlated with survival in patients undergoing surgical resection for epithelial MPM. Tumor-infiltrating T cells in MPMs are polarized to produce IL-4 and may provide endogenous activation signals to MPM tumor cells in situ. The IL-4/IL-4 receptor axis is a potential therapeutic target in human MPM. Clin Cancer Res; 18(6); 1568–77. 2012 AACR.

Introduction more aggressive tumors with significantly worse survival Malignant pleural mesothelioma (MPM) is an aggressive (2). tumor that arises from the mesothelial lining of the pleura. It recently has become evident that chronic inflammation It is a unique malignancy whose primary mode of spread is may predispose individuals to develop cancer. Inflamma- invasion of surrounding tissues and whose proclivity for tion promotes the proliferation and survival of malignant distant metastases is low. MPM is a highly fatal tumor that is cells, supports angiogenesis and metastasis, and subverts resilient to single modality treatment with surgery, chemo- immune surveillance and responses to chemotherapy (3). therapy, or radiation. With multimodality therapy in select- MPM, in particular, is a cancer whose origin and pathogen- ed patients, median survival is 19 months (1). Approxi- esis are closely aligned with inflammation. Occupational mately two-thirds of patients with MPM have epithelial exposure to asbestos is identified as the etiology of MPM in histology. The remaining one third of patients have non- 80% of cases (4). Asbestos fibers are thought to initiate a epithelial histology (sarcomatoid or biphasic), which are carcinogenic inflammatory reaction marked by an influx of phagocytes into the tumor and the release of an array of proinflammatory cytokines, such as TNF-a, that promote malignant transformation of the mesothelium via NF-kB– Authors' Affiliations: 1Division of Thoracic Surgery, and 2Department of Pathology, The Brigham & Women's Hospital, Boston, Massachusetts dependent mechanisms (5). In human MPM tumors, ongo- ing inflammation is represented by massive leukocyte infil- Note: Supplementary data for this article are available at Clinical Cancer Research Online (http://clincancerres.aacrjournals.org/). trates of T lymphocytes and macrophages whose densities are correlated with survival (6, 7). Corresponding Author: Bryan M. Burt, Division of Thoracic Surgery, Harvard Medical School, Brigham & Women's Hospital, 75 Francis Interleukin-4 receptor a (IL-4Ra) is the major subunit of Street, Boston, MA 02115. Phone: 617-840-7411; Fax: 617-566-6434; the IL-4 receptor. In addition to being present on immune E-mail: [email protected] cells such as B cells, T cells, and macrophages, IL-4Ra is doi: 10.1158/1078-0432.CCR-11-1808 expressed on a variety of cancer cell lines (8), and in situ on a 2012 American Association for Cancer Research. number of solid human tumors including non–small cell

1568 Clin Cancer Res; 18(6) March 15, 2012

IL-4Ra in Human Pleural MPM

were snap frozen on the day of resection. For flow cytometry Translational Relevance experiments, tumor and matched blood samples were Malignant pleural mesothelioma (MPM) is an aggres- obtained fresh and used on the day of surgery. Peripheral sive tumor of the pleura that arises in a background of blood mononuclear cells were isolated by density centrifu- chronic inflammation. The overall survival of patients gation over Ficoll-Paque Plus (GE Healthcare). To isolate with MPM is poor despite multimodality treatment tumor-infiltrating mononuclear cells, tumor tissues were strategies that include surgery, chemotherapy, and radi- minced into fine fragments in cold Hank’s balanced salt ation. In this study, we have identified interleukin-4 solution (HBSS) and digested in RPMI [American Type receptor a (IL-4Ra), the major subunit of the IL-4 Culture Collection (ATCC)] containing 10% fetal calf receptor, as a biomarker with prognostic correlation in serum (FCS; Gibco), 1 mg/mL Collagenase D (Roche Diag- MPM. In patients undergoing resection for MPM, high nostic), and 100 mg/mL DNase (Qiagen) at 37C for 3 tumor IL-4Ra expression was an independent predictor hours. The tissue slurry was passed through 70 and 40 mm of decreased survival. Furthermore, freshly isolated T filters and washed with cold HBSS. The cells were then cells infiltrating human MPM tumors were polarized to layered over a Ficoll-Paque density gradient and centrifuged produce IL-4, and IL-4 had direct biologic effects on at 1,000 g for 20 minutes at which time the interface layer MPM tumor cells that included promotion of STAT-6 containing tumor mononuclear cells was harvested. phosphorylation and inflammatory cytokine produc- MPM cell lines. Four human MPM cell lines were main- tion. Therefore, we believe that IL-4 endogenously pro- tained in RPMI media containing 10% FCS (Gibco). H2052 duced in the tumor microenvironment may promote cells (epithelial origin) were obtained from the ATCC. cancer inflammation and progression, and that inter- MS924 (epithelial) and MS428 cells (sarcomatoid) were ruption of the IL-4/IL-4 receptor axis may be a promising kindly provided by Dr. Jonathan A. Fletcher, MD, Depart- therapeutic target in human MPMs. ment of Pathology, Brigham & Women’s Hospital (Boston, MA). JMN cells (biphasic) were a gift from Dr. James Rheinwald (16).

Quantitative PCR lung cancer, pancreatic cancer, ovarian cancer, breast can- Quantitative PCR was carried out on MPM tumor tissues cer, prostate cancer, and head and neck cancer (9–12). as previously described (17). Briefly, tumors were selected Furthermore, IL-4 production is augmented in the immune for portions with highest tumor density (>50% tumor cells cells of patients with cancer, and serum and tissue levels of per high-power field) and stored snap frozen. Total RNA (2 IL-4 are correlated with tumor progression in man (13). In mg) was isolated using TRIzol reagent (Invitrogen Life Tech- MPM, Beseth and colleagues have shown IL-4Ra expression nologies) and reverse transcribed using Taq-Man Reverse on mesothelioma tumors from 13 patients using immuno- Transcription reagents (Applied Biosystems). Quantitative histochemistry (14). We set out to determine the clinical real-time PCR (RT-PCR) was carried out using a SYBR-Green significance and biologic function of IL-4Ra in human fluorometric-based detection system (Applied Biosystems) MPM. with a Stratagene MX 3000P device. The primer sequences for IL-4Ra (synthesized by Invitrogen Life Technologies) Materials and Methods used for RT-PCR were as follows: 50-TCATGGAT- Patients, tissues, and cell lines GACGTGGTCAGT-30 (forward) and 50-GTGTCGGAGA- Patients. Human tumor, blood, and effusion were CATTGGTGTG-30 (reverse). PCR amplification of cDNA acquired from patients undergoing extrapleural pneumo- samples was carried out in triplicate using appropriate nectomy (EPP) in accordance with Institutional Review negative controls and the melting curves were in agreement Board policy and following informed consent. Required with the expected amplified fragments. Ribosome 18S RNA criteria for EPP include disease confined to the ipsilateral was used as an internal control for normalization of data hemithorax and sufficient predicted postoperative pulmo- [50-ATGGCCGTTCTTAGTTGGTG-30 (forward) and 30- nary reserve. At our institution, EPP is more commonly CGCTGAGCCAGTCAGTGTAG-50 (reverse)]. carried out over pleurectomy/decortication for MPM and likely provides more complete resection, especially for Immunohistochemistry and tissue microarray patients with bulky disease. Patient cohorts for PCR and Immunostaining was carried out on a tissue microarray immunohistochemical experiments were based on previ- (TMA) of 58 resected MPM tumors as previously described ously published data sets constructed to represent institu- (7). Each tumor was represented in duplicate or quadru- tional experience (15) and their clinicopathologic charac- plicate cores. Six patients were excluded because of damage teristics are shown in Supplementary Tables S1 and S2. For to the tissue cores, resulting in a sample size of 52. In brief, flow cytometry experiments, tissues were obtained from the TMA slides were soaked in xylene, passed through the operating room in patients undergoing cytoreductive graded alcohols, and then placed in distilled water. The surgery for biopsy-proven MPMs. slides were next treated with citrate buffer in a steam Tissues. The pathologic diagnosis of MPM was con- pressure cooker (Decloaking Chamber; BioCare Medical) firmed in all cases. For PCR experiments, tumor samples and then washed in distilled water. All further steps were

www.aacrjournals.org Clin Cancer Res; 18(6) March 15, 2012 1569

Burt et al.

carried out at room temperature in a hydrated chamber. To concentration of 20 ng/mL. Cellular proliferation was mea- quench endogenous peroxidase activity, the slides were sured at indicated time points using an MTT cell prolifer- pretreated with peroxidase block (DAKO) for 5 minutes. ation assay (ATCC), and apoptosis was assessed via mea- A mouse monoclonal anti-human IL-4Ra antibody (H-4, surement of caspase-3/7 activity with a luminescence-based SC-28361; Santa Cruz Biotechnology) was applied at 1:100 assay (Promega). ELISA was used to assay the supernatant in DAKO diluent for 1 hour. Afterward, the slides were for cytokines IL-6, IL-8, and VEGF, 24 hours after the washed in 50 mmol/L Tris-Cl, pH 7.4, and detected with addition of IL-4. In-Cell ELISA assays were used to quantify Mouse Envision kit (DAKO) as per manufacturer’s instruc- changes in the concentration of STAT-6 and phosphorylated tions. After another washing step, immunoperoxidase stain- STAT-6, after 24 hours of culture in 20 ng/mL of IL-4. Final ing was developed using a DAB chromogen (DAKO) and values were normalized to cell number using Janus Green counterstained with hematoxylin. For image analysis, TMA reagent. To detect IL-4 in tumor tissues, snap-frozen MPM slides stained with IL-4Ra were scanned at 200 magnifi- tumors were homogenized in buffer containing Iscove’s cation using an Aperio ScanScope XT workstation (Aperio Medium (Gibco), 10% FCS, 0.5 mmol/L amino-n-caproic Technology, Inc.). The images were visualized and anno- acid (Sigma), and 0.05 mmol/L EDTA (Sigma). An ultra- tated using ImageScope software (version 10.0.35.1800; sensitive ELISA detection kit was used to detect IL-4 in the Aperio Technology). For each core, 3 regions of interest tissue homogenate. All ELISA kits were purchased from (ROI) were selected by a pathologist (S.J. Rodig) based on Invitrogen. areas containing predominantly tumor cells and thereby excluding the stromal compartment which harbors most of Statistical analysis the immune infiltrate (Supplementary Fig. S1). Aperio Statistical analysis was conducted using JMP statistical software (positive pixel count v9 algorithm) was used to software (version 8.0; SAS Institute). Distributions were deconvolute the brown (positive staining for the antigen) plotted to examine compliance with distributional assump- and blue (hematoxylin stain) colorations. The percentage of tions. Nonparametric tests were used to compare dichoto- the annotated area was scored by a software algorithm based mous variables in univariate analyses. The Wilcoxon test upon a scale of pixel intensity ranging from 0 to 256, with 0 was used to compare 2 groups of continuous variables representing black and 256 representing white. Staining and one-way ANOVA was used to compare 3 groups of intensity thresholds were set as follows: strong (0–100), continuous variables. Kaplan–Meier plots were used to not strong (101–256). A representation of strong staining is illustrate the cumulative proportion surviving as a func- shown in Supplementary Fig. S1A and a representation of tion of months since surgery, and differences in survival not strong staining is shown in Supplementary Fig. S1B. were tested for significance by the log-rank test. Cox pro- Data were collected for each ROI and the results were portional hazards regression was used to test the predictive averaged. Data from duplicate cores were similarly capacity of IL-4Ra expression in univariate and multi- averaged. variate models and to estimate 95% confidence intervals (CI) influencing survival. P values less than 0.05 were Flow cytometry and intracellular cytokine detection considered significant. Flow cytometry was carried out on mononuclear cell suspensions and cancer cell lines with a LSR II cytometer (BD Biosciences) using the following monoclonal antibo- Results dies: CD4-FITC, CD45-PE-Cy5, CD8-PE-Cy7, CD3-Pacific Increased expression of the IL-4Ra message is Blue (all from BD Biosciences); and IL-4Ra-APC (R&D predictive of poor survival in patients undergoing Systems). Matched immunoglobulin isotype antibodies resection for epithelial MPM and unstained cells were used as controls to set gates. Tumor expression of IL-4Ra was determined by RT- Detection of intracellular IL-4 was carried out on blood PCR of MPM tumors from 37 patients undergoing EPP. and tumor mononuclear cell suspensions 4 hours after Greater than median expression levels of IL-4Ra were activation with 5 ng/mL phorbol 12-myristate 13-acetate associated with significantly worse survival for patients (PMA; Sigma-Aldrich) and 1 mmol/L ionomycin (Sigma), in with epithelial (n ¼ 22), but not nonepithelial MPM (n ¼ the presence of 10 mg/mL Brefeldin A (BD Biosciences). 15; Fig. 1A). Univariate Cox proportional hazards model- Surface staining was completed and then the cells were fixed ing identified increased IL-4Ra expression as a predictor and permeabilized. Intracellular IL-4 was assayed using an of worse survival in patients with epithelial MPM; when allophycocyanin (APC)-conjugated anti-IL-4 antibody adjusted for age, gender, and stage, IL-4Ra expression (Intracellular Cytokine Detection kit; BD Biosciences). was retained as an independent predictor of survival Flow cytometry data were analyzed with FlowJo software (Table 1). Interestingly, a slightly higher level of IL-4Ra (TreeStar). expression was found in nonepithelial tumors than in epithelial tumors (Fig. 1B). The expression level of IL-4Ra Cell-based assays did not correlate with pathologic stage in the epithelial Five thousand cells were plated on 96-well cell culture group but could not be assessed in the nonepithelial plates in RPMI containing 10% FCS. Twenty-four hours group owing to the high distribution of stage III tumors later, recombinant human IL-4 (Invitrogen) was added at a in these patients (Fig. 1C).

1570 Clin Cancer Res; 18(6) March 15, 2012 Clinical Cancer Research

IL-4Ra in Human Pleural MPM

Figure 1. Expression of the IL-4Ra message predicts survival in human MPM. A, IL-4Ra mRNA expression was determined by RT-PCR in MPM tumor specimens obtained after EPP in patients with all (n ¼ 37), epithelial (EPI; n ¼ 22), and nonepithelial histology (NE; n ¼ 15). Patients were divided into groups of either greater than or less than median tumor IL- 4Ra expression. Expression of tumor IL-4Ra was compared between epithelial and nonepithelial tumors (B) and among tumors of varying pathologic stage (C). TNM, tumor- node-metastasis.

Increased expression of the IL-4Ra protein is predictive protein than nonepithelial tumors (27% 4% vs. 10% of poor survival in patients undergoing resection for 2%, Fig. 2D), and strong IL-4Ra staining was not associated epithelial MPM with pathologic stage (Fig. 2E). Because IL-4Ra is expressed To confirm the expression of IL-4Ra on tumor cells from on tumor-infiltrating macrophages in human MPM (7), we human MPM tumor specimens, we carried out flow cyto- compared IL-4Ra and CD68 staining on epithelial and metry on the CD45-negative cell fraction of tumor mono- nonepithelial mesothelioma tumors (Supplementary Fig. nuclear cell suspensions, where high frequencies of IL-4Ra– S1). While IL-4Ra strongly stained the tumor cell compart- positive cells were found (Fig. 2A). To determine the degree ment, there was less marked staining in the tumor stroma. of expression of the IL-4Ra protein on epithelial and non- Conversely, CD68 staining was predominantly found on epithelial MPM, and if expression of the IL-4Ra protein was stromal cells. associated with survival in MPM, we stained our TMA of tumors from 52 patients with an anti-IL-4Ra antibody (Fig. Tumor-infiltrating T cells in human MPM are polarized 2B). IL-4Ra staining was present on the tumors of 31 of 32 to produce IL-4 patients with epithelial tumors (97%) and 19 of 20 patients Because T lymphocytes are a chief source of IL-4 in with nonepithelial tumors (95%). Although a small, scat- tumors, we carried out intracellular cytokine analyses to tered fraction of cells in the stromal compartment of the determine whether tumor-infiltrating T cells in MPM were tumor stained positive for IL-4Ra, the great majority of IL- a potential source of endogenous IL-4. Mononuclear cells 4Ra staining was found on the tumor cells themselves. were freshly isolated from matched tumor and blood sam- Similar to the RT-PCR data, strong IL-4Ra staining was ples from 8 patients undergoing EPP for MPM (epithelial, associated with worse overall survival following EPP in n ¼ 4; nonepithelial, n ¼ 4). We found that high frequencies patients with epithelial MPM (Fig. 2C). A greater fraction of tumor-infiltrating CD3 T cells produced IL-4 after stim- of epithelial tumors showed strong staining for the IL-4Ra ulation (21% 7%) when compared with matched blood

Table 1. Association of IL-4Ra message with survival in MPM

Nonadjusted Adjusteda

Variable HR (95% CI) P HR (95% CI) P

All (n ¼ 37) 1.78 (1.32–2.47) <0.0001 1.81 (1.28–2.68) 0.0003 Epithelial (n ¼ 22) 2.97 (1.70–6.02) <0.0001 3.13 (1.68–7.15) <0.0001 Nonepithelial (n ¼ 15) 0.91 (0.50–1.66) 0.8 1.16 (0.57–2.51) 0.7

aAdjusted for age, gender, and stage.

www.aacrjournals.org Clin Cancer Res; 18(6) March 15, 2012 1571

Burt et al.

Figure 2. Expression of the IL-4Ra protein is associated with survival in human MPM. A, IL-4Ra expression on tumor cells from human MPM tumors was confirmed by flow cytometry on CD45-negative tumor mononuclear cells (n ¼ 4). IL-4Ra protein expression was determined by immunohistochemistry on MPM tumor specimens obtained after EPP for patients with all (n ¼ 52), epithelial (EPI; n ¼ 32), and nonepithelial histology (NE; n ¼ 20). B, representative cores from epithelial and nonepithelial MPM stained with IL-4Ra. C, Kaplan–Meier survival curves are shown for patients with either high or low tumor expression of IL-4Ra. D, the percentage of patients with high tumor expression of IL-Ra was compared between epithelial and nonepithelial histology. E, the percentage of patients with high tumor expression of IL-4Ra was stratified by pathologic stage. TNM, tumor-node-metastasis.

þ þ CD3 T cells from the same patients (4% 2%; Fig. 3A and the tumor, representing 63% 12% of tumor CD3 IL-4 T B), suggesting that T cells within MPM tumors are skewed to cells, and CD4 T cells were the major T-cell subset respon- produce IL-4. Although both CD4 and CD8 tumor-inﬁl- sible for IL-4 production in the blood, representing 56% þ þ trating T cells produced IL-4 (Fig. 3C), CD8 T cells were the 14% of blood CD3 IL-4 T cells (Fig. 3D). This difference major T-cell subset responsible for IL-4 production within was due mainly to a greater number of CD8 than CD4 T cells

Figure 3. Tumor-infiltrating T cells in human MPM are polarized to produce IL-4. Mononuclear cells were freshly isolated from matched blood and tumor specimens from patients undergoing surgical resection of MPM. Intracellular IL-4 production was determined in CD3 T cells by intracellular cytokine analysis after 4 hours of nonspecific stimulation with PMA and ionomycin. A, flow cytometric dot plots from one representative patient. B, summary data of matched þ þ tumor and blood CD3 IL-4 cells from 8 patients. C, representative contour plot of intratumoral CD4 and CD8 T-cell subsets and intracellular IL-4 production. D, CD4 and CD8 subset composition of CD3þIL-4þ T cells (n ¼ 5). E, the percentage of CD4 and CD8 T cells that stain positive for IL-4 (n ¼ 5). IL-4 concentrations (F) and IL-13 concentrations (G) in tumor (n ¼ 8), effusion (n ¼ 8), and serum (n ¼ 10) of patients with MPM was determined by ultrasensitive ELISA. Values represent pg/mg of tissue and pg/mL of effusion and serum.

1572 Clin Cancer Res; 18(6) March 15, 2012 Clinical Cancer Research

IL-4Ra in Human Pleural MPM

within the tumor, as a similar number of CD8 and CD4 T cells in each compartment stained positive for IL-4 (Fig. 3E). þ Similar numbers of IL-4 tumor-infiltrating T cells were found in epithelial and nonepithelial tumors (not shown). To further characterize the phenotype of tumor-infiltrating T cells in human MPMs, we assayed for T-cell production of IFN-g and IL-10. Interestingly, we found that tumor-infiltrating T cells in human MPM were significant producers of IFN-g in response to PMA/ionomycin when compared with matched circulating T cells (Supplementary Fig. S2A, 47% 4% tumor, 20% 12% blood, P ¼ 0.02). A small frequency of both tumor-infiltrating T cells and matched circulating T cells produced IL-10 in response to PMA/ionomycin stimulation (Supplemental Fig. S2B, 5% 1% tumor, 3% 1% blood, P ¼ 0.2). To determine whether the ligands for the IL- 4 receptor are present in MPM tumors, we used ELISA to detect IL-4 and IL-13 concentrations within the tumor tissue. Low levels of IL-4 were measured in tumor tissues (n ¼ 8), whereas IL-4 was undetectable in the serum (n ¼ 10) and effusion (n ¼ 8) of patients with MPM (Fig. 3F). IL- Figure 4. Human MPM cells respond to IL-4 by activation of STAT-6 13 was present in all compartments, with higher levels signaling. A, the presence of IL-4Ra on 4 human MPM cell lines was found in tumors and effusions than in serum (Fig. 3G). determined by flow cytometry. Shaded histograms represent IL-4Ra staining and empty histograms represent isotype controls. B, each human MPM cell line was cultured for 24 hours in the presence of STAT-6 signaling is activated in human MPM cancer 20 ng/mL of IL-4, and the level of STAT-6 and phosphorylated (p-) STAT-6 cells in response to IL-4 was determined by intracellular ELISA. These experiments were carried Because human MPM tumor cells express high in situ out a minimum of 3 times. levels of IL-4Ra and tumor-infiltrating T cells in MPM are skewed toward an IL-4–producing phenotype, we asked MPM cells. Culture in IL-4 resulted in increased production whether IL-4 had a biologic effect on MPM tumor cells. To of the angiogenic cytokine, VEGF, in H2052, MS924, and investigate the function of IL-4Ra on human MPM cancer JMN cells. Although all MPM cell lines made significant cells, we studied 4 human MPM cell lines. We first con- quantities of TGF-b, no increase in production was found in firmed that each of these cell lines showed cell surface response to IL-4, and no MPM cell lines made significant expression of IL-4Ra using flow cytometry (Fig. 4A). We quantities of TNF-a, macrophage colony-stimulating factor, then cultured MPM tumor cells in the presence of IL-4 to or monocyte chemotactic protein 1 (MCP-1) while resting determine whether IL-4 caused activation of STAT-6 signal- or in response to IL-4 stimulation (not shown). ing. Culture with IL-4 increased the levels of phosphorylated STAT-6 in each of the 4 MPM cells lines by 29% to 46%, suggesting that the IL-4Ra receptor complex on MPM Discussion tumor cells is functional (Fig. 4B). IL-4Ra is the major subunit of the IL-4 receptor and a primary IL-4–binding protein. After binding of IL-4 and Human MPM cancer cells respond to IL-4 via receptor complex formation, intracellular signaling ensues production of inflammatory and angiogenic cytokines through phosphorylation of STAT-6 (19, 20). IL-4Ra is IL-4 is a pleiotropic cytokine that has been shown to have expressed on a number of cancer cell lines derived from varied and sometimes opposing effects on tumors that nonhematologic human malignancies that include pancre- include both increasing and decreasing the proliferation atic adenocarcinoma, renal cell carcinoma, melanoma, and survival of malignant cells (18). To determine whether ovarian carcinoma, Kaposi sarcoma, glioblastoma, colon IL-4 could provide a direct survival benefit to human MPM carcinoma, gastric carcinoma, breast carcinoma, lung car- cells, we cultured MPM tumor cells in the presence of IL-4 cinoma, and head and neck cancers (8, 21). Moreover, and found no change in proliferation or apoptosis (Fig. 5A elevated expression of IL-4Ra has been found on several and B). To determine whether IL-4 had proinflammatory or human tumors in situ. For example, IL-4Ra has been anti-inflammatory effects on MPM tumors cells, we assayed detected in 66% to 79% of non–small cell lung cancer cell culture supernatants for production of various cyto- tumors (9), 60% of ovarian carcinomas (10), 77% of head kines including IL-6, IL-8, and VEGF. Addition of IL-4 and neck cancers (11), and is also present on pancreatic, resulted in statistically significant increases in production breast, bladder, and prostate carcinomas (12). of the inflammatory cytokine IL-6 on each of the 4 MPM cell In addition to its expression on tumor cells, IL-4Ra is lines (Fig. 5C). MS924 and JMN cells similarly increased detected on circulating and tumor-infiltrating myeloid production of the inflammatory cytokine IL-8, whereas no immune cells with immunosuppressive properties in can- change in IL-8 production was seen in H2052 or MS428 cer-bearing hosts. IL-4Ra is a marker of the recently

www.aacrjournals.org Clin Cancer Res; 18(6) March 15, 2012 1573

Burt et al.

Figure 5. Human MPM cells produce proinﬂammatory and angiogenic cytokines in response to IL-4. Four human MPM cell lines were cultured in the presence of 20 ng/mL of IL-4. On days 1, 3, and 5 thereafter, cellular proliferation was assayed by MTT (A), and apoptosis was assayed by determination of caspase-3/7 activity measured by relative light units (B). Twenty-four hours after the addition of IL-4, supernatants were assayed for IL- 6, IL-8, and VEGF via ELISA (C). These experiments were carried out a minimum of 3 times. RLU, relative light units.

described myeloid-derived suppressor cells in tumor-bear- expression of the entire tumor mass (tumor cells and ing mice and also identifies a fraction of circulating mono- stromal cells that include macrophages), whereas our nuclear and polymorphonuclear cells with T-cell suppres- immunohistochemical experiments were designed to focus sive function in patients with melanoma and colon cancer on tumor cell expression of IL-4Ra. To our knowledge, this (22, 23). In human MPM, we have found that IL-4Ra is is the first study to equate tumor expression of IL-4Ra with highly expressed on a population of tumor-infiltrating survival indices in patients with cancer. macrophages that displays an immunoregulatory pheno- The distinction of Th1 and Th2 immune responses has type (7). Given the high expression of IL-4Ra on multiple important implications for patients with cancer. Antitumor human cancers and the tumor-promoting effects of IL-4 on immune responses are sustained principally by tumor-infil- human cancer cells (18), it is likely that IL-4Ra plays a role trating lymphocytes consisting primarily of T cells. While in tumor progression in humans. For example, IL-4Ra Th1 responses typically signify the production of IL-2 and competent murine tumors exhibit increased tumor growth IFN-g with activation of cytotoxic T-cell responses, Th2 when compared with tumors from IL-4Ra–deficient coun- responses generally represent a cytokine profile including terparts (24). Herein, we show that patients with epithelial IL-4, IL-5, IL-6, and IL-13 (25), of which IL-4 is the proty- MPM whose tumors show high IL-4Ra expression have a pical Th2 cytokine. IL-4 has differential effects on immunity significant survival disadvantage after surgical resection. that could be advantageous to tumor growth. These include The reason for this difference between epithelial and none- inhibition of Th1 cell differentiation and cellular immunity, pithelial MPM is not exactly clear; however, nonepithelial abrogation of CD8 T-cell cytotoxicity, accumulation of Th2 MPM is a particularly aggressive tumor and the immune cytokines, and facilitation of Th2 CD4 T-cell development response to epithelial and nonepithelial MPM is distinct (26, 27). Most clinical studies support the finding of an (7). It should be noted that our PCR data represent IL-4Ra abnormal Th1 to Th2 ratio in patients with various cancers

1574 Clin Cancer Res; 18(6) March 15, 2012 Clinical Cancer Research

IL-4Ra in Human Pleural MPM

and it has been shown that the nature of the immune Taken together, the high expression of IL-4Ra on human response (Th1 vs. Th2) can affect outcome in patients with tumor cells, its association with protumor function, and its cancer. Specifically, high serum and tumor IL-4 levels, and correlation with survival reported here make it a promising high frequencies of IL-4–secreting T cells are elevated in therapeutic target. Puri’s group has targeted tumor IL-4Ra patients with cancer and parallel worse prognosis (28–30). with a chimeric protein consisting of a circularly permu- Our data show that the natural immune reaction to human tated IL-4 molecule that is fused to a truncated from of MPM involves tumor-infiltrating T cells skewed toward an Pseudomonas exotoxin. This agent has been shown to have IL-4–producing phenotype. potent cytotoxic activity in vitro against numerous IL-4Ra The direct effects of IL-4 on tumor cells are varied and expressing cell lines and in vivo in xenograft models of sometimes opposing. In human colon cancer cell lines, various human cancers including MPM (8, 14, 46, 47). treatment with IL-4 results in dose-dependent increases in Ultimately, this molecule has been brought to clinical trials cellular proliferation (31), and IL-4 is an autocrine growth and has shown encouraging results in patients with glio- factor made by pancreatic cancer cells (21). IL-4 protects blastoma (48). human prostate, breast, and cancer cell lines from chemo- In summary, MPM is a malignancy whose origin and therapy and death receptor–induced apoptosis through development are deeply rooted in inflammation. We have upregulation of antiapoptotic proteins cFLIP/FLAME-1 and found that IL-4Ra is highly expressed in situ by tumor cells Bcl-xL (12). Similar results are seen in murine in vivo tumor in human MPM tumors, tumor expression of IL-4Ra is models where IL-4 increases tumor growth, reduces apopto- predictive of survival in patients undergoing cytoreductive sis, and promotes metastases (24, 32). Conversely, IL-4 is surgery for epithelial MPM, and that MPM tumor cells required for the development of effective antitumor immu- respond to IL-4 via upregulation of STAT-6 signaling and nity (33). For instance, IL-4 can have a direct inhibitory effect inflammatory cytokine production. We hypothesize that IL- on proliferation of human cancer cells (34, 35), and IL-4 can 4 endogenously produced by MPM tumor-infiltrating T cells be a potent antitumor agent in some animal models (36, 37). provides stimulus to the IL-4 receptor on MPM tumor cells It has been suggested that the differential positive and neg- in vivo and promotes inflammatory and angiogenic signals ative effects of IL-4 on tumor development may be explained that stimulate tumorigenesis. Taken together, it seems that by the source of IL-4, with exogenously delivered IL-4 exert- mesothelioma tumors with high IL-4Ra expression are ing tumor-suppressing effects and endogenously produced clinically more aggressive cancers that have worse outcome IL-4 exerting tumor-promoting effects (18). after surgical resection. These data have important implica- Our data show that IL-4 did not have a direct effect on tions for future research and clinical applications. IL-4Ra is MPM tumor cell line proliferation or apoptosis. Activation a potentially useful marker of prognosis in patients under- of human MPM cells with IL-4 resulted in upregulation of going surgical resection and possibly also for patients STAT-6 phosporylation and an increase in production of undergoing definitive chemotherapy; however, this inflammatory and angiogenic cytokines including IL-6, IL- requires further investigation. Moreover, the IL-4/IL-4Ra 8, and VEGF. Each of these cytokines has been shown to be axis represents a promising target for directed therapy in produced by MPM cell lines where they act as potent man, with potential application for IL-4 and IL-4 receptor autocrine growth factors and are each detected at significant blocking agents and IL-4Ra–specific cytotoxins. levels in the effusions of patients with MPMs (38–42). IL-6 is a multifunctional, proinflammatory cytokine involved in Disclosure of Potential Conflicts of Interest acute phase inflammatory responses and additionally sti- No potential conflicts of interest were disclosed. mulates VEGF production by MPM cells (43). It has been Acknowledgments suggested that IL-6 may be responsible for the systemic The authors thank Dr. Christina Wei for technical assistance. manifestations of MPM that include fever, cachexia, and thrombocytosis (44). IL-8 is a member of the C-X-C che- Grant Support mokine family and has direct growth-potentiating and This study was supported by the International Mesothelioma Program of angiogenic effects on tumors. VEGF is a powerful mitogen the Brigham & Women’s Hospital. The costs of publication of this article were defrayed in part by the for the vascular endothelium, and in patients with MPM, payment of page charges. This article must therefore be hereby marked elevated serum levels of VEGF correlate with advanced advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate disease stage and worse survival (45). It is reasonable to this fact. surmise that each of these factors has tumor-promoting Received July 20, 2011; revised December 22, 2011; accepted December effects in the tumor microenvironment. 31, 2011; published OnlineFirst January 18, 2012.

References 1. Sugarbaker DJ, Flores RM, Jaklitsch MT, Richards WG, Strauss GM, 2. Herndon JE, Green MR, Chahinian AP, Corson JM, Suzuki Y, et al. Resection margins, extrapleural nodal status, and cell type Vogelzang NJ. Factors predictive of survival among 337 determine postoperative long-term survival in trimodality therapy of patients with mesothelioma treated between 1984 and 1994 malignant pleural mesothelioma: results in 183 patients. J Thorac by the Cancer and Leukemia Group B. Chest 1998;113:723– Cardiovasc Surg 1999;117:54–63. 31.

www.aacrjournals.org Clin Cancer Res; 18(6) March 15, 2012 1575

Burt et al.

3. Mantovani A, Allavena P, Sica A, Balkwill F. Cancer-related inflamma- 25. Abbas AK, Murphy KM, Sher A. Functional diversity of helper T tion. Nature 2008;454:436–44. lymphocytes. Nature 1996;383:787–93. 4. Carbone M, Kratzke RA, Testa JR, Elmishad AG, Chen Y, Liu Z, et al. 26. Murphy KM, Reiner SL. The lineage decisions of helper T cells. Nat Rev The pathogenesis of mesothelioma. Semin Oncol 2002;29:2–17. Immunol 2002;2:933–44. 5. Yang H, Bocchetta M, Kroczynska B, Elmishad AG, Chen Y, Liu Z, et al. 27. Kienzle N, Olver S, Buttigieg K, Grobes P, Janas ML, Baz A, et al. TNF-alpha inhibits asbestos-induced cytotoxicity via a NF-kappaB- Progressive differentiation and commitment of CD8þ T cells to a dependent pathway, a possible mechanism for asbestos-induced poorly cytolytic CD8low phenotype in the presence of IL-4. J Immunol oncogenesis. Proc Natl Acad Sci U S A 2006;103:10397–402. 2005;174:2021–9. 6. Anraku M, Cunningham KS, Yun Z, Tsao MS, Zhang L, Keshavjee S, 28. Asselin-Paturel C, Echchakir H, Carayol G, Gay F, Opolon P, Grunen- et al. Impact of tumor-infiltrating T cells on survival in patients with wald D, et al. Quantitative analysis of Th1, Th2 and TGF-beta1 cytokine malignant pleural mesothelioma. J Thorac Cardiovasc Surg 2008;135: expression in tumor, TIL and PBL of non-small cell lung cancer 823–9. patients. Int J Cancer 1998;77:7–12. 7. Burt BM, Rodig SJ, Tilleman TR, Elbardissi AW, Bueno R, Sugarbaker 29. Nakagomi H, Pisa P, Pisa EK, Yamamoto Y, Halapi E, Backlin K, DJ. Circulating and tumor-infiltrating myeloid cells predict survival in et al. Lack of interleukin-2 (IL-2) expression and selective expres- human pleural mesothelioma. Cancer 2011;117:5234–44. sion of IL-10 mRNA in human renal cell carcinoma. Int J Cancer 8. Kawakami K, Kawakami M, Puri RK. Overexpressed cell surface 1995;63:366–71. interleukin-4 receptor molecules can be successfully targeted for 30. Roussel E, Gingras MC, Grimm EA, Bruner JM, Moser RP. Predom- antitumor cytotoxin therapy. Crit Rev Immunol 2001;21:299–310. inance of a type 2 intratumoural immune response in fresh tumour- 9. Kawakami M, Kawakami K, Stepensky VA, Maki RA, Robin H, Muller W, infiltrating lymphocytes from human gliomas. Clin Exp Immunol et al. Interleukin 4 receptor on human lung cancer: a molecular target 1996;105:344–52. for cytotoxin therapy. Clin Cancer Res 2002;8:3503–11. 31. Koller FL, Hwang DG, Dozier EA, Fingleton B. Epithelial interleukin-4 10. Kioi M, Takahashi S, Kawakami M, Kawakami K, Kreitman RJ, Puri RK. receptor expression promotes colon tumor growth. Carcinogenesis Expression and targeting of interleukin-4 receptor for primary and 2010;31:1010–7. advanced ovarian cancer therapy. Cancer Res 2005;65:8388–96. 32. Kobayashi M, Kobayashi H, Pollard RB, Suzuki F. A pathogenic role of 11. Strome SE, Kawakami K, Alejandro D, Voss S, Kasperbauer JL, Th2 cells and their cytokine products on the pulmonary metastasis of Salomao D, et al. Interleukin 4 receptor-directed cytotoxin therapy for murine B16 melanoma. J Immunol 1998;160:5869–73. human head and neck squamous cell carcinoma in animal models. Clin 33. Schuler T, Qin Z, Ibe S, Noben-Trauth N, Blankenstein T. T helper cell Cancer Res 2002;8:281–6. type 1-associated and cytotoxic T lymphocyte-mediated tumor immu- 12. Conticello C, Pedini F, Zeuner A, Patti M, Zerilli M, Stassi G, et al. IL-4 nity is impaired in interleukin 4-deficient mice. J Exp Med 1999;189: protects tumor cells from anti-CD95 and chemotherapeutic agents via 803–10. up-regulation of antiapoptotic proteins. J Immunol 2004;172:5467–77. 34. Obiri NI, Hillman GG, Haas GP, Sud S, Puri RK. Expression of high 13. Shurin MR, Lu L, Kalinski P, Stewart-Akers AM, Lotze MT. Th1/Th2 affinity interleukin-4 receptors on human renal cell carcinoma cells and balance in cancer, transplantation and pregnancy. Springer Semin inhibition of tumor cell growth in vitro by interleukin-4. J Clin Invest Immunopathol 1999;21:339–59. 1993;91:88–93. 14. Beseth BD, Cameron RB, Leland P, You L, Varricchio F, Kreitman RJ, 35. Toi M, Bicknell R, Harris AL. Inhibition of colon and breast carcinoma et al. Interleukin-4 receptor cytotoxin as therapy for human malignant cell growth by interleukin-4. Cancer Res 1992;52:275–9. pleural mesothelioma xenografts. Ann Thorac Surg 2004;78:436–43. 36. Tepper RI, Levinson DA, Stanger BZ, Campos-Torres J, Abbas AK, 15. Gordon GJ, Rockwell GN, Jensen RV, Rheinwald JG, Glickman JN, Leder P. IL-4 induces allergic-like inflammatory disease and alters T Aronson JP, et al. Identification of novel candidate oncogenes and cell development in transgenic mice. Cell 1990;62:457–67. tumor suppressors in malignant pleural mesothelioma using large- 37. Golumbek PT, Lazenby AJ, Levitsky HI, Jaffee LM, Karasuyama H, scale transcriptional profiling. Am J Pathol 2005;166:1827–40. Baker M, et al. Treatment of established renal cancer by tumor 16. Demetri GD, Zenzie BW, Rheinwald JG, Griffin JD. Expression of cells engineered to secrete interleukin-4. Science 1991;254: colony-stimulating factor genes by normal human mesothelial cells 713–6. and human malignant mesothelioma cells lines in vitro. Blood 1989;74: 38. Monti G, Jaurand MC, Monnet I, Chretien P, Saint-Etienne L, Zeng L, 940–6. et al. Intrapleural production of interleukin 6 during mesothelioma and 17. De Rienzo A, Dong L, Yeap BY, Jensen RV, Richards WG, Gordon GJ, its modulation by gamma-interferon treatment. Cancer Res 1994;54: et al. Fine-needle aspiration biopsies for gene expression ratio-based 4419–23. diagnostic and prognostic tests in malignant pleural mesothelioma. 39. Galffy G, Mohammed KA, Dowling PA, Nasreen N, Ward MJ, Antony Clin Cancer Res 2011;17:310–6. VB. Interleukin 8: an autocrine growth factor for malignant mesothe- 18. Li Z, Chen L, Qin Z. Paradoxical roles of IL-4 in tumor immunity. Cell Mol lioma. Cancer Res 1999;59:367–71. Immunol 2009;6:415–22. 40. DeLong P, Carroll RG, Henry AC, Tanaka T, Ahmad S, Leibowitz MS, 19. Nelms K, Keegan AD, Zamorano J, Ryan JJ, Paul WE. The IL-4 et al. Regulatory T cells and cytokines in malignant pleural effusions receptor: signaling mechanisms and biologic functions. Ann Rev secondary to mesothelioma and carcinoma. Cancer Biol Ther 2005;4: Immunol 1999;17:701–38. 342–6. 20. Kawakami K, Leland P, Puri RK. Structure, function, and targeting of 41. Schmitter D, Lauber B, Fagg B, Stahel RA. Hematopoietic growth interleukin 4 receptors on human head and neck cancer cells. Cancer factors secreted by seven human pleural mesothelioma cell lines: Res 2000;60:2981–7. interleukin-6 production as a common feature. Int J Cancer 1992;51: 21. Prokopchuk O, Liu Y, Henne-Bruns D, Kornmann M. Interleukin-4 296–301. enhances proliferation of human pancreatic cancer cells: evidence for 42. Strizzi L, Catalano A, Vianale G, Orecchia S, Casalini A, Tassi G, et al. autocrine and paracrine actions. Br J Cancer 2005;92:921–8. Vascular endothelial growth factor is an autocrine growth factor in 22. Gallina G, Dolcetti L, Serafini P, De Santo C, Marigo I, Colombo MP, human malignant mesothelioma. J Pathol 2001;193:468–75. et al. Tumors induce a subset of inflammatory monocytes with immu- 43. Adachi Y, Aoki C, Yoshio-Hoshino N, Takayama K, Curiel DT, Nishi- nosuppressive activity on CD8þ T cells. J Clin Invest 2006;116: moto N. Interleukin-6 induces both cell growth and VEGF production in 2777–90. malignant mesotheliomas. Int J Cancer 2006;119:1303–11. 23. Mandruzzato S, Solito S, Falisi E, Francescato S, Chiarion-Sileni V, 44. Nakano T, Chahinian AP, Shinjo M, Tonomura A, Miyake M, Togawa Mocellin S, et al. IL4Ralphaþ myeloid-derived suppressor cell expan- N, et al. Interleukin 6 and its relationship to clinical parameters in sion in cancer patients. J Immunol 2009;182:6562–8. patients with malignant pleural mesothelioma. Br J Cancer 1998;77: 24. Li Z, Jiang J, Wang Z, Zhang J, Xiao M, Wang C, et al. Endogenous 907–12. interleukin-4 promotes tumor development by increasing tumor cell 45. Yasumitsu A, Tabata C, Tabata R, Hirayama N, Murakami A, Yamada resistance to apoptosis. Cancer Res 2008;68:8687–94. S, et al. Clinical significance of serum vascular endothelial growth

1576 Clin Cancer Res; 18(6) March 15, 2012 Clinical Cancer Research

IL-4Ra in Human Pleural MPM

factor in malignant pleural mesothelioma. J Thorac Oncol 2010;5: ligand-directed toxin, with decreased immunogenicity and 479–83. potent anti-mesothelioma activity. Br J Cancer 2009;101:1114– 46. Husain SR, Kreitman RJ, Pastan I, Puri RK. Interleukin-4 receptor- 23. directed cytotoxin therapy of AIDS-associated Kaposi's sarcoma 48. Rand RW, Kreitman RJ, Patronas N, Varricchio F, Pastan I, Puri RK. tumors in xenograft model. Nat Med 1999;5:817–22. Intratumoral administration of recombinant circularly permuted inter- 47. Stish BJ, Oh S, Chen H, Dudek AZ, Kratzke RA, Vallera DA. leukin-4-Pseudomonas exotoxin in patients with high-grade glioma. Design and modiﬁcation of EGF4KDEL 7Mut, a novel bispeciﬁc Clin Cancer Res 2000;6:2157–65.

www.aacrjournals.org Clin Cancer Res; 18(6) March 15, 2012 1577

Expression of Interleukin-4 Receptor Alpha in Human Pleural Mesothelioma Is Associated with Poor Survival and Promotion of Tumor Inflammation

Bryan M. Burt, Andrew Bader, Daniel Winter, et al.

Clin Cancer Res 2012;18:1568-1577. Published OnlineFirst January 18, 2012.

Updated version Access the most recent version of this article at: doi:10.1158/1078-0432.CCR-11-1808

Supplementary Access the most recent supplemental material at: Material http://clincancerres.aacrjournals.org/content/suppl/2012/01/18/1078-0432.CCR-11-1808.DC1

Cited articles This article cites 48 articles, 18 of which you can access for free at: http://clincancerres.aacrjournals.org/content/18/6/1568.full#ref-list-1

Citing articles This article has been cited by 2 HighWire-hosted articles. Access the articles at: http://clincancerres.aacrjournals.org/content/18/6/1568.full#related-urls

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 Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://clincancerres.aacrjournals.org/content/18/6/1568. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.