Interleukin 13 Receptor α2 in Pancreatic Ductal ... · in human pancreatic cancer cell lines....

Published OnlineFirst January 12, 2010; DOI: 10.1158/1078-0432.CCR-09-2015

Cancer Therapy: Preclinical Clinical
Cancer
Research
Interleukin 13 Mediates Signal Transduction throughInterleukin 13 Receptor α2 in Pancreatic DuctalAdenocarcinoma: Role of IL-13 Pseudomonas Exotoxinin Pancreatic Cancer Therapy
Takeshi Shimamura, Toshio Fujisawa, Syed R. Husain, Bharat Joshi, and Raj K. Puri

Abstract

Authors' Aof CellularResearch,

Note: SuppResearch O

T. Shimam

Current addCity Univer

CorresponBranch, DiEvaluation29B, RoomPhone: 301

doi: 10.115

©2010 Am

www.aacr

Down

Purpose: Interleukin-13 receptor α2 (IL-13Rα2) is a tumor antigen that is overexpressed in certainhuman tumors. However, its significance and expression in pancreatic cancer is not known. It is alsonot known whether IL-13 can signal through IL-13Rα2 in cancer.Experimental Design: The expression of IL-13Rα2 was assessed in pancreatic cancer samples by immu-

nohistochemistry and in cell lines by flow cytometry and reverse transcription-PCR. The role of IL-13Rα2was examined by IL-13–induced signaling in pancreatic cancer cell lines. IL-13Rα2–positive tumors weretargeted by IL-13PE cytotoxin in vitro and in vivo in an orthotopicmurinemodel of human pancreatic cancer.Results: Of the pancreatic tumor samples 71% overexpressed moderate to high-density IL-13Rα2 chain

compared with normal pancreatic samples. IL-13 induced transforming growth factor-β1 promoter activityin IL-13Rα2–positive tumor cells and in cells engineered to express IL-13Rα2 but not in IL-13Rα2–negativeor RNA interference knockdown cells. c-Jun and c-Fos of the AP-1 family of nuclear factors were activatedby IL-13 only in IL-13Rα2–positive cells. In the orthotopic mouse model, IL13-PE significantly decreasedtumor growth when assessed by whole-body imaging and prolonged the mean survival time. Similar re-sults were observed in mice xenografted with a surgically resected human pancreatic tumor sample.Conclusions: These results indicate that IL-13Rα2 is a functional receptor as IL-13 mediates signaling

in human pancreatic cancer cell lines. IL-13 causes transforming growth factor-β activation via AP-1pathway, which may cause tumor induced immunosuppression in the host. In addition, IL13-PEcytotoxin may be an effective therapeutic agent for the treatment of pancreatic cancer. Clin Cancer

Res; 16(2); 577–86. ©2010 AACR.

Interleukin-13 (IL-13) plays a central role in immune re-sponse to cancer, infection, and inflammation, and binds totwo chains, IL-13Rα1 and IL-13Rα2. IL-13 binds to IL-13Rα1, and then recruits IL-4Rα to form a functionalIL-13R complex and subsequently mediates signal transduc-tion through the Janus-activated kinase (JAK)/signal trans-ducers and activators of transcription 6 (STAT6) pathway(1). On the other hand, IL-13Rα2 binds IL-13 with high

ffiliation: Tumor Vaccines and Biotechnology Branch, Divisionand Gene Therapies, Center for Biologics Evaluation andFood and Drug Administration, Bethesda, Maryland

lementary data for this article are available at Clinical Cancernline (http://clincancerres.aacrjournals.org/).

ura and T. Fujisawa contributed equally.

ress for T. Shimamura: Gastroenterology Division, Yokohamasity Graduate School of Medicine, Japan.

ding Author: Raj K. Puri, Tumor Vaccines and Biotechnologyvision of Cellular and Gene Therapies, Center for Biologicsand Research, Food and Drug Administration, NIH Building2NN20 HFM-735, 29 Lincoln Drive, Bethesda, MD 20892.-827-0471; Fax: 301-827-0449; E-mail: [email protected].

8/1078-0432.CCR-09-2015

erican Association for Cancer Research.

journals.org

Researcon Decembclincancerres.aacrjournals.org loaded from

affinity but it does not mediate signal transduction throughthe JAK/STAT6 pathway although the IL-13–IL-13Rα2 com-plex is internalized after binding (2). In reconstitution stud-ies, it was shown that IL-13Rα2 does not induce activationof the STAT6 transcription factor. However, upon cotrans-fection with IL-4Rα and IL-13Rα1 mRNA, IL-13Rα2 inhib-ited IL-13–induced STAT6 activation (3). Recently, it hasbeen reported that IL-13 can signal through IL-13Rα2 in amurine macrophage cell line and that IL-13 signaling wasSTAT6-independent and involved the AP-1 pathway to in-duce activation of transforming growth factor (TGF)-β1promoter activity (4). Whether IL-13 signals in cancerthrough IL-13Rα2 is not known.To target IL-13 receptors, we have developed IL13-PE

composed of wild type IL-13 andmutated Pseudomonas exo-toxin (PE38) and single-chain FV-anti-IL-13Rα2 antibodyand PE38 (5–10). These fusion cytotoxins and immunotox-ins irreversibly ADP-ribosylate the diphthamide residue ofelongation factor 2, using NAD+ as a cofactor (5, 11). As aconsequence, protein synthesis is inhibited and celldeath occurs. IL13-PE38 and anti-IL-13Rα2-PE38 have po-tent antitumor activities in IL-13R–expressing tumor cellsin vi tro (5–9) and in vivo (10, 12–15). IL-13R is

577

h. er 12, 2020. © 2010 American Association for Cancer

http://clincancerres.aacrjournals.org/

Translational Relevance

Although many approaches are used to treat patientswith pancreatic ductal carcinoma (PDA), median sur-vival has not improved beyond six months. Therefore,newer effective approaches are urgently needed. Weshow that interleukin-13 receptor α2 (IL-13Rα2), ahigh-affinity receptor of the IL-13R complex, is overex-pressed in 71% of PDA and that IL-13 can signalthrough the AP-1 pathway in pancreatic tumor cellsvia IL-13Rα2 and produce transforming growth fac-tor-β1. Furthermore, we show that IL-13 cytotoxinmediates remarkable antitumor effects against humanPDA in orthotopic cancer models developed from celllines and patient-derived tumor. Because IL-13 cyto-toxin is being evaluated in the clinic for safety andefficacy in glioblastoma patients, the current resultsare relevant to future clinical trials in pancreatic can-cer therapy.

Shimamura et al.

578


overexpressed in certain types of human cancer cells, in-cluding glioblastoma, head and neck cancer, kidney cancer,ovarian cancer, certain types of medulloblastoma, andKaposi's sarcoma (6, 16–19). Targeted toxins provide aunique opportunity for tumor therapy, offering an advan-tage of unique specificity and limiting potential toxicity tonormal tissues. A bispecific cytotoxin that consists ofDipther-ia toxin, epidermal growth factor, and human IL-13 is shownto be effective in inducing tumor cell killing in vitro and in vivoagainst the MiaPaCa-2 cell line (20). Furthermore, IL-13PE38QQR and IL-4PE38QQR have also shown cytotoxiceffects in some of the pancreatic cell lines in vitro (21).In the present study, we show that IL-13R is overex-

pressed in human pancreatic cancer and that an alternatepathway of IL-13–induced signal transduction is opera-tional in pancreatic cancer cell lines directly involving IL-13Rα2. Furthermore, we have targeted IL-13Rα2–positivepancreatic cancer cell lines with the IL-13Rα2 targeted cy-totoxin IL13-PE in vitro and in orthotopic animal modelsof human pancreatic cancer in vivo.

Materials and Methods

Cell culture, reagents, and tissue specimens. Cells wereobtained from the American Type Culture Collection andfrom ScienCell. Recombinant IL13-PE38 was generated aspreviously described by fusing wild-type human IL-13 withtruncated Pseudomonas exotoxin in a prokaryotic expressionsystem (6–9). IL-13 was obtained from PeproTech.Immunohistochemistry and flow cytometry. Immunohis-

tochemistry was done as described previously (17). The re-sults were scored on the basis of the density of stainingand were scored as -, +/-, +, and ++, indicating negative,weak, moderate, and strong staining, respectively, withfield positive score ranging between 0% and 10%, 11%

Clin Cancer Res; 16(2) January 15, 2010

Researcon Decembclincancerres.aacrjournals.org Downloaded from

and 50%, and 51% and 100%. Tissue sections for IL-13Rα2 were evaluated by two pathologists independently.Expression of IL-13Rα2 in pancreatic cancer cell lines andHPDE cells was assessed by flow cytometry (22).Protein synthesis inhibition assay. In vitro cytotoxic activ-

ity of IL-13 cytotoxin was measured by the inhibition ofprotein synthesis as described earlier (6–9).Reverse transcription-PCR. Total RNA was isolated and

reverse transcription-PCR was done using gene-specificprimers as described previously (1).Stable transfection of pancreatic cancer cells. For the IL-

13Rα2 knockdown inMIA-PaCa-2 cells, retrovirus-mediatedRNA interferencewas donewith the pSuper RNA interferencesystem (Oligoengine) following the manufacturer's instruc-tions with minor modification. The target sequence forIL-13Rα2 was 5′-TGGATCATCAGAGAACAAGCC-3′. IL-13Rα2 overexpression in HPAF-II cells and green fluorescentprotein (GFP) expression in MIA-PaCa-2 and HS766Tcells were determined as described previously (23, 24).In vivo animal study. Severe combined immunodefi-

cient (SCID) mice and nude nu/nu mice between age 5and 6 wk were maintained in a barrier facility on high effi-ciency particulate air–filtered racks. All animal studieswere conducted under an approved protocol in accordancewith the principles and procedures outlined in the NIHGuide for the Care and Use of Laboratory Animals.Surgical orthotopic implantation of pancreatic tumors.

Two different pancreatic cancer cell lines expressing GFPprotein were injected s.c. into the right dorsal flank of nudemice. S.c. tumors grown for 14 d were harvested and cutinto pieces (∼3 × 3 × 3 mm) under aseptic conditions.For orthotopic tumor implant, the pancreas was carefullyexposed, and tumor chunks from nude mice and clinicaltissue were transplanted on the middle of the pancreas us-ing 6-0 Dexon surgical suture (Davis and Geck Inc.).Orthotopic pancreatic cancer model using a clinical tumor

sample. Primary pancreatic cancer specimens were obtainedfromapatient undergoing radical pancreatectomy at theNa-tional Cancer Institute under institutional review board–ap-proved protocol (sample kindly provided by Dr. RichardRoyal, Surgery Branch). Viable tumor tissue pieces from tu-mor specimenwere cut into small pieces (3 × 3 × 3mm) andimplanted in the pancreas of 5- to 6-week-old male SCIDmice. Primary cell lines were also established from the can-cer specimen for protein synthesis inhibition assay.Experimental design and monitoring tumor growth. Primary

tumor lesions were detected by external whole-body imag-ing on day 4 after transplantation. Once the tumors werevisualized, the mice were randomized into four or fivegroups of 10 mice each. The mice then received injectionsof vehicle (0.2% human serum albumin in PBS) or IL-13cytotoxin either by bolus i.p. injection (Bip; 10, 25, 50 μg/kg/d ×7) or continuous i.p. infusion (Cip; 50 μg/kg/d for7 d) beginning on day 5 for early tumor model or day 31for advanced metastatic model. Continuous administra-tion was done by loading a miniosmotic pump (ALZET)with 100 μL of IL-13 cytotoxin. Real-time determinationof tumor burden was done by quantifying fluorescent

Clinical Cancer Research



IL-13Rα2 as a Target for Therapy in Pancreatic Cancer


surface area as described previously (24). The vital organssuch as liver, kidney, spleen, lung, and heart from the trea-ted and untreated group of mice were harvested to evaluatethe general toxicity of the cytotoxin by H&E staining, andthe results were evaluated by two independent investigators.Luciferase assay for TGFβ1 promoter activity. A plasmid

containing the human TGFB1 promoter and luciferase re-porter gene was provided by J. Nam (US National CancerInstitute, NIH). Cells were transiently transfected with theluciferase reporter plasmid pTGF-β1-luciferase (250 ng)and the pSV-β-galactosidase vector (25 ng; Promega).Data are expressed as mean ± S.D. of triplicate determina-tions from three independent experiments.AP-1 activation assay. Cells were stimulated in the pres-

ence or absence of IL-13 and TGF-β1 (10 ng/mL each) for4 h. Nuclear extracts were collected using the Nuclear Ex-tract Kit (Active Motif) and tested for DNA binding activityusing the AP-1 family TransAM Kit (Active Motif).

www.aacrjournals.org


Statistical analysis. The difference in mean tumor vol-ume was analyzed by ANOVA. Survival curves were gener-ated by Kaplan-Meier method and compared by using thelog-rank test.

Results

Pancreatic ductal adenocarcinoma tissues and cancer celllines express IL-13Rα2. Tissue sections from 17 normalpancreas and 71 pancreatic ductal adenocarcinoma(PDA) specimens were analyzed by immunohistochemicalanalysis (Supplementary Table S1). Figure 1A shows rep-resentative normal pancreas and tumor specimens withnegative to strong staining for IL-13Rα2 in PDAs. Whenthe proportion of IL-13Rα2–positive cancer cells wascounted, 27 of 73 (37%) primary tumors classified intostrong expression group, 25 (34%) into the moderateexpression group, 16 (22%) into the weak expression

Fig. 1. Expression of IL-13Rα2 in PDA sections and pancreatic cancer cell lines. A, surgically resected normal pancreas and tumor tissues were stained withantihuman IL-13Rα2 polyclonal antibody. Representative results are shown from 71 tumor specimens. Magnification, ×200. Arrows, pancreatic cancer cells.B, RNA was isolated from normal pancreas and PDA sections, and gene expression was evaluated by reverse transcription-PCR. Renal cell carcinomacells (PM-RCC) were used as a positive control. C, RNA was isolated from pancreatic cancer and normal HPDE cell lines, and gene expression wasevaluated by reverse transcription-PCR analysis. D, cell surface expression of IL-13Rα2 was assessed by flow cytometry using phycoerythrin-conjugatedanti-IL-13Rα2 monoclonal antibody (red line). Staining with isotype-matched IgG served as a control (green line). PMRCC cells were used as apositive control.

Clin Cancer Res; 16(2) January 15, 2010 579



Shimamura et al.

580


group, and 5 (7%) into the negative expression group.Thus, 52 of 73 (71%) PDA samples expressed moderateto high-density IL-13Rα2. In contrast, only 4 of 17(24%) normal pancreas samples showed weak stainingfor IL-13Rα2 in normal acinar and ductal cells. Similarresults by reverse transcription-PCR analysis confirmedimmunohistochemistry findings (Fig. 1B).We also examined the expression of IL-13Rα2 mRNA by

reverse transcription-PCR in eight pancreatic cancer linesand one normal HPDE cell line. Four of eight PDA cell lines(SW1990, MIA-PaCa-2, KLM, and HS766T) showed a var-ied density of IL-13Rα2 mRNA expression, whereas fourPDA cell lines (Panc-1, Capan-1, HPAC, and HPAF-II)and the normal HPDE cell line showed no expression. Thisanalysis showed that the HS766T cell line expressed thehighest level of IL-13Rα2 mRNA, followed by the MIA-Pa-Ca-2 and KLM cell lines (Fig. 1C). Flow cytometric analysisconfirmed mRNA expression data and showed that IL-13Rα2 is expressed on the cell surface of three pancreaticcancer cell lines but not in normal HPDE cells (Fig. 1D).MIA-PaCa-2 pancreatic cancer cells were knocked down

for the IL-13Rα2 gene and HPAF-II cells were engineeredto overexpress the IL-13Rα2 gene using retroviral transduc-tion technique. Reverse transcription-PCR analysis con-firmed that IL-13Rα2 was sufficiently knocked down inMIA-PaCa-2 cells and HPAF-II cells began to overexpressIL-13Rα2 mRNA (Supplementary Fig. S1). However, themRNA levels of IL-13Rα1 and IL-4Rα showed no differ-ence between mock and transfected cells.IL-13 signaling in PDA through IL-13Rα2. To investigate

the functional significance of expression of IL-13Rα2 inPDA, IL-13–induced signal transduction was studied. Ithas been reported that a macrophage cell line transfectedwith IL-13Rα2 activates TGFB1 promoter activity throughthe AP-1 signal transduction pathway (4). Therefore, westimulated naturally IL-13Rα2–positive MIA-PaCa-2 cellswith IL-13 and observed that TGFB promoter activitywas induced. In contrast, no TGFB1 promoter activitywas induced by IL-13 in IL-13Rα2 knockdown MIA-Pa-Ca-2 cells (Fig. 2A). Gene transfer of IL-13Rα2 in HPAF-II cells also induced significant TGFB1 promoter activitywhen stimulated with IL-13. In contrast, TGFB1 promoteractivity was not induced in HPAF-II mock cells (IL-13Rα2negative; Fig. 2B). As IL-13Rα2 is not involved in signal-ing induced by IL-4 and TGF-β1, stimulation by these twocytokines did not increase TGFB1 promoter activity inboth MIA-PaCa-2 and HPAF-II cells, indicating that IL-13Rα2 is indeed a key signaling receptor necessary forTGF-β1 activation. These results suggest that IL-13 can sig-nal through the IL-13Rα2 chain in pancreatic cancer cells.We next examined the signaling pathway through IL-

13/IL-13Rα2 in pancreatic cancer cells. Mock, IL-13Rα2–expressing, and IL-13Rα2 knockdown MIA-PaCa-2 cellswere incubated with IL-13 or TGF-β1 for 4 hours and thenDNA binding assay was done to determine activation of theAP-1 family of nuclear factors. The stimulation of MIA-Pa-Ca-2mock cells with IL-13mainly induced the activation ofc-Jun and Fra-2 (Fig. 2C), but not the other AP-1 members



such as c-Fos, Fos B, Fra-1, JunB, or JunD. Interestingly, IL-13 did not induce activation of any AP-1 protein in IL-13Rα2 knockdown MIA-PaCa-2 cells. In contrast, TGF-β1did not induce activation of any AP-1 family members inboth mock control and IL-13Rα2 knockdown MIA-PaCa-2cells, which suggest that the AP-1 transcription factorpathway is involved in IL-13–induced signaling throughIL-13Rα2 leading to activation of TGFB1 promoter activ-ity in pancreatic cancer cells.IL-13Rα2–positive pancreatic cancer cell lines are sensitive

to IL-13 cytotoxin. IL13-PE inhibited protein synthesis ofpancreatic cancer cell lines in a concentration-dependentmanner. TheHS766T cell line was themost sensitive to thecytotoxin (IC50, 0.08 ng/mL), followed by Capan-1 (IC50,9 ng/mL) and MIA-PaCa-2 (IC50, 26 ng/mL; Fig. 3A). TheIC50 in the SW1990 cell line was >100 ng/mL (data notshown). Consistent with the lack of IL-13Rα2 mRNA ex-pression, the HPAF-II, Panc-1, KLM, and HPAC cell lineswere not sensitive to IL13-PE (IC50, ≥1,000 ng/mL; data notshown). Similarly, IL13-PE was not cytotoxic to the noncan-cerous cell lines HUVEC and HPDE (IC50, ≥1,000 ng/mL).The cytotoxic activity of IL13-PE to the HS766T cell line wasneutralized by incubation with an excess of IL-13, suggestingspecific cytotoxicity through binding of IL13-PE cytotoxin toIL-13Rα2 (Fig. 3B).We also examined the cytotoxicity of IL13-PE to IL-

13Rα2 knockdown MIA-PaCa-2 cells and HPAF-II cells en-gineered to express IL-13Rα2. As expected, IL-13Rα2knockdown MIA-PaCa-2 cells lost sensitivity to IL13-PE(IC50, ≥1,000 ng/mL; Fig. 3C), but IL-13Rα2–overexpres-sing HPAF-II cells became sensitive to IL13-PE (IC50,10 ng/mL; Fig. 3D). These results suggest that cytotoxic ac-tivity of IL-13 cytotoxin is mediated through IL-13Rα2.Effect of IL13-PE on pancreatic tumor growth assessed by

real-time imaging in an early orthotopic PDA model. Pancre-atic cancer cells were transfected with retroviral vector ex-pressing GFP, which after transfection revealed consistentbright GFP fluorescence in MIA-PaCa-2, HS766T, and IL-13Rα2 knockdown MIA-PaCa-2 cell lines. After transfec-tion, no significant difference in morphology, growth rate,and sensitivity to IL13-PE was observed (data not shown).Small primary MIA-PaCa-2 tumors were detectable four

days postorthotopic implantation by real-time whole-bodyimaging (average fluorescent area, 29.11 ± 3.34 mm2;Fig. 4A). These mice were i.p. injected with three differentdoses of IL13 cytotoxin for seven days. IL13-PE treatmentshowed regression of established tumors in a dose-dependent manner. At 50 μg/kg IL-13 cytotoxin dose, tu-mor growth was arrested for 10 days from the beginningof treatment (day 5) and then tumors began to grow slowly.At day 40, tumor size remained smaller compared with con-trol, and the difference in tumor size between no-treatmentcontrol and50μg/kg IL13-PE–treated animalswas statistical-ly significant (P < 0.0001). MIA-PaCa-2 cells with IL-13Rα2-knockdown grew linearly when implanted on the pancreas.However, in contrast to IL-13Rα2–positive MIA-Pa-Ca-2cells, treatment of IL-13Rα2-knockdown tumor-bearingmicewith IL13-PE did not inhibit tumor growth (Fig. 4B).






On the other hand, the treatment of HS766T orthotopictumor-bearing mice with IL13-PE (25 μg/kg and 50 μg/kg)showed a significant reduction of tumor growth. At day40, visible fluorescent area in treated groups was signifi-cantly lower compared with the nontreatment group(P = 0.004 and 0.003, respectively; Fig. 4C). As HS766Ttumor cells expressed higher levels of IL-13Rα2, thesetumors were more sensitive to the antitumor effect ofIL-13 cytotoxin compared with MIA-PaCa-2 tumors.Cip infusion enhances the antitumor effect of IL-13 cyto-

toxin compared with Bip injection. We compared the anti-tumor efficacy of IL13-PE delivered by Bip injection or Cipinfusion using micro-osmotic pumps in the early HS766Ttumor model. The fate of these animals was monitored forsurvival until day 94. Due to large tumor burden in thecorresponding control arm and ethical reasons stipulatedby the Institutional Animal Care Committee such asweight loss/cachexia and appearance, etc., all animals weresacrificed. Compared with Bip administration of 50 μg/kgIL-13 cytotoxin daily for 7 consecutive days, Cip of 50 μg/kg



(infused over 7 days) significantly suppressed tumorgrowth (P = 0.022) from the beginning of the treatmentuntil the end of the experiment (Fig. 5A and B). Similarly,median survival time of animals was significantly in-creased from 35 days to 42, 51, 68, and 78 days in theBip 10 μg/kg (P < 0.0012), Bip 25 μg/kg (P < 0.0001),Bip 50 μg/kg (P < 0.0001), and Cip 50 μg/kg groups(P < 0.0001), respectively (Fig. 5C). Compared with theBip 50 μg/kg group, a significant prolonged survival timewas observed in the Cip 50 μg/kg group (P = 0.046). Pro-longed survival benefit in the Cip 50 μg/kg group was223% compared with the nontreatment group. Increasein significant survival advantage correlated with tumorarea as detected by GFP fluorescence.H&E staining of tissue sections from treated animals

showed no toxicity attributable to IL13-PE in any of thegroups of mice. Evaluation of results in the liver, kidney,and spleen (Supplementary Fig. S3) of two mice from thecytotoxin or placebo groups suggested that this cytotoxindid not induce any organ toxicity in animals at this dose.

Fig. 2. IL-13Rα2 increases TGFβ1 promoter activity through AP-1 activation in pancreatic cancer cell lines. A and B, TGFB1 promoter activity wasdetermined by TGF-β1 luciferase assay in MIA-PaCa-2 mock cells and IL-13Rα2 knockdown MIA-PaCa-2 cells (A), and in HPAF-II mock cells andIL-13Rα2–overexpressed HPAF-II cells (B). Cells were treated with IL-13 (10 ng/mL), IL-4 (10 ng/mL), or TGF-β1 (2 ng/mL) for 24 h. Promoter activity isexpressed as ratio of activity under stimulated versus unstimulated conditions. C, AP-1 family of transcription factors was analyzed in MIA-PaCa-2IL-13Rα2–positive mock and IL-13Rα2 knockdown cells stimulated with IL-13 (10 ng/mL) or TGF-β1 (2 ng/mL). Only c-Jun and Fra-2 activities wereupregulated with IL-13 in MIA-PaCa-2 cells, but not in IL-13Rα2 knockdown cells.




Shimamura et al.

582


PDA clinical sample expresses IL-13Rα2 and is sensitive toIL13-PE. Surgically resected tumor was pathologically di-agnosed as moderately differentiated PDA (SupplementaryFig. S2A). This tumor section showed strong staining forIL-13Rα2 in the ductal adenocarcinoma cells and faintstaining of fibroblasts (Fig. 6A). We established a primarycell line from this tumor specimen to examine the antitumoractivity of IL-13 cytotoxin. This cell line expressing IL-13Rα2was highly sensitive to IL13-PE (IC50, 9 ng/mL), whereasfibroblast cells cultured from the same tumor sample werenot sensitive to IL13-PE (IC50, >1,000 ng/mL; Fig. 6B).IL-13 cytotoxin prolonged survival of mice in orthotopic

model of clinical PDA. The clinical sample was orthotopi-cally transplanted on the pancreas of SCID mice. All miceshowed growth of primary tumors andmetastasis to lymphnodes in peritoneum, hepatoduodenum ligament, and pa-ra-aortic areas. Seventy-five percent of these mice showedthe metastasis lesions to liver when they were sacrificed30 days after tumor implantation (SupplementaryFig. S2B). To assess the antitumor effect of IL13-PE in anearly tumor model and an advanced metastasis model,these mice were divided into six groups in each model on



day 5. In an early tumormodel, IL-13 cytotoxin treatment be-gan on day 5. Similar to the study in Fig. 5, the fate of theanimals was monitored for survival until day 105. Due tothe large tumor burden in the corresponding control armand the stipulated ethical reasons, all animals were sacrificed.As shown in Fig. 6C, themedian survival time of animals was45days in thenontreatment group,whereas itwas significant-ly increased to 64, 82, and 103 days in the Bip 10 μg/kg (P =0.0265), Bip 25 μg/kg (P = 0.0002), and Bip 50 μg/kg (P <0.0001) groups, respectively. The median survival time ofthe Cip 50 μg/kg group was >105 days (P < 0.0001).Mice with advanced metastasis were also treated with IL-

13 cytotoxin from day 31 through day 37 after tumorimplantation. As shown in Fig. 6D, the median survivaltime of animals was 58 and 60 days in the nontreatmentgroup and Bip 10 μg/kg group, respectively, whereas itwas significantly increased to 73, 84, and 92 days in theBip 25 μg/kg (P = 0.0274), Bip 50 μg/kg (P = 0.0001),and Cip 50 μg/kg (P = 0.0001) groups, respectively. Pro-longed survival time in the Cip 50 μg/kg group was 159%better than in the nontreatment group. Because of ethicalreasons of animal health in control arms, these experiments

Fig. 3. Cytotoxic activity of IL13-PE in pancreatic cancer cell lines and normal HPDE and HUVEC cells. Cells (1 × 104) were incubated with variousconcentrations of IL-13 cytotoxin (0-1,000 ng/mL). A, cytotoxic activity was assessed by protein synthesis inhibition assay in pancreatic cancer cell linesHS766T, MIA-PaCa-2, and Capan-1, and in normal HPDE and HUVEC cells. B, cells were pretreated with excess of IL-13 (1 μg/mL) prior to IL-13 cytotoxintreatment. IL-13 pretreatment inhibited the cytotoxic activity of IL13-PE cytotoxin against HS766T cells. C and D, cytotoxic activity of IL13-PEwas compared between MIA-PaCa-2 mock transfected and IL-13Rα2 knockdown cells and between HPAF-II mock transfected cells andIL-13Rα2–overexpressed cells.








were terminated. Thus, it is noteworthy that a significantimproved survival of animals was observed even in theadvanced metastatic PDA model.

Discussion

We show that 71% of PDA tumors express moderate tohigh-density IL-13Rα2 and that IL-13 can signal throughIL-13Rα2 in pancreatic cancer cell lines. IL-13 upregulatedTGFB1 promoter activity in MIA-PaCa-2 cells naturallyexpressing IL-13Rα2. This induction of TGFB1 promoteractivity is mediated by IL-13Rα2 as knockdown of theIL-13Rα2 gene by RNA interference eliminated IL-13–induced TGFB1 promoter activation. Another pancreaticcancer cell line, which is engineered to express IL-13Rα2by gene transfer, supported this conclusion. Knockdownor overexpression of IL-13Rα2 did not impact on IL-4Rαand IL-13Rα1 in both the MIA-PaCa-2 and HPAF-II celllines. These results suggest that only IL-13Rα2 is involvedin IL-13–induced signal transduction, not IL-4Rα and IL-13Rα1. The signal transduction through IL-13Rα2 is anovel observation as no previous study has shown thatIL-13 can signal through IL-13Rα2 in cancer cells.Our studies also show that IL-13 signaling through IL-

13Rα2 involved the AP-1 pathway, which was activated byIL-13 in a murine macrophage cell line (4). IL-13 causedactivation of the AP-1 family of transcription factors, inparticular Fra-2 and c-Jun in IL-13Rα2–positive pancreaticcancer cells. The pancreatic cancer cell line naturallyexpressing IL-13Rα2 or transfected to express IL-13Rα2showed AP-1 activation in response to IL-13, whereas cellknockdown of IL-13Rα2 did not show such activation.These results corroborated previous observations and con-firmed that IL-13 can mediate signal transduction andactivate the AP-1 pathway through IL-13Rα2 in pancreaticcancer cells. The significance of IL-13–induced TGF-β1activation in pancreatic cancer cells is not known. TGF-β1plays paradoxical roles in cancer development. It has a crit-ical function in the prevention of early-stage tumorigenesis(25). However, as genetic and epigenetic changes occur inthe context of normal epithelium cells, TGF-β1 evokes tu-morigenecity and finally promotes tumor metastasis (26).Friess et al. have shown that induction of TGF-β1 produc-tion in pancreatic cancer can contribute to the disease pro-gression (27). Thus, our results suggest that IL-13, throughIL-13Rα2, contributes to TGF-β1 production in pancreaticcancer and induces fibrosis, which may lead to tumor struc-ture formation and metastasis. Therefore, strategies toeliminate TGF-β1 production by tumors may offer noveltargets and opportunities for designing novel anticancertherapeutic approaches. In this regard, there are severalongoing clinical trials in which investigators are tryingto block the production of TGF-β1 (28).The expression of IL-13Rα2 sensitized PDA to the cytotox-

ic activity of IL-13 cytotoxin. IL13-PE was highly cytotoxic topancreatic cancer cell lines in vitro, but it was not cytotoxic toHPDE and HUVEC cells, which express no or low IL-13Rα2.Kornmann et al. also showed that IL13-PE is cytotoxic to

Fig. 4. Quantification of tumor volume by real-time imaging and effect ofIL13-PE on pancreatic tumor growth in an early tumor model. Selectivetumor GFP fluorescence facilitated real-time visualization of tumorburden in the live animals (representative images are shown in Fig. 5).Treatment was initiated on day 5. Images are captured twice a weekbeginning on day 5. Mice group 1 served as the negative control and didnot receive any treatment. Groups 2, 3, and 4 received Bip injection ofIL-13 cytotoxin at 10, 25, and 50 μg/kg, respectively, twice a day for 7 d.A, quantification of MIA-PaCa-2 tumor GFP fluorescence enabledreal-time determination and comparison of tumor load during the courseof each treatment and, therefore, permitted real-time comparison ofefficacy between groups. B, the effect of IL13-PE (50 μg/kg) on IL-13Rα2knockdown MIA-PaCa-2 tumor growth. C, the effect of IL13-PE on GFP+

HS766T tumor growth. Points, mean area of GFP fluorescence for liveanimals in each treatment group (n = 10 mice); bars, SD.




Shimamura et al.

584


pancreatic cancer cell lines in vitro (21). However, we showhere that IL13-PE is highly effective in mediating antitumoractivity in animal models of human pancreatic cancer. A sig-nificant prolonged survival was observed by IL13-PE admin-istration inmicewith early or advancedmetastatic disease. Inaddition, significant survival benefit was also observed inmice orthotopically implanted with the clinical pancreaticcancer sample. These results indicate that IL-13Rα2 may bea useful target for pancreatic cancer therapy.We adopted a whole-body imaging system to assess

the antitumor efficacy of IL13-PE. We were able to measurethe tumor size in live animals, which were implanted withhuman pancreatic tumor chunks derived from pancreatictumor cell lines or tumor tissue obtained after surgical re-section. Interestingly, the visible fluorescence, a measureof tumor size, correlated with the survival of animals trea-ted with IL13-PE. Cip administration of IL13-PE produced



the best tumor responses in both tumor models. Cipadministration was very well tolerated without any evi-dence of toxicity. These results show that the pancreaticcancer model established in our laboratory can be usefulin testing other therapeutic approaches for pancreatic can-cer therapy and that IL13-PE has remarkable antitumoractivity against pancreatic ductal adenocarcinoma.In the current study, we used wild-type human IL-13 to

make a fusion protein with PE38 and tested it in vitro and inanimal models of human PDA. However, in an ongoingproject in our laboratory, we have examined the effect ofmutation of IL-13 and fusing the mutated form to PE. Theresulting fusion protein IL13E13K-PE38 has been com-pared with wild-type (wt)IL-13-PE38 (29). In varioustumor cell lines including glioblastoma multiforme andIL-13Rα2–transfected tumor cell lines, IL13E13K-PE38bound to cells with 3 to 10 times higher affinity compared



.

Fig. 5. Cip infusion of IL-13cytotoxin enhanced the antitumoreffect compared with Bip injectionA, sequential in vivo imaging oftumor progression over time inearly pancreatic cancer model.Panels, a representative mousefrom each group. Treatment wasinitiated on day 5. Images werecaptured every 10 d beginning onday 4. Groups 1, 2, 3, and 4 are thesame as in Fig. 4. Group 5 receivedCip infusion of IL-13 cytotoxin(50 μg/kg) by mini-osmotic pumpsfor 7 d. B, quantification of HS766Ttumor with GFP in each group.C, Kaplan-Meier survival curves(n = 10 mice) after treatment withBip injection of IL-13 cytotoxin(10, 25, 50 μg/kg, twice a dayduring days 5-11) and Cip of IL-13cytotoxin (50 μg/kg, for 7 dbeginning day 5 delivered byminiosmotic pumps).




with wtIL13-PE38. However, IL13E13K-PE38 did not showhigher cytotoxicity compared with wtIL13-PE38 in brain tu-mors or any other cell lines tested. The antitumor activity ofIL13E13K-PE38 in nude mice with brain tumors was alsosimilar to wtIL13-PE38. Some improvement in antitumoractivity was observed when lower doses of IL13E13K-PE38were injected intratumorally in s.c. tumors. These results in-dicate that IL13E13K-PE38mediates similar cytotoxicity andantitumor activity towtIL13-PE38despite its improvedbind-ing affinity to IL-13 receptors.To further develop IL-13 cytotoxin for cancer therapy,

various preclinical toxicology studies have been done.Based on these preclinical studies, several phase I/II clini-cal trials have been completed to determine the safety andtolerability of IL13-PE in patients with malignant braintumors. In these clinical studies, IL13-PE cytotoxin wasadministered by convection enhanced delivery via intratu-moral or peritumoral catheters. These studies showed that



intratumoral and peritumoral infusion of IL13-PE waswell tolerated (30). On the basis of these results, a phaseIII PRECISE clinical trial was initiated in which two tothree catheters were placed peritumorally and IL13-PE ata concentration of 0.5 μg/mL was infused. This multicen-ter trial was completed, and the safety and efficacy of IL13-PE is being evaluated (31). This study is currently beingdesigned to extend to additional patient population toexamine the efficacy and tolerability of IL13-PE in glio-blastoma multiforme patients. Based on this clinical expe-rience, delivery of IL13-PE cytotoxin by intratumoral routemay be suited for PDA for better clinical outcome. An en-doscopic approach may be able to accomplish intratumor-al delivery of IL13-PE. IL13-PE may also be delivered i.p.similar to the preclinical animal model of PDA in additionto intratumor administration for metastatic disease. Addi-tional preclinical studies are needed to explore the poten-tial of IL13-PE in pancreatic cancer therapy.

Fig. 6. Effect of IL13-PE on early and advanced orthotopic pancreatic cancer models developed from a clinical PDA sample. A, expression of IL-13Rα2 inthe PDA clinical sample. The tissue specimen was stained with antihuman IL-13Rα2 polyclonal antibody. Magnification, ×200. B, cytotoxic activity ofIL-13 cytotoxin (0.1-1,000 ng/mL) to primary culture of tumor cells and fibroblasts as determined by protein synthesis inhibition assay. C, Kaplan-Meiersurvival curves (n = 10 mice) in the early pancreatic cancer model. Mice were treated with Bip by IL-13 cytotoxin (10, 25, 50 μg/kg, twice a day duringdays 5-11) or Cip by IL-13 cytotoxin (50 μg/kg, for 7 d beginning on day 5 with miniosmotic pumps). D, Kaplan-Meier survival curves (n = 10 mice) in theadvanced pancreatic cancer model. Mice in each group were treated with IL-13 cytotoxin during days 31 to 37.




Shimamura et al.

586


Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

We thank Drs. Michail Alterman and HideyukiNakashima for reviewing the manuscript and helpfulsuggestions, Dr. Brenton McCright and Ms. PamelaDover for help in microscopic examination of tumor



specimens and general help, and members of TumorVaccines and Biotechnology Branch, Division of Cellularand Gene Therapies, Center for Biologics Evaluation andResearch for their suggestions.The costs of publication of this article were defrayed in

part by the payment of page charges. This article musttherefore be hereby marked advertisement in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.Received 7/29/09; revised 10/22/09; accepted 10/27/09;

published OnlineFirst 1/12/10.

References
1. Murata T, Obiri NI, Debinski W, Puri RK. Structure of IL-13 receptor:
analysis of subunit composition in cancer and immune cells. Bio-chem Biophys Res Commun 1997;238:90–4.

2. Kawakami K, Takeshita F, Puri RK. Identification of distinct roles for adileucine and a tyrosine internalization motif in the interleukin (IL)-13binding component IL-13 receptor α 2 chain. J Biol Chem 2001;276:25114–20.

3. Kawakami K, Taguchi J, Murata T, Puri RK. The interleukin-13 recep-tor α 2 chain: an essential component for binding and internalizationbut not for interleukin-13-induced signal transduction through theSTAT6 pathway. Blood 2001;97:2673–9.

4. Fichtner-Feigl S, Strober W, Kawakami K, Puri RK, Kitani A. IL-13signaling through the IL-13α2 receptor is involved in induction ofTGF-beta1 production and fibrosis. Nat Med 2006;12:99–106.

5. Frankel AE, Kreitman RJ, Sausville EA. Targeted toxins. Clin CancerRes 2000;6:326–34.

6. Puri RK, Leland P, Obiri NI, et al. Targeting of interleukin-13 receptoron human renal cell carcinoma cells by a recombinant chimeric pro-tein composed of interleukin-13 and a truncated form of Pseudomo-nas exotoxin A (PE38QQR). Blood 1996;87:4333–9.

7. Debinski W, Obiri NI, Pastan I, Puri RK. A novel chimeric proteincomposed of interleukin 13 and Pseudomonas exotoxin is highly cy-totoxic to human carcinoma cells expressing receptors for interleukin13 and interleukin 4. J Biol Chem 1995;270:16775–80.

8. Joshi BH, Puri RK. Optimization of expression and purification of twobiologically active chimeric fusion proteins that consist of human in-terleukin-13 and Pseudomonas exotoxin in Escherichia coli. ProteinExpr Purif 2005;39:189–98.

9. Joshi BH, Kawakami K, Leland P, Puri RK. Heterogeneity in interleu-kin-13 receptor expression and subunit structure in squamous cellcarcinoma of head and neck: differential sensitivity to chimeric fusionproteins comprised of interleukin-13 and a mutated form of Pseudo-monas exotoxin. Clin Cancer Res 2002;8:1948–56.

10. Kioi M, Seetharam S, Puri RK. Targeting IL-13Rα2-positive cancerwith a novel recombinant immunotoxin composed of a single-chainantibody and mutated Pseudomonas exotoxin. Mol Cancer Ther2008;7:1579–87.

11. Reiter Y. Recombinant immunotoxins in targeted cancer cell therapy.Adv Cancer Res 2001;81:93–124.

12. Husain SR, Joshi BH, Puri RK. Interleukin-13 receptor as a uniquetarget for anti-glioblastoma therapy. Int J Cancer 2001;92:168–75.

13. Kawakami K, Kawakami M, Joshi BH, Puri RK. Interleukin-13 recep-tor-targeted cancer therapy in an immunodeficient animal model ofhuman head and neck cancer. Cancer Res 2001;61:6194–200.

14. Kawakami M, Kawakami K, Puri RK. Intratumor administration of in-terleukin 13 receptor-targeted cytotoxin induces apoptotic cell deathin human malignant glioma tumor xenografts. Mol Cancer Ther 2002;1:999–1007.

15. Kioi M, Kawakami M, Shimamura T, Husain SR, Puri RK. Interleukin-13 receptor α2 chain: a potential biomarker and molecular target forovarian cancer therapy. Cancer 2006;107:1407–18.

16. Joshi BH, Plautz GE, Puri RK. Interleukin-13 receptor α chain: a noveltumor-associated transmembrane protein in primary explants of hu-man malignant gliomas. Cancer Res 2000;60:1168–72.

17. Kawakami M, Kawakami K, Kasperbauer JL, et al. Interleukin-13 re-ceptor α2 chain in human head and neck cancer serves as a uniquediagnostic marker. Clin Cancer Res 2003;9:6381–8.

18. Husain SR, Obiri NI, Gill P, et al. Receptor for interleukin 13 on AIDS-associated Kaposi's sarcoma cells serves as a new target for a po-tent Pseudomonas exotoxin-based chimeric toxin protein. ClinCancer Res 1997;3:151–6.

19. Joshi BH, Leland P, Puri RK. Identification and characterization ofinterleukin-13 receptor in human medulloblastoma and targetingthese receptors with interleukin-13-pseudomonas exotoxin fusionprotein. Croat Med J 2003;44:455–62.

20. Vallera DA, Stish BJ, Shu Y, et al. Genetically designing a more po-tent antipancreatic cancer agent by simultaneously co-targeting hu-man IL13 and EGF receptors in a mouse xenograft model. Gut 2008;57:634–41.

21. Kornmann M, Kleeff J, Debinski W, Korc M. Pancreatic cancer cellsexpress interleukin-13 and -4 receptors, and their growth is inhibitedby Pseudomonas exotoxin coupled to interleukin-13 and -4. Antican-cer Res 1999;19:125–31.

22. Shimamura T, Royal RE, Kioi M, Nakajima A, Husain SR, Puri RK.Interleukin-4 cytotoxin therapy synergizes with gemcitabine in amouse model of pancreatic ductal adenocarcinoma. Cancer Res2007;67:9903–12.

23. Kawakami K, Terabe M, Kioi M, Berzofsky JA, Puri RK. Intratumoraltherapy with IL13–38 results in effective CTL-mediated suppressionof IL-13Rα2-expressing contralateral tumors. Clin Cancer Res 2006;12:4678–86.

24. Bouvet M, Wang J, Nardin SR, et al. Real-time optical imaging ofprimary tumor growth and multiple metastatic events in a pancreaticcancer orthotopic model. Cancer Res 2002;62:1534–40.

25. Engle SJ, Hoying JB, Boivin GP, Ormsby I, Gartside PS, DoetschmanT. Transforming growth factor beta1 suppresses nonmetastatic co-lon cancer at an early stage of tumorigenesis. Cancer Res 1999;59:3379–86.

26. Wakefield LM, Roberts AB. TGF-β signaling: positive and negativeeffects on tumorigenesis. Curr Opin Genet Dev 2002;12:22–9.

27. Friess H, Yamanaka Y, Buchler M, et al. Enhanced expression oftransforming growth factor β isoforms in pancreatic cancer cor-relates with decreased survival. Gastroenterology 1993;105:1846–56.

28. Fakhrai H, Mantil JC, Liu L, et al. Phase I clinical trial of a TGF-β an-tisense-modified tumor cell vaccine in patients with advanced glio-ma. Cancer Gene Ther 2006;13:1052–60.

29. Kioi M, Kawakami K, Puri RK. Analysis of antitumor activity of an in-terleukin-13 (IL-13) receptor-targeted cytotoxin composed of IL-13antagonist and Pseudomonas exotoxin. Clin Cancer Res 2004;10:6231–8.

30. Kunwar S, Prados MD, Chang SM, et al. Direct intracerebral deliveryof cintredekin besudotox (IL13–38QQR) in recurrent malignant glio-ma: a report by the Cintredekin Besudotox Intraparenchymal StudyGroup. J Clin Oncol 2007;25:837–44.

31. Kunwar S, Chang SM, Westphal M, et al. Phase III randomized trial ofCED of IL13-PE38QQR vs. gliadel wafers for recurrent glioblastoma.Neuro-Oncol. In press 2010.




2010;16:577-586. Published OnlineFirst January 12, 2010.Clin Cancer Res Takeshi Shimamura, Toshio Fujisawa, Syed R. Husain, et al. Pancreatic Cancer Therapy

Exotoxin inPseudomonasAdenocarcinoma: Role of IL-13 2 in Pancreatic DuctalαInterleukin 13 Receptor

Interleukin 13 Mediates Signal Transduction through

Updated version

10.1158/1078-0432.CCR-09-2015doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2010/01/18/1078-0432.CCR-09-2015.DC1

Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/16/2/577.full#ref-list-1

This article cites 30 articles, 19 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/16/2/577.full#related-urls

This article has been cited by 9 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://clincancerres.aacrjournals.org/content/16/2/577To request permission to re-use all or part of this article, use this link

Research. on December 12, 2020. © 2010 American Association for Cancerclincancerres.aacrjournals.org Downloaded from


http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-09-2015

http://clincancerres.aacrjournals.org/content/suppl/2010/01/18/1078-0432.CCR-09-2015.DC1

http://clincancerres.aacrjournals.org/content/16/2/577.full#ref-list-1

http://clincancerres.aacrjournals.org/content/16/2/577.full#related-urls

http://clincancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/16/2/577


Interleukin 13 Receptor α2 in Pancreatic Ductal ... · in human pancreatic cancer cell lines....

Documents

Transcript of Interleukin 13 Receptor α2 in Pancreatic Ductal ... · in human pancreatic cancer cell lines....