Analysis of interleukin-13 receptor α2 expression in human pediatric brain tumors§


  • Mariko Kawakami M.D., Ph.D.,

    1. Laboratory of Molecular Tumor Biology, Division of Cellular and Gene Therapies, Center for Biologics Evaluation and Research, Food and Drug Administration, Bethesda, Maryland
    Search for more papers by this author
  • Koji Kawakami M.D., Ph.D.,

    1. Laboratory of Molecular Tumor Biology, Division of Cellular and Gene Therapies, Center for Biologics Evaluation and Research, Food and Drug Administration, Bethesda, Maryland
    Search for more papers by this author
  • Satoru Takahashi M.D., Ph.D.,

    1. Department of Experimental Pathology and Tumor Biology, Nagoya City University Graduate School of Medical Sciences, Aichi, Japan
    Search for more papers by this author
  • Masato Abe M.D.,

    1. Department of Surgical Pathology, Fujita Health University School of Medicine, Aichi, Japan
    Search for more papers by this author
  • Raj K. Puri M.D., Ph.D.

    Corresponding author
    1. Laboratory of Molecular Tumor Biology, Division of Cellular and Gene Therapies, Center for Biologics Evaluation and Research, Food and Drug Administration, Bethesda, Maryland
    • Laboratory of Molecular Tumor Biology, Division of Cellular and Gene Therapies, Center for Biologics Evaluation and Research, Food and Drug Administration, NIH Building 29B, Room 2NN10, HFM-735, Bethesda, MD 20892
    Search for more papers by this author
    • Fax: (301) 827-0449

  • This article is a U.S. Government work and, as such, is in the public domain in the United States of America.

  • The current study was conducted as part of a collaboration between the Food and Drug Administration and NeoPharm Inc. under a Cooperative Research and Development Agreement.

  • §

    The opinions expressed herein do not necessarily reflect the views of the Food and Drug Administration or the U.S. Government.



Compared with normal brain tissue cells, human malignant glioma cells express higher levels of interleukin-13 receptor (IL-13R). However, whether this receptor is expressed in situ has not been carefully examined. With IL-13R–targeted cytotoxin (IL13-PE38QQR, comprising IL-13 and a mutated form of Pseudomonas exotoxin [PE]) being tested in three Phase I/II clinical trials for the treatment of adult human glioma, and with pediatric studies being planned, the authors set out to analyze pediatric brain tumor tissue specimens for the expression of IL-13R.


Using in situ hybridization and immunohistochemical staining, the authors examined 58 pediatric brain tumor specimens for expression of the predominant IL-13 binding and internalizing protein (IL-13Rα2) chain at the mRNA and protein levels.


Overall, approximately 83% of pediatric brain tumor samples expressed IL-13Rα2. One hundred percent (11 of 11) high-grade astrocytoma, 79% (26 of 33) low-grade astrocytoma, 67% (4 of 6) medulloblastoma, and 67% (2 of 3) ependymoma samples were positive for IL-13Rα2. Among IL-13Rα2–positive samples, 88% (42 of 48 samples) had positive expression in ≥ 50% of all tumor fields. The results obtained using both assays were consistent with each other.


The current study established that pediatric brain tumor specimens expressed the IL-13Rα2 chain. Because the IL-13Rα2 chain is a major binding component of the IL-13R complex, these results suggest that the targeting of IL-13R may represent a useful approach for the treatment of pediatric brain tumors. Cancer 2004. Published 2004 by the American Cancer Society.

Brain tumors in children comprise a variety of phenotypes, including medulloblastomas, pilocytic and fibrillary astrocytomas, ependymomas, and gliomas.1 In contrast to adult brain tumors, which typically are supratentorial, pediatric tumors are localized predominantly in the posterior fossa and brain stem.1 Because of these characteristics, optimal resection of the tumor is not possible, resulting in progressive disease. Therefore, additional therapeutic measures, including the use of chemotherapeutic and biologic agents, are urgently needed.

Recent studies have focused on the development of tumor-specific targeted agents for glioma therapy. For example, agents targeting interleukin-4 receptor (IL-4R), interleukin-13 receptor (IL-13R), epidermal growth factor receptor, transferrin receptor, and urokinase receptor are currently being developed.2–7 These tumor-specific agents show promising antitumor activity in vitro and in vivo against adult glioma in animal models. Studies conducted at our laboratories have identified overexpression of IL-4R and IL-13R in human brain tumor cells, whereas normal tissue samples obtained from the human brain expressed undetectable or low levels of IL-4R and IL-13R.8–13 The structure and function of these receptors in tumor cell lines have been extensively examined.14–19 The IL-13R complex in brain tumor specimens comprises the IL-4Rα, IL-13Rα1, and IL-13Rα2 chains.12, 18 Although the IL-13Rα2 chain binds IL-13 with high affinity and is subsequently internalized, it does not mediate signal transduction through the JAK/STAT pathway. In contrast, the IL-13Rα1 chain binds IL-13 weakly, but it forms a complex with the IL-4Rαchain, which mediates signal transduction through the JAK/STAT pathway.14

Although the functional significance of the expression of IL-4R and IL-13R on solid tumor cells in general and on glioma cells in particular is not clear, these receptors provide attractive targets for the delivery of specific cytotoxic agents. To target IL-13R, a chimeric fusion protein consisting of human IL-13 and a truncated form of Pseudomonas exotoxin (termed IL-13 cytotoxin) has been produced.10 This fusion protein is highly cytotoxic and selective for IL-13R–positive tumor cells in vitro and in animal models of human brain tumors.20, 21 Further studies conducted by our group have demonstrated that the IL-13Rα2 chain confers extreme sensitivity to IL-13 cytotoxin in vitro and in vivo.22–25 With the IL-13 cytotoxin being tested in three Phase I/II multicenter and multinational clinical trials and generating measurable tumor responses, and with pediatric studies being planned, we set out to investigate whether pediatric brain tumor specimens express the IL-13Rα2 chain.26–29 The information yielded by this investigation may be used to design a new clinical trial in which IL-13 cytotoxin is used to treat pediatric brain tumors.


Tissue Samples

Fifty-eight available pediatric brain tumor samples (pilocytic astrocytoma [n = 29], glioblastoma [n = 8], medulloblastoma [n = 6], low-grade astrocytoma [n = 7], giant cell tumor [n = 3], ependymoma [n = 3], mixed glioma with pilocytic astrocytoma and low-grade astrocytoma [n = 1], and ganglioglioma [n = 1]) were obtained from the National Cancer Institute Cooperative Human Tissue Network (CHTN), the Department of Experimental Pathology and Tumor Biology at the Nagoya City University Graduate School of Medical Sciences (Aichi, Japan), or the Department of Surgical Pathology at the Fujita Health University School of Medicine (Aichi, Japan). These samples were obtained after securing approval from the Food and Drug Administration's Research Involving Human Subjects Committee. Tissue samples were obtained during surgical resection and were fixed in formalin and embedded in paraffin. Each diagnosis was made by neuropathologists at the institution where the sample was collected. No tissue samples were retrieved from patients with recurrent disease.

In Situ Hybridization (ISH) Analysis

Tissue samples were analyzed for the expression of mRNA encoding the various IL-13R chains using ISH as described previously.30, 31 In brief, 5 μm sections mounted on positively charged slides (HistoServ, Gaithersburg, MD) were rehydrated by serial incubation with decreasing percentages of ethanol and RNase-free water and then enzymatically treated with 10 μg/mL proteinase K (Sigma-Aldrich, St Louis, MO) for 15 minutes at 37 °C. Sections were then acetylated with freshly prepared 0.25% acetic anhydride and 0.1 M triethanolamine buffer (pH 8.0) for 10 minutes. Next, tissue sections were hybridized overnight at 55 °C with 20 μL heat-denatured antisense or sense digoxigenin-labeled RNA probes at a concentration of 100 ng/mL in hybridization solution (10 mM Tris/HCl [pH 7.4], 600 mM NaCl, 1 mM ethylenediaminetetraacetic acid, 50% deionized formamide, 1X Denhardt solution, 10 mg/mL salmon sperm DNA, 10 mg/mL yeast tRNA, 0.25% sodium dodecyl sulfate, and 10X dextran sulfate) in a humidified chamber (all reagents obtained from Sigma-Aldrich) and subsequently washed in sodium chloride/sodium citrate (SCC) at concentrations ranging from 2X to 0.1X. Digoxigenin-labeled antisense or sense RNA probes were generated by using SP6 or T7 RNA polymerase in conjunction with an in vitro transcription system (Roche, Indianapolis, IN). Hybridization reactions were detected by immunofluorescent staining with fluorescein isothiocyanate or rhodamine-conjugated anti-digoxigenin antibodies (Roche). After washing, sections were dried and layered with ProLong Antifade mounting medium (Molecular Probes, Eugene, OR) and a cover slip. Slides were viewed on an Olympus IX70 fluorescence microscope using appropriate filters (Olympus Optical Co., Tokyo, Japan). Images were compiled from sets of three consecutive single optical sections using Spot Insight software (Version 3.2; Diagnostic Instruments, Sterling Heights, MI). Positive fluorescence density was scored as 1+, 2+, or 3+.

Immunohistochemical (IHC) Analysis

IHC was performed using the Vector ABC peroxidase kit according to the manufacturer's instructions (Vector Laboratories, Burlingame, CA). In brief, paraffin-embedded tissue sections were deparaffinized by xylene treatment and washed with an alcohol gradient (from 100% to 50%) and phosphate-buffered saline. Sections were incubated with monoclonal antibodies against human IL-4Rα (M57 [20 μg/mL]; kindly supplied by Immunex Corp., Seattle, WA), IL-13Rα1 (10 μg/mL; Diaclone, Besançon, France), IL-13Rα2 (10 μg/mL; Diaclone), or mouse immunoglobulin G1 (isotype control) for 18 hours at 4 °C. Slides were developed using 3,3′-diaminobenzidine substrate biotinylated peroxidase reagent (Vector Laboratories) and then lightly counterstained with hematoxylin. The intensity of positive staining was classified as 1+ (∼30%), 2+ (30–70%), or 3+ (> 70%), with the strongest observed intensity being defined as 100%. The percentage of positive fields in receptor-positive sections was assessed by viewing the entire field under the same magnification and counting stained tumor cells.

Statistical Analysis

The significance of differences in IL-13Rα2 chain positivity among various types of brain tumor samples was analyzed using the Student t test.


In Situ Hybridization

ISH analysis was performed for 36 pediatric brain tumor specimens (obtained from CHTN) to investigate the expression of IL-13Rα2 chain mRNA. As shown in Figure 1, randomly selected brain tumor specimens (Specimens 2584 and 4301) expressed IL-13Rα2 mRNA at a high level. In contrast, tissue samples hybridized with sense RNA exhibited only background staining. IL-13Rα2 mRNA expression levels varied among the 36 pediatric brain tumor samples. However, mRNA expression patterns were similar to the staining intensity patterns observed on IHC analysis (Table 1). Eighty-three percent of tumor samples were positive for IL-13Rα2 chain mRNA, exhibiting modest-to-high staining density. Pilocytic astrocytoma was the only tumor type that exhibited negative staining (23% [6 of 26 tissue samples]).

Figure 1.

In situ hybridization analysis of interleukin-13 receptor α2 chain mRNA. Randomly selected pediatric brain tumor specimens from two patients were analyzed by in situ hybridization. Antisense or sense RNA probes were used to hybridize tissue sections. Original magnification ×400.

Table 1. In Situ Expression of the IL-13Rα2 Chain in Pediatric Brain Tumor Sections by Immunohistochemical Methods
SpecimenPatient informationIL-13Rα2 staining
Age (yrs)GenderDiagnosisSiteIntensityb% positive fields
  • GBM: glioblastoma multiforme; M: male; F: female; IL-13Rα2: interleukin-13 receptor α2.

  • a

    Mixed glioma with oligodendroglioma and pilocytic astrocytoma.

  • b

    —: negative; +: ∼ 30% staining; ++: 30–70% staining; +++: > 70% staining (WHO classification).

28779FLow-grade astrocytomaRight optic nerve++80
434510MGBMCorpus callosum+/++90
540311MGiant cell tumorThird ventricle++90
666912MGiant cell tumorThird ventricle+/++90
3665MPilocytic astrocytomaHypothalamus0
34095MPilocytic astrocytomaCerebellum+50
57348FMixed gliomaaLeft temple+70
260611MPilocytic astrocytomaPosterior fossa+50
491220MPilocytic astrocytomaHypothalamus+40
718613FPilocytic astrocytomaCerebellum++70
71974FPilocytic astrocytomaCerebellum++50
49015FPilocytic astrocytomaPosterior fossa+50
7185MPilocytic astrocytomaPosterior fossa+80
120511FPilocytic astrocytomaCerebellum++50
376920FPilocytic astrocytomaPosterior fossa0
430113MPilocytic astrocytomaCerebellum++50
1002FPilocytic astrocytomaPosterior fossa++80
138414MPilocytic astrocytomaPosterior fossa+++90
149314FPilocytic astrocytomaSpinal cord+40
31634MPilocytic astrocytomaCerebellum++90
66675FPilocytic astrocytomaSpinal cord++90
25841FGangliogliomaBrain stem+++80
30288MPilocytic astrocytomaBrain stem+50
40009MPilocytic astrocytomaCerebellum++80
439613MPilocytic astrocytomaBrain stem++50
55284FPilocytic astrocytomaCerebellum++50
417213FPilocytic astrocytomaCerebellum+50
470111FPilocytic astrocytomaLeft tempoloparietal0
599618MGBMLeft frontoparietal++90
18411FPilocytic astrocytomaLeft tempoloparietal0
70096FGiant cell tumorLateral ventricle++90
255710FPilocytic astrocytomaLeft optic nerve0
53686FPilocytic astrocytomaLeft basal ganglia++50
626813FPilocytic astrocytomaLeft basal ganglia0
92005626FDesmoplastic medulloblastomaCerebellum0
1094215FGlioblastomaForth ventricle++80
30653120MGlioblastomaRight frontal+++90
850429814FGiant cell glioblastomaLeft temporal+++90
263212FPilocytic astrocytomaHypothalamus+40
153FPilocytic astrocytomaCerebellum+40
940210910FPilocytic astrocytomaCerebellum++60
234419FAstrocytoma Grade IIHypothalamus+60
296717MFibrillary astrocytomaLeft temple0
30553820MAstrocytoma Grade II–IIILeft temple++70
950603515MAstrocytoma Grade II–IIIRight thalamus++40
84026886FAstrocytoma Grade IIPons++90
84053602FAstrocytoma Grade IIIBilateral thalamus++50
15281MEpendymomaLeft frontal0
52676MEpendymomaLeft occipital+50
20592320FClear cell ependymoma Grade II–IIIRight occipital+80


IL-13Rα2 protein expression was confirmed by IHC analysis in 58 pediatric brain tumor samples. In addition, we analyzed the expression of the IL-13-binding proteins IL-13Rα1 and IL-4Rα. As shown in Figure 2, three randomly selected brain tumor specimens expressed IL-13Rα2 protein in a manner that was similar to what was observed on ISH analysis. Three tumor samples (glioblastoma [Specimen 4345], giant cell tumor [Specimen 6669], and pilocytic astrocytoma [Specimen 7197]) expressed IL-13Rα2 protein at moderate-to-high levels. Analysis of all 58 tissue samples indicated that 31 tumor specimens had strong staining for the IL-13Rα2 chain (30% to > 70%), 17 tumor specimens had moderate staining (∼30%), and 10 tumor specimens had no staining (Table 1). IL-13Rα2–positive tumor sections were further analyzed for the percentage of fields that were positively stained by viewing the entire tumor field and counting stained tumor cells. As shown in Table 1, strong and moderately stained tumor sections had positive staining for the IL-13Rα2 chain in 40–90% of all fields.

Figure 2.

Immunohistochemical analysis of interleukin-13 receptor (IL-13R) chains in pediatric brain tumor specimens. Tissue specimens were stained with hematoxylin and eosin (H & E) or with specific monoclonal antibodies against the IL-13Rα2 chain, the IL-13Rα1 chain, or the interleukin-4 receptor α (IL-4Rα) chain. Mouse immunoglobulin G1 (IgG1) served as a control. Original magnification × 400.

All 58 tissue samples were also stained to assess IL-4Rα and IL-13Rα1 chain expression levels. Three randomly selected samples are shown in Figure 2. Forty-four of the 58 (76%) samples exhibited moderate-to-strong staining for the IL-4Rα chain, and 54 of 58 (93%) samples exhibited weak-to-modest staining for the IL-13Rα1 chain.

Further analysis of staining intensity (Table 1) demonstrated that 4 of 6 (67%) medulloblastoma samples, 11 of 11 (100%) high-grade astrocytoma samples (including glioblastoma samples), 26 of 33 (79%) low-grade astrocytoma samples (including pilocytic astrocytoma samples), 2 of 3 (67%) ependymoma samples, and 5 of 5 (100%) other histologic brain tumor specimens were positive for the expression of the IL-13Rα2 chain (Table 2). There was no significant correlation between IL-13Rα2 chain expression in a given sample and the age or gender of the patient from whom the sample was obtained. Aside from the sensitivity of staining, heterogeneity in cell populations may have been responsible for the observation of differential expression. For example, a majority of the pilocytic astrocytoma samples that were analyzed contained interspersed normal cells.

Table 2. IL-13Rα2 Chain Expression in Various Histologic Samples of Pediatric Brain Tumors
DiagnosisNo. of patientsStaining intensity
  • IL-13Rα2: interleukin-13 receptor α2.

  • a

    Grade III and Grade IV (glioblastoma and glioblastoma multiforme).

  • b

    Grade I (pilocytic astrocytoma) and Grade II (fibrillary astrocytoma).

  • c

    Mixed glioma with oligodendroglioma and pilocytic astrocytoma.

High-grade astrocytomaa110110
Low-grade astrocytomab3371115
Giant cell tumor3003
Mixed gliomac1010

Although 83% of all pediatric brain tumor samples were positive for the IL-13Rα2 chain, there was significant heterogeneity with respect to protein expression within the tumor samples. Nevertheless, 40–90% of tumor cells within tissue sections stained positively for the IL-13Rα2 chain. Other tumor cells may also be positive. However, due to the sensitivity of the assay, their expression was not detected. This observation is consistent with the heterogeneous nature of brain tumors.


Our findings demonstrated that pediatric brain tumors express the IL-13Rα2 chain at extremely high levels. These results are in agreement with previous studies (ours and others') demonstrating IL-13Rα2 chain overexpression in adult glioblastoma multiforme primary culture cells and cell lines.12, 15, 19 In the current study, 29 of 58 pediatric brain tumor specimens were obtained from patients with pilocytic astrocytoma. Further studies involving larger numbers of samples, including primitive neuroectodermal tumor, choroid plexus tumor, and malignant germ cell tumor samples, are necessary to fully understand the expression profile of IL-13R in the pediatric brain tumor population. Although normal pediatric brain tissue specimens were not available in the current study and therefore were not examined, previous studies have shown that IL-13Rα2 is undetectable or expressed at low levels in normal brain-derived cells.10

Like adult gliomas, pediatric brain tumors express type I IL-13R, which comprises the IL-13Rα1, IL-13Rα2, and IL-4Rα chains.3, 18 Because of this configuration, it is possible that IL-13R molecules on pediatric brain tumor cells are also functional.

The significance of high IL-13Rα2 chain expression levels in brain tumor specimens remains unclear. However, it is possible that the IL-13Rα2 chain is associated with tumorigenicity.16 In addition, because expression of the IL-13Rα2 chain has been shown to sensitize human glioma cells to IL-13 cytotoxin in vitro and in vivo, it is hypothesized that this cytotoxin will also be active in pediatric patients with brain tumors. Because not all pediatric brain tumor cells were positive for the IL-13Rα2 chain, it is possible that IL-13 cytotoxin will not eliminate all malignant cells. However, it is noteworthy that in our preclinical studies, IL-13 cytotoxin has been highly effective in eliminating the majority of malignant cells in in vivo animal models despite the fact that the IL-13Rα2 gene is not expressed in all malignant cells after IL-13Rα2 plasmid injection. Tumor regression was attributed to infiltration by innate immune cells. Therefore, it is possible that in the clinical setting, IL-13 cytotoxin will be effective in eliminating most tumor cells, even if they do not express detectable levels of the IL-13Rα2 chain.24, 25 In addition, although pilocytic astrocytoma is a surgically curable disease, the targeting of IL-13R may represent an alternate treatment strategy for patients in whom complete surgical resection cannot be achieved. Because IL-13 cytotoxin has been shown to be well tolerated when administered directly to adult glioma or to normal brain tissue adjoining the tumor resection cavity,26–29 this agent may possess utility in the treatment of pediatric patients with brain tumors.


The authors thank Ms. Pamela Dover and Dr. S. Rafat Husain (Center for Biologics Evaluation and Research, Food and Drug Administration, Bethesda, MD) and Dr. Jeff Sherman (NeoPharm Inc., Lake Forest, IL) for their critical reading of the current article. The authors also thank the Cooperative Human Tissue Network, which is funded by the National Cancer Institute, for providing tumor tissue samples.