Overexpression of Interleukin-13 Receptor-α2 in Neuroendocrine Malignant Pheochromocytoma: A Novel Target for Receptor Directed Anti-Cancer Therapy

Abstract

Context: Pheochromocytomas and paragangliomas are rare catecholamine-secreting neuroendocrine tumors arising from the adrenal medulla and sympathetic tissues. When complete surgical resection is not an option, the treatment of pheochromocytoma is limited.

Objective: The objective of the study was to identify and characterize overexpression of IL-13 receptor-α2 (IL-13Rα2) gene expression in human and murine tumors and verify xenograft mouse pheochromocytoma cell (MPC)-derived tumor’s response to a selective cytotoxin.

Design/Setting/Patients: Expression of IL-13Rα2 was evaluated in a panel of 25 human pheochromocytoma clinical samples by RT-PCR and eight MPC tumors by indirect immunofluorescence assay and RT-PCR.

Intervention: The function of IL-13Rα2 in these tumor cells was examined by evaluating tumor sensitivity to a recombinant IL-13-Pseudomonas exotoxin (IL-13PE). Subcutaneous small and large MPC tumors in athymic nude mice (n = 10) were treated intratumorally with IL-13PE (100 μg/kg).

Main Outcome Measures: IC₅₀ and tumor size were measured.

Results: IL-13PE immunotoxin was highly cytotoxic to IL-13Rα2-overexpressing MPC cells (IC₅₀ <2.5 ng/ml) in vitro. Furthermore, IL-13PE was highly cytotoxic to sc tumors. Our results showed a statistically significant decrease in tumor size as early as 3 d after initial treatment and further suppressed growth of MPC tumors. All tumors displayed a histological evidence of necrosis in response to IL-13 immunotoxin without any adverse effects in host at this dose.

Conclusions: Human and murine neuroendocrine pheochromocytoma overexpress the IL-13Rα2 chain, and an IL-13PE-based receptor-directed anticancer approach may prove useful in treatment for metastatic pheochromocytoma patients.

Human and murine neuroendocrine pheochromocytoma tumors overexpress the IL-13Rα2 chain, and an IL-13PE-based receptor-directed anticancer approach may prove useful in treatment for metastatic pheochromocytoma patients.

Despite advances in diagnosis and imaging, pheochromocytoma remains one of the tumors for which no cure exists when metastatic disease is present. The prognosis of benign and malignant pheochromocytoma vastly differs. Benign tumors are generally amenable to complete surgical resection and have a 10-yr survival of up to 94% (1). Conversely, no curative treatment exists for metastatic pheochromocytoma, and the 10-yr survival rate is only 20% (1,2,3). Current regimens for treating metastatic pheochromocytoma are limited to combination chemotherapy, radioactive ¹³¹I-metaiodobenzylguanidine, local radiotherapy, and more recently radiofrequency ablation (3,4). Therefore, novel therapeutic approaches for treating metastatic malignant pheochromocytoma are urgently warranted.

Previously we demonstrated an overexpression of IL-13 receptor (IL-13R) in a number of cancers. IL-13Rα2 is overexpressed in primary brain tumors, head and neck cancer, renal cell carcinoma, ovarian carcinoma, and AIDS-related Kaposi sarcoma (5,6,7). Conversely, little or no IL-13Rα2 is expressed in normal immune cells or tissues (8). IL-13 binds with higher affinity to IL-13Rα2 compared with IL-13Rα1 chain. The IL-13R complex can be of three different types and its structure varies in different cell types (9). IL-13 predominantly mediates signaling through its heterodimer composed of the IL-4α and IL-13Rα1 chains and initiates signaling events through Janus kinase/signal transducer and activator of transcription-6 pathways (10,11,12). Recently IL-13 has been shown to mediate signaling through IL-13Rα2 chain in certain cell types (13). In this case, IL-13 signals through the AP-1 pathway.

To target tumor cells overexpressing IL-13R, we developed and produced an immunotoxin, consisting of IL-13 and truncated Pseudomonas exotoxin A (IL-13PE). IL-13PE produces IL-13R-specific cytotoxicity both in vitro and in vivo tumor models (8). The mechanism of IL-13PE in receptor-positive cells is extensively studied. After endocytosis of IL-13PE, the A chain translocates to the cytosol, cleaves ribosomal RNA, and thereby inhibits protein synthesis and cells are killed (9,10).

In this study, we demonstrate overexpression of IL-13Rα2 at the mRNA and protein levels in animal and human pheochromocytoma tumor specimens. We also demonstrate that IL-13R-targeted cytotoxin IL-13PE is highly cytotoxic to pheochromocytoma tumors in vitro and in vivo in a sc mouse pheochromocytoma cell (MPC) tumor model. Thus, IL-13Rα2 is a novel target on pheochromocytoma that can be targeted by IL-13PE for cancer therapy.

Materials and Methods

Recombinant immunotoxin and cell lines

Recombinant IL-13PE was produced and purified as previously described (14). Mouse pheochromocytoma cells (MPC 4/30/PRR), developed from NF1 knockout mice, were generously provided by Dr. Arthur S. Tischler (15). PM-RCC is a IL-13Rα2-positive human renal cell carcinoma and T98G is a IL-13Rα2-negative human glioma cell line (8,10). Cells were grown in appropriate medium as described previously (8,10,15).

Cell viability assay

The IC₅₀ of IL-13PE against the above cell lines was determined by trypan blue dye exclusion. Briefly, MPC, PM-RCC, and T98G were seeded at 5 × 10⁶ cells in 25-mm³ Corning flasks (Corning, NY) with increasing concentration of IL-13PE (0, 0.1, 1, 10, 100, and 1,000 ng/ml) at 37 C for 24 h. The cells were incubated for an additional 4 d before viable cells were counted by trypan blue dye exclusion technique and data expressed as percent positive cells.

RT-PCR

Twenty-five human samples from patients evaluated and operated for pheochromocytoma were collected at the National Institutes of Health (Bethesda, MD) and three normal human adrenal tissues (Rotterdam, The Netherlands) after institutional review board approval. Pheochromocytoma was confirmed by histology of surgically resected tumors. Adrenal medulla tissue was separated consisting of no more than 50% cortex. RNA was extracted using TRIzol (Invitrogen, Carlsbad, CA) followed by RNeasy minikit (QIAGEN, Valencia, CA). RT-PCR analysis was performed for IL-13Rα2 and β-actin (internal control). The optimal conditions for human IL-13Rα2 RT-PCR and primers have been published previously (16). The following primers were used for mouse IL-13Rα2: 5′-CGC-ATT-TGT-CAG-AGC-ATT-GT-3′ (sense) and 5′-CCA-AGC-CCT-CAT-ACC-AGA-AA-3′ (antisense). All studies were done in triplicate for validation.

Histology and immunohistochemistry

Immunofluorescence assay (IFA) for IL-13Rα2 and tyrosine hydroxylase was performed as previously described (5). A biotin-streptavidin based IFA was used for red and green fluorescence and the sections were viewed in a Nikon fluorescence microscope (×200; Tokyo, Japan).

Animal studies

Six- to 10-wk-old female athymic nude mice (NCr-ν) were obtained from Taconic, Inc. (Germantown, MD). Tumors were established by sc injection of MPC cells (5 × 10⁶). Mice were grouped by approximate tumor size (large, >50 mm³, and small, <50 mm³) before treatment. Tumor size was calculated by measuring three diameters using a caliper multiplying π/6 by height, length, and width and evaluated for statistical significance. Animals were treated with 100 μg/kg IL-13PE or PBS intratumorally (IT) in three axes for 3 consecutive days and were monitored for health and behavior for 1 wk.

Statistical analysis

The statistical significance of data were calculated by unpaired t test.

Results

Overexpression of IL-13Rα2 in human and murine pheochromocytoma

RNA microarray results were obtained from Brouwers et al. (17). Ninety-eight human pheochromocytoma (from 90 patients) compared with normal human adrenal medulla polyA reference and 20 MPC-derived tumors (10 sc and 10 liver metastasis) compared with parental MPC (our unpublished data) were examined for IL-13Rα2 mRNA expression. These results showed increased levels of IL-13Rα2 expression compared with normal human adrenal medulla. The expression ratio in benign tumors was median of 2.074 (interquartile range 1.599, 3.504) compared with malignant lesions median of 2.635 (interquartile range 1.992, 3.236) (P = 0.81). Mouse MPC-derived sc tumors revealed a ratio of 2.10 (sc tumor compared with MPC) and mouse MPC-derived liver metastasis revealed a ratio of 2.06 compared with MPC. The ratio of liver lesions compared with sc lesions was not significant (P = 0.92). Ratios were calculated based on the mean background subtracted intensities.

IL-13Rα2 expression was confirmed in mouse sc tumors (n = 4) and mouse liver tumors (n = 4) (Fig. 1A) as well as 25 clinical samples of pheochromocytoma by RT-PCR (Fig. 1B). Positive expression was observed in all tumors. Conversely, normal adrenal medulla tissue did not reveal expression of IL-13Rα2 (supplemental Table S1, published as supplemental data on The Endocrine Society’s Journals Online Web site at http://jcem.endojournals.org). Normal human adrenal tissue immunostained for IL-13Rα2 also confirmed absence or low expression of protein (our unpublished data).

Figure 1.

A, RT-PCR reveals increased relative levels of IL-13Rα2 in sc and liver tumors as assessed and normalized to β-actin. B, RT-PCR reveals levels of IL-13Rα2 mRNA expression in normal adrenal medulla, positive control (PM-RCC) and a variety of tumor samples across various genotype and tumor behavior backgrounds. Relative levels of IL-13Rα2 mRNA were assessed and normalized to β-actin. Expression of IL-13Rα2 in MPC (C) and mouse sc tumor tissue (D) by immunohistochemical staining reveals positive expression of IL-13Rα2. E, Cytotoxicity assay in MPC cells, negative control cells (T98G), and positive control cells (PM-RCC). TH, Tyrosine hydroxylase.

Sensitivity of MPC cell line and tumors to IL-13PE immunotoxin

Both MPC cells (Fig. 1C) and in vivo tumors (Fig. 1D) demonstrated IL-13Rα2 expression by IFA for plasma membrane as well as intracytoplasmic immunostaining. Biological function of overexpressed IL-13R was evaluated by treating with IL-13PE. As shown in Fig. 1E, MPC cells revealed moderately high sensitivity to IL-13PE (IC₅₀ <2.5 ng/ml), compared with no sensitivity in receptor-negative T98G (IC₅₀ >1000 ng/ml, P < 0.001). Receptor-positive PM-RCC served as a positive control with IC₅₀ = 0.23 ng/ml.

IL-13PE inhibits pheochromocytoma tumor growth in vivo

We developed an in vivo mouse model of pheochromocytoma by sc implanting 5 × 10⁶ MPC cells. The tumors developed in 10–14 d and were subgrouped by tumor size (n = 10 large, n = 10 small). The animals were treated (n = 5) with IL-13PE (100 μg/kg) or placebo IT. IL-13PE treatment resulted in suppression of established tumors compared with placebo after 3 d initial treatment of 75.1 ± 11.3 mm³ (mean ± sem) vs. control of 336.8 ± 57.7 mm³ (P = 0.0021) and through the observation period of 105.2 ± 58.5 mm³ compared with 1113.2 ± 540.9 mm³ in control at d 9 (P = 0.0032) (Fig. 2A). The study was repeated with smaller tumors and showed comparable results of 46.06 ± 3.900 mm³ vs. control 74.89 ± 14.02 mm³ (P = 0.0540) at d 3 and 44.21 ± 5.084 mm³ compared with 221.4 ± 51.98 mm³ in control at d 9 (P = 0.0095) (Fig. 2B). Additionally, IL-13PE treatment demonstrated a histological response in these tumors without any adverse effects (Fig. 2C).

Figure 2.

Intratumoral treatment of sc implanted mouse pheochromocytoma tumors with IL-13PE immunotoxin. Female nude mice were injected sc with 5 × 10⁶ MPC. A, IL-13PE was injected IT in large (95 ± 17 mm³, average ± sem) and small (53 ± 4 mm³) tumors with 100 μg dose per kilogram, three doses on consecutive d 1, 2, and 3 (arrows). Significance compared with control by unpaired t test. *, P < 0.05; **, P < 0.01; ***, P < 0.001. IL-13PE treatment regressed tumors in treated tumors (B) and hematoxylin and eosin staining of paraffin sections (×20 and ×60) (C) also revealed necrotic regions and nuclear fragmentation in treated tumors compared with untreated control.

Discussion

Our study provides direct evidence that IL-13Rα2 is overexpressed in pheochromocytoma tumors. Furthermore, we show IL-13PE immunotoxin may be a novel therapeutic approach in patients with IL-13Rα2-positive metastatic pheochromocytoma. We identified an overexpression of mRNA for IL-13Rα2 in murine and human pheochromocytoma tumors compared with normal adrenal tissues. These results are in corroboration with a previously published study of expression profiling on pheochromocytoma, which confirmed overexpression of IL-13Rα2 (18). IFA studies in MPC cells and tumors demonstrate increased expression of the IL-13Rα2, which suggest that overexpressed mRNA for IL-13Rα2 were translated into an immunoreactive protein.

The biological activity of IL-13Rα2 chain was examined in vitro and in vivo in an animal model of pheochromocytoma. In vitro, MPC demonstrated a dose-dependent sensitivity to IL-13PE. Similarly, IL-13PE caused regression of established tumors in vivo in mice. Previous studies have shown IL-13Rα2-negative tumors in vivo did not respond to IL-13PE treatment (19,20,21). These results provide a proof of principle by demonstrating that IT injections of IL-13PE, which binds IL-13Rα2 with high specificity, reduced tumor growth followed by complete regression in most animals. Nevertheless, the pheochromocytoma xenograft models can be limited when extrapolated to more clinically relevant orthotopic or transgenic models (22,23).

Currently therapeutic options available to pheochromocytoma patients are limited. Therefore, novel therapeutic agents and approaches are needed. Various preclinical models have been tested by administration of IL-13PE by different routes, and its safety and efficacy against different human cancers have been examined (9,24). On the basis of these studies, several clinical trials were initiated in patients with malignant brain tumors (10). In the present study, we identified overexpression of IL-13Rα2 in pheochromocytoma, another class of endocrine tumors. Thus, IL-13Rα2 overexpression may serve as a novel tumor target for IL-13PE-based receptor targeted therapy of pheochromocytoma. IL-13PE-based approach may serve beneficial with restricted adverse effects because it mediates its anticancer effects exclusively after binding to IL-13Rα2 in tumors. Thus, one would expect a low amount of toxicity to normal tissues because they do not express this receptor chain. Indeed in clinical trials for malignant glioma, no systemic toxicity was seen and no toxicity to normal brain was observed at the optimum dose (25,26).

However, to treat large metastatic pheochromocytomas, it may be necessary to optimize effective delivery of the drug, which may require needle repositioning and geographic overlap between successive zones for perfusion in the tumors. It is likely that tumors greater than 300 mm³ may be associated with a greater failure rate, most likely due to underperfusion of the tumors and failure to treat tissue beyond the tumor to create an adequate tumor-free margin. Therefore, for large tumors, a continuous intratumor delivery method may need to be tested (6). In addition, for metastatic pheochromocytoma, a combination approach of local administration and systemic delivery may be needed. In that regard, it has been reported that IL-13PE is well tolerated at a dosage of 2 μg/kg infused iv every alternate day for three injections.

In conclusion, the present study identified IL-13 receptor overexpression in neuroendocrine murine and human pheochromocytoma. Furthermore, this overexpression could be exploited for IL-13PE-based receptor-directed therapy, which targets IL-13Rα2. These results should prove beneficial in clinical settings because IT injections of IL-13PE reduced tumor growth followed by complete regression in most animals.