Interleukin-13 receptor alpha 2 cooperates with EGFRvIII signaling to promote glioblastoma multiforme

Abstract

The interleukin-13 receptor alpha2 (IL-13Rα2) is a cancer-associated receptor overexpressed in human glioblastoma multiforme (GBM). This receptor is undetectable in normal brain which makes it a highly suitable target for diagnostic and therapeutic purposes. However, the pathological role of this receptor in GBM remains to be established. Here we report that IL-13Rα2 alone induces invasiveness of human GBM cells without affecting their proliferation. In contrast, in the presence of the mutant EGFR (EGFRvIII), IL-13Rα2 promotes GBM cell proliferation in vitro and in vivo. Mechanistically, the cytoplasmic domain of IL-13Rα2 specifically binds to EGFRvIII, and this binding upregulates the tyrosine kinase activity of EGFRvIII and activates the RAS/RAF/MEK/ERK and STAT3 pathways. Our findings support the "To Go or To Grow" hypothesis whereby IL-13Rα2 serves as a molecular switch from invasion to proliferation, and suggest that targeting both receptors with STAT3 signaling inhibitor might be a therapeutic approach for the treatment of GBM.

Fig. 1

GBM patients co-expressing EGFR and IL-13Rα2 correlate to poor survival where the overexpression of IL-13Rα2 alone leads to enhance cell migration but not proliferation. Kaplan−Meier survival analysis of a all gliomas patients; b GBM patients from REMBRANDT database from National Cancer Institute (USA). Patients overexpressing EGFR mRNA by 2-fold (blue) with high (red), intermediate (yellow) and low (green) levels of IL-13Rα2 expression were shown. The log-rank p-values were indicated. c Kaplan−Meier survival plots for patients expressing high YKL-40 mRNA levels TCGA. High IL-13Rα2 expression group (red) and low IL-13Rα2 expression group (blue) were determined by aggregating all patients whose z-score normalized expression was above or below 0, respectively (Log-rank test p-value = 0.0374). Immunoblotting analysis showed the expression of EGFR and IL-13Rα2 protein levels were determined from d a panel of 10 patient-derived GBM e and the isogenic cell lines generated from Gli36 glioma cells. Pan-actin or β tubulin served as internal loading controls. f Cell proliferation and g Cell cycle analysis were performed with Gli36 and Gli36.IL-13Rα2 cells h Soft agar colony formation assay was performed, Gli36.EGFRvIII was used as a positive control. i In vitro migration and j invasion assays were determined in Gli36 and Gli36.IL-13Rα2 cells. All data are represented as mean ± SEM, unpaired t-test **p < 0.01; ***p < 0.001; NS not significant

Fig. 2

Ectopic expression of IL-13Rα2 promotes glioma invasion. a In vitro migratory capacity of control and IL-13Rα2-RNAi treated U87MG cells was determined. Percent of migrated cells was normalized to CTRL-RNAi. b U87MG cells were transfected with non-specific siRNA (CTRL-RNAi) or IL-13Rα2 specific siRNA (IL-13Rα2-RNAi). Cell proliferation was subsequently determined, and the percent of proliferation was normalized to CTRL-RNAi day 1. c In vitro migratory capacity of control and IL-13Rα2-RNAi treated U87MG cells, stimulated with 1 µg ml⁻¹ YKL40 or 20 ng ml⁻¹ IL-13 for 18 h, was determined using wound-healing migration assay. All data are represented as mean ± SEM. Unpaired t-test *p < 0.05; ***p < 0.001; NS not significant. d Immunoblotting experiment showing upregulation of MMP-2, and vimentin in Gli36.IL-13Rα2 cells. Densitometry quantification was done for the indicated proteins by normalizing to pan-actin as the internal loading control. Ratios were indicated below each blot. e Mice were implanted with either Gli36-GFP or Gli36.IL-13Rα2-GFP cells intracranially. Tumors were collected from representative mice implanted with Gli36.IL-13Rα2-GFP showed the invasive phenotype (left panel) when compared to the contralateral normal brain parenchyma (right panel) by haematoxylin and eosin (H&E) staining. Red arrows indicated glioma tumor at the invasive front f Representative images of mouse brain transplanted with Gli36- GFP cells and Gli36-IL-13Rα2-GFP, counterstained with DAPI (blue). The top panel shows the contralateral hemisphere; bottom panel shows tumor-bearing hemisphere of the mouse brain (N, normal; T, tumor). Scale bar, 50 μm. g Immunofluorescence red staining of MMP-2, vimentin or isotypic control in mice bearing either Gli36.IL-13Rα2-GFP or Gli36-GFP. Scale bar, 50 μm. h Kaplan−Meier survival curves of mice bearing Gli36 and Gli36.IL-13Rα2 tumors. NS not significant

Fig. 3

Co-expression of IL-13Rα2 and EGFRvIII enhances glioma transformation. a Cell proliferation of Gli36.EGFRvIII and Gli36.IL-13Rα2/EGFRvIII. b The cell cycle profile was compared between Gli36.EGFRvIII and Gli36.IL-13Rα2/EGFRvIII using flow cytometry analysis. c Soft agar colony formation assay of the indicated cells was performed, including the non-transforming Gli36 cells as a negative control. d IL-13Rα2 silencing (as validated by immunoblot presented as an insert) significantly reduced the proliferation of Gli36.IL-13Rα2/EGFRvIII cells to a level similar to the Gli36.EGFRvIII cells. e In vitro migratory capacity of Gli36.EGFRvIII and Gli36.IL-13Rα2/EGFRvIII cells. f U87MG.EGFRvIII cells were transfected with non-specific siRNA (CTRL-RNAi) or IL-13Rα2 specific siRNA (IL-13Rα2-RNAi). Cell proliferation was determined with CCK-8 assay in U87.EGFRvIII. The percent proliferation was normalized to CTRL-RNAi day 1. g In vitro migratory capacity of control and IL-13Rα2-RNAi treated U87MG.EGFRvIII was determined. Percent of migrated cells was normalized to CTRL-RNAi. All data are represented as mean ± SEM. ANOVA with Tukey's multiple comparison tests *p < 0.05; ***p < 0.001; NS not significant. h Kaplan−Meier survival curves of mice bearing Gli36.EGFRvIII and Gli36.IL-13Rα2/EGFRvIII tumors *p < 0.0292

Fig. 4

Oncogenic signaling of IL-13Rα2 increases tyrosine kinase activities and promote cell proliferation through activation of the RAS/RAF/MEK/ERK signaling cascade. a Cell lysates from Gli36, Gli36.IL-13Rα2, Gli36.EGFRvIII, and Gli36.EGFRvIII/IL-13Rα2 post-knockdown with scrambled or IL-13Rα2 siRNA were examined for the total levels of tyrosine phosphorylation using anti-phosphotyrosine antibodies. b CCK-8 proliferation assay was performed with and without 10 μM AG1478 treatments. All data are represented as mean ± SEM. Unpaired t-test ***p < 0.001 c Endogenous expression levels of the RAS/RAF/MEK/ERK signaling were examined in the indicated cells with and without AG1478 treatment. RAS activation was determined by either d Raf-1 RBD agarose beads pull-down assay or e ELISA assay. Endogenous protein expression levels of f Total and p-C-RAF, Total and p-MEK/p-ERK g Total and p-STAT3 h PTEN, total and p-PI3K p85α and total and p-AKT were examined in the indicated cells. Normal human astrocytes transfected with EGFRvIII, IL-13Rα2 or co-expressing both receptors were examined for endogenous expression levels of i Total and p-C-RAF, Total and p-MEK/p-ERK, j Total and p-STAT3, k PTEN, total and p-PI3K p85α and total and p-AKT. For all immunoblots, pan-actin served as internal loading controls, and band densitometry quantifications for the proteins were performed using ImageJ (NIH). The value derived from densitometry quantification is obtained from normalizing each of the signaling proteins against actin in a single experiment, and presented as a ratio of phosphorylate form over total protein. Each of these was performed at least two independent times

Fig. 5

Deletion of the cytoplasmic domain of IL-13Rα2 resulted in a loss of physical interaction with EGFRvIII and enhanced proliferation is abolished. a Whole-cell lysates prepared from stable cell line Gli36.IL-13Rα2/EGFRvIII cells were used for immunoprecipitation with anti-IL-13Rα2 antibody, then immunoprobed with an anti-EGFR antibody. IgG served as control while unprecipitated extracts serve as input. b Similar cell lysates were reverse immunoprecipitated with anti-EGFR antibody, then immunoprobed with an anti-IL13Rα2antibody. Lysates from Gli36.EGFRvIII served as additional control c Gli36.IL-13Rα2/EGFRvIII cell lysates were immunoprecipitated with anti-EGFR antibody, then immunoprobed with anti-Grb antibody. To further examine the domains of interaction, IL-13Rα2 and EGFR mutants were used. Gli36.EGFRvIII cells were first transfected with pIRESneo2 (Vector), IL-13Rα2 full length (Wild-type) and IL-13Rα2 Cyt tail deleted constructs (Mutant) and then analyzed by d cell proliferation assay at the indicated time points, f co-immunoprecipitation, and h PLA assays. Findings were validated using Gli36.IL-13Rα2 cells transiently transfected with vector (CTRL), full length/wild-type EGFRvIII, DK, and DY3 mutants. e proliferation outputs, g co-immunoprecipitation, i and PLA assay were performed. j represent the corresponding positive and negative controls

Fig. 6

Enhanced cellular proliferation mediated by IL-13Rα2 is specific to EGFRvIII, and not WT EGFR. a U251-E6 or c U251-E18 cells were treated with or without tetracycline (Tet). At indicated time points, immunoblot analysis was carried out. Gli36, Gli36.EGFRvIII cell lysates were included as negative or positive controls for EGFRvIII, respectively. Growth kinetics of b U251-E6 and d U251-E18 was determined by CCK-8 assay. Percent cell viability was normalized to day 1 (without induction). All data are represented as mean ± SEM. Unpaired t-test ***p < 0.001, NS. not significant. e Co-immunoprecipitation was performed in stable cell lines Gli36.IL-13Rα2/wtEGFR as well as U251MG-E6 (i.e. wtEGFR) cells at 48 h post tetracycline induction with the indicated antibodies. Gli36.IL-13Rα2/EGFRvIII served as positive controls. f The interaction between endogenous wtEGFR and IL-13Rα2 was shown in primary wtEGFR-positive GBM patient tumor derived from Mayo clinic, and IgG served as positive and negative controls respectively. Knockdown of IL-13Rα2 in cell line or patient-derived GBM samples expressing g wtEGFR or h EGFRvIII

Fig. 7

IL-13Rα2 mediate greater tumorigenic potential with EGFRvIII, and not WT EGFR. a Tumor volume b and tumor weight of tetracycline regulatable U251 gliomas (U251-E6 and U251-E18 was examined in vivo. Bars depict the mean values and error bars represent 95% confidence intervals. P-values were calculated using ANOVA with Tukey’s multiple comparison test *p < 0.05; **p < 0.01; ***p < 0.001. Photomicrographs of represented collected tumors are shown. c Immunoblot analysis of proteins from U251-E6 and -E18 tumor lysates in the presence or absence of tetracycline with the indicated antibodies. One representative tumor under each of the uninduced and induced conditions was shown. U251MG whole-cell lysate served as positive control for IL-13Rα2. d Kaplan−Meier survival curves of mice bearing U251-E6 and U251-18 tumors **p < 0.0039. Kaplan−Meier survival plots for patients expressing e high EGFR mRNA levels (excluding EGFRvIII) or f high EGFRvIII mRNA levels from TCGA database. High IL-13Rα2 expression (red) and low IL-13Rα2 expression (blue) were determined by aggregating all patients whose z-score normalized expression was above or below 0, respectively. g Schematic model showing signal transduction pathway co-induced by IL-13Rα2 and EGFRvIII. Overexpression of IL-13Rα2 in human gliomas increases cell migration and invasion through the activation of MMP-2, vimentin. Amplification of EGFRvIII promotes the co-interaction of both receptors mediating an increase in tyrosine kinase activities and a preferential activation of RAS-MEK-ERK and STAT3 pathways leading to aberrant cellular proliferation