Inositol-requiring enzyme 1alpha is a key regulator of angiogenesis and invasion in malignant glioma

Abstract

Inositol-requiring enzyme 1 (IRE1) is a proximal endoplasmic reticulum (ER) stress sensor and a central mediator of the unfolded protein response. In a human glioma model, inhibition of IRE1alpha correlated with down-regulation of prevalent proangiogenic factors such as VEGF-A, IL-1beta, IL-6, and IL-8. Significant up-regulation of antiangiogenic gene transcripts was also apparent. These transcripts encode SPARC, decorin, thrombospondin-1, and other matrix proteins functionally linked to mesenchymal differentiation and glioma invasiveness. In vivo, using both the chick chorio-allantoic membrane assay and a mouse orthotopic brain model, we observed in tumors underexpressing IRE1: (i) reduction of angiogenesis and blood perfusion, (ii) a decreased growth rate, and (iii) extensive invasiveness and blood vessel cooption. This phenotypic change was consistently associated with increased overall survival in glioma-implanted recipient mice. Ectopic expression of IL-6 in IRE1-deficient tumors restored angiogenesis and neutralized vessel cooption but did not reverse the mesenchymal/infiltrative cell phenotype. The ischemia-responsive IRE1 protein is thus identified as a key regulator of tumor neovascularization and invasiveness.

Fig. 1.

Histopathology distinguishes two glioma phenotypes after xenotransplantation of U87ctrl and U87dn cells in the mouse brain. IHC labeling was developed on coronal sections at days 26–53 of tumor development. (A) H&E staining. (B) Reticulin fibers staining. (C) Labeling using anti-sm-actin antibodies. (D) Colabeling using anti-Ki-67 (pink color, nuclei) and anti-GFAP (orange color, cytoplasm) antibodies. (Scale bars: 100 μm.) (E) Kaplan-Meier survival analysis after implantation of U87wt cells, U87ctrl cells (two clones), and U87dn cells (three clones). n, number of mice. Results in exp. 1 (curves and Inset) and 2 (Inset only) are representative of four independent experiments. 1C5 vs. T1P5, P < 0.0001; 2A4 or 2D3 vs. wt or T2P4, P < 0.0001; T2P4 vs. wt, P < 0.135.

Fig. 2.

Analysis of the vascular bed of IRE1-dn tumors using immunofluorescence and IVM. Tumors were allowed to develop for 28–45 d after intracerebral implantation. (A) U87dn cells coopt blood vessels. (a and b) Immunofluorescent labeling using anti-vimentin (tumor cells). (Scale bars: 200 μm.) (c and d) Edges of growing tumors, as depicted by detection of vimentin and of CD31 (endothelial cells). (Scale bars: 50 μm.) (e–g) Detail of U87dn cell invasion alongside blood vessels. (Scale bars: 50 μm.) (B) Labeling of the vascular bed in tumors using anti-CD31 and anti-endoglin (proliferating endothelial cells) antibodies. Asterisks focus on the avascular bulks of U87dn tumors. (Scale bars: 100 μm.) (C) Transwell plate migration assay. U87 cells were allowed to migrate for 3 h with or without 10% FBS. Results are means ± SD from triplicate of three independent experiments (*P < 0.01). (D and E) IVM analyses. (D) Kinetic measurements of total and of functional vessel densities in U87ctrl- (△) and U87dn- (●) derived tumors (n = 4 in each group) implanted s.c. or intracerebrally (i.c.) in nude mice. Values are means ± SD (E) Images of i.c. tumor microcirculation. Tumor vascularization (arrows) is predominant in U87ctrl tumors and is reduced in U87dn tumors. Physiologic microvessels (low permeability, bright intravascular fluorescent signal) are readily visualized in U87dn tumors because of the significant reduction of tumor vessels. Physiological and organized angioarchitecture (arterioles, capillaries, and venules) and vessels of constant diameters (stars pointing at physiological cortical microvessels) are observed in dn tumors.

Fig. 3.

Angiogenesis vs. invasive phenotypes of U87 cell-derived tumors in the chicken egg model. U87 cells were deposited onto the chicken CAM and tumors were allowed to grow for 4 d. (A) Representative views of the CAM and of U87wt-, U87ctrl-, and U87dn-derived tumors delimited by plastic rings. (Scale bar: 2 mm.) (B) Histogram repartition in percent of the different tumor phenotypes (U87wt, n = 15; U87ctrl, n = 26; U87dn, n = 14) according to criteria defined in SI Materials and Methods. Left, size of the tumors; Center, percentage of invasive tumors; Right, degree of tumor vascularization. Results are mean values ± SD (*P < 0.01; **P < 0.005; ***P < 0.001). ns, not significant. (C) Transversal section of U87ctrl and U87dn tumors at day 4. Blood vessels (SNA lectin) and glioma cells (vimentin) were shown. (Left) Vertical presentation of CAM and tumor masses. (Right) Higher magnifications of boxed area in Left. (D) Gene expression analysis. Results are means ± SD of triplicate measures obtained from at least two independent analyzes (n ≥ 5 eggs for each condition). (E) Protein expression analyses. Results are expressed as picograms of the cytokines relative to nanograms of human B2M. Data represent mean values ± SD (n ≥ 5 eggs for each condition).

Fig. 4.

Gene and protein expression analyses in U87 cell-derived tumors in mice brains. (A) LCM analysis. U87 cells were implanted intracerebrally in mice and brains were removed 4–8 wk after implantation. Comparative gene expression analyses are represented as fold increase and are means ± SD of triplicate experiments. Most values were obtained by using the SYBR Green dye detection procedure, the TaqMan approach being also used for IL-6, IL-8, and VEGF-A. exp., experiments; NC, no change; → 0, No Ct value obtained with U87dn tumors; → ∞, value > 3,000. (B) Immunolabeling of SPARC in U87ctrl- and U87dn- gliomas. DAPI-labeled nuclei are in blue. Dashed lines show borders between tumor (t.) and normal brain (n.) tissues. (Scale bars: 50 μm.) (C) Protein quantification by ELISA. Pooled brain extracts were analyzed (n = 5 for each condition), and results are reported in picograms of cytokines per nanogram of B2M. Data represents means ± SD. ELISA were performed twice with similar results.