Necrotic cells influence migration and invasion of glioblastoma via NF-κB/AP-1-mediated IL-8 regulation

Abstract

Glioblastoma multiforme (GBM) is the most common primary intracranial tumor in adults and has poor prognosis. Diffuse infiltration into normal brain parenchyma, rapid growth, and the presence of necrosis are remarkable hallmarks of GBM. However, the effect of necrotic cells on GBM growth and metastasis is poorly understood at present. In this study, we examined the biological significance of necrotic tissues by exploring the molecular mechanisms underlying the signaling network between necrotic tissues and GBM cells. The migration and invasion of the GBM cell line CRT-MG was significantly enhanced by treatment with necrotic cells, as shown by assays for scratch wound healing and spheroid invasion. Incubation with necrotic cells induced IL-8 secretion in CRT-MG cells in a dose-dependent manner. In human GBM tissues, IL-8 positive cells were mainly distributed in the perinecrotic region, as seen in immunohistochemistry and immunofluorescence analysis. Necrotic cells induced NF-κB and AP-1 activation and their binding to the IL-8 promoter, leading to enhanced IL-8 production and secretion in GBM cells. Our data demonstrate that when GBM cells are exposed to and stimulated by necrotic cells, the migration and invasion of GBM cells are enhanced and facilitated via NF-κB/AP-1 mediated IL-8 upregulation.

Figure 1. Necrotic cells increase migration of CRT-MG cells.

(a) Representative micrographs of the scratch wound healing assay in CRT-MG cells treated with necrotic CRT-MG cells incubated for 0, 24, and 48 h. scale bar = 100 μm. Migration activity was measured by calculating the area that advanced from boundary lines of scratch to cell-free space for 48 h. Data are presented as the fold induction compared with each untreated control cells (right). **P < 0.01 vs. control. (b) Cell lysates of CRT-MG cells treated or untreated with necrotic cells for 24 h were analyzed using a chemokine array. The duplicate spots corresponding to IL-8 are indicated by a solid box (left), and reference spots positive control are indicated by a dashed line box. The intensity of spots corresponding to IL-8 and positive control was quantified using Image J software and subtracted from the background, then expressed as a ratio to the positive control. The value of IL-8 in the untreated control was set as 1. **P < 0.01 vs. control. NC, necrotic cells. (c) The scratch wound healing assay was performed using CRT-MG cells treated with necrotic CRT-MG cells in the presence or absence of either anti-IL-8 neutralizing antibody (2.5 μg/ml) or control nonspecific goat-IgG (2.5 μg/ml). Data shown are representative of at least three experiments. ***P < 0.001 vs. control, ^##P < 0.01 vs. necrotic cells. (d) CRT-MG cells treated with different ratios of necrotic cells (NC) or necrotic cells only (NC only) for 24 h as indicated. After incubation, supernatants from each condition were collected, and IL-8 protein levels were measured by ELISA. ***P < 0.001 vs. control. (e) Quantitative real-time PCR (qRT-PCR) analysis was performed for CRT-MG cells incubated in the absence or presence of necrotic CRT-MG cells for 24 h. Data were presented as fold induction compared with control cells. **P < 0.01, ***P < 0.001 vs. control.

Figure 2. Necrotic cells induce IL-8 expression and cell invasion.

(a) Immunohistochemical-l staining for IL-8 protein expression in the perinecrotic region was performed for 12 cases of the grade IV glioblastoma. Boxed areas of the necrotic border (1, 2, original magnification, ×100) are shown enlarged in the insets (3, 4, original magnification, ×400). The tumor cells distant from necrosis were IL-8 negative (arrow). PN, perinecrotic area; N, necrosis. (b) Double staining of IL-8 and GFAP in glioblastoma tissues. Tissues were immunostained with primary antibodies against IL-8 and the astrocyte marker GFAP. Nuclei were stained with DAPI. The experiment was repeated at least three times, with similar results (scale bar = 20 μm). Boxed areas of necrotic border are shown as a Z-stack projection image. YZ and XZ cross-sections are presented in the right and bottom panels (IL-8, red; GFAP, gray; Merged, pink) through the merged XY image showings the position of XZ or YZ cross-sections (pink) (10; scale bar = 10 μm; N, necrosis). (c) Invasion of CRT-MG cell clones stably transfected with construct IL-8p-d2EGFP was analyzed with a spheroid invasion assay. Spheroids were grown in the presence or absence of necrotic cells for up to 4 days and imaged by microscopy. Cell invasion was quantified by calculating the invading area as indicated in the Methods (right). The extent of invasion in the untreated control was set as 1 and data presented as the fold induction compared with untreated control cells. **P < 0.01, ***P < 0.001 vs. control. NC, necrotic cells. Data represent of three independent experiments. Scale bar represents 200 μm for all panels.

Figure 3. Necrotic cells induce phosphorylation of p38, JNK, and IκBα and IL-8 induction.

(a) CRT-MG cells were either untreated (control) or treated with necrotic cells (NC) for 24 h, and 430 μg of cell lysate were used for the phosphokinase assay as indicated in Methods. The duplicate spots corresponding to the increased phosphorylation of p38, JNK, and c-Jun and decreased phosphorylation of FAK are highlighted in the solid box (left). The dashed line box indicates reference spots. The intensity of spots corresponding to p38, JNK, c-Jun, FAK, and the positive control was quantified using Image J software and subtracted from the background, then expressed as a ratio to the positive control. *P < 0.05 vs. each control. n.s., not significant. (b) Cell lysates from the CRT-MG cells treated with necrotic cells for 24 h were analyzed by immunoblotting to confirm the phosphokinase array results. Tubulin was used as the loading control. (c) The human IL-8 promoter AP-1 and NF-κB binding elements are indicated. (d–g) The nuclear extracts (NEs) from CRT-MG cells treated with or without necrotic cells were incubated with a radiolabeled DNA probe for the human IL-8 promoter and consensus oligonucleotides and subjected to an electro mobility shift assay. NEs were analyzed with probe for either AP-1 consensus oligonucleotide (d), AP-1 sequence within the IL-8 promoter (e), NF-κB consensus oligonucleotide (f), or NF-κB sequence within the IL-8 promoter (g). The competition assay was performed by adding a 100-fold molar excess of cold probe. Anti-c-Fos, p50 or -p65 antibodies were added to test the specificity of interaction. Data shown represent at least three experiments. F.P, free probe; C, competition; SS, supershift.

Figure 4. NF-κB, JNK, and p38 inhibitors reduced IL-8 expression, cell migration, and invasion in CRT-MG cells.

CRT-MG cells were pretreated with an NF-κB inhibitor (BAY 11-7082 [BAY]), a JNK inhibitor (SP600125 [SP]), or a p38 inhibitor (SB203580 [SB]) for 30 min and then exposed to necrotic cells for 24 h. (a) Indicated supernatants from each conditions were collected, and IL-8 protein levels were measured by ELISA. The data represent three independent experiments. (b) Scratch wound healing assay in CRT-MG cells treated with necrotic CRT-MG cells and inhibitors incubated for 24 h. Migration activity was measured by calculating the area that advanced from boundary lines of the scratch to the cell-free space. Data are presented as the fold induction compared with each untreated control cells. (c) Invasion of CRT-MG cells was analyzed with a spheroid invasion assay. Spheroids were grown in the presence or absence of necrotic cells and inhibitors. Cell invasion was quantitated by calculating the invading area as indicated in Methods. The extent of invasion in the untreated control was set as 1; data are presented as the fold induction compared with untreated control cells. These parameters were analyzed in seven independent spheroids per condition. *P < 0.05, **P < 0.01, ***P < 0.001 vs. control, ^###P < 0.001 vs. necrotic cells.

Figure 5. Proposed model for the effect of necrosis/necrotic cells in regulating glioblastoma invasion.

Exposure to necrotic cells induces activation of NF-κB and AP-1 via phosphorylation of p38, JNK and IκBα and their binding to the IL-8 promoter, leading to increased IL-8 production and secretion, migration and invasion of glioblastoma cells.