Journal: Cell Death & Disease
Article Title: Interleukin-8/CXCR2 signaling regulates therapy-induced plasticity and enhances tumorigenicity in glioblastoma
doi: 10.1038/s41419-019-1387-6
Figure Lengend Snippet: a Limiting dilution neurosphere assays were performed on two cell lines—GBM43 and GBM6—after treatment with 50 ng/ml of IL-8. Stem cell frequency for GBM43 with IL-8 35.7, lower limit 50.7 and upper limit 25.1 as compared with no IL-8 118.6, lower limit 169.2 and upper limit 83.1, p = 0.0156; for GBM6 with IL-8 27.1, lower limit 39.2 and upper limit 18.8 as compared with no IL-8 60.7, lower limit 85 and upper limit 18.8, p = 0.001. b To determine the ability of IL-8 to influence cellular plasticity, we employed a reporter cell line in which RFP expression is controlled the OCT4 promoter. Cells were treated with 50 ng/ml of IL-8, and RFP expression was monitored by FACS over 6 days. Treatment increased both Oct4 and Sox2. Bars represent means from three independent experiments and error bars represent the standard deviation. Multiple Student's t tests were performed. ** p < 0.01, *** p < 0.001. c The network of GIC-promoting genes in patient tumors, we selected the top GIC-associated genes activated during TMZ therapy (Fig. and Supplementary Table ) and correlated with their levels with IL-8 mRNA using the GlioVis data portal for visualization and analysis of brain tumor expression database (gliovis.bioinfo.cnio.es; dataset LeeY) 16–18. IL-8 expression was significantly correlated with critical GIC-associated genes including KLF4, CD44, HIF1A, HIF2A, Myc, and Twist expression (Fig. 5c). The IL-8 expression in different anatomical location and potential colocalization of these GIC-specific genes with areas of high IL-8 transcript level (Fig. 5c, heatmap). d Immunoblot analysis of endogenous glioma-initiating cell-associated transcription factors expression upon stimulation with escalation dose of IL-8 (0–100 ng/ml) for 24 h. Protein extracts of IL-8-treated PDX lines GBM43 and GBM6 were immunoblotted with antibody against several GIC markers, including c-myc, Sox2, Nanog, KLF4, OCT4, or an antibody against β-actin as a control for equal loading. e GBM43 PDX cells were treated with neutralizing antibody against IL-8 or control IgG antibody (100 ng/ml) prior to treatment with DMSO or 50 µM TMZ. Neutralizing antibody was added every day for 8 days,and protein extracts from this experiment were immunoblotted with antibody against c-myc, Sox2, OCT4, or an antibody against β-actin as a control for equal loading. f Schematic diagram of experiment design for in vivo testing. Top graph, U251 cells were infect with lentivirus (Sigma Mission shRNA) shRNA against IL-8 or scrambled shRNA (control) with 10 infectious unit/cell. In total, 2 × 10 5 transduced cells were stereotactically injected into the right hemisphere of the brain of athymic nude mice ( n = 8 per group, four males and four females). Two weeks after implantation, two groups of mice, control, and knock down, were treated with vehicle treated (DMSO, top curve) or TMZ (2.5 mg/kg) intraperitoneally. Survival curves were obtained by the Kaplan–Meier method, and overall survival time was compared between groups using log-rank test. All statistical tests were two-sided. Bottom graph, to examine the role of IL-8 in GBM progression in a more clinically relevant manner, next the same method was used to knockdown the IL-8 expression in GBM43 PDX line. In total, 1.5 × 10 5 cells were injected stereotactically into the right hemisphere of the brain of athymic nude mice ( n = 8 per group, four males and four females). Survival curves were obtained by the Kaplan–Meier method, and overall survival time was compared between groups using log-rank test
Article Snippet: The following antibodies were employed: mouse anti-IL-8 (R&D), rabbit anti-c-Myc (Cell Signaling, Danvers, MA, USA), rabbit anti-Sox2 (Cell Signaling), rabbit anti-Nanog (Cell Signaling), rabbit anti-LIN28A (Cell Signaling), rabbit anti-KLF4 (Cell Signaling), mouse anti-β-actin (Abgent, San Diego, CA, USA), rabbit anti-EZH2 (Cell Signaling), mouse anti-phosphoEZH2 (AbCam, Cambridge, UK), rabbit anti-Bmi1 (Cell Signaling), rabbit anti-ring1A (Cell Signaling), rabbit anti-SUZ12 (Cell Signaling), rabbit anti-H3 (Cell Signaling), rabbit anti-H3K4methyl3 (Cell Signaling), rabbit anti-H3K27acetyl (Cell Signaling), rabbit anti-H3K27methyl3 (Cell Signaling), rabbit anti-H3K36methyl3 (Cell Signaling), and rabbit anti-H2A-Ubiquitin (Cell Signaling).
Techniques: Expressing, Standard Deviation, Western Blot, In Vivo, shRNA, Injection
Proteintech is a verified supplier 
![a PDX GBM43 cells were treated with TMZ (50 μM), and RNA collected after 4 or 8 days. Microarray analysis using Affymetrix platform was performed. Gene set enrichment analysis (GSEA) for genes known to support and maintain the GIC phenotype and demonstrated significance increase after 8 days [day 4 FDR q 0.82 and day 8 FDR 0.08 and FWER p-value 0.046, TMZ-treated cells as compared with DMSO controls]. b Whole-genome ChIP-seq analysis of H3K27 trimethylation (H3K27me) and H3k27 acetylation mark (H2K27ac) on PDX GBM43 following 4 days treatment with TMZ or vehicle control. Top two tracks represent the BET file showing significant enrichment of histone mark relative to DMSO control. Bottom two lanes represent WIG file showing significant enrichment peak. Analysis of three established astrocyte enhancer tracks for the <t>IL-8</t> gene showed no changes in H3K27me3; however, they did show significant change in #3 enhancer region [Chr. 4 :74783222–74783418, fold enrichment 3.2 relative to IgG input. p < 0.0001. FDR = 0.004]. Right bottom inset image shows zoomed in view of the #3 enhancer region. Left bottom inset image represents overlayed peaks for TMZ and DMSO control. c Expression of IL-8 mRNA after exposure to 50 µM TMZ was determined by quantitative real-time polymerase chain reaction (qPCR) after treatment with TMZ across 8 days. All IL-8 values were normalized to glyceraldehyde 3-phosphate dehydrogenase (GAPDH). Bars represent means from three independent experiments and error bars represent the standard deviation. Multiple Student's t tests were performed. ** p < 0.01, *** p < 0.001. d Immunohistochemistry was performed on mouse brains with intracranial xenografts of GBM43. 1.5 × 10 5 GBM43 PDX cells were implanted to establish orthotropic xenograft tumors. Animals received 2.5 mg/kg of either DMSO or TMZ for five consecutive days, beginning 7 days after tumor implantation. Animals were killed 5 days after the cessation of treatment, and whole brains were extracted, flash frozen, then sectioned (8 µm) and analyzed by immunofluorescence. 4′-6-diamidino-2-phenylindole (DAPI) stained DNA (blue) in the nuclei and allophycocyanin-conjugated secondary (orange) antibody was used against primary antibody for IL-8. Dotted lines represent the edge of the tumor based on cell density and morphology. e–f PDX GBM xenografts were harvested and immediately plated in either mild differentiation media (DMEM containing 1% FBS) or GIC maintenance media (neurobasal supplemented with FGF and EGF). Cells were then treated with either DMSO or TMZ (50 μM). After 4 days, IL-8 levels were determined by ELISA. g GBM cells were treated with physiologically relevant dose of TMZ (50 μM), carmustine (BCNU, 100 μM), or equimolar DMSO. After 24 h, conditioned media was collected and IL-8 levels were quantified by ELISA. Bars represent means from three independent experiments and error bars represent the standard deviation. Multiple Student's t tests were performed. ** p < 0.01, *** p < 0.001](https://pub-med-central-images-cdn.bioz.com/pub_med_central_ids_ending_with_1047/pmc06441047/pmc06441047__41419_2019_1387_Fig1_HTML.jpg)
![a IL-8 mRNA expression levels were analyzed using the Affymetrix U133a platform on the Cancer Genome Atlas (TCGA) for different WHO grades of glioma. GBM (Grade IV glioma) patients had higher IL-8 expression level than low-grade (Grade II, Grade III) glioma patients. b All GBM patients were stratified into IL-8-upregulated and IL-8-downregulated groups based on IL-8 gene expression using quartile (Q1, Q3) as split points. High expression of IL-8 correlated with reduced median survival. Survival curves were generated via the Kaplan–Meier method and compared by log-rank test. ** p < 0.01. Multivariate stepwise Cox proportional hazards model with stepwise variable selection was conducted to examine whether IL-8 could be an independent factor for predicting survival with major clinical variables adjusted. This analysis confirmed that IL-8 was an independent prognostic factor for survival in all GBM patients (HR [95% CI]: all GBM 1.07 [1.01, 1.14], p = 0.0467). c Index (95% CI) or C statistics are provided. c Within grade IV glioma subtypes, proneural GBM patients had the lowest level of IL-8 expression. Boxplots represent means and interquartile range. One-way ANOVAs with Bonferroni correction for the multiple comparisons were performed. * p < 0.05, *** p < 0.001. d Patients with proneural GBM were stratified into IL-8-upregulated and IL-8-downregulated groups based on IL-8 gene expression using quartile (Q1, Q3) as split points. Kaplan–Meier survival curves and multivariate stepwise Cox proportional hazards models were generated as in B. e In TCGA database (U133a) GBM patients with IL-8-downregulated also have a longer time to recurrence, compared with IL-8-upregulated patients. f The Ivy Glioblastoma Atlas Project (IVY GAP) was employed to determine the location of IL-8 in glioblastoma samples. Each column represents the data for one biopsy from a tumor. Microdissection for the noted anatomically portions of the tumor and subsequent mRNA extraction and expression analysis demonstrated that IL-8 is upregulated in the perinecrotic zone and pseudopalisading cells. Heatmap illustrates most significantly and differential expressed genes with a false discovery rate < 0.01. mRNA expression in each anatomical compartment were compared. IL-8 was significantly upregulated in the perinecrotic zone. Bars represent means from three independent experiments and error bars represent the standard deviation. Multiple Student's t tests were performed. *** p < 0.001. g Brain tumor samples from primary biopsies or surgical resections were stained for IL-8 at the Northwestern Brain Tumor Tissue Bank. Histological and morphological analysis confirm that IL-8 is present in the perinecrotic zone and pseduopalisading cells. Scale bar 50 microns](https://pub-med-central-images-cdn.bioz.com/pub_med_central_ids_ending_with_1047/pmc06441047/pmc06441047__41419_2019_1387_Fig2_HTML.jpg)
