Fig. 3

Microglia loss-of-function model with transcriptional profiling of astrocytes. a Illustration of the workflow to set-up a human slice model combined with microglia depletion and tumor injection. Entry cortex was taken from the operation theatre, sliced within 10 min into 300 µm slices and cultured in serum-free conditions. In a 3 days time-course, slices are incubated with 1 mg/ml of Chlodronate to deplete microglia or control condition. 20,000 primary serum-free cultured ZsGreen tagged tumor cells were injected. After 4 days of tumor growth, astrocytes and microglia were purified for RNA-seq analysis (Detailed workflow is given in the Supplementary Fig. 6) b Representative staining of IBA-1 confirmed a robust depletion of microglia in human slices without loss of NeuN expression in neurons. c Representative staining of tumor injection in a time dependent manner. d Analysis of differentially expressed genes of purified astrocytes in Microglia(+) or Microglia(−) condition. Genes are ordered based on fold-change of gene expression between the control (Tumor(−)) and tumor injection (Tumor(+)), with blue indicating an > = 12 fold higher expressed in control samples and red an > = 12 fold higher expression in tumor injected samples. Lines indicate the differences of the fold-change rank (top 30 genes) between Microglia(+) and Microglia(−) condition. The R-code and detailed description is given in the source data. e Gene Set Expression Analysis (GSEA) of ranked gene expression of Microglia(+) or Microglia(−) condition indicate the enrichment differences of the JAK/STAT pathway. f Cytokine protein level in all conditions. Exact values, as well as statistical analysis are given in the source file. g Representative immunostaining of tumor infiltration after 4 days of culture and quantification h, i. Exact values of cell numbers are given in the source file. j Gating strategy to purify astrocytes from FACS data. Detailed plots for gating are given in the Supplementary Fig. 8. k FACS data analyzed by T-SNE, colors indicate the experimental conditions. I T-SNE map with STAT3-P (left) and Ki-67 (right) intensity, colors indicate the intensity (red: high intensity, blue: low intensity). P-values are determined by one-way ANOVA (f, h, i) adjusted by Benjamini–Hochberger (f, h, i) or False-Discovery Rate (e) for multiple testing. Data is given as mean ± standard deviation