Extended Data Fig. 5: Neutrophil-triggered pseudohypoxia is required for the metastasis property of THY1+ CSCs.
a, GO enrichment analysis of notably upregulated genes (log2FC > 1, p.adj < 0.01) in HIF-signaturelow THY1+ CSC versus HIF-signaturelow THY1− tumour cell-related regions. b, Heatmap of neutrophil infiltration indices in TCGA cancer tissues with low or high THY1+ CSC hypoxia signature score, calculated by GSVA using specific markers (log2FC > 1, min.pct > 0.2) defining THY1⁺ CSC-related hypoxic regions in the spatial transcriptome analysis. c, Efficiency of THY1-knockdown (shTHY1) or overexpression (THY1-OE) in Huh-7 cells (n = 5). d, Efficiency of HIF1A-GFP-overexpression (HIF1α-GFP) in Huh-7 cells (n = 3; scale bar, 10 µm), wild-type GFP was used as control. e, Metastasis score of THY1+ CSCs in patients with HCC with high versus low neutrophil infiltration (n = 94 patients). f, Gating strategies for FACS of neutrophils from mouse hepatoma tumours. g, Depleting efficiency of αLy6G antibody on neutrophils in mouse hepatoma model (n = 5). Data are from three (d) or five (c) independent experiments. Data are presented as mean ± s.e.m. of three (g) independent experiments. Statistical significance was determined using over-representation analysis based on the hypergeometric test, with false discovery rate correction applied for multiple comparisons (a), the Mann–Whitney U test (b,e), with false discovery rate correction applied for multiple comparisons (b), or two-way ANOVA with Tukey’s post hoc test (g). Violin plots show data distribution with median and interquartile range (e).
