Fig 2: HIF-2α overexpression increases tumorigenicity and resistance to PTX by activating the Wnt and Notch pathways in vivo. a Green fluorescent protein (GFP) expression was detected in xenograft mice stably transfected with NC-cDNA and HIF-2α-cDNA MCF7 cells by small animal imaging. b HIF-2α expression was measured by western blot in xenograft tumor tissues. c Average tumor volumes were measured in xenograft mice every two days. d Images of resected MCF7 tumor tissues and average tumor weight at the end of indicated treatment. e TUNEL measurement of apoptosis in xenograft mice using fluorescent staining after treatment with or without PTX. Red fluorescence cells were considered positive (magnification × 600). Bar standards are 20 μm. f Expression of Wnt pathway-related proteins was detected by western blot in different treatment groups. g Expression of Notch pathway-related proteins was detected by western blot in different treatment groups. h Expression of c-Myc, OCT4 and Nanog was determined by western blot in different treatment groups. i Correlation analysis of the EPAS1 and Wnt, Notch pathway key factors in TCGA dataset. AXIN2, NOTCH1 and MYC display significant positive correlation with EPAS1 expression (p < 0.0001) in TCGA datasets. Data are shown as the mean ± SD (n = 6), *P < 0.05, **P < 0.01, ***P < 0.0001 compared with the NC-cDNA group. ##P < 0.01 compared with the NC-cDNA + PTX group

Fig 3: Effect of altered EPAS1 expression levels on LUAD metastasis. (A) TCGA analysis of the expression levels of EPAS1 in paired LUAD tissues. (B) TCGA analysis of the expression levels of EPAS1 in unpaired LUAD tissues. (C) ROC curve used to obtain a cut-off value of EPAS1 in LUAD. (D) LUAD patients, categorized based on the cut-off value of EPAS1. (E) Kaplan-Meier analysis of the association between EPAS1 expression and DFS as well as tumor recurrence. (F) Kaplan-Meier analysis of the association between EPAS1 expression and OS as well as poor survival.

Fig 4: Induction of pro-differentiation and proliferation proteins by Wnt3a/Rspo2 treatment. (A) Immunoblotting of BMP4 and EPAS1 after 48 hours of treatment shows increases in SK-N-BE(2)-C and SH-SY5Y cell-lines, but not in SK-N-AS. (B) Immunoblotting of CCND1, E2F1 and phospho-RB shows increases in SK-N-AS, but not in SK-N-BE(2)-C and SH-SY5Y cell-lines.

Fig 5: Identification of EPAS1 as a direct target of miR-182-5p in NSCLC cells. (A) Heatmap of 26 differentially expressed genes regulated by miR-182-5p in LUAD. (B) Pearson correlation analysis of the correlation between miR-182-5p and EPAS1 expressions in LUAD tissues. (C) Binding sites of miR-182-5p to WT or Mut 3' UTR of EPAS1. (D) Luciferase activities of WT or Mut 3' UTR of EPAS1 after co-transfection with miR-182-5p mimics and WT or Mut 3'UTR of EPAS1 in NSCLC cells. (***p<0.001).