Extended Data Fig. 7: Squamous reprogramming with hypoxia and Wnt/R-spondin deprivation.
a, The expression (left) and enrichment (right) of Hallmark hypoxia genes in the TCGA PAAD dataset. n = 31 (Squamous), 53 (Pancreatic progenitor), 27 (immunogenic) and 38 (ADEX). Box plots represent the median (centre line), upper and lower quartiles (box limits), and 1.5 × interquartile ranges (whiskers). b, Enrichment of 171 hypoxia-associated genes in ICGC and TCGA datasets. Statistics, one-way ANOVA followed by Tukey’s post-hoc test. Box plots represent the median (centre line), upper and lower quartiles (box limits), and 1.5 × interquartile ranges (whiskers). c, Comparative gene expression analysis of hypoxia-cultured KCTS#2-GATA6KO organoids vs the normoxia-cultured counterpart using edgeR. Genes with fold change > 4, adjusted P value < 0.05 are shown in red, and those with fold change < 0.25, adjusted P value < 0.05 are shown in blue. d, The growth and morphology of hypoxia-reprogrammed KCTS#2-GATA6KO organoids following a re-exposure to the indicated conditions (left) and viability measurement with ATP luminescence (right). After 2 months of culture in a Wnt/R-spondin-deficient hypoxic condition and acquisition of PASC features, KCTS#2-GATA6KO organoids were plated as single cells and cultured in the indicated conditions for seven days. Data are shown as mean ± s.d. Each dot represents one well. e, GATA6 expression in the PC organoid RNA-seq data. f–j, The morphology and p63 expression in the indicated lines following a two-month culture in normoxia or hypoxia. Wnt+ lines were also cultured in the presence or absence of R-spondin. Two Wnt+ lines (PC5 and PC29) showed upregulation of p63 in a hypoxic R-spondin-lacking condition, whereas the other lines did not. M, marker. Scale bar, 100 μm. For each image, similar result was obtained at least three times from different time points. Immunoassay data derive from single capillary runs.
