Fig. 2: Loss of Gata4 in vivo is associated with reduction of inflammatory pathways and results in reduced tumor-infiltrating lymphocytes.

a RNA-seq was performed on bulk tumors and included both tumor and stromal cells. Individual tumors were micro-dissected from mouse lung tissue. GSEA on differentially expressed genes between sgGata4-targeted tumors and sgCtrl-targeted tumors is shown; genesets of interest were derived from the curated genesets from the Reactome database. Genesets of interest that were depleted in sgGata4-targeted tumors versus sgCtrl-targeted tumors are shown in red. Genesets of interest, which were enriched in sgGata4-targeted tumors versus sgCtrl-targeted tumors are shown in green; 10 out of the top 16 differentially expressed genesets are shown. All genesets displayed have a statistically significant FDR < 0.01. b Sets of genes whose expression correlates with the abundance of different subsets of immune cells (see Supplementary Data 1) were analyzed for their expression in sgGata4-targeted tumors versus sgCtrl-targeted tumors collected 16 weeks P.I. from the autochthonous KPC lung cancer model. Representative genes characterizing each immune cell type are shown with a heatmap representing the corresponding FDR q-value derived from the RNA-seq analysis of tumors. c Quantification of IHC of CD3 staining of serially sectioned lung lobes collected 16 weeks P.I. Error bars represent the mean ∓ SD and statistics derived from two-sided t-test. Data are derived from n = 4 mice totalling n = 139 tumors for sgCtrl and n = 6 mice totalling  n = 262 tumors for sgGata4. d Quantification of IHC of CD8 staining of serially sectioned lung lobes collected 16 weeks P.I. Error bars represent the mean ∓ SD, and statistics were derived from a two-sided t-test. Data are derived from n = 4 mice totalling n = 183 tumors for sgCtrl and n = 6 mice totalling n = 171 tumors for sgGata4.