Fig 2: E47 knockout mice are protected from GC-induced side effects. a i.p. glucose tolerance test (GTT) and b i.p. pyruvate tolerance test (PTT) of E47-/- and WT mice on Dex. Data were analyzed by ANOVA and Bonferroni’s multiple comparison test and are shown as mean ± SEM. Asterisks indicate significance, *P < 0.05, n = 10 per genotype (GTT); n = 7 (WT) & 9 (E47-/-) (PTT). c qRT-PCR of gluconeogenic genes in Dex-treated livers, normalized to U36b4; data are mean ± SEM *P < 0.05, Student’s t test. n = 6 (WT) & 5–6 (E47-/-). d Liver sections stained with Oil Red O and hematoxylin and eosin (H&E) after Cort or vehicle treatment. ×20 magnification; black scale bar: 100 µm, representative image from n = 3. e Liver and plasma triglycerides on Cort. Data are mean ± SEM, n = 6 (WT) and 4–6 (E47-/-). f Volcano plot and GO analysis of deregulated genes (blue = downregulated; red = upregulated, FC 1.3, P < 0.05) in livers of E47-/- mice on Dex, n = 2 (WT) & 3 (E47-/-). For GO analysis a base mean cutoff > 200 was used. g qRT-PCR of metabolic genes on Dex, normalized to U36B4. Data are mean ± SEM, *P < 0.05, Student’s t test, n = 5–7 (WT) & 4–7 (E47-/-). h Volcano plot and GO analysis of deregulated genes (blue = downregulated; red = upregulated, FC 1.3, P < 0.05) in livers of E47-/- mice on Cort, n = 4 per genotype. For GO analysis a base mean cutoff > 200 was used. i qRT-PCR of differentially expressed genes on Cort, normalized to U36b4; data are mean ± SEM *P < 0.05, Student’s t test. n = 4 per genotype

Fig 3: The ß-catenin-TCF3 axis repressed ATF3 expression in GC cells.a RT-qPCR analysis of TCF3 mRNA expression in our GC sample set. b, c Correlation between the mRNA levels of ß-catenin and TCF3 in the TCGA STAD specimens and the GEO GSE66229 dataset. d, e Correlations between the mRNA levels of TCF3 and ATF3 in the TCGA STAD specimens (d) and the GEO GSE66229 dataset (e). f Correlation between the mRNA levels of ß-catenin and TCF3 in our sample set. g, h Correlations between the mRNA levels of ATF3 and ß-catenin (g) or TCF3 (h) in our sample set. i, j qPCR and western blot analysis of TCF3 mRNA (i) and protein expression (j-i, -ii), respectively, in ß-catenin-knockdown BGC-803 and MKN-45 cells. k The ChIP assay was performed to detect ß-catenin occupancy in the TCF3 promoter region, and the IgG group was used as a negative control. l TCF3 promoter activity was measured by luciferase reporter assays in HEK293 cells cotransfected with ß-catenin or empty vector. Data are presented as the mean ± s.d. n = 3. m, n RT-qPCR and western blot analysis of ATF3 mRNA (m) and protein expression (n-i, -ii), respectively, in TCF3-knockdown MGC-803 and MKN-45 cells. o ChIP assays were performed to detect TCF3 occupancy in the ATF3 promoter region, and the IgG group was used as a negative control. p ATF3 promoter activity was measured by luciferase reporter assays in HEK293 cells cotransfected with TCF3 or empty vector. The data are presented as the mean ± s.d. q The protein expression levels of TCF3 and ATF3 were detected in consecutive paraffin sections of tissues from the GC TMA G10 database by IHC staining. r Correlation between the IHC scores of ATF3 and TCF3 in the GC TMA G10 specimens as detected by IHC staining. *p < 0.05, **p < 0.01 and ***p < 0.001 by Student’s t test compared with the paired normal or control group.

Fig 4: E47 is required for efficient chromatin binding of GR, FoxO1 and Mediator. a ChIP-MS for GR in Dex-injected control livers. Pale blue: chromatin remodelers (GO Biol. Process); green: NR coregulators; dark blue: sequence specific DNA binding transcription factor activity (GO Mol. Function); dark grey: significant outliers derived by Fisher’s exact test (FDR < 0.05, s0 = 1).; n = 3 biological replicates. b ChIP-MS for GR in Dex-injected E47?LKO livers. Pale blue: chromatin remodelers (GO Biol. Process); green: NR coregulators; dark blue: sequence specific DNA binding transcription factor activity (GO Mol. Function); dark grey: significant outliers derived by Fisher’s exact test (FDR < 0.05, s0 = 1).; n = 3 biological replicates. c ChIP-qPCR in Dex-treated E47-/- and wild-type livers shows binding of GR, Med1 and FoxO1 at metabolic promoters and enhancers. The data are shown as fold enrichment over IgG and normalized to a negative control region. Data are mean ± SEM, *P < 0.05, **P < 0.01 Student’s t test, n = 6, ns not significant

Fig 5: Genomic binding of GR and E2A overlaps in mouse liver. a Bioinformatic motif analyses of the hepatic GR cistrome detect enriched transcription factor consensus sites in GR ChIP sequences. b Area-proportional Venn diagram depicting the overlap between the GR and E2A liver cistromes. Numbers of called peaks in at least one sample are shown (n = 2 biol. replicates). c Analogous motif analyses of the hepatic E2A ChIP sequences. d Functional annotation of shared GR-E2A target genes by GO analysis, assigning common ChIP peaks to the nearest coding gene. e Representative ChIP-Seq tracks displaying co-occupancy of GR and E2A at metabolic promoters and enhancers. Y-axes show normalized read counts