Fig. 3

Dexamethasone (DEX) injection induces rhythmic clock gene expression in B16 tumors in vivo. a, b Clock gene expression in the lung and B16 lung tumors (a) (n = 21–22; cosine-wave regression: Bmal1: tumor: p > 0.05, lung: p < 0.01, n = 22, 3–4 mice/time point; Per1: tumor: p > 0.05, lung: p < 0.05; Per2: tumor: p < 0.05, lung: p < 0.0001; two-way ANOVA: interaction: p < 0.0001, group: p < 0.0001, Time: p < 0.0001) and subcutaneous (s.c.) tumors (b) (n = 15–16, 2–3 mice/time point, cosine-wave regression, F-test: all genes: p > 0.05). c Bioluminescence of a cultured Bmal1-Luc tumor slice treated repeatedly with DEX (grey bars). d Snapshots of Bmal1-Luc bioluminescence of single B16 cells in a lung tumor slice over 44 h in the absence of (left panel) or over 48 h after DEX treatment (right panel) from Additional file 6. e–i Clock gene expression in subcutaneous (s.c.) tumors after repeated intra-tumoral DEX or phosphate-buffered saline (PBS) injection (n = 9–10, 2–3 mice/time point; two-way ANOVA: time effect: Per2 and Cry1: p < 0.05, Per1 and Nr1d1: p < 0.01, Bmal1: p < 0.001; group effect: Per1 and Cry1: p < 0.05, Per2 and Nr1d1: p < 0.001, Bmal1: p < 0.05; posthoc test, group effect: *p < 0.0.5, **p < 0.01, ***p < 0.001). j BMAL1 protein expression in DEX- or PBS-treated tumors generated by s.c. injection (n = 11–12, 3 mice/time point, cosine-wave regression, F-test: PBS: p > 0.05, DEX: p < 0.01; two-way ANOVA: group p < 0.0001, time: p < 0.01) treated repeatedly for 8–11 days. Significant rhythms are illustrated with fitted cosine curves, otherwise data are connected by straight lines between data points, indicating no significant circadian rhythms. Data are represented as mean ± standard error of the mean. For details of statistics, see Additional file 1