Fig. 3

Rhythmic expression of KLF9 in normal MCF10A cells is abolished in MDA-MB-231 TNBC line. A–C MCF10A and D–F MDA-MB-231 cells were pulsed with 1 μM CORT for 2 h to synchronize circadian gene expression prior to collection of RNA every 4 h. A Circadian expression of KLF9 (blue) was antiphase with the expression of BMAL1 (ARNTL, black), as assayed through RT-qPCR. B, C A cosine-wave curve was fitted to the expression data of B BMAL1 and C KLF9 to determine the period, phase shift, and goodness-of-fit r² of the oscillation. D No apparent phasic relationship exists between BMAL1 and KLF9 expression in MDA-MB-231 cells. E Cosine-wave regression determined oscillation parameters of BMAL1, F whereas the KLF9 transcript levels remained relatively unchanged, and no period was detected. G, H MCF10A and I, J MDA-MB-231 cells stably overexpressing CLOCK were transiently transfected with a transient BMAL1 expression vector prior to analysis of gene expression after 24 h. Forced expression of CLOCK and BMAL1 induced the expression of G, I KLF9 and H, J direct CLOCK/BMAL1 target PER1 in both cell lines (Student’s t-test; P < 0.05). K MCF10A cells were transiently transfected with CLOCK and BMAL1 expression vectors along with the eKSM luciferase reporter and Renilla luciferase normalization control. Ectopic expression of CLOCK and BMAL1 induced eKSM luciferase activity relative to the pGL4.23 empty vector negative control lines (Student’s t-test; P < 0.001). Dashed lines indicate confidence bands demarcating the likely location of the true curve at 95% confidence level. Bars represent mean ± SEM with asterisks above the mean indicating significant differences. All experiments were performed with N ≥ 3 replicates