Abstract
Circadian disruption accelerates malignant growth and shortens survival, both in experimental tumor models and cancer patients. In previous experiments, tumor circadian disruption was rescued with seliciclib, an inhibitor of cyclin-dependent kinases (CDKs). This effect occurred at a selective dosing time and was associated with improved antitumor activity. In the current study, seliciclib altered robust circadian mRNA expression of the clock genes Rev-erbα, Per2, and Bmal1 in mouse liver following dosing at zeitgeber time (ZT) 3 (i.e., 3 h after the onset of the 12 h light span), when mice start to rest, but not at ZT19, near the middle of the 12 h dark span, when mice are most active. However, liver exposure to seliciclib, as estimated by the liver area under the concentration × time curve (AUC), was ∼80% higher at ZT19 than at ZT3 (p = 0.049). Circadian clock disruption was associated with increased serum liver enzymes and modified glycogen distribution in hepatocytes, as revealed by biochemical determinations and optic and electronic microscopy. The extent of increase in liver enzymes was most pronounced following dosing at ZT3, as compared to ZT19 (p < 0.04). Seliciclib further up-regulated the transcriptional activity of c-Myc, a cell cycle gene that promotes cell cycle entry and G1-S transition (p < 0.001), and down-regulated that of Wee1, which gates cell cycle transition from G2 to M (p < 0.001). These effects did not depend upon drug dosing time. Overall, the results suggest the circadian time of seliciclib delivery is more critical than the amount of drug exposure in determining its effects on the circadian clock. Seliciclib-induced disruption of the liver molecular clock could account for liver toxicity through the resulting disruption of clock-controlled detoxification pathways. Modifications of cell cycle gene expression in the liver likely involve other mechanisms. Circadian clocks represent relevant targets to consider for optimization of therapeutic schedules of CDK inhibitors. (Author correspondence: [email protected])
ACKNOWLEDGMENTS
We thank Alper Okyar for his precious help in the interpretation of pharmacokinetics results and Gérard Pierron for his guidance in the electron microscopy studies. We also thank Christophe Desterke for his advice in molecular biology experiments and Virginie Hossard for excellent technical assistance.
This article was supported by the European Union through contracts LSHG-CT-2006-037543 (Data processing in chronobiological studies, Temporal genomics for tailored chronotherapeutics, TEMPO) and LSHB-CT-2004-005137 (Biosimulation: a new tool for drug development, BIOSIM); the Association Internationale pour la Recherche sur le Temps Biologique et la Chronothérapie (ARTBC International), hospital Paul Brousse, Villejuif, France; and Universita' “G.D'Annunzio”, Chieti, Italy; as well as by the Swedish Children Cancer Society and the Swedish Cancer Foundation.