Systems Chronotherapeutics

Abstract

Chronotherapeutics aim at treating illnesses according to the endogenous biologic rhythms, which moderate xenobiotic metabolism and cellular drug response. The molecular clocks present in individual cells involve approximately fifteen clock genes interconnected in regulatory feedback loops. They are coordinated by the suprachiasmatic nuclei, a hypothalamic pacemaker, which also adjusts the circadian rhythms to environmental cycles. As a result, many mechanisms of diseases and drug effects are controlled by the circadian timing system. Thus, the tolerability of nearly 500 medications varies by up to fivefold according to circadian scheduling, both in experimental models and/or patients. Moreover, treatment itself disrupted, maintained, or improved the circadian timing system as a function of drug timing. Improved patient outcomes on circadian-based treatments (chronotherapy) have been demonstrated in randomized clinical trials, especially for cancer and inflammatory diseases. However, recent technological advances have highlighted large interpatient differences in circadian functions resulting in significant variability in chronotherapy response. Such findings advocate for the advancement of personalized chronotherapeutics through interdisciplinary systems approaches. Thus, the combination of mathematical, statistical, technological, experimental, and clinical expertise is now shaping the development of dedicated devices and diagnostic and delivery algorithms enabling treatment individualization. In particular, multiscale systems chronopharmacology approaches currently combine mathematical modeling based on cellular and whole-body physiology to preclinical and clinical investigations toward the design of patient-tailored chronotherapies. We review recent systems research works aiming to the individualization of disease treatment, with emphasis on both cancer management and circadian timing system-resetting strategies for improving chronic disease control and patient outcomes.

Graphical abstract

Fig. 1.

The CTS. The CTS is composed of a central pacemaker located in the SCN that displays autonomous circadian oscillations, but is also entrained by external cues such as light or socioprofessional activities. The SCN further generate rhythmic physiologic signals exerting a control on the autonomous molecular clocks present in each nucleated cell, which, in turn, induce oscillations in the expression of a large number of genes involved in key intracellular processes.

Fig. 2.

Multiscale longitudinal assessment of the CTS. Recent technologies allow for the recording of biomarkers of the CTS at multiple scales: in Per2::luc Hepa1-6 cell culture, imposed temperature cycles (A) and Per2 bioluminescence measured by Lumicycle (B); in individual B6D2F1 male mice entrained in LD12:12, body temperature recorded by telemetry (C); and, in Per2::luc animals, bioluminescence recorded in RT-Bio (D); in individual young male healthy volunteers, skin temperature recorded though new thoracic wearable sensors (E, In Casa project), and individual Per2 mRNA level in peripheral blood mononuclear cells (F, dots are data from Teboul et al., 2005; dotted line is the best-fit cosinor model). Longitudinal measurement over several days of molecular biomarkers is currently not available in the clinics, and multiscale systems approaches aim at predicting from preclinical results and clinical investigations the patient-specific dynamical information needed for treatment personalization. Time is expressed in days. Zero represents midnight (clock hours) on the first day of experiment.

Fig. 3.

Circadian control of drug PK-PD. The CTS regulates drug transport at various levels, including intestinal absorption, intracellular uptake and efflux, and renal and intestinal excretion. Similarly, the amount of protein and metabolite binding to drugs in the plasma varies according to circadian time. Regarding PD, most systems of the organism are under the control of the CTS at the molecular, cellular, and physiologic levels. They can either be altered in specific diseases and impact on drug chronoefficacy or be involved in drug tolerability as targets of dose-limiting toxicities.

Fig. 4.

Molecular pathways of the circadian clock control on the cell cycle machinery (adapted from Gérard and Goldbeter, 2012). Several molecular processes along the cell cycle are regulated by the clock. At the early G1 phase, the BMAL1/CLOCK heterodimer downregulates Myc transcription to prevent its overexpression. In response to double-strand DNA damage, PER1 directly interacts with ataxia telangiectasia mutated and CHK2 to control G1 checkpoint. DNA damage induced by γ-radiation activates ataxia telangiectasia mutated/CHK2-mediated G1/S and G2/M checkpoints via p53 and p21. DNA damage induced by UV radiation leads to activation of ATR/CHK1-mediated intra-S checkpoint. In S phase, CRY2/TIM complex directly interacts with ATR/CHK1. In the G2 phase, PER-mediated ataxia telangiectasia mutated/CHK2/p53 signaling in response to DNA double-strand breaks leads to activation of G2/M checkpoint. BMAL1/CLOCK-activated Wee1 expression leads to activation of G2/M checkpoint.

Fig. 5.

A systems medicine approach for personalized chronotherapeutics. Individual patient data—such as measurements of circadian biomarkers, gene polymorphism, patient general characteristics, or disease history—are input into the systems medicine algorithm that computes personalized chrono-infusion schemes. The algorithm is developed through a multiscale pipeline integrating mathematical and experimental investigations in cell culture, in laboratory animals, and in patient populations. Results in multiple cell lines, animal strains, and patient subgroups allow for the reliable design of the personalization framework.

Fig. 6.

Relationship between hepatic chronomodulated delivery schedules and plasma PK profiles in metastatic cancer patients from the Optiliv trial. Irinotecan (A–C), Oxaliplatin (D–F), and 5-FU (G–H) were administered according to specified infusion patterns in the hepatic artery of 11 cancer patients. For irinotecan and oxaliplatin, a significant delay was observed between the administration peak time and the plasma PK curves of both administered agents and corresponding active metabolite or ultrafiltrate concentration. On the opposite, 5-FU plasma concentration closely followed the infusion profile for all patients. Large interpatient variability was observed in the plasma PK of all measured quantities [data from Lévi et al. (2017)].

Fig. 7.

Multiscale systems chronopharmacology to personalize irinotecan chronotherapy. An in vitro study of irinotecan chronopharmacology led to the design of a cellular chronoPK-PD model (A) incorporating multitype experimental data, including the extra- and intracellular concentrations of active metabolite SN 38 and irinotecan-induced apoptosis after irinotecan exposure at three CTs (B–D) Dulong et al., 2015). This cellular investigation provided the basis for a mouse study and the development of a whole-body model of irinotecan chronoPK-PD explicitly incorporating the cellular model in relevant organs (E) (Ballesta et al., 2012). The model was first developed for B6D2F1 male mice in which several chronopharmacology datasets were available, including plasma and colon chronoPK profiles of SN38 after irinotecan at best and worst time of tolerability (F–G). The next step will consist in fitting intestinal chronotoxicity data available for the same mouse category (H) (Li et al., 2013). Dots or bars represent experimental results, and solid lines represent best-fit models.

Fig. 8.

Optimizing irinotecan chronotherapy in Caco-2 cell culture (adapted from Ballesta et al., 2011). (A) Predicted drug cytotoxicity in synchronized cells with respect to exposure duration and circadian time of beginning of exposure. The cumulative dose was set to 500 mM/h. (B) Optimal exposure schemes following the strategy of maximizing efficacy in unsynchronized cells considered as cancer cells, under a constraint of maximal allowed toxicity in synchronized cells, considered as healthy cells. The toxicity threshold was varied (y-axis), and corresponding optimal schemes consisted in administering the optimal cumulative dose (written in green) over 3h40 to 7h10, starting between CT2 and CT3.

Fig. 9.

Main outcomes of cancer patients according to circadian functioning status, estimated through assessment of rest-activity or salivary marker rhythms. Left panel: estimated 2-year survival rate (mean + 95% confidence limit) in a total of 1077 cancer patients (Cohen et al., 2012; Innominato et al., 2012; Lévi et al., 2014; Schrepf et al., 2015; Sephton et al., 2013; Sephton et al., 2000). Right panel: global quality of life domain (mean + standard error of the mean), derived from the EORTC QLQ-C30 questionnaire, completed by a total of 237 patients with mCRC (Innominato et al., 2009b).

Fig. 10.

Available behavioral and pharmacological strategies to restore circadian rhythms in patients. The timing and regularity of bright light exposure, physical exercise, social and family life, and sleep-wake routine affect the function of the central clock, whereas the fasting-feeding schedule impacts on metabolism-linked peripheral clocks. Novel synthetic agonists or modified-release formulations of the hormones melatonin (e.g., Ramelteon, Tasimelteon, Agomelanine) and cortisol (e.g., delayed-release prednisone) can be used to target the central clock, as well as the peripheral oscillators in tissues equipped with the specific receptors. Several drugs used to treat psychiatric conditions, thus affecting neuronal function in the central nervous system, such as lithium, selective serotonin-uptake inhibitors, anxiolytic and hypnotic GABA agonists, or novel sleep inducers like the orexin-antagonist Suvorexant, can directly or indirectly modulate the function of the central clock, because SCN neurons are equipped with receptors of these drugs, or receive neuronal input from other brain areas affected by these drug classes. Finally, recent small molecules targeting core proteins of the circadian molecular clock modify their activity and thus impact circadian functions; these include REV-ERB agonist, CK1 inhibitor, SIRT1 agonist, and CRY activator.