Circadian oscillations alter drug absorption, distribution, metabolism, and excretion (ADME) as well as intracellular signaling systems, target molecules (e.g., receptors, transporters, and enzymes), and gene transcription. There is a positive influence of drug dosing-time on the efficacy of depression therapy. On the other hand, antidepressants have also demonstrated to modulate circadian rhythmicity and sleep–wake cycles.
Chronopharmacological studies with antidepressants are often performed in vivo. Light is a strong zeitgeber in rodents, which display an active phase during the night. For this reason, studies are usually performed under a 12h00 light–12h00 dark cycle, during which drugs are administered at different zeitgeber times (ZT). The ZT0 corresponds to the moment that lights are turned on and ZT12 when lights are turned off . In contrast, chronotherapy in humans needs to consider zeitgeber factors other than light, namely mealtime, oxygen levels, temperature, and exercise . Pharmacokinetic alterations at different drug dosing times are associated with intrinsic circadian rhythms in several tissues that are also involved in physiological ADME (Figure 1).
Since antidepressants are orally administered, their bioavailability depends not only on their physicochemical properties, such as lipophilicity, but also on physiological processes of the digestive system modulated by circadian rhythms that may be disrupted in depressed patients .
Although the mechanism of action of trimipramine is not fully characterized, it is known to be a TCA that differs from others, particularly because it does not directly inhibit reuptake transporters, namely the 5-HT transporter (SERT) or NE reuptake transporter (NET) . Chronopharmacokinetic studies in humans treated with two oral formulations of the trimipramine (solution and tablet) showed that tablets seem to be less affected by dosing-time (Table 1) . Importantly, drug absorption remains faster after solution administration than tablets administration. Solutions demonstrated significantly higher maximum concentration (Cmax), lower time to reach Cmax (tmax), and lower mean residence time (MRT) in the morning (under fasting conditions) than at night (after a meal).
|Antidepressant or Active Metabolite||Subjects||Study Design||Daily Dose (mg)||Duration (Days)||Formulation||Time of Administration||Plasma Pharmacokinetic Parameters||Ref.|
|tmax (h)||Cmax (mg/L)||AUC (mg.h/L)||t1/2β
|ka (h−1)||MRT (h)|
|Amitriptyline||10 healthy subjects (♂), 22–31 years old.||Crossover||50||21||Injectable solution||9h00||3.2 *||96.1||1270||15.7||-||0.36 *||-|||
|21h00||4.4 *||72.8||1224||17.2||-||0.25 *||-|
|Nortriptyline||10 healthy subjects (♂), 22–30 years old.||Crossover||100||14||Oral formulation: 25 mg capsules||9h00||6.2||32||730||15.0||-||-||-|||
|Trimipramine||12 healthy subjects (6 ♀, 6 ♂), 22–37 years old.||Crossover||100||15||Oral formulation: 100 mg tablet||8h00||2.5||37.8||362||10.9||-||-||10.8|||
|Oral formulation: solution||8h00||1.5 *||48.2 *||372||9.9||-||-||9.8 *|
|20h00||2.5 *||28.8 *||322||11.1||-||-||11.8 *|
|Sertraline||10 healthy subjects (♂), 18–45 years old.||Crossover||100||1||Oral formulation: 100 mg tablet||Morning||7.0||24.5||0.664||20.0||0.0347||-||-|||
Intestinal permeability shows daily rhythms due to variations of expression of tight junction proteins that regulate the epithelial paracellular pathway . In the small intestine, mRNA levels of occludin and claudin-3 were higher during the late dark phase than the late light phase and inversely associated with paracellular permeability data .
From another point of view, it is important to bear in mind that several antidepressant drugs are known as substrates of efflux transmembrane ATP-binding cassette (ABC) proteins, including P-glycoprotein (P-gp; ABCB1; MDR1)  and Breast Cancer Resistant Protein (BCRP; ABCG2) . These transporters are expressed in several tissues, namely the intestine, kidney, liver, and blood–brain barrier (BBB). They reduce the bioavailability, facilitate the elimination, and hamper the access of compounds to the brain, including antidepressant drugs . The expression of P-gp in the intestine is modulated by proline- and acid-rich basic leucine zipper (PAR bZIP) proteins, particularly hepatic leukemia factor (HLF), whose expression is regulated by core oscillator components . The HLF and E4 promoter binding protein-4 (E4BP4), a putative antagonist of PAR bZIP proteins, respectively, increase and decrease the mRNA levels and expression of P-gp . In the mouse intestine, Mdr1a mRNA levels exhibit a significant 24 h daily variation, increasing during the light phase with a peak at ZT12, when the lights go off . Similarly, in the rat jejunum, P-gp mRNA varies 5.4-fold with the circadian time . Consistent with the daily rhythmicity observed in total protein with a peak level at ZT8, the P-gp function in the intestine is significantly higher at ZT12 than ZT0 . Additionally, feeding patterns and gender also influence the expression and activities of the ABC transporters during the day, since circadian amplitudes of mRNA and protein levels of P-gp in the ileum are larger in female mice than in male mice .
The cardiovascular system is susceptible to circadian rhythmicity since blood pressure, heart rate, and plasma protein levels display daily oscillations . The blood pressure of mammals has a 24 h cycle, with a peak during their active phase and a 10–20% slope at rest phase, along with a 17% daily fluctuation of total plasma proteins . Particularly for antidepressant drugs that are highly protein-bound, distribution is strongly affected by the presence of plasma proteins, mainly albumin and α1-acid glycoprotein . A study performed in rats demonstrated that plasma protein levels are higher in the active phase . Furthermore, in healthy humans, α1-acid glycoprotein plasma levels are also increased during the active phase . Hence, these protein fluctuations may be important to adjust the dosing-time of highly protein-bound antidepressants since differences of unbound fraction and tissue exposure can affect drug efficacy and toxicity . A single intragastric administration of amitriptyline in rats (64 mg/kg) at six different time-points revealed highest bioavailability at the end of the active phase (ZT22) . In fact, the total drug exposure was higher in the beginning of the rest phase, even though the ka after intragastric administration was unaltered during the light–dark cycle . Exposure in liver and lung, given by area under the concentration–time curve (AUC), showed a 24 h oscillation (peak at ZT4). Moreover, multi-dosing for 10 days demonstrated significantly higher AUC values in liver, lung, and kidney tissues in the light than in the dark phase . These results may be linked to the reduction of plasma protein levels during the light phase, since amitriptyline is a highly protein-bound drug (95%) .
The therapeutic targets of antidepressant drugs are located in the central nervous system (CNS), whose access is limited by blood-CNS barriers, including the BBB and the blood-cerebrospinal fluid (CSF) barrier. Circadian rhythms have an important role in BBB homeostasis and integrity, since the deletion of the clock component BMAL1 leads to BBB hyperpermeability . Additionally, a strong circadian gene expression of Per2 in the choroid plexus is responsible for the adjustment of the SCN clock and brain homeostasis through the CSF .
One of the main causes behind antidepressant inefficacy is the overexpression of P-gp and BCRP by the endothelial cells of BBB of depressed patients  and the efflux of xenobiotics across the BBB of mammals is known to be regulated by circadian rhythms . Nonetheless, there is not much data concerning diurnal variations in the transcript and protein levels of efflux transporters in the brain and the studies performed until today are conflicting. Pulido et al. observed an inverse association between P-gp activity and the active phase of wild-type mice . The authors reported a diurnal oscillation of P-gp transcription (peak at ZT12) in the brain, which increased during the light phase and decreased during the dark phase. It was modulated by PAR bZip transcription factors, specifically albumin D box-binding protein (DBP), thyrotroph embryonic factor (TEF), and HLF . Accordingly, the accumulation of a P-gp substrate (rhodamine-123) in the brain was higher in the beginning of the light phase . In opposition, Zhang et al. did not find circadian alterations of mRNA or protein levels of P-gp in the BBB of mice or humans . However, its activity was modulated by clock genes. In mice, higher efflux activity was observed during the active phase, but transcription levels and protein expression were unaffected. The magnesium transporter transient receptor potential cation channel, subfamily M, member 7 (TRPM7) contributed to efflux rhythms in human cultured brain endothelial cells . The BMAL1 modulates transcript and protein expression of TRPM7 during active periods, which enhances intracellular magnesium levels and promotes efflux activity. It was observed that BBB permeability in Drosophila is regulated by 24 h cyclically expressed gap junctions .
Liver enzyme activity and hepatic blood flow are two time-dependent processes that drive hepatic drug metabolism . Lipophilic drugs, such as antidepressants, are metabolized in the liver into more hydrophilic polar metabolites in three phases . Phase I functionalization reactions are mostly performed by the cytochrome P450 (CYP) enzyme superfamily, specifically the isoforms CYP1A2, CYP2C9, CYP2C19, CYP2D6, and CYP3A4 . After phase II conjugation reactions, metabolites may be excreted into the bile or transported into the systemic circulation by efflux transporters (phase III). The PAR bZip and an atypical nuclear receptor expressed abundantly in the liver, named as small heterodimer partner (SHP), play a pivotal role in CYP activity 24 h oscillations in mice, controlling daily variations of detoxification and drug metabolism . In addition, in mice, Zhang et al. described that the mRNA levels of phase I enzymes increase in the dark phase, while phase II enzymes rise in the light phase . Analyzing serum-shocked cells of a human hepatocellular carcinoma cell line (HepG2), a 24 h oscillation in the expression of CYP3A4 and CYP2D6 was revealed . Transcript levels seem to be considerably more affected than protein levels when circadian rhythms are disrupted. The rhythmicity of liver proteins can be explained by rhythmic mRNAs, and translational and post-translational regulation and feeding behavior .
Metabolic oscillations appear to have a weaker role in the chronopharmacokinetics of antidepressants. For instance, the elimination half-life time (t1/2β) of sertraline did not change after administration to humans in the morning or evening .
Different results of chronopharmacodynamic studies with antidepressants can be related with fluctuations of the pharmacokinetic parameters or daily variations of the expression of antidepressant drug targets affected by circadian rhythms (Figure 2). Distinct mechanisms of action of antidepressant drugs may lead to different chronopharmacological profiles, and dosing-time can influence both therapeutic and toxic effects.
Before chronopharmacological studies, it is important to ensure the entrainment of circadian rhythms in healthy mice to a standard light/dark cycle. Phenotyping circadian rhythms in mice requires measurements of their activity in free running conditions , quantifying circadian clock or clock-related genes , or monitoring sleep behavior .
The most common behavioral tests in rodents to evaluate antidepressant effects are the forced swimming test (FST) and the tail suspension test (TST). Both present high predictive reliability and validity but different sensitivity .
Specifically, NET inhibition increases climbing, whereas inhibition of SERT selectively increases swimming . However, for a reliable comparison with clinical data, it is advisable to investigate therapeutic effects following chronic treatment in rodents . It is important to refer that reduction of immobility is interpreted as an antidepressant effect, if it does not increase general locomotor activity, which could be interpreted as a false positive result . This is related with the fact that antidepressants and psychostimulants, respectively, reduce and increase the locomotor activity of rodents in new environments . Moreover, behavioral experiments at different times of the day strongly affect the obtained results . The FST experiments display different results if performed during the dark or light phase, since rodents are more active during the dark phase . Kelliher et al. noticed that rats were more agitated or worsened when taken from the swim apparatus during light phase .
Seasonal variations seem to influence the efficacy of antidepressants. In rats, FST revealed significant seasonal fluctuations of the antidepressant effect for TCAs, namely amineptine, amitriptyline, desipramine, imipramine, and mianserin . Maximal reduction of immobility in FST was observed in March for all aforementioned antidepressants. The desipramine and mianserin anti-immobility effect was also evaluated in a light–dark cycle and was not circadian-dependent (Table 2) . Nomifensine, a NE-dopamine reuptake inhibitor, was equally effective throughout the year with no seasonal variations, but it revealed a circadian-dependent effect, more pronounced during the light phase (peak at ZT7) (Table 2) . The interpretation of these results is complicated by the low number of experimental replicates (n = 1 or 2). Furthermore, the controlled conditions of rats excluded the influence off the light–dark cycle, humidity and temperature variations that occur during the seasons. Therefore, the observed seasonal effect may be linked with internal mechanisms, namely the daily amplitude of subtype receptor 5-HT1A binding in the rat brain, which is higher in March than December .
|Antidepressant||Species (Gender)||Dose (mg/Kg)||Initial of Experiment after Administration (h)||Route||Zeitgeber
Time (ZT) Administrations
|Test||24 h Rhythm Variation||Observations|
|Amitriptyline||ICR mice (male)||15||0.5||Intraperitoneal||ZT2, ZT6, ZT10, ZT14, ZT18, ZT22||FST||Yes||Lowest immobility at ZT14.||-|||
|Bupropion||C57BL/6 mice (male)||20||1||Intraperitoneal||ZT1, ZT7, ZT13, ZT19||TST||No, but significantly different between ZT||Lowest immobility at ZT1.||No significant differences between dosing times in plasma and brain.|||
|Desipramine||CD-COBS rats (male)||20||24, 5 and 1||Intraperitoneal||ZT3, ZT7, ZT11, ZT15, ZT19, ZT23||FST||No||-||-|||
|Fluoxetine||C57BL/6 mice (male)||30||1||Intraperitoneal||ZT1, ZT7, ZT13, ZT19||TST||Yes||Lowest immobility at ZT1.||No significant differences between dosing times in plasma and brain.|||
|Locomotor activity||Yes||Lowest locomotion activity at ZT1|
|Fluvoxamine||ICR mice (male)||30||0.5||Intraperitoneal||ZT2, ZT6, ZT10, ZT14, ZT18, ZT22||FST||Yes||Lowest immobility at ZT14.||ZT2 > ZT14 in plasma, significantly different after 1h of drug injection. No differences in brain.|||
|ZT2, ZT14||Locomotor activity||No||No effect|
|Imipramine||C57BL/6 mice (male)||30||1||Intraperitoneal||ZT1, ZT7, ZT13, ZT19||TST||Yes||Lowest immobility at ZT13.||No significant differences between dosing times in plasma and brain|||
|Wistar Hannover rats (male)||30||1||Intraperitoneal||ZT1, ZT13||FST||Yes||Lowest immobility and highest climbing at ZT1.||ZT1 > ZT13 for imipramine and desipramine in plasma but not significantly different|||
|10 for 2 weeks||1||Intraperitoneal||ZT1, ZT13||FST||Yes||Lowest immobility and highest climbing at ZT1.||-|
|30 for 2 weeks||1||Intraperitoneal||ZT1, ZT13||FST||No||-||-|
|Mianserin||CD-COBS rats (male)||15||24, 5 and 1||Intraperitoneal||ZT3, ZT7, ZT11, ZT15, ZT19, ZT23||FST||No||-||-|||
|Milnacipran||Wistar Hannover rats (male)||60||1||Oral||ZT1, ZT13||FST||Yes||Lowest immobility and highest swimming at ZT1.||No significant differences between dosing times in plasma and brain|||
|Nomifensine||CD-COBS rats (male)||5||24, 5 and 1||Intraperitoneal||ZT3, ZT7, ZT11, ZT15, ZT19, ZT23||FST||Yes||Lowest immobility at ZT7||-|||
|Venlafaxine||C57BL/6 mice (male)||30||1||Intraperitoneal||ZT1, ZT7, ZT13, ZT19||TST||Yes||Lowest immobility at ZT7.||No significant differences between dosing times in plasma and brain|||
|Locomotor activity||Yes||Lowest locomotion activity at ZT7.|
The effect of light–dark cycles on pharmacodynamic studies with antidepressants has also been investigated. In the modified FST, the immobility time of mice treated with SSRI fluvoxamine showed a significant 24 h rhythm, with the lowest immobility at the beginning of the dark phase (ZT14) (Figure 3) without increasing locomotor activity (Table 2) .
Dual-action antidepressants may also have different chronopharmacological profiles. In FST, the immobility time of mice treated with TCA amitriptyline exhibited a 24 h rhythm variation with a peak of antidepressant effect at ZT14 (Table 2 and Figure 3), identical to the previously mentioned results for fluvoxamine .
In general, data from the animal studies herein discussed demonstrate that antidepressants display distinct chronopharmacodynamic outcomes (Figure 3) which should be specifically evaluated. Other behavioral experiments (sucrose preference test or elevated plus maze) could be additionally performed to increase result reliability . Moreover, the use of depressive mice and the monitorization of circadian rhythms could facilitate the implementation of clinical studies in humans. Transferability of data from mice to humans regarding circadian rhythms is challenging. For instance, although mice are nocturnally active and humans are diurnally active, both secrete melatonin during the nighttime . Therefore, the translation of chronopharmacological results of antidepressants from mice to humans needs particular care.
|Antidepressant||Subjects||Study Design||Daily Dose (mg)||Duration (Days)||Time Administrations||Pharmacodynamic||Ref.|
|Test||24 h Rhythm Variation||Observations|
|Amitriptyline||10 healthy (♂) subjects. Range age: 22–31 years old.||Crossover||50||21||9h00
|Antimuscarinic action (saliva flow) and sedation effect by self-rating scales||Yes||Highest salivary flow and lowest sedative effect at 21h00|||
|Clomipramine||40 patients with MDD (15 ♀, 25 ♂). Range age: 18–65 years old.||Crossover||150||28||8h20
|HRSD and BDRS||Yes||Lowest depressive symptoms at 12h20|||
|Lofepramine||30 patients with MDD (22 ♀, 8 ♂). Range age: 25–60 years old.||Parallel||210||21||8h00
|HRSD and CSRS||Yes||Lowest depressive symptoms at 24h00|||
|Antidepressant||Pre-Clinical Studies||Clinical Studies||References|