Physiology & Behavior 110–111 (2013) 42–50

Contents lists available at SciVerse ScienceDirect

Physiology & Behavior

j ourna l homepage: www.e lsev ie r .com/ locate /phb

The effects of combining serotonin reuptake inhibition and 5-HT7 receptor blockadeon circadian rhythm regulation in rodents

Ligia Westrich a,⁎, Jeffrey Sprouse b, Connie Sánchez a

a Lundbeck Research USA, 215 College Avenue, Paramus, NJ 07652, USAb Sprouse Consulting, LLC, 513 East 82nd Street, New York, NY 10028, USA

H I G H L I G H T S

► Inhibition of serotonin transporter and 5-HT7 receptor lengthens circadian period.► Serotonin transporter and 5-HT7 receptor blockade delay phase of circadian activity.► 5-HT7 receptor activation modulates circadian period length.

⁎ Corresponding author at: Fairleigh Dickinson UniversPark Avenue, Florham Park, NJ 09732, USA. Tel.: +1 201

E-mail address: westrich@fdu.edu (L. Westrich).

0031-9384/$ – see front matter © 2012 Elsevier Inc. Allhttp://dx.doi.org/10.1016/j.physbeh.2012.12.009

a b s t r a c t

a r t i c l e i n f o

Article history:Received 18 May 2012Received in revised form 1 August 2012Accepted 20 December 2012Available online 28 December 2012

Keywords:Serotonin receptorSerotonin transporterCircadian rhythmsSuprachiasmatic nucleusPeriodPER2::LUCIFERASE

Disruption of circadian rhythms may lead to mood disorders. The present study investigated the potential ther-apeutic utility of combining a 5-HT7 antagonist with a selective serotonin (5-HT) reuptake inhibitor (SSRI), thestandard of care in depression, on circadian rhythm regulation. In tissue explants of the suprachiasmatic nucleus(SCN) from PER2::LUC mice genetically modified to report changes in the expression of a key clock protein, theperiod length of PER2 bioluminescence was shortened in the presence of AS19, a 5-HT7 partial agonist. Thisreduction was blocked by SB269970, a selective 5-HT7 antagonist. The SSRI, escitalopram, had no effect aloneon period length, but a combination with SB269970, yielded significant increases. Dosed in vivo, escitalopramhad little impact on the occurrence of activity onsets in rats given access to running wheels, whether the drugwas given acutely or sub-chronically. However, preceding the escitalopram treatment with a single acute doseof SB269970 produced robust phase delays, in keeping with the in vitro explant data. Taken together, these find-ings suggest that the combination of an SSRI and a 5-HT7 receptor antagonist has a greater impact on circadianrhythms than that observedwith either agent alone, and that such amultimodal approachmay be of therapeuticvalue in treating patients with poor clock function.

© 2012 Elsevier Inc. All rights reserved.

1. Introduction

Virtually all organisms exhibit circadian rhythms. Highly conservedprocesses are in place to ensure that the timingof daily internal rhythmsmatches those of the solar day with the primary clock mechanism inmammals originating in the suprachiasmatic nucleus (SCN). So-calledretinorecipient neurons receive direct retinal input and dictate thetiming of light to an adjacent population of pacemaker cells [1,2].Other oscillators located within the brain, as well as those distributedwidely in the periphery, follow this lead in a highly orchestratedmanner[3,4]. The capacity to oscillate is the result of a sequence of transcription-al/translational feedback loops within the oscillator cells with key pro-teins (PER2, among others) cycling with a period length of 24 h [5–7].Non-photic input can also serve as zeitgebers or “time-givers” to oscilla-tor function and the SCN also receives input from a number of midbrain

ity, School of Pharmacy, 230694 5752.

rights reserved.

and forebrain projections [8,9]. Of these, the serotonergic (5-HT) systemis of particular interest in view of its dense innervation of the SCN [10].

While a number of 5-HT receptors have been implicated in SCN func-tion, the 5 HT7 subtype has received a great deal of attention. Evidencefor the presence of 5-HT7 receptors in the SCN is clear from anatomicaland functional studies [11–16], but the precise nature of their actionsis not fully understood [17]. Most experimental work suggests thatwithin the SCN 5-HT7 receptors have a primary role of limiting light-induced changes in the phase of rest:activity rhythms [15,18]. Theirfunction in terms of daily adaptation to the prevailing light:dark cycleis more obscure, although 5-HT7 receptor activation is known to leadto advances in phase [16,19]. What is perhaps clearer is the potentialvalue of the 5-HT7 receptor as a target for novel therapeutic approachesto mental disorders [20]. The concept of 5-HT7 receptor antagonists asantidepressants, antipsychotics, sleep aids and cognitive enhancers hasreceived substantial experimental support [21–24] and, in fact, a 5-HT7 antagonist/SSRI combination has been shown to enhance the in-crease in cognitive ability observed with the antagonist alone [25]. Tothe extent that 5-HT7 receptor-mediated modulation of circadian

43L. Westrich et al. / Physiology & Behavior 110–111 (2013) 42–50

rhythm function underlies such clinical utility, a better understanding ofits action within the SCN might benefit the field.

Disruptions in circadian rhythm regulation are thought to lead toincreased risk of cardiovascular disease, cancer, diabetes and obesity,among other health-related consequences [26–28], and there is a grow-ing link to mood disorders [29]. Perception of mood follows a circadianpattern in healthy individuals and loss of synchrony may lead to mal-adaptive changes [30]. Unipolar and bipolar depression have beenmost commonly linked to the circadian clock and a new generation ofantidepressants appears poised to take advantage of this connection[31–33]. Perhaps more speculatively, the question arises as to the roleof traditional pharmacotherapies— the selective serotonin reuptake in-hibitors (SSRIs) and the serotonin norepinephrine reuptake inhibitors(SNRIs)— in modulating circadian rhythm function. Given their currentwidespread use, there is a growing need to add to the handful of studiesthat have examined these agents alone or in combinationwith selectivemodulators that might fine-tune changes in net 5-HT neurotransmis-sion [34–38].

5-HT7 receptors represent an interesting target in such multimodalcombinations. Selective antagonists of this receptor subtype appear tobe robustly active in behavioral measures of antidepressant-like re-sponses when administered with an SSRI [39–41]. A combined 5-HT7receptor antagonist/SSRI effect on cortical 5-HT release may underliethese observations, although the nature of this interaction in otherbrain regions is unknown. A goal of the present investigation was togain a better understanding of the respective roles of 5-HT7 receptorsand SSRIs in SCN function. To this end, tissue explants frommice genet-ically modified to provide a bioluminescent measure of mPER2 cyclingwere used as a measure of oscillator function, as determining the pat-tern of clock gene expression has proven to be a useful means to mon-itor changes in circadian period length and amplitude [9,42]. Wheelrunning activity was used as a means to assess circadian behavior atthe whole animal level.

2. Materials and methods

2.1. Animals

Three to six-month-old male homozygous mPER2::LUCIFERASE(PER2::LUC) knock-in mice [43] purchased from Jackson Laboratory,Bar Harbor, ME were housed in a 12 h:12 h light:dark (12:12 LD)cycle for at least 2 weeks prior to dissection and tissue collection.Two-month-old Sprague–Dawley rats purchased from Charles RiverLaboratories (Wilmington, MA) were housed individually with free ac-cess to running wheels (model 80820, Lafayette Instruments, Lafayette,IA) for the duration of the experiment. All animals were maintainedunder controlled environmental conditions (22±2 °C; lights on at06:00 to 18:00 h) with food and water available ad libitum. All animalprotocols were approved by the Lundbeck Research USA, Inc. Institu-tional Animal Care and Use Committee.

2.2. Drug treatments

For the in vitro studies, AS19, a reported 5-HT7 receptor partial ago-nist [7,39], and SB269970, a 5-HT7 receptor antagonist [44] (Tocris Bio-science, Ellisville, MO) were dissolved in DMSO and sterile water,respectively; the SSRI escitalopram (H. LundbeckA/S, DK)was dissolvedin sterile water. At the time of tissue treatment, drugs were diluted intothe culture media to final concentrations. For the acute and subchronicin vivo treatments, drugs were dissolved as follows: AS19 in 20%β-cyclodextrin (Sigma-Aldrich, St. Louis, MO), and SB267790 andescitalopram in saline or 5% β-cyclodextrin. Prior to administration, pHwas adjusted toward pH 5–6 with 0.1 N NaOH; more physiological pHvalues caused compounds to precipitate.

Drugs and vehicles were administered subcutaneously in a volumeof 2 ml/kg body weight to investigate the acute effects of serotonergic

modulation, or via osmotic minipumps (ALZET Osmotic Pumps, Cuper-tino, CA) to investigate phasic effects. Once daily dosing was timed tooccur at either CT6 (Circadian Time 6) or ZT11 (Zeitgeber Time 11).Prior to the dose, animals were switched from the prevailing 12:12 LDcycle to a period of constant darkness beginning at either the time oflights-on (ZT0, or 6 h before the “subjective” midday treatment atCT6) or switched at the time of lights-off (ZT12 or shortly after theZT11 treatment during LD). This method was employed to minimizelight cues and thus allow the expression of endogenous rhythms. Ani-mals were kept in constant darkness for approximately 10 days follow-ing acute drug administration; 14 days of constant darkness wereallowed after implantation of osmotic minipumps.

Dosing of escitalopram at CT6 was chosen as earlier studies hadshown this time to be sensitive to serotonergicmodulation [14,45]; dos-ing at ZT11 was chosen to mimic the early morning dose commonlyused by patients and to be less disruptive to the normal sleep–wakecycle of the rodents.

2.3. Tissue preparation

Tissue explants containing the SCN from PER2::LUC knock-in micewere prepared according to published methods [43]. Briefly, animalswere anesthetized with isoflurane, brains were rapidly removed and300 μm coronal sections through the SCN were cut on a vibratome(Campden Instruments, Lafayette, IN) in chilledHank's balanced salt so-lution (Invitrogen, Carlsbad, CA). Sections were trimmed by hand toinclude the SCN and aminimum of surrounding tissue and immediatelytransferred to culture inserts (Millipore, Billerica, MA) in 35 mmdishes(BD Falcon, Franklin Lakes, NJ). Dishes contained 1.2 ml Dulbecco'smodified Eagle's medium (Sigma-Aldrich, St. Louis, MO) supplementedwith 1× B27 (Invitrogen), 4 mM L-glutamine (Invitrogen), 25 mMglucose (Sigma), 4.2 mM NaHCO3 (Sigma), 10 mM HEPES (Sigma),25 units/ml penicillin-G sodium (Invitrogen), 34 μM streptomycin sul-fate (Invitrogen), and 100 μM beetle luciferin (Promega, Madison, WI,USA) and were sealed with vacuum grease (Dow, Midland, MI) andglass coverslip. Two SCN tissue sectionswere collected froma single an-imal; multiple samples of the same tissue from a single animal wereplaced in different treatment groups.

2.4. Bioluminescence monitoring

SCN tissue cultures were maintained at 35.8 °C, and integrated bio-luminescence was collected for 60 s every 10 min with a commercialluminometer (LumiCycle, Actimetrics, Wilmette, IL) before and aftertreatments. Bioluminescence analysis was performed using LumiCycleAnalysis software. Data from the dissection day (Day 0) and from thedays with media changes or day 1 of treatment were excluded fromthe analysis. Bioluminescence traces were de-trended by baselinesubtraction of a 24-hour moving average and then smoothed with a2-hour running average. De-trended and smoothed data were analyzedfor periodicity with a χ2 periodogram (Lumicycle Analysis software;Actimetrics Inc., Evanston, IL). Chi-square tests for the period of Per2-luc expression were considered significant at pb0.001. The period ofthe PER2::LUC expression rhythms was determined with a sine-fitcurve [30];mean values obtained in the control (vehicle) groups rangedfrom 24.94±0.10 to 25.15±0.14 (Figs. 1 and 2), in keeping with nu-merous published reports (e.g. [19,21,28]). Aminimumof 3 days of bio-luminescent data was included in the analysis. Period values werecompared with one-way analysis of variance (ANOVA) followed byNewman–Keuls multiple comparison test. All comparisons were con-sidered significant at pb0.05.

2.5. Locomotor activity rhythm monitoring

Dailywheel-running activity ofmale Sprague–Dawley ratswasmon-itored with Running Wheel Activity Software (AWM V12.0, Lafayette

BA

0 0.01 0.1 1 1023

24

25

26

* **

AS19 (μM)

Per

iod

len

gth

(h

)

23

24

25

26

*

Per

iod

len

gth

(h

)

0 10 30 100 30023

24

25

26

SB269970 (μM)

Per

iod

len

gth

(h

)

AS19 (0.01 μM); τ=24.50 h AS19 (0.01 μM) + SB269970 (100 μM); τ=25.44 h

Bio

lum

ines

cenc

e(d

etre

nded

, cps

)

Time (days)

DC

Fig. 1. 5-HT7 receptor-mediated shortening of the period length of PER2 bioluminescence in tissue explants of suprachiasmatic nucleus (SCN) frommPER2::LUC transgenicmalemice. (A) Con-centration–response curve for AS19, a 5-HT7 receptor partial agonist on period length (τ) showing a significant shortening at the two lowest concentrations tested (mean±SEM; ANOVA, [F(4,50)=5.448, p=0.0010], **pb0.001; *pb0.05 vs. control); (B) concentration–response for SB269970, a 5-HT7 receptor antagonist, revealing no significant changes in τ (mean±SEM; ANOVA,[F(4, 34)=1.278, p=0.2979]); (C) period shortening by AS19 (0.01 μM) and its blockade by SB269970 (100 μM) (mean±SEM; ANOVA, [F(3, 28)=4.324, p=0.0126], *pb0.01 vs. control,b0.05 vs. SB269970 alone); (D) representative records of the daily cycling of mPER2 bioluminescence in the presence of AS19 (0.01 μM; τ=24.50 h; black line) and the lengthening of τ inthe presence of the combination of AS19 (0.01 μM) and SB269970 (100 μM; τ=25.44 h; gray line).

23

24

25

26

27

*

Per

iod

len

gth

(h

ou

rs)

Escitalopram (60 μM); τ=24.00 h

Escitalopram (60 μM) + SB269970 (100 μM); τ=25.50 h

Bio

lum

ines

cenc

e(d

etre

nded

, cps

)

Time (days)

BA

Fig. 2. Combined dosing of a selective serotonin reuptake inhibitor (SSRI) and a 5-HT7 receptor antagonist on the period length ofmPER2 bioluminescence in tissue explants of suprachiasmaticnucleus (SCN) frommPER2::LUC. (A) Effect of escitalopram (60 μM) or SB269970 (100 μM) alone, revealing no significant change in τ but a significant lengthening following dosing with thecombination (mean±SEM; ANOVA, [F(3, 31)=4.978, p=0.0062], *pb0.05 vs. control and escitalopram alone). (B) Daily cycling of PER2 bioluminescence in the presence of escitalopram(60 μM; τ=24.00 h; black line) and the lengthening of τ in the presence of the combination of escitalopram (60 μM) and SB269970 (100 μM; τ=25.50 h; gray line).

44 L. Westrich et al. / Physiology & Behavior 110–111 (2013) 42–50

45L. Westrich et al. / Physiology & Behavior 110–111 (2013) 42–50

Instruments, Lafayette, IN) via activity wheel counters (Model 86061)interfaced with computers. Total revolutions of the activity wheel werecontinuously recorded in 2-min epochs. Collected data were split usingAWM software to obtain interval count data only for further analysisby Clocklab software (Actimetrics). Cumulative activity was representedin actograms fromwhich activity onsets and associated circadian param-eters were calculated. Least-squares lines were fitted to the onsets ofactivity before and after dosing. The difference between the intersec-tions of the two lines on the first day after treatment was used to calcu-late the shift in phase (by convention a minus symbol (−) was used toindicate delays and a positive symbol (+) to indicate advances). Statis-tical analyses of the phase changes comprised an ANOVA followed by aNewman–Keuls multiple comparison test. pb0.05 values were consid-ered significant.

2.6. Functional assay for the rat 5HT7 receptor

The functional potency of SB269970 at the rat 5-HT7 receptorwas de-termined using awhole cell cAMPHiRangeHTRF assay (Cisbio Bioassays,Bedford,MA, two step protocol, 62AM6PEB). Briefly, polyclonal HEK-293cell lines expressing the rat 5-HT7 receptor were re-suspended in stimu-lation buffer (1× Hank's balanced salt solution with CaCl2 and MgCl2,5 mM HEPES pH 7.4, 0.075% BSA) and counted on a Guava EasyCyte(Millipore, Billerica, MA). In each well of a 384-well Optiplate, the fol-lowing were added: 5 μl of cells (30,000 cells), 2.5 μl drug, 2.5 μl 5-HT,5 μl of cAMP-D2 conjugate and 5 μl cryptate conjugate as recommendedby the kit manufacturer. Reconstituted anti-cAMP-cryptate and cAMP-D2 lyophilized reagents were used at 1/5 dilutions using the lysis bufferprovided in the kit. Plates were incubated with SB269970 (8× stock) for30 min (final concentration of 1 μM). The plateswere then incubated for30 min with 5-HT (at EC80 corresponding to 60 nM final) at 37 °C, andan additional 10 min at room temperature. Following addition of theconjugates, the plates were incubated for 1 h at room temperature be-fore the cell cAMP contentwasmeasuredusingEnVision 2104MultilabelReader (Perkin Elmer, Waltham, MA). Results were calculated from theratio of absorbance at 665 nm/620 nm and expressed in Delta F. Calcu-lated Delta F {[(standard or sample ratio−ratio of negative control)/ratio of negative control]×100}was used by interpolation from the stan-dard curve to calculate cAMP concentrations. All data points were nor-malized and analyzed by sigmoidal dose response curve-fitting usingGraph Pad Prism 4 to determine IC50. The Cheng–Prusoff equation wasused to calculate Kb: Kb=IC50/[1+(A/EC50)], where A=concentrationof reference agonist that is being inhibited and EC50=relative EC50 of thereference agonist.

2.7. Bioanalysis

One hour following acute dosing of escitalopram and SB269970brainswere rapidly removed and frozen, while trunk bloodwas collect-ed in EDTA-containing tubes and spun down to obtain plasma. Frozenbrains were weighed and homogenized in 3× (weight/volume) of ho-mogenization buffer consisting of 50% water, 30% 2-propanol and 20%DMSO. A 150 μl internal standard solution was added to 50 μl homoge-nized brain/plasma sample. The samples were mixed and centrifuged.The supernatant was injected directly into the ThermoFinnigam Quan-tum Ultra LC/MS/MS (liquid chromatography–tandemmass spectrom-etry) system for analysis.

3. Results

3.1. Combined effects of 5-HT7 receptor blockade and serotonin reuptakeinhibition on circadian rhythm regulation in SCN tissue explants frommPER2::LUC mice

Incubation of SCN tissue explantswith AS19, a selective 5-HT7 recep-tor partial agonist, produced a shortening of the period (τ) of mPER2

bioluminescence by approximately 1 h at the two lower concentrationstested (0.01 and 0.1 μM;pb0.05). In contrast, the two higher concentra-tions (1 and 10 μM)yielded no significant change in τ (Fig. 1A). In a sep-arate series of experiments, SB269970, a selective 5-HT7 receptorantagonist, had no effect on τ within a 30-fold range of concentrations(10–300 μM) (Fig. 1B). When dosed in combination with an active con-centration of AS19 (0.01 μM), SB269970 (100 μM) reduced the effectsof the agonist on period shortening (Fig. 1C and D), since the value ofτ with the combination was not significantly different from control ex-plants. For these experiments a high concentration of SB269970was se-lected to ensure complete blockade of 5-HT7 receptors in thepresence ofAS19, a compound with a much higher affinity for this 5-HT receptorsubtype (Ki~30 nM vs. 0.6–4.6 nM; [44,46,47]). Dampening of the bio-luminescence amplitude over time followingAS19 or SB269970was notapparent from visual examination of the records.

Incubation of SCN explants with escitalopram, a highly selectiveserotonin reuptake inhibitor, yielded no change in PER2 expressionwhen dosed at a high concentration (60 μM) relative to its potency ininhibiting 5-HT transport in vitro (IC50=2.1 nM; [7]) (Fig. 2). Whenescitalopram (60 μM) was co-incubated with SB269970 (100 μM),however, therewas a lengthening of τ by approximately 1 h, a value sig-nificantly greater inmagnitude thanwith either agent alone (Fig. 2A andB; pb0.05).

3.2. Combined effects of 5-HT7 receptor blockade and serotonin reuptakeinhibition on circadian rhythm regulation in freely behaving rats withaccess to running wheels

Plotting of wheel running behavior into daily actograms revealedevidence of entrainment of the normal rest:activity rhythms to the pre-vailing light:dark cycle (12:12 LD) within 2 weeks of exposure to thewheels. Shortly thereafter, a single dose of escitalopram (2.5 mg/kgs.c.) was given at ZT11 and the animals were released into constantdarkness (DD) at the time of lights-off on the day of dosing (Fig. 3A).The switch to DD, in this instance, was incorporated into the paradigmso as to limit the masking effects of light that normally limit locomotorbehavior. Extrapolation of the activity onsets duringDDback to the timeof escitalopram dosing yielded no change in phase (ΔΦ=−0.59±0.12h; p>0.05). Similarly timed dosing with SB269970 (10 mg/kg s.c.)alone also produced no change in the occurrence of activity onsets(Fig. 3C).

When the two agents were combined in paired injections at ZT11, asignificant phase delay emerged, in keepingwith the period lengtheningobserved with the combination in vitro (Fig. 3B and C). The magnitudeof this delay with the dosed combination was significantly differentfrom that observed with vehicle control or either of the active agentsalone (pb0.05). Dosing timed to occur at CT6 (or circadian time 6, dur-ing DD, but 6 h after lights-on would have normally occurred) yieldedvery similar results as the ZT11 dosing (Fig. 3D).

Subchronic dosing with escitalopram in DD (10 mg/kg/day s.c. viaosmotic minipumps for 14 days) yielded no significant change inwheel running onsets (Fig. 4C). A single dose of SB269970 (30 mg/kgs.c.), timed at ZT11 at the beginning of subchronic escitalopram dosing,produced a significant delay in activity onset (pb0.05; Fig. 4B and C). Nochanges in the occurrence of phase were observed with the SB269970dosing that preceded subchronic dosing of vehicle in DD (Fig. 4A andC). Although free-running periods were not examined in all animalsprior to minipump implantation, in those experiments for which re-cords are available (e.g., Fig. 4A and B), there appeared to be a lengthen-ing of period due to the escitalopram/SB269970 combination.

3.3. Exposures of escitalopram and SB269970 achieved with in vivo dosingto rats

A satellite group of rats dosed with escitalopram (2.5 mg/kg s.c.) tomimic the acute experiments on rest:activity rhythms achieved high

A

ΔΔ Φ = -1.50h

Escitalopramand SB267790

B

Δ Φ = -0.25h

Escitalopram

Constant darkness

-2.0

-1.5

-1.0

-0.5

0.0

0.5

-2.0

-1.5

-1.0

-0.5

0.0

0.5

*

Ph

ase

chan

ge

(ho

urs

)

C

*Ph

ase

chan

ge(

ho

urs

)

D summarized data at CT6summarized data at ZT11

Fig. 3. Combined dosing of a selective serotonin reuptake inhibitor (SSRI) and a 5-HT7 receptor antagonist on daily running wheel behavior as revealed in double-plotted actograms. (A) Rep-resentative actogram showing the effect of a single acute injection of escitalopram (2.5 mg/kg s.c.) at ZT11 (zeitgeber time 11 or 11 h after lights-on or 1 h before lights-off in a 12:12 light:dark(LD) cycle) following entrainment to the LD cycle as calculated by extrapolation of activity onsets in constant darkness (DD) back to the time of dosing (change in phase (ΔΦ)=−0.25 h); (B)effect of combined doses of escitalopram (2.5 mg/kg s.c.) and SB269970 (10 mg/kg s.c.) at ZT11, yielding amore substantial phase change in activity onsets (ΔΦ=−1.50 h); (C) summarizeddatawith all injections occurring at ZT11 and revealing a significantly greater phase delay following combineddosingwith escitalopramand SB269970 (mean±SEM;ANOVA, [F(3, 21)=7.797,p=0.0011]; *pb0.001 vs. vehicle;b0.05 vs. all other treatment groups); (D) summarized datawith all injections occurring at CT6 (circadian time6, duringDDbut 6 hafter lights-onwouldhavenormally occurred) (mean±SEM; ANOVA, F(3, 12)=3.950, p=0.0358; *pb0.05 vs. vehicle and SB269970 alone).

46 L. Westrich et al. / Physiology & Behavior 110–111 (2013) 42–50

brain levels of drug 1 h post dose (Fig. 5A). Transporter occupancies atthis exposure were also high, exceeding 90% based on displacement ofradiolabeled compound, [3H]DASB (Alan Pehrson, personal communi-cation); occupancies following sub-chronic dosing of escitalopram viaminipumps (10 mg/kg/day s.c.) were equally robust (data not shown).

Brain exposures 1 h following acute dosing with SB269970 (1–30 mg/kg s.c.) were low relative to that observed in plasma (Fig. 5B).Such levelsmay be sufficient to achieve significant 5-HT7 receptor occu-pancy, however, based on the potency of SB269970 in reducing agonist-elicited increases in cAMP production in HEK293 cells (IC50=2.7 nMor0.95 ng/ml; Ki=0.3 nM or 0.11 ng/ml; Fig. 5C).

4. Discussion

Serotonergic modulation of circadian rhythms has been extensivelystudied in a number of experimental paradigms. In vitro approacheshave ranged from monitoring the electrical activity of cultured SCNslices to bioluminescence measures in PER::LUC explanted tissues (asin the present work); in vivo approaches have included work in rats,mice and hamsters, measuring both the effects on period length andon light-induced shifts in behavioral onsets. Taken as a whole, the re-sults have been mixed with some research groups providing ample ev-idence of rhythm regulation (e.g. [35,37]) and others, surprisingly little(e.g. [48,49]). Studies that have focused more closely on the effects ofantidepressants (citalopram, paroxetine, venlafaxine, fluvoxamine, flu-oxetine, sertraline) [35,37,38,50] constitute only a small subset of thetotal. A persistent question that underlies much of this work can be

simply stated: Why would the SCN, which receives dense serotonergicinnervation involving a host of 5-HT receptor subtypes, be so resistantto the changes in tone afforded by the selective reuptake inhibitors?As the results of the present study suggest, the answer may lie in themix of receptors activated indirectly by reuptake blockade.

Monitoring the oscillation in clock gene expression, as enabled bythe SCN explants from the mPER2::LUC mice, is a simple means todetect changes in various circadian parameters [9,42]. As shown inFig. 1, the effect of the 5-HT7 receptor partial agonist AS19 on the periodlength of mPER2 oscillation was concentration-dependent at lowerdoses (b0.1 μM), yielding a shortening of period, with higher doses(>0.1 μM) having little apparent effect. A shortening was expected —

earlier work monitoring the effect of 8-OH-DPAT, a mixed 5-HT1A/7 re-ceptor agonist, on the firing rate of cells within cultured SCN slicesshowed phase advances, an effect sensitive to 5-HT7 receptor antago-nists [51]. The reasons for the loss of a period-shortening effect whenAS19was given at higher concentrations are not known, but may be re-lated to its level of intrinsic efficacy at the 5-HT7 receptor or its selectiv-ity relative to other 5-HT receptor subtypes. As to the former, AS19 isreported to be a 5-HT7 receptor partial agonist (77% intrinsic efficacy;[46,47]) and such a profile might add to the existing degree of 5-HTtone at low concentrations of the partial agonist, but detract from it asexposure rises and its low intrinsic activity prevails. Alternatively, thereported affinity of AS19 for 5-HT1A receptors (Ki~75 nM vs. 4.6 nM;[39]) might mitigate its action at 5-HT7 receptors, as the two subtypeshave opposing effects on intracellular signaling cascades [16,52]. Re-gardless of the mechanisms that shape the concentration–response

BA

-1.5

-1.0

-0.5

0.0

0.5

*Ph

ase

chan

ge

(ho

urs

)

Saline

SB269970

Saline Pumpx 14 days

Constant darkness

ΔΔ Φ = -0.45h

Saline

SB267790

EscitalopramPump x 14 days

Δ Φ = -0.98h

C

Fig. 4. Subchronic dosing of an SSRI combinedwith a single acute dose of a 5-HT7 receptor antagonist on daily runningwheel behavior as revealed in double-plotted actograms. (A) Represen-tative actogram showing the effect of a single acute dose of SB269970 at ZT11 followed by subchronic vehicle dosing via minipump during constant darkness (DD) and revealing no apparentchange in the phase of activity onsets (ΔΦ=−0.45 h); (B) representative actogram showing the effect of a single acute dose of SB269970 at ZT11 followed by subchronic escitalopram dosingvia minipump (10 mg/kg/day s.c.) during DD and revealing amore substantial phase delay (ΔΦ=−0.98 h); (C) summarized data showing a significantly greater phase delay following com-bination dosing (mean±SEM; ANOVA [F(3, 11)=5.438, p=0.0154]; *pb0.05 vs. vehicle and all other treatment groups). Note that in the two actograms shown (A & B) an earlier period ofconstant darkness (DD) enabled an assessment of free-running period length both prior to and after drug treatment (although thiswas undertaken in too few animals for statistical analyses). Ineach case, the period of DD was preceded by a dose of saline at CT6 as part of pilot work for later studies. The summary data (C) are derived only from the periods of active drug treatments.

47L. Westrich et al. / Physiology & Behavior 110–111 (2013) 42–50

curve, the effect of AS19 at the lower concentrations (0.01 μM) is likelymediated by 5-HT7 receptors as the decrease in period length was re-duced by SB269970, a 5-HT7 receptor antagonist. Confirmation of thiseffect at a lower concentration of SB269970would strengthen this argu-ment as would blockade by a second, structurally unrelated, 5-HT7antagonist.

Having thus defined a 5-HT7 receptor-active dose of SB269970 in thein vitro explant paradigm, it is interesting to note that blockade of thisreceptor subtype does not, by itself, elicit a change in period length,find-ings similar to those reported previously [53]. One possible explanationfor this lack of an effect again relates to 5-HT tone in the explant culture,presumably too low for an antagonist to yield a sufficient change in neu-rotransmission. When tone is increased by the presence of an SSRI, thecombination leads to a significant increase in period length, the oppositeof that observed with AS19. In this sense, the present findings are con-sistent with reports of 5-HT7-receptor-mediated phase advances [16]and suggest that an antagonist may have effects as well if tone is suffi-ciently high. Conversely, the lack of an effect by an antagonist mightbe a reasonable sign that serotonin tone is endogenously (or exogenous-ly) low. It is surprising that escitalopramwhen given alone was ineffec-tive in modulating period length as previous studies had shown aperiod-shortening effect for SSRIs in tissue culture [37]. The differencesreported here might suggest a distinguishing property of escitalopramas the earlier work had chosen to focus primarily on sertraline. Alterna-tively, the differences in experimental protocols (primarily rat-1 fibro-blasts as in the case of [25] vs. SCN explants in the present study;

mPER1 expression in [37] vs. mPER2 in the present study) might havehad their impact.

Such observations in the in vitro explant system also appear to holdwhen roughly equivalent studies are conducted in vivo. Acute doses ofSB269970 (10 mg/kg s.c.) given at ZT11 yielded no significant changein activity onsets as the animals entered 10 days of constant darkness(Fig. 3A). Escitalopram also produced no discernable change in thephase of locomotor rhythms at the dose tested (2.5 mg/kg s.c.). Dosingat ZT11 was chosen to mimic the morning dose commonly used bypatients and to be less disruptive to the normal sleep–wake cycle of noc-turnal animals. That a phase change could have been observed experi-mentally is confirmed by earlier work by our group employing 8-OH-DPAT [54]. In contrast to single dosing, the net effect of combiningescitalopram and SB269970 amounted to a phase delay larger in magni-tude than any other treatment group, essentially recapitulating the invitro explant results and generalizing such results across (mouse andrat) species. Dosing the combination at CT6, a time noted by others toproduce a robust effect on phase with 5-HT agonists [45,55] producednearly identical results (Fig. 3D) as did sub-chronic inhibition of the5-HT transporter combined with acute blockade of the 5-HT7 receptor(Fig. 4). Based on these observations, the effect of phasic and tonic sero-tonergic modulation appeared to be much the same. The modest effectof chronic SSRI dosing on period length as reported by others [14] wasnot observed, possibly due to differences in exposure. Decreases in5-HT7 receptor sensitivity that might occur early on as a result of contin-uous transporter inhibition [56] did not appear to be sufficient to alter

A

0

100

200

300

400

500

0

1000

2000

3000

4000

5000

1 10 30SB269970 (mg/kg)

Bra

in E

xpo

sure

(n

g/g

)

Plasm

a Co

ncn

etration

(ng

/ml)

Brain

Plasma

B

2.5 2.5Escitalopram(mg/kg)

0

1000

2000

3000

4000

5000

Bra

in E

xpo

sure

(n

g/g

)

0

250

500

750

1000 Plasm

a Exp

osu

re (ng

/ml)Brain Plasma

-12 -11 -10 -9 -8 -7 -6 -5 -40

25

50

75

100

Log[SB269970], M

% c

AM

P r

esp

on

se in

5-H

T7R

-HE

K29

3 ce

llsC

Fig. 5. Brain and plasma exposures of drugs following in vivo dosing to satellite groups of rats. (A) Levels of escitalopram in brain and plasma 1 h following a single acute dose (2.5 mg/kg s.c.;mean±SEM, N=8); (B) levels of SB269970 in brain and plasma 1 h following single acute doses (1–30 mg/kg s.c.; mean±SEM, N=4–8); (C) potency of SB269970 in blockingagonist-stimulated increases in cAMP production in HEK293 cells (N=3).

48 L. Westrich et al. / Physiology & Behavior 110–111 (2013) 42–50

activity onsets. In other words, the decrement in 5-HT7 receptor-specificneurotransmission that might follow desensitization of this subtype wasmodest compared to that achieved with pharmacologic antagonism.

Given the high exposures achieved and the evidence of near com-plete transporter/receptor occupancy, it is tempting to infer synergy(as opposed to additivity) in interpreting our results. The precise mech-anism that would account for synergy between 5-HT transporter inhibi-tion and 5-HT7 receptor blockade is not known.At best, one cannote thatthe net effect of an increase in 5-HT tone afforded by an SSRI alone ismodest in terms of circadian rhythm regulation and only becomes im-portant when the 5-HT7 receptor subtype is no longer functional, thatis, when tone is increased on all of the 5-HT receptor systems with theexception of 5-HT7. Enhanced increases in net 5-HT release have beenreported for citalopram when it is combined with SB269970 [39], butthe exact nature of this interaction is not known, nor is its potential im-portance within the SCN. In fact, previous attempts to impact the sup-pression of light-induced phase advances by citalopram failed to find arole for SB269970 or 5-HT7 antagonism [36].

The range of 5-HT receptor subtypes that might be activated in theperi-SCN by an SSRI is impressive — 5-HT1A, 5-HT1B, 5-HT2C, 5-HT3,5-HT5A, 5-HT7 — and this list essentially expands to include all of thesubtypes when inputs to and outputs from the SCN are factored in[14,17,51,55,57,58]. There is also the possibility that the site of actionmay be some distance from the SCN. For example, the 5-HT-containingneurons of the raphe nuclei provide substantial input to the SCN and ac-tivation/blockade of 5-HT7 cell body autoreceptors can impact non-photic phase-resetting responses [13,59,60]. In either case, the 5-HT7 re-ceptor would appear to oppose the actions of the other 5-HT receptorsubtypes, such that limiting its functionwith an antagonist would disin-hibit some or all of the remaining subtypes. Situated postsynaptically oncell bodies, the 5-HT7 receptor subtype is unique in being positivelycoupled to cAMP production and to that extent may oppose, for exam-ple, 5-HT1A receptors in terms of intracellular second messenger cas-cades as noted earlier [16]. An interesting follow-up study might be to

incorporate a 5-HT1A receptor antagonist in combination with AS19 todetermine whether the resulting loss of 5-HT1A receptor function addsto (or synergizes with) the effect size observed with the 5-HT7 receptoragonist alone.

The potential clinical advantages of a combined SSRI plus 5-HT7receptor antagonist are interesting to consider. Circadian rhythmdistur-bances in patients that lead to poor synchrony and an altered moodstate would naturally represent a primary target. The prevailing theorypurports that an abnormal phase relationship between the circadiansystemand sleep could be involved in the pathogenesis and in themain-tenance of depression such that a delay in circadian rhythms (or an ad-vance in the sleep–wake cycle) might effectively treat mood symptoms[33,61]. In fact, sleep phase advancement has been shown to alleviatedepressive mood in the majority of patients after 2–3 weeks of therapy[62]. Theoretically, a delay in the circadian phase to achieve properalignment (that did not appreciably alter sleep onset) would alsoimprove patient symptoms. An important caveat with respect to thepresent findings is the artificial nature of the LD cycle employed:where-as the effects on phase were examined under the constant darkness inthe animal studies, patients obviously experience normal LD cycles.

The actions of SSRIs in this subset of patients would likely be minimalin light of the present studies butmight rise in importance in combinationwith a 5-HT7 receptor antagonist. In terms of reported clinical findings,the atypical antipsychotics risperidone and olanzapine have beenshown to improve the antidepressant potential of SSRIs in unipolar andbipolar depressed patients [63–67] with possible mechanisms relatingto their actions at a variety of receptors, most often the 5-HT2A receptorsubtype. 5-HT7 receptor binding is also typical of this class of agents[23,68] and efficacy related to this aspect of their overall pharmacologymight also reveal important synergies in patients. On this last note, it is in-teresting that olanzapine addition to SSRI therapy in treatment-resistantpatients has been shown to improve sleep continuity along with depres-sive symptoms [69], an action thatmight be closely linked to SCN functiongiven its role in sleep regulation [26,31]. Certainly, many receptor

49L. Westrich et al. / Physiology & Behavior 110–111 (2013) 42–50

subtypes may account for the effects observed with compounds such asolanzapine that involve multiple and complex pharmacologies, but thepresent findings suggest a new and potentially important role for 5-HT7receptors in therapies designed to improve mood.

In conclusion, the combination of an SSRI and a 5-HT7 receptorantagonist has the potential to produce an effect on circadian rhythmregulation greater in magnitude than that observed with either agentalone. Such a multimodal approach, by targeting many but not all 5-HT receptor subtypes for an improvement in neurotransmission, mayyield changes in clock function that serve to restore proper rhythms inthose patients suffering from poor synchrony.

References

[1] Antle MC, Silver R. Orchestrating time: arrangements of the brain circadian clock.Trends Neurosci 2005;28:145–51.

[2] Moore RY. Organization of the primate circadian system. J Biol Rhythms 1993;8:S3–9[Suppl.].

[3] Cuninkova L, Brown SA. Peripheral circadian oscillators: interesting mechanismsand powerful tools. Ann N Y Acad Sci 2008;1129:358–70.

[4] Stratmann M, Schibler U. Properties, entrainment, and physiological functions ofmammalian peripheral oscillators. J Biol Rhythms 2006;21:494–506.

[5] Hastings M, O'Neill JS, Maywood ES. Circadian clocks: regulators of endocrine andmetabolic rhythms. J Endocrinol 2007;195:187–98.

[6] Reppert SM,Weaver DR. Molecular analysis of mammalian circadian rhythms. AnnuRev Physiol 2001;63:647–76.

[7] Roenneberg T, Chua EJ, Bernardo R, Mendoza E. Modelling biological rhythms. CurrBiol 2008;18:R826–35.

[8] Mrosovsky N. Locomotor activity and non-photic influences on circadian clocks. BiolRev Camb Philos Soc 1996;71:343–72.

[9] Weinert D. Ontogenetic development of themammalian circadian system. ChronobiolInt 2005;22:179–205.

[10] Morin LP. Serotonin and the regulation of mammalian circadian rhythmicity. AnnMed 1999;31:12–33.

[11] Duncan MJ, Franklin KM. Expression of 5-HT7 receptor mRNA in the hamsterbrain: effect of aging and association with calbindin-D28K expression. Brain Res2007;1143:70–7.

[12] Duncan MJ, Short J, Wheeler DL. Comparison of the effects of aging on 5-HT7 and5-HT1A receptors in discrete regions of the circadian timing system in hamsters.Brain Res 1999;829:39–45.

[13] Glass JD, Grossman GH, Farnbauch L, DiNardo L. Midbrain raphe modulation ofnonphotic circadian clock resetting and 5-HT release in the mammaliansuprachiasmatic nucleus. J Neurosci 2003;23:7451–60.

[14] Pickard GE, Smith BN, Belenky M, Rea MA, Dudek FE, Sollars PJ. 5-HT1Breceptor-mediated presynaptic inhibition of retinal input to the suprachiasmaticnucleus. J Neurosci 1999;19:4034–45.

[15] Smith BN, Sollars PJ, Dudek FE, PickardGE. Serotonergicmodulation of retinal input tothe mouse suprachiasmatic nucleus mediated by 5-HT1B and 5-HT7 receptors. J BiolRhythms 2001;16:25–38.

[16] Sprouse J, Reynolds L, Li X, Braselton J, Schmidt A. 8-OH-DPAT as a 5-HT7 agonist:phase shifts of the circadian biological clock through increases in cAMP production.Neuropharmacology 2004;46:52–62.

[17] Lovenberg TW, Baron BM, de Lecea L, Miller JD, Prosser RA, Rea MA, et al. A noveladenylyl cyclase-activating serotonin receptor (5-HT7) implicated in the regulationof mammalian circadian rhythms. Neuron 1993;11:449–58.

[18] Ying SW, Rusak B. 5-HT7 receptors mediate serotonergic effects on light-sensitivesuprachiasmatic nucleus neurons. Brain Res 1997;755:246–54.

[19] Ehlen JC, Grossman GH, Glass JD. In vivo resetting of the hamster circadian clockby 5-HT7 receptors in the suprachiasmatic nucleus. J Neurosci 2001;21:5351–7.

[20] Leopoldo M, Lacivita E, Berardi F, Perrone R, Hedlund PB. Serotonin 5-HT7 receptoragents: structure–activity relationships and potential therapeutic applications incentral nervous system disorders. Pharmacol Ther 2011;129:120–48.

[21] Mullins UL, Gianutsos G, Eison AS. Effects of antidepressants on 5-HT7 receptorregulation in the rat hypothalamus. Neuropsychopharmacology 1999;21:352–67.

[22] Roberts AJ, Krucker T, Levy CL, Slanina KA, Sutcliffe JG, Hedlund PB. Mice lacking5-HT receptors show specific impairments in contextual learning. Eur J Neurosci2004;19:1913–22.

[23] Roth BL, Craigo SC, Choudhary MS, Uluer A, Monsma Jr FJ, Shen Y, et al. Binding oftypical and atypical antipsychotic agents to 5-hydroxytryptamine-6 and5-hydroxytryptamine-7 receptors. J Pharmacol Exp Ther 1994;268:1403–10.

[24] Thomas DR, Melotto S, Massagrande M, Gribble AD, Jeffrey P, Stevens AJ, et al.SB-656104-A, a novel selective 5-HT7 receptor antagonist, modulates REM sleep inrats. Br J Pharmacol 2003;139:705–14.

[25] Nikiforuk A. Selective blockade of 5-HT7 receptors facilitates attentional set-shiftingin stressed and control rats. Behav Brain Res 2012;226:118–23.

[26] Huang W, Ramsey KM, Marcheva B, Bass J. Circadian rhythms, sleep, and metabo-lism. J Clin Invest 2011;121:2133–41.

[27] Takeda N, Maemura K. Circadian clock and cardiovascular disease. J Cardiol 2011;57:249–56.

[28] Yu EA, Weaver DR. Disrupting the circadian clock: gene-specific effects on aging,cancer, and other phenotypes. Aging (Albany NY) 2011;3:479–93.

[29] Kronfeld-SchorN, Einat H. Circadian rhythms and depression: human psychopathol-ogy and animal models. Neuropharmacology 2011;62:101–14.

[30] Boivin DB. Influence of sleep–wake and circadian rhythmdisturbances in psychiatricdisorders. J Psychiatry Neurosci 2000;25:446–58.

[31] Harvey AG. Sleep and circadian functioning: critical mechanisms in the mood disor-ders? Annu Rev Clin Psychol 2011;7:297–319.

[32] Hickie IB, Rogers NL. Novel melatonin-based therapies: potential advances in thetreatment of major depression. Lancet 2011;378:621–31.

[33] Westrich L, Sprouse J. Circadian rhythm dysregulation in bipolar disorder. Curr OpinInvestig Drugs 2010;11:779–87.

[34] Duncan MJ, Hester JM, Hopper JA, Franklin KM. The effects of aging and chronicfluoxetine treatment on circadian rhythms and suprachiasmatic nucleus expressionof neuropeptide genes and 5-HT1B receptors. Eur J Neurosci 2010;31:1646–54.

[35] GannonRL,MillanMJ. Evaluation of serotonin, noradrenaline anddopamine reuptakeinhibitors on light-induced phase advances in hamster circadian activity rhythms.Psychopharmacology (Berl) 2007;195:325–32.

[36] Gannon RL, Peglion JL, Millan MJ. Differential influence of selective 5-HT5A vs 5-HT1A,5-HT1B, or 5-HT2C receptorblockadeupon light-inducedphase shifts in circadian activ-ity rhythms: interaction studieswith citalopram. Eur Neuropsychopharmacol 2009;19:887–97.

[37] Nomura K, Castanon-Cervantes O, Davidson A, Fukuhara C. Selective serotoninreuptake inhibitors and raft inhibitors shorten the period of Period1-driven circadianbioluminescence rhythms in rat-1 fibroblasts. Life Sci 2008;82:1169–74.

[38] Possidente B, Lumia AR, McEldowney S, Rapp M. Fluoxetine shortens circadian peri-od for wheel running activity in mice. Brain Res Bull 1992;28:629–31.

[39] Bonaventure P, Kelly L, Aluisio L, Shelton J, Lord B, Galici R, et al. Selective block-ade of 5-hydroxytryptamine (5-HT)7 receptors enhances 5-HT transmission,antidepressant-like behavior, and rapid eye movement sleep suppression inducedby citalopram in rodents. J Pharmacol Exp Ther 2007;321:690–8.

[40] Hedlund PB, Huitron-Resendiz S, Henriksen SJ, Sutcliffe JG. 5-HT7 receptor inhibitionand inactivation induce antidepressantlike behavior and sleep pattern. Biol Psychia-try 2005;58:831–7.

[41] Mnie-Filali O, Faure C, Lambas-Senas L, El Mansari M, Belblidia H, Gondard E, et al.Pharmacological blockade of 5-HT7 receptors as a putative fast acting antidepressantstrategy. Neuropsychopharmacology 2011;36:1275–88.

[42] Yamazaki S, Takahashi JS. Real-time luminescence reporting of circadian gene ex-pression in mammals. Methods Enzymol 2005;393:288–301.

[43] Yoo SH, Yamazaki S, Lowrey PL, Shimomura K, Ko CH, Buhr ED, et al. PERIOD2::LUCIFERASE real-time reporting of circadian dynamics reveals persistent circadi-an oscillations in mouse peripheral tissues. Proc Natl Acad Sci U S A 2004;101:5339–46.

[44] Hagan JJ, Price GW, Jeffrey P, Deeks NJ, Stean T, Piper D, et al. Characterization ofSB-269970-A, a selective 5-HT(7) receptor antagonist. Br J Pharmacol 2000;130:539–48.

[45] Edgar DM, Miller JD, Prosser RA, Dean RR, Dement WC. Serotonin and the mamma-lian circadian system: II. Phase-shifting rat behavioral rhythmswith serotonergic ag-onists. J Biol Rhythms 1993;8:17–31.

[46] Bosker FJ, Folgering JH, Gladkevich AV, Schmidt A, van der Hart MC, Sprouse J, et al.Antagonism of 5-HT(1A) receptors uncovers an excitatory effect of SSRIs on 5-HTneuronal activity, an action probably mediated by 5-HT(7) receptors. J Neurochem2009;108:1126–35.

[47] Vela Hernandez JM, Torrens-Jover A, Buschmann H, Romero-Alonzo L. Use of 5-HT7receptor agonists for the treatment of pain. International application publishedunder the patent cooperation treaty (PCT) WO 2008/000495 A1 2008.

[48] Antle MC, Ogilvie MD, Pickard GE, Mistlberger RE. Response of the mouse circadiansystem to serotonin 1A/2/7 agonists in vivo: surprisingly little. J Biol Rhythms2003;18:145–58.

[49] Klemfuss H, Kripke DF. Antidepressant and depressogenic drugs lack consistenteffects on hamster circadian rhythms. Psychiatry Res 1994;53:173–84.

[50] Sprouse J, Braselton J, Reynolds L. Fluoxetinemodulates the circadian biological clockvia phase advances of suprachiasmatic nucleus neuronal firing. Biol Psychiatry2006;60:896–9.

[51] Sprouse J, Reynolds L, Braselton J, Schmidt A. Serotonin-induced phase advances ofSCN neuronal firing in vitro: a possible role for 5-HT5A receptors? Synapse 2004;54:111–8.

[52] Hoyer D, Hannon JP, Martin GR. Molecular, pharmacological and functional diversityof 5-HT receptors. Pharmacol Biochem Behav 2002;71:533–54.

[53] Gannon RL. 5HT7 receptors in the rodent suprachiasmatic nucleus. J Biol Rhythms2001;16:19–24.

[54] Westrich L, Bisulco S, Sprouse J. Modulation of GSK-3β kinase phosphorylationstate as a determinant of circadian behavior: effects of lithium and 8-OH-DPAT.Program No. 869.7. Neuroscience Meeting Planner. Chicago, IL: Society for Neuro-science; 2009 [Online].

[55] Prosser RA,DeanRR, EdgarDM,HellerHC,Miller JD. Serotonin and themammalian cir-cadian system: I. In vitro phase shifts by serotonergic agonists and antagonists. J BiolRhythms 1993;8:1–16.

[56] Sleight AJ, Carolo C, Petit N, Zwingelstein C, Bourson A. Identification of5-hydroxytryptamine7 receptor binding sites in rat hypothalamus: sensitivity tochronic antidepressant treatment. Mol Pharmacol 1995;47:99–103.

[57] Graff C, Challet E, Pevet P,Wollnik F. 5-HT3 receptor-mediated photic-like responsesof the circadian clock in the rat. Neuropharmacology 2007;52:662–71.

[58] Roca AL, Weaver DR, Reppert SM. Serotonin receptor gene expression in the ratsuprachiasmatic nuclei. Brain Res 1993;608:159–65.

[59] Duncan MJ, Congleton MR. Neural mechanisms mediating circadian phase reset-ting by activation of 5-HT(7) receptors in the dorsal raphe: roles of GABAergicand glutamatergic neurotransmission. Brain Res 2010;1366:110–9.

50 L. Westrich et al. / Physiology & Behavior 110–111 (2013) 42–50

[60] DuncanMJ, Grear KE, Hoskins MA. Aging and SB-269970-A, a selective 5-HT7 receptorantagonist, attenuate circadian phase advances induced by microinjections of seroto-nergic drugs in the hamster dorsal raphe nucleus. Brain Res 2004;1008:40–8.

[61] Dallaspezia S, Benedetti F. Chronobiological therapy for mood disorders. Expert RevNeurother 2011;11:961–70.

[62] Wehr TA, Wirz-Justice A, Goodwin FK, Duncan W, Gillin JC. Phase advance of thecircadian sleep–wake cycle as an antidepressant. Science 1979;206:710–3.

[63] Croxtall JD, Scott LJ. Olanzapine/fluoxetine: a review of its use in patients withtreatment-resistant major depressive disorder. CNS Drugs 2010;24:245–62.

[64] Deeks ED, Keating GM. Spotlight on olanzapine/fluoxetine in acute bipolar depres-sion. CNS Drugs 2008;22:793–5.

[65] Hirose S, Ashby Jr CR. An open pilot study combining risperidone and a selectiveserotonin reuptake inhibitor as initial antidepressant therapy. J Clin Psychiatry2002;63:733–6.

[66] Ostroff RB, Nelson JC. Risperidone augmentation of selective serotonin reuptake in-hibitors in major depression. J Clin Psychiatry 1999;60:256–9.

[67] Shelton RC, Stahl SM. Risperidone and paroxetine given singly and in combinationfor bipolar depression. J Clin Psychiatry 2004;65:1715–9.

[68] Rauly-Lestienne I, Boutet-Robinet E, AilhaudMC, Newman-Tancredi A, Cussac D. Dif-ferential profile of typical, atypical and third generation antipsychotics at human5-HT7a receptors coupled to adenylyl cyclase: detection of agonist and inverse ago-nist properties. Naunyn Schmiedebergs Arch Pharmacol 2007;376:93–105.

[69] Sharpley AL, AttenburrowME, Hafizi S, Cowen PJ. Olanzapine increases slow wavesleep and sleep continuity in SSRI-resistant depressed patients. J Clin Psychiatry2005;66:450–4.