Neurochemistry International 58 (2011) 78–84

Contents lists available at ScienceDirect

Neurochemistry International

journal homepage: www.elsevier.com/locate/neuint

Effect of co-administration of varenicline and antidepressants on extracellular monoamine concentrations in rat prefrontal cortex

Hans Rollema a,*, Gary G. Wilson b, Theodore C. Lee c, Joost H.A. Folgering d,1, Gunnar Flik d,1 a Neuroscience, Pfizer Global Research and Development, Groton, CT, USA b Worldwide Safety Strategy, Pfizer Inc, New York, NY, USA c Medical Affairs, Pfizer Inc, New York, NY, USA d Brains On-Line BV, Groningen, The Netherlands

ARTICLE INFO ABSTRACT

Article history: Since a substantial proportion of smokers have comorbid mood disorders, the cessation aid Received 27 October 2010 varenicline might occasionally be prescribed to patients who are simultaneously treated with Accepted 28 October 2010 antidepressants. Given that varenicline is a selective nicotinic receptor partial and Available online 5 November 2010 not a substrate or inhibitor of drug metabolizing enzymes, pharmacokinetic interactions with various classes of antidepressants are highly unlikely. It is, however, conceivable that varenicline may have a Keywords: pharmacodynamic effect on antidepressant-evoked increases in central monoamine release. Interac- Varenicline tions resulting in excessive transmitter release could cause adverse events such as serotonin syndrome, Antidepressants while attenuation of monoamine release could impact the clinical efficacy of antidepressants. To Monoamine release Microdialysis investigate this we examined whether varenicline administration modulates the effects of the selective Serotonin syndrome serotonin reuptake inhibitor sertraline and the monoamine oxidase inhibitor clorgyline, given alone and combined, on extracellular concentrations of the monoamines serotonin, , and in rat brain by microdialysis. Given the important role attributed to cortical monoamine release in serotonin syndrome as well as antidepressant activity, the effects on extracellular monoamine concentrations were measured in the medial prefrontal cortex. Responses to maximally effective doses of sertraline or clorgyline and of sertraline plus clorgyline were the same in the absence as in the presence of a relatively high dose of varenicline, which by itself had no significant effect on cortical monoamine release. This is consistent with the binding profile of varenicline that has insufficient affinity for receptors, enzymes, or transporters to inhibit or potentiate the pharmacologic effects of antidepressants. Since varenicline neither diminished nor potentiated sertraline- or clorgyline-induced increases in levels, combining varenicline with serotonergic antidepressants is unlikely to cause excessive serotonin release or to attenuate antidepressant efficacy via effects on cortical serotonin, dopamine or norepinephrine release. ß 2010 Elsevier Ltd. All rights reserved.

1. Introduction population. The World Health Organization predicts that by 2030 more than 8 million deaths per year will be caused by smoking- dependence and depression are illnesses that have related illnesses, while depression is one of the leading causes of serious health consequences and affect a large proportion of the disability affecting more than 120 million people worldwide (U¨ stu¨ n et al., 2004; World Health Organization, 2008). There is substantial epidemiologic evidence that smokers are more likely to Abbreviations: aCSF, artificial cerebrospinal fluid; DA, dopamine; MAO-I, mono- have mood or personality disorders, depressive symptoms, or amine oxidase inhibitor; mPFC, medial prefrontal cortex; MRM, multiple-reaction- clinical depression and that the smoking prevalence is approxi- monitoring; nAChR, nicotinic ; NE, norepinephrine; SSRI, mately 50% among psychiatric outpatients with depression selective serotonin reuptake inhibitor; SNRI, selective norepinephrine and serotonin reuptake inhibitor; 5-HT, serotonin; s.c., subcutaneous; MS/MS, tandem (Kalman et al., 2005; Paperwalla et al., 2004; Ziedonis et al., mass spectrometry; TCAs, tricyclic antidepressants. 2008). Data from the first wave of the US National Epidemiologic * Corresponding author at: Department of Neuroscience, Pfizer Global Research Survey on and Related Conditions indicate that, among and Development, Eastern Point Road, MS 8220-4457, Groton, CT 06340, USA. nicotine-dependent individuals, 21.1% had a mood disorder (Grant Tel.: +1 860 441 6374; fax: +1 860 715 5382. et al., 2004). The US National Comorbidity Study demonstrated E-mail address: hans.rollema@pfizer.com (H. Rollema). 1 Address: Brains On-Line BV, L.J. Zielstraweg 1, 9713 GX Groningen, The that smoking prevalence in those with a psychiatric diagnosis Netherlands. (41.0%) is about two times the rate seen in those without a

0197-0186/$ – see front matter ß 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.neuint.2010.10.015 H. Rollema et al. / Neurochemistry International 58 (2011) 78–84 79 psychiatric diagnosis (22.5%) (Lasser et al., 2000). An association and to have maximal effect on dopamine release in nucleus between smoking and occurrence of depressive symptoms has also accumbens (Coe et al., 2005; Rollema et al., 2007). Monoamine been reported in other population-based studies of adults in the release was measured in the medial prefrontal cortex (mPFC), a United States (Breslau et al., 1998; Kinnunen et al., 2006), Finland brain area that plays an important role in serotonin syndrome (Korhonen et al., 2007), Norway (Klungsøyr et al., 2006), and The (Gillman, 2006) as well as in the effects of antidepressants (Artigas Netherlands (Cuijpers et al., 2007). and Adell, 2007). In addition, we compiled in vitro data of Treating and depression are effective varenicline to examine whether its binding affinities or inhibitory ways to improve health, and a substantial number of smokers are potencies for relevant receptors, enzymes, or transporters would thus likely to receive antidepressant and be sufficient to inhibit or potentiate the pharmacologic effects of medication simultaneously. Given the possibility of pharmacoki- antidepressants at therapeutic doses. netic or pharmacodynamic drug–drug interactions, it is important to investigate whether and to what extent treatments for smoking 2. Materials and methods cessation could interact with co-administered antidepressants. 2.1. Materials Varenicline was recently introduced as a novel efficacious Varenicline tartrate and sertraline HCl were synthesized at Pfizer (Groton, CT, smoking cessation aid that acts as an a4b2 nicotinic acetylcholine USA); clorgyline HCl and all other chemicals were purchased from Sigma (St. Louis, receptor (nAChR) (Coe et al., 2005; Rollema et al., MO, USA) unless indicated otherwise. 2007). Since varenicline is virtually not metabolized, has low , is excreted unchanged renally, and does 2.2. In vivo microdialysis not inhibit or induce the activity of the major cytochrome P450 Male Wistar rats (280–350 g; Harlan, Horst, The Netherlands) were individually enzymes (Burstein et al., 2007; Faessel et al., 2010; Obach et al., housed in plastic cages (30 cm  30 cm  40 cm) under a 12:12 h light/dark cycle 2006), pharmacokinetic interactions between varenicline and (starting at 7:00 a.m.) and had access to food and water ad libitum. Experiments antidepressant drugs are highly unlikely. It was for instance were conducted in accordance with the declarations of Helsinki and were approved shown that concomitant administration of varenicline with the by the Institutional Animal Care and Use Committee of the University of Groningen, The Netherlands. Rats were anesthetized using isoflurane (2%, 800 ml/min O2). dopamine (DA) and norepinephrine (NE) reuptake inhibitor, Bupivacain/epinephrine was used for local anesthesia and fynadine (0.1%, 1 ml/kg) , in a randomized, placebo-controlled study in smokers for pre/peri operative analgesia. The animals were placed in a stereotaxic frame did not result in clinically relevant effect on bupropion’s (Kopf Instruments, Tujunga, CA, USA), and I-shaped probes (Hospal membrane, (Faessel et al., 2010). 4 mm exposed surface; Brainlink, Groningen, The Netherlands) were inserted into the mPFC. Coordinates for the tips of the probes were posterior (AP) +3.4 mm to However, both varenicline and antidepressant agents affect bregma, lateral +0.8 mm to midline, and ventral À5.0 mm to dura; the toothbar was monoaminergic neurotransmission and it is conceivable that a set at À3.3 mm (Paxinos and Watson, 1982). Experiments were performed 24–48 h pharmacodynamic interaction could cause either a reduction or an after probe implantation. On the day of the experiment, the probes were connected increase in the antidepressant-induced release of the monoamine with flexible PEEK tubing to a microperfusion pump (Syringe pump UV 8301501, DA, NE, and serotonin (5-HT). Pharmacody- TSE, Karlsruhe, Germany) and perfused with artificial cerebrospinal fluid (aCSF) containing 147 mM NaCl, 3.0 mM KCl, 1.2 mM CaCl2, and 1.2 mM MgCl2, at a flow namic interactions with the nAChR partial agonist varenicline rate of 1.5 ml/min. Microdialysis samples were collected for 15-min periods into could occur at nAChRs, since most antidepressant monoamine mini-vials containing 7.5 ml of 0.02 M formic acid, using an automated fraction reuptake inhibitors have weak nAChR antagonist properties (Fryer collector (CMA 142, CMA Microdialysis, Solna, Sweden), and stored at À80 8C and Lukas, 1999; Hennings et al., 1999; Szasz et al., 2007), but can pending analysis. At least four pre-drug basal samples were collected before compounds were administered (varenicline, 1 mg/ml free-base in saline; clorgyline, also be mediated via other, unknown mechanisms. Interactions 4 mg/ml free-base in saline; sertraline 17.8 mg/ml free-base in 20% solutol; all that would result in large increases in monoamine release might be compounds and vehicles were administered s.c. in a volume of 1 ml/kg s.c.). Effects associated with adverse events, such as serotonin syndrome via of each compound alone on monoamine dialysate levels were determined by accumulation of extracellular 5-HT levels and excessive activation administering varenicline, sertraline, clorgyline or corresponding vehicles and of central nervous system and peripheral serotonergic receptors collecting 15-min samples for 4.5 h. For the combination experiments varenicline and sertraline, which have similar half lives in the rat of approximately 4–5 h (Boyer and Shannon, 2005; Gillman, 2006; Shioda et al., 2004; Sun- (Obach et al., 2006; Tremaine et al., 1989) were administered together and Edelstein et al., 2008). On the other hand, a reduction of clorgyline was administered 2.5 h later. For each of the five treatment groups the antidepressant-induced central monoamine release could impact appropriate vehicles were given at the same time points (see Table 1) and six the clinical efficacy of monoaminergic antidepressants. To animals were used per group. After the experiment, rats were euthanized and brains were removed and stored for 3 days in a 4% (w/v) solution of investigate this possibility in a preclinical in vivo model, we paraformaldehyde. The position of each probe was histologically verified by measured the effects of varenicline treatment on basal and making coronal sections of the brain (Paxinos and Watson, 1982). antidepressant-induced elevated extracellular concentrations of 5-HT, DA, and NE in rat brain. Since the primary purpose of the 2.3. Neurotransmitter assays study was to examine the potential impact on serotonin syndrome, which is mostly associated with combinations of serotonergic Microdialysate concentrations of 5-HT, NE, and DA were determined by HPLC combined with tandem mass spectrometry (MS/MS) detection. Dialysate samples antidepressants, we used the selective serotonin reuptake inhibi- (22.5 ml) were collected in 0.3 ml polypropylene vials containing 7.5 mlof20mM tor (SSRI) sertraline and the monoamine oxidase inhibitor (MAO-I) formic acid. After addition of 4 ml of internal standards (deuterated 5-HT, NE, and clorgyline. Both compounds were given at doses that are known to DA) to each vial, 40 ml of the derivatization reagent SymDAQ (Klein et al., 2009)was maximally inhibit 5-HT transport and MAO, respectively (Artigas added with an autosampler (SIL-10AD vp, Shimadzu, Kyoto, Japan). Two minutes later, 50 ml of the mixture was injected onto the HPLC system and analytes were and Adell, 2007; Segal et al., 1992), and were combined with a separated over a reversed phase Phenomenex 100  3.0 mm (2.5 mm) column at varenicline dose known to result in levels above the human 30 8C, using a linear gradient of acetonitrile/0.1% formic acid at a flow rate of 0.3 ml/ therapeutic exposure (Faessel et al., 2010; Rollema et al., 2009b) min. The starting acetonitrile concentration (5%) was increased to 30% after 4 min,

Table 1 Varenicline (VAR) and/or sertraline and/or appropriate vehicles (saline or 20% solutol) were administered at t = 0 min and either clorgyline or vehicle (saline) was administered at t = 150 min.

Dosing at: VAR alone; VAR + MAO-I SSRI alone; SSRI + MAO-I MAO-I alone VAR + SSRI + MAO-I Vehicle

t = 0 min Solutol + varenicline Sertraline + saline Solutol + saline Sertraline + varenicline Solutol + saline t = 150 min Clorgyline Clorgyline Clorgyline Clorgyline Saline 80 H. Rollema et al. / Neurochemistry International[()TD$FIG] 58 (2011) 78–84 to 70% after 6.5 min, and decreased to 5% after 7 min. A post-column make-up flow 5-HT - effects of test compounds alone 300 A of mobile phase at 0.15 ml/min was added to the column effluent to enhance Vehicle ionization efficiency. The flow was diverted to waste for 2.5 min to avoid source T Clorgyline contamination and then switched to the MS, consisting of an API 4000 MS/MS H 250 Sertraline detector and a Turbo Ion Spray interface (Applied Biosystems, Ijssel, The

+ SEM Varenicline Netherlands). The acquisitions were performed in positive ionization mode with 200 the ion spray voltage set at 3 kV and a probe temperature of 200 8C, using multiple- reaction-monitoring (MRM). The collision gas (nitrogen) pressure was held at 150 2 psig. Data were calibrated and quantified using the AnalystTM data system Extracellular 5- (Applied Biosystems, version 1.4.2). Mass transitions were: 5-HT Q1 422.2 and Q3 % of basal 100 478.4; 5-HT-d4 Q1 426.2 and Q3 482.4; NE Q1 415.3 and Q3 371.7; NE-d6 Q1 421.3 and Q3 377.7; DA Q1 399.2 and Q3 355.4; DA-d4 Q1 403.2 and Q3 359.4. Var, Sert, or Veh Clorg or Veh 50 -90 -60 -30 0 30 60 90 120 150 180 210 240 270 2.4. Data analysis Time (minutes)

Four consecutive pre-treatment microdialysis samples with less than 50% NE - effects of test compounds alone 300 B variation were taken as baseline and the mean concentration was set at 100%. Drug Vehicle effects were expressed as percentages of basal level (mean Æ SE) within the same Clorgyline subject. Statistical analysis was performed using Sigmastat for Windows (SPSS E 250

N Sertraline Corporation, Chicago, IL, USA). Varenicline + SEM Treatment effects were compared with baseline and between treatments, using 200 two-way analysis of variance (ANOVA) for repeated measurements followed by Student Newman Keuls post hoc test. Treatment and time were the main effects. 150 The level of statistical significance was set at p < 0.05. Extracellular % of basal 100 2.5. In vitro profile Var, Sert, or Veh Clorg or Veh In vitro data on the interaction of varenicline with selected receptors, 50 transporters, and enzymes were taken from the literature or generated by standard -90 -60 -30 0 30 60 90 120 150 180 210 240 270 methods (indicated in Table 4). Time (minutes)

DA - effects of test compounds alone 300 C 3. Results Vehicle Clorgyline 3.1. Microdialysis of monoamines 250 Sertraline A M

D Varenicline 200 3.1.1. Basal and drug-evoked extracellular monoamine + SE concentrations 150 Basal concentrations of 5-HT, NE, and DA in microdialysates Extracellular from rat mPFC did not differ between treatment groups, with mean % of basal 100 values (not corrected for recovery) of 5.30 Æ 0.48 (n = 30), Var, Sert, or Veh Clorg or Veh 16.55 Æ 1.31 (n = 30), and 7.65 Æ 0.91 fmol/30 ml sample (n = 30), 50 respectively. Effects of vehicle and drug treatments on extracellular -90 -60 -30 0 30 60 90 120 150 180 210 240 270 Time (minutes) monoamine concentrations are expressed as percentages of baseline levels Æ SEM and shown as time courses in Figs. 1–4. Results of Fig. 1. Time courses for the effects of vehicle (grey squares), varenicline (1 mg/kg, statistical analyses are summarized in Tables 2 and 3. black diamonds), sertraline (17.8 mg/kg, open squares), and clorgyline (4 mg/kg, grey circles) on extracellular concentrations of 5-HT (A), NE (B), and DA (C) in rat prefrontal cortex. Data are expressed as percentage baseline Æ SEM. Arrows indicate 3.1.2. Effects of administration of test compounds alone on vehicle or drug administration at t = 0 for vehicle or varenicline or sertraline, or at extracellular monoamine levels t = 2.5 h for clorgyline (to mimic conditions of the combination experiments). Varenicline, given at 1 mg/kg, caused no significant increases in Statistical significance of the drug effects are listed in Table 1. the extracellular levels of 5-HT, NE, and DA in rat mPFC compared [()TD$FIG] with vehicle treatment (Fig. 1, statistical evaluation in Table 2). Sertraline, when given alone at a dose of 17.8 mg/kg, increased 5-HT - effects of drug combinations cortical 5-HT levels approximately 2.5-fold within 1 h after dosing, 1800 while at this dose sertraline caused a small, significant increase in Varenicline + Sertraline + Clorgyline 1600 Vehicle + Sertraline + Clorgyline NE levels to about 1.5-fold of baseline levels 2 h after administra- 1400 Varenicline + Vehicle + Clorgyline tion and had no significant effect on extracellular DA levels (Fig. 1 Vehicle + Vehicle + Clorgyline + SEM 1200 and Table 2). Clorgyline administration (4 mg/kg) alone effectively al s 1000 inhibited MAO and caused the expected increases in extracellular 800 levels of all three monoamines, 5-HT to 1.5-fold, NE to 2.5-fold, and Clorg

% of ba 600 Extracellular 5-HT Var, Sert, or Veh DA to 2.5-fold of basal levels (Fig. 1 and Table 2). 400 200 3.1.3. Effect of varenicline on sertraline- and clorgyline-induced 0 extracellular 5-HT levels -90 -60 -30 0 30 60 90 120 150 180 210 240 270 Varenicline (1 mg/kg s.c.) had no significant effect on the Time (minutes) increases in extracellular 5-HT levels in rat mPFC produced either Fig. 2. Time courses for the effects of pretreatment with vehicle or 1 mg/kg by sertraline or by clorgyline alone. Administration of 4 mg/kg varenicline on increases in extracellular 5-HT concentrations in rat prefrontal clorgyline to rats pretreated with 17.8 mg/kg sertraline increased cortex produced by sertraline (17.8 mg/kg s.c.), clorgyline (4 mg/kg s.c.), or co- extracellular 5-HT concentrations to more than 10-fold of baseline administration of sertraline (17.8 mg/kg) and clorgyline (4 mg/kg). Data are expressed as percentages of baseline Æ SEM. Arrows indicate drug administrations at levels. This massive 5-HT increase was not affected by pretreat- t = 0 (varenicline, or sertraline or vehicle) and at t = 150 min (clorgyline). Time courses ment with varenicline and time courses for the effects of sertraline, for the effects of controls (vehicle + vehicle + vehicle treatment) are shown in Fig. 1 and clorgyline, or sertraline with clorgyline on 5-HT in the absence and not included in Figs. 2–4. Statistical significance of the drug effects are listed in Table 2. [()TD$FIG] H. Rollema et al. / Neurochemistry International 58 (2011) 78–84 81

NE - effects of drug combinations 3.1.5. Effect of varenicline on sertraline- and clorgyline-induced 800 Varenicline + Sertraline + Clorgyline extracellular DA levels 700 Vehicle + Sertraline + Clorgyline Varenicline and sertraline, which when given separately do not Varenicline + Vehicle + Clorgyline 600 Vehicle + Vehicle + Clorgyline affect extracellular DA concentrations, had also no significant

+ SEM 500 effect when administered together. The small clorgyline-induced 400 Clorg DA increases appeared to be affected by varenicline pretreatment. Var, Sert, or Veh 300 We have no reasonable explanation for this observation, although Extracellular NE Extracellular % of basal 200 it could be due to the MAO-A inhibitory activity of clorgyline that has an indirect and variable effect on DA release via 5-HT and/or NE 100 increases. Importantly, varenicline administration showed a trend 0 -90 -60 -30 0 30 60 90 120 150 180 210 240 270 to reducing the 2.5-fold DA increase produced by co-administra- Time (minutes) tion of sertraline and clorgyline, but this effect was not significant (Fig. 4, F1,72 = 0.762; p = 0.403). Fig. 3. Time courses for the effects of pretreatment with vehicle or varenicline (1 mg/kg s.c.) on increases in extracellular NE concentrations in rat prefrontal cortex produced by sertraline (17.8 mg/kg s.c.), clorgyline (4 mg/kg s.c.), or co- 3.2. In vitro binding profile of varenicline administration of sertraline (17.8 mg/kg) and clorgyline (4 mg/kg). For further legends see Fig. 2. The in vitro binding profile (Table 4) shows that varenicline is

highly selective for a4b2 containing nAChRs (Ki = 0.4 nM) and [()TD$FIG] binds with at least 200-fold lower affinity to other nAChR subtypes

(Ki  85 nM). Varenicline does not have significant affinity for DA- effects of drug combinations other ion channels or for neurotransmitter receptors, uptake sites 500 Varenicline + Sertraline + Clorgyline and enzymes, including targets of common antidepressant drugs, Vehicle + Sertraline + Clorgyline such as the monoamine transporters, the metabolizing enzyme 400 Varenicline + Vehicle + Clorgyline Vehicle + Vehicle + Clorgyline MAO, and the various 5-HT, NE, and DA receptor subtypes (Ki values >1 mM), with the exception of its affinity for the 5-HT3A 300 Clorg receptor (Ki = 0.35 mM). Var, Sert, or Veh 200

Extracellular DA 4. Discussion % of basal + SEM % of basal + 100 Since a large proportion of smokers have comorbid depression 0 and smoking prevalence is approximately 50% among psychiatric -90 -60 -30 0 30 60 90 120 150 180 210 240 270 Time (minutes) outpatients with depression (Grant et al., 2004; Kalman et al., 2005; Paperwalla et al., 2004; Ziedonis et al., 2008), nicotine Fig. 4. Time courses for the effects of pretreatment with vehicle or varenicline dependence and depression will sometimes be treated simulta- (1 mg/kg s.c.) on increases in extracellular DA concentrations in rat prefrontal neously with a smoking cessation aid and an antidepressant, which cortex produced by sertraline (17.8 mg/kg s.c.), clorgyline (4 mg/kg s.c.), or co- administration of sertraline (17.8 mg/kg) and clorgyline (4 mg/kg). For further theoretically at least could lead to either pharmacokinetic or legends see Fig. 2. pharmacodynamic drug–drug interactions. This has not yet been investigated in preclinical models for the recently introduced smoking cessation aid varenicline, a partial agonist at the a4b2 the presence of varenicline were essentially the same (Fig. 2 and nAChR (Cahill et al., 2007; Coe et al., 2005; Rollema et al., 2007). Table 3). Since its pharmacokinetic profile (Burstein et al., 2007; Faessel et al., 2010; Obach et al., 2006) makes it highly unlikely that 3.1.4. Effect of varenicline on sertraline- and clorgyline-induced varenicline will affect the disposition of concomitantly adminis- extracellular NE levels tered antidepressants, this study investigated potential pharma- Varenicline did not significantly change the increases in cortical codynamic interactions in an animal model with two extracellular NE levels produced by either sertraline alone or antidepressants that act via different mechanisms. The results clorgyline alone. Co-administration of sertraline and clorgyline show that varenicline neither reduces nor further increases caused a robust 5-fold increase in extracellular NE concentrations elevated monoamine neurotransmitter levels produced by admin- that was not modulated by pretreatment with varenicline, again istration of an SSRI and an MOA-I, or by combined SSRI and MOA-I resulting in similar time-response courses for the effects of administration, consistent with its in vitro profile. sertraline, clorgyline, or sertraline with clorgyline on NE in the Currently, antidepressant pharmacotherapy is aimed at im- absence and the presence of varenicline (Fig. 3 and Table 3). proving dysfunctional monoaminergic neurotransmission by chronically elevating extracellular monoamine levels through Table 2 multiple mechanisms, such as inhibition of monoamine re-uptake Statistical analysis of the effects of drug alone treatment vs. vehicle treatment. by tricyclic antidepressants (TCAs), selective inhibition of seroto- Comparison Time (min) Analyte Fp nin or serotonin and norepinephrine reuptake by SSRIs or SNRIs, and inhibition of monoamine metabolism by MAO-Is (Slattery Varenicline vs. vehicle 0–150 5-HT F1,93 = 1.296 p = 0.281 ns

NE F1,94 = 4.204 p = 0.067 ns et al., 2004). The neurochemical effects of these antidepressants

DA F1,93 = 0.796 p = 0.393 ns have been well characterized in numerous in vivo intracerebral Sertraline vs. vehicle 0–150 5-HT F1,98 = 12.880 p = 0.005** microdialysis studies by assessment of the effects on extracellular NE F1,98 = 6.303 p = 0.031* transmitter levels in rodent brain areas, such as hippocampus and DA F = 0.066 p = 0.802 ns 1,97 prefrontal cortex (Artigas and Adell, 2007). Varenicline is known to Clorgyline vs. vehicle 150–270 5-HT F1,80 = 7.617 p = 0.020*

NE F1,80 = 84.187 p < 0.001** increase DA release in the nucleus accumbens and striatum, which DA F1,97 = 34.695 p < 0.001** is key to its clinical efficacy as a smoking cessation aid (Coe et al., ns: no significant effect compared with vehicle treatment. 2005; Rollema et al., 2007), but does not significantly modulate DA *p < 0.05, **p < 0.01 significant effect compared with vehicle treatment. or NE release in the prefrontal cortex (Coe et al., 2005; Rollema 82 H. Rollema et al. / Neurochemistry International 58 (2011) 78–84

Table 3 Statistical analysis of effects of varenicline vs. vehicle treatment on antidepressant-induced extracellular monoamine concentrations.

Comparison Time (min) Analyte Fp

Varenicline + sertraline vs. vehicle + sertraline 0–150 5-HT F1,93 = 0.0002 p = 0.988 ns

NE F1,94 = 2.705 p = 0.131 ns

DA F1,93 = 13.315 p = 0.587 ns

Varenicline + vehicle + clorgyline vs. vehicle + vehicle + clorgyline 150–270 5-HT F1,72 < 0.001 p = 0.996 ns

NE F1,72 = 0.0375 p = 0.851 ns

DA F1,72 = 10.257 p = 0.011 s

Varenicline + sertraline + clorgyline vs. vehicle + sertraline + clorgyline 0–270 5-HT F1,168 = 0.501 p = 0.495 ns

NE F1,169 = 0.037 p = 0.852 ns

DA F1,165 = 0.365 p = 0.559 ns ns: varenicline treatment did not produce a significant effect compared with vehicle treatment. s: varenicline treatment produced a significant effect compared with vehicle treatment.

et al., 2007, 2009b). However, it is conceivable that pharmacologic with previous reports, which showed that varenicline only interactions with antidepressants could result in reductions or produces robust increases in cortical DA and NE release after a increases in antidepressant-induced elevated monoamine neuro- high dose of 10 mg/kg (Rollema et al., 2009b). This dose is transmitters. Most classes of antidepressant drugs have been associated with very high brain concentrations of >1 mM, which shown to act as weak nAChR antagonists (Fryer and Lukas, 1999; can interact with several other nAChR subtypes than a4b2 Hennings et al., 1999; Szasz et al., 2007), and might thus interact nAChRs. The lack of effects of varenicline on cortical monoamine with varenicline at nAChRs, while other, unknown mechanisms release is also consistent with its in vitro properties, since could also result in pharmacodynamic interactions and modula- varenicline has very low affinity for the 5-HT transporter or 5- tion of monoamine release. HT receptors and can thus neither increase 5-HT by blocking 5-HT We applied the microdialysis technique to examine the effects reuptake or by interacting with central 5-HT receptors. Although of varenicline alone and combined with two antidepressants that varenicline has modest affinity for 5-HT3A receptors (Ki = 0.35 mM) have a different mechanism of action, on extracellular levels of at which it is a full agonist (EC50 =1mM), the therapeutic unbound monoamines in rat prefrontal cortex, an area that is believed to be varenicline brain concentrations are predicted to be approximately involved in the underlying mechanism of serotonin syndrome and 50 nM and thus insufficient for activating central 5-HT3A receptors. in the action of antidepressants. We used sertraline and clorgyline: Furthermore, varenicline does not bind to DA receptors sertraline is an efficacious SSRI (Cipriani et al., 2009) that has, like (Ki > 1 mM) that can modulate 5-HT release, and does not inhibit other SSRIs, weak nAChR antagonist activity (Fryer and Lukas, the enzyme that metabolizes 5-HT, MAO-A (IC50 > 1 mM). 1999), and clorgyline is a potent MAO-I with antidepressant Sertraline produced the well-documented robust and long- efficacy (Lipper et al., 1979). The effect of co-administration of an lasting increase in extracellular 5-HT levels in rat brain (Artigas SSRI and an MAO-I was studied because that combination is most and Adell, 2007; Sprouse et al., 1996). Adding clorgyline to frequently associated with serotonin syndrome (Gillman, 2006). sertraline-treated rats significantly further elevated 5-HT levels For this study we selected a dose of each compound known to be and produced the expected increases in extracellular DA and NE as maximally effective in previous microdialysis studies in rodents a consequence of MAO inhibition (Segal et al., 1992; Waldmeier (Artigas and Adell, 2007; Coe et al., 2005; Rollema et al., 2009b; et al., 1976), confirming numerous previous reports on the effects Segal et al., 1992; Sprouse et al., 1996; Waldmeier et al., 1976). We of antidepressants on extracellular neurotransmitter levels. used a relatively high dose of 1 mg/kg varenicline that is associated Administration of 1 mg/kg varenicline did not significantly change with peak rat plasma concentrations of about 150 ng/ml at 30 min the elevated extracellular monoamine levels induced by sertraline (Rollema et al., 2009b), which, with a half life in rats of 4.5 h (Obach alone, by clorgyline alone, or by sertraline followed by clorgyline et al., 2006), decline to about 40 ng/ml after 3 h. The 1 mg/kg dose treatment. The lack of an effect of varenicline on antidepressant- results in higher rat plasma concentrations throughout the induced transmitter increases is again consistent with its in vitro duration of the microdialysis experiment than steady state profile of a highly selective a4b2 nAChR partial agonist that has therapeutic plasma levels in patients on 1 mg BID, which are poor affinity for the receptors, transporters, or enzymes that can about 10 ng/ml (Faessel et al., 2010). affect central momoamine levels (Table 4). Varenicline by itself caused no significant changes in extracel- These results have both safety and therapeutic implications. lular levels of 5-HT, NE, or DA in rat prefrontal cortex, in agreement Most importantly, the data do not suggest safety concerns related to exaggerated neurotransmitter release when combining vareni- Table 4 cline with serotonergic antidepressants. A well-known potential Binding affinities and inhibitory potencies of varenicline for ion channels, receptors, side effect of excessive 5-HT release is serotonin syndrome, a uptake sites, and enzymes. Ki or IC50 values >1or10mM indicate no significant potentially life-threatening adverse reaction to therapeutic drug inhibition of radioligand binding at 1 or 10 mM varenicline. Data from Rollema et al. use or inadvertent interactions between drugs, resulting in 5-HT (2007, 2009b) and on file (Pfizer Inc.). accumulation and excess activation of serotonergic receptors Ion channels Ki (nM) Receptors Ki (mM) Uptake sites Ki (mM) (Gillman, 2006; Shioda et al., 2004; Sun-Edelstein et al., 2008). A

a4b2 nAChR 0.4 Dopamine D1–5 >1DA >1 wide variety of drugs and drug combinations have been associated a3b2 nAChR 86 NEa1,2;NEb >10 NE >1 with this condition, including MAOIs, TCAs, SSRIs, SNRIs, opiate a6 nAChR 111 5HT1,2,4,6 >1 5-HT >1 analgesics, antibiotics, anti-emetics, and drugs that inhibit a7 nAChR 125 5-HT 0.35 >1 3 cytochrome P450 isoforms. However, serotonin syndrome has a1 nAChR 8200 Muscarinic M1–5 >10 GABA >1 Glutamate >1 Glu >1 been most frequently observed after combined administration of K, Na, Ca >1000 GABA >1 antidepressant drugs that act via increased serotonergic neuro- GABACl >1000 Cannabinoid CB1,2 >10 Enzymes IC50 (mM) transmission, in particular co-administration of SSRIs and MAOIs Glutamate >1000 Neurokinin NK1 >10 MAO-A >1 (Gillman, 2006; Shioda et al., 2004; Sun-Edelstein et al., 2008). Histamine H >10 Cyt-P450 >30 1–3 Previous microdialysis studies on the neurochemistry of serotonin H. Rollema et al. / Neurochemistry International 58 (2011) 78–84 83 syndrome have shown a close correlation between the magnitude interaction between varenicline and the SSRI or MAO-I class of of cortical extracellular 5-HT increases produced by antidepressant antidepressants is unlikely and that varenicline has little risk of drugs and the severity of serotonin syndrome in patients taking either triggering adverse events by excessive neurotransmitter these drugs (Shioda et al., 2004). Since the present study shows release or reducing the efficacy of concomitant therapy with that varenicline does not increase central 5-HT levels and does not serotonergic antidepressants. Taken together with clinical and have an additive or synergistic effect on 5-HT increases produced post-marketing data, the available evidence at this time does not by an SSRI, by an MAO-I, or by an SSRI and MAO-I combination, demonstrate that varenicline is likely to have pharmacodynamic or varenicline is unlikely to cause or exacerbate serotonin syndrome. pharmacokinetic interactions with serotonergic antidepressant While a recent report of a potential association between antidepressants; nor does the data support the view that varenicline and development of serotonin syndrome was based on varenicline plays a role in serotonin syndrome reported in some seven cases of suspected serotonin syndrome in patients taking patients. varenicline that were spontaneously reported to the World Health Organization’s VigiBase database (Johannson and Hill, 2009), in all Acknowledgements but one of the reported cases, varenicline was used in combination with antidepressants associated with serotonin syndrome. In Microdialysis studies were performed by JHAF and GF (Brains addition to varenicline’s lack of effect on central extracellular 5- On-Line, The Netherlands) and were funded by a Pfizer grant. JHAF HT levels, clinical data provide further support that varenicline is and GF report no other potential conflicts of interest. HR, GGW, and unlikely to have played a role in the reported cases of serotonin TCL are employees of Pfizer Inc, own Pfizer stock, and have Pfizer syndrome, but rather was coincidentally taken by patients already stock options. Editorial support was provided by Brenda Smith, at higher risk of the event due to underlying comorbid illness and PhD and Abegale Templar, PhD of UBC Scientific Solutions, and was antidepressant medication. In 10 randomized, double-blind, funded by Pfizer Inc. placebo-controlled phase II–IV Pfizer-sponsored varenicline trials completed through end of March 2009 (Tonstad et al., 2010), and three open-label varenicline trials (Aubin et al., 2008; Pfizer Inc., References 2007; Tonstad et al., 2006), which excluded patients with depression requiring antidepressant treatment, there were no Artigas, F., Adell, A., 2007. The use of brain microdialysis in antidepressant drug research. In: Westerink, B., Cremers, T.I.F.H. (Eds.), Handbook of Microdialysis. reports of serotonin syndrome amongst over 5000 varenicline Academic Press, pp. 527–544. patients. In addition, a keyword search in the literature (key- Aubin, H.J., Bobak, A., Britton, J.R., Oncken, C., Billing Jr., C.B., Gong, J., Williams, K.E., words: ‘‘varenicline’’ and ‘‘serotonin syndrome’’ or ‘‘serotonin Reeves, K.R., 2008. Varenicline versus transdermal for smoking cessation: results from a randomised open-label trial. Thorax 63, 717–724. toxicity’’) through January 2010 revealed no other published Boyer, E.W., Shannon, M., 2005. The serotonin syndrome. N. Engl. J. Med. 352, 1112– reports of serotonin syndrome or serotonin toxicity associated 1120. with varenicline use. Breslau, N., Peterson, E.L., Schultz, L.R., Chilcoat, H.D., Andreski, P., 1998. Major depression and stages of smoking. A longitudinal investigation. Arch. Gen. The pharmacotherapeutic significance of this study is related to Psychiatry 55, 161–166. the potential impact of varenicline on the clinical efficacy of co- Burstein, A.H., Clark, D.J., O’Gorman, M., Willavize, S.A., Brayman, T.G., Grover, G.S., administered SSRI and MAO-I antidepressants. Since varenicline Walsky, R.L., Obach, R.S., Faessel, H.M., 2007. Lack of pharmacokinetic and pharmacodynamic interactions between a smoking cessation therapy, vareni- does not attenuate increases in central extracellular 5-HT levels cline, and warfarin: an in vivo and in vitro study. J. Clin. Pharmacol. 47, 1421– induced by those antidepressants, it is unlikely to reduce the action 1429. of serotonergic antidepressant drugs, e.g., via blunting of the 5-HT Cahill, K., Stead, L.F., Lancaster, T., 2007. Nicotine receptor partial for smoking cessation. Cochrane Database Syst. Rev. CD006103. response. On the contrary, several preclinical studies have shown Cipriani, A., Furukawa, T.A., Salanti, G., Geddes, J.R., Higgins, J.P., Churchill, R., that a4b2 nAChR partial agonists can augment the effect of Watanabe, N., Nakagawa, A., Omori, I.M., McGuire, H., Tansella, M., Barbui, classical antidepressants, most likely via reduction of C., 2009. Comparative efficacy and acceptability of 12 new-generation anti- signaling (Picciotto et al., 2008; Shytle et al., 2002). The data also depressants: a multiple-treatments meta-analysis. Lancet 373, 746–758. Coe, J.W., Brooks, P.R., Vetelino, M.G., Wirtz, M.C., Arnold, E.P., Huang, J., Sands, S.B., indicate that the recent observation that varenicline augments Davis, T.I., Lebel, L.A., Fox, C.B., Shrikhande, A., Heym, J.H., Schaeffer, E., Rollema, sertraline’s antidepressant-like activity in a preclinical murine H., Lu, Y., Mansbach, R.S., Chambers, L.K., Rovetti, C.C., Schulz, D.W., Tingley III, antidepressant model (Rollema et al., 2009a) is unlikely due to a F.D., O’Neill, B.T., 2005. Varenicline: an alpha4beta2 nicotinic receptor partial agonist for smoking cessation. J. Med. Chem. 48, 3474–3477. synergistic effect on SSRI-induced 5-HT increase, but rather a Cuijpers, P., Smit, F., ten Have, M., de Graaf, R., 2007. Smoking is associated with consequence of varenicline’s interaction with a4b2 nAChRs. first-ever incidence of mental disorders: a prospective population-based study. Finally, the lack of effect on evoked increases in cortical DA and 102, 1303–1309. Faessel, H.M., Obach, R.S., Rollema, H., Ravva, P., Burstein, A.H., 2010. A review of the NE release suggests that varenicline would not have an effect on clinical pharmacokinetics and pharmacodynamics of varenicline for smoking the efficacy of other types of antidepressants that act via elevated cessation. Clin. Pharmacokinetics 49, 799–816. DA and/or NE levels. However, since the present experiments were Fryer, J.D., Lukas, R.J., 1999. Antidepressants noncompetitively inhibit nicotinic acetylcholine receptor function. J. Neurochem. 72, 1117–1124. only conducted with single doses of a typical SSRI and MAO-I in the Gillman, P.K., 2006. A review of serotonin toxicity data: implications for the prefrontal cortex, further studies would be needed to examine mechanisms of antidepressant drug action. Biol. Psychiatry 59, 1046–1051. potential interactions with NE or DA reuptake-inhibitors. Grant, B., Hasin, D., Chou, S., Stinson, F., Dawson, D., 2004. Nicotine dependence and psychiatric disorders in the United States: results from the national epidemio- logic survey on alcohol and related conditions. Arch. Gen. Psychiatry 61, 1107– 5. Conclusions 1115. Hennings, E.C., Kiss, J.P., De Oliveira, K., Toth, P.T., Vizi, E.S., 1999. Nicotinic In view of the high comorbidity of nicotine dependence with acetylcholine receptor antagonistic activity of monoamine uptake blockers in rat hippocampal slices. J. Neurochem. 73, 1043–1050. mood disorders, this study investigated potential interactions of Johannson, J., Hill, R., 2009. Serotonin syndrome following varenicline use. Drug Saf. the smoking cessation agent varenicline with two antidepressant 32, 932. drugs, an SSRI (sertraline) and an MAOI (clorgyline), in a preclinical Kalman, D., Morissette, S.B., George, T.P., 2005. Co-morbidity of smoking in patients with psychiatric and substance use disorders. Am. J. Addict. 14, 106–123. animal model. Varenicline by itself has no effect on the cortical Kinnunen, T., Haukkala, A., Korhonen, T., Quiles, Z.N., Spiro III, A., Garvey, A.J., 2006. release of 5-HT, NE, and DA and when combined with one or with Depression and smoking across 25 years of the Normative Aging Study. Int. J. both antidepressant drugs, it did not cause significant effects on Psychiatry Med. 36, 413–426. Klein, T., Hofland, C.M., Dremencov, E., Cremers, T.I.F.H., 2009. SymDAQ derivatiza- the increased neurotransmitter levels induced by sertraline and/or tion for single-run profiling of MA and AA in microdialysate, plasma and clorgyline. These results demonstrate that a pharmacodynamic cerebrospinal fluid by LC–MS/MS. Soc. Neuroscience Abstr. FF95. 84 H. Rollema et al. / Neurochemistry International 58 (2011) 78–84

Klungsøyr, O., Nyga˚ rd, J.F., Sørensen, T., Sandanger, I., 2006. smoking and Segal, D.S., Kuczenski, R., Okuda, C., 1992. Clorgyline-induced increases in presyn- incidence of first depressive episode: an 11-year, population-based follow-up aptic DA: changes in the behavioral and neurochemical effects of study. Am. J. Epidemiol. 163, 421–432. using in vivo microdialysis. Pharmacol. Biochem. Behav. 42, 421–429. Korhonen, T., Broms, U., Varjonen, J., Romanov, K., Koskenvuo, M., Kinnunen, T., Shioda, K., Nisijima, K., Yoshino, T., Kato, S., 2004. Extracellular serotonin, dopamine Kaprio, J., 2007. Smoking behaviour as a predictor of depression among Finnish and glutamate levels are elevated in the hypothalamus in a serotonin syndrome men and women: a prospective cohort study of adult twins. Psychol. Med. 37, animal model induced by and fluoxetine. Prog. Neuropsycho- 705–715. pharmacol. Biol. Psychiatry 28, 633–640. Lasser, K., Boyd, J.W., Woolhandler, S., Himmelstein, D.U., McCormick, D., Bor, D.H., Shytle, R.D., Silver, A.A., Lukas, R.J., Newman, M.B., Sheehan, D.V., Sanberg, P.R., 2000. Smoking and mental illness: a population-based prevalence study. JAMA 2002. Nicotinic acetylcholine receptors as targets for antidepressants. Mol. 284, 2606–2610. Psychiatry 7, 525–535. Lipper, S., Murphy, D.L., Slater, S., Buchsbaum, M.S., 1979. Comparative behavioral Slattery, D.A., Hudson, A.L., Nutt, D.J., 2004. Invited review: the evolution of effects of clorgyline and pargyline in man: a preliminary evaluation. Psycho- antidepressant mechanisms. Fundam. Clin. Pharmacol. 18, 1–21. pharmacology (Berl.) 62, 123–128. Sprouse, J., Clarke, T., Reynolds, L., Heym, J., Rollema, H., 1996. Comparison of the Obach, R.S., Reed-Hagen, A.E., Krueger, S.S., Beth, J., Obach, B.J., O’Connell, T.N., effects of sertraline and its metabolite desmethylsertraline on blockade of Zandi, K.S., Miller, S., Coe, J.W., 2006. Metabolism and disposition of varenicline, central 5-HT reuptake in vivo. Neuropsychopharmacology 14, 225–231. a selective a4b2 acetylcholine receptor partial agonist, in vivo and in vitro. Sun-Edelstein, C., Tepper, S.J., Shapiro, R.E., 2008. Drug-induced serotonin syn- Drug. Metab. Disp. 34, 121–130. drome: a review. Expert. Opin. Drug Saf. 7, 587–596. Paperwalla, K.N., Levin, T.T., Weiner, J., Saravay, S.M., 2004. Smoking and depression. Szasz, B.K., Vizi, E.S., Kiss, J.P., 2007. Nicotinic acetylcholine receptor antagonistic Med. Clin. North Am. 88, 1483–1494 x–xi. property of the selective dopamine uptake inhibitor, GBR-12909 in rat hippo- Paxinos, G., Watson, C., 1982. The Rat Brain in Stereotaxic Coordinates. Academic campal slices. Neuroscience 145, 344–349. Press, San Diego. Tonstad, S., Davies, S., Flammer, M., Russ, C., Hughes, J., 2010. Psychiatric adverse Pfizer Inc., 2007. A study of the efficacy and safety of varenicline in Japanese events in randomized, double-blind, placebo-controlled clinical trials of var- smokers motivated to quit smoking. enicline: a pooled analysis. Drug Saf. 33, 289–301. Picciotto, M.R., Addy, N.A., Mineur, Y.S., Brunzell, D.H., 2008. It is not ‘‘either/or’’: Tonstad, S., Tønnesen, P., Hajek, P., Williams, K.E., Billing, C.B., Reeves, K.R., Var- activation and desensitization of nicotinic acetylcholine receptors both con- enicline Phase 3 Study Group, 2006. Effect of maintenance therapy with tribute to behaviors related to nicotine addiction and mood. Prog. Neurobiol. 84, varenicline on smoking cessation: a randomized controlled trial. JAMA 296, 329–342. 64–71. Rollema, H., Chambers, L.K., Coe, J.W., Glowa, J., Hurst, R.S., Lebel, L.A., Lu, Y., Tremaine, L.M., Welch, W.M., Ronfeld, R.A., 1989. Metabolism and disposition of the Mansbach, R.S., Mather, R.J., Rovetti, C.C., Sands, S.B., Schaeffer, E., Schulz, 5-hydroxytryptamine uptake blocker sertraline in the rat and dog. Drug Metab. D.W., Tingley III, F.D., Williams, K.E., 2007. Pharmacological profile of the Dispos. 17, 542–550. alpha4beta2 nicotinic acetylcholine receptor partial agonist varenicline, an U¨ stu¨ n, T.B., Ayuso-Mateos, J.L., Chatterji, S., Mathers, C., Murray, C.J., 2004. Global effective smoking cessation aid. Neuropharmacology 52, 985–994. burden of depressive disorders in the year 2000. Br. J. Psychiatry 184, 386–392. Rollema, H., Guanowsky, V., Mineur, Y.S., Shrikhande, A., Coe, J.W., Seymour, P.A., Waldmeier, P.C., Delini-Stula, A., Maıˆtre, L., 1976. Preferential deamination of Picciotto, M.R., 2009a. Varenicline has antidepressant-like activity in the forced dopamine by an A type monoamine oxidase in rat brain. Naunyn Schmiede- swim test and augments sertraline’s effect. Eur. J. Pharmacol. 605, 114– bergs Arch. Pharmacol. 292, 9–14. 116. World Health Organization, 2008. WHO Report on the Global Epidemic: Rollema, H., Hajo´ s, M., Seymour, P.A., Kozak, R., Majchrzak, M.J., Guanowsky, V., The MPOWER Package. World Health Organization, Geneva. Horner, W.E., Chapin, D.S., Hoffmann, W.E., Johnson, D.E., McLean, S., Freeman, Ziedonis, D., Hitsman, B., Beckham, J.C., Zvolensky, M., Adler, L.E., Audrain-McGo- J., Williams, K.E., 2009b. Preclinical pharmacology of the alpha4beta2 nAChR vern, J., Breslau, N., Brown, R.A., George, T.P., Williams, J., Calhoun, P.S., Riley, partial agonist varenicline related to effects on reward, mood and cognition. W.T., 2008. Tobacco use and cessation in psychiatric disorders: National Insti- Biochem. Pharmacol. 78, 813–824. tute of Mental Health report. Nicotine Tob. Res. 10, 1691–1715.