Journal of Pharmacological and Toxicological Methods 70 (2014) 120–127

Contents lists available at ScienceDirect

Journal of Pharmacological and Toxicological Methods

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

Original article

A vapourized Δ9-tetrahydrocannabinol (Δ9-THC) delivery system part I:Development and validation of a pulmonary cannabinoid route ofexposure for experimental pharmacology studies in rodents

Laurie A. Manwell a,b,c,d,⁎, Armen Charchoglyan c, Dyanne Brewer c, Brittany A. Matthews a,Heather Heipel a, Paul E. Mallet a,⁎a Department of Psychology, Wilfrid Laurier University, Waterloo, ON N2L3C5, Canadab Department of Psychology, University of Guelph, Guelph, ON N1G2W1, Canadac Mass Spectrometry Facility, University of Guelph, Guelph, ON N1G2W1, Canadad Centre for Addiction and Mental Health, Social Aetiology of Mental Illness Program, University of Toronto, ON M5T1R8, Canada

Abbreviations: (Δ9-THC), Δ9-tetrahydrocannabindroxy-Δ9-tetrahydrocannabinol; (AEA), arachidonoarachidonoylglycerol; (APCI), atmospheric pressurecoefficient of variation; (CPA), conditioned place averspreference; (DIs), deuterated internal standards; (E(ECB), endocannabinoid; (FAAH), fatty acid amide hydroacid; (LLE), liquid–liquid extraction; (HPLC), high perform(LOQ), limit of quantification; (LOD), limit of detection; (mass spectrometry; (SPE), solid-phase extraction; (SIM),⁎ Corresponding authors at: Unit 1111, 33 Russell Street

Tel.: +1 416 535 8501x77632.E-mail addresses: laurie.manwell@camh.ca, laurieman

http://dx.doi.org/10.1016/j.vascn.2014.06.0061056-8719/© 2014 Elsevier Inc. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 25 February 2014Accepted 4 June 2014Available online 25 June 2014

Keywords:Δ9-tetrahydrocannabinol (THC)11-hydroxy-Δ9-tetrahydrocannabinol (11-OH-THC)Whole bloodPlasmaLiquid-chromatography/mass spectrometry (LC/MS)Pulmonary administrationRat

Introduction:Most studies evaluating the effects of Δ9-tetrahydrocannabinol (Δ9-THC) in animal models admin-ister it via a parenteral route (e.g., intraperitoneal (IP) or intravenous injection (IV)), however, the common routeof administration for human users is pulmonary (e.g., smoking or vapourizing marijuana). A vapourized Δ9-THCdelivery system for rodents was developed and used to compare the effects of pulmonary and parenteral Δ9-THC administration on blood cannabinoid levels and behaviour.Methods: Sprague–Dawley rats were exposed to pulmonary Δ9-THC (1, 5, and 10 mg of inhaled vapour) deliv-ered via a Volcano® vapourizing device (Storz and Bickel, Germany) or to parenteral Δ9-THC (0.25, 0.5, 1.0, and1.5 mg/kg injected IP). Quantification of Δ9-THC and its psychoactive metabolite, 11-hydroxy-Δ9-THC (11-OH-Δ9-THC), in blood was determined by liquid chromatography/mass spectrometry (LC/MS). In order to verify thepotential for the vapourization procedure to produce a robust conditioned place preference (CPP) or conditionedplace avoidance CPA, classical conditioning procedures were systematically varied by altering the exposure time(10 or 20 min) and number of exposed rats (1 or 2) while maintaining the same vapourization dose (10 mg).

Results: Blood collected at 20 min intervals showed similar dose-dependent and time-dependent changes inΔ9-THC and 11-OH-Δ9-THC for both pulmonary and parenteral administration ofΔ9-THC. However, vapourizedΔ9-THC induced CPP under certain conditions whereas IP-administered Δ9-THC induced CPA.Discussion: These results support and extend the limited evidence (e.g., in humans, Naef et al., 2004; in rodents,Niyuhire et al., 2007) that Δ9-THC produces qualitatively different effects on behaviour depending uponthe route of administration.

© 2014 Elsevier Inc. All rights reserved.

1. Introduction

Studies of the rewarding and addictive properties of cannabinoidswhich use rodents as animal models of human behaviour have often

ol; (11-OH-Δ9-THC), 11-hy-yl ethanolamide; (2-AG), 2-chemical ionization; (CV),

ion; (CPP), conditioned placeSI), electrospray ionization;lase; (GABA), γ-aminobutyricance liquid-chromatography;

LCMS), liquid chromatographyselected ion monitoring., Toronto, ON,M5S 2S1, Canada.

well@gmail.com (L.A. Manwell).

failed to replicate many findings from human studies. Since animalstudies typically employ parenteral routes of administration (e.g., intra-venous (IV) or intraperitoneal (IP) injection), whereas humans typicallysmoke cannabis, it is possible that the discrepancies may be related todifferent pharmacokinetics of parenteral and pulmonary routes ofadministration. Studying the effects of cannabinoids in animal modelsto better understand the development of drug use, abuse, and depen-dence in humans should account for the fact that marijuana is typicallyinhaled by human users, either by combustion (smoking) orvapourization. More than 400 different constituents of Cannabis sativahave been identified, at least 60 of which are phytocannabinoids(Turner, Bouwsma, Billets, & Elsohly, 1980; Turner, Elsohly, & Boeren,1980); a small proportion of these are classified as psychoactive, mean-ing that these compounds cross the blood–brain barrier (BBB), actingdirectly at the level of the central nervous system (CNS), and alterbrain function in ways that affect consciousness, mood, perception

121L.A. Manwell et al. / Journal of Pharmacological and Toxicological Methods 70 (2014) 120–127

and behavior (Segal & Duffy, 1999). Of these, the primary psychoactivecannabinoid is Δ9-tetrahydrocannabinol (Δ9-THC) which exerts its ef-fects in the central nervous system via binding to cannabinoid receptors(CBRs) (Devane, Dysarz, Johnson, Melvin, & Howlett, 1988; Dewey,1986; Mechoulam, 1970). Pulmonary administration (inhalation) ofcannabis produces the greatest bioavailable levels of Δ9-THC, withdose-dependent blood levels peakingwithinminutes and subjective ef-fects occurring almost immediately; in comparison, subjective effectsare delayed from 30 to 120min after both parenteral (injected) and en-teral (oral) administration, which result in significantly lower peakblood levels due to considerable first-pass liver metabolism, where itis degraded by enzymes of the cytochrome P450 system (Matsunaga,Iwawaki, Watanabe, et al., 1995; reviewed in Grotenhermen, 2003;Koob & Le Moal, 2006). During metabolism, hydroxylation of Δ9-THCproduces 11-OH-Δ9-THC, which is also psychoactive; although theelimination half-life for Δ9-THC is estimated within the range of20–60 h, the elimination half-life for 11-OH-Δ9-THC is much longer,lasting up to 5–6 days (reviewed in Grotenhermen, 2003; Koob & LeMoal, 2006). Tolerance to the effects of Δ9-THC has been demonstratedexperimentally in both animals and humans; pharmacodynamictolerance is largely attributed to brain neuroadaptation (i.e., changesin the endocannabinoid (ECB) system), as opposed to pharmacokineticchanges (i.e., changes in disposition, absorption, or metabolism of theparent compound) (Dewey, 1986; Hunt & Jones, 1980; Maykut, 1985).Tolerance that develops rapidly after a few doses diminishes quickly,whereas tolerance to large doses persists, even after drug use hasstopped. Cessation after chronic marijuana use can produce a with-drawal syndrome in humans – with a typical onset of 1–3 days post-cessation – including symptoms of irritability, restlessness, mildagitation, sleep electroencephalography (EEG) disturbance, insomnia,nausea and cramping (O'Brien, 1996). Psychological and physicaldependence have been reported in both animals and humans afterrepeated administration of marijuana (Kobayashi et al., 1999; Koob &Le Moal, 2006).

Since the combustion of cannabis releasesmany compounds in addi-tion to Δ9-THC, including other cannabinoids (e.g., cannabidiol) andmany toxins and carcinogens (e.g., anthrocyclines, nitrosamines, poly-cyclic aromatic hydrocarbons, terpenes, and vinyl chloride) (Aldingtonet al., 2008; Berthiller et al., 2008; Hashibe et al., 2005; Roth et al.,2001; Sarafian et al., 1999; Turner, Bouwsma et al., 1980; Turner,Elsohly et al., 1980; Voirin et al., 2006; Zhang et al., 1999; also reviewedin Reece, 2009) which may exert their own effects, vapourization ofpure Δ9-THC provides a more accurate assessment of the direct ef-fects of Δ9-THC and its primary psychoactive metabolite, 11-OH-Δ9-THC. In the current study, a novel delivery system of vapourized Δ9-THCwas developed and assessed for its pharmacokinetic, pharmacody-namic, and behavioural effects in rodents. A commercially availablevapourizer, commonly used by cannabis smokers, was used to assessthe effects of pulmonary administration of Δ9-THC and directly com-pared to parenteral administration of Δ9-THC.

2. Materials and method

2.1. Animals

Male Sprague–Dawley rats (Charles River, Canada) (N = 148),weighing 200–300 g, were used in all experiments. Rats were pair-housed in standard plastic shoebox cages (45 × 25 × 20 cm3) main-tained at 21–22 °C in a colony room on a 12-h reversed light–darkcycle (lights off at 0700 h) and fed standard rat chow (Harlan 8640)and water ad libitum. Testing was conducted during the dark cycle.Experimental procedures followed the Canadian Council on AnimalCare guidelines and were approved by the Wilfrid Laurier UniversityAnimal Care Committee. Rats were acclimatized to the colony and han-dling procedures prior to experimentation.

2.2. Drugs and analytical chemicals

Δ9-THC (Dronabinol, N98% purity) was obtained from THC PharmGmbH (Frankfurt, Germany). For experiments involving pulmonary (in-haled) drug administration,Δ9-THCwas prepared in ethanol at concen-trations of 4, 8, 20, 40, and 80 mg/ml and 250 μl of each was applied tosmall steel wool pads (liquid pads, Storz & Bickel, Tuttlingen, Germany),yielding final amounts of 1, 2, 5, 10, and 20 mg/pad. The ethanol wasthen allowed to evaporate before vapourization. For experiments in-volving i.p. administration, Δ9-THC was dissolved in a small volume ofethanol and then mixed with TWEEN-80 (polyoxyethylene sorbitanmonooleate; ICN Biomedicals). The ethanol was evaporated under astream of nitrogen gas, and the suspension was then mixed with phys-iological saline. The final vehicle contained 15 μl TWEEN-80 per 2 mlsaline. Δ9-THC was administered in a volume of 1 ml/kg body weight.

Internal standards for Δ9-THC (1.0 mg/ml), 11-OH-Δ9-THC(100 μg/ml), Δ9-THC-D3 (100 μg/ml), and 11-OH-Δ9-THC-D3(100 μg/ml) were obtained from Cerilliant (Round Rock, Texas,USA). Each standard was diluted in methanol to working concentra-tions of 5 to 600 ng/ml and stored at 4 °C. All solvents were of analyt-ical or HPLC grade; water, acetonitrile, methanol, ethyl acetate, andhexane were obtained from VWR and Fisher Scientific. All bloodsamples were collected in ethylenediaminetetraacetic acid (EDTA)treated glass test tubes (BD Vacutainer tubes). Solid phase extraction(SPE) silica-bonded (Si) C18 (Bond Elut) flash columns were obtainedfrom Varian Canada Inc. (Mississauga, Ontario, Canada).

2.3. Vapourization administration apparatus

A Volcano® vapourizing device (Storz and Bickel, GmbH and Co.,Tuttlingen, Germany) was used as described by Hazekamp, Ruhaak,Zuurman, van Gerven, and Verpoorte (2006). Briefly, Δ9-THC (1, 5, or10 mg/pad) was vapourized (at heat setting 9, approximately 226 °C)and collected into detachable plastic balloons (approximately 8 l),which were then fitted with a release valve that expelled the vapourover 10 s into enclosed plastic boxes (45 × 25 × 20 cm3) containingtwo rats (randomized to drug treatment group) that inhaled the vapourfor 5 min. Animals were removed 5 min later and either sacrificed atregular intervals for determination of blood cannabinoid levels orplaced in the place conditioning apparatus for behavioral testing. Thisdevice has been previously reported to deliver N50% of the loaded Δ9-THC into the balloon in a reproduciblemannerwith a pulmonary uptakecomparable to smoking of cannabis in humans (Hazekamp et al., 2006).

2.4. Place conditioning apparatus

The conditioning apparatus consisted of a black acrylic rectangularbox (60 × 25 × 25 cm3) with a wire-mesh lid. During conditioning trials,tactile cues on both sides of the boxwere identical. During the pretest andchoice tests, one side of the chamber had a plasticfloorwith raised bumpsand the other side had a plastic floor with holes (counterbalanced); asmall plastic smoothfloor, defined as a neutral zone (9× 25 cm), separat-ed the two cued floors. The amount of time (s) each rat spent on each ofthe floorswas recorded and subsequently analyzed by the ANY-maze be-havioural video-tracking software (Stoelting, Wood Dale, Illinois). Pre-tests did not indicate a significant difference between time spent onthe plastic bumps or holes floors indicating that the apparatus providesan unbiased test of conditioned preference/aversion.

2.5. Experimental procedures

2.5.1. Pharmacokinetic analyses of Δ9-THC and 11-OH-Δ9-THC in wholeblood after vapourized and IP administration of Δ9-THC

2.5.1.1. Cannabinoid extraction procedure. Calibration standards wereprepared fresh in methanol daily and stored on ice. After pulmonary

122 L.A. Manwell et al. / Journal of Pharmacological and Toxicological Methods 70 (2014) 120–127

or parenteral administration of Δ9-THC, blood samples were obtainedby cardiac puncture performed on rats (N = 40) anesthetized withisoflurane gas. Immediately upon collection, 1 ml of blood was addedto iced test tubes, already containing standards inmethanol, then gentlymixed and allowed to stand for approximately 1 h on ice (all steps wereconducted on ice unless otherwise indicated). Control and calibrationsamples were prepared by spiking 1 ml drug-free rat blood with stan-dard solutions: concentrations of Δ9-THC and 11-OH-Δ9-THC standardsranged from 0.5 to 600 ng/ml, whereas DIs were consistently added inamounts of 45, 75 or 150 ng/ml depending upon the range of standardconcentrations (e.g., low, medium, or high range). After equilibration,blood samples were acidified with 200 μl of 10% acetic acid and gentlyvortex mixed for 15 s. Proteins were precipitated by adding 2 ml aceto-nitrile, gently vortex mixing for 30 s, equilibrating for 3 min, andcentrifuging for 20 min at 10,000 ×g at 4 °C. All supernatant fractionswere removed using glass pipettes and cannabinoids extracted by LLE.At approximately a 2:1 (v/v) amount, 4 ml of 0.1 M acetate buffer(pH 4.5) was added to the supernatant fraction, vortex mixed for 15 s,and followed by the addition of 3 ml of hexane/ethyl acetate (95:5, v/v),vortex mixed for 30 s, and centrifuged for 20 min at 10,000 ×g at 4 °C.Supernatant fractions were removed, placed into small glass test tubes,and then evaporated at room temperature under nitrogen. Residueswere reconstituted in 3 ml hexane and cannabinoids were selectivelyseparated by SPE. Bond Elut columns were preconditioned with three3 ml volumes of hexane and samples were applied to columns usingglass pipettes and then washed with another three 3 ml volumes ofhexane. Δ9-THC was eluted from columns with three 3 ml volumes ofhexane/ethyl acetate (95:5, v/v) and 11-OH-Δ9-THC was eluted withthree 3ml volumes of hexane/ethyl acetate (75:25, v/v); an interveningfraction of three 3 ml volumes of hexane/ethyl acetate (90:10, v/v) wasused to remove contaminants. All fractions were kept separate for theremaining steps, including LC/MS analysis. Eluted samples were evapo-rated at room temperature under nitrogen and reconstituted in 200 μlmethanol.

2.5.1.2. Liquid chromatography. Chromatography was performed oneither aWaters YMC Pro C18 (2.0 × 150mm, 3 μm) column or aWatersYMC ODS-AQ (2.0 × 150 mm, 3 μm) column using Agilent 1100system's binary pumps equipped with an autosampler. The gradientstarted at a 70/30 v/v methanol/water mobile flow rate of 0.5 ml/minand a nitrogen flow rate of 5.00 ml/min. After 1 min it linearly changedto 100%methanolwhich took 6min and continued for another 6min. Atmin 14 of the run, the solvent ratio linearly returned to the initialconditions over the course of 5 min and equilibrated for another5 min. The eluent was additionally monitored with a single wavelengthdetector setup at 205 nm.

2.5.1.3. Mass spectrometry. The LC/MS system consisted of an Agilent1100 Series — LC–MSD with a single quadrupole mass analyzer. AnAgilentmultimode ionization sourcewas used to provide the combinedAPCI and ESI. The drying gas nitrogen flow was set at 5 l/min at 300 °Cand the nebulizer gas pressure was set at 40 psi. Capillary voltage wasset at 2000 V, with the corona current at 1 μA, in order to keep optimalconditions for ESI and APCI. Themass spec was set in negative ionmodeto scan from 200 to 400 m/z. A post-run extracted ion chromatogramwas reconstructed at m/z 313 [Δ9-THC-H]−, m/z 329 [11-OH-Δ9-THC-H]−, m/z 316 [Δ9-THC-D3-H]−, and m/z 332 [11-OH-Δ9-THC-D3-H]−.

2.5.1.4. Validation. For determination of intra- and inter-day accuracyand precision of the assays, calibration standards were added to 1 mlof whole blood using two ranges of Δ9-THC and 11-OH-Δ9-THC (4.5, 9,18, 37.5, 75, 150, 300, 600 ng/ml; and 0.5, 1, 2, 5, 10, 2, 50, 100,200 ng/ml) and consistent amounts of Δ9-THC-D3 and 11-OH-Δ9-THC-D3 (45, 75, or 150 ng/ml) depending upon the range. Accuracy wasexpressed as the mean % [(mean measured concentration)∕(expectedconcentration) × 100]. Precision was calculated as intra- and inter-day

coefficient of variation [% CV = (S.D. ∕ mean) × 100]. The LOQ wasdetermined as defined by Valiveti and Stinchcomb (2004) and was theconcentration of Δ9-THC and 11-OH-Δ9-THC that could be obtainedwith acceptable precision (%CV b 10). Matrix effects (co-elution ofcompounds that could affect analyte ionization) were determined byextracting “blank” whole blood and reconstituting with methanol con-taining a known amount of the analytes. Reconstituted extracts wereanalyzed and their peak areas were compared to those of the analytesadded directly to pure methanol. Initial analyte recoveries were per-formed in triplicate, for every experimental run, in whole blood spikedwith Δ9-THC and 11-OH-Δ9-THC and DIs then compared with analytepeak areas of Δ9-THC and 11-OH-Δ9-THC and DIs added directly tomethanol.

2.5.2. Place conditioning

2.5.2.1. Habituation. Rats were habituated to the vapour administrationcompartments for 5 min on four days prior to experimentation tobecome familiar with the apparatus and reduce confinement stress.Rats assigned to the vapourized Δ9-THC groups were habituated tothe exposure chamber three days preceding experimentation in in-creasing intervals of time: 30 s, 2 min, and 5 min respectively. Ratswere not, however, habituated to the place conditioning chambers.The vehicle group controlled for any novelty-induced behaviour.

2.5.2.2. Comparison of vapourized and IP administered Δ9-THC-inducedplace conditioning versus morphine-induced place conditioning. Rats(N= 72) were handled for six consecutive days before commencingexperimental testing in 6 equal groups (n = 12) as follows: vehicle,5.0 mg/kg morphine, 1.5 mg/kg IP THC, 2 mg vapourized Δ9-THC,10mgvapourizedΔ9-THC, and 20mgvapourizedΔ9-THC. Rats assignedto vapourized vehicle or Δ9-THC groups were exposed to 8 l of controlvapour (0 mg Δ9-THC) for 5 min in the exposure chambers. Ratsassigned to the IP injection controlled groups were habituated with asaline injection (1 ml/kg IP) one day prior to experimentation.

The experiment was conducted over 18 days in three phases: pre-conditioning, conditioning, and testing. The preconditioning phaseconsisted of a 15 min drug-free session during which each rat wasplaced in the test chamber equipped with both types of floors: bumpsand holes, separated by a smooth plastic strip placed in the middle ofthe floor types. At preconditioning, time spent on the bumps versusholes floor types was analyzed to confirm that rats had no bias for afloor type. Rats assigned to the inhaled Δ9-THC group were placed inthe test chamber immediately after the 5 min vapour exposure. Ratsinjected with 5.0 mg/kg morphine or 1.5 mg/kg THC were placed inthe chambers 5 min and 30 min post-injection respectively. Timespent on each floor type was tracked and recorded with ANY-maze.The conditioning phase consisted of four 30-min sessions, each lastingtwo days, separated by 1–2 days of non-testing between conditioningcycles. Each conditioning cycle consisted of a drug testing day and avehicle day. The same floor type was consistently paired with eitherthe drug (i.e., the drug-paired floor or zone) or vehicle (i.e., the vehicle-paired floor or zone) administration throughout the conditioning phase.The test phase was identical to the preconditioning session where eachrat received a 15-min drug-free test session, and time spent on eachfloor type was recorded. The order of treatment, session of drug versusvehicle, and floor types were all counterbalanced across rats for the dura-tion of the experiment.

2.5.2.3. Comparison of three vapourized Δ9-THC place conditioningprocedures. In order to verify that the vapourized Δ9-THC procedurehad the potential to produce a CPP or CPA, the procedure was sys-tematically varied by altering the exposure time (10 or 20 min) andnumber of exposed rats (1 or 2) while maintaining the samevapourization dose (10 mg). Rats (N = 36) were habituated to vapourexposure chambers and activity monitors for 3 days prior to

123L.A. Manwell et al. / Journal of Pharmacological and Toxicological Methods 70 (2014) 120–127

experimentation. The experiment consisted of 3 phases, precondition-ing, conditioning, and testing. The preconditioning phase consisted ofa 15-min no-drug session, where rats were placed in activity monitorswith both types of floors (bumps and holes) separated by a smoothplastic strip. Time spent on each floor type was analyzed to confirmthat there was no predetermined bias towards one particular floor tex-ture. During the conditioning phase, three different procedures wereemployed. In Procedure 1 (n = 12), one rat was exposed to 16 l ofvapourized Δ9-THC (10 mg) or vehicle per vapour exposure chamber,over 10 min (4 l administered every 2.5 min; 1.6 l/min). For Procedure2 (n= 12), one rat was exposed to 16 l of vapourized Δ9-THC (10 mg)or vehicle per chamber, over 20 min (4 l administered every5 min; 0.8 l/min). While in Procedure 3 (n = 12), two rats wereexposed to 8 l of vapourized Δ9-THC (10 mg) or vehicle per chamber,over 10 min (2 l was administered every 2.5 min; 0.8 l/min). Aftereach exposure, rats were placed immediately into activity monitors.The conditioning cycle consisted of eight 30-min sessions per day, withdrug and vehicle administration occurring on even and odd days respec-tively. The same floor types were consistently pairedwith either drug orvehicle throughout conditioning and time spent on each floor type wasrecorded. The test phase consisted of a 15-min no-drug session, whererats were placed in activity monitors with both types of floors (bumpsand holes) separated by a smooth plastic strip. The order of treatment,drug or vehicle sessions, andfloor types were counterbalanced through-out the experiment.

3. Analyses and results

3.1. Pharmacokinetic analyses of Δ9-THC and 11-OH-Δ9-THC in wholeblood after vapourized and IP administration of Δ9-THC

Themass spectra ofΔ9-THC (m/z 313.2) andΔ9-THC-D3 (m/z 316.0)and themass spectra of 11-OH-Δ9-THC (m/z 329.2) and11-OH-Δ9-THC-D3 (m/z 332.4) were identified and no interfering peaks were observedin conjunction with the drug peaks. For analytes spiked directly intopure methanol only, retention times were very similar on bothcolumns: on the Waters YMC ODS-AQ column, retention times forΔ9-THC, Δ9-THC-D3, 11-OH-Δ9-THC, and 11-OH-Δ9-THC-D3 were12.580, 12.556, 10.887, and 10.874 min, respectively, and on theWaters Pro YMC column, retention times for Δ9-THC, Δ9-THC-D3,11-OH-Δ9-THC, and 11-OH-Δ9-THC-D3 were 12.231, 12.249, 10.686,and 10.573 min, respectively. Similarly, for analytes spiked directlyinto whole blood, retention times were similar for both columns:retention times for Δ9-THC, Δ9-THC-D3, 11-OH-Δ9-THC, and 11-OH-Δ9-THC-D3 on the Waters YMC ODS-AQ column were 12.580,12.556, 10.887, and 10.874 min, respectively, and on the Waters ProYMC C18 column were 12.355, 12.279, 10.598, and 10.571 min, respec-tively. Both columns have similar octadecylsilane stationary phasesand behave similarly in terms of peak resolution and column bleeding.Calibration graphs were constructed using a linear regression of thepeak-areas of a) Δ9-THC product ions divided by Δ9-THC-D3 ions andb) 11-OH-Δ9-THC product ions divided by 11-OH-Δ9-THC-D3 ions. Stan-dard calibration curves for both Δ9-THC and 11-OH-Δ9-THC in wholeblood were linear over the range of 5 to 600 ng/ml (means of n = 3each). Mean absolute recoveries from whole blood of Δ9-THC and 11-OH-Δ9-THC were 108.96% (%CV = 5.94, n = 6) and 100.85% (%CV =7.77, n=6), respectively; for both cannabinoids, the limits of quantifica-tion (LOQ) and detection (LOD) were 5 ng/ml and 2 ng/ml, respectively.No significant matrix effects were observed for the analytes in wholeblood; extracts from blank whole blood samples showed no interferingpeaks during method development (total n N 100) or in batch-to-batchstability analyses (minimum n= 3 for each run). The mean CV of peakareas of reconstituted samples was b10% for both Δ9-THC and 11-OH-Δ9-THC indicating that the analytes' ionization was not affected by co-eluting compounds. Results of the intra-day and inter-day validation as-says for Δ9-THC and 11-OH-Δ9-THC indicated that no significant

degradation of samples occurred for samples left in the autosampler atroom temperature for at least 24 h. Representative blood concentrationsare shown in Table 1 and Figs. 1 and 2; whole blood collected at 20 and40 min from rats receiving Δ9-THC either through pulmonary adminis-tration (1, 5, and 10 mg inhaled vapour) or parenteral administration(0.25, 0.50, 1.0, and 1.5 mg/kg injected i.p.) showed both dose-dependent and time-dependent concentration changes in Δ9-THC andits metabolite 11-OH-Δ9-THC. Replication of these procedures andresults are reported in Manwell (2013).

3.2. Place conditioning

3.2.1. Comparison of vapourized and IP administered Δ9-THC-inducedplace conditioning versus morphine-induced place conditioning

For place conditioning, time spent in the drug-paired zone, timespent in the vehicle-paired zone, and time spent in the center zonewere all analyzed using a two-factor (group by phase) analysis of vari-ance (ANOVA) for the vapourized THC treatments, where phase repre-sented pretest versus posttest and group represented the four inhaleddoses: Vehicle, 2 mg, 10 mg, and 20 mg vapourized Δ9-THC. Plannedcomparisons examined CPP separately for each level of treatmentusing a paired-sample t test. The six contrasts of interest (α = 0.05)were: (1) 5.0 mg/kg morphine vs. vehicle, (2) 1.5 mg/kg Δ9-THC vs.vehicle, (3) Vehicle vs. vehicle, (4) 2 mg Δ9-THC vs. vehicle, (5) 10 mgΔ9-THC vs. vehicle, (6) 20 mg Δ9-THC vs. vehicle. Planned comparisonsexamined each of the six treatment groups using a two-way repeatedmeasures (treatment by day) ANOVA. ANOVAs and planned compari-sons were conducted using SPSS 13 for Mac OS X.

Fig. 3 shows data for place conditioning induced by vapourizedΔ9-THC, injected Δ9-THC, morphine and vehicle controls. As expected,administration of morphine (5 mg/kg, IP) during conditioning inducedCPP; morphine-treated animals showed a significant increase in thepreference for the drug-paired floor compared to the vehicle-pairedfloor. Rats showed a significant increase in time spent on the drug-paired zone frompre-test to post-test [t(10)= 5.47, p b 0.001], a signif-icant decrease in time spent on the vehicle-paired floor from pre-test topost-test [t(10) = 5.84, p b 0.001], and no significant difference be-tween time spent on the center (neutral) floor from pre-test to post-test [t(10) = 0.84, p N 0.05]. The significant CPP induced by 5 mg/kgmorphine suggested that the procedural parameters used in thispresent study were sufficiently sensitive for detection of place pref-erence for this positive control. Also, as expected, Δ9-THC injection(1.5 mg/kg, IP) significantly decreased preference for the drug-pairedfloor; rats showed a significant decrease in time spent on the drug-paired side from pre-test to post-test [t(11) = 3.11, p b 0.05], a signifi-cant increase in time spent on the vehicle-paired floor from pre-testto post-test [t(11) = 2.66, p b 0.05], and no significant differencebetween time spent on the center (neutral) floor from pre-test topost-test.

Two-way ANOVAs conducted for time spent on the drug-paired side,time spent on the vehicle-paired floor, and time spent on the center(neutral) floor revealed no significant effects of phase or a phase bygroup interaction for vapourized Δ9-THC treatment groups. The lack ofsignificant difference in time spent on the center floor between pre-test and post-test demonstrates that there was no bias for the novelcenter floor present at posttest, since the centre zonewas absent duringconditioning trials. Planned comparisons examined CPP separately foreach level of inhaled treatment group using a paired-sample t-test fortime spent on the drug-paired floor, time spent on the vehicle-pairedfloor, and time spent on the center (neutral) floor. However, no signifi-cant effects of phase were seen for all levels of the vapourized Δ9-THCgroups. Consequently, there was no difference between post-test andpre-test at all levels of vapourized Δ9-THC groups, thus all vapourizedΔ9-THC groups failed to produce either a place avoidance or a placepreference.

Table 1Comparison of Δ9-THC and 11-OH-Δ9-THC concentrations in whole blood after Δ9-THC administration. Each row represents the concentrations of Δ9-THC and 11-OH-Δ9-THC recoveredfrom whole blood derived from a single rat exposed to one of seven doses.

Pulmonary administration (1, 5, and 10 mg vapourized) Parenteral administration (1.0, 1.5, and 2.0 mg/kg injected i.p.)

Dose(mg/pad)

Time(min)

Recovered(ng/ml/kg)

Concentration(ng/ml/kg)

Dose(mg/kg)

Time(min)

Recovered(ng/ml/kg)

Concentration(ng/ml/kg)

Δ9-THC 11-OH-Δ9-THC Δ9-THC 11-OH-Δ9-THC

1 20 157.50 86.45 0.25 20 151.37 353.7040 155.67 120.27 40 173.17 244.86

5 20 223.06 173.19 0.50 20 210.54 357.5740 209.38 219.86 40 244.47 267.34

10 20 402.31 496.26 1.0 20 244.46 410.8740 301.06 233.04 40 395.33 294.82

1.5 20 349.33 445.9540 410.89 402.96

124 L.A. Manwell et al. / Journal of Pharmacological and Toxicological Methods 70 (2014) 120–127

3.2.2. Comparison of three vapourizedΔ9-THC place conditioning proceduresFor place conditioning, time spent in the drug paired zone was ana-

lyzed by a two-factor (procedure × phase) ANOVA, where phase refersto the preconditioning versus test phase. Planned comparisons exam-ined place preference for each procedure using paired-sample t tests.Planned two-way repeated measures (treatment × day) ANOVA com-parisons were conducted for each procedure. Bonferroni-adjustedt-tests were used to compare conditioning days. Paired samples t-testswere used to compare treatment across days. Analyses were conductedusing SPSS 13.0 for Windows (SPSS Inc., Chicago, IL) (α = 0.05).

Time (min)4020

11-O

H-T

HC

Con

cent

ratio

n (n

g/m

l/kg)

0

100

200

300

400

500

600 THC 10 mg vapourized THC 5 mg vapourized THC 1 mg vapourized

Time (min)4020

THC

Con

cent

ratio

n (n

g/m

l/kg)

0

100

200

300

400

500

600 THC 10 mg vapourized THC 5 mg vapourized THC 1 mg vapourized

Fig. 1. Dose- and time-dependent concentrations of Δ9-THC (top) and 11-OH-Δ9-THC(bottom) recovered from whole blood derived from a single rat exposed to one of threedoses of vaporized Δ9-THC (1, 5, or 10 mg) at a 20 min interval.

Fig. 4 shows place conditioning for each of the vapourized Δ9-THCprocedures. Procedure 1 produced a clear place preference, while Proce-dure 2 demonstrated a small place avoidance. Procedure 3 did notinduce conditioned place preference or avoidance. The two-factor(procedure x phase) ANOVA revealed no significant main effect ofprocedure, F(2, 33) = 1.01, ns, nor phase, F(1, 33) b 1, ns. The proce-dure × phase interaction was also not significant, F(1, 33) =2.09, ns.Since no groups indicated a preference or aversion to any floor types

Time (min)4020

THC

Con

cent

ratio

n (n

g/m

l/kg)

0

100

200

300

400

500

600

THC 0.25 mg/kg injected (i.p.) THC 0.50 mg/kg injected (i.p.) THC 1.00 mg/kg injected ( i.p.) THC 1.50 mg/kg injected (i.p.)

Time (min)4020

11-O

H-T

HC

Con

cent

ratio

(ng/

ml/k

g)

0

100

200

300

400

500

600THC 0.25 mg/kg injected (i.p.) THC 0.50 mg/kg injected (i.p.) THC 1.00 mg/kg injected (i.p.)THC 1.50 mg/kg injected (i.p.)

Fig. 2. Dose- and time-dependent concentrations of Δ9-THC (top) and 11-OH-Δ9-THC(bottom) recovered from whole blood derived from a single rat exposed to one of fourdoses of injected Δ9-THC (0.25, 0.50, 1.00, or 1.50 mg/kg) at a 20 min interval.

Fig. 3. Mean (±SEM) change in time (s) spent on the drug-paired side from the precondi-tioning session to the test session. VEH = vehicle; 5 MOR= dose of injected morphine inmg/kg; 1.5 THC = dose of injected Δ9-THC in mg/kg; 2 THC = dose of vapourized Δ9-THCfor 2mg; 10 THC=dose of vapourizedΔ9-THC for 10mg; and 20 THC=dose of vapourizedΔ9-THC for 20 mg. ⁎⁎p b 0.001, ⁎p b 0.05, different from the vehicle control group.

125L.A. Manwell et al. / Journal of Pharmacological and Toxicological Methods 70 (2014) 120–127

during the preconditioning phase, it was hypothesized a priori thatone or more groups would show a preference or aversion for thedrug-paired floor compared to the vehicle paired floor after condi-tioning. Thus, planned comparisons using dependent t-tests were con-ducted for each procedure to determine if there were any direct effectsof place conditioning. Procedure 1 produced a significant place prefer-ence as shown by a significant increase in time spent on the drug-paired side from the pre-conditioning to the test phases, t(11) =2.28,p b .05. Neither Procedure 2 nor 3 resulted in significant place prefer-ences, t(11) = 1.76, ns, t(11) = 1.87, ns, respectively. In order todetermine if there was a stable baseline of time spent of the drug pairedside across the three procedures, a one-way ANOVA was used to com-pare groups. Results revealed no significant preference across proceduregroups, F(2, 33) b 1, ns.

4. Discussion

The development of a novel Δ9-THC vapourization method forrodent behavioural neuropharmacology had two main objectives.First, we developed a sensitive and selective method of simultaneousdetection and quantification Δ9-THC and its psychoactive metabolite

Fig. 4.Mean (±SEM) change in time (s) spent on the drug paired side from precondition-ing phase to test phase. Procedure 1 (n = 12) = 10 min exposure of 16 l of 10 mgvapourized Δ9-THC to one rat per vapour exposure chamber; Procedure 2 (n = 12) =20 min exposure of 16 l of 10 mg vapourized Δ9-THC to one rat per vapour exposurechamber; Procedure 3 (n = 12) = 10 min exposure of 8 l of 10 mg vapourized Δ9-THCto two rats per vapour exposure chamber. *p b 0.05, significantly different from vehicle.

11-OH-Δ9-THC in blood after exposure to vapourized or IP administeredΔ9-THC. Second, we employed this method to assess the behaviouraleffects of vapourized and IP administeredΔ9-THC onplace conditioning.Our findings demonstrate that the route of administration of Δ9-THCproduces different levels of Δ9-THC and 11-OH-Δ9-THC in blood anddifferent behavioural effects for place conditioning in rodents. We spe-cifically chose time points of 20 and 40min post-exposure for determi-nation of blood cannabinoid levels because these are key time points inthe behavioural experiments reported here in Part I and in Part II. Forexample, in the behavioural conditioning studies (e.g., Figs. 3 and 4),rats were exposed to Δ9-THC (pulmonary or parenteral) and testedapproximately 10 to 20min later for approximately 20 to 30min; there-fore, the 20 and 40min time points cover two key periods in the behav-ioural experiments.

We developed a reliable method of analyzing levels of Δ9-THC and11-OH-Δ9-THC in small amounts of whole blood from rats followingΔ9-THC administration via vapourization or IP-administration. Thismethod has high specificity, sensitivity, and selectivity for detectingpsychoactive cannabinoids in whole blood and plasma, using a min-imized protocol for extraction, detection, and quantification thatproduces an efficient pharmacokinetic analysis for the purposes ofcomparing biological levels of psychoactive cannabinoids in studiesof animal behaviour. Representative blood concentrations collected at20 min time intervals from rats receiving either vapourized Δ9-THC(1, 5, and 10 mg inhaled vapour) or IP-administered Δ9-THC (0.25,0.5, 1.0, and 1.5 mg/kg injected IP) showed both dose-dependentand time-dependent concentration changes in Δ9-THC and its metabo-lite 11-OH-Δ9-THC. These dose- and time-dependent changes in Δ9-THCare comparable to those reported by Wilson, Varvel, Harloe, Martin,and Lichtman (2006) in mice. It should be noted that doses required toproduce similar behaviours can vary between species on an order ofmagnitude of ten times ormore; amongst rodents,mice typically requirehigher dosages than rats because they have a higher rate of drugmetab-olism in general (reviewed in Murray & Bevins, 2010). Here, the objec-tive was to determine the initial dose(s) of parenterally-administeredΔ9-THC that would produce blood levels of both Δ9-THC and 11-OH-Δ9-THC equivalent to those produced by pulmonary-administered Δ9-THC. There were roughly equivalent mean blood concentrations afterexposure to vapourized and injected Δ9-THC; rats receiving eitherpulmonary-administered Δ9-THC (1, 5, and 10 mg inhaled vapour)or parenterally-administered Δ9-THC (0.25, 0.50, 1.0, and 1.5 mg/kginjected i.p.) showed both dose-dependent and time-dependent con-centration changes in Δ9-THC and its metabolite 11-OH-Δ9-THC. Ourresults are consistent with the sparse and narrow studies reported inthe literature evaluating pulmonary and/or parenteral administrationofΔ9-THC in rodents; of these studies, onlyWilson et al. (2002) directlycompared pulmonary and parenteral routes of administration.

First, similar levels of Δ9-THC and 11-OH-Δ9-THC were found inwhole blood in male Sprague–Dawley rats when there was an approxi-mate 6-fold increase in the amount of Δ9-THC that was deliveredthrough pulmonary administration compared to parenteral administra-tion. Wilson et al. (2002) also demonstrated that an approximate 6-foldincrease produced similar results inmalemice: at 20min post-exposure,meanΔ9-THCwhole blood levels were 409± 86 and 1132± 240 ng/mlfor 20 and 60mg of pulmonary-administeredΔ9-THC and 365±39 and1324 ± 38 ng/ml for 3 and 10 mg/kg parenterally-administeredΔ9-THC, respectively. However,Wilson et al. (2002) also reported a dis-sociation between Δ9-THC levels found in blood and brain that wasdependent upon the route of administration; although Δ9-THC bloodand brain levels were roughly equivalent following pulmonary admin-istration, Δ9-THC levels were 2–3-fold greater in brain than in bloodafter parenteral administration. Althoughwe only assessed blood levelsofΔ9-THC and 11-OH-Δ9-THC in rats, it is quite possible that brain levelsof these cannabinoids varied more widely than blood levels, similar toWilson et al.'s (2002) results in mice. In their conclusions, Wilsonet al. (2002, p. 264) acknowledged that, in earlier studies also in mice,

126 L.A. Manwell et al. / Journal of Pharmacological and Toxicological Methods 70 (2014) 120–127

employing parenteral routes of administration either increased ordecreased Δ9-THC in brain compared to blood; they cautioned that thedifference in results across studies, even in the same species, could beattributed in part to “different extraction procedures, drug absorptiontimes, the use of radiolabeled Δ9-THC rather than actual drug, data anal-yses and other methodological considerations” making direct compari-sons difficult.

Second, our results are consistent with those reported by Valiveti andStinchcomb (2004); the methods of LC/MS quantification of Δ9-THC inExperiment 1 were modeled primarily upon the methods reported intheir paper. Valiveti and Stinchcomb (2004) evaluated levels of Δ9-THCand 11-OH-Δ9-THC in rat and guinea pig plasma (and not whole blood)after parenteral administration of Δ9-THC; they did not assess the effectsof pulmonary administration of Δ9-THC. Valiveti and Stinchcomb (2004)reported maximal levels of Δ9-THC (194 ng/ml) and 11-OH-Δ9-THC(10 ng/ml) in rat plasma immediately after parenteral administration(e.g., an intravenous bolus dose) of 1 mg/kg Δ9-THC which plateauedafter approximately 1 h and maintained a steady state for approximately6 h. Using the same methodology, they reported maximal levels ofΔ9-THC (197.5 ng/ml) in guinea pig plasma which also plateauedafter 1 h, but reported nometabolites for up to 6 h, noting that guineapigs metabolize Δ9-THC differently than other mammals (Valiveti &Stinchcomb, 2004). Valiveti and Stinchcomb (2004) did not report onthe strain or sex of the rodents used in their experiments.

Third, our results for male Sprague–Dawley rats reported here arealso consistent with those for male Sprague–Dawley rats reported byTseng, Harding, and Craft (2004) using parenterally-administeredradiolabeled Δ9-THC. Tseng et al. (2004), who tested both male andfemale Sprague–Dawley rats, administered 5 μCi/10 mg/ml [3H]-Δ9-THC parenterally (i.p.) and reported blood serum levels of totalradiolabeled cannabinoids (Δ9-THC+ metabolites) that were equiv-alent for males and females of 2000 and 1000 DPM/ml serum at 15and 30 min post-administration respectively. In addition, brainlevels of Δ9-THC and metabolites in males peaked within 15 minand maintained a steady state similar to blood levels (Tseng et al.,2004). However, there was a significant dissociation between bloodand brain levels for females only: a) brain levels of Δ9-THC and metab-olites were significantly higher in females than males, and b) brainlevels in females peaked at 120 min and then steadily decreased(Tseng et al., 2004). When evaluated separately, levels of Δ9-THCwere not significantly different in brain tissue for males and females;the difference was attributable to the metabolite 11-OH-Δ9-THCwhich peaked once for males (at 30 min) and twice for females (at 20and 120 min) (Tseng et al., 2004).

Finally, in studies in humans, there is also significant variabilityacross routes of administration. For example, Naef, Russman, Petersen-Felix, and Brenneisen (2004) reported that in males and females, pul-monary administration of a Δ9-THC liquid aerosol (0.053 mg/kg) pro-duced peak plasma levels of Δ9-THC in the range of 18.7 ± 7.4 ng/mlfor approximately 10 to 20 min after inhalation, and then rapidly de-creased; however, plasma levels after parenteral (i.v.) administrationof Δ9-THC (0.053 mg/kg) were more variable, ranging from 81.6 to640.6 ng/ml (mean of 271.5 ± 61.1 ng/ml) within 5 min after injection(Naef et al., 2004). In addition, Naef et al. (2004) reported significantlylower levels of 11-OH-Δ9-THC after pulmonary administration, puta-tively due to lack of hepatic first pass metabolism. Naef et al. (2004)did not report any significant sex differences. The results of the currentstudy also highlight some of the problems arising from using differentadministration and extraction methods and important interspeciesmetabolic differences thatmake it difficult to relate the pharmacologicaland toxicological effects of Δ9-THC across species of rodents and evenextrapolate to humans (reviewed in Grotenhermen, 2003). Interspeciesdifferences in metabolism may be attributable to factors such as size,respiratory rate, body fat ratio, site of Δ9-THC metabolism and/or stor-age, and differences in isoenzymes of the cytochrome P450 (CYP) com-plex, which is largely responsible for hydroxylation and oxidation of

Δ9-THC in the liver (reviewed in Grotenhermen, 2003; Murray &Bevins, 2010). There are also important sex differences in the phar-macokinetic and behavioural effects of Δ9-THC between males andfemales (reviewed in Fattore and Fratta, 2010).

Our secondobjectivewas to directly compare thebehavioural effectsof vapourized and IP-administeredΔ9-THConplace conditioning in rats.In the place conditioning studies, the same procedures used to producea morphine-induced CPP also produced an IP-administered Δ9-THC-induced CPA, both of which are consistent with previous research(Cheer, Kendall, & Mardsen, 2000; Mallet & Beninger, 1998; Mucha,Bucenieks, O'Shaughnessy, & van der Kooy, 1982; Parker & Gillies,1995; Robinson, Hinder, Pertwee, & Riedel, 2003). It should be notedthat the CPA induced by IP-administeredΔ9-THC (1.5 mg/kg) is not con-sistent with Sañudo-Peña et al. (1997) or Vann et al. (2008) who failedto find a CPA (or CPP) at similar doses, and is indicative of the highly in-consistent findings of studies using parenterally-administered cannabi-noids. The same procedure used in the morphine and IP-administeredΔ9-THC place conditioning experiment failed to produce a vapourizedΔ9-THC-induced CPA or CPP at any of the three doses used. Becausethe methods used to employ a pulmonary route of Δ9-THC administra-tion are different than those used for parenteralΔ9-THC administration,a separate study was conducted comparing three different methods ofvapourizedΔ9-THC, all at the same dose (10mg),whichwas themediandose used in the prior place conditioning experiment. Thus, in order toverify that the vapourized Δ9-THC procedure had the potential to pro-duce a CPP or CPA, the procedure was systematically varied by alteringthe exposure time (10 or 20 min) and number of rats (1 or 2) whilemaintaining the same vapourization dose (10 mg). Procedures 1 and 2both exposed a single rat to 16 l of vapour from derived from a 10 mgdose ofΔ9-THCover a period of either 10min or 20min; Procedure 1 pro-duced a CPP, whereas Procedure 2 demonstrated a small non-significantCPA. Procedure 3wasmore similar to the procedure used in thefirst placeconditioning study: 8 l of vapour derived from a 10 mg dose of Δ9-THCwas expelled into a small chamber containing two rats over a period of10 min rather than just 5 min in the earlier experiment. Procedure 3replicated the failure to produce a CPP or CPA found in the firstplace conditioning experiment. Thus, the spatial (e.g., 1 or 2 rats perbox) and temporal (e.g., 5, 10 or 20 min exposure) parameters of thepulmonary administration method are as consequential as the dose ofΔ9-THC administered. The present results suggest that vapourized Δ9-THC administration can produce behavioural effects that are qualitative-ly different from those induced by IP administration in rodents.

Competing interests

The authors declare that they have no competing interests.

Author contributions

LM wrote the manuscript, designed and performed thevapourized and IP-administered cannabinoid quantification and experi-ments under the supervision of AC and DB, and designed and performedthe vapourized and IP-administered cannabinoid behavioural experi-ments under the supervision of PM; BM and HH performed the placeconditioning experiments; PM supervised all experiments and co-wrote the manuscript.

Acknowledgments

This work was performed with grants from Wilfrid Laurier (to PM)and the Natural Sciences and Engineering Research Council of Canada(to LM, BM, and PM). We would like to thank Dr. Masoud Jelokhani-Niakari of the Department of Chemistry at Wilfrid Laurier Universityfor providing comments on an earlier version of this article.

127L.A. Manwell et al. / Journal of Pharmacological and Toxicological Methods 70 (2014) 120–127

References

Aldington, S., Harwood, M., Cox, B., Weatherall, M., Beckert, L., Hansell, A., et al. (2008).Cannabis and Respiratory Disease Risk Group. Cannabis use and risk of lung cancer:A case–control study. European Respiratory Journal, 31, 280–286.

Berthiller, J., Straif, K., Boniol, M., Voirin, N., Benhaim-Luzon, V., Ayoub,W. B., et al. (2008).Cannabis smoking and risk of lung cancer in men: A pooled analysis of three studiesin Maghreb. Journal of Thoracic Oncology, 3, 1398–1403.

Cheer, J. F., Kendall, D. A., & Mardsen, C. A. (2000). Cannabinoid receptors and reward inthe rat: A conditioned place preference study. Psychopharmacology, 151, 25–30.

Devane, W. A., Dysarz, F. A. I., Johnson,M. R., Melvin, L. S., & Howlett, A.C. (1988). Determi-nation and characterization of a cannabinoid receptor in the rat brain. MolecularPharmacology, 34, 605–613.

Dewey, W. L. (1986). Cannabinoid pharmacology. Pharmacology Review, 38, 151–178.Grotenhermen, F. (2003). Pharmacokinetics and pharmacodynamics of cannabinoids.

Clinical Pharmacokinetics, 42, 327–360.Hashibe, M., Straif, K., Tashkin, D. P., Morgenstern, H., Greenland, S., & Zhang, Z. F. (2005).

Epidemiological review of marijuana use and cancer risk. Alcohol, 35, 265–275.Hazekamp, A., Ruhaak, R., Zuurman, L., van Gerven, J., & Verpoorte, R. (2006). Evaluation

of vaporizing device (Volcano®) for the pulmonary administration of tetrahydrocan-nabinol. Journal of Pharmaceutical Sciences, 95, 1308–1317.

Hunt, C. A., & Jones, R. T. (1980). Tolerance and disposition of tetrahydrocannabinol inman. Journal of Pharmacology and Experimental Therapeutics, 215, 35–44.

Kobayashi, H., Suzuki, T., Kamata, R., Saito, S. -Y., Sata, I., Tsuda, S., et al. (1999). Recentprogress in the neurotoxicology of natural drugs associated with dependence oraddiction, their endogenous agonists and receptors. Journal of Toxicological Sciences,24, 1–16.

Koob, G. F., & Le Moal, M. (2006). Opioids (chapter 4) and cannabinoids (chapter 7). Neuro-biology of addiction. London, UK: Elsevier Inc, 121–172 (289-338).

Mallet, P. E., & Beninger, R. J. (1998). Δ9-Tetrahydrocannabinol, but not endogenouscannabinoid receptor ligand anandamide, produces conditioned place avoidance.Life Sciences, 62, 2431–2439.

Manwell, L. A. (2013). Role of the endocannabinoid system in extinction of learned behavioursmotivated by opioid-induced reward and aversion in rats. (PhD Dissertation). NationalLibrary of Canada.

Matsunaga, T., Iwawaki, Y., Watanabe, K., et al. (1995). Metabolism of delta-9-tetrahydro-cannabinol by cytochrome P450 isozymes purified from hepatic microsomes of mon-keys. Life Sciences, 56, 2089–2095.

Maykut, M.O. (1985). Health consequences of acute and chronic marihuana use. Progressin Neuropsychopharmacology and Biological Psychiatry, 9, 209–238.

Mechoulam, R. (1970). Marihuana chemistry. Science, 168, 1159–1166.Mucha, R. F., Bucenieks, P., O'Shaughnessy, M., & van der Kooy, D. (1982). Drug reinforce-

ment studied by the use of place conditioning. Brain Research, 243, 91–105.Murray, J. E., & Bevins, R. A. (2010). Cannabinoid conditioned reward and aversion:

Behavioral and neural processes. American Chemical Society Chemical Neuroscience,1, 265–278.

Naef, M., Russman, S., Petersen-Felix, S., & Brenneisen, R. (2004). Development and pharma-cokinetic characterization of pulmonal and intravenous delta-9-tetrahydrocannabinol(THC) in humans. Journal of Pharmaceutical Sciences, 1176–1184.

Niyuhire, F., Varvel, S. A., Martin, B. R., & Lichtman, A. H. (2007). Exposure to marijuanasmoke impairs memory retrieval in mice. Journal of Pharmacology and ExperimentalTherapeutics, 322, 1067–1075.

O'Brien, C. P. (1996). Drug addiction and drug abuse. In J. G. Hardman, L. E. Lombird, P. B.,R. W. Ruddon, & A. E. Gillman (Eds.), In Goodman and Gilman's the pharmacologicalbasis of therapeutics (pp. 557–577) (9th ed.). New York: McGraw-Hill.

Parker, L. A., & Gillies, T. (1995). Δ9-THC-induced place and taste aversions in Lewis andSprague–Dawley rats. Behavioral Neuroscience, 109, 71–78.

Reece, A. S. (2009). Review: Chronic toxicology of cannabis. Clinical Toxicology, 47,517–524.

Robinson, L., Hinder, L., Pertwee, R. G., & Riedel, G. (2003). Effects of delta9-THC and WIN-55,212-2 onplace preference in thewatermaze in rats. Psychopharmacology, 166, 40–50.

Roth, M.D., Marques-Magallanes, J. A., Yuan, M., Sun, W., Tashkin, D. P., & Hankinson, O.(2001). Induction and regulation of the carcinogen-metabolizing enzyme CYP1A1bymarijuana smoke and delta(9)-tetrahydrocannabinol. American Journal of RespiratoryCell and Molecular Biology, 24, 339–344.

Sañudo-Peña, M. C., Tsou, K., Delay, E. R., Hohman, A. G., Force, M., & Walker, J. M. (1997).Endogenous cannabinoids as an aversive or counter-rewarding system in the rat.Neuroscience Letters, 223, 125–128.

Sarafian, T. A., Magallanes, J. A., Yuan, M., Sun, W., Tashkin, D. P., & Roth, M.D. (1999). Oxida-tive stress produced by marijuana smoke. An adverse effect enhanced by cannabinoids.American Journal of Respiratory Cell and Molecular Biology, 20, 1286–1293.

Segal, B., & Duffy, L. K. (1999). Biobehavioral effects of psychoactive drugs. In R. J. M.Niesink, R. M.A. Jaspers, L. M. W. Kornet, & J. M. Van Ree (Eds.), Drugs of abuse andaddiction: Neurobehavioral toxicology. (pp. 24–63)Boca Raton, FL: CRC Press (Ch. 2).

Tseng, A. H., Harding, J. W., & Craft, R. M. (2004). Pharmacokinetic factors in sex differ-ences in Δ9-tetrahydrocannabinol-induced behavioral effects in rats. BehaviouralBrain Research, 154, 77–83.

Turner, C. E., Bouwsma, O. J., Billets, S., & Elsohly, M.A. (1980a). Constituents of Cannabissativa L. XVIII: Electron voltage selected ion monitoring study of cannabinoids.Biomedical Mass Spectrometry, 7, 247–256.

Turner, C. E., Elsohly, M.A., & Boeren, E. G. (1980b). Constituents of Cannabis sativa L. XVII:A review of the natural constituents. Journal of Natural Products, 43, 169–234.

Valiveti, S., & Stinchcomb, A. L. (2004). Liquid chromatographic–mass spectrometricquantitation of Δ9-tetrahydrocannabinol and two metabolites in pharmacokineticstudy plasma samples. Journal of Chromatography B, 803, 243–248.

Vann, R. E., Gamage, T. F., Warner, J. A., Marshall, E. M., Taylor, N. L., Martin, B. R., et al.(2008). Divergent effects of cannabidiol on the discriminative stimulus and placeconditioning effects of Δ9-tetrahydrocannabinol. Drug and Alcohol Dependence, 94,191–198.

Voirin, N., Berthiller, J., Benhaim-Luzon, V., Boniol, M., Straif, K., Ayoub, W. B., et al. (2006).Risk of lung cancer and past use of cannabis in Tunisia. Journal of Thoracic Oncology, 1,577–579.

Wilson, D. M., Peart, J., Martin, B. R., Bridgen, D. T., Byron, P. R., & Lichtman, A. H. (2002).Physiochemical and pharmacological characterization of a Δ9-THC aerosol generatedby a metered dose inhaler. Drug and Alcohol Dependence, 67, 259–267.

Wilson, D. M., Varvel, S. A., Harloe, J. P., Martin, B. R., & Lichtman, A. H. (2006). SR141716(Rimonabant) precipitates withdrawal in marijuana-dependent mice. Pharmacology,Biochemistry and Behavior, 85, 105–113.

Zhang, Z. F., Morgenstern, H., Spitz, M. R., Tashkin, D. P., Yu, G. P., Marshall, J. R., et al.(1999). Marijuana use and increased risk of squamous cell carcinoma of the headand neck. Cancer Epidemiology, Biomarkers and Prevention, 9, 1071–1078.