Ee

LKa

Ab

c

d

e

a

ARR2AA

KEAVCIV

1

ad

paietSt

lD

0h

Cell Calcium 55 (2014) 104– 118

Contents lists available at ScienceDirect

Cell Calcium

jo u rn al h om epage: www.elsev ier .com/ locate /ceca

ffects of endogenous cannabinoid anandamide onxcitation–contraction coupling in rat ventricular myocytes

ina T. Al Kurya, Oleg I. Voitychukd, Ramiz M. Alia, Sehamuddin Galadari c,eun-Hang Susan Yange, Frank Christopher Howarthb, Yaroslav M. Shubad, Murat Oza,∗

Laboratory of Functional Lipidomics, Department of Pharmacology, College of Medicine and Health Sciences, UAE University, Al Ain, Abu Dhabi, Unitedrab EmiratesDepartment of Physiology, College of Medicine and Health Sciences, UAE University, Al Ain, Abu Dhabi, United Arab EmiratesDepartment of Biochemistry, College of Medicine and Health Sciences, UAE University, Al Ain, Abu Dhabi, United Arab EmiratesBogomoletz Institute of Physiology and International Center of Molecular Physiology, National Academy of Sciences of Ukraine, Kyiv-24, UkraineDepartment of Biological Sciences, Schmid College of Science and Engineering, Chapman University, One University Drive, Orange, CA 92866, USA

r t i c l e i n f o

rticle history:eceived 19 September 2013eceived in revised form5 November 2013ccepted 26 December 2013vailable online 5 January 2014

eywords:ndocannabinoidnandamideentricular myocytesontraction

a b s t r a c t

A role for anandamide (N-arachidonoyl ethanolamide; AEA), a major endocannabinoid, in the cardio-vascular system in various pathological conditions has been reported in earlier reports. In the presentstudy, the effects of AEA on contractility, Ca2+ signaling, and action potential (AP) characteristics wereinvestigated in rat ventricular myocytes. Video edge detection was used to measure myocyte short-ening. Intracellular Ca2+ was measured in cells loaded with the fluorescent indicator fura-2 AM. AEA(1 �M) caused a significant decrease in the amplitudes of electrically evoked myocyte shortening andCa2+ transients. However, the amplitudes of caffeine-evoked Ca2+ transients and the rate of recovery ofelectrically evoked Ca2+ transients following caffeine application were not altered. Biochemical studiesin sarcoplasmic reticulum (SR) vesicles from rat ventricles indicated that AEA affected Ca2+-uptake andCa2+-ATPase activity in a biphasic manner. [3H]-ryanodine binding and passive Ca2+ release from SR vesi-cles were not altered by 10 �M AEA. Whole-cell patch-clamp technique was employed to investigate the

ntracellular calciumentricular action potential

effect of AEA on the characteristics of APs. AEA (1 �M) significantly decreased the duration of AP. Theeffect of AEA on myocyte shortening and AP characteristics was not altered in the presence of pertussistoxin (PTX, 2 �g/ml for 4 h), AM251 and SR141716 (cannabinoid type 1 receptor antagonists; 0.3 �M)or AM630 and SR 144528 (cannabinoid type 2 receptor antagonists; 0.3 �M). The results suggest thatAEA depresses ventricular myocyte contractility by decreasing the action potential duration (APD) in a

B1 an

manner independent of C

. Introduction

Endocannabinoids belong to a family of polyunsaturated fattycid-based compounds that mimic most of the effects of tetrahy-rocannabinol, the active ingredient of the marijuana plant

Abbreviations: AP, action potential; APD, action potential duration; APD60, actionotential duration at 60% level of repolarization; AMP, amplitude; AA, arachidoniccid; APIII, antipyrylazo III; BSA, bovine serum albumin; DMSO, dimethylsulphox-de; DHP, dihydropyridine; NAEs, N-acylethanolamines; AEA, N-arachidonoylthanolamide, anandamide; NEM, N-ethylmaleimide; NO, nitric oxide; NT, normalyrode; PTX, pertussis toxin; RCL, resting cell length; metAEA, R-methanandamide;R, sarcoplasmic reticulum; THALF, time from peak to half; TPK, time to peak; TRP,ransient-receptor potential.∗ Corresponding author at: Department of Pharmacology and Therapeutics, Col-

ege of Medicine and Health Sciences, UAE University, P.O. Box 17666, Al Ain, Abuhabi, United Arab Emirates. Tel.: +971 3 7137523; fax: +971 3 7672033.

E-mail address: murat oz@uaeu.ac.ae (M. Oz).

143-4160/$ – see front matter © 2014 Elsevier Ltd. All rights reserved.ttp://dx.doi.org/10.1016/j.ceca.2013.12.005

d CB2 receptors.© 2014 Elsevier Ltd. All rights reserved.

Cannabis sativa. The most widely studied endogenous cannabi-noids are N-arachidonoyl ethanolamide (AEA) or anandamide and2-arachidonylglycerol [1,2]. In recent years extensive researchfocusing on the biological actions of these molecules has revealedthe existence of an endocannabinoid system that regulates sev-eral physiological processes and pathological conditions [3]. It wassuggested that the endocannabinoid system consists of at leastthe endocannabinoid receptors (such as CB1 and CB2 cannabi-noid receptors), the enzymes regulating the synthesis (such asphospholipase-D, and monoacylglycerol), and the degradation(such as fatty-acid amide hydrolase and lipases) processes, andthe proteins involved in their transport across the biologicalmembranes [1,3]. The CB1 receptors are located in the brainand several peripheral tissues including the heart and the vas-

culature. The CB2 receptors, on the other hand, are expressedprimarily in the immune system but recently their presence inthe brain, myocardium, and smooth muscle cells have also beendemonstrated [3].

dx.doi.org/10.1016/j.ceca.2013.12.005http://www.sciencedirect.com/science/journal/01434160http://www.elsevier.com/locate/cecahttp://crossmark.crossref.org/dialog/?doi=10.1016/j.ceca.2013.12.005&domain=pdfmailto:murat_oz@uaeu.ac.aedx.doi.org/10.1016/j.ceca.2013.12.005

Calciu

chrbsarodrolo

cftAebtoel

ishsm

2

2

(siNCvhsatN7uE0cGgmiC

2

tmf

L.T. Al Kury et al. / Cell

Emerging evidence suggests a role for endocannabinoids in theardiovascular system in various pathological conditions, such asypertension, myocardial infarction and heart failure (for recenteviews [4,5]). AEA has profound and rather complex actions onlood pressure and cardiac function. For example, AEA has beenhown to decrease arterial blood pressure in a triphasic manner bycting on the contractility of smooth muscle, the release of neu-otransmitters from peripheral nerve endings, and the activationf autonomic reflex pathways [6]. Both cannabinoid receptor-ependent and independent mechanisms have been shown to playoles in AEA inhibition of smooth muscle contraction dependingn the type of vascular structure, the presence of intact endothe-ium, the metabolic products of endocannabinoids and the activityf metabolizing enzymes [6–8].

Compared with a large body of information available on the vas-ular effects of endocannabinoids, surprisingly few studies haveocused on the role of endocannabinoids in the regulation of con-ractility and Ca2+ signaling in cardiac muscle. Experiments withEA performed in isolated Langendorff rat hearts and in isolated,lectrically stimulated atrial appendages from human, rat, and rab-it [9–12] have revealed a negative inotropic effect of cannabinoidshat may underlie the ability of AEA to decrease cardiac output asbserved in studies performed in vivo (for reviews [4,7,8]). How-ver, mechanisms underlying these cardiac actions of AEA remainargely unknown.

In the present study, we have hypothesized that the negativenotropic actions of anandamide are due to altered Ca2+ homeo-tasis and AP characteristics in ventricular myocytes. Thus, weave investigated the actions of AEA on contractile properties, Ca2+

ignaling and AP waveforms in acutely dissociated rat ventricularyocytes.

. Materials and methods

.1. Ventricular myocyte isolation

Ventricular myocytes were isolated from adult male Wistar rats264 ± 19 g) according to previously described techniques [13]. Thistudy was carried out in accordance with the recommendationsn the Guide for the Care and Use of Laboratory Animals of theational Institutes of Health. The protocol was approved by theommittee on the Ethics of Animal Experiments of the UAE Uni-ersity. Briefly, the animals were euthanized using a guillotine andearts were removed rapidly and mounted for retrograde perfu-ion according to the Langendorff method. Hearts were perfused at

constant flow of 8 ml g heart−1 min−1 and at 36–37 ◦C with a solu-ion containing (mM): 130 NaCl, 5.4 KCl, 1.4 MgCl2, 0.75 CaCl2, 0.4aH2PO4, 5 HEPES, 10 glucose, 20 taurine, and 10 creatine set to pH.3 with NaOH. When the heart had stabilized perfusion was contin-ed for 4 min with Ca2+-free isolation solution containing 0.1 mMGTA, and then for 6 min with cell isolation solution containing.05 mM Ca2+, 0.75 mg/ml collagenase (type 1; Worthington Bio-hemical Corp., USA) and 0.075 mg/ml protease (type X1V; Sigma,ermany). Ventricles were excised from the heart, minced andently shaken in collagenase-containing isolation solution supple-ented with 1% BSA. Cells were filtered from this solution at 4-min

ntervals and resuspended in isolation solution containing 0.75 mMa2+.

.2. Measurement of ventricular myocyte shortening

Ventricular myocytes were allowed to settle on the glass bot-om of a Perspex chamber mounted on the stage of an inverted

icroscope (Axiovert 35, Zeiss, Germany). Myocytes were super-used (3–5 ml/min) with normal tyrode (NT) containing (mM): 140

m 55 (2014) 104– 118 105

NaCl, 5 KCl, 1 MgCl2, 10 glucose, 5 HEPES, and 1.8 CaCl2 (pH 7.4).Shortening of myocytes was recorded using a video edge detectionsystem (VED-114, Crystal Biotech, USA). Resting cell length (RCL)and amplitude of shortening (expressed as a % of resting cell length)were measured in electrically stimulated (1 Hz) myocytes main-tained at 35–36 ◦C. Data were acquired and analyzed with SignalAverager software v 6.37 (Cambridge Electronic Design, UK). Exper-imental solutions were prepared from stock immediately prior toeach experiment.

2.3. Measurement of intracellular Ca2+ concentration

Myocytes were loaded with the fluorescent indicator fura-2 AM(F-1221, Molecular Probes, USA) as described previously [13]. Inbrief, 6.25 �l of a 1 mM stock solution of fura-2 AM (dissolved indimethylsulphoxide) was added to 2.5 ml of cells to give a finalfura-2 concentration of 2.5 �M. Myocytes were shaken gently for10 min at 24 ◦C (room temperature). After loading, myocytes werecentrifuged, washed with NT to remove extracellular fura-2 andthen left for 30 min to ensure complete hydrolysis of the intracel-lular ester. To measure intracellular Ca2+concentration, myocyteswere alternately illuminated by 340 and 380 nm light using amonochromator (Cairn Research, UK) which changed the excita-tion light every 2 ms. The resulting fluorescence emitted at 510 nmwas recorded by a photomultiplier tube and the ratio of the emit-ted fluorescence at the two excitation wavelengths (340/380 ratio)was calculated to provide an index of intracellular Ca2+ concentra-tion. Resting fura-2 ratio, TPK Ca2+ transient, THALF decay of the Ca2+

transient, and the amplitude of the Ca2+ transient were measuredin electrically stimulated (1 Hz) myocytes.

2.4. Measurement of sarcoplasmic reticulum Ca2+ content

Sarcoplasmic reticulum (SR) Ca2+ release was assessed usingpreviously described techniques [13,14]. After establishing steadystate Ca2+ transients in electrically stimulated (1 Hz) myocytesmaintained at 35–36 ◦C and loaded with fura-2, stimulation waspaused for a period of 5 s. Caffeine (20 mM) was then applied for10 s using a solution switching device customized for rapid solutionexchange. Electrical stimulation was resumed and the Ca2+ trans-ients were allowed to recover to steady state. SR-releasable Ca2+

was assessed by measuring the area under the curve of the caffeine-evoked Ca2+ transient. Fractional release of SR Ca2+ was assessedby comparing the amplitude of the electrically evoked steady stateCa2+ transients with that of the caffeine-evoked Ca2+ transient andrefilling of SR was assessed by measuring the rate of recovery ofelectrically evoked Ca2+ transients following application of caffeine.

2.5. Assessment of myofilament sensitivity to Ca2+

In some cells shortening and fura-2 ratio were recorded simul-taneously. Myofilament sensitivity to Ca2+ was assessed fromphase-plane diagrams of fura-2 ratio versus cell length by mea-suring the gradient of the fura-2-cell length trajectory during laterelaxation of the twitch contraction. The position of the trajectoryreflects the relative myofilament response to Ca2+ and hence, canbe used as a measure of myofilament sensitivity to Ca2+ [15].

2.6. Preparation of cardiac sarcoplasmic reticulum vesicles

SR vesicles were obtained from rat ventricles by minor mod-ifications of methods described earlier [16]. Adult male Wistar

rats were anesthetized with intraperitoneal injection of sodiumpentobarbital (100 mg/kg) and hearts were quickly removed andrinsed in Ca2+-free Tyrode solution. The ventricles were blotted onfilter paper to remove excess solution and homogenized in cold

1 Calciu

s14wi1N−rbswE

2

1ddaK(m7r

2

pt(CAwAhtn

2

pe7w(0elbwestomos

2

5

06 L.T. Al Kury et al. / Cell

aline solution of the following composition (mM): 154 NaCl; and0 Trizma-maleate; pH 6.8. The homogenate was centrifuged at000 × g for 30 min and the pellet was discarded. The supernatantas centrifuged at 10,000 rpm (8000 × g) for 30 min, and the result-

ng supernatant was recentrifuged at 47,000 rpm (100,000 × g) for h. The final pellet was resuspended in storage solution (mM): 154aCl; 10 Trizma-maleate; and 300 sucrose, pH: 6.8 then stored at80 ◦C until used. The whole procedure was carried out in a cold

oom, at 4 ◦C and in the presence of protease inhibitors (mM: 500enzamidine; 1 leupeptine; 1 pepstatin A and 200 phenylmethyl-ulphonyl fluoride). Membrane preparations for cardiac SR vesiclesere generous gift from Dr. Susan Dunn (University of Alberta,

dmonton, Canada).

.7. Measurements of Ca2+ uptake rate

SR Ca2+ uptake was measured with a spectrophotometer (Cory00, Varian; Walnut Creek, CA, USA) using the Ca2+-sensitiveye antipyrylazo III (APIII; Sigma–Aldrich, St. Louis, MO, USA), asescribed previously [16]. SR membrane vesicles (100 �g) weredded to 1 ml phosphate buffer medium containing (in mM): 100H2PO4, 3 MgCl2, 2 ATP, 0.02 ruthenium red, and 0.1 APIII, pH 7.1

37 ◦C). Ca2+ uptake was initiated by addition of 10 �M CaCl2 to theedium and measured as changes in absorbance of APIII between

10 and 790 nm. Ruthenium red (10 �M) was used to block Ca2+

elease from the SR.

.8. Measurements of Ca2+-ATPase activity

ATPase activity was measured by minor modifications of cou-led enzyme reaction method described earlier [17,18]. Briefly,he reaction medium used for the assay of Ca2+-ATPase containedmM) 20 mM MOPS, pH 7.1, 80 mM KCI, 5 mM MgCl2, 0.1 mMaCl2, 0.1 mM EGTA, 100 �g of SR membrane protein/ml, and 2 �M23187 (Ca2+ ionophore) was incubated for 5 min at 37 ◦C. A23187as added to the reaction to render the vesicles permeable to Ca2+.t this point 2.5 mM [�-32P] ATP was added, and the amount ofydrolysis was measured by a colorimetric method [17]. Followinghe addition of ATP (3 mM) serial samples were taken for determi-ation of Pi as described earlier [18].

.9. Measurements of [3H]-ryanodine binding

[3H]-ryanodine binding was assayed by some modifications ofreviously described methods [19]. Briefly, ventricles were homog-nized in five volumes of 300 mM sucrose and 10 mM imidazole (pH.1 at 4 ◦C) and aliquots of homogenate (about 100 �g of protein)ere incubated at 37 ◦C in a buffer containing 25 mM imidazole

pH 7.4 at 37 ◦C), 1 M KCl, 0.2–20 nM [3H]-ryanodine (7.4 Ci/mmol),. 9 mM EGTA, and 1 mM CaCl2. After 60 min, 0.4 ml aliquots ofach sample were filtered under vacuum through Watman cellu-ose nitrate filters with pores of 0.45 �M (presoaked in washinguffer containing 25 mM imidazole and 1 M KCl), and washed twiceith 5 ml of ice-cold washing buffer. The filters were dried and

xtracted in 5 ml of HydroflourTM (National Diagnostics, FL, USA)cintillation fluid before counting for 3H. Triplicate 50-�l samples ofhe incubation mixtures were also counted directly for estimationsf total binding. Nonspecific binding was estimated from paralleleasurements of binding in the presence of 10 �M unlabeled ryan-

dine. The radioactivity was measured in Beckman LS 6000IC liquidcintillation counter.

.10. Measurement of Ca2+ release

SR vesicles (50 �g/ml) were passively loaded by incubation in �l of 45CaCl2 (10 mCi/ml, 42.7 mCi/mg; NEN Life Science Products,

m 55 (2014) 104– 118

Inc.) in 1 ml of (mM): 100 KCl; 20 PIPES dipotassium salt; 2 CaCl2;at pH 7.1 at room temperature for 3 h. 45Ca2+ efflux was initiatedby 50-fold dilution of 20 �l of 45Ca2+-loaded SR vesicles intorelease buffer (mM): 100, KCl; 20 PIPES dipotassium salt; 1 EGTA;1 CaCl2; at 37 ◦C followed, after 10 s, by filtration through HA filters(0.45 �m, Millipore Corp., Bedford, MA). Filters were rinsed with5 ml ice-cold wash buffer (mM): 100 KCl; 20 PIPES dipotassiumsalt; 1 EGTA; 1 CaCl2; 10 MgCl2; and 20 �M ruthenium red. Filterswere allowed to air-dry overnight, and radioactivity remainingwithin the membranes was counted in Beckman LS 6000IC liquidscintillation counter. Stock solutions of A23187 (Calbiochem Corp.,San Diego, CA, USA) were made in 100% DMSO; final concentrationof DMSO was 0.12% (v/v). In control experiments, stock solutionsof DMSO up to 0.3% (v/v) and ethanol up to 0.1% (highest concen-tration used 0.07%) did not affect control conditions in Ca2+ uptake,ryanodine binding, and 45Ca2+ release from SR vesicles.

2.11. Measurement of action potentials

Action potentials (APs) were measured using whole-cell patchclamp technique. In our recordings, only rod shaped quiescentmyocytes with clear cross-striations were used. Recordings weremade with an Axopatch 200B amplifier (Molecular Devices, Down-ington, PA, USA) coupled to an A/D interface (Digidata 1322;Molecular Devices, Downington, PA, USA). Patch pipettes werefabricated from filamented GC150TF borosilicate glass (HarvardApparatus, Holliston, MA, USA) on a horizontal puller (SutterInstruments Co., Novato, CA, USA). Electrode resistances rangedfrom 2.0 to 3.0 M�, and seal resistances were 1–5 G�. After gigaseal formation, the membrane was ruptured with gentle suc-tion to obtain whole cell current-clamp configuration. APs wereelicited in current-clamp mode by 4-ms, 0.9–1 nA injections ofsquare current pulses at a frequency of 0.2 Hz. Basic extracellu-lar solutions used for electrophysiological recordings contained(in mM): 144 NaCl, 5.4 KCl, 1.8 CaCl2, 1.2 MgCl2, 1 NaH2PO4, 10HEPES, 10 glucose, and pH 7.4 (adjusted with NaOH). Recordingpatch pipettes were filled with intracellular solution containing(in mM): 150 KCl, 10 NaCl, 120 aspartate, 5 MgCl2, 0.1 CaCl2,1.1 EGTA, 10 HEPES, 4 Mg-ATP, 5 sucrose, and pH 7.2 (adjustedwith HCl). Experiments were performed at room temperature(22–24 ◦C). Changes of external solutions and application of drugswere performed using a multi-line perfusion system driven bya micro-pump with a common outflow connected to the cellchamber. In some experiments, applications of drugs were per-formed using multi-barrel puffing micropipette with a commonoutflow positioned in close proximity to the cell under investiga-tion. In these experiments, the cell was continuously superfusedwith the solution via a puffing pipette to reduce possible arti-facts related to the switch from a static to a moving solution andvice versa. Complete external solution exchange was achieved in

Calciu

wwruwaSocecp

(CClt

2

fdftwtsfPt

3

3

e6ortsAwsi

oFfewt(ottaP

tao

L.T. Al Kury et al. / Cell

as routinely included in extracellular solution. In agreementith earlier studies in rat cardiomyocytes [21,23], AP waveforms

emained unchanged in the presence of ethanol concentrationsp to 0.07% (v/v). AM251, AM630, methanandamide, and URB597ere purchased from Tocris (Ellisville, MO, USA). Other reagents

nd chemicals used in our experiments were purchased fromigma–Aldrich (St. Louis, MO, USA). Drugs were dissolved in DMSOr ethanol. Stocks were kept at −20 ◦C until their use and final con-entrations for ethanol and DMSO used in the experiments did notxceed 0.07% (v/v). At low concentrations, DMSO had no signifi-ant effect on the control conditions during shortening or actionotential recordings.

Electrophysiological data were analyzed using pClamp 8.0Molecular Devices, Downington, PA, USA), Origin 8.0 (Origin Laborp., Northampton, MA, USA) and Mat Lab R2011a (MathWorksorp., Natick, MA, USA) software. APD was measured at 60% of repo-

arization from AP amplitude. Cardiac myocytes were classified toypes based on the shape of the AP in current-clamp mode.

.12. Data analysis

Each experiment was performed on several myocytes from dif-erent batches. Since the results of identical experiments fromifferent groups were qualitatively similar, the data were pooledor statistical purposes with “n” denoting the total number of cellsested for a particular data point. The results of the experimentsere expressed as mean ± standard error of the mean (S.E.M.). Sta-

istical analysis was performed using the paired t-test (within theame cell analysis) and ANOVA tests (for analysis of data from dif-erent cells). On all graphs (*) denote statistical significance with

< 0.05, between specified values, or if not specified to the respec-ive control.

. Results

.1. Effects of anandamide on ventricular myocyte shortening

In normal tyrode (NT) solutions, amplitudes of myocyte short-ning in response to electrical stimulation (stimulated at 1 Hz,0 pulses delivered every 1 min) gradually decreased to 85–80%f controls during experiments lasting up to 20 min. No furtherun down of the shortening amplitudes was observed in NT solu-ion containing 0.007% ethanol (used as vehicle in 1 �M AEAolution; data not shown; n = 7–10; compared to 0 time point;NOVA, F(1,15) = 0.018; P > 0.05) and, unless it was stated other-ise, ethanol (in the concentration that was used in AEA containing

olution) was also included routinely in control (NT) solutions dur-ng shortening and Ca2+ transient experiments.

In the first set of experiments, we tested the effect of AEAn the contractility of acutely isolated rat ventricular myocytes.ig. 1A shows typical records of shortening in a myocyte super-used with either NT (in the absence of AEA, NT contained 0.007%thanol in all experiments) or NT + 1 �M AEA and during washoutith NT. The amplitude of shortening was significantly (paired

-test; n = 12–14; P < 0.05) reduced up to 47.3 ± 2.6% of controlsFig. 1B) when the concentration of AEA was increased in the rangef 1 nM to 10 �M compared to NT (for 0.01 �M AEA; paired t-est, n = 9; t(8) = 6.09; P < 0.05). Synthetic cannabinoid WIN55,212-2ested at 1 �M concentration did not cause a significant alter-tion on the amplitudes of myocyte shortening (paired t-test, n = 7;

> 0.05).

The negative inotropic effect (decrease of shortening ampli-

udes) by AEA could be due to degradation products of AEA such asrachidonic acid. For this reason, we have investigated the effectsf methanandamide (metAEA), a non-hydrolyzable analog of AEA

m 55 (2014) 104– 118 107

[24] and URB597, an inhibitor of fatty acid amide hydrolase (FAAH,enzyme that metabolize AEA) [25]. In cardiomyocytes pretreatedfor 10 min with 0.1–10 �M metAEA (Fig. 1C), the extent of inhi-bition was not significantly different from that of AEA (n = 7–8,ANOVA, F(1,13) = 5.9, P > 0.05). Similarly, pretreatment with 1 �MURB597 for 45 min at 37 ◦C did not alter the extent of AEA inhibi-tion (Fig. 1D). In the presence and absence of URB597 treatment,AEA (1 �M) inhibited myocyte shortening to 68.2 ± 4.8% of controlsand 62.7 ± 8.4% of controls, respectively. There were no statisticallysignificant differences in the inhibitory effect of AEA between con-trol (NT + 0.007% ethanol after 45 min pretreatment) and URB597pretreated cells (n = 9–11, ANOVA, F(1,18) = 7.8, P > 0.05). We havealso tested whether cyclooxygenase products of AEA metaboliteswould mediate the observed actions of this compound. The resultsindicated that the extent of shortening amplitudes of cardiomy-ocytes was not significantly different after incubating the cellswith 30 �M indomethacin, a cyclooxygenase inhibitor, for 30 min(n = 8–11, ANOVA, F(1,17) = 6.1, P > 0.05). Among other contractionparameters measured, resting cell length (RCL) was not signifi-cantly altered (paired t-test; n = 12–14; P > 0.05, data not shown)by 10 min superfusion with AEA (1 nM–1 �M). However, increasingthe concentration of AEA to 10 �M caused a small but statisticallysignificant reduction of RCL in about 60% of cells tested (pairedt-test; n = 12–14; P < 0.05, data not shown).

In the next series of experiments we have tested whether theeffect of AEA is mediated by the activation of cannabinoid (CB)receptors. For this purpose, we have first tested the effects of estab-lished antagonists of CB1 and CB2 receptors on the AEA inhibitionof shortening amplitudes in cardiomyocytes. Two structurally dif-ferent CB1 receptor antagonists (AM251 with a Ki of 7.5 nM [26];Fig. 2A and SR141716 with a Ki of 1.8 nM [27]; Fig. 2B) and twoCB2 receptor antagonists (AM630 with a Ki of 32.1 nM [26] andSR144528 with Ki of 0.6 nM [27]) were tested. At 300 nM concentra-tion, these antagonists were not able to reverse the inhibitory effectof AEA on the shortenings amplitudes of cardiomyocytes (n = 8–12,ANOVA, P > 0.05).

Since the activation of both CB1 and CB2 receptors are medi-ated by Gi/o subtypes of G-proteins [26], we have also tested theeffect of inhibitors of Gi/o proteins such as pertussis toxin (PTX) andN-ethylmaleimide (NEM) on AEA induced inhibition of cardiomy-ocyte shortening (Fig. 2E and F). Preincubation of cardiomyocytesin either PTX (2 �g/ml, 3 h at 37 ◦C) or NEM (50 �M for 30 min at37 ◦C; [28]) did not alter the extent of AEA effect on cardiomy-ocyte shortening (n = 9–12, ANOVA, P > 0.05). However, in positivecontrol experiments, preincubation of cardiomyocytes in PTX orNEM effectively blocked the inhibitory effect of clenbuterol, a �2adrenoreceptor agonist on cardiomyocyte shortening (supplementFig. 1).

3.2. Effects of anandamide on intracellular Ca2+ levels

In the second set of experiments we have investigated the effectsof 10 min bath application of 1 �M AEA on the resting intracellularCa2+ levels and on the amplitudes and kinetics of Ca2+ transientselicited by electrical-field stimulation. Typical records of Ca2+ trans-ients in a myocyte superfused with either NT or NT + 1 �M AEAand during washout with NT are shown in Fig. 3A. The effects of1 �M AEA on resting fura-2 ratio, TPK Ca2+ transient, THALF decayof the Ca2+ transient, and AMP of Ca2+ transient are shown inFig. 3B–E, respectively. Although, AEA has been shown to alter intra-cellular Ca2+ levels in various types of cells (for reviews [29,30]),application of 1 �M AEA for 10 min did not cause a significant alter-

ation in resting fura-2 ratio and TPK Ca2+ transient (paired t-test;n = 21–24 cells, P > 0.05). However, THALF decay of the Ca2+ tran-sient and AMP of the Ca2+ transient were significantly reduced by1 �M AEA to 118.9 ± 5.5 ms and 0.243 ± 0.032 fura-2 ratio units

108 L.T. Al Kury et al. / Cell Calcium 55 (2014) 104– 118

Fig. 1. Effects of AEA, metAEA and preincubation with URB597 or indomethacin on AEA induced inhibition of ventricular myocyte shortening. (A) Typical records of shorteningin an electrically stimulated (1 Hz) ventricular myocyte superfused with either NT (containing the vehicle, 0.007% ethanol) or NT + 1 �M AEA and during washout with NT. (B)Bar graph showing the mean amplitudes (AMP) of shortening expressed as a percentage of control values in vehicle (NT + 0.007% ethanol), and in presence of AEA (1 nM–10 �M).Myocytes were maintained at 35–36 ◦C and superfused with AEA for 10 min. Data are shown as means ± S.E.M., n = 12–14 cells. *Indicates statistically significant differenceat the level of P < 0.05. (C) Bar graph showing the mean amplitudes (AMP) of shortening expressed as a percentage of control values in vehicle (NT + 0.007% ethanol), and inpresence of metAEA (1 nM–10 �M). Data are shown as means ± S.E.M., n = 7–8 cells. *Indicates statistically significant difference at the level of P < 0.05. (D) Bar graph showingthe effect of AEA on the mean amplitudes (AMP) of shortening expressed as percentage of control values in NT containing 0.007% ethanol or after 45 min incubation with1 �M URB597 at 37 ◦C. The S.E.M. values are indicated on top of each column. Data are mean ± S.E.M., n = 7–11 cells. (E) Bar graph showing the effect of AEA on the meana ontainT –10 ce

cPc

3t

mimlu

mplitudes (AMP) of shortening expressed as percentage of control values in NT che S.E.M. values are indicated on top of each column. Data are mean ± S.E.M., n = 9

ompared to 129.3 ± 5.2 ms (paired t-test, n = 24 cells, t(23) = 3.73; < 0.05) and 0.326 ± 0.024 fura-2 ratio units (paired t-test, n =24ells, t(23) = 5.79; P < 0.05) in controls, respectively.

.3. Effect of anandamide on sarcoplasmic reticulum Ca2+

ransport

The effect of 1 �M AEA on Ca2+ transport was investigated inyocytes exposed to 20 mM caffeine. Fig. 4A shows a typical record

llustrating the protocol used in these experiments. Initially, theyocyte was electrically stimulated (ES) at 1 Hz. Electrical stimu-

ation was then turned off for 5 s. Caffeine was then applied for 10 ssing a rapid solution-exchanger device. Electrical stimulation was

ing 0.007% ethanol or after 30 min incubation with 30 �M indomethacin at 37 ◦C.lls.

then restarted, and the recovery of intracellular Ca2+ was recordedduring a period of 60 s. SR Ca2+ content was assessed by measuringcaffeine-evoked Ca2+ release (area under the caffeine-evoked Ca2+

transient) and fractional release of Ca2+ by comparing the ampli-tude of the electrically evoked steady-state Ca2+ transients withthat of the caffeine-evoked Ca2+ transient in the presence of eitherNT alone or NT with 1 �M AEA. Fractional release of SR Ca2+ was notsignificantly altered in 1 �M AEA compared to NT (0.749 ± 0.024in AEA versus 0.753 ± 0.028 in controls; paired t-test; n = 23 cells,

t(22) = 0.166; P > 0.05; Fig. 4B). The area of caffeine-evoked Ca2+

transient (Fig. 4C) and recovery of the Ca2+ transient during elec-trical stimulation following application of caffeine (Fig. 4D) werealso not significantly altered in myocytes exposed to 1 �M AEA

L.T. Al Kury et al. / Cell Calcium 55 (2014) 104– 118 109

0

20

40

60

80

100

N-ethylmaleimide trea ted

Pertuss is tox in trea ted

vehicletrea ted

SR14452 8

AEA - - + + AM630 - + + -

AEA AEA

AM63 0 DC

B

% A

MP

shor

teni

ng( n

orm

aliz

ed to

NT

)

A

***

AM25 1

0

20

40

60

80

100

*

% A

MP

shor

teni

ng( n

orm

aliz

ed to

NT

)

SR14171 6

0

20

40

60

80

100

% A

MP

sho

rten

ing

( nor

mal

ized

to N

T)

**

0

20

40

60

80

100

AEA - - + + SR144528 - + + -

% A

MP

shor

teni

ng( n

orm

aliz

ed to

NT

)

AEA - - + + AM251 - + + -

**

0

20

40

60

80

100 P>0.05

% A

MP

sho

rten

ing

(nor

mal

ized

to N

T)

0

20

40

60

80

100

vehicletrea ted

AEA - - + + SR141716 - + + -

FE

% A

MP

sho

rten

ing

( nor

mal

ized

to N

T)

P>0.05

Fig. 2. Effects of cannabinoid receptor antagonists, pertussis toxin and N-ethylmaleimide on AEA-induced inhibition of cardiomyocyte shortening. Bar graphs showing themean amplitudes (AMP) of shortening expressed as a percentage of control values in vehicle (NT + 0.007% ethanol), and (A) in presence of CB1 receptor antagonist AM251(0.3 �M), (B) CB1 receptor antagonist SR141716 (0.3 �M), (C) CB2 receptor antagonist AM630 (0.3 �M), and (D) CB2 receptor antagonist SR144528 (0.3 �M). Data are shownas means ± S.E.M., n = 9–11 cells. *Indicates statistically significant difference at the level of P < 0.05. (E) Bar graph showing the effect of AEA on the mean amplitudes (AMP) ofshortening expressed as a percentage of control values in vehicle (NT + distilled water), and pertussis toxin treated cells. (F) Bar graph showing the effect of AEA on the meanamplitudes (AMP) of shortening expressed as a percentage of control values in vehicle and N-ethylmaleimide treated cells. Data are shown as means ± S.E.M., n = 9–11 cells.*

mP

3

att[perbgo(s

Indicates statistically significant difference at the level of P < 0.05.

yocytes compared to control cells (paired t-test; n = 21–23 cells, > 0.05).

.4. Effect of anandamide on myofilament sensitivity to Ca2+

The effects of AEA on myofilament sensitivity to Ca2+ werelso investigated. These experiments tested whether AEA decreaseshe mechanical responses by altering the affinity of the contrac-ile machinery of the ventricular myocytes to intracellular Ca2+

15]. A typical record of myocyte shortening and fura-2 ratio andhase-plane diagrams of fura-2 ratio versus cell length in myocytesxposed to NT are shown in Fig. 5A. The gradient of the trajectoryeflects the relative myofilament response to Ca2+ and hence, haseen used as a measure of myofilament sensitivity to Ca2+ [15]. The

radients of the fura-2-cell length trajectory during late relaxationf the twitch contraction measured during the periods 500–600 msFig. 5B), 500–700 ms (Fig. 5C) and 500–800 ms (Fig. 5D) were notignificantly altered in AEA compared to NT (containing 0.007%

ethanol) suggesting that myofilament sensitivity to Ca2+ is notreduced by AEA (AEA-treatment was compared to NT containing0.007% ethanol, paired t-test; n = 23–30 cells; P > 0.05).

3.5. Effect of anandamide on Ca2+-uptake and Ca2+-ATPaseactivity in cardiac sarcoplasmic reticulum vesicles

Decreases in THALF decay and AMP of the Ca2+ transient byAEA during twitch responses, can be due to previously reportedeffects of endocannabinoids on Ca2+ release and uptake processesin myocytes and neurons ([31,32]; for reviews [29,30]). For thispurpose, we first measured Ca2+ uptake by SR vesicles isolatedfrom rat ventricles. The net Ca2+ uptake is the sum of the SR Ca2+

influx (mainly by the activity of Ca2+-ATPase) and Ca2+ leak from

SR (through ryanodine-sensitive channels). Our experiments werecarried out in the presence of 10 �M ruthenium red, to inhibitcompletely the activity of cardiac Ca2+ release channel. Underthese conditions, the net Ca2+ uptake closely correlates with the

110 L.T. Al Kury et al. / Cell Calcium 55 (2014) 104– 118

Fig. 3. Effects of AEA on amplitude and time-course of intracellular Ca2+ in ventricular myocytes. (A) Typical records of Ca2+ transients in an electrically stimulated (1 Hz)ventricular myocyte superfused with either NT or NT + 1 �M AEA and during washout with NT; scale bar indicates 0.2 fura-2 ratio unit (RU). Also shown resting fura-2r ecay ow s meaP

Scusdu1(wcF(smawp1e

atio (340/380 nm; B), time to peak (TPK) Ca2+ transient (C), time to half (THALF) dere maintained at 35–36 ◦C and superfused with AEA for 10 min. Data are shown a

< 0.05.

R Ca2+-ATPase activity. Fig. 6A illustrates that AEA, in the con-entration range 0.1–1 �M, causes a significant increase of Ca2+

ptake rate, while elevating the concentrations of AEA to 10 �Mlowed down the uptake rate significantly (Fig. 6A). Experimentalata were fitted by a single exponential function from the rate ofptake and time constants � were presented in control (treated0 min with vehicle, 0.07% ethanol) and in the presence of AEAFig. 6A and B). In the presence of 0.1 and 1 �M AEA, � valuesere significantly decreased (AEA-treatment was compared to NT

ontaining 0.007% ethanol; n = 8–9 samples from 3 experiments,(1,13) = 14.9; ANOVA, P < 0.05), whereas, 10 �M AEA increasedn = 8–9, F(1,13) = 17.2; ANOVA, P < 0.05) � values (Fig. 6B). Sub-equently, we have measured the activity of Ca2+-ATPase in SRembranes (Fig. 6C and D). In the presence of 0.1 and 1 �M AEA,

ctivity of Ca2+-ATPase was significantly increased (AEA-treatment

as compared to NT containing 0.007% ethanol; n = 8–9 sam-les from 3 experiments, F(1,20) = 24.1; ANOVA, P < 0.05), whereas,0 �M AEA decreased (n = 8–9, F(1,20) = 7.3; ANOVA, P < 0.05) thenzyme activity (Fig. 6C and D).

f the Ca2+ transient (D) and amplitude (AMP) of the Ca2+ transient (E). Myocytesns ± SEM, n = 21–24 cell. *Indicates statistically significant difference at the level of

3.6. Effect of anandamide on [3H]-ryanodine binding andCa2+-release in cardiac sarcoplasmic reticulum vesicles

Ryanodine binding is widely used method to investigate theinteraction between drugs and SR release channels. In Fig. 7A,saturation binding curves for [3H]-ryanodine were presentedin the presence and absence of the AEA. As the [3H]-ryanodineconcentration was increased from 0 to 20 nM, ryanodine bindingto SR vesicles was rapidly increased and saturated at 10–20 nM.At a concentration of 10 �M, AEA caused a significant inhibition ofthe specific binding of [3H]-ryanodine. In controls and in presenceof 0.1, 1 and 10 �M AEA, maximum binding activities (Bmax)were 352 ± 37, 365 ± 42, 339 ± 29 and 289 ± 24 pmol/mg protein,respectively. In controls and in presence of 0.1, 1 and 10 �MAEA, KD values were 2.2 ± 0.6, 1.8 ± 0.7, 2.5 ± 0.7 and 2.3 ± 0.4 nM,

respectively. There was no statistically significant differencebetween control and AEA treated groups with respect to Bmax andKD values (ANOVA, n = 9–12, P > 0.05). Effect of AEA on saturationbinding was further analyzed by Scatchard analysis (Fig. 7B), which

L.T. Al Kury et al. / Cell Calcium 55 (2014) 104– 118 111

F he effa of Ca2

e ± S.E.

y21o

Fat

ig. 4. Effect of AEA on sarcoplasmic reticulum Ca2+. (A) Typical record illustrating t rat ventricular myocyte. Also shown are mean amplitudes of SR fractional releasevoked intracellular Ca2+ after application of caffeine (D). Data are shown as means

ielded Bmax values of 289 ± 31, 293 ± 28, 289 ± 33, 263 ± 36 and01 ± 24 pmol/mg protein in controls and in presence of 0.1, 1 and0 �M AEA, respectively. KD values for controls and in presencef 0.1, 1 and 10 �M AEA were 0.9 ± 0.2, 0.9 ± 0.1, 1.0 ± 0.2 and

ig. 5. Effect of AEA on myofilament sensitivity to Ca2+. (A) Typical record of myocyte short myocytes exposed to NT. The arrow indicates the region where the gradient was measurrajectory during late relaxation of the twitch contraction during the periods 500–600 (B),

ects of electrical stimulation (ES) and rapid application of caffeine on fura-2 ratio in+. (B) Area under the caffeine-evoked Ca2+ transient (C) and recovery of electricallyM., n = 21–23 cells.

0.6 ± 0.1 nM, respectively. Although Bmax values tend to decreaseslightly with increasing AEA concentrations, there were no sta-tistically significant difference between controls and 0.1 �M and1 �M AEA treated groups (ANOVA, n = 11–14, P < 0.05). However,

ening and fura-2 ratio and phase-plane diagrams of fura-2 ratio versus cell length ined. B-D shows the effect of 1 �M AEA on the mean gradient of the fura-2-cell length500–700 (C) and 500–800 ms (D). Data are shown as means ± S.E.M., n = 25–30 cells.

112 L.T. Al Kury et al. / Cell Calcium 55 (2014) 104– 118

0

2

4

6

8

10

AEA 10 μM

Ca2

+ (

μM)

AEA 0.1 μM

0

50

100

150

200

*

Ca2

+-A

TPa

se a

ctiv

ity(%

of

cont

rol)

*

0 2 4 6 8 100

5

10

15

20

25 Con trol AEA 0.1 μM AEA 1 μM AEA 10 μM

AEA0.1 μM

100 ms AEA0.1 μM

Time (min)

AEA10 μM

Con trol

DC

AEA10 μM

AEA1 μM

Con trol

AT

P hy

drol

yzed

(μM

Pi/m

g pr

otei

n)

0

50

100

150

200

250

*

*

AEA1 μM

BA

τ o

f C

a2+ u

ptak

e (s

)

Fig. 6. Effects of AEA on Ca2+ uptake and Ca2+-ATPase activity in cardiac SR vesicles. (A) Effect of AEA on the uptake of Ca2+ from SR vesicles. Ca2+ uptake by SR membraneswas measured in control conditions and in the presence of 0.1, 1 and 10 �M AEA. Uptake of Ca2+ was initiated by addition of CaCl2 (10 �M). (B) Effect of AEA on the rate of Ca2+

uptake by SR vesicles. Experimental data were fitted by a single exponential function, from which the time constant (�) of Ca2+ uptake was calculated. Average time constantsof SR Ca2+ uptake in control conditions, and in the presence of 0.1, 1 and 10 �M AEA were presented in the bar graph. Data are the mean ± S.E.M., of 8–9 measurements.* ct of A 2+

C fit ofm value

tA

ov4

dvC(

3m

ate(cmgpnawftp

Indicates statistical difference from control values at the level of P < 0.05. (C) Effea2+-ATPase activity. Slope of each hydrolysis curve was calculated from the lineareans ± S.E.M. of 8–9 measurements. *Indicates statistically different from control

he difference in Bmax and KD values between controls and 10 �MEA was statistically significant (ANOVA, n = 11–14, P > 0.05).

In the next set of experiments, we have tested the effect of AEAn the passive release of Ca2+ from SR. For this purpose, cardiac SResicles were loaded with 45Ca2+ and effect of AEA was tested on the5Ca2+ content (Fig. 7C). AEA, in the concentration range 0.1–10 �M,id not cause any alteration in the 45Ca2+ content of cardiac SResicles loaded with 45Ca2+. However, incubation in 2 �M A23187,a2+ ionophore, effectively reduced 45Ca2+ content of SR vesiclesANOVA, n = 8–9, F(1,15) = 25.6; P < 0.05).

.7. Effects of anandamide on the action potentials of ventricularyocytes

Generation of the cardiac AP requires a specific temporalctivation pattern of several ion channels. AEA has been showno influence the functional properties of several of these channelsither directly or indirectly thereby affecting cardiac excitabilityfor a review [29]). In this set of experiments, patch-clampedardiomyocytes were exposed to the AEA while continuouslyonitoring their Vrest and APs in the current clamp mode. The

eneration of APs was evoked by 0.9–1 nA depolarizing currentulses of 4 ms duration. Since the intracellular pipette solution didot contain Ca2+-chelating agents, the generation of each AP wasccompanied by myocyte contraction. Therefore, current pulses

ere applied at a frequency of 0.2 Hz. During a typical experiment

ollowing protocols were employed: first, whole-cell configura-ion was established and 4–5-min dialysis of the myocytes withipette solution was allowed to ensure the equilibrium conditions

EA on the activity of Ca -ATPase in SR vesicles. (D) Effect of AEA on the slope of the data, and the slopes were presented as percent of controls. Data are shown ass at the level of P < 0.05.

between the intracellular pipette solution and intracellular milieu.Subsequent to achieving stable recordings of baseline electricalactivity (Vrest and AP parameters), myocytes were exposed to AEAfor 10–15 min and subsequently it was washed out.

In our initial experiments, effects of AEA were tested on thepassive membrane properties of cardiomyocytes. The passive prop-erties of the ventricular cells from controls were not significantlydifferent from those of the AEA treated cells. Resting membranepotentials (mean ± SEM) were −76.3 ± 1.7 and −74.2 ± 1.6 mV incontrol (n = 11) and AEA treated (n = 14) myocytes, respectively.The mean cell capacitance in the control group was 109.6 ± 12.8 pF,whereas in the AEA treated cells was 104.7 ± 11.6 pF. Theinput resistance was 82.3 ± 13.4 M� in the control cells and85.6 ± 11.8 M� in AEA-treated cells. In control cells, these passivemembrane properties were not altered significantly in experi-ments lasting up to 25–30 min. In 9 control cells measured, restingmembrane potentials, cell capacitance, and input resistance after25 min of experiment were −73.9 ± 3.8 mV, 114.3 ± 12.7 pF, and83.4 ± 11.7 M�, respectively. These values were not significantlydifferent than control values obtained within the first 5 min ofpatch-clamp experiment (n = 9; paired t-test, P > 0.05).

In agreement with earlier studies in rat ventricular myocytes(for a review [33]) and according to their duration of APs, twomain types of waveforms were observed; endocardial (with longAP durations, Fig. 8A; 44.6 ± 2.9 ms; n = 23) and epicardial (with

short AP durations; Fig. 8B; 14.9 ± 1.6; n = 26) myocytes. Restingmembrane potentials were −77.6 ± 1.3 mV in endocardial cells, and−78.5 ± 1.4 mV in epicardial cells. Similar to earlier findings [33],the amplitudes of APs (122.6 ± 8.9 mV versus 110.7 ± 7.1 mV) and

L.T. Al Kury et al. / Cell Calciu

0 5 10 15 200

100

200

300

Con trol AEA 0.1 μM AEA 1 μM AEA 10 μM

[3H]-Ryanodine (nM)

[3H

]-R

yano

dine

bou

nd (

fmol

/mg)

0

10

20

30

40

A2318 7 2 μM

AEA10 μM

AEA1 μM

AEA0.1 μM

[3H]- Ryanod ine boun d (fmol/ mg)

C

Ca2

+ c

onte

nt in

SR

ves

icle

s(n

mo l

/mg

prot

ein)

Con tro l

*

0 10 0 20 0 30 0

100

200

300

B

AB

ound

/Fre

e(f

mol

/mg/

nM)

Fig. 7. Effects of AEA on [3H]-ryanodine binding and Ca2+ release from SR vesicles.(A) Effect of AEA (0.1–10 �M) on the saturation binding of [3H]-ryanodine to SRvesicle membranes. (B) Scatchard plots derived from the saturation binding exper-iments. Ratio of specific [3H]-ryanodine binding to the free [3H]-ryanodine wasplotted against the [3H]-ryanodine concentration used in saturation experiments.Data were from 3 separate experiments (3–4 trials in each experiment). Each datapoint represents the mean of 9–14 trials. (C) Effect of AEA on Ca2+ release from SRmembranes. The loaded vesicles were preincubated for 30 min in AEA (0.1–10 �M)o8

dd(

erocRept

tudes without causing significant alteration to the time course

r A23187 (2 �M) prior to the release assay. Data are shown as means ± S.E.M., from to 9 measurements.

V/dtmax values (177.3 ± 8.4 V/s versus 141.6 ± 7.2 V/s) in endocar-ial cells were significantly higher than those of epicardial cellsANOVA, n = 23–26, P < 0.05).

In the concentration range 0.1–1 �M, AEA consistently short-ned the duration of AP in both cell types (measured at 60% ofepolarization, APD60; Fig. 8A and B) with a hyperpolarizing shiftn Vrest. Changes in AP shortening in response to AEA (1 �M) appli-ation were noticeable within 10–15 s (insets to Fig. 8A and B).ecoveries were usually partial and required longer time. Similar

ffect of AEA was observed in 5 cells recorded at physiological tem-eratures (35–36 ◦C). APD60 decreased from 33 ± 4 ms in controlso 21 ± 3 ms in the presence of AEA (ANOVA, n = 5, P < 0.05).

m 55 (2014) 104– 118 113

Effects of AEA on Vrest (Fig. 8C), and APD60 (Fig. 8F) reacheda statistically significant level at 1 �M AEA (ANOVA, n = 8–11,P < 0.05). However, AP amplitude (Fig. 8D) and maximal rate of rise(dV/dtmax) of AP (Fig. 7E) were not altered significantly in both typesof cells at 0.1 �M and 1 �M AEA (ANOVA, n = 10–12, P > 0.05).

At higher concentrations (10 �M), AEA caused a 5–10 mVdepolarization (Fig. 9A–C; ANOVA, n = 18, F(1,16) = 8.3; P < 0.05 inendocardial cells, and ANOVA, n = 24, F(1,22) = 0.6; P < 0.05 in epi-cardial cells) and decreased significantly the AP amplitudes anddV/dtmax in the majority of both types of myocytes (Fig. 9D and E;ANOVA, n = 18–24, P < 0.05). AEA (10 �M) also caused a significantdecrease in APD60 (Fig. 9F). However, in a subgroup of endocardial(5 out of 18) and epicardial (4 out of 24) cells, AEA, although causeda significant depolarization and caused a significant decrease in APamplitude and maximum rate of rise of AP (supplement Figs. 3 and4), it did not alter APD60 values significantly (supplement Fig. 4).

In the next series of experiments, we have investigated theeffects of CB1 receptor antagonist AM251 (0.3 �M) and CB2 recep-tor antagonist AM630 (0.3 �M) on AEA-induced changes in the APduration of endocardial and epicardial myocytes. The effect of AEAon APD60 remained unaltered in the presence of 0.3 �M AM251(ANOVA, n = 5–7, P > 0.05; Fig. 10A). Similarly, pretreatment withCB2 receptor antagonist AM630 (0.3 �M) did not cause a significanteffect on AEA-induced changes in AP duration in both endocar-dial and epicardial myocytes (ANOVA, n = 5–7, P > 0.05; Fig. 10B).Actions of cannabinoid receptors (CB1 and CB2) are mediated by theactivation of Gi/o proteins sensitive to PTX treatments [3]. There-fore, we have tested whether PTX pretreatment (2 �g/ml, 3 h at37 ◦C) alters AEA-induced changes in AP duration in endocardialand epicardial myocytes. The results of experiments were pre-sented in Fig. 10C. PTX did not cause a significant alterations inthe effects of AEA on both endocardial and epicardial myocytes(ANOVA, n = 5–7, P > 0.05; Fig. 10C).

4. Discussion

The results of this study indicate for the first time that impairedCa2+ signaling underlies the negative inotropic actions of AEA inrat ventricular myocytes and that direct interaction of AEA withion channel(s) shaping APs, rather than the activation of knowncannabinoid receptors, mediates, at least in part, the effects of AEAon myocyte contractility.

Administration of AEA causes complex hemodynamic changesinvolving phases of both increased and decreased blood pressureas well as changes in heart rate and contractility (for reviews[4,7]). The results of earlier studies suggested that the vascularactions of endocannabinoids involve multiple systems and com-plex set of cellular and molecular mechanisms [6,7]. In additionto receptor-mediated and direct actions of endocannabinoids onmuscular structures, neuronal and endothelial cells have also beenshown to be influenced by AEA and its metabolic products. Further-more, the presence of off-target sites has been reported to causeexperimental variations observed in these studies [34].

Use of video edge detection in individual myocytes has sev-eral advantages over in vivo experiments and traditional in vitrosystems such as Langendorff-perfused heart preparation, sinceit allows measurement of contractility at single-cell level in arelatively isolated environment and excludes the influence ofautonomic nerve endings, gap-junctions, neurotransmitter uptakesystem, and coronary perfusion status. In our experiments, AEAcaused a significant reduction in the maximal shortening ampli-

of myocyte contraction. These findings provide evidence that thenegative-inotropic effect of AEA results from a direct interactionof AEA with ventricular myocytes, rather than actions of AEA on

114 L.T. Al Kury et al. / Cell Calcium 55 (2014) 104– 118

Fig. 8. Effect of AEA on the excitability of ventricular myocytes. Representative recordings show the APs in controls (dark gray area), in the presence of 1 �M AEA (light grayarea) and after washout (striped area) in the ventricular endocardial (A) and epicardial (B) myocytes; the insets on panels A and B show the time course of the action potentialduration (APD60) and resting potential (Vrest) changes in response to AEA application (indicated by horizontal bars). (C)–(F) show summary of AEA effects on amplitude andshape of the AP in cardiomyocytes; quantification of the changes in Vrest (C), AP amplitude (D), AP maximal rate of rise (E) and AP duration (F), characterized byAPD60 incontrols (dark gray bars) and in response to 0.1 �M (cross hatched bars) or 1 �M AEA (light gray bars). Data are shown as means ± S.E.M., from 8 to 12 myocytes for eachg

nr

iotaHbCksti

bccmdiwpamd

roup.

erve endings and neurotransmitter uptake systems that have beeneported in various neuronal structures [35,36].

Negative-inotropic actions of AEA might be attributed to thempaired release of Ca2+ from the SR. In fact, AEA and other vari-us cannabinoid receptor agonists have been reported to modulatehe ryanodine sensitive intracellular Ca2+ stores and Ca2+-ATPasectivity in various cell types [31,37–39], for recent reviews [30,40].owever, binding of ryanodine to SR membranes was not alteredy AEA. Similarly, in the presence of AEA (0.1 and 1 �M), passivea2+ release remained unchanged. Furthermore, the amplitude andinetics of caffeine-induced Ca2+ release from intracellular Ca2+

tores were not changed by AEA. Collectively, these results indicatehat ryanodine-sensitive intracellular Ca2+ stores are not involvedn negative-inotropic effects of AEA observed in this study.

However decreases in THALF decay and AMP of the Ca2+ transienty AEA during twitch responses can be due to increased uptake ofytosolic Ca2+ to SR. In fact, AEA, in 0.1 and 1 �M concentrations,aused a significant increase in Ca2+-ATPase activity in cardiac SRembranes suggesting that increased Ca2+ uptake contributes to

ecrease in Ca2+ transients. At higher concentrations, AEA (10 �M)nhibited the activity of this enzyme (Fig. 6B). Interestingly, NAEs

ith varying carbon chain lengths [37] and fatty acid-based com-

ounds such as arachidonic acid (metabolic product of AEA) haslso been shown to modulate the activity of Ca2+-ATPase in a similaranner in cardiac and skeletal SR membranes [41]. It is likely that

ecreased resting cell length found at high concentrations (10 �M)

of AEA is due to increased resting Ca2+ levels in some of the car-diomyocytes.

However, methanandamide (metAEA), a non hydrolyz-able analog of AEA [24], also showed similar biphasicactions, suggesting that AEA, but not its metabolic prod-ucts produces observed actions on Ca2+-ATPase (supplementFig. 2). Potentiation of the Ca2+-ATPase activity without alteringCa2+ release and ryanodine-binding to the Ca2+ release channelcan account for the decrease in THALF decay and amplitude of theCa2+ transient caused by AEA (1 �M). Decreases in THALF decayand amplitudes of the Ca2+ transient by low concentration of AEA(1 �M) during twitches, but not caffeine-induced Ca2+ transients,may suggest that compared to caffeine-induced responses, fastCa2+ transients during electrical stimulations are relatively moresensitive to alterations in Ca2+-ATPase activity.

Activation of cannabinoid receptors alters the levels of sec-ond messengers such as cAMP, cGMP and protein kinase C [30,42]which are known to be involved in tuning the Ca2+ sensitivity ofthe contractile proteins. However, sensitivity of contractile pro-teins to intracellular Ca2+ remained unchanged in the presenceof AEA suggesting that phosphorylation and de-phosphorylationof the contractile proteins do not play a significant role in

negative-inotropic actions of AEA. An earlier study in isolated ratatria demonstrated that AEA caused negative inotropic effects bydecreasing cAMP and increasing nitric oxide (NO) levels [11]. How-ever, AEA still decreased contractile performance in the presence

L.T. Al Kury et al. / Cell Calcium 55 (2014) 104– 118 115

Fig. 9. Effects of high AEA concentration (10 �M) on the excitability of ventricular myocytes. Representative recordings of APs in controls (dark gray area) and in the presenceof 10 �M AEA (light gray area) and after washout (striped area) in the ventricular endocardial (A) and epicardial (B) myocytes; the insets on panels A and B show the timecourse of the APD60 and resting potential (Vrest) changes in response to AEA application (indicated by horizontal bars). (C)–(F) show summary of AEA effects on amplitudea P ampc as m

oaascrAlICsu

acete(dAutncnit

r

nd shape of the AP in cardiomyocytes; quantification of the changes in Vrest (C), Aontrols (dark gray bars) and in response to 10 �M (light gray bars). Data are shown

f L-NAME, a NO synthase inhibitor, excluding a NO-mediated neg-tive inotropic effect on human atrial muscle [10]. Similarly, innother study in rat isolated heart, the negative inotropic actions ofynthetic cannabinoid HU-210 were not correlated with the intra-ellular concentrations of cAMP and cGMP [43]. Collectively, theseesults, in agreement with our findings, suggest that the effects ofEA on myocyte contractility are not related to changes in intracel-

ular Ca2+ release machinery or sensitivity of myofilaments to Ca2+.n addition, in the presence of AEA, resting levels of intracellulara2+ and cell length of ventricular myocytes remained unaltereduggesting that AEA does not significantly affect Ca2+ homeostasisnder resting conditions.

During excitation–contraction coupling, alterations in themplitudes and kinetics of cardiac AP are closely associated withorresponding changes in the contractility of myocytes. Earlierlectrophysiological studies on rat ventricular myocytes indicatedhat there are two distinctly different groups of cells; displayingither epicardial (short duration) or endocardial (long duration) APsfor a review [33]). In our study, low AEA concentrations (0.1 �M),id not cause alterations in amplitudes and kinetics of APs (Fig. 8).t low AEA concentrations, there was a slight decrease in Vrest val-es, which reached a statistically significant level at 1 �M AEA. Athis concentration, AEA decreased durations of APs without sig-ificantly affecting the amplitudes and dV/dtmax of APs. At higheroncentrations (10 �M), AEA-induced changes in APD accompa-ied with depolarization of the Vrest and decreases of dV/dtmax

n the endocardial and epicardial ventricular myocytes, suggestinghat AEA acts on multiple ion channels with different potencies.

In cardiac muscle, extracellular Ca2+ required to trigger Ca2+

elease from SR enters through L-type voltage-dependent Ca2+

litude (D) AP maximal rate of rise (E) and AP shape (F), characterized by APD60 ineans ± S.E.M., from 13 to 20 myocytes.

channels opened during the AP. Our unpublished results usingvoltage-clamp mode of whole-cell patch clamp technique indi-cate that, in agreement with the changes in the amplitudesand dV/dtmax of APs, AEA (1–10 �M) caused significant inhi-bition of voltage-dependent Na+ and L-type Ca2+ channels incardiomyocytes. Collectively, these results suggest that duringexcitation–contraction coupling, shortening of AP due to the inhi-bition of L-type Ca2+ channels and decreases in Ca2+-induced Ca2+

release from sarcoplasmic reticulum causes negative-inotropiceffect of AEA reported in earlier studies. In line with this hypoth-esis, although caffeine induced contractures and myofilamentsensitivity to Ca2+ remained unchanged, electrically induced Ca2+

transients were significantly depressed by AEA; further suggest-ing that Ca2+-induced Ca2+ release was impaired in the presence ofAEA.

Although, this is the first patch clamp study investigating AEAaction on the cardiac APs, an earlier report using intracellularrecording methods in rat papillary muscle fibers reported that AEA,in the concentration range of as low as 1–100 nM potently inhibitedAP durations in an AM251 sensitive-manner, suggesting that theactivation of CB1 receptors mediates the negative inotropic actionsof AEA [44]. However, under our experimental conditions, changeson neither amplitudes nor kinetics of epicardial and endocardialAPs were detectable until 1 �M concentration of AEA. Importantly,the effects of AEA on the duration of both types of APs werenot reversed in the presence of CB receptor antagonists tested;

AM251 and AM630 (Fig. 10). In addition, AEA continued to affectAPD after PTX-pretreatment (Fig. 10 and supplement Fig. 4). It islikely that differences in methods (patch clamp versus intracellu-lar sharp electrode recording resulting in lower resting membrane

116 L.T. Al Kury et al. / Cell Calciu

Fig. 10. Effects of cannabinoid receptor antagonists and pertussis toxin treatmentson AEA-induced changes in myocyte excitability. (A) Effect of CB1 receptor antag-onist AM251 (1 �M) on AEA-induced changes in AP duration in endocardial andepicardial myocytes. APD60 was presented in control (dark gray bars) and in thepresence of 1 �M AEA (light gray bars), 1 �M AM251 (dark gray cross hatched bars),and 1 �M AEA + 1 �M AM251 (light gray hatched bars). (B) Effect of CB2 receptorantagonist AM630 (1 �M) on AEA-induced changes in AP duration in endocardialand epicardial myocytes. APD60 was presented in control (dark gray bars) and inthe presence of 1 �M AEA (light gray bars), 1 �M AM630 (dark gray cross hatchedbars), and 1 �M AEA + 1 �M AM630 (light gray hatched bars). (C) Effect of pertus-sis toxin (PTX) treatments on AEA-induced changes in AP duration in endocardialand epicardial myocytes. Changes are presented in control (dark gray bars), in pres-ence of 1 �M AEA (light gray bars), in control + PTX (dark gray cross hatched bars),and in presence of 1 �M AEA + PTX (light gray hatched bars). Data are shown asmeans ± S.E.M., from 5 to 7 myocytes of each type; *indicate statistically significantdifference at the level of P < 0.05.

m 55 (2014) 104– 118

potentials) and preparations (ventricular myocytes versus intactmuscle fibers with nerve endings and gap junction connections)used in these studies may account for some of the discrepancies.In earlier studies AEA has been reported to inhibit noradrenalinerelease from atrium subjected to electrical field stimulation [30]and enhance vagal activity (for a review [6]) in in vivo and in in vitromuscle preparations. However, this uptake mechanism is not likelyto be involved in our studies on isolated ventricular myocytes.

Involvement of cannabinoid-receptors in the negative inotropicactions of cannabinoids has been reported in several earlier studies[9–12]. However, the results of these investigations have not beenconclusive (for reviews [6–8]). Both cannabinoid receptor depend-ent and independent mechanisms have been suggested [6]. WhileFord et al. [9] showed in rat cardiac muscle, that the inhibitoryeffect of AEA on contractility was not reversed in the presenceof CB1 (SR141716A) and CB2 (SR144528) receptor antagonists,Bonz et al. [10] reported that AEA, metAEA, and HU-210 decreasedcontractile performance in human atrial muscle via activation ofCB1 receptors. In another study in rat atria, AEA was suggestedto have negative and positive inotropic effects mediated by theactivation of CB1 and CB2 receptors, respectively [11]. Under ourexperimental conditions, two structurally different CB1 antago-nists AM251 (0.3 �M) and SR141716 (0.3 �M) were not able toreverse the inhibitory effect of AEA on cardiomyocyte shorten-ing. Similarly, two different CB2 antagonists AM630 (0.3 �M) andSR144528 (0.3 �M) failed to antagonize AEA-induced decrease ofcardiomyocyte shortening. However, at higher concentrations suchas 1 �M, these antagonists themselves showed inhibitory actionson cardiomyocyte shortening (n = 9–12; data are not shown). To ourknowledge, negative inotropic actions of relatively high concen-trations of AM251 and AM630 have not been reported previously.However, negative inotropic actions of SR141716 and SR144528on the contractile functions of isolated rat heart have also beenattributed to direct actions of SR141716 and SR144528 on thecontractility of cardiomyocytes [45]. Earlier studies on cardiac mus-cle and other preparations also indicate that cannabinoid receptorantagonists with different chemical structures can have off-targetbinding sites ([9,29,46,47], for a review [29]). In summary, basedon the insensitivity of the effect of AEA on myocyte shorteningto CB1 and CB2 antagonists, as well as to the pretreatments withPTX and NEM, it is likely that AEA decreases myocyte shorteningand shortens AP duration in a manner that is independent of CB1and CB2 cannabinoid receptors. In agreement with these results,AEA, at similar or higher concentrations, has been shown to inhibitdirectly the functions of voltage-gated Na+ channels in neuronalstructures [48–50], L-type Ca2+ channels [51,52] and various typesof K+ channels (for a review [34]).

AEA belongs to a group of fatty acid-based molecules calledlong-chain N-acylethanolamines (NAEs) which are produced abun-dantly in response to tissue necrosis and cellular stress [53,54].In fact, accumulation of NAEs was first observed in experimentalmyocardial infarction induced by ligation of coronary arteries incanine heart ([37,55], for a review [56]). It was demonstrated thatNAE content increases up to 500 nmol/g (approximately 500 �M) ininfarcted areas of canine heart during ischemia [55]. Although AEAconstitutes minor (1–3%) portion of total NAE levels [56], the resultsof this study may have important implications regarding the con-tractile responses of ventricular myocytes to ischemia and cellularstress [53,54,56]. We have previously reported that major NAEsspecies such as N-stearylethanolamine and N-oleoylethanolamineproduced during ischemia have significant effects on the ampli-tudes and kinetics of APs and accompanying ionic currents that

could account for the negative inotropic actions of these com-pounds on ventricular myocytes [57]. Similar to AEA, other NAEs,metabolic degradation products of NAEs, and structurally relatedcompounds have also been shown to modulate functions of

Calciu

vKmmioo

fiseasfitashat

C

A

sLmmt

A

t

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

L.T. Al Kury et al. / Cell

oltage-gated Ca2+ [57,51,58,59], Na+[48,51], and ATP modulated+ channels [60]. In this context, it is important to note that metAEA,etabolically stable analog of AEA, also decreased the shortening ofyocytes. Furthermore, AEA continued to inhibit myocyte shorten-

ng after pretreatment of these cells with URB597, a potent inhibitorf fatty acid amide hydrolase, suggesting that degradation productsf AEA are not involved in observed effects of this compound.

Shortening of AP duration by AEA can be beneficial or harm-ul, depending on the underlying pathology. Thus, during acuteschemia, in which the duration of the cardiac APD is alreadyhortened, a further decrease should be proarrhythmic [61]. How-ver, the APD shortening should be beneficial in preventing thoserrhythmias caused by triggered activities observed in conditionsuch as heart failure [61,62]. In conclusion, the results indicateor the first time that AEA inhibits myocyte contractility by act-ng on multiple mechanisms. In the present study, AEA decreasedhe duration of APs and modulated the activity of Ca2+-ATPase in

CB1 and CB2 receptor-independent manner. Considering mas-ive release of various NAEs including AEA during ischemia andypoxic conditions, further understanding of their target proteinsnd action mechanisms would help in the development of betterreatment modalities for these pathological conditions.

onflict of interest

Authors have no conflict of interest in this study.

cknowledgments

This study was supported by the United Arab Emirates Univer-ity Research Funds. Research in our laboratory is also supported byABCO partner of Sigma–Aldrich. We cordially thank Mr. Muham-ad A. Qureshi for his technical help in myocyte isolation andyocyte shortening experiments. This article has been submitted

o fulfill in part the thesis requirements for Ms. Lina Al Kury.

ppendix A. Supplementary data

Supplementary data associated with this article can be found, inhe online version, at http://dx.doi.org/10.1016/j.ceca.2013.12.005.

eferences

[1] V. Di Marzo, L. De Petrocellis, T. Bisogno, The biosynthesis, fate and pharma-cological properties of endocannabinoids, Handb. Exp. Pharmacol. 168 (2005)147–185.

[2] L.O. Hanus, R. Mechoulam, Novel natural and synthetic ligands of the endo-cannabinoid system, Curr. Med. Chem. 17 (2010) 1341–1359.

[3] R.G. Pertwee, A.C. Howlett, M.E. Abood, et al., International Union of Basic andClinical Pharmacology. LXXIX. Cannabinoid receptors and their ligands: beyondCB(1) and CB(2), Pharmacol. Rev. 62 (2010) 588–631.

[4] S. Batkai, P. Pacher, Endocannabinoids and cardiac contractile function: patho-physiological implications, Pharmacol. Res. 60 (2009) 99–106.

[5] F. Montecucco, V. Di Marzo, At the heart of the matter: the endocannabinoidsystem in cardiovascular function and dysfunction, Trends Pharmacol. Sci. 33(2012) 331–340.

[6] B. Malinowska, M. Baranowska-Kuczko, E. Schlicker, Triphasic blood pressureresponses to cannabinoids: do we understand the mechanism? Br. J. Pharmacol.165 (2012) 2073–2088.

[7] M.D. Randall, D.A. Kendall, S. O’Sullivan, The complexities of the cardiovascularactions of cannabinoids, Br. J. Pharmacol. 142 (2004) 20–26.

[8] V.E. Mendizabal, E. Dler-Graschinsky, Cannabinoids as therapeutic agents incardiovascular disease: a tale of passions and illusions, Br. J. Pharmacol. 151(2007) 427–440.

[9] W.R. Ford, S.A. Honan, R. White, C.R. Hiley, Evidence of a novel site mediating

anandamide-induced negative inotropic and coronary vasodilatator responsesin rat isolated hearts, Br. J. Pharmacol. 135 (2002) 1191–1198.

10] A. Bonz, M. Laser, S. Kullmer, et al., Cannabinoids acting on CB1 recep-tors decrease contractile performance in human atrial muscle, J. Cardiovasc.Pharmacol. 41 (2003) 657–664.

[

[

m 55 (2014) 104– 118 117

11] L. Sterin-Borda, C.F. Del Zar, E. Borda, Differential CB1 and CB2 cannabinoidreceptor-inotropic response of rat isolated atria: endogenous signal transduc-tion pathways, Biochem. Pharmacol. 69 (2005) 1705–1713.

12] Z. Su, L. Preusser, G. Diaz, et al., Negative inotropic effect of a CB2 agonist A-955840 in isolated rabbit ventricular myocytes is independent of CB1 and CB2receptors, Curr. Drug Saf. 6 (2011) 277–284.

13] F.C. Howarth, M.A. Qureshi, E. White, Effects of hyperosmotic shrinking on ven-tricular myocyte shortening and intracellular Ca2+ in streptozotocin-induceddiabetic rats, Pflugers Arch. 444 (2002) 446–451.

14] J.W. Bassani, W. Yuan, D.M. Bers, Fractional SR Ca release is regulated by trig-ger Ca and SR Ca content in cardiac myocytes, Am. J. Physiol. 268 (1995)C1313–C1319.

15] H.A. Spurgeon, W.H. DuBell, M.D. Stern, et al., Cytosolic calcium and myofil-aments in single rat cardiac myocytes achieve a dynamic equilibrium duringtwitch relaxation, J. Physiol. 447 (1992) 83–102.

16] A.V. Zima, J.A. Copello, L.A. Blatter, Differential modulation of cardiac and skele-tal muscle ryanodine receptors by NADH, FEBS Lett. 547 (2003) 32–36.

17] S.M. Dunn, Voltage-dependent calcium channels in skeletal muscle transversetubules. Measurements of calcium efflux in membrane vesicles, J. Biol. Chem.264 (1989) 11053–11060.

18] Y. Sagara, G. Inesi, Inhibition of the sarcoplasmic reticulum Ca2+ transportATPase by thapsigargin at subnanomolar concentrations, J. Biol. Chem. 266(1991) 13503–13506.

19] R. Zucchi, S. Ronca-Testoni, G. Yu, P. Galbani, G. Ronca, M. Mariani, Effect ofischemia and reperfusion on cardiac ryanodine receptors—sarcoplasmic reti-culum Ca2+ channels, Circ. Res. 74 (1994) 271–280.

20] V. Lopez-Miranda, E. Herradon, M.T. Dannert, A. Alsasua, M.I. Martin, Anan-damide vehicles: a comparative study, Eur. J. Pharmacol. 505 (2004) 151–161.

21] R.S. Danziger, M. Sakai, M.C. Capogrossi, H.A. Spurgeon, R.G. Hansford, E.G.Lakatta, Ethanol acutely and reversibly suppresses excitation–contraction cou-pling in cardiac myocytes, Circ. Res. 68 (1991) 1660–1668.

22] L.M. Delbridge, P.J. Connell, P.J. Harris, T.O. Morgan, Ethanol effects on car-diomyocyte contractility, Clin. Sci. (Lond.) 98 (2000) 401–407.

23] M. Bebarova, P. Matejovic, M. Pasek, et al., Effect of ethanol on action potentialand ionic membrane currents in rat ventricular myocytes, Acta Physiol. (Oxf.)200 (2010) 301–314.

24] V. Abadji, S. Lin, G. Taha, et al., (R)-Methanandamide: a chiral novel anandamidepossessing higher potency and metabolic stability, J. Med. Chem. 37 (1994)1889–1893.

25] S. Kathuria, S. Gaetani, D. Fegley, et al., Modulation of anxiety through blockadeof anandamide hydrolysis, Nat. Med. 9 (2003) 76–81.

26] R.G. Pertwee, The pharmacology of cannabinoid receptors and their ligands: anoverview, Int. J. Obes. (Lond.) 30 (2006) S13–S18.

27] D. Shire, B. Calandra, M. Bouaboula, et al., Cannabinoid receptor interactionswith the antagonists SR 141716A and SR 144528, Life Sci. 65 (1999) 627–635.

28] M. Oz, K.H. Yang, T.S. Shippenberg, L.P. Renaud, M.J. O’Donovan, Cholecys-tokinin B-type receptors mediate a G-protein-dependent depolarizing actionof sulphated cholecystokinin ocatapeptide (CCK-8s) on rodent neonatal spinalventral horn neurons, J. Neurophysiol. 98 (2007) 1108–1114.

29] R. Baur, J. Gertsch, E. Sigel, The cannabinoid CB1 receptor antagonists rimon-abant (SR141716) and AM251 directly potentiate GABA(A) receptors, Br. J.Pharmacol. 165 (2012) 2479–2484.

30] C.E. Goodfellow, M. Glass, Anandamide receptor signal transduction, Vitam.Horm. 81 (2009) 79–110.

31] S.Y. Zhuang, D. Bridges, E. Grigorenko, et al., Cannabinoids produce neuro-protection by reducing intracellular calcium release from ryanodine-sensitivestores, Neuropharmacology 48 (2005) 1086–1096.

32] Y.A. Mahmmoud, M. Gaster, Uncoupling of sarcoplasmic reticulum Ca2+-ATPaseby N-arachidonoyl dopamine. Members of the endocannabinoid family as ther-mogenic drugs, Br. J. Pharmacol. 166 (2012) 2060–2069.

33] C. Antzelevitch, S. Sicouri, S.H. Litovsky, et al., Heterogeneity within the ven-tricular wall. Electrophysiology and pharmacology of epicardial, endocardial,and M cells, Circ. Res. 69 (1991) 1427–1449.

34] M. Oz, Receptor-independent actions of cannabinoids on cell membranes: focuson endocannabinoids, Pharmacol. Ther. 111 (2006) 114–144.

35] E.J. Ishac, L. Jiang, K.D. Lake, K. Varga, M.E. Abood, G. Kunos, Inhibition of exo-cytotic noradrenaline release by presynaptic cannabinoid CB1 receptors onperipheral sympathetic nerves, Br. J. Pharmacol. 118 (1996) 2023–2028.

36] M. Oz, V. Jaligam, S. Galadari, G. Petroianu, Y.M. Shuba, T.S. Shippenberg, Theendogenous cannabinoid, anandamide, inhibits dopamine transporter functionby a receptor-independent mechanism, J. Neurochem. 112 (2010) 1454–1464.

37] D.E. Epps, F. Mandel, A. Schwartz, The alteration of rabbit skeletal sarcoplasmicreticulum function by N-acylethanolamine, a lipid associated with myocardialinfarction, Cell Calcium 3 (1982) 531–543.

38] J.V. Mombouli, G. Schaeffer, S. Holzmann, G.M. Kostner, W.F. Graier,Anandamide-induced mobilization of cytosolic Ca2+ in endothelial cells, Br. J.Pharmacol. 126 (1999) 1593–1600.

39] G.K. Rao, N.E. Kaminski, Cannabinoid-mediated elevation of intracellular cal-cium: a structure–activity relationship, J. Pharmacol. Exp. Ther. 317 (2006)820–829.

40] P.L. De, V. Di Marzo, Role of endocannabinoids and endovanilloids in Ca2+

signalling, Cell Calcium 45 (2009) 611–624.41] C. Dettbarn, P. Palade, Arachidonic acid-induced Ca2+ release from isolated

sarcoplasmic reticulum, Biochem. Pharmacol. 45 (1993) 1301–1309.42] D.G. Demuth, A. Molleman, Cannabinoid signalling, Life Sci. 78 (2006) 549–

563.

http://dx.doi.org/10.1016/j.ceca.2013.12.005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0005http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0010http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0015http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0020http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0020http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0020http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0020http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0020http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0020http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0020http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0020http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0020http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0020http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0020http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0020http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0020http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0020http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0020http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0020http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0020http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0020http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0020http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0025http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0030http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0035http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0040http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0040http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0040http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0040http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0040http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0040http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0040http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0040http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0040http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0040http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0040http://refhub.elsevier.com/S0143-4160(13)00158-9/sbref0040http://refhub.elsevier.com/S014