Protective activation of the endocannabinoid system during ischemia in dopamine neurons

www.elsevier.com/locate/ynbdi

Neurobiology of Disease 24 (2006) 15 – 27

Protective activation of the endocannabinoid system during

ischemia in dopamine neurons

Miriam Melis,a,b,*,1 Giuliano Pillolla,a,b,1 Tiziana Bisogno,c Alberto Minassi,c

Stefania Petrosino,c Simona Perra,a,b Anna Lisa Muntoni,a,d Beat Lutz,e Gian Luigi Gessa,a,b,c

Giovanni Marsicano,e Vincenzo Di Marzo,c and Marco Pistisa,b

aCentre of Excellence ‘‘Neurobiology of Addiction’’, University of Cagliari, Italyb‘‘B.B. Brodie’’ Department of Neuroscience, University of Cagliari, 09042 Monserrato (CA), ItalycC.N.R. Endocannabinoid Research Group, Institute of Biomolecular Chemistry, Pozzuoli (NA), ItalydC.N.R. Institute of Neuroscience, c/o ‘‘B.B. Brodie’’ Department of Neuroscience, University of Cagliari, 09042 Monserrato (CA), ItalyeDepartment of Physiological Chemistry, Johannes Gutenberg University, Mainz, Germany

Received 2 January 2006; revised 6 April 2006; accepted 24 April 2006

Available online 8 June 2006

Endocannabinoids act as neuroprotective molecules promptly released

in response to pathological stimuli. Hence, they may represent one

component of protection and/or repair mechanisms mobilized by

dopamine (DA) neurons under ischemia. Here, we show that the

endocannabinoid 2-arachidonoyl-glycerol (2-AG) plays a key role in

protecting DA neurons from ischemia-induced altered spontaneous

activity both in vitro and in vivo. Accordingly, neuroprotection can be

elicited through moderate cannabinoid receptor type-1 (CB1) activa-

tion. Conversely, blockade of endocannabinoid actions through CB1

receptor antagonism worsens the outcome of transient ischemia on DA

neuronal activity. These findings indicate that 2-AG mediates neuro-

protective actions by delaying damage and/or restoring function of DA

cells through activation of presynaptic CB1 receptors. Lastly, they

point to CB1 receptors as valuable targets in protection of DA neurons

against ischemic injury and emphasize the need for a better

understanding of endocannabinoid actions in the fine control of DA

transmission.

D 2006 Elsevier Inc. All rights reserved.

Keywords: CB1; Dopamine; Endocannabinoid; Ischemia; Midbrain; Neuro-

protection; Retrograde signal

0969-9961/$ - see front matter D 2006 Elsevier Inc. All rights reserved.

doi:10.1016/j.nbd.2006.04.010

Abbreviations: AEA, anandamide; ATZ, 3-amino-2,4tiazole; 2-AG, 2-

arachidonoyl-glycerol; CB1, cannabinoid type 1 receptor; DA, dopamine;

DSE, depolarization-induced suppression of excitation; EPSCs, excitatory

post-synaptic currents; MCS, mercaptosuccinate; VTA, ventral tegmental

area.

* Corresponding author. ‘‘B.B. Brodie’’ Department of Neuroscience,

University of Cagliari, Cittadella Universitaria, 09042 Monserrato (CA),

Italy. Fax: +39 070 6754320.

E-mail address: myriam@unica.it (M. Melis).1 These authors contributed equally to this work.

Available online on ScienceDirect (www.sciencedirect.com).

Introduction

Endocannabinoids are a family of lipid molecules implicated in

several functions spanning from neuromodulation to neuroprotec-

tion (Mechoulam et al., 2002; van der Stelt et al., 2002;

Mechoulam and Lichtman, 2003; Piomelli, 2003; Lutz, 2004;

van der Stelt and Di Marzo, 2005). Anandamide (AEA) and 2-

arachidonoyl-glycerol (2-AG), produced ‘‘on demand’’ (Di Marzo

et al., 1999), are the best characterized members of this family.

These two lipid signaling molecules are synthesized and metab-

olized by different enzymes and display distinct pharmacological

profiles, as well as regional diversity (De Petrocellis et al., 2004).

AEA and 2-AG have been found in the ventral tegmental area

(VTA) (Marinelli et al., 2003), where they serve as retrograde

messengers (Melis et al., 2004a,b) acting at cannabinoid receptor

type 1 (CB1), which is localized on both excitatory and inhibitory

synapses (Marinelli and Mercuri, 2004).

The principal neurons of the VTA contain dopamine (DA) and

play a dominant role in motivation, reward-related behaviors, and

cognition (Schultz, 2002). Prolonged depolarization of DA neurons

induces a Ca2+-dependent transient release of endocannabinoids

which, by acting retrogradely, activate presynaptic CB1 receptors

and suppress glutamate release (Melis et al., 2004a). However, DA

neurons rarely undergo prolonged depolarization suggesting that

endocannabinoids might serve to limit pathological excitation of

VTA DA neurons. Accordingly, both endogenous and exogenous

cannabinoids have been reported to be neuroprotective following

many cerebral insults as well as ischemia (Mechoulam et al., 2002;

Franklin et al., 2003; Panikashvili et al., 2005; van der Stelt and Di

Marzo, 2005). Despite the fact that hypoxic– ischemic brain injury

accounts for changes of the DA system, and ultimately for dystonia

(Burke et al., 1992) or post-ischemic Parkinsonism (Critchley,

1929; De Reuck et al., 2001), to date there are no studies

M. Melis et al. / Neurobiology of Disease 24 (2006) 15–2716

investigating the electrophysiological responses of DA cells to

ischemia, and only a few examined the response to either hypoxia

or hypoglycemia (Hausser et al., 1991; Jiang et al., 1994; Mercuri

et al., 1994a,b; Guatteo et al., 1998a,b; Marinelli et al., 2000, 2001;

Geracitano et al., 2005).

Additionally, despite the potential significance of a better

understanding of the mechanisms underlying endocannabinoid

signaling in the VTA, several key pieces of information are yet

unknown. Primarily, are endocannabinoids released from VTA DA

cells under pathological scenarios? Can exogenous application of

cannabinoids protect DA neurons from ischemic-induced changes

in firing activity? Is CB1 receptor activation involved in the

defense of DA cells? Which of the two major endocannabinoids is

mostly involved in these putative protective mechanisms?

Ergo, we investigated endocannabinoid signaling in the VTA in

models of transient ischemia both in vitro and in vivo.

Materials and methods

All experiments were carried out in strict accordance with the

care and use of animals approved by the American Physiological

Society and EEC Council Directive of 24 November 1986 (86/

609). All efforts were made to minimize pain and suffering and to

reduce the number of animals used.

In vitro recordings

The preparation of VTA slices was as described previously

(Johnson and North, 1992). Briefly, male Sprague–Dawley rats

(14–28 d, Harlan Nossan, San Pietro al Natisone, Italy) or

FAAH�/� and FAAH+/+ mice (see below for detailed information)

were anesthetized with halothane and killed. A block of tissue

containing the midbrain was rapidly dissected and sliced in the

horizontal plane (300 and 230 Am for rat and mouse slices,

respectively) with a vibratome (Campden Instruments Ltd.) in ice-

cold low-Ca2+ solution containing (in mM) 126 NaCl, 1.6 KCl, 1.2

NaH2PO4, 1.2 MgCl2, 0.625 CaCl2, 18 NaHCO3, and 11 glucose.

Slices were transferred to a holding chamber with artificial

cerebrospinal fluid (ACSF, 37-C) saturated with 95% O2 and 5%

CO2 containing (in mM) 126 NaCl, 1.6 KCl, 1.2 NaH2PO4, 1.2

MgCl2, 2.4 CaCl2, 18 NaHCO3, and 11 glucose. Slices (two per

animal) were allowed to recover for at least 1 hr before being

placed (as hemislices) in the recording chamber and superfused

with the ACSF (37-C) saturated with 95% O2 and 5% CO2. Cells

were visualized with an upright microscope with infrared

illumination (Axioskop FS 2 plus, Zeiss), and whole-cell current-

and voltage-clamp recordings (one per hemislice) were made by

using an Axopatch 200B amplifier (Axon Instruments, CA).

Current-clamp experiments were made with electrodes filled with

a solution containing the following (in mM) 144 KCl, 10 HEPES,

3.45 BAPTA, 1 CaCl2, 2.5 Mg2ATP, and 0.25 Mg2GTP (pH 7.2–

7.4, 275–285 mosM).

DSE experiments were made with electrodes filled with a

solution containing the following (in mM): 117 Cs methansulfonic

acid, 20 HEPES, 0.4 EGTA, 2.8 NaCl, 5 TEA-Cl, 2.5 Mg2ATP,

and 0.25 Mg2GTP (pH 7.2–7.4, 275–285 mosM). Picrotoxin (100

AM) was added to the ACSF only for recording excitatory post-

synaptic currents (EPSCs), to block GABAA receptor-mediated

inhibitory PSCs. Experiments were begun only after series

resistance had stabilized (typically 15–40 MV). Series and input

resistance were monitored continuously on-line with a 5 mV

depolarizing step (25 ms). Data were filtered at 2 kHz, digitized at

10 kHz, and collected on-line with acquisition software (pClamp

8.2, Axon Instruments, CA). DA neurons from the anterior VTA

were identified by the presence of a large Ih current (Johnson and

North, 1992) that was assayed immediately after break-in, using a

series of incremental 10-mV hyperpolarizing steps from a holding

potential of �70 mV. A bipolar stainless steel stimulating electrode

(FHC, USA) was placed 100 Am rostral to the recording electrode

and was used to stimulate at a frequency of 0.1 Hz. AMPA-

mediated EPSCs were recorded at �70 mV. The amplitudes of

AMPA EPSCs were calculated by taking a 1-ms window around

the peak of the EPSC and comparing this with the 5-ms window

immediately before the stimulation artifact. The magnitude of DSE

was measured as percentage of the mean amplitude of consecutive

EPSCs after the voltage step (+40 mV, 0.5–10 s) relative to that of

five EPSCs before the step (Melis et al., 2004a). To induce

ischemia, the control solution was substituted with an aglycemic (0

glucose) ACSF, saturated with 95% N2–5% CO2 instead of 95%

O2–5% CO2. The period of ischemia was for 7 min. Each slice

received only a single ischemia protocol and drug exposure.

Averaged data from different experiments are presented as mean TSEM. Statistical significance was assessed using one- or two-way

analysis of variance (ANOVA) for repeated measures followed

either by Dunnett’s or t test (with Welch’s correction), where

appropriate.

In vivo recordings

Rats (male Sprague–Dawley, 250–350 g, Harlan, San Pietro al

Natisone, Italy) were anesthetized with urethane (1.5 g/kg, i.p.),

and their femoral vein was cannulated for intravenous administra-

tion of pharmacological agents. Rats were then prepared for

transient forebrain ischemia (Pulsinelli and Brierley, 1979;

Pulsinelli et al., 1982; Volpe et al., 1984). Briefly, the common

carotids were exposed and encircled with silastic ligatures, and the

vertebral arteries were permanently occluded by electrocautery.

Then rats were placed in the stereotaxic apparatus (Kopf, Tujunga

CA, USA) with their body temperature maintained at 37 T 1-C by a

heating pad. Thereafter, the scalp was retracted, and one burr hole

was drilled above the VTA (AP + 2.0 mm from lambda, L 0.3–0.6

mm from midline) for the placement of a recording electrode. VTA

was localized according to the stereotaxic atlas of Paxinos and

Watson (1997). Single unit activity of neurons located in the VTA

(V 7.2–8.0 mm from the cortical surface) was recorded extracel-

lularly with glass micropipettes filled with 2% pontamine sky blue

dissolved in 0.5 M sodium acetate (impedance 2–5 MV). Single

unit activity was filtered (bandpass 500–5000 Hz), and individual

spikes were isolated by means of a window discriminator (Digi-

timer, Hertfordshire, UK), displayed on a digital storage oscillo-

scope (Tektronics, Marlow, UK) and recorded by a VCR.

Experiments were sampled on line and off line by a computer

connected to CED 1401 interface (Cambridge Electronic Design,

Cambridge, UK). Recording electrodes were slowly lowered into

the VTA via a micromanipulator (Narishige, Tokyo, Japan). Single

units were isolated and identified according to the already

published criteria (Guyenet and Aghajanian, 1978; Grace and

Bunney, 1983, 1984; Ungless et al., 2004). Bursts were defined as

the occurrence of two spikes at interspike interval <80 ms and

terminated when the interspike interval exceeded 160 ms (Grace

and Bunney, 1984). Baseline firing rates were obtained for at least

M. Melis et al. / Neurobiology of Disease 24 (2006) 15–27 17

5 min, then SR141716A, or vehicle, was administered intrave-

nously in a single dose (1 mg/kg, i.v.; 1 ml/kg). URB597 was

administered (0.5 mg/kg, i.p.) immediately after the anesthesia.

Thereafter, forebrain ischemia was produced by tightening the

carotid ligatures. The carotid ligatures were released after 7 min.

During ischemia, DA neurons were continuously recorded.

Only one cell was recorded per rat.

At the end of each recording section, DC current (10 AA for 15

min) was passed through the recording electrode in order to eject

Pontamine sky blue, which allowed the identification of the

recorded cells. Brains were removed and fixed in 8% formalin

solution. The position of the electrodes was microscopically

identified on serial sections (60 Am) stained with Cresyl violet.

Data obtained were always normalized to pre-ischemic baseline

levels and analyzed utilizing one-way ANOVA or two-way

ANOVA where appropriate. Post hoc multiple comparisons were

made using the Dunnett’s test.

Pharmacology

WIN55,212-2, AM281, AP-5, SKF96365 were purchased from

Tocris Cookson (UK). CNQX, catalase, mercaptosuccinate, ATZ

and other chemicals were purchased from Sigma Aldrich (Italy).

URB597 was purchased from Alexis (Vincibiochem, Italy). AA-

5HT and THL were provided by Dr. V. Di Marzo (Endocannabi-

noid Research Group, Institute of Biomolecular Chemistry,

Pozzuoli, Italy). O-3640 was a generous gift from Drs. B.R.

Martin and R. Razdan (Organix Inc., Woburn, MA). SR141716A

was a generous gift from Dr. G. Le Fur (Sanofi-Aventis Recherche,

France).

In vitro recordings

For extracellular application, drugs were applied in known

concentrations to the superfusion medium. For intracellular

application, drugs were applied in known concentrations to the

internal solution. All the drugs were dissolved in DMSO. The final

concentration of DMSO was <0.01%.

In vivo recordings

SR141716Awas emulsified in 1% Tween 80, and URB597 was

dissolved in DMSO. Drugs were then diluted in saline solution and

sonicated.

FAAH knockout mice

FAAH knockout mice were generated as described elsewhere

(Cravatt et al., 2001). Homozygous mutant mice (FAAH�/�) and

wild-type (FAAH+/+) littermates derived from heterozygous

matings were genotyped as described (Azad et al., 2004). In the

present study, juvenile mutant male mice aged between 3 and 5

weeks were used. Animals were housed in groups under standard

laboratory conditions (12-h light/dark cycle) with food and water

available ad libitum. In all cases, the experimenter was blind to the

genotype.

Biochemical assays and endocannabinoid measurements

Enzymatic assays for fatty acid amide hydrolase (FAAH)

activity and diacylglycerol lipase (DAG lipase) activity were

carried out on membrane fractions prepared from VTA slices, as

described in Bisogno et al. (2006). Slices were obtained as

described above (see in vitro recordings), then transferred to the

recording chamber where they were subjected to the ischemia

protocol. Slices were incubated in the presence of URB597 (10

nM) or O-3640 (5 AM) for at least 1 h before the ischemia protocol.

Activity data are reported as pmols of [14C]AEA or sn-1-

[14C]oleoyl-2-arachidonoyl-glycerol hydrolyzed per milligram of

protein and per minute, for FAAH and DAGL, respectively. The

levels of AEA and 2-AG in pre-purified lipid extracts from control

or ischemic VTA slices, or sections of tissue containing the

mesencephalon from in vivo experiments, were measured as

described by Marsicano et al. (2003), by using isotope-dilution/

liquid chromatography-mass spectrometry. Concentrations are

reported as pmol/mg extracted lipids, in the case of VTA slices,

or pmol/g wet tissue weight, in the case of brains. To obtain brain

sections, animals were decapitated immediately after the ischemia

protocol. Brains were rapidly removed, and sections of mesen-

cephalon dissected were immediately frozen in liquid nitrogen and

stored at � 80-C until assayed. Means T SEM were compared by

using analysis of variance followed by the Bonferroni’s post hoc

test.

Results

Endocannabinoids protect DA neurons during ischemia in vitro

We first investigated the effects of an ischemic insult on VTA

DA cells. To mimic ischemia, a temporary lack of oxygen and

glucose (0 glucose, O2 deprivation, 7 min) was applied replacing

oxygenated ACSF. Under current-clamp mode, VTA DA neuronal

response to ischemia was a membrane depolarization ranging from

4 to 20 mV (mean 10.3 T 3 mV; n = 9 Figs. 1A and B) that

caused, after an initial increase, an irreversible block of firing

activity (6 out of 9 cells; Fig. 1A). These changes occurred within

7 min from the beginning of ischemia (7.4 T 1 min, Figs. 1A and

C), and only three cells recovered the spontaneous firing activity

within 15 min from reperfusion with oxygenated ACSF containing

glucose (11 mM). Since endocannabinoids are able to decrease

excitation of VTA DA cells (Melis et al., 2004a), we hypothesized

that a similar on demand activation of the endocannabinoid system

might occur in response to an ischemic insult. To begin to explore

this possibility, we examined the effects of ischemia in the

presence of the CB1 receptor antagonist AM281 (500 nM). As

compared to controls, we observed a larger and irreversible

membrane depolarization (mean 23.9 T 1 mV; n = 5, P < 0.002

versus control; Figs. 1A and B), resulting in a faster block of

spontaneous activity (2 T 0.3 min, P < 0.005 versus control; Figs.

1A and C) that never resumed firing activity even long after

reperfusion (¨30 min).

AEA and 2-AG are the main endocannabinoids in the brain

(Freund et al., 2003) and are synthesized and degraded through

different pathways (Piomelli, 2003; De Petrocellis et al., 2004).

Since 2-AG is the most abundant endocannabinoid in tissues

(Mechoulam et al., 1995; Sugiura et al., 1995; Stella et al., 1997)

and is also the retrograde messenger for train-induced suppression

of excitation in the VTA (Melis et al., 2004b), we first examined its

role during ischemia. 2-AG is synthesized from diacylglycerol

(DAG), which is then hydrolyzed to the 2-monoacylglycerol

selectively by a sn-1-DAG lipase (Bisogno et al., 1997; Stella et

al., 1997; Di Marzo et al., 1998; Bisogno et al., 2003). 2-AG

signaling is then terminated by the enzyme monoglyceride lipase

Fig. 1. Endocannabinoids protect VTA DA neurons during ischemia in vitro. (A) The voltage traces depict the response of a VTA dopaminergic cell to ischemia

(0 glucose, 0 oxygen, 7 min) under control conditions (top left), in the presence of either the CB1 receptor antagonist, AM281 (bottom left), the sn-a-DAG

lipase inhibitor, O-3640 (top right), or the FAAH inhibitor, URB597 (bottom right). The ischemia led to a membrane depolarization, which was followed by a

depolarization block (complete cessation of spontaneous activity). The dashed line indicates the basal voltage membrane. (B) The bar graph shows the peak of

change of voltage membrane during the ischemia protocol, taken at the end of the protocol, under control conditions, in the presence of either AM281, O-3640

or URB597. The numbers above the columns indicate the number of neurons used for each determination. The asterisks represent statistically significant

difference from control (*P < 0.05, **P < 0.005). O-3640 and URB597 were applied by means of the recording pipette. (C) The graph shows the latency

between the beginning of the ischemia protocol (time 0) and the onset of depolarization block under control conditions and in the presence of either AM281, O-

3640 or URB597 for each of the experiments in panel B. The asterisks represent statistically significant difference from control (*P < 0.005, **P < 0.03).

M. Melis et al. / Neurobiology of Disease 24 (2006) 15–2718

(MGL; Dinh et al., 2002). Therefore, to clarify whether or not 2-

AG activates CB1 receptors following an ischemic injury, we filled

the postsynaptic cell with O-3640 (5 AM), an inhibitor of sn-1-

DAG lipase activity (Bisogno et al., 2006), and observed that

ischemia produced a larger and permanent membrane depolariza-

tion (mean 20.6 T 2.2 mV; n = 6, P < 0.03 versus control; Fig. 1B)

resulting in a faster and irreversible block of spontaneous activity

(3.5 T 0.7 min, P < 0.03 versus control; Fig. 1C). Given that AEA

is relatively abundant in the VTA (Marinelli et al., 2003), we next

examined its possible contribution to the electrophysiological

changes induced by an ischemic insult. AEA formation occurs

through a two-step process (Di Marzo et al., 1994; Sugiura et al.,

1995; Cadas et al., 1997; Okamoto et al., 2004), and its signaling is

terminated by the enzyme FAAH (Hillard et al., 1995; Ueda et al.,

1995; Cravatt et al., 1996). Therefore, we filled the postsynaptic

DA cell with URB597 (10 nM), an inhibitor of FAAH activity

(Kathuria et al., 2003) under transient ischemia. Ischemia effects

on DA cells were similar to control conditions, these effects being

a change in voltage membrane of 11.7 T 0.9 mV (n = 5, P > 0.05

versus control; Fig. 1B) and onset of depolarization of 6.8 T 0.6

min (n = 5, P > 0.05 versus control; Fig. 1C).

Endocannabinoids are produced during ischemic conditions in

VTA slices

In order to confirm the involvement of 2-AG, we next measured

whether this endocannabinoid is produced following ischemia in

VTA slices identical to those used for the in vitro recordings. As

shown in Table 1, 2-AG but not AEA levels were significantly

increased (2-fold, P < 0.05) under ischemic conditions. This

phenomenon was not due to increase in DAG lipase activity (Table

2), and it is likely to be caused by increased availability of DAG

precursors for 2-AG, since the enzyme most responsible for DAG

formation, PI-PLC, is known to be up-regulated during ischemia

(Strosznajder et al., 1990). We also assessed the effect of the FAAH

and DAG lipase inhibitors used during the in vitro recordings on

FAAH and DAG lipase activity in VTA slices. We found that

URB597 (10 nM) potently decreased FAAH activity in slices, thus

Table 1

Effect of ischemia on endocannabinoid levels

AEA 2-AG

VTA slice—control

(pmol/mg extracted lipids)

1.0 T 0.1 (N = 4) 16.4 T 0.7 (N = 4)

VTA slice—ischemic

(pmol/mg extracted lipids)

1.5 T 0.3 (N = 4) P > 0.05 vs. respective control 35.3 T 7.2 (N = 4) P < 0.05 vs. respective control

Mesencephalon—control

(pmol/g tissue weight)

58.0 T 7.2 (N = 4) 5120 T 460 (N = 4)

Mesencephalon—ischemic

(pmol/g tissue weight)

47.4 T 3.2 (N = 4) P > 0.05 vs. respective control 9040 T 1300 (N = 4) P < 0.05 vs. respective control

Data are means T SEM and were compared by ANOVA followed by the Bonferroni’s test. As the amount of VTA slices was too little to be weighed without

defrosting the samples, amounts are normalized to mg of extracted lipids in this case.

M. Melis et al. / Neurobiology of Disease 24 (2006) 15–27 19

demonstrating that the lack of effect of this compound observed

above was not due to ineffectiveness as a FAAH inhibitor. On the

other hand, O-3640 (5 AM) did not significantly decrease DAG

lipase activity, although a trend for a decrease versus vehicle-

treated slices was indeed observed (P = 0.15, Table 2). This latter

finding was not surprising in view of the fact that the inhibitor,

possibly because of its scarce permeability through the cell

membrane, has been reported to be most effective only when

introduced directly into the cell (Bisogno et al., 2003), i.e. by

filling the recording pipette (present results).

Endocannabinoids protect DA neurons during ischemia in vivo

To determine if this endocannabinoid neuroprotective signal is

a relevant phenomenon in vivo, we studied VTA DA neuronal

responses to ischemia in anesthetized rats. We performed a model

of reversible forebrain ischemia (Toda et al., 2002) consisting in

electrocauterizing the vertebral arteries and then subjecting the

rats to 7 min of forebrain ischemia by occluding the carotid

arteries with silastic ligatures. We recorded VTA DA cell

spontaneous activity during ischemia and recirculation and

observed a reversible decrease in firing rate (maximum decrease

to 59.4 T 9.7% of baseline firing rate; n = 8, P < 0.001, ANOVA

and Dunnett’s test; Figs. 2A and B) and burst firing (maximum

decrease to 9.3 T 3.8% of baseline burst firing; n = 5, P < 0.01,

ANOVA and Dunnett’s test, data not shown), which recovered to

baseline levels with recirculation. This observation is consistent

with an initial DA neuronal suffering observed after transient

forebrain ischemia, which is followed by more persistent damage

(Volpe et al., 1995). If endocannabinoids were released under

these circumstances, they would modulate VTA DA cell

responses to the ischemic injury. Spontaneous activity of DA

cell during ischemia in the presence of the CB1 antagonist

SR141716A (SR, 1 mg/kg, i.v.) displayed a significantly different

course (two-way ANOVA, F(treatment)1,170 = 5.07, P < 0.05).

Table 2

FAAH and DAG lipase activities under control and ischemic conditions in VTA

VTA slice FAAH (pmol � mg protein�1 � min�1)

Vehicle/Control 780 T 44 (N = 4)

Vehicle/Ischemia 943 T 131 (N = 4)

Inhibitor/Control 30 T 7 (N = 4) P < 0.001 vs. vehicle/co

Inhibitor/Ischemia 40 T 5 (N = 4) P < 0.001 vs. vehicle/isc

Effect of specific inhibitors.

Data are means T SEM and were compared by ANOVA followed by the Bonferr

‘‘Inhibitor’’ was URB597 (10 nM) in the case of FAAH, and O-3640 (5 AM) in

Hence, firing rate was increased (peaking at 156.8 T 31% of

baseline firing, n = 9, P < 0.05, ANOVA and Dunnett’s test,

Figs. 2A and C), as well as burst firing (maximum increase to

190 T 40% of baseline burst firing, n = 8; P < 0.05, data not

shown). These findings suggest an enhanced susceptibility of DA

neurons to excitatory inputs and concurrent depolarization. We

then tested whether an increase in brain AEA content would

modify the outcome of ischemia in vivo. The FAAH inhibitor

URB597, at a dose (0.5 mg/kg, i.p.) previously shown to elevate

AEA but not 2-AG levels in rat brain (Kathuria et al., 2003), was

administered immediately following anesthesia (¨1–2 h before

recordings). Consistent with our in vitro results, pretreatment with

URB597 did not change the time course of DA neuron firing rate

during ischemia and reperfusion, as compared to controls (two-

way ANOVA, F1,115 = 0.09, P > 0.05) (Fig. 2A). Accordingly,

we found that ischemia is accompanied by a significant (P <

0.05) elevation of 2-AG (¨2-fold), but not AEA, levels in brain

sections containing the mesencephalon (Table 1), similarly to

what observed above in VTA slices.

Activation of CB1 receptors protects VTA DA neurons during

ischemia

These results suggest that endocannabinoids might prevent

excessive membrane depolarization and delay the onset of

depolarization block through activation of CB1 receptors. If this

is the case, then exogenous cannabinoids, through activation of

CB1 receptors, might also serve as neuroprotective agents. Bath

application of low doses (3–30 nM) of the CB1 receptor agonist

WIN55,212-2 (WIN) during ischemia resulted in a significant

reduction of ischemia-induced membrane depolarization (WIN 3

nM: mean 1.1 T 1.3 mV, n = 6, P < 0.003 versus control, Figs.

3A and B; WIN 10 nM: mean 1.4 T 2 mV; n = 6, P < 0.006

versus control; WIN 30 nM: mean 4.7 T 1.5 mV; n = 6, P < 0.02

versus control; Fig. 3B). Additionally, WIN delayed the onset of

slices

DAG lipase (pmol � mg protein�1 � min�1)

100 T 15 (N = 4)

84 T 11 (N = 4)

ntrol 81 T 16 (N = 4) P = 0.15 vs. vehicle/control

hemia 117 T 13 (N = 4) P > 0.2 vs. vehicle/ischemia

oni’s test.

the case of DAG lipase.

M. Melis et al. / Neurobiology of Disease 24 (2006) 15–2720

depolarization block (WIN 10 nM: 12.7 T 2 min, n = 6, P < 0.05

versus control; WIN 30 nM: 11.4 T 0.6 min, n = 6, P < 0.03

versus control; Fig. 3C). Remarkably, in the presence of WIN 3

nM, of all cells tested, none ceased to fire spontaneously (v2 =

43.24, P < 0.0002; Fig. 3D). In the presence of WIN 10 nM two

out of six cells never ceased to fire spontaneously, and two that

underwent depolarization block recovered the spontaneous

activity within 2 min of reperfusion (Fig. 3D). However, dose

range and maximum efficacy at which WIN acted as neuro-

protective agent were extremely narrow. In fact, a lower and two

higher concentrations of WIN (1, 100 nM and 1 AM) neither did

significantly reduce ischemia-induced depolarization (WIN 1 nM:

mean 15.7 T 4 mV, n = 5, P > 0.5 versus control; WIN 100 nM:

mean 16.7 T 1 mV, n = 5, P > 0.5 versus control; WIN 1 AM:

mean 17.8 T 2 mV, n = 6, P > 0.1 versus control, Fig. 3B) nor

significantly delayed the onset of depolarization block (WIN 1

nM: mean 7.3 T 2.5 min, n = 5, P > 0.5 versus control, WIN 100

nM: 8 T 2 min, n = 5, P > 0.3 versus control; WIN 1 AM: 3.5 T1 min, n = 6, P = 0.05 versus control; Fig. 3C). Importantly,

although the ischemia-induced effects on VTA DA cells in the

presence of WIN 1 nM, 100 nM and 1 AM resembled those

induced under control conditions (Figs. 3B and C), only one out

of five cells treated with WIN 1 nM resumed firing after

reperfusion, whereas all the cells treated with WIN 100 nM and 1

AM underwent an irreversible depolarization block (Fig. 3D).

Taken together, these data suggest that manipulation of CB1

receptors might influence the outcome of ischemia: activation of

CB1 receptors exerts biphasic effects, with lower doses being

more effective than higher, whereas their blockade is detrimental

(Fig. 3E).

2-AG protects DA neurons during ischemia through a DSE-like

mechanism

Excessive glutamate release resulting in Ca+2 overload into

cells is the excitotoxic event that initiates ischemic damage in the

brain (Benveniste et al., 1984; Roettger and Lipton, 1996; Dirnagl

et al., 1999). Ionotropic glutamate receptors, notably NMDA and

AMPA, together with voltage-gated Ca+2 channels, allow Ca+2

entry. To test for DA neuronal depolarization dependence on

excessive extracellular glutamate acting on postsynaptic AMPA

and NMDA receptors, we applied the ischemia protocol in the

presence of both D-AP5 (100 AM) and CNQX (10 AM), NMDA

and AMPA receptor antagonist respectively. D-AP5 + CNQX

prevented ischemia-induced dysfunction of DA neurons, with the

change in membrane polarization of 3.8 T 2 mV (n = 6, P < 0.02

versus control, Figs. 4A and B), and three out of six cells

continuously firing (v2 = 22.9, P < 0.02, Fig. 4C). These results

support the hypothesis that CB1 receptor blockade, by preventing

2-AG actions, allows an excessive glutamate release under

ischemic conditions, which results in cell depolarization and cease

of spontaneous activity (Figs. 1A–C). If so, we would expect the

bath application of D-AP5 + CNQX to counteract AM281 effects.

Fig. 2. Endocannabinoids protect VTA DA neurons during ischemia in

vivo. Time–response curves displaying the effects of ischemia on the

spontaneous firing rate (A), expressed as percentage of baseline values.

VTA DA cells were recorded from urethane anesthetized rats. The animals

were prepared for the bilateral global brain ischemia with the four vessel

occlusion protocol (vertebral arteries were cauterized and common carotids

were isolated and loosely tied, see Materials and methods). Baseline firing

activity of the cell was recorded for 5 min, and then both common carotids

were occluded determining the complete interruption of blood supply to the

brain. The ischemia protocol was preceded by either CB1 receptor

antagonist SR141716A administration (SR, 1 mg/kg, i.v., closed circles)

or FAAH inhibitor URB597. There is a highly significant difference

between control and SR group [firing rate: F(treatment)1,170 = 5.07, P <

0.05, two-way ANOVA. (B and C) Representative firing rate histograms of

VTA DA neurons recorded from urethane anesthetized rats. Following the

recording of the baseline firing activity of the cell for 5 min, ischemia was

induced. Under these circumstances, we observed a slow progressive

decline in the firing rate of the cell (B). In panel C, the ischemia protocol

was preceded by the administration of the CB1 antagonist SR (1 mg/kg,

i.v.), which was per se ineffective. In this example, the neuron was strongly

excited following induction of ischemia and displayed an enhanced bursting

pattern of firing activity. Data are expressed as mean T SEM.*P < 0.05,

one-way ANOVA followed by Dunnett’s test.

Fig. 3. CB1 receptor-mediated effects on ischemia outcome on VTA DA neurons. (A) The voltage traces depict the response of a VTA dopaminergic cell to

ischemia (0 glucose, 0 oxygen, 7 min) under control conditions (top), in the presence of the CB1 receptor agonist WIN (3 nM, bottom). In the presence of WIN

(3 nM), ischemia led to a small membrane depolarization, which did not lead to the transient depolarization block (complete cessation of spontaneous activity)

usually observed under control conditions. The dashed line indicates the basal voltage membrane. (B) The bar graph shows the peak of change of voltage

membrane during the ischemia protocol, taken at the end of the protocol, under control conditions or in the presence of increasing concentrations of WIN. The

numbers above the columns indicate the number of neurons used for each determination. The asterisks represent statistically significant difference from control

(*P < 0.05). (C) The graph shows the latency between beginning of the ischemia protocol (time 0) and the onset of depolarization block under control

conditions and in the presence of the different concentrations of WIN for each of the experiments in panel B. The asterisks represent statistically significant

difference from control (*P < 0.05). (D) The graph shows that the treatment with WIN affects the percentage of cells that do not stop firing and re-start firing

after re-oxygenation. Note, no cells underwent a depolarization block when WIN was applied at the concentration of 3 nM. On the other hand, no cells restarted

their spontaneous activity in the presence of the high doses of WIN tested (100 nM and 1 AM) (v2 = 43.24, P < 0.0002). (E) Inverted ‘‘U’’-shaped function for

the role of CB1 receptor activation in neuroprotective actions. When either CB1 receptor activation is below the optimal range, as may occur when either

endocannabinoid synthesis/release and/or CB1 receptor number/function is impaired, or above the optimal range, their neuroprotective actions are impaired. In

such scenarios, CB1 receptor agonists or antagonists, respectively, may restore endocannabinoid actions to the optimal range. The protective activation of the

endogenous cannabinoid system during ischemia is represented by ‘‘0’’, whereas ‘‘0 + CB1 blockade’’ represents when endocannabinoid actions are prevented

by blocking CB1 receptors (by applying AM281).

M. Melis et al. / Neurobiology of Disease 24 (2006) 15–27 21

Accordingly, in the presence of D-AP5, CNQX and AM281, DA

cells underwent a small change in membrane polarization under

ischemia (mean 4.9 T 3 mV, n = 5, P < 0.05 versus control, and P <

0.0005 versus AM281 alone Fig. 4B). Additionally, four out of five

cells never ceased their spontaneous activity (P < 0.02, Chi square,

Fig. 4C).

Ischemia/Reperfusion injury, as well as ionotropic glutamate

receptor activation, can increase the synthesis of reactive oxygen

species (Carriedo et al., 2000; Avshalumov et al., 2003) and

contribute to cellular dysfunction and death (White et al., 2000;

Schaller and Graf, 2004). Among reactive oxygen species,

hydrogen peroxide is produced by midbrain DA neurons in an

activity-dependent manner (Chen et al., 2002), modulates synaptic

transmission (Chen et al., 2001) as well as DA neuronal

excitability (Avshalumov et al., 2005), and prevents metabolic

stress-induced dysfunction following hypoxia (Geracitano et al.,

2005). To determine if hydrogen peroxide is produced by VTA DA

cells in response to an ischemic insult, we filled the postsynaptic

DA cell with mercaptosuccinate (MCS; 0.3 mM) and 3-amino-

1,2,4-triazole (ATZ; 1 mM), glutathione peroxidase, and catalase

inhibitor respectively. Thus, when hydrogen peroxide degradation

was prevented, we observed a small change in voltage membrane

(mean 4.8 T 1 mV, n = 6, P < 0.02 versus control, Fig. 4B) with no

significant influence on delaying the onset of depolarization block

M. Melis et al. / Neurobiology of Disease 24 (2006) 15–2722

(9.8 T 1 min, P > 0.05 versus control). Since hydrogen peroxide

may also be a diffusible messenger (Avshalumov et al., 2003), an

intriguing possibility is that its actions might be at presynaptic

terminals, where it might stimulate 2-AG production. In other

words, hydrogen peroxide might be the true retrograde messenger

during ischemia. To test this hypothesis, we prevented its

intracellular elimination by catalase and glutathione peroxidase

inhibitors (by filling the postsynaptic cell with ATZ + MCS) and

bath applied catalase (500 UI/ml) to block its extracellular actions.

Under these conditions, the ischemia outcome on DA neurons was

similar (mean 2.8 T 2 mV, n = 6, P < 0.02 versus control, and P >

0.05 versus MCS + ATZ, Fig. 4B) to that induced in absence of

catalase (6.1 T 2 min, P > 0.05 versus control, data not shown).

These data suggest that the production of hydrogen peroxide is

unlikely to be involved in endocannabinoid-mediated protection of

DA neurons against excessive neuronal activation after ischemic

insults.

2-AG plays a key role in DSE

The events occurring during ischemia that involve endocanna-

binoid production/action strongly resemble a form of short-term

synaptic plasticity, previously termed depolarization-induced sup-

pression of excitation (DSE; Kreitzer and Regehr, 2001). In the

VTA, DSE depends upon a rise of intracellular Ca2+ leading to the

release of endocannabinoids that, by acting retrogradely, activate

presynaptic CB1 receptors, and suppress glutamate release (Melis et

al., 2004a). To date, which of the two most studied endocannabi-

noids is released during long-lasting depolarization remains

unknown. Using the recording electrode, we filled the postsynaptic

cell with either tetrahydrolipstatin (THL; 500 nM) or O-3640 (5

AM), two inhibitors of sn-1-DAG lipase activity (Bisogno et al.,

2003, 2006). Both THL (n = 7) and O-3640 (n = 9) partially

prevented 10 s DSE (Fig. 5A; EPSCs amplitude 79.9 T 8 and 89.8 T10% of baseline for THL and O-3640, respectively; P < 0.002 and

P < 0.0002 respectively versus control). Similar results were

obtained when slices were perfused with U73122 (5 AM), an

inhibitor of PI-PLC activity (Bleasdale et al., 1990) (Fig. 5A;

EPSCs amplitude 78.7 T 18% of baseline, n = 7; P < 0.04 versus

control), or the cells filled with LiCl (5 mM), which prevents IP3

formation (Fig. 5A; EPSCs amplitude 74.3 T 11% of baseline, n = 7;

P < 0.02 versus control). These results suggest that 2-AG plays a

Fig. 4. CB1-mediated effects on ischemia outcome on VTA DA neurons

involve glutamate release. (A) The voltage traces depict the response of a

VTA dopaminergic cell to ischemia (0 glucose, 0 oxygen, 7 min) under

control conditions (top), in the presence of the NMDA and AMPA receptor

antagonists (AP5 + CNQX, bottom). In the presence of AP5 + CNQX,

ischemia led to a small membrane depolarization, which was not followed

by a transient depolarization block (complete cessation of spontaneous

activity). The dashed lines indicate the basal voltage membrane. (B) The

bar graph shows the peak of change of voltage membrane during the

ischemia protocol, taken at the end of the protocol, under control

conditions, in the presence of NMDA and AMPA receptor antagonists

(AP5 + CNQX) alone or plus AM281 (***P < 0.0005 versus AM281). The

bar graph also shows the peak of change of voltage membrane during the

ischemia protocol under control conditions, and in the presence of

glutathione peroxidase and catalase inhibitors, MCS and ATZ respectively

(MCS + ATZ), alone or in the presence of catalase (MCS + ATZ + catalase

P > 0.05 versus MCS + ATZ). The numbers above the columns indicate the

number of neurons used for each determination. The asterisks represent

statistically significant difference from control (***P < 0.0005, *P < 0.05).

(C) The graph shows that the blockade of AMPA and NMDA receptors

affects the percentage of cells that do not stop firing and re-start firing after

reperfusion (v2 = 22.9, P < 0.02). Note, fewer cells underwent a permanent

depolarization block when AM281 was applied in the presence of AP5 and

CNQX. On the other hand, no significant effect was observed when

endogenous hydrogen peroxide levels were manipulated.

M. Melis et al. / Neurobiology of Disease 24 (2006) 15–27 23

key role in DSE, and its formation is likely to occur downstream of

the PLC cascade. To next ascertain whether or not AEA contributes

to DSE, we filled the postsynaptic DA cell with either AA-5HT or

URB597, two inhibitors of FAAH activity, during DSE induction, at

doses (5 AM and 10 nM, respectively) previously shown to inhibit

FAAH in vitro (Bisogno et al., 1998; Kathuria et al., 2003). DSE

was not changed in the presence of either drug (Fig. 5B). In fact,

DSE induced by different depolarizing pulses (3–10 s) in the

presence of either AA-5HT (being EPSCs amplitude 81.2 T 10, 78.7T 6 and 40.2 T 8% of baseline for 3, 5 and 10 s, respectively; n = 10

for each condition, P > 0.5) or URB597 (being EPSCs amplitude

90.2 T 9, 68.6 T 10 and 45.1 T 13% of baseline for 3, 5, and 10 s,

respectively; n = 6 for each condition) did not significantly differ

from control. Additionally, since FAAH�/� mice have been

reported to have a severely impaired AEA degradation and 10-

fold higher levels of brain AEA, but not 2-AG (Cravatt et al., 2001),

the effect of 3-, 5-, and 10-s depolarizations was tested in mice

lacking this enzyme (FAAH�/�) and in their respective wild-type

littermates (FAAH+/+). However, the magnitude of DSE in FAAH�/

� mice was comparable to that of FAAH+/+ mice (Fig. 5C; FAAH+/

+: EPSCs amplitude 103 T 6, 73.5 T 6 and 43.5 T 5% of baseline for

3, 5, and 10 s, respectively; n = 4 for each condition; FAAH�/�:

EPSCs amplitude 95.9 T 7, 76.3 T 3 and 47.4 T 2% of baseline for 3,

5, and 10 s, respectively; n = 6 for each condition; P > 0.5 versus

FAAH+/+). Taken together, these observations argue against a

possible role of AEA in this scenario.

Discussion

We have reported here that the endocannabinoid system might

serve as a neuroprotective mechanism to prevent/reduce ischemia-

induced VTA DA cell dysfunction. We also provide evidence for 2-

AG being the neuroprotective molecule released as a physiological

response to mitigate ischemia/reperfusion injury. We ascribe this

action to the activation of presynaptic CB1 receptors, which results

in a suppression of excessive glutamate released under ischemic

conditions. This on demand activation of the endocannabinoid

system might represent one of the neuroprotective mechanisms

reducing DA neuronal damage during episodes of energy

deprivation, which are often ischemic in nature.

We found that ischemia-induced DA cell membrane depolar-

ization involves 2-AG production, whose effects ultimately limit

changes in voltage membrane and delay the onset of depolarization

block through CB1 receptor activation. Several observations

support our conclusion that 2-AG, and not AEA, protects DA

neurons from ischemic damage. Firstly, 2-AG, but not AEA, levels

are elevated in VTA slices or brain sections containing the

Fig. 5. Contribution of 2-AG and AEA to DSE. (A) Magnitude of EPSCs

amplitude after 10 s DSE for THL (n = 7), O-3640 (n = 9), U73122

(n = 7), and LiCl (n = 7) plotted as the percentage of baseline before the

train. The asterisks represent statistically significant difference from

control (*P < 0.02, **P < 0.002, ***P < 0.0005). Representative traces

for each condition from a single experiment are shown (grey traces

represent the EPSC after the depolarizing step for each condition are

shown). Scale bar: 20 ms, 100 pA. (B) Averaged data for DSE induced

by depolarizing pulses with a duration of 3, 5, and 10 s are plotted for all

conditions. EPSCs amplitude was normalized to the averaged value

(dotted line) before depolarization. Each symbol represents the averaged

value obtained from different cells under normal conditions (white circles,

n = 10), in the presence of URB597 (grey circles, n = 6) or AA-5HT

(black circles, n = 10). Representative traces for each condition from a

single experiment are shown (grey traces represent the EPSC after the

depolarizing step for AA-5HT, whereas grey dotted traces represent the

EPSC after the depolarizing step for URB597). Scale bar: 20 ms, 100 pA.

(C) Averaged data for DSE induced by depolarizing pulses with duration

of 3, 5 and 10 s are plotted for FAAH+/+ and FAAH�/� mice. EPSCs

amplitude was normalized to the averaged value (dotted line) before

depolarization. Each symbol represents the averaged value obtained from

different cells from FAAH+/+ (open circles, n = 4) and FAAH�/� (closed

circles, n = 6). Representative traces for each condition from a single

experiment are shown (grey traces represent the EPSC after the

depolarizing step in FAAH�/� mice. Scale bar: 20 ms, 100 pA.

M. Melis et al. / Neurobiology of Disease 24 (2006) 15–2724

mesencephalon following ischemia in vitro and in vivo, respec-

tively. Secondly, when 2-AG synthesis is inhibited during an

ischemic episode, DA neurons undergo larger and irreversible

depolarization in vitro. On the other hand, when AEA degradation

is prevented, we do not observe any protection against the ischemic

insult both in vitro and in vivo. Conversely, 2-AG neuroprotective

actions are prevented when CB1 receptors are blocked—both in

vitro and in vivo. More specifically, in vivo DA cells respond to

bilateral transient general brain ischemia with a progressive decline

of their spontaneous activity, which is, conversely, strongly

enhanced (displaying an increased firing rate and bursting activity)

when endocannabinoid actions are prevented by the CB1 receptor

antagonist SR141716A. Therefore, the endocannabinoid system

might represent a significant component of protective mechanisms

engaged by these cells during ischemic insults and be part of

intrinsic defensive mechanisms which protect and preserve

dopaminergic neuronal function. Based on these experimental

results, and in agreement with other data present in literature

(Schmid et al., 1995; Nagayama et al., 1999; Panikashvili et al.,

2001; van der Stelt et al., 2001a; Mechoulam et al., 2002;

Parmentier-Batteur et al., 2002; Panikashvili et al., 2005), we

hypothesize that 2-AG helps protecting DA neurons from ischemic

injury. A prediction of this hypothesis is that positive modulation

of CB1 receptors should mimic endocannabinoid actions. Accord-

ingly, we observed that WIN exerted neuroprotective actions on

DA cells subjected to ischemia. Paradoxically, at higher doses

tested (0.1, 1 AM) WIN effects were rather detrimental, resembling

those induced by ischemia when either CB1 receptors were

blocked or 2-AG synthesis inhibited. Taken together, these results

support the idea that CB1 receptor agonists exert biphasic (Sulcova

et al., 1998; Cernak et al., 2004) effects, with low doses being the

most effective. This observation is in accordance with similar

neuroprotective effects exerted by exogenous cannabinoids in

diverse models of in vitro and in vivo ischemia (Shen and Thayer,

1998; Leker et al., 1999; Nagayama et al., 1999; Sinor et al., 2000;

Lavie et al., 2001; van der Stelt et al., 2001a,b; Panikashvili et al.,

2001; Franklin et al., 2003; Cernak et al., 2004; Kim et al., 2004;

Panikashvili et al., 2005), with more severe effects induced by

cerebral ischemia on CB1-deficient mice (Parmentier-Batteur et al.,

2002), and with increased endocannabinoid levels during severe

ischemia (Schmid et al., 1995). Additionally, our findings provide

an alternative explanation to the recent observation that the CB1

receptor agonist HU210 produces DA neuronal death in rat

mesencephalic cultures (Kim et al., 2005). Indeed, the detrimental

effects observed in the present study at the highest doses tested

(WIN 0.1, 1 AM) are consistent with, and might explain why,

HU210, proven to be more potent than WIN at CB1 receptors

(Breivogel et al., 2001), induced cell death at higher concentrations

(2–3 AM; Kim et al., 2005). Moreover, differences might exist

between the effects of on demand production of endocannabinoids

and the administration of CB1 agonists. For instance, on demand

localized activation of the endocannabinoid system was recently

shown to exert a key role in protection against excitotoxic seizures

(Marsicano et al., 2003), whereas systemic treatments with high

doses of CB1 agonists, or generalized and congenital enhance-

ments of endocannabinoid levels, showed a paradoxical worsening

effect under the same conditions (Clement et al., 2003). Therefore,

the present results draw further attention to CB1 receptor agonist

dosage as a possible confounding factor underlying the discrep-

ancies found in literature (for review, see Mechoulam et al., 2002;

Sarne and Keren, 2004).

We also provide evidence for the role of excessive glutamate

released during an ischemic insult as one of the major components

contributing to excitotoxicity which leads to DA neuronal

dysfunction. Indeed, the outcome of ischemia improves when

glutamate actions are prevented by blocking both AMPA and

NMDA receptors. Importantly, the neuroprotective effects resulting

from concomitant blockade of AMPA, NMDA and CB1 receptors

imply that activation of CB1 receptors by 2-AG prevents

glutamate-mediated consequences on DA cells during ischemia.

Additionally, the observation that endogenously produced hydro-

gen peroxide during an ischemic insult is not the signaling

molecule which triggers 2-AG-mediated neuroprotection supports

our view that 2-AG production is a response to reduce ischemia/

reperfusion injury. Our current hypothesis is that ischemia induces

DA cell depolarization, leading to excessive Ca2+ entry through

voltage-gated Ca2+ channels and ionotropic glutamate receptors, 2-

AG synthesis via PLC and sn-1-DAG lipase, 2-AG release, and

activation of presynaptic CB1 receptors on glutamatergic termi-

nals. Noteworthy, this scenario strongly resembles endocannabi-

noid-mediated DSE of VTA DA neurons (Melis et al., 2004a).

Accordingly, the use of long-lasting depolarizing steps, which are

routinely applied to postsynaptic cells to induce DSE and cannot be

regarded as physiological, has so far led to the assumption that

endocannabinoids might be released in response to pathological

stimuli or during periods of intense synaptic stimulation (for a

review, see Diana and Marty, 2004). Indeed, we found that, in the

VTA, DSE is elicited by prolonged depolarization, thus suggesting

a neuroprotective role for these signals (Melis et al., 2004a). It is

worth mentioning that we cannot identify definitively the identity

of endocannabinoid(s) other than 2-AG involved in DSE. Indeed,

the ineffectiveness shown by both FAAH inhibitors tested, as well

as the comparable DSE observed in FAAH�/� and FAAH+/+ mice,

argue against the possibility that AEA plays a role in DSE.

Conversely, a more plausible interpretation of the available data

might be that 2-AG plays a major role in DSE, given that this form

of short-term synaptic plasticity was similarly attenuated when

either PLC activity was inhibited, IP3 formation prevented, or sn-

1-DAG lipase inhibited. This is consistent with the fact that 2-AG

concentrations in the mesencephalon are 100 times higher than

AEA (Bisogno et al., 1999), that 2-AG is selectively produced by

VTA DA neurons to shape their own firing pattern and activity

both in vitro and in vivo (Melis et al., 2004b), and that 2-AG

protects DA cells during ischemia (present results). Lastly, the

similarity of the results obtained by applying long-lasting

depolarizing steps and ischemia points to DSE as a useful model

to study these lipids as modulators of neuronal communication.

Whether they serve as retrograde synaptic signaling molecules

involved in fine-tuning neuronal activity, or to protect neurons

from ischemic injury/energy deprivation, the present results

highlight the importance of an efficient endocannabinoid system

for a functional dopaminergic transmission.

The pathophysiological relevance of the present findings is that

disabilities of the endocannabinoid system at regulating dopami-

nergic transmission might be of considerable importance in the

pathogenesis of many neurological and psychiatric disorders.

Accordingly, since a dysfunctional endocannabinoid signal has

been reported in diverse neurodegenerative and psychiatric

disorders (e.g., Parkinson’s disease, schizophrenia, etc.) (for

reviews, see Glass, 2001; Romero et al., 2002; Brotchie, 2003;

van der Stelt and Di Marzo, 2003), a pharmacological manipu-

lation aimed at potentiating this signal might prove useful as

M. Melis et al. / Neurobiology of Disease 24 (2006) 15–27 25

symptomatic treatment and/or neuroprotective therapy to reduce

cell suffering/death in the early stages of these pathologies.

Particularly, inhibitors of MGL (Saario et al., 2005; Nithipatikom

et al., 2005; Makara et al., 2005), or of the putative endocanna-

binoid transporter (Marsicano et al., 2003; de Lago et al., 2005),

which would enhance endogenous 2-AG contents, might prove as

valuable targets, and be therapeutically useful. Further studies will

be needed to answer several questions that arise. Particularly, it is

important to know whether other mechanisms underlie (endo)can-

nabinoid actions, and whether timing and extent of ischemia affect

endocannabinoid tone and neuroprotective actions. If this is the

case, endocannabinoids might form the basis for the development

of new neuroprotective drugs useful in stroke and other neuro-

degenerative pathologies.

Acknowledgments

We thank B. Schilstrom, S. Marinelli, and M. A. Ungless for

many useful comments on the manuscript, W. T. Dunn III for

proof-reading the manuscript, K. Monory for genotyping, and B.F.

Cravatt for providing the FAAH-deficient mice, Drs. B.R. Martin

and R. Razdan for the gift of O-3640, and Dr. G. Le Fur for the gift

of SR141716A. This work was supported by the grants from Ass.

Igiene e Sanita (R.A.S.), COFIN 2003 (MIUR), Deutsche

Forschungsgemeinschaft (DFG) (LU755/1-3), and by a scholarship

from the Hertie Foundation (to B.L.).

References

Avshalumov, M.V., Chen, B.T., Marshall, S.P., Pena, D.M., Rice, M.E.,

2003. Glutamate-dependent inhibition of dopamine release in striatum

is mediated by a new diffusible messenger, H2O2. J. Neurosci. 23 (7),

2744–2750.

Avshalumov, M.V., Chen, B.T., Koos, T., Tepper, J.M., Rice, M.E., 2005.

Endogenous hydrogen peroxide regulates the excitability of midbrain

dopamine neurons via ATP-sensitive potassium channels. J. Neurosci.

25 (17), 4222–4231.

Azad, S.C., Monory, K., Marsicano, G., Cravatt, B.F., Lutz, B.,

Zieglgansberger, W., Rammes, G., 2004. Circuitry for associative

plasticity in the amygdala involves endocannabinoid signaling. J.

Neurosci. 24 (44), 9953–9961.

Benveniste, H., Drejer, J., Schousboe, A., Diemer, N.H., 1984. Elevation of

the extracellular concentrations of glutamate and aspartate in rat

hippocampus during transient cerebral ischemia monitored by intracere-

bral microdialysis. J. Neurochem. 43 (5), 1369–1374.

Bisogno, T., Sepe, N., Melck, D., Maurelli, S., De Petrocellis, L., Di Marzo,

V., 1997. Biosynthesis, release and degradation of the novel endo-

genous cannabimimetic metabolite 2-arachidonoylglycerol in mouse

neuroblastoma cells. J. Biochem. 322 (Pt 2), 671–677.

Bisogno, T., Melck, D., De Petrocellis, L., Bobrov, M.Y., Gretskaya,

N.M., Bezuglov, V.V., Sitachitta, N., Gerwick, W.H., Di Marzo,

V., 1998. Arachidonoylserotonin and other novel inhibitors of fatty

acid amide hydrolase. Biochem. Biophys. Res. Commun. 248 (3),

515–522.

Bisogno, T., Berrendero, F., Ambrosino, G., Cebeira, M., Ramos, J.A.,

Fernandez-Ruiz, J.J., Di Marzo, V., 1999. Brain regional distribu-

tion of endocannabinoids: implications for their biosynthesis and

biological function. Biochem. Biophys. Res. Commun. 256 (2),

377–380.

Bisogno, T., Howell, F., Williams, G., Minassi, A., Cascio, M.G., Ligresti,

A., Matias, I., Schiano-Moriello, A., Paul, P., Williams, E.J., Gang-

adharan, U., Hobbs, C., Di Marzo, V., Doherty, P., 2003. Cloning of the

first sn1-DAG lipases points to the spatial and temporal regulation of

endocannabinoid signaling in the brain. J. Cell. Biol. 163, 463–468.

Bisogno, T., Cascio, M.G., Saha, B., Mahadevan, A., Urbani, P.,

Minassi, A., Appendino, G., Saturnino, C., Martin, B., Razdan, R.,

Di Marzo, V., 2006. Development of the first potent and specific

inhibitors of endocannabinoid biosynthesis. Biochem Biophys Acta.

1761, 205–212.

Bleasdale, J.E., Thakur, N.R., Gremban, R.S., Bundy, G.L., Fitzpatrick,

F.A., Smith, R.J., Bunting, S., 1990. Selective inhibition of receptor-

coupled phospholipase C-dependent processes in human platelets and

polymorphonuclear neutrophils. Pharmacol. Exp. Ther. 255 (2),

756–768.

Breivogel, C.S., Griffin, G., Di Marzo, V., Martin, B.R., 2001. Evidence for

a new G protein-coupled cannabinoid receptor in mouse brain. Mol.

Pharmacol. 60 (1), 155–163.

Brotchie, J.M., 2003. CB1 cannabinoid receptor signalling in Parkinson’s

disease. Curr. Opin. Pharmacol. 3 (1), 54–61.

Burke, R.E., Macaya, A., DeVivo, D., Kenyon, N., Janec, E.M., 1992.

Neonatal hypoxic– ischemic or excitotoxic striatal injury results in a

decreased adult number of substantia nigra neurons. Neuroscience. 50

(3), 559–569.

Cadas, H., di Tomaso, E., Piomelli, D., 1997. Occurrence and

biosynthesis of endogenous cannabinoid precursor, N-arachidonoyl

phosphatidylethanolamine, in rat brain. J. Neurosci. 17 (4),

1226–1242.

Carriedo, S.G., Sensi, S.L., Yin, H.Z., Weiss, J.H., 2000. AMPA exposures

induce mitochondrial Ca(2+) overload and ROS generation in spinal

motor neurons in vitro. J. Neurosci. 20 (1), 240–250.

Cernak, I., Vink, R., Natale, J., Stoica, B., Lea, IV, P.M., Movsesyan, V.,

Ahmed, F., Knoblach, S.M., Fricke, S.T., Faden, A.I., 2004. The dark

side of endocannabinoids: a neurotoxic role for anandamide. J Cereb

Blood Flow Metab. 24 (5), 564–578.

Chen, B.T., Avshalumov, M.V., Rice, M.E., 2001. H(2)O(2) is a novel,

endogenous modulator of synaptic dopamine release. J. Neurophysiol.

85 (6), 2468–2476.

Chen, B.T., Avshalumov, M.V., Rice, M.E., 2002. Modulation of

somatodendritic dopamine release by endogenous H(2)O(2): suscepti-

bility in substantia nigra but resistance in VTA. J. Neurophysiol. 87 (2),

1155–1158.

Clement, A.B., Hawkins, E.G., Lichtman, A.H., Cravatt, B.F., 2003.

Increased seizure susceptibility and proconvulsant activity of ananda-

mide in mice lacking fatty acid amide hydrolase. J. Neurosci. 23 (9),

3916–3923.

Cravatt, B.F., Giang, D.K., Mayfield, S.P., Boger, D.L., Lerner, R.A.,

Gilula, N.B., 1996. Molecular characterization of an enzyme that

degrades neuromodulatory fatty-acid amides. Nature. 384 (6604),

83–87.

Cravatt, B.F., Demarest, K., Patricelli, M.P., Bracey, M.H., Giang, D.K.,

Martin, B.R., Lichtman, A.H., 2001. Supersensitivity to anandamide

and enhanced endogenous cannabinoid signaling in mice lacking

fatty acid amide hydrolase. Proc. Natl. Acad. Sci. U S A. 98 (16),

9371–9376.

Critchley, M., 1929. Arteriosclerotic parkinsonism. Brain 52, 23–83.

de Lago, E., Petrosino, S., Valenti, M., Morera, E., Ortega-Gutierrez, S.,

Fernandez-Ruiz, J., Di Marzo, V., 2005. Effect of repeated systemic

administration of selective inhibitors of endocannabinoid inactivation

on rat brain endocannabinoid levels. Biochem. Pharmacol. 70 (3),

446–452.

De Petrocellis, L., Cascio, M.G., Di Marzo, V., 2004. The endocannabinoid

system: a general view and latest additions. Br. J. Pharmacol. 141 (5),

765–774.

De Reuck, J., Siau, B., Decoo, D., Santens, P., Crevits, L., Strijckmans, K.,

Lemahieu, I., 2001. Parkinsonism in patients with vascular dementia:

clinical, computed- and positron emission-tomographic findings. Cere-

brovasc. Dis. 11 (1), 51–58.

Diana, M.A., Marty, A., 2004. Endocannabinoid-mediated short-term

synaptic plasticity: depolarization-induced suppression of inhibition

M. Melis et al. / Neurobiology of Disease 24 (2006) 15–2726

(DSI) and depolarization-induced suppression of excitation (DSE). Br.

J. Pharmacol. 142, 9–19.

Di Marzo, V., Fontana, A., Cadas, H., Schinelli, S., Cimino, G., Schwartz,

J.C., Pomelli, D., 1994. Formation and inactivation of endogenous

cannabinoid anandamide in central neurons. Nature 372, 686–691.

Di Marzo, V., Melck, D., Bisogno, T., De Petrocellis, L., 1998.

Endocannabinoids: endogenous cannabinoid receptor ligands with

neuromodulatory action. Trends Neurosci. 21, 521–528.

Di Marzo, V., De Petrocellis, L., Bisogno, T., Melck, D., 1999. Metabolism

of anandamide and 2-arachidonoylglycerol: an historical overview and

some recent developments. Lipids. 34 (Suppl), S319–S325.

Dinh, T.P., Freund, T.F., Piomelli, D., 2002. A role for monoglyceride lipase

in 2-arachidonoylglycerol inactivation. Chem. Phys. Lipids. 121 (1–2),

149–158.

Dirnagl, U., Iadecola, C., Moskowitz, M.A., 1999. Pathobiology of

ischaemic stroke: an integrated view. Trends Neurosci. 22 (9), 391–397.

Franklin, A., Parmentier-Batteur, S., Walter, L., Greenberg, D.A., Stella, N.,

2003. Palmitoylethanolamide increases after focal cerebral ischemia and

potentiates microglial cell motility. J. Neurosci. 23 (21), 7767–7775.

Freund, T.F., Katona, I., Piomelli, D., 2003. Role of endogenous

cannabinoids in synaptic signaling. Physiol. Rev. 83 (3), 1017–1066.

Geracitano, R., Tozzi, A., Berretta, N., Florenzano, F., Guatteo, E., Viscomi,

M.T., Chiolo, B., Molinari, M., Bernardi, G., Mercuri, N.B., 2005.

Protective role of hydrogen peroxide in oxygen-deprived dopaminergic

neurones of the rat substantia nigra. J. Physiol. 568 (Pt. 1), 97–110.

Glass, M., 2001. The role of cannabinoids in neurodegenerative diseases.

Prog. Neuropsychopharmacol. Biol. Psychiatry. 25 (4), 743–765.

Grace, A.A., Bunney, B.S., 1983. Intracellular and extracellular electro-

physiology of nigral dopaminergic neurons: I. Identification and

characterization. Neuroscience 10, 301–315.

Grace, A.A., Bunney, B.S., 1984. The control of firing pattern in nigral

dopamine neurons: burst firing. J. Neurosci. 4, 2877–2890.

Guatteo, E., Federici, M., Siniscalchi, A., Knopfel, T., Mercuri, N.B.,

Bernardi, G., 1998a. Whole cell patch-clamp recordings of rat midbrain

dopaminergic neurons isolate a sulphonylurea- and ATP-sensitive

component of potassium currents activated by hypoxia. J. Neuro-

physiol. 79 (3), 1239–1245.

Guatteo, E., Mercuri, N.B., Bernardi, G., Knopfel, T., 1998b. Intracellular

sodium and calcium homeostasis during hypoxia in dopamine

neurons of rat substantia nigra pars compacta. J. Neurophysiol. 80

(5), 2237–2243.

Guyenet, P.G., Aghajanian, G.K., 1978. Antidromic identification of

dopaminergic and other output neurons of the rat substantia nigra.

Brain Res. 150, 69–84.

Hausser, M.A., de Weille, J.R., Lazdunski, M., 1991. Activation by

cromakalim of pre- and post-synaptic ATP-sensitive K+ channels in

substantia nigra. Biochem. Biophys. Res. Commun. 174 (2), 909–914.

Hillard, C.J., Wilkison, D.M., Edgemond, W.S., Campbell, W.B., 1995.

Characterization of the kinetics and distribution of N-arachidonyletha-

nolamine (anandamide) hydrolysis by rat brain. Biochem. Biophys.

Acta. 1257 (3), 249–256.

Jiang, C., Sigworth, F.J., Haddad, G.G., 1994. Oxygen deprivation activates

an ATP-inhibitable K+ channel in substantia nigra neurons. J. Neurosci.

14 (9), 5590–5602.

Johnson, S.W., North, R.A., 1992. Two types of neurone in the rat ventral

tegmental area and their synaptic inputs. J. Physiol. (Lond) 450, 455–468.

Kathuria, S., Gaetani, S., Fegley, D., Valino, F., Duranti, A., Tontini, A.,

Mor, M., Tarzia, G., La Rana, G., Calignano, A., Giustino, A., Tattoli,

M., Palmery, M., Cuomo, V., Piomelli, D., 2003. Modulation of anxiety

through blockade of anandamide hydrolysis. Nat. Med. 9 (1), 76–81.

Kim, S.H., Won, S.J., Mao, X.O., Jin, K., Greenberg, D.A., 2004.

Involvement of protein kinase A in cannabinoid receptor-mediated

protection from oxidative neuronal injury. J. Pharmacol. Exp. Ther.

Kim, S.R., Leeda, Y., Chung, E.S., Oh, U.T., Kim, S.U., Jin, B.K., 2005.

Transient receptor potential vanilloid subtype 1 mediates cell death of

mesencephalic dopaminergic neurons in vivo and in vitro. J. Neurosci.

25 (3), 662–671.

Kreitzer, A.C., Regehr, W.G., 2001. Retrograde inhibition of presynaptic

calcium influx by endogenous cannabinoids at excitatory synapses onto

Purkinje cells. Neuron. 29 (3), 717–727.

Lavie, G., Teichner, A., Shohami, E., Ovadia, H., Leker, R.R., 2001. Long

term cerebroprotective effects of dexanabinol in a model of focal

cerebral ischemia. Brain Res. 901 (1–2), 195–201.

Leker, R.R., Shohami, E., Abramsky, O., Ovadia, H., 1999. Dexanabinol;

a novel neuroprotective drug in experimental focal cerebral ischemia.

J. Neurol. Sci. 162 (2), 114–119.

Lutz, B., 2004. On-demand activation of the endocannabinoid system in the

control of neuronal excitability and epileptiform seizures. Biochem.

Pharmacol. 68 (9), 1691–1698.

Makara, J.K., Mor, M., Fegley, D., Szabo, S.I., Kathuria, S., Astarita,

G., Duranti, A., Tontini, A., Tarzia, G., Rivara, S., Freund, T.F.,

Piomelli, D., 2005. Selective inhibition of 2-AG hydrolysis enhances

endocannabinoid signaling in hippocampus. Nat. Neurosci. 8 (9),

1139–1141.

Marinelli, S., Bernardi, G., Giacomini, P., Mercuri, N.B., 2000. Pharmaco-

logical identification of the K(+) currents mediating the hypoglycemic

hyperpolarization of rat midbrain dopaminergic neurones. Neurophar-

macology. 39 (6), 1021–1028.

Marinelli, S., Federici, M., Giacomini, P., Bernardi, G., Mercuri,

N.B., 2001. Hypoglycemia enhances ionotropic but reduces meta

botropic glutamate responses in substantia nigra dopaminergic

neurons. J Neurophysiol. 85 (3), 1159–1166.

Marinelli, S., Di Marzo, V., Berretta, N., Matias, I., Maccarrone, M.,

Bernardi, G., Mercuri, N.B., 2003. Presynaptic facilitation of gluta-

matergic synapses to dopaminergic neurons of the rat substantia nigra

by endogenous stimulation of vanilloid receptors. J. Neurosci. 23 (8),

3136–3144.

Marinelli, S., Mercuri, N.B., 2004. Cannabinoid and vanilloid receptors are

activated by N-arachidonyl-dopamine in rat midbrain dopaminergic

neurons. 2004 Symposium on the Cannabinoids, Burlington, Vermont,

International Cannabinoid Research Society, pp. 140.

Marsicano, G., Goodenough, S., Monory, K., Hermann, H., Eder, M.,

Cannich, A., Azad, S.C., Cascio, M.G., Gutierrez, S.O., van der Stelt,

M., Lopez-Rodriguez, M.L., Casanova, E., Schutz, G., Zieglgansberger,

W., Di Marzo, V., Behl, C., Lutz, B., 2003. CB1 cannabinoid receptors

and on-demand defense against excitotoxicity. Science. 302 (5642),

84–88.

Mechoulam, R., Ben-Shabat, S., Hanus, L., Ligumsky, M., Kaminski, N.E.,

Schatz, A.R., Gopher, A., Almog, S., Martin, B.R., Compton, D.R.,

1995. Identification of an endogenous 2-monoglyceride, present in

canine gut, that binds to cannabinoid receptors. Biochem. Pharmacol.

50, 83–90.

Mechoulam, R., Spatz, M., Shohami, E., 2002. Endocannabinoids and

neuroprotection. Sci. STKE. 129, RE5.

Mechoulam, R., Lichtman, A.H., 2003. Neuroscience. Stout guards of the

central nervous system. Science. 302 (5642), 65–67.

Melis, M., Pistis, M., Perra, S., Muntoni, A.L., Pillolla, G., Gessa,

G.L., 2004a. Endocannabinoids mediate presynaptic inhibition of

glutamatergic transmission in rat ventral tegmental area dopa-

mine neurons through activation of CB1 receptors. J. Neurosci.

24, 53–62.

Melis, M., Perra, S., Muntoni, A.L., Pillolla, G., Lutz, B., Marsicano,

G., Di Marzo, V., Gessa, G.L., Pistis, M., 2004b. Prefrontal cortex

stimulation induces 2-arachidonoyl-glycerol-mediated suppression of

excitation in dopamine neurons. J. Neurosci. 24 (47), 10707–10715.

Mercuri, N.B., Bonci, A., Johnson, S.W., Stratta, F., Calabresi, P.,

Bernardi, G., 1994a. Effects of anoxia on rat midbrain dopamine

neurons. J. Neurophysiol. 71 (3), 1165–1173.

Mercuri, N.B., Bonci, A., Calabresi, P., Stratta, F., Bernardi, G.,

1994b. Responses of rat mesencephalic dopaminergic neurons to

a prolonged period of oxygen deprivation. Neuroscience. 63 (3),

757–764.

Nagayama, T., Sinor, A.D., Simon, R.P., Chen, J., Graham, S.H., Jin, K.,

Greenberg, D.A., 1999. Cannabinoids and neuroprotection in global and

M. Melis et al. / Neurobiology of Disease 24 (2006) 15–27 27

focal cerebral ischemia and in neuronal cultures. J. Neurosci. 19 (8),

2987–2995.

Nithipatikom, K., Endsley, M.P., Isbell, M.A., Wheelock, C.E., Hammock,

B.D., Campbell, W.B., 2005. A new class of inhibitors of 2-

arachidonoylglycerol hydrolysis and invasion of prostate cancer cells.

Biochem. Biophys. Res. Commun. 332 (4), 1028–1033.

Okamoto, Y., Morishita, J., Tsuboi, K., Tonai, T., Ueda, N., 2004.

Molecular characterization of a phospholipase D generating ananda-

mide and its congeners. J. Biol. Chem. 279 (7), 5298–5305.

Panikashvili, D., Simeonidou, C., Ben-Shabat, S., Hanus, L., Breuer,

A., Mechoulam, R., Shohami, E., 2001. An endogenous cannabi-

noid (2-AG) is neuroprotective after brain injury. Nature. 413

(6855), 527–531.

Panikashvili, D., Shein, N.A., Mechoulam, R., Trembovler, V., Kohen, R.,

Alexandrovich, A., Shohami, E., 2005 (Dec). The endocannabinoid 2-

AG protects the blood–brain barrier after closed head injury and

inhibits mRNA expression of proinflammatory cytokines. Neurobiol

Dis 16 (Epub ahead of print).

Parmentier-Batteur, S., Jin, K., Mao, X.O., Xie, L., Greenberg, D.A., 2002.

Increased severity of stroke in CB1 cannabinoid receptor knock-out

mice. J. Neurosci. 22 (22), 9771–9775.

Paxinos, G., Watson, C., 1997. The Rat Brain in Stereotaxic Coordinates.

Academic Press, San Diego.

Piomelli, D., 2003. The molecular logic of endocannabinoid signalling. Nat.

Rev. Neurosci. 4 (11), 873–874.

Pulsinelli, W.A., Brierley, J.B., 1979. A new model of bilateral hemispheric

ischemia in the unanesthetized rat. Stroke. 10 (3), 267–272.

Pulsinelli, W.A., Brierley, J.B., Plum, F., 1982. Temporal profile of

neuronal damage in a model of transient forebrain ischemia. Ann.

Neurol. 11 (5), 491–498.

Roettger, V., Lipton, P., 1996. Mechanism of glutamate release from rat

hippocampal slices during in vitro ischemia. Neuroscience. 75 (3),

677–685.

Romero, J., Lastres-Becker, I., de Miguel, R., Berrendero, F., Ramos, J.A.,

Fernandez-Ruiz, J., 2002. The endogenous cannabinoid system and the

basal ganglia. biochemical, pharmacological, and therapeutic aspects.

Pharmacol. Ther. 95 (2), 137–152.

Saario, S.M., Salo, O.M., Nevalainen, T., Poso, A., Laitinen, J.T., Jarvinen,

T., Niemi, R., 2005. Characterization of the sulfhydryl-sensitive site in

the enzyme responsible for hydrolysis of 2-arachidonoyl-glycerol in rat

cerebellar membranes. Chem. Biol. 12 (6), 649–656.

Sarne, Y., Keren, O., 2004. Are cannabinoid drugs neurotoxic or neuro-

protective? Med Hypotheses 63 (2), 187–192.

Schaller, B., Graf, R., 2004. Cerebral ischemia and reperfusion: the

pathophysiologic concept as a basis for clinical therapy. J. Cereb.

Blood Flow Metab. 24 (4), 351–371.

Schmid, P.C., Krebsbach, R.J., Perry, S.R., Dettmer, T.M., Maasson, J.L.,

Schmid, H.H., 1995. Occurrence and postmortem generation of

anandamide and other long-chain N-acylethanolamines in mammalian

brain. FEBS Lett. 375 (1–2), 117–120.

Schultz, W., 2002. Getting formal with dopamine and reward. Neuron 36,

241–263.

Shen, M., Thayer, S.A., 1998. Cannabinoid receptor agonists protect

cultured rat hippocampal neurons from excitotoxicity. Mol. Pharmacol.

54 (3), 459–462.

Sinor, A.D., Irvin, S.M., Greenberg, D.A., 2000. Endocannabinoids protect

cerebral cortical neurons from in vitro ischemia in rats. Neurosci. Lett.

278 (3), 157–160.

Stella, N., Schweitzer, P., Piomelli, D., 1997. A second endogenous

cannabinoid that modulates long-term potentiation. Nature 388,

773–778.

Strosznajder, J., Goracci, G., Gaiti, A., 1990. Synaptic vesicle-bound

phospholipase(s) acting on phosphatidylinositol exhibit(s) high suscep-

tibility to brain ischemia. Neurosci Lett. 114, 329–332.

Sugiura, T., Kondo, S., Sukagawa, A., Nakane, S., Shinoda, A., Itoh, K.,

Yamashita, A., Waku, K., 1995. 2-Arachidonoylglycerol: a possible

endogenous cannabinoid receptor ligand in brain. Biochem. Biophys.

Res. Commun. 215, 89–97.

Sulcova, E., Mechoulam, R., Fride, E., 1998. Biphasic effects of

anandamide. Pharmacol. Biochem. Behav. 59 (2), 347–352.

Toda, S., Ikeda, Y., Teramoto, A., Hirakawa, K., Uekusa, K., 2002. Highly

reproducible rat model of reversible forebrain ischemia-modified four-

vessel occlusion model and its metabolic feature. Acta Neurochir.

(Wien) 144 (12), 1297–1304.

Ueda, N., Kurahashi, Y., Yamamoto, S., Tokunaga, T., 1995. Partial

purification and characterization of the porcine brain enzyme

hydrolyzing and synthesizing anandamide. J. Biol. Chem. 270 (40),

23823–23827.

Ungless, M.A., Magill, P.J., Bolam, J.P., 2004. Uniform inhibition of

dopamine neurons in the ventral tegmental area by aversive stimuli.

Science 303 (5666), 2040–2042.

van der Stelt, M., Veldhuis, W.B., van Haaften, G.W., Fezza, F., Bisogno, T.,

Bar, P.R., Veldink, G.A., Vliegenthart, J.F., Di Marzo, V., Nicolay, K.,

2001a. Exogenous anandamide protects rat brain against acute neuronal

injury in vivo. J. Neurosci. 21 (22), 8765–8771.

van der Stelt, M., Veldhuis, W.B., Bar, P.R., Veldink, G.A., Vliegenthart,

J.F., Nicolay, K., 2001b. Neuroprotection by Delta9-tetrahydrocannab-

inol, the main active compound in marijuana, against ouabain-induced

in vivo excitotoxicity. J. Neurosci. 21 (17), 6475–6479.

van der Stelt, M., Veldhuis, W.B., Maccarrone, M., Bar, P.R., Nicolay, K.,

Veldink, G.A., Di Marzo, V., Vliegenthart, J.F., 2002. Acute neuronal

injury, excitotoxicity, and the endocannabinoid system. Mol. Neurobiol.

26 (2–3), 317–346.

van der Stelt, M., Di Marzo, V., 2003. The endocannabinoid system in the

basal ganglia and in the mesolimbic reward system: implications for

neurological and psychiatric disorders. Eur. J. Pharmacol. 480 (1–3),

133–150.

van der Stelt, M., Di Marzo, V., 2005. Cannabinoid receptors and their role

in neuroprotection. Neuromolecular Med. 7 (1–2), 37–50.

Volpe, B.T., Pulsinelli, W.A., Tribuna, J., Davis, H.P., 1984. Behavioral

performance of rats following transient forebrain ischemia. Stroke. 15

(3), 558–562.

Volpe, B.T., Blau, A.D., Wessel, T.C., Saji, M., 1995. Delayed

histopathological neuronal damage in the substantia nigra compacta

(nucleus A9) after transient forebrain ischaemia. Neurobiol. Dis. 2 (2),

119–127.

White, B.C., Sullivan, J.M., DeGracia, D.J., O’Neil, B.J., Neumar, R.W.,

Grossman, L.I., Rafols, J.A., Krause, G.S., 2000. Brain ischemia and

reperfusion: molecular mechanisms of neuronal injury. J. Neurol. Sci.

179 (S 1–2), 1–33.