Upregulation of CB1 Receptors and Agonist-Stimulated [35S]

Upregulation of CB1 receptors and agonist-stimulated [35S]GTPcS binding in the prefrontal cortex of depressed suicide victims BL Hungund1,2,3,5, KY Vinod2,5, SA Kassir1, BS Basavarajappa1,2, R Yalamanchili2, TB Cooper1,2,3, JJ Mann1,3 and V Arango1,3,4 1New York State Psychiatric Institute, New York, NY, USA; 2Nathan Kline Institute for Psychiatric Research, Orangeburg, NY, USA; 3Department of Psychiatry, College of Physicians and Surgeons, Columbia University, New York, NY, USA; 4Department of Anatomy and Cell Biology, College of Physicians and Surgeons, Columbia University, New York, NY, USA

Endogenous and exogenous cannabinoids (CBs) acting through the CB1 receptors have been implicated in the regulation of several behavioral and neuroendocrine functions. Modulation of endocannabinoidergic system by ethanol in mouse brain, and the association of suicide and mood disorders with alcoholism suggest possible involvement of the cannabinoidergic system in the pathophysiology of depression and suicide. Therefore, the present study was

undertaken to examine the levels of CB1 receptors and mediated signaling in the dorsolateral prefrontal cortex (DLPFC) of subjects with major depression who had died by suicides 3 35 (depressed suicides, DS). [ H]CP-55,940 and CB1 receptor-stimulated [ S]GTPcS binding sites were analyzed in membranes obtained from DLPFC of DS (10) and matched normal controls (10). Upregulation (24%, Po0.0001) of CB1 receptor density (Bmax) was observed in DS (644.6748.8 fmol/mg protein) compared with matched controls (493.3752.7 fmol/mg protein). However, there was no significant alteration in the affinity of receptor (DS; 1.1470.08 vs 7 control; 1.12 0.10 nM). Higher density of CB1 receptors in DS (38%, Po0.001) was also 35 demonstrated by Western blot analysis. The CB1 receptor-stimulated [ S]GTPcS binding was significantly greater (45%, Po0.001) in the DLPFC of DS compared with matched controls. The observed upregulation of CB1 receptors with concomitant increase in the CB1 receptor- mediated [35S]GTPcS binding suggests a role for enhanced cannabinoidergic signaling in the prefrontal cortex of DS. The cannabinoidergic system may be a novel therapeutic target in the treatment of depression and/or suicidal behavior. Molecular Psychiatry (2004) 9, 184–190. doi:10.1038/sj.mp.4001376

35 Keywords: CB1 receptor; [ S]GTPgS binding; depression; suicide; prefrontal cortex

Introduction of the cannabinoidergic system in health and disease is just beginning to emerge.5,6 Cannabinoid (CB) receptors received their name Significant progress has been made in characteriz- as those receptors that bind cannabinoidergic ing CB receptors both centrally and peripherally, and drugs, such as 9-tetrahydrocannabinol (D9-THC), in studying the role of second messenger systems at derived from Cannabis sativa and its biologically the cellular level. To date, two types of CB receptors, 9 7,8 active synthetic analogs. D -THC is the major CB1 and CB2, have been identified and have been psychoactive component in marijuana extracts shown to belong to G-protein coupled receptor and can produce a multiplicity of effects in (GPCR) family. Two endogenous cannabimimetic humans, including alterations in mood, percep- substances, characterized to be N-arachidonyl etha- tion, cognition and memory.1–4 Marijuana nolamide (AEA/anandamide) and 2-arachidonylgly- is currently the most widely abused drug second cerol (2-AG), were discovered and shown to act as 9,10 to alcohol. However, the functional significance agonists for CB receptors. The CB1 receptor is distributed primarily in neural tissue, whereas CB2 receptor is expressed mainly in the peripheral immune system.11,12 These receptors exhibit seven

Correspondence: Dr BL Hungund, Nathan Kline Institute for transmembrane domains, linked to Gi/o protein to 13 Psychiatric Research, Bldg 39, 140 Old Orangeburg Road, inhibit adenylyl cyclase. Interestingly, the CB1 Orangeburg, NY 10962, USA; receptor is one of the most abundant neuromodula- E-mail: [email protected] tory receptors in the brain and is expressed predomi- 5BLH and KYV contributed equally to this work. Received 20 January 2003; revised 7 March 2003; accepted 10 nantly in the cerebral cortex, hippocampus, March 2003 cerebellum and basal ganglia.14–16 CB1 receptors-mediated signaling in brain of DS BL Hungund et al 185 Numerous studies have implicated alterations in The study was conducted with approval from the several receptors and G-protein function in the NYSPI Institutional Review Board. Dorsolateral pre- pathophysiology of various neurological and psychia- frontal cortex postmortem samples (Brodmann area 9) tric disorders. Decreased CB1 receptor binding in from 10 normal controls (age range: 15–79 years) were neurodegenerative diseases related to extrapyramidal studied with a matched group of 10 subjects who had function has been reported.17 Alterations in the a lifetime diagnosis of major depression and died by serotonergic (for a review see, Arango et al18) and suicide (age range: 13–77 years). The groups com- adrenergic19 receptors in the pathophysiology of prised pairs of depressed suicides (DS) and control depression and suicidal behavior are well documen- cases matched for age, sex, postmortem interval (PMI) 20 ted. Recently, Dean et al, reported increased CB1 and ethnic group. There were nine pairs of caucasians receptor density in dorsolateral prefrontal cortex and one pair of African-Americans (2nd pair in Table (DLPFC) in schizophrenia. 1). This distribution reflects the ethnic make-up of The increased levels of endocannabinoids and Allegheny County, where the samples were collected. downregulation of CB1 receptor in response to There were no significant differences in age, sex and chronic ethanol intake in mouse brain suggest PMI distribution between DS and control subjects. modulation of the endocannabinoidergic system by The demographic variables, such as sex, age, PMI and alcohol21,22 (for a review, see Hungund et al23). The cause of death, as well as toxicology results, are mood and cognition altering ability of exogenous summarized in Table 1. The Coroner determined the cannabinoids and alcohol, and the association be- cause of death and reached the verdict of suicide. tween depression, suicide and alcohol abuse raise the Toxicological analyses were performed on all the question whether endogenous cannabinoidergic sys- cases, ruling out recent consumption of substance of tem plays any role in the etiology of depression and abuse or psychoactive medication except in three suicidal behavior. Therefore, to address this question, samples where an anxiolytic drug (n ¼ 1) and ethanol we studied the density of CB1 receptors and CB1 (n ¼ 3) were detected. Two individuals received receptor-mediated [35S]GTPgS binding in prefrontal lidocaine as part of resuscitation efforts at the cortex of subjects with major depression who had emergency room. All cases were free of neuropathol- died by suicides. Further studies of the role of the ogy. Both suicides and controls were examined cannabinoidergic system in various neuropsychiatric psychiatrically by structured interviews with family disorders would be of great interest. members and/or close friends. The psychiatric diag- noses were made according to DSM-III-R criteria.25 The psychological autopsies revealed that all suicide Materials and methods victims had a lifetime diagnosis of major depression Human brain tissue and all controls were free of psychopathology. The Brain samples of prefrontal cortex were obtained from samples were coded to mask investigators to the autopsy material derived from the brain tissue diagnostic group of all subjects. All the assays were collection of the Department of Neuroscience at the carried out in a paired design under the same New York State Psychiatric Institute (NYSPI) and experimental conditions. Columbia University, NY, USA. All tissue used in this study was provided by the Allegheny County Coroner Membrane preparation in accordance with protocols approved by the Brain tissue (B1 g) was homogenized in 20 volumes

Institutional Review Board of the University of of ice-cold TME buffer (50 mM Tris-HCl, 3 mM MgCl2 Pittsburgh. Brains were collected and bisected at and 1 mM of EDTA, pH 7.4) containing 0.32 M sucrose autopsy. The right hemispheres were cut coronally and freshly added protease inhibitor cocktail. The into 1.5-cm thick sections. Blocks were placed on a homogenate was centrifuged at 1000 g for 10 min at glass slide, immersed in freon (À201C) and stored at 41C. The resulting supernatant was then centrifuged À801C in tightly sealed, thick plastic bags until at 22 000 g for 20 min. The pellet was dissolved in sectioning. After a control and suicide were matched, TME buffer and recentrifuged at 22 000 g for 20 min. coronal sections (20 mm) from the hemicerebrum were The final pellet, dissolved in TME buffer, was made taken from a level just anterior to the genu of the aliquots and stored at –801C until the assay. corpus callosum with a large format Leica Cryopoly- cut cryostat. Interleaved sections every 200 mm were Determination of protein content sectioned at 50 mm and stained with cresylecht violet The protein content of the membrane fraction was for cytoarchitectonics. Once the sectioning was determined by Lowry’s method26 using bovine serum completed, Brodmann area 9 was identified using albumin (BSA) as the standard. Protein content of the gyral and sulcal landmarks, cytoarchitecture and a membrane, also normalized by silver staining, was standardized coronal atlas (Robert Perry and Edward used for Western blot analysis. Bird, personal communication), as previously described.24 Tissue from Brodmann area 9 (B1.5 g) was [3H]CP-55,940 binding assay dissected frozen, the white matter was removed as An aliquot of membrane (100 mg protein) was incu- much as possible, and tissue returned immediately to bated with TME buffer, [3H]CP-55,940 (0.05–5.0 nM) –801C until membrane preparation. and 0.1% fatty acid-free BSA in silicone-treated test

Molecular Psychiatry CB1 receptors-mediated signaling in brain of DS BL Hungund et al 186 Table 1 Demographic characteristics of DS and their matched control subjects

Diagnosis Sex (M/F) Age at death (years) PMI (h) Cause of death Toxicology

1. MDD M 13 18 Suicide, NaOH poisoning None Control M 15 16 Homicide, GSW None 2. MDD M 14 16 Suicide, hanging None Control M 17 14 Homicide, GSW Alcohol 3. MDD M 17 17 Suicide, GSW None Control M 16 18 Accident, asphyxia None 4. MDD M 25 18 Suicide, GSW Alcohol Control M 25 18 Accident, explosion Alcohol 5. MDD F 28 19 Suicide, fall from height None Control F 27 15 MVA, pedestrian None 6. MDD F 34 04 Suicide, hanging None Control F 38 07 Homocide, GSW None 7. MDD F 43 12 Suicide, poisoning ANX, OPIA, ACE Control F 42 16 Homicide, GSW None 8. MDD F 61 11 Suicide, hanging None Control F 56 18 MVA, passenger None 9. MDD M 64 19 Suicide, GSW LIDO Control M 66 19 Natural, cardiovascular None 10. MDD M 77 18 Suicide, hanging None Control M 79 10 Natural, cardiovascular LIDO

MDD (10) 6M/4F 37.677.2 15.971.4 Control (10) 6M/4F 38.177.1 15.071.2

The data are presented as mean7SEM. MDD, major depressive disorder; GSW, gun shot wound; MVA, motor vehicle accident; PMI, postmortem interval (rounded to near hours); ANX, anxiolytic; LIDO, lidocaine; OPIA, opiates; ACE, acetaminophen; Low level of alcohol was detected in three cases (0.02–0.03%).

tubes for 1 h at 371C. The nonspecific binding of reaction mixture was performed as described for radioligand was defined by CP-55,940 (10 mM). The [3H]CP-55,940 binding assay. The radioactivity was reaction was terminated by the addition of 2 ml ice- measured by liquid scintillation spectroscopy at an cold termination buffer (0.1% BSA in 50 mM Tris- efficiency of 95% for 35S. HCl, pH 7.4). The reaction mixture was rapidly filtered through polyethyleneimine (0.1%) pretreated Western blot analysis glass fiber filters using a Brandel 24-position cell Briefly, aliquots of membrane protein (30 mg), sepa- harvester (Brandel, Gaithersburg, MD, USA). Filters, rated by 10% polyacrylamide gel, were electrophor- washed three times with the termination buffer, were etically transferred to nitrocellulose membrane. The transferred to scintillation vials containing 5 ml of membrane was treated with blocking buffer (TTBS scintillation cocktail (ICN biochemicals, USA), and (10 mM Tris, 0.9% NaCl; 1% Tween 20 containing 3% were incubated overnight at room temperature. The milk powder) of pH 7.4) for 1 h at room temperature.

radioactivity was measured by liquid scintillation The membrane was incubated with human anti-CB1 spectroscopy (Beckman) at an efficiency of 47% for receptor antibody (1 : 500) overnight at 41C. The blot tritium. was washed three times with TTBS and then incubated with alkaline phosphatase-conjugated 35 CB1 receptor-stimulated [ S]GTPgS binding assay anti-IgG for 1 h at room temperature. After washing

The functional coupling between CB1 receptor and the blot for 3–4 times with TTBS, the immunoreactive G-protein was assessed by [35S]GTPgS binding assay band was visualized by CDP-star reagent. The blot as described previously27 with minor modification. was reprobed with a-tubulin antibody to ensure equal Briefly, an aliquot of membrane (50 mg protein) was protein loading. incubated in assay buffer (TME buffer and 0.1% fatty acid-free BSA and 100 mM NaCl) containing GDP Data and statistical analysis 35 (40 mM), and [ S]GTPgS (0.05 nM) in silicone-treated The Bmax (maximal binding sites) and Kd (apparent test tubes for 1 h at 371C. The CB1 receptor agonist, dissociation constant) values were determined from CP-55,940 (1 mM), was used to study CB1 receptor- saturation isotherms using nonlinear regression ana- stimulated [35S]GTPgS binding. The nonspecific bind- lysis to fit the data to the single-site binding equation ing of radioligand was determined in the presence of (Prism; GraphPad software). The density and affinity

10 mM GTPgS. The termination and filtration (without of CB1 receptor was expressed as fmol/mg protein and presoaking the filters in polyethyleneimine) of nM, respectively. The CB1 receptor-stimulated

Molecular Psychiatry CB1 receptors-mediated signaling in brain of DS BL Hungund et al 187 [35S]GTPgS binding expressed as fmol/mg protein’ is 600 a percentage of stimulation over the basal activity. 400

Statistical analysis performed using nonparametric 600 DS 200 analysis of variance (Mann–Whitney U) and para- Bound / Free Control metric (paired Student ‘t’-test). Differences were 0 400 0 100 200 300 400 500 600 considered to be significant at Po0.05. Immunoblots Bound were analyzed using the NIH image software program. (fmol/mg protein) 7 Data are expressed as mean SEM from two to three 200 experiments, each run in at least duplicate unless otherwise indicated. 0 Chemicals Specific binding 0 1 2 3 4 5 [3H]CP-55,940 (nM) [35S]GTPgS and [3H]CP-55,940 were purchased from DuPont NEN (Boston, MA, USA). Fatty acid-free BSA, Figure 1 The saturation binding of [3H]CP-55,940 (0.5– protease inhibitor cocktail, GDP and GTPgS were 5.0 nM) to prefrontal cortical membrane of DS and matched procured from Sigma Co (St Louis, MO, USA). Glass control. The inset represents the Scatchard transformation fiber filters (GF/B) were purchased from Brandel Inc. of the same binding data. (Gaithersburg, MD, USA). CP-55,940 was a gift from Pfizer Pharmaceutical (Groton, CT, USA). Human 1000 anti-CB1 receptor was obtained from Biosource Inter- Control nationals (California, CA, USA). Anti-a-tubulin DS monoclonal antibody was from Amersham Bioscience 750 (Piscataway, NJ, USA). Alkaline phosphatase-conju- *** gated anti-IgG was obtained from Promega (Madison, receptor WI, USA). CDP-star chemiluminescence kit was 1 purchased from Tropix (Bedford, MA, USA). Other 500 chemicals, of analytical grade, were purchased from

standard commercial sources. (fmol/mg protein)

Density CB 250

Results 0 The density of CB1 receptor 3 A saturation analysis suggests that [ H]CP-55,940 Figure 2 The density of CB1 receptor was estimated in binding is saturable below 5.0 nM concentration. prefrontal cortical membranes of DS (10) and matched The nonspecific binding was about 15% of total controls (10). Data are mean7SEM of two to three experi- [3H]CP-55,940 binding. A Scatchard analysis of the ments, each assayed in duplicate. ***Po0.0001. binding data indicates a monophasic binding of radioligand, and Hill’s coefficient of near unity 35 CB1 receptor-stimulated [ S]GTPgS binding suggests the binding of radioligand to a single class 35 of receptor at the concentration used. A representa- The CB1 receptor-stimulated [ S]GTPgS binding was tive saturation isotherm and Scatchard plot is shown used to assess the coupling efficiency between a in Figure 1. receptor and its G-protein. Using the CB1 receptor agonist, CP-55,940 stimulated-[35S]GTPgS binding as The average density (Bmax)ofCB1 receptor in DLPFC of normal control subjects was 493.3752.7 fmol/mg the outcome measure, maximum stimulation of [35S]GTPgS binding was observed when cortical protein. The apparent dissociation constant (Kd) was 1.1270.10 nM, suggesting a high affinity of the membranes were incubated with 1 mM CP-55,940 and 40 mM of GDP (data not shown). The increase in receptor for the radioligand. All the comparisons 35 shown below utilized Mann–Whitney U and paired CB1 receptor-stimulated [ S]GTPgS binding was 45% greater in cortical membranes of DS (31.174.7 fmol/ ‘t’-tests. Greater density of CB1 receptor was observed in the DLPFC of DS (644.6748.8 fmol/mg protein; mg protein; Po0.001) compared with matched controls (16.972.5 fmol/mg protein) (Figure 4). However, 24%, Po0.0001) compared with matched controls 35 (Figure 2). However, there was no difference in the no significant group difference in basal [ S]GTPgS affinity of receptor for radioligand (DS; 1.1470.08 vs binding was observed. control; 1.1270.10 nM), suggesting an upregulation of the density of receptor in the absence of altered Discussion affinity of the receptor. A significant increase (38%, Po0.001) in CB1 receptor immunoreactivity was also Although our understanding of clinical aspects of demonstrated by Western blot analysis. A representa- depression has advanced, the precise underlying tive CB1 receptor immunoblot of a DS and matched neurobiological basis of this disorder remains to be control (3A) and levels of CB1 receptor immunoreac- elucidated. Disturbances in pre- and postsynaptic tivity of all the subjects (3B) are shown in Figure 3. proteins in DS have been reported.28 A number of

Molecular Psychiatry CB1 receptors-mediated signaling in brain of DS BL Hungund et al 188 a studies have found differences in serotonergic and adrenergic receptors in the prefrontal cortex of DS victims.18,19 Recently, Gurevich et al,29 suggested that Control DS alterations in the postmodification regulation of gene kDa expression of serotonin might play a role in the etiology of major depression. It is likely that the 183 pathobiology of depression cannot be attributed to 114 dysfunction in a single neurotransmitter pathway. 81 Therefore, the search for other neurochemical ab- 64 normalities associated with depression is continuing. Recent studies from our laboratory have suggested 50 CB1R the participation of cannabinoidergic system in alcoholism and related behaviors.21–23 Existence of 38 comorbidity between alcoholism and depression led us to investigate the role of cannabinoid signaling in 26 depression. Indeed, the present study for the first time

reveals greater CB1 receptor density and coupling α−α− Tubulin between these receptors and Gi-protein in DLPFC of DS subjects. CB1 receptor immunoblot analysis found more CB receptor protein immunoreactivity, sub- b 1 800 stantiating the radioligand binding results. Control Many transmembrane signaling processes of extra- DS cellular hormone and neurotransmitters are mediated 600 by receptor interaction with heterotrimeric (a,b,g) ** guanosine nucleotide binding proteins (G-protein). The receptor activation alters the conformation of 400 G-proteins leading to the exchange of GDP by GTP on Ga-subunit. The conformational change promotes the dissociation of G-protein into active Ga-GTP and

(Arbitrary number) 30 Gbg-subunits. These two subunits later regulate receptor immunoreactivity 200 1 the activity of several effector molecules within the CB cell. Recently, a nonhydrolyzable GTP analogue, 35 0 [ S]GTPgS, has been employed to asses the coupling efficacy of several neurotransmitter receptors and G- Figure 3 (a) A representative immunoblot of the CB1 proteins in cortical membranes of human postmortem receptor (top) and the same blot was reporbed with a- 27 tubulin (bottom) to ensure equivalent total protein loading. brain. (b) Levels of CB receptor immunoreactivity in prefrontal The results of the present study suggest greater CB1 1 35 cortical membranes of DS (10) and normal controls (10) receptor-stimulated [ S]GTPgS binding in DLPFC of reexpressed in mean7SEM of arbitrary densitometric units. DS compared to matched controls. The observed **Po0.001. increase in [35S]GTPgS binding may be due to more CB1 receptors. Interestingly, we observed low percentage of stimulation of CB1 receptor-mediated 50 [35S]GTPgS binding either in DS or matched controls.

Control Despite the high density of CB1 receptors, the reason 35 DS for low agonist-stimulated [ S]GTPgS binding is not 40 ** known at this time. However, lower efficiency of CB1 receptor coupling to Gi-protein, compared to other 31 30 GPCRs has been suggested. This is borne out from direct comparison between the efficacy of cannabinoids and opiates in which opiates signaling was 20 % Stimulation found to be 20-fold more efficient than cannabinoid

(fmol/mg protein) 32 signaling. Therefore, it is speculated that the CB1 10 receptor signaling functions as a subtle, fine-tuning mechanism for cells. The high density of receptors

makes the CB1 receptor highly sensitive to agonists; 0 however, the poor coupling efficiency ensures that 31 35 overactivation of the system will not occur. Figure 4 CB1 receptor-stimulated [ S]GTPgS binding was done in prefrontal cortical membranes of DS (10) and The consequence of elevated CB1 receptor-mediated matched controls (10). Data, presented as percentage of signaling in the pathophysiology of depression is not stimulation over the basal, are mean7SEM values of two to known. Abnormalities in cAMP signaling in depres- three experiments, each assayed in triplicate. **Po0.001. sive disorders have been reported. Dowlatshahi

Molecular Psychiatry CB1 receptors-mediated signaling in brain of DS BL Hungund et al

33 189 et al, found decreased cAMP signaling in the brain may be assumed that the observed elevated CB1 of depressed suicides. The increased CB1 receptor receptor and mediated signaling may be a pathologi- density and its mediation in [35S]GTPgS binding cal consequence of depression and/or schizophrenia. suggest the sensitization of cannabinoidergic signal- However, the reported elevation of endocannabinoids ing, which may lead to the decreased cAMP content in CSF of schizophrenics38 reflects the overall of the cell as these receptors are negatively coupled metabolism of brain, rather than region-specific to AC. alteration. To understand the overall status of the Recent studies have suggested age-dependent al- endocannabinoidergic system in depression and terations in many neurotransmitter receptors. Aged other psychiatric illnesses, the study on endocanna- rats exhibited a marked decrease in CB1 receptors and binoid levels in different brain regions is essential its mediated [35S]GTPgS binding sites in rat brain.34 In and such studies are currently underway. this study, we also observed (data not shown) reduced Regulation of receptor sensitization and desensiti- receptor density associated with increasing age in zation is a complex phenomenon. The consequence of normal control subject, suggesting that receptor losses elevated CB1 receptor-mediated response observed in are related to the aging process. This observation is this study is not known. The hyperactivity of consistent with previous reports.15,16 cannabinoidergic signaling could be an adaptive Several studies have suggested that age, sex, PMI feedback in response to the decreased levels of and psychoactive drug medications may be respon- endocannabinoids. The mechanism, physiological sible for the alterations in neurotransmitter receptors role and regulation of endocannabinoidergic system and G-proteins. The brain samples analyzed in this are yet to be understood. Recently, it has been shown study were well matched with regards to sex, age, that endocannabinoids are involved in retrograde ethnic background and postmortem interval. Suicide, signaling, and CB1 receptor activation suppresses however, is often associated with major depression,35 neurotransmitter release by inhibiting a calcium- and postmortem studies are often unable to resolve dependent step in vesicle release,41 thus decreasing whether the observed abnormalities are due to the the local release of synaptic vesicles42–44 However, we presence of major depression or whether they reflect cannot rule out increased endocannabinoid levels, abnormalities that characterize suicidal behavior. In which, if combined with observed hyperactivity of this study, we are unable to tease out the effect of CB1 receptor-mediated signaling, and hence elevated suicide vs the effect of depression. Although suicid- retrograde cannabinoidergic neurotransmission in the ality is often associated with multiple depressive pathophysiology of depression or suicidal behavior. symptoms, future studies should test for the differ- In summary, the upregulation of CB1 receptors with ences between suicide victims with a history of major concomitant increase in the CB1 receptors-mediated depression and nondepressed suicides and or de- [35S]GTPgS binding strongly suggests a role for the pressed and normal subjects who died by similar participation of abnormal endocannabinoidergic neu- cause of death. The next question whether the rotransmission in the etiology of depression and observed abnormalities in DS victims are a conse- suicide. The pharmacological manipulation of endo- quence of pathobiology or antidepressant medication cannabinoid system may serve as a new therapeutic is of particular relevance. However, in the present target in the treatment of depression. study, only three patients had medications or alcohol at the time of death and no psychoactive drugs were Acknowledgements detected in the remaining patients. Therefore, the present findings in brains of DS are related to the This study was supported by Grants AA13003 and illness, be it suicide or major depression, rather than NARSAD independent investigator award (BLH); to antemortem drug treatment. AA09004 and MH40210 (VA); MH62185 (JJM). The It has been suggested that cannabis use aggravates preliminary findings of this study were presented at existing psychosis.36,37 Two endocannabinoids, which Neuroscience meeting. We thank Dr Veeranna, Centre act on CB receptor, anadamide and palmitoylethano- for Dementia Research, NKI, for his technical advice. lamide, were shown to be increased in the cerebrosp- 38 inal fluid (CSF) of schizophrenics. Increased CB1 receptor density in the DLPFC of schizophrenia has References 20 also been recently reported. It has been suggested 1 Dewey WL. Cannabinoid pharmacology. Pharmacol Rev 1986; 38: that the clinical signs of chronic cannabis consump- 151–178. tion may resemble negative symptoms of schizophre- 2 Hollister LE. Health aspects of cannabis. Pharmacol Rev 1986; 38: nia.39 Several common symptomatologies do exist 1–20. 3 Abood ME, Martin BR. Neurobiology of marijuana abuse. Trends between schizophrenia and mood disorders. From a Pharmacol Sci 1992; 13: 201–206. neuopharmacological standpoint, the psychoses of 4 Iverson LL. The Science of Marjuana. Oxford University Press: NY, schizophrenia and the mania of bipolar disorder can 2000, p 36. both be treated with antipsychotic drugs. Some 5 Piomelli D, Giuffrida A, Calignano A, Rodriguez de Fonseca F. The endocannabinoid system as a target for therapeutic drugs. Trends prominent negative symptoms of schizophrenia such Pharmacol Sci 2000; 21: 218–224. as affective flattering, alogia and avolition are most 6 Pertwee R. Cannabinoids and multiple sclerosis. Pharmacol Ther commonly observed in depression.40 Therefore, it 2002; 95:165.

Molecular Psychiatry CB1 receptors-mediated signaling in brain of DS BL Hungund et al 190 7 Matsuda LA, Lolait SJ, Brownstein MJ, Young AC, Bonner TI. the ventrolateral prefrontal cortex of suicide victims. Brain Res Structure of a cannabinoid receptor and functional expression of 1995; 688: 121–133. the cloned cDNA. Nature 1990; 346: 561–564. 25 Kelly TM, Mann JJ. Validity of DSM-III-R diagnosis by psycholo- 8 Munro S, Thomas KL, Abu-Shaar M. Molecular characterization of gical autopsy: a comparison with antemortem diagnosis. Acta a peripheral receptor for cannabinoids. Nature 1993; 365: 61–65. Psychiatr Scand 1996; 94: 337–343. 9 Devane WA, Hanus L, Breuer A, Pertwee RG, Stevenson LA, 26 Lowry OH, Rosebrough NJ, Farr AG, Randall RJ. Protein Griffin G et al. Isolation and structure of a brain constituent that measurement with folin phenol reagent. J Biol Chem 1951; 193: binds to the cannabinoid receptor. Science 1992; 258: 1946–1949. 265–275. 10 Sugiura T, Konda S, Sukagawa A, Nakane S, Shinoda A, Itoh K et 27 Gonzalez-Maeso J, Rodriguez-Puertas R, Gabilondo J, Meana JJ. al. 2-Arachidonoylglycerol: a possible endogenous cannabinoid Characterization of receptor mediated [35S]GTPgS binding to receptor ligand in brain. Biochem Biophys Res Commun 1995; 215: cortical membrane from post-mortem human brain. Eur J Pharma- 89–97. col 2000; 390: 25–36. 11 Pertwee RG, Stevenson LA, Elrick DB, Mechoulam R, Corbett AD. 28 Sawada K, Young CE, Barr AM, Longworth K, Takahashi S, Arango Inhibitory effects of certain enantiomeric cannabinoids in the V et al. Altered immunoreactivity of complexin protein in mouse vas deferens and the myenteric plexus preparation of prefrontal cortex in severe mental illness. Mol Psychiatry 2002; guinea-pig small intestine. Br J Pharmacol 1992; 105: 980–984. 7: 484–492. 12 Kaminski NE, Abood ME, Kessler FK, Martin BR, Schatz AR. 29 Gurevich I, Tamir H, Arango V, Dwork AJ, Mann JJ, Schmauss C. Identification of a functionally relevant cannabinoid receptor on Altered editing of serotonin 2C receptor pre-mRNA in the mouse spleen cells that is involved in cannabinoid-mediated prefrontal cortex of depressed suicide victims. Neuron 2002; 34: immune modulation. Mol Pharmacol 1992; 42: 736–742. 346–356.

13 Howlett AC, Qualy JM, Khachatrian LL. Involvement of Gi in the 30 Gilman AG. G proteins: transducers of receptor-generated signals. inhibition of adenylate cyclase by cannabimimetic drugs. Mol Ann Rev Biochem 1987; 56: 615–649. Pharmacol 1986; 29: 307–313. 31 Kearn CS, Greenberg MJ, DiCamelli R, Kurzawa K, Hillard CJ. 14 Herkenham M, Lynn AB, Johnson MR, Melvin LS, deCost BR, Rice Relationships between ligand affinities for the cerebellar canna-

KC. Characterization and localization of cannabinoid receptors in binoid receptor CB1 and the induction of GDP/GTP exchange. rat brain; a quantitative in vitro autoradiographic study. J Neurosci J Neurochem 1999; 72: 2379–2387. 1991; 16: 8057–8066. 32 Sim LJ, Selley DE, Childers SR. In vitro autoradiography of 15 Glass M, Dragunow M, Faull RL. Cannabinoid receptors in the receptor-activated G proteins in rat brain by agonist-stimulated human brain: a detailed anatomical and quantitative autoradio- guanylyl 50-[gamma-[35S]thio]-triphosphate binding. Proc Natl graphic study in the fetal, neonatal and adult human brain. Acad Sci USA 1995; 92: 7242–7246. Neuroscience 1997; 77: 299–318. 33 Dowlatshahi D, MacQueen GM, Wang JF, Reiach JS, Young LT. G 16 Westlake TM, Howlett AC, Bonner TI, Matsuda LA, Herkenham M. Protein-coupled cyclic AMP signaling in postmortem brain Cannabinoid receptor binding and messenger RNA expression in of subjects with mood disorders: effects of diagnosis, suicide, human brain: an in vitro receptor autoradiography and in situ and treatment at the time of death. J Neurochem 1999; 73: hybridization histochemistry study of normal aged and Alzhei- 1121–1126. mer’s brains. Neuroscience 1994; 63: 637–652. 34 Romero J, Berrendero F, Garcia-Gil L, de la Cruz P, Ramos JA, 17 Glass M, Faull RL, Dragunow M. A significant loss of cannabinoid Fernandez-Ruiz JJ. Loss of cannabinoid receptor binding and receptors in the substantia nigra in Huntington’s disease. Neu- messenger RNA levels and cannabinoid agonist-stimulated roscience 1993; 56: 523–527. [35S]guanylyl-50O-(thio)-triphosphate binding in the basal ganglia 18 Arango V, Underwood MD, Mann JJ. Serotonin brain circuits of aged rats. Neuroscience 1998; 84: 1075–1083. involved in major depression and suicide. Prog Brain Res 2002; 35 Mann JJ. The neurobiology of suicide. Nat Med. 1998; 1; 25–30. 136: 443–453. 36 Hollister LE. Health aspect of cannabis: revisited. Int J Neuropsy- 19 Gonzalez-Maeso J, Rodriguez-Puertas R, Meana JJ, Garcia-Sevilla chopharmacol 1998; 1: 71–80. JA, Guimon J. Neurotransmitter receptor-mediated activation of G- 37 Mathers DC, Ghodse AH. Cannabis and psychotic illness. Br J proteins in the brains of suicide victims with mood disorders: Psychiatry 1992; 161: 648–653.

selective supersensitivity of a2A-adrenoceptors. Mol Psychiatry 38 Leweke FM, Giuffrida A, Wurster U, Emrich HM, Piomelli D. 2002; 7: 755–767. Elevated endogenous cannabinoids in schizophrenia. Neuroreport 20 Dean B, Sundram S, Bradbury R, Scarr E, Copolov D. Studies on 1999; 10: 1665–1667. [3H]CP-55940 binding in the human central nervous system: 39 Pani L, Gessa GL. The substituted benzamides and their clinical regional specific changes in density of cannabinoid-1 receptors potential on dysthymia and on the negative symptoms of associated with schizophrenia and cannabis use. Neuroscience schizophrenia. Mol Psychiatry 2002; 7: 247–253. 2001; 103: 9–15. 40 Kupfer DJ, Detre T, Koral J, Fajans PA. Comment on the 21 Basavarajappa BS, Hungund BL. Chronic ethanol increases the ‘amotivational syndrome’ in marijuana smokers. Am J Psychiatry cannabinoid receptor agonist anandamide and its precursor N- 1973; 130: 1319–1321. arachidonoylphosphatidyl ethanolamine in SK-N-SH cells. 41 Kreitzer AC, Regehr WG. Retrograde inhibition of presynaptic J Neurochem 1999; 72: 522–528. calcium influx by endogenous cannabinoids at excitatory 22 Basavarajappa BS, Hungund BL. Down-regulation of cannabinoid synapses onto Purkinje cells. Neuron 2001; 29: 717–727. receptor agonist-stimulated [35S]GTPgS binding in synaptic plas- 42 Takahashi KA, Linden DJ. Cannabinoid receptor modulation of ma membrane from chronic ethanol exposed mouse. Brain Res synapses received by cerebellar Purkinje cells. J Neurophysiol 1999; 64: 429–436. 2000; 83: 1167–1180. 23 Hungund BL, Basavarajappa BS, Vadasz C, Kunos G, Rodriguez de 43 Kreitzer AC, Regehr WG. Cerebellar depolarization-induced Fonseca F et al. Ethanol, endocannabinoids, and the cannabinoi- suppression of inhibition is mediated by endogenous cannabi- dergic signaling system. Alcohol Clin Exp Res 2002; 26: 565–574. noids. J Neurosci 2001; 21: RC174. 24 Arango V, Underwood MD, Gubbi AV, Mann JJ. Localized 44 Wilson RI, Nicoll RA. Endocannabinoid signaling in the brain. alterations in pre- and postsynaptic serotonin binding sites in Science. 2002; 296: 678–682.

Molecular Psychiatry