|Home | About | Journals | Submit | Contact Us | Français|
Several lines of evidence link cannabinoid (CB) type 1 (CB1) receptor-mediated endogenous CB (eCB) signaling to the etiology of alcohol dependence (AD). However, to date, only peripheral measures of eCB function have been collected in living humans with AD and no human in vivo data on the potentially critical role of the brain CB1 receptor in AD have been published. This is an important gap in the literature, because recent therapeutic developments suggest that these receptors could be targeted for the treatment of AD.
Medication-free participants were scanned during early abstinence 4 weeks after their last drink. Using positron emission tomography (PET) with a high resolution research tomograph and the CB1 receptor selective radiotracer [11C]OMAR, we determined [11C]OMAR volume of distribution (VT) values, a measure of CB1 receptor density, in a-priori selected brain regions in men with AD (N=8, age 37.4± 7.9 years; 5 smokers) and healthy control (HC) men (N=8, age 32.5± 6.9 years; all non-smokers). PET images reconstructed using the MOLAR algorithm with hardware motion correction were rigidly aligned to the subject-specific magnetic resonance (MR) image, which in turn was warped to an MR template. Time-activity curves (TACs) were extracted from the dynamic PET data using a priori selected regions of interest delineated in the MR template space.
In AD relative to HC, [11C]OMAR VT values were elevated by ~20% (p=.023) in a circuit, including the amygdala, hippocampus, putamen, insula, anterior and posterior cingulate cortices and orbitofrontal cortex. Age, body mass index or smoking status did not influence the outcome.
These findings agree with preclinical evidence and provide the first, albeit still preliminary in vivo evidence suggesting a role for brain CB1 receptors in AD. The current study design does not answer the important question of whether elevated CB1 receptors are a pre-existing vulnerability factor for AD or whether elevations develop as a consequence of AD.
Almost 2 in 3 alcoholics resume drinking after engaging in alcohol treatment (Brown et al., 1995; Hunt et al., 1971), highlighting the substantial need to improve treatments for this patient population. At a neurobiological level, studies show impairments in decision-making in alcohol-dependent (AD) patients (Dom et al., 2006), which is associated with altered functions in cortico-limbic-striatal circuitry including the amygdala, hippocampus, anterior and posterior cingulate cortices, insula, and subregions of the striatum, i.e. putamen (Goldstein et al., 2004; Makris et al., 2008). At least three sets of factors may contribute to high alcohol relapse rates. First, individual differences in the positive, reinforcing properties of alcohol may increase risk of alcoholism and possibly alcohol relapse (Schuckit and Smith, 1996). Second, stimuli previously associated with alcohol use and its physiological and subjective effects may become paired with alcohol and serve as “conditioned cues” that can increase alcohol craving and subsequent alcohol use (O'Brien et al., 1998). Finally, differences in stress responsiveness may contribute to the development and maintenance of alcohol addiction (Brown et al., 1990; Miller et al., 1996; Sinha, 2001). All three factors can be linked to the endogenous cannabinoid (eCB) system and its attending CB1 receptor (Serrano and Parsons, 2011), and a growing body of evidence points to an involvement of CB1 receptors in the acquisition and maintenance of AD (Vinod and Hungund, 2006).
The behavioral effects of CB1 receptor knock-out (Hungund et al., 2003; Naassila et al., 2004; Poncelet et al., 2003; Thanos et al., 2005) and pharmacological manipulation of CB1 receptor function (Femenia et al., 2010; Maccioni et al., 2010; Maccioni et al., 2008; Malinen and Hyytia, 2008) result in reduced voluntary alcohol intake. In contrast, administration of a CB1 receptor agonist (Colombo et al., 2002; Gallate et al., 1999; Vinod et al., 2008) enhances alcohol consumption. These findings suggest that upregulation of CB1 receptor-mediated G- protein signaling in a circuit involving the amygdala, hippocampus, ventromedial prefrontal and orbitofrontal cortex, insula and striatum (Sullivan and Pfefferbaum, 2005) might contribute to the increased alcohol consummatory behavior in patients with AD.
Positron emission tomography (PET) imaging is a direct, sensitive and straightforward means of probing the functional neurochemistry of humans and assessing molecular targets in the brain in vivo. The recent development of a CB1 receptor selective radiotracer, designated [11C]OMAR (Horti et al., 2006; Wong et al., 2010), facilitates in vivo assessment of [11C]OMAR volumes of distribution (VT), a measure of total binding in the brain, and proportional to CB1 receptor density. Given the higher rates of AD among men compared with women (Nolen- Hoeksema and Hilt, 2006) and important sex differences in the biological underpinnings of AD (Fox et al., 2009), we measured CB1 receptors using PET and [11C]OMAR in men with AD during early abstinence defined as 4 weeks after their last drink and in healthy control (HC) subjects. Based on preclinical data, we hypothesized that CB1 receptor binding will be increased in men with AD compared to matched HC participants.
Eight men with AD, all medication-free in early abstinence 4 weeks after their last drink and eight male age- and body mass index (BMI)-matched HC participants without a history of psychiatric or substance use disorder were recruited through public advertisement. After giving informed consent, participants were evaluated and diagnosed using Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) criteria and the Structured Clinical Interview for DSM-IV (SCID) (First et al., 1996). AD patients did not have any additional axis-I diagnosis and were treatment engaged. They were participating in standard 3 times weekly substance abuse counseling and had been abstinent from alcohol for four weeks at the time of the PET scan. Abstinence was confirmed with regular breathalyzer and urine toxicology tests at each clinic visit to ensure abstinence before the PET scan. Signs and symptoms of alcohol withdrawal were quantified with the Clinical Institute Withdrawal Assessment of Alcohol Scale, Revised (CIWA-Ar). AD patients were past the acute alcohol-withdrawal period at the time of the PET scan. All participants were evaluated by physical examination, electrocardiogram, standard laboratory tests, urine analysis and toxicology and were free of significant medical or neurological conditions. Cannabis users were excluded given recent evidence showing reversible and regionally selective downregulation of brain CB1 receptors in human subjects who chronically smoke cannabis (Hirvonen et al., 2011). No participants received medication for at least 4 weeks prior to PET scanning. The protocol was approved by the Yale University School of Medicine Human Investigation Committee, the Human Subjects Subcommittee of the Veterans Affairs Connecticut Healthcare System, the Magnetic Resonance Research Center and the Yale New Haven Hospital Radiation Safety Committee.
[11C]OMAR was prepared at high specific activity (212± 95 MBq/nmol at end of synthesis) by methods previously described but adopted for synthesis using the TRACERlab FXC Pro automated synthesis module (GE Healthcare, Milwaukee, WI) (Horti et al., 2006). The radiotracer (651± 69 MBq, 45± 27 ng/kg) was infused as a slow bolus over 1 minute through the antecubital vein. The radioactivity concentration of blood from the radial artery was measured continuously using an integrated peristaltic pump and calibrated radioactivity detector (PBS101, Veenstra Instruments, Joure, The Netherlands) for the first 7 min after radiotracer administration with discrete sample manually obtained at 3, 5, 15, 30, 60, and 90 minutes post-injection. Manual samples were measured on a gamma counter (Wizard 1480, Perkin-Elmer, Waltham, Massachusetts, United States) to determine radioactivity concentration in whole blood and in plasma. The whole blood-plasma ratio was calculated and used to scale the continuous whole blood concentration data measured by the continuous blood counter used at the beginning of the scan. Selected discrete blood sampled were analyzed for the fraction of unchanged [11C]OMAR and its radiometabolites using a column-switching high performance liquid chromatography assay (Hilton et al., 2000).
Listmode emission data were collected for 120 minutes using the HRRT PET scanner (Siemens Medical Systems, Knoxville, Tennessee, United States). Subject motion was measured using the Polaris Vicra optical tracking system (Northern Digital Inc., Waterloo, Ontario, Canada). Vicra recordings were used to correct listmode data in the Motion-compensation Ordered subsets expectation maximization Algorithm for resolution-recovery Reconstruction (Carson et al., 2003) which produced dynamic images corrected for motion, photon scatter and attenuation, random coincidences, scanner dead-time, and detector inhomogeneity. The PET image sequences were registered to subject-specific T1-weighted magnetic resonance images acquired on a 3 Tesla Trio imaging system (Siemens Medical Systems, Erlangen, Germany). Anatomical MR images were in turn non-linearly warped to an MR template where regions of interest (ROIs) were defined (Tzourio-Mazoyer et al., 2002). These regions (Figure 1., Table 2.) constitute a neural circuit consistently implicated in the pathophysiology of AD, and are known to contain high levels of CB1 receptors. Regional time activity curves (TACs) were extracted from the dynamic PET data and analyzed using multilinear analysis (Ichise et al., 2002) with metabolite-corrected arterial input functions and cutoff time of t*=30 minutes. This kinetic analysis yields regional estimates of total VT, the equilibrium ratio of radioligand in tissue relative to arterial plasma (Innis et al., 2007), which reflects binding of [11C]OMAR to CB1 receptors plus nondisplaceable binding. The mulilinear analysis method used here was found to yield VT values in excellent agreement with full compartmental modeling (Normandin et al., 2010a). The cutoff time of t*=30 min was the shortest that adequately fit data from all regions, and exhibited the best test-retest reproducibility in an initial cohort of healthy subjects (Normandin et al., 2010b). Since no region without specific binding has been found for [11C]OMAR, binding potential cannot be determined.
Independent sample t-tests were used to compare continuous clinical and demographic variables between AD and HC. Data were normally distributed as determined by visual inspection and the Kolmogorov-Smirnov D test. Chi-square was used in the case of dichotomous variables. [11C]OMAR VT values were subjected to a repeated measures analysis of variance (rmANOVA) with group as a between-subject factor and brain region as a within- subject factor. Significant results for the omnibus test were followed-up with post-hoc analyses using ANOVA with regional [11C]OMAR VT as dependent variable and group as predictor variable to determine significant differences between groups in individual ROIs. Tests of association between continuous variables were performed using Pearson’s product-moment correlations. All tests were performed two-tailed, with results considered significant at p<0.05. Means and standard deviations are reported. All statistical analyses were conducted using SPSS version 16.0 (SPSS Inc, Chicago, IL, USA).
Participants in the AD and HC groups were matched for age and body mass index (BMI); smoking frequency was greater in the AD compared to HC group. As expected, AD patients relative to HC had a significantly earlier age at their first drink (15.8 ± 2.4 versus 21.9 ± 3.7 years) and had a mean illness duration of 16.6 ± 9.0 years. Participants in the AD group experienced low levels of alcohol withdrawal at the time of the PET scan as indicated by a mean CIWA score of 4.6 ± 4.2 (Table 1.).
[11C]OMAR VT values within the circuit were significantly higher in the AD group compared to the HC group (main effect of group, F=6.56; df=1,14; p=0.023). Post-hoc testing revealed elevated [11C]OMAR VT values in the a priori selected ROIs in patients with AD compared to the HC group (Figure 1., upper panel; Table 2.) which is evident also in an example comparing amygdala [11C]OMAR VT in a patient with AD relative to a HC subject (Figure 1., lower panel). Including smoking status as a co-variate in the rmANOVA did not influence the outcome. Exploratory analyses in other regions outside of the circuit including caudate, pallidum, parietal cortex or thalamus did not yield any significant between-group differences.
There were no associations between [11C]OMAR VT values and age or BMI in either group. There were no associations between [11C]OMAR VT values and CIWA score, age at first drink or illness duration in the AD group.
Herein we demonstrate for the first time in vivo elevated VT values of [11C]OMAR in chronic AD patients during early abstinence relative to matched HC subjects. Age at onset of drinking, number of drinks or illness duration did not correlate with [11C]OMAR VT. We interpret these findings as reflecting elevations in CB1 receptor density in AD patients.
Regional VT values showed good agreement with the known density and distribution of CB1 receptors in human autoradiography studies (Glass et al., 1997; Herkenham et al., 1990). Relatively lower hippocampus [11C]OMAR VT values despite highest [3H]CP55,940 binding shown in the autoradiography studies was also reported by other groups in their in vivo PET studies using different CB1 receptor specific radioligands (Burns et al., 2007; Terry et al.; Terry et al., 2010; Terry et al., 2009; Van Laere et al., 2008; Wong et al., 2010) and could be explained by the limited availability of the hippocampus CB1 receptor for in vivo binding as suggested by data showing that approximately 85% of these receptors are located in intracellular vesicles (Burns et al., 2007; Coutts et al., 2001).
AD is characterized by episodes of alcohol deprivation alternating with episodes of increased alcohol intake. Therefore, the timing of the PET studies in relation to alcohol consumption and deprivation is important to consider. Consistent with the alcohol deprivation phase in animals exposed to chronic alcohol intake which is associated with an upregulation of CB1 receptor-mediated G-protein signaling in response to reduced eCB levels (Vinod et al., 2006), we found elevated [11C]OMAR VT values in AD subjects, indicating a possible upregulation of CB1 receptors. Similarly to animal studies, CB1 receptor upregulation could be associated with enhanced CB1 receptor mediated G-protein signaling. We recently showed decreased serum anandamide levels in AD patients (Mangieri et al., 2009) and, although no significant correlation exists between anandamide concentrations in cerebrospinal fluid and serum (Koethe et al., 2009), it can be speculated that the CB1 receptor upregulation in AD is a compensatory mechanism for a decrease in eCB tone during early abstinence. This change may represent a vulnerability factor for the development of craving and relapse. It is unclear, however, if these alterations are a state marker of early abstinence or represent a trait characteristic for AD. Animal models would suggest that we describe herein a state marker for early abstinence because CB1 receptor function (Ortiz et al., 2004) as well as expression (Basavarajappa et al., 1998; Basavarajappa and Hungund, 2005) were reduced immediately after chronic alcohol intake. Future PET studies in actively drinking AD patients will answer this important question.
Our study does not answer if elevated CB1 receptors are a pre-existing vulnerability factor for AD per se or result from alterations in other neurochemical systems that have been linked to the etiology of AD. For example, interactive effects between reduced striatal dopamine 2 and elevated CB1 receptor availability were reported recently in animal models of AD (Thanos et al., 2011) and were substantiated with our human in vivo data showing elevated binding in the putamen suggesting complex system interactions moderating the reinforcing and addictive effects of alcohol in AD. Future longitudinal PET studies need to validate these animal models clarifying if changes of CB1 receptor-mediated G-protein signaling change in different phases of alcohol intake and abstinence as well as if altered CB1 receptor expression in AD represents a change per se in the presence and absence of alterations in other neurochemical systems linked to AD.
In addition, it needs to be determined if women with AD show similar changes in CB1 receptor binding, importantly because dopamine and CB1 receptor interactions have only been studied in male mice to date (Thanos et al., 2011) and sex-dependent differences in eCB and dopaminergic functions do not allow to necessarily generalize our results for both sexes.
In conclusion, we report for the first time elevated CB1 receptor binding in AD. Chronic alcohol use is seen as persistent challenges to the organism, leading to an altered set-point across multiple systems. This hypothesis is consistent with evidence of adaptations in brain reward and stress circuits that could contribute to important aspects of addictive processes like craving or urges, decreases in self-control, and a compulsive engagement in unhealthy behaviors, all of which characterize patients with AD. Our study directly implicates the CB1 receptor in AD and suggests a mechanism for novel treatment development for AD (Bailey and Neumeister, 2011).
The authors acknowledge the excellent work of the staff of the Yale PET Center and the nursing support from Michelle San Pedro, R.N. and Brenda Breault, R.N., B.S.N. for their contributions with patient care during the PET scans.
The project described was supported by the National Institutes of Health (NIH) Roadmap for Medical Research Common Fund, through the following grants: UL1-DE019586 and the PL1-DA024859, the National Institute on Alcohol Abuse and Alcoholism (NIAAA) through the following awards: RAA017540B and RAA017540Y (American Reinvestment and Recovery Act of 2009), and RAA018329B. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH or NIAAA.
Conflict of Interest/Financial Disclosures
The authors report no potential conflict of interest relevant to this article.