The present study describes region‐specific differences in the expression of key elements within the ECS between individuals with AUD and C. These differences were observed in two brain regions within the mesocorticolimbic circuit, the PFC and NAc: (1) compared to C, individuals with AUD showed higher CNR1 and lower CNR2 expression in the PFC and NAc; (2) GPR55 expression was higher in the PFC but lower in the NAc; and (3) FAAH expression was lower in the PFC and higher in the NAc, whereas MGLL expression was lower in the NAc, with no group differences observed in the PFC.

These results are highly innovative for two main reasons. First, the samples analyzed are unique, AUD patients with a long history of alcohol consumption (35.5 ± 8.7 years) and no misuse of other addictive drugs. These samples enable the assessment of potential molecular differences associated with chronic alcohol use. Second, to our knowledge, this is the first study to simultaneously measure gene expression of cannabinoid receptors (CNR1, CNR2), the extended ECS receptor GPR55 and the enzymes FAAH and MGLL in brain regions of the mesocorticolimbic system in AUD patients. The study focused on the PFC and NAc, which are brain regions that play a crucial role in different aspects of alcohol addiction. The NAc is closely involved in reward, conditioning and habit formation [40, 41]. Dysfunctions of the PFC were related to the development of physical and social consequences induced by craving and compulsive drug‐seeking [42]. This may help identify whether differences in these targets are associated with chronic alcohol use.

Our results showed higher levels of CNR1 expression in the PFC and NAc of AUD patients compared with C. These findings generally match post‐mortem human studies on AUD. An increase in CB1R protein was observed in the PFC of AUD patients, whereas no changes in mRNA levels were detected [32]. Additionally, enhanced CB1R‐mediated G‐protein signaling has been previously detected in the PFC of alcoholic suicide victims [43]. Therefore, our results support earlier reports suggesting that chronic alcohol use may be associated with elevated CNR1 expression and signaling in cortical and limbic regions. Such alterations could influence dopaminergic signaling [29, 44] and synaptic plasticity, mechanisms that have been involved in alcohol reinforcement and relapse [4, 45].

However, our findings differ from PET studies reporting reduced CB1R availability in the brains of individuals with AUD [29, 30]. This discrepancy may reflect methodological differences between post‐mortem molecular analyses, which measure gene expression, and PET imaging, which assesses receptor binding and availability. Other factors, such as duration of abstinence before death or scanning, receptor internalization or post‐transcriptional regulation, could also contribute to these differences.

Conversely, lower levels of CNR2 expression were found in the PFC and NAc of individuals with AUD compared to C. These results are in agreement with previous studies of our lab that reported increased vulnerability to alcohol consumption in mice deficient in CB2R [7]. Moreover, CB2R mediates neuroprotective and anti‐inflammatory processes that may be impaired in AUD [46, 47, 48]. Therefore, the lower CNR2 expression identified in this study may reflect impaired regulatory mechanisms involved in neuroinflammation and neuronal integrity in individuals with chronic alcohol consumption [49, 50].

The fact that both cannabinoid receptor expressions differed in opposite directions between individuals with AUD and C may indicate a potential cooperation/interaction between them, as supported in previous studies. In animal models, a downregulation of CB2R and CB1R has been reported in the amygdala [51] and striatum [52, 53] following repeated ethanol withdrawal. Moreover, pharmacological blockade of CB2R has been shown to modify the Cnr1 expression in the NAc of mice exposed to ethanol self‐administration [54].

Interestingly, significant differences in GPR55 were also found, with higher levels in the PFC and lower levels in the NAc of AUD compared to C. Although there are no direct post‐mortem data on GPR55 expression in AUD brains, peripheral and in vitro studies suggest that ethanol induces GPR55 upregulation via an epigenetic mechanism [33]. Our finding of higher GPR55 expression in the PFC and lower in the NAc suggests a region‐specific role for this orphan G‐protein‐coupled receptor, a possibility warranting further investigation in human post‐mortem tissue.

This study confirms previous research that highlights the close interaction between cannabinoid receptors and GPR55. It has been demonstrated that CB1R and CB2R form heteromers in the rat striatum and that activated microglia exhibit bidirectional cross‐antagonism, negative cross‐talk in resting microglia and positive cross‐talk in activated microglia. Additionally, CB1R‐CB2R heteromers appear to help reduce neurotoxicity and neuroinflammation caused by psychostimulants [55, 56]. Recent studies also show that CB1R and GPR55 form heteromers in the striatum and human embryonic kidney (HEK293) cells, affecting their function [57]. Moreover, CB2R and GPR55 collaborate to regulate the immune response [24, 58] and play key roles in suicide behavior [59]. Additionally, modulating GPR55, CB1R and CB2R is partly linked to lowering alcohol intake, vulnerability and relapse caused by cannabidiol (CBD) in mice [60]. Although further research is necessary to clarify the molecular mechanisms involved, it is plausible that cannabinoid receptors and GPR55 jointly influence pathways underlying alcohol dependence.

Furthermore, we found lower FAAH expression in the PFC and higher expression in the NAc, whereas MGLL expression did not differ in the PFC, but was lower in the NAc. These region‐specific differences may reflect alterations in ECS signaling, affecting the availability of anandamide (AEA) and 2‐AG, which are critical for synaptic function and plasticity [22, 61]. Previous studies have linked associations between differences in FAAH and MGLL expression and behavioral domains such as anxiety, stress response and addictive behaviors, all relevant to the pathophysiology of AUD [4, 62]. However, the results on FAAH expression in human post‐mortem studies are inconsistent or understudied. Erdozain et al. [32] reported no consistent bidirectional change in FAAH protein levels in the PFC, although downstream functional markers were altered. In line with our findings, PET imaging studies have shown lower FAAH availability in the brains of heavy‐drinking young adults and in individuals during early abstinence [34, 35]. Together, these findings suggest that alterations in FAAH may potentially influence emotional and reward‐related processes to alcohol. Our results provide evidence of region‐specific differences in FAAH and MGLL expression that have not been previously described in human post‐mortem tissue.

Additionally, exploratory Spearman’s correlations revealed significant associations within the AUD group. Brain pH correlated positively with CNR1 and negatively with GRP55 expression in the PFC. However, no significant correlations were observed in the C group, and mean brain pH values were comparable between groups (AUD: 6.6 ± 0.2; C: 6.7 ± 0.2). Importantly, no significant correlations were found between gene expression levels and the RIN or PMI. These findings suggest that differences in gene expression are unlikely because of sample degradation or other post‐mortem factors.

Furthermore, CNR2 expression in the NAc was inversely associated with the age of drinking onset, suggesting that earlier initiation of alcohol use may be linked to lower CNR2 expression levels in this brain region. These correlations should be interpreted with caution, given the modest sample size and the exploratory nature of the analysis. Still, they may provide insights into the relationship between clinical variables and ECS‐related gene expression in AUD. Conversely, correlations that did not reach statistical significance should be interpreted as a lack of statistical evidence for association, rather than evidence of the absence of a relationship.

In summary, this descriptive post‐mortem study reveals region‐specific differences in ECS gene expression between individuals with AUD and C. These molecular differences may reflect alterations in ECS‐related pathways associated with AUD and contribute to a better understanding of its neurobiological mechanisms. However, causal relationships cannot be inferred from these observational data.

One limitation of the study is that several individuals in the AUD group presented co‐morbid depression, whereas none of the C cases did. Because depression is also linked to the ECS, this comorbidity may have influenced the findings and should be addressed in future studies with depression‐matched C groups. Another limitation of this study is the relatively small sample size and the inclusion of only male subjects, which may limit the applicability of the findings to broader populations. Only two brain regions were analyzed because of tissue availability. Future studies should include additional areas, such as the ventral tegmental area, to provide a more comprehensive understanding of ECS alterations in AUD.