Review article
Dopamine and opioid systems adaptation in alcoholism revisited: Convergent evidence from positron emission tomography and postmortem studies

https://doi.org/10.1016/j.neubiorev.2018.09.010Get rights and content

Highlights

  • A hypodopaminergic model is commonly proposed to underly relapse to alcohol drinking.

  • Addiction leads to highly dynamic changes in dopamine and opioid systems.

  • Existing data support both hypo- and hyperdopaminergic states during abstinence.

  • A revised model of opioid-dopamine interaction is proposed in protracted abstinence.

Abstract

A major hypothesis in the addiction field suggests deficits in dopamine signaling during abstinence as a driving mechanism for the relapsing course of the disorder. Paradoxically, blockade of mu-opioid receptors (MORs) intended to suppress dopamine release and alcohol reward is a widely used treatment for preventing relapse in alcohol use disorder (AUD). To elucidate this apparent discrepancy, we systematically survey the literature on experimental studies in AUD subjects and animal models, which assessed striatal dopamine levels and D1, D2-like receptor, dopamine transporter and MOR via positron emission tomography (PET) and ex vivo receptor binding assays. The reported evidence indicates a changing dopaminergic signaling over time, which is associated with concomitant alterations in MOR, thus suggesting a highly dynamic regulation of the reward system during abstinence. Such a view can reconcile the various evidences from in vivo and postmortem studies, but makes developing an effective pharmacological intervention that specifically targets either dopamine receptors or the transporter system a daunting task.

Introduction

More than 2 billion people worldwide regularly drink alcohol. Alcohol is a contributing cause of more than 200 diseases and leads to approximately 3.3 million deaths per year globally (World Health Organization, 2014). The major disease burden results from alcohol use disorders (AUD), including alcohol dependence, which affect approximately 15% of regular drinkers. Alcohol related harm is largely determined by the amount consumed. In fact, AUD patients with very high drinking risk (>100 g / day) lose 22 years in life expectancy (Rehm, 2018; van den Brink et al., 2018).

Therefore, reducing alcohol related harm is a major public health priority. Unfortunately, treatment options are limited. Oral naltrexone (NTX) - a non-selective opioid antagonist - is a prototypical example and provides proof-of-concept for successful AUD pharmacotherapy, but suffers from modest effect size and low impact on clinical practice (Jonas et al., 2014a).

It is suggested that NTX blocks mu-opioid receptors (MORs) within the dopaminergic reward system and thereby produces its anti-relapse effects (Altshuler et al., 1980; Heilig et al., 2011). Here we will discuss why this simple explanation falls short and has to be revised. First, we will briefly summarize the current understanding of the function of dopamine and opioid systems in mediating alcohol reward. We then describe the clinical implication of this interaction for the treatment of AUD using opioid receptor antagonists and the interpretation of molecular neuroimaging studies by positron emission tomography (PET). The original conclusions drawn from these studies are seemingly discordant with recent findings on MOR and dopamine receptor levels in postmortem brain tissue from AUD subjects. To shed light on this conundrum, we survey the literature on striatal dopamine and mu-opioid receptor levels in AUD subjects, as well as rodent models of AUD, with the aim of achieving a broader and less biased view of the current data. An overarching objective of this review is to present a framework for translating the considerable advances recently made in preclinical research on dopamine and MOR in alcohol addiction into therapeutic outcomes for patients.

It is a major challenge to understand why some individuals develop AUD, whereas others do not. Addiction results from a complex interplay of genetic (Goldman et al., 2005; Hiroi and Agatsuma, 2005), developmental, and environmental factors (Volkow and Li, 2005). The hallmark of an addictive disorder is its chronic relapsing course characterized by a strong resistance to change an apparent self-destructive behavior. Initially controllable drug consumption relies on subjective effects induced by the drug, such as the increase of positive subjective feelings or the relief of negative affective states. Although alcohol affects many neurochemical systems, its effects are primarily mediated by the brain’s reward system with dopamine and endogenous opioids playing major roles (Spanagel, 2009). These neurotransmitter systems interact to facilitate the salience and pleasurable effects of many drugs, including alcohol. During this initial stage, drug use is largely goal-directed towards the well-being or optimized performance of the individual. Continued regular use can lead to the formation of habits or routines, whereby taking a drug becomes uncoupled from its intended goal of well-being and is largely compelled by particular external cues. In some individuals, this process progresses to compulsive drug seeking and taking which no longer can be controlled or limited (Fig. 1). Withdrawal symptoms and negative emotional states emerge when drug use is discontinued. In this stage, strong drug desire or urges (“cravings”) may occur, which can be induced by drug-associated (conditioned) cues (O’Brien et al., 1988), or by small amounts of the drug (“drug priming”), or by stress; these all contribute to relapse during attempts at abstinence (Goldstein and Volkow, 2002; Koob and Le Moal, 1997). While cravings are initially related to the rewarding effects of the drug, their nature seems to shift more and more towards a craving for relief of aversive states (Heinz et al., 2003). The emergence of negative affective symptoms at the expense of an increasingly compromised reward system during progression of AUDs has been termed the ‘dark side’ of alcohol addiction (Koob and Le Moal, 2005). Several authors conceptualized this process as an allostatic dysregulation of the reward system (Koob and Le Moal, 1997; Siegel, 1983; Solomon, 1977).

Allostasis – literally ‘stability through change’ according to Sterling and Eyer (1988) describes a process by which homeostasis can only be maintained by shifting the set point around which it is regulated. The application of the allostasis concept has provided a powerful research framework with support from clinical observations and animal models. Chronic intermittent exposure (CIE) to ethanol vapor in rodents leads to severe withdrawal symptoms, escalated ethanol self-administration, high motivation to consume the drug revealed by progressive ratio breakpoints, enhanced alcohol seeking, and reduced sensitivity of self-administration to punishment (for review see (Meinhardt and Sommer, 2015)). Reduced positive reinforcement in this model has been deduced from examination of reward thresholds in intracranial self-stimulation experiments during acute withdrawal from alcohol (Koob, 2013; Schulteis and Liu, 2006; Schulteis et al., 1995). Some human findings can be interpreted as supporting a reduced impact of positive reinforcement in AUD according to the allostatic model. The most direct evidence for this process comes from a series of studies by Ramchandani et al., which directly showed ventral striatum activation to alcohol challenge in light social drinkers, attenuation of this signal in heavy social drinkers (some of whom met criteria for diagnosis), and lack of this response in hospitalized treatment-seeking patients (Gilman et al., 2012, 2008; Spagnolo et al., 2014). Also, exposure to alcohol-related stimuli was found to be correlated with reward craving or neural cue reactivity in reward-related brain areas (e.g., ventral striatum) in non-dependent alcohol consumers, while there was no such association in heavy drinkers with AUD (Bujarski et al., 2017; Bujarski and Ray, 2014; Vollstadt-Klein et al., 2012).

Despite its plausibility and experimental support, the allostatic model has some apparent weaknesses. Firstly, the main prediction of the allostatic model – craving and drinking being correlated or driven by negative reinforcement and affect – still lacks experimental support in both humans and animal models. Secondly, the proposed transition into a new functional state is difficult to operationalize. So far, no definitions of “dark side symptoms” or proxies thereof and means for their quantification have been proposed that could be widely applied to the study of brain or behavior in a translational manner. Lastly, the strong emphasis on negative affect by the proponents of the allostasis model seems to disregard the role of positive reinforcement as a driving force for alcohol consumption once an addiction has developed. As we discuss below, this does not reflect clinical reality, where patients are highly heterogeneous and differ in their account of reward and positive incentives as drivers of relapse behavior. To summarize, the allostatic model predicts a progressive decline of dopamine and opioid mediated reward function in AUD, but it is less clear to what extent they still contribute to alcohol drinking and relapses in later stages of the disorder.

Activation of the mesolimbic pathway is thought to reinforce the pursuit or consumption of rewards and to mediate the addictive properties of drugs (Di Chiara et al., 2004; Robinson and Berridge, 2003; Wise, 2004). Studies in experimental animals have indeed demonstrated that alcohol activates the mesolimbic dopamine circuitry (Di Chiara and Imperato, 1988; Spanagel, 2009). A hallmark of this activation is the release of dopamine from neurons originating in the ventral tegmental area (VTA) of the midbrain into striatal target areas particularly in the nucleus accumbens (Acb). VTA dopaminergic (∼ 60%) and GABAergic (∼ 30%) neurons project to the striatum and other brain areas including prefrontal cortex and amygdala (Sesack and Grace, 2010). There, dopamine binds to its targets - dopamine receptors of mainly the D1 and D2 type - which consequently leads to cellular events that modulate neuronal activity. In the striatum, GABAergic medium spiny neurons (MSNs) constitute the major cell type and account for ∼ 90% of striatal neurons (Meredith, 1999). These neurons encode information in two parallel pathways: 1) directly to the mesencephalic output nuclei of the basal ganglia and 2) indirectly to the midbrain through projections to the pallidum. The direct pathway is characterized by D1-containing MSNs and the indirect pathway predominantly by D2-containing MSNs (Gerfen, 1992; Gerfen et al., 1990; Surmeier et al., 1996). This architecture was mainly established for the dorsal striatum. The use of modern circuit dissection methods revealed that the output of the ventral striatal reward circuit is considerably more nuanced: D1-MSNs convey information not only directly to the ventral mesencephalon, but also by a considerable proportion to the ventral pallidum, and thus signal indirectly both to the ventral mesencephalon, but also to the ventral medial thalamus (Kupchik et al., 2015; Sesack and Grace, 2010). These findings indicate that, in contrast to the canonical view, D1- and D2-MSNs of the ventral striatum/nucleus accumbens may inhibit or disinhibit thalamic activity, not based on their genetic and neurochemical characteristics, but depending on their projection pattern.

Activation of striatal D1 receptors facilitates signaling through induction of long-term potentiation (LTP) on glutamatergic synapses (Gerfen and Surmeier, 2011; Grace et al., 2007). In contrast, D2 stimulation blocks signaling by induction of long-term depression (LTD) (Shen et al., 2008). These processes are crucial for reward- and aversion learning, respectively (Danjo et al., 2014; Hikida et al., 2010, 2013; Kravitz et al., 2012). How the dopamine signal and the interplay of D1 and D2 receptors encode reward and related addictive properties remains controversial. In fact, neurochemical lesions show that mesolimbic dopamine activity, while contributing to alcohol reinforcement, is not essential to maintain alcohol intake (Kiianmaa et al., 1979; Rassnick et al., 1993). D1 receptors seem to be the major effector in encoding alcohol reward processes, while D2 receptors counteract these processes both at the post- and presynaptic level. Available dopamine levels are further controlled by presynaptic reuptake of dopamine via the dopamine transporter (DAT).

Dopamine neurotransmission in the mesolimbic system is further controlled by MORs located on GABAergic interneurons in the VTA, which exert inhibitory tone onto dopamine neurons. MSNs of the direct pathway synapse onto non-dopaminergic interneurons in the VTA and the interneuron terminals have been proven to be sensitive to MOR agonists (Xia et al., 2011). Presynaptic activation of these MORs results in hyperpolarization via G-protein dependent voltage-gated potassium channel pathways (Johnson and North, 1992; Zhang et al., 2015). This removes GABAergic inhibition on dopaminergic neurons. The activity of MORs is crucial for the maintenance of baseline dopamine levels, as well as for the firing of the dopamine neuron response to activating cues (Spanagel et al., 1992). It has been proposed that blockade of the MOR suppresses ethanol reinforcement by attenuating ethanol-stimulated mesolimbic dopamine activity (Herz, 1997). Alcohol induces the release of endogenous opioid peptides and systemic pharmacological blockade of MOR largely prevents accumbal dopamine release induced by alcohol intake (Tanda and Di Chiara, 1998). However, while dopamine release evoked by other rewards such as morphine, nicotine and highly palatable food can be efficiently blocked by site-specific injection of MOR antagonists into the VTA, this was not demonstrated for alcohol in this study. Later rodent studies are ambiguous regarding the role of VTA in mediating alcohol – opioid – dopamine interactions. For example, VTA microinjections of NTX suppress alcohol-evoked dopamine release into the nucleus accumbens, but do so in a highly delayed manner compared to the immediate effect of VTA-NTX on morphine-evoked dopamine release (Ramachandra et al., 2011; Valenta et al., 2013). In this case the authors concluded that MOR activation may not be necessary for the initiation of mesolimbic dopamine release, but rather, is required for the maintenance of ethanol-stimulated dopamine release. Furthermore, selective knockout of Oprm1 in GABAergic forebrain neurons, leaving MOR in the midbrain (including the VTA) and hindbrain intact, resulted in reduced alcohol intake (Ben Hamida et al., 2017). Thus, MOR in the forebrain, specifically, are able to account for the attenuation or elimination of alcohol intake observed in global MOR knockout mice (Hall et al., 2001; Roberts et al., 2000). Further support for this view comes from experiments with intra-Acb injection of MOR agonists resulting in increased accumbal dopamine (Spanagel et al., 1990; Yoshida et al., 1999). In the Acb, MORs are located on corticostriatal terminals, extrasynaptically on MSN dendrites (Gracy et al., 1997; Svingos et al., 1997), as well as on pre-synaptic GABAergic afferents (Svingos et al., 1997; Wang and Pickel, 1998). Endogenous opioid peptides acting on accumbal MORs can exert rewarding effects (Koob et al., 1986; Simmons and Self, 2009). Of the various striatal MOR populations, reward behaviors are driven by MORs located on the direct-pathway D1-MSN (Cui et al., 2014). Further, it has been demonstrated that pharmacological blockade by NTX disrupts a key step in the cascade of alcohol triggered cellular events, namely the phosphorylation of DARPP32 (dopamine and cAMP-regulated phosphoprotein, Mr 32 kD) in striatal neurons (Bjork et al., 2010). Furthermore, a recent in vivo neuroimaging study in rats revealed robust blockade of alcohol drinking related activity within the Acb by NTX (Dudek et al., 2016). In these experiments, neuronal activity was assessed by manganese enhanced magnetic resonance imaging (MEMRI), a method that reports the intracellular accumulation of manganese – a paramagnetic contrast agent that can enter the neurons based on its calcium-like properties. Alcohol consumption increased the manganese influx into the Acb and its projections, while accumulation was blocked by NTX. In conclusion, the specific site of action of MOR blockade with respect to alcohol drinking remains somewhat elusive. A simplified scheme of MOR actions within the neurocircuitry between VTA and Acb is shown in Fig. 2.

In humans, the binding of a receptor ligand can be monitored by PET. Displacement of a radioligand provides an indirect measure of transmitter release. Using the D2-like antagonist [11C]-raclopride, some studies found support for alcohol-evoked dopamine release in the Acb of healthy social drinkers (Boileau et al., 2003; Ramchandani et al., 2011; Urban et al., 2010), but this effect was only found in male subjects. In contrast, studies by Yoder et al., 2007, Yoder et al., 2016 could only detect alcohol-evoked dopamine release in the ventral striatum of alcohol dependent subjects, not healthy social drinkers. Finally, a [11C]-carfentanil displacement study showed that alcohol intake indeed results in release of endogenous opioids (Mitchell et al., 2012).

Taken together, the reviewed human and animal literature supports the idea that blockade of MOR could disrupt alcohol-evoked striatal dopamine release and consequently consumption, although the exact mechanism of action is not well understood.

Interestingly, both in animals and humans, this effect was observed using NTX long before the above mentioned mechanisms were elucidated (Altshuler et al., 1980; Egli, 2005; Heilig et al., 2011; Ross et al., 1976). Ultimately, the findings from these studies led to the development and approval of NTX as a medication for relapse prevention in alcoholism (O’Malley et al., 1992; Volpicelli et al., 1992). Over the years, the efficacy of NTX has been supported by several meta-analyses (Bouza et al., 2004; Chamorro et al., 2012; Jonas et al., 2014a; Rosner et al., 2010, 2008). These studies provide robust evidence for NTX’s efficacy accumulated from many thousands of patients, but they also highlight its challenges: namely, modest effect size, with the estimated number of patients needed to treat (NNT) ranging around 10 (Jonas et al., 2014a). Furthermore, several human laboratory studies have examined NTX effects on alcohol responses and self-administration during short-term medication protocols. A meta-analysis of these studies supports the view of NTX as an efficacious treatment and, here, no differences in effect size depending on study population were found (Hendershot et al., 2017). To improve clinical outcomes, the use of extended release NTX has been suggested. After a successful 12-week clinical trial demonstrating efficacy in reducing the number of heavy drinking days per week and the number of drinks consumed (Ciraulo et al., 2008), the U. S. Food and Drug Administration (FDA) approved extended release naltrexone (XR-NTX, injected once per month) for the treatment of alcohol dependence in 2006. A first meta-analysis showed beneficial effects of XR-NTX versus acamprosate and oral naltrexone (Hartung et al., 2014). In experimental animals, effects of chronic NTX are variable. For example, NTX treatment delivered via osmotic minipumps actually increased alcohol relapse behavior in rats (Holter and Spanagel, 1999). A recent study also found tolerance in reducing alcohol drinking, as well as the development of supersensitivity of opioid receptors (Korpi et al., 2017).

A different approach to improving opioid antagonist therapy was suggested early on by Sinclair (1990). He postulated, based on animal data, that drinking alcohol while on antagonist treatment may extinguish alcohol’s reinforcing properties (Sinclair, 2001). This idea was later supported by observations in patients. Based on this work, a structured treatment approach was developed using the non-selective opioid antagonist nalmefene as an intermittent, “as needed” form of treatment as opposed to continuous daily administration. In practice, patients are asked to take their medications only when they feel at risk of returning to heavy drinking (Karhuvaara et al., 2007). Applying this treatment schedule, the efficacy of nalmafene in reducing alcohol consumption in dependent patients was successfully demonstrated, leading to its approval in Europe.(Gual et al., 2013; Mann et al., 2013). Nalmefene also offers an improved clinical profile, including prolonged duration of action and reduced hepatotoxicity. In contrast to NTX, nalmefene acts as an antagonist at mu- and delta-opioid receptors (M- and DOR), while exhibiting partial agonistic activity at kappa opioid receptors (KOR) (Bart et al., 2005). Nalmefene‘s affinity for KOR and DOR is higher than that of NTX. Administration of nalmefene into the Acb of alcohol-dependent rats reduced self-administration to a higher degree than in non-dependent rats and this effect was attributed to KOR-mediated mechanisms (Nealey et al., 2011; Walker and Koob, 2008). It is assumed that nalmefene could be superior to NTX treatment by also acting at hypersensitized KORs and thereby reducing negative affective states. Similarly to NTX, a recent meta-analysis found the clinical efficacy of nalmefene to be limited (Palpacuer et al., 2015), but no comparative studies between both treatments are available so far.

Despite the support obtained from meta-analyses for the efficacy of opioid antagonist treatment in AUD, the average effect size is small, which clearly affects clinical utility and impact (Mark et al., 2003). Efforts have been undertaken to determine the source of the observed variance in efficacy of MOR antagonists on alcohol outcomes and to identify predictors for a positive treatment response. Low effect sizes could reflect heterogeneity among patients, as some individuals seem to improve dramatically, while others show no response to NTX pharmacotherapy (Heilig et al., 2011). Accordingly, there is demand for personalized treatment approaches. Indeed, it has been reported that a family history of alcoholism (due to environmental and/or genetic factors) positively influences the therapeutic response to NTX (Jonas et al., 2014a, b; King et al., 1997; Krishnan-Sarin et al., 2007; Rubio et al., 2005), while in individuals without a family history of alcohol dependence, NTX can increase alcohol drinking (Krishnan-Sarin et al., 2007).

A genetic factor that has received extensive attention in human and animal studies for its potential contribution to the effects of NTX treatment is a functional variant within the MOR gene locus (OPRM1: rs1799971_A > G, often referred to as A118 G). We will only briefly summarize this line of research here because it has been excellently reviewed recently by Heilig et al. (2011). This single nucleotide polymorphism is located within exon 1 of the OPRM1 gene and encodes a non-synonymous substitution (Asn40Asp) in the extracellular N-terminal loop of MOR, resulting in loss of a glycosylation site (Beyer et al., 2004; Bond et al., 1998). The precise molecular consequences of this mutation remain unclear, but animal studies indicate that agonist-specific differences in opioid sensitivity associated with the OPRM1 A118 G polymorphism are not due to altered receptor function, but rather are determined by the number of receptor-binding sites and the size of the receptor reserve (Robinson et al., 2015). With respect to alcohol, studies in animal models, the human laboratory, and some (but not all) clinical trials all implicate the OPRM1 118 G allele in elevated alcohol reward and enhanced therapeutic response to naltrexone (reviewed in (Chamorro et al., 2012; Garbutt et al., 2014; Heilig et al., 2011; Ray et al., 2012). However, human genetic studies aimed at identifying risk variants suffer from two important potential confounds, i.e. linkage disequilibrium with other variants and stratification bias. These problems can be overcome by genetically modified mouse models, for example by replacing the endogenous mouse Oprm1 exon 1 with the corresponding human sequences encoding either the A- or G allele of rs1799971 (Ramchandani et al., 2011). By differing only in a single nucleotide, these ‘humanized’ h/mOPRM1-118 A A and h/mOPRM1-118 G G mouse lines allow for examination of the functional consequences of each variant in isolation. 118 G G mice replicate important phenotypes associated with the G allele in humans (Bilbao et al., 2015). Importantly, using in vivo microdialysis to measure alcohol-induced DA release in the ventral striatum of these mice, enhanced dopamine responses to alcohol in the 118 G G line compared to 118 A A mice was found, which corresponds to the findings in human G-allele carriers measured by [11C]-raclopride displacement. Other relevant translational findings included increased nicotine reward and decreased analgesic efficacy of morphine (Bernardi et al., 2016; Mahmoud et al., 2011). Importantly, using these mice, we were able to establish a pharmacogenetic role of this variant by demonstrating increased sensitivity of the G-allele MOR to the effects of NTX and nalmefene treatment on reducing alcohol intake (Bilbao et al., 2015). This is of particular interest among individuals of European or Asian ancestry, where its frequency is 15% to 30% and 40% to 50%, respectively, while the 118 G allele frequency is low (1% to 3%) among individuals of African or Hispanic ancestry (Gelernter et al., 1999; Pang et al., 2009; Tan et al., 2003).

Another important factor involved in the response to NTX treatment could be the amount of craving experienced in response to alcohol-associated stimuli. A recent meta-analysis of human laboratory studies found that, under controlled experimental conditions, NTX relative to placebo reduced the extent of subjective craving and the amount of alcohol consumption (Hendershot et al., 2017). Neural cue reactivity to drug associated stimuli measured by functional magnetic resonance imaging (fMRI) is widely used as a proxy marker for craving, although its specificity for a pathological process has been questioned given the high spatial overlap of the neural response to drug and natural cues demonstrated by meta-analysis (Noori et al., 2016). Nevertheless, it was shown that alcohol cues can induce brain activity in the ventral striatum of both healthy consumers and AUD patients (Schacht et al., 2013a; Vollstadt-Klein et al., 2012; Yalachkov et al., 2012). Patients with high cue reactivity seem to respond favorably to NTX treatment (Mann et al., 2014). Functional MRI studies also identified significant interactions between naltrexone treatment and reduction of mesolimbic cue reactivity in the ventral striatum (Schacht et al., 2013b, 2017) and several prefrontocortical areas (Lukas et al., 2013). Notably, patients with a greater NTX effect on ventral striatal cue-reactivity during early treatment showed better long-term clinical outcomes (Schacht et al., 2017), which seems to be in concordance with the NTX findings on subjective craving responses. Interestingly, the OPRM1 A118G genotype did not significantly moderate the NTX effects observed in the study by Schacht et al., though G-allele carriers who received NTX relapsed faster once the medication was discontinued. Similarly, two other studies found that a higher cue-induced reactivity of the ventral striatum before the start of NTX treatment predicted longer relapse-free periods, while a persistent elevation of cue reactivity in the ventral striatum after two weeks of treatment was associated with an increased relapse risk (Mann et al., 2014; Reinhard et al., 2015). Collectively, changes in neural alcohol cue reactivity during early abstinence could comprise a potential biomarker of treatment response.

Another neurophysiological moderator that seems to be positively associated with NTX treatment response is a preference for sweet tastes. Animal studies demonstrate that the hedonic response to sweet taste reflects in part activity of the MOR system (Leventhal et al., 1995; Pecina and Berridge, 2005) and that the preference for sweet taste and ethanol seems to be genetically correlated (Belknap et al., 1993; Blizard and McClearn, 2000; Carroll et al., 2008). In humans, a sweet-liking phenotype has been shown to be associated with the familial risk for alcohol use disorders (Kampov-Polevoy et al., 2003; Pepino and Mennella, 2007). In a recent clinical trial, the combination of a sweet-liking phenotype and a high level of craving for alcohol were associated with a strong response to NTX (Garbutt et al., 2016).

Section snippets

The state of the reward system in AUD

Taking the above mentioned findings together, NTX treatment seems to be most effective under conditions associated with a high functioning dopaminergic reward system (e.g., a strong response to positive alcohol cues manifested as cravings or high neural reactivity), extreme sweet-liking, or increased reward sensitivity mediated by the 118 G-allele. Obviously, these factors may not act in isolation, but show complex interactions in moderating NTX effects on clinical outcomes.

Clinical studies

The number of human neuroimaging studies by PET and SPECT is relatively low; only those that reported unchallenged baseline values and had a control group with healthy subjects were included. Similarly, postmortem studies had to include an AUD and a non-AUD control group. Since the number of studies and subjects is rather small, we included subjects regardless of whether they were intoxicated or not at the time of death. In most studies, both in vivo and postmortem, the duration of abstinence

Proposed model of opioid-dopamine interaction during abstinence

Opioid-dopamine systems interaction contributes to the rewarding effects of alcohol. Thus, as described in the introduction, in non-dependent individuals, acute alcohol intake results in activation of MORs by endogenous opioids including β-endorphin (Mitchell et al., 2012; Olive et al., 2001). One important mechanism seems to be the presynaptic activation of MORs on GABAergic afferents in the VTA resulting in elevated activity of MSNs and disinhibition of dopaminergic neurons, thereby

Conclusions and outlook

The present survey of the literature is a testimony to the enormous efforts undertaken by the field to elucidate mechanisms of alcohol reward, how these mechanisms change during the course of AUD and how these alterations can best be targeted for treatment. Despite the seemingly variable outcomes of the different experiments, the reviewed data highlight the wide-spread dysregulation of dopamine and endogenous opioid systems during abstinence in alcohol dependence in humans and rats. These

Acknowledgements

This work was supported by the Bundesministerium für Bildung und Forschung (BMBF; FZ: 01ZX1503; e:Med program, FKZ: 01ZX1311A (Spanagel et al., 2013)), Deutsche Forschungsgemeinschaft (DFG Center Grant SFB1134) and the European Union’s Horizon 2020 Program 668863-SyBil-AA. Portions of this manuscript, including some of the figures, were included in the Ph.D. dissertation of co-author Natalie Hirth (Hirth, 2015). The authors thank the Sydney Brain Bank, especially Clive Harper, Donna Sheedy, and

References (340)

  • R. Ciccocioppo et al.

    Buprenorphine reduces alcohol drinking through activation of the nociceptin/orphanin FQ-NOP receptor system

    Biol. Psychiatry

    (2007)
  • G. Di Chiara et al.

    Dopamine and drug addiction: the nucleus accumbens shell connection

    Neuropharmacology

    (2004)
  • M. Dudek et al.

    Modulation of nucleus accumbens connectivity by alcohol drinking and naltrexone in alcohol-preferring rats: a manganese-enhanced magnetic resonance imaging study

    Eur. Neuropsychopharmacol.

    (2016)
  • D. Economidou et al.

    Dysregulation of nociceptin/orphanin FQ activity in the amygdala is linked to excessive alcohol drinking in the rat

    Biol. Psychiatry

    (2008)
  • G. Faedda et al.

    Pharmacology of binding of 3H-SCH-23390 to D-1 dopaminergic receptor sites in rat striatal tissue

    Biochem. Pharmacol.

    (1989)
  • C.R. Gerfen

    The neostriatal mosaic: multiple levels of compartmental organization

    Trends Neurosci.

    (1992)
  • E. Gil et al.

    Effects of chronic treatment with ethanol and withdrawal of ethanol on levels of dopamine, 3,4-dihydroxyphenylacetic acid and homovanillic acid in the striatum of the rat. Influence of benzodiazepines, barbiturate and somatostatin

    Neuropharmacology

    (1992)
  • E. Gili-Martin et al.

    Effects of chronic ethanol treatment and ethanol withdrawal on [3H]SCH23390 binding to rat striatal membranes

    Neuropharmacology

    (1997)
  • A. Gjedde et al.

    Mapping neuroreceptors at work: on the definition and interpretation of binding potentials after 20 years of progress

    Int. Rev. Neurobiol.

    (2005)
  • Y. Goto et al.

    The Yin and Yang of dopamine release: a new perspective

    Neuropharmacology

    (2007)
  • A.A. Grace et al.

    Regulation of firing of dopaminergic neurons and control of goal-directed behaviors

    Trends Neurosci.

    (2007)
  • W.C. Graham et al.

    Autoradiographic studies in animal models of hemi-parkinsonism reveal dopamine D2 but not D1 receptor supersensitivity. I. 6-OHDA lesions of ascending mesencephalic dopaminergic pathways in the rat

    Brain Res.

    (1990)
  • A. Gual et al.

    A randomised, double-blind, placebo-controlled, efficacy study of nalmefene, as-needed use, in patients with alcohol dependence

    Eur. Neuropsychopharmacol.

    (2013)
  • D. Guilloteau et al.

    Exploration of the dopamine transporter: in vitro and in vivo characterization of a high-affinity and high-specificity iodinated tropane derivative (E)-N-(3-iodoprop-2-enyl)-2beta-carbomethoxy-3beta-(4’-m ethylph enyl)nortropane (PE2I)

    Nucl. Med. Biol.

    (1998)
  • A.L. Gundlach et al.

    125I-Spiperone: a novel ligand for D2 dopamine receptors

    Life Sci.

    (1984)
  • H. Hall et al.

    Synthesis and binding properties of the fluorinated substituted benzamide [3H]NCQ 115, a new selective dopamine D2 receptor ligand

    Eur. J. Pharmacol.

    (1991)
  • S. Aalto et al.

    Intravenous ethanol increases dopamine release in the ventral striatum in humans: PET study using bolus-plus-infusion administration of [(11)C]raclopride

    J. Cereb. Blood Flow Metab.

    (2015)
  • A. Abi-Dargham et al.

    Increased prefrontal cortical D(1) receptors in drug naive patients with schizophrenia: a PET study with [(1)(1)C]NNC112

    J. Psychopharmacol. (Oxford)

    (2012)
  • H.C. Akunne et al.

    Studies of the biogenic amine transporters. III. Demonstration of two binding sites for [3H]GBR12935 and [3H]BTCP in rat caudate membranes

    J. Pharmacol. Exp. Ther.

    (1994)
  • D.S. Albrecht et al.

    Effects of smoking on D(2)/D(3) striatal receptor availability in alcoholics and social drinkers

    Brain Imaging Behav.

    (2013)
  • A.C. Andorn et al.

    [3H]Spiroperidol binding in rat striatum: two high-affinity sites of differing selectivities

    J. Neurochem.

    (1980)
  • S. Apaydin et al.

    Effect of arachidonic acid on specific binding of [3H]naloxone to opioid receptors

    Acta Neurobiol. Exp. (Wars)

    (2000)
  • F. Arai et al.

    The presence and role of the dopamine DA-2 receptor in the human decidua

    J. Obstet. Gynaecol. Res.

    (2000)
  • S. Barak et al.

    Glial cell line-derived neurotrophic factor reverses alcohol-induced allostasis of the mesolimbic dopaminergic system: implications for alcohol reward and seeking

    J. Neurosci.

    (2011)
  • G. Bart et al.

    Nalmefene induced elevation in serum prolactin in normal human volunteers: partial kappa opioid agonist activity?

    Neuropsychopharmacology

    (2005)
  • I. Bazov et al.

    Dynorphin and kappa-opioid receptor dysregulation in the dopaminergic reward system of human alcoholics

    Mol. Neurobiol.

    (2018)
  • J.K. Belknap et al.

    Voluntary consumption of ethanol in 15 inbred mouse strains

    Psychopharmacology

    (1993)
  • S. Ben Hamida et al.

    Mu opioid receptors in GABAergic neurons of the forebrain promote alcohol reward and drinking

    Addict. Biol.

    (2017)
  • S. Bender et al.

    The efficacy of the dopamine D2/D3 antagonist tiapride in maintaining abstinence: a randomized, double-blind, placebo-controlled trial in 299 alcohol-dependent patients

    Int. J. Neuropsychopharmacol.

    (2007)
  • R.E. Bernardi et al.

    A gene-by-sex interaction for nicotine reward: evidence from humanized mice and epidemiology

    Transl. Psychiatry

    (2016)
  • A. Beyer et al.

    Effect of the A118G polymorphism on binding affinity, potency and agonist-mediated endocytosis, desensitization, and resensitization of the human mu-opioid receptor

    J. Neurochem.

    (2004)
  • K. Bjork et al.

    Ethanol-induced activation of AKT and DARPP-32 in the mouse striatum mediated by opioid receptors

    Addict. Biol.

    (2010)
  • D.A. Blizard et al.

    Association between ethanol and sucrose intake in the laboratory mouse: exploration via congenic strains and conditioned taste aversion

    Alcohol. Clin. Exp. Res.

    (2000)
  • I. Boileau et al.

    Alcohol promotes dopamine release in the human nucleus accumbens

    Synapse

    (2003)
  • J.W. Boja et al.

    Expression of a single dopamine transporter cDNA can confer two cocaine binding sites

    Neuroreport

    (1992)
  • J.W. Boja et al.

    High-affinity binding of [125I]RTI-55 to dopamine and serotonin transporters in rat brain

    Synapse

    (1992)
  • C. Bond et al.

    Single-nucleotide polymorphism in the human mu opioid receptor gene alters beta-endorphin binding and activity: possible implications for opiate addiction

    Proc. Natl. Acad. Sci. U. S. A.

    (1998)
  • C. Bouza et al.

    Efficacy and safety of naltrexone and acamprosate in the treatment of alcohol dependence: a systematic review

    Addiction

    (2004)
  • I. Brand et al.

    Global ethanol-induced enhancements of monoaminergic neurotransmission: a meta-analysis study

    Alcohol. Clin. Exp. Res.

    (2013)
  • S. Bujarski et al.

    Functional significance of subjective response to alcohol across levels of alcohol exposure

    Addict. Biol.

    (2017)
  • Cited by (0)

    View full text