|Home | About | Journals | Submit | Contact Us | Français|
Acetylcholine (ACh), the first neurotransmitter discovered, participates in many CNS functions including sensory and motor processing, sleep, nociception, mood, stress response, attention, arousal, memory, motivation and reward. These diverse cholinergic effects are mediated by nicotinic (nAChR) and muscarinic (mAChR) type cholinergic receptors. The goal of this review is to synthesize a growing literature that supports the potential role of ACh as a treatment target for stimulant addiction. ACh interacts with the dopaminergic reward system in the ventral tegmental area (VTA), nucleus accumbens (NAc) and prefrontal cortex (PFC). In the VTA, both nAChR and mAChR stimulate the dopaminergic system. In the NAc, cholinergic interneurons integrate cortical and subcortical information related to reward. In the PFC, the cholinergic system contributes to the cognitive aspects of addiction. Preclinical studies support a facilitative role of nicotinic agonists in the development of stimulant addiction. Muscarinic agonists seem to have an inhibitory role depending on the subtype of mAChR. In human studies acetylcholine esterase (AChE) inhibitors, which increase synaptic ACh levels, have shown promise for the treatment of stimulant addiction. Further studies testing the efficacy of cholinergic medications for stimulant addiction are warranted.
Acetylcholine (ACh) is the founding member of the class of biochemicals known as neurotransmitters. Expressed in both the peripheral (PNS) and central nervous systems (CNS), ACh is essential to diverse bodily functions and processes. In the PNS, cholinergic transmission is essential in muscle contraction and autonomic nervous system functions,[2, 3] while in the CNS, it plays a number of important roles, including sensory and motor processing, sleep, nociception, mood, stress response, attention, arousal, memory, motivation and reward.[4-6] A large body of evidence also supports the role of ACh in initiation and maintenance of addictive processes. The goal of this paper is to synthesize the growing literature that supports a prominent part of ACh in both the initiation and maintenance of addiction to stimulant drugs of abuse, specifically cocaine and amphetamines, as well as the potential role of ACh as a treatment target for stimulant addiction. Our review complements an excellent paper by Williams and Adinoff  which provided a comprehensive review of the cholinergic system and cocaine addiction, with a focus on neural circuits and neurobiology. In this paper, first we briefly overview the cholinergic system [for more extensive reviews see, 3, 5-7, 8, 9, 10]. Next, we review the nascent clinical literature that shows the emerging significance of the cholinergic system as a treatment target in stimulant addiction.
In this section, we review ACh biosynthesis and metabolism, receptor types, and distribution in the CNS.
The neurotransmitter ACh is synthesized from acetyl-CoA and choline by choline acetyltransferase (ChAT), an enzyme primarily expressed by cholinergic neurons . ChAT is a specific marker for identifying cholinergic neurons in the central and peripheral nervous systems. Following its release to the synaptic cleft, ACh is hydrolyzed by the enzyme acetylcholinesterase (AChE). This enzyme is the target of a group of medications called AChE inhibitors, which increase levels of acetylcholine available at both nicotinic (nAChR) and muscarinic (mAChR) type cholinergic receptors.
While the cholinergic projection neurons are located in the basal forebrain and the brainstem, cholinergic interneurons are located mainly in the striatum and nucleus accumbens (NAc). Basal forebrain cholinergic neurons are located in the medial septal nucleus, the vertical and horizontal limb nuclei of Broca, and the nucleus basalis of Meynert. These neurons project mainly to the cerebral cortex, hippocampus and amygdala. The brainstem cholinergic neurons are located in pedunculopontine tegmental and the laterodorsal tegmental nuclei and project to the brainstem and midbrain including the ventral tegmental area (VTA), where dopaminergic neurons are found.
In the synaptic cleft, ACh interacts with nAChR or mAChR, named based on their differential sensitivity to the exogenous ligands, respectively muscarine and nicotine. These receptors differ both in their structure and function.
mAChR are G-protein coupled, can be either excitatory or inhibitory, and are relatively slower acting compared to nAChR.[3, 13] mAChR represent approximately 90 percent of the cholinergic receptors. There are 5 different muscarinic receptors that are divided in two groups based on their G-protein coupling mechanisms: M1, M3 and M5 vs. M2 and M4.[2, 14] The M1, M3 and M5 receptors are coupled to the Gq/G11 family, whereas the M2 and M4 receptors prefer Gi/Go type of G-proteins. The M1, M3, and M5 mAChR are found post-synaptically and they activate several effectors including phospholipase C, intracellular calcium, inositol triphosphatase, and MAPK (mitogen activated protein kinase) . The M2 and M4 mAChR are found both pre- and post-synaptically and inhibit adenylyl cyclase and voltage-operated calcium channels and activate MAP kinases and G-protein-activated inwardly rectifying potassium channels (GIRK) channels . These signaling cascades are essential for many neuronal functions including neuronal differentiation, survival, and synaptic plasticity, which also play an important role in addiction.[16-18] The M1, M4 and M5 subtype mAChR are predominantly expressed in the CNS, whereas the M2 and M3 mAChR subtypes are found both in the PNS and CNS.
The M1 mAChR is widely expressed in the brain including the cerebral cortex, hippocampus and striatum. The M1 receptors, the predominant mAChR in the CNS, are implicated in learning and memory processes. M1 agonists are potential treatment agents for Alzheimer's disease and other cognitive disorders. The M3 receptors are less densely distributed in the CNS and its function is not well-characterized. The M5 receptors are expressed on the dopaminergic neurons in the substantia nigra and the VTA (ventral tegmental area). Stimulation of M5 receptors facilitates dopamine (DA) release in the NAc. The M5 type mAchR are potential treatment targets for stimulant addiction, Parkinson's disease and schizophrenia.
Pre-synaptically located M2 and M4 mAChR regulate release of acetylcholine and other neurotransmitters including DA [6, 22]. The M2 receptors are widely expressed in the CNS, including the brainstem, thalamus, cortex, hippocampus and striatum, where they control ACh release. Blockage of M2 receptors, by inhibiting autoreceptor functioning, increases cholinergic activity. The M2 receptors are potential treatment targets for Alzheimer's disease and pain. The M4 mAChR are found in the midbrain, cortex, hippocampus, and striatum. The stimulation of M4 mAChR is thought to decrease the activity of dopaminergic VTA neurons, leading to reduced DA release in the NAc. The M4 receptor is potential treatment targets for schizophrenia, Parkinson's disease and possibly drug addiction.
nAChR are ligand-gated ion channels, permeable to sodium, potassium, and calcium ions. These receptors show relatively fast responses, within milliseconds, and are excitatory.[8, 13] Most nAChR in the CNS are presynaptically located and they modulate the release of acetylcholine, DA, serotonin, glutamate, GABA, and norepinephrine. nAChR can also be postsynaptically located as those on the dopaminergic neurons in the VTA. nAChR are pentameric combinations of 12 subunits (α2-α10 and β2-β4). Most nicotinic receptors contain α4, β2 or α7 subunits. β2-containing receptors are critical for the addictive as well as cognitive performance-enhancing properties of nicotine.[24, 25] Nicotine withdrawal has been shown to reduce brain reward function in rats, which may also be mediated by α4, β2 or α7 subunits 
ACh acts on mAChR and nAChR. mAChR are G-protein coupled and are slower acting compared to nAChR. nAChR are ligand-gated ion channels and mediate excitatory, fast responses. Cholinergic neurons are either projection neurons or interneurons. While the cholinergic projection neurons are located in the basal forebrain and the brainstem, cholinergic interneurons are located mainly in the striatum and NAc.
Next, we review the distribution of ACh neurons and receptors in the CNS, with emphasis on cholinergic modulation of essential brain reward circuitry.
The main components of the brain reward circuit are thought to include the mesocorticolimbic DA system originating from the VTA of the midbrain and targeting a number of limbic and cortical structures, including the NAc, amygdala, and PFC. DA release, especially in the NAc, is a critical event mediating the rewarding effects of stimulant drugs. Stimulants enhance dopaminergic activity by either facilitating release (e.g., amphetamines) and/or blocking uptake (e.g., cocaine) of DA. DA release in the NAc is modulated by GABA, glutamate and ACh. Glutamate, the main excitatory neurotransmitter in the brain, plays an important role in learning and memory. GABA is the main inhibitory neuron in the brain and the output neurons in the NAc (medium spiny neurons) are GABAergic.
Within the brain reward circuit, the cholinergic and mesolimbic dopaminergic systems interact closely in the VTA, NAc and PFC.
Cholinergic projections to VTA dopaminergic neurons derive from brainstem cholinergic neurons.[31, 32] In the VTA, β2 containing nAChR increase the activity of dopaminergic neurons, which leads to increased DA release in terminal regions including NAc and PFC. In addition, activation of α7-nicotinic receptors on glutamate terminals arising from PFC increase glutamate release in the VTA, which excites DA neurons in the VTA to release DA in the NAc. The resultant DA release in the NAc is critical for the rewarding and addictive effects of nicotine. The M5 mAChR are found on the dopaminergic neurons and similar to nAChR, their activation increases dopaminergic activity. Thus, both types of cholinergic receptors, mAChR and nAChR, stimulate dopaminergic neurons in the VTA.
DA release in the NAc is a critical event mediating the rewarding effects of stimulant drugs. In the NAc, ACh is released from the cholinergic interneurons, which have extensive dendritic and axonal connections with other cells in the in NAc.[30, 34] An early hypothesis posited that the cholinergic and dopaminergic neurotransmitters act antagonistically in the NAc, however, more recent evidence supports a more complementary interaction between these neurotransmitters. While D1 dopaminergic receptor activation increases ACh release in the NAc, D2 dopaminergic activation reduces ACh release. The activity of medium spiny neurons (the main output neurons of the NAc) is also modulated by both cholinergic and dopaminergic input, which may be excitatory or inhibitory depending on the receptor subtypes that are stimulated: the D1 dopaminergic and M1 mAChR are excitatory while D2 dopaminergic and M4 mAChR are inhibitory. Cholinergic interneurons receive dopaminergic input from the VTA, and glutamatergic input mainly from the PFC, hippocampus, and amygdala. This DA and glutamate convergence on cholinergic interneurons may provide a mechanism for DA-mediated reward to be associated with glutamate-mediated learning and contextual information. Accordingly, cholinergic interneurons control the translation of reward signals into contextually appropriate behavior.
The functions attributed to the PFC include attention, working memory, response inhibition, affective processing, decision making and goal-directed behaviors. Compulsive drug use is the hallmark of drug addiction; it is characterized by behavioral inflexibility and more specifically, decreased ability to disengage from the processing of drug-related stimuli , both functions that are thought to be governed by the PFC. Many studies have demonstrated that chronic stimulant use is associated with a range of PFC dysfunction including poor sustained attention, response inhibition, and decision making.[40-45] In neuroimaging studies, stimulant users show decreased PFC activity and this decrease is associated with craving for cocaine in cocaine-dependent individuals. In line with these findings, many preclinical studies, using animal models of relapse, have demonstrated that the PFC controls stress- and cocaine-induced reinstatement of cocaine self-administration.[47-49] The contribution of the cholinergic system in these processes is not well understood.
Amphetamines and cocaine increase ACh release in the PFC.[50, 51] Increased ACh release in the PFC accompanies development of psychomotor sensitization to amphetamines, suggesting participation of the prefrontal cholinergic system in neuronal adaptations leading to stimulant addiction. The contribution of nAChR and mAChR in these processes is not well characterized. In one study, local application of the muscarinic antagonist, scopolamine, to the PFC, enhanced response to intravenous cocaine self-administration in rats, supporting the role of mAChR in stimulant reinforcement. Prefrontal cortical nicotinic receptors are thought to be involved in the cognitive enhancing properties of nicotine, although the exact mechanisms remain to be elucidated.
To summarize, the cholinergic system interacts with the dopaminergic reward system at three levels: VTA, NAc, and PFC. In the VTA, both nAChR and mAChR stimulate the dopaminergic system. In the NAc, cholinergic interneurons integrate the cortical and subcortical information related to reward. In the PFC, the cholinergic system contributes to the cognitive control of addictive processes, although the neurobiological mechanisms remain to be elucidated.
The cholinergic system has been most extensively examined in connection to dementia and other cognitive disorders. However, preclinical and human studies examining the possible role of the cholinergic system in stimulant addiction have begun accumulate. These studies will be summarized below.
Many preclinical studies have also demonstrated that ACh modulates the rewarding effects of stimulants. It has been demonstrated that ablation of cholinergic neurons in the NAc through administration of a cholinergic immune toxin results in mice showing greater sensitivity and preference to cocaine. In contrast, enhancement of cholinergic transmission by treatment with the AChE inhibitor, physostigmine, decreased cocaine self-administration in Rhesus monkeys. Another AChE inhibitor, galantamine, which also allosterically potentiates nAChR, inhibited cocaine-induced hyperlocomotion in mice. Galantamine also inhibited the dextroamphetamine-induced unrest, arousal, and stereotypy in Cebus monkeys. Similarly, donepezil, another AChE inhibitor, reduced locomotor sensitivity and preference to cocaine in mice. In rats, donepezil, as well as nicotine, treatment also reduced methamphetamine-seeking behavior. Donepezil's effects on methamphetamine-seeking behavior were blocked by systemic mecamylamine, an nAChR antagonist, but not by a mAChR antagonist, scopolamine, suggesting that the inhibitory effects of donepezil were mediated through nAChR.
Further pharmacological studies using drugs selective for mAChR or nAChR support the conclusion that both receptor types participate in stimulant effects. In rats, treatment with repeated nicotine administration enhanced cocaine self-administration. Interestingly, the nAChR antagonist, mecamylamine, reduced, but did not prevent escalation of cocaine self-administration by rats. These findings suggest that nAChR play an important role in the transition to cocaine addiction.
In other studies, mAChR medications have been examined in various models of drug addiction. The mAChR agonist, oxotremorine, has been shown to reduce amphetamine-induced locomotor activity, DA release in NAc, and cocaine self-administration in rats,[50, 62, 63] and mice. On the other hand, atropine, an mAChR antagonist, reduced cocaine self-administration in Rhesus monkeys trained to self-administer cocaine. Similarly, scopolamine, another mAChR antagonist, enhanced cocaine induced locomotor activity, and reduced cocaine self-administration in Rhesus monkeys.
The individual role of mAChR subtypes in stimulant addiction was examined using selective gene targeting approach. Mice lacking M1 receptors showed elevated DA levels in the striatum and enhanced locomotor activity in response to dextroamphetamine. In another study, compared to wild type, M1-deficient mice showed a reduced conditioned place preference for cocaine or morphine. Similar findings were obtained with the M1 antagonist pirenzepine, further supporting the role of M1 receptors in stimulant addiction.
M4 deficient mice displayed elevated DA basal values and enhanced DA response to dextroamphetamine in the NAc. Consistent with these findings, in another study, two M4 positive allosteric modulators have been shown to reverse amphetamine-induced locomotion. These findings support the possibility that M4 mAChR agonists may also be potential treatment agents for stimulant addiction.
The M5 deficient mice, compared to the wild-type, showed reduced cocaine self-administration and withdrawal-induced anxiety. These findings suggest that M5 receptors might be a potential target for stimulant addiction. However, no ligands selective for M5 mAChR are available yet.
There have been few human studies examining cholinergic medications as potential treatments for stimulant addiction. Janowsky et al. first examined whether AChE inhibitors modulate amphetamine response several decades ago. In that study, physostigmine, an AChE inhibitor, attenuated the subjective effects of a stimulant medication, methylphenidate, in bipolar and schizophrenic patients. These findings were extended in more recent studies. De La Garza et al. examined the effects of rivastigmine (1.5 or 3 mg/day), an AChE inhibitor, on intravenous methamphetamine responses (30 mg/day) in 23 methamphetamine dependent humans. In that study, 3 mg of rivastigmine treatment for 11 days attenuated the methamphetamine-induced diastolic blood pressure increases as well as subjective ratings of “desire” and “anxious”. The same study also examined methamphetamine self-administration behavior, which was not affected by rivastigmine treatment. These findings are promising, and warrant further studies examining AChE inhibitors as potential treatments for stimulant addiction.
We are aware of one clinical trial in which a AChE inhibitor was evaluated for cocaine addiction. In that 10-week study, donepezil at 10 mg was well tolerated but did not change cocaine use behavior. The study had a small sample size (only 17 subjects assigned to donepezil) providing limited statistical power to test the efficacy of donepezil for cocaine addiction. Only one dose of donepezil dose was evaluated. However, those treated with donepezil did show significant reductions in craving and other indexes of severity of addiction to cocaine and other drugs. Other AChE inhibitors have not been evaluated as potential treatment agents for stimulant addiction.
Two human studies have evaluated whether nAChR agonists can change typical subjective and physiological responses to stimulants. In one study, 7 cigarettes smokers, who occasionally used cocaine, were treated with 14 mg nicotine or placebo patch followed by intranasal cocaine (0.9 mg/kg or placebo). Nicotine patch, compared to placebo, attenuated cocaine-induced “high” and “stimulated” and increased the latency to detect cocaine effects without affecting physiological responses or pharmacokinetics of cocaine. However, in another study, the 21 mg nicotine patch, compared to placebo, did not affect the subjective or reinforcing effects of intravenous cocaine (15 or 30 mg/70 kg). These inconsistent findings could be due to differences in subject population (dependent vs. occasional cocaine users) as well as route of cocaine administration (intranasal vs. intravenous). Since both studies used a single dose of active nicotine patch (14 or 21 mg), dose-dependent effects of nicotine on cocaine responses were not examined. Future studies addressing these limitations will provide a clearer picture on how nAChR agonists affect typical stimulant effects in humans.
The nAChR antagonist, mecamylamine, has also been examined for cocaine addiction. A single 2.5 mg mecamylamine treatment reduced cue-induced cocaine craving in 23 cocaine users. A randomized, double-blind, placebo-controlled study examined the efficacy of mecamylamine (6 mg/day) or placebo transdermal patches for a 16-week as a treatment of cocaine addiction. In that study, mecamylamine did not reduce cocaine use or craving for cocaine. However, the small sample size, a total of 35 subjects, offered very limited statistical power. Further, only one dose of mecamylamine was tested, and only the transdermal route of administration was used.
In a human laboratory study, Penetar and colleagues examined the effects of the conjoint administration of cocaine with a mAChR antagonist, benztropine, in occasional cocaine users. Benztropine, an antimuscarinic and antihistamine medication, is commonly used to alleviate the extrapyramidal side effects (stiffness) from typical antipsychotics. Benztropine at 1, 2 or 4 mg, given as a single acute dose did not affect subjective or physiological effects from 0.9 mg/kg intranasally administered cocaine. Given that benztropine has significant antihistamine effects, it remains to be determined whether selective mAChR antagonists, like scopolamine or atropine, would alter cocaine's effects in humans.
To summarize, preclinical studies support a facilitative role of nAChR in the development of stimulant addiction. Among mAChR, M1, M4 and M5 subtypes seem to be potential targets for stimulant addiction. As summarized in Table 1, there have been only a few studies to evaluate cholinergic medications for treatment of stimulant addiction
Based on the preclinical and clinical studies summarized above, the cholinergic medications that have potential for the treatment of stimulant addiction include AChE inhibitors, as well as medications targeting nAChR and mAChR.
As mentioned above, AChE inhibitors increase the synaptic concentrations of ACh, which leads to increased stimulation of both nAChR and mAChR. Several AChE inhibitors, including tacrine, rivastigmine, donepezil, and galantamine are available for clinical use for the treatment of dementia.[81-83] Other disorders characterized with cognitive impairment, including Parkinson's disease, traumatic brain injury and schizophrenia are also potential treatment targets [84-86] for AChE inhibitors. AChE inhibitors differ in their pharmacological as well as side effect profiles. The most common adverse effects of AChE inhibitors include diarrhea, nausea, vomiting, loss of appetite and dizziness. Tacrine, the first AChE inhibitor that received FDA approval in the US, has limited use due to hepatotoxicity and short half-life. Donepezil and rivastigmine are more potent AChE inhibitors than galantamine. Galantamine binds to nicotinic receptors, especially α7 and α4β2 subtypes, and enhances responses to acetylcholine, effectively resulting in DA release in the mesocorticolimbic pathway. To date, only one study has evaluated galantamine as treatment for addiction. In a sample of 114 alcohol-dependent individuals, galantamine was associated with significant reductions in smoking compared with placebo. Interestingly, the galantamine effect occurred even though smoking was not specifically targeted. In addition, given the promising human laboratory findings with physostigmine and rivastigmine in combination with amphetamines, [72, 73] further studies evaluating the efficacy of different AChE inhibitors for stimulant addiction are warranted.
Another potential treatment target for AChE inhibitors is cognitive impairment associated with long-term stimulant use. Cocaine and methamphetamine users have well-documented impairment in attention, working memory, decision-making and response inhibition. [40-44, 45]. Studies also suggest that cognitive impairments seen in stimulant users were not reversible with short-term abstinence. Importantly, these cognitive impairments seems to predict poor treatment outcome in stimulant users who are in treatment.[90-92] Whether AChE inhibitors improve cognitive performance in stimulant users has not been examined. Thus, AChE inhibitors, by improving cognitive impairments found in stimulant users, may prove to be useful in treating stimulant addicted individuals.
An important development in nicotinic pharmacotherapies is the availability of partial nAChR agonists for clinical use. Cytisine, a plant alkaloid has been used for smoking cessation in Eastern Europe, is a partial α4β2 nAChR agonist. A synthetic variation of cytisine is varenicline, a partial agonist at the α4β2 nAChR, has recently been marketed for smoking cessation. Varenicline inhibits nicotine's ability to activate nAChR and release DA and may not cause rapid desensitization at the nAChR, which limits the therapeutic utility of nAChR agonists. Several other partial nAChR agonists, including dianicline and ispronicline, are undergoing human studies for smoking cessation and treatment of dementia. Partial nAChR agonists may also have value for stimulant addiction pharmacotherapy, given the role of nicotinic receptors in stimulant effects. Varenicline and other partial nAChR agonists remain to be evaluated for stimulant addiction. The most common adverse effects of varenicline include nausea, vomiting, difficulty sleeping, headache and vivid dreams. Less common but more serious adverse effects include abnormal behavior and suicidal thoughts .
Another plant alkaloid, lobeline, has antagonistic effects on the α3β2 and α4β2 nAChR but has nicotinic agonistic effects as well. Lobeline has been evaluated for smoking cessation, with no evidence of efficacy observed. Lobeline also increases DA release from dopaminergic synapses, and blocks the DA transporter, similar to stimulants. In preclinical studies, lobeline blocks the behavioral and neurochemical effects of stimulants. Based on these preclinical findings, lobeline has been suggested to be a potential treatment for stimulant addiction. Lobeline is currently being tested in humans as a treatment of methamphetamine addiction.
Development of medications targeting the M1 mAChR has been an important goal, especially for the treatment of Alzheimer's disease and other cognitive disorders. Muscarinic agonists that have reached human clinical studies included arecoline, carbachol, cevimeline, xanomeline, milameline and sabcomeline. Unfortunately, none of these medications moved into clinical practice for the treatment of Alzheimers' disease or other cognitive disorders primarily due to their affinity to other mAChR subtypes, especially to peripheral M2 and M3 mAChR, which mediates adverse effects like diarrhea and nausea which limits their clinical use.
The difficulty in developing M1 selective compounds has been attributed to high sequence similarity across subtypes, especially in their ACh binding sites. An alternative strategy has been to develop compounds that selectively bind to their target mAChR subtypes through allosteric mechanisms. Compounds selective for M1, M2 and M4 mAChR, with more favorable side effect profiles than nonselective mAChR have been developed.[102-105]. The availability of such compounds will also have greatly facilitate the development of novel treatments for stimulant addiction.
As summarized above, preclinical and clinical studies support the potential utility of cholinergic system as a target to develop medications for stimulant addiction. Further studies are warranted to test the efficacy of AChE inhibitors as well as nicotinic and muscarinic medications of the treatment of stimulant addiction.
The authors were supported by career development awards, K02-DA021304 (MS) and K01-DA-019446 (MM), by RO1DA019885, RO1DA020752, RO1DA014537 and the Veterans Administration Mental Illness Research, Education and Clinical Center (MIRECC).