Orexin receptor blockers: A tool for lowering alcohol intake and alcohol addictive behavior in the light of preclinical studies

At the moment, in the pharmacological treatment of alcoholism there are only a few clinically effective synthetic drugs, i.e., disulfiram, acamprosate, naltrexone, and recently, nalmefene. These drugs are known to reduce the craving and a relapse into alcohol of dependent patients [1]. Since alcoholism is one of the most prevalent neuropsychiatric diseases, it has enormous health and socioeconomic impact [2]. Therefore, the studies on the introduction of new drugs with anti-alcoholic properties seem to be a high priority.

Alcohol addiction, according to the current state of knowledge, is closely related to the brain reward and punishment system—a complicated network of neurons that convey stimulation to specific brain structures, constituting the anatomical basis of the neural mechanisms underlying reward-seeking and motivation [3, 4, 5]. It is known that the mesocorticolimbic system, containing, among other things, such substructures as the ventral tegmental area (VTA) and nucleus accumbens (NAc), is essential for dopamine neurotransmission [6, 7]. The evidence confirming the role of dopamine in ethanol reward shows that it increases the dopamine release in NAc [5, 8] and in the prefrontal cortex [9]. Acting on dopaminergic systems, these drugs, both excitatory and inhibitory, influence alcohol consumption [10, 11, 12, 13].

The dopaminergic neurons propagating neurotransmission, through the above-mentioned and other substructures of the mesolimbic system and their dopaminergic receptors, are not the only crucial components contributing to the complexity of cognitive and behavioral reactions leading to ethanol reward-seeking. The participation of other neurotransmission systems, affecting the mesolimbic neurons and modulating the action of dopamine, is also associated with the AUD-related processes. For example, a release of endogenous opioids from NAc is linked to an increase in alcohol consumption [14]. As it is currently proposed, rewarding effects of alcohol are the result of a number of coexisting brain mechanisms, which has been already described in more detail in other publications [3, 13, 15, 16, 17]. Thus, a significant amount of data indicate the functional connection of the mesolimbic system with glutaminergic, GABA, acetylcholine, and serotonin neuronal networks, which act in VTA or NAc modulating compulsive alcohol or drug intake [13, 16].

Over the years, a growing number of studies on the role of orexins (OXs), also known as hypocretins, have been conducted, particularly in a brain reward system [18, 19, 20]. One of the potential targets of alcohol addiction therapy may be the effect on the receptor level of these neuropeptides [18, 20]. Although it is known that OXs are able to trigger responses in the central nervous system (CNS) as well as in many other tissues, the responses are often different, surprising, and unexpected [21]. A number of experimental studies and functional analyses of their receptors (OXRs) have been documented, including the use of more or less selective ligands [18]. However, the exact role of OXs in the etiopathogenesis of alcohol dependence still remains not fully understood.

The aim of this mini-review is to systematize the current state of knowledge in the field of usage of OXR antagonists in preclinical behavioral studies, along with the description of major experimental in vivo models and the xenobiotics focused on a potential AUD pharmacotherapy.

OXs are neuropeptides synthesized entirely by neurons of the lateral hypothalamic area (LH)—the lateral, perifornical, and dorsomedial hypothalamus, as well as in regions of the subthalamus including the zona incerta and subthalamic nuclei [22, 23].

There are two OX proteins, Orexin A (OX-A) and B (OX-B), originally formed as a result of cleavage of prepro-orexin during their maturation [22, 23]. The amino acid sequence is similar to the gut hormone secretin, and it is strongly conservative among a number of animal species (rat, mouse, dog, cow, sheep, pig) and humans [18]. Interestingly, this group of proteins has not been found in invertebrates [20, 24]. OX-A consists of 33 amino acids (3562 Da), it contains a N-terminal pyroglutamyl residue, two intramolecular disulfide bridges between Cys6–Cys12 and Cys7–Cys14, and also a C-terminal amidation [22, 23]. OX-B (2937 Da) consists of 28 amino acids that build a linear peptide with a C-terminal amidation and probably forms two alpha helices. The human OX-B differs from the pig and dog orexine by one amino acid and by two amino acids in rats and mice. This protein is 46% amino acid-compatible with OX-A. The C-terminus determines mostly the similarity in amino acid sequence, whereas the N-terminal end is more variable in that matter [24]. Although OXs are formed only in the hypothalamus [23], OXs system fibers form connections with the above mentioned VTA and NAc regions of the brain mesolimbic system [18, 20, 24]. However, no significant affinity to neuropeptide Y, secretin, or other peptides was observed, despite their convergence in amino acid composition [19, 23].

These proteins act via their two appropriate G-protein-coupled receptors: orexin 1 and 2 (OX1R and OX2R, respectively) [23], distributed throughout the central nervous system [20, 22, 23]. OX-A has an equal affinity for OX1R and OX2R, while OX-B preferentially binds to OX2R [23, 25]. In other words, OX1R is considered to be more selective for OX-A, requiring half of the maximally effective concentration (EC₅₀) 30 nM of OX-A but EC₅₀ 2500 nM of OX-B, whereas OX2R binds both peptides equally [25], requiring a concentration of EC₅₀ 34 nM of OX-A and EC₅₀ 60 nM of OX-B for a half-maximum response [23].

The receptor peptides are detected quite widely throughout the brain. Their projections are most highly concentrated in different limbic regions involved in arousal and emotion [18]. OX1R is more densely located in the brain regions involved in motivation: the medial prefrontal cortex (mPFC) and amygdala, whereas OX2R is located in areas associated with regulation of sleep and arousal, such as the hypothalamic and brainstem nuclei [25]. The expression of OX1R is also observed in several other regions, for example, the infralimbic cortex, hippocampus, paraventricular thalamic nucleus, ventromedial hypothalamic nucleus, dorsal raphe nucleus, and locus coeruleus. In turn, OX2R transcripts are detected in the cerebral cortex, hippocampus, septal nuclei, raphe nuclei, dorsomedial nucleus, paraventricular nucleus, medial thalamic groups, and ventral premammillary nucleus [26].

The two receptors interact with the Gq, Gs, and Gi proteins [27] and are identified as located pre- and postsynaptically [18, 28]. For instance, there are a couple of regions where neurons show direct postsynaptic responses to applied OXs; however, with their transcript levels no higher than background levels (e.g., the substantia nigra pars reticulata, medial dorsal nucleus of the thalamus), suggesting that functional receptors may be present in neurons having low mRNA levels or in regions with low densities of expressing neurons. This implies that functionally important groups of orexin-responding neurons have not yet been discovered [21]. The main activity of OXs in CNS cells is described as capable of producing a slow and long-lasting depolarization that can be large enough to initiate firing, or—if the neuron is already firing— to increase its firing rate. OXs are therefore considered as excitatory neuropeptides because their receptors coupled to effectors produce depolarization. In addition to participation in mediating of neurons membrane depolarization, it is believed that native OXRs alter the integrative properties of their target neurons by modulating postsynaptic ion channels and thus strongly influence how synaptic inputs from other sources get converted into firing [21]. Functional experimental studies have shown the presence of OXRs at or near presynaptic terminals of some glutamate- and GABA-releasing neurons. This ability to release glutamate and/or GABA by increasing miniature postsynaptic current frequency has been observed in CNS [29], although the underlying signaling mechanisms have not been investigated. In some structures, OXs preferentially release GABA or glutamate [29], typically in conjunction with postsynaptic actions. Presumably, this allows OXs to amplify or attenuate information transmission arising from particular afferents [21]. These relationships may be the reason for finding a reduction of OX-A levels in patients addicted to ketamine, a drug whose pharmacological profile involves blocking NMDA receptors in the glutamatergic system [30].

Looking at the basic mechanisms of intracellular signal transduction and the differences between the above discussed, it should be mentioned that OX1R is thought to couple to Gq, while OX2R is considered to signal through Gq or Gi/Go. However, the coupling mechanisms seem to be cell type specific and have not been fully described in neurons [24, 31].

Apart from their main occurrence in brain structures involved in food and/or drug-seeking motivation—especially in the limbic part of the reward system—there is a growing amount of evidence that they are also involved in the regulation of sleep and emotional states [18, 20, 24]. At the beginning of studies on the OX system, most of the research back then focused on determining its impact on the regulation of food-related behaviors. OX neurons activities occur during long-term fast, restricted nutrition and resulting hypoglycemia. The activation of these neurons is reduced after food intake. OXR expression is elevated after 48 to 72 hours without food [23, 32, 33]. With a limited access to nutrition, OX neurons are more active just prior to planned food availability, and the OX gene knock-out mice show reduced awakening and motor activity during the food waiting period [34].

It has become clear that these neuropeptides are involved in many other behaviors, many of which are used to promote survival. Apart from the above-mentioned feeding-related activities, the stimulation of the OX system also affects other aspects of behavior like stress [35] and arousal [36].

In the behavioral experiments to explain the role of OXs and their receptors in the course of alcohol addiction, there are three dominant models of alcohol consumption by rodents:

In the light of the current state of knowledge, the proportion of P-strain rats (who prefer drinking alcohol) prevails in the studies involving OXRs antagonists. They are considered more relevant to the studies on ethanol drinking behavior due to the fact that these animals have been selectively bred in accordance with the procedures aimed at strengthening the ethanol effects [39]. Thus, they are less sensitive to the calming and aversive effects of ethanol and more sensitive to its stimulating effects, they have a genetic predisposition to alcohol abuse and alcohol dependence, and they can be used in many models of alcohol preference [40]. For that purpose, not only male P-strain rats but females also have been studied, as in the study by Anderson [4], in which a home cage with a two-bottle choice and the progressive operant responding (lever presses) were applied. Alcohol consumption and preference seemed to be higher in females than in males, in particular in the first stages of this process, which, over time, might be equalized at a similar level in both sexes [41].

Some papers highlight the differences in the amount of alcohol consumed or the effects of alcohol using different strains of rats, for example: Long-Evans [42, 43], Wistar [44, 45] and Sprague-Dawley [46, 47].

In addition to the above-mentioned rats, especially P-strain, male C57BL mice are also used for the OXR antagonist investigations in the studies on alcoholism [4, 48, 49, 50, 51]. Another mouse strain—DBA—is used with much less frequency due to the fact that their daily ethanol consumption capacity is small [45, 52]. For example, the C57BL mice can consume 1.68 g/kg of ethanol, while DBA ones—only 0.25 g/ kg [53].

Due to the fact that animals principally do not consume large amounts of alcohol willingly, the administration of alcohol is very often forced [38, 54]. Therefore, more frequent use of animals with a genetically higher ethanol consumption feature has been postulated, so there will be a greater chance of success in the study.

SB-334867—a selective OX1R antagonist—was the most effective in three standard models of alcohol drinking in mice [4, 48, 49, 50, 51, 53, 55] and in rats [4, 43, 44, 46, 47, 53, 56, 57, 58]. The lowest effective dose after the i.p. administration was 3 mg/kg in C57BL/6 mice [49] and 5 mg/kg in rats [48, 51, 57]. The alcohol-conditioned place preference was disturbed after SB-334867 treatment [52]. Moreover, the compound induced the reduction of alcohol drinking relapse [59]. Another selective OX1R antagonist—GSK1059865— also significantly reduced the alcohol consumption in C57BL/6J mice [60]. Besides, OXR antagonists were directly administered to the VTA and NAc substructures. For example, the cannulation of SB-334867 into VTA [55, 61], CeA (central nucleus of the amygdala) [55], NAc core [44], mNAsh (medial Nucleus Accumbens Shell), and mPFC [50] reduced the alcohol consumption in all behavioral paradigms. Almorexant, a nonselective blocker of both types of OXRs, reduced the consumption of ethanol but also the sugar intake in all three models, both in mice and rats, in the dose range of 15–100 mg/kg, i.p., [4, 62]. Its infusion into VTA significantly reduced the animals response to the operant progressive responding ratio [62].

Apart from the above-mentioned studies, there is some evidence indicating the modulation of the reward system by OXs in the increase of extracellular dopamine [63] or the reduction of cocaine-induced dopamine levels in NAc [64] due to the intra-VTA OX1R antagonist administration.

Compared with studies on OX1R antagonists, the number of studies in the context of assessing the use of OX2R antagonists in the possibility of their influence on ethanol drinking is relatively poor. For example, a selective OX2R antagonist—TCS-OX2-29—nonspecifically reduced alcohol consumption after the cannulation into a PVT [42] and NAc core [56]. In another study, TCS-OX2-29 was not effective [50]. Another selective OX2R antagonist—LSN2424100— administered i.p. decreased the self-administration of ethanol, but selectively—only in the binge-like drinking paradigm [4]. JNJ10397049 attenuated the ethanol self-administration, reinstatement, and place preference in mice, but at a lower degree than the positive control, naltrexone [45]. Apart from Shoblock et al. [45], no one has done any appropriate research to determine the effect of OX2R blockers on dopamine circulation or the expression of dopaminergic receptors in the limbic system. According to these authors, the assessment of the JNJ10397049 activity showed that the ethanol-induced increase in dopamine level in the NAc was not changed after the OX2R blockade, which was a surprising result [45].

The table summarizes the state of behavioral preclinical experiments using antagonists of OXRs along with a brief description of experimental details [Table 1].

There are many factors to be considered when targeting the OX system as a treatment against alcohol motivation. It is postulated that changes in the OXRs’ activity in specific brain structures, associated directly with pathologically enhanced motivation, drive compulsive alcohol consumption. Modulation of activity of the reward system, which is closely related to addiction, may play a major role in inhibiting alcohol consumption with the administration of OXR antagonists, but the accurate role of OXs in AUD still remains not fully understood [18, 20]. A number of experiments have been undertaken to assess the effectiveness of OXR antagonists and to evaluate the behavioral and physiological consequences of their usage [18, 20, 24]. There are also other studies confirming the possible use of OXR blockers in the treatment of nicotine, cocaine, or morphine addictions [20, 65, 66].

In most experiments related to the effects of alcohol, the activity of SB-334867 is specific, but almost all OX2R and OX1/2R antagonists exhibit a similar trend in the decrease of alcohol (and sugar) intake, and sometimes they may indicate a problem with excessive sedation. The dual antagonists almorexant and suvorexant especially show such a tendency. Therefore, these two compounds represent possible groups of new hypnotics. Moreover, suvorexant has a rather low addictive potential in comparison with zolpidem [67]. Suvorexant shows overall sleep-promoting effects in the ethanol-treated group after 24 hours following alcohol withdrawal in rats [68]. So it appears that the sleep-promoting action may be an additional benefit because sleep disorders are often associated with addiction and treating them improves the abstinence rates [69]. For example, the occurrence of insomnia in people with AUD is an important factor causing the recurrence of ethanol drinking [70]. The majority of alcohol addicts suffer from insomnia and they often self-medicate this condition with alcohol [70]. The additional hypnotic effect of substances that reduce ethanol consumption can be considered as an advantage in some cases.

Furthermore, SB-334867 seems to have an impact on cognitive function and anxiety, which could lead to doubts as to whether this substance significantly reduces alcohol consumption in various models of this addiction [20]. However, blocking OX1R with SB-334867 has a minimal effect on the inhibition of measured response in a rat model of the so-called stop-signal reaction time task (a different experimental model designed to verify whether a given xenobiotic affects cognitive function or attention) [71], which clearly indicates its impact on drug-seeking behavior without disturbance of cognitive functions.

Of all the studies described above, the use of naltrexone as a positive control was found only in Shoblock et al. [45] and Anderson et al. [4]. The researchers not only checked the effectiveness of the substances studied but also compared the effects they produce with the drug mostly used in the treatment of alcoholism.

ORXs signaling via OX1Rs has a strong effect on lowering the alcohol consumption. SB-334867 (its activity and selectivity on OX1Rs) was the most frequently studied compound, showing the desired behavioral effects. The use of selective OX2Rs antagonists, i.e., TCS-OX2-29 or LSN2424100, did not produce such effective results; they either nonspecifically reduced alcohol consumption [42, 56] or were active only in the binge-like drinking paradigm [4].

The verification of behavioral and physiological consequences of using OXR antagonists (mostly targeting OX1Rs) provides crucial knowledge regarding their pharmacological profile in the field of experimental alcohol addiction in vivo models. Such knowledge may be valuable, so increasing the number of experimental studies should be postulated. Last but not least, it would be beneficial to expand the scope of currently published clinical trials involving OXR antagonists and to successfully treat those aspects of AUD that coexist with other behavioral or physiological disorders, such as insomnia or anxiety.