Recent perspectives on orexin/hypocretin promotion of addiction-related behaviors

Highlights

• An attempt to synthesize recent work addressing orexin and addiction.

• Orexin/Ox1R importance for more motivating substances and/or overcoming cost/effort.

• Ox1R importance for individuals with higher motivation for intoxicants.

• Orexin importance for anxiety/stress responding, perhaps centrally for higher salience.

• Differences in orexin regulation of behavior in females and males.

Abstract

The neuropeptide hypocretin/orexin plays a broad and important role in physiological functions ranging from addiction, stress, and anxiety to sleep, energy metabolism, and homeostatic regulation. A number of recent reviews addressing the importance of orexin for different addictive behaviors, especially the contribution of orexin-1-receptors (Ox1Rs) in responding for intoxicants in higher-motivation individuals and situations, and orexin-2-receptor (Ox2Rs) in stress-related aspects of addictive responding. This may parallel the importance of more lateral orexin neurons in the hypothalamus for reward and more medial for stress and arousal. However, there is clearly also some crossover, which may reflect, in part, where positive and negative conditioning (reward- and relief-seeking) are both present concurrently in established addiction, and also where orexin signaling can differ in subregions of a particular brain region. Here, we attempt to examine and synthesize some of the most recent work addressing orexin functions in addiction, including a particular role for Ox1Rs for driving responding in higher-motivation individuals and under higher levels of effort. While there are some commonalities across addictive substances addressed here (alcohol, cocaine, opiates), there are also some differences, which may relate to several factors including the speed of intoxication with a given substance. Together, recent findings have shed important insight and clues into what a more unified role of Ox1Rs might entail, and critical areas for future work. In addition, these many studies support the development of Ox1R blockers for use in humans to counteract addiction and other disorders of motivation.

This article is part of the special issue on Neuropeptides.

Introduction

Here, we examine and attempt to synthesize some of the most recent work addressing the function of the neuropeptide orexin, especially studies related to addiction, and to identify critical areas where additional information is needed. One particular focus is on the particular importance of orexin for responding for more motivating substances, especially when overcoming cost and higher effort are required. Similarly, orexin receptors are particularly impactful in individuals with higher motivation for and intake of intoxicants. Orexins are important for anxiety and stress responding, perhaps suggesting a role in certain higher salience behaviors, rather than positive or negative events per se. Also interesting are differences in orexin regulation of behavior in females and males, e.g. where females have higher orexin levels and differential regulation of orexin-related behaviors. We also try to address the molecular bases of differential orexin effects, including changes in the number of orexin-expressing neurons. Finally, we briefly address the targeting of orexin receptors as possible therapy to reduce the impact of addiction.

The neuropeptide hypocretin/orexin has been written about extensively, with a number of recent reviews addressing the importance of orexin for different addictive behaviors as well as negative affective states (e.g. Johnson et al., 2012, Barson and Leibowitz, 2016; James et al., 2017a, James et al., 2017b, Walker and Lawrence, 2017, Anderson et al., 2018, Moorman, 2018, Summers et al., 2018, James and Aston-Jones, 2020). More generally, orexins play an important role in physiological functions ranging from addiction, stress, and anxiety to sleep, energy metabolism, and homeostatic regulation (reviewed in de Lecea et al., 1998, Mahler et al., 2014, Brown et al., 2015b, Li et al., 2016, James et al., 2017a, James et al., 2017b). Orexins are synthesized in a specific subset of cells localized to the lateral hypothalamus area (LHA) and regions just medial (dorsomedial hypothalamus, perifornical area), while orexin-containing fibers are broadly distributed across the brain allowing very wide potential impact. Also, LHA are proposed to mediate reward behaviors, while more medial orexin neurons may be more important for stress and arousal (Harris et al., 2005, Aston-Jones et al., 2010, Freeman and Aston-Jones, 2018). Further, orexins act through two classes of receptors, orexin-1-receptor (Ox1R) and orexin-2-receptor (Ox2R). Ox2Rs are more broadly distributed across the brain, while Ox1Rs are more specifically localized within key limbic regions (e.g. D'Almeida et al., 2005), and, in parallel, Ox2Rs are related more to sleep and arousal while Ox1Rs are more related to addiction, reward and motivation, although there clearly is some crossover (de Lecea et al., 1998, Mahler et al., 2014, Brown et al., 2015b, Li et al., 2016, James et al., 2017a, James et al., 2017b) (and see below). Fig. 1.

One central finding is that orexin signaling is of particular importance for driving many motivation- and addiction-related behaviors (Aston-Jones et al., 2010, Mahler et al., 2012, Mahler et al., 2014, Boutrel et al., 2013, Barson and Leibowitz, 2016, James et al., 2017b. More specifically, Ox1Rs are linked to ostensibly higher-motivation behaviors, with less impact on intake involving less reinforcing substances and/or simpler response requirements. For example, using SB336847 (SB), a generally validated Ox1R-targeted compound (see Summers et al., 2018), Ox1Rs are shown to be critical for cost-resistant but not cost-free responding for cocaine and some opiates (described in detail below). In addition, Ox1R inhibition reduces progressive-ratio responding for cocaine and high-fat food but not simpler food pellets (Borgland et al., 2009), although response levels are similar for food and cocaine. Also, work from our lab and others finds greater Ox1R importance in individuals that drink higher levels of alcohol, with parallel studies for cocaine and opiates (see below). While many critical questions remain, a central Ox1R role in higher motivation suggests that orexin receptor blockers may reflect useful therapeutic interventions for addiction and other motivational disorders in humans (Khoo and Brown, 2014, Li et al., 2016, James et al., 2017b, Walker and Lawrence, 2017, Perrey and Zhang, 2018).

One area of mixed findings is the impact of Ox1R inhibition on responding for natural rewards. In some studies, responding for high-fat and/or -sweet substances requires Ox1Rs (e.g., Borgland et al., 2009, Srinivasan et al., 2012, Anderson et al., 2014, Cason and Aston-Jones, 2014, Olney et al., 2015), while in others it does not (e.g., Lopez et al., 2016, Matzeu and Martin-Fardon, 2018, Fragale et al., 2019). Indeed, one group found that cocaine but not highly palatable foods increases cFos in orexin neurons (Matzeu and Martin-Fardon, 2018). Also, Jupp et al. (2011) show that SB reduces alcohol and sucrose responding under a simpler response requirement (fixed-ratio 3) but only impact alcohol (not sucrose) responding under progressive ratio responding (and somewhat converse to the general link of Ox1R to higher motivation). Further, many findings indicate that Ox1Rs in particular brain regions are required for drug/alcohol but not sucrose intake. Local Ox1R inhibition with SB in cortex reduces nicotine but not food intake (Hollander et al., 2008) and alcohol but not sucrose seeking (Brown et al., 2015a), and SB in VTA reduces alcohol seeking, cocaine intake, and morphine CPP but not locomotion (Aston-Jones et al., 2010, James et al., 2011, Mahler et al., 2013, Brown et al., 2015a). Furthermore, i. c.v. administration of SB decreases alcohol but not saccharin or food intake (Carvajal et al., 2015). Also, higher-drinking selected mice have higher anxiety and compulsion-like alcohol responding, with no difference in saccharin preference or sweet or bitter taste reactivity (Wolstenholme et al., 2011; Radwanska and Kaczmarek, 2012; Bahi, 2013). Thus, one critical question is how higher motivation for addictive substances might differ from higher motivation for “natural” rewards. One speculation is that natural reward circuitry taps into more basic homeostatic regulatory systems, while addiction involves dysregulation of emotional processing rather than physiological homeostasis per se. In addition, sensory, attentional, response, and reward aspects of the context present could all influence whether orexins are required to respond. For example, the presence or absence of food deprivation might also be important, since it increases orexin cell activity (e.g. Soya et al., 2017, Yeoh et al., 2019; but see Cason and Aston-Jones, 2013). Thus, orexin importance for natural versus drug rewards remains an area needing strong clarification.

One avenue of particular interest is the identification of behaviors that are sensitive or insensitive to orexin receptor inhibition, which could provide important clues to understanding more specifically the central role of orexin, including Ox1Rs. There are a few caveats to this approach, including that the majority of Ox1R studies utilize SB, which, although widely used, has the possibility of off-target effects especially at higher doses. However, some groups have argued that the behavioral specificity of SB indicates a selective Ox1R effect, e.g. if there are limited effects on simpler motor behaviors but more selective effects on particular aspects of motivated responding. As discussed below, this approach has resulted in some mixed findings, but are nonetheless overall quite encouraging. Also, some central findings have been replicated with additional Ox1R specific drugs for alcohol intake (e.g. Lopez et al., 2016), or with Ox1R shRNA for cocaine intake (Bernstein et al., 2018), increasing confidence in the selectivity of SB in targeting Ox1R. In addition, since SB can impact Ox2Rs at higher concentrations (Scammell and Winrow, 2011), studies can show a behavioral effect of SB but not an Ox2R blocker (e.g., Lei et al., 2016a, Lei et al., 2016b), also increasing confidence in the selectively of any observed SB effect (at least in relation to different orexin receptors).

Also of interest is the relative dose effect of Ox1R inhibition, i.e., where some behaviors are only inhibited by higher levels of SB, while others can be inhibited by much lower SB concentrations. For example, compulsion-like responding for alcohol is inhibited by lower SB doses (systemically) relative to those that reduce alcohol-only intake (Lei et al., 2016a, see below). We interpret such a finding to indicate that the behavior is more dependent on the particular signaling pathway (in this case orexin), with fewer parallel signaling systems to help support the behavior. However, one caveat when evaluating SB studies is that the solubility of commercially available SB may have changed across time. As strange as it may seem, studies after the 2000s require DMSO in order to get SB to be soluble, while studies before this time report solubility in vehicle without DMSO (discussed in Lei et al., 2019). While the exact basis of this remains unclear, it represents an important consideration when comparing the results of earlier versus later studies using SB. However, despite these different concerns, studies using systemic SB and other Ox1R blockers have shown interesting and selective patterns of behavioral disruption (described in detail below).

Another, more theoretical, caveat when searching for a unifying role for orexin is that many studies involve systemic drug administration. In particular, given the wide range of states orexin is implicated in, systemic inhibition might decrease both activating and inactivating pathways, and thus have no clear net effect. While this seems more complicated, some experimental findings (not related to orexins) suggest such a possibility. For example, Yun et al. (2004) find that dopamine receptor blockers in the nucleus accumbens (NAc) strongly reduce responding to a predictive cue, while glutamate receptor blockers or tetrodotoxin in the NAc do not reduce cue responding. One explanation for such a finding is that the NAc contains direct and indirect output pathways that act in opposition (e.g. as suggested by Calipari et al., 2016, and others). Thus, strong inhibition would remove the impact of both output pathways, resulting in no net effect, while a more selective disruption of signaling (dopamine receptors) would impact only one of the two outputs and thus is able to have a behavioral effect. Similarly, we find that inhibiting NMDA receptors within the NAc disrupts compulsion-like alcohol drinking, with no effect on alcohol-only intake (Seif et al., 2013, Seif et al., 2015), but that strong NAc suppression with muscimol/baclofen did not reduce either form of drinking (unpublished findings). Given the diversity of sites in the brain where orexin acts, and the ability to regulate both appetitive behaviors and anxiety/stress related avoidant behaviors (see Introduction), some caution may be warranted for studies with systemic receptor inhibitors.

Compulsion-like drives, where consumption continues despite negative consequences, is central to allowing much of excessive human intake (Larimer et al., 1999, Anton, 2000, Tiffany and Conklin, 2000, Koob and Volkow, 2010, Hopf and Lesscher, 2014, Everitt and Robbins, 2016, Voon et al., 2015). The recent DSM-5 emphasizes disordered choosing (deciding to continue intake despite known harm) as central to the diagnosis of alcohol use disorder (DSM–5, 2013; Grant et al., 2015). This cost/effort-resistant aspect of intake is likely critical for driving addiction for many substances, in concert with other factors such as greater valuation of rewards (Hogarth, 2020). There have been different considerations of compulsion, some of which consider compulsion a stronger form of devaluation-resistant habit (Everitt and Robbins, 2016), while others consider compulsion more of a goal-directed process where habit mechanisms may be recruited in the service of overcoming and ignoring conflict associated with action (see Tiffany and Conklin, 2000, Naqvi and Bechara, 2010, Hopf and Lesscher, 2014, Hopf, 2017). There is still some divergence in consideration of whether consequence-resistant actions involve excessive insensitivity to cost information (e.g., Claus et al., 2013, Reiter et al., 2016, Sebold et al., 2017) or not (Hogarth and Hardy, 2018, Hogarth, 2020). However, in either case, some aspect of cost discounting seems to be a consistent consideration for many addicts (references above). Thus, even with these caveats, there is considerable interest in understanding the underlying circuitry and molecular signaling that allows aversion-resistant drives to have such power in promoting intake. Also, since compulsion-like determination likely involves high motivation (since the drive to consume persists despite negative consequences), it is also logical that Ox1Rs might be particularly important for promoting aversion-resistant responding (as discussed below). Furthermore, tolerance of greater relative effort (as described below) is likely to be conceptually related to cost-resistant responding, and thus a more specific role for Ox1Rs might also be predicted for higher effort conditions.