The evolving story of orexin biology: the hits keep coming.

In the span of just 11 years since their discovery, the study of the orexins (hypocretins) has not only provided insight into the biology of sleep/wakefulness, but also demonstrated the importance of the development of new pharmacologic tools and genetic models with which to understand basic physiologic mechanisms and provide potential strategies for the treatment of human pathologies. Highlights from recently published novel approaches and findings are reviewed here.

Introduction and context

Discovered almost simultaneously by two independent groups, orexin A and orexin B (also known as hypocretin-1 and hypocretin-2) are post-translational products of a gene expressed almost exclusively in neurons of the hypothalamus [1,2], where functionally distinct subpopulations of orexin neurons have been hypothesized to reside [3]. Orexin neurons in the lateral hypothalamic area have been suggested to be important in hunger and reward, while those in the dorsomedial hypothalamus are more related to arousal state [3]; however, their exact roles remain open for debate [4]. Two cognate receptors have been identified, orexin type 1 receptor (OX1R), which couples primarily via a Gq/11, and orexin type 2 receptor (OX2R), which couples via Gi or Gq/11, and perhaps a Gs protein. In addition to their originally described orexigenic actions, the peptides act pharmacologically to increase arousal [5,6], elevate sympathetic tone [7,8], stimulate the neuroendocrine mechanisms controlling stress hormone secretion [9-12], and activate reward pathways [3]. However, their hallmark actions, which are clearly physiologic in nature, are related to sleep/wakefulness [13]. Compromises in the production of orexin or OX2R result in a behavioral phenotype similar to human narcolepsy/cataplexy [14,15] and indeed many human cases of narcolepsy appear to be the result of selective degeneration of orexin-producing neurons in the lateral hypothalamic/perifornical area of the brain [16].

Major recent advances

The potential for therapeutic use of orexin or orexin analogs to treat the debilitating symptoms of narcolepsy has driven the study of these pluripotent peptides. In the process, much has been learned about more than simply the biology of sleep/wakefulness; many novel approaches for their study have been developed. One that cannot adequately be discussed here is the innovative use of genetic targeting of the channelrhodopsin-2 gene to orexin (hypocretin) neurons followed by optogenetic photostimulation, which resulted in clear evidence that the orexin (hypocretin)-producing neurons control the transition to wakefulness from slow wave and rapid eye movement (REM) sleep [17]. It is impossible in a brief review to cite inclusively all the outstanding manuscripts in the field. Instead, the current focus is on two or three biologic effects of the peptides, recent advances in the study of those actions, and the innovative approaches employed by investigators in the field to elucidate the full biologic spectrum and potential therapeutic uses of the orexins. For more comprehensive coverage of the subject, the readers may consult a recent, excellent review by one of the peptides’ discoverers, T Sakurai [13].

Which orexin receptor mediates the arousal effects?

Deletion of the orexin gene results in an animal with a behavioral phenotype similar to human narcolepsy/cataplexy [14]. Similarly, compromise of the OX2R gene [15] produces a narcoleptic/cataplectic phenotype. In normal rats, central administration of a selective OX2R ligand enhanced wakefulness while reducing REM and non-REM sleep [18]. An innovative approach to the identification of the orexin receptor subtype responsible for maintenance of wakefulness was recently employed by Shiromani and colleagues [19]. They identified OX2R expression in enhanced green fluorescent protein-positive GABAergic (γ-aminobutyric acid-producing) neurons of the ventrolateral periaquaductal gray (vlPAG) of a transgenic animal and then targeted those neurons for destruction using a saporin hypocretin-2 (orexin B) conjugate. This not only identified the vlPAG as a potential site of action for the arousal effects of orexin, but also demonstrated the importance of the OX2R in the inhibitory control of REM sleep. Additional experiments in this excellent manuscript suggest the importance of the vlPAG in the action of the orexins is limited to REM sleep, but that it is not a substrate for the cataplexy of orexin absence.

Is there hope for development of a therapeutic strategy for orexin replacement?

In humans degeneration of orexin-producing neurons can cause narcolepsy/cataplexy [16]. Thus, the full phenotype of human narcolepsy might require loss of the orexin-producing neuron, and consequently the absence of not only orexin, but also co-localized peptides like galanin, dynorphin, or neuronal activity-related pentraxin. For this reason Yanagisawa and colleagues developed the orexin-ataxin-3 transgenic mouse and rat models [20,21], in which, over time, the orexin-producing neurons degenerate and the behavioral phenotype eventually appears. Proof that the behavior was due to the loss of orexin came in studies using these animals with subsequent central administration of high doses of orexin or ectopic expression in brain of a prepro-orexin transgene [22]. Thus, it appeared possible that ligands selective for OX2R might be developed that could cross the blood-brain-barrier and replace the missing endogenous orexin, rescuing the phenotype. However, with the exception of one report in nonhuman primates [23], little evidence for therapeutic penetration of peripherally administered orexin has been forthcoming.

An alternative would be gene transfer within brain. Proof of concept was recently published by Liu et al. [24]. Using a replication defective herpes simplex virus-1 amplicon-based vector, these authors transferred the gene encoding orexin into the lateral hypothalamus of orexin knockout mice. They demonstrated transient neuronal expression and, importantly, a temporally correlated improvement of the behavioral phenotype. As the authors openly conclude, gene transfer worked in the orexin knockout mice, where orexin expression has been lost but not the orexin-producing neurons. The technique now needs to be applied to the more relevant animal model, the orexin-ataxin-3 transgenic mouse (or rat), in which the orexin-producing neurons have been lost, just as is the case for the brains of humans suffering from narcolepsy/cataplexy.

What about the opposite side of the coin? Could orexin antagonists be used to treat insomnia?

Almorexant, a dual OX1R/OX2R antagonist has been demonstrated to act after peripheral administration to promote sleep in a variety of species, including humans [25]. Recently, Dugovic and colleagues [26] demonstrated that the effects of almorexant were in all likelihood due to the compound’s ability to block OX2R, as similar sleep-promoting effects (decreased latency for persistent sleep and increased non-REM and REM sleep time) were observed in rats treated with a selective OX2R antagonist.

What other effects of orexin might appear during therapeutic use of agonists or antagonists of the receptors?

As mentioned already, the orexins are pluripotent neuromodulators. Two areas of current research point to impressive additional actions of the peptides. Orexin knockout mice not only demonstrate a narcoleptic phenotype, but also with aging develop obesity, which at first seems counterintuitive since the peptide pharmacologically stimulates food intake. Indeed, knockout mice eat less but gain more weight than age-matched wild-type mice. Once again, the Yanagisawa and Sakurai labs have employed state of the art genetic approaches to examine this apparent paradox [27]. They have demonstrated recently that transgenic overexpression of orexin protects against high-fat-diet-induced obesity, and the attendant insulin insensitivity, by promoting energy expenditure. Surprisingly, these mice actually ate less as well. Combining the genetic background of the orexin overexpressing transgene with either OX1R or OX2R knockout animals, these authors demonstrated that OX2R is essential for the prevention of diet-induced obesity. Leptin sensitivity was also improved in these animals. This may indicate the important interaction of metabolic cues and orexin, as might be predicted from the intriguing findings of Burdakov and colleagues [28] that orexin neurons in lateral hypothalamus respond directly to fluctuating levels of plasma and cerebrospinal fluid glucose. Thus, selective activation of OX2R in humans may not only ameliorate the sleep disturbances of orexin cell loss, but also alleviate some of the consequences of the metabolic-syndrome-like phenotype of many narcoleptic humans [29].

Can manipulation uniquely of the OX1R be rewarding?

Finally, much interest has been generated in the possibility that the orexins play an important role in reward behaviors [3]. Orexin appears to influence morphine-induced place preference [30]. A role for OX1R in reward/addictive behaviors has been demonstrated by these authors as well [3]. Antagonism of OX1R, but not OX2R, decreased cue-induced cocaine seeking, following extinction or abstinence [31]. Thus, one could envision the potential use of orexin antagonists for the modification of reward-based behaviors [3]. However, OX1R has also been implicated as the receptor mediating the autonomic and neuroendocrine actions of the orexins [12] and, therefore, selectivity of effect may be a difficult issue to control. The realization of the important actions of the orexins on reward-based and addictive behaviors may provide insight into the neuronal mechanisms responsible for the expression of those behaviors.

Is arousal the overriding effect of the orexins or can the peptides exert biological actions independent of arousal state?

Another way of stating the question would be: is the orexigenic action of the peptide secondary to arousal? If so, why do orexin knockout mice become obese even though they eat less? We were never overly impressed by the magnitude of food consumption observed in our ad libitum fed rats following orexin administration, but we were impressed by the sympatho-stimulatory effects of those doses and the increases in spontaneous locomotor activity that ensued [5,6,8]. Are the increased autonomic function and locomotor activities a reflection of activation of stress pathways in the brain [9-12]? Does activation of stress pathways initially induce and subsequently suppress food intake? It is clear from the outstanding work of Yanagisawa and colleagues [27] that “…orexin gain of function promotes energy expenditure while loss of function promotes energy conservation”. As those authors speculated [27], the sympatho-stimulatory action of the orexins may partly explain those observations.

Future directions

Clearly the development of orally active orexin analogs is a goal of many investigators. However, the simple fact that the orexins are ‘pluripotent’ neuropeptides raises design problems that may prove daunting, if not insurmountable. While some antagonists are in development, the manipulation of central orexin levels has, to date, been limited to direct administration or transgenic delivery. Orexin analogs for therapeutic usage must have receptor selectivity especially when targeting OX2R, since activation of OX1R, at least in experimental animals, causes a stress response (i.e., elevated mean arterial pressure and stress hormone secretion). The promise of application of OX2R selective ligands for the treatment of metabolic syndrome is great, but again bioavailability and selectivity are major hurdles to be cleared. Will these orexin analogs be of clinical value for addiction therapies and eating disorders? Here too, selectivity will be an issue. In any case, the study of orexin biology has been and still is opening many new insights into human physiology and driving the development of novel scientific approaches.