The Perfect Hypnotic?

See allHide authors and affiliations

Science  05 Apr 2013:
Vol. 340, Issue 6128, pp. 36-38
DOI: 10.1126/science.1237998

This article has a correction. Please see:

The market for hypnotics is big business, with 10 to 15% of the population in the United States suffering from chronic insomnia (1, 2). Although the pathology is unknown, and likely heterogeneous, patients with insomnia are in a state of hyperarousal (even during the day), suggesting that wake-promoting systems are hyperactive (3). The search for the ideal hypnotic has been marked by cycles of exuberance followed by disappointment, as adverse side effects associated with each new class of drug have emerged following wide use. Although insomnia therapy increasingly uses cognitive behavioral therapy, the enormous number of sufferers mandates new pharmacological approaches. Uslaner et al. (4) report the prospect of a new class of compound with a new mode of action that may usher in a new era for insomnia treatment, with the potential for fewer side effects.

At the beginning of the 20th century, chloral hydrate, meprobamate, and barbiturates were touted as nonaddictive miracle tranquilizers, but in the 1950s, their potential for tolerance and severe addiction was recognized (5). At high dose, they lead to pulmonary arrest and death, outcomes that gained further notoriety with the deaths of celebrities Marilyn Monroe and Jimi Hendrix. Broad inhibition of brain activity through high-dose barbiturates can even induce a “flat” electroencephalogram mimicking brain death. Barbiturates activate a chloride channel receptor for gamma-aminobutyric acid [type A (GABAA)]. GABA is the main inhibitory neurotransmitter, produced by 15 to 20% of all brain neurons.

Hope for a safe, effective hypnotic reemerged with chlordiazepoxide, the first benzodiazepine [(BDZ); a two-benzene ring structure linked by a third, diazepine ring]. BDZs are safer, rarely causing respiratory depression, even when “overdosed,” and more than 40 derivatives (including the prototypical diazepam) have been synthesized. Prescribed variously for sleep induction or maintenance, as anticonvulsants, muscle relaxants, or as antianxiety (anxiolytic) agents, these drugs were proclaimed as the perfect solution. But they too fell out of favor when long-acting compounds were found to induce residual sedation, memory loss, and addiction. BDZs bind to the GABAA receptor at a site that is different from that for barbiturates, yielding similar but less extensive effects (5, 6) The residual effects, however, spurred development of short–half-life hypnotics in the 1980s (e.g., triazolam). Although effective, these compounds manifested other problems: rebound insomnia on cessation and occasionally a state of “confusional arousal” in which the patient may be acting out unaware, half asleep but moving around, not unlike sleep-walking (this led to legal cases). Triazolam was withdrawn from many markets and/or the dose decreased.

On the horizon?

A compound that blocks the wake-promoting effects of orexin hormones may usher in a new era for insomnia treatment.


Then came the “Z-drugs”—zaleplon, zolpidem, and zopiclone (and eszopiclone)—short–half-life compounds with lower effective doses and improved side-effect profiles. Although lacking the three-ring BDZ structure, they recognize the BDZ-binding site of the GABAA receptor and manifest receptor subtype specificity such that some have primarily hypnotic (zolpidem) rather than both hypnotic and anxiolytic (eszopiclone) effects (6). Although safer than classic BDZs, occasional problems with dependence, tolerance, and “confusional arousals” are still reported with Z-drugs in at-risk populations. These BDZ-like compounds are now the “gold standard” treatment for insomnia, although they are “scheduled” substances by the U.S. Food and Drug Administration (their use and distribution are tightly controlled because of abuse potential) and viewed with some suspicion by doctors and patients.

In parallel with this drug evolution, sedatives (e.g., chlorpromazine) were introduced to treat agitation in psychotic patients in the 1950s. These compounds block receptors for the neurotransmitter dopamine, although sedative effects mostly correlate with their ability to block H1 histaminergic receptors that relay wake-promoting histamine signals within the brain. Antihistamines were developed to treat allergies but soon found utility as remedies for insomnia. More “natural” sedatives, such as the hormone melatonin, have a modest hypnotic effect, typically insufficient for severe insomnia, although typical doses far exceed normal physiological amounts. Melatonin or melatonin agonists (such as ramelteon) activate cognate receptors in the brain to provide a “darkness” signal that may be useful, when combined with light therapy, to adjust circadian rhythms. Similarly, wary of addiction to GABAergic drugs, many patients are prescribed sedative antidepressants that block the H1 histamine or the 5-hydroxytryptamine (5HT2) serotonin receptors. These compounds were never developed for insomnia, and have long half-lives, thus causing residual daytime sedation (5).

Uslaner et al. (4) explored the effects of a new class of hypnotic compounds that are antagonists of the two receptors for orexin, so-called dual orexin receptor antagonists (DORAs). Orexins (also called hypocretins) are key neurotransmitters of arousal in the central nervous system. Genetic and pharmacological studies pointed to modulation of the orexin system as potential therapy for insomnia and other sleep disorders (7, 8). DORAs have fewer effects on daytime performance tests, and no rebound insomnia on cessation (9), although there are residual dose-dependent sedative effects (10), and high doses administered during the day can impair human performance. Uslaner et al. (4) compared the effects of DORA-22 with those of diazepam (a BDZ hypnotic), and eszopiclone and zolpidem (BDZ-like hypnotics) in rats and monkeys and concluded that DORA-22 promotes sleep at doses that do not impair cognition and memory, in contrast to BDZ and BDZ-like compounds. The experimental task studied included novel object recognition by rats, with associated changes in hippocampal Arc expression, a measure of synaptic plasticity believed to correlate with memory retention. These measures were all impaired by diazepam, zolpidem, and eszopiclone but not by DORA-22. Similarly, the ability of monkeys to remember and match colors after a small delay (a working memory task) or to react rapidly to the appearance of an object on a screen (a measure of attention) was not impaired by DORA-22, whereas all GABAergic treatments had substantial detrimental effects. What is needed next are parallel studies in humans to demonstrate generalization to a clinical population.

Unlike the broad inhibitory effects of GABA on brain activity, orexins produce selective wake-promoting signals. Although orexins are produced by only 70,000 neurons in the hypothalamus, these send widespread anatomical projections, thereby exciting other wake-promoting systems such as the histaminergic tuberomammillary nucleus, the adrenergic locus coeruleus, and various cholinergic and aminergic cell groups. Blocking orexin may thus be closer to treating the underlying issue of excess alertness in insomnia compared to promoting sleep by inhibiting brain activity.

Discovered in 1998 (11, 12), the role of orexins (there are two types, A and B, processed from the same pre-propeptide) in sleep emerged when it was discovered that impaired orexin signaling results in the sleep disorder narcolepsy in animal models and humans (7). Narcolepsy is characterized by sleepiness and abnormal rapid eye movement (REM) sleep, so that patients not only fall asleep easily but also experience symptoms where REM sleep intermingles with wakefulness, resulting in dream-like hallucinations or episodes of muscle paralysis when awake (sleep paralysis and cataplexy). Narcolepsy is caused by an autoimmune attack against neurons that express orexin.

Orexins act on G protein–coupled receptors called hypocretin receptor 1 and 2 (also called OX1R and OX2R, respectively). Although OX2R is more prominently involved in narcolepsy, both receptors may regulate sleep, although this could be species dependent. The potential for orexin antagonists to increase REM sleep and dreaming, or even narcolepsy-like symptoms, has been a concern (8), but has not been observed to a substantial degree in clinical trials with DORA molecules (13). A small but significant increase in REM sleep and decreased REM latency were noted, and sleep may be less consolidated than after GABAergic hypnotics (9, 10, 13). As the beneficial effects of REM versus non-REM sleep on restoration of wakefulness, memory, synaptic plasticity, and mood are highly debated (14, 15), it is uncertain whether this difference in REM versus non-REM profile may be beneficial.

Are DORAs the perfect hypnotics? Only long-term use in large numbers of insomnia patients will reveal whether these drugs will be preferred to GABAergic hypnotics, and whether they produce rare complications, including narcolepsy-like symptoms in predisposed individuals (16). Side effects are expected with any active treatment, and the benefit-risk ratio must be considered. There is minimal tolerance for side effects and risk in the treatment of insomnia, so distinct drugs offering multiple modes of action, especially complementary ones, will greatly benefit patients.

References and Notes

  1. The U.S. Food and Drug Administration is currently reviewing an application for the use of a DORA in insomnia, based on the result of a phase III clinical trial;

Navigate This Article