The endocannabinoids are a family of bioactive lipids that activate CB 1 cannabinoid receptors in the brain and exert intense emotional and cognitive effects. Here, we have examined the role of endocannabinoid signaling in psychotic states by measuring levels of the endocannabinoid anandamide in cerebrospinal fluid (CSF) of acute paranoid-type schizophrenic patients. We found that CSF anandamide levels are eight-fold higher in antipsychotic-naïve first-episode paranoid schizophrenics (n ¼ 47) than healthy controls (n ¼ 84), dementia patients (n ¼ 13) or affective disorder patients (n ¼ 22). Such an alteration is absent in schizophrenics treated with 'typical' antipsychotics (n ¼ 37), which antagonize dopamine D 2 -like receptors, but not in those treated with 'atypical' antipsychotics (n ¼ 34), which preferentially antagonize 5HT 2A receptors. Furthermore, we found that, in nonmedicated acute schizophrenics, CSF anandamide is negatively correlated with psychotic symptoms (r S ¼ À0.452, P ¼ 0.001). The results suggest that anandamide elevation in acute paranoid schizophrenia may reflect a compensatory adaptation to the disease state.

There is much evidence that progesterone has hypnotic anesthetic properties. In this vehicle-controlled study, we examined the effects of three doses of progesterone (30, 90, and 180 mg/kg) administered intraperitoneally at light onset on sleep in rats. Progesterone dose dependently shortened non-rapid eye movement sleep (NREMS) latency, lengthened rapid eye movement sleep (REMS) latency, decreased the amount of wakefulness and REMS, and markedly increased pre-REMS, an intermediate state between NREMS and REMS. Progesterone also elicited dose-related changes in sleep state-specific electroencephalogram (EEG) power densities. Within NREMS, EEG activity was reduced in the lower frequencies (< or = 7 Hz) and was enhanced in the higher frequencies. Within REMS, EEG activity was markedly enhanced in the higher frequencies. The effects were maximal during the first postinjection hours. The concentrations of progesterone and the progesterone metabolites 3 alpha-hydroxy-5 alpha-pregnan-20-one and 3 alpha-hydroxy-5 beta-pregnan-20-one, both positive allosteric modulators of gamma-aminobutyric acid A (GABAA) receptors, were determined at different time intervals after vehicle and 30 or 90 mg/kg progesterone. Progesterone administration resulted in dose-dependent initially supraphysiological elevations of progesterone and its metabolites in the plasma and brain, which were most prominent during the first hour postinjection. The effects of progesterone on sleep closely resemble those of agonistic modulators of GABAA receptors such as benzodiazepines and correlate well with the increases in the levels of its GABAA agonistic metabolites. These observations suggest that the hypnotic effects of progesterone are mediated by the facilitating action of its neuroactive metabolites on GABAA receptor functioning.

Recent studies in the rat demonstrated that systemic administration of muscimol and THIP, both selective GABAA receptor agonists, elevates slow wave activity in the EEG during non-rapid eye movement (NREM) sleep. In this placebo-controlled study, we assessed the influence of an oral dose of 20 mg THIP on nocturnal sleep in young healthy humans. Compared to placebo, THIP increased slow wave sleep by about 25 min. Spectral analysis of the EEG within NREM sleep revealed significant elevations in the lower frequencies (< 8 Hz) and reductions in the spindle frequency range (approximately 10-16 Hz). In accordance with previous findings in the rat, these data imply that GABAA agonists promote deep NREM sleep, without suppressing REM sleep. These effects are opposite to those induced by agonistic modulators of GABAA receptors such as benzodiazepines and are at variance with established mechanisms according to which GABAA agonists and modulatory agonists would have similar effects. The sleep response to GABAA agonists is highly similar to that evoked by sustained wakefulness, suggesting that GABAA receptors may be implicated in the homeostatic regulation of sleep.

To assess the influence of the gamma-aminobutyric acid (GABA)A receptor on sleep and sleep EEG, rats were injected intraperitoneally with vehicle, two doses of muscimol (0.2 and 0.4 mg/kg), a selective GABAA agonist, and midazolam (3 mg/kg), a benzodiazepine-GABAA agonist. EEG and EMG recordings were made for 6 or 8 hours. Muscimol dose-dependently increased the amount of nonrapid eye movement sleep (nonREMS) and REMS. The higher dose of muscimol enhanced EEG activity over almost the entire frequency range (0.5-25 Hz), including delta (0.5-4 Hz) and sigma (11-16 Hz) activity, within nonREMS and in the frequencies over 10 Hz within REMS. Midazolam also increased the amount of nonREMS. However, most of the other effects of midazolam contrasted the effects of muscimol: midazolam decreased REMS, reduced low frequency (< or = 11 Hz) EEG activity within nonREMS, and enhanced the activity in higher frequencies during both nonREMS and REMS. These data demonstrate the involvement of GABAA receptors in the regulation of sleep-wake behavior as well as in the generation of spindles and delta waves during nonREMS. The effects of these two GABAA agonists indicate that activation of different binding sites on the GABAA receptor complex differentially affect sleep states and sleep EEG.