The sleep-inducing effect of prostaglandin D2 (PGD2) was studied in five conscious male rhesus monkeys (Macaca mulatta) maintained in a 12-hr light/dark cycle. PGD2 was infused into the lateral or the third ventricle of the cerebrum slowly and continuously for 6 hr in the light period. Infusion of PGD2 into the lateral ventricle at 15-2250 pmol/min induced natural sleep as identified by electroencephalogram, electromyogram, electrooculogram, body temperature, heart rate, and animal behavior. Although sensitivity to PGD2 was slightly different among individual animals, the amount of total sleep time increased maximally up to 3-to 4-fold over the control level. PGD2 infused into the third ventricle induced effects similar to those observed for the lateral ventricular route, but infusion into the third ventricle was about 1000 times more effective than infusion into the lateral ventricle. In three monkeys, PGD2 increased the amount of sleep in a dose-dependent manner. Bell-shaped dose-esponse curves were observed for the other two monkeys. Infusion of prostaglandin E2 or F2,, into the lateral ventricle caused sedation but slightly reduced the amount of slow-wave sleep and produced increases in heart rate and body temperature. These findings suggest that endogenous PGD2 may be involved in the regulation ofsleep by acting on the brain structures surrounding the third ventricle in the rhesus monkey.Prostaglandin D2 (PGD2) is one of the major prostanoids in the mammalian brain including the human (1-3), and its metabolism (4, 5) and physiological activities (6, 7) in the central nervous system have been delineated in detail. Previous work in our laboratory showed that microinjection of PGD2 into the preoptic area of conscious rats, or intracerebroventricular administration ofas little as 60 fmol ofPGD2 per min to rats, induced sleep that was indistinguishable from natural sleep on the basis of electroencephalogram (EEG), electromyogram (EMG), and behavior (8, 9). The sleepinducing effect of PGD2 was dose-dependent and specific for PGD2 (8, 9). Other prostaglandins were much less effective or were totally inactive (8, 9); in fact, prostaglandin E2 (PGE2) was inhibitory (10 (14). For recordings of EEGs, stainless steel screw electrodes were implanted on the dura at frontal, parietal, and occipital cortices. Similar electrodes were placed across the outer canthi for recordings of electrooculograms (EOGs). The stainless steel wire EMG electrodes were inserted subcutaneously into the posterior cervical muscles. For the infusion oftest compounds into the cerebral ventricles, stainless steel guide cannulae (outer diameter, 0.8 mm) were implanted into the lateral and the third ventricles as described below according to the stereotaxic atlas of Snider and Lee (15). By continuous x-ray monitoring of Angio-Conray (sodium iothalamate) contrast medium, which was injected from the cannula, the position of the cannula, as well as the location of the lateral and third ventricles, was identified. In one monkey, we implanted ano...