HIGHLIGHTS New Brain Lipids that Induce Sleep Thomas Kolter and Konrad Sandhoff * In search of endogenous factors with sleep-inducing properties, Lerner et al. isolated a substance from the cerebrospinal fluid of cats that had been deprived of sleep for 22 hours.['] In a further publication the unknown compound was characterized as the aniide I of an o-9-unsaturated fatty acid with a ( Z ) double bond."] The substance, which could not be detected in cats under control conditions, was isolated by reversed-phase HPLC in amounts of 0.1 to 5 pmol per 100 pL of cerebrospinal fluid. Mass spectrometric analysis revealed 4 3 304.2614 for the [ M + Na]' ion, which best fits the formula C,,H,,NO having two double-bond equivalents. The ready loss of a neutral 17Da fragment from the [ M HI+ ion suggests the cleavage of a primary amino group as ammonia. Based on these and further spectroscopic data (UV, 'HNMR), the substance was initially proposed to be a "cerebrodiene", for example 2.[11The postulated structure suggested that 2 may arise biosynthetically from sphingosine 3 or other sphingolipids and thus attracted consid- + erable attention, since several intermediates of sphingolipid metabolism are currently discussed as potential signaling mole~uIes.1~1 The final assignment of structure I to the unknown compound succeeded by comparison of the spectroscopic data to those of model substances.['] The position of the double bond was determined after ozonolysis and its configuration by IR and ' H N M R spectroscopy. Moreover, it could be shown in a funcK S.indholT. Dr. T. Kolter f u r Ol-ganische Chemie und Biochemie der Universitlit (icl-liai-d-L)i~in;i_rk-Sr~-asse 1. D-53121 Bonn (Germany) Telefiis: In1 code + (218)73777X [ * ] Prol'. D r Iiislitu1 tional assay that intraventricular injection of 10 nmol(2.8 pg) of synthetic 1 induced sleep in laboratory animals. Variation of double-bond configuration, position, and alkyl chain length resulted in decreasing extent and duration of the observed effect. Further, it was shown that membrane fractions isolated from brain rapidly degrade 1 into oleic acid. The crucial assumption in this study is an idea that originated at the beginning of our century. Henri Pieron['] postulated the existence of a sleep-inducing substance that is formed in the course of the day, accumulates, and is destroyed during sleep. To verify this hypothesis, Pieron collected cerebrospinal fluid from dogs that had been deprived of sleep for several days. He injected the fluid into the ventricles of other dogs, which then promptly fell asleep for two to six hours. In recent years several substances with sleep-promoting factors have been discovered,"] including the delta sleep inducing peptide.['] which was isolated from the blood of sleeping rabbits and led to sleeplike behavior after intraventricular infusion into rabbit brain (&slow wave sleep). Prostaglandin D2l8,91 also meets most of the requirements of a sleep substance:["] 1 ) The substance must induce sleep. 2) The induced sleep must be natural. 3) A dose-response curve must be demonstrated. 4) Sleep must be followed by appropriate wakefulness. 5) The substance must be present in the brain. 6) There should be binding sites for the substance in the central nervous system. 7) Receptor blockers should prevent sleep induction. 8) Direct application to regions of high binding should induce sleep. 9) The substance should have appropriate circadian changes. 10) Inhibition of synthesis of the substance should reduce sleep. 11) Metabolic pathways should be known. 12) Relevant enzymes should be present in appropriate regions. 13) The substance should accumulate with prolonged wakcfulness. 14) Its concentration should decrease with prolonged sleep. 15) The substance should be present in several species. Additional substances[6. 1'1 discussed as potential sleep promoting factors include interleukin 1, interferon 1x2. lipopolysaccharides, muramyl peptides, serotonin, tumor necrosis factor, and vasoactive intestinal peptide. The relationship between substances apparently acting as neurotransmitters and neuromodulators and the phenomenon of sleep is far from clear. Sleep is one of several circadian rhythms, which, genetically controlled, are synchronized with the change ofday and night as the most important environmental factors." '1 Several stages of sleep can be distinguished, which follow each other in a predictable manner and appear to be controlled by different neurochemical systems."21 The stages of sleep (and of other circadian HIGHLIGHTS processes) are affected differently by drugs in pharmacological intervention." 31 During a typical night's sleep. the adult human progresses through different sleeping periods in a predictable manner, which can be distinguished electrophysiologically (slow wave sleep) and by the occurrence of rapid eye movements (REM sleep). Slow wave sleep and REM sleep alternate four to six times a night with increasing duration of REM sleep but decreasing duration and depth of slow wave sleep."21 Slow wave sleep is characterized by progressively longer wavelengths and amplitudes in the electroencephalogram (EEG). During slow wave sleep, humans pass through up to four levels, which are accompanied by successive lower heart frequency, decreasing blood pressure, and deeper sleep. After about 45 minutes these levels are retraced in the opposite order. After a total of about 90 minutes several abrupt physiological changes occur. and the EEG becomes similar but not identical to that of the waking state. Different names are used for this state: REM sleep. paradoxical sleep, active sleep, and desynchronized sleep. The active EEG patterns are coupled with a loss of muscle tone and rapid eye movements occur. According to the criterion of arousability by external influences. REM sleep is the deepest; however, spontaneous waking occurs most easily. When humans are awakened from REM sleep, dreams are much more easy remembered than when they are awakened from slow wave sleep. On the other hand, definite nightmares occur predominantly in stages three and four of slow wave sleep. the so-called delta sleep.1121Barbiturates and ethanol suppress REM sleep, while benzodiazepines, for instance, reduce the fourth and deepest stage of slow wave sleep to a much greater extent than R E M sleep. Just as unsuccessful as the search for a universal sleep-inducing substance was the search for a single sleep center in the brain.["] Stimulation of neurons at different anatomical locations induced sleep. Selective destruction of these cells led to insomnia in laboratory animals. Neurons of the brain stem that use serotonin as neurotransmitter seem to responsible for induction and maintenance of slow wave sleep. Injection of serotonin into these regions induces sleep, and application of pmuchlorophenylalanine, an inhibitor of serotonin biosynthesis, causes insomnia as expected. However. after one week of daily injection of the inhibitor it is observed that in spite of low serotonin levels sleeping behavior normalizes and reaches 70 % of the control values. Let us return to the relatively simple structure of 1, which might play a role in the complex phenomenon of sleep. Several questions are still unanswered. It remains to be clarified how the substance is metabolized, how the metabolism is regulated. and whether it can be attributed to regions of the brain that are responsible for the induction and maintenance of sleep.['01 The isolation of another long-chain carboxylic acid amide with biological function from brain has attracted attention to this class of compounds. In a screening process, anandamide 4, tee ethanol amide of arachidonic acid has been identified as an endogenous ligand of the cannabinoid receptor ( K , = 52 nM)."41 In 3 functional assay it gave the same physiological response and behavior as A9-tetrahydrocannabinol ( 5 ) . the active con- H 4 y 3 5 stituent of hashish and marijuana. Brain membranes are able to synthesize 4 from ethanolamine and arachidonic acid, which should be accessible from phospholipids of the plasma membrane on action of phospholipase D or A2.['" It remains to be seen whether more than a structural similarity exists between 1 and 4.and further, whether the findings of the authors"'21 will contribute to the answer of the still-open question of how sleep "functions". Since nature, and maybe also sleep, is organized hierarchically, the question of why we need to sleep at all can presumably not be answered on the molecular level alone German version: Atigcw . Clteni. 1995. 107. 2559 2560 Keywords: brain lipids . neurotransmitters sphingosines -~ ( I ] R. A . Lerner. G . Siuzdak. 0 . Prospero-Garcia. S. J. Henriksen. D. L. Boger. B. F Cravatt. f r / J l ' .Nrirl. Ai.ad. S(.i. L'SA 1994. 91. 9505-9508.  B. E Cr:ivatt. 0. Prospero-Garcia. G. Siuzdak. N. B. Gilula. S. J Henriksen. D. L. Boger, R . A. Lerner. Scicwe(, 1995, 168. 1506- 1509.  L. Jaenicke. (%ciriic, Unscrer Zivr 1995. 29. 21. i ) A . Olivera. S. Spiegel. N~orirrr1993.365, 557-560; b) R. Kolesnick. MctI C / i c f i i . ; V ~ w o p i d i i i / .1994. 21. 287 297: c) R. H . Michell. M. J. 0 .Wakelam. Crtrr. Bid. 1994.4. 370 373: d) R. Kolesnick. D. W. Golde. Crll1994. 77, 3 2 5 - 3 2 8 : e) Y. A. Hannun. J. B i d C'hem. 1994. 169. 3125 312X: f) Y A. Hannun. L. M . Oheid. 7i.ctir/.> Biot/ic~nr See. 1995. 111. 7 3 ~ ~ 7 7 . [ 5 ] H. Pieron. Lr P r o h l c h ~P/tj.siologiqitc [ h i S ~ I ~ I I ~ Paris. I P I I , Masson. 1913. cited in ref. (6. 121.  A B. Borbely. I Tobler. Phvsiol. RIY. 1989, 69. 605 670.  G A. Schoenenberger. M. Monnier. P r o ( . Nor/. A c d . Sci. L'SA 1977. 74. 12x2 1286. [XI R. Ueno. Y. lshikawa. T. Nakayam'i. 0. Hayashi, Btoc/ien$. Biophi c. Rr\. ~ ~ ~ J ~ l 1982. ~ ~ ~ llay. l t ~576-l . 582.  0 . Hayaishi. FASEB J. 1991. S. 2575-2581. [lo] W. C. Dement. presented at the International Symposium on Endogenous Sleep Factors. Honolulu. HI. November 7-11, 1988; see also S. Nishino. E. Migot. B. Fruhstorfei-.W. C. Dement. 0 . Hayaishi. Proc. Narl. A d . Sct. L:SA 1989. 86. 2483-2487, cited in ref. 191. (1 I ] .I. S. Takahashi. Armit R P K Nc,urn.sct. 1995, 18. 531 553.  D. D. Kelly in Priwip/r.v ( J / hrcrrrui Sciwiie (Eds.: E. R. Kandel. J. H. Schwarz. T. M. Jessel), Elsevier. Nets York. 1991. Chapter 51, pp. 792- 804 1131 J D. Miller. .4riirri. R ~ Med. I C'lwfi? 1992. 27. 11 -19. (14) W. A . Devane. L. HanuG. A. Breuer. R. Ci. Pertwee. L. S. Stevenson, G. Oriffin. D. Gibson. A . Mandelbaum. A . Etinger, R. Mechoulam. S<knce 1992. 258, 1946 1949. [IS] W, A . Devane. J. Axelrod. Prot-. N d A c d Sci. C:SA 1994, 91. 6698-6701.