SLEEP PHYSIOLOGY Prostaglandin D Synthase (β-trace) in Healthy Human Sleep Wolfgang Jordan, MD1; Hayrettin Tumani, MD, PhD2; Stefan Cohrs, MD1; Sebastian Eggert1; Andrea Rodenbeck, DSc1; Edgar Brunner, MD, PhD3; Eckart Rüther, MD, PhD1; Göran Hajak, MD, PhD1,4 1Department of Psychiatry and Psychotherapy, University of Göttingen; 2Department of Neurology, Universitiy of Ulm; 3Department of Medical Statistics, University of Göttingen; 4Department of Psychiatry and Psychotherapy, University of Regensburg, Germany Study Objectives: The prostaglandin D system plays an important role in Results: Serum L-PGDS concentrations showed marked time-dependent animal sleep. In humans, alterations in the prostaglandin D system have changes with evening increases and the highest values at night (P< been found in diseases exhibiting sleep disturbances as a prominent .0005). This nocturnal increase was suppressed during total sleep depri- symptom, such as trypanosoma infection, systemic mastocytosis, bacteri- vation (P< .05), independent of external light conditions and melatonin Downloaded from https://academic.oup.com/sleep/article/27/5/867/2708439 by guest on 28 September 2021 al meningitis, major depression, or obstructive sleep apnea. Assessment secretion. Rapid eye movement sleep deprivation had no impact on cir- of this system’s activity in relation to human physiologic sleep was the tar- culating L-PGDS levels. get of the present study. Conclusions: The circadian L-PGDS pattern and its suppression by total Design: Serum concentrations of lipocalin-type prostaglandin D synthase sleep deprivation indicate an interaction of the prostaglandin D system (L-PGDS, former β-trace), and plasma levels of the pineal hormone mela- and human sleep regulation. L-PGDS measurements may well provide tonin were measured in 20 healthy humans (10 women, 10 men; aged: new insights into physiologic and pathologic sleep regulation in humans. 23.3 ± 2.39 years) at 4-hour intervals over a period of 5 days and nights, Abbreviations: CSF,cerebrospinal fluid; L-PGDS, lipocalin-type which included physiologic sleep, rapid eye movement sleep deprivation, prostaglandin-D-synthase; PGD2, prostaglandin D2; REM, rapid eye and total sleep deprivation. In addition, the serum L-PGDS and plasma movement; SWS, slow wave sleep melatonin levels of 6 subjects were determined under conditions of bright Key Words: Lipocalin-type prostaglandin D synthase (EC 5.3.99.2), β- white (10,000 lux) or dark red light (< 50 lux) in a crossover design during trace, PGD2, melatonin, human physiologic sleep regulation, sleep depri- total sleep deprivation. Nocturnal blood sampling was performed by a vation, hypersomnia through-the-wall tube system. L-PGDS was measured by an automated Citation: Jordan W; Tumani H; Cohrs S et al. Prostaglandin D synthase immunonephelometric assay, and melatonin was analyzed by direct (β-trace) in healthy human sleep. SLEEP 2004;27(5):867-874. radioimmunoassay. INTRODUCTION Additionally, the impact of the prostaglandin D system on sleep regulation could also be demonstrated in primates.17 In humans, PROSTAGLANDINS ARE UBIQUITOUSLY DISTRIBUTED the prostaglandin D system seems to be involved in several dis- IN VIRTUALLY ALL MAMMALIAN TISSUES AND eases presenting hypersomnia as a prominent symptom, such as ORGANS.Prostaglandin D (PGD ) is unique among the 2 2 African sleeping sickness,18 systemic mastocytosis,19 or severe prostaglandins in being present in relatively high concentrations obstructive sleep apnea syndrome.20 Alterations in the in the mammalian brain. Among the 3 enzymes catalyzing the prostaglandin D system have been also associated with bacterial conversion of prostaglandin H2 to PGD2, the lipocalin-type meningitis21 or major depression,22,23 which are usually accom- prostaglandin D synthase (L-PGDS) (Enzyme Commission panied by sleep disturbances. However, the prostaglandin D sys- Number: EC 5.3.99.2) is responsible for the biosynthesis of tem has not yet been investigated in human physiologic sleep. PGD2 in the brain. L-PGDS itself is mainly synthesized in the Therefore, serum L-PGDS and plasma melatonin, a known deter- choroid plexus, leptomeninges, and oligodendrocytes of the cen- minant of the sleep-wake rhythm, were simultaneously measured tral nervous system and is secreted into the cerebrospinal fluid in healthy subjects during physiologic sleep and in conditions of (CSF).1,2 Based on the results of amino-acid sequencing, L- total and rapid eye movement (REM) sleep deprivation. PGDS has been identified as β-trace.3,4 β-trace is the second most abundant protein in human CSF after albumin, representing 8% METHODS of the total CSF protein.4 A 32:1 CSF-serum concentration gradi- ent indicates that the β-trace protein in the serum originates pre- Study Subjects 5 dominantly from the brain. The study was approved by the local Research Ethical A large number of studies has demonstrated a crucial role for Committee. A total of 20 medical students (10 women, 10 men; 2,6-16 the prostaglandin D system in rodent sleep regulation. mean age ± SD 23.3 ± 2.39 years) were included in the study after obtaining their written informed consent. All were nor- Disclosure Statement motensive and found to be healthy by medical history and phys- No significant financial interest/other relationship to disclose. ical examination. In particular, they had no history of smoking, drug intake, affective disorder, allergy, hypertension, or pul- Submitted for publication September 2003 monary, renal, neurologic, or cerebrovascular disease. Routine Accepted for publication February 2004 laboratory tests included blood cell count, coagulating properties, Address correspondence to: Dr. W. Jordan, Department of Psychiatry and electrolytes, liver enzymes, creatinine, and urea. Normal renal Psychotherapy, Georg-August-University of Göttingen, von Sieboldstr. 5, 37075 function was assessed by serum and urine creatinine and noctur- Göttingen, Germany; Tel: 49 551 396761; Fax: 49 551 393887; nal creatinine clearance. Absence of any inflammation was con- E-mail: [email protected] SLEEP, Vol. 27, No. 5, 2004 867 Prostaglandin D Synthase (β-trace) in Healthy Human Sleep—Jordan et al firmed by laboratory results. Screening of volunteers involved Behring Nephelometer Analyzer (BNA). The test kit employs blood-pressure determination, electrocardiogram, electroen- affinity-purified polyclonal rabbit antibodies coupled to latex cephalogram, and polysomnography. Subjects with more than 10 particles and is sensitive enough (analytical sensitivity of 0.05 sleep-related breathing disturbances or more than 5 periodic leg mg/L) to reliably measure serum concentrations. The interassay movements per hour were excluded from the investigation. coefficient of variation of a control sample was 4.7% at 0.262 ± During the entire study period, subjects stayed in the sleep lab- 0.012 mg/L. The validity data of the L-PGDS assay have been oratory and were given a protein-carbohydrate balanced lunch reported previously.20,21 and a light morning and evening meal, which was a standardized Plasma melatonin concentrations were measured by direct feature of the study. The mealtimes remained unchanged during radioimmunoassay using a highly specific antibody (GS 704- total sleep deprivation; this included overnight fasting until after 6483) from Guildhay antisera (Guildford, United Kingdom). The the morning blood sampling. detection limit of this assay (at which 5% of the ligand is dis- placed) was 3.33 pg/mL. Intraassay and interassay variance were Study Design 6% and 12%, respectively.26 Concentrations of serum L-PGDS and plasma melatonin were Downloaded from https://academic.oup.com/sleep/article/27/5/867/2708439 by guest on 28 September 2021 Analysis measured in 20 healthy subjects at 4-hour intervals (4:00 PM, 8:00 PM, midnight, 4:00 AM, 8:00 AM, and noon) over a period Sleep stages and arousal data were manually determined in 30- of 5 days and nights covering physiologic sleep (baseline), REM second epochs and added to the data stored on a personal com- sleep deprivation, and total sleep deprivation, as well as a recov- puter system. Standard parameters included time in bed, sleep ery night between the nights of sleep deprivation. Sleep was period time (time from sleep onset to final awakening), total polysomnographically monitored every night except for the night sleep time (sleep period time minus time spent awake), sleep effi- of total sleep deprivation. Overnight cardiorespiratory ciency (total sleep time divided by the time in bed), percentage polysomnography followed international guidelines using stan- wake of sleep period time, sleep latencies to stage 2 sleep and dard methods.24,25 All recordings were made on a polysomno- slow wave sleep (SWS), REM latency (time from first epoch of graph (Nihon Kohden; Tokyo, Japan) and simultaneously stored stage 2 sleep until the first epoch of REM sleep), percentage of on a personal computer system (Sagura Polysomnograph 2000, sleep stages (stage 1, stage 2, SWS, REM sleep) of total sleep Sagura Medizintechnik GmbH, 63165 Mühlheim, Germany). time, REM sleep duration, and total number of transitions from Each session started with an adaptation night with the complete sleep stages. In order to get more information about the efficien- montage for polysomnographic recordings, followed by baseline cy of the REM sleep deprivation induced by acoustic stimuli, sleep during which sequential blood sampling started. Nocturnal REM sleep duration per number of transitions were