Doi:10.1016/s0006-3223(02)01740-7

Diurnal Variation of Cerebrospinal Fluid Hypocretin-1(Orexin-A) Levels in Control and Depressed Subjects Ronald M. Salomon, Beth Ripley, John S. Kennedy, Benjamin Johnson,Dennis Schmidt, Jamie M. Zeitzer, Seiji Nishino, and Emmanuel Mignot Background: Hypocretins, excitatory neuropeptides at
Introduction
monoaminergic synapses, appear to regulate humansleep-wake cycles. Undetectable cerebrospinal fluid hypo- Among pathophysiologic findings in depression, rela- cretin-1 levels are seen in narcolepsy, which is frequently tionships with sleep and circadian abnormalities have associated with secondary depression. Shortened rapid eye movement latency is observed in both narcolepsy and depression. Cerebrospinal fluid hypocretin-1 levels have not been reported in mood disorders. early in the course of the disease include sleep distur- Methods: We examined hypocretin-1 levels in 14 control
and 15 depressed subjects. Cerebrospinal fluid was drawn especially rapid eye movement (REM) deprivation have continuously in supine subjects for 24 hours with an indwelling intrathecal catheter under entrained light-dark conditions. Depressed subjects were studied before and after 5 weeks of sertraline (n ϭ 10, three nonresponders) or bupropion (n ϭ 5, two nonresponders). and decreased REM latency in depression is highly Results: Hypocretin-1 levels varied slightly (amplitude
10%) but significantly across the diurnal cycle in control creased amplitudes of behavioral, physiologic, and neu- subjects, with amplitude significantly reduced in depres- roendocrine circadian measures and disrupted responses of sion (3%). Levels were lowest at midday, surprising for ahypothetically wake-promoting peptide. Mean hypocretin the circadian pacemaker to the light-dark cycle are ob- levels trended higher in depressive than in control sub- jects. Hypocretin-1 levels decreased modestly but signifi- cantly after sertraline (Ϫ14%) but not bupropion. Conclusions: Our results are consistent with previous phys-
sion is associated with diurnal changes in hypothalamic- iologic findings in depression indicating dampened diurnal pituitary-adrenal axis activity and other endocrine abnor- variations in hypocretin-1. The finding that sertraline but not bupropion slightly decreased cerebrospinal fluid hypocre- tin-1 indicates a serotoninergic influence on hypocretin tone. Biol Psychiatry 2003;54:96 –104 2003 Society of Biolog- hypocretin-1 and hypocretin-2, are processed from aprepropeptide encoded by a single gene Key Words: Depression, hypocretin, sertraline, bupro-
tuberal region of the hypothalamus project to the entire neuraxis, including the spinalcord Extremely dense and almost From the Department of Psychiatry (RMS, DS) and Anesthesiology (BJ), Vander- bilt University School of Medicine, Nashville, Tennessee; the Stanford Uni- invariably excitatory projections are noted to aminergic versity Center for Narcolepsy (BR, JMZ, SN, EM), Stanford University School cell groups (e.g., adrenergic locus coeruleus, serotonergic of Medicine, Palo Alto, California; the Department of Psychiatry (JSK),Indiana University School of Medicine and Lilly Research Laboratories, Eli raphe´ nuclei, histaminergic tuberomammillary nucleus, Lilly and Company (JSK), Indianapolis, Indiana.
dopaminergic substantia nigra, and ventral tegmental area; Presented in part in meeting abstract at the annual conference of the Associated Professional Sleep Societies, Seattle, Washington, June 2002.
Address reprint requests to Ronald M. Salomon, Vanderbilt University School of Medicine, Department of Psychiatry, VAV 2200, 1500 21st Ave S, Nashville,TN 37212, or to Emmanuel Mignot, M.D., Ph.D., Stanford University Center for Narcolepsy, 701 Welch Road B, basement room 145, Palo Alto, CA ment links transmitter systems of each of the respective Received April 23, 2002; revised August 7, 2002; accepted September 4, 2002.
nuclei to the pathophysiology of depression (Janowsky and Overstreet 1995; Maes and Meltzer 1995; Schatzberg ter (GCRC) from 1990 to present. Patients showing any sugges- tion of less than optimal cognition and thought processes were excluded because of inability to provide consent. Additionally, seven healthy subjects (four of them male, 27.7 Ϯ 4.86 years old) a disabling disorder characterized by daytime sleepiness, were studied repeatedly on separate days at different hours acrossthe day and night with standard lumbar puncture procedures with cataplexy, and extremely short REM sleep latency the approval of the Stanford University institutional review At the Vanderbilt University site, subjects were recruited by and may consolidate wakefulness and reduce sleep. Mono- newspaper advertisement. Depressed patients were in medically aminergic tone, which is high during wakefulness, de- stable condition, were 18 to 65 years old, had a DSM-IV major creases during sleep and REM sleep and may be driven by depressive episode, and had been free of antidepressant drugs for at least 2 months. Exclusion criteria included high suicidal risk (e.g., previous severe suicide attempt), past or present psychosis, reduce monoaminergic activity, leading to daytime sleep- current tobacco use, and the existence of any other primary iness and abnormally short REM sleep latency psychiatric diagnosis. Subjects were also free of abnormalities on physical examination, electrocardiogram, or extensive laboratory Narcolepsy is frequently associated with depression evaluation (including hepatitis screens and pregnancy testing).
Eligibility required agreement from the subject to comply with all study requirements. All subjects were reimbursed for incon- enhance monoaminergic activity and affect sleep physiol- venience and expenses and income lost, and depressed patients ogy. The hypocretin system is activated by sleep (or REM) received free treatment for 8 study weeks. In the case of the limited number of healthy volunteers studied at Stanford Uni- antidepressant effects Finally, hypocretins versity, subjects were recruited by word of mouth, compensated activate the hypothalamic-pituitary-adrenal axis for their effort, and included only if they were free of any medical illness and psychotropic treatment.
These relationships led us to study the role of hypocretins At both sites the study and potential risks were fully explained in depression. To do so we examined lumbar cerebrospinal and all questions were answered before informed consent was fluid (CSF) hypocretin-1 levels in 14 control and 15 signed, typically after several days of consideration by the depressed subjects (before and after antidepressant ther- subject along with family members and close friends. Discomfort and inconvenience, most notably spinal headache, and potentialrisk for serious adverse outcomes, such as the potential forparalysis or death as a result of treatment-resistant bacterialmeningitis resulting from participating in the CSF collection Methods and Materials
procedure, were all thoroughly discussed.
Fourteen control subjects (six of them male, 41 Ϯ 4 years old [allexpressions are mean Ϯ SD]) and 15 depressive subjects (five of In the continuous 24-hour CSF sampling studies, healthy subjects them male, 39 Ϯ 3 years old) were studied with continuous CSF were studied on only one occasion, whereas depressed subjects sampling. Diagnosis was determined by psychiatric interview were studied before and after 5 weeks of antidepressant therapy and confirmed by the Structured Clinical Interview for DSM-IV with sertraline (n ϭ 10, three nonresponders) or bupropion (n ϭ Axis I Disorders and Structured Clinical Interview for DSM-IV 5, two nonresponders). All subjects consumed a controlled, Axis II Disorders according to DSM-IV balanced monoamine diet provided by the outpatient GCRC for three daily meals for 3 days before each catheterization. During item Hamilton Rating Scale for Depression (HRSD) score was the entire preadmission and inpatient study, subjects were not permitted to consume methylxanthine-containing foods, such as subjects with bipolar type I, four with bipolar type II, and others caffeinated beverages, chocolate, and artichokes. Diets were with unipolar illness or a first depressive episode (one patient).
designed by the nutritionist according to individual taste prefer- On average, patients described 2.5 discrete depressive episodes, ences to maximize compliance. Balanced meals contained ap- although this may be a low estimate because many episodes were proximately 50% carbohydrates, 20% protein, and 30% fat. On described as prolonged and thus may have represented several in admission into the GCRC, meals were given only at 6:30 AM and series without full interepisode recovery, and most patients 9:30 AM, yielding the same total daily intake. All subjects described onset early in life with only recent diagnosis. None had remained strictly supine during the precatheterization (12 hours), illness considered refractory to treatment in previous episodes.
catheterization (48 hours), and postcatheterization (24 hours) The studies were approved by the Vanderbilt University institu- periods and were not allowed to rise for any reason (including tional review board and the committee of the National Institutes being required to toilet in the supine position). Thus the total time of Health–supported Vanderbilt General Clinical Research Cen- that each subject was required to be recumbent in bed was a Table 1. Depressed Patient Protocol and Rating Scale Results For physical symptom checklist, baseline symptoms were rated before first CSF study as 1, none or negligible; 2, mild; 3, moderate; 4, severe. There was no clinically or statistically significant worsening during the study from baseline at any point, although there were frequently significant improvements.
minimum of 84 hours for each of the two sampling periods. After screening and initial GCRC intake scores at week 8 of treatment the 24-hour baseline period reported here, participants received (study week 10), and (2) an HRSD total score at the final week dietary restrictions or medications for other study purposes; 8 assessment of 8 or less. According to these criteria, only five of however, the second 24-hour collection period is not relevant to 15 subjects were nonresponders, three with sertraline and two this report. Patients were encouraged to change position fre- quently. Room lights were turned off at 11:00 PM and back on at6:15 AM before breakfast at 6:30 AM. CSF collection was adapted Monitoring of Side Effects and Adverse Events from a previously reported procedure Forhealthy subjects, intrathecal catheters were placed at 6:00 PM, Weekly outpatient and twice-daily inpatient ratings for study- and collection began at 6:30 PM. For depressed subjects, intra- related adverse events were recorded and analyzed. There was no thecal catheters were in place by 7:30 AM, and collections began statistically or clinically significant worsening relative to base- at 8:00 AM. CSF samples were collected into polypropylene vials line physical symptom checklist, which included headache, held in a chilled fraction collector at 4°C and moved to dry ice at constipation, poor memory, nausea, drowsiness, blurred vision, increased appetite, difficulty starting urination, trouble concen- In a separate study, seven healthy subjects had lumbar punc- trating, nightmares, difficulty sitting still, irregular heartbeat, tures in the early afternoon, the middle of the night, and the early diarrhea, frequent urination, dry mouth, decreased appetite, morning, with samplings being separated by at least 2 weeks.
tremors or shakiness, skin rash, ringing in ears, sweating, fainting Subjects were instructed to continue their normal activity sched- or light-headedness, poor muscle coordination, and muscle stiff- ule and were allowed to sleep in the laboratory in preparation for ness. A total physical symptom checklist score was also gener- their nocturnal and early morning spinal taps. Not all subjects ated and did not vary across the study although a small completed all three studies because of headaches (n ϭ 3) or other improvement was noted in treated depressed subjects (data not intervening circumstances (n ϭ 3).
shown). Importantly, there were also no time points with groupdifferences for headache. Three days after withdrawal of the CSFcatheter, all depressed subjects were free of headache. Most Antidepressant Treatment and Definition of subjects had extremely limited symptom changes through the course of the study. Severe headaches were unusual, any head- As shown in depressed patients were assigned to 8 ache was infrequent, and all postspinal headaches resolved with weeks of open-label, outpatient treatment with either sertraline a local blood patch administered between 24 and 72 hours after (Zoloft, all of the first five patients plus five randomly assigned the catheter removal. A blood patch procedure was offered and of the other 10 depressed subjects) or bupropion (Wellbutrin-XR, performed after seven of the 30 catheterizations that were all five randomly assigned). Antidepressant treatment was initi- ated in the morning of the third day after the completion ofpretreatment lumbar intrathecal catheterization. All patients were free of spinal headaches before beginning therapy and reportedtaking all their medications. Sertraline therapy began with 50 mg CSF hypocretin-1 levels were measured in duplicate with an each morning for 7 days and increased to 100 milligrams each established direct radioimmunoassay in 10-min samples (50 ␮L day for 7 weeks. Bupropion was given as extended release ϫ 2; The CSF hypocretin-1 assay was highly 150-mg tablets each morning for the first 7 days and then 150 reliable, showing stability in samples measured repeatedly from milligrams each morning and each day at 1:00 PM for 7 weeks.
samples kept frozen (Ϫ80°C) for as long as 10 years, after 72 The HRSD was the primary mood outcome measure. An hours at room temperature, or repeatedly thawed and frozen HRSD decrease of 50% or reaching 8 or less defined remission Specificity was shown (r ϭ .99, p Ͻ .001) at 5 weeks. After 8 weeks of treatment, end point remission was by linear correlation of exogenous hypocretin-1 (0 –1200 pg/mL) defined similarly as meeting both a priori criteria: (1) decrease on in 1 mL CSF to the obtained level measurements the HRSD total score of at least 50% from the mean of HRSD Measured levels were at least 15 times greater than Table 2. Sinusoidal Curve Fitting in Control and Depressed Subjects Data are mean Ϯ SD. Times are given as 24-hour clock times.
minimal detectable levels In 11 subjects (six .001). Contrary to an expected increasing level gradient control subjects and five depressed subjects before and after across serial samples that is often reported in monoamine treatment), all samples were assayed. In the remaining eight metabolite measures, levels of hypocretin declined very control subjects and 10 depressed (before and after) subjects, one gradually during the first few hours of sampling (average 8.8% in 3.5 hours). This differs from the monoamineeffect in that our opening fluid samples were reserved for culture and microscopy as part of a safety procedure, and To study overall diurnal variation, data smoothing was per- the hypocretin gradient was observed over a longer period.
formed by averaging 10-min hypocretin-1 values across 2 hour Sinusoid curve-fitting results are indicated in and periods. Data were normalized within subjects by dividing each A 24-hour sine wave accurately predicted the data point by the mean 24-hour hypocretin value of the subject form of the data from control subjects (r2 ϭ .92), with an and then averaging across subjects within treatment groups.
amplitude of 6% and a peak occurring several hours after Group data were fitted with a 24-hour cosine-wave function (y ϭ a ϩ {b ϫ cos[(2␲/24) ϫx ϩ c]}) with a nonlinear, least-squares depressed subjects, both before and after treatment, a sine fitting method (Levenberg-Marquardt algorithm, Microcal Ori- wave poorly predicted the overall oscillation of lumbar gin v.6.0; Microcal Software, Northampton, MA). The phases of hypocretin (r2 ϭ .32 before treatment, r2 ϭ .42 after hypocretin maxima and hypocretin fluctuation amplitude (half treatment), and the amplitude of wave was apparently peak to trough) were estimated for each group.
Daytime and nighttime values were also compared. In each reduced, with nonoverlapping 95% confidence intervals individual subject mean hypocretin levels during the day between control subjects and depressed subjects before or (11 AM– 6:00 PM) and night (11 PM– 6:00 AM) were calculated, after treatment. There may have been a small recovery of together with mean 24-hour individual hypocretin-1 levels. The amplitude in the depressed patients after treatment, though 7-hour day and night windows were selected according to such an effect is not statistically evident from the sine- light-dark schedule (lights were off from 11 PM– 6:00 AM).
fitting data. In an additional limited set of ambulatory Day-night differences in mean levels (11 PM– 6:00 AM minus control subjects who underwent multiple lumbar punctures 11AM– 6:00 PM) for each subject were considered an approxima-tion of the amplitude on each experimental night. We compareddaytime and nighttime hypocretin levels within each of the threesubject groups (two-tailed, paired Student t tests), daytime andnighttime levels, and day-night differences between groups(single-factor ANOVA) and the effect of antidepressant treat-ment (two-factor ANOVA). Statistical analyses were performedon a personal computer with Excel 2000 (v. 9.0.2720; MicrosoftCorporation, Redmond WA), with the exception of the two-factor ANOVAs that were performed with the Java script foundon http://faculty.vassar.edu/lowry/anova2u.html Diurnal Variation of Hypocretin-1 Levels inDepressed and Control Subjects Overall the concentration of hypocretin-1 obtained fromhuman lumbar CSF did not vary significantly with age or Figure 1. Sinusoidal fitted curves to 2-hour interval measure- gender (not shown), in agreement with previous reports. In ments of CSF hypocretin-1 levels from lumbar catheterization in control subjects, hypocretin levels varied very slightly but control (squares), depressed (triangles), and treated (circles) consistently and significantly across the diurnal cycle (p Ͻ Basal Levels of Hypocretin-1 in Depressed andControl Subjects Basal 24-hour, daytime, and nighttime levels did not differbetween supine control subjects and untreated patients, butslightly higher levels were observed in depressed subjects,especially during the day (daytime concentrations in con-trol subjects 251.7 Ϯ 11.5 pg/mL and in pretreatmentdepressed subjects 275.8 Ϯ 11.5 pg/mL, nighttime con-centrations in control subjects 274.0 Ϯ 10.6 pg/mL and inpretreatment depressed 280.9 Ϯ 12.2 pg/mL, mean-24hour concentrations in control subjects 265.0 Ϯ 11.0pg/mL and in pretreatment depressed subjects 281.2 Ϯ11.8 pg/mL, differences not significant; Thesevalues were also similar to previously reported data inambulatory control subjects and markedly higher thanlevels observed in narcolepsy. As suspected from thesinusoid analysis, there was a highly significant differencebetween daytime (11:00 AM– 6:00 PM) and night time(11:00 PM– 6:00 AM) hypocretin-1 concentrations in con-trol subjects (p Ͻ .001, paired two-tailed t test). Theday-night difference was lost in depressive subjects before(p ϭ .32, paired two-tailed t test) or after (p ϭ .24, pairedtwo-tailed t test) treatment.
Effects of Antidepressant Treatment on Hypocretin- Figure 2. Seven healthy volunteers had spinal taps at different times of the day, with five of them having at least two taps atdifferent times of the day (singleton values not shown). Symbols Drug effects were associated with changes in hypocretin connected by solid lines indicate the concentration of hypocre- levels, whereas mood responses were not. Hypocretin tin-1 (Hcrt-1) in the CSF of single spinal taps for each individual.
levels decreased modestly but significantly both overall Multiple samplings in the same individual are connected with with sertraline treatment: (Ϫ14%, p Ͻ .01), both during solid lines. The averages of the samples obtained in the early the day (Ϫ14%, p Ͻ .05) and during the night (Ϫ14%, afternoon, early night, and late night are shown as diamonds, p Ͻ .005) but not with bupropion treatment: (overall with x and y axis error bars representing SEM, and are connected Ϫ3%, p ϭ .59; daytime Ϫ4%, p ϭ .35; nighttime Ϫ1%, by a dashed line. The time of day is indicated on the x axis, and p ϭ .91; all p values from two-tailed, paired t tests; the dark period is approximated by the black bar (10:00 PM– Absolute concentrations of hypocretin during the day, night, or 24-hour sampling period, as well as the differ-ence between day and night (a measure of circadian at different times of day, levels did not vary significantly amplitude), were affected neither by treatment nor out- but were generally slightly higher during the nighttime, come (two-factor ANOVA p Ͼ .38 for all comparisons in agreement with the continuous sampling data set and interactions; Pretreatment to posttreatment changes in hypocretin concentrations during the day, Table 3. Mean Ϯ SEM Hypocretin Levels in Control Subjects and Depressed Patients before and after Treatments There were significant differences (p Ͻ .05) between depressed and control groups and in depressed group before and after treatment as indicated.
aSignificant difference between depressed and control group by t test.
bSignificant difference between before and after treatment by paired t test.
Table 4. Mean Ϯ SEM Hypocretin Levels (in picograms per milliliter) in Depressed Patients and Treatment Responses aSignificant (p Ͻ .05, paired t test) effects from pretreatment to posttreatment.
bSignificant (p Ͻ .05, paired t test) drug effects from pretreatment to posttreatment.
night, or 24-hour sampling period were also not affected part of the active day period, just before the transition to by treatment or outcome (two-factor ANOVA p Ͼ .05 for tion that hypocretin may be an important wake promotingsignal, opposing sleep debt in the second part of the activephase Discussion
In our human lumbar CSF study, highest values This study presents data from subjects confined to bed rest were observed around 2:00 AM (3 hours after lights out), for 12 hours before the lumbar puncture and also from which is several hours later than in the squirrel monkey.
subjects allowed ad libitum activity until the time of the Furthermore, changes were of much smaller magnitude lumbar puncture. Under condition of constant bed rest, than in the nonhuman mammalian studies. This temporal continuously sampled (for 24 hours) CSF hypocretin-1 difference most likely reflects an anatomically delayed levels fluctuated moderately but significantly across the and dilution-dampened oscillation of cortical hypocretin, 24-hour cycle (Ͻ15%; Similar variations were with actual brain fluctuations being more consistent with observed in the small number of subjects in the individual the monkey data. Consistent with this hypothesis, studies tap study indicating that normal activity did not have shown a 90- to 120-minute delay in equilibration dramatically alter this profile. The observation that fluc- between higher CSF compartments and the lumbar sac tuations in levels across the 24-hour period were much smaller than differences observed between groups of Intense projections from the hypothalamus to the spinal healthy and narcoleptic patients is of practical importance.
cord have been reported (van den Pol 1999), but even if Lower CSF hypocretin levels (hypocretin-1 levels Ͻ110 spinal release occurs, the cell bodies located in the lateral pg/mL) have been shown to be diagnostic for narcolepsy hypothalamic area are likely to be active during the active period. This would not explain the discrepancy between indicate that the time of day at which sampling is done is unlikely to interfere with this test. The data should be Hypocretins are uniquely positioned for involvement in considered preliminary because of the relatively small depression. Whereas hypocretin cell bodies are all local- number of patients and the differences in times of collec- ized within the perifornical area, extremely dense, almost tion initiation between control subjects and depressed invariably excitatory projections are noted in aminergic cell groups (e.g., adrenergic locus coeruleus, serotonergic Surprisingly, the direction of the hypocretin diurnal raphe´ nuclei, histaminergic tuberomammillary nucleus, variation observed in the CSF was opposite to that dopaminergic substantia nigra, and ventral tegmental area expected for a wake-promoting peptide. Highest levels were observed in the middle of the night, while subjects were mostly asleep. In rats, both cisternal CSF have reported higher levels in the latter portion of hypocretin system in narcolepsy was first demonstrated in the active phase (dark period). In diurnal squirrel monkeys (Saimiri sciureus), cisterna magna CSF levels of hypocre- Hypocretin deficiency causes human narcolepsy tin (ranging from 170 – 430 pg/mL, similar to the human samples) were also found to be highest during the latter disabling disorder characterized by daytime sleepiness, cataplexy, and extremely short REM sleep latency with decreased hypocretin levels, suggesting a small but significant serotonergic influence on hypocretin tone after 5 weeks of treatment. The finding that bupropion, a Hypocretin release is higher during the active phase in rats dopaminergic and adrenergic reuptake blocker with sig- not modify hypocretin levels was rather surprising. The Whereas decreased hypocretin tone in narcolepsy is fact that sertraline is a more potent REM-suppressing associated with depression, our data indicate no dramatic decrease in baseline CSF hypocretin values in depression.
significant in explaining this difference. Alternatively, If anything, slightly higher hypocretin values were found changes caused by bupropion may be observed either in this small number of depressed subjects This acutely, after initial dosing, or more delayed in recovery.
finding indicates that hypocretin deficiency is an unlikely Stimulant medications such as amphetaminelike dopamine cause for depression. Slightly increased hypocretin levels releasing agents and reuptake inhibitors have not dramat- in depression may rather represent disturbed sleep and ically modified CSF hypocretin levels in patients with activity in depression or compensatory mechanisms. Stud- ies with larger numbers are, however, needed to expand on acute changes in hypocretin are observed in preclinical studies of dopamine releasing agents. Studies including In contrast with the lack of striking baseline hypocretin narcoleptic subjects treated with serotonin reuptake inhib- level differences, we found significantly decreased diurnal itors may be needed to study this effect.
variation in depressed subjects versus control subjects.
In conclusion, our studies of CSF hypocretin-1 levels in These results are consistent with previous findings in depression indicate that a decrease in mean hypocretin depression indicating that diurnal physiologic measures release is not a likely cause of depression. In control subjects, CSF hypocretin-1 levels were found to vary slightly but significantly across the 24-hour study period, source of the observed diminution in signal amplitude with higher levels observed at night. In depression, re- cannot be ascribed definitively to any of the involved sites.
duced amplitude of diurnal variation was observed. Addi- This observation does, however, provide evidence in tional studies in depression are needed to expand on these depression that diminished circadian rhythms of behav- iors, physiologic measures, and peripheral neuroendocrinefunctions can also be observed centrally. Evidence sug-gests dampened monoamine metabolite fluctuation in Supported by grants from the National Institutes of Health (Grants NS depression (Salomon et al 1998, unpublished data). Sleep 23724, NS 33797, and MH40041) to EM, by an investigator-initiatedgrant from Pfizer, by grants from National Alliance for Research in Schizophrenia and Depression and the Theodore and Vada Stanley caused by decreased hypocretin fluctuation. Interestingly, Foundation to RMS, and by a General Clinical Research Grant from the however, relief of depressive symptoms was not correlated National Institutes of Health National Center for Research Resources with restored diurnal rhythmicity in this small sample.
(Grant MO1RR00095) to Vanderbilt University Medical Center, Nash- Still, a slight improvement in amplitude was noted in We thank our research assistants, Linda Todd, Barbee Smith, and treated subjects. This is highly vulnerable to error because Kerry Hook, the General Clinical Research Center nursing staff, and all of the small sample size and the relatively mild to moderate depressive episodes in these patients and thuswill require replication; however, our results suggest thatdepressive mood and diurnal variation in hypocretins may References
be independent characteristics. Chronic treatment studies American Psychiatric Association (1994): Diagnostic and Sta- in depression may lead to more significant changes in tistical Manual of Mental Disorders, 4th ed. Washington, DC: hypocretin diurnal variation, especially if the restoration of sleep patterns from direct soporific effects is avoided and Arborelius L, Owens MJ, Plotsky PM, Nemeroff CB (1999): The the delayed normalization of sleep is related to peptidergic role of corticotropin-releasing factor in depression and anxi-ety disorders. J Endocrinol 160:1–12.
Benca RM, Obermeyer WH, Thisted RA, Gillin JC (1992): Sleep We also explored whether antidepressant treatment and psychiatric disorders. A meta-analysis. Arch Gen Psychi- modified CSF hypocretin levels in patients with depres- sion. A possible pharmacologic effect not related to Beersma DG, Van den Hoofdakker RH, Van Berkestijn JW antidepressant response was observed. We found that (1983): Circadian rhythms in affective disorders. Body tem- treatment with sertraline but not bupropion was associated perature and sleep physiology in endogenous depressives. In: Van Praag HM, Mendlewicz J, editors. Advances in Biolog- Jaszberenyi M, Bujdoso E, Pataki I, Telegdy G (2000): Effects of ical Psychiatry, 11. Basel: Karger, 114 –127.
orexins on the hypothalamic-pituitary-adrenal system. J Neu- Beuckmann CT, Yanagisawa M (2002): Orexins: From neu- roendocrinol 12:1174 –1178.
ropeptides to energy homeostasis and sleep/wake regulation.
Kennedy JS, Polinsky RJ, Johnson B, Loosen P, Enz A, Laplanche R, et al (1999): Preferential cerebrospinal fluid Borbely AA, Wirz-Justice A (1982): Sleep, sleep deprivation and acetylcholinesterase inhibition by rivastigmine in humans.
depression: A hypothesis derived from a model of sleep J Clin Psychopharmacol 19:513–521.
regulation. Hum Neurobiol 1:205–210.
Kilduff TS, Peyron C (2000): The hypocretin/orexin ligand- Chemelli RM, Willie JT, Sinton CM, Elmquist JK, Scammell T, receptor system: Implications for sleep and sleep disorders.
Lee C, et al (1999): Narcolepsy in orexin knockout mice: Trends Neurosci 23:359 –365.
Molecular genetics of sleep regulation. Cell 98:437–451.
Kleitman N (1939): Sleep and Wakefulness. Chicago: University Daniels E, King MA, Smith IE, Shneerson JM (2001): Health- related quality of life in narcolepsy. J Sleep Res 10:75–81.
Kripke DF, Mullaney DJ, Atkinson MS, Wolf C (1987): Circa- de Lecea L, Kilduff TS, Peyron C, Gao X, Foye PE, Danielson dian rhythm disorders in manic-depressives. Biol Psychol PE, et al (1998): The hypocretins. Hypothalamus-specific peptides with neuroexcitatory activity. Proc Natl Acad Sci Kupfer DJ, Ehlers CL, Frank E, Grochocinski VJ, McEachran AB (1991): EEG sleep profiles and recurrent depression. Biol Detre TP, Himmelhoch JM, Swartzburg M, Anderson CM, Byck R, Kupfer DJ (1972): Hypersomnia and manic-depressive Kuru M, Ueta Y, Serino R, Nakazato M, Yamamoto Y, Shibuya disease. Am J Psychiatry 128:1303–1305.
I, et al (2000): Centrally administered orexin/hypocretin Di Chiro G, Hammock MK, Bleyer WA (1976): Spinal descent activates HPA axis in rats. Neuroreport 11:1977–1980.
of cerebrospinal fluid in man. Neurology 26:1–8.
Lin L, Faraco J, Li R, Kadotani H, Rogers W, Lin X, et al (1999): Duncan WC, Johnson KA, Sutin E, Wehr TA (1998): Disruption The sleep disorder canine narcolepsy is caused by a mutation of the activity-rest cycle by MAOI treatment. Dependence on in the hypocretin (orexin) receptor 2 gene. Cell 98:365–376.
light and a secondary visual pathway to the circadian pace- Maes M, Meltzer HY (1995): The serotonin hypothesis of major maker. Brain Res Bull 45:457–465.
depression. In: Bloom F, Kupfer D, editors. Psychopharma- Estabrook IV, McCarthy MT, Ko E, Chou TC, Chemelli RM, cology. The Fourth Generation of Progress. New York: Yanagisawa M, et al (2001): Fos expression in orexin neurons varies with behavioral state. J Neurosci 21:1656 –1662.
Marcus JN, Aschkenasi CJ, Lee CE, Chemelli RM, Saper CB, First MB, Spitzer RL, Gibbon M, Williams JB (1996a): Struc- Yanagisawa M, et al (2001): Differential expression of orexin tured Clinical Interview for DSM-IV Axis I Disorders (SCID- receptors 1 and 2 in the rat brain. J Comp Neurol 435:6 –25.
I/P), 2nd ed. New York: Biometrics Research Department, Mazure C, Nelson JC, Price LH (1986): Reliability and validity New York State Psychiatric Institute.
of the symptoms of major depressive illness. Arch Gen First MB, Spitzer RL, Gibbon M, Williams JB, Benjamin L (1996b): Structured Clinical Interview for DSM-IV Axis II Mignot E (2001): A commentary on the neurobiology of Disorders (SCID-II), 2nd ed. New York: Biometrics Research the hypocretin/orexin system. Neuropsychopharmacology Department, New York State Psychiatric Institute.
Fujiki N, Yoshida Y, Ripley B, Honda K, Mignot E, Nishino S Mignot EM, Lammers GJ, Ripley B, Okun M, Nevsimalova S, (2001): Changes in CSF hypocretin-1 (orexin A) levels in rats Overeem S, et al (2002): The role of CSF hypocretin across 24 hours and in response to food deprivation. Neuro- measurement in the diagnosis of narcolepsy and hypersomnia.
Gillin JC, Buchsbaum M, Wu J, Clark C, Bunney W Jr (2001): Nishino S, Mao J, Sampathkumaran R, Shelton J, Mignot E Sleep deprivation as a model experimental antidepressant (1998): Increased dopaminergic transmission mediates the treatment: Findings from functional brain imaging. Depress wake-promoting effects of CNS stimulants. Sleep Res Online Healy D (1987): Rhythm and blues. Neurochemical, neurophar- Nishino S, Ripley B, Overeem S, Lammers GJ, Mignot E (2000): macological and neuropsychological implications of a hy- Hypocretin (orexin) deficiency in human narcolepsy. Lancet pothesis of circadian rhythm dysfunction in the affective disorders. Psychopharmacology 93:271–285.
Nishino S, Ripley B, Overeem S, Nevsimalova S, Lammers GJ, Holsboer F (2001): Stress, hypercortisolism and corticosteroid Vankova J, et al (2001): Low cerebrospinal fluid hypocretin receptors in depression. Implications for therapy. J Affect (orexin) and altered energy homeostasis in human narcolepsy.
Hungs M, Mignot E (2001): Hypocretin/orexin, sleep and nar- Nowell PD, Buysse DJ (2001): Treatment of insomnia in patients colepsy. Bioessays 23:397–408.
with mood disorders. Depress Anxiety 14:7–18.
Janowsky SJ, Overstreet DH (1995): The role of acetylcholine Peyron C, Faraco J, Rogers W, Ripley B, Overeem S, Charnay Y, mechanisms in mood disorders. In: Bloom F, Kupfer D, et al (2000): A mutation in a case of early onset narcolepsy editors. Psychopharmacology. The Fourth Generation of and a generalized absence of hypocretin peptides in human Progress. New York: Raven Press, 945–956.
narcoleptic brains. Nat Med 6:991–997.
Peyron C, Tighe DK, van den Pol AN, de Lecea L, Heller HC, Van den Hoofdakker RH (1994): Chronobiological theories of Sutcliffe JG, et al (1998): Neurons containing hypocretin nonseasonal affective disorders and their implications for (orexin) project to multiple neuronal systems. J Neurosci treatment. J Biol Rhythms 9:157–183.
Van den Hoofdakker RH, Beersma DG (1988): On the contri- Reynolds CF, Gillin JC, Kupfer DJ (1987): Sleep and affective bution of sleep wake physiology to the explanation and the disorders. In: HY Meltzer, editor. Psychopharmacology: The treatment of depression. Acta Psychiatr Scand Suppl 341:53– Third Generation of Progress. New York: Raven Press, 647– van den Pol AN (1999): Hypothalamic hypocretin (orexin).
Ringel BL, Szuba MP (2001): Potential mechanisms of the sleep Robust innervation of the spinal cord. J Neurosci 19:3171– therapies for depression. Depress Anxiety 14:29 –36.
Ripley B, Overeem S, Fujiki N, Nevsimalova S, Uchino M, Yesavage J, et al (2001): CSF hypocretin/orexin levels in Willie JT, Chemelli RM, Sinton CM, Yanagisawa M (2001): To narcolepsy and other neurological conditions. Neurology eat or to sleep? Orexin in the regulation of feeding and wakefulness. Annu Rev Neurosci 24:429 –458.
Russell SH, Small CJ, Dakin CL, Abbott CR, Morgan DG, Willner P (1995): Dopaminergic mechanisms in depression. In: Ghatei MA, et al (2001): The central effects of orexin-A in Bloom F, Kupfer D, editors. Psychopharmacology. The the hypothalamic-pituitary-adrenal axis in vivo and in vitro in Fourth Generation of Progress. New York: Raven Press, male rats. J Neuroendocrinol 13:561–566.
Sakurai T, Amemiya A, Ishii M, Matsuzaki I, Chemelli RM, Winokur A, Gary KA, Rodner S, Rae-Red C, Fernando AT, Tanaka H, et al (1998): Orexins and orexin receptors: A Szuba MP (2001): Depression, sleep physiology, and antide- family of hypothalamic neuropeptides and G protein-coupled pressant drugs. Depress Anxiety 14:19 –28.
receptors that regulate feeding behavior. Cell 92:573–585.
Wirz-Justice A (1995): Biological rhythms in mood disorders. In: Sakurai T, Moriguchi T, Furuya K, Kajiwara N, Nakamura T, Bloom F, Kupfer D, editors. Psychopharmacology. The Yanagisawa M, et al (1999): Structure and function of human Fourth Generation of Progress. New York: Raven Press, prepro-orexin gene. J Biol Chem 274:17771–17776.
Schatzberg A, Schildkraut JJ (1995): Recent studies on norepi- Wu JC, Bunney WE (1990): The biological basis of an antide- nephrine systems in mood disorders. In: Bloom F, Kupfer D, pressant response to sleep deprivation and relapse. Review editors. Psychopharmacology. The Fourth Generation ofProgress. New York: Raven Press, 911–920.
and hypothesis. Am J Psychiatry 147:14 –21.
Shelton RC, Winn S, Ekhatore N, Loosen PT (1993): The effects Yoshida Y, Fujiki N, Nakajima T, Ripley B, Matsumura H, of antidepressants on the thyroid axis in depression. Biol Yoneda H, et al (2001): Fluctuation of extracellular hypocre- tin-1 (orexin A) levels in the rat in relation to the light-darkcycle and sleep-wake activities. Eur J Neurosci 14:1075– Steiner M, Werstiuk ES, Seggie J (1987): Dysregulation of neuroendocrine crossroads: Depression, circadian rhythmsand the retina—a hypothesis. Prog Neuropsychopharmacol Young EA, Haskett RF, Murphy-Weinberg V, Watson SJ, Akil Biol Psychiatry 11:267–278.
H (1991): Loss of glucocorticoid fast feedback in depression.
Taheri S, Zeitzer JM, Mignot E (2002): The role of hypocretins Arch Gen Psychiatry 48:693–699.
(orexins) in sleep regulation and narcolepsy. Ann Rev Neu- Zeitzer JM, Buckmaster CL, Parker KJ, Hauck CM, Lyons DM, Mignot E (2002): Diurnal variation of hypocretin-1 in the Thannickal TC, Moore RY, Nienhuis R, Ramanathan L, Gulyani cisternal cerebrospinal fluid of a diurnal sleep-consolidating S, Aldrich M, et al (2000): Reduced number of hypocretin primate, Saimiri sciureus. Soc Res Biol Rhythms 8th Annual neurons in human narcolepsy. Neuron 27:469 –474.

Source: http://www.geneqol-consortium.org/wp-content/uploads/sites/9/2013/07/Salomon_Orexin-diurnal-variation-and-depression_Biol-Psychiatry-2003.pdf