- Email: [email protected]
Assessment of diurnal variation of cerebrospinal fluid tryptophan and 5-hydroxyindoleacetic acid in healthy human females
Life
scicnea,vol.
60, No. 12 pp. wmm7,1997 r*lu?eluvierseiacc1nc. -_. PrintcdinthcUSA. AMrightsremcd am-324x/97 $1761)t .oo 7
PII SOOU-3205(97)ooo21-0
ELSEVIER
OF DIURNAL VARIATION OF CEREBROSPINAL FLUID TRYPTOPHAN AND 5-HYDROXYINDO LEACETIC ACID IN IEALTHY HUMAN FEMALES
ASSESSMENT
Paul D. Kirwin,l*4, George M. Anderson.2, Phillip B. Chappel12, Lloyd Saberski?, James F. Leckman,l~~, Thomas D. Geracioti,s, George R. Heninger,1*4, Lawrence H. Price,1*4, Christopher J. McDougle,l*2*4 Departments of Psychiatty,t Child Study Center,2 and Anesthesiology3 Yale University School of Medicine Clinical Neuroscience Research Unit4 Abraham Ribicoff Research Facilities Connecticut Mental Health Center DepartmentaZPsychiatry~ University of Cincinnati School of Medicine Veterans Administration Medical Center (Received in final fom January 6,1997)
Summary The role of serotonin (5-HT) in the pathogenesis and treatment of major neuropsychiatric disorders, including mood and anxiety disorders, continues to be the subject of extensive research. Previous studies examinin g central 5-I-IT functioning measured cerebrospinal fluid (CSF) levels of 5hydroxyindoleacetic acid (5-HIAA) by using single or multiple lumbar punctures. A number of investigators have demonstrated the feasiblity of continuous CSF sampling via an indwelling lumbar catheter to study CSF neurochemistry in healthy subjects and patients with neuropsychiatric illness. Four healthy female volunteers, aged 21-34 years, underwent continuous CSF sampling. CSF was collected at a constant rate of 1 ml every 10 minutes over a 30-hour period, with levels of tryptophan (TRP) and 5-HIAA measured every hour. Plasma was also obtained hourly for TRP determination. The results of this study indicate that CSF 5-I-I&4, CSF TRP, and plasma TRP levels showed variation over time, but failed to show diurnal fluctuation. Intraindividual coefficients of variationdetermined for CSF 5-HIAA, CSF TRP, and plasma TRP ranged from 9.2 to 14.9%, 8.8 to 14.6%, and 14.7 to 19.0%, respectively. Continuous CSF sampling is safe and feasible in humans, and may prove useful for studies of central 5-I-II neurotransmission in neuropsychiatric illness. key Wd:
cerebrospinal fluid, serotonin, neuropsychiatric disorders, Shydroxyindoleacetic acid
The role of serotonin (5-hydroxytryptamine [5HT]) in the pathogenesis and treatment of major neuropsychiatric disorders, including mood (1) and anxiety (2) disorders, continues to be the subject of intensive research. Investigators have taken two primary approaches to assess central 5HT functioning in patients. Numerous studies have utilized 5-I-H precursors and receptor agonists and antagonists as pharmacological probes to explore hypothesized 5-I-IT dysfunctions (3). Corresponding Author: Christopher J. McDougle, M.D., Connecticut Mental Health Center, Clinical Neuroscience Research Unit, Room 333B, 34 Park Street, New Haven, CT 06519, Phone: (203) 789-7333, FAX: (203) 789-7651, E-mail: [email protected]
900
Hourly Human CSF 5-HIAA and Tryptophan
Vol. 60, No. 12, 1997
However, the most commonly employed paradigm for assessing the S-HT system has involved the measurement of cerebrospinal fluid (CSF) levels of the principle 5-HT metabolite 5hydroxyindoleacetic acid (5-HIAA). Nearly all of the studies of CSF 5-HIAA have examined metabolite concentrations in single lumbar CSF samples. This cross-sectional approach does not address the possible rhythmic secretion of neurochemicals into the CSF (45). Nor does it allow for an accounting of the changes which might occur in response to the stress of the lumbar puncture itself (6). The continuous sampling of CSF offers a possible strategy for dealing with these issues. Continuous lumbar spinal drainage has been safely used by neurosurgeons and anesthesiologists to administer spinal anesthesia, to treat CSF fistulas, and to control elevated ventricular pressure (7,8). In 1988, Bruce and Oldfield (9) described a modification of this technique which provides a less traumatic means of performing sequential CSF sampling. A number of investigators have now employed the sequential CSF sampling strategy to study CSF neurochemistry in healthy subjects and patients with neuropsychiatric illness. Investigations to characterize CSF diurnal variation of melatonin in neurosurgical patients and healthy volunteers (lo), corticotropin-releasing hormone (CRH) in healthy volunteers (1 l), oxytocin in neurosurgical patients (12) and healthy volunteers (13), and norepinephrine (NE) in healthy volunteers (14) have been conducted. In addition, sequential sampling of CSF for shorter periods of time have examined NE in depressed patients and normal controls (15), cholecystokinin in depression and alcoholism (16) CRH in depressed patients and healthy volunteers (17), and CRH, NE, 3-methoxy-4-hydroxyphenylglycol (MHPG), 5-HIAA, and L-tryptophan (TRP) in alcoholic patients and healthy controls (18). The purpose of the current study was to characterize the diurnal variation of CSF 5-HIAA and the amino acid precursor of 5-HT, TRP, in healthy human subjects. Four healthy adult female subjects underwent continuous CSF sampling over a period of 30 hours via an indwelling lumbar catheter. We determined hourly levels of CSF 5-HIAA and TRP, and compared the CSF measures to plasma levels of TRP measured over the same period.
Methods Subjects The study was approved by the Human Investigation Committee of the Yale University School of Medicine. Four female subjects, free from medical or psychiatric illness, gave voluntary, written informed consent to participate as paid volunteers in the study. The subjects ranged in age from 2 1 to 34 years (mean& S.D.=26.5 f 5.5 years) and weighed 54 to 67.5 kg (mean f S.D.=59.1 + 6.5 kg). Subjects were accepted into the study only if they had no prior history of medical, neurologic, or psychiatric illness. Normal laboratory values of electrolytes, serum creatinine and blood urea nitrogen, complete blood count and differential, clotting factors, thyroid indices, liver function tests, fasting glucose, and a negative serum pregnancy test were required prior to participation in the study. Normal urinalysis and a negative urine drug toxicology screen were also required. In addition, all subjects received a thorough physical examination, including complete neurologic and fundoscopic exam. All subjects had normal electrocardiograms before entering the study. Although information about the specific phase of the menstrual cycle was not obtained, none of the subjects were actively menstruating at the time of the study. Cerebrospinal Fluid And Plasma Collection Subjects followed a low-monoamine diet for three days prior to entering the General Clinical Research Center at Yale-New Haven Hospital for the study. Subjects continued on this diet, consisting of meals at 0800 hr (307 mg TRP), 1200 hr (727 mg TRP), and 1700 hr (811 mg TRP), until the 30-hour CSF sampling ceased. Subjects remained in the hospital for a total of four days. They were admitted to the hospital at 1600 hr (Day 1) on the day prior to the lumbar puncture to control diet and exercise, and to insure adequate intravenous (I.V.) hydration with 0.9% normal saline, which ran at 250 cc per hour from 2000 hr (Day 1) until 1700 hr (Day 4). At 0800 hr (Day 2), after a night of bedrest and I.V. hydration, a catheter was placed in the lumbar subarachnoid space in a modification of the method of Bruce and Oldfield (9) previously
Vol. 60, No. 12, 1997
Hourly Human CSF 5-HIAA and Tryptophan
901
described by Geracioti et al. (15-18). Subjects were placed in the lateral decubitus pOSitiOn in their hospital bed and, after application of intradermal lidocaine anesthesia, an 18-gauge Touhy needle was inserted through the L3-L4 or LA-L5 interspace. After entry into the subarachnoid space a 20gauge nylon catheter was advanced cephalad 5 to 10 cm and secured externally with Tegaderm tape and capped. Later, the subarachnoid catheter was extended with 0.3 mm diameter sterilized silicone tubing, which was attached to a peristaltic pump. The total dead space in this system was estimated to be 1.7 ml. Three hours after placement of the catheter CSF sampling began (at 1100 h0. CSF was continuously withdrawn into test tubes at a rate of 0.1 ml per minute, estimated to be 2533% of the normal CSF production rate (7). Every 10 minutes, the 1 ml of CSF that had been collected was separated into 500 p.1aliquots and frozen on dry ice at the bedside. During the 30hour collection period, the CSF was aliquoted into 180 serial samples. We chose to report the results of hourly CSF samples, so as to correspond to the hourly plasma measurements which were obtained through an I.V. catheter secured into a forearm vein. The subarachnoid catheter was withdrawn at 1700 hr on Day 3 of the study, following 30 hours of serial sampling. The subjects then remained at bedrest, with I.V. hydration for the next 24 hours until 1700 hr on Day 4. Subjects remainep in the hospital an additional 3 hours to monitor for side effects, while they ambulated ad lib turn. Each subject was discharged at 2000 hr on Day 4 of the study, with 12-, 24- and 72-hour follow-up phone calls after discharge. Metabolite and Amino Acid Assays Levels of CSF 5-HIAA and TRP were determined using a modified version of a previously described high performance liquid chromatography (HPLC) method with combined fluorometric and amperometric detection (19). Plasma concentrations of total TRP were determined, after perchloric acid deproteination, using an HPLC-fluorometric method (20). Day-to-day coefficients of variation for the CSF and plasma analyses ranged from 5-9%. Data Analysis Concentrations of CSF 5-HIAA and TRP, and plasma TRP levels were analyzed for change over time using a repeated measures analysis of variance (ANOVA). The measures were then averaged across 6-hour intervals, and the ANOVAs were repeated. Cosinor analysis (21) was performed to test for the presence of a 24-hour sinusoidal rhythm. Individual subjects were tested by linear regression equations, using the equations Xl=COS((2p/24)t) and X2=SIN((2p/24)t), where t indicates the time of sampling, as independent factors. In addition, the data for each subject were normalized (z-scores) using their own mean and S.D. (y’=( y-m)/sy). The mean of these normalized values was then fitted to a cosinor model. Pearson correlations were performed to determine if there was an intra-individual relationship between CSF TRP and Plasma TRP. The correlations were repeated incorporating a 4-hour time lag to account for the possible delay between the two compartments. All statistical analyses were performed using SPSS for the Macintosh, version 6.1.
Results A repeated-measures ANOVA across the four subjects showed a trend toward significant hourly variation in CSF 5-HIAA levels (F (29,87)=1.76, p
902
Hourly Human CSF 5-HIAA and Tryptophan
Vol. 60, No. 151997
CLOCK TIME (HOURS) Fig. 1 Mean (iS.E.) concentrations of lumbar CSF 5-HL4A observed in hourly samples obtained from 4 healthy female subjects. The shaded area displays nightime hours (Clock Time 2300-0700).
CLOCK TIME (HOURS) Fig. 2 Mean (&.E.) concentrations of lumbar CSF TRF’ observed in hourly samples obtained from 4 healthy female subjects. The shaded urea displays nightime hours (Clock Time 2300-0700).
Vol. 60, No. 151997
Hourly Human CSF 5-HIAA and Tryptophan
903
Analysis of CSF TRP levels showed significant hourly variation across subjects (F (29,87)=2.84, p
CLOCK TIME (HOURS) Fig. 3 Mean @.E.) concentrations of PLASMA TRP observed in hourly samples obtained from 4 healthy female subjects. The shaded urea displays nightime hours (Clock Time 2300-0700).
A correlation analysis was performed on CSF TRP concentrations and plasma TRP levels for each of the four subjects over the 30-hour sampling period to determine the intra-individual relationship between CSF and plasma levels of TRP. Only one of the four subjects had a significant correlation between CSF and plasma TRP levels (~~0.01). However, when the correlation analysis incorporated a 4-hour time lag to compensate for possible delays between the two compartments (1 l), 2 of the 4 subjects displayed a significant correlation between CSF and plasma TRP levels (~46, p=O.O2; ~0.67, p=O.OOl). A comparison of intra-individual CSF 5HIAA and CSF TRP levels revealed no consistent pattern of association between the two variables.
904
Hourly Human CSF 5-HL4A and Tryptophan
Vol. 60, No. 12,1997
All four subjects tolerated the procedure well. Three of the four women developed a postural headache 24-48 hours following the procedure. The headaches responded well to bedrest, oral hydration and acetaminophen. One woman developed mild, transient motion sickness 24 hours after cessation of the study which resolved within 12 hours. No subjects experienced any further neurologic or infectious complications.
Discussion The results of this preliminary study suggest that a significant diurnal fluctuation in lumbar CSF 5-HIAA levels does not exist in healthy human females. This is in contrast to the findings of Nicoletti et al. (22) who reported a circadian variation in ventricular CSF 5-HIAA in 5 male neurosurgical patients. Their study reported a steady rise in ventricular CSF 5-HIAA beginning at 0600 hr, reaching peak levels at 2400 hr, and returning to initial levels by 0600 hr the next morning. In our study, no appreciable diurnal change in lumbar CSF 5-HIAA levels occurred over the 30-hour period of sampling. The present study and that of Nicoletti and colleagues (22) differ in several important aspects. Our study included healthy women volunteers, whereas the previous study investigated male neurosurgical patients. It is also unclear what effect the presence of a central nervous system disease state in the neurosurgical patients may have had on the CSF 5-HIAA levels. Furthermore, we sampled lumbar CSF whereas Nicoletti et al. sampled ventricular CSF. While controversy exists regarding the ratio of spinal cord vs. brain contribution to lumbar CSF 5-HIAA levels, there is ample evidence that lumbar CSF 5-HIAA levels correlate well with ventricular CSF 5-HIAA levels (23,24). Human studies have shown concentration gradients for 5-HIAA, with levels 3-5 times higher in cistemal compared to lumbar CSF, but with a high correlation between the two sites (25,26). In addition, a postmortem study found that concentrations of cerebral cortex 5HIAA correlated positively with lumbar CSF levels in the same individuals (27). Given these findings, it is difficult to ascribe the difference between our results and those of Nicoletti et al. solely to the different site of CSF sampling. It should be noted that the sleep-wake cycle in our subjects was inconsistent throughout the study, with some subjects napping during the day, and others experiencing interrupted sleep at night. It appears that the neurosurgical patients of Nicoletti et al. followed the standard sleep-wake cycle, but the study description lacks clear documentation of this fact. A number of animal studies have provided data relevant to the issue of diurnal variation in 5-HT turnover and functioning. Polygraph recordings of 5-HT neuronal activity in the cat dorsal raphe nucleus (DRN) have clearly demonstrated markedly greater activity during waking hours, with relative quiescence during slow-wave and rapid-eye-movement sleep (28). Complementary data have been obtained from in vivo dialysis studies of the extracellular fluid (ECF) concentration of 5HT in the cat DRN (29), anterior hypothalamus, and caudate nucieus (30). The dialysis studies have found much lower rates of central 5-HT release during sleep in comparison to awake animals (29,30). The picture with respect to brain, ECF, and CSF levels of 5-HIAA during sleep and wakefulness is much less clear. In monkeys, Yan and colleagues (3 1) have reported that lumbar CSF 5-HIAA rose steadily from 1500 hr to a peak at 2100 hr, with a subsequent drop in levels during night time hours. In the rat, Young and colleagues (32) observed somewhat higher cisternal CSF 5-HIAA levels during the night. This was consistent with the study of Hery et al. (33) which reported higher levels of rat brain 5-HIAA at night, the period of greatest activity in nocturnal species such as the rat. In contrast, others have reported increases in cisternal CSF 5-HJAA levels in cats during sleep (34). Furthermore, results from tissue analysis of rat brain have revealed that opposite patterns in the light/dark phase variation of 5-HT metabolism may exist in different regions of the brain (35). Thus, while central 5-HT release can be postulated to exhibit a diurnal rhythm in humans, the preclinical data are inconclusive as to what sort of variation might be expected to occur in human lumbar CSF 5-HIAA.
Vol. 60, No. 12.1997
Hourly Human CSF 5-HIAA and Tryptophan
905
Although some variation in CSF TRP was observed, a significant diurdal fluctuation over 24hours did not occur. To our knowledge, there are no other available data regarding possible diurnal fluctuation in human lumbar CSF TRP levels. Diurnal changes in plasma TRP also did not occur in our subjects. This is consistent with previous studies of plasma TRP in humans (34-38). The lack of diurnal variation in plasma TRP levels parallels the lack of diurnal change in CSF TRP levels. It is known that meals can cause a transient rise in plasma levels of large neutral amino acids, including TRP (39), and several studies in humans have reported generally lower plasma TRP levels in the early morning hours which rise to peak levels during the late afternoon and evening hours. The fluctuation seen here. and in the previous studies indicates that meals contribute significantly to the daily variation in plasma TRP levels. Plasma TRP and CSF TRP levels were highly correlated in 2 of the 4 subjects. This relationship is not surprising given previous studies in non-human primates showing effects on CSF TRP when plasma TRP is either increased (40) or lowered (41). In addition, studies in rats have shown that increasing plasma TRP by intraperitoneal injection of TRP causes substantial elevations in CSF TRP and 5-HT synthesis (42-45). We found no apparent association between CSF TRP and CSF 5-HIAA. However, this finding is not inconsistent with earlier work which found relatively weak correlations for basal levels of CSF TRP and CSF 5-HIAA in rats (46) and in humans (47). In summary, the results of this preliminary study suggest that there is no significant diurnal variation in lumbar CSF J-HIAA or CSF TRP levels in healthy human females. Rather, we found the measures to be relatively stable over the course of the 30-hour collection period. Because of the small sample size of our study, however, caution in making definitive conclusions from the data is warranted. The apparent stability of human lumbar CSF 5-HIAA may be an advantage when assessing the effects of perturbing central 5-HT functioning; alternatively, it may indicate that the measure is relatively unresponsive to changes in central 5-HT turnover or functioning. Further challenge studies are planned and should provide answers to these questions and other issues concerning central %-IT.
Acknowledgments This work was supported in part by grants MH30929, MH49351, and MH25642 from the National Institute of Mental Health, by NIH General Clinical Research Center Grant RROO125, by the State of Connecticut Department of Mental Health and Addiction Services, by a Young Investigator Award from the National Alliance for Research on Schizophrenia and Depression (Dr. McDougle), and by the Stanley Foundation Research Grant Program (Drs. Heninger and Price). The investigators would like to acknowledge the expert assistance of Debora Mordowanec, R.N., of the Clinical Neuroscience Research Unit, and Laura Hall, M.S. and Ilaria Borghese, M.S. of the Laboratory of Developmental Neurochemistry (Yale Child Study Center), for CSF collection and processing; and the clinical expertise of the General Clinical Research Center, Yale-New Haven Hospital. Sally Vegso, M.S. and Kathryn Czarkowski, B.S. assisted in data analysis and management. Elizabeth Kyle, A.S. expertly prepared the manuscript.
References 1. 2.
P.L. DELGADO, L.H. PRICE, H.L. MILLER, R.M. SALOMON, G.K. AGHAJANIAN, G.R. HENINGER and D.S. CHARNEY, Arch. Gen. Psychiatry 51 865-874 (1994). L.H. PRICE, A.W. GODDARD, L.C. BARR and W.K. GOODMAN, Psychopharmacology: The Fourth Generation of Progress, F.E. Bloom and D.J. Kupfer (Eds), 1311-1323, Raven Press, New York (1995).
906
3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. ;:: 22. 23. 24.
27. it30: 31. 32. 33.
Hourly Human CSF 5-HLtA and Tryptophan
Vol. 60, No. 12,1997
L.H. PRICE, D.S. CHARNEY, P.L. DELGADO and G.R. HENINGER, Am. J. Psychiatry 148 1518-1525 (1991). M. PERLOW, M.H. EBERT, E.K. GORDON, M.G. ZIEGLER, C.R. LAKE and T.N. CHASE, Brain Res. 139 101-l 13 (1978). N.H. KALIN, S.T. SHELTON, C.M. BARKSDALE and M.S. BROWNFIELD, Brain Res. 426 385-391 (1987). P. CHAPPELL, M. RIDDLE, G. ANDERSON, L. SCAHILL, M. HARDIN, D. WALKER, D. COHEN and J. LECKMAN, Biol. Psychiatry 36 35-43 (1994). R.J. WHITE, G. DAKTERS, D. YASHON and M.S. ALBIN, J. Neurosurg. 30 264-269 (1969). J. MCCALLUM, J.C. MAROON and P.J. JANNETTA, J. Neurosurg. 42 434-437 (1975). J.N. BRUCE and E.H. OLDFIELD, Neurosurgery 23 788-790, 1988. J. BRUCE, L. TAMARKIN, C. RIEDEL, S. MARKEY and E. OLDFIELD, J. Clin. Endocrinol. Metab. 72 8 19-823 (1991). M.A. KLING, M.D. DEBELLIS, D.K. O’ROURKE, S.J. LISTWAK, T. D. GERACIOTI, I.E. MCCUTCHEON, K.T. LALOGERAS, E.H. OLDFIELD and P.W. GOLD, J. Clin. Endocrinol. Metab. 79 233-239 (1994). J.A. AMICO, R. TENICELA, J. JOHNSTON and A.G. ROBINSON, J. Clin. Endocrinol. Metab. 57 947-95 1 (1983). M.A. KLING, M.D. DE BELLIS, S.J. LISTWAK, T.D. GERACIOTI, D.K. O’ROURKE, K.T. KALOGERAS, E.H. OLDFIELD, G.P. CHROUSOS and P.W. GOLD, Neuropsychopharmacology 10 201s (1994). M.A. KLING, M.D. DE BELLIS, D. HU, T.D. GERACIOTI, D.S. GOLDSTEIN, E.H. OLDFIELD and P.W. GOLD, Sot. Neurosci. Abst. 21 1236 (1995). T.D. GERACIOTI, D. SCHMIDT, N.N. EKHATOR, R. SHELTON, W. PARRIS, P.T. LOOSEN and M.H. EBERT, Depression 1 149-155 (1993). T.D. GERACIOTI, W.E. NICHOLSON, D.N. ORTH, N.N. EKHATOR and P.T. LOOSEN, Brain Res. 629 260-268 (1993). T.D. GERACIOTI, D.N. ORTH, N.N. EKHATOR, B. BLUMENKOPF and P.T. LOOSEN, J. Clin. Endocrinol. Metab. 74 1325- 1330 (1992). T.D. GERACIOTI, P.T. LOOSEN, M.H. EBERT, N.N. EKHATOR, W.E. NICHOLSON and D.N. ORTH, Neuroendocrinology 60 635-642, 1994. G.M. ANDERSON, J.G. YOUNG and D.J. COHEN, J. Chromatogr. 142 501-505 (1979). G.M. ANDERSON, F.C. FEIBEL and D.J. COHEN, Life Sci. 40 1063-1070 (1987). W. NELSON, Y.L. TONG, J.-K. LEE and F. HALBERG, Chronobiologia 6 305-323 (1979). F. NICOLETTI, R. RAFFAELE, A. FALSAPERLA and R. PACI , Eur. Neurol. 20 9-12 (1981). J.H. WOOD, Neurology 30 645-65 1 (1980). L. SIEVER, H. KRAEMER, R. SACK, P. ANGWIN, P. BERGER, V. ZARCONE, J. BARCHAS, H.K.H. BRODIE, Dis. Nerv. Sys. 36 13-16 (1975). I. DEGRELL and E. NAGY, Biol. Psychiatry 27 891-896 (1990). A.T.B. MOIR, G.W. ASHCROFT, T.B.B. CRAWFORD, D. ECCLESTON, H.C. GULDBERG, Brain 93.357-368 (1970). M. STANLEY, L. TRASKMAN-BENDZ and K. DOROVINI-ZIS, Life Sci. 37 12791286 (1985). B.L. JACOBS and C.A. FORNAL, Trends Neurosci. 16 346-352 (1993). C.M. PORTAS and R.W. MCCARLEY, Brain Res. 648 306-312 (1994). L.O. WILKINSON, S.B. AUERBACH and B.L. JACOBS, J. Neurosci. ll(9) 27322741 (1991). D. YAN, T. URANO, M.H. PIETRASZEK 1 CUTMnVAMA v TTELuTDA V ._____---*-, _. KOJIMA, K. SAKAKIBARA, K. SL____ ERTZ4WA, Y. TAKADA and A. TAKADA, Life Sci. 52 745-749 (1993). S.N. YOUNG, d.M. ANDERSON and W.C. PURDY, J. Neurochem. 34 309-315 (1980). F. HERY, G. CHOUVET, J.P. KAN, J.F. PUJOL and J. GLOWINSKI, Brain Res. 123 137-145 (1977). Y-*I*.Av
‘LXl.l_,
LX.
“lAI”I”I.iT,
I.
Vol. 60, No. 12,1!3W
34. ;:: 37. 38. 39. 40. 41. t:: fi: 46. 47.
Hourly Human CSF 5-HIAA and Tryptophan
907
M. RADULOVACKI, R.L. BUCKINGHAM, E.H. CHEN and R. KOVACEVIC, Brain Res. 129 371-374 (1977). L. PONCET, L. DENORGY and M. JOUVET, J. Neural. Transm. 92 137-150 (1993). R.J. WURTMAN, C.M. ROSE, C. CHOU and F.F. LARIN, N. Engl. J. Med. 279 17 l175 (1968). M.H. PIETRASZEK, S. TAKAHADHI, Y. TAKADA, K. OHARA, H. INATOMI, N. KONDO, K. OHARA and A. TAKADA, Thrombosis Res. 64 243-252 (1991). H. DAM, E.T. MELLERUP and O.J. RAFAELSEN, Acta Psychiatr. Stand. 69 190-196 (1984). T. ERIKSSON, L. VOOG, J. WALLINDER and T.E. ERIKSSON, J. Psychiat. Res. 23 241-249 (1989). S.N. YOUNG and F.R. ERVIN, J. Neurochem. 42 1570-1573 (1984). S.N. YOUNG, F.R. ERVIN, R.O. PIHL and P. FINN, Psychopharm. 98 508-5 11 (1989). \-- -- ,J.D. FERNSTOM and R.J. WURTMAN, Science 173 149-154 (1971). G. GESSA, G. BIGGIO and A. TAGLIAMONTE, Fedn. Proc. 312168-2172 (1972). P.J. KNOTT and G. CURZON, Nature 239 452-453 (1972). S.N. YOUNG, G.M. ANDERSON and W.C. PURDY, J. Neurochem. 34 309-315 (1980). G. CURZON, B.D. KANTAMANENI, J.R. BARTLETT and P.K. BRIDGES, J. Neurochem. 26 613-615 (1976). S.N. YOUNG, D. GAUTHIER, G.M. ANDERSON and W.C. PURDY, J. Neurol. Neurosur. Psychiatry 43 438-445 (1980).
scicnea,vol.
60, No. 12 pp. wmm7,1997 r*lu?eluvierseiacc1nc. -_. PrintcdinthcUSA. AMrightsremcd am-324x/97 $1761)t .oo 7
PII SOOU-3205(97)ooo21-0
ELSEVIER
OF DIURNAL VARIATION OF CEREBROSPINAL FLUID TRYPTOPHAN AND 5-HYDROXYINDO LEACETIC ACID IN IEALTHY HUMAN FEMALES
ASSESSMENT
Paul D. Kirwin,l*4, George M. Anderson.2, Phillip B. Chappel12, Lloyd Saberski?, James F. Leckman,l~~, Thomas D. Geracioti,s, George R. Heninger,1*4, Lawrence H. Price,1*4, Christopher J. McDougle,l*2*4 Departments of Psychiatty,t Child Study Center,2 and Anesthesiology3 Yale University School of Medicine Clinical Neuroscience Research Unit4 Abraham Ribicoff Research Facilities Connecticut Mental Health Center DepartmentaZPsychiatry~ University of Cincinnati School of Medicine Veterans Administration Medical Center (Received in final fom January 6,1997)
Summary The role of serotonin (5-HT) in the pathogenesis and treatment of major neuropsychiatric disorders, including mood and anxiety disorders, continues to be the subject of extensive research. Previous studies examinin g central 5-I-IT functioning measured cerebrospinal fluid (CSF) levels of 5hydroxyindoleacetic acid (5-HIAA) by using single or multiple lumbar punctures. A number of investigators have demonstrated the feasiblity of continuous CSF sampling via an indwelling lumbar catheter to study CSF neurochemistry in healthy subjects and patients with neuropsychiatric illness. Four healthy female volunteers, aged 21-34 years, underwent continuous CSF sampling. CSF was collected at a constant rate of 1 ml every 10 minutes over a 30-hour period, with levels of tryptophan (TRP) and 5-HIAA measured every hour. Plasma was also obtained hourly for TRP determination. The results of this study indicate that CSF 5-I-I&4, CSF TRP, and plasma TRP levels showed variation over time, but failed to show diurnal fluctuation. Intraindividual coefficients of variationdetermined for CSF 5-HIAA, CSF TRP, and plasma TRP ranged from 9.2 to 14.9%, 8.8 to 14.6%, and 14.7 to 19.0%, respectively. Continuous CSF sampling is safe and feasible in humans, and may prove useful for studies of central 5-I-II neurotransmission in neuropsychiatric illness. key Wd:
cerebrospinal fluid, serotonin, neuropsychiatric disorders, Shydroxyindoleacetic acid
The role of serotonin (5-hydroxytryptamine [5HT]) in the pathogenesis and treatment of major neuropsychiatric disorders, including mood (1) and anxiety (2) disorders, continues to be the subject of intensive research. Investigators have taken two primary approaches to assess central 5HT functioning in patients. Numerous studies have utilized 5-I-H precursors and receptor agonists and antagonists as pharmacological probes to explore hypothesized 5-I-IT dysfunctions (3). Corresponding Author: Christopher J. McDougle, M.D., Connecticut Mental Health Center, Clinical Neuroscience Research Unit, Room 333B, 34 Park Street, New Haven, CT 06519, Phone: (203) 789-7333, FAX: (203) 789-7651, E-mail: [email protected]
900
Hourly Human CSF 5-HIAA and Tryptophan
Vol. 60, No. 12, 1997
However, the most commonly employed paradigm for assessing the S-HT system has involved the measurement of cerebrospinal fluid (CSF) levels of the principle 5-HT metabolite 5hydroxyindoleacetic acid (5-HIAA). Nearly all of the studies of CSF 5-HIAA have examined metabolite concentrations in single lumbar CSF samples. This cross-sectional approach does not address the possible rhythmic secretion of neurochemicals into the CSF (45). Nor does it allow for an accounting of the changes which might occur in response to the stress of the lumbar puncture itself (6). The continuous sampling of CSF offers a possible strategy for dealing with these issues. Continuous lumbar spinal drainage has been safely used by neurosurgeons and anesthesiologists to administer spinal anesthesia, to treat CSF fistulas, and to control elevated ventricular pressure (7,8). In 1988, Bruce and Oldfield (9) described a modification of this technique which provides a less traumatic means of performing sequential CSF sampling. A number of investigators have now employed the sequential CSF sampling strategy to study CSF neurochemistry in healthy subjects and patients with neuropsychiatric illness. Investigations to characterize CSF diurnal variation of melatonin in neurosurgical patients and healthy volunteers (lo), corticotropin-releasing hormone (CRH) in healthy volunteers (1 l), oxytocin in neurosurgical patients (12) and healthy volunteers (13), and norepinephrine (NE) in healthy volunteers (14) have been conducted. In addition, sequential sampling of CSF for shorter periods of time have examined NE in depressed patients and normal controls (15), cholecystokinin in depression and alcoholism (16) CRH in depressed patients and healthy volunteers (17), and CRH, NE, 3-methoxy-4-hydroxyphenylglycol (MHPG), 5-HIAA, and L-tryptophan (TRP) in alcoholic patients and healthy controls (18). The purpose of the current study was to characterize the diurnal variation of CSF 5-HIAA and the amino acid precursor of 5-HT, TRP, in healthy human subjects. Four healthy adult female subjects underwent continuous CSF sampling over a period of 30 hours via an indwelling lumbar catheter. We determined hourly levels of CSF 5-HIAA and TRP, and compared the CSF measures to plasma levels of TRP measured over the same period.
Methods Subjects The study was approved by the Human Investigation Committee of the Yale University School of Medicine. Four female subjects, free from medical or psychiatric illness, gave voluntary, written informed consent to participate as paid volunteers in the study. The subjects ranged in age from 2 1 to 34 years (mean& S.D.=26.5 f 5.5 years) and weighed 54 to 67.5 kg (mean f S.D.=59.1 + 6.5 kg). Subjects were accepted into the study only if they had no prior history of medical, neurologic, or psychiatric illness. Normal laboratory values of electrolytes, serum creatinine and blood urea nitrogen, complete blood count and differential, clotting factors, thyroid indices, liver function tests, fasting glucose, and a negative serum pregnancy test were required prior to participation in the study. Normal urinalysis and a negative urine drug toxicology screen were also required. In addition, all subjects received a thorough physical examination, including complete neurologic and fundoscopic exam. All subjects had normal electrocardiograms before entering the study. Although information about the specific phase of the menstrual cycle was not obtained, none of the subjects were actively menstruating at the time of the study. Cerebrospinal Fluid And Plasma Collection Subjects followed a low-monoamine diet for three days prior to entering the General Clinical Research Center at Yale-New Haven Hospital for the study. Subjects continued on this diet, consisting of meals at 0800 hr (307 mg TRP), 1200 hr (727 mg TRP), and 1700 hr (811 mg TRP), until the 30-hour CSF sampling ceased. Subjects remained in the hospital for a total of four days. They were admitted to the hospital at 1600 hr (Day 1) on the day prior to the lumbar puncture to control diet and exercise, and to insure adequate intravenous (I.V.) hydration with 0.9% normal saline, which ran at 250 cc per hour from 2000 hr (Day 1) until 1700 hr (Day 4). At 0800 hr (Day 2), after a night of bedrest and I.V. hydration, a catheter was placed in the lumbar subarachnoid space in a modification of the method of Bruce and Oldfield (9) previously
Vol. 60, No. 12, 1997
Hourly Human CSF 5-HIAA and Tryptophan
901
described by Geracioti et al. (15-18). Subjects were placed in the lateral decubitus pOSitiOn in their hospital bed and, after application of intradermal lidocaine anesthesia, an 18-gauge Touhy needle was inserted through the L3-L4 or LA-L5 interspace. After entry into the subarachnoid space a 20gauge nylon catheter was advanced cephalad 5 to 10 cm and secured externally with Tegaderm tape and capped. Later, the subarachnoid catheter was extended with 0.3 mm diameter sterilized silicone tubing, which was attached to a peristaltic pump. The total dead space in this system was estimated to be 1.7 ml. Three hours after placement of the catheter CSF sampling began (at 1100 h0. CSF was continuously withdrawn into test tubes at a rate of 0.1 ml per minute, estimated to be 2533% of the normal CSF production rate (7). Every 10 minutes, the 1 ml of CSF that had been collected was separated into 500 p.1aliquots and frozen on dry ice at the bedside. During the 30hour collection period, the CSF was aliquoted into 180 serial samples. We chose to report the results of hourly CSF samples, so as to correspond to the hourly plasma measurements which were obtained through an I.V. catheter secured into a forearm vein. The subarachnoid catheter was withdrawn at 1700 hr on Day 3 of the study, following 30 hours of serial sampling. The subjects then remained at bedrest, with I.V. hydration for the next 24 hours until 1700 hr on Day 4. Subjects remainep in the hospital an additional 3 hours to monitor for side effects, while they ambulated ad lib turn. Each subject was discharged at 2000 hr on Day 4 of the study, with 12-, 24- and 72-hour follow-up phone calls after discharge. Metabolite and Amino Acid Assays Levels of CSF 5-HIAA and TRP were determined using a modified version of a previously described high performance liquid chromatography (HPLC) method with combined fluorometric and amperometric detection (19). Plasma concentrations of total TRP were determined, after perchloric acid deproteination, using an HPLC-fluorometric method (20). Day-to-day coefficients of variation for the CSF and plasma analyses ranged from 5-9%. Data Analysis Concentrations of CSF 5-HIAA and TRP, and plasma TRP levels were analyzed for change over time using a repeated measures analysis of variance (ANOVA). The measures were then averaged across 6-hour intervals, and the ANOVAs were repeated. Cosinor analysis (21) was performed to test for the presence of a 24-hour sinusoidal rhythm. Individual subjects were tested by linear regression equations, using the equations Xl=COS((2p/24)t) and X2=SIN((2p/24)t), where t indicates the time of sampling, as independent factors. In addition, the data for each subject were normalized (z-scores) using their own mean and S.D. (y’=( y-m)/sy). The mean of these normalized values was then fitted to a cosinor model. Pearson correlations were performed to determine if there was an intra-individual relationship between CSF TRP and Plasma TRP. The correlations were repeated incorporating a 4-hour time lag to account for the possible delay between the two compartments. All statistical analyses were performed using SPSS for the Macintosh, version 6.1.
Results A repeated-measures ANOVA across the four subjects showed a trend toward significant hourly variation in CSF 5-HIAA levels (F (29,87)=1.76, p
902
Hourly Human CSF 5-HIAA and Tryptophan
Vol. 60, No. 151997
CLOCK TIME (HOURS) Fig. 1 Mean (iS.E.) concentrations of lumbar CSF 5-HL4A observed in hourly samples obtained from 4 healthy female subjects. The shaded area displays nightime hours (Clock Time 2300-0700).
CLOCK TIME (HOURS) Fig. 2 Mean (&.E.) concentrations of lumbar CSF TRF’ observed in hourly samples obtained from 4 healthy female subjects. The shaded urea displays nightime hours (Clock Time 2300-0700).
Vol. 60, No. 151997
Hourly Human CSF 5-HIAA and Tryptophan
903
Analysis of CSF TRP levels showed significant hourly variation across subjects (F (29,87)=2.84, p
CLOCK TIME (HOURS) Fig. 3 Mean @.E.) concentrations of PLASMA TRP observed in hourly samples obtained from 4 healthy female subjects. The shaded urea displays nightime hours (Clock Time 2300-0700).
A correlation analysis was performed on CSF TRP concentrations and plasma TRP levels for each of the four subjects over the 30-hour sampling period to determine the intra-individual relationship between CSF and plasma levels of TRP. Only one of the four subjects had a significant correlation between CSF and plasma TRP levels (~~0.01). However, when the correlation analysis incorporated a 4-hour time lag to compensate for possible delays between the two compartments (1 l), 2 of the 4 subjects displayed a significant correlation between CSF and plasma TRP levels (~46, p=O.O2; ~0.67, p=O.OOl). A comparison of intra-individual CSF 5HIAA and CSF TRP levels revealed no consistent pattern of association between the two variables.
904
Hourly Human CSF 5-HL4A and Tryptophan
Vol. 60, No. 12,1997
All four subjects tolerated the procedure well. Three of the four women developed a postural headache 24-48 hours following the procedure. The headaches responded well to bedrest, oral hydration and acetaminophen. One woman developed mild, transient motion sickness 24 hours after cessation of the study which resolved within 12 hours. No subjects experienced any further neurologic or infectious complications.
Discussion The results of this preliminary study suggest that a significant diurnal fluctuation in lumbar CSF 5-HIAA levels does not exist in healthy human females. This is in contrast to the findings of Nicoletti et al. (22) who reported a circadian variation in ventricular CSF 5-HIAA in 5 male neurosurgical patients. Their study reported a steady rise in ventricular CSF 5-HIAA beginning at 0600 hr, reaching peak levels at 2400 hr, and returning to initial levels by 0600 hr the next morning. In our study, no appreciable diurnal change in lumbar CSF 5-HIAA levels occurred over the 30-hour period of sampling. The present study and that of Nicoletti and colleagues (22) differ in several important aspects. Our study included healthy women volunteers, whereas the previous study investigated male neurosurgical patients. It is also unclear what effect the presence of a central nervous system disease state in the neurosurgical patients may have had on the CSF 5-HIAA levels. Furthermore, we sampled lumbar CSF whereas Nicoletti et al. sampled ventricular CSF. While controversy exists regarding the ratio of spinal cord vs. brain contribution to lumbar CSF 5-HIAA levels, there is ample evidence that lumbar CSF 5-HIAA levels correlate well with ventricular CSF 5-HIAA levels (23,24). Human studies have shown concentration gradients for 5-HIAA, with levels 3-5 times higher in cistemal compared to lumbar CSF, but with a high correlation between the two sites (25,26). In addition, a postmortem study found that concentrations of cerebral cortex 5HIAA correlated positively with lumbar CSF levels in the same individuals (27). Given these findings, it is difficult to ascribe the difference between our results and those of Nicoletti et al. solely to the different site of CSF sampling. It should be noted that the sleep-wake cycle in our subjects was inconsistent throughout the study, with some subjects napping during the day, and others experiencing interrupted sleep at night. It appears that the neurosurgical patients of Nicoletti et al. followed the standard sleep-wake cycle, but the study description lacks clear documentation of this fact. A number of animal studies have provided data relevant to the issue of diurnal variation in 5-HT turnover and functioning. Polygraph recordings of 5-HT neuronal activity in the cat dorsal raphe nucleus (DRN) have clearly demonstrated markedly greater activity during waking hours, with relative quiescence during slow-wave and rapid-eye-movement sleep (28). Complementary data have been obtained from in vivo dialysis studies of the extracellular fluid (ECF) concentration of 5HT in the cat DRN (29), anterior hypothalamus, and caudate nucieus (30). The dialysis studies have found much lower rates of central 5-HT release during sleep in comparison to awake animals (29,30). The picture with respect to brain, ECF, and CSF levels of 5-HIAA during sleep and wakefulness is much less clear. In monkeys, Yan and colleagues (3 1) have reported that lumbar CSF 5-HIAA rose steadily from 1500 hr to a peak at 2100 hr, with a subsequent drop in levels during night time hours. In the rat, Young and colleagues (32) observed somewhat higher cisternal CSF 5-HIAA levels during the night. This was consistent with the study of Hery et al. (33) which reported higher levels of rat brain 5-HIAA at night, the period of greatest activity in nocturnal species such as the rat. In contrast, others have reported increases in cisternal CSF 5-HJAA levels in cats during sleep (34). Furthermore, results from tissue analysis of rat brain have revealed that opposite patterns in the light/dark phase variation of 5-HT metabolism may exist in different regions of the brain (35). Thus, while central 5-HT release can be postulated to exhibit a diurnal rhythm in humans, the preclinical data are inconclusive as to what sort of variation might be expected to occur in human lumbar CSF 5-HIAA.
Vol. 60, No. 12.1997
Hourly Human CSF 5-HIAA and Tryptophan
905
Although some variation in CSF TRP was observed, a significant diurdal fluctuation over 24hours did not occur. To our knowledge, there are no other available data regarding possible diurnal fluctuation in human lumbar CSF TRP levels. Diurnal changes in plasma TRP also did not occur in our subjects. This is consistent with previous studies of plasma TRP in humans (34-38). The lack of diurnal variation in plasma TRP levels parallels the lack of diurnal change in CSF TRP levels. It is known that meals can cause a transient rise in plasma levels of large neutral amino acids, including TRP (39), and several studies in humans have reported generally lower plasma TRP levels in the early morning hours which rise to peak levels during the late afternoon and evening hours. The fluctuation seen here. and in the previous studies indicates that meals contribute significantly to the daily variation in plasma TRP levels. Plasma TRP and CSF TRP levels were highly correlated in 2 of the 4 subjects. This relationship is not surprising given previous studies in non-human primates showing effects on CSF TRP when plasma TRP is either increased (40) or lowered (41). In addition, studies in rats have shown that increasing plasma TRP by intraperitoneal injection of TRP causes substantial elevations in CSF TRP and 5-HT synthesis (42-45). We found no apparent association between CSF TRP and CSF 5-HIAA. However, this finding is not inconsistent with earlier work which found relatively weak correlations for basal levels of CSF TRP and CSF 5-HIAA in rats (46) and in humans (47). In summary, the results of this preliminary study suggest that there is no significant diurnal variation in lumbar CSF J-HIAA or CSF TRP levels in healthy human females. Rather, we found the measures to be relatively stable over the course of the 30-hour collection period. Because of the small sample size of our study, however, caution in making definitive conclusions from the data is warranted. The apparent stability of human lumbar CSF 5-HIAA may be an advantage when assessing the effects of perturbing central 5-HT functioning; alternatively, it may indicate that the measure is relatively unresponsive to changes in central 5-HT turnover or functioning. Further challenge studies are planned and should provide answers to these questions and other issues concerning central %-IT.
Acknowledgments This work was supported in part by grants MH30929, MH49351, and MH25642 from the National Institute of Mental Health, by NIH General Clinical Research Center Grant RROO125, by the State of Connecticut Department of Mental Health and Addiction Services, by a Young Investigator Award from the National Alliance for Research on Schizophrenia and Depression (Dr. McDougle), and by the Stanley Foundation Research Grant Program (Drs. Heninger and Price). The investigators would like to acknowledge the expert assistance of Debora Mordowanec, R.N., of the Clinical Neuroscience Research Unit, and Laura Hall, M.S. and Ilaria Borghese, M.S. of the Laboratory of Developmental Neurochemistry (Yale Child Study Center), for CSF collection and processing; and the clinical expertise of the General Clinical Research Center, Yale-New Haven Hospital. Sally Vegso, M.S. and Kathryn Czarkowski, B.S. assisted in data analysis and management. Elizabeth Kyle, A.S. expertly prepared the manuscript.
References 1. 2.
P.L. DELGADO, L.H. PRICE, H.L. MILLER, R.M. SALOMON, G.K. AGHAJANIAN, G.R. HENINGER and D.S. CHARNEY, Arch. Gen. Psychiatry 51 865-874 (1994). L.H. PRICE, A.W. GODDARD, L.C. BARR and W.K. GOODMAN, Psychopharmacology: The Fourth Generation of Progress, F.E. Bloom and D.J. Kupfer (Eds), 1311-1323, Raven Press, New York (1995).
906
3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. ;:: 22. 23. 24.
27. it30: 31. 32. 33.
Hourly Human CSF 5-HLtA and Tryptophan
Vol. 60, No. 12,1997
L.H. PRICE, D.S. CHARNEY, P.L. DELGADO and G.R. HENINGER, Am. J. Psychiatry 148 1518-1525 (1991). M. PERLOW, M.H. EBERT, E.K. GORDON, M.G. ZIEGLER, C.R. LAKE and T.N. CHASE, Brain Res. 139 101-l 13 (1978). N.H. KALIN, S.T. SHELTON, C.M. BARKSDALE and M.S. BROWNFIELD, Brain Res. 426 385-391 (1987). P. CHAPPELL, M. RIDDLE, G. ANDERSON, L. SCAHILL, M. HARDIN, D. WALKER, D. COHEN and J. LECKMAN, Biol. Psychiatry 36 35-43 (1994). R.J. WHITE, G. DAKTERS, D. YASHON and M.S. ALBIN, J. Neurosurg. 30 264-269 (1969). J. MCCALLUM, J.C. MAROON and P.J. JANNETTA, J. Neurosurg. 42 434-437 (1975). J.N. BRUCE and E.H. OLDFIELD, Neurosurgery 23 788-790, 1988. J. BRUCE, L. TAMARKIN, C. RIEDEL, S. MARKEY and E. OLDFIELD, J. Clin. Endocrinol. Metab. 72 8 19-823 (1991). M.A. KLING, M.D. DEBELLIS, D.K. O’ROURKE, S.J. LISTWAK, T. D. GERACIOTI, I.E. MCCUTCHEON, K.T. LALOGERAS, E.H. OLDFIELD and P.W. GOLD, J. Clin. Endocrinol. Metab. 79 233-239 (1994). J.A. AMICO, R. TENICELA, J. JOHNSTON and A.G. ROBINSON, J. Clin. Endocrinol. Metab. 57 947-95 1 (1983). M.A. KLING, M.D. DE BELLIS, S.J. LISTWAK, T.D. GERACIOTI, D.K. O’ROURKE, K.T. KALOGERAS, E.H. OLDFIELD, G.P. CHROUSOS and P.W. GOLD, Neuropsychopharmacology 10 201s (1994). M.A. KLING, M.D. DE BELLIS, D. HU, T.D. GERACIOTI, D.S. GOLDSTEIN, E.H. OLDFIELD and P.W. GOLD, Sot. Neurosci. Abst. 21 1236 (1995). T.D. GERACIOTI, D. SCHMIDT, N.N. EKHATOR, R. SHELTON, W. PARRIS, P.T. LOOSEN and M.H. EBERT, Depression 1 149-155 (1993). T.D. GERACIOTI, W.E. NICHOLSON, D.N. ORTH, N.N. EKHATOR and P.T. LOOSEN, Brain Res. 629 260-268 (1993). T.D. GERACIOTI, D.N. ORTH, N.N. EKHATOR, B. BLUMENKOPF and P.T. LOOSEN, J. Clin. Endocrinol. Metab. 74 1325- 1330 (1992). T.D. GERACIOTI, P.T. LOOSEN, M.H. EBERT, N.N. EKHATOR, W.E. NICHOLSON and D.N. ORTH, Neuroendocrinology 60 635-642, 1994. G.M. ANDERSON, J.G. YOUNG and D.J. COHEN, J. Chromatogr. 142 501-505 (1979). G.M. ANDERSON, F.C. FEIBEL and D.J. COHEN, Life Sci. 40 1063-1070 (1987). W. NELSON, Y.L. TONG, J.-K. LEE and F. HALBERG, Chronobiologia 6 305-323 (1979). F. NICOLETTI, R. RAFFAELE, A. FALSAPERLA and R. PACI , Eur. Neurol. 20 9-12 (1981). J.H. WOOD, Neurology 30 645-65 1 (1980). L. SIEVER, H. KRAEMER, R. SACK, P. ANGWIN, P. BERGER, V. ZARCONE, J. BARCHAS, H.K.H. BRODIE, Dis. Nerv. Sys. 36 13-16 (1975). I. DEGRELL and E. NAGY, Biol. Psychiatry 27 891-896 (1990). A.T.B. MOIR, G.W. ASHCROFT, T.B.B. CRAWFORD, D. ECCLESTON, H.C. GULDBERG, Brain 93.357-368 (1970). M. STANLEY, L. TRASKMAN-BENDZ and K. DOROVINI-ZIS, Life Sci. 37 12791286 (1985). B.L. JACOBS and C.A. FORNAL, Trends Neurosci. 16 346-352 (1993). C.M. PORTAS and R.W. MCCARLEY, Brain Res. 648 306-312 (1994). L.O. WILKINSON, S.B. AUERBACH and B.L. JACOBS, J. Neurosci. ll(9) 27322741 (1991). D. YAN, T. URANO, M.H. PIETRASZEK 1 CUTMnVAMA v TTELuTDA V ._____---*-, _. KOJIMA, K. SAKAKIBARA, K. SL____ ERTZ4WA, Y. TAKADA and A. TAKADA, Life Sci. 52 745-749 (1993). S.N. YOUNG, d.M. ANDERSON and W.C. PURDY, J. Neurochem. 34 309-315 (1980). F. HERY, G. CHOUVET, J.P. KAN, J.F. PUJOL and J. GLOWINSKI, Brain Res. 123 137-145 (1977). Y-*I*.Av
‘LXl.l_,
LX.
“lAI”I”I.iT,
I.
Vol. 60, No. 12,1!3W
34. ;:: 37. 38. 39. 40. 41. t:: fi: 46. 47.
Hourly Human CSF 5-HIAA and Tryptophan
907
M. RADULOVACKI, R.L. BUCKINGHAM, E.H. CHEN and R. KOVACEVIC, Brain Res. 129 371-374 (1977). L. PONCET, L. DENORGY and M. JOUVET, J. Neural. Transm. 92 137-150 (1993). R.J. WURTMAN, C.M. ROSE, C. CHOU and F.F. LARIN, N. Engl. J. Med. 279 17 l175 (1968). M.H. PIETRASZEK, S. TAKAHADHI, Y. TAKADA, K. OHARA, H. INATOMI, N. KONDO, K. OHARA and A. TAKADA, Thrombosis Res. 64 243-252 (1991). H. DAM, E.T. MELLERUP and O.J. RAFAELSEN, Acta Psychiatr. Stand. 69 190-196 (1984). T. ERIKSSON, L. VOOG, J. WALLINDER and T.E. ERIKSSON, J. Psychiat. Res. 23 241-249 (1989). S.N. YOUNG and F.R. ERVIN, J. Neurochem. 42 1570-1573 (1984). S.N. YOUNG, F.R. ERVIN, R.O. PIHL and P. FINN, Psychopharm. 98 508-5 11 (1989). \-- -- ,J.D. FERNSTOM and R.J. WURTMAN, Science 173 149-154 (1971). G. GESSA, G. BIGGIO and A. TAGLIAMONTE, Fedn. Proc. 312168-2172 (1972). P.J. KNOTT and G. CURZON, Nature 239 452-453 (1972). S.N. YOUNG, G.M. ANDERSON and W.C. PURDY, J. Neurochem. 34 309-315 (1980). G. CURZON, B.D. KANTAMANENI, J.R. BARTLETT and P.K. BRIDGES, J. Neurochem. 26 613-615 (1976). S.N. YOUNG, D. GAUTHIER, G.M. ANDERSON and W.C. PURDY, J. Neurol. Neurosur. Psychiatry 43 438-445 (1980).