In vivo influences on cerebrospinal fluid amino acid levels

Neurochem. Int. Vol. 12, No. I, pp. 25-31, 1988

Printed in Great Britain. All rights reserved

0197-0186/88 $3.00+ 0.00 Copyright © 1988 PergamonJournals Ltd

IN VIVO I N F L U E N C E S ON CEREBROSPINAL F L U I D AMINO ACID LEVELS JOHANNES KORNHUBER*t, MALTE E. KORNHUBER, GERT M. HARTMANN and ANSELM W. KORNHUBER Department of Neurology, University of Ulm, Steinh6velstr.9, 7900 Ulm, F.R.G. (Received 27 March 1987; accepted 6 July 1987)

Abstract--The concentration of 19 amino acids and ethanolamine in two independent groups (n = 36, n = 19) of normal human cerebrospinal fluid (CSF) samples were measured by high performance liquid chromatography. Age, sex, time of lumbar puncture, position during lumbar puncture, protein and glucose content of the CSF were monitored and the influence of these parameters on CSF amino acid levels was determined. Hypotheses formulated after observing measurements from the first group of CSF samples were tested against the second group of CSF samples using conservative statistics. The main finding was a positive correlation between CSF glucose and CSF glutamate levels.

Examination of cerebrospinal fluid (CSF) constitutes an important approach to the neurochemistry of the human central nervous system (CNS). Many suspected neurotransmitters are amino acids whose concentrations can be detected in the CSF. While the origin of the CSF amino acids is not entirely clear, it is suggested that at least some of them are released from the nervous tissue and therefore may reflect to some extent its functional activity. Therefore, determination of amino acids in CSF has been performed in a number of investigations on healthy subjects (Perry and Jones, 1961; Hagenfeldt et al., 1984), control patients (Gjessing et al., 1974; Plum et al., 1974; Perry et al., 1975; Goodnick et al., 1980; Ferraro and Hare, 1984; Kruse et al., 1985) and experimental patients suffering from a variety of psycho- and neuropathological states [e.g. Huntington's chorea (Perry et al., 1973; Oepen et al., 1982); epilepsy (Mutani et al., 1974; Plum, 1974; Engelsen and Elsayed, 1984); Parkinson's disease (Lakke and Teelken, 1976; Iijima et al., 1978); Alzheimer's disease (Smith et al., 1985)]. Little, however, is known about physiological factors that might influence the amino acid composition of the CSF. The influence of age and sex has been recorded for a series of amino acids, but so far no consistent results have emerged (for references see discussion). The aim of the present *Present address: Department of Psychiatry, University of Wiirzburg, Ffichsleinstrafle 15, 8700 Wiirzburg, F.R.G. tAddress correspondence to: J. Kornhuber, Nervenklinik der Universit/it, F/ichsleinstraBe 15, 8700 Wiirzburg, F.R.G.

study was to determine whether CSF amino acid levels vary according to time of day, age, sex, patient position at time of lumbar puncture, glucose and protein concentration. A preliminary account of these results has been published (Kornhuber et al., 1986). EXPERIMENTAL PROCEDURES

CSF samples were taken from patients with low back pain and unspecific neurological symptoms like headache and dizziness. Spinal tap was performed between 8 a.m. and 5 p.m. in either sitting or lying position. Only clear, colourless CSF was used. The first 2 ml of a 7 ml sample were used for amino acid analysis. This 2ml sample was cooled on ice, deproteinised (microcollodium filtration; Sartorius, Grttingen, F.R.G.) and frozen to - 8 0 ° C for storage within 15 min after CSF collection. Using the residual 5ml of the original sample routine laboratory analysis was performed within 3 hours after CSF collection. Only CSF specimens with normal laboratory parameters (glucose: 1.9-5.0 mmol/l; total protein: 150-450mg/1; lactate: 1.2-2.1 mmol/1; chloride: 120-130mmol/1; white cell count: up to 4/#1) were used. Two independent samples of CSF specimens were collected: group A from March 1985 to August 1985; group B from October 1985 to December 1985. Methodological procedure was identical in both groups. The levels of 19 amino acids and ethanolamine were determined by HPLC using a recent modification (Zettlmeil31 et al., 1986) of the pre25

JOHANNES

el ell.

KORNHUBER

a

b~

Q

b-

tO 0

~0 0 t~

o5

tO b-

GO

d

04

co tO

b

r~

Q

e~

rO (Xl rO

r<

~-

re)

~2

b-

.

I~tO

rr)~

•

Fig. 1. (a) A chromatogram of a standard amino acid solution. The starting point is on the left side. The retention time in min is given above each peak. (b). A typical chromatogram of a lumbar CSF specimen. EA = ethanolamine.

CSF amino acids column o-phthaldialdehyde-derivatisation method (Lindroth and Mopper, 1979). In brief, 100 #1 of the o-phthaldialdehyde solution (Pierce) was mixed for one minute with 50/~1 of freshly thawed C S F at room temperature. Reaction was stopped by addition of 8 5 0 / d of a sodium acetate/boric acid buffer. The mixture was run through a precolumn (RP-18, Shandon ODS Hypersil, 5/~m, 17 x 5 ram) and the derivatised amino acids were separated on a reversedphase column (RP-18, Shandon O D S Hypersil, 5 #m, 125 x 5 m m ) using an eluent containing a 50raM sodium acetate/25 m M boric acid buffer and methanol at a flow rate of 1 ml/min. The fluorescence was monitored with a Shimadzu RF-530 fluorescence detector, excitation 340 nm, emission 455 nm. The concentration of an amino acid in the C S F was calculated from the ratio of its peak area (Shimadzu C - R 3 A recording integrator) compared to the peak area for this amino acid in the external standard. Within a 3 day period of measurement variability was less than 5% Citrultine, taurine, ethanolamine and ornithine were purchased from Serva, and all other amino acids from Pierce. Nonparametric statistics were applied throughout (rank correlation coefficient, M a n n - W h i t n e y U-test, Kruskal-Wallis one-way analysis of variance, Person's c h i - s q u a r e test). As pointed out above, two independent samples of C S F specimens were collected and analysed using the same methodology. The formal statistical reasoning for using two separate groups was that results from group A with levels of significance below 0.05 (two-tailed) were used as hypotheses, and these hypotheses only were tested in group B. The existence of specific hypotheses allowed the use of one-tailed probability tests on the data from group B. Testing multiple hypotheses unfortunately raises the problem that some results are accepted as significant by chance alone. This potential problem may be treated in a conservative manner by dividing the level of significance by the number of hypotheses (according to Bonferroni). This regimen was applied to each factor investigated (e.g. age, sex, glucose) since the set of hypotheses of each factor in question could be regarded as being independent from that of other factors. The statistical procedure is exemplified in Table 3b. RESULTS Typical chromatograms from a standard solution and a human C S F sample are shown in Fig. 1. The mean concentrations and standard deviations of 19 free amino acids and ethanolamine measured in the

27

C S F samples of group A and B are listed in Table 1. These values are within the range of previously published normal values (Gjessing et al., 1972; Ferraro and Hare, 1984; Kruse et al., 1985). Parameters investigated with respect to amino acid levels in C S F are characterised in Table 2. There was a significant positive correlation between C S F glucose and glutamate levels (Fig. 2, Table 3a and Table 3b). The correlation coefficients and levels of significance of glucose versus the different amino acids are given in Table 3a for both groups. Overall, there was no significant correlation between single amino acids and time of day, sex, age, position at time of spinal tap or C S F total protein (Table 4). However, regarding group A for example, there were significant correlations between the different factors of influence and some amino acids which could not be confirmed in group B and vice versa (Table 4). For time of spinal tap there was no significant influence on any C S F amino acid in group A or group B. For this reason these results were not included in Table 4. DISCUSSION In contrast to the extensive work on C S F amino acid levels in a number of pathological states, little is known about physiological factors that might influence the normal C S F amino acid composition. Previous reports have dealt with the influence of single parameters (e.g. age, sex) on the amino acid levels in CSF. In the present study the influence of age, sex, time and position of lumbar puncture, Table 1. Amino acid and ethanolamine concentrations in the CSF samples of group A and B Group A Group B (n = 36) (n = 19) Glutamate 2.11 + 0.54 1.84 + 0.37 Asparagine 6.17 + 1.42 5.01 + 0.82 Serine 21.38 + 5.59 19.95 + 2.60 Glutamine 445.54 _+83.47 443.04 _+30.00 Histidine 15.37 + 3.59 9.85 + 2.00 Glycine 5.81 + 1.84 5.36 + 1.59 Threonine 24.99 + 7.25 20.91 + 4.00 Citrulline 2.01 _+0.81 1.91 + 0.59 Taurine 6.22 + 1.72 6.88 _+1.55 Arginine 20.53 + 3.73 20.54 + 2.70 Alanine 27.86 + 8.70 39.42 + 12.84 Tyrosine 10.93 + 2.31 17.72 + 2.79 Ethanolamine 17.71 _+4.57 9.30 + 1.72 Methionine 3.29 + 1.69 2.01 + 0.44 Phenylalanine 16.25 + 4.56 12.73 _+3.47 Valine 18.33 _+4.50 16.33 + 3.27 Isoleucine 3.79 + 1.25 4.43 + 1.34 Leucine 14.91 + 4.00 8.46 + 2.27 Ornithine 3.99 + 1.02 3.70 _ 0.62 Lysine 22.84 + 5.01 19.07 + 3.74 Values are given as mean _+SD (nmol/ml).

28

JOHANNES KORNHUBER el a/. Table 2. Pattern of the investigated groups A and B (cf text) with respect to the analysed factors with possible influence on CSF amino acids

n Sex (male/female) Age in years (range) Patient positionat spinal tap (lying/sitting) Time of spinal tap (morning/midday/afternoon)* CSF glucose(mmol/l) CSF protein(rag/l)

Group A 36 26/10

Group B 19 13/6

39.44+ 14.83 20-72

35.47 _+15.58 19 65

25/t 1 18/10/8

10/9 8/7/4

3.64 + 0.37 309.17 _+79.87

3.67 + 0.22 293.16+ 60.I0

*Morning8.00-11.30am; midday 11.31am-2.30pm; afternoon2.31-5.00pm. Values are expressed as mean_+SD where possible.

glucose and protein concentration on amino acid and ethanolamine levels in human lumbar CSF has been investigated, using conservative statistical reasoning. The predominant finding in this investigation is the relation between glucose and glutamate levels in the CSF. The cause for this correlation is not yet clear. Glucose is an important metabolic precursor of CNS glutamate (Gaitonde et al., 1964; Fonnum, 1984) including the releasable glutamate pool (Potashner, 1978; Hamberger et al., 1979; Ward et al., 1983). A net efflux of glutamate from brain to blood (Pardridge, 1979) indicates that a relatively high amount of newly synthesised glutamate is continuously released from the brain. CSF glutamate

00Q Q

eoe-e

o

o

o

eo

8

ol •

o e *

o ".

"SQ

• !

8o ° o

i~o

I

*;'l

D

Dillon (niniel/ll

Fig. 2. Plot of glucose versus glutamate values for the CSF samples of group A (closed circles) and group B (open circles).

may, therefore, originate in large part from CNS glutamate pools that are influenced indirectly by glucose. In agreement with the positive correlation between CSF glucose and glutamate, a decrease of brain glutamate levels is found in insulin-induced hypoglycemia in rats (Lewis et al., 1974; Engelsen and Fonnum, 1983) and mice (Gorell et al., 1976). Furthermore, also brain levels of glutamine and alanine decrease with hypoglycemia (Lewis et al., 1974). In this study we found no association between CSF glucose and alanine or glutamine because of the conservative statistical reasoning. However, looking at alanine in Table 3a there are weak, but consistent correlations between CSF alanine and glucose in both groups of patients. Further experiments designed to clarify these interactions are necessary. In addition, CSF amino acids should be measured in diabetic patients. There are several reports on changes in amino acid content in CNS tissue of different age (Timiras et al., 1973; Price et al., 1981; Rajeswari and Radha, 1984). Furthermore, in spinal fluid significant but not consistent changes in amino acid levels with aging have been published (Gjessing et al., 1974; Lakke and Teelken, 1976; McGale et al., 1977; Goodnick et al., 1980; Ferraro and Hare, 1985). In our investigation there are significant correlations between age and some amino acids in group A which cannot be confirmed in group B and vice versa. However, no significant correlation between age and a single amino acid remains if hypotheses are derived from one group and subsequently tested using the complementary group. This result reflects the recent inconsistent findings. Therefore, we feel that the influence of age on amino acid levels in CSF is of minor relevance. Sex-related influences on CSF amino acid levels

29

CSF amino acids Table 3a. Correlations between CSF glucose and CSF amino acid concentrations Group A

Group B

r

P

r

(two-tailed) Glutamate Asparagine Serine Glutamine Histidine Glycine Threonine Citrulline Taurine Arginine Alanine Tyrosine Ethanolamine Methionine Phenylalanine Valine Isoleucine Leucine Ornithine Lysine

0.40 0.23 0.21 0.32 0.32 - 0.13 0.37 0.27 0.10 0.24 0.41 0.34 0.06 0.17 0.38 0.28 0.31 0.37 0.25 0.39

P (two-tailed)

0.015 0.181 0.226 0.056 0.055 0.440 0.026 0.115 0.548 0.166 0.013 0.044 0.707 0.313 0.023 0.104 0.069 0.025 0.135 0.018

0.58 0.57 0.04 0.47 0.40 0.35 0.40 0.59 0.13 0.36 0.48 0.21 0.03 0.17 0.38 0.17 0.22 0.26 -0.03 0.27

0.010 0.010 0.862 0.042 0.086 0.138 0.089 0.008 0.606 0.129 0.037 0.378 0.902 0.482 0.110 0.477 0.360 0.277 0.899 0.266

Rank correlations (r) and levels of significance (P) for the amino acids versus glucose for group A and group B.

Table 3b. The table demonstrates the procedure of statistical reasoning (cf methods) using data of Table 3a Group A

Group B

r

P (two-tailed)

Glutamate Threonine Alanine Tyrosine Phenylalanine Leucine Lysine

0.40 0.37 0.41 0.34 0.38 0.37 0.39

r

P (one-tailed)

0.015 0.026 0.013 0.044 0.023 0.025 0.018

0.58 0.40 0.48 0.21 0.38 0.26 0.27

0.005* 0.045 0.019 0.189 0.055 0.139 0.133

Results are regarded as significant if P < 0.05. Seven of the 20 performed tests in group A have a P < 0.05 (two-tailed) and only these correlations are shown in this table. These 7 correlations were used as hypotheses in group B. The tests in group B were performed one-tailed (using P/2) because the hypothesis included the sign of the correlation. The levels of significance had to be adjusted because of multiple testing. The adjusted level of significance (according to Bonferroni) for 7 hypotheses is 0.05/7 = 0.00714. According to this procedure only glutamate remains being significantly correlated with glucose.

have previously been reported (Lakke and Teelken, 1976; McGale et al., 1977; Goodnick et al., 1980; Hagenfeldt et al., 1984; Ferraro and Hare, 1985). A comparison of these studies again reveals no consistent relation between gender and amino acid levels. A similar inconsistency is derived from the present data. Nevertheless, future studies in this field should address the influence of different stages during the menstrual cycle, since amino acid levels in different rat brain nuclei appear to be correlated with different stages of the estrous cycle (Frankfurt et al., 1984).

It has been proposed that amino acid content (Choma et al., 1979; Ross et al., 1980) and release (Dodd and Bradford, 1974; Perlow et al., 1979) from mammalian CNS may exhibit diurnal rhythms. No consistent influence of time of lumbar puncture on amino acid levels in CSF has been detected. The existence of a diurnal rhythm is not, however, disproved by our data, since the CSF specimens have been collected during day time only. In cases of blood-CSF barrier dysfunction (i.e. enhanced CSF/serum albumin ratio) with enhanced

30

JOHANNES KORNHUBER el a[.

Table 4. Coefficients of correlation between CSF amino acids and factors of possible influence (CSF protein, age. sex. patient position al spinal tap) are tabulated for both group A and group B CSF protein A Glutamate Asparagine Serine Glutamine Histidine Glycine Threonine Citrulline Taurine Arginine Alanine Tyrosine Ethanolamine Methionine Phenylalanine Valine tsoleucine Leucine Ornithine Lysine

0.03 0.28 0. I 1 0.29 0.33 0.17 0.22 0.21 0.19 0.33 0.27 0.35 0.15 0.23 0.31 0.36 0.09 0.10 0.03 0.28

Age B

0.21 0.22 - 0.18 0.30 0.28 0.34 0.02 0.50 0.15 0.17 0.40 - 0.20 0.04 0.13 0.16 0.21 - 0.14 0.02 0.16 0.21

Patient position at spinal tap

Sex

A

B

0.15 0.44 0.43 0.32 0.29 0.33 0.20 0.23 0.44 0.51 0.27 0.28 0.19 0.46 0.44 t).31 0.16 0.13 0.16 0.44

0.23 0.18 0.48 I).29 0.30 0.21 0.01 0.18 0.10 0.01 0.12 0.32 0.55 0.35 - 0.38 0.07 0.57 0.51 0.01 0.33

A

B

A

B

0.30 I).06 0.08 0.40 0.14 0.08 0.02 0.27 0.05 0.18 0.32 0.31 0.28 0.13 0.21 0.27 0.38 0.40 0.29 I).26

0.19 0.08 0.04 0.04 0.13 0.25 0.55 0. I I 0.42 0.1 I 0.22 I).22 0.45 0.01 I).34 I).07 0.21 0.36 0.02 0.08

0.15 0.23 0.31 0.32 0.27 0.09 023 0.16 0.10 0.24 0.10 0.26 0.02 0.22 0.1 I 0.16 0.02 0.10 0.19 0.15

0.26 0.0 I 0.13 0.01 0.13 0.33 I).22 0.44 0.18 I).13 1).2I 0.06 0.47 I).1)3 0.18 0.10 0.24 0.29 0.119 0. I 1

For age and CSF protein rank correlations are given. For the dichotomous variables, gender and patient position at spinal tap, product-moment correlation coefficients are tabulated (positive correlation coefficients indicate higher CSF amino acid levels in females or in sitting position at spinal tap). C S F total protein, there m a y be e n h a n c e d C S F a m i n o acid c o n c e n t r a t i o n s ( K r u s e et al., 1985). In this s t u d y we d e m o n s t r a t e t h a t u n d e r n o r m a l c i r c u m stances, there is n o c o r r e l a t i o n b e t w e e n a m i n o acid c o n c e n t r a t i o n s a n d the C S F total protein. T h e intent o f o u r investigation w a s to d e t e r m i n e the influence o f different in vivo variables o n the c o n c e n t r a t i o n o f individual a m i n o acids a n d e t h a n o l a m i n e in h u m a n C S F . Since multiple c o m p a r i s o n s have been p e r f o r m e d c o n s e r v a t i v e statistical p r o c e d u r e s have been used to avoid acceptance o f s o m e results by c h a n c e alone. As a c o n s e q u e n c e , true b u t w e a k f a c t o r s o f influence ( s h o w n as weak b u t c o n s i s t e n t c o r r e l a t i o n s in T a b l e 3a) m a y have been rejected by the statistical limits. O u r results s h o w that the significance o f altered C S F a m i n o acid levels for different disease states m a y be i n t e r p r e t e d w i t h o u t r e g a r d to the effect o f the m a j o r i t y o f the variables we tested. Acknowledgements--We would like to thank Dr Kriebel for generous support and Dr Wirschin for his cooperation with respect to CSF sampling under myelography. Furthermore, we are indepted to H. ZettlmeiBl and Dr Blome for the facility to carry out the HPLC measurements and Dr Erickson for reading the manuscript. J.K is a recipient of a D F G scholarship. REFERENCES

Choma P. P., Puri S. K. and Volicer L. (1979) Circadian rhythm of cyclic nucleotide and GABA levels in the rat brain. Pharmacology 19, 307 314.

Dodd P. R. and Bradford H. F. (1974) Release of amino acids from the chronically-superfused mammalian cerebral cortex. J. Neurochem. 23, 289 292. Engelsen B. and Elsayed S. (1984) Increased concentration of aspartic acid in the cerebrospinal fluid of patients with epilepsy and trigeminal neuralgia: An effect of medication? Acta Neurol. Scand. 69, 71~76. Engelsen B. and Fonnum F. (1983) Effects of hypoglycemia on the transmitter pool and the metabolic pool of glutamate in rat brain. Neurosci. Lett. 42, 317 322. Ferraro T. N. and Hare T. A. 11984) Triple-column ionexchange physiological amino acid analysis with fluorescent detection: Baseline characterization of human cerebrospinal fluid. Anal. Biochem. 143, 82-94. Ferraro T. N. and Hare T. A. 11985) Free and conjugated amino acids in human CSF: Influence of age and sex. Brain Res. 338, 53 60. Fonnum F. (1984) Glutamate: A neurotransmitter in mammalian brain. J. Neurochem. 42, 1 11. Frankfurt M., Fuchs E. and Wuttke W. 11984) Sex differences in 7-aminobutyric acid and glutamate concentrations in discrete rat brain nuclei. Neurosci. Lett. 50, 245 -250. Gaitonde M. K., Marchi S. A. and Richter D. (1964) The utilization of glucose in the brain and other organs of the cat. Proc. R. Soc. B 160, 124-136. Gjessing L. R., Gjesdahl P., Dietrichson P. and Presthus J. 11974) Free amino acids in the cerebrospinal fluid in old age and in Parkinson's disease. Eur. Neurol. 12, 33 37. Gorell J. M., Dolkart P. H. and Ferrendelli J. A. (1976) Regional levels of glucose, amino acids, high energy phosphates, and cyclic nucleotides in the central nervous system during hypoglycemic stupor and behavioral recovery. J. Neurochem. 27, 1043 1049. Goodnick P. J., Evans H. E., Dunner D. L. and Fieve R. R. 11980) Amino acid concentrations in cerebrospinal

CSF amino acids fluid: Effects of aging, depression, and probenecid. Biol. Psychiat. 15, 557-563. Hagenfeldt L., Bjerkenstedt L., Edman G., SedvaU G. and Wiesel F.-A. (1984~ Amino acids in plasma and CSF and monoamine metabolites in CSF: Interrelationship in healthy subjects. J. Neurochem. 42, 833-837. Hamberger A. C., Chiang G. H., Nylen E. S., Scheff S. W. and Cotman C. W. (1979) Glutamate as a CNS transmitter. I. Evaluation of glucose and glutamine as precursors for the synthesis of preferentially released glutamate. Brain Res. 168, 513-530. Iijima K., Takase S., Tsumuraya K., Endo M. and Itahara K. (1978) Changes in free amino acids of cerebrospinal fluid and plasma in various neurological diseases. Tohoku J. Exp. Med. 126, 133 150. Kornhuber J., Kornhuber M. E., Hartmann G. M. and Kornhuber A. W. (1986) Positive Korrelation zwischen Glucose und Glutamat im Liquor cerebrospinalis des Menschen. Dtsch. med. Wschr. 111, 1422. Kruse T., Reiber H. and Neuhoff V. (1985) Amino acid transport across the human blood-CSF barrier. An evaluation graph for amino acid concentrations in cerebrospinal fluid. J. Neurol. Sci. 70, 129-138. Lakke J. P. W. F. and Teelken A. W. (1976) Amino acid abnormalities in cerebrospinal fluid of patients with parkinsonism and extrapyramidal disorders. Neurology 26, 489~,93. Lewis L. D., Ljunggren B., Norberg K. and Siesj6 B. K. (1974) Changes in carbohydrate substrates, amino acids and ammonia in the brain during insulin-induced hypoglycemia. J. Neurochem. 23, 659~71. Lindroth P. and Mopper K. (1979) High performance liquid chromatographic determination of subpicomole amounts of amino acides by precolumn fluorescence derivatization with o-phthaldialdehyde. Anal. Chem. 51, 1667-1674. McGale E. H. F., Pye I. F., Stonier C., Hutchinson E. C. and Aber G. M. (1977) Studies of the inter-relationship between cerebrospinal fluid and plasma amino acid concentrations in normal individuals. J. Neurochem. 29, 291-297. Mutani R., Monaco F., Durelli L. and Delsedime M. (1974) Free amino acids in the cerebrospinal fluid of epileptic subjects. Epilepsia 15, 593-597. Oepen G., Cramer H., Bernasconi R. and Martin P. (1982) Huntington's disease--Imbalance of free amino acids in the cerebrospinal fluid of patients and offspring at-risk. Arch. Psychiatr. Nervenkr. 231, 131-140. Pardridge W. M. (1979) Regulation of amino acid availability to brain: Selective control mechanisms for glutamate, In Glutarnic Acid: Advances in Biochemistry and Physiology, (Filer L. J. Jr, Garattini S., Kare M. IL,

31

Reynolds W. A. and Wurtman R. J., eds), pp. 125-137. Raven Press, New York. Perlow M. J., Enna S. J., O'Brien P. J., Hoffman H. J. and Wyatt R. J. (1979) Cerebrospinal fluid gammaaminobutyric acid: Daily pattern and response to haloperidol. J. Neurochem. 32, 265-268. Perry T. L., Hansen S. and Kennedy J. (1975) CSF amino acids and plasma-CSF amino acid ratios in adults. J. Neurochem. 24, 587-589. Perry T. L., Hansen S., Lesk D. and Kloster M. (1973) Amino acids in plasma, cerebrospinal fluid and brain of patients with Huntington's chorea. Adv. Neurol. 1, 603~18. Perry T. L. and Jones R. T. (1961) The amino acid content of human cerebrospinal fluid in normal individuals and in mental defectives. J. clin. Invest. 40, 1363-1372. Plum C. M. (1974) Free amino acid levels in the cerebrospinal fluid of normal humans and their variation in cases of epilepsy and Spielmeyer-Vog~Batten disease. J. Neurochem. 23, 595-600. Potashner S. J. (1978) The spontaneous and electrically evoked release, from slices of guinea-pig cerebral cortex, of endogenous amino acids labelled via metabolism of D-[U-14C]glucose. J. Neurochem. 31, 177 186. Price M. T., Olney J. W. and Haft R. (1981) Age-related changes in glutamate concentration and synaptosomal glutamate uptake in adult rat striatum. Life Sci. 28, 1365-1370. Rajeswari T. S. and Radha E. (1984) Metabolism of the glutamate group of amino acids in rat brain as a function of age. Mechanisms Ageing Dev. 24, 139-149. Ross F. H., Sermons A. L. and Walker C. A. (1980) Twenty-four hour daily rhythms of glycine in the rodent brain. Life Sci. 26, 1019-1022. Smith C. C. T., Bowen D. M., Francis P. T., Snowden J. S. and Neary D. (1985) Putative amino acid transmitters in lumbar cerebrospinal fluid of patients with histologically verified Alzheimer's dementia. J. Neurol. Neurosurg. Psychiat. 48, 469~471. Timiras P. S., Hudson D. B. and Oklund S. (1973) Changes in central nervous system free amino acids with development and aging, In: Progress in Brain Research, (Ford D. H., ed.), Vol. 40, pp. 267-275. Elsevier, Amsterdam. Ward H. K., Thanki C. M. and Bradford H. F. (1983) Glutamine and glucose as precursors of transmitter amino acids: Ex vivo studies. J. Neurochem. 40, 855-860. Zettlmeil31 H., Blome J. and Kornhuber H. H. (1986) A sensitive, fast, highly selective method for the determination of amino acids and biogenic amines in the cerebrospinal fluid and other body fluids and tissues. Arch. Ital. Biol. 124, 129-132.