- Email: [email protected]
Comparative analysis of indices of central dopaminergic functions in man
Life Sciences, Vol° 32, pp. 2667-2676 Printed in the U.S.A.
Pergamon Pres~
COMPARATIVE ANALYSIS OF INDICES OF CENTRAL DOPAMINERGIC FUNCTIONS IN MAN Gyorgy Bagdy, Mildly Aratd*, Krisztina~ and Marton I.K. Fekete z
Baraczka 1
*National Institute for Nervous and Mental Diseases, Budapest 27, Pf. i. 1281 Hungary. iDept. Psychiatry, Semmelweis Medical School, Budapest. 2Institute Experimental Medicine, Hungarian Academy of Sciences, Budapest, Hungary. (Received in final form March 24, 1983) SUMMA~ Various postulated indices of central dopaminergic activity - cerebrospinal fluid (CSF) dopamine (DA), dihydroxy-phenylacetic acid (DOPAC), homovanillic acid (HVA), noradrenaline (NA), plasma NA, serum prolactin, serum dopamine-~-hydroxylase (DBH), and platelet monoamine oxidase (MAO) activity - were measured in 30 drug-free inpatients. The mean values and the ranges were similar to those described in the literature. Plasma NA showed significant positive correlation with age. Significant positive correlation was found between CSF DA and its metabolites DOPAC and HVA. Serum DBH activity showed a slight but significant inverse correlation with CSF DA and its two metabolites. CSF NA showed a significant positive correlation with CSF DOPAC, but only in females. Serum DBH activity had no significant correlation either with CSF or with plasma NA levels. These findings suggest that either CSF HVA or DOPAC and DA may be useful indicators of DA metabolism in humans. Serum DBH activity may be in relationship with the central dopaminergic functions. Abnormalities of dopamine (DA) neurotransmission have been postulated as the underlying pathophysiological mechanism in several meuropsychiatric diseases e.g. Parkinson's disease (i), Huntington's chorea (2), schizophrenia (3). The following possibilities exist for the assessment of central dopamin~ ergic functions in clinical studies: measurement of DA and of its metabolite~ in different body fluids, determination of catecholamine regulatory enzymes activity and the measurement of serum hormone levels primarily under dopaminergic control. However, the relationships of these indicators and their drug. induced changes have scarcely been studied. The aim of the present study was to investigate the correlation between postulated indices of central do~aminergic mechanism: cerebrospinal fluid (CSF) DA, dihydroxy-phenylacetic acid (DOPAC), homovanillic acid (HVA), noradrenaline (NA), plasma NA, serum prolac tin ~RL), serum dopamine-O-hydroxylase (DBH) and platelet monoamine oxidase (MAO) activity. Determination of CSF DA in clinical studies has not been described in the literature (4). This may be due to technical reasons, but in addition there is not sufficient evidence that the DA level is a suitable index of dopaminer * Reprint request t0:-M. Arat6, M.D., P h . D . , Hamilton Psychiatric P.O. Box 585, Hamilton, Ontario. Canada. L8N 3K7 0024-3205/83/232667-10503.00/0 Copyright (c) 1983 Pergamon Press Ltd.
Hospital,
2668
Dopaminergic Indices in Man
Vol. 32, No. 23, 1983
gic functions. Clinical studies of the level of DOPAC in the CSF have been made only sporadically, probably also for technical reasons. Further aspects have also to be considered: a. DOPAC is an intermediate DA metabolite and therefore could be a more reliable index of DA metabolism than either DA itself or the final metabolic product of DA HVA. b. However, DOPAC is only a minor metabolite in man~ c., and constitutes only one pathway of DA catabolism. The measurement of HVA is the most frequently used and accepted approach . It should, however, be borne in mind that its actual level is not only determined by its formation, but also by its elimination. This problem can be overcome to a certain extent by the use of the probenecid technique (5). DBH is the enzyme catalyzing the conversion of DA to NA. Experimental studies have demonstrated a correlation between DBH activity and noradrenergic functions. However, recent human studies have failed to support this finding, and accordingly serum DBH activity cannot be considered as an index of peripheral sympathetic activity (6, 7). Our previous study suggested a relationship between serum DBH activity and central dopaminergic functions (8). We have measured NA in the CSF and plasma because NA is the step succeeding DA in catecholamine metabolism, and it is postulated that the above mentioned "dopaminergic" diseases are accompanied by disturbance of NA metabolism (9, i0), and noradrenergic system modulates dopaminergic functions (Ii). MAO the enzyme responsible for the oxidative deamination of monoamines is an important factor in neurotransmitter regulation. Platelet MAO activity is associated with the activity of the type B MAO in the brain (12). The relationship of platelet MAO activity to serum PRL suggests a correlation between platelet ~ 0 activity and central dopaminergic functions (13). Tonic dopaminergic inhibition plays a fundamental role in the control of PRL secretion (14). Thus, determination of serum PRL levels may provide indirect information about central dopaminergic activity. Different parameters have been selected throughout the literature and therefore the results are difficult to compare. Levels of the mentioned parameters have never been determined in the same population, The aim of the present study was to seek evidence of correlations of the postulated dopaminergic indices, to find out how many are interrelated and to what degree, and to clarify whether the individual indices could be substituted for each other. Material and methods Subjects Inpatients of neurologic departments in whom lumbar puncture was performed for diagnostic purposes were studied with the patients' consent. This mixed group consisted of patients with illnesses thought not to affect catecholamine metabolism (e.g. cervical spondylosis, herniated disk, peripheral neuralgia, etc.), and patients who were admitted for neurological assessment but in whom no abnormality was detectable. Only CSF samples with normal laboratory values (cell counts, protein, glucose, colloid lability tests) were evaluated. The patients were drug-free, apart from 9 females and 1 male who received small doses of benzodiazepines, mainly i0 mg nitrazepam, as night sedation. Thirty patients (24 females and 6 males) complied with these requirements and their ages ranged between 24 and 62 years. All samples were obtained within 3 days, between 8 A~ and 9 AM after a i0 hour fast and bed rest. The CSF was collected in plastic tubes and immediately frozen in dry ice. Two ml aliquots were used: - the first for routine laboratory tests,
Vol. 32, No. 23, 1983
Dopaminergic Indices in Man
2669
the second for DA, DOPAC and NA determination and third for the measurement of HVA. Blood amples for the assessment of plasma NA, serum PRL, DBH and platelet MAO were taken at the same time as the CSF samples. Methods CSF DA, DOPAC, NA and plasma NA levels were determined by a radioenzymatic method described by Fekete et al. (15). Twenty-five m l of plasma containing EGTA and glutathione and 50 ul of CSF were incubated in the presence of catechol-O-methyl-transferase (COMT) and 3H-S-adenosylmethionine. The Omethylated products were extracted with organic solvents in the presence of sodium tetraphenylborate (16) and chromatographed. The radioactivity of the authentic spots was measured and the concentration of COMT substrates (DA, DOPAC, NA) in the CSF calculated. The O-methylated derivates of NA were oxidized as described by Peuler and Johnson (17). The sensitivity of the method varied between 1-50 pg/reaction, corresponding to concentrations of 0.02-0.08 ng/ml CSF. Homovanillic acid was measured by the methods of Westerink and Korf (18) with some modification (19). Serum DBH activity was determined according to the method of Nagatsu and Udenfriend (20). Platelet MAO activity was assayed as follows: Aliquots of 8.5 ml venous blood were collected in plastic tubes containing 1.5 ml ACD solution. After standing for 15 min at room temperature the samples were centrifuged at i00 g for 15 min. The supernatant was drawn off and cooled. After recentrifugation at 200 g for 15 min the supernatants were pooled. Platelet rich plasma was centrifuged at 1500 g for 20 min, the supernatant aspired, and the platelets were washed with 2 ml saline. This was followed by centrifugation at 1500 g for 20 min. The supernatant was discarded, the platelets were taken up in 0.05 M, pH 7.4 phosphate buffer solution, and homogenized. Measurements were made using 14C-tyramine as substrate in a concentration of 1.3x10 -4 M/I. InGubation was carried out in the presence of EDTA at pH 7.4 and 37 ° C for 20 minutes. Protein concentration measured by the method of Lowry (21) was 5001 0 0 0 ~ g / m l and showed linear correlation with enzyme activity. The formed aldehydes were extracted with ethylacetate. Measurements were performed in a Packard liquid scintillation spectrometer after the addition of a dioxane cocktail. Intraassay scattering was 3.9%. Prolactin was determined by radioimmunoassay using a kit - IRE. For the amines, their metabolites and PRL values are given in ng/ml. DBH activity was expressed as nmol octopamine /min/ml serum. MAO activity is given as nmol/mg protein/hour. All neurochemical measurements were performed in one assay, the amines and their metabolites were measured within one we~k from sampling. Results Table I presents the ranges, means~SEM of the postulated dopaminergic indices measured and a correlation matrix is presented in Table II. The mean values of the measured parameters in patients taking benzodiazepines (n=lO) did not differ from those of the others. Plasma NA levels correlated positively with age, but the relationship was significant only in females (r=0.59, p (O.O1). A significant positive correlation was found between CSF DA and its metabolites (DOPAC and HVA) and between the two metabolites (Fig. i).
2670
Dopaminergic Indices in Man
Vol. 32, No. 23, 1983
TABLE I Mean Values and Ranges of Dopaminergic Indices CSF DOPAC Females n=
23
CSF DA
CSF NA
CSF HVA
plasma NA
serum P RL
21
24
21
24
24
platelet MAO
serum DB H
24
24
mean ±SEM
1.91 0.14
2.81 0.23
0.16 0.01
19.9 1.3
0.30 0.03
6.5 1.3
12.9 1.3
27.0 3.4
range
0.33.98
1.36.3
0.060.32
ii31
0.080.73
0.528
3.529.6
1.576.2
6
6
6
6
6
6
6 21.6 5.2
Males 6
n= mean
+_SEM range
3.12 0.68
2.92 0.39
0.16 0.03
22.9 4.1
0.23 0.02
7.3 3.3
8.7 1.2
1.65.5
1.54.2
0.030.25
12.534.5
0.150.30
2.323.5
3.711.8
8.137
Concentrations of amines (DA,NA), metabolites (DOPAC,HVA) and PRL given as ng/ml. Activities of MAO and DBH given as nmol/mg protein/hour and nmol octopamine /min/ml serum, respectively.
CSF
DOPAC
nglml 67
females •
males
n=20
n=6
r = 0,75
r = 0,78
o O
p< 0,001
0
• •
•
I
00
0
•
CSF
HVA
FIG. I. Significant positive correlation between CSF HVA and DOPAC.
Vol. 32, No. 23, 1983
Dopaminergic Indices in Man
2671
TABLE II Correlation matrix of dopaminergic indices Females
n=24
CSF DA
O. 7 2 x x x
CSF NA
0.62 x x x
0.35
CSF HVA
0.75 xxx
0.58 xx
plasma NA
0.09
0.18
-0.01
0.42
serum PRL
-0.20
-0.i0
-0.08
0.07
0.39
platelet MAO
-0.18
0.14
-0.21
-0.33
-0.16
0.05
-0.45 x
-0.39
-0.38
-0.07
-0.03
-0.23
CSF NA
CSF HVA
plasma NA
serum PRL
platelet MAO
serum DBH
-0.47 x
CSF DOPAC Males CSF DA
-CSF DA
0.38
n=6 0.89 xx
CSF NA
-0.40
CSF HVA
0.78
-0.52
0.87 x
-0.53
plasma NA
-0.20
-0.ii
0.73
0.05
serum pRL
0.04
0.03
0.56
-0.28
0.23
platelet MAO
-0.i8
-0.35
0.ii
-0.72
0.44
serum DBH
-0.56
-0.74 x: p 40.05;
0.29
0.07 -0.91 xx -0.25 0.35 xx: p ~0.02; xxx: p < 0 . 0 1
0.86 x
Serum DBH activity in females showed a significant inverse correlation with CSF DA (Fig. 2) and DOPAC and inverse tendency with CSF HVA. Similar tendencies were found also in males. Due to the small number of males their data are given only as complementary information. Significant positive correlation was found between CSF NA and CSF DOPAC and positive tendency between CSF NA and CSF HVA levels, but only in females.
2672
Dopaminergic
CSF
Indices
in M a n
Vol.
32, No.
23,
DA
ng/ml 7,
females
•
mele8
0
n=21
n=6
r = -0,45
r : ~0,74
p-O,O5
5-
O
41
O
•
ee
•
•
0
o
• o
•
0
serum DBH
o
~'o
2"o
3'o
4i~
s3
FIG. Significant plem
inverse
correlation
6'0
7'o
8'0 (u)
2.
between
serum DBH activity
and CSF DA
NA
0.7-
0.6~
e•
0,5
N.S.
0,4
o,3
o •
o
o
•
0
~2 0 oI
0,1
serum DBH
0
15
25
35
4o
~ FIG.
No significant correlation between NA level, n=30, 24 females - I,
i0
7o
~
(u)
3. serum DBH activity 6 males - o.
and plasma
1983
Vol. 32, No. 23, 1983
Dopaminergic
Indices in Man
2673
No relationship was found between serum DBH activity and the NA level of either CSF or plasma (Fig. 3). Discussion The means and ranges of the CSF DA metabolites reported here correspond to published data (22,23). The values of CSF and plasma NA are also similar to those described in the literature (24,25,26,27,28). It is a limitation of these data that the used assay procedure measures only the free and not the conjugated catecholamines and metabolites, and it has recently been shown that catecholamines in plasma are for the major part present as conjugated derivatives (29), and conjugates are present also in human CSF (30). However, in spite of individual variations, statistically significant correlations were found between the concentrations of free and conjugated DOPAC and HVA in CSF (30). The high level of CSF DA compared to CSF NA was surprising. A hypothetic explanation to it may be that the main source of CSF DA is the DA-rich nucleus caudatus lying closest to CSF channels, and the DA/NA ratio is very high in this region (31). The mean values of serum DBH activity in the present study correspond to published data (20,32,33). The platelet MAO activity values were in accordance with the results of our earlier study: normal female controls, n=34: 12.6J6.1 (SD), males, n=70: i0.5~4. Serum PRL levels were in the normal range, with the exception of 1 female and 1 male patient whose PRL values were above the normal limit (27.7 and 23.5 ng/ml respectively). Since the ages of the subjects ranged from 24 to 62 years, possible age correlations were calculated. Significant age dependence was found only in plasma NA levels of females. This data is in harmony with the finding of Ziegler et al. (34). CSF HVA showed rising and CSF DOPAC showed decreasing tendency with age. CSF DA and its two metabolites (DOPAC and HVA) were significantly positively correlated. This finding suggests that not only baseline HVA and DOPAC,but also CSF DA level may be a suitable index of central DA activity. It remains to be seen if these relationships held under all circumstances for example drug treatment. However, owing to the great individual variation of the data it seems to be more informative to have several measures of dopaminergic activity. The significant negative correlation between serum DBH activity and CSF DA and its metabolites has not been previously reported. The correlations were significant only in females, but similar tendencies were also found in males. This finding supports our earlier hypothesis that serum DBH activity may be related to central dopaminergic mechanisms (8). Serum DBH activity is intraindividually stable, and is mainly genetically determined and DA turnover shows dynamic changes. Therefore the observed correlation implies a role of DBH activity in central dopaminergic regulation. Another possible explanation is that like serum DBH activity, basal DA turnover is under genetic control (12), and both may be controlled by a common genetic factor. The significant positive correlation between CSF NA and DOPAC may support this hypothesis. Antelman and Caggiula (ii) demonstrated that NA modulates dopaminergic functions. It is conceivable that DBH regulates dopaminergic functions by this mechanism. Inhibition of DBH with disulflram, both in experimental animals (ii) and clinical studies (35), leads to dopaminergic overactivity and related behavioural changes and symptoms. This evidence of
2674
Dopaminergic Indices in Man
Vol. 32, No. 23, 1983
an inverse relationship between DBH and DA activity taken together with evidence of a negative correlation between DBH and indices of DA activity found in the present study may be relevant in studies into the pathogenesis of schizophrenia where both underactivity of NA and overactivity of DA have been implicated. For example Wise and Stein (36) demonstrated decreased DBH activity in the brains of schizophrenic patients in post mortem studies. Fujita et al. (37) and Arato et al. (33) have found decreased DBH activity in large populations of paranoid schizophrenic patients. Sternberg et al. (38) obtained good therapeutic response to neuroleptic treatment in schizophrenic patients with low CSF DBH activity, while the prognosis in the presence of high DBH activity was poor. Perenyi et al. (39) found serum DBH activity also to be of predictive value: the responders to neuroleptic therapy showing no extrapyramidal symptoms in their schizophrenic population had significantly lower serum DBH activity than the non-responders developing extrapyramidal symptoms. According to Lerner et al. (40) and our unpublished data, basal serum DBH and CSF DBH activities show significant positive correlation. In contrast to some earlier presumptions, numerous recent studies have fai~ ed to establish any relationships between DBH activity and plasma NA levels (41,42,43). This lack of relationship was confirmed in the present study as we have found no correlation between serum DBH activity and NA of plasma and CSF, even if making corrections of plasma NA values for age. These findings confirm that serum DBH activity is not an indicator of peripheral noradrenergic activity. Change in serum PRL level is a good index of the effects of DA antagonist and agonist drugs (44). However, no correlation can be expected to be between baseline serum PRL level and the CSF indicators of DA turnover (45) as PRL is controlled by the tuberoinfundibular dopaminergic system, while CSF DA and its metabolites derive mainly from the nigrostriatal system (4). Our data did not support the finding of Oreland et al. (12) that there is a positive correlation between platelet MAO activity and CSF concentration of HVA. According to our data platelet MAO activity rather correlates negatively to CSF HVA. This contradictory result requires further confirmation. Acknowledgement We thank Dr. I.N. Ferrier for his helpful comments on our text.
References I. 2. 3. 4. 5. 6. 7.
8. 9.
W.BIRKMAYER and O. HORNYKIEWICZ, Wien. klin. Wschr. 73 787-788 (1961). A. BARBEAU, Psych. Forum 4 8-15 (1973). A. RANDRUP and I. MUNKVAD, Psychopharmacol. ii 300-310 (1973). D.P. VAN KAMMER and D.E. STERNBERG, Neurobiology of .Cerebrospinal Fluid, ed. by J.H. Wood, 7.9-742, Plenum Publ. Corp., New York (1980). H.M. VAN PRAAG, Biol. Psychiat. 12 101-131 (1977). R.M. WEINSHILBOUM, Pharmacol. Re~. 30 133-166 (1979). J.M. STOLK, J.H.HURST, M.J. FRIEDMAN, P.O. HARRIS, D.A. VAN RIPPER and B.C. NISULA, E__nzymes and Neurotransmitters in Mental Disease, ed. by E. Usdin, T.L. Sourkes and M.B.H. Youdim, 171-192, John Wiley and Sons Ltd., Chichester (1980). G. BAGDY AND M. ARATO, Intergrative Neurohumoral Mechanisms~ ed by E. Endroczl, Akad~mia Kiad6, Budapest (In press). I.J. FARLEY, K.S. PRICE, E. MCCULLOGH, J.H.N. DECK, W. HODYNSKI and O. HORNYKIEWICZ, Science 200 456-458 (1978).
Vol. 32, No. 23, 1983
i0. ii. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.
29. 30. 31. 32. 33. 34. 35. 36. 37. 38.
39. 40. 41.
42. 43.
Dopaminergic Indices in Man
2675
C.R. LAKE, D.E. STERNBERG, J.C. BALLENGER, M.G. ZIEGLER, R.M. POST, l.J. KOPIN and W.E. BUNNEY, Science 207 331-333 (1980). S.M. ANTELMAN and A.R. CAGGIULA, Science 195 646-653.(1977). L. ORELAND, A. WIBERG, M. ASBERG, L. TRASKMAN, L. SJOSTRAND, P. THOREN, L. BERTILSSON and G. TYBRING, Psychiat. Res. 4 21-29 (1981). J.E. KLEINMAN, S. POTKIN, A. ROGOL, M.S. BUCHSBAUM, D.L. MURPHY, J.C. GILLIN, H.A. NASRALLAH and R.J. WYATT, Science 206 479-481 (1979). R.M. MACLEOD, Frontiers in Neuroendocrinology, ed by L. Martini and W. Ganong, 169-194, Raven Press, New York (1976). M.I.K. FEKETE, B. KANYICSKA and J.P. HERMAN, Life Sci. 23 1549-1556 (1978). M. DA PRADA and G. ZURCHER, Life Sci. 19 1164-1174 (1976). J.D. PEULER and G.A. JOHNSON, Life Sci. 21 625-636 (1977). B.H.C. SWESTERINK and J. KORF, Biochem. Med. 12 106-115 (1975). C.M. BANKI, J. Neurochem. 28 235-237 (1977). T. NAGATSU and S. UDENFRIEND, Clin. Chem. 18 980-983 (1972). O.H. LOWRY, N.J. ROSEBROUGH, A.L. FARR and R.J. RANDALL, J. Biol. Chem. 193 265-275 (1951). G. SEDVALL, B. FYRO, B. GULLBERG, H. NYBACK, F.A. WIESEL and B. WODEHELOODT, Brit. J. Psychiat, 136 366-374 (1980). S. WILK and M. STANLEY, Psychopharmacol. 57 77-81 (1978). R. LAKE, M.G. ZIEGLER and D.L. MIIRPHY, Arch. Gen. Psychiat. 34 553-556 (1977). G. PLANZ and R. PLANZ, Experientia 35 207 (1979). U.C.R. GOMES, B.C. SHANLEY, L. POTGIETER and J.T. ROUX, Brit. J. Psychiat. 137 346-351 (1980). R.M. POST, C.R. LAKE, D.C. JIMERSON, W.E. BUNNEY, J.H. WOOD, M.G. ZIEGLER and F.K. GOODWIN, Am. J. Psychiatry 135 907-912 (1978). D.P. VAN KAMMEN, D.E, STERNBERG, C.R. LAKE and P. LERNER, Frontiers of H0rmone Researqh, ed. by T.B. Van Wimersma Greidanus, 198-212, Karger, Basel (1982). M. DA PRADA, Trends Pharmacol. Sci. 1 157-159 (1980). P.A. BERGER, K.F. FAULL, J. KILKOWSK~, P.J. ANDERSON, H. KRAEMER, K.L. DAVIS and J.D. BARCHAS, Am. J. Psychiatry 137 174-180 (1980). P. NYBER, R. ADOLFSSON, N.-E. ANDEN and B. WINBLAD, Acta Neurol. Scandinav. 65 267-273 (1982). R.A. STONE, J.C. GUNNELS, R.R. ROBINSON, S.M. SHANBERG and N. KIRSCHNER, Circ. Res. 34 47-56 (1974). r M. ARATO, G. BAGDY and A. PERENYI, Biol. Psychiat. 17 853-854 (1982). M.G. ZIEGLER, C.R. LAKE and I.J. KOPIN, Nature 261 333-335 (1976). E. HARTMANN and M. KELLER-TESCHKE, Biol. Psychiat. 14 455-462 (1979). C.D. WISE and L. STEIN, Science 181 344-347 (1973). K. FUJITA, T. ITO, K. MARUTA, H. BEPPU, Y. NAKAGAMI and Y. KATO, J. Neur~ chem. 30 1569-1572 (1978). D.E. STERNBERG, D.P. VAN KAMMEN, J.C. BALLENGER, P. LERNER, S.R. MARDER and R.M. POST, Annual Meeting Amer. Psychiat. Assoc. New Res. Abstr. 32 (1978). A. PERENYI, G. BAGDY and M. ARATO, 10th Danube Symp. Psychiat. Magdeburg, Abstr. 112 (1982). P. LERNER, F.K. GOODWIN, D.P. VAN KAMMEN, R.M. POST, L.F. MAJOR, J.C. BALLENGER and W. LOVENBERG, Biol. Psychiat. 13 685-694 (1978). M. TANAKA, T. SANO, Y. KOHNO, T. YOKOTA, Y. MIZUKI and N. NAGASAKI, Prophylactic Approac h to Hypertensive Diseases, ed. by Y. Yamori, 321328, Raven Press, New York, (1979). B.A. BERKOWITZ, R. HEAD, T. JOH and J. HEMSTEAD, J. Pharmacol. Exptl. Ther. 213 18-23 (1980). S. SAARIKOSI, K. SELANDER, E. NUKALA and P. PYSTYNEN, Prostaglandins 20 487-492 (1980).
2676
44. 45.
Dopaminergic
Indices in Man
Vol. 32, No. 23, 1983
G. LANGER, E.J. SACHAR, R.S. NATHAN, M.A. TABRIZI, J.M. PEREL and F.S. HALPERN, Psychopharmacol. 65 161-164 (1979). K.J. ZANDER, B. FISCHER, R. ZIMMER and M. ACKENHEIL, Psychopharmacol. 73 43-47 (1981).
Pergamon Pres~
COMPARATIVE ANALYSIS OF INDICES OF CENTRAL DOPAMINERGIC FUNCTIONS IN MAN Gyorgy Bagdy, Mildly Aratd*, Krisztina~ and Marton I.K. Fekete z
Baraczka 1
*National Institute for Nervous and Mental Diseases, Budapest 27, Pf. i. 1281 Hungary. iDept. Psychiatry, Semmelweis Medical School, Budapest. 2Institute Experimental Medicine, Hungarian Academy of Sciences, Budapest, Hungary. (Received in final form March 24, 1983) SUMMA~ Various postulated indices of central dopaminergic activity - cerebrospinal fluid (CSF) dopamine (DA), dihydroxy-phenylacetic acid (DOPAC), homovanillic acid (HVA), noradrenaline (NA), plasma NA, serum prolactin, serum dopamine-~-hydroxylase (DBH), and platelet monoamine oxidase (MAO) activity - were measured in 30 drug-free inpatients. The mean values and the ranges were similar to those described in the literature. Plasma NA showed significant positive correlation with age. Significant positive correlation was found between CSF DA and its metabolites DOPAC and HVA. Serum DBH activity showed a slight but significant inverse correlation with CSF DA and its two metabolites. CSF NA showed a significant positive correlation with CSF DOPAC, but only in females. Serum DBH activity had no significant correlation either with CSF or with plasma NA levels. These findings suggest that either CSF HVA or DOPAC and DA may be useful indicators of DA metabolism in humans. Serum DBH activity may be in relationship with the central dopaminergic functions. Abnormalities of dopamine (DA) neurotransmission have been postulated as the underlying pathophysiological mechanism in several meuropsychiatric diseases e.g. Parkinson's disease (i), Huntington's chorea (2), schizophrenia (3). The following possibilities exist for the assessment of central dopamin~ ergic functions in clinical studies: measurement of DA and of its metabolite~ in different body fluids, determination of catecholamine regulatory enzymes activity and the measurement of serum hormone levels primarily under dopaminergic control. However, the relationships of these indicators and their drug. induced changes have scarcely been studied. The aim of the present study was to investigate the correlation between postulated indices of central do~aminergic mechanism: cerebrospinal fluid (CSF) DA, dihydroxy-phenylacetic acid (DOPAC), homovanillic acid (HVA), noradrenaline (NA), plasma NA, serum prolac tin ~RL), serum dopamine-O-hydroxylase (DBH) and platelet monoamine oxidase (MAO) activity. Determination of CSF DA in clinical studies has not been described in the literature (4). This may be due to technical reasons, but in addition there is not sufficient evidence that the DA level is a suitable index of dopaminer * Reprint request t0:-M. Arat6, M.D., P h . D . , Hamilton Psychiatric P.O. Box 585, Hamilton, Ontario. Canada. L8N 3K7 0024-3205/83/232667-10503.00/0 Copyright (c) 1983 Pergamon Press Ltd.
Hospital,
2668
Dopaminergic Indices in Man
Vol. 32, No. 23, 1983
gic functions. Clinical studies of the level of DOPAC in the CSF have been made only sporadically, probably also for technical reasons. Further aspects have also to be considered: a. DOPAC is an intermediate DA metabolite and therefore could be a more reliable index of DA metabolism than either DA itself or the final metabolic product of DA HVA. b. However, DOPAC is only a minor metabolite in man~ c., and constitutes only one pathway of DA catabolism. The measurement of HVA is the most frequently used and accepted approach . It should, however, be borne in mind that its actual level is not only determined by its formation, but also by its elimination. This problem can be overcome to a certain extent by the use of the probenecid technique (5). DBH is the enzyme catalyzing the conversion of DA to NA. Experimental studies have demonstrated a correlation between DBH activity and noradrenergic functions. However, recent human studies have failed to support this finding, and accordingly serum DBH activity cannot be considered as an index of peripheral sympathetic activity (6, 7). Our previous study suggested a relationship between serum DBH activity and central dopaminergic functions (8). We have measured NA in the CSF and plasma because NA is the step succeeding DA in catecholamine metabolism, and it is postulated that the above mentioned "dopaminergic" diseases are accompanied by disturbance of NA metabolism (9, i0), and noradrenergic system modulates dopaminergic functions (Ii). MAO the enzyme responsible for the oxidative deamination of monoamines is an important factor in neurotransmitter regulation. Platelet MAO activity is associated with the activity of the type B MAO in the brain (12). The relationship of platelet MAO activity to serum PRL suggests a correlation between platelet ~ 0 activity and central dopaminergic functions (13). Tonic dopaminergic inhibition plays a fundamental role in the control of PRL secretion (14). Thus, determination of serum PRL levels may provide indirect information about central dopaminergic activity. Different parameters have been selected throughout the literature and therefore the results are difficult to compare. Levels of the mentioned parameters have never been determined in the same population, The aim of the present study was to seek evidence of correlations of the postulated dopaminergic indices, to find out how many are interrelated and to what degree, and to clarify whether the individual indices could be substituted for each other. Material and methods Subjects Inpatients of neurologic departments in whom lumbar puncture was performed for diagnostic purposes were studied with the patients' consent. This mixed group consisted of patients with illnesses thought not to affect catecholamine metabolism (e.g. cervical spondylosis, herniated disk, peripheral neuralgia, etc.), and patients who were admitted for neurological assessment but in whom no abnormality was detectable. Only CSF samples with normal laboratory values (cell counts, protein, glucose, colloid lability tests) were evaluated. The patients were drug-free, apart from 9 females and 1 male who received small doses of benzodiazepines, mainly i0 mg nitrazepam, as night sedation. Thirty patients (24 females and 6 males) complied with these requirements and their ages ranged between 24 and 62 years. All samples were obtained within 3 days, between 8 A~ and 9 AM after a i0 hour fast and bed rest. The CSF was collected in plastic tubes and immediately frozen in dry ice. Two ml aliquots were used: - the first for routine laboratory tests,
Vol. 32, No. 23, 1983
Dopaminergic Indices in Man
2669
the second for DA, DOPAC and NA determination and third for the measurement of HVA. Blood amples for the assessment of plasma NA, serum PRL, DBH and platelet MAO were taken at the same time as the CSF samples. Methods CSF DA, DOPAC, NA and plasma NA levels were determined by a radioenzymatic method described by Fekete et al. (15). Twenty-five m l of plasma containing EGTA and glutathione and 50 ul of CSF were incubated in the presence of catechol-O-methyl-transferase (COMT) and 3H-S-adenosylmethionine. The Omethylated products were extracted with organic solvents in the presence of sodium tetraphenylborate (16) and chromatographed. The radioactivity of the authentic spots was measured and the concentration of COMT substrates (DA, DOPAC, NA) in the CSF calculated. The O-methylated derivates of NA were oxidized as described by Peuler and Johnson (17). The sensitivity of the method varied between 1-50 pg/reaction, corresponding to concentrations of 0.02-0.08 ng/ml CSF. Homovanillic acid was measured by the methods of Westerink and Korf (18) with some modification (19). Serum DBH activity was determined according to the method of Nagatsu and Udenfriend (20). Platelet MAO activity was assayed as follows: Aliquots of 8.5 ml venous blood were collected in plastic tubes containing 1.5 ml ACD solution. After standing for 15 min at room temperature the samples were centrifuged at i00 g for 15 min. The supernatant was drawn off and cooled. After recentrifugation at 200 g for 15 min the supernatants were pooled. Platelet rich plasma was centrifuged at 1500 g for 20 min, the supernatant aspired, and the platelets were washed with 2 ml saline. This was followed by centrifugation at 1500 g for 20 min. The supernatant was discarded, the platelets were taken up in 0.05 M, pH 7.4 phosphate buffer solution, and homogenized. Measurements were made using 14C-tyramine as substrate in a concentration of 1.3x10 -4 M/I. InGubation was carried out in the presence of EDTA at pH 7.4 and 37 ° C for 20 minutes. Protein concentration measured by the method of Lowry (21) was 5001 0 0 0 ~ g / m l and showed linear correlation with enzyme activity. The formed aldehydes were extracted with ethylacetate. Measurements were performed in a Packard liquid scintillation spectrometer after the addition of a dioxane cocktail. Intraassay scattering was 3.9%. Prolactin was determined by radioimmunoassay using a kit - IRE. For the amines, their metabolites and PRL values are given in ng/ml. DBH activity was expressed as nmol octopamine /min/ml serum. MAO activity is given as nmol/mg protein/hour. All neurochemical measurements were performed in one assay, the amines and their metabolites were measured within one we~k from sampling. Results Table I presents the ranges, means~SEM of the postulated dopaminergic indices measured and a correlation matrix is presented in Table II. The mean values of the measured parameters in patients taking benzodiazepines (n=lO) did not differ from those of the others. Plasma NA levels correlated positively with age, but the relationship was significant only in females (r=0.59, p (O.O1). A significant positive correlation was found between CSF DA and its metabolites (DOPAC and HVA) and between the two metabolites (Fig. i).
2670
Dopaminergic Indices in Man
Vol. 32, No. 23, 1983
TABLE I Mean Values and Ranges of Dopaminergic Indices CSF DOPAC Females n=
23
CSF DA
CSF NA
CSF HVA
plasma NA
serum P RL
21
24
21
24
24
platelet MAO
serum DB H
24
24
mean ±SEM
1.91 0.14
2.81 0.23
0.16 0.01
19.9 1.3
0.30 0.03
6.5 1.3
12.9 1.3
27.0 3.4
range
0.33.98
1.36.3
0.060.32
ii31
0.080.73
0.528
3.529.6
1.576.2
6
6
6
6
6
6
6 21.6 5.2
Males 6
n= mean
+_SEM range
3.12 0.68
2.92 0.39
0.16 0.03
22.9 4.1
0.23 0.02
7.3 3.3
8.7 1.2
1.65.5
1.54.2
0.030.25
12.534.5
0.150.30
2.323.5
3.711.8
8.137
Concentrations of amines (DA,NA), metabolites (DOPAC,HVA) and PRL given as ng/ml. Activities of MAO and DBH given as nmol/mg protein/hour and nmol octopamine /min/ml serum, respectively.
CSF
DOPAC
nglml 67
females •
males
n=20
n=6
r = 0,75
r = 0,78
o O
p< 0,001
0
• •
•
I
00
0
•
CSF
HVA
FIG. I. Significant positive correlation between CSF HVA and DOPAC.
Vol. 32, No. 23, 1983
Dopaminergic Indices in Man
2671
TABLE II Correlation matrix of dopaminergic indices Females
n=24
CSF DA
O. 7 2 x x x
CSF NA
0.62 x x x
0.35
CSF HVA
0.75 xxx
0.58 xx
plasma NA
0.09
0.18
-0.01
0.42
serum PRL
-0.20
-0.i0
-0.08
0.07
0.39
platelet MAO
-0.18
0.14
-0.21
-0.33
-0.16
0.05
-0.45 x
-0.39
-0.38
-0.07
-0.03
-0.23
CSF NA
CSF HVA
plasma NA
serum PRL
platelet MAO
serum DBH
-0.47 x
CSF DOPAC Males CSF DA
-CSF DA
0.38
n=6 0.89 xx
CSF NA
-0.40
CSF HVA
0.78
-0.52
0.87 x
-0.53
plasma NA
-0.20
-0.ii
0.73
0.05
serum pRL
0.04
0.03
0.56
-0.28
0.23
platelet MAO
-0.i8
-0.35
0.ii
-0.72
0.44
serum DBH
-0.56
-0.74 x: p 40.05;
0.29
0.07 -0.91 xx -0.25 0.35 xx: p ~0.02; xxx: p < 0 . 0 1
0.86 x
Serum DBH activity in females showed a significant inverse correlation with CSF DA (Fig. 2) and DOPAC and inverse tendency with CSF HVA. Similar tendencies were found also in males. Due to the small number of males their data are given only as complementary information. Significant positive correlation was found between CSF NA and CSF DOPAC and positive tendency between CSF NA and CSF HVA levels, but only in females.
2672
Dopaminergic
CSF
Indices
in M a n
Vol.
32, No.
23,
DA
ng/ml 7,
females
•
mele8
0
n=21
n=6
r = -0,45
r : ~0,74
p-O,O5
5-
O
41
O
•
ee
•
•
0
o
• o
•
0
serum DBH
o
~'o
2"o
3'o
4i~
s3
FIG. Significant plem
inverse
correlation
6'0
7'o
8'0 (u)
2.
between
serum DBH activity
and CSF DA
NA
0.7-
0.6~
e•
0,5
N.S.
0,4
o,3
o •
o
o
•
0
~2 0 oI
0,1
serum DBH
0
15
25
35
4o
~ FIG.
No significant correlation between NA level, n=30, 24 females - I,
i0
7o
~
(u)
3. serum DBH activity 6 males - o.
and plasma
1983
Vol. 32, No. 23, 1983
Dopaminergic
Indices in Man
2673
No relationship was found between serum DBH activity and the NA level of either CSF or plasma (Fig. 3). Discussion The means and ranges of the CSF DA metabolites reported here correspond to published data (22,23). The values of CSF and plasma NA are also similar to those described in the literature (24,25,26,27,28). It is a limitation of these data that the used assay procedure measures only the free and not the conjugated catecholamines and metabolites, and it has recently been shown that catecholamines in plasma are for the major part present as conjugated derivatives (29), and conjugates are present also in human CSF (30). However, in spite of individual variations, statistically significant correlations were found between the concentrations of free and conjugated DOPAC and HVA in CSF (30). The high level of CSF DA compared to CSF NA was surprising. A hypothetic explanation to it may be that the main source of CSF DA is the DA-rich nucleus caudatus lying closest to CSF channels, and the DA/NA ratio is very high in this region (31). The mean values of serum DBH activity in the present study correspond to published data (20,32,33). The platelet MAO activity values were in accordance with the results of our earlier study: normal female controls, n=34: 12.6J6.1 (SD), males, n=70: i0.5~4. Serum PRL levels were in the normal range, with the exception of 1 female and 1 male patient whose PRL values were above the normal limit (27.7 and 23.5 ng/ml respectively). Since the ages of the subjects ranged from 24 to 62 years, possible age correlations were calculated. Significant age dependence was found only in plasma NA levels of females. This data is in harmony with the finding of Ziegler et al. (34). CSF HVA showed rising and CSF DOPAC showed decreasing tendency with age. CSF DA and its two metabolites (DOPAC and HVA) were significantly positively correlated. This finding suggests that not only baseline HVA and DOPAC,but also CSF DA level may be a suitable index of central DA activity. It remains to be seen if these relationships held under all circumstances for example drug treatment. However, owing to the great individual variation of the data it seems to be more informative to have several measures of dopaminergic activity. The significant negative correlation between serum DBH activity and CSF DA and its metabolites has not been previously reported. The correlations were significant only in females, but similar tendencies were also found in males. This finding supports our earlier hypothesis that serum DBH activity may be related to central dopaminergic mechanisms (8). Serum DBH activity is intraindividually stable, and is mainly genetically determined and DA turnover shows dynamic changes. Therefore the observed correlation implies a role of DBH activity in central dopaminergic regulation. Another possible explanation is that like serum DBH activity, basal DA turnover is under genetic control (12), and both may be controlled by a common genetic factor. The significant positive correlation between CSF NA and DOPAC may support this hypothesis. Antelman and Caggiula (ii) demonstrated that NA modulates dopaminergic functions. It is conceivable that DBH regulates dopaminergic functions by this mechanism. Inhibition of DBH with disulflram, both in experimental animals (ii) and clinical studies (35), leads to dopaminergic overactivity and related behavioural changes and symptoms. This evidence of
2674
Dopaminergic Indices in Man
Vol. 32, No. 23, 1983
an inverse relationship between DBH and DA activity taken together with evidence of a negative correlation between DBH and indices of DA activity found in the present study may be relevant in studies into the pathogenesis of schizophrenia where both underactivity of NA and overactivity of DA have been implicated. For example Wise and Stein (36) demonstrated decreased DBH activity in the brains of schizophrenic patients in post mortem studies. Fujita et al. (37) and Arato et al. (33) have found decreased DBH activity in large populations of paranoid schizophrenic patients. Sternberg et al. (38) obtained good therapeutic response to neuroleptic treatment in schizophrenic patients with low CSF DBH activity, while the prognosis in the presence of high DBH activity was poor. Perenyi et al. (39) found serum DBH activity also to be of predictive value: the responders to neuroleptic therapy showing no extrapyramidal symptoms in their schizophrenic population had significantly lower serum DBH activity than the non-responders developing extrapyramidal symptoms. According to Lerner et al. (40) and our unpublished data, basal serum DBH and CSF DBH activities show significant positive correlation. In contrast to some earlier presumptions, numerous recent studies have fai~ ed to establish any relationships between DBH activity and plasma NA levels (41,42,43). This lack of relationship was confirmed in the present study as we have found no correlation between serum DBH activity and NA of plasma and CSF, even if making corrections of plasma NA values for age. These findings confirm that serum DBH activity is not an indicator of peripheral noradrenergic activity. Change in serum PRL level is a good index of the effects of DA antagonist and agonist drugs (44). However, no correlation can be expected to be between baseline serum PRL level and the CSF indicators of DA turnover (45) as PRL is controlled by the tuberoinfundibular dopaminergic system, while CSF DA and its metabolites derive mainly from the nigrostriatal system (4). Our data did not support the finding of Oreland et al. (12) that there is a positive correlation between platelet MAO activity and CSF concentration of HVA. According to our data platelet MAO activity rather correlates negatively to CSF HVA. This contradictory result requires further confirmation. Acknowledgement We thank Dr. I.N. Ferrier for his helpful comments on our text.
References I. 2. 3. 4. 5. 6. 7.
8. 9.
W.BIRKMAYER and O. HORNYKIEWICZ, Wien. klin. Wschr. 73 787-788 (1961). A. BARBEAU, Psych. Forum 4 8-15 (1973). A. RANDRUP and I. MUNKVAD, Psychopharmacol. ii 300-310 (1973). D.P. VAN KAMMER and D.E. STERNBERG, Neurobiology of .Cerebrospinal Fluid, ed. by J.H. Wood, 7.9-742, Plenum Publ. Corp., New York (1980). H.M. VAN PRAAG, Biol. Psychiat. 12 101-131 (1977). R.M. WEINSHILBOUM, Pharmacol. Re~. 30 133-166 (1979). J.M. STOLK, J.H.HURST, M.J. FRIEDMAN, P.O. HARRIS, D.A. VAN RIPPER and B.C. NISULA, E__nzymes and Neurotransmitters in Mental Disease, ed. by E. Usdin, T.L. Sourkes and M.B.H. Youdim, 171-192, John Wiley and Sons Ltd., Chichester (1980). G. BAGDY AND M. ARATO, Intergrative Neurohumoral Mechanisms~ ed by E. Endroczl, Akad~mia Kiad6, Budapest (In press). I.J. FARLEY, K.S. PRICE, E. MCCULLOGH, J.H.N. DECK, W. HODYNSKI and O. HORNYKIEWICZ, Science 200 456-458 (1978).
Vol. 32, No. 23, 1983
i0. ii. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28.
29. 30. 31. 32. 33. 34. 35. 36. 37. 38.
39. 40. 41.
42. 43.
Dopaminergic Indices in Man
2675
C.R. LAKE, D.E. STERNBERG, J.C. BALLENGER, M.G. ZIEGLER, R.M. POST, l.J. KOPIN and W.E. BUNNEY, Science 207 331-333 (1980). S.M. ANTELMAN and A.R. CAGGIULA, Science 195 646-653.(1977). L. ORELAND, A. WIBERG, M. ASBERG, L. TRASKMAN, L. SJOSTRAND, P. THOREN, L. BERTILSSON and G. TYBRING, Psychiat. Res. 4 21-29 (1981). J.E. KLEINMAN, S. POTKIN, A. ROGOL, M.S. BUCHSBAUM, D.L. MURPHY, J.C. GILLIN, H.A. NASRALLAH and R.J. WYATT, Science 206 479-481 (1979). R.M. MACLEOD, Frontiers in Neuroendocrinology, ed by L. Martini and W. Ganong, 169-194, Raven Press, New York (1976). M.I.K. FEKETE, B. KANYICSKA and J.P. HERMAN, Life Sci. 23 1549-1556 (1978). M. DA PRADA and G. ZURCHER, Life Sci. 19 1164-1174 (1976). J.D. PEULER and G.A. JOHNSON, Life Sci. 21 625-636 (1977). B.H.C. SWESTERINK and J. KORF, Biochem. Med. 12 106-115 (1975). C.M. BANKI, J. Neurochem. 28 235-237 (1977). T. NAGATSU and S. UDENFRIEND, Clin. Chem. 18 980-983 (1972). O.H. LOWRY, N.J. ROSEBROUGH, A.L. FARR and R.J. RANDALL, J. Biol. Chem. 193 265-275 (1951). G. SEDVALL, B. FYRO, B. GULLBERG, H. NYBACK, F.A. WIESEL and B. WODEHELOODT, Brit. J. Psychiat, 136 366-374 (1980). S. WILK and M. STANLEY, Psychopharmacol. 57 77-81 (1978). R. LAKE, M.G. ZIEGLER and D.L. MIIRPHY, Arch. Gen. Psychiat. 34 553-556 (1977). G. PLANZ and R. PLANZ, Experientia 35 207 (1979). U.C.R. GOMES, B.C. SHANLEY, L. POTGIETER and J.T. ROUX, Brit. J. Psychiat. 137 346-351 (1980). R.M. POST, C.R. LAKE, D.C. JIMERSON, W.E. BUNNEY, J.H. WOOD, M.G. ZIEGLER and F.K. GOODWIN, Am. J. Psychiatry 135 907-912 (1978). D.P. VAN KAMMEN, D.E, STERNBERG, C.R. LAKE and P. LERNER, Frontiers of H0rmone Researqh, ed. by T.B. Van Wimersma Greidanus, 198-212, Karger, Basel (1982). M. DA PRADA, Trends Pharmacol. Sci. 1 157-159 (1980). P.A. BERGER, K.F. FAULL, J. KILKOWSK~, P.J. ANDERSON, H. KRAEMER, K.L. DAVIS and J.D. BARCHAS, Am. J. Psychiatry 137 174-180 (1980). P. NYBER, R. ADOLFSSON, N.-E. ANDEN and B. WINBLAD, Acta Neurol. Scandinav. 65 267-273 (1982). R.A. STONE, J.C. GUNNELS, R.R. ROBINSON, S.M. SHANBERG and N. KIRSCHNER, Circ. Res. 34 47-56 (1974). r M. ARATO, G. BAGDY and A. PERENYI, Biol. Psychiat. 17 853-854 (1982). M.G. ZIEGLER, C.R. LAKE and I.J. KOPIN, Nature 261 333-335 (1976). E. HARTMANN and M. KELLER-TESCHKE, Biol. Psychiat. 14 455-462 (1979). C.D. WISE and L. STEIN, Science 181 344-347 (1973). K. FUJITA, T. ITO, K. MARUTA, H. BEPPU, Y. NAKAGAMI and Y. KATO, J. Neur~ chem. 30 1569-1572 (1978). D.E. STERNBERG, D.P. VAN KAMMEN, J.C. BALLENGER, P. LERNER, S.R. MARDER and R.M. POST, Annual Meeting Amer. Psychiat. Assoc. New Res. Abstr. 32 (1978). A. PERENYI, G. BAGDY and M. ARATO, 10th Danube Symp. Psychiat. Magdeburg, Abstr. 112 (1982). P. LERNER, F.K. GOODWIN, D.P. VAN KAMMEN, R.M. POST, L.F. MAJOR, J.C. BALLENGER and W. LOVENBERG, Biol. Psychiat. 13 685-694 (1978). M. TANAKA, T. SANO, Y. KOHNO, T. YOKOTA, Y. MIZUKI and N. NAGASAKI, Prophylactic Approac h to Hypertensive Diseases, ed. by Y. Yamori, 321328, Raven Press, New York, (1979). B.A. BERKOWITZ, R. HEAD, T. JOH and J. HEMSTEAD, J. Pharmacol. Exptl. Ther. 213 18-23 (1980). S. SAARIKOSI, K. SELANDER, E. NUKALA and P. PYSTYNEN, Prostaglandins 20 487-492 (1980).
2676
44. 45.
Dopaminergic
Indices in Man
Vol. 32, No. 23, 1983
G. LANGER, E.J. SACHAR, R.S. NATHAN, M.A. TABRIZI, J.M. PEREL and F.S. HALPERN, Psychopharmacol. 65 161-164 (1979). K.J. ZANDER, B. FISCHER, R. ZIMMER and M. ACKENHEIL, Psychopharmacol. 73 43-47 (1981).