- Email: [email protected]
Simultaneous monitoring of dopamine release in rat frontal cortex, nucleus accumbens and striatum: Effect of drugs, circadian changes and correlations with motor activity
Neuroscience Vol. 16, No. I, pp. 49-55, 1985 Printed in Great Britain
0306-4522/85 163.00+ 0.00 Pergamon Press Lti 0 1985 IBRC
SIMULTANEOUS MONITORING OF DOPAMINE RELEASE IN RAT FRONTAL CORTEX, NUCLEUS ACCUMBENS AND STRIATUM: EFFECT OF DRUGS, CIRCADIAN CHANGES AND CORRELATIONS WITH MOTOR ACTIVITY R. D. O’NEILL*t and M. FILLENZ University Laboratory of Physiology, Parks Road, Oxford OX1 3PT, U.K. Abstract-Changes in homovanillic acid concentration, recorded in viuo by voltammetry with carbonpaste electrodes, were used as an index of dopamine release. With electrodes implanted in frontal cortex, nucleus accumbens and striatum, the pattern of dopamine release was monitored simultaneously in the three brain regions together with the rats’ motor activity. Changes in response to the systemic administration of dopamine-receptor agonists and antagonists were used as an index of feedback control of dopamine release. The relationships between dopamine release and motor activity, as well as that between dopamine release in the different brain regions, were investigated by calculating correlation coefficients for data collected over 24-h periods. The results show that dopamine release in frontal cortex is subject to little feedback regulation, that there is no nocturnal increase and no correlation with motor activity. Dopamine release in accumbens and striatum follows a very different pattern. There was a high correlation between dopamine release in these two regions on both sides of the brain; the correlation between the accumbens and the ipsilateral striatum was the highest. Dopamine release in both regions shows evidence of considerable feedback regulation, a nocturnal increase and a high correlation with
motor activity. The importance of the accumbens in relation to the level of motor activity is supported by the finding that the correlation coefficientbetween motor activity and dopamine release in this structure was significantly higher than that between activity and release in the striatum.
Two groups of dopaminergic cell bodies in the mammalian midbrain, A9 and AlO, give rise to the nigrostriatal, mesolimbic and mesocortical dopaminergic pathways which project to the striatum, nucleus accumbens and frontal cortex, respectively. Studies of each of these systems, using a variety of experimental techniques, suggest that they are associated with different functions and that the release of dopamine in the three target areas is subject to different regulatory mechanisms.’ So far, however, there have been few direct comparisons of the three systems in the same animal. With the development of microcomputer-based linear sweep voltammetry with carbon-paste electrodes,“.14 it is now possible to monitor continuously changes in dopamine release in discrete brain regions of the unrestrained rat over long periods by detecting changes in the extracellular concentration of the dopamine metabolite, homovanillic acid (HVA).‘2.‘6 The studies undertaken to identify the HVA signal in striatum have been reported fully.” We now describe an HVA signal in frontal cortex and accumbens and use the dopamine-synthesis inhibitor, a-methyl-ptyrosine, to confirm its identity; the effects of dopamine agonists and antagonists on the release
were also investigated. It has been suggested that there is asymmetry and mutual inhibition between the two nigrostriatal dopaminergic projections.9 To investigate this and explore other possible relationships between release of dopamine in these forebrain regions, correlation coefficients were calculated for the circadian changes in the HVA signal recorded in frontal cortex, accumbens and striatum. Finally we looked at the relationship between circadian changes in motor activity and dopamine release in each brain region. EXPERIMENTAL PROCEDURES Carbon-paste electrodes were prepared and implanted as previously described. ” Briefly, carbon paste (2.8 g carbon powder in I ml silicone oil) was packed into 300~pm (o.d.) Teflon-coated silver wire. Male Sprague-Dawley rats (initial weight 270-400 g) were stereotaxically implanted, under chloral hydrate anaesthesia, with these electrodes; silver wires, placed in the cortex, were used for the referen& and auxiliary electrodes. Rats were typically implanted with four carbon-paste working electrodes either in frontal cortex, nucleus accumbens, striatum and hippocampus or bilaterally in accumbens and bilaterally in striatum. (The results from the hippocampal electrodes are not relevent to the present study and no further mention is made of the hippocampus.) The co-ordinates, with the head level between bregma and lambda, were as follows: frontal cortexAP 3.0 (from bregma), L 1.3 (from bregma) and DV I.5 (from skull); accumbens-AP 1.5, L 1.5, DV6.5; and striatum-AP -0.5, L 3.0 and DV 4.8. The animals were given at least 2 days to recover before being placed in the recording cages and connected to the microcomputercontrolled equipment.14
*To whom all correspondence should be addressed. tBeit Memorial Research Fellow. Abbreviations : HVA, 3PPP, homovanillic acid; 3-(3-hydroxyphenyl)-N,n-propylpiperidine. 49
50
R. D. O’NEILLand M. A. Voltammogram
from
frontal
FILLENZ
cortex
0 B(i) 4
E(M)
E(il)
cortex: Id3=0.13+0.06
wcteur
I rontal
-
nA (n=@)
3
r 13~0.21
f0.02
nA (n=14)
clcoumbenw
4
nA
3 nA
nA
2
2
2
3
1
0
1 430
0
’
01
430 SOOmU 570 500 mU 570 Fig. I. (A) Example of a voltammogram, recorded in the frontal cortex of an unrestrained rat, using microcomputer-controlled linear sweep voltammetry with a carbon-paste electrode. A linearly increasing potential (range O-650 mV with respect to an Ag reference electrode) is applied at a sweep rate of 5 mV/s to the implanted carbon-paste electrode, and the oxidation current produced recorded as a function of potential. When the background current, which is measured in situ by scanning continually, is substracted from voltammograms recorded at 1Zmin intervals, a difference volt~mogmm consisting of three separate peaks is obtained. Peak 1 is due to the oxidation of ascorbic acid, peak 2 to uric acid and the dopamine metabolite HVA is responsible for peak 3. Changes in the height of each peak are proportional to ehanges in the extracellular concentration of the corresponding compound around the electrode tip. (B) Peak 3 recorded in three brain regions of the same rat. (i) Frontal cortex: average peak potential ( V3)= 490 &-10 mV; average peak height (h3) = 0.13 f 0.05 nA (n = 9). (ii) Nucleus accumbens: V, = 490 + 5 mV; h, = 0.42 -(_0.04 nA (n = 14). (iii) Striatum: V, = 480 f 6 mV; h, = 0.21 + 0.02 nA (n = 14). 500 mU
570
430
Linear sweep voltammograms were recorded at 12-min intervals at a rate of 5 mV/s between 0 and 650 mV. The background current for each electrode was measured in situ before each experiment as described previously.‘3 Eight electrodes (i.e. two rats) were scanned simultaneously and the voltammograms stored on magnetic disc. Voltammograms recorded from the three brain regions usually consisted of three separate peaks; in some rats peak 3 was not observed in frontal cortex. The height of peak 3 (h3) was measured as indicated in Fig. 1.Between the electrochemical recordings (for approx. 9min duration) the total motor activity of each rat was monitored using a Doppler-shift microwave device linked to the PI0 of the interface.”
Drugs were administered i.p. using either 0.9% saline or distilled water (for apomorphine only) as the carrier. The significance of the effects of drugs on h, in each brain region was calculated using paired r-tests by comparing the average pre- and post-injection values for each electrode. The significance of the difference in effects between brain regions was calculated by comparing the percentage changes in the same rat. The mean and SEM are quoted; n is the number of electrodes. A group of 8 rats were given no drugs and were used to record simultaneously the circadian variation in motor activity and h, in frontal cortex, accumbens and striatum. To reduce the fluctuations inherent in recording these small
Monitoring dopamine release in rat forebrain using voltammetry
51
Table 1. Percentage changes in the homovanillic acid signal recorded in three brain regions Brain region Frontal cortex
Haloperidol
Nocturnal increase
AMPT
Apomorphine
-94+3(4) (P < 0.001)
-24*6(4) (P < 0.05)
50 k 17 (6) (P < 0.05)
20* 19(9) NS
Accumbens
-96*2(S) (P < 0.001)
-63 k 7(6) (P < 0.001)
306 k 55 (6) (P < 0.005)
48k7(14) (P < 0.001)
Striatum
-98 f I (5) (P < 0.001)
-74*8(7) (P < 0.001)
365 rt. 81 (9) (P < 0.005)
52 f 5 (14) (P < 0.001)
The effects of a-methyl-p-tyrosine (AMPT, 250 mg/kg i.p.). apomorphine (2 mg/kg i.p.) and haloperidol(0.5 mg/kg i.p.) on the HVA signal recorded in three brain regions. The results are expressed as a percentage change compared with the average pre-injection value. The nocturnal increase is the percentage increase compared with the average day-time value. Numbers in brackets are the number of electrodes. Significance calculated using r-tests on data from each electrode. NS = not significant. currents (< I nA) and clarify the systematic relationships between the variables, a weighted moving average of the original data was calculated.” Correlation coefficients between h, in the different regions and between h, and motor activity were calculated for each 24-h period (120 points). Cross-correlations were also calculated by comparing, e.g. m(t + 1) with h,(r), where h,(t) is the height of peak 3 at time t (l2-min intervals) and m(r) is the value of the movement counter following that scan. Cross-correlations were calculated in the range m(r - 5) to m(r + IO). Data were recorded from each rat for a period of 2 or 3 days; the n value for these results are the number of determinations. RESULTS Voltammograms
recorded
in frontal
cortex,
nu-
cleus accumbens and striatum consisted of three separate peaks (Fig. 1). The average potential of peak 3 was: frontal cortex = 490 f 10 mV, accumbens = 490 + 5 mV and striatum = 480 + 6 mV. The average height of peak 3 (h,) recorded in the three brain regions was: frontal cortex = 0.13 f 0.05 nA (n = 9), accumbens = 0.42 f 0.4 nA (n = 14) and striatum = 0.21 & 0.02 (n = 14) (n = number of electrodes). There was no significant difference in h, recorded in frontal cortex and striatum; hJ in accumbens, however, was significantly greater than that in the other two regions (P < 0.01, see Fig. 1). The identity of peak 3 recorded in striatum with homovanillic acid (HVA, a dopamine metabolite) is established; we used the dopamine-synthesis inhibitor a-methyl-p-tyrosine (Sigma) to determine the extent of the contribution of HVA to h, recorded in frontal cortex and accumbens. a-Methyl-p-tyrosine (250mg/ kg i.p.) caused a decrease in /I~ in the three regions: in frontal cortex by 94 + 3% (n = 4, P < O.OOl), in accumbens by 96 + 2% (n = 5, P < 0.001) and in striatum by 98 + 1% (n = 5, P < 0.001). To compare the contribution made by dopamine in regulating its own release, the effects on hJ in the three regions of the dopamine-receptor agonist, apomorphine (Sigma), and antagonist, haloperidol (Searle), were investigated. Apomorphine (2 mg/kg i.p.) caused a decrease in h, in the three regions: in frontal cortex by 24 f 6% (n = 4, P < 0.05), in accum-
bens by 63 + 7% (n = 6, P < 0.001) and in striatum by 74 + 8% (n = 7, P < 0.001). There was no significant difference in the percentage change recorded in accumbens and striatum; the difference between the effect in frontal cortex and that in accumbens and striatum was significant with P c 0.05 and 0.01, respectively. Haloperidol(O.5 mg/kg i.p.) increased h, in the three brain regions: in frontal cortex by 50 f 17% (n = 6, P < 0.05) in accumbens by 306 k 55% (n = 6, P < 0.005) and in striatum by 365 + 81% (n = 9, P -e0.005). As with apomorphine, there was no significant difference between the effects in accumbens and striatum; the difference between the effect in frontal cortex compared with that in accumbens and striatum was significant with P c 0.01and 0.05, respectively. The effects of these drugs are summarized in Table 1. To investigate the effect of 3-(3-hydroxyphenyl)N,n-propylpiperidine (3PPP) on dopamine receptors in accumbens and striatum, the two enantiomers of this drug were injected into rats implanted with electrodes bilaterally in accumbens and striatum. The results are shown in Table 2; (+)3PPP and (-)3PPP (1 and 25 mg/kg i.p.) produced similar effects in both brain regions, and although the average effect in striatum was greater than that in accumbens for all doses, this difference was significant only for (-)3PPP at 1 mg/kg. Circadian changes in extracellular HVA in the three brain regions were measured in 8 rats which had not been given drugs. The minimum values of hJ (recorded during daytime, typically between 10.00 and 12.00 h) and the maximum values (recorded during the night-time, active period, between 04.00 and 06.00 h) were as follows: frontal cortex, 0.13 + 0.05/O. 16 f 0.07, increase = 20 + 19% (n = 9, not significant); accumbens, 0.42 f 0.04/0.62 + 0.06, increase = 48 + 7% (n = 14, P < 0.001); and striatum, 0.21 + 0.02/0.32 &-0.04, increase = 52 &-5% (n = 14, P < 0.001). There was no significant difference between the percentage increase observed in accumbens and striatum (Table 1).
R. D. @NEILL and M. FILLENZ
52
Table 2. Effect of 3-(3-hydroxyphenyl)-N,n-propylpiperidine
on the homovaniilic acid signal
(+)Isomer Accumbens
1w/kg
Accumben: - )lsomer Striatum
Striatum
-22 f So/, tl=8 -46 + 10% n=6 - 8 i 2% n=4 - 38 i 7% n=3 P < 0,05 P < 0.01 P < o,05 P < o.05
25 mg/kg
n=S -58i80i,p
-75,120/,Pn<&
67+ 13&,n<;;2 132&35?,,n<=o;5
The effects of the two enantiomers of 3-(3-hydroxyphenyl)-N,n-propylpiperidine (3PPP) on the HVA signal recorded in nucleus accumbens and striatum. Results expressed as a percentage of the pre-injection value; n is the number of electrodes.
The intercorrelations for circadian changes in the HVA signal in accumbens and striatum are shown in Table 3. There was no significant difference between values calculated for accumbens vs contralateral accumbens, striatum vs contralateral striatum or accumbens vs contralateral striatum. However, the correlation coefficient calculated for accumbens vs ipsilateral striatum was significantly greater than all contralateral values (P < 0.01). For frontal cortex, only correlations with contralateral accumbens and ipsilateral striatum are available: 0.46 &-0.10 (n = 13) and 0.42 f 0.12 (n = 13), respectively. In all cases, the maximum values of the intercorrelation were obtained when the time shift was zero. We next looked at the correlation between motor activity and the HVA signal in the three regions (Table 4). There was no significant correlation between motor activity and HVA in frontal cortex. The correlation for HVA in accum~ns vs motor activity (0.68 f 0.03, n = 25) was significantly greater than that for striatal HVA (0.56 & 0.04, n = 24) with P < 0.01. The maximum correlation was obtained when the motor activity data were shifted by
24 f 7 min with respect to the HVA signal in accumbens and by 30 i 7min with respect to the striatal HVA signal; there was no significant difference between these two shifts. The same data were also analysed by averaging the 24-h recordings of motor activity and the HVA data from each brain region. Cross-correlations were calculated for these averaged time courses: in frontal cortex, r = 0.267, shift = -48 mitt; in accumbens, r = 0.926, shift = + 36 mitt; and in striatum, r = 0.832, shift = + 36 min (Fig. 2).
DISCUSSION Voltammetry in vivo was initially developed as a technique to monitor changes in the concentration of transmitter substances in the extracellular fluid. Howdrug-induced* and electrically ever, although stimulated’ changes in dopamine concentration have been reported, it has become clear that the concentration of transmitters, under normal conditions, is too low to measure in situ with the present electro-
Table 3. Correlation coefficients for changes in the homovanillic acid signals Correlation coefficients
Brain regions Accumbens Striatum vs Accumbens Accumbens
vs accumbens (contralateral) striatum (contralateral) vs striatum (contralateral) vs striatum (ipsilateral)
0.61 k 0.60 + 0.67 + 0.78 k
O.lOfn 0.09 (n 0.03 (n 0.05 (n
= = = =
6) 6) 24) IO)*
Correlation coefficients for the circadian variation in the HVA signal in nucleus ac~umbens and striatum calculated for each day’s data (120 points); n = number of determinations. Maximum correlations were obtained when the time shift was zero. *Significantly greater than contralateral values, P < 0.01.
Table 4. Correlation
Brain region Frontal cortex Nucleus accumbens Striatum
coefficients for motor activity vs homovanillic acid signals Correlation coefficient 0.1 kO.4 0.68 * 0.03* 0.56 f 0.04
Time shift (min) 24 * 7 30*7
n-value 13 25 24
Correlation coefficients for circadian changes in total motor activity vs the HVA signal in frontal cortex, nucleus accumbens and striatum. Values calculated for each day’s data (120 points); n = number of determinations. Time shift refers to the displacement of the motor-activity data which gave the maximum correlation; i.e. the voltammetric response lagged behind changes in motor activity. *Greater than that for striatum. P < 0.01.
53
Monitoring dopamine release in rat forebrain using voltammetry
changes
Circadian
in HVA and motor activity
0.4 A. nucleus
accumber
19
0.3
100
WA
~ h3lnA
,ounts
0.2 r SO.926
shift ~38
0.
200 min f
f
motor
activity
DARK
0.0
2
15
18
El
24
J
6
5
12t
0
Time of day
0.4 6. striatum
400
HVA
0.3
t
h3/nA
counts
0.2 rr0.832 shift -36
0.
200
mln
I motor
activity
DARK
0.0
2
15
18
21
,24
J
6
9
32t
0
Time of day Fig. 2. Averaged time course of the circadian changes in total motor activity and the HVA signal in the nucleus accumbens (A, n = 25) and striatum (B, n = 24). Left ordinate is the height of peak 3 (h,) in nA, right ordinate is the activity count recorded between electrochemical recordings; abscissa is the time of day in hours. 120 points at It&n intervals. Shift is the displacement of the movement data needed for m~mum correlation.
54
R. D.
O’NEILL and
chemical techniques. Many voltammetric methods can, however, detect transmitter metabolites such as the dopamine metabolite, 3,4-dihydroxyphenylacetic acid” and the serotonin metabolite, S-hydroxy indoleacetic acid.4.6 We have shown that peak 3 recorded using linear sweep voltammetry with carbon-paste electrodes in rat striatum is due to yet another metabolite of dopamine, homovanillic acid (HVA).“.” Voltammograms recorded using this technique consist of three separate peaks in areas with a significant dopaminergic input such as frontal cortex, accumbens and striatum (Fig. l), but only peak 1 (ascorbic acid) and peak 2 (uric acid) are seen in occipital cortex, hippocampus and cerebellum, where dopamine transmission is minimal; moreover, the potential of peak 3 in the three brain regions is the same (Fig. 1). These observations indicate that peak 3 in frontal cortex and accumbens is likewise due to HVA and the conclusion is supported by the greater than 95”/, reduction in the peak height caused by the dopa~ne-synthesis inhibitor, ~-methyl-p-tyrosine (Table I>. (Contributions from the methylated metabolites of noradrenaline in frontal cortex and accumbens can be neglected since these are sulphate conjugated, and therefore are non-electroactive.) Dopamine gives rise to four metabolic end-products: the free and conjugated forms of 3,4-dihydroxyphenylacetic acid and HVA. Westerink and Korf (1976)17 found that, following certain pharmacological manipulations, free 3,4-dihydroxyphenylacetic acid and HVA did not change in parallel and the authors suggested that 3,4-dihydroxyphenylacetic acid levels were a more accurate reflection of dopamine release. However, no such comparison in the absence of drugs has been made. Changes in the extracellular concentration of striatal HVA show a high correlation with motor activity and druginduced changes in the HVA signal, up to I h after injection, are similar to the known effects of the drugs on dopamine release. We have therefore concluded that, under physiological conditions, changes in HVA reflect changes in the release of dopamine;” it is reasonable to assume that the HVA signals recorded in frontal cortex and accumbens are an index of dopamine release in their respective areas. We have previously reported the effects of drugs on striatai HVA;” we have repeated some of these experiments in rats with electrodes in the three brain regions to investigate regional differences. The lack of correlation between the size of the HVA signal (peak 3) and dopamine content, or even dopamine turnover, in different brain regions is due to a number of factors. The electrode monitors the HVA concentration in its immediate vicinity and not in the brain region as a whole. The distribution of dopamine terminals in the frontal cortex, in contrast to that in the striatum, is non-uniform in that it is confined to certain layers. As a consequence, an HVA signal is always present in recordings from but is occasionally undetectable in striatum”
M.
FILLENZ
frontal cortex. This leads to a bias in the sampling since zero values are not included in the mean peak height. The finding that the dopamine-receptor agonist, apomorphine, and antagonist, haloperidol, produced large changes in dopamine release in accumbens and striatum but only very small changes in frontal cortex (Table I ), supports the view that there is much Iess feedback inhibition of dopamine release in frontal cortex.’ Since neither apomorphine nor haloperidoi are selective for pre- or postsynaptic dopamine receptors, they do not allow the distinction to be made between long-loop feedback mediated through postsynaptic receptors and local feedback through presynaptic autoreceptors. However, the failure of dopamine agonists to reverse the dopamine accumulation following gamma-butyrolactone administration,’ or to reduce neuronal firing’ provides evidence that a sub-population of mesocortical dopaminergic neurones lack both presynaptic and somatodendritic autoreceptors. Although the experiments with apomorphine and ha~operido~ showed no significant difference between the effects in accumbens and striatum, it is noteworthy that the average percentage change in accumbens was less than that in striatum in both cases; this would suggest that the autoregulation of dopamine release in accumbens is weaker than in striatum. This conclusion is supported by the results of the administration of the two isomers of 3PPP. At both doses of these two isomers, the average effect in accumbens was less than that in striatum and at 1 mg/kg (-)3PPP the difference was statistically significant (Table 2). These results do not support the suggestion that 3PPP has a preferential action in accumbens.” They show that the effects of the drug in accumbens, as in striatum,” are consistent with the suggestion” that at low doses (+)3PPP acts as a presynaptic dopamine-receptor agonist, and as a postsynaptic agonist at higher doses; (-)3PPP, however, appears to have presynaptic dopamine-receptor agonist properties at low doses, but behaves as a postsynaptic antagonist at high doses. The failure to observe a significant nocturnal increase in dopamine release in frontal cortex (Table I), together with the higher correlation between motor activity and dopamine release in accumbens and striatum compared with frontal cortex (Table 4), suggests a different role for dopamine release in the three brain regions in relation to motor activity. The significantly higher correlation for motor activity vs dopamine release in accumbens compared with striatum supports the view that dopamine release in accumbens is more important in controlling activity levels, whereas release in striatum may have more to do with patterns of movement.5 The correlations between dopamine release in accumbens and in striatum (Table 3) are interesting on two counts. First, the high correlation for each structure with its contralateral counterpart, which suggests that mutual inhibjtion9 does not play an
Monitoring dopamine release in rat forebrain using voltammetry important role under normal conditions. Second, the significantly greater correlation, compared with all contralateral values, for accumbens vs ipsilateral striatum. This result may reflect the amount of turning movements which take place in the small recording cages (25 x 25 x 25 cm3) leading to preferential release of dopamine in the two structures on the side contralateral to the direction of turning. Summary
This study illustrates some new approaches in the study of transmitter release in the unrestraint ani-
55
mal made possible by voltammetry in vivo. The results show that, as expected from pharmacological studies, changes in the release of dopamine in nucleus accumbens and striatum, but not in frontal cortex, are positively correlated with changes in the level of motor activity. Acknowledgements-The
work was supported in part by the Medical Research Council. We are grateful to the E. P. Abrahams Cephalosporin Fund for a grant to obtain additional equipment. We thank Astra Lakemedel, Sweden, for the gift of 3PPP, Dr. J. F. Stein for helpful discussions and Mr. M. R. BIoom~eId for art work.
REFERENCES 1. Bannon M. J. and Roth R. H. (1983) Pharmacology of mesocortical dopamine neurones. Pharmuc. Rev. 35, 53-68. 2. Bannon M. J., Michaud R. L. and Roth R. H. (1981) Mesocorticai dopamine neurons: lack of autoreceptors modulating dopamine synthesis. ~0~~~. Pharmac. 19, 270-275. 3. Bannon M: J., Chiodo L. A., Roth R. H. and Bunney B. S. (1982) Mesocortical dopamine neurons. I. Electrophysiological and biochemical evidence for the absence of autoreceptors in a subpopulation. Sot. Neurosci. Abstr. 8, 480. 4. Cespuglio R., Faradji H., Riou F., Buda M., Gonon F., Pujol J. F. and Jouvet M. (1981) Differential pulse voltammetry in brain tissue. II. Detection of J-hydroxyindoleacetic acid in the rat striatum. Bruin. Res. 223, 299-311. 5. Costall B., Naylor R. J. and Nohria V. (1979) Hyperactivity response to apomorphine and amphetamine in the mouse: the importance of the nucleus accumbens and caudate-putamen. J. Phurm. Pkur~c. 31, 259-261. 6. Crespi F., Sharp T., Maidment N. and Marsden C. (1983) Differential pulse voltammetry in t&o--evidence that uric acid contributes to the indole oxidation peak. Neurosci. Lett. 43, 203-208. 7. Ewing A. G., Bigelow J. C. and Wightman R. M. (1983) Direct in viva monitoring of dopamine released from two striatal compartments. Science, Wash. 221, 169-171, 8. Gonon F. G., Navarre F. and Buda M. J. (1984) In viuo monitoring of dopamine release in the rat brain with differential normal pulse voltammetry. Anulyt. Chem. 56, 573-575. 9. Greenfield S. A., Cheramy A., Leviel V. and Glowinski J. (1980) In vivo release of acetylcholinesteras in the cat substantiae nigrae and caudate nuclei. Nature 284, 355-357. 10. Hjorth S., Carisson A., Clark D., Svensson K., Wikstrom H., Sanchez D., Lindberg P., Hacksell U., Arvidsson L.-E., Johansson A. and Nilsson J. L. G. (1983) Central dopamine receptor agonist and antagonist actions of the enantiomers of 3-PPP. Psychopharmacology 81, 89-99. Il. O’Neill R. D. and Fillenz M. (1985) Detection of homovanillic acid in vivo using microcomputer-controlled voltammetry: simultaneous monitoring of rat motor activity and striatal dopamine release. Neuroscience 14, 753-763. 12. O’Neill R. D., Fillenz M. and Albery W. J. (1982) Circadian changes in homovanillic acid and ascorbate in the rat striatum using micropro~s~r-controll~ voltammetry. Neurosci. L&t. 34, 189-193. 13. O’Neill R. D., Fiftenz M. and Albery W. J. (1983) The development of tinear sweep voltammetry with carbon-paste electrodes in viva. J. NeHrosci. Meth. 8, 263-273. 14. O’Neill R. D., Fillenz M., Albery W. J. and Goddard N. J. (1983) The monitoring of ascorbate and monoamine transmitter metabolites in the striatum of the unanaesthetised rat using microprocessor-controlled voltammetry. Neuroscience 9, 81-93.
IS. O’Neill R. D., Grunewald R. A., Fillenz M. and Albery W. J. (1982) Linear sweep voltammetry with carbon-paste electrodes in the rat striatum. Neuroscience 7, 1945-1954. 16. O’NeilI R. D., Grunewald R. A., Fillenz M. and Albery W. J. (1983) The effect of unilateral cortical lesions on the circadian changes in rat striata1 ascorbate and homovaniliic acid levels measured in viuo using voltammet~. Neurosci. Left. 42, 105-I 10.
17. Westerink B. H. C. and Korf J. (1976) Turnover of acid dopamine metabolites in striatal and mesolimbic tissue of the rat brain. Eur. J. Pharmac. 37, 249-255. (Accepfed 3 April 1985)
N.S.C. W-D
0306-4522/85 163.00+ 0.00 Pergamon Press Lti 0 1985 IBRC
SIMULTANEOUS MONITORING OF DOPAMINE RELEASE IN RAT FRONTAL CORTEX, NUCLEUS ACCUMBENS AND STRIATUM: EFFECT OF DRUGS, CIRCADIAN CHANGES AND CORRELATIONS WITH MOTOR ACTIVITY R. D. O’NEILL*t and M. FILLENZ University Laboratory of Physiology, Parks Road, Oxford OX1 3PT, U.K. Abstract-Changes in homovanillic acid concentration, recorded in viuo by voltammetry with carbonpaste electrodes, were used as an index of dopamine release. With electrodes implanted in frontal cortex, nucleus accumbens and striatum, the pattern of dopamine release was monitored simultaneously in the three brain regions together with the rats’ motor activity. Changes in response to the systemic administration of dopamine-receptor agonists and antagonists were used as an index of feedback control of dopamine release. The relationships between dopamine release and motor activity, as well as that between dopamine release in the different brain regions, were investigated by calculating correlation coefficients for data collected over 24-h periods. The results show that dopamine release in frontal cortex is subject to little feedback regulation, that there is no nocturnal increase and no correlation with motor activity. Dopamine release in accumbens and striatum follows a very different pattern. There was a high correlation between dopamine release in these two regions on both sides of the brain; the correlation between the accumbens and the ipsilateral striatum was the highest. Dopamine release in both regions shows evidence of considerable feedback regulation, a nocturnal increase and a high correlation with
motor activity. The importance of the accumbens in relation to the level of motor activity is supported by the finding that the correlation coefficientbetween motor activity and dopamine release in this structure was significantly higher than that between activity and release in the striatum.
Two groups of dopaminergic cell bodies in the mammalian midbrain, A9 and AlO, give rise to the nigrostriatal, mesolimbic and mesocortical dopaminergic pathways which project to the striatum, nucleus accumbens and frontal cortex, respectively. Studies of each of these systems, using a variety of experimental techniques, suggest that they are associated with different functions and that the release of dopamine in the three target areas is subject to different regulatory mechanisms.’ So far, however, there have been few direct comparisons of the three systems in the same animal. With the development of microcomputer-based linear sweep voltammetry with carbon-paste electrodes,“.14 it is now possible to monitor continuously changes in dopamine release in discrete brain regions of the unrestrained rat over long periods by detecting changes in the extracellular concentration of the dopamine metabolite, homovanillic acid (HVA).‘2.‘6 The studies undertaken to identify the HVA signal in striatum have been reported fully.” We now describe an HVA signal in frontal cortex and accumbens and use the dopamine-synthesis inhibitor, a-methyl-ptyrosine, to confirm its identity; the effects of dopamine agonists and antagonists on the release
were also investigated. It has been suggested that there is asymmetry and mutual inhibition between the two nigrostriatal dopaminergic projections.9 To investigate this and explore other possible relationships between release of dopamine in these forebrain regions, correlation coefficients were calculated for the circadian changes in the HVA signal recorded in frontal cortex, accumbens and striatum. Finally we looked at the relationship between circadian changes in motor activity and dopamine release in each brain region. EXPERIMENTAL PROCEDURES Carbon-paste electrodes were prepared and implanted as previously described. ” Briefly, carbon paste (2.8 g carbon powder in I ml silicone oil) was packed into 300~pm (o.d.) Teflon-coated silver wire. Male Sprague-Dawley rats (initial weight 270-400 g) were stereotaxically implanted, under chloral hydrate anaesthesia, with these electrodes; silver wires, placed in the cortex, were used for the referen& and auxiliary electrodes. Rats were typically implanted with four carbon-paste working electrodes either in frontal cortex, nucleus accumbens, striatum and hippocampus or bilaterally in accumbens and bilaterally in striatum. (The results from the hippocampal electrodes are not relevent to the present study and no further mention is made of the hippocampus.) The co-ordinates, with the head level between bregma and lambda, were as follows: frontal cortexAP 3.0 (from bregma), L 1.3 (from bregma) and DV I.5 (from skull); accumbens-AP 1.5, L 1.5, DV6.5; and striatum-AP -0.5, L 3.0 and DV 4.8. The animals were given at least 2 days to recover before being placed in the recording cages and connected to the microcomputercontrolled equipment.14
*To whom all correspondence should be addressed. tBeit Memorial Research Fellow. Abbreviations : HVA, 3PPP, homovanillic acid; 3-(3-hydroxyphenyl)-N,n-propylpiperidine. 49
50
R. D. O’NEILLand M. A. Voltammogram
from
frontal
FILLENZ
cortex
0 B(i) 4
E(M)
E(il)
cortex: Id3=0.13+0.06
wcteur
I rontal
-
nA (n=@)
3
r 13~0.21
f0.02
nA (n=14)
clcoumbenw
4
nA
3 nA
nA
2
2
2
3
1
0
1 430
0
’
01
430 SOOmU 570 500 mU 570 Fig. I. (A) Example of a voltammogram, recorded in the frontal cortex of an unrestrained rat, using microcomputer-controlled linear sweep voltammetry with a carbon-paste electrode. A linearly increasing potential (range O-650 mV with respect to an Ag reference electrode) is applied at a sweep rate of 5 mV/s to the implanted carbon-paste electrode, and the oxidation current produced recorded as a function of potential. When the background current, which is measured in situ by scanning continually, is substracted from voltammograms recorded at 1Zmin intervals, a difference volt~mogmm consisting of three separate peaks is obtained. Peak 1 is due to the oxidation of ascorbic acid, peak 2 to uric acid and the dopamine metabolite HVA is responsible for peak 3. Changes in the height of each peak are proportional to ehanges in the extracellular concentration of the corresponding compound around the electrode tip. (B) Peak 3 recorded in three brain regions of the same rat. (i) Frontal cortex: average peak potential ( V3)= 490 &-10 mV; average peak height (h3) = 0.13 f 0.05 nA (n = 9). (ii) Nucleus accumbens: V, = 490 + 5 mV; h, = 0.42 -(_0.04 nA (n = 14). (iii) Striatum: V, = 480 f 6 mV; h, = 0.21 + 0.02 nA (n = 14). 500 mU
570
430
Linear sweep voltammograms were recorded at 12-min intervals at a rate of 5 mV/s between 0 and 650 mV. The background current for each electrode was measured in situ before each experiment as described previously.‘3 Eight electrodes (i.e. two rats) were scanned simultaneously and the voltammograms stored on magnetic disc. Voltammograms recorded from the three brain regions usually consisted of three separate peaks; in some rats peak 3 was not observed in frontal cortex. The height of peak 3 (h3) was measured as indicated in Fig. 1.Between the electrochemical recordings (for approx. 9min duration) the total motor activity of each rat was monitored using a Doppler-shift microwave device linked to the PI0 of the interface.”
Drugs were administered i.p. using either 0.9% saline or distilled water (for apomorphine only) as the carrier. The significance of the effects of drugs on h, in each brain region was calculated using paired r-tests by comparing the average pre- and post-injection values for each electrode. The significance of the difference in effects between brain regions was calculated by comparing the percentage changes in the same rat. The mean and SEM are quoted; n is the number of electrodes. A group of 8 rats were given no drugs and were used to record simultaneously the circadian variation in motor activity and h, in frontal cortex, accumbens and striatum. To reduce the fluctuations inherent in recording these small
Monitoring dopamine release in rat forebrain using voltammetry
51
Table 1. Percentage changes in the homovanillic acid signal recorded in three brain regions Brain region Frontal cortex
Haloperidol
Nocturnal increase
AMPT
Apomorphine
-94+3(4) (P < 0.001)
-24*6(4) (P < 0.05)
50 k 17 (6) (P < 0.05)
20* 19(9) NS
Accumbens
-96*2(S) (P < 0.001)
-63 k 7(6) (P < 0.001)
306 k 55 (6) (P < 0.005)
48k7(14) (P < 0.001)
Striatum
-98 f I (5) (P < 0.001)
-74*8(7) (P < 0.001)
365 rt. 81 (9) (P < 0.005)
52 f 5 (14) (P < 0.001)
The effects of a-methyl-p-tyrosine (AMPT, 250 mg/kg i.p.). apomorphine (2 mg/kg i.p.) and haloperidol(0.5 mg/kg i.p.) on the HVA signal recorded in three brain regions. The results are expressed as a percentage change compared with the average pre-injection value. The nocturnal increase is the percentage increase compared with the average day-time value. Numbers in brackets are the number of electrodes. Significance calculated using r-tests on data from each electrode. NS = not significant. currents (< I nA) and clarify the systematic relationships between the variables, a weighted moving average of the original data was calculated.” Correlation coefficients between h, in the different regions and between h, and motor activity were calculated for each 24-h period (120 points). Cross-correlations were also calculated by comparing, e.g. m(t + 1) with h,(r), where h,(t) is the height of peak 3 at time t (l2-min intervals) and m(r) is the value of the movement counter following that scan. Cross-correlations were calculated in the range m(r - 5) to m(r + IO). Data were recorded from each rat for a period of 2 or 3 days; the n value for these results are the number of determinations. RESULTS Voltammograms
recorded
in frontal
cortex,
nu-
cleus accumbens and striatum consisted of three separate peaks (Fig. 1). The average potential of peak 3 was: frontal cortex = 490 f 10 mV, accumbens = 490 + 5 mV and striatum = 480 + 6 mV. The average height of peak 3 (h,) recorded in the three brain regions was: frontal cortex = 0.13 f 0.05 nA (n = 9), accumbens = 0.42 f 0.4 nA (n = 14) and striatum = 0.21 & 0.02 (n = 14) (n = number of electrodes). There was no significant difference in h, recorded in frontal cortex and striatum; hJ in accumbens, however, was significantly greater than that in the other two regions (P < 0.01, see Fig. 1). The identity of peak 3 recorded in striatum with homovanillic acid (HVA, a dopamine metabolite) is established; we used the dopamine-synthesis inhibitor a-methyl-p-tyrosine (Sigma) to determine the extent of the contribution of HVA to h, recorded in frontal cortex and accumbens. a-Methyl-p-tyrosine (250mg/ kg i.p.) caused a decrease in /I~ in the three regions: in frontal cortex by 94 + 3% (n = 4, P < O.OOl), in accumbens by 96 + 2% (n = 5, P < 0.001) and in striatum by 98 + 1% (n = 5, P < 0.001). To compare the contribution made by dopamine in regulating its own release, the effects on hJ in the three regions of the dopamine-receptor agonist, apomorphine (Sigma), and antagonist, haloperidol (Searle), were investigated. Apomorphine (2 mg/kg i.p.) caused a decrease in h, in the three regions: in frontal cortex by 24 f 6% (n = 4, P < 0.05), in accum-
bens by 63 + 7% (n = 6, P < 0.001) and in striatum by 74 + 8% (n = 7, P < 0.001). There was no significant difference in the percentage change recorded in accumbens and striatum; the difference between the effect in frontal cortex and that in accumbens and striatum was significant with P c 0.05 and 0.01, respectively. Haloperidol(O.5 mg/kg i.p.) increased h, in the three brain regions: in frontal cortex by 50 f 17% (n = 6, P < 0.05) in accumbens by 306 k 55% (n = 6, P < 0.005) and in striatum by 365 + 81% (n = 9, P -e0.005). As with apomorphine, there was no significant difference between the effects in accumbens and striatum; the difference between the effect in frontal cortex compared with that in accumbens and striatum was significant with P c 0.01and 0.05, respectively. The effects of these drugs are summarized in Table 1. To investigate the effect of 3-(3-hydroxyphenyl)N,n-propylpiperidine (3PPP) on dopamine receptors in accumbens and striatum, the two enantiomers of this drug were injected into rats implanted with electrodes bilaterally in accumbens and striatum. The results are shown in Table 2; (+)3PPP and (-)3PPP (1 and 25 mg/kg i.p.) produced similar effects in both brain regions, and although the average effect in striatum was greater than that in accumbens for all doses, this difference was significant only for (-)3PPP at 1 mg/kg. Circadian changes in extracellular HVA in the three brain regions were measured in 8 rats which had not been given drugs. The minimum values of hJ (recorded during daytime, typically between 10.00 and 12.00 h) and the maximum values (recorded during the night-time, active period, between 04.00 and 06.00 h) were as follows: frontal cortex, 0.13 + 0.05/O. 16 f 0.07, increase = 20 + 19% (n = 9, not significant); accumbens, 0.42 f 0.04/0.62 + 0.06, increase = 48 + 7% (n = 14, P < 0.001); and striatum, 0.21 + 0.02/0.32 &-0.04, increase = 52 &-5% (n = 14, P < 0.001). There was no significant difference between the percentage increase observed in accumbens and striatum (Table 1).
R. D. @NEILL and M. FILLENZ
52
Table 2. Effect of 3-(3-hydroxyphenyl)-N,n-propylpiperidine
on the homovaniilic acid signal
(+)Isomer Accumbens
1w/kg
Accumben: - )lsomer Striatum
Striatum
-22 f So/, tl=8 -46 + 10% n=6 - 8 i 2% n=4 - 38 i 7% n=3 P < 0,05 P < 0.01 P < o,05 P < o.05
25 mg/kg
n=S -58i80i,p
-75,120/,Pn<&
67+ 13&,n<;;2 132&35?,,n<=o;5
The effects of the two enantiomers of 3-(3-hydroxyphenyl)-N,n-propylpiperidine (3PPP) on the HVA signal recorded in nucleus accumbens and striatum. Results expressed as a percentage of the pre-injection value; n is the number of electrodes.
The intercorrelations for circadian changes in the HVA signal in accumbens and striatum are shown in Table 3. There was no significant difference between values calculated for accumbens vs contralateral accumbens, striatum vs contralateral striatum or accumbens vs contralateral striatum. However, the correlation coefficient calculated for accumbens vs ipsilateral striatum was significantly greater than all contralateral values (P < 0.01). For frontal cortex, only correlations with contralateral accumbens and ipsilateral striatum are available: 0.46 &-0.10 (n = 13) and 0.42 f 0.12 (n = 13), respectively. In all cases, the maximum values of the intercorrelation were obtained when the time shift was zero. We next looked at the correlation between motor activity and the HVA signal in the three regions (Table 4). There was no significant correlation between motor activity and HVA in frontal cortex. The correlation for HVA in accum~ns vs motor activity (0.68 f 0.03, n = 25) was significantly greater than that for striatal HVA (0.56 & 0.04, n = 24) with P < 0.01. The maximum correlation was obtained when the motor activity data were shifted by
24 f 7 min with respect to the HVA signal in accumbens and by 30 i 7min with respect to the striatal HVA signal; there was no significant difference between these two shifts. The same data were also analysed by averaging the 24-h recordings of motor activity and the HVA data from each brain region. Cross-correlations were calculated for these averaged time courses: in frontal cortex, r = 0.267, shift = -48 mitt; in accumbens, r = 0.926, shift = + 36 mitt; and in striatum, r = 0.832, shift = + 36 min (Fig. 2).
DISCUSSION Voltammetry in vivo was initially developed as a technique to monitor changes in the concentration of transmitter substances in the extracellular fluid. Howdrug-induced* and electrically ever, although stimulated’ changes in dopamine concentration have been reported, it has become clear that the concentration of transmitters, under normal conditions, is too low to measure in situ with the present electro-
Table 3. Correlation coefficients for changes in the homovanillic acid signals Correlation coefficients
Brain regions Accumbens Striatum vs Accumbens Accumbens
vs accumbens (contralateral) striatum (contralateral) vs striatum (contralateral) vs striatum (ipsilateral)
0.61 k 0.60 + 0.67 + 0.78 k
O.lOfn 0.09 (n 0.03 (n 0.05 (n
= = = =
6) 6) 24) IO)*
Correlation coefficients for the circadian variation in the HVA signal in nucleus ac~umbens and striatum calculated for each day’s data (120 points); n = number of determinations. Maximum correlations were obtained when the time shift was zero. *Significantly greater than contralateral values, P < 0.01.
Table 4. Correlation
Brain region Frontal cortex Nucleus accumbens Striatum
coefficients for motor activity vs homovanillic acid signals Correlation coefficient 0.1 kO.4 0.68 * 0.03* 0.56 f 0.04
Time shift (min) 24 * 7 30*7
n-value 13 25 24
Correlation coefficients for circadian changes in total motor activity vs the HVA signal in frontal cortex, nucleus accumbens and striatum. Values calculated for each day’s data (120 points); n = number of determinations. Time shift refers to the displacement of the motor-activity data which gave the maximum correlation; i.e. the voltammetric response lagged behind changes in motor activity. *Greater than that for striatum. P < 0.01.
53
Monitoring dopamine release in rat forebrain using voltammetry
changes
Circadian
in HVA and motor activity
0.4 A. nucleus
accumber
19
0.3
100
WA
~ h3lnA
,ounts
0.2 r SO.926
shift ~38
0.
200 min f
f
motor
activity
DARK
0.0
2
15
18
El
24
J
6
5
12t
0
Time of day
0.4 6. striatum
400
HVA
0.3
t
h3/nA
counts
0.2 rr0.832 shift -36
0.
200
mln
I motor
activity
DARK
0.0
2
15
18
21
,24
J
6
9
32t
0
Time of day Fig. 2. Averaged time course of the circadian changes in total motor activity and the HVA signal in the nucleus accumbens (A, n = 25) and striatum (B, n = 24). Left ordinate is the height of peak 3 (h,) in nA, right ordinate is the activity count recorded between electrochemical recordings; abscissa is the time of day in hours. 120 points at It&n intervals. Shift is the displacement of the movement data needed for m~mum correlation.
54
R. D.
O’NEILL and
chemical techniques. Many voltammetric methods can, however, detect transmitter metabolites such as the dopamine metabolite, 3,4-dihydroxyphenylacetic acid” and the serotonin metabolite, S-hydroxy indoleacetic acid.4.6 We have shown that peak 3 recorded using linear sweep voltammetry with carbon-paste electrodes in rat striatum is due to yet another metabolite of dopamine, homovanillic acid (HVA).“.” Voltammograms recorded using this technique consist of three separate peaks in areas with a significant dopaminergic input such as frontal cortex, accumbens and striatum (Fig. l), but only peak 1 (ascorbic acid) and peak 2 (uric acid) are seen in occipital cortex, hippocampus and cerebellum, where dopamine transmission is minimal; moreover, the potential of peak 3 in the three brain regions is the same (Fig. 1). These observations indicate that peak 3 in frontal cortex and accumbens is likewise due to HVA and the conclusion is supported by the greater than 95”/, reduction in the peak height caused by the dopa~ne-synthesis inhibitor, ~-methyl-p-tyrosine (Table I>. (Contributions from the methylated metabolites of noradrenaline in frontal cortex and accumbens can be neglected since these are sulphate conjugated, and therefore are non-electroactive.) Dopamine gives rise to four metabolic end-products: the free and conjugated forms of 3,4-dihydroxyphenylacetic acid and HVA. Westerink and Korf (1976)17 found that, following certain pharmacological manipulations, free 3,4-dihydroxyphenylacetic acid and HVA did not change in parallel and the authors suggested that 3,4-dihydroxyphenylacetic acid levels were a more accurate reflection of dopamine release. However, no such comparison in the absence of drugs has been made. Changes in the extracellular concentration of striatal HVA show a high correlation with motor activity and druginduced changes in the HVA signal, up to I h after injection, are similar to the known effects of the drugs on dopamine release. We have therefore concluded that, under physiological conditions, changes in HVA reflect changes in the release of dopamine;” it is reasonable to assume that the HVA signals recorded in frontal cortex and accumbens are an index of dopamine release in their respective areas. We have previously reported the effects of drugs on striatai HVA;” we have repeated some of these experiments in rats with electrodes in the three brain regions to investigate regional differences. The lack of correlation between the size of the HVA signal (peak 3) and dopamine content, or even dopamine turnover, in different brain regions is due to a number of factors. The electrode monitors the HVA concentration in its immediate vicinity and not in the brain region as a whole. The distribution of dopamine terminals in the frontal cortex, in contrast to that in the striatum, is non-uniform in that it is confined to certain layers. As a consequence, an HVA signal is always present in recordings from but is occasionally undetectable in striatum”
M.
FILLENZ
frontal cortex. This leads to a bias in the sampling since zero values are not included in the mean peak height. The finding that the dopamine-receptor agonist, apomorphine, and antagonist, haloperidol, produced large changes in dopamine release in accumbens and striatum but only very small changes in frontal cortex (Table I ), supports the view that there is much Iess feedback inhibition of dopamine release in frontal cortex.’ Since neither apomorphine nor haloperidoi are selective for pre- or postsynaptic dopamine receptors, they do not allow the distinction to be made between long-loop feedback mediated through postsynaptic receptors and local feedback through presynaptic autoreceptors. However, the failure of dopamine agonists to reverse the dopamine accumulation following gamma-butyrolactone administration,’ or to reduce neuronal firing’ provides evidence that a sub-population of mesocortical dopaminergic neurones lack both presynaptic and somatodendritic autoreceptors. Although the experiments with apomorphine and ha~operido~ showed no significant difference between the effects in accumbens and striatum, it is noteworthy that the average percentage change in accumbens was less than that in striatum in both cases; this would suggest that the autoregulation of dopamine release in accumbens is weaker than in striatum. This conclusion is supported by the results of the administration of the two isomers of 3PPP. At both doses of these two isomers, the average effect in accumbens was less than that in striatum and at 1 mg/kg (-)3PPP the difference was statistically significant (Table 2). These results do not support the suggestion that 3PPP has a preferential action in accumbens.” They show that the effects of the drug in accumbens, as in striatum,” are consistent with the suggestion” that at low doses (+)3PPP acts as a presynaptic dopamine-receptor agonist, and as a postsynaptic agonist at higher doses; (-)3PPP, however, appears to have presynaptic dopamine-receptor agonist properties at low doses, but behaves as a postsynaptic antagonist at high doses. The failure to observe a significant nocturnal increase in dopamine release in frontal cortex (Table I), together with the higher correlation between motor activity and dopamine release in accumbens and striatum compared with frontal cortex (Table 4), suggests a different role for dopamine release in the three brain regions in relation to motor activity. The significantly higher correlation for motor activity vs dopamine release in accumbens compared with striatum supports the view that dopamine release in accumbens is more important in controlling activity levels, whereas release in striatum may have more to do with patterns of movement.5 The correlations between dopamine release in accumbens and in striatum (Table 3) are interesting on two counts. First, the high correlation for each structure with its contralateral counterpart, which suggests that mutual inhibjtion9 does not play an
Monitoring dopamine release in rat forebrain using voltammetry important role under normal conditions. Second, the significantly greater correlation, compared with all contralateral values, for accumbens vs ipsilateral striatum. This result may reflect the amount of turning movements which take place in the small recording cages (25 x 25 x 25 cm3) leading to preferential release of dopamine in the two structures on the side contralateral to the direction of turning. Summary
This study illustrates some new approaches in the study of transmitter release in the unrestraint ani-
55
mal made possible by voltammetry in vivo. The results show that, as expected from pharmacological studies, changes in the release of dopamine in nucleus accumbens and striatum, but not in frontal cortex, are positively correlated with changes in the level of motor activity. Acknowledgements-The
work was supported in part by the Medical Research Council. We are grateful to the E. P. Abrahams Cephalosporin Fund for a grant to obtain additional equipment. We thank Astra Lakemedel, Sweden, for the gift of 3PPP, Dr. J. F. Stein for helpful discussions and Mr. M. R. BIoom~eId for art work.
REFERENCES 1. Bannon M. J. and Roth R. H. (1983) Pharmacology of mesocortical dopamine neurones. Pharmuc. Rev. 35, 53-68. 2. Bannon M. J., Michaud R. L. and Roth R. H. (1981) Mesocorticai dopamine neurons: lack of autoreceptors modulating dopamine synthesis. ~0~~~. Pharmac. 19, 270-275. 3. Bannon M: J., Chiodo L. A., Roth R. H. and Bunney B. S. (1982) Mesocortical dopamine neurons. I. Electrophysiological and biochemical evidence for the absence of autoreceptors in a subpopulation. Sot. Neurosci. Abstr. 8, 480. 4. Cespuglio R., Faradji H., Riou F., Buda M., Gonon F., Pujol J. F. and Jouvet M. (1981) Differential pulse voltammetry in brain tissue. II. Detection of J-hydroxyindoleacetic acid in the rat striatum. Bruin. Res. 223, 299-311. 5. Costall B., Naylor R. J. and Nohria V. (1979) Hyperactivity response to apomorphine and amphetamine in the mouse: the importance of the nucleus accumbens and caudate-putamen. J. Phurm. Pkur~c. 31, 259-261. 6. Crespi F., Sharp T., Maidment N. and Marsden C. (1983) Differential pulse voltammetry in t&o--evidence that uric acid contributes to the indole oxidation peak. Neurosci. Lett. 43, 203-208. 7. Ewing A. G., Bigelow J. C. and Wightman R. M. (1983) Direct in viva monitoring of dopamine released from two striatal compartments. Science, Wash. 221, 169-171, 8. Gonon F. G., Navarre F. and Buda M. J. (1984) In viuo monitoring of dopamine release in the rat brain with differential normal pulse voltammetry. Anulyt. Chem. 56, 573-575. 9. Greenfield S. A., Cheramy A., Leviel V. and Glowinski J. (1980) In vivo release of acetylcholinesteras in the cat substantiae nigrae and caudate nuclei. Nature 284, 355-357. 10. Hjorth S., Carisson A., Clark D., Svensson K., Wikstrom H., Sanchez D., Lindberg P., Hacksell U., Arvidsson L.-E., Johansson A. and Nilsson J. L. G. (1983) Central dopamine receptor agonist and antagonist actions of the enantiomers of 3-PPP. Psychopharmacology 81, 89-99. Il. O’Neill R. D. and Fillenz M. (1985) Detection of homovanillic acid in vivo using microcomputer-controlled voltammetry: simultaneous monitoring of rat motor activity and striatal dopamine release. Neuroscience 14, 753-763. 12. O’Neill R. D., Fillenz M. and Albery W. J. (1982) Circadian changes in homovanillic acid and ascorbate in the rat striatum using micropro~s~r-controll~ voltammetry. Neurosci. L&t. 34, 189-193. 13. O’Neill R. D., Fiftenz M. and Albery W. J. (1983) The development of tinear sweep voltammetry with carbon-paste electrodes in viva. J. NeHrosci. Meth. 8, 263-273. 14. O’Neill R. D., Fillenz M., Albery W. J. and Goddard N. J. (1983) The monitoring of ascorbate and monoamine transmitter metabolites in the striatum of the unanaesthetised rat using microprocessor-controlled voltammetry. Neuroscience 9, 81-93.
IS. O’Neill R. D., Grunewald R. A., Fillenz M. and Albery W. J. (1982) Linear sweep voltammetry with carbon-paste electrodes in the rat striatum. Neuroscience 7, 1945-1954. 16. O’NeilI R. D., Grunewald R. A., Fillenz M. and Albery W. J. (1983) The effect of unilateral cortical lesions on the circadian changes in rat striata1 ascorbate and homovaniliic acid levels measured in viuo using voltammet~. Neurosci. Left. 42, 105-I 10.
17. Westerink B. H. C. and Korf J. (1976) Turnover of acid dopamine metabolites in striatal and mesolimbic tissue of the rat brain. Eur. J. Pharmac. 37, 249-255. (Accepfed 3 April 1985)
N.S.C. W-D