Release of endogenous glutamate from rat cerebellar synaptosomes: Interactions with adenosine and ethanol

Life Sciences, Vol. 44, pp. 1625-1635 Printed in the U.S.A.

Pergamon Press

RELEASE OF ENDOGENOUS GLUTAMATE FROM RAT CEREBELLAR SYNAPTOSOMES: INTERACTIONS WITH ADENOSINE AND ETHANOL Mike Clark' and M. Saeed Dar* Department of Pharmacology, School of Medicine 27858 East Carolina University, Greenville, NC (Received in final form Zlarch 23, 1989) Summarv The effects of ethanol and adenosine receptor agonist R-PIA and antagonist theophylline on release of endogenous glutamate were tested in rat cerebellar synaptosomal preparation. Release was carried out for 5 to 60 set after which time the released glutamate was separated from the synaptosomal membranes by rapid filtration. The amount of released glutamate in the filtrate was measured by an enzyme-l-inked fluorometric assay. Basal endogenous glutamate release was estimated as 3.7 + 0.3 nmol/mg protein/5 set and was stimulated by high Kt. Glutamate release consisted of an initial rapid phase for the first 10 set that was followed by a relatively slower phase. Both Ca*+dependent and Ca +-independent glutamate release were observed which suggested the involvement of both neuronal and glial constituents of the synaptosomal preparation, respectively. Pharmacologically relevant concentrations of ethanol (25-100 mM) caused a trend toward a dose-dependent inhibition of glutamate release. R-PIA and theophylline inhibited and stimulated, respectively, basal release of glutamate and R-PIA-inhibited release was blocked by theophylline. Ethanol (25 mM) blocked the stimulatory effect of theophylline and the results of experiments following the inclusion of adenosine deaminase suggested the involvement of adenosine in this effect of ethanol. The results support our previous findings that suggest an involvement of cerebellar adenosine in the motor disturbing effects of acute ethanol and extend those findings by indicating that ethanol inhibits glutamate release from granule cells of the cerebellar cortex through an adenosine-sensitive mechanism. Abundant data now exist that suggest a neurotransmitter role in the CNS for the excitatory amino acid glutamate (1,2). The granule cells of the cerebellar cortex appear to use glutamate as the excitatory transmitter (3,4) releasing the amino acid from their parallel gibers in the molecular layer (5). Synaptosomal release of glutamate is Ca -dependent and stimulated by

'Current address: BPB, NIMH, Bldg 10, Rm 3N212, Bethesda, MD *To whom all correspondence should be sent.

0024-3205189 $3.00 + .OO Copyright (c) 1989 Pergamon Press plc

20892

1626

Glutamate Release, Adenosine and Ethanol

Vol. 44, No. 22, 1989

Kt-depolarization in both $.hecerebral (6) and cerebellar (4,5) cortices. However, Kt-stimulated, Ca +-independent release also occurs and was suggested to involve a different storage pool for glutamate (5,6). The Nat channel activator veratridine stimulated glutamate release from cultured cerebellar granule cells in a manner that was blocksd by tetrodotoxin (7) which suggested that glutamate release can occur in a Ca +-independent manner. Pull and McIlwain (8) observed that glutamate potently increased purinergic efflux from slices of guinea pig cortex. Similarly, Jhamandas and Dumbrille (9) found that application of glutamate to the exposed cerebral and c rebellar cortices of anesthetized rats markedly stimulated the release of [ sHladenosine derivatives. Adenosine is well accepted to act as a neuromodulator in the CNS by inhibiting neuronal activity and neurotransmitter release and modulating CAMP levels (10,11,12). Adenosine and its analogues inhibit release of glutamate from slices of rat hippocampus (13,14) and dentate gyrus (15) as well as from cultured cerebellar neurons (16,17). Several studies (18,19,20,21,22,23) suggested that central adenosine is involved in the CNS effects of ethanol. We recently observed a functional correlation between ethanol-induced motor disturbances and increased density of adenosine A binding sites in rat csrebellar cortex (21). Additionally, we found that 25 mM ethanol inhibits [ Hladenosine uptake (24) and concurrently enhances release of endogenous adenosine (25) from rat cerebellar synaptosomes. In view of 1) the opposing actions of glutamate and adenosine on each other's release, 2) the apparent involvement of adenosine in the CNS effects of ethanol and 3) the finding (26) that acute ethanol lowers cerebellar tissue levels of glutamate, the present study was carried out to investigate the possibility of an adenosine-mediated interaction between ethanol and cerebellar synaptosomal glutamate release. Methods Preparation of Svnaotosomes: Synaptosomes were prepared according to the method of Gray and Whittaker (27) as modified by White (28). Four male Sprague-Dawley rats (150-200 g; Charles River, Raleigh, NC) were decapitated. The cerebellae were pooled, homogenized in 10 ml of 0.32 M sucrose and centrifuged at 1000 g for 5 min. The pellet was resuspended in 15 ml of 0.32 M sucrose and recentrifuged. All sucrose solutions were buffered to pH 7.5 with 5 mM Tris-HCl and maintained at 4'C. All centrifugations were carried out at 4Oc. The two supernatants were combined and subjected to centrifugation at 12,000 g for 20 min. The resulting pellet was resuspended in 12.2 ml of 0.32 M sucrose, layered over a discontinuous sucrose density gradient (0.8 M and 1.2 M) and centrifuged for 60 min at 150,000 g. The 0.8:1.2 M interface was removed, diluted to 40 ml with 0.32 M sucrose and centrifuged at 20,000 x g for 20 min. The resulting pellet (synaptosomes) was resuspended in 1.1 ml of incubation solution (pH 7.5-7.6) which contained in final concentration (mM): 120 NaCl, 4.75 KCl, 1.18 MgSO4, 26 NaHC03, 1.2 KH2P04, 1.77 CaC12, 5.5 glucose and 58.5 sucrose. Except for necessary mixing (2-3 set), the synaptosomal suspension was kept on ice throughout the experiment. An aliquot of this suspension was taken for protein analysis by the method of Lowry et al. (29). Release of Endoqenous Glutamate bv Cerebellar Svnaotosomes: Release of endogenous glutamate was initiated by adding 40 ul (200 to 280 ug of protein) of synaptosomal preparation (maintained at O'C) to tubes containing 960 ul of incubation solution maintained at either 37'C or O'C. The release was allowed to continue for 5, 10, 30 or 60 set at either 37'C (to obtain total values) or O°C (to obtain blank values). Identical constituents were assayed at the two temperatures for strict controls. After the incubation period, the incubation solution (which now contained the released glutamate) was rapidly separated

Vol. 44, No. 22, 1989

Glutamate Release, Adenosine ani'Ethano1

1627

from the synaptosomes by vacuum filtration through Whatman GF/C ,glass microfibre filters. For the various release experiments conducted, the incubation medium of appropriate tubes (totals and blanks) contained (in final con entration) 50 mM KCl, 0.5 mM ethyleneglycol-bis-(beta-amino-ethyl-ether)-N,N F tetraacetic acid (EGTA; Sigma, St Louis, MO), 10 nM N6-(R-2-phenylisopropyl)adenosine (R-PIA; Sigma), 10 nM N6-(S-2-phenylisopropyl)-adenosine (S-PIA; Sigma), 2 IU/ml adenosine deaminase (ADA;EC 3.5.4.4.; Sigma), 5 uM theophylline (Sigma), and/or 25, 50, 75, or 100 mM ethanol. Measurement of Released Glutamate: Glutamate content of the filtrate 7 (i.e., the released glutamate) was measured by an enzyme-linked fluorometric assay according to the method reported by Nicholls et al. (30). L-Glutamic dehydrogenase (GDH; EC 1.4.1.3; Sigma, Type G-2626) was the enzyme used to convert glutamate to 2-oxyglutarate. This oxidation reaction required the presence of an electron acceptor. Nicotinamide adenine dinucleotide phosphate (NADP+ monosodium salt; Sigma) was utilized as the electron acceptor in these assays. Fluorescence of NADPH (a product of the reaction directly proportional to the amount of glutamate oxidized) was measured at an excitation wavelength of 360 nm and an emission wavelength of 470 nm (uncorrected values) to determine the amount of glutamate present in the cuvette. A series of experiments was conducted to establish optimum assay conditions. All glutamate assays were carried out in a 2 ml final volume. Separate pilot experiments were conducted in which either different amounts of GDH (2, 3.7, 7.4 and 14.8 units) were used to determine the optimum amount of GDH for the completion of the glutamate conversion reaction or various incubation times (30, 60, 90, 120, 150 and 270 min) were tested for optimal conversion reaction for various amounts of glutamate. The concentration range of glutamate selected (0.25 to 10 nmoles) was based on the amount of synaptosomal glutamate released and measured by Nicholls et al. (30) under similar assay conditions. The reaction (or incubation) was conducted at room temperature (22-24'C). Data (not shown) from pilot experiments revealed that both 2 and 3.7 units of GDH required 90 min for completion of reaction, while the reaction was complete within 30 min in the presence of either 7.4 or 14.8 units of GDH. Therefore, 7.4 units (10 ul) of GDH and a reaction time of 30 min were used in all subsequent glutamate measurement experiments. For all assays, the final concentration of NADP+ was 1 mM which was reported to be sufficient for adequate conversion of greater than 25 nmoles of glutamate (30). Glutamate standards were prepared in incubation solution, while NADP+ was prepared in distilled water. For measurement of released glutamate, 300 ul of filtrate, glutamate standard, or incubation solution (for blank), 100 ul of NADP+ and 1.6 ml of distilled water were placed in a cuvette. Ten microliters (7.4 units) of GDH were subsequently added to initiate the reaction. The solutions were well mixed immediately following the addition of GDH. After the 30 min incubation, fluorescence of NADPH was measured (Model 430 Spectrofluorometer; G.K. Turner Associates, Palo Alto, CA). Standards (0.5, 1, 2, 3, 4 and 5 nmoles) of glutamate were run with each assay. The relative fluorescent intensity (RFI) of NADPH formed was linear over the range of 0.5 to 5 nmoles of glutamate (data not shown). The RF1 of filtrate from samples fell within this linear range of standards. Analvsis of Data: The time course of glutamate release was statistically tested by a two-way ANOVA with Newman-Keuls post hoc analysis for a significant interaction between incubation time and treatment (presence or absence of KCl). This was followed by one-way ANOVA at appropriate incubation times to test for statistically significant increases in basal release by KCl. Unless stated otherwise, the rest of the glutamate release data was analyzed with one-way ANOVA and Newman-Keuls post hoc comparisons. Data were considered statistically significant at pcO.05.

1628

Glutamate Release, Adenosine and Ethanol

Vol. 44, No. 22, 1989

Results For clarity, an experiment consisted of a single batch of synaptosomes prepared from the pooled cerebellae of four rats. In each experiment, six tubes with identical contents were assayed (3 at 37'C and 3 at O'C) for each variable (e.g., incubation time or drug concentration). Each experiment was repeated as indicated in the figure legends using a new batch of synaptosomes to obtain sample numbers large enough for statistical analyses. A preliminary experiment was conducted to determine the relationship between the concentration of synaptosomal membrane protein and release of endogenous glutamate (data not shown). The amount of protein ranged from 61 ug to 368 ug in a single assay with linearity observed for all protein quantities. Based on this information the synaptosomal protein range of 200 to 280 ug was selected for all subsequent glutamate release experiments.

22-r 20 --

16--

7

14--

’

12 -10 --

"

,

I

,

0 -6 -rota1

* 4 --

a Icvcold

2 --

.mt

0-l

0

I 1

,I

1I

I

I

,I

I

10

20

30

40

50

60

70

Time bed Fig. 1. Time course of endogenous glutamate release from rat cerebellar synaptosomes at either 37OC (total) or 0' (ice-cold). Net release was calculated as the difference between total and icecold blank. The results are the mean +S.E.M. of four separate experiments done in triplicate. Figure 1 shows the relationship between the incubation time (in set) and release of endogenous glutamate. The release was composed of an initial rapid phase followed by a relativel{ slower phase. No release occurred during the 60 set incubation period at 0 C as noted by the zero-slope of the Ice-cold line of Figure 1. Figure 2 shows that KC1 enhanced glutamate release by 85, 38 and 46% at 5, 10 and 30 set, respectively. Release was not significantly different in the presence or absence of KC1 with a 60-set incubation period. Since KCl-stimulated release of glutamate was most pronounced in the initial rapid phase, all subsequent assays were conducted with a 5 set incubation period.

vol. 44, No. 22, 1989

1629

Glutamate Release, Adenosine and Ethanol

.

control

. 50 WI KC1

0’ 0

10

20

30

40

50

60

70

Time (set) Time course of basal (control) release of endogenous FIG. 2. glutamate from rat cerebellar synaptosomes and the effects of 50 mM KCL on it. The results are the mean t.J.E.M. of foyg to six sepapt0.01 compared pt0.05 and rate experiments done in triplicate. to release at respective control time. An experiment was conducted (and repeated six times) to investigate the dose-response relationship between basal release of endogenous glutamate and various pharmacologically relevant concentrations (25-100 mM) of ethanol. Basal release in the absence of ethanol was estimated to be 3.7 + 0.3 nmol/mg protein/5 sec. Linear regression analysis showed that ethanol produced a dose-dependent inhibition of release. However, this inhibition was not statistically significant as analyzed by ANOVA. Endogenous glutamate release was 95, 86, 78 and 70% of basal (control) release at 25, 50, 75 and 100 mM ethanol, respectively. The effects of various compounds on the release of endogenous glutamate in the presence or absence of 25 mM ethanol are shown in Figure 3. Although ethanol inhibited KCl-stimulated release by 13X, this effect did not reach statistical significance. On the other hand, ethanol profoundly2tlocked the release stimulated by EGTA in an incubation medium coe)aining Ca (Fig. 3). However, in an incubation medium that contained no Ca or EGTA, basal glutamate release was 51% (2.2 + 0.7 nmol$ng protein/5 set; pt0.05) of control (basal release in the presence of Ca ; data not shown). The adenosine agonist R-PIA significantly reduced basal release of glutamate by 41% (2.2 + 0.4 nmol/mg protein/5 set) and 54% (1.7 + 0.4 nmol/mg protein/5 set) in the absence and presence of 25 mM ethanol, respectively (Fig. 3). However, the increased inhibition when ethanol was included in the incubation medium with R-PIA was not statistically different from the effect of R-PIA alone. Con-

1630

Vol. 44, No. 22, 1989

Glutamate Release, Adenosine and Ethanol

versely, adenosine antagonist theophylline markedly stimulated glutamate release to 5.9 + 0.9 nmol/mg protein/5 sec. This effect of theophylline was reduced to control value by 25 mM ethanol (3.7 + 0.6 nmol/mg protein/5 set; Fig. 3). Figure 4 shows the effects of the two diastereoisomers of PIA on release of endogenous glutamate. The R-isomer inhibited release by 35% whereas the Sisomer used stimulated release 76% above basal value. The concentration of each isomer was the same (10 nM). Theophylline completely blocked R-PIA's inhibitory action with no diminuition of its stimulatory effect on glutamate release (Fig. 4). The effects of ADA on release of endogenous glutamate are also shown in Fig. 4. ADA alone enhanced release (39%). Additionally, ADA reduced the inhibitory effect of 25 mM ethanol on theophylline-stimulated release of glutamate to 11%. 9-

a -c

n

7 --

.

: D 2 I P k

5--

\x

4--

6--

LT

2 1 I

o-

EN.1

1I ++

1 3 ;

1

NC1

EElA

R-Pm

rrnaph

FIG. 3. Effect of various compounds on the basal release of

endogenous glutamate in the presence or absence of 25 mM ethanol. Each bar represents the mean +S.E.M. Numbers in parentheses repre~~0.05 and fS;nt the number of separate experiments in triplicate. pt0.01 vs basal release. 'p
Endogenous glutamate release studies 'ndicate a neurotransmitter role for the amino acid in the cerebellum (31). CaPt-dependent, Kt- stimulated endogenous glutamate release in the cerebellum was demonstrated by using slices (4,32), synaptosomes (5,6,30) and cultured cerebellar granule cells (33). The results of the present study show that endogenous glutamfte release from rat cesgbellar synaptosom$! fraction is stimulated by high K and consists of both -independen$+basal release. It was previously demonCa -dependent strated that Ca hvddkzendent and Ca -independent release of glutamate occurs via neuronal and glial cells, respectively (34,35). It is possible that the increased release observed with the inclusion of EGTA in the incubation medium (Fig. 3) represents glutamate release from glial contaminants of the synaptosomal preparation. Support for this notion is taken frgm the work of Levi et al. (34) who observed that Kt-stimulated release of D-[ H]asp$ate from containing or frsshly isolated, glia-enriched fractions was identical in Ca Ca +-free incubation solution whereas that from gliaJ+cultures was substanbut containing 1 mM tially greater in incubation medium containing no Ca

Vol. 44, No. 22, 1989

Glutamate

Release, Adenosine and Ethanol

1631

EGTA. Although the differential results obtained by Levi et al. (34) may be due to the use of different glial preparations, our observed increased glutamate release in the presence of EGTA (Fig. 3) suggests the possibility that EGTA stimulates endogenous glutamate release. This EGTA-stimulated release rnn; i;v;l;;2qlial cells. Corroboratively, basal glutamate release was lowered - and EGTA-free iftubation medium (data not shown). This observation is consistent with a Ca -dependent release by the neuronal constituents of the synaptosomal preparation. On the other hand, it is possible that the observed EGTA-stimulated release does not involve calcium chelation. The marked inhibitory effect of ethanol on EGTA-stimulated glutamate release (Fig. 3) may involve both an a&nof;;;e-mediated component (see below) and a direct action of ethanol on Ca .

*

I-

S

T+R

T+E

T+E+A

Fig. 4. Effect of the two diastereoisomers of PIA and of AUA on endogenous glutamate release. Each bar represents the mean +S.E.M. of three separate experiments performed in triplicate. Basal iB), /U$ I@, R-PIA (R), S-PIA (S), theophylline (T), ethanol (E); ~~0.05 pt0.01 vs basal release using Student's T-test. As noted by the complete inhibition of release at O°C (Fig. l), basal endogenous glutamate release was temperature-dependent. Basal release was rapid during the first 10 set after which time it decreased to a relatively slower rate. About 3.7 nmol/mg protein of glutamate were released within the initial 5 set and Kt-stimulated release was maximal within 30 set (Fig. 2). Nicholls et al. (30) observed similar kinetics for the release of endogenous glutamate from guinea-pig cerebral cortical synaptosomes. They found that about 0.3 nmol/mg protein of L-glutamate were released within 5 set and that rapid, Kt-stimulated release reached the maximum rate within 15 sec. The slight but insignificant inhibitory action of ethanol on Kt-stimulated glutamate release (Fig. 3) could result from ethanol-induced accumulation of adenosine with subsequent adenosine-mediated inhibition of glutamate release (see below).

1632

Glutamate Release, Adenosine and Ethanol

Vol. 44, No. 22, 1989

Ethanol added in vitro resulted in a trend toward dose-dependent inhibition of glutamate release (data not shown). This inhibition may be mediated via the neuromodulator adenosine. Indeed, we observed that 25 mM ethanol increased basal adenosine release (50%) within 5 set (25). Adenosine was implicated in the CNS effects of ethanol (18,19,20) and was shown to inhibit the release of glutamate from slices of rat hippocampus (13,14) and dentate gyrus (15) as well as from cultured cerebellar neurons (16,17). Adenosine receptor agonist R-PIA inhibited basal glutamate release (fig. 3). We found that acute treatment of rats with ethanol (1.5 g/kg, i.p.) caused a significant increase in the maximum number of adenosine AI binding sites in the cerebellar cortex within 15 min of ethanol treatment (21). It is plausible that, since the synaptosomes in the present study were exposed to ethanol only for 5 set, the membranes did not have sufficient exposure time to ethanol to elicit a large increase of AI receptors. Indeed, the magnitude of adenosine's neuromodulatory action was suggested (36) to be regulated by the number of adenosine AI receptors rather than the adenosine concentrations per se. Hence, the effect of ethanol on glutamate release, via adenosine-mediated inhibition, was minimal (i.e., not reaching statistical significance). Likewise, ethanol's potentiation of R-PIA-induced inhibition of glutamate release did not reach statistical significance (Fig. 3). Moreover, adenosine receptor antagonist theophylline markedly increased glutamate release (59%), an effect that was completely antagonized by ethanol (Fig. 3). This strongly suggests that adenosine exerts a tonic inhibitory action on the release of glutamate because theophylline, a competitive antagonist at adenosine receptors, probably increased glutamate release via adenosine receptor blockade. Furthermore,theophylline completely blocked R-PIA-induced inhibition of glutamate release (Fig. 4). Our hypothesis that ethanol's inhibition of glutamate release is mediated by increased adenosine release was substantiated by the ADA antagonism of ethanol's inhibitory action on theophylline-stimulated glutamate release. ADA did not completely block this effect of ethanol on theophylline-stimulated glutamate release. The increased endogenous glutamate release by ADA alone (Fig.4) was relatively less compared to that produced by theophylline most likely because the synaptosomes were incubated with ADA for only 5 sec. This short incubation time was necessary to be consistent with the experimental procedure and to remain in the rapid phase of glutamate release as mentioned in the Results section (Fig. 2). If the synaptosomes were preincubated with ADA to remove endogenous adenosine, then sufficient glutamate release might have occurred to prevent the rapid phase of release due to depletion of the glutamate stores. Nevertheless, the antagonism of ethanol's interaction with the adenosine receptor antagonist theophylline, observed in the presence of ADA, strongly suggested an adenosine component. The differential effects of the two diastereoisomers of PIA on glutamate release was indeed interesting (Fig. 4). R-PIA is known to be approximately 40-fold more potent than S-PIA for binding to adenosine A receptors in brain iastereoisomers is membranes (37). However, stereoselectivity of these two a* non-demonstrable at the adenosine A receptor (38). Additionally, R-PIA has a 20-fold greater potency than S-PIA for potentiating ethanol-induced motor incoordination in rats when injected intracerebroventricularly (Clark and Dar, unpublished observations). In the present study, we found that R-PIA inhibited endogenous glutamate release while same concentration of S-PIA significantly increased the release (Fig. 4). These findings indicate that the adenosine receptor mediating these responses is of the A subtype. Moreover, the AI receptors would be expected to be involved since i hey are highly concentrated in the cerebellum while the A2 receptors are limited mainly to the

Vol. 44, No. 22, 1989

Glutamate Release, Adenosine and Ethanol

1633

striatum and olfactory tubercle (39). Higher concentrations of S-PIA may, similar to R-PIA, inhibit the glutamate release since much greater concentrations of S-PIA are needed to observe similar pharmacological effects as lower concentrations of R-PIA (37; Clark and Dar, unpublished observations). Additionaly, this effect of S-PIA may very well represent the biphasic nature of adenosine. Adenosine has a biphasic pharmacological profile on some behaviors in that low doses actually increase motor activity while higher doses are depressant (40,41). Theophylline-stimulated release was slightly greater when R-PIA was included in the incubation medium (Fig. 4). This finding seems paradoxical at first note. However, the theophylline and R-PIA would be competing for binding at the adenasine receptors. Since a substantially greater number of theophylline molecules are present than that of R-PIA, the percentage of available R-PIA molecules bound at any one point in time would be far less than that in the absence of theophylline. It follows that the biphasic nature of adenosine and its analogues is responsible for this apparent potentiation of theophylline-stimulated glutamate release by R-PIA. The amount of R-PIA bound in the presence of theophylline would be comparable to that of a substantially lower concentration of R-PIA alone in which a stimulatory effect of R-PIA might be expected. This is similar to the observed stimulation elicited by S-PIA discussed above. Again, the data support a biphasic nature of adenosine's pharmacological profile (40,41). Interestingly, the behavioral stimulant property of theophylline could be due to the end result of increased release of glutamate mediated through the antagonism of adenosine binding sites by theophylline. Dolphin and Archer (15) also reported that theophylline enhanced the release of glutamate but not GABA from slices of rat dentate gyrus. These findings suggest that a specific interaction between adenosine and glutamate may occur. Data obtained by Dolphin and Prestwich (16) suggested that adenosine-mediated inhibition of glutamate release occurs by activation of AI receptors. Adenosine receptor agonists depressed the Kt-evoked release while adenosine receptor antagonist, 8-phenyltheophylline alone, stimulated in hippocampal CA1 slices the release of glutamate suggesting that endogenous adenosine was released to exert this tonic modulatory effect on glutamate release (42). Autoradiographic (43) and other (44) evidence indicates that adenosine A binding sites are located on axon terminals of the parallel fibers of granu 1e cells in the molecular layer of the cerebellar cortex. From an autoradiographic study (carried out separately and to be reported elsewhere), we found that a single acute treatment with ethanol (1.5 g/kg, i.p.) markedly increased the density of adenosine AI binding sites in the molecular and granular layers of the rat cerebellar cortex. This observation is in agreement with the results of binding studies done using tissue homogenates of the cerebellar cortex (21). Adenosine selectively blocks parallel-fiber-mediated synaptic potentials in rat cerebellar cortex (45). Corroboratively, Braas et al. (46) found high adenosine immunoreactivity in the Purkinje cells and their dendrites which extend into the molecular layer of the cerebellum and receive input from the parallel fibers of the granule cells. Since the granule cells apparently use glutamate as the excitatory transmitter (3,4,5), the data obtained from the present investigation support our hypothesis that acute ethanol-induced motor incoordination is partially mediated by an adenosine-mediated inhibition of glutamate release from granule cells in the cerebellar cortex of the rat. This adenosinemediated inhibition probably involves the adenosine AI receptors upregulated by acute ethanol (21).

Glutamate Release, Adenosine and Ethanol

1634

Vol. 44, No. 22, 1989

Acknowledoments This work was partially supported by the North Carolina Alcoholism Research Authority Grant No. 8605 and by Sigma Xi, The Scientific Research Society Grant No. 94813. References

:: 3. 4.

7. 8. 9. 10.

11.

14. 15. 16. 17. 18. 19. 20. ;:: E: E: 27. E: 30. ::: 33. 34. 35.

M. ERECINSKA, Biochem. Pharmacol. 36:3547-3555 (1987). J.C. WATKINS and R.H. EVANS, Ann. Rev. Pharmacol. Toxicol. a:165-204

(1981).

A.B. YOUNG, M.L. OSTER-GRANITE, R.M. HERNDON, and S.H. SNYDER, Brain Res 73:1-13 (1974). R.S. FLINT, M.A. REA and W.J. MCBRIDE, J. Neurochem. x:1425-1430

(1981).

M.E. SANDOVAL and C.W. COTMAN, Neurosci. 3:199-206 (1978). J. SANCHEZ-PRIETO, T.S. SIHRA and D.G. NICHOLLS, J. Neurochem. fi:58-64 (1987). A. KINGSBURY, V. GALL0 and R. BALAZS, Neurosci. Lett. Supplement a:S236 (1985). I. PULL and H. McILWAIN, J. Neurochem. 2:695-700 (1985). K. JHAMANDAS and A. DUMBRILLE, Can. J. Physiol. Pharmacol. %:1262-1278 (1980). T.V. DUNWIDDIE, Intl. Rev. Neurobiol. a:63-139 (1985). P.J. MARANGOS and J.P. BOULENGER, Neurosci. Biobehav. Rev. 9:421-430 (1985). S.H. SNYDER, Ann. Rev. Neurosci. 8:103-124 (1985). R. CORRADETTI, G. LO CONTE, F. MORONI, M.B. PASSANI and G. PEPEU, Eur. J Pharmacol . 104: 19-26 (1984). J. FASTBOM AND B.B. FREDHOLM, Acta Physiol. Stand. m:121-123 (1985). A.C. DOLPHIN and E.R. ARCHER, Neurosci. Lett. 43:49-54 (1983). A.C. DOLPHIN and S.A. PRESTWICH, Nature =:148-150 (1985). J. DREJER, A. FRANDSEN, T. HONORE and A. SCHOUSBOE, Neurochem. Intl. u:77-81 (1987). M.S. DAR, S.J. MUSTAFA and W.R. WOOLES, Life Sci. s:1363-1374 (1983). W.R. PROCTOR and T.V. DUNWIDDIE, Science m:519-521 (1984). 8.B. FREDHOLM, N.R. ZAHNISER, G.R. WEINER, W.R. PROCTOR and T.V. DUNWIDDIE, Eur. J. Pharmacol. 111:133-136 (1985). M. CLARK and M.S. DAR, Pharmacol. Biochem. Behav. x:155-161 (1988). M.S. DAR and W.R. WOOLES, Life Sci. 3:1429-1439 (1986). M.S. DAR, J. Pharm. Pharmacol. 40:482-487 (1988). M. CLARK and M.S> DAR, Alcoholism: Clin. Exp. Res. in press (1989). M. CLARK and M.S. DAR, J. Neurochem. in press (1989). I.A. SYTINSKY, B.M. GUZIKOV, M.V. GOMANKO, V.P. EREMIN and N.N. KONOVALOVA, J. Neurochem. a:43-48 (1975). E.G. GRAY and V.P. WHITTAKER, J. Anat. x:79-88 (1962). T.D. WHITE, J. Neurochem. a:1037-1042 (1975). O.H. LOWRY, N.J. ROSEBROUGH, A.L. FARR and R.J. RANDALL, J. Biol. Chem. w:265-275 (1951). D.G. NICHOLLS, T.S. SIHRA and J. SANCHEZ-PRIETO, J. Neurochem. 49:50-57 (1987). L. GIULIO and V.GALLO, Neurochem. Res. u:1627-1642 (1986). J.N. REYNOLDS and W.J. RACZ, Can. J. Physiol. Pharmacol. fi:791-798 (1987). V. GALLO, M.T. CIOTTI, A. COLETTI, F. ALOISI and G. LEVI, Proc. Natl. Acad. Sci USA 79:7919-7923 (1982). G. LEVI, R.D. GORDON, V. GALLO, G.P. WILDIN and R. BALAZS, Brain Res. m:425-445 (1982). F. FONNUM, J. Neurochem. Q:l-11 (1984).

Vol. 44, No. 22, 1989

36. 37.

Glutamate Release, Adenosine and Ethanol

1635

K.S. LEE, M. REDDINGTON, P. SCHUBERT and G. KREUTZBERG, Brain Res.

m:156-159 (1983). J.W. DALY, Purines: Pharmacolow and Phvsiolooical Roles, ed. T.W. Stone, pp. 5-15, MacMillan Press, London (1985). J.A. RIBEIRO and A.M. SEBASTIAO, Prog. Neurobiol. 26:179-209 (1986).

JARVIS and M. WILLIAMS< Pharmacol. Biochem. Behav. 30:707-714 3;:M.F. (1988). 40.

41. 42. 43. t:: 46.

V.L. COFFIN, J.A. TAYLOR, J.W. PHILLIS, H.J. ALTMAN and R.A. BARRACO, Neurosci. Lett. g:91-98 (1984). J.J. KATIMS, Z. ANNAU AND S.H. SNYDER, J. Pharmacol. Exp. Ther. m:167173 (1983). S.P. BURKE AND J.V. NADLER, J. Neurochem. 51:1541-1550, (1988). R.R. GOODMAN, M.J. KUHAR, L. HESTER and S.H. SNYDER, Science m:967-969

(1983). W.J. WOJCIKand N.H. NEFF, J. Neurochem. q1:759-763 (1983).

J.D. KOCSIS, D.L. ENG and R.B. BHISITKUL, Proc. Natl. Acad. Sci USA

a:6531-6534

(1984).

K.M. BRAAS, A.C. NEWBY, V.S. WILSON and S.H. SNYDER, J. Neurosci. &1952-

1961 (1986).