- Email: [email protected]
Effects of baclofen and nitrendipine on ethanol withdrawal responses in the rat
Neuropharmacology Vol. 30, No. 2, pp. 183-190, 1991 Printed in Great Britain. All rights reserved
0028-3908/91 $3.00 + 0.00 Copyright © 1991 Pergamon Press plc
EFFECTS OF BACLOFEN A N D NITRENDIPINE ON ETHANOL WITHDRAWAL RESPONSES IN THE RAT SANDRA E. FILE, A . ZHARKOVSKY a n d K . GULATI Psychopharmacology Research Unit, UMDS Division of Pharmacology, London University, Guy's Hospital, London SE1 9RT, U.K.
(Accepted II October 1990)
Summary--Withdrawal of rats from 5 weeks of a liquid ethanol diet (10%), resulted in anxiogenic responses in the social interaction and elevated plus-maze tests of anxiety. The rats withdrawn from ethanol also showed increased aggression, tremor and rearing. Baclofen (1.25 and 2.5 mg/kg), but not nitrendipine (25-100 mg/kg), reversed the anxiogenic withdrawal responses, without having any effect in control animals and without having significant sedative effects. Baclofen reduced the enhanced aggression during withdrawal of ethanol, but this may have reflected a more general anti-aggressive action. Baclofen (2.5 mg/kg) reduced the withdrawal tremor. Nitrendipine (100 mg/kg) significantly reduced withdrawal tremor, but this dose was sedative, so this was likely to be a non-specific effect. It is proposed that the anxiogenic response during withdrawal of ethanol is due to a reduced GABA function, involving both GABAA and GABAB receptors. Key words---ethanol withdrawal, baclofen, nitrendipine, GABA, anxiety, tremor, aggression.
Chronic exposure of animals or humans to ethanol results in the development of physical dependence, that is manifested by a characteristic withdrawal syndrome upon the cessation of the intake of ethanol. The withdrawal syndrome in rats is characterized by increased anxiety, tremor, muscle rigidity and seizure activity (Victor, 1970; Freund, 1975; File, Baldwin and Hitchcott, 1989). It has been suggested that the development of a withdrawal syndrome, after chronic exposure to ethanol, might be explained by diminished y-aminobutyric acid (GABA-ergic) neurotransmission (Ticku, 1980, 1989). Various parameters of GABAergic neurotransmission are altered during withdrawal of ethanol including an increase in Ko of the low affinity GABA A receptor sites (Ticku, 1980), and reduced concentrations of GABA in various regions of the brain (Sytinsky, Guzikov, Gomanko, Erenien and Konovalova, 1975). Benzodiazepines and barbiturates are used as sedative drugs in the treatment of the ethanol-withdrawal syndrome, where they might compensate for the decreased GABA-ergic inhibition (Volicer, 1980). y-Aminobutyric acid receptors have been classified into GABA A and GABAB subtypes, according to their pharmacological and anatomical properties (Bowery, Hill, Hudson, Double, Middlemiss, Shaw and Turnbull, 1980; Bowery, Hill and Hudson, 1982). Previous studies have shown that the GABAA-receptor agonists, GABA, muscimol and 4,5,6,7-tetrahydro isoxazolo[5,4-c]pyridin-3-ol (THIP), injected intracisternally, reduced the severity of audiogenic seizures but had no effect on tremor in ethanol-withdrawn rats (Frye, McCown and Breese, 1983). Less is known about the effect of the GABAB receptor agonist, baclofen on the
signs of ethanol withdrawal, but in one study it slightly reduced the tremor in ethanol-withdrawn monkeys (Tarika and Winger, 1980) and the GABAB receptor antagonist, phaclofen reversed the anticonvulsant effect of ethanol against picrotoxin-induced seizures (Mehta and Ticku, 1990). Current evidence suggests that voltage-operated calcium channels also might be involved in the development of tolerance during chronic administration of ethanol and the ethanol-withdrawal syndrome (Gandhi and Ross, 1989, for review). Calcium channel antagonists, such as nitrendipine, prevented the development of tolerance to the ataxic and general anaesthetic actions of ethanol, when they were given concurrently with ethanol (Little and Dolin, 1987; Dolin and Little, 1989). Calcium channel antagonists, nitrendipine and nimodipine, abolished convulsions due to withdrawal of ethanol (Little, Dolin and Halsey, 1986) and inhibited apomorphine-induced aggressive reactions, in ethanol-withdrawn rats (Pucilowski and Kostowski, 1988). However, not all behavioural effects of withdrawal of ethanol were affected by administration of nitrendipine; it was without effect on the anxiogenic response, as measured in the social interaction test in ethanol-withdrawn rats (File et al., 1989)i Recent studies showed that baclofen and the dihydropyridines have some common mechanism of action. Nitrendipine, at nanomolar concentrations, inhibits the binding of [3H]baclofen to GABA B receptors (AI-Dahan and Thalman, 1989). Baclofen interacting with GABAB receptors, inhibits the increase in voltage-dependent [Ca2+]i in synaptosomes from the brain of the rat (Stirling, Cross, Robinson 183
184
SANDRAE. FILEet al.
and Green, 1989), which probably accounts for the observed inhibitory effect of baclofen on release of GABA, 5-hydroxytryptamine and noradrenaline (Bowery et al., 1980; Gray and Green, 1987). Because of the evidence that different neurochemical mechanisms might underlie different withdrawal responses, in the present study, the effects of baclofen and nitrendipine were compared on anxiogenic responses, aggression and tremor in ethanol-withdrawn rats.
METHODS
Animals
Male hooded Lister rats (Olac, Bicester, U.K.), with initial weight 180-200g, were housed in a room with a 13 hr light/l 1 hr dark cycle (lights on at 06.00 hr). Animals were housed in groups of five, until 5 days before testing, when they were housed individually. Three days prior to behavioural testing the rats were allocated to test partners on the basis of their weight (_+ 25 g). Members of a test pair always had the same chronic and acute treatment with drug. Drugs
Baclofen (Ciba Geigy) was dissolved in warm distilled water, nitrendipine (Sigma) was suspended in distilled water, with 2-4 drops of Tween-20 and injected intraperitoneally (i.p.) in a volume of 2 ml/kg body weight. Treatment with ethanol
Thirty rats were randomly allocated to each of the ethanol and control diets. All the rats received 20% w/v liquid chocolate Complan diet; water was freely available in a separate bottle. The ethanol group received additions of absolute ethanol to the Complan. This was 3% w/v on day 1 and was increased in 1% steps over the next 7 days. Rats were then maintained on a diet containing 10% ethanol for a further 4 weeks (giving a final daily dose 15.2 + 0.9 g/kg/day). The intake of liquid diet of the control group was restricted to maintain a weight gain comparable to that of the ethanol-treated rats. Apparatus
The social interaction test arena was a wooden box 60 x 60 cm, with 35 cm high walls and was lit by dim light (35 radiometric lux). A camera was mounted vertically above the arena and the rats were observed from a video monitor in the adjacent room. Infrared photocells were mounted in the walls, 4.5 and 12.5 cm from the floor and the interruption of these beams provided automated measures of locomotor activity and rearing, respectively. The output from the photocells and the scores of the observers were entered into a microcomputer. The plus-maze was made of wood and consisted of two opposite open arms, 50 x 10 cm and two opposite
enclosed arms, 50 x 10 x 40 cm with an open roof. The arms were connected by a central square, 10 x 10 cm, thus the maze formed the shape of a plus sign. It was elevated to a height of 50 cm above the floor. The holeboard was a wooden box 60 x 60 x 35 cm with four holes, each 6.5 cm in diameter, equally spaced on the floor. Head dipping was measured by the interruption of infra-red beams from cells located immediately beneath the edges of the holes; locomotor activity and rearing were measured by the interruption of infrared beams from cells located in the walls of box, 4.5 and 12.5cm, respectively, from the floor. Behavioural testing
The Complan diet containing ethanol was replaced by an equivalent amount of Complan without ethanol 7.5hr before testing. All behavioural tests were performed between 0700 and 1200hr and the first test was 30 min after the acute injection of vehicle or baclofen. Initially, animals from both the control and the chronic ethanol diet were randomly assigned to receive intraperitoneal (i.p.) injections of vehicle or baclofen (2.5 or 5mg/kg). However, after testing 3 pairs of rats with 5 mg/kg, it was clear that this dose was too sedative; therefore, the remaining rats originally allocated to this dose were tested with 1.25 mg/kg. The animals were then given behavioural tests in the following order: social interaction, plusmaze and holeboard tests. After testing, the rats were maintained on their previous ethanol or control liquid diets for a further 7 days. They were subsequently withdrawn again from their diets for another 7.5 hr and randomly assigned to receive an intraperitoneal injection of either saline or nitrendipine (25, 50 or 100 mg/kg) 30 min before testing. The animals were retested in the social interaction and plus-maze tests and tremor was again assessed. Social interaction test
The low light, familiar test condition was used; thus, the rats were familiarized with the test arena for 7.5 min on the 2 days prior to testing. The rats were tested in an order randomized for previous chronic treatment and for acute treatment with drug. The behaviour of the rats was videorecorded to permit rescoring. An observer blind to the treatment with drug scored separately two categories of behaviour: social investigation (sniffing, following and grooming the partner) and aggression (wrestling, boxing and aggressive grooming). At the end of each trial the test arena was thoroughly cleaned. Plus-maze test
Each rat was placed in the central square of the plus-maze, facing an enclosed arm. An observer who was blind to the treatment with drug scored the times spent on the open and the enclosed arms of plusmaze. Each test lasted 5 min and at the end of each trial the maze was thoroughly cleaned.
Ethanol withdrawal: baclofen and nitrendipine Control
2oo
185
Ethanol- withdrawal
T
N
I0o
~
so
Veh.
t.25
2.5
Veh.
Baclofen
t.25
2.5
(llg/kg)
bclofen
Fig. 1. Mean ( + S E M ) time spent in social investigation by rats pre-treated with the control diet or withdrawn from the ethanol diet, 30rain after injection with vehicle or baclofen (1.25 or 2.5 mg/kg). I"P <0.05 compared with vehicle-treated control group. *P < 0.05 compared with vehicle-treated ethanol-withdrawal group.
Holeboard test
Statistical analysis
Rats, naive to the holeboard apparatus, were placed singly in the centre o f the holeboard and the following measures were t a k e n d u r i n g a 5 m i n trial: n u m b e r of head dips, time spent head dipping, locomotor activity a n d rearing. A t the end o f each trial, any foecal boluses were r e m o v e d a n d the box was wiped clean.
The data, with the exception of the t r e m o r scores, were analysed with analysis of variance ( A N O V A ) followed, where appropriate, with D u n c a n ' s multiple range tests between individual groups. The t r e m o r scores were analysed using K r u s k a l - W a l l i s H-test.
RESULTS
Ethanol withdrawal-induced tremor This was evaluated using the m e t h o d of Frye, M c C o w n a n d Breese (1983). T r e m o r scores were determined by lifting the rats vertically by the tail a n d scoring the following reactions: a score of three was assigned to rats showing a n immediate forelimb extension a n d violent generalized forelimb-whole body tremor; a score of two was assigned to rats t h a t showed this reaction when they were rotated 180 ° a r o u n d the axis of the tail a n d a score of one to rats that failed to display forelimb extension but showed clearly distinct forelimb tremor, when lifted by the tail a n d rotated 180 ° . The rats were scored on two occasions, once immediately before a n d once after the social interaction test; a m e a n o f these two scores was taken for the statistical analysis.
Experiment 1 Effect of withdrawal of ethanol. W i t h d r a w a l from chronic t r e a t m e n t with ethanol resulted in a significant decrease in the time spent in social investigation [F(I,21) = 28.9, P <0.0001], see Figure 1, without any change in m o t o r activity of rearing [F(1,21) < 1.0 in b o t h cases]. This indicated a n anxiogenic withdrawal response. Withdrawal from ethanol also caused a significant increase in aggression, b o t h in the number of aggressive acts [ F ( 1 , 2 1 ) = 21.6, P < 0.000], see Figure 2 a n d in the time spent in aggression [F(I,21) = 7.3, P = 0.01]. W i t h d r a w a l from ethanol resulted in a n anxiogenic response in the plus-maze, s h o w n by a significant decrease [P < 0.01] in the percentage of time spent in the open arms, see Figure 3,
Table 1. Effect of baclofen on locomotor activity in holeboard of control (CON) and chronic ethanol-withdrawn (EW) rats. The data are means ± SEM Baclofen (mg/kg) Behavioural measure Vehicle 1.25 2.5 Time head-dipping
CON EW
33.6 + 6.9 30.4 +_7.7
27.7 + 5.7 25.4 + 2.8
21.8 _+2.7 17.6 ± 3.7
Number of head dips
CON EW
28.7 + 4.2 27.8 _+5.0
26.2 __.3.0 28.8 + 4.2
23.4 + 4.7 24.5 ± 7.0
Locomotor activity
CON EW
472 + 156 532± 114
465 _+ 173 614+ 186
338 ± 87 418± 141
Rearing
CON EW
15.1 ± 3.1 28.2 + 5.3*
13.1 + 2.2 33.3 + 8.7
7.7 ± 0.12 19.2 ± 4.3
*P <
0.05 compared with vehicle-treated control group.
186
SANDRA
Control
E. FILE et al. Ethanol-withdrawal ÷÷÷
T 8
~o ti
4 3
! 0 Veh.
t.25 2.5 Baclofen
Veh.
1.25
2.5
(q/kg)
Baclofen
Fig. 2. Mean ( _ SEM) number of aggressive acts in control and ethanol-withdrawn rats, treated acutely with vehicle or baclofen (1.25 or 2.5 mg/kg), t t t P < 0.001 compared with vehicle-treated control group. *P < 0.05, **P < 0.01 compared with vehicle-treated ethanol-withdrawal group.
Table 2. Median tremor scores of rats withdrawn from ethanol and treated with vehicleor baclofen (1.25 or 2.5 mg/kg) or nitrendipine(25, 50 or 100 mg/kg) Median Range Experiment 1 Vehicle Baclofen 1.25 mg/kg 2.5 mg/kg
2.5
1-3
2.0 1.0"
1-2 0-2
without any change in total arm entries. The only measure to change in the holeboard, on withdrawal from ethanol, was rearing, which was significantly increased [ F ( 1 , 4 8 ) = 15.3, P <0.001], see Table I. Withdrawal from ethanol induced a significant [P < 0.001] tremor, see Table 2. Effect of baclofen. Baclofen was without effect in the control rats, but significantly reversed the decrease in social interaction shown by the ethanol-withdrawn animals (pretreatment x baclofen interaction, F(2,21) = 6.5, P < 0.01). As can be seen from Figure l, the reversal was due to the smaller dose o f baclofen. There was no effect o f baclofen on l o c o m o t o r activity or the n u m b e r o f rears made by pairs o f rats in the social interaction test [ F ( 2 . 2 1 ) < 1.0 in both cases].
Experiment 2 Vehicle 3.0 1-3 Nitrendipine 25 mg/kg 2.0 23 50 mg/kg 2.0 2-3 100 mg/kg 1.0" 0-1 *Significantly different from vehicle group, see text for details.
Ethanol..tthdrnal
Control 50 1
E 40 NN
_=
l
3O
2o m
,0
0 Yeh.
1.25
2.5
Baclofen
Veh.
i.25
2.5
(Jg/kg)
Baclofen
Fig. 3. Mean (+SEM) percentage time spent on the open arms of the plus-maze by rats pre-treated with the control diet or withdrawn from the ethanol diet, 30 min after injection with vehicle or baclofen (1.25 or 2.5mg/kg). "M'P<0.01 compared with vehicle-treated control group. **P <0.01 compared with vehicle-treated ethanol-withdrawal group.
187
Ethanol withdrawal: baclofen and nitrendipine Table 3. Mean (+ SEM) time spent in social interaction (sec), number of rears made in the social interaction test and percentage of time spent on the open arms of the plus-maze, by control (CON) and ethanol-withdrawn (EW) rats treated with vehicle or nitrendipine Nitrendipine (mg/kg) Vehicle 25 50 100 CON 100.0 _ 9.3 85.7 ± 22.8 78.5 ± 28.1 37.0 _ 13.6 Social EW 47.4** _ 12.8 31.7 ± 9.1 66.5 ± 9.5 31.0 +_5.0 CON 73.8 ± 3.5 62.3 ± 26.9 59.3 ± 19.6 35.0* + 16.2 Rears EW 82.2 ± 6.6 52.0 ± 15.7 76.5 ± 9.6 24.0~t _+8.3 Percentage time
CON EW
21.2 + 8.4 8.3 ± 2.0
17.9 ± 7.1 2.7 ± 1.3
11.5 ± 3.5 4.6 _ 3.2
2.4* + 1.5 3.1 + 3.1
*P < 0.05, **P < 0.001, significantlydifferent from vehicle-treatedvehicle-treatedcontrol group. t~'P < 0.01, significantlydifferent from vehicle-treatedwithdrawal group.
Baclofen dose-dependently reversed the number of aggressive acts shown by the ethanol-withdrawn rats [F(2,21) = 11.3, P < 0.001], see Figure 2 and the time spent in aggressive behaviour was significantly [P < 0.05] reduced by the larger dose. There was also a trend for baclofen to reduce the time spent in aggression [F(2,21)= 3.0]. Baclofen dose-dependently [P < 0.01] reversed the decrease in the percentage of time spent on the open arms of the plus-maze, shown by the ethanolwithdrawn animals, but was without significant effect in the control groups, see Figure 3. Baclofen was without significant effect on the number of rears made in the holeboard [F(2.48)= 2.5] and did not reverse the increase in rearing shown by the ethanol-withdrawn group, see Table 1. However, baclofen reduced the ethanol-withdrawal tremor [Kruskal-Wallis H - - 1 4 , P < 0.0001], but only the largest dose had a significant effect, see Table 2.
Experiment 2 Once again, withdrawal from ethanol resulted in significant anxiogenic effects, indicated by decreased social interaction [ F ( 1 , 2 2 ) = 7 . 4 , P <0.01] and a decreased percentage of time spent on the open arms [F(I,52) = 5.6, P < 0.05], see Table 3. These decreases were unaccompanied by significant decreases in locom o t o r activity or rearing in the social interaction test or by a change in the total arm entries in the plusmaze. There was also a significant increase in tremor on withdrawal of ethanol [H = 12.7, P < 0.001], see Table 2. Treatment with nitrendipine did not reverse the decreased social interaction [F(3,22)= 1.3] or the decrease in the percentage of time spent on the open arms, see Table 3. There was a trend towards a reduction of social interaction, caused by nitrendipine [F(3,22) = 2.6] and a trend towards reduced locomotion [F(3,22) = 2.4] but these did not reach overall significance. However, nitrendipine did significantly reduce rearing [F(3,22) = 4.7, P = 0.01]; this was due to the largest dose, see Table 3. There was also a trend towards a reduction by nitrendipine of the percentage of time spent on the open arms but this did not reach significance [F(3,52) = 1.9]. Nitrendipine significantly reversed the tremor, induced by withdrawal of alcohol
[H = 11, P < 0.001] but the reduction was significant only for the largest dose, see Table 2.
DISCUSSION It has previously been reported (File et al., 1989) that withdrawal from chronic treatment with ethanol produced an anxiogenic response in the social interaction test. In the present two experiments, this anxiogenie withdrawal response was detected in both the social interaction and the plus-maze tests. Small doses of baclofen reversed these anxiogenic responses, although baclofen was without anxiolytic effect in control rats. Baclofen also reversed the ethanolwithdrawal tremor, at a dose that was not generally sedative. In rhesus monkeys, baclofen (3 and 5.6 mg/ kg) produced only slight reductions in tremor and these doses were also sedative (Tarika and Winger, 1980). It therefore seems that, in some species, it might be difficult to separate the anti-tremor and sedative effects of baclofen. Baclofen reduced the increased aggression shown by the ethanol-withdrawn rats. However, there was a trend towards a reduction of the time spent in aggression in control animals, even though these levels were very low. This raises the possibility that baclofen might have an anti-aggressive action, that is not just limited to the conditions of withdrawal of ethanol. Baclofen failed to reduce the increased rearing in the holeboard, shown by the ethanol-withdrawn group, which suggests that this withdrawal response is mediated by a different mechanism from that underlying increased anxiety. It has long been suggested that different biochemical changes may mediate different withdrawal responses to ethanol and there is some evidence that noradrenergic pathways mediate the hyperactivity during withdrawal of ethanol (Trzaskowski, Pucilowski, Dyr, Kostowski and Hauptmann, 1986); there is also evidence that noradrenergic pathways mediate the hyperactivity during withdrawal of benzodiazepines (Kulchandy and Kulkarni, 1986). In contrast to the results with baclofen, the dihydropyridine calcium channel antagonist, nitrendipine, failed to reverse the anxiogenic withdrawal response. The social interaction scores in Experiment 2 were
188
SANDRAE. FILE et al.
smaller than in Experiment 1 but, in both cases, the decrease with withdrawal of ethanol was about 50%. It is unlikely that the failure to find reversal with nitrendipine was due to the differences in scores or to the fact that it was the second time the animals had been withdrawn from ethanol. In a previous study (File et al., 1989) it was also found that nitrendipine failed to reverse the anxiogenic withdrawal response to ethanol and in that study, the animals were tested after their first withdrawal from ethanol and their scores were very close to those found in Experiment 1. Thus, it seems that, in contrast to the ability of calcium channel antagonists to protect against tremor and seizures after withdrawal of ethanol (Little et al., 1986), nitrendipine was unable to reverse anxiogenic withdrawal responses. This argues for different mechanisms mediating these two withdrawal responses. Nitrendipine (100mg/kg) did reduce withdrawal tremor, but this dose was clearly sedative and therefore this is likely to be a non-specific effect. Baclofen seemed to have a more specific effect against ethanolwithdrawal tremor, at least in the rats used here and this finding is similar to that of Bone, Majchrowicz, Martin, Linnoila and Nutt (1989), showing diazepam to be superior to calcium channel antagonists. The reversals with diazepam and baclofen implicate both GABAA and GABA B receptors in the ethanolwithdrawal tremor. Since both the G A B A a agonist, baclofen (Experiment l) and the benzodiazepine antagonist, flumazenil (File et al., 1989) reversed the increased anxiety during withdrawal of ethanol, it seems that this response was the result of a decreased GABA function. This has also been suggested by Idemudia, Bhadra and Lal (1989), because both bicuculline and pentylenetetrazole enhanced the anxiogenic response in withdrawal of ethanol. During withdrawal of ethanol Ticku (1980) found a decreased affinity of the low affinity G A B A A receptors in the brain of the rat. However, this was not confirmed by Ulrichsen, Clemmesen, Barry and Hemmingsen (1988) and nor is there any change in binding at benzodiazepine or picrotoxinin sites (Rastogi, Thyagarajah, Clothier and Ticku, 1986; Tamborska and Marangos, 1986; Ulrichsen et aL, 1988). Ethanol stimulates the uptake of CI- into synaptoneurosomes (Suzdak, Schwartz, Skolnick and Paul, 1986) and cultured spinal cord neurones (Mehta and Ticku, 1988) and the increased chloride flux is blocked by picrotoxin (Mehta and Ticku, 1988; McQuilkin and Harris, 1990), but not by baclofen (Mehta and Ticku, 1990); however, the effects on chloride flux, during withdrawal of ethanol have not been reported. The reversal by flumazenil of the anxiogenic withdrawal response, suggests that during withdrawal of ethanol there is a change in the coupling of the GABAA-benzodiazepine-chloride channel receptor complex. Flumazenil was also able to reverse the anxiogenic response during withdrawal of benzodiazepines (Baldwin and File, 1988, 1989) and this can be explained either by antagonism of an endogen-
ous inverse agonist ligand or by a resetting of the benzodiazepine receptors to a neutral state (File and Hitchcock, 1990). It seems that flumazenil may also be able to reset to GABAA-benzodiazepine receptor complex during withdrawal of ethanol. A general change in GABA function has been proposed to underlie the symptoms of withdrawal of ethanol and benzodiazepine (Cowen and Nutt, 1982) and, although the present data on anxiety responses would support this hypothesis, it is unlikely to apply to all withdrawal responses. In particular, it is important to distinguish between the increased anxiety and the increased seizures seen during withdrawal of ethanol. Whereas baclofen and flumazenil can reverse the former they do not reverse the latter (Tarika and Winger, 1980; Adinoff, Majchrowicz, Martin and Linnoila, 1986; Mehta and Ticku, 1989). Indeed, the benzodiazepine inverse agonist Ro 15-4503 (ethyl 8-azido-5,6-dihydro-5-methyl-6-oxo-4H-imidazol[1,5- a][1,4] benzodiazepine- 3- carboxylate), becomes proconvulsant during withdrawal of ethanol (Lister and Karanian, 1987; Mehta and Ticku, 1989). Recent evidence implicates the N M D A (N-methyI-D-aspartate) receptor complex in ethanol-withdrawal seizures: they can be antagonised by the N M D A receptor antagonist MK 801, (+)-5-methyl- 10,1 l-dihydro-5Hdibeazo[a,d] cyclohepten-5,10-imine (Grant, Valverius, Hudsmith and Tabakoff, 1990; Morrisett, Rezvani, Overstreet, Janowsky, Wilson and Swartzwelder, 1990) and enhanced by N M D A (Grant et al., 1990). At the doses of baclofen that were used in this study, there is no evidence for its acting at GABAA receptors and therefore the results suggest that during withdrawal of ethanol, GABAB receptors may also be involved in the anxiogenic response. Whilst baclofen is not anxiolytic in control animals there is evidence that the combination of muscimol and baclofen is anxiolytic (Gray, 1983). Since by its action at presynaptic GABA receptors, baclofen would reduce the release of GABA, its reversal of the anxiogenic response must be either by a presynaptic action at heteroreceptors (e.g. Gray and Green, 1987) or by an action at post-synaptic GABA B receptors. Acute administration of ethanol has been reported to increase recurrent inhibition in the dentate gyrus of the hippocampus (Weisner, Henriksen and Bloom, 1987) and a post-synaptic GABA B site has been identified in the hippocampus (Dutar and Nicoll, 1988). In this region release of GABA from interneurones interacts with both GABA A and GABA Breceptors, to generate a fast and slow inhibitory postsynaptic potentials (i.p.s.p.), respectively and larger concentrations were needed to activate the GABA B receptors, which would explain why this system might not be active under normal circumstances. Whilst it can be concluded that baclofen reversed the anxiogenic responses, measured in the social interaction and plus-maze tests during withdrawal of ethanol, further experiments are needed to firmly establish that these effects of baclofen are indeed due
Ethanol withdrawal: baclofen and nitrendipine tO a n action of G A B A B receptors. It is also i m p o r t a n t to determine w h e t h e r it is post-synaptic G A B A B receptors t h a t decrease in n u m b e r or affinity d u r i n g withdrawal of e t h a n o l or w h e t h e r baclofen was exerting its effects t h r o u g h a n action at heteroreceptors. Acknowledgement--This study was supported by a grant from the Wellcome Trust. REFERENCES
Adinoff B., Majchrowicz E., Martin P. R. and Linnoila M. 0986) The benzodiazepine antagonist Ro 15-1788 does not antagonize the ethanol withdrawal syndrome. Biol. Psychiat. 21: 643-649. AI-Dahan M. I. and Thalman R. H. (1989) Effects of dihydropyridine calcium channel ligands on rat brain ~,aminobutyric acidB receptors. J. Neurochem. 53: 982-985. Baldwin H. A. and File S. E. (1988) Reversal of increased anxiety during benzodiazepine withdrawal: evidence for an anxiogenic endogenous ligand for the benzodiazepine receptor. Brain Res. Bull. 20: 603-606. Baldwin H. A. and File S. E. (1989) Flumazenil prevents the development of chlordiazepoxide withdrawal in rats tested in the social interaction test of anxiety. Psychopharmacology 97: 424-426. Bone G. H. A., Majchrowicz E., Martin P. R., Linnoila H. and Nutt D. J. (1989) A comparison of calcium antagonists and diazepam in reducing ethanol withdrawal tremors. Psychopharmacology 99: 386-388. Bowery N. G., Hill D. R., Hudson A. L., Doble A., Middlemiss D. N., Shaw J. and Turnbull M. (1980) (-)-Baclofen decreases neurotransmitter release in the mammalian CNS by an action at a novel GABA receptor. Nature 283: 92-94. Bowery N. G., Hill D. R. and Hudson A. L. (1982) Characteristics of GABA Breceptor binding sites on whole brain synaptic membranes. Br. J. Pharmac. 78: 191-206. Cowen P. J. and Nutt D. J. (1982) Abstinence symptoms after withdrawal of tranquillising drugs: is there a common neurochemical mechanism? Lancet 360-362. Dolin S. J. and Little H. J. (1989) Are changes in neuronal calcium channels involved in ethanol tolerance? J. Pharmac. exp. Ther. 25t1: 985-991. Dutar P. and Nicoll R. A. (1988) A physiological role for GABA A receptors in the central nervous system. Nature 332: 156-158. File S. E., Baldwin H. A. and Hitchcott P. K. (1989) Flumazenil but not nitrendipine reverses the increased anxiety during ethanol withdrawal in the rat. Psychopharmacology 98:262 264. File S. E. and Hitchcott P. K. (1990) A theory of benzodiazepine dependence that can explain whether flumazenil will enhance or reverse the phenomena. Psychopharmacology lOt: 525 532. Freund G. (1975) Induction of physical dependence on alcohol in rodents. In: Biochemial Pharmacology of Ethanol (Majchrowicz E., Ed.), pp. 211-235. Plenum Press, New York. Frye G. D., McCown T. J. and Breese G. R. (1983) Differential sensitivity of ethanol withdrawal signs in the rat to aminobutyric acid (GABA) mimetics: blockade of audiogenic seizures but not forelimb tremors. J. Pharmac. exp. Ther. 226:720 725. Gandhi C. R. and Ross D. H. (1989) Influence of ethanol on calcium, inositol phospholipids and intracellular signalling mechanisms. Experientia 45: 407-412. Grant K. A., Valverius P., Hudspith M. and Tabakoff B. (1990) Ethanol withdrawal seizures and the NMDA receptor complex. Eur. J. Pharmac. 176: 289-296. Gray J. A. (1983) Gamma-aminobutyrate, the benzodiazepines and the septohippocampal system. In: Benzo-
189
diazepines Divided (Trimble M., Ed.), pp. 101-128. Wiley, New York. Gray J. A. and Green A. R. 0987) GABA e receptor mediated inhibition of potassium evoked release of endogenous 5-hydroxytryptamine from mouse frontal cortex. Br. J. Pharmac. 91" 517-522. Idemudia S. O., Bhadra S. and Lal H. 0989) The pentylenetetrazol-like interoceptive stimulus produced by ethanol withdrawal is potentiated by bicuculline and picrotoxinin. Neuropsychopharmacology 2:115-122. Kunchandy J. and Kulkarni S. K. (1986) Reversal by alpha 2 agonists of diazepam hyperactivity in rats. Psychopharmacology 90: 198-202. Lister R. G. and Karanian I. W. (1987) Ro 15-4513 induces seizues in DBA/2 mice undergoing alcohol withdrawal. Alcohol 4:409-411. Little H. J., Dolin S. J. and Halsey M. J. (1986) Calcium channel antagonists decrease the ethanol withdrawal syndrome. Life Sci. 39: 2054-2065. Little H. J. and Dolin S. J. (1987) Lack of tolerance to ethanol after concurrent administration of nitrendipine. Br. J. Pharmac. 92: 606P. McQuilkin S. J. and Harris R. A. (1990) Factors affecting actions of ethanol on GABA-activated chloride channels. Life Sci. 46: 527-541. Mehta A. K. and Ticku M. K. 0988) Ethanol potentiation of GABAergic transmission in cultured spinal cord neurons involves 7-aminobutyric acidA-gated chloride channels. J. Pharmac. exp. Ther. 246: 558-564. Mehta A. K. and Ticku M. K. (1989) Chronic ethanol treatment alters the behavioral effects of Ro 15-4513, a partially negative ligand for benzodiazepine binding sites. Brain Res. 489: 93-100. Mehta A. K. and Ticku M. K. (1990) Are GABAB receptors involved in the pharmacological effects of ethanol? Eur. J. Pharmac. 182: 473-480. Morrisett R. A., Rezvani A. H., Overstreet D., Janowsky D. S., Wilson W. A. and Swartzwelder H. S. (1990) MK-801 potently inhibits alcohol withdrawal seizures in rats. Eur. J. Pharmac. 176: 103-105. Pucilowski O. and Kostowski W. (1988) Diltiazem suppresses apomorphine induced fighting and proaggressive effect of withdrawal from chronic ethanol or haloperidol in rats. Neurosci. Lett. 93: 96-100. Rastogi S. K., Thyagarajan R., Clothier J. and Ticku M. K. (1986) Effect of chronic treatment of ethanol on benzodiazepine and picrotoxin sites on the GABA receptor complex in regions of the brain of the rat. Neuropharmacology 25:1179-1184. Stirling J. M., Cross A. J., Robinson T. N. and Green A. R. (1989) The effect of GABA a receptor agonists and antagonists on potassium stimulated [Ca:+]i in rat brain synaptosomes. Neuropharmacology 28: 699-704. Suzdak P. D., Schwartz R. D., Skolnick P. and Paul S. M. (1986) Ethanol stimulates 7-aminobutyric acid receptormediated chloride transport in rat brain synaptoneurosomes. Neurobiology 83: 4071-4075. Sytinski I. A., Guzikov B. M., Gomanko M. V., Erenian V. P. and Konovalova N. N. (1975) The gamma-aminobutyric acid (GABA) system in brain during acute and chronic ethanol intoxication. J. Neurochem. 25: 43-48. Tamborska E. and Marangos P. J. (1986) Brain benzodiazepine binding sites in ethanol dependent and withdrawal states. Life Sci. 38: 465-472. Tarika J. S. and Winger G. (1980) The effect of ethanol, pentobarbital and baclofen on ethanol withdrawal in the rhesus monkeys. Psychopharmacology 70: 201-208. Ticku M. K. (1980) The effects of acute and.chronic ethanol administration and its withdrawal on 7-aminobutyric acid receptor binding in rat brain. Br. J. Pharmac. 70: 403-410. Ticku M. K. (1989) Ethanol and the benzodiazepine-GABA receptor-ionophore complex. Experientia 45: 413-417.
190
SANDRAE. FILE et al.
Trzaskowska E., Pucilowski O., Dyr W., Kostowski W. and Hauptmann M. (1986) Suppression of ethanol tolerance and dependence in rats treated with DSP-4, a noradrenergic neurotoxin. Drug Alcohol Dependence 18: 349-353. Ulrichsen J., Clemmesen L., Barry D. and Hemmingsen R. (1988) The GABA/benzodiazepine receptor chloride channel complex during repeated episodes of physical ethanol dependence in the rat. Psychopharmacology 96: 227-231.
Victor M. (1970) The alcohol withdrawal syndrome: theory and practice. Postgrad. Med. 47: 68-72. Volicer L. (1980) GABA levels and receptor binding after acute and chronic ethanol administration. Brain Res. Bull. 5: Suppl. 2, 809-813. Wiesner J. B., Henriksen S. J. and Bloom F. E. (1987) Ethanol enhances recurrent inhibition in the dentate gyrus of the hippocampus. Neurosci. Lett. 79: 169-173.
0028-3908/91 $3.00 + 0.00 Copyright © 1991 Pergamon Press plc
EFFECTS OF BACLOFEN A N D NITRENDIPINE ON ETHANOL WITHDRAWAL RESPONSES IN THE RAT SANDRA E. FILE, A . ZHARKOVSKY a n d K . GULATI Psychopharmacology Research Unit, UMDS Division of Pharmacology, London University, Guy's Hospital, London SE1 9RT, U.K.
(Accepted II October 1990)
Summary--Withdrawal of rats from 5 weeks of a liquid ethanol diet (10%), resulted in anxiogenic responses in the social interaction and elevated plus-maze tests of anxiety. The rats withdrawn from ethanol also showed increased aggression, tremor and rearing. Baclofen (1.25 and 2.5 mg/kg), but not nitrendipine (25-100 mg/kg), reversed the anxiogenic withdrawal responses, without having any effect in control animals and without having significant sedative effects. Baclofen reduced the enhanced aggression during withdrawal of ethanol, but this may have reflected a more general anti-aggressive action. Baclofen (2.5 mg/kg) reduced the withdrawal tremor. Nitrendipine (100 mg/kg) significantly reduced withdrawal tremor, but this dose was sedative, so this was likely to be a non-specific effect. It is proposed that the anxiogenic response during withdrawal of ethanol is due to a reduced GABA function, involving both GABAA and GABAB receptors. Key words---ethanol withdrawal, baclofen, nitrendipine, GABA, anxiety, tremor, aggression.
Chronic exposure of animals or humans to ethanol results in the development of physical dependence, that is manifested by a characteristic withdrawal syndrome upon the cessation of the intake of ethanol. The withdrawal syndrome in rats is characterized by increased anxiety, tremor, muscle rigidity and seizure activity (Victor, 1970; Freund, 1975; File, Baldwin and Hitchcott, 1989). It has been suggested that the development of a withdrawal syndrome, after chronic exposure to ethanol, might be explained by diminished y-aminobutyric acid (GABA-ergic) neurotransmission (Ticku, 1980, 1989). Various parameters of GABAergic neurotransmission are altered during withdrawal of ethanol including an increase in Ko of the low affinity GABA A receptor sites (Ticku, 1980), and reduced concentrations of GABA in various regions of the brain (Sytinsky, Guzikov, Gomanko, Erenien and Konovalova, 1975). Benzodiazepines and barbiturates are used as sedative drugs in the treatment of the ethanol-withdrawal syndrome, where they might compensate for the decreased GABA-ergic inhibition (Volicer, 1980). y-Aminobutyric acid receptors have been classified into GABA A and GABAB subtypes, according to their pharmacological and anatomical properties (Bowery, Hill, Hudson, Double, Middlemiss, Shaw and Turnbull, 1980; Bowery, Hill and Hudson, 1982). Previous studies have shown that the GABAA-receptor agonists, GABA, muscimol and 4,5,6,7-tetrahydro isoxazolo[5,4-c]pyridin-3-ol (THIP), injected intracisternally, reduced the severity of audiogenic seizures but had no effect on tremor in ethanol-withdrawn rats (Frye, McCown and Breese, 1983). Less is known about the effect of the GABAB receptor agonist, baclofen on the
signs of ethanol withdrawal, but in one study it slightly reduced the tremor in ethanol-withdrawn monkeys (Tarika and Winger, 1980) and the GABAB receptor antagonist, phaclofen reversed the anticonvulsant effect of ethanol against picrotoxin-induced seizures (Mehta and Ticku, 1990). Current evidence suggests that voltage-operated calcium channels also might be involved in the development of tolerance during chronic administration of ethanol and the ethanol-withdrawal syndrome (Gandhi and Ross, 1989, for review). Calcium channel antagonists, such as nitrendipine, prevented the development of tolerance to the ataxic and general anaesthetic actions of ethanol, when they were given concurrently with ethanol (Little and Dolin, 1987; Dolin and Little, 1989). Calcium channel antagonists, nitrendipine and nimodipine, abolished convulsions due to withdrawal of ethanol (Little, Dolin and Halsey, 1986) and inhibited apomorphine-induced aggressive reactions, in ethanol-withdrawn rats (Pucilowski and Kostowski, 1988). However, not all behavioural effects of withdrawal of ethanol were affected by administration of nitrendipine; it was without effect on the anxiogenic response, as measured in the social interaction test in ethanol-withdrawn rats (File et al., 1989)i Recent studies showed that baclofen and the dihydropyridines have some common mechanism of action. Nitrendipine, at nanomolar concentrations, inhibits the binding of [3H]baclofen to GABA B receptors (AI-Dahan and Thalman, 1989). Baclofen interacting with GABAB receptors, inhibits the increase in voltage-dependent [Ca2+]i in synaptosomes from the brain of the rat (Stirling, Cross, Robinson 183
184
SANDRAE. FILEet al.
and Green, 1989), which probably accounts for the observed inhibitory effect of baclofen on release of GABA, 5-hydroxytryptamine and noradrenaline (Bowery et al., 1980; Gray and Green, 1987). Because of the evidence that different neurochemical mechanisms might underlie different withdrawal responses, in the present study, the effects of baclofen and nitrendipine were compared on anxiogenic responses, aggression and tremor in ethanol-withdrawn rats.
METHODS
Animals
Male hooded Lister rats (Olac, Bicester, U.K.), with initial weight 180-200g, were housed in a room with a 13 hr light/l 1 hr dark cycle (lights on at 06.00 hr). Animals were housed in groups of five, until 5 days before testing, when they were housed individually. Three days prior to behavioural testing the rats were allocated to test partners on the basis of their weight (_+ 25 g). Members of a test pair always had the same chronic and acute treatment with drug. Drugs
Baclofen (Ciba Geigy) was dissolved in warm distilled water, nitrendipine (Sigma) was suspended in distilled water, with 2-4 drops of Tween-20 and injected intraperitoneally (i.p.) in a volume of 2 ml/kg body weight. Treatment with ethanol
Thirty rats were randomly allocated to each of the ethanol and control diets. All the rats received 20% w/v liquid chocolate Complan diet; water was freely available in a separate bottle. The ethanol group received additions of absolute ethanol to the Complan. This was 3% w/v on day 1 and was increased in 1% steps over the next 7 days. Rats were then maintained on a diet containing 10% ethanol for a further 4 weeks (giving a final daily dose 15.2 + 0.9 g/kg/day). The intake of liquid diet of the control group was restricted to maintain a weight gain comparable to that of the ethanol-treated rats. Apparatus
The social interaction test arena was a wooden box 60 x 60 cm, with 35 cm high walls and was lit by dim light (35 radiometric lux). A camera was mounted vertically above the arena and the rats were observed from a video monitor in the adjacent room. Infrared photocells were mounted in the walls, 4.5 and 12.5 cm from the floor and the interruption of these beams provided automated measures of locomotor activity and rearing, respectively. The output from the photocells and the scores of the observers were entered into a microcomputer. The plus-maze was made of wood and consisted of two opposite open arms, 50 x 10 cm and two opposite
enclosed arms, 50 x 10 x 40 cm with an open roof. The arms were connected by a central square, 10 x 10 cm, thus the maze formed the shape of a plus sign. It was elevated to a height of 50 cm above the floor. The holeboard was a wooden box 60 x 60 x 35 cm with four holes, each 6.5 cm in diameter, equally spaced on the floor. Head dipping was measured by the interruption of infra-red beams from cells located immediately beneath the edges of the holes; locomotor activity and rearing were measured by the interruption of infrared beams from cells located in the walls of box, 4.5 and 12.5cm, respectively, from the floor. Behavioural testing
The Complan diet containing ethanol was replaced by an equivalent amount of Complan without ethanol 7.5hr before testing. All behavioural tests were performed between 0700 and 1200hr and the first test was 30 min after the acute injection of vehicle or baclofen. Initially, animals from both the control and the chronic ethanol diet were randomly assigned to receive intraperitoneal (i.p.) injections of vehicle or baclofen (2.5 or 5mg/kg). However, after testing 3 pairs of rats with 5 mg/kg, it was clear that this dose was too sedative; therefore, the remaining rats originally allocated to this dose were tested with 1.25 mg/kg. The animals were then given behavioural tests in the following order: social interaction, plusmaze and holeboard tests. After testing, the rats were maintained on their previous ethanol or control liquid diets for a further 7 days. They were subsequently withdrawn again from their diets for another 7.5 hr and randomly assigned to receive an intraperitoneal injection of either saline or nitrendipine (25, 50 or 100 mg/kg) 30 min before testing. The animals were retested in the social interaction and plus-maze tests and tremor was again assessed. Social interaction test
The low light, familiar test condition was used; thus, the rats were familiarized with the test arena for 7.5 min on the 2 days prior to testing. The rats were tested in an order randomized for previous chronic treatment and for acute treatment with drug. The behaviour of the rats was videorecorded to permit rescoring. An observer blind to the treatment with drug scored separately two categories of behaviour: social investigation (sniffing, following and grooming the partner) and aggression (wrestling, boxing and aggressive grooming). At the end of each trial the test arena was thoroughly cleaned. Plus-maze test
Each rat was placed in the central square of the plus-maze, facing an enclosed arm. An observer who was blind to the treatment with drug scored the times spent on the open and the enclosed arms of plusmaze. Each test lasted 5 min and at the end of each trial the maze was thoroughly cleaned.
Ethanol withdrawal: baclofen and nitrendipine Control
2oo
185
Ethanol- withdrawal
T
N
I0o
~
so
Veh.
t.25
2.5
Veh.
Baclofen
t.25
2.5
(llg/kg)
bclofen
Fig. 1. Mean ( + S E M ) time spent in social investigation by rats pre-treated with the control diet or withdrawn from the ethanol diet, 30rain after injection with vehicle or baclofen (1.25 or 2.5 mg/kg). I"P <0.05 compared with vehicle-treated control group. *P < 0.05 compared with vehicle-treated ethanol-withdrawal group.
Holeboard test
Statistical analysis
Rats, naive to the holeboard apparatus, were placed singly in the centre o f the holeboard and the following measures were t a k e n d u r i n g a 5 m i n trial: n u m b e r of head dips, time spent head dipping, locomotor activity a n d rearing. A t the end o f each trial, any foecal boluses were r e m o v e d a n d the box was wiped clean.
The data, with the exception of the t r e m o r scores, were analysed with analysis of variance ( A N O V A ) followed, where appropriate, with D u n c a n ' s multiple range tests between individual groups. The t r e m o r scores were analysed using K r u s k a l - W a l l i s H-test.
RESULTS
Ethanol withdrawal-induced tremor This was evaluated using the m e t h o d of Frye, M c C o w n a n d Breese (1983). T r e m o r scores were determined by lifting the rats vertically by the tail a n d scoring the following reactions: a score of three was assigned to rats showing a n immediate forelimb extension a n d violent generalized forelimb-whole body tremor; a score of two was assigned to rats t h a t showed this reaction when they were rotated 180 ° a r o u n d the axis of the tail a n d a score of one to rats that failed to display forelimb extension but showed clearly distinct forelimb tremor, when lifted by the tail a n d rotated 180 ° . The rats were scored on two occasions, once immediately before a n d once after the social interaction test; a m e a n o f these two scores was taken for the statistical analysis.
Experiment 1 Effect of withdrawal of ethanol. W i t h d r a w a l from chronic t r e a t m e n t with ethanol resulted in a significant decrease in the time spent in social investigation [F(I,21) = 28.9, P <0.0001], see Figure 1, without any change in m o t o r activity of rearing [F(1,21) < 1.0 in b o t h cases]. This indicated a n anxiogenic withdrawal response. Withdrawal from ethanol also caused a significant increase in aggression, b o t h in the number of aggressive acts [ F ( 1 , 2 1 ) = 21.6, P < 0.000], see Figure 2 a n d in the time spent in aggression [F(I,21) = 7.3, P = 0.01]. W i t h d r a w a l from ethanol resulted in a n anxiogenic response in the plus-maze, s h o w n by a significant decrease [P < 0.01] in the percentage of time spent in the open arms, see Figure 3,
Table 1. Effect of baclofen on locomotor activity in holeboard of control (CON) and chronic ethanol-withdrawn (EW) rats. The data are means ± SEM Baclofen (mg/kg) Behavioural measure Vehicle 1.25 2.5 Time head-dipping
CON EW
33.6 + 6.9 30.4 +_7.7
27.7 + 5.7 25.4 + 2.8
21.8 _+2.7 17.6 ± 3.7
Number of head dips
CON EW
28.7 + 4.2 27.8 _+5.0
26.2 __.3.0 28.8 + 4.2
23.4 + 4.7 24.5 ± 7.0
Locomotor activity
CON EW
472 + 156 532± 114
465 _+ 173 614+ 186
338 ± 87 418± 141
Rearing
CON EW
15.1 ± 3.1 28.2 + 5.3*
13.1 + 2.2 33.3 + 8.7
7.7 ± 0.12 19.2 ± 4.3
*P <
0.05 compared with vehicle-treated control group.
186
SANDRA
Control
E. FILE et al. Ethanol-withdrawal ÷÷÷
T 8
~o ti
4 3
! 0 Veh.
t.25 2.5 Baclofen
Veh.
1.25
2.5
(q/kg)
Baclofen
Fig. 2. Mean ( _ SEM) number of aggressive acts in control and ethanol-withdrawn rats, treated acutely with vehicle or baclofen (1.25 or 2.5 mg/kg), t t t P < 0.001 compared with vehicle-treated control group. *P < 0.05, **P < 0.01 compared with vehicle-treated ethanol-withdrawal group.
Table 2. Median tremor scores of rats withdrawn from ethanol and treated with vehicleor baclofen (1.25 or 2.5 mg/kg) or nitrendipine(25, 50 or 100 mg/kg) Median Range Experiment 1 Vehicle Baclofen 1.25 mg/kg 2.5 mg/kg
2.5
1-3
2.0 1.0"
1-2 0-2
without any change in total arm entries. The only measure to change in the holeboard, on withdrawal from ethanol, was rearing, which was significantly increased [ F ( 1 , 4 8 ) = 15.3, P <0.001], see Table I. Withdrawal from ethanol induced a significant [P < 0.001] tremor, see Table 2. Effect of baclofen. Baclofen was without effect in the control rats, but significantly reversed the decrease in social interaction shown by the ethanol-withdrawn animals (pretreatment x baclofen interaction, F(2,21) = 6.5, P < 0.01). As can be seen from Figure l, the reversal was due to the smaller dose o f baclofen. There was no effect o f baclofen on l o c o m o t o r activity or the n u m b e r o f rears made by pairs o f rats in the social interaction test [ F ( 2 . 2 1 ) < 1.0 in both cases].
Experiment 2 Vehicle 3.0 1-3 Nitrendipine 25 mg/kg 2.0 23 50 mg/kg 2.0 2-3 100 mg/kg 1.0" 0-1 *Significantly different from vehicle group, see text for details.
Ethanol..tthdrnal
Control 50 1
E 40 NN
_=
l
3O
2o m
,0
0 Yeh.
1.25
2.5
Baclofen
Veh.
i.25
2.5
(Jg/kg)
Baclofen
Fig. 3. Mean (+SEM) percentage time spent on the open arms of the plus-maze by rats pre-treated with the control diet or withdrawn from the ethanol diet, 30 min after injection with vehicle or baclofen (1.25 or 2.5mg/kg). "M'P<0.01 compared with vehicle-treated control group. **P <0.01 compared with vehicle-treated ethanol-withdrawal group.
187
Ethanol withdrawal: baclofen and nitrendipine Table 3. Mean (+ SEM) time spent in social interaction (sec), number of rears made in the social interaction test and percentage of time spent on the open arms of the plus-maze, by control (CON) and ethanol-withdrawn (EW) rats treated with vehicle or nitrendipine Nitrendipine (mg/kg) Vehicle 25 50 100 CON 100.0 _ 9.3 85.7 ± 22.8 78.5 ± 28.1 37.0 _ 13.6 Social EW 47.4** _ 12.8 31.7 ± 9.1 66.5 ± 9.5 31.0 +_5.0 CON 73.8 ± 3.5 62.3 ± 26.9 59.3 ± 19.6 35.0* + 16.2 Rears EW 82.2 ± 6.6 52.0 ± 15.7 76.5 ± 9.6 24.0~t _+8.3 Percentage time
CON EW
21.2 + 8.4 8.3 ± 2.0
17.9 ± 7.1 2.7 ± 1.3
11.5 ± 3.5 4.6 _ 3.2
2.4* + 1.5 3.1 + 3.1
*P < 0.05, **P < 0.001, significantlydifferent from vehicle-treatedvehicle-treatedcontrol group. t~'P < 0.01, significantlydifferent from vehicle-treatedwithdrawal group.
Baclofen dose-dependently reversed the number of aggressive acts shown by the ethanol-withdrawn rats [F(2,21) = 11.3, P < 0.001], see Figure 2 and the time spent in aggressive behaviour was significantly [P < 0.05] reduced by the larger dose. There was also a trend for baclofen to reduce the time spent in aggression [F(2,21)= 3.0]. Baclofen dose-dependently [P < 0.01] reversed the decrease in the percentage of time spent on the open arms of the plus-maze, shown by the ethanolwithdrawn animals, but was without significant effect in the control groups, see Figure 3. Baclofen was without significant effect on the number of rears made in the holeboard [F(2.48)= 2.5] and did not reverse the increase in rearing shown by the ethanol-withdrawn group, see Table 1. However, baclofen reduced the ethanol-withdrawal tremor [Kruskal-Wallis H - - 1 4 , P < 0.0001], but only the largest dose had a significant effect, see Table 2.
Experiment 2 Once again, withdrawal from ethanol resulted in significant anxiogenic effects, indicated by decreased social interaction [ F ( 1 , 2 2 ) = 7 . 4 , P <0.01] and a decreased percentage of time spent on the open arms [F(I,52) = 5.6, P < 0.05], see Table 3. These decreases were unaccompanied by significant decreases in locom o t o r activity or rearing in the social interaction test or by a change in the total arm entries in the plusmaze. There was also a significant increase in tremor on withdrawal of ethanol [H = 12.7, P < 0.001], see Table 2. Treatment with nitrendipine did not reverse the decreased social interaction [F(3,22)= 1.3] or the decrease in the percentage of time spent on the open arms, see Table 3. There was a trend towards a reduction of social interaction, caused by nitrendipine [F(3,22) = 2.6] and a trend towards reduced locomotion [F(3,22) = 2.4] but these did not reach overall significance. However, nitrendipine did significantly reduce rearing [F(3,22) = 4.7, P = 0.01]; this was due to the largest dose, see Table 3. There was also a trend towards a reduction by nitrendipine of the percentage of time spent on the open arms but this did not reach significance [F(3,52) = 1.9]. Nitrendipine significantly reversed the tremor, induced by withdrawal of alcohol
[H = 11, P < 0.001] but the reduction was significant only for the largest dose, see Table 2.
DISCUSSION It has previously been reported (File et al., 1989) that withdrawal from chronic treatment with ethanol produced an anxiogenic response in the social interaction test. In the present two experiments, this anxiogenie withdrawal response was detected in both the social interaction and the plus-maze tests. Small doses of baclofen reversed these anxiogenic responses, although baclofen was without anxiolytic effect in control rats. Baclofen also reversed the ethanolwithdrawal tremor, at a dose that was not generally sedative. In rhesus monkeys, baclofen (3 and 5.6 mg/ kg) produced only slight reductions in tremor and these doses were also sedative (Tarika and Winger, 1980). It therefore seems that, in some species, it might be difficult to separate the anti-tremor and sedative effects of baclofen. Baclofen reduced the increased aggression shown by the ethanol-withdrawn rats. However, there was a trend towards a reduction of the time spent in aggression in control animals, even though these levels were very low. This raises the possibility that baclofen might have an anti-aggressive action, that is not just limited to the conditions of withdrawal of ethanol. Baclofen failed to reduce the increased rearing in the holeboard, shown by the ethanol-withdrawn group, which suggests that this withdrawal response is mediated by a different mechanism from that underlying increased anxiety. It has long been suggested that different biochemical changes may mediate different withdrawal responses to ethanol and there is some evidence that noradrenergic pathways mediate the hyperactivity during withdrawal of ethanol (Trzaskowski, Pucilowski, Dyr, Kostowski and Hauptmann, 1986); there is also evidence that noradrenergic pathways mediate the hyperactivity during withdrawal of benzodiazepines (Kulchandy and Kulkarni, 1986). In contrast to the results with baclofen, the dihydropyridine calcium channel antagonist, nitrendipine, failed to reverse the anxiogenic withdrawal response. The social interaction scores in Experiment 2 were
188
SANDRAE. FILE et al.
smaller than in Experiment 1 but, in both cases, the decrease with withdrawal of ethanol was about 50%. It is unlikely that the failure to find reversal with nitrendipine was due to the differences in scores or to the fact that it was the second time the animals had been withdrawn from ethanol. In a previous study (File et al., 1989) it was also found that nitrendipine failed to reverse the anxiogenic withdrawal response to ethanol and in that study, the animals were tested after their first withdrawal from ethanol and their scores were very close to those found in Experiment 1. Thus, it seems that, in contrast to the ability of calcium channel antagonists to protect against tremor and seizures after withdrawal of ethanol (Little et al., 1986), nitrendipine was unable to reverse anxiogenic withdrawal responses. This argues for different mechanisms mediating these two withdrawal responses. Nitrendipine (100mg/kg) did reduce withdrawal tremor, but this dose was clearly sedative and therefore this is likely to be a non-specific effect. Baclofen seemed to have a more specific effect against ethanolwithdrawal tremor, at least in the rats used here and this finding is similar to that of Bone, Majchrowicz, Martin, Linnoila and Nutt (1989), showing diazepam to be superior to calcium channel antagonists. The reversals with diazepam and baclofen implicate both GABAA and GABA B receptors in the ethanolwithdrawal tremor. Since both the G A B A a agonist, baclofen (Experiment l) and the benzodiazepine antagonist, flumazenil (File et al., 1989) reversed the increased anxiety during withdrawal of ethanol, it seems that this response was the result of a decreased GABA function. This has also been suggested by Idemudia, Bhadra and Lal (1989), because both bicuculline and pentylenetetrazole enhanced the anxiogenic response in withdrawal of ethanol. During withdrawal of ethanol Ticku (1980) found a decreased affinity of the low affinity G A B A A receptors in the brain of the rat. However, this was not confirmed by Ulrichsen, Clemmesen, Barry and Hemmingsen (1988) and nor is there any change in binding at benzodiazepine or picrotoxinin sites (Rastogi, Thyagarajah, Clothier and Ticku, 1986; Tamborska and Marangos, 1986; Ulrichsen et aL, 1988). Ethanol stimulates the uptake of CI- into synaptoneurosomes (Suzdak, Schwartz, Skolnick and Paul, 1986) and cultured spinal cord neurones (Mehta and Ticku, 1988) and the increased chloride flux is blocked by picrotoxin (Mehta and Ticku, 1988; McQuilkin and Harris, 1990), but not by baclofen (Mehta and Ticku, 1990); however, the effects on chloride flux, during withdrawal of ethanol have not been reported. The reversal by flumazenil of the anxiogenic withdrawal response, suggests that during withdrawal of ethanol there is a change in the coupling of the GABAA-benzodiazepine-chloride channel receptor complex. Flumazenil was also able to reverse the anxiogenic response during withdrawal of benzodiazepines (Baldwin and File, 1988, 1989) and this can be explained either by antagonism of an endogen-
ous inverse agonist ligand or by a resetting of the benzodiazepine receptors to a neutral state (File and Hitchcock, 1990). It seems that flumazenil may also be able to reset to GABAA-benzodiazepine receptor complex during withdrawal of ethanol. A general change in GABA function has been proposed to underlie the symptoms of withdrawal of ethanol and benzodiazepine (Cowen and Nutt, 1982) and, although the present data on anxiety responses would support this hypothesis, it is unlikely to apply to all withdrawal responses. In particular, it is important to distinguish between the increased anxiety and the increased seizures seen during withdrawal of ethanol. Whereas baclofen and flumazenil can reverse the former they do not reverse the latter (Tarika and Winger, 1980; Adinoff, Majchrowicz, Martin and Linnoila, 1986; Mehta and Ticku, 1989). Indeed, the benzodiazepine inverse agonist Ro 15-4503 (ethyl 8-azido-5,6-dihydro-5-methyl-6-oxo-4H-imidazol[1,5- a][1,4] benzodiazepine- 3- carboxylate), becomes proconvulsant during withdrawal of ethanol (Lister and Karanian, 1987; Mehta and Ticku, 1989). Recent evidence implicates the N M D A (N-methyI-D-aspartate) receptor complex in ethanol-withdrawal seizures: they can be antagonised by the N M D A receptor antagonist MK 801, (+)-5-methyl- 10,1 l-dihydro-5Hdibeazo[a,d] cyclohepten-5,10-imine (Grant, Valverius, Hudsmith and Tabakoff, 1990; Morrisett, Rezvani, Overstreet, Janowsky, Wilson and Swartzwelder, 1990) and enhanced by N M D A (Grant et al., 1990). At the doses of baclofen that were used in this study, there is no evidence for its acting at GABAA receptors and therefore the results suggest that during withdrawal of ethanol, GABAB receptors may also be involved in the anxiogenic response. Whilst baclofen is not anxiolytic in control animals there is evidence that the combination of muscimol and baclofen is anxiolytic (Gray, 1983). Since by its action at presynaptic GABA receptors, baclofen would reduce the release of GABA, its reversal of the anxiogenic response must be either by a presynaptic action at heteroreceptors (e.g. Gray and Green, 1987) or by an action at post-synaptic GABA B receptors. Acute administration of ethanol has been reported to increase recurrent inhibition in the dentate gyrus of the hippocampus (Weisner, Henriksen and Bloom, 1987) and a post-synaptic GABA B site has been identified in the hippocampus (Dutar and Nicoll, 1988). In this region release of GABA from interneurones interacts with both GABA A and GABA Breceptors, to generate a fast and slow inhibitory postsynaptic potentials (i.p.s.p.), respectively and larger concentrations were needed to activate the GABA B receptors, which would explain why this system might not be active under normal circumstances. Whilst it can be concluded that baclofen reversed the anxiogenic responses, measured in the social interaction and plus-maze tests during withdrawal of ethanol, further experiments are needed to firmly establish that these effects of baclofen are indeed due
Ethanol withdrawal: baclofen and nitrendipine tO a n action of G A B A B receptors. It is also i m p o r t a n t to determine w h e t h e r it is post-synaptic G A B A B receptors t h a t decrease in n u m b e r or affinity d u r i n g withdrawal of e t h a n o l or w h e t h e r baclofen was exerting its effects t h r o u g h a n action at heteroreceptors. Acknowledgement--This study was supported by a grant from the Wellcome Trust. REFERENCES
Adinoff B., Majchrowicz E., Martin P. R. and Linnoila M. 0986) The benzodiazepine antagonist Ro 15-1788 does not antagonize the ethanol withdrawal syndrome. Biol. Psychiat. 21: 643-649. AI-Dahan M. I. and Thalman R. H. (1989) Effects of dihydropyridine calcium channel ligands on rat brain ~,aminobutyric acidB receptors. J. Neurochem. 53: 982-985. Baldwin H. A. and File S. E. (1988) Reversal of increased anxiety during benzodiazepine withdrawal: evidence for an anxiogenic endogenous ligand for the benzodiazepine receptor. Brain Res. Bull. 20: 603-606. Baldwin H. A. and File S. E. (1989) Flumazenil prevents the development of chlordiazepoxide withdrawal in rats tested in the social interaction test of anxiety. Psychopharmacology 97: 424-426. Bone G. H. A., Majchrowicz E., Martin P. R., Linnoila H. and Nutt D. J. (1989) A comparison of calcium antagonists and diazepam in reducing ethanol withdrawal tremors. Psychopharmacology 99: 386-388. Bowery N. G., Hill D. R., Hudson A. L., Doble A., Middlemiss D. N., Shaw J. and Turnbull M. (1980) (-)-Baclofen decreases neurotransmitter release in the mammalian CNS by an action at a novel GABA receptor. Nature 283: 92-94. Bowery N. G., Hill D. R. and Hudson A. L. (1982) Characteristics of GABA Breceptor binding sites on whole brain synaptic membranes. Br. J. Pharmac. 78: 191-206. Cowen P. J. and Nutt D. J. (1982) Abstinence symptoms after withdrawal of tranquillising drugs: is there a common neurochemical mechanism? Lancet 360-362. Dolin S. J. and Little H. J. (1989) Are changes in neuronal calcium channels involved in ethanol tolerance? J. Pharmac. exp. Ther. 25t1: 985-991. Dutar P. and Nicoll R. A. (1988) A physiological role for GABA A receptors in the central nervous system. Nature 332: 156-158. File S. E., Baldwin H. A. and Hitchcott P. K. (1989) Flumazenil but not nitrendipine reverses the increased anxiety during ethanol withdrawal in the rat. Psychopharmacology 98:262 264. File S. E. and Hitchcott P. K. (1990) A theory of benzodiazepine dependence that can explain whether flumazenil will enhance or reverse the phenomena. Psychopharmacology lOt: 525 532. Freund G. (1975) Induction of physical dependence on alcohol in rodents. In: Biochemial Pharmacology of Ethanol (Majchrowicz E., Ed.), pp. 211-235. Plenum Press, New York. Frye G. D., McCown T. J. and Breese G. R. (1983) Differential sensitivity of ethanol withdrawal signs in the rat to aminobutyric acid (GABA) mimetics: blockade of audiogenic seizures but not forelimb tremors. J. Pharmac. exp. Ther. 226:720 725. Gandhi C. R. and Ross D. H. (1989) Influence of ethanol on calcium, inositol phospholipids and intracellular signalling mechanisms. Experientia 45: 407-412. Grant K. A., Valverius P., Hudspith M. and Tabakoff B. (1990) Ethanol withdrawal seizures and the NMDA receptor complex. Eur. J. Pharmac. 176: 289-296. Gray J. A. (1983) Gamma-aminobutyrate, the benzodiazepines and the septohippocampal system. In: Benzo-
189
diazepines Divided (Trimble M., Ed.), pp. 101-128. Wiley, New York. Gray J. A. and Green A. R. 0987) GABA e receptor mediated inhibition of potassium evoked release of endogenous 5-hydroxytryptamine from mouse frontal cortex. Br. J. Pharmac. 91" 517-522. Idemudia S. O., Bhadra S. and Lal H. 0989) The pentylenetetrazol-like interoceptive stimulus produced by ethanol withdrawal is potentiated by bicuculline and picrotoxinin. Neuropsychopharmacology 2:115-122. Kunchandy J. and Kulkarni S. K. (1986) Reversal by alpha 2 agonists of diazepam hyperactivity in rats. Psychopharmacology 90: 198-202. Lister R. G. and Karanian I. W. (1987) Ro 15-4513 induces seizues in DBA/2 mice undergoing alcohol withdrawal. Alcohol 4:409-411. Little H. J., Dolin S. J. and Halsey M. J. (1986) Calcium channel antagonists decrease the ethanol withdrawal syndrome. Life Sci. 39: 2054-2065. Little H. J. and Dolin S. J. (1987) Lack of tolerance to ethanol after concurrent administration of nitrendipine. Br. J. Pharmac. 92: 606P. McQuilkin S. J. and Harris R. A. (1990) Factors affecting actions of ethanol on GABA-activated chloride channels. Life Sci. 46: 527-541. Mehta A. K. and Ticku M. K. 0988) Ethanol potentiation of GABAergic transmission in cultured spinal cord neurons involves 7-aminobutyric acidA-gated chloride channels. J. Pharmac. exp. Ther. 246: 558-564. Mehta A. K. and Ticku M. K. (1989) Chronic ethanol treatment alters the behavioral effects of Ro 15-4513, a partially negative ligand for benzodiazepine binding sites. Brain Res. 489: 93-100. Mehta A. K. and Ticku M. K. (1990) Are GABAB receptors involved in the pharmacological effects of ethanol? Eur. J. Pharmac. 182: 473-480. Morrisett R. A., Rezvani A. H., Overstreet D., Janowsky D. S., Wilson W. A. and Swartzwelder H. S. (1990) MK-801 potently inhibits alcohol withdrawal seizures in rats. Eur. J. Pharmac. 176: 103-105. Pucilowski O. and Kostowski W. (1988) Diltiazem suppresses apomorphine induced fighting and proaggressive effect of withdrawal from chronic ethanol or haloperidol in rats. Neurosci. Lett. 93: 96-100. Rastogi S. K., Thyagarajan R., Clothier J. and Ticku M. K. (1986) Effect of chronic treatment of ethanol on benzodiazepine and picrotoxin sites on the GABA receptor complex in regions of the brain of the rat. Neuropharmacology 25:1179-1184. Stirling J. M., Cross A. J., Robinson T. N. and Green A. R. (1989) The effect of GABA a receptor agonists and antagonists on potassium stimulated [Ca:+]i in rat brain synaptosomes. Neuropharmacology 28: 699-704. Suzdak P. D., Schwartz R. D., Skolnick P. and Paul S. M. (1986) Ethanol stimulates 7-aminobutyric acid receptormediated chloride transport in rat brain synaptoneurosomes. Neurobiology 83: 4071-4075. Sytinski I. A., Guzikov B. M., Gomanko M. V., Erenian V. P. and Konovalova N. N. (1975) The gamma-aminobutyric acid (GABA) system in brain during acute and chronic ethanol intoxication. J. Neurochem. 25: 43-48. Tamborska E. and Marangos P. J. (1986) Brain benzodiazepine binding sites in ethanol dependent and withdrawal states. Life Sci. 38: 465-472. Tarika J. S. and Winger G. (1980) The effect of ethanol, pentobarbital and baclofen on ethanol withdrawal in the rhesus monkeys. Psychopharmacology 70: 201-208. Ticku M. K. (1980) The effects of acute and.chronic ethanol administration and its withdrawal on 7-aminobutyric acid receptor binding in rat brain. Br. J. Pharmac. 70: 403-410. Ticku M. K. (1989) Ethanol and the benzodiazepine-GABA receptor-ionophore complex. Experientia 45: 413-417.
190
SANDRAE. FILE et al.
Trzaskowska E., Pucilowski O., Dyr W., Kostowski W. and Hauptmann M. (1986) Suppression of ethanol tolerance and dependence in rats treated with DSP-4, a noradrenergic neurotoxin. Drug Alcohol Dependence 18: 349-353. Ulrichsen J., Clemmesen L., Barry D. and Hemmingsen R. (1988) The GABA/benzodiazepine receptor chloride channel complex during repeated episodes of physical ethanol dependence in the rat. Psychopharmacology 96: 227-231.
Victor M. (1970) The alcohol withdrawal syndrome: theory and practice. Postgrad. Med. 47: 68-72. Volicer L. (1980) GABA levels and receptor binding after acute and chronic ethanol administration. Brain Res. Bull. 5: Suppl. 2, 809-813. Wiesner J. B., Henriksen S. J. and Bloom F. E. (1987) Ethanol enhances recurrent inhibition in the dentate gyrus of the hippocampus. Neurosci. Lett. 79: 169-173.