Alcohol & Alcoholism. Vol. 26. No 5/6. pp. 575-584. 1991 Printed in Great Britain
0735-0)14/91 $3.00 + 0.00 Pergamon Press pic © 1991 Medical Council on Alcoholism
CHANGES IN DRINKING BEHAVIOUR AND CHOLINERGIC BINDING SITES INDUCED BY INTERMITTENT LONG-TERM ETHANOL TREATMENT IN THE MALE RAT G. WAHLSTROM* and A. NORDBERGt 'Department of Pharmacology, University of Umea, S-901 87 Umea, Sweden; and fDepartment of Pharmacology, University of Uppsala, Biomedicum, Box 591, S-751 24 Uppsala, Sweden (Received2\ March 1991; accepted 15 August 1991) Abstract — Male rats were treated with ethanol for approximately 1 year. One group of rats (Group 1) could, for 24 hr each week, choose between ethanol (10%, w/v) and water as drinking fluid. At the end of each choice period, 2.0 g/kg ethanol was given ip. A second group (Group 2) had the same choice as Group 1, but was given a saline injection. A third group (Group 3) could at all times choose between ethanol and water. Saline was injected ip once a week. A control group (Group 4) was not exposed to ethanol. During the following evaluation period, lasting 30 weeks, Groups 1, 2 and 3 had, except for a 3 week period, continuous access to ethanol. Different concentrations of ethanol were tested. Group 4 had access to 30% ethanol only for one week. At the end of the experiment, nicotinic and muscarinic binding sites (total, measured with QNB, Ml and M2) were determined in brain cortex, striatum and medulla + midbrain. During the treatment, voluntary ethanol intake in Group 1 was inhibited compared to Group 2 with a maximum around week 5 and a return to the level of Group 2 after 15-20 weeks. After a 3 week interruption of access to ethanol during the evaluation period, intake of ethanol was increased in Groups 1, 2 and 3. The increase was independent of prior intake in Groups 1 and 2 and proportional to prior ethanol intake in Group 3, where the largest increase was seen in rats with the largest intake prior to interruption. During testing of different concentrations (5 and 20%) no change in individual intake was seen. Thus a short abstinence from ethanol was a way to increase voluntary ethanol intake with a more specific effect than a mere change of concentration. No differences between groups were seen in muscarinic or nicotinic binding in brain. The muscarinic receptor binding in striatum and medulla + midbrain positively correlated with ethanol intake in Group 3. Th nicotinic receptor binding in the cortex of Groups 1 and 2 positively correlated with water intake recorded during the last week of the evaluation period. The muscarinic Ml and M2 binding sites were positively correlated in the striatum in all groups, but differences between the groups were also seen in the correlation pattern within or between brain parts.
INTRODUCTION Detoxification which is equivalent to a forced abstinence is the natural first step in treatment of alcoholism in the human. After such a detoxification, a short-term temporal improvement with total abstinence from ethanol is often seen, but the number of relapses is usually very high during the following months (Horowitz et al., 1970; Schuckit, 1979; Shaw, 1980; Longabaugh, 1988). Two probably related basic questions could be asked with regard to this restarted ethanol intake; why did the relapse occur; and does ethanol intake
after relapse differ from that prior to detoxification? The complex first question can at present only be elucidated by a clinical approach, while the more straightforward second question can be studied experimentally provided that the clinical prerequisites for starting a detoxification such as dependence and a long-term ethanol exposure are included in the experimental model. The main object of the present paper was to study this question with a recently developed technique of alcoholization which includes both prerequisites, There is a long period of intermittent ethanol exposure (treatment) and this treatment is
575
576
G. WAHLSTROM and A. NORDBERG
known to induce empirical signs of dependence in the male rat (Wahlstrom, 1983, 1985; Wahlstrom and Nordberg, 1987; Wahlstrom et al., 1988). The definition of dependence used here is an oral voluntary ethanol intake which is independent of the concentration of the offered ethanol solution. Voluntary intake of ethanol corresponding to this definition has been seen during the evaluation period after long-term treatments which consist of an approximately 24 hr each week restricted ethanol exposure (Wahlstrom 1983, 1985; Wahlstrom and Nordberg 1987; Wahlstrom et al., 1988). The reason for the use of such a treatment is the Nordic social pattern of ethanol drinking in non-alcoholic subjects, where a heavy consumption is seen during the weekends with only a fairly small intake during weekdays (Simpura, 1987). Since one of the main features of the present model is a stable voluntary intake of ethanol during the evaluation period after the end of the treatment, the present experiment was ended with a neurochemical analysis. The idea is to search for correlations which could indicate possible causal relations between changes in neurochemical variables and voluntary ethanol intake. As recently discussed by Kalant (1990), this is a complex field, where basic information certainly is needed. Due to earlier experience (Wahlstrom and Nordberg, 1988) the present analyses were focused on cholinergic mechanisms.
Design of experiment
80
TREATMENT PERIOD
WEEKS
EVALUATION PERIOD
Fig. 1. Schematical representation of the design of the experiment. The number of the groups and the symbols used are given to the left and the number of participating animals to the right (N). Treatments are denoted as follows: in Group 1, oblique thick lines indicate choice between ethanol (10%, w/v) and water for 24 hr immediately followed by an ethanol injection (2.0 g/kg ip) each week; in Group 2, oblique thin lines indicate choice between ethanol (10%) and water for 24 hr immediately followed by a saline injection (ip) each week; in Group 3, horizontal lines indicate a continuous random choice between water and ethanol (10%) as drinking fluid: and in Group 4 open bar indicates neither ethanol exposure nor any injections. In the evaluation period, the same symbols as in the treatment period of Groups 3 and 4 indicate the same ethanol choice conditions. No injections were given during the evaluation period.
age of the rats at the start of the experiment was approximately 6 weeks.
Design of experiment The basic design of the experiment (Fig. 1) was similar to that used in previous experiMETHODS ments (Wahlstrom 1983, 1987; Wahlstrom et Animals al., 1988). In the present experiment, the Male Sprague-Dawley rats (MokSPRD, animals were divided into 4 groups. Groups 1 Mollegaards Breeding Center Ltd, Li. Skens- and 2 participated in a treatment period folved, Denmark) were housed individually in an lowed by an evaluation period. During the animal room with a reversed light/dark sche- treatment period (week 1-54; Fig. 1), these dule (light on 19:00-07:00 hr) and a tempera- groups were given one ip injection/week of ture around 24°C. Each cage was supplied with either ethanol (Group 1:2.0 g/kg, 10%, w/v in 2 drinking bottles, 1 of which always contained saline) or saline (Group 2). For 24 hr prior to drinking water. In the other bottle, there was, the injection, Groups 1 and 2 had the choice when appropriate (see experimental design), between ethanol (10%, w/v) or water as the an ethanol solution. Food (commercial rat drinking fluid. During the treatment, except in pellets R 3, Ewos AB, Sweden) was available the last week, ethanol was offered in the bottle ad libitum. After 1 week of acclimatization, the from which the rat had taken most water treatment, as shown in Fig. 1, was started. The during the previous 6 days. The reverse proce-
ALCOHOL CONSUMPTION AND CHOLINERGIC BINDING SITES
dure performed during the last week of the treatment (week 54) has been denoted critical test (Wahlstrom et al., 1988). During the evaluation period (week 55-84; Fig. 1) Groups 1 and 2 had continuous access to ethanol in a standard 2 bottle choice, except for a 3-week period (weeks 60-62; Fig.l), when water was offered as the only drinking fluid in both bottles. The position of the ethanol bottle was altered each week in a random fashion. The standard ethanol concentration in the evaluation period was 10% (w/v), but during some parts of the evaluation period different concentrations of ethanol were offered for periods of 3 weeks as the choice against water (5% in weeks 66-68, 20% in weeks 72-74 and 30% in weeks 78-80). During the first days of week 85, ethanol was withdrawn in the afternoon and the rats were killed by decapitation the following morning (after approximately 12 hr without ethanol). Group 3 had a different ethanol exposure, compared with Groups 1 and 2. During the treatment period (weeks 1-54; Fig. 1) these rats had continuous access to ethanol (10%, w/v) in a standard 2 bottle choice. The position of the ethanol solution was randomly allocated each week. Saline was injected ip once a week. During the evaluation period (weeks 55-85; Fig. 1), these rats had the same schedule of exposure to ethanol as Groups 1 and 2. They were killed by decapitation after approximately 12 hr without ethanol. Group 4 consisted of untreated controls. The only ethanol exposure of this group consisted of a choice between water and ethanol (30%, w/v) during week 83. They were killed by decapitation during the first days of week 85 in the same manner as the other 3 groups. At autopsy, body weight and the weight of brain, liver, kidney, adrenal glands and testes were measured. The brain was dissected on ice. Cortex, striatum and medulla + midbrain were kept frozen (-80°C) until analysed. Neurochemical methods Crude P2 fractions were prepared by homogenizing the brain tissue in 0.32 M sucrose solution (20 vol) and centrifuging at 2800 rpm at 4°C for 10 min. The supernatant was centrifuged at 12,000 rpm at 4°C for 15 min and the resulting pellet (P2 fraction) was dissolved
577
in 0.32 M sucrose solution to a final protein concentration of 0.1-0.4 mg/ml. The protein concentration was determined according to Lowry et al. (1951). P2 fractions were incubated with 3H-quinuclidinyl benzilate (3HQNB, 0.2 nM, spec. act. 42 Ci/mmol, Amersham, UK) in 50 mM Na-K phosphate buffer (pH 7.4) to a final volume of 1.0 ml in order to measure total muscarinic binding. In order to differentiate between Ml and M2 receptor binding sites, carbachol (10~4 M) (Ml) or pirenzepine (3 x 10~7 M) (M2) were added to the incubation media in some experiments according to the method described by Corte's et al. (1986). The incubation was performed at 25°C for 90 min and was terminated by filtration of the tissue homogenate through Whatman's GF/C glass filters, presoaked in a 0.05% polyethylenimine (PEI) solution. The filters were washed with 4 ml portions of cold buffer, dried and placed in a scintillation tube containing 5 ml HP/B scintillation fluid and counted in a scintillation counter (Nuclear Chicago®). Non-specific binding was defined in the presence of atropine 10~4 M. In order to measure nicotinic receptors, P2 fractions were incubated with 5 nM (-)3H-nicotine (SA 80 Ci/ mmol, Isolab, Uppsala, Sweden) in 50 mM Tris-HCl buffer (pH 8.0) at 4°C for 40 min. The samples were then filtered through Whatman GF/C glass filters presoaked with a 0.05% PEI solution. The filters were washed 3 times with the assay buffer and the radioactivity entrapped in the filters was counted. The specific binding was calculated by subtracting the values for nonspecific binding in the presence of 10~3 M unlabelled (-)nicotine from the total binding. All samples were assayed in triplicate. Statistical methods Conventional statistical methods were used. Differences between 2 groups were tested with Student's r-test; differences between more groups were initially tested with a one-way analysis of variance. Parametric statistical values on correlation (r) and regression (b) were calculated. A P < 0.05 was used as a basic level of significance. NS denotes values outside this range. N denotes number of observations and DF degrees of freedom.
G. WAHLSTROM and A. NORDBERG
578
RESULTS Treatment period Ethanol intake in Groups 1, 2 and 3 during the treatment period is shown in Fig. 2. The ethanol intake for 24 hr each week in Group 2, which each week was given the saline injections, increased slightly during the first 10 weeks of treatment. After that, there was a gradual reduction down to a level around 4 g/kg/day at the end of the treatment. The corresponding intake in Group 1, which each week was given the ethanol injections, decreased the first week after the initial ethanol injection. The decrease was maximum around week 5, followed during the following 10 weeks by a gradual return to the control level represented by Group 2. During the critical test in week 54, the intake in Groups 1 and 2 was similar to that seen in the previous week (Fig. 2 and Table 1). In Group 3, with the continuous choice, intake during the treatment period was stable at a level around 2 g/kg/day. To compare total ethanol exposure in the different groups, a week is a convenient time unit. Calculated in this manner total ethanol exposure in Group 1 was (after week 15 in the treatment period) approximately 6 g/kg/week, in Group 2 ap-
u. O
:•
•
• . - :
-.
•
UJ
DURATION OF TREATMENT
Fig. 2. Daily voluntary ethanol intake during the treatment period. The treatment schedule is given in Fig. 1. In Groups 1 (•) and 2 (•) the intake represents 24 hr of choice between ethanol and water each week. In Group 3 (A) the intake represents average daily doses of ethanol for each week calculated on data from a continuous choice between ethanol and water.
proximately 4 g/kg/week and in Group 3 approximately 14 g/kg/week. Thus total exposure in Group 3 was more than twice the total exposure in Groups 1 and 2, though the time of the exposure was restricted in these groups. Evaluation period
Data on ethanol and fluid intake from the last week of the treatment period and from some weeks in the evaluation period are given in Table 1. It is evident that, during the last week of the treatment period (week 54), voluntary daily intake of ethanol in Groups 1 and 2 recorded for 24 hr was twice as large as intake recorded continuously in Group 3. During the first weeks of the evaluation period, when all groups had continuous access to ethanol, this difference was reduced by decreased intake in Groups 1 and 2. The new level of intake in these latter two groups was established around week 58 (Table 1). The difference (1.0-1.5 g/kg/day) between Groups 1 and 2 on the one hand and Group 3 on the other, was at this stage not significant for individual weeks, but with the exception of some experimentally induced changes (period without access to ethanol and testing with 30% ethanol, see below) was fairly stable during the remaining part of the evaluation period. If voluntary ethanol intake for 24 hr recorded in Groups 1 and 2 in week 54 (critical test at the end of treatment) is compared with daily ethanol intake recorded for 7 days in week 55 (first week of evaluation period), a highly significant correlation is obtained (r = 0.81, b = 0.67, DF = 14, P < 0.001). The corresponding value in Group 3 between ethanol intake recorded for 7 days both in week 54 and in week 55 was not significant (r = 0.22, b = 0.11, DF = 6, NS). Thus the 24 hr intake recorded in Groups 1 and 2 during the last week of the treatment period is clearly related to intake at the beginning of the continuous choice situation in the evaluation period, although the intake was reduced. The corresponding relation in Group 3, which had continuous access to ethanol without interruption, appeared to be much less pronounced. The effect of a 3-week interruption of access to ethanol on individual drinking pattern was studied during the evaluation period. Weeks
ALCOHOL CONSUMPTION AND CHOLINERGIC BINDING SITES
579
Table 1. Ethanol, water and total fluid intake (total) during the last week (W) of the treatment period (W 54) and during the evaluation period (W 55, W 58, W 65 and W 84). Thirty per cent ethanol was offered to Groups 1-3 in W 78-80 and to Group 4 in W 83. W 85 only consists of the last 12 hr of the evaluation period without ethanol exposure. D = 24 hr, N = number of animals, NS = not significant, % = per cent ethanol in offered solution (w/v) Voluntary ethanol intake (g/kg) Group
N
—
W 84
W 84
W 84
W 85
10%
30%
10%
7 D
7-21 D
7 D
Water 7 D
Total 7D
Water 0.5 D
2.88 0.38 3.11 0.42 1.94 0.25
1.77 0.23 2.40 0.44 1.45 0.21 0.92 0.21
2.31 0.26 2.40 0.32 1.51 0.30
65 3
88.4
7.7
9.8
3.8
61.1
85.1
34.4
8.4
7.9
3.9
—
82.7
82.7
27.2
5.8
5.8
2.3
2.37
0.79
NS
NS
W 54
W 55
W 58
W 65
10%
10%
10%
1 D
7 D
7 D
4.02 0.32 4.46 0.86 1.89 0.52
3.24 0.37 3.41 0.68 1.65 0.17
2
7
3
7
4
10
—
—
2.93 0.47 3.02 0.53 1.60 0.25 .—
F
2,18 3,27
4.88
4.56
3.35
3.03
< 0.05
< 0.05
NS
NS
1
P
7
Fluid intake (g/kg)
60-62 in the evaluation period, when intake in Groups 1 and 2 had stabilized, were selected as the period without any ethanol exposure. If individual ethanol intake during the last 3 weeks prior to interruption are related to the corresponding intake recorded during the first 3 weeks after interruption the results seen in Fig. 3 (A, B and C) were obtained. It is clear that intake prior to interruption is strongly correlated to that after interruption in all 3 groups. However, there are also some differences between Groups 1 and 2 on the one hand, and Group 3 on the other. The intercept with the ordinate is significantly different from zero both in Groups 1 (0.85 ± 0.26) and 2 (0.90 ± 0.32), but clearly not in Group 3 (-0.14 ± 0.20). The regression coefficient (b in Fig. 3 A, B and C) in all 3 groups was significantly different from 0 but significantly different from 1.0 only in Group 3 (P < 0.02). This suggests that the re-established drinking of ethanol after interruption occurred with a fixed increase, corresponding to the deviation in the ordinate, in all individuals in Groups 1 and 2. In Group 3, which had continuous choice throughout the treatment period, there was a proportional and surprisingly linear increase with the largest increase in the individuals which prior to
42.5
8.4
7.1
3.1
82.8
97.9
42.3
2.73 5.19 < 0.01
NS
5.74 < 0.01
interruption showed the largest intake (Fig. 3C). The range of average total fluid intake (calculated each week) was (during weeks 5765) in Group 1: 83.5-91.5 g/kg/day; in Group 2: 78.9-86.8 g/kg/day; in Group 3: 83.0-90.3 g/ kg/day; and in Group 4: 73.2-78.1 g/kg/day. The standard error of the mean of these averages varied between 4-10 g/kg/day and there were no significant changes within the groups. Thus there was no change in total fluid intake within the groups in the weeks prior to, during or after the period without access to ethanol. During the evaluation period, 3 tests with different ethanol concentrations were performed for periods of 3 weeks in Groups 1, 2 and 3. The first 2 tests were performed during weeks 66-68 (5% instead of 10% ethanol) and during weeks 72-74 (20% instead of 10%). If the average voluntary individual intake from 2 control weeks on 10% ethanol prior to and after the changed concentration are compared with the corresponding average intake recorded during the 3-week test period, correlations are obtained indicating dependence on ethanol (Wahlstrom, 1987). When tested with 5% (data not shown), these correlations were 0.77 (Group 1), 0.90 (Group 2) and 0.96
580
G. WAHLSTROM and A. NORDBERG
GROUP 1 ^kg
UJ
GROUP 3
GROUP 2
A.
y
5 . r=0.97 b=0.77
y
y
OC
m
t3 •
0735-0)14/91 $3.00 + 0.00 Pergamon Press pic © 1991 Medical Council on Alcoholism
CHANGES IN DRINKING BEHAVIOUR AND CHOLINERGIC BINDING SITES INDUCED BY INTERMITTENT LONG-TERM ETHANOL TREATMENT IN THE MALE RAT G. WAHLSTROM* and A. NORDBERGt 'Department of Pharmacology, University of Umea, S-901 87 Umea, Sweden; and fDepartment of Pharmacology, University of Uppsala, Biomedicum, Box 591, S-751 24 Uppsala, Sweden (Received2\ March 1991; accepted 15 August 1991) Abstract — Male rats were treated with ethanol for approximately 1 year. One group of rats (Group 1) could, for 24 hr each week, choose between ethanol (10%, w/v) and water as drinking fluid. At the end of each choice period, 2.0 g/kg ethanol was given ip. A second group (Group 2) had the same choice as Group 1, but was given a saline injection. A third group (Group 3) could at all times choose between ethanol and water. Saline was injected ip once a week. A control group (Group 4) was not exposed to ethanol. During the following evaluation period, lasting 30 weeks, Groups 1, 2 and 3 had, except for a 3 week period, continuous access to ethanol. Different concentrations of ethanol were tested. Group 4 had access to 30% ethanol only for one week. At the end of the experiment, nicotinic and muscarinic binding sites (total, measured with QNB, Ml and M2) were determined in brain cortex, striatum and medulla + midbrain. During the treatment, voluntary ethanol intake in Group 1 was inhibited compared to Group 2 with a maximum around week 5 and a return to the level of Group 2 after 15-20 weeks. After a 3 week interruption of access to ethanol during the evaluation period, intake of ethanol was increased in Groups 1, 2 and 3. The increase was independent of prior intake in Groups 1 and 2 and proportional to prior ethanol intake in Group 3, where the largest increase was seen in rats with the largest intake prior to interruption. During testing of different concentrations (5 and 20%) no change in individual intake was seen. Thus a short abstinence from ethanol was a way to increase voluntary ethanol intake with a more specific effect than a mere change of concentration. No differences between groups were seen in muscarinic or nicotinic binding in brain. The muscarinic receptor binding in striatum and medulla + midbrain positively correlated with ethanol intake in Group 3. Th nicotinic receptor binding in the cortex of Groups 1 and 2 positively correlated with water intake recorded during the last week of the evaluation period. The muscarinic Ml and M2 binding sites were positively correlated in the striatum in all groups, but differences between the groups were also seen in the correlation pattern within or between brain parts.
INTRODUCTION Detoxification which is equivalent to a forced abstinence is the natural first step in treatment of alcoholism in the human. After such a detoxification, a short-term temporal improvement with total abstinence from ethanol is often seen, but the number of relapses is usually very high during the following months (Horowitz et al., 1970; Schuckit, 1979; Shaw, 1980; Longabaugh, 1988). Two probably related basic questions could be asked with regard to this restarted ethanol intake; why did the relapse occur; and does ethanol intake
after relapse differ from that prior to detoxification? The complex first question can at present only be elucidated by a clinical approach, while the more straightforward second question can be studied experimentally provided that the clinical prerequisites for starting a detoxification such as dependence and a long-term ethanol exposure are included in the experimental model. The main object of the present paper was to study this question with a recently developed technique of alcoholization which includes both prerequisites, There is a long period of intermittent ethanol exposure (treatment) and this treatment is
575
576
G. WAHLSTROM and A. NORDBERG
known to induce empirical signs of dependence in the male rat (Wahlstrom, 1983, 1985; Wahlstrom and Nordberg, 1987; Wahlstrom et al., 1988). The definition of dependence used here is an oral voluntary ethanol intake which is independent of the concentration of the offered ethanol solution. Voluntary intake of ethanol corresponding to this definition has been seen during the evaluation period after long-term treatments which consist of an approximately 24 hr each week restricted ethanol exposure (Wahlstrom 1983, 1985; Wahlstrom and Nordberg 1987; Wahlstrom et al., 1988). The reason for the use of such a treatment is the Nordic social pattern of ethanol drinking in non-alcoholic subjects, where a heavy consumption is seen during the weekends with only a fairly small intake during weekdays (Simpura, 1987). Since one of the main features of the present model is a stable voluntary intake of ethanol during the evaluation period after the end of the treatment, the present experiment was ended with a neurochemical analysis. The idea is to search for correlations which could indicate possible causal relations between changes in neurochemical variables and voluntary ethanol intake. As recently discussed by Kalant (1990), this is a complex field, where basic information certainly is needed. Due to earlier experience (Wahlstrom and Nordberg, 1988) the present analyses were focused on cholinergic mechanisms.
Design of experiment
80
TREATMENT PERIOD
WEEKS
EVALUATION PERIOD
Fig. 1. Schematical representation of the design of the experiment. The number of the groups and the symbols used are given to the left and the number of participating animals to the right (N). Treatments are denoted as follows: in Group 1, oblique thick lines indicate choice between ethanol (10%, w/v) and water for 24 hr immediately followed by an ethanol injection (2.0 g/kg ip) each week; in Group 2, oblique thin lines indicate choice between ethanol (10%) and water for 24 hr immediately followed by a saline injection (ip) each week; in Group 3, horizontal lines indicate a continuous random choice between water and ethanol (10%) as drinking fluid: and in Group 4 open bar indicates neither ethanol exposure nor any injections. In the evaluation period, the same symbols as in the treatment period of Groups 3 and 4 indicate the same ethanol choice conditions. No injections were given during the evaluation period.
age of the rats at the start of the experiment was approximately 6 weeks.
Design of experiment The basic design of the experiment (Fig. 1) was similar to that used in previous experiMETHODS ments (Wahlstrom 1983, 1987; Wahlstrom et Animals al., 1988). In the present experiment, the Male Sprague-Dawley rats (MokSPRD, animals were divided into 4 groups. Groups 1 Mollegaards Breeding Center Ltd, Li. Skens- and 2 participated in a treatment period folved, Denmark) were housed individually in an lowed by an evaluation period. During the animal room with a reversed light/dark sche- treatment period (week 1-54; Fig. 1), these dule (light on 19:00-07:00 hr) and a tempera- groups were given one ip injection/week of ture around 24°C. Each cage was supplied with either ethanol (Group 1:2.0 g/kg, 10%, w/v in 2 drinking bottles, 1 of which always contained saline) or saline (Group 2). For 24 hr prior to drinking water. In the other bottle, there was, the injection, Groups 1 and 2 had the choice when appropriate (see experimental design), between ethanol (10%, w/v) or water as the an ethanol solution. Food (commercial rat drinking fluid. During the treatment, except in pellets R 3, Ewos AB, Sweden) was available the last week, ethanol was offered in the bottle ad libitum. After 1 week of acclimatization, the from which the rat had taken most water treatment, as shown in Fig. 1, was started. The during the previous 6 days. The reverse proce-
ALCOHOL CONSUMPTION AND CHOLINERGIC BINDING SITES
dure performed during the last week of the treatment (week 54) has been denoted critical test (Wahlstrom et al., 1988). During the evaluation period (week 55-84; Fig. 1) Groups 1 and 2 had continuous access to ethanol in a standard 2 bottle choice, except for a 3-week period (weeks 60-62; Fig.l), when water was offered as the only drinking fluid in both bottles. The position of the ethanol bottle was altered each week in a random fashion. The standard ethanol concentration in the evaluation period was 10% (w/v), but during some parts of the evaluation period different concentrations of ethanol were offered for periods of 3 weeks as the choice against water (5% in weeks 66-68, 20% in weeks 72-74 and 30% in weeks 78-80). During the first days of week 85, ethanol was withdrawn in the afternoon and the rats were killed by decapitation the following morning (after approximately 12 hr without ethanol). Group 3 had a different ethanol exposure, compared with Groups 1 and 2. During the treatment period (weeks 1-54; Fig. 1) these rats had continuous access to ethanol (10%, w/v) in a standard 2 bottle choice. The position of the ethanol solution was randomly allocated each week. Saline was injected ip once a week. During the evaluation period (weeks 55-85; Fig. 1), these rats had the same schedule of exposure to ethanol as Groups 1 and 2. They were killed by decapitation after approximately 12 hr without ethanol. Group 4 consisted of untreated controls. The only ethanol exposure of this group consisted of a choice between water and ethanol (30%, w/v) during week 83. They were killed by decapitation during the first days of week 85 in the same manner as the other 3 groups. At autopsy, body weight and the weight of brain, liver, kidney, adrenal glands and testes were measured. The brain was dissected on ice. Cortex, striatum and medulla + midbrain were kept frozen (-80°C) until analysed. Neurochemical methods Crude P2 fractions were prepared by homogenizing the brain tissue in 0.32 M sucrose solution (20 vol) and centrifuging at 2800 rpm at 4°C for 10 min. The supernatant was centrifuged at 12,000 rpm at 4°C for 15 min and the resulting pellet (P2 fraction) was dissolved
577
in 0.32 M sucrose solution to a final protein concentration of 0.1-0.4 mg/ml. The protein concentration was determined according to Lowry et al. (1951). P2 fractions were incubated with 3H-quinuclidinyl benzilate (3HQNB, 0.2 nM, spec. act. 42 Ci/mmol, Amersham, UK) in 50 mM Na-K phosphate buffer (pH 7.4) to a final volume of 1.0 ml in order to measure total muscarinic binding. In order to differentiate between Ml and M2 receptor binding sites, carbachol (10~4 M) (Ml) or pirenzepine (3 x 10~7 M) (M2) were added to the incubation media in some experiments according to the method described by Corte's et al. (1986). The incubation was performed at 25°C for 90 min and was terminated by filtration of the tissue homogenate through Whatman's GF/C glass filters, presoaked in a 0.05% polyethylenimine (PEI) solution. The filters were washed with 4 ml portions of cold buffer, dried and placed in a scintillation tube containing 5 ml HP/B scintillation fluid and counted in a scintillation counter (Nuclear Chicago®). Non-specific binding was defined in the presence of atropine 10~4 M. In order to measure nicotinic receptors, P2 fractions were incubated with 5 nM (-)3H-nicotine (SA 80 Ci/ mmol, Isolab, Uppsala, Sweden) in 50 mM Tris-HCl buffer (pH 8.0) at 4°C for 40 min. The samples were then filtered through Whatman GF/C glass filters presoaked with a 0.05% PEI solution. The filters were washed 3 times with the assay buffer and the radioactivity entrapped in the filters was counted. The specific binding was calculated by subtracting the values for nonspecific binding in the presence of 10~3 M unlabelled (-)nicotine from the total binding. All samples were assayed in triplicate. Statistical methods Conventional statistical methods were used. Differences between 2 groups were tested with Student's r-test; differences between more groups were initially tested with a one-way analysis of variance. Parametric statistical values on correlation (r) and regression (b) were calculated. A P < 0.05 was used as a basic level of significance. NS denotes values outside this range. N denotes number of observations and DF degrees of freedom.
G. WAHLSTROM and A. NORDBERG
578
RESULTS Treatment period Ethanol intake in Groups 1, 2 and 3 during the treatment period is shown in Fig. 2. The ethanol intake for 24 hr each week in Group 2, which each week was given the saline injections, increased slightly during the first 10 weeks of treatment. After that, there was a gradual reduction down to a level around 4 g/kg/day at the end of the treatment. The corresponding intake in Group 1, which each week was given the ethanol injections, decreased the first week after the initial ethanol injection. The decrease was maximum around week 5, followed during the following 10 weeks by a gradual return to the control level represented by Group 2. During the critical test in week 54, the intake in Groups 1 and 2 was similar to that seen in the previous week (Fig. 2 and Table 1). In Group 3, with the continuous choice, intake during the treatment period was stable at a level around 2 g/kg/day. To compare total ethanol exposure in the different groups, a week is a convenient time unit. Calculated in this manner total ethanol exposure in Group 1 was (after week 15 in the treatment period) approximately 6 g/kg/week, in Group 2 ap-
u. O
:•
•
• . - :
-.
•
UJ
DURATION OF TREATMENT
Fig. 2. Daily voluntary ethanol intake during the treatment period. The treatment schedule is given in Fig. 1. In Groups 1 (•) and 2 (•) the intake represents 24 hr of choice between ethanol and water each week. In Group 3 (A) the intake represents average daily doses of ethanol for each week calculated on data from a continuous choice between ethanol and water.
proximately 4 g/kg/week and in Group 3 approximately 14 g/kg/week. Thus total exposure in Group 3 was more than twice the total exposure in Groups 1 and 2, though the time of the exposure was restricted in these groups. Evaluation period
Data on ethanol and fluid intake from the last week of the treatment period and from some weeks in the evaluation period are given in Table 1. It is evident that, during the last week of the treatment period (week 54), voluntary daily intake of ethanol in Groups 1 and 2 recorded for 24 hr was twice as large as intake recorded continuously in Group 3. During the first weeks of the evaluation period, when all groups had continuous access to ethanol, this difference was reduced by decreased intake in Groups 1 and 2. The new level of intake in these latter two groups was established around week 58 (Table 1). The difference (1.0-1.5 g/kg/day) between Groups 1 and 2 on the one hand and Group 3 on the other, was at this stage not significant for individual weeks, but with the exception of some experimentally induced changes (period without access to ethanol and testing with 30% ethanol, see below) was fairly stable during the remaining part of the evaluation period. If voluntary ethanol intake for 24 hr recorded in Groups 1 and 2 in week 54 (critical test at the end of treatment) is compared with daily ethanol intake recorded for 7 days in week 55 (first week of evaluation period), a highly significant correlation is obtained (r = 0.81, b = 0.67, DF = 14, P < 0.001). The corresponding value in Group 3 between ethanol intake recorded for 7 days both in week 54 and in week 55 was not significant (r = 0.22, b = 0.11, DF = 6, NS). Thus the 24 hr intake recorded in Groups 1 and 2 during the last week of the treatment period is clearly related to intake at the beginning of the continuous choice situation in the evaluation period, although the intake was reduced. The corresponding relation in Group 3, which had continuous access to ethanol without interruption, appeared to be much less pronounced. The effect of a 3-week interruption of access to ethanol on individual drinking pattern was studied during the evaluation period. Weeks
ALCOHOL CONSUMPTION AND CHOLINERGIC BINDING SITES
579
Table 1. Ethanol, water and total fluid intake (total) during the last week (W) of the treatment period (W 54) and during the evaluation period (W 55, W 58, W 65 and W 84). Thirty per cent ethanol was offered to Groups 1-3 in W 78-80 and to Group 4 in W 83. W 85 only consists of the last 12 hr of the evaluation period without ethanol exposure. D = 24 hr, N = number of animals, NS = not significant, % = per cent ethanol in offered solution (w/v) Voluntary ethanol intake (g/kg) Group
N
—
W 84
W 84
W 84
W 85
10%
30%
10%
7 D
7-21 D
7 D
Water 7 D
Total 7D
Water 0.5 D
2.88 0.38 3.11 0.42 1.94 0.25
1.77 0.23 2.40 0.44 1.45 0.21 0.92 0.21
2.31 0.26 2.40 0.32 1.51 0.30
65 3
88.4
7.7
9.8
3.8
61.1
85.1
34.4
8.4
7.9
3.9
—
82.7
82.7
27.2
5.8
5.8
2.3
2.37
0.79
NS
NS
W 54
W 55
W 58
W 65
10%
10%
10%
1 D
7 D
7 D
4.02 0.32 4.46 0.86 1.89 0.52
3.24 0.37 3.41 0.68 1.65 0.17
2
7
3
7
4
10
—
—
2.93 0.47 3.02 0.53 1.60 0.25 .—
F
2,18 3,27
4.88
4.56
3.35
3.03
< 0.05
< 0.05
NS
NS
1
P
7
Fluid intake (g/kg)
60-62 in the evaluation period, when intake in Groups 1 and 2 had stabilized, were selected as the period without any ethanol exposure. If individual ethanol intake during the last 3 weeks prior to interruption are related to the corresponding intake recorded during the first 3 weeks after interruption the results seen in Fig. 3 (A, B and C) were obtained. It is clear that intake prior to interruption is strongly correlated to that after interruption in all 3 groups. However, there are also some differences between Groups 1 and 2 on the one hand, and Group 3 on the other. The intercept with the ordinate is significantly different from zero both in Groups 1 (0.85 ± 0.26) and 2 (0.90 ± 0.32), but clearly not in Group 3 (-0.14 ± 0.20). The regression coefficient (b in Fig. 3 A, B and C) in all 3 groups was significantly different from 0 but significantly different from 1.0 only in Group 3 (P < 0.02). This suggests that the re-established drinking of ethanol after interruption occurred with a fixed increase, corresponding to the deviation in the ordinate, in all individuals in Groups 1 and 2. In Group 3, which had continuous choice throughout the treatment period, there was a proportional and surprisingly linear increase with the largest increase in the individuals which prior to
42.5
8.4
7.1
3.1
82.8
97.9
42.3
2.73 5.19 < 0.01
NS
5.74 < 0.01
interruption showed the largest intake (Fig. 3C). The range of average total fluid intake (calculated each week) was (during weeks 5765) in Group 1: 83.5-91.5 g/kg/day; in Group 2: 78.9-86.8 g/kg/day; in Group 3: 83.0-90.3 g/ kg/day; and in Group 4: 73.2-78.1 g/kg/day. The standard error of the mean of these averages varied between 4-10 g/kg/day and there were no significant changes within the groups. Thus there was no change in total fluid intake within the groups in the weeks prior to, during or after the period without access to ethanol. During the evaluation period, 3 tests with different ethanol concentrations were performed for periods of 3 weeks in Groups 1, 2 and 3. The first 2 tests were performed during weeks 66-68 (5% instead of 10% ethanol) and during weeks 72-74 (20% instead of 10%). If the average voluntary individual intake from 2 control weeks on 10% ethanol prior to and after the changed concentration are compared with the corresponding average intake recorded during the 3-week test period, correlations are obtained indicating dependence on ethanol (Wahlstrom, 1987). When tested with 5% (data not shown), these correlations were 0.77 (Group 1), 0.90 (Group 2) and 0.96
580
G. WAHLSTROM and A. NORDBERG
GROUP 1 ^kg
UJ
GROUP 3
GROUP 2
A.
y
5 . r=0.97 b=0.77
y
y
OC
m
t3 •
