Animal models of self-administration are essential for understanding the neurochemical systems and pathways responsible for the initiation and maintenance of ethanol drinking (McBride and Li, 1998). Drinking procedures with rodents have shown that drugs from a number of pharmacological classes can attenuate ethanol self-administration. Opioid antagonists (Myers and Critcher, 1982; Myers and Lankford, 1986;Froehlich et al., 1990; June et al., 1998; Stromberg et al., 1998), serotonin agonists and antagonists (Silvestre et al., 1998; Wilson et al., 1998), γ-aminobutyric acid (GABA) agonists (Samson and Grant, 1985; Rassnick et al., 1993a; Hodge et al., 1996; Roberts et al., 1996) and dopamine agonists and antagonists (Rassnick et al., 1993b; McBride and Li, 1998) all reduce ethanol intake. However, many of the drugs that have shown promising results in rodents have not proven effective in humans (Garbutt et al., 1999; Wiesbeck et al., 1999).
One factor that may influence the clinical utility of a pretreatment drug is selectivity. Lack of selectivity (i.e. side effects) are a major factor in patient non-compliance and should be assessed in preclinical animal studies. In addition, a pretreatment drug that attenuates ethanol drinking in animals only at doses that also suppress other consummatory behaviours may be acting via mechanisms not directly related to ethanol's reinforcing effects. In contrast, a drug that suppresses ethanol drinking without altering other reinforced behaviour is more likely to be interacting selectively with ethanol reinforcement mechanisms. Multiple schedule procedures have proven useful for examining the selectivity of pretreatment drugs for reducing cocaine and heroin self-administration (Woolverton and Virus, 1989; Kleven and Woolverton, 1993; Mello et al., 1993). Multiple schedules typically employ alternating periods of drug and non-drug reinforcer availability to determine whether a pretreatment selectively reduces drug self-administration or non-selectively decreases responding for both drug and non-drug reinforcers. In a multiple schedule procedure it is also possible to match closely rates of drug and non-drug reinforced responding, in order to minimize rate-dependent effects that might interfere with the determination of pretreatment selectivity (Lucki, 1983; Sanger and Blackman, 1976).
Most multiple schedule studies have found that pretreatment drugs that reduce drug-reinforced responding are not selective, in that they decrease responding for an alternative reinforcer at the same doses that decrease drug self-administration (Woolverton and Virus, 1989; Kleven and Woolverton, 1993; Mello et al., 1993). However, at least three multiple schedule studies, one in rodents (Shoaib et al., 1998) and two in monkeys, have shown that drug self-administration can be selectively reduced compared with a comparison non-drug reinforcer (Glowa and Fantegrossi, 1997; Wojnicki et al., 1999). In the ethanol self-administration area, only three studies have examined the multiple schedule procedure and only two have used a multiple schedule procedure to examine the selectivity of pretreatment drugs. Slawecki et al. (1997) found that amphetamine, naloxone, morphine and haloperidol all reduced ethanol plus sucrose self-administration at lower doses than were required to reduce sucrose responding. However, the authors reported that none of the drugs tested was totally selective for reducing ethanol intake. In a similar study, Shelton and Balster (1997) found that ethanol itself, but not N -methyl-d-aspartate (NMDA) antagonists or GABA positive modulators, selectively reduced ethanol drinking compared with a saccharin control.
The physiology of primates is much closer to that of humans than is the physiology of rodents, making them a valuable tool for bridging the gap between rodents and humans in the field of drug abuse research (Higley and Bennett, 1999). Primate ethanol self-administration models have been used for many years to delineate successfully the conditions under which ethanol can serve as a reinforcer (Meisch, 1987). However, in comparison with rodents, relatively little research has focused on developing ethanol self-administration procedures in primates specifically designed to determine pretreatment drug selectivity and no studies have been published on multiple schedule models of ethanol self-administration in primates.
There were three primary goals of the present study. The first was to develop a multiple schedule procedure of ethanol self-administration in cynomolgus monkeys that would result in high levels of ethanol intake, producing blood ethanol concentrations above the legal limit for impaired driving in most of the USA of 80 mg/dl level. In rodent drinking procedures, blood ethanol concentrations seldom reach above 80 mg/dl, even when total oral ethanol doses of up to 1.5 g/kg are self-administered (Czachowski et al., 1999a). To attain this goal in the present study we chose to use a multiple schedule with four 15-minute access periods, two of which allowed access to ethanol and two of which allowed access to an alternative reinforcer. With this schedule it was hoped that the animals would have adequate access to ethanol to drink to satiation and for central nervous system (CNS) effects of ethanol to become manifest, while at the same time providing a sufficiently long session for the effect of pretreatment drugs to be assessed. In the first experiment, ethanol concentration was manipulated to determine empirically the ethanol concentration that produced the combination of the greatest volume of ethanol intake as well as the greatest total self-administered ethanol dose. In the second study, delivery of non-contingent food pellets under three fixed-time (FT) schedules were examined for their ability to produce enhancement of ethanol intake beyond that which could be generated by optimizing ethanol concentration. The second major goal of the study was to examine the use of a sugar-free Tang powder (General Foods) solution as an alternate reinforcer in the multiple schedule. Towards this goal, experiment three was conducted to determine if ethanol and Tang were self-administered at similar rates and patterns, and to ensure that both solutions were serving as reinforcers when compared with vehicle.
The third major goal of the study was to use the multiple schedule to examine the selectivity of a drug pretreatment for reducing ethanol intake compared with the Tang control. Based on the data from prior rodent and primate studies, it was hypothesized that ethanol itself would be the most likely to produce selective effects on ethanol self-administration (Karoly et al., 1978; Petry, 1995; Shelton and Balster, 1997). Doses of ethanol ranging from 0.25 to 1.5 g/kg were administered intragastrically (i.g.) both 30 and 60 min prior to the self-administration session and examined for their effect on subsequent ethanol and Tang self-administration. These i.g. ethanol doses and pretreatment times were chosen based on i.g. ethanol discrimination studies conducted in our laboratory showing that ethanol administration at both time points results in CNS-mediated discriminative stimulus effects (Grant et al., 2000). Throughout the study, blood ethanol concentrations (BECs) were measured to determine whether BEC levels generated in the self-administering monkeys would be above the 80 mg/dl level considered to be threshold for intoxicated driving in most of the USA.
Six adult cynomolgus monkeys (two male, four female) were used as subjects. The animals had previously undergone an ethanol self-administration induction procedure in which they had been adapted over the course of many weeks to drink increasing volumes of ethanol, culminating in access to 4% ethanol for 22 h/day for 6 months (Vivian et al., 2001). Upon removal from that study, the monkeys were individually housed in standard four-unit (1 m2/unit) primate cages with removable dividing panels. The female monkeys were pair-housed for approximately 2 h each day. A 12/12 h light–dark cycle was in effect for the duration of the study. The male animals weighed 6.55 and 7.95 kg at the onset of the study. The female monkeys weighed 3.32, 3.62, 3.54 and 3.68 kg. The animals were fed sufficient monkey chow and fruit daily, at least one hour after the end of the last experimental session, to maintain these weights. Water was freely available except during experimental sessions.
Self-administration sessions were conducted in a separate two-unit primate cage. One side wall of each cage was modified to accept a work panel. Each panel had two independent liquid delivery systems, consisting of a motor-driven retractable brass Lixit spout connected via 11 mm Tygon tubing to a 2000 ml Nalgene bottle. The Lixit only allowed liquid flow when a central pin in the fluid spout was deflected. Each bottle was situated on an electronic balance (Ohaus) which measured fluid consumed by weight displacement. Three 2.8 W stimulus lamps (red, amber, green) were located 12 cm above each spout. Between the two spouts was a recessed receptacle with a white stimulus lamp and a horizontal dowel manipulandum. Schedule conditions and data recording were accomplished by Macintosh 7600 computer connected to a National Instruments interface and controlled by LabView instrumentation control software.
General self-administration training and testing
Monkeys were tested 5 days/week (Monday–Friday). On Saturday, a bottle containing 4% w/v (weight/volume) ethanol was hung on each animal's homecage for 1 h. On Sunday, the animals did not receive access to ethanol. Experimental sessions were 60 min in duration and consisted of four 15-minute components. In experiments 1 and 2 water was available during the first and third 15-minute components and ethanol was available during the second and fourth 15-minute components from the opposite spout (i.e. water, ethanol, water, ethanol). The ethanol and water spouts were counterbalanced across animals. In experiments 2, 3 and 4, a 4% w/v sugar-free Tang powder (General Foods) solution was available to four of the monkeys during the first and third 15-minute components from one of the two spouts in alternation with 4% ethanol. The fifth monkey had access to 6% w/v Tang in alternation with ethanol. The availability of liquid from one spout (right/left) was signalled by the illumination of the amber stimulus lamp above that spout. A 1 g banana-flavoured food pellet (P.J. Noyes) was also delivered non-contingently to each animal at the start of each component to signal component changes and facilitate drinking in experiments 3 and 4. Food pellet delivery schedule was varied in experiment 2.
The available solution was accessed by the monkey inserting its hand into the recess between the two levers and pulling the dowel. Closure of a switch resulted in the illumination of the white and green stimulus lamps above the active spout and the insertion of that spout into the cage for the duration of the dowel pull. Release of the dowel resulted in the green and white stimulus lamp being extinguished and the retraction of the spout.
Experiment 1: Effect of increasing ethanol concentration on ethanol and water self-administration
To examine the effect of increasing ethanol concentration availability on ethanol intake, four monkeys (4991, 4992, 4993 and 4996) were allowed to self-administer ethanol concentrations of 2%, 4%, 6% or 8% in alternation with water under the multiple schedule. Ethanol concentration was changed when two criteria were met. First, a minimum of six session were conducted and, second, the standard deviation of ethanol intake for a monkey during the last three sessions was less than 20% of the mean self-administered ethanol intake during those three sessions. This stability criteria was also used in experiments 2 and 3. Blood samples for BEC determination were collected immediately following the end of the ethanol self-administration session from every animal on one day within the six-day period at each ethanol concentration. Samples were analysed for BECs as detailed in the section ‘Blood ethanol concentration analysis’ below.
Experiment 2: Effect of pellet FT schedule on ethanol and water self-administration
Four monkeys (4991, 4992, 4993 and 4996) were allowed to self-administer 4% ethanol and water under the multiple schedule. The 4% ethanol concentration was chosen because it resulted in the highest mean ethanol intake in volume and g/kg. In successive blocks of at least six sessions, or until stability was reached, the effect of food pellet delivery under a range of FT schedules was examined. Food pellets (1 g) were presented non-contingently in the following sequential order: no pellets, FT 180 s, FT 300 s and FT 900 s.
Experiment 3: Effect of water substitution for 4% ethanol and/or Tang on self-administration in the multiple schedule
The monkeys were allowed to self-administer 4% ethanol and Tang under the multiple schedule with a FT 900 s schedule of food pellet delivery until behaviour was stable. The effect of replacing ethanol and Tang with water was then examined to determine if ethanol and Tang were serving as reinforcers. This experiment also served to assess whether the self-administration of one of the reinforcers could be altered without affecting self-administration of the second reinforcer. In the first condition, the Tang solution was replaced with water for blocks of at least six sessions or until stable. Subsequently ethanol was replaced with water, and then both ethanol and Tang were replaced with water for at least six sessions or until stability was reached. A minimum of six baseline 4% ethanol and Tang self-administration sessions were conducted between each water substitution test block.
Experiment 4: Effect of acute and chronic ethanol pretreatment on ethanol and Tang self-administration
To examine the consequences of non-contingent ethanol administration on subsequent ethanol and Tang self-administration, increasing doses of ethanol (0.25, 0.5, 0.75, 1 and 1.5 g/kg) were administered by i.g. gavage to three female (4991, 4992, 4993) and two male (4996, 5498) monkeys prior to daily self-administration sessions. Each ethanol dose was tested twice, with a minimum of 2 days between ethanol pretreatment doses. Ethanol pretreatment times of 30 and 60 min were examined. After the determination of the acute dose–effect curves, each of the monkeys was administered 1 g/kg ethanol (60 min pretreatment) prior to 15 consecutive self-administration sessions. Ethanol gavage was accomplished by first seating the monkey in a primate restraint chair (Plaslabs, Lansing, MI, USA). A 5 French infant feeding tube was then fed into the nostril, passed down the oesophagus and into the stomach. The ethanol solution was subsequently infused directly into the stomach and the feeding tube was removed. The monkey was then placed into the test chamber and the session started after the appropriate (30 or 60 min) time-out interval.
Blood ethanol concentration analysis
Blood samples (20 μl) were collected from the saphenous vein. The blood samples were sealed in air-tight vials containing 500 μl of distilled water and 20 μl of 200 mg/dl isopropanol internal standard. Samples were stored at −4°C until analysed. BECs were analysed using a Hewlett Packard 5890 Series II gas chromatograph with flame ionizing detector, headspace autosampler and Hewlett Packard 3392A integrator.
The ethanol drinking solution was prepared from 95% ethanol w/v diluted with tap water to a 4% w/v concentration. Sugar-free Tang powder (General Foods) was diluted in distilled water to produce a 4% w/v (monkeys 4991, 4992, 4993, 5498) or 6% w/v (monkey 4996) solution. Gavaged ethanol solutions were prepared by diluting 95% ethanol with tap water to a 20% w/v ethanol solution.
Total ethanol and water self-administration (experiments 1 and 2) or ethanol and Tang self-administration (experiments 3 and 4) per session were collected for individual monkeys in all experiments. Total ethanol intake in g/kg (± SEM) and ml/kg (± SEM) were calculated for individual animals and for the group. Water and Tang intake were converted to ml/kg to control for differences in monkey weights. In experiment 1, mean (± SEM) ethanol intake (ml/kg) and total self-administered ethanol dose (g/kg) were calculated for the last 6 days at each ethanol concentration for individual monkeys and for the group. The data were subjected to individual one-way, repeated measures analysis of variance (ANOVA) using ethanol concentration as the factor, to determine if self-administered ethanol concentration significantly changed ethanol intake. In experiment 2, the ethanol and water self-administration data were analysed using separate one-way, repeated measures ANOVAs to determine if pellet delivery frequency influenced ethanol and water intake significantly. Tukey post hoc tests were performed to assess differences between individual pellet intervals when the main effect reached statistical significance (P < 0.05). In experiment 3, group mean (± SEM) liquid intake (ml/kg) was calculated for the final 6 days at each water substitution condition. Individual paired t -tests (StatView 4.5, Abacus Concepts, Berkeley, CA, USA) were performed to compare ethanol and Tang intake under control conditions to liquid intake when water was substituted for the individual solutions (4% ethanol or Tang). Bonferroni corrections were used to control for multiple comparisons. In experiment 4, ethanol pretreatment tests were conducted no more than once every third day, provided the SEM of the previous three non-drug days was no greater than 20% of the mean ethanol intake for those three days. Individual one-way repeated measures ANOVAs were performed on ethanol and Tang intake to determine if acute ethanol gavage decreased ethanol and Tang intake. Significant main effects were further analysed using Dunnett's multiple comparison tests to assess differences in individual ethanol pretreatment doses from the water pretreatment control. To determine if chronic 1.0 g/kg ethanol gavage altered ethanol or Tang self-administration, 5 days of water gavage and 15 subsequent days of ethanol pretreatment, divided into three five-day blocks were subjected to a one-way repeated measures ANOVA. Significant main effects were analysed using Dunnett's multiple comparison tests to assess if individual five-day blocks of ethanol pretreatment differed significantly from the water pretreatment control period.
Experiment 1: Effect of increasing ethanol concentration on ethanol and water self-administration
Figure 1 shows the effect of increasing ethanol concentrations (2%, 4%, 6% and 8%) on total ethanol intake for the group of four monkeys as well as for each individual monkeys. The upper left panel of Figure 1 shows mean group ethanol intake (ml/kg) as a function of increasing ethanol concentration. There was no significant main effect of ethanol concentration on group ethanol intake. However, there was a general negative relationship between self-administered ethanol concentration and mean ethanol intake in ml/kg from 4% to 8%. At the 4% ethanol concentration, mean group ethanol intake was 26.23 ± 5.98 ml/kg). At the 8% ethanol concentration, mean group ethanol intake was 11.93 ± 1.86 ml/kg, but this decrease in intake was not statistically significant. There were substantial individual differences among monkeys (Figure 1, upper right four panels). Monkeys 4991 and 4992 decreased ethanol intake volumes as ethanol concentration increased. Ethanol intake volumes were highest at the 4% concentration for monkeys 4993 and 4996.
The lower left panel of Figure 1 shows the group mean of total self-administered ethanol dose (g/kg) as ethanol concentration was increased. There was no significant main effect of ethanol concentration on total group self-administered ethanol dose. Mean group self-administered ethanol dose at the 2% concentration was 0.49 ± 0.25 g/kg. Mean group ethanol intake doubled to 1.05 ± 0.24 g/kg when the ethanol concentration was increased to 4% and remained at similar levels at the 6% and 8% ethanol concentrations. BECs taken immediately after the self-administration session paralleled ethanol intake (Figure 1, lower left panel) with a mean group BEC of 23.88 ± 9.74 mg/dl at the 2% ethanol concentration, reaching a mean group BEC of 89.13 ± 16.40 mg/dl at the 4% ethanol concentration. The four panels on the lower right of Figure 1 show the effect of increasing ethanol concentration on total self-administered ethanol dose for each of the four monkeys. Two of the four monkeys (4991, 4993) showed little change in total self-administered ethanol dose between 4% and 8%. Monkey 4992 showed decreases in self-administered ethanol dose at doses of 6% and 8% ethanol compared with 4% ethanol. Monkey 4996 showed increases in total self-administered ethanol dose at successively higher ethanol concentrations.
Experiment 2: Effect of pellet ET schedule on ethanol and water self-administration
Since maximal ethanol intake was generated by 4% ethanol, that concentration was chosen as the basis for the subsequent experiments. In experiment 2, water was available in alternation with ethanol under the multiple schedule. The upper left panel of Figure 2 shows the effect of increased frequency of non-contingent food pellet delivery on 4% ethanol self-administration. There was a significant main effect [F(*,**) = 49.5, P < 0.001] of pellet frequency on ethanol intake. When no pellets were delivered a group mean of 26.23 ± 5.98 ml/kg of ethanol was self-administered. At the shortest pellet delivery interval of FT 180 s a group mean of 39.45 ± 5.40 ml/kg of ethanol was consumed. Pellet intervals of FT 300 s and FT 180 s resulted in significantly greater group mean ethanol intake (P < 0.05) than when no food pellets were delivered. When ethanol intake is plotted for individual monkeys as a function of food pellet interval (Figure 2, upper right panels), mean ethanol intake increased in all four animals, although the magnitude of the increase varied among individual monkeys. The lower left panel of Figure 2 shows group mean water intake in the multiple schedule as pellet delivery frequency was increased. Group mean water intake was very low (1.49 ± 1.03 ml/kg) under conditions in which no food pellets were delivered. As pellet frequency increased, group mean water intake was greater, but this effect did not reach statistically significance [F(*,**) = 1.87, NS]. When examined individually, two of the subjects (4991, 4992) showed large increases in water intake. In particular, 4992 increased water intake from 3.29 ml/kg when no pellets were delivered to 92.75 ml/kg at FT 180 s. The other two monkeys (4993, 4996) showed almost no water intake at all of the pellet FT intervals. At the pellet interval of FT 900 s, and when no pellets were delivered, ethanol served as a reinforcer when compared with water for all four monkeys (Figure 2, right panels).
Experiment 3: Effect of water substitution for 4% ethanol and/or Tang on self-administration in the multiple schedule
Figure 3 shows the effect of replacing first Tang, then 4% ethanol, then both Tang and 4% ethanol with water. The first two bars show baseline group mean 4% ethanol and Tang intake. Under baseline conditions, group mean 4% ethanol intake was 44.97 ± 5.06 ml/kg and Tang intake was 57.45 ± 6.65 ml/kg. Replacing Tang with water resulted in a significant reduction (P < 0.05) in liquid intake from the Tang spout compared with the Tang baseline, without any significant change in ethanol intake. Similarly, replacing 4% ethanol with water resulted in a significant reduction in liquid intake from the 4% ethanol spout compared with the previous 4% ethanol baseline, without significantly altering Tang self-administration. Replacing both 4% ethanol and Tang with water resulted in a significant reduction (P < 0.05) in liquid intake from both spouts.
Experiment 4: Effect of ethanol pretreatment on ethanol and Tang self-administration
Under water gavage pretreatment conditions, the patterns of ethanol and Tang self-administration under the multiple schedule were similar across 15-minute components of availability (Table 1). More than 75% of total 4% ethanol and Tang were self-administered during the first Tang and ethanol components. A much smaller amount of ethanol and Tang were self-administered in the second two 15-minute components. The results of increasing doses of non-contingent i.g. ethanol administered 30 min prior to ethanol and Tang self-administration are shown in Figure 4. Under i.g. water pretreatment baseline conditions, a group mean of 1.29 ± 0.14 g/kg ethanol was self-administered (upper panel). As the pretreatment dose of ethanol increased, group mean 4% ethanol self-administration did not significantly decrease [F(*,**) = 2.60, P = 0.057]. Individually, four of five monkeys showed lower ethanol intakes following the 1.5 g/kg ethanol pretreatment than following water pretreatment (Table 2). The lower panel of Figure 4 shows the effect of i.g. ethanol pretreatment on Tang self-administration. A group mean of 38.85 ml/kg of Tang was self-administered under 30-minute water pretreatment baseline conditions and there was no significant main effect of 30-minute pretreatment of i.g. ethanol on Tang self-administration [F(*,**) = 0.67, NS]. Individual monkeys showed a wide range of Tang intake at i.g. water baseline, ranging from 12.89 ± 8.05 ml/kg for monkey 4996, to 75.45 ± 3.84 ml/kg for monkey 5498 (Table 2). When examined individually, none of the monkeys showed dose-dependent changes in Tang intake as a function of non-contingent i.g. ethanol.
Figure 5 shows the results of 60minute pretreatment with increasing doses of non-contingent i.g. ethanol on group mean 4% ethanol and Tang self-administration. Mean BECs immediately before the self-administration session are also shown, as well as mean BECs measured after the end of the self-administration session. Mean 4% ethanol intake in the i.g. water baseline condition was 1.66 ± 0.18 g/kg, compared with the 1.29 ± 0.14 g/kg baseline intake in the 30-minute ethanol pretreatment condition (upper panel). There was a significant main effect of ethanol dose on ethanol intake [F(*,**) = 27.53, P < 0.01]. However, the reduction in ethanol intake was not dose-dependent and post hoc analysis indicated that only the 1 g/kg ethanol dose significantly reduced ethanol intake by 24% compared with the water control. Mean BEC before the session varied directly with i.g. ethanol pretreatment dose. Mean BEC increased from a low of 11.13 ± 1.49 mg/dl at the 0.25 mg/kg ethanol pretreatment dose, to a high of 120.25 ± 7.88 mg/dl at the 1.5 g/kg i.g. ethanol dose (Figure 5, upper panel). Mean BECs measured immediately following self-administration also increased as i.g. pretreatment dose increased (Figure 5, upper panel). The greatest mean BEC of 199.50 ± 9.97 mg/dl was obtained immediately following the self-administration sessions in which the 1.5 g/kg i.g. ethanol pretreatment dose was tested. Individually, monkeys 4991, 4996 and 4993 showed the greatest reductions in ethanol intake (Table 3). Monkeys 5498 and 4993 showed no decreases in ethanol self-administration at any ethanol pretreatment doses. There was no significant effect of 60-minute pretreatments of i.g. ethanol on Tang self-administration (lower panel). Four of five monkeys (4991, 4992, 4993 and 4996) showed decreases in Tang self-administration at more than one i.g. ethanol pretreatment dose, but none of the animals showed consistent dose-dependent reductions in Tang self-administration (Table 3).
Table 4 shows the results of five consecutive days of i.g. water gavage at a 60minute pretreatment time, 15 consecutive days of 1.0 g/kg i.g. ethanol gavage and an additional 5 days of no gavage post-treatment self-administration days of 4% ethanol and Tang self-administration (n = 5 monkeys). A group mean of 48 ± 3 ml/kg of 4% ethanol and 61 ± 8 ml/kg of Tang were self-administered following i.g. water pretreatment. Repeated, daily 1.0 g/kg ethanol gavage significantly reduced ethanol self-administration by 23% [F(*,**) = 6.56, P < 0.001]. Post hoc analysis indicated that ethanol self-administration was significantly (P < 0.05) decreased following the second and third consecutive five-day block of ethanol pretreatment. Mean BECs 60 min after the 1.0 g/kg i.g. ethanol pretreatment were stable across each five-day period with a mean of 99 ± 5 mg/dl, 103 ± 6 mg/dl and 97 ± 6 mg/dl for days 1–5, 6–10 and 11–15, respectively. Mean BEC assessed following 1 g/kg ethanol pretreatment combined with the self-administered ethanol was 160 ± 8 mg/dl, 134 ± 6 mg/dl and 139 ± 6 mg/dl in days 1–5, 6–10 and 11–15 of ethanol pretreatment, respectively. Unlike the effect of i.g. ethanol pretreatment on ethanol self-administration, there was no significant effect of daily, repeated 1.0 g/kg i.g. ethanol pretreatment on Tang self-administration [F(*,**) = 0.47, NS]. Figure 6 shows mean group 4% ethanol and Tang intake for each day of the chronic ethanol pretreatment study. Mean ethanol intake was stable during the initial water baseline and then decreased over the course of the first three ethanol treatment days. Mean ethanol intake remained stable for the remainder of ethanol pretreatment, but quickly returned to initial water baseline levels following the discontinuation of ethanol pretreatment. Tang intake showed no trends before, during or after i.g. ethanol pretreatment.
The results from the present study show that a multiple schedule of ethanol self-administration can be successfully established in both male and female cynomolgus monkeys. Under the multiple schedule, monkeys self-administered over 1 g/kg of ethanol when ethanol availability was restricted to two 15-minute access periods. When ethanol consumption was measured as a function of ethanol concentration available, there was no significant change in self-administration, although the greatest mean ethanol intake, in terms of volume consumed and total dose self-administered, occurred at the 4% ethanol concentration. Self-administered ethanol produced BECs in excess of 80 mg/dl at the 4% ethanol concentration as measured immediately after the session. A BEC of 80 mg/dl in humans is the legal level for intoxicated driving in most of the USA. When the ethanol concentration was increased to 6% and then 8%, total self-administered ethanol volume decreased, but total self-administered ethanol dose remained stable. These results are consistent with other oral ethanol self-administration studies using concurrent ethanol and water access, showing that the greatest volumes of ethanol are self-administered at low concentrations and successively higher self-administered ethanol concentration result in decreased liquid intake (Henningfield and Meisch, 1978, 1979). The present results are also similar to rodent studies, showing that the greatest ethanol preference in rats occurs at ethanol concentrations in the 4% range (Richter and Campbell, 1940). Ethanol intake varied among animals from 1 g/kg to 2 g/kg ethanol, but drinking behaviour for individual monkeys was very stable across sessions. There were insufficient subjects to assess gender differences in ethanol self-administration, although there appeared to be no distinct trends in intake. Indeed, of the two highest drinkers, one was male (5498) and the other was female (4992).
Intermittent presentation of food pellets to animals allowed access to a liquid has been shown to result in greatly enhanced drinking (Falk and Tang, 1988). This procedure, known as schedule-induced polydipsia, has been used in both rodents and primates to induce ethanol consumption (Holman and Meyers, 1968; Mello and Mendelson, 1971; Falk et al., 1972). The ability of schedule induction to increase ethanol intake under a multiple schedule of ethanol and water self-administration was examined in the present study by presenting food pellets at FT intervals of 900, 300 and 180 s. The results showed that as a group, increases in the frequency of food pellet delivery significantly enhanced ethanol consumption (Figure 2). When assessed on an individual animal basis, all four monkeys showed increased ethanol intake at one or more FT interval. Although the group data showed that there was no significant effect of food pellet presentation on water delivery, two of the four monkeys (4991, 4992) showed greatly enhanced water intake at shorter pellet delivery intervals. In fact, water intake increased to a much greater degree than did ethanol intake in these two monkeys. These results are in contrast to the findings in rodents in which 5% ethanol is self-administered in excess of water under schedule induction (Tang and Falk, 1977). A number of factors could be responsible for these differences, including FT intervals, species of subjects, assessment of individual subject effects versus group means, or the multiple schedule itself. Drug intake in excess of vehicle is the most accepted method of determining if that drug is serving as a reinforcer, suggesting that ethanol was no longer serving as a reinforcer for two of the monkeys when drinking was schedule induced. As a consequence, the pellet delivery interval was decreased to FT 900 s, resulting in the delivery of one food pellet at the onset of each of the four self-administration components. This allowed pellet delivery to serve as a discriminative stimulus for component changes, without significantly altering ethanol or water self-administration.
At least three studies, two examining ethanol, have shown that multiple schedules are relatively free from reinforcer interactions (Shoaib et al., 1998; Slawecki et al., 1998;Czachowski et al., 1999b). The present study examined if 4% ethanol and Tang self-administration were also free from reinforcer interactions (e.g. independent) by sequentially replacing each of the reinforcers with water and determining if that manipulation altered self-administration of the remaining reinforcer (Figure 3). The results showed that replacing Tang with water did not significantly alter 4% ethanol self-administration. Likewise, although replacing ethanol with water did increase mean intake of Tang, the increase in Tang consumption was not significant. Replacement of ethanol and Tang with water resulted in self-administration declining significantly compared with the ethanol/Tang baseline. These results indicate that 4% ethanol and Tang self-administration were indeed independent and that 4% ethanol and Tang were serving as reinforcers under the multiple schedule. Ethanol and Tang were self-administered in similar volumes and in similar temporal patterns, but over 75% of total fluid intake was in the first two components and much smaller liquid volumes self-administered in the latter two components. This suggests that our choice of a four-component multiple schedule of liquid delivery may offer no distinct advantage over two components. One possible theoretical advantage of the second two components in future studies would be to determine if any reinitiation of drinking occurred following an initial suppression of behaviour by a pretreatment drug.
The primary goal of the present study was to determine if a multiple schedule of ethanol and Tang self-administration would provide a useful baseline for examining the selectivity of pretreatment drug effects. It might be predicted that the most effective drug for reducing ethanol self-administration would be ethanol itself. A prior multiple schedule study in rats found that of the drugs tested, only i.p. ethanol pretreatment could reduce ethanol drinking selectively (Shelton and Balster, 1997). Non-contingent ethanol has also been shown to reduce ethanol self-administration in a concurrent ethanol and sucrose access paradigm in rats (Petry, 1995) and in rhesus monkeys self-administering ethanol intravenously (Karoly et al., 1978). We examined whether ethanol self-administration could be attenuated by non-contingent administration of ethanol by administering i.g. ethanol prior to the self-administration session. In the present study, ethanol self-administration was not significantly reduced by i.g. ethanol administered 30 min prior to the self-administration session, although BECs at the start of the self-administration session were substantial. Specifically, the mean BEC, 30 min after 1.0 g/kg i.g. ethanol was 60.4 ± 9.89 mg/dl. In contrast, 1.0 g/kg i.g. ethanol administered 60 min prior to the self-administration significantly reduced ethanol self-administration. This effect, while statistically significant, may not be meaningful since there was no evidence of dose-responsiveness in the 60-minute ethanol pretreatment dose–effect curve, nor did the higher 1.5 g/kg ethanol pretreatment dose suppress ethanol drinking. It is also unlikely that the increased BEC level at the start of the self-administration session in the 60-minute ethanol pretreatment condition (82.75 ± 7.09 mg/dl) could account for the significant effect. Similar BECs were attained following 30 min pretreatment with 1.5 g/kg ethanol, which did not significantly reduce ethanol self-administration. Tang self-administration was not significantly altered by non-contingent ethanol administered at either the 30 or 60 min pretreatment time.
Doses of ethanol higher than those tested may have been necessary to suppress ethanol and/or Tang drinking significantly, but technical factors prohibited testing these doses. Specifically, the i.g. gavage volume at the 1.5 g/kg ethanol dose was quite substantial and the volume of 20% ethanol necessary to administer even greater doses were prohibitively large for a single bolus administration. Technical issues notwithstanding, it is unlikely that greater ethanol pretreatment doses would have been effective for a number of reasons. First, the highest dose of 1.5 g/kg of i.g. ethanol resulted in a mean BEC of 120 mg/dl, well above the BEC level that typically was reached following intervening baseline self-administration sessions. When combined with the ethanol self-administered during the session, BEC levels increased to 199.5 mg/dl. Secondly, the animals were showing overt signs of impairment following sessions in which the higher i.g. ethanol doses were administered, leading us to conclude that even greater doses, could they have been tested, would have been likely to reduce rates through non-selective motor impairment. Lastly, we have successfully trained an ethanol discrimination in cynomolgus monkeys using 1 g/kg i.g. ethanol administered at both 30 and 60 min presession, indicating that these doses and pretreatment times are sufficient to produce CNS-mediated discriminative stimulus effects of ethanol (Grant et al., 2000). Taken together, these findings would suggest that sufficient ethanol was present at the start of the self-administration session to produce CNS effects similar to those occurring after control ethanol and Tang self-administration sessions.
It is possible that the lack of an acute ethanol pretreatment effect on ethanol self-administration was due to insufficient association of the CNS consequences of combined non-contingent and contingent ethanol administration. In the present study, the highest mean BEC and greatest pharmacological effects of combined non-contingent and self-administered ethanol were likely to have occurred well after the end of the self-administration session (Green et al., 1999). Consequently, the majority of the feedback from the CNS effects of the combined ethanol dose could only be manifested in behavioural changes in subsequent self-administration sessions. In the prior rodent study showing decreases in ethanol self-administration by non-contingent ethanol, ethanol was administered non-contingently for six consecutive days (Shelton and Balster, 1997). The primate study in the literature, showing attenuation of ethanol self-administration by non-contingent ethanol, employed intravenous ethanol self-administration which was likely to have produced peak BECs and CNS effects rapidly enough to have affected ongoing behaviour (Karoly et al., 1978). A second possibility was that the acute pharmacological effects of ethanol were insufficient to overcome the behavioural control over drinking generated by the multiple schedule itself.
Pretreatment drugs used to suppress drug self-administration are generally administered chronically and chronic drug effects may produce a distinctly different pattern of results in self-administration studies than the same drugs given acutely (for review, see Mello and Negus, 1996). To examine the possibility that repeated i.g. ethanol administration was necessary to attenuate ethanol drinking, we administered 1.0 g/kg i.g. ethanol for 15 consecutive days. Chronic non-contingent ethanol did not significantly reduce ethanol intake until days 6–10 of daily administration (Table 4). Days 11–15 of non-contingent ethanol gavage resulted in slightly greater reductions in ethanol drinking than days 6–10. In contrast to ethanol self-administration, Tang drinking was not altered following non-contingent ethanol gavage, indicating that the reduction of ethanol drinking by non-contingent ethanol was selective (Table 4). These results suggest that substantial experience with non-contingent ethanol is necessary before monkeys will begin to make compensatory reductions in their own self-administered ethanol dose. The data from the chronic pretreatment study are at odds with our acute ethanol pretreatment data, showing that a 60-minute pretreatment with 1 g/kg i.g. ethanol resulted in a significant reduction in ethanol self-administration. To address this contradiction, we compared the final day of water pretreatment in the chronic study with the first day of ethanol pretreatment in the chronic study. These data should be similar to the previous acute data at the 1.0 g/kg i.g. dose. The final day of water gavage in the chronic ethanol experiment resulted in a total self-administered ethanol volume of 49.6 ± 5.4 ml/kg. The first day of 1.0 g/kg ethanol pretreatment resulted in a total self-administered ethanol volume of 44.5 ± 6.2 ml/kg. These results suggest, as we have previously stated, that our findings at the 1 g/kg ethanol dose in the acute condition, while statistically significant, are not reliable given their lack of dose-responsiveness.
Interestingly, despite 15 days of experience with daily, non-contingent 1 g/kg i.g. ethanol, the monkeys in the present study did not reduce their ethanol self-administration sufficiently to produce BEC levels comparable with those achieved following self-administration alone (Table 3). These finding parallel the results of non-contingent ethanol injections in rats, in that ethanol doses well in excess of those self-administered were required to reduce ethanol drinking significantly (Shelton and Balster, 1997). The present results suggest that ethanol self-administration only partially controlled the pharmacological effects of ethanol. It is possible, even likely, that ethanol self-administration is to some degree controlled by other factors such as the taste of ethanol. Since the taste of ethanol would not be affected by non-contingent ethanol gavage, drinking would be unlikely to be altered by pretreatment with i.g. ethanol. Although this hypothesis could partially explain the present findings, they do not fully explain why non-contingent ethanol administration became more effective over repeated daily administrations. Nor can taste be used as an explanation for the data from other studies showing that non-contingent delivery of intravenously self-administered drugs which have no taste, such as cocaine, produce finding similar to those in the present study. For instance, constant non-contingent infusion of cocaine under some conditions will selectively decrease cocaine self-administration relative to food (Glowa and Fantegrossi, 1997), but under other conditions has no effect, or even increases in cocaine self-administration (Glowa and Fantegrossi, 1997; Panlilio et al., 1998). Taken as a whole, these findings would suggest that there is a complex interaction between pharmacological and behavioural factors underlying the effects of non-contingent drug administration on later drug self-administration behaviour. Factors such as self-administered unit dose have been shown to play an important role in studies with cocaine, but additional studies will be necessary to understand these interactions with ethanol more fully (Glowa and Fantegrossi, 1997).
In summary, these findings show that a multiple schedule of ethanol and Tang self-administration can be successfully trained in cynomolgus monkeys and that the monkeys drink sufficient ethanol to produce blood ethanol concentrations above 80 mg/dl. Ethanol intakes ranged from four to eight standard drinks per day and were consumed over an extremely short period of time. Ethanol and Tang both served as reinforcers, and self-administration of the ethanol and Tang were independent and occurred at similar rates. Non-contingent acute administration of ethanol, across a wide dose range, did not attenuate either ethanol or Tang self-administration, despite producing BECs of near 200 mg/dl. On the other hand, repeated daily pretreatment with ethanol selectively reduced ethanol drinking by 23% without affecting Tang self-administration, suggesting that chronic pretreatment may be necessary to show drug effects under a multiple schedule procedure. Additional studies are ongoing to assess the selectivity and efficacy of other pretreatment drugs using the multiple schedule procedure.
In 1st sentence of “Introduction” McBride and Li, 1998 not in refs. Further down same para, Roberts et al., 1996 also not in refs.
In the “Introduction” section, in the para starting “There were three primary goals of the present study.”, does abbreviation FT = fixed-time?
In “Discussion” section, in para starting “At least three studies, two examining ethanol, …”, should Slawecki et al., 1998 be Slawecki et al., 1997? And in para starting “The primary goal of the present study…”, should Koroly et al., 1978 be Karoly et al., 1978 as in refs?
In “Methods/subjects” section, should “1 meter/square/unit” be “1 m2/unit” or “1 m/square/unit”?
In “Results/Experiment 2”, editor had queried degrees of freedom – please add: “[F(x,xx) = 49.5, P < 0.001]”
In the “Discussion” section in the para starting “It is possible that the lack of an acute ethanol pretreatment effect …”, is the following sentence, with added commas, OK?:
“The primate study in the literature, showing attenuation of ethanol self-administration by non-contingent ethanol, employed intravenous ethanol self-administration which was likely to have produced peak BECs and CNS effects rapidly enough to have affected ongoing behaviour (Karoly et al., 1978).”
Samson HH, Grant KA (1985). Chlordiazepoxide effects on ethanol self-administration: dependence on concurrent conditions. J Exp Ana Beh 43:353–364. Shopuld this be J Exp Behav?
Wilson AW, Neill JC, Costall B (1998). An investigation into the effects of 5-HT agonists and receptor antagonists on ethanol self-administration in the rat. Alcohol 16:249–470. Are you sure page range is correct? Seems like a very long paper.
This work was supported by NIAAA grants AA11997 and AA10254.
1. Czachowski CL, Samson HH, Denning CE ( 1999). Blood ethanol
concentrations in rats drinking sucrose/ethanol
solutions. Alcohol Clin Exp Res 23:1331–1335.
2. Czachowski CL, Samson HH, Denning CE ( 1999). Independent ethanol
- and sucrose-maintained responding on a multiple schedule
of reinforcement. Alcohol Clin Exp Res 23:398–403.
3. Falk JL, Tang M ( 1988). What schedule-induced polydipsia can tell us about alcoholism. Alcohol Clin Exp Res 12:577–585.
4. Falk JL, Samson HH, Winger G ( 1972). Behavioral maintenance of high concentrations of blood ethanol
and physical dependence in the rat. Science 177:811–813.
5. Froehlich JC, Harts J, Lumeng L, Li TK ( 1990). Naloxone attenuates voluntary ethanol
intake in rats selectively bred for high ethanol
preference. Pharmacol Biochem Behav 35:385–390.
6. Garbutt JE, West SL, Carey TS, Lohr KN, Crews FT ( 1999). Pharmacological treatment of alcohol dependence: a review of the evidence. JAMA 281:1318–1325.
7. Glowa JR, Fantegrossi WE ( 1997). Effects of dopaminergic drugs on food- and cocaine-maintained responding. IV: Continuous cocaine infusions. Drug Alcohol Depend 45:71–79.
8. Grant KA, Waters CA, Green-Jordan K, Azarov A, Szeliga KT ( 2000). Characterization of the discriminative stimulus effects of GABA(A) receptor ligands in Macaca fascicularis
monkeys under different ethanol
training conditions. Psychopharmacology 152:181–188.
9. Green KL, Szeliga KT, Bowen CA, Kautz MA, Azarov AV, Grant KA ( 1999). Comparison of ethanol
metabolism in male and female cynomolgus macaques (Macaca fascicularis
). Alcohol Clin Exp Res 23:611–616.
10. Henningfield JE, Meisch RA ( 1978). Ethanol
drinking by rhesus monkeys as a function of concentration. Psychopharmacology 57:133–136.
11. Henningfield JE, Meisch RA ( 1979). Ethanol
drinking by rhesus monkeys with concurrent access to water. Pharmacol Biochem Behav 10:777–782.
12. Higley JD, Bennett AJ ( 1999). Central nervous system serotonin and personality as variables contributing to excessive alcohol consumption in non-human primates. Alcohol Alcohol 34:402–418.
13. Hodge CW, Haraguchi M, Chappelle AM, Samson HH ( 1996). Effects of ventral tegmental microinjections of the GABAA agonist muscimol on self-administration
and sucrose. Pharmacol Biochem Behav 53:971–977.
14. Holman RB, Meyers RD ( 1968). Ethanol
consumption under conditions of psychogenic polydipsia. Physiol Behav 3:369–371.
15. June HL, Cason CR, Chen SH, Lewis MJ ( 1998). Buprenorphine alters ethanol self-administration
in rats: dose–response and time-dependent effects. Psychopharmacology 140:29–37.
16. Kleven MS, Woolverton WL ( 1993). Effects of three monoamine uptake inhibitors on behavior maintained by cocaine or food presentation in rhesus monkeys. Drug Alcohol Depend 31:149–158.
17. Karoly AJ, Winger G, Ikomi F, Woods JH ( 1978). The reinforcing property of ethanol
in the rhesus monkey
. II. Some variables related to the maintenance of intravenous ethanol
-reinforced responding. Psychopharmacology 58:19–25.
18. Lucki I ( 1983). Rate-dependent effects of amphetamine on responding under random-interval schedules of reinforcement in the rat. Pharmacol Biochem Behav 18:195–201.
19. Meisch RA ( 1987). Factors controlling drug reinforced behavior. Pharmacol Biochem Behav 27:367–371.
20. Mello NK, Mendelson JH ( 1971). Evaluation of a polydipsia technique to induce alcohol consumption in monkeys. Physiol Behav 7:827–836.
21. Mello NK, Negus SS ( 1996). Preclinical evaluation of pharmacotherapies for treatment of cocaine and opioid abuse using drug self-administration
procedures. Neuropsychopharmacology 14:375–424.
22. Mello NK, Kamien JB, Lukas SE, Drieze J, Mendelson JH ( 1993). The effects of nalbuphine and butorphanol treatment on cocaine and food self-administration
in Rhesus monkeys. Neuropsychopharmacology 8:44–55.
23. Myers RD, Critcher EC ( 1982). Naloxone alters alcohol drinking induced in the rat by tetrahydropapaveroline (THP) infused ICV. Pharmacol Biochem Behav 16:827–836.
24. Myers RD, Lankford MF ( 1996). Suppression of alcohol preference in high alcohol drinking rats: efficacy of amperozide versus naltrexone. Neuropsychopharmacology 14:139–149.
25. Panlilio LV, Goldberg SR, Gilman JP, Jufer R, Cone EJ, Schindler CW ( 1998). Effects of delivery rate and non-contingent infusion of cocaine on cocaine self-administration
in rhesus monkeys. Psychopharmacology 137:253–258.
26. Petry NM ( 1995). Ro 15-4513 selectively attenuates ethanol
, but not sucrose, reinforced responding in a concurrent access procedure; comparison to other drugs. Psychopharmacology 121:192–203.
27. Rassnick S, D'Amico E, Riley E, Koob GF ( 1993). GABA antagonist and benzodiazepine partial inverse agonist reduce motivated responding for ethanol
. Alcohol Clin Exp Res 17:124–130.
28. Rassnick S, Pulvirenti L, Koob G ( 1993). SDZ-205,152, a novel dopamine receptor agonist, reduces oral ethanol self-administration
in rats. Alcohol 10:127–132.
29. Richter CP, Campbell KH ( 1940). Alcohol taste thresholds and concentrations of solution preferred by rats. Science 91:507–508.
30. Samson HH, Falk JL ( 1974). Alteration of fluid preference in ethanol
dependent animals. J Pharmacol Exp Ther 190:365–376.
31. Samson HH, Grant KA ( 1985). Chlordiazepoxide effects on ethanol self-administration
: dependence on concurrent conditions. J Exp Ana Beh 43:353–364.
32. Sanger DJ, Blackman DE ( 1976). Rate-dependent effects of drugs: a review of the literature. Pharmacol Biochem Behav 4:73–83.
33. Shelton KL, Balster RL ( 1997). Effects of gamma-aminobutyric acid agonists and N
-methyl- d -aspartate antagonists on a multiple schedule
and saccharin self-administration
in rats. J Pharmacol Exp Ther 280:1250–1260.
34. Shoaib M, Swanner LS, Beyer CE, Goldberg SR, Schindler CW ( 1998). The GABAB agonist baclofen modifies cocaine self-administration
in rats. Behav Pharmacol 9:195–206.
35. Silvestre JS, Palacios JM, Fernandez AG, O'Neill MF ( 1998). Comparison of effects of a range of 5-HT receptor modulators on consumption and preference for a sweetened ethanol
solution in rats. J Psychopharmacol 12:168–176.
36. Slawecki CJ, Hodge CW, Samson HH ( 1997). Dopaminergic and opiate agonists and antagonists differentially decrease multiple schedule
responding maintained by sucrose/ethanol
and sucrose. Alcohol 14:281–294.
37. Stromberg MF, Casale M, Volpicelli L, Volpicelli JR, O'Brien CP ( 1998). A comparison of the effects of the opioid antagonists naltrexone, naltrindole, and beta-funaltrexamine on ethanol
consumption in the rat. Alcohol 15:281–289.
38. Tang M, Falk JL ( 1977). Ethanol
dependence as a determinant of fluid preference. Pharmacol Biochem Behav 7:471–474.
39. Vivian JA, Green HL, Young JE, Majersky LS, Thomas BW, Shively CA, Nader MA, Grant KA ( 2001). Induction and maintenance of ethanol self-administration
in cynomolgus monkeys (Macaca fascicularis
): long-term characterization of sex and individual differences. Alcohol Clin Exp Res 25:1–11.
40. Wiesbeck GA, Weijers HG, Chick J, Naranjo CA, Boening J ( 1999). Ritanserin in relapse prevention in abstinent alcoholics: results from a placebo-controlled double-blind international multicenter trial. Ritanserin in Alcoholism Work Group. Alcohol Clin Exp Res 23:230–235.
41. Wilson AW, Neill JC, Costall B ( 1998). An investigation into the effects of 5-HT agonists and receptor antagonists on ethanol self-administration
in the rat. Alcohol 16:249–470.
42. Wojnicki FHE, Rothman RB, Rice KC, Glowa JR ( 1999). Effects of phentermine on responding maintained under multiple fixed-ratio schedules of food and cocaine presentation in the rhesus monkey
. J Pharmacol Exp Ther 288:550–560.
43. Woolverton WL, Virus RM ( 1989). The effects of a D1 and a D2 dopamine antagonist on behavior maintained by cocaine or food. Pharmacol Biochem Behav 32:691–697.
Keywords:© 2001 Lippincott Williams & Wilkins, Inc.
ethanol; operant; multiple schedule; self-administration; alcohol treatment; monkey