Stimulus-controlled responding for ethanol in AA and Wistar rats

Alcohol, Val. 8. pp. 229-234. ¢ Pergamon Press plc, 1991. Printed in the U.S.A.

-

Stimulus-Controlled Responding for Ethanol in AA and Wistar Rats P. H Y Y T I , ~ A N D J. D. S I N C L A I R

Research Laboratories, Alko Ltd., P.O.B. 350, SF-O0101 Helsinki, Finland

R e c e i v e d 31 July 1990; A c c e p t e d 7 N o v e m b e r 1990

HYYTI,~, P. AND J. D. SINCLAIR. Stimulus-controlled responding for ethanol in AA and Wistar rats. ALCOHOL 8(3) 229-234, 1991.--A method for establishing stimulus control of ethanol responding was developed. After acquisition of lever pressing for oral ethanol, rats of the high-drinking AA (Alko, Alcohol) line and of the moderate-drinking Wistar strain were subjected to alternating 20-rain alcohol access periods signaled by a stimulus light, and 40-rain nonaccess periods with no light. Ethanol responding during access periods progressively irIcreased and decreased during nonaccess. These changes were faster in the AAs than the Wistars, probably related to differential reinforcement from ethanol. In a second experiment, rats responding under stimulus control were given periods of alcohol deprivation of 3. 6, 12. and 24 h, indicated by a stimulus light. Deprivations shorter than 24 h increased the first-hour intake after renewed access by the AAs, but the Wistars showed no increase until after a 24-h deprivation. The results show stimulus control of ethanol responding and demonstrate the applicability of the procedure for causing ethanol responding to occur at a time chosen by the experimenter. Alcohol-preferring rats

Alcohol deprivation

Stimulus control

PREVIOUS studies have shown that rats will acquire an operant response such as lever pressing that is reinforced by giving alcohol to drink (5, 12, 17). The extensive research on operant conditioning with other reinforcers has developed a variety of procedures which can be applied also in studies involving ethanol-reinforced responding. A useful operant conditioning technique involves a discriminative stimulus which signals when pressing the lever produces the reinforcing agent. Our initial goal in the present study was to see whether ethanol responding could be placed reliably under stimulus control (21), i.e.. whether an antecedent stimulus would eventually determine the probability of the occurrence of the response to obtain alcohol. Stimulus control was established in two phases. After being trained to lever press for oral ethanol using a procedure developed in previous studies in our laboratory (6), the rats were subjected to a discrimination training procedure (21) during which ethanol reinforcement was available in the presence of a discriminative stimulus, an illuminated stimulus light, but not when the light was extinguished. If successful, this procedure would lead to a decrease in the rate of ethanol responding during the negative stimulus, while maintaining or increasing responding during the positive one. In addition, we wished to examine whether this acquisition of stimulus control over ethanol responding differs in AA and Wistar rats. The AA line has been developed by selebtive outbreeding for high voluntary alcohol drinking in a free-choice situation (3); Wistar rats represent the genetic variability of the foundation population, and on the average show only moderate levels of alcohol drinking. The AA rats rapidly acquire an operant response for oral ethanol, and increase their responding when the number of responses (fixed-ratio) for each ethanol reinforcement is increased

Ethanol self-administration

(6,15). Wistar rats are less likely to lever press for alcohol using identical procedures, and their responding is not affected by increases in the fixed-ratio schedule, suggesting less reinforcement from ethanol (6). The second experiment here represents a demonstration of one application of the stimulus-controlled procedure in alcohol research. In the two-bottle choice paradigm, depriving rats of alcohol for a week or more usually produces a temporary increase in ethanol consumption when access is again allowed (19). AA rats (and also rats of the alcohol-preferring P line) show, however, a "short alcohol-deprivation effect," i.e., they increase their consumption after only a few hours without alcohol, whereas Wistar rats show an increase only after a day or more of deprivation (18). We now attempted to replicate this finding with AAs and Wistars lever pressing for alcohol, using the stimulus light to indicate to the animals when alcohol was available. The second experiment also acts as a control for the first. In the first experiment, the selective responding could have been caused either by the rats' lever pressing when the stimulus light was on, or by their learning a temporal discrimination and responding during the regularly occurring access periods. The temporal cue was eliminated in the deprivation phase of the second experiment. Consequently, if responding began at a high level when the stimulus light was first illuminated, it would indicate stimulus control. GENERAL METHOD

Animals Eighteen male AA rats of the F54 generation and 6 male Wistar:Han/HY rats (Department of Laboratory Animals, Univer-

229

230

HYYTIA AND SINCLAIR

sity of Helsinki) were used in this study. At the age of three months, the animals were placed in individual stainless steel wire cages, in a room with controlled humidity and temperature (22°C±2 °) and a 12-h light-dark schedule (06:00-18:00 lights on). The rats were given a choice between 10% (v/v) ethanol solution and tap water, both in graduated 100 ml Richter tubes. Standard powdered Ewos R3 rat diet (Srdert~ilje, Sweden) was always available from a glass jar.

Apparatus Six identical operant chambers (20 cm wide x 28 cm long × 21 cm high, Lafayette Instrument Co., model 80001 ), each equipped with two levers located 9.5 cm apart, 7 cm above the grid floor, were used for the experimental sessions. A single lever press activated a solenoid-driven graduated liquid dispenser delivering either a 0.1 ml drop of 10% (v/v) ethanol or water to one of the drinking spouts between the levers, 2 cm above the floor. There was no detectable evaporation from the sealed liquid dispensers. The beginning of ethanol access periods was signaled by illuminating the light with a blue lens cap located above the alcohol lever. Half of the chambers had the ethanol lever on the right; half had it on the left. In addition to being able to obtain alcohol and water by lever pressing, the rats always had free access to tap water from a separate 100 ml Richter tube and to food from a glass jar, both placed at the opposite end of the chamber. The operant chambers were enclosed in separate sound-attenuation cubicles (35 cm wide x 62 cm long × 38 cm high), equipped with a ventilation fan and a house light with a 12-h light-dark cycle (06:00-18:00 lights on). Scheduling the alcohol access periods and recording of the responses on the levers were accomplished using a microcomputer (IBM XT), connected to the chambers via a LVB interface (Med Associates, Inc.), and controlled by OPN software (20).

Acquisition of Lever Pressing for Ethanol After a 48-day free choice between 10% (v/v) ethanol and water in their home cages, the rats were placed in operant chambers where they were housed continually for 8 days. The mean ( ~ SEM) weights of the AAs and Wistars were 347 ± 6 g and 4 6 0 ± 9 g, respectively. During this acquisition period, the animals could obtain ethanol and water by lever pressing on a concurrent FRI:FRI schedule during daily 23-h sessions. The stimulus light above the ethanol lever was not illuminated during the initial training. No shaping or additional training was used. Each daily 23-h session started at 13:00 and ended at 12:00 the next day when the sound-attenuation cubicles were opened. Between 12:00 and 13:00 the session data was stored on the fixed disk of the computer, the fluid levels were recorded, the dispensers, Richter tubes, and food jars were refilled, and the animals were removed from the chambers to record their weights. Since there were only 6 operant chambers available, the initial training period and the experiments were conducted sequentially in batches of 6 animals. The first two batches consisted of 3 AA and 3 Wistar rats, and the two remaining batches of 6 AAs. All 4 batches were subjected sequentially to training for stimulus control (Experiment 1), but only the 6 AAs and 6 Wistars from the first two batches were used for investigating the short alcoholdeprivation effect (Experiment 2). The rest of the AAs were used in other experiments. After completion of each experimental phase in operant chambers, the animals were returned to their home cages with free access to 10% (v/v) ethanol and water. EXPERIMENT I: ACQUISITION OF STIMULUS CONTROL

Procedure The rats had access to ethanol and water on a concurrent FRI:

FRI schedule as they had had during the initial training, but the availability of ethanol was signaled by illumination of the blue light above the alcohol lever. During the first two sessions, the stimulus light was illuminated constantly, indicating unrestricted access to ethanol. The following 12 23-h sessions each consisted of 23 signaled 20-min access periods, which alternated with 40min nonaccess periods during which pressing the ethanol lever had no consequence. The sessions always began with an access period, and ended with nonaccess. Water was always available from pressing the second lever on a nonrestricted basis and also from a Richter tube. The operant chambers were serviced daily as described before, and the 12-h light-dark cycle in the sound-attenuation cubicles was retained. The number of responses on ethanol and water levers were recorded separately for the 20-min and 40-rain components.

Results Both AA and Wistar rats learned to lever press for ethanol during the 8-day acquisition period (data not shown). The AAs responded significantly more for alcohol than water during the first 23-h training session, t(l 7) = 2.80, p<0.05; the Wistars' responses for alcohol were not significantly greater than those for water until the 5th day, t(5)=3.19, p<0.05. The mean ± S E M number of daily alcohol responses by the AAs and the Wistars over the last 4 days of training were 3 2 5 ± 10 and 101 _ 15. and the number of water responses 10±1 and 2 8 ± 8 , respectively. During this period, the mean intake of ethanol of the AAs was 6 . 9 2 ± 0 . 2 2 g/kg/day, and that of the Wistars, 1.83±0.31 g/kg/day. Following the acquisition period the rats were subjected to alternating 20-min signaled alcohol access periods and 40-min nonaccess periods, throughout the daily 23-h sessions. The mean response rate of AA rats during the access periods was already significantly greater than the rate during the nonsignaled, nonreinforced 40-min periods during the first day, t(17) = 6.76, p<0.001, and then increased over the 12-day training period, while the rate during the nonsignaled periods rapidly decreased to the level of water responding (Fig. 1). The two-way repeated measures analysis of variance showed a significant effect for the difference between the rates across the training days, F( 1,34) = 274.94, p<0.001, and for the interaction between the rates and days, F(11,374)= 14.11, p<0.001. Restricting ethanol access to 20-min periods decreased the total ethanol consumption compared to nonrestricted conditions: the level of stabilized intake during the last 4 days of training (5.36±0.22 g/kg/day) was significantly lower than that during the initial acquisition period, t(17)=7.41, p<0.001. The response rate of the Wistars during the access periods was significantly lower than that of the AAs, F( 1,22) = 74.16, p<0.001, and did not differ significantly from the rate of nonaccess periods, F(I,10)= 1.37, NS, over the 12-day training (Fig. 1). There was, however, a significant interaction between the rates and treatment days, F( I 1,110) = 2.74, p < 0 . 0 I, showing a differential change of the rates during the training, which suggests that the stimulus acquired control of the Wistars' responding. The Wistars' mean ethanol intake during last 4 days of the training (0.75__-0.30 g/kg/day) was significantly lower than that during the acquisition period, t(5)= 4.05, p<0.01. As is clear in Fig. 1, responses of the AA rats during the signal had already separated from the nonsignaled responding during the first day. Figure 2 shows a more detailed analysis of the apparent establishment of stimulus control by the AAs during this first 23-h session. Although the result is unavoidably affected by the diurnal rhythm of lever pressing, e.g., by greater activity during the dark than the light phase [see (6)], a general tendency can

STIMULUS-CONTROLLED ETHANOL RESPONDING

AA

231

WISTAR

1.0

0.6•

05-

-4 ~ - -

_Z

0.4.

Ethanol acoe .¢~ Ethanol nonacCeSs Walef Ethanol basel,no

•

Access Nonaccess

0.8 z

¢.,q 0.30.2

0.1

W 0.6 U') ¢a~

K**

ILl

.,"/,

,'r7

rr 2

4

6 DAYS

8

10

12

2

4

6

8

10

12

DAYS

FIG. I. Acquisition of stimulus control by 18 AA and 6 Wistar rats during 12 daily 23-h sessions. Shown are the mean rates of responding for ethanol during the 20-min access periods with an illuminated stimulus light, during the 40-min nonaccess periods with no stimulus light, and the total daily rate of responding for water. Ethanol and water were available on a concurrent FRI:FRI schedule. The water lever was always operative, and water could also be obtained from a Richter tube. The baseline represents the mean rate of responding for ethanol during the last 4 days of the previous acquisition period. In both strains the rate of ethanol responding increased progressively during" ethanol access and decreased during nonaccess to the level of water responding; the AAs, however, showed greater changes.

be seen. The AAs gradually increased their rate of responding during the signaled access periods, and, contrary to the normal diurnal activity rhythm, showed high responding also during the first hours of the light phase. During the nonaccess periods, however, there was a progressive decrease following the high responding at the beginning of the dark phase. A two-way repeated measures analysis of variance revealed significant effects for the difference between the rates during the positive and negative stimuli over all the 8 3-h blocks, F ( I , 3 4 ) = 4 7 . 6 0 , p < 0 . 0 0 1 , the change of rates over the blocks, F(7,238)= 10.71, p < 0 . 0 0 1 , and the interaction between the different rates and the blocks of the session, F(7,238) = 5.93, p < 0 . 0 0 1 . EXPERIMENT 2: SHORT ALCOHOL.DEPRIVATION EFFECT

o._ rr U.I >

z/j 0.4

//,

/// z_/., ///

W __a 0 . 2

"// "~"

0 13-15 15-18 18-21 21-24 24-03 03-06 06-09 09-12

TIME (HOUR) FIG. 2. Acquisition of stimulus control by 18 AA rats during the first 23-h session of discrimination training. The bars represent the rates of responding for ethanol (+--SEM)during the 20-rain access, and the '40min nonaccess periods. The session began with a 20-min access period which may have contributed to the differential responding during the first 2 hours (13-15). During the next 3 hours (15-18) there was no difference between access and nonaccess rates of responding, but differential responding developed over the subsequent 5 3-h periods. The isolation cubicles had a 12-h light-dark schedule: at 18:00 lights off. at 06:00 lights on. Comparing each animal's rate of responding during access to that during nonaccess: *p<0.05. **p<0.01. ***p<0.001. with a matchedpair t-test.

sponse patterns in detail, responses on ethanol and water levers were recorded at 2-min intervals between 04:00 and 06:00. The same hours of the last of the three intervening days of continuous access provided control data. During the rest of the 23-h session, the number of lever presses was recorded hourly. The chambers were serviced daily as described earlier.

Procedure Stimulus-dependent ethanol responding was used as a method to examine the short alcohol-deprivation effect in AA and Wistar rats. Between establishment of stimulus control and measurement of the short alcohol-deprivation effect, the animals were housed in their home cages. After being returned to the operant chambers, the rats were first retrained for stimulus control under the restricted (20-min access/40-min nonaccess hourly) schedule for a week. Then the animals were subjected to a series of short periods of alcohol deprivation (3, 6, 12, and 24 h), interspersed with 3 days of continuous (23 h a day) access. Each deprivation duration was tested twice. The order of deprivations for the first batch (of 3 AAs and 3 Wistars) was 3, 12, 6, 12, 3, 6, 24, and 24 h, and for the second batch 6, 12, 3, 24, 3, 12, 6, and 24 h. The beginning of deprivation was signaled by extinguishing the stimulus light above the ethanol lever. Access to water from pressing the second lever and from the Richter tube was not restricted. Each deprivation ended at 04:00, 2 hours before the end of the dark phase, when the stimulus light was illuminated again. According to our previous findings, the rate of responding at this time point is similar to the mean nocturnal responding of both AA and Wistars rats (6). In order to examine the postdeprivation re-

Results The results from the two batches were similar, and thus the combined means are presented. The first-hour ethanol intake after 3, 6, 12, and 24 hours of alcohol deprivation showed a different pattern in AA and Wistar rats (Fig. 3). The AAs earned significantly more ethanol after the 3-h, t(5)= 4.84, p < 0 . 0 1 , and 12-h, t(5) = 4.52, p < 0 . 0 I, deprivation compared to the consumption of the same hour during unrestricted access. In contrast to the AAs, the Wistars did not increase their alcohol intake significantly until after a 24-h deprivation, t(5) = 2.85, p < 0 . 0 5 . A twoway repeated measures analysis of variance showed significant effects for the difference between the strains, F(1,10)=23.99, p < 0 . 0 0 1 , the change with the duration of deprivation, F(4,40)= 3.82, p < 0 . 0 5 , and the interaction between strains and the deprivation duration, F ( 4 , 4 0 ) = 4 . 4 4 , p < 0 . 0 1 . The rats started lever pressing immediately after illumination of the stimulus light after alcohol deprivation (Fig. 4). The rate of responding was highest during the first minutes of access, then gradually decreased, thus creating a negatively accelerated response pattern. Most of responding occurred during the first 30 minutes. The AAs did not lever press for water after alcohol dep-

232

HYYTI,~ AND SINCLAIR

[] •

,,

E

!i

AFTER 3-HR DEPRIVATION

1.5 AA Wistar

1.0 m

AFTER 6-HR DEPRIVATION AA

....

Wator

AA E.....

o! . . . . . . . . . . . . . . . . . .

0.5

; \^ i

to

._m__ 0

3

T

6

12

24

HOURS OF DEPRIVATION FIG. 3. The effects of short periods of alcohol deprivation on the amount of ethanol earned in the first hour of renewed access (between 04:00 and 05:00) in AA and Wistar rats. Each bar represents the mean (--SEM) ethat:lol intake of 6 rats. The 0 deprivation control represents the consumption during the corresponding hour during the last of the 3 intervening days of unrestricted access. Ethanol volume was calculated by multiplying the number of responses during the first access hour with the ethanol drop size. Each deprivation duration was tested twice. Comparing each rat's first hour intake to that of the control hour, the AAs increased significantly their intake after 3 (**p<0.01) and 12 (**p<0.01) hours of deprivation, and the Wistars after the 24-h deprivation (#p<0.05).

DISCUSSION

If a change in a particular stimulus is always followed by a change in the probability or the rate of a particular response, the stimulus is considered to exercise some control over that response.

60 0

A F T E R 12-HR D E P R I V A T I O N

e

AA

10

i

WISTAR

__

20 30 40 :50 MIN AFTER DEPRIVATION

60

AFTER 24-HR DEPRIVATION

'

AA

i! , i

0

rivation at all. The Wistars, however, started responding for ethanol, but then had also relatively high levels of water responding. The number of lever presses for ethanol by Wistars during the first hour of access was significantly greater than that for water only after the 24-h deprivation, t ( 5 ) = 3.18, p < 0 . 0 5 . Since ethanol responding was recorded in 2-min bins after deprivation, it was possible to analyze the size and the number of ethanol bouts during the first postdeprivation access hour. An ethanol bout was defined as a set of at least five lever presses for ethanol terminated by at least a 2-min period of no responding. The volume of 10% ethanol per bout was determined by multiplying the number of lever presses per bout with the drop size. Alcohol deprivations shorter than 24 h tended to increase both the bout size and the number of bouts by the AAs during the first hour of renewed access as compared to the control hours. The bout size after the 12-h deprivation (0.43 ---0.04 g/kg) was significantly greater than that during the control hours, 0.31 --- 0.03 g/kg, t ( 5 ) = 2 . 5 9 , p < 0 . 0 5 , and the number of bouts increased from 1.42---0.39 to 2.42-+0.44, t ( 5 ) = 3 . 6 3 , p < 0 . 0 5 . In contrast, after the 24-h deprivation the AA rats took almost all their alcohol in a large, single bout (0.51 -+0.11 g/kg): 5 of the AAs had this bout during the first hour, but one had it during the second hour of renewed access. The response pattern of the Wistars after the 24-h deprivation was similar to that of the AAs: they drank their ethanol in a single bout (0.17---0.05 g/kg) that was not, however, significantly larger than the bouts during the control hours, 0 . 1 2 ± 0 . 0 3 g/kg, t ( 5 ) = 1.53, NS.

~ 30 40 50 MIN AE'I"ER DEPRIVATION

A

.Vv\j.~

W TAR

10

20

:30

40

SO

MIN AFTER DEPRIVATION

60 0

10

20

30

40

MIN AFTER DEPRIVATION

FIG. 4. Patterns of lever pressing for ethanol and water during the first hour after specified periods of alcohol deprivation in AA and Wistar rats. Each data point represents the mean number of ethanol or water responses by 6 animals during a 2-min period. Most animals started responding for ethanol during the first 2 minutes of reinstated access, and most responses occurred at the beginning of the access hour. The AAs showed the highest initial response rates after the 3- and 12-h deprivation, and the Wistars after the 6- and 24-h deprivation. The AAs did not respond for water, but the Wistars had also a burst of water responding, usually following the first ethanol responses.

The term stimulus control has been used as a general expression for such relationships between changes in external stimuli and behavior (8,21). The present study demonstrates two aspects of stimulus-controlled responding for oral ethanol in alcohol-preferring AA and moderate-drinking Wistar rats: first, the acquisition of stimulus control by these strains; and second, the effect of short periods of alcohol deprivation on ethanol-reinforced stimulus-controlled responding. An alternation of ethanol access periods with a discriminative stimulus (the stimulus light) and ethanol nonaccess periods with no light produced stimulus control for ethanol responding in both strains: the response rate during the access periods increased, while that during nonaccess declined to the level of sporadic water responding. The AAs, however, acquired stimulus control much faster than the Wistars. Generally, the speed at which discrimination learning takes place and the precision of performance has been shown to depend upon many factors including the subject's sensory system and prior experience, the physical properties of the stimuli, the nature

STIMULUS-CONTROLLED ETHANOL RESPONDING

233

of the response, and the nature of the reinforcer (8). Since we did not observe any differences in the " s a l i e n c e " of the light stimulus between the AAs and the Wistars, and they had had similar training, the differential acquisition of discrimination by these rat strains could be attributed to the fact that ethanol appears to be a more efficient reinforcer for the AAs than the Wistars. This explanation is supported by the AAs" faster initial acquisition of the lever-pressing response, their higher overall response rates, and the increase of ethanol responding under increasing fixed-ratio schedules seen in AAs, but not Wistars (6). However, as has been noted in this and earlier studies (18), the AAs tend to increase their alcohol drinking even after short periods of forced abstinence, which may also have contributed to their rapidly increasing alcohol responding during the hourly access periods. It is generally found that depriving an organism of the opportunity to emit a habitual response will increase the probability of that response. Consequently, interrupting continuous access to alcohol increases alcohol intake when unrestricted access is reinstated in rats (18.19), monkeys (16), and humans (2). Substantial increases in ethanol consumption are also obtained when alcohol availability is restricted to alternate days (1, 4, 14, 24h or to repeated access periods of different lengths within daily 24-h sessions (9-1 I, 13). In Experiment 2, the animals were on continuous alcohol availability, except that once every 3 days stimulus light was turned off and the ethanol lever was made inoperative for a specific number of hours. The temporal cue that had been available in the first experiment (access for 20 min out of every hour) was no longer present. The duration of alcohol deprivation also could not be used as a temporal cue since it varied randomly. Furthermore, if ethanol responding had not been under stimulus control, it would probably have been extinguished during the 3- to 24hour periods of nonaccess. Responding during these periods was rare, indicating further that the rats were not continually checking whether the lever was operative. Nevertheless, when the stimulus light was illuminated to indicate the end of deprivation, the rats began responding immediately. These results show clearly that the rats had learned to respond on the basis of the stimulus light, and that the selective responding in the first experiment was not controlled by the temporal sequence. In agreement with the previous two-bottle choice study (18), the AAs and the Wistars had a different pattern of increases as a function of the deprivation duration: the first-hour alcohol drinking by the AAs increased after deprivation shorter than 24 h, but

the Wistars were not affected until after a 24-h deprivation. The short alcohol-deprivation effect is also shown by P rats (18), and the first-hour response rates of the AAs observed in the present study were similar to those reported for P rats that had been trained to respond for ethanol during daily 30-min sessions under a continuous reinforcement schedule (23). In agreement with previous studies that have restricted alcohol access to repeated periods of different lengths within 24-h sessions (9-11, 13), increasing deprivation duration increased the size of ethanol bouts. The sizes of the observed bouts were almost identical with ethanol doses that have been reported to have discriminative stimulus properties (26), to increase dopamine release in the nucleus accumbens (7), and to stimulate spontaneous locomotive activity in rats (22), which has been suggested to express the positive reinforcing effects of ethanol (23). The response patterns of the Wistar rats were less conspicuous than those by the AAs, and the initially smaller ethanol bouts were not increased by deprivation. Finally, contrary to the response patterrl by the AAs, the Wistars also had a burst of water responding after the first ethanol bout. This might indicate that the stimulus light served as an initiating stimulus for reinforcement in general for the Wistars. or that the reinforcing and discriminative value of ethanol for them is rather low. In summary, the present results underline the role of genotype in determining the pattern of ethanol-reinforced, behavior across different environmental variables. The differences between AA and Wistar rats in acquisition of stimulus control for ethanol responding and in the magnitude and the pattern of responding after alcohol deprivation are probably manifestations of the differential value of ethanol as a reinforcer for these strains. The results also suggest that stimulus control could be a useful tool in alcohol research, as illustrated by the study here on short alcohol-deprivation effects. Furthermore, since rats reliably begin responding for ethanol immediately when the signal lamp is illuminated, the procedure could be used for studying the effects of short-acting pharmacological agents on alcohol drinking. In contrast, e.g., to placing an alcohol bottle on the cage at the end of a deprivation period, the signal for renewed access from the lamp is invariant and standardized. The procedure provides detailed records and temporal patterns. Finally, it could be combined with other operant procedures, such as increasing the number of lever presses per reinforcement in order to distinguish the motivation for ethanol from the capacity to consume it.

REFERENCES 1. Amit, Z.; Stem, M. H.; Wise, R. A. Alcohol preference in the laboratory rat induced by hypothalamic stimulation. Psychopharmacologia 17:367-377; 1970. 2. Burish, T. G.: Maisto, S. A.; Cooper, A. M.; Sobell. M. B. Effects of voluntary short-term abstinence from alcohol on subsequent drinking patterns of college students. J. Stud. Alcohol 42:1013-1020; 1981. 3. Eriksson, K. Genetic selection for voluntary alcohol consumption in the albino rat. Science 159:739-741; 1968. 4. Holloway, F. A.; Bird, D. C.; Devenpon, J. A. Periodic availability: Factors affecting alcohol selection in rats. Alcohol 1:19-25; 1984. 5. Hyyti~i, P.; Sinclair, J. D. Demonstration of lever pressing for oral ethanol by rats with no prior training or ethanol experience. Alcohol 6:161-164; 1989. 6. Hyytili, P.; Sinclair, J. D. Differential reinforcement and diurnal rhythms of lever pressing for ethanol in AA and Wistar rats. Alcohol.: Clin. Exp. Res. 14:375-379; 1990. 7. Imperato, A.; Di Chiara, G. Preferential stimulation of dopamine release in the nucleus accumbens of freely moving rats by ethanol. J. Pharmacol. Exp. Ther. 239:219-228; 1986.

8. Mackintosh, N. J. Stimulus control: attentional factors. In: Honig, W. K.; Staddon, J. E. R., eds. Handbook of operant behavior. Englewood Cliffs, NJ: Prentice-Hall; 1977:481-513. 9. Marcucella, H.; Munro. I. Patterns of ethanol and water consumption as a function of restricted access and feeding condition. Psychopharmacology (Berlin) 89:145-149; 1986. 10. Marcucella, H.; Munro, I. Ethanol consumption of free feeding animals during restricted access. Alcohol Drug Res. 7:405--414; 1987. 11. Marcucella, H.; Munro, I.; MacDonall, J. S. Patterns of ethanol consumption as a function of the schedule of ethanol access. J. Pharmacol. Exp. Ther. 230(3):658.--664; 1984. 12. Meisch, R. A.; Thompson, T. Ethanol intake in the absence of concurrent food reinforcement. Psychopharmacology (Berlin) 22:72-79; 1971. 13. Murphy, J. M.; Gatto, G. J.; Waller, M. B.; McBride, W. J.; Lumeng, L.; Li, T.-K. Effects of scheduled access on ethanol intake by the alcohol-preferring (P) line of rats. Alcohol 3:331-336; 1986. 14. Pinel, J. P. J.; Huang, E. Effects of periodic withdrawal on ethanol and saccharin selection in rats. Physiol. Behav. 16:693-698; 1976. 15. Ritz, M. C.; George, F. R.; Meisch, R. A. Ethanol self-administra-

234

16. 17. 18. 19. 20. 21.

HYYTI,~ A N D S I N C L A I R

tion in ALKO rats: II. Effects of selection and fixed-ratio size. Alcohol 6:235-239; 1989. Sinclair, J. D. The alcohol-deprivation effect in monkeys. Psychon. Sci. 25:21-22; 1971. Sinclair, J. D. Rats learning to work for alcohol. Nature 249:590592; 1974. Sinclair, J. D.; Li, T.-K. Long and short alcohol deprivation: Effects on AA and P alcohol-preferring rats. Alcohol 6:505-509; 1989. Sinclair, J. D.; Senter, R. J. Increased preference for ethanol in rats following alcohol deprivation. Psychon. Sci. 80):11-12; 1967. Spencer, D. G.; Emmett-Oglesby, M. W. Parallel processing strategies in the application of microcomputers to the behavioral laboratory. Behav. Res. Methods Instrum. 17:294-300; 1985. Terrace, H. S. Stimulus control. In: Honig, W. K., ed. Operant behavior: areas of research and application. Englewood Cliffs, NJ: Prentice-Hall; 1966:271-344.

22. Waller, M. B.; Murphy, J. M.; McBride, W. J.; Lumeng, L.; Li, T.-K. Effect of low dose ethanol on spontaneous motor activity in alcohol-preferring and -nonpreferring lines of rats. Pharmacol. Biochem. Behav. 24:617-623; 1986. 23. Weiss, F.; Mitchiner, M.: Bloom, F. E.; Koob, G. F. Free-choice responding for ethanol versus water in alcohol preferring (P) and unselected Wistar rats is differentially modified by naloxone, bromocriptine, and methysergide. Psychopharmacology (Berlin) 101: 178-186; 1990. 24. Wise, R. A. Voluntary ethanol intake in rats following exposure to ethanol on various schedules. Psychopharmacologia 29:203-210; 1973. 25. Wise, R. A.; Bozarth, M. A. A psychomotor stimulant theory of addiction. Psychol. Rev. 94(4):469--492; 1987. 26. York, J. L. A comparison of the discriminative stimulus effects of ethanol, barbital, and phenobarbital in rats. Psychopharmacology (Berlin) 60:19-23; 1978.