Psychopharmacology (199t) 105:101-106 003331589100175R
Psychopharmacology © Springer-Verlag t 991
Chronic opioids impair acquisition of both radial maze and Y-maze choice escape J a m e s W . Spain and Gordon C. N e w s o m
Department of Biomedical Sciences, University of Illinois College of Medicine at Rockford, 1601 Parkview Avenue, Rockford, IL 61107, USA Received August 16, 1990 / Final version January 15, 1991
Abstract. Chronic morphine impaired acquisition of two dissimilar behavioral tasks. In the radial maze, the performance of saline-treated and morphine-treated groups diverged with the latter failing to improve despite extensive training. In contrast, rats treated with naltrexone became skilled in the procedure 2-4 times as rapidly as saline controls. Withdrawal of treatment significantly improved performance o f morphine-treated rats, with no change for rats treated with saline or naltrexone. When a second group of rats was extensively trained prior to instituting chronic morphine treatment, performance scores were not affected, suggesting that morphine does not impair spatial working memory despite subjective evidence o f other gross behavioral effects, such as ataxia. In the Y-maze choice escape task, acquisition of a response strategy was significantly impaired in rats that had been previously treated with morphine for 17-21 days, despite clear indications that morphinetreated rats were sensitive to the aversive stimulus. Key words: Opioids - Learning - W o r k i n g m e m o r y Radial arm maze Y-maze choice escape - Tolerance
Numerous studies suggest that opioids are involved in cognitive processes (see Squire and Davis 1981). In general, agonists tend to impair while antagonists facilitate performance. Specifically, it has been shown that performance in shock motivated active or passive avoidance tests can be impaired by morphine (Castellano 1975; Izquierdo and McGaugh 1985), fl-endorphin (Izquierdo and McGaugh 1985), met-enkephalin (Introini et al. 1985), or leu-enkephalin (Martinez et al. 1986). Agonist-induced impairment has been shown at doses that are lower than those required to provide analgesia (Squire and Davis 198t). In contrast, antagonists tend to enhance retention of shock motivated avoidance, with naloxone (Izquierdo 1979; Flood et at. 1987), naltrexone Offprint requests to: J.W. Spain
(Gallagher 1982), and diprenorphine (Gallagher 1982) shown to be effective. In appetitively motivated tasks, such as the radial arm maze, a similar pattern has been demonstrated with naloxone enhancing performance (Gallagher et al. 1983; Spain and Newsom 1989; Canli et al. 1990). Importantly, effects produced in this type of test demonstrate that the action o f opioids is not simply to modulate nociceptive thresholds or block the effects of fl-endorphin released during electrical shock (Messing 1988). Furthermore, enhanced performance in the radial maze is counter to expectations based on naloxone's propensity to reduce food and water intake (Messing 1988). While effects on aversively motivated tasks could potentially be explained by a shift in reward value or affect (Squire and Davis 1981), results from appetitive studies do not support this interpretation. The goal o f these studies was to define further the role of opioids in cognitive processes by using a unique chronic treatment schedule. Through the induction o f tolerance, we were able to use a substantial dose of morphine while virtually eliminating interference produced by the gross behavioral effects associated with acute administration,
Materials and methods
Rats were obtained from Sasco Labs (Omaha, NE) and allowed 1 week to acclimate to our vivarium. Throughout all experiments, rats were moderately food restricted (10 g/day) to motivate food seeking behavior. Morphine sulfate was obtained from Merck Chemical (Rockway, N J) and naltrexone hydrochloride from Sigma Chem. Co. (St Louis, MO). Experiment 1. The effects of chronic treatment with morphine or naltrexone were determined for radial arm maze performance. The eight arm maze (Olton 1987), housed in a sound attenuated room, was constructed in our laboratory. The bulk of the maze is composed of transparent acrylic, permitting spatial orientation to the external surroundings. The major external landmarks include several tables, an assortment of laboratory apparatus (clutter), and a seated investigator logging the animal's choices. The central corn-
102 partment is a transparent regular octagon, 30 cm wide and 20 cm tall. Radiating from each side of the octagon is one arm of the maze. The arms are transparent acrylic tubes 90 cm long with 9 cm internal diameter. Each arm is fitted into a 6 cm long opaque polyvinyl sleeve attached to the exterior of the central octagon. The ability to quickly replace or m.ove arms of the maze allowed us to rule out the hypothesis that an arm of the maze is chosen on the basis of olfactory information. The caps at the end of each arm are also removable, composed of opaque PVC, and fitted with a small dish to hold the appetitive reward (one quarter of a Fruit Loop, Kelloggs Cereals). This reward has been found to be an exceptionally strong reinforcer. On day 1, both food restriction and drug treatment were begun. Rats (n= 6/group) were treated twice daily with vehicle (normal saline), morphine (20 mg/kg), or nattrexone (5 mg/kg), once at least 2 h before entering the maze and once 4-6 h after testing. On day 2 and 3, rats were twice placed into the baited maze for 15 min for acclimation to the apparatus. For the remainder of the week (days 4-6), the dose of morphine was raised to 30 mg/kg and rats were given one to two trials per day during which they were allowed only 8 choices (Strijkstra and Bolhuis 1987). The recording investigator considered that the rat had made a choice if it entered an arm and all four paws were no longer on the floor of the central enclosure (Strijkstra and Bolhuis 1987). Throughout the study, we recorded the sequence of arm visited and the timing of the visits. Later analysis provided the number of errors (entry of an arm where the bait had been previously consumed) and the speed at which the task was performed. During the third stage of the study (weeks 2-6), rats were allowed to retrieve four rewards, were moved from the maze to a holding cage for a 2 rain delay, and then returned to the maze to attempt recovery of the remaining four rewards (Gallagher et al. 1983; Strijkstra and Bolhuis 1987). Also during the second week, the dose of morphine was elevated to 40 mg/kg, remaining at this level until week 7. For the final stage (weeks 7-8), drug treatments were withdrawn while the other test parameters remained as before.
Experiment 2. The effects of chronic morphine treatment on radial arm maze performance of highly trained rats was determined. Twenty-four drug naive rats were extensively trained (6 weeks) on the maze with a 2-rain delay, then separated on the basis of performance criteria into two similar groups (n = 10) with the four worst performers eliminated. One group was assigned chronic morphine treatments (40 mg/kg, IP, twice daily) while the other received an equal volume of vehicle. Rats were tested in the maze 2 h after the first administration each day. Experiment 3. The effect of morphine on Y-maze choice escape performance was determined using rats that had been chronically treated. This test examines the learning of response patterns motivated by an aversive stimulus. The apparatus is a three arm maze (Y-maze) built according to the specifications of Jackson et al. (1980) with each arm being 22.5 cm long, 11 cm wide, 13.7 cm high, and 120 degrees apart. A small microcomputer (AIM-65) controlled experiments and recorded data, utilizing strategically placed photocells to monitor the position of the rat within the maze. A scrambled foot shock (1 mA) was delivered to the grid floor of the maze by a solid state shocker/distributor (El3-10, Coulbourn Instruments, Inc.). Background noise was masked by 80 db white noise. Each trial began when amperage was applied to the grid floor and terminated when the rat entered the arm to the left of that which it most recently occupied. If the rat made an incorrect choice (right arm), it was required to turn left when exiting the incorrect choice. If the rat again turned right, the trial continued until a left choice was made or 60 s elapsed. Each rat received 100 trials with a 45 s intertrial separation period. Each rat executed this set of trials only once and the investigator running the apparatus was blind as to treatment group. The percent incorrect choices for the first choice of each trial and the performance (total trial time/number of arms visited) were tallied for each of ten trial blocks (100 trials divided
into 10 blocks of 10 trials each). Rats used in this study had been treated twice daily for 17-21 days with either morphine (40 mg/kg, n = 19) or vehicle ( n - 15) with final administration 2 h before testing.
Experiment 4. Tolerance to morphine-induced analgesia was quantified. Rats chronically treated with morphine or saline were subjected to two tests of morphine-induced analgesia (thermal tail-flick latency and foot-shock vocalization threshold) on days 0, 3, 6, 9, 13 and 17. One set of rats received morphine while the other was treated with vehicle. Each treatment group was further subdivided into animals tested 20 min or 120 min after administration with n = 5 for each subgroup. On day 0, analgesia was measured by both tests following 20 mg/kg morphine, IP, or vehicle. On the third day of chronic twice daily treatments with the previous test dose, the dose was elevated to 40 mg/kg to remain at that level for the remainder of the experiment. The threshold for CNS perception of nociceptive stimuli was determined using the foot-shock vocalization response (Levine et al. 1984) either 20 or 120 min after injection of saline or morphine. Rats were placed in an apparatus whose floor is a grid acting as a conduit for scrambled electric current (solid state shocker/distributor, E 13-16, Coulbourn Instruments, Inc.) at a computercontrolled amperage. A microphone was positioned 40 cm above the grid with its output amplified, rectified, and input through an analog/digital converter (Metrobyte, DAS-8) to an IBM-AT. Custom software controlled the entire testing process after placement of the rat in the apparatus. The current was stepped (+0.2 mA increments with a random interval between) until vocalization is detected. The thermal tail-flick latency test (Yoburn et al. 1984; Berge et al. 1988) was also used to quantify morphine-induced analgesia. At either 20 or 120 min after administration, heat was applied to a spot on the tail 5 cm from the distal end. Movement of the tail is detected by a photocell with latency of the response recorded in tenths of a second (Muromaehi, MK 330). If a response was not detected within 10 s, the test was terminated and a score of > 10 recorded. Restraint stress can greatly influence results of this test, probably through activation of the endogenous opioid system (Porro and Fields 1988). We have found that highly reproducible results are obtained when rats are restrained in a cloth towel rather than a transparent plastic restraint cage.
Results It has been reported that performance in the radial arm m a z e relies o n w o r k i n g m e m o r y recall o f a list o f p r e v i o u s s p a t i a l o r i e n t a t i o n s w i t h r e s p e c t to e x t e r n a l s u r r o u n d i n g s ( S t r i j k s t r a a n d B o l h u i s 1987) a n d is n o t b a s e d o n i n t r a m a z e cues o r c h a i n i n g ( W a l l a c e et al. 1980). O u r f i n d i n g s c o n f i r m this h y p o t h e s i s . R e m o v i n g the r a t after r e c o v e r y o f f o u r r e w a r d s , r o t a t i n g t h e m a z e 180 °, a n d r e l o c a t i n g r e w a r d s to their p r e v i o u s s p a t i a l l o c a t i o n s d i d n o t r e d u c e the n u m b e r o f r e w a r d s r e c o v e r e d w h e n the r a t is r e t u r n e d to the m a z e to finish the task. T h i s suggests t h a t rats d o n o t c h o o s e a r m s o n t h e basis o f i n t r a m a z e cues.
Experiment 1 D u r i n g the first w e e k o f c h r o n i c t r e a t m e n t w i t h saline, m o r p h i n e , o r n a l t r e x o n e , t h e r e were n o a p p a r e n t differences i n p e r f o r m a n c e b e t w e e n g r o u p s (Fig. 1). I n d e e d , m o r p h i n e - t r e a t e d rats t e n d e d to r e c o v e r m o r e r e w a r d s t h a n the o t h e r g r o u p s . W e h a v e d e t e r m i n e d , u s i n g p r o b a -
103 2.0
1.5
tO
w
0 rr 1.0
Y
rr
lU 0.5
x
START
0.0
0
1
WITHDRAWAL
DELAY
,
~
,
~
,
2
3
4
5
6
7
8
DRUG
9
WEEKS
Fig. 1. Eight arm radial maze performance of rats chronically treated with vehicle (e), morphine (o), or naltrexone (~). Points represent the mean errors/trial ( _+SEM) determined from 5-8 trials/ week for each of 6 animals/group. A 2-rain delay between recovery of rewards 4 and 5 was instituted at the start of week 2. Withdrawal of drug treatment occurred at the start of week 7
bility theory, that the mean errors ( ± SEM) in a random walk with eight choices would be 2.748 ± 0.090 for 100 trials. Experimental performance of all groups during the first week averages approximately 0.8 errors/trial, substantially better than a random walk. During the second week when a 2 min delay was introduced between choices 4 and 5, performance of all groups fell equally to average of approximately 1.5 errors. Following the delay, probable errors incurred during a random walk with four choices would be 2.345±0.081 (n=100). Due to the induction of tolerance to the effects of morphine, subjective observation of behavior during this phase of testing could not differentiate test groups. During weeks 3-6, there was a clear divergence of the three treatment groups. The performance of morphine-treated rats did not improve, saline-treated rats required approximately 3 weeks to approach their optimal performance, while naltrexone-treated rats achieved their optimal performance within I week. Withdrawal of treatment at the start of week 7 did not affect performance of salinetreated or naltrexone-treated, but the performance of morphine-treated rats was substantially improved in week 8, nearly equalling that of saline-treated rats. Weekly mean errors for each individual rat were subjected to statistical analysis. Two-way analysis of variance with replication was performed on three groups of data determined by conditions within the experiment. When data from weeks 1-8 were analyzed, significant differences could be attributed to both treatment group (P = 0.033) and week ( P < 0.001). Moreover, an interaction between treatment group and week was detected (P < 0.005). Analysis of weeks 2-6 (during treatment and with a 2 rain delay) again indicated significant differences in both treatment group ( P = 0.024) and week ( P < 0.001) with a trend toward an interaction between group and week (P = 0.071). Similarly, analysis of weeks 7-8 (withdrawal period) indicated significant differences in both treatment group (P--0.031) and week ( P = 0.003) with a trend toward an interaction between group and week (P = 0.088).
One-way analysis of variance with unequal groups followed by the Tukey post-hoc test was used to determine if statistical differences occur between weekly averages for the three groups on a given week or between weekly performances of a given group. One week after the delay was instituted, performance of naltrexonetreated rats had significantly improved to only 0.6 errors, not to improve further in the subsequent 5 weeks (Fig. 1). The level of performance of naltrexone-treated rats was also significantly better than both saline-treated and morphine-treated rats (P=0.001) and remained so throughout the study (except with respect to salinetreated rats on weeks 4 and 6 with P < 0.1 and ns, respectively). The performance of saline-treated rats was significantly improved during weeks 3 and 4. Morphinetreated rats demonstrated significant differences (P
Psychopharmacology © Springer-Verlag t 991
Chronic opioids impair acquisition of both radial maze and Y-maze choice escape J a m e s W . Spain and Gordon C. N e w s o m
Department of Biomedical Sciences, University of Illinois College of Medicine at Rockford, 1601 Parkview Avenue, Rockford, IL 61107, USA Received August 16, 1990 / Final version January 15, 1991
Abstract. Chronic morphine impaired acquisition of two dissimilar behavioral tasks. In the radial maze, the performance of saline-treated and morphine-treated groups diverged with the latter failing to improve despite extensive training. In contrast, rats treated with naltrexone became skilled in the procedure 2-4 times as rapidly as saline controls. Withdrawal of treatment significantly improved performance o f morphine-treated rats, with no change for rats treated with saline or naltrexone. When a second group of rats was extensively trained prior to instituting chronic morphine treatment, performance scores were not affected, suggesting that morphine does not impair spatial working memory despite subjective evidence o f other gross behavioral effects, such as ataxia. In the Y-maze choice escape task, acquisition of a response strategy was significantly impaired in rats that had been previously treated with morphine for 17-21 days, despite clear indications that morphinetreated rats were sensitive to the aversive stimulus. Key words: Opioids - Learning - W o r k i n g m e m o r y Radial arm maze Y-maze choice escape - Tolerance
Numerous studies suggest that opioids are involved in cognitive processes (see Squire and Davis 1981). In general, agonists tend to impair while antagonists facilitate performance. Specifically, it has been shown that performance in shock motivated active or passive avoidance tests can be impaired by morphine (Castellano 1975; Izquierdo and McGaugh 1985), fl-endorphin (Izquierdo and McGaugh 1985), met-enkephalin (Introini et al. 1985), or leu-enkephalin (Martinez et al. 1986). Agonist-induced impairment has been shown at doses that are lower than those required to provide analgesia (Squire and Davis 198t). In contrast, antagonists tend to enhance retention of shock motivated avoidance, with naloxone (Izquierdo 1979; Flood et at. 1987), naltrexone Offprint requests to: J.W. Spain
(Gallagher 1982), and diprenorphine (Gallagher 1982) shown to be effective. In appetitively motivated tasks, such as the radial arm maze, a similar pattern has been demonstrated with naloxone enhancing performance (Gallagher et al. 1983; Spain and Newsom 1989; Canli et al. 1990). Importantly, effects produced in this type of test demonstrate that the action o f opioids is not simply to modulate nociceptive thresholds or block the effects of fl-endorphin released during electrical shock (Messing 1988). Furthermore, enhanced performance in the radial maze is counter to expectations based on naloxone's propensity to reduce food and water intake (Messing 1988). While effects on aversively motivated tasks could potentially be explained by a shift in reward value or affect (Squire and Davis 1981), results from appetitive studies do not support this interpretation. The goal o f these studies was to define further the role of opioids in cognitive processes by using a unique chronic treatment schedule. Through the induction o f tolerance, we were able to use a substantial dose of morphine while virtually eliminating interference produced by the gross behavioral effects associated with acute administration,
Materials and methods
Rats were obtained from Sasco Labs (Omaha, NE) and allowed 1 week to acclimate to our vivarium. Throughout all experiments, rats were moderately food restricted (10 g/day) to motivate food seeking behavior. Morphine sulfate was obtained from Merck Chemical (Rockway, N J) and naltrexone hydrochloride from Sigma Chem. Co. (St Louis, MO). Experiment 1. The effects of chronic treatment with morphine or naltrexone were determined for radial arm maze performance. The eight arm maze (Olton 1987), housed in a sound attenuated room, was constructed in our laboratory. The bulk of the maze is composed of transparent acrylic, permitting spatial orientation to the external surroundings. The major external landmarks include several tables, an assortment of laboratory apparatus (clutter), and a seated investigator logging the animal's choices. The central corn-
102 partment is a transparent regular octagon, 30 cm wide and 20 cm tall. Radiating from each side of the octagon is one arm of the maze. The arms are transparent acrylic tubes 90 cm long with 9 cm internal diameter. Each arm is fitted into a 6 cm long opaque polyvinyl sleeve attached to the exterior of the central octagon. The ability to quickly replace or m.ove arms of the maze allowed us to rule out the hypothesis that an arm of the maze is chosen on the basis of olfactory information. The caps at the end of each arm are also removable, composed of opaque PVC, and fitted with a small dish to hold the appetitive reward (one quarter of a Fruit Loop, Kelloggs Cereals). This reward has been found to be an exceptionally strong reinforcer. On day 1, both food restriction and drug treatment were begun. Rats (n= 6/group) were treated twice daily with vehicle (normal saline), morphine (20 mg/kg), or nattrexone (5 mg/kg), once at least 2 h before entering the maze and once 4-6 h after testing. On day 2 and 3, rats were twice placed into the baited maze for 15 min for acclimation to the apparatus. For the remainder of the week (days 4-6), the dose of morphine was raised to 30 mg/kg and rats were given one to two trials per day during which they were allowed only 8 choices (Strijkstra and Bolhuis 1987). The recording investigator considered that the rat had made a choice if it entered an arm and all four paws were no longer on the floor of the central enclosure (Strijkstra and Bolhuis 1987). Throughout the study, we recorded the sequence of arm visited and the timing of the visits. Later analysis provided the number of errors (entry of an arm where the bait had been previously consumed) and the speed at which the task was performed. During the third stage of the study (weeks 2-6), rats were allowed to retrieve four rewards, were moved from the maze to a holding cage for a 2 rain delay, and then returned to the maze to attempt recovery of the remaining four rewards (Gallagher et al. 1983; Strijkstra and Bolhuis 1987). Also during the second week, the dose of morphine was elevated to 40 mg/kg, remaining at this level until week 7. For the final stage (weeks 7-8), drug treatments were withdrawn while the other test parameters remained as before.
Experiment 2. The effects of chronic morphine treatment on radial arm maze performance of highly trained rats was determined. Twenty-four drug naive rats were extensively trained (6 weeks) on the maze with a 2-rain delay, then separated on the basis of performance criteria into two similar groups (n = 10) with the four worst performers eliminated. One group was assigned chronic morphine treatments (40 mg/kg, IP, twice daily) while the other received an equal volume of vehicle. Rats were tested in the maze 2 h after the first administration each day. Experiment 3. The effect of morphine on Y-maze choice escape performance was determined using rats that had been chronically treated. This test examines the learning of response patterns motivated by an aversive stimulus. The apparatus is a three arm maze (Y-maze) built according to the specifications of Jackson et al. (1980) with each arm being 22.5 cm long, 11 cm wide, 13.7 cm high, and 120 degrees apart. A small microcomputer (AIM-65) controlled experiments and recorded data, utilizing strategically placed photocells to monitor the position of the rat within the maze. A scrambled foot shock (1 mA) was delivered to the grid floor of the maze by a solid state shocker/distributor (El3-10, Coulbourn Instruments, Inc.). Background noise was masked by 80 db white noise. Each trial began when amperage was applied to the grid floor and terminated when the rat entered the arm to the left of that which it most recently occupied. If the rat made an incorrect choice (right arm), it was required to turn left when exiting the incorrect choice. If the rat again turned right, the trial continued until a left choice was made or 60 s elapsed. Each rat received 100 trials with a 45 s intertrial separation period. Each rat executed this set of trials only once and the investigator running the apparatus was blind as to treatment group. The percent incorrect choices for the first choice of each trial and the performance (total trial time/number of arms visited) were tallied for each of ten trial blocks (100 trials divided
into 10 blocks of 10 trials each). Rats used in this study had been treated twice daily for 17-21 days with either morphine (40 mg/kg, n = 19) or vehicle ( n - 15) with final administration 2 h before testing.
Experiment 4. Tolerance to morphine-induced analgesia was quantified. Rats chronically treated with morphine or saline were subjected to two tests of morphine-induced analgesia (thermal tail-flick latency and foot-shock vocalization threshold) on days 0, 3, 6, 9, 13 and 17. One set of rats received morphine while the other was treated with vehicle. Each treatment group was further subdivided into animals tested 20 min or 120 min after administration with n = 5 for each subgroup. On day 0, analgesia was measured by both tests following 20 mg/kg morphine, IP, or vehicle. On the third day of chronic twice daily treatments with the previous test dose, the dose was elevated to 40 mg/kg to remain at that level for the remainder of the experiment. The threshold for CNS perception of nociceptive stimuli was determined using the foot-shock vocalization response (Levine et al. 1984) either 20 or 120 min after injection of saline or morphine. Rats were placed in an apparatus whose floor is a grid acting as a conduit for scrambled electric current (solid state shocker/distributor, E 13-16, Coulbourn Instruments, Inc.) at a computercontrolled amperage. A microphone was positioned 40 cm above the grid with its output amplified, rectified, and input through an analog/digital converter (Metrobyte, DAS-8) to an IBM-AT. Custom software controlled the entire testing process after placement of the rat in the apparatus. The current was stepped (+0.2 mA increments with a random interval between) until vocalization is detected. The thermal tail-flick latency test (Yoburn et al. 1984; Berge et al. 1988) was also used to quantify morphine-induced analgesia. At either 20 or 120 min after administration, heat was applied to a spot on the tail 5 cm from the distal end. Movement of the tail is detected by a photocell with latency of the response recorded in tenths of a second (Muromaehi, MK 330). If a response was not detected within 10 s, the test was terminated and a score of > 10 recorded. Restraint stress can greatly influence results of this test, probably through activation of the endogenous opioid system (Porro and Fields 1988). We have found that highly reproducible results are obtained when rats are restrained in a cloth towel rather than a transparent plastic restraint cage.
Results It has been reported that performance in the radial arm m a z e relies o n w o r k i n g m e m o r y recall o f a list o f p r e v i o u s s p a t i a l o r i e n t a t i o n s w i t h r e s p e c t to e x t e r n a l s u r r o u n d i n g s ( S t r i j k s t r a a n d B o l h u i s 1987) a n d is n o t b a s e d o n i n t r a m a z e cues o r c h a i n i n g ( W a l l a c e et al. 1980). O u r f i n d i n g s c o n f i r m this h y p o t h e s i s . R e m o v i n g the r a t after r e c o v e r y o f f o u r r e w a r d s , r o t a t i n g t h e m a z e 180 °, a n d r e l o c a t i n g r e w a r d s to their p r e v i o u s s p a t i a l l o c a t i o n s d i d n o t r e d u c e the n u m b e r o f r e w a r d s r e c o v e r e d w h e n the r a t is r e t u r n e d to the m a z e to finish the task. T h i s suggests t h a t rats d o n o t c h o o s e a r m s o n t h e basis o f i n t r a m a z e cues.
Experiment 1 D u r i n g the first w e e k o f c h r o n i c t r e a t m e n t w i t h saline, m o r p h i n e , o r n a l t r e x o n e , t h e r e were n o a p p a r e n t differences i n p e r f o r m a n c e b e t w e e n g r o u p s (Fig. 1). I n d e e d , m o r p h i n e - t r e a t e d rats t e n d e d to r e c o v e r m o r e r e w a r d s t h a n the o t h e r g r o u p s . W e h a v e d e t e r m i n e d , u s i n g p r o b a -
103 2.0
1.5
tO
w
0 rr 1.0
Y
rr
lU 0.5
x
START
0.0
0
1
WITHDRAWAL
DELAY
,
~
,
~
,
2
3
4
5
6
7
8
DRUG
9
WEEKS
Fig. 1. Eight arm radial maze performance of rats chronically treated with vehicle (e), morphine (o), or naltrexone (~). Points represent the mean errors/trial ( _+SEM) determined from 5-8 trials/ week for each of 6 animals/group. A 2-rain delay between recovery of rewards 4 and 5 was instituted at the start of week 2. Withdrawal of drug treatment occurred at the start of week 7
bility theory, that the mean errors ( ± SEM) in a random walk with eight choices would be 2.748 ± 0.090 for 100 trials. Experimental performance of all groups during the first week averages approximately 0.8 errors/trial, substantially better than a random walk. During the second week when a 2 min delay was introduced between choices 4 and 5, performance of all groups fell equally to average of approximately 1.5 errors. Following the delay, probable errors incurred during a random walk with four choices would be 2.345±0.081 (n=100). Due to the induction of tolerance to the effects of morphine, subjective observation of behavior during this phase of testing could not differentiate test groups. During weeks 3-6, there was a clear divergence of the three treatment groups. The performance of morphine-treated rats did not improve, saline-treated rats required approximately 3 weeks to approach their optimal performance, while naltrexone-treated rats achieved their optimal performance within I week. Withdrawal of treatment at the start of week 7 did not affect performance of salinetreated or naltrexone-treated, but the performance of morphine-treated rats was substantially improved in week 8, nearly equalling that of saline-treated rats. Weekly mean errors for each individual rat were subjected to statistical analysis. Two-way analysis of variance with replication was performed on three groups of data determined by conditions within the experiment. When data from weeks 1-8 were analyzed, significant differences could be attributed to both treatment group (P = 0.033) and week ( P < 0.001). Moreover, an interaction between treatment group and week was detected (P < 0.005). Analysis of weeks 2-6 (during treatment and with a 2 rain delay) again indicated significant differences in both treatment group ( P = 0.024) and week ( P < 0.001) with a trend toward an interaction between group and week (P = 0.071). Similarly, analysis of weeks 7-8 (withdrawal period) indicated significant differences in both treatment group (P--0.031) and week ( P = 0.003) with a trend toward an interaction between group and week (P = 0.088).
One-way analysis of variance with unequal groups followed by the Tukey post-hoc test was used to determine if statistical differences occur between weekly averages for the three groups on a given week or between weekly performances of a given group. One week after the delay was instituted, performance of naltrexonetreated rats had significantly improved to only 0.6 errors, not to improve further in the subsequent 5 weeks (Fig. 1). The level of performance of naltrexone-treated rats was also significantly better than both saline-treated and morphine-treated rats (P=0.001) and remained so throughout the study (except with respect to salinetreated rats on weeks 4 and 6 with P < 0.1 and ns, respectively). The performance of saline-treated rats was significantly improved during weeks 3 and 4. Morphinetreated rats demonstrated significant differences (P
