Physiology&Behavior,Vol. 51, pp. 381-390. PergamonPress plc, 1992. Printed in the U.S.A.
0031-9384/92 $5.00 + .00
Perceptual Bisection in Rats: The Effects of Physostigmine, Scopolamine and Pirenzepine I DAVID SHURTLEFF 2
Thermal Stress Adaptation Program, Naval Medical Research Institute, Bethesda, MD 20889-5055 T H O M A S G. R A S L E A R
Department of Microwave Research, Walter Reed Army Institute of Research, Washington, DC 20307-5100 RAYMOND
F. G E N O V E S E
AND LARRY SIMMONS
Department of Medical Neurosciences, Walter Reed Army Institute of Research, Washington, DC 20307-5100 R e c e i v e d 4 M a r c h 1991 SHURTLEFF, D., T. G. RASLEAR, R. F. GENOVESE AND L. SIMMONS. Perceptualbisection in rats: The effects ofphysostigmine, scopolamine and pirenzepine. PHYSIOL BEHAV 51(2) 381-390, 1992.--The effects of cholinergic drugs on three different perceptual bisection tasks were studied in rats. Physostigmine (0.056-0.56 mg/kg), a reversible anticholinesterase, produced dose-dependent decrements in discriminability (A'), but did not affect the bisection point (BP) in visual duration, auditory duration, and auditory intensity bisection tasks. This finding is consistent with results previously obtained in an auditory duration bisection task with an irreversible anticholinesterase, diisopropyl phosphorfluoridate. Scopolamine (0.075--0.422 mg/kg), a muscarinic cholinergic-receptor antagonist, produced dose-dependent decrements in both A' and BP in visual and auditory duration bisection tasks. The behavioral antagonism between physostigmine (0.56 mg/kg) and scopolamine (0.075-0.237 mg/kg) was studied in the visual and auditory duration bisection tasks. The BP was not affected by physostigmine alone or in combination with scopolamine, except at the largest dose of scopolamine, which produced a reliable decrement in the BP. A', however, was equally decreased by physostigmine alone and all combinations of physostigmine and scopolamine. Pirenzepine (1, 3 and 10 mg/kg), a selective high-affinity M t muscarinic antagonist, had no effect on A' or the BP in the duration bisection tasks, suggesting changes in perception produced by muscarinic antagonists do not involve the M 1 receptor subtype. The similar drug effects in different sensory modalities (visual and auditory) and perceptual systems (subjective duration and loudness) suggest that cholinergic drugs may affect perceptual mechanisms responsible for sensory coding, such as the output of a neural generator. Physostigmine Rats
Pirenzepine
Scopolamine
Cholinergic
RECENTLY, we reported that the organophosphorus anticholinesterase diisopropyl phosphorofluoridate (DFP) disrupts time discrimination, but not bisection performance, in a temporal bisection procedure (25). Since we had previously demonstrated that DFP produces a long-term disruption of circadian activity and temperature patterns (10, 22, 23), we hypothesized that an organophosphorns anticholinesterase, such as DFP, alters circadian patterns and time perception by directly affecting an "internal clock" (23,26). However, low doses of the anticholinesterase physostigmine have been shown to improve discrimination performance on a
Temporal discrimination
Loudness discrimination
temporal bisection task (15) and on a visual discrimination task (31). Unlike DFP, physostigmine produces a carbamylated acetylcholinesterase (ACHE) enzyme, which prevents the normal enzyme-catalyzed hydrolysis of acetylcholine for several hours (27). DFP, however, is an irreversible anticholinesterase, in that it results in a more stable phosphorylated enzyme, resulting in permanent deactivation of AChE (27). Thus ACHE activity occurs only after the synthesis of new enzyme. It is possible that the differential effects on behavior seen with DFP and physostigmine are determined by their differential effects on ACHE. Alternatively, the effects of anticholinesterase compounds on
~The research described in this report was conducted in compliance with the Animal Welfare Act and other federal statutes and regulations relating to animals and experiments involving animals and adheres to the principles stated in the Guide for the Care and Use of Laboratory Animals, NIH publication 85-23. The views of the author(s) do not purport to reflect the position of the Department of the Army, Department of the Navy or the Department of Defense, (para 4-3, AR 360-5). Portions of this research were presented at the annual meetings of the Eastern Psychological Association in Philadelphia, PA, April 1990; and The American Psychological Society in Dallas, TX, June 1990. 2Requests for reprints should be addressed to Dr. David Shurtleff, Thermal Stress Adaptation Program, MS-11, Naval Medical Research Institute, Bethesda, MD 20889-5055.
381
382
SHURTLEFF, RASLEAR, GENOVESE AND 51MM~)NS
time perception may be dose related, as has been demonstrated by Warburton and Brown (31) for physostigmine with visual discrimination performance. The following experiments examine these possibilities by determining the effects of a range of doses of physostigmine on performance in a temporal bisection procedure using both visual and auditory stimuli. To further investigate the role of the cholinergic system on temporal bisection and the internal clock, a range of doses of the musarinic antagonist scopolamine was administered. To examine the potential behavioral antagonism between physostigmine and scopolamine, various doses of scopolamine were combined with a behaviorally effective dose of physostigmine. We also examined the effects of the muscarinic antagonist pirenzepine on time perception. Pirenzepine has a high affinity for the M~ muscarinic receptor subtype and a low affinity for the M 2 receptor subtype (32). Although highly hydrophilic-systemic administration of pirenzepine has been shown to be behaviorally active (6,33). Finally, to determine the generality of the effects observed with physostigmine in the temporal bisection task, Experiment 2 examined the effects of physostigmine on performance in an auditory intensi~ bisection task. This is of interest because the perception of loudness can be modeled by a process similar to that used in animal timing models [cf. (8,13)]. That is. '~counting models" of loudness perception relate the loudness of an auditory intensity to the number of neural impulses that have occurred within a fixed time period (1). Similarly, recent animal models of time perception (8) postulate an internal clock, or pacemaker, the output of which is accumulated (counted) and used to estimate the duration of an event. The similar theoretical mechanisms of duration and loudness perception suggest that similar drug effects should also be apparent in the auditory intensity bisection task. EXPERIMENT 1 A temporal bisection task was used to explore the perceptual effects of several cholinergic drugs. Separate groups of animals performed the task with visual and auditory durations as stimuli so that sensory-specific effects of the drugs would be discernable from effects specific to duration perception. As previously described (25,26), performance on this task is analyzed within the context of a simple, heuristic, information processing model of animal timing (2, 3, 8). The model contains a pacemaker, or clock, which produces a mean rate of output (ticks/unit time) and an associated variance (variability in clock rate across observation periods). Clock variance can arise from a variety of sources, either inherent to the pacemaker or contributed by other functional elements of the model. For instance, the model contains a memory function which allows a direct comparison of previous and current samples of the clock output. If the memory function becomes impaired, variance is added to the clock output and is indistinguishable from inherent variability. Our analysis provides two different measures of clock function. A' is a nonparametric signal detection theory measure of discriminability (9), and depends upon both the mean clock rate and the variance of the clock [see (25)]. The bisection point (BP) is an estimate of the perceptual midpoint of the temporal interval under consideration, and is a variance-free measure of clock function. Thus the pattern of drug effects on these two measures provides model-relevant information concerning drug action on time perception. Specifically, drugs that affect the mean rate of the clock should produce a change in A' and BP, while drugs that affect the variance of the clock should only produce a change in A'.
METHOD
Subjects Twenty male, albino, Sprague-Dawley rats. 90-t 10 day~ old at the start of the experiment, served. Following a short period of adaptation to the laboratory and individual housing, the rats were reduced to 80% of their free-feeding weight. Water was freely available in the home cages and supplemental chow was provided, as needed, to maintain body weight. A [2-h light:12h dark cycle (lights on at 0600) was maintained throughout the experiment. Two rats did not complete the experiment, and their data was used in the analysis only for the portion of the experiment they did complete.
Apparatus Five similarly constructed operant chambers (Coulbourn Instruments, Model El0-10) were used. Each chamber consisted of a sound- and light-attenuating enclosure that contained a Plexiglas and metal cage with a grid floor, two response levers. a houselight and cue lights, a Sonalert (2.9 kHz), a food magazine, and a pellet feeder which dispensed 45-mg Bio-serv (Frenchtown, NJ) food pellets. The cages measured 30.6 cm (L) × 32.8 (H) × 24 cm (W). The response levers were mounted 7 cm above the cage floor and 16.5 cm apart on the same wall as the food magazine, which was centered between the levers 2.5 cm above the cage floor. A 1.12-W incandescent lamp, mounted at the top of the wall above the levers, served as a houselight. The enclosures were ventilated by fans, and were located in a temperature- and humidity-controlled chamber. A PDP 11/73 computer was used to control the experiment and record data.
Procedure General. The expermental session for each subject occurred at approximately the same time of day, 5 days per week excluding holidays. No water was available in the experimental chambers. Discrimination training. A two-choice, discrete trial paradigm was used, in which rats were trained to discriminate between two durations that defined the temporal interval to be bisected (0.5 and 5.0 s). For half of the rats, the houselight served as the discriminative stimulus, and for the rest, a 90-dB tone was used. Responses were effective for 10 s following termination of the discriminative stimulus. A left or right lever response was reinforced with a single food pellet following termination of the 5.0- or 0.5-s stimulus, respectively. If no response was made during the 10-s interval, the trial terminated and a null response was recorded. The intertriai interval was 10 s, during which time responses had no effect. Each stimulus occurred with equal probability on each trial. A session consisted of 320 trials, of which the first 20 were warm-up, and every correct response was reinforced. Performance on these trials was used to determine the condition the rat would experience for the next 300 trials. During discrimination training, if the animals produced a minimum of 90% correct responses during the warm-up, the remainder of the session consisted of noncorrected discrimination training, in which the probability of reinforcement was set at a predetermined level. At the end of training and during the remainder of the experiment, the probability of reinforcement was 0.25. Otherwise, a correction procedure was in effect in which an incorrect response to a discriminative stimulus resulted in that stimulus being presented again, 0.2 s following the incorrect response. The proba-
PERCEPTUAL BISECTION AND CHOLINERGICS
bility of reinforcement for a correct response was the same as in the noncorrection procedure. Generalization testing. A maintained generalization procedure was used during the test phase. Five new durations, intermediate to the training stimuli, were presented in addition to the training stimuli. The five new stimuli were equally spaced on a logarithmic scale and were of the following durations: 0.74, 1.1, 1.62, 2.4 and 3.54 s. All seven stimuli were presented in random order throughout the session. Responses to the test stimuli were never reinforced, but responses to the training stimuli continued to be reinforced. Testing sessions occurred on Tuesdays and Fridays. Discrimination training was conducted on the remaining three days of the week. Drug administration. Physostigmine salicylate, scopolamine hydrobromide, (United States Army Medical Research Institute of Chemical Defense, Aberdeen, MD) and pirenzepine hydrochloride (Research Biochemicals, Inc., Natick, MA) were dissolved in 0.9% saline, and 0.9% saline was used for vehicle injections. All injections were administered intraperitoneally (IP) in a volume of 1.0 ml/kg b.wt. Drug solutions were prepared on the day of injection, and all doses are expressed as the salt form. Physostigmine was administered 15 min prior to the beginning of the session. Scopolamine and pirenzepine were administered 45 min prior to the beginning of the session. Phase 1. Saline was administered on the first and last days of generalization testing. On test days 2-6, physostigmine was administered according to the following ascending series: 0.056, 0.100, 0.178, 0.316, and 0.56 mg/kg. Over the next 5 test days, the same doses of physostigmine were administered in descending order. Phase 2. Saline was administered on the first day of generalization testing. On test days 2-6, varying doses of scopolamine were administered in the following mixed order: 0.133, 0.075, 0.10, 0.237 and 0.422 mg/kg. Phase 3. Physostigmine and scopolamine were administered in combination prior to generalization testing. The dose of physostigmine was always 0.56 mg/kg. The doses of scopolamine administered concurrently with physostigmine were 0.075, 0.100, 0.133, and 0.237 mg/kg. In addition to these drug combinations, 0.56 mg/kg physostigmine was administered alone prior to one of the test sessions, and prior to two separate test sessions, saline was administered alone. These treatment conditions were administered in a mixed order. Phase 4. On the first test day, rats were administered 3.0 mg/kg pirenzepine. Saline was administered on test day 2. On test days 3 and 4, rats were administered 10 and 1.0 mg/kg pirenzepine, respectively.
Data Analysis The discriminability between training stimuli on generalization test days was computed for each rat using the following formula: A' = {[(HIT - FA) + (HIT - FA)2]/[4*HIT*(1 FA)]} + 0.5. HIT refers to the probability of a response on the lever appropriate for the long stimulus (i.e., a " l o n g " response), given the long stimulus was presented, and FA represents the probability of a response to the lever appropriate for a long stimulus, given the short stimulus was presented. Values of this index can range from 0.5 (no discrimination) to 1.0 (perfect discrimination). To estimate the BP, the psychometric function relating the proportion of " l o n g " responses to the seven stimulus durations presented during generalization tests was constructed. The BP was determined by interpolation of the stimulus duration to which, during generalization testing, the rat was equally likely
383 I
I
I
i
I
I / / ~
90 (/3 Z © [3_
7o
~
5o
zo
30 ~
~
~
~
P:Ys:~tig~ig~kg)
10
w [3_
J i // 0.4
J
i
0.6
1.0
M 1.6
[] 0.316 II i 0.560 t 2.5
4.0
I 6.3
STIMULUS DURATION (s)
FIG. 1. Mean percent long responses (averaged across determinations) as a function of tone duration and physostigmine dose. In order to produce a linear function, the Y axis uses a probability scale and the x axis uses a log scale. The two solid lines on the graph were determined by the method of least squares and represent the best linear approximation of the saline and 0.56-mg/kg physostigmine conditions.
to respond long or short. The index is used to determine if the drugs altered the perceived duration of the stimuli presented. A longer BP duration suggests stimuli were perceived as shorter than usual, whereas a shorter BP suggests stimuli were perceived as longer than usual. In addition, time to complete the session during generalization testing was calculated in order to assess possible changes in ability to perform the task following drug administration. Significant differences between measures were assessed using repeated-measures ANOVA and Scheffe tests. RESULTS
Phase 1 Figure 1 presents the percent " l o n g " responses (responses to the lever, which produced reinforcers if the training stimulus was 5 s) in generalization testing as a function of tone duration and physostigmine dose. There were orderly changes in duration judgments as the tone duration increased at all drug doses. For all doses, it appears that the BP is approximately constant (2.5 s) as dose increases, but the psychometric functions are less steep with increasing dose, suggesting a decrease in discriminability (A'). Figure 2 shows mean changes in A' as a function of the dose of physostigmine. For both light and tone durations, A' changed significantly as a function of physostigmine dose, F(5,209)= 25.09, p0.05. The BP did not differ between modalities, F(1,178)=3.41, p>0.06, or drug determinations (first vs. second), F(1,209)=0.51, p>0.40. Figure 4 shows time to complete the session as a function of the dose of physostigmine for both determinations. Except for the initial determinations of 0.10, 0.178, 0.316 and 0.56 mg/kg, rats completed all the trials in a session in less than the maximum time allowed. Consequently, analysis showed a significant effect between determinations, F(1,209)= 34.18, p
0031-9384/92 $5.00 + .00
Perceptual Bisection in Rats: The Effects of Physostigmine, Scopolamine and Pirenzepine I DAVID SHURTLEFF 2
Thermal Stress Adaptation Program, Naval Medical Research Institute, Bethesda, MD 20889-5055 T H O M A S G. R A S L E A R
Department of Microwave Research, Walter Reed Army Institute of Research, Washington, DC 20307-5100 RAYMOND
F. G E N O V E S E
AND LARRY SIMMONS
Department of Medical Neurosciences, Walter Reed Army Institute of Research, Washington, DC 20307-5100 R e c e i v e d 4 M a r c h 1991 SHURTLEFF, D., T. G. RASLEAR, R. F. GENOVESE AND L. SIMMONS. Perceptualbisection in rats: The effects ofphysostigmine, scopolamine and pirenzepine. PHYSIOL BEHAV 51(2) 381-390, 1992.--The effects of cholinergic drugs on three different perceptual bisection tasks were studied in rats. Physostigmine (0.056-0.56 mg/kg), a reversible anticholinesterase, produced dose-dependent decrements in discriminability (A'), but did not affect the bisection point (BP) in visual duration, auditory duration, and auditory intensity bisection tasks. This finding is consistent with results previously obtained in an auditory duration bisection task with an irreversible anticholinesterase, diisopropyl phosphorfluoridate. Scopolamine (0.075--0.422 mg/kg), a muscarinic cholinergic-receptor antagonist, produced dose-dependent decrements in both A' and BP in visual and auditory duration bisection tasks. The behavioral antagonism between physostigmine (0.56 mg/kg) and scopolamine (0.075-0.237 mg/kg) was studied in the visual and auditory duration bisection tasks. The BP was not affected by physostigmine alone or in combination with scopolamine, except at the largest dose of scopolamine, which produced a reliable decrement in the BP. A', however, was equally decreased by physostigmine alone and all combinations of physostigmine and scopolamine. Pirenzepine (1, 3 and 10 mg/kg), a selective high-affinity M t muscarinic antagonist, had no effect on A' or the BP in the duration bisection tasks, suggesting changes in perception produced by muscarinic antagonists do not involve the M 1 receptor subtype. The similar drug effects in different sensory modalities (visual and auditory) and perceptual systems (subjective duration and loudness) suggest that cholinergic drugs may affect perceptual mechanisms responsible for sensory coding, such as the output of a neural generator. Physostigmine Rats
Pirenzepine
Scopolamine
Cholinergic
RECENTLY, we reported that the organophosphorus anticholinesterase diisopropyl phosphorofluoridate (DFP) disrupts time discrimination, but not bisection performance, in a temporal bisection procedure (25). Since we had previously demonstrated that DFP produces a long-term disruption of circadian activity and temperature patterns (10, 22, 23), we hypothesized that an organophosphorns anticholinesterase, such as DFP, alters circadian patterns and time perception by directly affecting an "internal clock" (23,26). However, low doses of the anticholinesterase physostigmine have been shown to improve discrimination performance on a
Temporal discrimination
Loudness discrimination
temporal bisection task (15) and on a visual discrimination task (31). Unlike DFP, physostigmine produces a carbamylated acetylcholinesterase (ACHE) enzyme, which prevents the normal enzyme-catalyzed hydrolysis of acetylcholine for several hours (27). DFP, however, is an irreversible anticholinesterase, in that it results in a more stable phosphorylated enzyme, resulting in permanent deactivation of AChE (27). Thus ACHE activity occurs only after the synthesis of new enzyme. It is possible that the differential effects on behavior seen with DFP and physostigmine are determined by their differential effects on ACHE. Alternatively, the effects of anticholinesterase compounds on
~The research described in this report was conducted in compliance with the Animal Welfare Act and other federal statutes and regulations relating to animals and experiments involving animals and adheres to the principles stated in the Guide for the Care and Use of Laboratory Animals, NIH publication 85-23. The views of the author(s) do not purport to reflect the position of the Department of the Army, Department of the Navy or the Department of Defense, (para 4-3, AR 360-5). Portions of this research were presented at the annual meetings of the Eastern Psychological Association in Philadelphia, PA, April 1990; and The American Psychological Society in Dallas, TX, June 1990. 2Requests for reprints should be addressed to Dr. David Shurtleff, Thermal Stress Adaptation Program, MS-11, Naval Medical Research Institute, Bethesda, MD 20889-5055.
381
382
SHURTLEFF, RASLEAR, GENOVESE AND 51MM~)NS
time perception may be dose related, as has been demonstrated by Warburton and Brown (31) for physostigmine with visual discrimination performance. The following experiments examine these possibilities by determining the effects of a range of doses of physostigmine on performance in a temporal bisection procedure using both visual and auditory stimuli. To further investigate the role of the cholinergic system on temporal bisection and the internal clock, a range of doses of the musarinic antagonist scopolamine was administered. To examine the potential behavioral antagonism between physostigmine and scopolamine, various doses of scopolamine were combined with a behaviorally effective dose of physostigmine. We also examined the effects of the muscarinic antagonist pirenzepine on time perception. Pirenzepine has a high affinity for the M~ muscarinic receptor subtype and a low affinity for the M 2 receptor subtype (32). Although highly hydrophilic-systemic administration of pirenzepine has been shown to be behaviorally active (6,33). Finally, to determine the generality of the effects observed with physostigmine in the temporal bisection task, Experiment 2 examined the effects of physostigmine on performance in an auditory intensi~ bisection task. This is of interest because the perception of loudness can be modeled by a process similar to that used in animal timing models [cf. (8,13)]. That is. '~counting models" of loudness perception relate the loudness of an auditory intensity to the number of neural impulses that have occurred within a fixed time period (1). Similarly, recent animal models of time perception (8) postulate an internal clock, or pacemaker, the output of which is accumulated (counted) and used to estimate the duration of an event. The similar theoretical mechanisms of duration and loudness perception suggest that similar drug effects should also be apparent in the auditory intensity bisection task. EXPERIMENT 1 A temporal bisection task was used to explore the perceptual effects of several cholinergic drugs. Separate groups of animals performed the task with visual and auditory durations as stimuli so that sensory-specific effects of the drugs would be discernable from effects specific to duration perception. As previously described (25,26), performance on this task is analyzed within the context of a simple, heuristic, information processing model of animal timing (2, 3, 8). The model contains a pacemaker, or clock, which produces a mean rate of output (ticks/unit time) and an associated variance (variability in clock rate across observation periods). Clock variance can arise from a variety of sources, either inherent to the pacemaker or contributed by other functional elements of the model. For instance, the model contains a memory function which allows a direct comparison of previous and current samples of the clock output. If the memory function becomes impaired, variance is added to the clock output and is indistinguishable from inherent variability. Our analysis provides two different measures of clock function. A' is a nonparametric signal detection theory measure of discriminability (9), and depends upon both the mean clock rate and the variance of the clock [see (25)]. The bisection point (BP) is an estimate of the perceptual midpoint of the temporal interval under consideration, and is a variance-free measure of clock function. Thus the pattern of drug effects on these two measures provides model-relevant information concerning drug action on time perception. Specifically, drugs that affect the mean rate of the clock should produce a change in A' and BP, while drugs that affect the variance of the clock should only produce a change in A'.
METHOD
Subjects Twenty male, albino, Sprague-Dawley rats. 90-t 10 day~ old at the start of the experiment, served. Following a short period of adaptation to the laboratory and individual housing, the rats were reduced to 80% of their free-feeding weight. Water was freely available in the home cages and supplemental chow was provided, as needed, to maintain body weight. A [2-h light:12h dark cycle (lights on at 0600) was maintained throughout the experiment. Two rats did not complete the experiment, and their data was used in the analysis only for the portion of the experiment they did complete.
Apparatus Five similarly constructed operant chambers (Coulbourn Instruments, Model El0-10) were used. Each chamber consisted of a sound- and light-attenuating enclosure that contained a Plexiglas and metal cage with a grid floor, two response levers. a houselight and cue lights, a Sonalert (2.9 kHz), a food magazine, and a pellet feeder which dispensed 45-mg Bio-serv (Frenchtown, NJ) food pellets. The cages measured 30.6 cm (L) × 32.8 (H) × 24 cm (W). The response levers were mounted 7 cm above the cage floor and 16.5 cm apart on the same wall as the food magazine, which was centered between the levers 2.5 cm above the cage floor. A 1.12-W incandescent lamp, mounted at the top of the wall above the levers, served as a houselight. The enclosures were ventilated by fans, and were located in a temperature- and humidity-controlled chamber. A PDP 11/73 computer was used to control the experiment and record data.
Procedure General. The expermental session for each subject occurred at approximately the same time of day, 5 days per week excluding holidays. No water was available in the experimental chambers. Discrimination training. A two-choice, discrete trial paradigm was used, in which rats were trained to discriminate between two durations that defined the temporal interval to be bisected (0.5 and 5.0 s). For half of the rats, the houselight served as the discriminative stimulus, and for the rest, a 90-dB tone was used. Responses were effective for 10 s following termination of the discriminative stimulus. A left or right lever response was reinforced with a single food pellet following termination of the 5.0- or 0.5-s stimulus, respectively. If no response was made during the 10-s interval, the trial terminated and a null response was recorded. The intertriai interval was 10 s, during which time responses had no effect. Each stimulus occurred with equal probability on each trial. A session consisted of 320 trials, of which the first 20 were warm-up, and every correct response was reinforced. Performance on these trials was used to determine the condition the rat would experience for the next 300 trials. During discrimination training, if the animals produced a minimum of 90% correct responses during the warm-up, the remainder of the session consisted of noncorrected discrimination training, in which the probability of reinforcement was set at a predetermined level. At the end of training and during the remainder of the experiment, the probability of reinforcement was 0.25. Otherwise, a correction procedure was in effect in which an incorrect response to a discriminative stimulus resulted in that stimulus being presented again, 0.2 s following the incorrect response. The proba-
PERCEPTUAL BISECTION AND CHOLINERGICS
bility of reinforcement for a correct response was the same as in the noncorrection procedure. Generalization testing. A maintained generalization procedure was used during the test phase. Five new durations, intermediate to the training stimuli, were presented in addition to the training stimuli. The five new stimuli were equally spaced on a logarithmic scale and were of the following durations: 0.74, 1.1, 1.62, 2.4 and 3.54 s. All seven stimuli were presented in random order throughout the session. Responses to the test stimuli were never reinforced, but responses to the training stimuli continued to be reinforced. Testing sessions occurred on Tuesdays and Fridays. Discrimination training was conducted on the remaining three days of the week. Drug administration. Physostigmine salicylate, scopolamine hydrobromide, (United States Army Medical Research Institute of Chemical Defense, Aberdeen, MD) and pirenzepine hydrochloride (Research Biochemicals, Inc., Natick, MA) were dissolved in 0.9% saline, and 0.9% saline was used for vehicle injections. All injections were administered intraperitoneally (IP) in a volume of 1.0 ml/kg b.wt. Drug solutions were prepared on the day of injection, and all doses are expressed as the salt form. Physostigmine was administered 15 min prior to the beginning of the session. Scopolamine and pirenzepine were administered 45 min prior to the beginning of the session. Phase 1. Saline was administered on the first and last days of generalization testing. On test days 2-6, physostigmine was administered according to the following ascending series: 0.056, 0.100, 0.178, 0.316, and 0.56 mg/kg. Over the next 5 test days, the same doses of physostigmine were administered in descending order. Phase 2. Saline was administered on the first day of generalization testing. On test days 2-6, varying doses of scopolamine were administered in the following mixed order: 0.133, 0.075, 0.10, 0.237 and 0.422 mg/kg. Phase 3. Physostigmine and scopolamine were administered in combination prior to generalization testing. The dose of physostigmine was always 0.56 mg/kg. The doses of scopolamine administered concurrently with physostigmine were 0.075, 0.100, 0.133, and 0.237 mg/kg. In addition to these drug combinations, 0.56 mg/kg physostigmine was administered alone prior to one of the test sessions, and prior to two separate test sessions, saline was administered alone. These treatment conditions were administered in a mixed order. Phase 4. On the first test day, rats were administered 3.0 mg/kg pirenzepine. Saline was administered on test day 2. On test days 3 and 4, rats were administered 10 and 1.0 mg/kg pirenzepine, respectively.
Data Analysis The discriminability between training stimuli on generalization test days was computed for each rat using the following formula: A' = {[(HIT - FA) + (HIT - FA)2]/[4*HIT*(1 FA)]} + 0.5. HIT refers to the probability of a response on the lever appropriate for the long stimulus (i.e., a " l o n g " response), given the long stimulus was presented, and FA represents the probability of a response to the lever appropriate for a long stimulus, given the short stimulus was presented. Values of this index can range from 0.5 (no discrimination) to 1.0 (perfect discrimination). To estimate the BP, the psychometric function relating the proportion of " l o n g " responses to the seven stimulus durations presented during generalization tests was constructed. The BP was determined by interpolation of the stimulus duration to which, during generalization testing, the rat was equally likely
383 I
I
I
i
I
I / / ~
90 (/3 Z © [3_
7o
~
5o
zo
30 ~
~
~
~
P:Ys:~tig~ig~kg)
10
w [3_
J i // 0.4
J
i
0.6
1.0
M 1.6
[] 0.316 II i 0.560 t 2.5
4.0
I 6.3
STIMULUS DURATION (s)
FIG. 1. Mean percent long responses (averaged across determinations) as a function of tone duration and physostigmine dose. In order to produce a linear function, the Y axis uses a probability scale and the x axis uses a log scale. The two solid lines on the graph were determined by the method of least squares and represent the best linear approximation of the saline and 0.56-mg/kg physostigmine conditions.
to respond long or short. The index is used to determine if the drugs altered the perceived duration of the stimuli presented. A longer BP duration suggests stimuli were perceived as shorter than usual, whereas a shorter BP suggests stimuli were perceived as longer than usual. In addition, time to complete the session during generalization testing was calculated in order to assess possible changes in ability to perform the task following drug administration. Significant differences between measures were assessed using repeated-measures ANOVA and Scheffe tests. RESULTS
Phase 1 Figure 1 presents the percent " l o n g " responses (responses to the lever, which produced reinforcers if the training stimulus was 5 s) in generalization testing as a function of tone duration and physostigmine dose. There were orderly changes in duration judgments as the tone duration increased at all drug doses. For all doses, it appears that the BP is approximately constant (2.5 s) as dose increases, but the psychometric functions are less steep with increasing dose, suggesting a decrease in discriminability (A'). Figure 2 shows mean changes in A' as a function of the dose of physostigmine. For both light and tone durations, A' changed significantly as a function of physostigmine dose, F(5,209)= 25.09, p0.05. The BP did not differ between modalities, F(1,178)=3.41, p>0.06, or drug determinations (first vs. second), F(1,209)=0.51, p>0.40. Figure 4 shows time to complete the session as a function of the dose of physostigmine for both determinations. Except for the initial determinations of 0.10, 0.178, 0.316 and 0.56 mg/kg, rats completed all the trials in a session in less than the maximum time allowed. Consequently, analysis showed a significant effect between determinations, F(1,209)= 34.18, p
