- Email: [email protected]
Eyeblink conditioning in rats
Physiology& Behavior,Vol. 48, pp. 755-758. ~ Pergamon Press plc, 1990. Printed in the U.S.A.
0031-9384/90 $3.00 + .00
RAPID COMMUNICATION
Eyeblink Conditioning in Rats NESTOR A. SCHMAJUK
A N D B E T H A. C H R I S T I A N S E N
Department of Psychology, Northwestern University, Evanston, IL 60208 R e c e i v e d 26 J u n e 1990
SCHMAJUK, N. A. AND B. A. CHRISTIANSEN. Eyeblink conditioning in rats. PHYSIOL BEHAV 48(5) 755-758, 1990.--Acquisition of eyeblink conditioned responses to a tone was studied in restrained male albino rats. Animals were placed in a modified restraining cage with their heads immobilized by surgically implanting a small bolt at the top of the skull and fastening it to the cage. Eyeblink responses were measured by a phototransistor which detected changes in the level of light reflected by eyelid closure. The conditioned stimulus was a 2,000-Hz, 90-dB, 500-ms tone; the unconditioned stimulus was a 4-psi, 150-ms air puff directed at the cornea; and the interstimulus interval was 350 ms, with a 60-s intertrial interval. Animals received 50 trials daily. All conditioned animals reached a criterion of 80% conditioned responses by Day 10. Pseudoconditioned control animals showed few conditioned responses by Day 10. The topography of the conditioned responses was comparable to those responses found with the nictitating membrane response in the rabbit. The data suggest that eyeblink conditioning in the rat can be used as a preparation to examine the attributes of classical conditioning as well as the neuroanatomicai basis of associative learning. Classical conditioning
Eyeblink conditioning
Albino rat
INTRODUCED by Gormezano, Schneiderman, Deaux and Fuentes (7), the rabbit nictitating membrane response (NMR) preparation has become a prevalent and extensively used technique in classical conditioning (8). While some NMR studies have dealt with advancing classical conditioning theory, others have explored the effects of brain ablations on associative learning, and yet others have focused on the neural substrates of classical conditioning. The success of the rabbit NMR preparation is due to a combination of (a) the passive behavior of the rabbit during restraint, (b) the reliability of the extension of the NM under stimulation by an air puff, (c) the inability of the rabbit to extend its membrane for long periods of time and, consequently, to avoid the air puff (7), and (d) the possibility of an accurate temporal recording of the NMR that permits establishing correlations between neural firing and behavior. Many of the positive aspects of the rabbit NMR preparation also apply to eyeblink conditioning. Although possibly the most common experimental subject, the rat has rarely been used in eyeblink conditioning procedures (I, 4, 6, 9, 11). The rat is particularly appealing as an experimental animal, because of the massive amount of information that has been accumulated regarding its neuroanatomical, neurophysiological, and behavioral features. No less important is the factor that rats are inexpensive to procure and maintain. Motivated by these considerations, the present study presents a new technique to train rats and shows that reliable eyeblink conditioning is obtained with this new preparation.
tained from Harlan Breeding Farms. Animals were housed in individual cages with food and water ad lib, with a 12:12 light-dark cycle. Rats weighed 200-250 g at surgery.
Surgical Procedures Rats were anesthetized with sodium pentobarbital (65 mg/kg, administered intraperitoneally). Atropine sulfate (1 mg/kg) was injected to inhibit secretions in the bronchi that might have blocked respiration. After the rat was anesthetized, the head was shaved and the rat placed in a stereotaxic instrument. The skull was exposed by making an incision down the midline, starting slightly anterior to the level of the eyes and ending just posterior to the ears. The skin was retracted laterally and the skull scraped clean of overlying muscles and connective tissue and allowed to dry. Four 1-mm holes were drilled on the skull, two at the level of the eyes and two at the level of the ears. Each hole was large enough to accommodate a stainless steel machine screw, 1 m m in diameter, and 3 m m in length (Small Parts Inc.). Cyanoacrylate cement (Super Glue) was placed on the skull and around each of the four screws so that a thin layer covered the entire exposed area. A stainless steel bolt, 13 m m in length, 8 m m at the base of the bolt, and 4 mm at the shaft was placed in between the four screws, perpendicular to the skull with the shaft of the bolt extending upwards. Dental acrylic was used to build a base around the screws and the bolt until 6-7 m m of the shaft of the bolt remained exposed. The surrounding skin was pulled up over the acrylic and allowed to dry. Animals were given postoperative care and allowed to recover for one week before habituation began.
METHOD
Subjects
Apparatus
The subjects were 60-day-old male Sprague-Dawley rats, ob-
Animals were restrained in a 108 × 48 mm Fisher Scientific
755
756
SCHMAJUK AND CHRISTIANSEN
100
Percentage CR
CR amplitude (mm) 5
!D
Percentage CRs
I
CR emplitude I
I
80
60
4O
2O
FIG. 1. A rat in the restraining cage (R), with phototransistor and lightemitting diode (D) mounted on a bracket arranged for recording the eyeblink, and tubing for delivering air puff (A).
restraining cage (Fig. 1). The cage was made of clear acrylic and had slots at the top through which an adjustable tailgate could be fit to prevent the rat from backing out. The rat's tail extended through a slot in the tailgate. Another slot extended the length of the cage along the bottom for animal waste. A small metal sheet with a hole in the middle served as the place of attachment for the restraining bolt on the rat's head. Additional pockets and air holes provided for adequate ventilation. A flexible tubing 3.5 m m in internal diameter, installed in a hole located in the front of the cage, served to deliver an airpuff to the eye. Eyeblink was measured, through a 2 × 2 cm window on the side of the cage, with a phototransistor that detects alteration in reflectance of infrared light from a light-emitting diode aimed at the eye (5). The output of the photodiode device was set at 200 mV for each mm of eyelid extension. The restraining cage was placed in an acoustically insulated conditioning chamber, 56 × 37 × 33 cm high, with a circulating fan and a speaker. Behavioral Procedures After the animal was placed in the restraining cage, head movements were eliminated by fastening the restraining bolt to the cage and securing it with a stainless steel nut. The cage with its tailgate in place served to eliminate body movements. Each animal was habituated to restraint in a darkened conditioning chamber for 50 minutes on 4 consecutive days prior to conditioning or pseudoconditioning. Habituation and training procedures were conducted at the same time on successive days. Throughout the experiment, a 72 dB background noise was maintained by the circulating fan in the conditioning chamber. After habituation, the rat was trained in a delay conditioning paradigm. The conditioned stimulus (CS) was a 500-ms, 2,000Hz, 92-dB tone generated by a Hewlett-Packard audio oscillator and amplified by a Realistic amplifier. The unconditioned stimulus (US) was a 150-ms, 4-psi air puff of compressed nitrogen controlled by a solenoid valve and delivered to the cornea through a flexible tubing with a 3.5-mm orifice placed 5 mm from the eye. The interstimulus interval (ISI) was 350 ms in duration. Execution of each trial and data collection were controlled by a Brain Wave computer system. During each trial, the eyeblink response was recorded for 400 ms before CS onset and 1,100 ms after the US offset (see Fig. 3). Simulus onset and offset were controlled by the microcomputer through an interface connected to the solenoid and tone generator. Fifty conditioning trials were administered on consecutive days until each rat reached a criterion of 80% total CRs on a given day. Each daily session lasted approximately 50 minutes. The intertrial interval (ITI) varied
0
-
0 2
3
4
5
6
7
8
9
10
Tenths to criterion FIG. 2. Percentage of CRs and CR amplitude during the CS-US period plotted as a function of Tenths to Criterion (TTC).
pseudorandomly from 40 s to 70 s and averaged 60 s. During pseudoconditioning, control animals received 50 unpaired puffs and 50 unpaired tones for a total of 100 presentations. The tones and puffs were presented pseudorandomly with the ITI varying from 20 to 50 s and averaging 30 s. All other parameters were the same as those for the experimental animals. After each animal reached 80% CR responding on a given day, extinction followed on the next day. Extinction consisted of 50 CS-alone presentations, delivered during consecutive days until rats achieved a criterion of 1 CR in a block of 10 consecutive trials. Six experimental animals were run until each one reached a criterion of 80% CRs on a given day. Six control animals were run, each for the same amount of trials as those taken by a matched experimental animal to achieve criterion. CRs were defined as those eyeblink responses with (a) onset latency longer than 100 ms and (b) amplitude greater than the mean plus 10 times the standard deviation of the amplitude of the response generated during the 400-ms pre-CS period [this criterion is equivalent to a 0.50 mm (_+0.28 S.D.) extension of the eyelid]. RESULTS
Two distinct distributions of responses, with onset latencies of 52 ms (-+20 S.D.) and 261 (_+57 S.D.), were observed. Responses with onset latencies shorter than 100 ms were classified as alpha responses and found on 4% of the total trials. When they did occur, these short-latency responses had an amplitude of 0.95 mm (-+0.49 S.D.) and a time course of 55 ms (_+ 10 S.D.). Alpha responses were completely absent in some animals. All rats achieved a criterion of 8 CRs in a block of l0 consecutive trials in 253 trials (-+ 135 S.D.). Training continued until each rat reached a criterion of 80% CRs on a given day (or 40 out of 50 possible CRs). This second criterion was met in 408 trials (-+ 80 S.D.). Control animals showed a total of 2. i% CRs (-+ 1.7 S.D.), with a maximum of 12% CRs on a given day. The extinction criterion of 10% CR rate in a block of l0 trials was reached in 38 trials (-+ 12 S.D.). Figure 2 shows a Vincentized bar diagram indicating CR amplitude and percentage of CRs as a function of Tenths to Criterion (TTCs). Analysis of variance of the percentage of CRs revealed a significant effect of TTCs, F(9,90) = 16.39, p < 0 . 0 0 0 1 , treatment, F ( 1 , 9 0 ) = 2 5 . 7 8 , p < 0 . 0 0 0 5 , and interaction, F(9,90) = 12.47, p < 0 . 0 0 0 1 . When animals reached the 80%
EYEBLINK CONDITIONING IN RATS
757
Eyeblink responses
E
10
t-
8
,mO
"t-
0
6
(/) ¢-
t--
4
CS
7-]
US I
0
I
I
I I[ I 1It
1I
I
I I
I
I
1
II I
2
sec
FIG. 3. Typical eyeblink responses as a function of Tenths to Criterion during acquisitionand during the first day of extinction.
CR responding criterion, they showed a CR amplitude of 2.82 mm (-+ 1.60 S.D.). CR amplitude and percentage of CRs were strongly correlated, r(18)= .96, p<0.001. Figure 3 shows the topography of typical eyeblinks on different TTCs. On TlvCs 1 and 2, rats rarely display CRs. During TFC 4, the unconditioned response (UR) extends after the US offset and small CRs start to appear. As TTCs progress, CR amplitude increases and CR latency decreases. CR latency initially just preceded the US with a mean latency of 309 ms (--_30 S.D.) and moved towards the onset of the CS with a mean latency of 214 ms (-+48 S.D.) over increasing trials. A t-test reveals that CR onset latency decreases significantly from the beginning to the end of training, t(10) = 3.72, p<0.01. During extinction, CRs decreased in amplitude and peaked approximately at the time of the US presentation.
DISCUSSION
The present study shows that eyeblink conditioning can be reliably obtained in the restrained rat. The rate of acquisition of eyeblink conditioning in rats seems to be somewhat slower than that of the rabbit NMR. Whereas rats achieved a criterion of 80% eyeblink CR responding on a given day in roughly 400 trials, rabbits took an approximately similar number of trials to reach comparable criterion (7). However, when an acquisition criterion
of 8 CRs in a block of 10 trials was adopted, rats achieved criterion in nearly 250 trials, while rabbits attained the comparable level of responding in around 120 trials (3). While acquisition might be slower in rats, they seem to extinguish faster than rabbits. An extinction criterion of 10% CRs in a block of 10 trials was reached by rats in around 40 trials, while rabbits reach a less stringent criterion in 500 trials (2). However, these comparisons should be taken with caution because the rat and rabbit studies described here differ in a number of important conditioning parameters including ISI, ITI, CS intensity and frequency, as well as in the number of daily trials. Figure 3 shows that the topography of the eyeblink response in the rat is remarkably similar to the NMR topography displayed by rabbits [see (10,12)]. In a typical classical conditioning paradigm, the rabbit acquires the conditioned NMR with an orderly sequence of changes: percentage of NMRs generated in each session increases, CR amplitude increases, and CR latency decreases. At the beginning of training, the first CRs are initiated just before the US, but initiation moves to progressively earlier portions of the CS-US interval with an asymptotic latency occurring at about the midpoint of the ISI. The maximal response amplitude (CR peak) tends to be located around the time of the US occurrence. In the present study similar results were obtained. The percentage of eyeblink responses increased over trials as did the CR amplitude. Further, the CR latency initially just preceded the US and over increasing trials moved towards the onset of the CS. Lastly, the peak amplitude of the CR usually coincided with the time of the US presentation. The resemblance between the eyeblink responses in the rat and NMR topography in the rabbit extend, not only to the CS period, but also to the US period. For instance, the UR was on numerous occasions followed by a small opening of the eye in response to the onset of the puff, a phenomenon also reported in the rabbit NMR preparation and attributed to the pressure of the airpuff which forces the eyelid open (5). Recently, Skelton (11) described a new technique for eyelid conditioning in the unrestrained rat. Electromyographic eyelid CRs were recorded by subdermally implanting stainless steel wires under the skin of the upper eyelid. A shock US was delivered through bipolar electrodes subdermally implanted to positions caudal and dorsocaudal to the eye. Our method seems to offer some advantages over Skelton's (11) procedure. First, contrasting with data obtained in unrestrained rats, our animals showed very few alpha responses. Second, because Skelton's approach uses electromyography to record eyeblink responses, it cannot sample the eyeblink during the US period due to the large stimulation artifact. In comparison, our approach uses a noninvasive photoelectric technique to record eyeblink responses which permits the recording of the UR during the US presentation. Third, Skelton reported that US intensities which elicited only eyeblinks did not support conditioning, and, therefore, US intensities were adjusted to elicit both a strong eyeblink and a rapid head jerk away from the US site. In contrast with Skelton's method, US intensities which only elicited eyeblinks did support conditioning in our procedure. In summary, the restrained rat eyeblink preparation described in the present paper seems to be a promising experimental technique. As in the case of the rabbit NMR preparation, positive features of this preparation are (a) the passive behavior of the rat during restraint, (b) the reliability of the extension of the eyelid under stimulation by an air puff, (c) the inability of the rat to close its eye for long periods of time and, consequently, to avoid the air puff, (d) the almost negligible spontaneous blinking, and (e) the possibility of an accurate temporal recording of the eyeblink that permits establishing correlations between neural firing and behavior.
758
SCHMAJUK AND CHRISTIANSEN
ACKNOWLEDGEMENTS We are indebted to Drs. John Disterhoft, Peter Frey and Richard Deyo for their technical assistance throughout the progress of this project. We are specially grateful to Dr. Tres Thompson for providing the computer program used to control the experiment and store data on the Brain Wave computer system. We would like also to thank Drs. M. Glickstein, I.
Gormezano, D. G. Lavond, J. W. Moore, M. M. Patterson, R. F. Thompson and C. Weiss for their valuable comments on an earlier version of this manuscript. Travis Seymour provided assistance in different aspects of the experiment. This project was supported in part by BRSG SO7 RR07028-22 awarded by the Biomedical Research Support Grant Program, Division of Research Resources, National Institutes of Health, to Northwestern University.
REFERENCES 1. Adams, R. M.; Zhang, A. A.; Lavond, D. G. Neural unit activity in cerebellar interpositus nucleus models classically conditioned eyelid responses in the rat. Soc. Neurosci. Abst. 15:890; 1989. 2. Berger, T. W.; Orr, W. B. Hippocampectomy selectively disrupts discrimination reversal conditioning of the rabbit nictitating membrane response. Behav. Brain Res. 8:49-68; 1983. 3. Berger, T. W.; Thompson, R. F. Neuronal plasticity in the limbic system during classical conditioning of the rabbit nictitating membrane response. I. The hippocampus. Brain Res. 145:323-346; 1978. 4. Biel, W. C.; Wickens, D. D. The effects of vitamin B~ deficiency on the conditioning of eyelid responses in the rat. J. Comp. Physiol. Psychol. 32:329-340; 1941. 5. Disterhoft, J. F.; Kwan, H. H.; Lo, W. D. Nictitating membrane conditioning to tone in the immobilized albino rat. Brain Res. 137: 127-143; 1977. 6. Ebel, H. C. A restraining device for use in the measurement of eyelid responses in laboratory rats. J. Exp. Anal. Behav. 9:605-606; 1966.
7. Gormezano, I.; Schneiderman, N.; Deaux, E. G.; Fuentes, I. Nictitaring membrane: classical conditioning and extinction in the albino rabbit. Science 138:33-34; 1962. 8. Gormezano, I.; Kehoe, E. J.; Marshall, B. S. Twenty years of classical conditioning research with the rabbit. Prog. Psychobiol. Physiol. Psychol. 10:197-275; 1983. 9. Hughes, B.; Schlosberg, H. Conditioning in the white rat. IV. The conditioned lid reflex. J. Exp. Psychol. 23:641-650; 1938. 10. Schneiderman, N. Interstimulus interval function of the nictitating membrane response of the rabbit under delay versus trace conditioning. J. Comp. Physiol. Psychol. 62:397-402; 1966 11. Skelton, R. W. Bilateral cerebellar lesions disrupt conditioned eyelid responses in unrestrained rats. Behav. Neurosci. 102:586-590; 1988. 12. Smith, M. C. CS-US interval and US intensity in classical conditioning of the rabbit's nictitating membrane response. J. Comp. Physiol. Psychol. 66:679-687; 1968.
0031-9384/90 $3.00 + .00
RAPID COMMUNICATION
Eyeblink Conditioning in Rats NESTOR A. SCHMAJUK
A N D B E T H A. C H R I S T I A N S E N
Department of Psychology, Northwestern University, Evanston, IL 60208 R e c e i v e d 26 J u n e 1990
SCHMAJUK, N. A. AND B. A. CHRISTIANSEN. Eyeblink conditioning in rats. PHYSIOL BEHAV 48(5) 755-758, 1990.--Acquisition of eyeblink conditioned responses to a tone was studied in restrained male albino rats. Animals were placed in a modified restraining cage with their heads immobilized by surgically implanting a small bolt at the top of the skull and fastening it to the cage. Eyeblink responses were measured by a phototransistor which detected changes in the level of light reflected by eyelid closure. The conditioned stimulus was a 2,000-Hz, 90-dB, 500-ms tone; the unconditioned stimulus was a 4-psi, 150-ms air puff directed at the cornea; and the interstimulus interval was 350 ms, with a 60-s intertrial interval. Animals received 50 trials daily. All conditioned animals reached a criterion of 80% conditioned responses by Day 10. Pseudoconditioned control animals showed few conditioned responses by Day 10. The topography of the conditioned responses was comparable to those responses found with the nictitating membrane response in the rabbit. The data suggest that eyeblink conditioning in the rat can be used as a preparation to examine the attributes of classical conditioning as well as the neuroanatomicai basis of associative learning. Classical conditioning
Eyeblink conditioning
Albino rat
INTRODUCED by Gormezano, Schneiderman, Deaux and Fuentes (7), the rabbit nictitating membrane response (NMR) preparation has become a prevalent and extensively used technique in classical conditioning (8). While some NMR studies have dealt with advancing classical conditioning theory, others have explored the effects of brain ablations on associative learning, and yet others have focused on the neural substrates of classical conditioning. The success of the rabbit NMR preparation is due to a combination of (a) the passive behavior of the rabbit during restraint, (b) the reliability of the extension of the NM under stimulation by an air puff, (c) the inability of the rabbit to extend its membrane for long periods of time and, consequently, to avoid the air puff (7), and (d) the possibility of an accurate temporal recording of the NMR that permits establishing correlations between neural firing and behavior. Many of the positive aspects of the rabbit NMR preparation also apply to eyeblink conditioning. Although possibly the most common experimental subject, the rat has rarely been used in eyeblink conditioning procedures (I, 4, 6, 9, 11). The rat is particularly appealing as an experimental animal, because of the massive amount of information that has been accumulated regarding its neuroanatomical, neurophysiological, and behavioral features. No less important is the factor that rats are inexpensive to procure and maintain. Motivated by these considerations, the present study presents a new technique to train rats and shows that reliable eyeblink conditioning is obtained with this new preparation.
tained from Harlan Breeding Farms. Animals were housed in individual cages with food and water ad lib, with a 12:12 light-dark cycle. Rats weighed 200-250 g at surgery.
Surgical Procedures Rats were anesthetized with sodium pentobarbital (65 mg/kg, administered intraperitoneally). Atropine sulfate (1 mg/kg) was injected to inhibit secretions in the bronchi that might have blocked respiration. After the rat was anesthetized, the head was shaved and the rat placed in a stereotaxic instrument. The skull was exposed by making an incision down the midline, starting slightly anterior to the level of the eyes and ending just posterior to the ears. The skin was retracted laterally and the skull scraped clean of overlying muscles and connective tissue and allowed to dry. Four 1-mm holes were drilled on the skull, two at the level of the eyes and two at the level of the ears. Each hole was large enough to accommodate a stainless steel machine screw, 1 m m in diameter, and 3 m m in length (Small Parts Inc.). Cyanoacrylate cement (Super Glue) was placed on the skull and around each of the four screws so that a thin layer covered the entire exposed area. A stainless steel bolt, 13 m m in length, 8 m m at the base of the bolt, and 4 mm at the shaft was placed in between the four screws, perpendicular to the skull with the shaft of the bolt extending upwards. Dental acrylic was used to build a base around the screws and the bolt until 6-7 m m of the shaft of the bolt remained exposed. The surrounding skin was pulled up over the acrylic and allowed to dry. Animals were given postoperative care and allowed to recover for one week before habituation began.
METHOD
Subjects
Apparatus
The subjects were 60-day-old male Sprague-Dawley rats, ob-
Animals were restrained in a 108 × 48 mm Fisher Scientific
755
756
SCHMAJUK AND CHRISTIANSEN
100
Percentage CR
CR amplitude (mm) 5
!D
Percentage CRs
I
CR emplitude I
I
80
60
4O
2O
FIG. 1. A rat in the restraining cage (R), with phototransistor and lightemitting diode (D) mounted on a bracket arranged for recording the eyeblink, and tubing for delivering air puff (A).
restraining cage (Fig. 1). The cage was made of clear acrylic and had slots at the top through which an adjustable tailgate could be fit to prevent the rat from backing out. The rat's tail extended through a slot in the tailgate. Another slot extended the length of the cage along the bottom for animal waste. A small metal sheet with a hole in the middle served as the place of attachment for the restraining bolt on the rat's head. Additional pockets and air holes provided for adequate ventilation. A flexible tubing 3.5 m m in internal diameter, installed in a hole located in the front of the cage, served to deliver an airpuff to the eye. Eyeblink was measured, through a 2 × 2 cm window on the side of the cage, with a phototransistor that detects alteration in reflectance of infrared light from a light-emitting diode aimed at the eye (5). The output of the photodiode device was set at 200 mV for each mm of eyelid extension. The restraining cage was placed in an acoustically insulated conditioning chamber, 56 × 37 × 33 cm high, with a circulating fan and a speaker. Behavioral Procedures After the animal was placed in the restraining cage, head movements were eliminated by fastening the restraining bolt to the cage and securing it with a stainless steel nut. The cage with its tailgate in place served to eliminate body movements. Each animal was habituated to restraint in a darkened conditioning chamber for 50 minutes on 4 consecutive days prior to conditioning or pseudoconditioning. Habituation and training procedures were conducted at the same time on successive days. Throughout the experiment, a 72 dB background noise was maintained by the circulating fan in the conditioning chamber. After habituation, the rat was trained in a delay conditioning paradigm. The conditioned stimulus (CS) was a 500-ms, 2,000Hz, 92-dB tone generated by a Hewlett-Packard audio oscillator and amplified by a Realistic amplifier. The unconditioned stimulus (US) was a 150-ms, 4-psi air puff of compressed nitrogen controlled by a solenoid valve and delivered to the cornea through a flexible tubing with a 3.5-mm orifice placed 5 mm from the eye. The interstimulus interval (ISI) was 350 ms in duration. Execution of each trial and data collection were controlled by a Brain Wave computer system. During each trial, the eyeblink response was recorded for 400 ms before CS onset and 1,100 ms after the US offset (see Fig. 3). Simulus onset and offset were controlled by the microcomputer through an interface connected to the solenoid and tone generator. Fifty conditioning trials were administered on consecutive days until each rat reached a criterion of 80% total CRs on a given day. Each daily session lasted approximately 50 minutes. The intertrial interval (ITI) varied
0
-
0 2
3
4
5
6
7
8
9
10
Tenths to criterion FIG. 2. Percentage of CRs and CR amplitude during the CS-US period plotted as a function of Tenths to Criterion (TTC).
pseudorandomly from 40 s to 70 s and averaged 60 s. During pseudoconditioning, control animals received 50 unpaired puffs and 50 unpaired tones for a total of 100 presentations. The tones and puffs were presented pseudorandomly with the ITI varying from 20 to 50 s and averaging 30 s. All other parameters were the same as those for the experimental animals. After each animal reached 80% CR responding on a given day, extinction followed on the next day. Extinction consisted of 50 CS-alone presentations, delivered during consecutive days until rats achieved a criterion of 1 CR in a block of 10 consecutive trials. Six experimental animals were run until each one reached a criterion of 80% CRs on a given day. Six control animals were run, each for the same amount of trials as those taken by a matched experimental animal to achieve criterion. CRs were defined as those eyeblink responses with (a) onset latency longer than 100 ms and (b) amplitude greater than the mean plus 10 times the standard deviation of the amplitude of the response generated during the 400-ms pre-CS period [this criterion is equivalent to a 0.50 mm (_+0.28 S.D.) extension of the eyelid]. RESULTS
Two distinct distributions of responses, with onset latencies of 52 ms (-+20 S.D.) and 261 (_+57 S.D.), were observed. Responses with onset latencies shorter than 100 ms were classified as alpha responses and found on 4% of the total trials. When they did occur, these short-latency responses had an amplitude of 0.95 mm (-+0.49 S.D.) and a time course of 55 ms (_+ 10 S.D.). Alpha responses were completely absent in some animals. All rats achieved a criterion of 8 CRs in a block of l0 consecutive trials in 253 trials (-+ 135 S.D.). Training continued until each rat reached a criterion of 80% CRs on a given day (or 40 out of 50 possible CRs). This second criterion was met in 408 trials (-+ 80 S.D.). Control animals showed a total of 2. i% CRs (-+ 1.7 S.D.), with a maximum of 12% CRs on a given day. The extinction criterion of 10% CR rate in a block of l0 trials was reached in 38 trials (-+ 12 S.D.). Figure 2 shows a Vincentized bar diagram indicating CR amplitude and percentage of CRs as a function of Tenths to Criterion (TTCs). Analysis of variance of the percentage of CRs revealed a significant effect of TTCs, F(9,90) = 16.39, p < 0 . 0 0 0 1 , treatment, F ( 1 , 9 0 ) = 2 5 . 7 8 , p < 0 . 0 0 0 5 , and interaction, F(9,90) = 12.47, p < 0 . 0 0 0 1 . When animals reached the 80%
EYEBLINK CONDITIONING IN RATS
757
Eyeblink responses
E
10
t-
8
,mO
"t-
0
6
(/) ¢-
t--
4
CS
7-]
US I
0
I
I
I I[ I 1It
1I
I
I I
I
I
1
II I
2
sec
FIG. 3. Typical eyeblink responses as a function of Tenths to Criterion during acquisitionand during the first day of extinction.
CR responding criterion, they showed a CR amplitude of 2.82 mm (-+ 1.60 S.D.). CR amplitude and percentage of CRs were strongly correlated, r(18)= .96, p<0.001. Figure 3 shows the topography of typical eyeblinks on different TTCs. On TlvCs 1 and 2, rats rarely display CRs. During TFC 4, the unconditioned response (UR) extends after the US offset and small CRs start to appear. As TTCs progress, CR amplitude increases and CR latency decreases. CR latency initially just preceded the US with a mean latency of 309 ms (--_30 S.D.) and moved towards the onset of the CS with a mean latency of 214 ms (-+48 S.D.) over increasing trials. A t-test reveals that CR onset latency decreases significantly from the beginning to the end of training, t(10) = 3.72, p<0.01. During extinction, CRs decreased in amplitude and peaked approximately at the time of the US presentation.
DISCUSSION
The present study shows that eyeblink conditioning can be reliably obtained in the restrained rat. The rate of acquisition of eyeblink conditioning in rats seems to be somewhat slower than that of the rabbit NMR. Whereas rats achieved a criterion of 80% eyeblink CR responding on a given day in roughly 400 trials, rabbits took an approximately similar number of trials to reach comparable criterion (7). However, when an acquisition criterion
of 8 CRs in a block of 10 trials was adopted, rats achieved criterion in nearly 250 trials, while rabbits attained the comparable level of responding in around 120 trials (3). While acquisition might be slower in rats, they seem to extinguish faster than rabbits. An extinction criterion of 10% CRs in a block of 10 trials was reached by rats in around 40 trials, while rabbits reach a less stringent criterion in 500 trials (2). However, these comparisons should be taken with caution because the rat and rabbit studies described here differ in a number of important conditioning parameters including ISI, ITI, CS intensity and frequency, as well as in the number of daily trials. Figure 3 shows that the topography of the eyeblink response in the rat is remarkably similar to the NMR topography displayed by rabbits [see (10,12)]. In a typical classical conditioning paradigm, the rabbit acquires the conditioned NMR with an orderly sequence of changes: percentage of NMRs generated in each session increases, CR amplitude increases, and CR latency decreases. At the beginning of training, the first CRs are initiated just before the US, but initiation moves to progressively earlier portions of the CS-US interval with an asymptotic latency occurring at about the midpoint of the ISI. The maximal response amplitude (CR peak) tends to be located around the time of the US occurrence. In the present study similar results were obtained. The percentage of eyeblink responses increased over trials as did the CR amplitude. Further, the CR latency initially just preceded the US and over increasing trials moved towards the onset of the CS. Lastly, the peak amplitude of the CR usually coincided with the time of the US presentation. The resemblance between the eyeblink responses in the rat and NMR topography in the rabbit extend, not only to the CS period, but also to the US period. For instance, the UR was on numerous occasions followed by a small opening of the eye in response to the onset of the puff, a phenomenon also reported in the rabbit NMR preparation and attributed to the pressure of the airpuff which forces the eyelid open (5). Recently, Skelton (11) described a new technique for eyelid conditioning in the unrestrained rat. Electromyographic eyelid CRs were recorded by subdermally implanting stainless steel wires under the skin of the upper eyelid. A shock US was delivered through bipolar electrodes subdermally implanted to positions caudal and dorsocaudal to the eye. Our method seems to offer some advantages over Skelton's (11) procedure. First, contrasting with data obtained in unrestrained rats, our animals showed very few alpha responses. Second, because Skelton's approach uses electromyography to record eyeblink responses, it cannot sample the eyeblink during the US period due to the large stimulation artifact. In comparison, our approach uses a noninvasive photoelectric technique to record eyeblink responses which permits the recording of the UR during the US presentation. Third, Skelton reported that US intensities which elicited only eyeblinks did not support conditioning, and, therefore, US intensities were adjusted to elicit both a strong eyeblink and a rapid head jerk away from the US site. In contrast with Skelton's method, US intensities which only elicited eyeblinks did support conditioning in our procedure. In summary, the restrained rat eyeblink preparation described in the present paper seems to be a promising experimental technique. As in the case of the rabbit NMR preparation, positive features of this preparation are (a) the passive behavior of the rat during restraint, (b) the reliability of the extension of the eyelid under stimulation by an air puff, (c) the inability of the rat to close its eye for long periods of time and, consequently, to avoid the air puff, (d) the almost negligible spontaneous blinking, and (e) the possibility of an accurate temporal recording of the eyeblink that permits establishing correlations between neural firing and behavior.
758
SCHMAJUK AND CHRISTIANSEN
ACKNOWLEDGEMENTS We are indebted to Drs. John Disterhoft, Peter Frey and Richard Deyo for their technical assistance throughout the progress of this project. We are specially grateful to Dr. Tres Thompson for providing the computer program used to control the experiment and store data on the Brain Wave computer system. We would like also to thank Drs. M. Glickstein, I.
Gormezano, D. G. Lavond, J. W. Moore, M. M. Patterson, R. F. Thompson and C. Weiss for their valuable comments on an earlier version of this manuscript. Travis Seymour provided assistance in different aspects of the experiment. This project was supported in part by BRSG SO7 RR07028-22 awarded by the Biomedical Research Support Grant Program, Division of Research Resources, National Institutes of Health, to Northwestern University.
REFERENCES 1. Adams, R. M.; Zhang, A. A.; Lavond, D. G. Neural unit activity in cerebellar interpositus nucleus models classically conditioned eyelid responses in the rat. Soc. Neurosci. Abst. 15:890; 1989. 2. Berger, T. W.; Orr, W. B. Hippocampectomy selectively disrupts discrimination reversal conditioning of the rabbit nictitating membrane response. Behav. Brain Res. 8:49-68; 1983. 3. Berger, T. W.; Thompson, R. F. Neuronal plasticity in the limbic system during classical conditioning of the rabbit nictitating membrane response. I. The hippocampus. Brain Res. 145:323-346; 1978. 4. Biel, W. C.; Wickens, D. D. The effects of vitamin B~ deficiency on the conditioning of eyelid responses in the rat. J. Comp. Physiol. Psychol. 32:329-340; 1941. 5. Disterhoft, J. F.; Kwan, H. H.; Lo, W. D. Nictitating membrane conditioning to tone in the immobilized albino rat. Brain Res. 137: 127-143; 1977. 6. Ebel, H. C. A restraining device for use in the measurement of eyelid responses in laboratory rats. J. Exp. Anal. Behav. 9:605-606; 1966.
7. Gormezano, I.; Schneiderman, N.; Deaux, E. G.; Fuentes, I. Nictitaring membrane: classical conditioning and extinction in the albino rabbit. Science 138:33-34; 1962. 8. Gormezano, I.; Kehoe, E. J.; Marshall, B. S. Twenty years of classical conditioning research with the rabbit. Prog. Psychobiol. Physiol. Psychol. 10:197-275; 1983. 9. Hughes, B.; Schlosberg, H. Conditioning in the white rat. IV. The conditioned lid reflex. J. Exp. Psychol. 23:641-650; 1938. 10. Schneiderman, N. Interstimulus interval function of the nictitating membrane response of the rabbit under delay versus trace conditioning. J. Comp. Physiol. Psychol. 62:397-402; 1966 11. Skelton, R. W. Bilateral cerebellar lesions disrupt conditioned eyelid responses in unrestrained rats. Behav. Neurosci. 102:586-590; 1988. 12. Smith, M. C. CS-US interval and US intensity in classical conditioning of the rabbit's nictitating membrane response. J. Comp. Physiol. Psychol. 66:679-687; 1968.