Brain Behav. Evol. 13: 116-124 (1976)
Frontal Cortex and Response Suppression in the Rat R. C. W ilcott , B. A. Sabol and R. P. Y urcheshen Department of Psychology, Case Western Reserve University, Cleveland, Ohio
Key Words. Frontal cortex • Electrical stimulation • Response suppression • Rat Abstract. Parts of the rat’s neocortex were mapped for sites where electrical stimulation, square waves at 10/sec, will suppress a bar-press response for food. Effective sites were found in the frontal pole and over most of the frontal dorso lateral cortex. Within these regions the strongest inhibitory influences were at sites in the frontal pole and adjacent frontal cortex, and at sites along the midline. These data generally agree with results of previous cortical ablation studies, but they suggest that inhibitory processes are more widespread in the dorsolateral cor tex than these studies indicated.
In monkeys, dogs and cats, increased skeletal and autonomic activity, and hyperreactivity, are usually observed following removal of parts of the frontal cortex including the prefrontal area [B ru tkow ski , 1965], These results seem to be due to release of a general inhibition normally maintained by parts of the frontal cortex. In the rat, removal of the frontal poles will produce increased activ ity, particularly as measured by a running wheel [R ichter and H aw kes , 1939; C ampbell and L ynch , 1969]. The same effects have been ob served following removal of anterior medial or dorsal rhinal sulcus cor tex [K olb , 1974]. Based on recent studies of thalamic projections, these latter areas may represent ‘prefrontal’ cortex in the rat more closely than the frontal poles [L eonard , 1968]. These studies suggest that parts of the frontal cortex in the rat, apparently corresponding to those parts
Downloaded by: King's College London 137.73.144.138 - 1/21/2019 12:32:32 AM
Introduction
Wilcott/S abol/Y urcheshen
117
identified in higher mammals, also have an inhibitory influence on be havior. Electrical stimulation has also been used to study inhibitory processes in the frontal cortex. Stimulation in the rat’s dorsolateral frontal cortex, with electrical square waves at a frequency of 0.5-30/sec, will markedly reduce the heart rate response to a noxious stimulus [G laser and G r if fin , 1962], All sites tested anterior to the bregma were effective, while stimulation at posterior sites had no effect on the heart rate response. In the cat, stimulation at a relatively low frequency (5-15/sec) in the pre frontal cortex will completely suppress a bar-press response for food, and partially suppress skin potential responses [W ilcott and B radley , 1970; W ilcott , 1974]. The cat’s dorsolateral cortex has also been part ly mapped for sites where stimulation will suppress the bar press re sponse [W ilcott and Sabol , 1976]. Results generally agree with inhibi tory areas of activity, etc., established by the ablation method [L ang worth and R ich ter , 1939], but they seem to show two things more ac curately. These are the extent of inhibitory areas in the dorsolateral frontal cortex, and the relative magnitude of inhibition produced at dif ferent areas. Because response suppression observed in these stimulation studies mainly occurs during the time stimulation is presented, response sup pression could have been due to extraneous factors such as production of competing responses, or disruption of neural tissue at the site of stim ulation. Available evidence, however, is opposed to these alternate inter pretations [W ilcott and B radley , 1970; W ilcott and Sabol, 1976]. The stimuation technique, particularly at a low frequency, seems to be a useful method for exploring inhibitory processes in the cortex. As documented below, it has been found that low-frequency stimula tion in the rat’s cortex will suppress a bar-press response. The aim of this study was to map parts of the rat’s dorsolateral cortex and frontal poles for sites where stimulation will suppress bar-pressing for food. Method
Electrodes and Stimulation Electrodes were obtained from Plastics Products Co. They consisted of pairs of fine stainless steel wires, insulated with Formvar varnish, wound together and cou-
Downloaded by: King's College London 137.73.144.138 - 1/21/2019 12:32:32 AM
Subjects Results are reported on 21 male Holtzman rats, weight ranging from about 400 to 420 g at the time of surgery.
W ilcott /S abol/Y urchesmen
ns
2
© ©
c
3 4
\
\
J
• o' 6 7 8 9 21 ' ©€ o • 0 10 11 22 12 18 6 0 O O © U 15 23 24 16 O A • o o\ 25 38 26 ^ 7 \ 17 0 O O O A ,-A' O \ 18 28 29 30 3 ? / 4 0 31 \ O \ 'o 0 O '41 \ 33 34 19 32 O O A A A A\ B- »2 0 .35 42 36 43 44 45 461 A A 'O A 47
32
48
49
A 51
A 50
à
A 53
52
A 54
A 55
A 59
A 60
A 56 A 58 A 61 \63 A A
J
A
65
A 57/
66
tJ
/
/
62/ A /
■ ¿ 67.
pled to a small molded socket. The wires were cut so that only their tips were ex posed. Rats were anesthetized with pentobarbital sodium. The dorsolateral cortex electrodes were implanted slightly within the cortex under direct visual control with the aid of a stereotaxic instrument. Four pairs of electrodes were implanted in each rat, two in the cortex on each side. Their positions were predetermined from sites marked on a drawing of the rat cortex using the bregma as a reference point. At necropsy brains were examined grossly for positions of electrodes and sites marked on a brain diagram (fig. 1). The lateral position of electrode sites was measured from the midline and stereotaxic measurement data corrected in a few cases. However, it was decided that the stereotaxic measurement was the best ap proximation of the anterior or posterior positions of an electrode site and this was used. Preliminary observations indicated that the depth of electrodes at the dorsolat eral cortex was not critical. Response suppression could be obtained just as well with electrode tips slightly protruding into white matter as with electrode tips in gray matter. The probable reason for this is that response suppression is produced by stimulation of cells and/or fibers making connections with subcortical struc tures [Wilcott and Sabol, 1965]. As a spot check on electrode depths at these
Downloaded by: King's College London 137.73.144.138 - 1/21/2019 12:32:32 AM
F ig.l. Electrode sites within the frontal pole (left) and in the dorsolateral cor tex (right). Closed circles are sites where response suppression was 75% or better, at open circles response suppression was 18-62% and always significant, and at triangle sites response suppression was 14% or less and never significant. The cross section at the frontal pole is about 6 mm anterior to the bregma. The dashed lines at the dorsolateral cortex mark the approximate anterior and posterior mar gins of motor area. B = Bregma.
Cortex and Response Suppression
119
sites, three electrode sites were selected where good response suppression was ob tained (75°/o or better), and three sites where response suppression was not signifi cant (below 14%), and frozen brain sections prepared. For the first group, tips of one pair of electrodes were in gray matter, a second pair was slightly within white matter, and the third pair were at the borderline position. For the second group, two pairs of electrodes were in gray matter and one pair was slightly in white matter. It appeared that electrode depth was not a factor in the results obtained. Electrodes were also implanted within the frontal pole. Their positions were determined from frozen sections at necropsy. Unilateral stimulation was done with a Grass S4 stimulator feeding through a Grass SIU5 stimulus isolation unit and a Grass constant current unit. Square wave pulses were 2 msec in duration and at a frequency of 10/sec. In the mapping study with the cat, bilateral stimulation, electrodes at corre sponding sites in the two cortices, was used [W ilcott and Sabol, 1976]. This pro duced stronger response suppression than unilateral stimulation and thereby af forded greater confidence in the results. But, with the rat, unilateral stimulation produced substantial response suppression and bilateral stimulation did not appear to be necessary. Despite the use of unilateral stimulation in the rat, stimulus cur rent requirements were consdierably less than those needed in the cat. This differ ence may be due to greater spread of current in rat cortex with a given current intensity.
Procedures Stimulation was presented continuously for 25 sec and the time between trials ranged from 2-5 min. Current ¡tensity was adjusted at each site separately and was near the maximum level that never produced movements, arousal responses, or a seizure. This point was ascertained at each cortical site by gradually increas ing stimulus current, usually in steps of 0.01 or 0.02 mA, until movements, etc., were observed. The current was then reduced to a level where it seemed certain that no movements, etc., would be produced during repeated testing. If move ments, etc., were not observed, a maximum current of 0.25 mA was used. Movements consisted of one or more of the following. Repeated movements of one or both front legs, head, or other parts of the body at the frequency of stimu lation, slow turning of the head to the contralateral side, or extension of one or both hind legs. Movements usually began at stimulation onset, but sometimes were not observed until stimulation was on for several seconds.
Downloaded by: King's College London 137.73.144.138 - 1/21/2019 12:32:32 AM
Bar-Press Response About 2 weeks after surgery the weight of each rat was reduced to 80% of its normal level with ad libitum feeding, and they were maintained at this level dur ing testing. A Gerbrand’s Skinner box and pellet feeder were used. They were housed in a Leheigh Valley Chamber fitted with a one-way window. The behavior of rats in the Skinner box could be closely observed at all times. White noise was introduced into the chamber to mask extraneous sounds. Rats were established on a 6/1 fixed ratio, and responses and reinforcements were recorded by two pens of a Grass polygraph (fig. 2). Because rats were on a fixed ratio, the rate of bar pressing was measured as the number of reinforcements.
120
W ilcott/S abol/Y urcheshen
UUIIII JU-iliJ
An arousal response began with an abrupt halt to all activity followed by rap id searching movements about the Skinner box. When stimulation was terminated, the rat usually again became motionless, but then continued searching movements for a short time before resuming bar pressing. In some cases, stimulation produc ing an arousal response was repeated and this response pattern was found to habi tuate. During a seizure, produced at only one electrode site, the rat fell to one side and displayed rapid jerking movements of the limbs. In this case, testing was not resumed for 2 days. During response suppression the behavior of rats appeared to be normal except that they did not press the bar or else responded at a reduced rate. When they were not bar-pressing, they stayed in the vicinity of the bar and groomed them selves or moved about that side of the Skinner box. An alerting stimulus, such as tapping on the chamber window, produced what appeared to be an appropriate orienting response. When a complete or nearly complete halt to responding was produced, rats would still readily eat from the feeder when it was operated exter nally. After stimulation was terminated rats resumed bar-pressing immediately or within 5-15 sec. Their behavior during response suppression appeared to be basi cally the same as observed in cats [W ilcott , 1974], except that they were more active. This was expected because rats are usually more active than cats.
Downloaded by: King's College London 137.73.144.138 - 1/21/2019 12:32:32 AM
Fig. 2. Effects of low-frequency stimulation in the cortex on the rate of bar-pressing and rate of reinforcement. In each section the upper trace is bar pressing, the middle is brain stimulation (25 sec.), and the lower is reinforcement. Section A displays a complete halt to bar-pressing and reinforcement in one rat, and section B shows a partial suppression (50°/o) in bar-pressing and reinforcement in another rat.
Cortex and Response Suppression
121
The magnitude of response suppression was determined by comparing the number of reinforcements during each 25-sec stimulation trial with the number of reinforcements during the 25-sec interval immediately preceding that trial. Mean values under these two conditions were then computed and a percent difference obtained. For electrode sites in the dorsolateral cortex means were computed from 12 trials, and for electrodes in the frontal pole they were computed from 6 trials. If the percent difference between means was below 30°/o for electrode sites at the dorsolateral cortex t tests were computed. This could not be done for data from frontal pole sites because of the small number, but considerable response suppres sion was always obtained at these sites and t tests were unnecessary.
Data from 67 sites were obtained from 21 rats. Sites are marked and numbered at the frontal pole and dorsolateral cortex in figure 1. Table I contains the mean and range of reinforcement rates for the 25-sec inter vals immediately before and during cortical stimulation, and the percent difference between these two conditions for each of the 67 sites. Exam ples of the effects of cortical stimulation on the rate of bar-pressing and reinforcement are shown in figure 2. Figure 1 also summarizes these results as a percent response suppres sion (percent difference in table I). At closed-circle sites, response sup pression was 75% or better (range 75-100%), and at open-circle sites response suppression ranged from 18 to 62% and was always significant (p
Frontal Cortex and Response Suppression in the Rat R. C. W ilcott , B. A. Sabol and R. P. Y urcheshen Department of Psychology, Case Western Reserve University, Cleveland, Ohio
Key Words. Frontal cortex • Electrical stimulation • Response suppression • Rat Abstract. Parts of the rat’s neocortex were mapped for sites where electrical stimulation, square waves at 10/sec, will suppress a bar-press response for food. Effective sites were found in the frontal pole and over most of the frontal dorso lateral cortex. Within these regions the strongest inhibitory influences were at sites in the frontal pole and adjacent frontal cortex, and at sites along the midline. These data generally agree with results of previous cortical ablation studies, but they suggest that inhibitory processes are more widespread in the dorsolateral cor tex than these studies indicated.
In monkeys, dogs and cats, increased skeletal and autonomic activity, and hyperreactivity, are usually observed following removal of parts of the frontal cortex including the prefrontal area [B ru tkow ski , 1965], These results seem to be due to release of a general inhibition normally maintained by parts of the frontal cortex. In the rat, removal of the frontal poles will produce increased activ ity, particularly as measured by a running wheel [R ichter and H aw kes , 1939; C ampbell and L ynch , 1969]. The same effects have been ob served following removal of anterior medial or dorsal rhinal sulcus cor tex [K olb , 1974]. Based on recent studies of thalamic projections, these latter areas may represent ‘prefrontal’ cortex in the rat more closely than the frontal poles [L eonard , 1968]. These studies suggest that parts of the frontal cortex in the rat, apparently corresponding to those parts
Downloaded by: King's College London 137.73.144.138 - 1/21/2019 12:32:32 AM
Introduction
Wilcott/S abol/Y urcheshen
117
identified in higher mammals, also have an inhibitory influence on be havior. Electrical stimulation has also been used to study inhibitory processes in the frontal cortex. Stimulation in the rat’s dorsolateral frontal cortex, with electrical square waves at a frequency of 0.5-30/sec, will markedly reduce the heart rate response to a noxious stimulus [G laser and G r if fin , 1962], All sites tested anterior to the bregma were effective, while stimulation at posterior sites had no effect on the heart rate response. In the cat, stimulation at a relatively low frequency (5-15/sec) in the pre frontal cortex will completely suppress a bar-press response for food, and partially suppress skin potential responses [W ilcott and B radley , 1970; W ilcott , 1974]. The cat’s dorsolateral cortex has also been part ly mapped for sites where stimulation will suppress the bar press re sponse [W ilcott and Sabol , 1976]. Results generally agree with inhibi tory areas of activity, etc., established by the ablation method [L ang worth and R ich ter , 1939], but they seem to show two things more ac curately. These are the extent of inhibitory areas in the dorsolateral frontal cortex, and the relative magnitude of inhibition produced at dif ferent areas. Because response suppression observed in these stimulation studies mainly occurs during the time stimulation is presented, response sup pression could have been due to extraneous factors such as production of competing responses, or disruption of neural tissue at the site of stim ulation. Available evidence, however, is opposed to these alternate inter pretations [W ilcott and B radley , 1970; W ilcott and Sabol, 1976]. The stimuation technique, particularly at a low frequency, seems to be a useful method for exploring inhibitory processes in the cortex. As documented below, it has been found that low-frequency stimula tion in the rat’s cortex will suppress a bar-press response. The aim of this study was to map parts of the rat’s dorsolateral cortex and frontal poles for sites where stimulation will suppress bar-pressing for food. Method
Electrodes and Stimulation Electrodes were obtained from Plastics Products Co. They consisted of pairs of fine stainless steel wires, insulated with Formvar varnish, wound together and cou-
Downloaded by: King's College London 137.73.144.138 - 1/21/2019 12:32:32 AM
Subjects Results are reported on 21 male Holtzman rats, weight ranging from about 400 to 420 g at the time of surgery.
W ilcott /S abol/Y urchesmen
ns
2
© ©
c
3 4
\
\
J
• o' 6 7 8 9 21 ' ©€ o • 0 10 11 22 12 18 6 0 O O © U 15 23 24 16 O A • o o\ 25 38 26 ^ 7 \ 17 0 O O O A ,-A' O \ 18 28 29 30 3 ? / 4 0 31 \ O \ 'o 0 O '41 \ 33 34 19 32 O O A A A A\ B- »2 0 .35 42 36 43 44 45 461 A A 'O A 47
32
48
49
A 51
A 50
à
A 53
52
A 54
A 55
A 59
A 60
A 56 A 58 A 61 \63 A A
J
A
65
A 57/
66
tJ
/
/
62/ A /
■ ¿ 67.
pled to a small molded socket. The wires were cut so that only their tips were ex posed. Rats were anesthetized with pentobarbital sodium. The dorsolateral cortex electrodes were implanted slightly within the cortex under direct visual control with the aid of a stereotaxic instrument. Four pairs of electrodes were implanted in each rat, two in the cortex on each side. Their positions were predetermined from sites marked on a drawing of the rat cortex using the bregma as a reference point. At necropsy brains were examined grossly for positions of electrodes and sites marked on a brain diagram (fig. 1). The lateral position of electrode sites was measured from the midline and stereotaxic measurement data corrected in a few cases. However, it was decided that the stereotaxic measurement was the best ap proximation of the anterior or posterior positions of an electrode site and this was used. Preliminary observations indicated that the depth of electrodes at the dorsolat eral cortex was not critical. Response suppression could be obtained just as well with electrode tips slightly protruding into white matter as with electrode tips in gray matter. The probable reason for this is that response suppression is produced by stimulation of cells and/or fibers making connections with subcortical struc tures [Wilcott and Sabol, 1965]. As a spot check on electrode depths at these
Downloaded by: King's College London 137.73.144.138 - 1/21/2019 12:32:32 AM
F ig.l. Electrode sites within the frontal pole (left) and in the dorsolateral cor tex (right). Closed circles are sites where response suppression was 75% or better, at open circles response suppression was 18-62% and always significant, and at triangle sites response suppression was 14% or less and never significant. The cross section at the frontal pole is about 6 mm anterior to the bregma. The dashed lines at the dorsolateral cortex mark the approximate anterior and posterior mar gins of motor area. B = Bregma.
Cortex and Response Suppression
119
sites, three electrode sites were selected where good response suppression was ob tained (75°/o or better), and three sites where response suppression was not signifi cant (below 14%), and frozen brain sections prepared. For the first group, tips of one pair of electrodes were in gray matter, a second pair was slightly within white matter, and the third pair were at the borderline position. For the second group, two pairs of electrodes were in gray matter and one pair was slightly in white matter. It appeared that electrode depth was not a factor in the results obtained. Electrodes were also implanted within the frontal pole. Their positions were determined from frozen sections at necropsy. Unilateral stimulation was done with a Grass S4 stimulator feeding through a Grass SIU5 stimulus isolation unit and a Grass constant current unit. Square wave pulses were 2 msec in duration and at a frequency of 10/sec. In the mapping study with the cat, bilateral stimulation, electrodes at corre sponding sites in the two cortices, was used [W ilcott and Sabol, 1976]. This pro duced stronger response suppression than unilateral stimulation and thereby af forded greater confidence in the results. But, with the rat, unilateral stimulation produced substantial response suppression and bilateral stimulation did not appear to be necessary. Despite the use of unilateral stimulation in the rat, stimulus cur rent requirements were consdierably less than those needed in the cat. This differ ence may be due to greater spread of current in rat cortex with a given current intensity.
Procedures Stimulation was presented continuously for 25 sec and the time between trials ranged from 2-5 min. Current ¡tensity was adjusted at each site separately and was near the maximum level that never produced movements, arousal responses, or a seizure. This point was ascertained at each cortical site by gradually increas ing stimulus current, usually in steps of 0.01 or 0.02 mA, until movements, etc., were observed. The current was then reduced to a level where it seemed certain that no movements, etc., would be produced during repeated testing. If move ments, etc., were not observed, a maximum current of 0.25 mA was used. Movements consisted of one or more of the following. Repeated movements of one or both front legs, head, or other parts of the body at the frequency of stimu lation, slow turning of the head to the contralateral side, or extension of one or both hind legs. Movements usually began at stimulation onset, but sometimes were not observed until stimulation was on for several seconds.
Downloaded by: King's College London 137.73.144.138 - 1/21/2019 12:32:32 AM
Bar-Press Response About 2 weeks after surgery the weight of each rat was reduced to 80% of its normal level with ad libitum feeding, and they were maintained at this level dur ing testing. A Gerbrand’s Skinner box and pellet feeder were used. They were housed in a Leheigh Valley Chamber fitted with a one-way window. The behavior of rats in the Skinner box could be closely observed at all times. White noise was introduced into the chamber to mask extraneous sounds. Rats were established on a 6/1 fixed ratio, and responses and reinforcements were recorded by two pens of a Grass polygraph (fig. 2). Because rats were on a fixed ratio, the rate of bar pressing was measured as the number of reinforcements.
120
W ilcott/S abol/Y urcheshen
UUIIII JU-iliJ
An arousal response began with an abrupt halt to all activity followed by rap id searching movements about the Skinner box. When stimulation was terminated, the rat usually again became motionless, but then continued searching movements for a short time before resuming bar pressing. In some cases, stimulation produc ing an arousal response was repeated and this response pattern was found to habi tuate. During a seizure, produced at only one electrode site, the rat fell to one side and displayed rapid jerking movements of the limbs. In this case, testing was not resumed for 2 days. During response suppression the behavior of rats appeared to be normal except that they did not press the bar or else responded at a reduced rate. When they were not bar-pressing, they stayed in the vicinity of the bar and groomed them selves or moved about that side of the Skinner box. An alerting stimulus, such as tapping on the chamber window, produced what appeared to be an appropriate orienting response. When a complete or nearly complete halt to responding was produced, rats would still readily eat from the feeder when it was operated exter nally. After stimulation was terminated rats resumed bar-pressing immediately or within 5-15 sec. Their behavior during response suppression appeared to be basi cally the same as observed in cats [W ilcott , 1974], except that they were more active. This was expected because rats are usually more active than cats.
Downloaded by: King's College London 137.73.144.138 - 1/21/2019 12:32:32 AM
Fig. 2. Effects of low-frequency stimulation in the cortex on the rate of bar-pressing and rate of reinforcement. In each section the upper trace is bar pressing, the middle is brain stimulation (25 sec.), and the lower is reinforcement. Section A displays a complete halt to bar-pressing and reinforcement in one rat, and section B shows a partial suppression (50°/o) in bar-pressing and reinforcement in another rat.
Cortex and Response Suppression
121
The magnitude of response suppression was determined by comparing the number of reinforcements during each 25-sec stimulation trial with the number of reinforcements during the 25-sec interval immediately preceding that trial. Mean values under these two conditions were then computed and a percent difference obtained. For electrode sites in the dorsolateral cortex means were computed from 12 trials, and for electrodes in the frontal pole they were computed from 6 trials. If the percent difference between means was below 30°/o for electrode sites at the dorsolateral cortex t tests were computed. This could not be done for data from frontal pole sites because of the small number, but considerable response suppres sion was always obtained at these sites and t tests were unnecessary.
Data from 67 sites were obtained from 21 rats. Sites are marked and numbered at the frontal pole and dorsolateral cortex in figure 1. Table I contains the mean and range of reinforcement rates for the 25-sec inter vals immediately before and during cortical stimulation, and the percent difference between these two conditions for each of the 67 sites. Exam ples of the effects of cortical stimulation on the rate of bar-pressing and reinforcement are shown in figure 2. Figure 1 also summarizes these results as a percent response suppres sion (percent difference in table I). At closed-circle sites, response sup pression was 75% or better (range 75-100%), and at open-circle sites response suppression ranged from 18 to 62% and was always significant (p
