Cat-P300 Present After Association Cortex Ablution
Department af Physiology, Brain Research hstitute and Mental Retardation Research Center, UCDI. Medical Center, Los Angeies, CA 90024
~~S~~, J. B., L. W. DICKERSON, S. SONG AND J. S. ~UC~A~~. ~fft*P~~~rese~r after association cortex ablation. BRAIN RES BULL w4) 551-560, 1990.--The cat-P300 is a positive endogenous potential, larger to a stimulus when rare than when frequent, with a latency of 2@3-500 msec. The role of poiysensory association cortex, postulated to be imponatrt in human P300 generation, was assessed in the cat. EEG was recorded in 13 awake cats from a skull screw at the vertex. Stimuli included frequent IP= O.gQ 1 kHz and rare (P=O. IO>2 kwz tone pulses with probabilities ~unterba~~~ed across 26&riat b&ks. After I2 ~o~~~ve sessions, bilateral ablations were made of pericruciate cortex f4 cats), anterior lateral and medial sup~y~~~~ gyri (4 cats) and all 3 areas (5 cats). Postoperatively, all 13 cats showed a P300 across $2 recording sessions. Thus polysensory association cortex is not essential for generation of the cat-PJOO. Cat-P300
Auditory PJOO
Endogenous potential
Polysensory association cortex
THE human P300 potential is of interest to both cognitive scientists and clinicians. It appears to be an elec~ophysiologica~ correlate of complex processes such as directed attention, stimulus discrimination, sequential information processing, short-term memory and decision making (36). Its amplitude varies with stimulus ~obab~ity (relative frequency and ~quential structure) and meaning (stimulus complexity, stimulus value) and with i~o~~on transmission (di~mination difficulty, allocation of attention) (25). Abnormal P300 latency and morphology occur in cognitively impaired groups such as schizopbreaic and demented patients (4, 14, 35, 45). Despite interest among psychologists and neurologists, the neural origin of the human long-latency potentials is not known. Because of the cognitive correlates of the P300, many researchers have assumed that it is generated by neoeortex, Postulated neocortical generators of the hu~n P300 have included frontal cortex (8,9,27, 28,51), temporal cortex (37), and centroparietaI or temporoparietal association cortex (13, 36, 40, 42, 47). Multiple subcortical sites (50) and thalamic sites have also been proposed as P300 generators (4153). Other evidence suggests involvement of the hip~~pa~ formation and limb& system in P3OOgeneration (18,29,31,44). Surface topography of the P300 is unchanged, however, after unilateral temporal lobectomy which includes the hippocampus (52), and hippocampal P3OO”smay not be volume-conducted to the surface (42). Postulated generator systems can be more readily explored in an animal P3OOmodel than in the human brain. Such studies have provided evidence for neocortkd and/or limbie system invohe-
Cortical ablation
ment. The rn~gina~ (i.e., lateral) gyrus, s~prasyl~~~ gyrus and hip~~pus have all been postulated as generators of the P3OOin the cat, based on intracraniahy recorded potentials showing polarity reversals (30,48). P3OO-Ike potentials have also been recorded from the anterior thalamus and cingulate cortex in she rabbit (12). Frontal cortex (3) and/or Iocus coeruleus (33) have been postulated as necessary components for P300 generation in the monkey. A P300-like potential in the monkey remains after bilateral ablation of the hippocampus (32). Over the past several years we have been studying a P3OO-like potential in the awake cat (6,7, 19). As in the human, the P300 in the cat is a positive potential, with a latency of 200-500 msec, which is signi~~~t~y larger when elicited by rare or omitted stimuli and is task dependent. The task we have used recently is an ins~men~ly conditioned eye blink in response to a tone, which is followed by eyelid shock, if the cat does not blink at a criterion level. Studies with monkeys typically use lever pulling or pushing tasks (2,32), while studies with humans usually use a counting task. Although the nature of these tasks is very different, the p~adigm we use appears to keep the cat alefi and paying attention to the auditory stimuli. The e~~~rne~t to a stimulus when it occurs rarely is fairly smaI1, 4 p,V (32) or >I7 pV (34), or humans, 12 WV (38). It is not surprising that a cognitive potential would be smaller in cats than primates, but the cat P3OO has many similarities with the human. Far example, in aged cats (>lO years old), as in aged humans, the P3OOshows increased Iatency and decreased amplitude (19). We therefore considered the
‘Requests for mPrmts should k addressedt0 Dr. Jean B. Harrison, Department of Physiology, UCLA Medical Center, Los Angeles, CA 9q~~24_1751.
552
HARRISON.
TABLE 1
Group I: PCA
Group II: ALA + MSA
Group III: ALA f MSA f PCA
Cat
Preoperative
Postoperative
4 10 11 12
p300* p300* P300 -
p300* p300* P300 P300
1 3 13 2
P300*
p300* P300 -
P300” p300* p300*
5 6 7 8 9
p3OQ* P300 P300 p300*
p300* p300* p300* P300
SONG AND BUCHWALD
cat model to be appropriate for lesion and depth recording studies of possible P300 generator substrates. We have therefore investigated both limbic and neocortical systems. We found that the cat-P300 disappears after septal lesions, which strongly suggests participation of the limbic-septo-hippocampal system (20, 22, 26). On the other hand, we found that bilateral ablation of primary auditory cortex had no effect on the cat-P300 elicited by auditory stimuli (21). The possibility remains, however, that other neocorticai systems contribute significantly. Several forebrain pathways convey auditory information to neocortex in the cat. The classical primary auditory pathway, called the “lemniscal line” (15 17) system or the “cochleotonic” (1) system, projects from the central nucleus of the inferior colhculus to the ventral division of the medial geniculate body to the primary auditory cortex. It appears to be concerned with the spectral and spatial attributes of acoustic stimuli (24). We have shown, as noted above, that primary auditory cortex is not necessary for the cat-P300 either as the generator per se or as a relay in the pathway to the generator. A different, nonspecific system projects auditory information from the brainstem reticular formation and midbrain tegmentum through the intralaminar and medial thalamus to three neocortical polysensory association areas: the pericruciate area (PCA), the anterior lateral gyrus (ALA) and the medial suprasylvian gyrus @ISA) (24). These areas appear to be related to attention and to be concerned with the significance rather than the physical parameters of the stimuli (24). Thus, such a polysensory system would seem to be an appropriate candidate generator for the P300 which can be elicited by auditory, visual or somatosensory stimuli. The
SUMMARY OF CAT-P300 RESPONSES BEFORE AND AFTER ASSOCIATION CORTEX ABLATION
Lesion
DICKERSON,
Key: P300: Ram response > frequent response in grand averages. *: Ram response > frequent response @ frequent response.
TABLE 2 SUMMARY OF ASSOCIATION CORTEX ABLATIONS
PCA
Anterior Area (parts of areas 1,53,18,19)
Posterior Area (mostly area 7)
R
L
R
L
R
L
R
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0 0
0 0
0
x
0 0
0
0 0
0 0
3. 0
34
‘h Y,
0
‘55
X X
?h X
X X X X
v2
% %
X X 34 X
0
0 0
‘A
X
0
0
X X X
%
0
0
0
0
0
‘vi
0
%
0
0
0
0
0
‘h
Posterior Sigmoid Gyms (area 4)
CAT
L
R
L
R
L
Group I 4 10 11 12
X X X X
X X
X
x
0
X X X
X
X
x
0
0 0 0 0
0 0 X
x
X X X % X
Group III 5 6 I 8 9
0 0 0
X X X X %
X X ‘/2 0
vi
Midectosylvian Gyms (parts of areas 50.22)
Anterior Area (mostly area 5)
Anterior Sigmoid Gyms (area 6)*
Group II 1 3 13 2
AI
MSA
ALA
vi
X
vi
X
X
x
x
x
x
X
%
x x
x x
% X
X 3/d
vi
x
x
X
L: left. R: right. X: ablation of designated area. 0: no damage. 1%:$5 ablated. *See Reinoso-Su&rez, F. To~~phi~her
z/ 0 0 X 0 % 0 0
hiiatlas der Katze. Darmstadt: E. Merck AG; 1961.
0
553
CAT-P300 PRESENT AFTER CORTICAL ABLATION
CAT#4
---
420
Right
-
RARE FREQUENT
80
TIME (msec)
A20
PIG. 1. Continued presence of cat-P300 after bilateral PCA ablation (cat No. 4). Stippled areas in the diagrams represent ablated regions projected onto brain surface and coronal sections. PCA is completely ablated bilaterally; ALA and MSA are intact. As illustrated by the grand averages of 12 preand 12 postlesion recording sessions (upper right in this and succeeding figures), the response to the rare 2 kHz tone is greater than the frequent 2 kHz tone response. This difference is significant in the 293-364-msec latency range preoperatively and in the 293-761-msec latency range postoperatively. (Note that in this animal, there is an evoked potential sequence triggered by the offset of the 520 msec tone stimuli. These potentials have not been included in the data set of the present paper.) In this and other figures, 25 units on the ordinate are equal to 1.9 WV.
association cortex could integrate information from several sensory systems with information about the rarity and relevance of the stimuli. Moreover, as noted above, association cortex in both frontal and parietal lobes has been postulated as a generator of the human P300. The object of the present study was, therefore, to investigate the role of the cortical association areas in the generation of the cat P300. This role could be actual generation of the voltages recorded at the vertex, or it might be integration of information at a step in the pathway before actual generation. METHOD
Thirteen adult cats obtained from the UCLA Viva&m facility were used in these experiments. Under general anesthesia (sodium pentobarbital, 35 mg/kg) a screw electrode was implanted at the vertex, and two aluminum sleeves were cemented to the skull with dental acrylic. After recovery from anesthesia, evoked responses
were recorded during daily sessions from the awake cat, restrained in a canvas bag, with the head secured by rods placed through the implanted aluminum sleeves. All procedures were carried out in accordance with NIH standards, and all protocols were approved by the UCLA Animal Use Committee. Free-field acoustic stimuli were delivered at intervals of 1.5 set from a speaker centered at a constant location 33 cm from the ears in a sound-isolation chamber. In the typical protocol three acoustic stimuli were used. One, a 4 kHz 96 dB SPL (te. 20 FPa) tone pulse with 520 msec duration occurred randomly (P=O. 10) and served as a conditioned stimulus (CS+). The CS+ was followed immediately by eyelid shock, when the rectified and integrated orbicularis oculi EMG conditioned response did not reach a specified level during the CS+ period. The purpose of this instrumental conditioning procedure was to establish a discriminated conditioned blink response to the CS+, but not to the other acoustic stimuli, and thus focus the cat’s attention on the acoustic
554
HARRISON, DICKERSON, SONG AND BUCHWALD
CATS 1
PRE-LESION
FIG. 2. Continued presence of cat-P300 after bilateral ALA and MSA ablations (cat No. 1). The ALA is ablated on the left and partially ablated on the right; the anterior portion of MSA and the anterior % of the posterior MSA are ablated biiaterally. The prelesion response to the rare loud click is significantly greater than that to the frequent loud click across the latency range of 312-544 msec. Postlesion, the rare response is significantly larger across the 259-496msec latency range, although both postoperative responses are smaller
than the preopetativeones.
stimuli. The other two acoustic stimuli were also randomly ordered with one stimulus “frequent” (P=O.SO) and the other stimulus “rare” (P=O.lO) during each of two 260-t& blocks. The probabilities were reversed in two other alternated 260~trial blocks. For nine cats the rare/frequent stimuli were tone pulses of 1kHz and 2 kHz, while for four cats the stimuli were loud and soft clicks. (For these four cats and one other, the probability of the “rare” stimulus was 0.15, and for the CS+ 0.05. Three of these cats were con~tion~ with a classical paradigm, in which the shock was delivered on every trial.) The “rare” and “frequent” evoked responses produced by these counterbalanced stimuli were used for P300 analyses. (The “rare” response to the CS+ was not analyzed as the CS+ could not be counterbalanced.) EEG from the vertex electrode, referenced to a subcutaneous
electrode at the neck, was amplified 20,000~ with filters set to pass 0.1 Hz to 3 kHz. Separate averages, each with 500 time points over the IS-set ~ters~~us interval, were made of the EEG responses to the rare and frequent stimuli. After 12 daily preoperative recording sessions, polysensory cortical areas PCA, or ALA + MSA, or ALA + MSA + PCA were ablated bilaterally by aspiration under pentobarbital anesthesia (35 mg/kg IP) using aseptic surgical procedure. The animals were allowed to recover from anesthesia for one to two days, then the EEG was recorded over 12 postoperative sessions. The cats were then deeply anesthetized with pentobarbital and w with normal saline followed bylO% buffered fonnalin. The brain was removed from the skull and photographed. Frozen brain sections were cut at 80 PM. Every 5th section through the
555
CAT-P300 PRESENT AFTER CORTICAL ABLATION PRE-LESION
POST- LESION
---- RARE -
FREQUENT
FIG. 3. Continued presence of cat-P300 after bilateral MSA and small ALA ablations (cat No. 3). The ALA is intact on the left and only the lateral r/2ablated on the right; the anterior MSA and anterior % of the Posterior
MSA are bilaterally ablated. Prelesion the response to the rare stimulus is significantly greater than that to the frequent stimulus across the 201-425-msec latency range. Postlesion, the rare response is significantly larger across the latency range of 293-596 msec, although the absolute amplitude of both the rare response and the frequent response is less than preopcratively.
lesioned area was stained with thionine, and the extent of the lesions was reconstructed on relevant brain atlas sections. Grand averages of the “rare” and the “frequent” responses were computed and plotted for each cat before and after the cortical lesions. We compared responses to the identical stimulus when it occurred rarely and frequently and did not analyze the response to the target conditioned stimulus. Behavioral observations as well as EMG recordings on many cats confiied that eye blinks did not occur except with the conditioned stimulus. Thus the “frequent” and “ram” responses were not contaminated by motor components or artifacts. Data points were measured with respect to 0 DC input at the analog-to-digital converter. To reduce
the amount of data in this study, we analyzed the averaged responses in each 21%trial block to the “best stimulus” only, e.g., 1 kHz or 2 kHz, depending on which elicited a larger or more reliable P300 potential (see below) for any given cat. Thus for each cat, across 12 recording sessions, we had 48 blocks of averaged responses. We also omitted data in the O-40-msec and lOCKI-1.500~msec latency periods from our analysis and averaged each 3 successive data points in the 4&1000-msec period, so each response, averaged over one stimulus block, was reduced to 103 data points. Since we regularly saw a response to the frequent stimulus as well as the rare, we did not consider the entire “rare” response to
HARRISON. DICKERSON, SONG AND BUCHWALD
CAT*2
PRE-LESION
i
1
PbSTW0lJ.l I i i
i
--
RARE
FIG. 4. Enhancement of cat-P300 after biiateral ALA and MSA ablations (cat No. 2). The ALA and anteiar MSA are ablated bilaterally as well as a small portion of the posterior FCA. F’reqeratively, the rare and frequent responses are similar, i.e., there is no cat-P?oO. Fostoperatively, the fzve response is significantly larger than the frequent in the 41%738~msec latency range. (Note the tone c&et response in this cat, simiiar ta that of cat No. 4.1
be the cat-P3XA Instead, we considered only the enhancement of the response to a stimulus when it was rare as the endogenous potential, Paired r-tests (two-tailed) compared each of these 103 averaged data points for the “rare” and “frequent” response blocks in each session for each cat. In all cases, our conservative requirement for a significant difference between “rare” and ‘*frequent” responses was a series of five successive time points with p values
Department af Physiology, Brain Research hstitute and Mental Retardation Research Center, UCDI. Medical Center, Los Angeies, CA 90024
~~S~~, J. B., L. W. DICKERSON, S. SONG AND J. S. ~UC~A~~. ~fft*P~~~rese~r after association cortex ablation. BRAIN RES BULL w4) 551-560, 1990.--The cat-P300 is a positive endogenous potential, larger to a stimulus when rare than when frequent, with a latency of 2@3-500 msec. The role of poiysensory association cortex, postulated to be imponatrt in human P300 generation, was assessed in the cat. EEG was recorded in 13 awake cats from a skull screw at the vertex. Stimuli included frequent IP= O.gQ 1 kHz and rare (P=O. IO>2 kwz tone pulses with probabilities ~unterba~~~ed across 26&riat b&ks. After I2 ~o~~~ve sessions, bilateral ablations were made of pericruciate cortex f4 cats), anterior lateral and medial sup~y~~~~ gyri (4 cats) and all 3 areas (5 cats). Postoperatively, all 13 cats showed a P300 across $2 recording sessions. Thus polysensory association cortex is not essential for generation of the cat-PJOO. Cat-P300
Auditory PJOO
Endogenous potential
Polysensory association cortex
THE human P300 potential is of interest to both cognitive scientists and clinicians. It appears to be an elec~ophysiologica~ correlate of complex processes such as directed attention, stimulus discrimination, sequential information processing, short-term memory and decision making (36). Its amplitude varies with stimulus ~obab~ity (relative frequency and ~quential structure) and meaning (stimulus complexity, stimulus value) and with i~o~~on transmission (di~mination difficulty, allocation of attention) (25). Abnormal P300 latency and morphology occur in cognitively impaired groups such as schizopbreaic and demented patients (4, 14, 35, 45). Despite interest among psychologists and neurologists, the neural origin of the human long-latency potentials is not known. Because of the cognitive correlates of the P300, many researchers have assumed that it is generated by neoeortex, Postulated neocortical generators of the hu~n P300 have included frontal cortex (8,9,27, 28,51), temporal cortex (37), and centroparietaI or temporoparietal association cortex (13, 36, 40, 42, 47). Multiple subcortical sites (50) and thalamic sites have also been proposed as P300 generators (4153). Other evidence suggests involvement of the hip~~pa~ formation and limb& system in P3OOgeneration (18,29,31,44). Surface topography of the P300 is unchanged, however, after unilateral temporal lobectomy which includes the hippocampus (52), and hippocampal P3OO”smay not be volume-conducted to the surface (42). Postulated generator systems can be more readily explored in an animal P3OOmodel than in the human brain. Such studies have provided evidence for neocortkd and/or limbie system invohe-
Cortical ablation
ment. The rn~gina~ (i.e., lateral) gyrus, s~prasyl~~~ gyrus and hip~~pus have all been postulated as generators of the P3OOin the cat, based on intracraniahy recorded potentials showing polarity reversals (30,48). P3OO-Ike potentials have also been recorded from the anterior thalamus and cingulate cortex in she rabbit (12). Frontal cortex (3) and/or Iocus coeruleus (33) have been postulated as necessary components for P300 generation in the monkey. A P300-like potential in the monkey remains after bilateral ablation of the hippocampus (32). Over the past several years we have been studying a P3OO-like potential in the awake cat (6,7, 19). As in the human, the P300 in the cat is a positive potential, with a latency of 200-500 msec, which is signi~~~t~y larger when elicited by rare or omitted stimuli and is task dependent. The task we have used recently is an ins~men~ly conditioned eye blink in response to a tone, which is followed by eyelid shock, if the cat does not blink at a criterion level. Studies with monkeys typically use lever pulling or pushing tasks (2,32), while studies with humans usually use a counting task. Although the nature of these tasks is very different, the p~adigm we use appears to keep the cat alefi and paying attention to the auditory stimuli. The e~~~rne~t to a stimulus when it occurs rarely is fairly smaI1, 4 p,V (32) or >I7 pV (34), or humans, 12 WV (38). It is not surprising that a cognitive potential would be smaller in cats than primates, but the cat P3OO has many similarities with the human. Far example, in aged cats (>lO years old), as in aged humans, the P3OOshows increased Iatency and decreased amplitude (19). We therefore considered the
‘Requests for mPrmts should k addressedt0 Dr. Jean B. Harrison, Department of Physiology, UCLA Medical Center, Los Angeles, CA 9q~~24_1751.
552
HARRISON.
TABLE 1
Group I: PCA
Group II: ALA + MSA
Group III: ALA f MSA f PCA
Cat
Preoperative
Postoperative
4 10 11 12
p300* p300* P300 -
p300* p300* P300 P300
1 3 13 2
P300*
p300* P300 -
P300” p300* p300*
5 6 7 8 9
p3OQ* P300 P300 p300*
p300* p300* p300* P300
SONG AND BUCHWALD
cat model to be appropriate for lesion and depth recording studies of possible P300 generator substrates. We have therefore investigated both limbic and neocortical systems. We found that the cat-P300 disappears after septal lesions, which strongly suggests participation of the limbic-septo-hippocampal system (20, 22, 26). On the other hand, we found that bilateral ablation of primary auditory cortex had no effect on the cat-P300 elicited by auditory stimuli (21). The possibility remains, however, that other neocorticai systems contribute significantly. Several forebrain pathways convey auditory information to neocortex in the cat. The classical primary auditory pathway, called the “lemniscal line” (15 17) system or the “cochleotonic” (1) system, projects from the central nucleus of the inferior colhculus to the ventral division of the medial geniculate body to the primary auditory cortex. It appears to be concerned with the spectral and spatial attributes of acoustic stimuli (24). We have shown, as noted above, that primary auditory cortex is not necessary for the cat-P300 either as the generator per se or as a relay in the pathway to the generator. A different, nonspecific system projects auditory information from the brainstem reticular formation and midbrain tegmentum through the intralaminar and medial thalamus to three neocortical polysensory association areas: the pericruciate area (PCA), the anterior lateral gyrus (ALA) and the medial suprasylvian gyrus @ISA) (24). These areas appear to be related to attention and to be concerned with the significance rather than the physical parameters of the stimuli (24). Thus, such a polysensory system would seem to be an appropriate candidate generator for the P300 which can be elicited by auditory, visual or somatosensory stimuli. The
SUMMARY OF CAT-P300 RESPONSES BEFORE AND AFTER ASSOCIATION CORTEX ABLATION
Lesion
DICKERSON,
Key: P300: Ram response > frequent response in grand averages. *: Ram response > frequent response @ frequent response.
TABLE 2 SUMMARY OF ASSOCIATION CORTEX ABLATIONS
PCA
Anterior Area (parts of areas 1,53,18,19)
Posterior Area (mostly area 7)
R
L
R
L
R
L
R
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0 0
0 0
0 0
0
x
0 0
0
0 0
0 0
3. 0
34
‘h Y,
0
‘55
X X
?h X
X X X X
v2
% %
X X 34 X
0
0 0
‘A
X
0
0
X X X
%
0
0
0
0
0
‘vi
0
%
0
0
0
0
0
‘h
Posterior Sigmoid Gyms (area 4)
CAT
L
R
L
R
L
Group I 4 10 11 12
X X X X
X X
X
x
0
X X X
X
X
x
0
0 0 0 0
0 0 X
x
X X X % X
Group III 5 6 I 8 9
0 0 0
X X X X %
X X ‘/2 0
vi
Midectosylvian Gyms (parts of areas 50.22)
Anterior Area (mostly area 5)
Anterior Sigmoid Gyms (area 6)*
Group II 1 3 13 2
AI
MSA
ALA
vi
X
vi
X
X
x
x
x
x
X
%
x x
x x
% X
X 3/d
vi
x
x
X
L: left. R: right. X: ablation of designated area. 0: no damage. 1%:$5 ablated. *See Reinoso-Su&rez, F. To~~phi~her
z/ 0 0 X 0 % 0 0
hiiatlas der Katze. Darmstadt: E. Merck AG; 1961.
0
553
CAT-P300 PRESENT AFTER CORTICAL ABLATION
CAT#4
---
420
Right
-
RARE FREQUENT
80
TIME (msec)
A20
PIG. 1. Continued presence of cat-P300 after bilateral PCA ablation (cat No. 4). Stippled areas in the diagrams represent ablated regions projected onto brain surface and coronal sections. PCA is completely ablated bilaterally; ALA and MSA are intact. As illustrated by the grand averages of 12 preand 12 postlesion recording sessions (upper right in this and succeeding figures), the response to the rare 2 kHz tone is greater than the frequent 2 kHz tone response. This difference is significant in the 293-364-msec latency range preoperatively and in the 293-761-msec latency range postoperatively. (Note that in this animal, there is an evoked potential sequence triggered by the offset of the 520 msec tone stimuli. These potentials have not been included in the data set of the present paper.) In this and other figures, 25 units on the ordinate are equal to 1.9 WV.
association cortex could integrate information from several sensory systems with information about the rarity and relevance of the stimuli. Moreover, as noted above, association cortex in both frontal and parietal lobes has been postulated as a generator of the human P300. The object of the present study was, therefore, to investigate the role of the cortical association areas in the generation of the cat P300. This role could be actual generation of the voltages recorded at the vertex, or it might be integration of information at a step in the pathway before actual generation. METHOD
Thirteen adult cats obtained from the UCLA Viva&m facility were used in these experiments. Under general anesthesia (sodium pentobarbital, 35 mg/kg) a screw electrode was implanted at the vertex, and two aluminum sleeves were cemented to the skull with dental acrylic. After recovery from anesthesia, evoked responses
were recorded during daily sessions from the awake cat, restrained in a canvas bag, with the head secured by rods placed through the implanted aluminum sleeves. All procedures were carried out in accordance with NIH standards, and all protocols were approved by the UCLA Animal Use Committee. Free-field acoustic stimuli were delivered at intervals of 1.5 set from a speaker centered at a constant location 33 cm from the ears in a sound-isolation chamber. In the typical protocol three acoustic stimuli were used. One, a 4 kHz 96 dB SPL (te. 20 FPa) tone pulse with 520 msec duration occurred randomly (P=O. 10) and served as a conditioned stimulus (CS+). The CS+ was followed immediately by eyelid shock, when the rectified and integrated orbicularis oculi EMG conditioned response did not reach a specified level during the CS+ period. The purpose of this instrumental conditioning procedure was to establish a discriminated conditioned blink response to the CS+, but not to the other acoustic stimuli, and thus focus the cat’s attention on the acoustic
554
HARRISON, DICKERSON, SONG AND BUCHWALD
CATS 1
PRE-LESION
FIG. 2. Continued presence of cat-P300 after bilateral ALA and MSA ablations (cat No. 1). The ALA is ablated on the left and partially ablated on the right; the anterior portion of MSA and the anterior % of the posterior MSA are ablated biiaterally. The prelesion response to the rare loud click is significantly greater than that to the frequent loud click across the latency range of 312-544 msec. Postlesion, the rare response is significantly larger across the 259-496msec latency range, although both postoperative responses are smaller
than the preopetativeones.
stimuli. The other two acoustic stimuli were also randomly ordered with one stimulus “frequent” (P=O.SO) and the other stimulus “rare” (P=O.lO) during each of two 260-t& blocks. The probabilities were reversed in two other alternated 260~trial blocks. For nine cats the rare/frequent stimuli were tone pulses of 1kHz and 2 kHz, while for four cats the stimuli were loud and soft clicks. (For these four cats and one other, the probability of the “rare” stimulus was 0.15, and for the CS+ 0.05. Three of these cats were con~tion~ with a classical paradigm, in which the shock was delivered on every trial.) The “rare” and “frequent” evoked responses produced by these counterbalanced stimuli were used for P300 analyses. (The “rare” response to the CS+ was not analyzed as the CS+ could not be counterbalanced.) EEG from the vertex electrode, referenced to a subcutaneous
electrode at the neck, was amplified 20,000~ with filters set to pass 0.1 Hz to 3 kHz. Separate averages, each with 500 time points over the IS-set ~ters~~us interval, were made of the EEG responses to the rare and frequent stimuli. After 12 daily preoperative recording sessions, polysensory cortical areas PCA, or ALA + MSA, or ALA + MSA + PCA were ablated bilaterally by aspiration under pentobarbital anesthesia (35 mg/kg IP) using aseptic surgical procedure. The animals were allowed to recover from anesthesia for one to two days, then the EEG was recorded over 12 postoperative sessions. The cats were then deeply anesthetized with pentobarbital and w with normal saline followed bylO% buffered fonnalin. The brain was removed from the skull and photographed. Frozen brain sections were cut at 80 PM. Every 5th section through the
555
CAT-P300 PRESENT AFTER CORTICAL ABLATION PRE-LESION
POST- LESION
---- RARE -
FREQUENT
FIG. 3. Continued presence of cat-P300 after bilateral MSA and small ALA ablations (cat No. 3). The ALA is intact on the left and only the lateral r/2ablated on the right; the anterior MSA and anterior % of the Posterior
MSA are bilaterally ablated. Prelesion the response to the rare stimulus is significantly greater than that to the frequent stimulus across the 201-425-msec latency range. Postlesion, the rare response is significantly larger across the latency range of 293-596 msec, although the absolute amplitude of both the rare response and the frequent response is less than preopcratively.
lesioned area was stained with thionine, and the extent of the lesions was reconstructed on relevant brain atlas sections. Grand averages of the “rare” and the “frequent” responses were computed and plotted for each cat before and after the cortical lesions. We compared responses to the identical stimulus when it occurred rarely and frequently and did not analyze the response to the target conditioned stimulus. Behavioral observations as well as EMG recordings on many cats confiied that eye blinks did not occur except with the conditioned stimulus. Thus the “frequent” and “ram” responses were not contaminated by motor components or artifacts. Data points were measured with respect to 0 DC input at the analog-to-digital converter. To reduce
the amount of data in this study, we analyzed the averaged responses in each 21%trial block to the “best stimulus” only, e.g., 1 kHz or 2 kHz, depending on which elicited a larger or more reliable P300 potential (see below) for any given cat. Thus for each cat, across 12 recording sessions, we had 48 blocks of averaged responses. We also omitted data in the O-40-msec and lOCKI-1.500~msec latency periods from our analysis and averaged each 3 successive data points in the 4&1000-msec period, so each response, averaged over one stimulus block, was reduced to 103 data points. Since we regularly saw a response to the frequent stimulus as well as the rare, we did not consider the entire “rare” response to
HARRISON. DICKERSON, SONG AND BUCHWALD
CAT*2
PRE-LESION
i
1
PbSTW0lJ.l I i i
i
--
RARE
FIG. 4. Enhancement of cat-P300 after biiateral ALA and MSA ablations (cat No. 2). The ALA and anteiar MSA are ablated bilaterally as well as a small portion of the posterior FCA. F’reqeratively, the rare and frequent responses are similar, i.e., there is no cat-P?oO. Fostoperatively, the fzve response is significantly larger than the frequent in the 41%738~msec latency range. (Note the tone c&et response in this cat, simiiar ta that of cat No. 4.1
be the cat-P3XA Instead, we considered only the enhancement of the response to a stimulus when it was rare as the endogenous potential, Paired r-tests (two-tailed) compared each of these 103 averaged data points for the “rare” and “frequent” response blocks in each session for each cat. In all cases, our conservative requirement for a significant difference between “rare” and ‘*frequent” responses was a series of five successive time points with p values
