The role of the inferior prefrontal convexity in performance of delayed nonmatching-to-sample.

Findings in an earlier study (Bachevalier, J. and Mishkin, M. Behav. Brain Res. 20, 249-261, 1986) indicated that ventromedial prefrontal cortex, which receives thalamic projections from the limbo-diencephalic system, is critical for visual recognition; whereas the dorsolateral prefrontal cortex, which receives no such thalamic projections, makes little or no contribution to this type of memory. In the present study, we examined the role in recognition of another prefrontal area outside the thalamic projection zone of the limbo-diencephalic system, namely, the inferior prefrontal convexity. In the first experiment, monkeys with lesions of this cortex (Group IC) were found to be impaired in relearning delayed nonmatching-to-sample (DNMS), but not on the subsequent DNMS performance test in which memory demands were greatly increased. In a second experiment, monkeys with combined lesions of the inferior and dorsolateral prefrontal cortex (Group LAT) were also found to be impaired in relearning DNMS, but in this case they were impaired, in addition, on the subsequent performance test. Neither group (IC or LAT) showed retardation in acquiring visual discrimination habits. Analysis of the DNMS behavior of both groups suggested that their deficits on this task were due not to a loss in recognition memory, but to various forms of perseverative interference. The results support the view that the inferior prefrontal convexity, like the dorsolateral prefrontal cortex, lies outside the limbo-diencephalic memory system not only anatomically but also functionally.

Neither perirhinal/entorhinal nor hippocampal lesions impair short-term auditory recognition memory in dogs.

Visual, tactile, and olfactory recognition memory in animals is mediated in part by the perirhinal/entorhinal (or rhinal) cortices and, possibly, the hippocampus. To examine the role of these structures in auditory memory, we performed rhinal, hippocampal, and combined lesions in groups of dogs trained in auditory delayed matching-to-sample with trial-unique sounds. The sample sound was presented through a central speaker and, after a delay, the sample sound and a different sound were played alternately through speakers placed on either side of the animal; the animal was rewarded for responding to the side emitting the sample sound. None of the lesion groups showed significant impairment in comparison either to their own preoperative performance or to the performance of intact control dogs. This was the case both for relearning the delayed matching rule at a delay of 1.5 s and for task performance at variable delays ranging from 10 to 90 s. From these findings we suggest that the tissue critical for auditory recognition memory is located outside both the perirhinal/entorhinal cortices and the hippocampus.

Cognitive functions of the temporal lobe in the dog: a review.

1 The current paper reviews the role of temporal lobe structures in learning and different kinds of memory, with an emphasis on behavioral tasks that re auditory stimuli. 2 The effects of lesions to structures in the temporal lobe were examined in separate groups of dogs, which were trained on an auditory spatial delayed response, or in a trial-unique auditory delayed match to sample recognition task. 3 Spatial memory was impaired after bilateral hippocampal lesions. On the other hand, neither an anterior temporal lesion or rhinal cortical injury nor combined lesion to the hippocampus and the anterior temporal lobe, affected postoperative retraining and performance of the spatial task. 4 Auditory recognition memory task was not impaired after a hippocampal and/or rhinal cortex lesion. However, postoperative retraining of the task was impaired after a lesion to auditory association areas. 5 These results confirm the role of the hippocampus in spatial memory in the dog. On the other hand, the organization of auditory recognition functions within the temporal lobe appears to be different from those described for visual recognition functions.

Effects of the anterior temporal lobe lesions, separate or combined with hippocampal damage, on spatial delayed responses guided by auditory stimulus.

Seventeen dogs were trained in a three-choice auditory spatial delayed response task, guided by auditory stimulus, at a 10 s delay to a criterion of 90% correct responses in 90 consecutive trials. Four dogs then received bilateral anterior temporal lobe lesions (AT), 6 dogs received hippocampal lesions (H), and 7 dogs served as controls (C). Group C reached postoperative criterion immediately while groups AT and H needed additional training. When subsequently tested at longer delays and with distractions, the group H animals performed more poorly than either the AT or C animals. Further, the group H dogs were again impaired when they retrained at a 10 s delay. In the second phase, the group H and AT animals received a second lesion forming a group (HAT) with bilateral lesions to both the hippocampus and the anterior temporal lobe. Unexpectedly, dogs from group HAT were unimpaired in either postoperative retraining or during performance task and distractions. The results emphasize the importance of the hippocampus in spatial delayed response with an acoustic cue. Effect of combined lesions after extensive training is discussed. Data might support the view, that the hippocampus plays time limited role in memory storage.

The method of training dogs in auditory recognition memory tasks with trial-unique stimuli.

Three adults dogs were trained in a auditory recognition delayed-matching-to-sample (DMS) task. The experimental setting consisted of one central speaker located in front of the dogs head, two side speakers with nearby response pedals and one rotary food delivery system. Three hundred twenty natural sounds were used as trial-unique stimuli. Sample stimuli were always given through the central speaker. After the delay of 1.5 s, both sample and testing stimuli were activated alternately through the two side speakers. Bar-press response toward the sample stimulus was rewarded by food. The DMS training was continued until attaining a criterion 90% correct responses in 90 consecutive trials. After a control pause, the dogs were retrained to the criterion, and then they were given performance tasks with delays extended to 10-, 30-, 60- and finally to 90-s, in blocks of 90 trials. Dogs required about 1,000 trials of auditory recognition memory training in order to reach the criterion. Their behavior was also stable after the control pause. The dogs performance declined gradually with extended delays reaching an average of 63.4% for the delay of 90 s. Results indicate that the DMS task with auditory stimuli alternating during the testing stage of trial, is a promising method for testing auditory recognition memory.

Influence of the varieties of differentiation training and prefrontal lesions on retention and reversal learning of avoidance responding in dogs.

The postoperative retention and reversal of go, no-go avoidance reflex differentiation with symmetrical and asymmetrical reinforcement were studied in dogs with prefrontal lesions and unoperated controls. The general pattern of the influence of such experimental variables as type of differentiation task an,d quality and arrangement of conditioned stimuli' was similar to that oberved preoperatively. However, prefrontal surgery changed the interactions between these experimental variables. The modifying role of the medial and the lateral parts of prefrontal cortex on the retention of differentiation and on reversal learning is discussed.

Avoidance responding in dogs trained in symmetrical or asymmetrical go, no-go differentiation.

The acquisition and the consolidation processes of avoidance responding in go, no-go differentiation with asymmetrical and symmetrical reinforcement procedures were studied in 49 male mongrel dogs. Differentiation training procedures with asymmetrical and symmetrical reinforcement were contrasted by the occurrence of painful shock on negative trials. The quality of conditioned stimuli and their relative saliency exerted strong effects on the rapidity of learning and the number of commision errors when the "asymmetrical" procedure was used, whereas these effects were strongly attenuated under "symmetrical" procedure of reinforcement. Dogs trained in "symmetrical" go, no-go differentiation showed characteristic responses executed with shorter latencies and larger percentages of errors on negative trials and numerous extra-and intertrial responding in early stages of differentiation learning. Retention tests showed a greater stability of acquired differential responding trained under "symmetrical" than under "asymmetrical" procedure of reinforcement. The data indicate that the strength of the secondary punishing effect of CS ;prolongation is directly related to the saliency of the stimulus. The relations between the primary and secondary punishing effects in the two types of differentiation task, and problem of signalling and arousing properties of stimuli used in training, were discussed.

Effect of experimental setting on learning and performance of Auditory Delayed Matching-to-Sample task in dogs.

Twelve dogs were trained in a new task for auditory recognition memory: auditory Delayed Matching-to-Sample (DMS). The animals were tested in two experimental settings using approach (Setting 1) or bar-press (Setting 2) responses. At the early stages of training, the learning took more trials in Setting 2, which was caused by different instrumental response and/or different relationship among manipulanda, stimuli, and reward in these two settings. The performance of the final task did not differ between settings and showed a gradual decline with extended delays. No differences were found in responding patterns or in dynamics of learning. Therefore, we conclude that the auditory DMS, trained in either setting, offers a valuable and reliable tool for studies of neural substrate of auditory recognition memory. The detailed analyses of the dogs' behaviour will allow to evaluate the subtle effects of experimental manipulations in future experiments, and for many reasons the data obtained from these two settings may be combined in further analyses.

The role of contextual and sporadic stimuli in escape conditioning in dogs.

Instrumental escape bar-pressing responses were rapidly acquired in four male dogs and showed great resistance despite a long pause in experimental sessions. More intertrial responses were performed after short- than after long-latency escape responses. In the next stage of the experiment shock trials were discontinued and on the defensive context novel accoustical stimuli were introduced. The auditory stimuli provoked instrumental bar-press responses that were somewhat dependent on the intensity of stimuli.

Thalamic and amygdaloid connections of the auditory association cortex of the superior temporal gyrus in rhesus monkey (Macaca mulatta).

Thalamic and amygdaloid connections of three association auditory areas (AA1, AA2, AA3) of the superior temporal gyrus (STG) were investigated. In order to define the projections of the particular areas, injections of fluorescent tracers were made in three monkeys. Distribution of labeling indicates that area AA1 differs from areas AA2 and AA3 in patterns of both thalamo-cortical and amygdalo-cortical connections. Area AA1 receives its predominant inputs from the ventral and dorsal nuclei of the medical geniculate body (MGB). The amygdaloid projection to the area AA1 originates from the basal nuclei, whereas input from the lateral nucleus was not found. The characteristic thalamic projections to areas AA2 and AA3 originate from the dorsal MGB nucleus and the polymodal nuclei of the posterior thalamus. The density of projections from the dorsal nucleus gradually decreases from area AA1 to area AA3 while projections from the Plm, Sg and Lim nuclei increase in the same direction. Areas AA2 and AA3 are the source of strong connections with the lateral nucleus of amygdala, which density increases progressively when injections shift from area AA2 to AA3. The basal and accessory basal nuclei are the source of a less significant amygdalofugal projections to both cortical areas. Thus, our experimental data indicate that influence of the polymodal thalamic nuclei increases substantially in the direction of the higher order association areas. The strong relation of the same cortical areas with the lateral amygdaloid nucleus might suggest that areas AA2 and AA3, in addition to auditory input are the site of transfer of complex sensory information to the amygdala.

Effects of hippocampal lesions on spatial delayed responses in dog.

Thirteen dogs were trained to perform spatial delayed responses to auditory cues in a three-choice Nencki testing apparatus with a delay of 0 s and then 10 s with a criterion of 90% correct responses in 90 consecutive trials. Then six dogs received bilateral surgical removal of the hippocampus via the cortex of the suprasylvian gyrus (without additional injury to the entorhinal and parahippocampal cortex). Three dogs received control surgical ablation of the suprasylvian gyrus, which was damaged in ablation of the hippocampus, and four dogs served as intact controls. After the surgery or rest period, the dogs were tested for their retention (10-s delay), and then they were given additional tests with extended delays (30, 60, and 120 s) and with distractions during the 60-s delay period. In comparison with both control groups, dogs with hippocampal ablations had moderately impaired postoperative retention, as evidenced by the elevated numbers of errors on criterion. In subsequent stages of testing with extended delays, the impairment was greater and was significantly correlated with the extent of injury to the hippocampus. These data, together with an analysis of the animals' responses to the three-choice situation, indicate that in dogs lesions of the hippocampus impair spatial memory.

Fos imaging reveals differential neuronal activation of areas of rat temporal cortex by novel and familiar sounds.

To provide information about the possible regions involved in auditory recognition memory, this study employed an imaging technique that has proved valuable in the study of visual recognition memory. The technique was used to image populations of neurons that are differentially activated by novel and familiar auditory stimuli, thereby paralleling previous studies of visual familiarity discrimination. Differences evoked by novel and familiar sounds in the activation of neurons were measured in different parts of the rat auditory pathway by immunohistochemistry for the protein product (Fos) of the immediate early gene c-fos. Significantly higher counts of stained neuronal nuclei (266 +/- 21/mm2) were evoked by novel than by familiar sounds (192 +/- 17/mm2) in the auditory association cortex (area Te3; AudA). No such significant differences were found for the inferior colliculus, primary auditory cortex, postrhinal cortex, perirhinal cortex (PRH), entorhinal cortex, amygdala or hippocampus. These findings are discussed in relation to the results of lesion studies and what is known of areas involved in familiarity discrimination for visual stimuli. Differential activation is produced by novel and familiar individual stimuli in sensory association cortex for both auditory and visual stimuli, whereas the PRH is differentially activated by visual but not auditory stimuli. It is suggested that this latter difference is related to the nature of the particular auditory and visual stimuli used.