Kazuo Hikosaka

Behavioral reactions reflecting differential reward expectations in monkeys.

Abstract. Learning theory emphasizes the importance of expectations in the control of instrumental action. This study investigated the variation of behavioral reactions toward different rewards as an expression of differential expectations of outcomes in primates. We employed several versions of two basic behavioral paradigms, the spatial delayed response task and the delayed reaction task. These tasks are commonly used in neurobiological studies of working memory, movement preparation, and event expectation involving the frontal cortex and basal ganglia. An initial visual instruction stimulus indicated to the animal which one of several food or liquid rewards would be delivered after each correct behavioral response, or whether or not a reward could be obtained. We measured the reaction times of the operantly conditioned arm movement necessary for obtaining the reward, and the durations of anticipatory licking prior to liquid reward delivery as a Pavlovian conditioned response. The results showed that both measures varied depending on the reward predicted by the initial instruction. Arm movements were performed with significantly shorter reaction times for foods or liquids that were more preferred by the animal than for less preferred ones. Still larger differences were observed between rewarded and unrewarded trials. An interesting effect was found in unrewarded trials, in which reaction times were significantly shorter when a highly preferred reward was delivered in the alternative rewarded trials of the same trial block as compared to a less preferred reward. Anticipatory licks preceding the reward were significantly longer when highly preferred rather than less preferred rewards, or no rewards, were predicted. These results demonstrate that behavioral reactions preceding rewards may vary depending on the predicted future reward and suggest that monkeys differentially expect particular outcomes in the presently investigated tasks.

Long- and short-range reward expectancy in the primate orbitofrontal cortex

The orbitofrontal cortex (OFC) is important in motivation and emotion. We previously reported reward expectancy‐related delay activities during a delayed reaction time task in primate OFC neurons. To further investigate the significance of the OFC in motivational operations, we examined pre‐instruction, baseline activities of OFC neurons in relation to reward expectancy during the delayed reaction time task. In this task, an instruction cue indicated whether reward would be present or absent in the trial. Each set of four consecutive trials constituted one block within which three different kinds of rewards and one trial with no reward were given in a fixed order that differed from the monkeys reward preference. We identified two types of OFC neurons with reward expectancy‐related pre‐instruction activities: Step‐type neurons showed stepwise changes (increase or decrease) in pre‐instruction activity toward the trial with a particular outcome, which usually was the most or least attractive within a block; Pref‐type neurons showed pre‐instruction activity changes according to the monkeys preference for each trials outcome. We propose that Step‐type and Pref‐type neurons are related to long‐range and short‐range reward expectancies of a particular outcome, respectively. The OFC is considered to play important roles in goal‐directed behaviour by adjusting the motivational level toward a certain (current or future) outcome of a particular motivational significance based on the two kinds of reward expectancy processes. Impairments in goal‐directed behaviour by OFC patients may be caused by a lack of long‐range expectancy or by a deficit in compromising between short‐range and long‐range expectancies.

Default Mode of Brain Activity Demonstrated by Positron Emission Tomography Imaging in Awake Monkeys: Higher Rest-Related than Working Memory-Related Activity in Medial Cortical Areas

Human neuroimaging studies have demonstrated the presence of a “default system” in the brain, which shows a “default mode of brain activity,” i.e., greater activity during the resting state than during an attention-demanding cognitive task. The default system mainly involves the medial prefrontal and medial parietal areas, including the anterior and posterior cingulate cortex. It has been proposed that this default activity is concerned with internal thought processes. Recently, it has been indicated that chimpanzees show high metabolic levels in these medial brain areas during rest. Correlated low-frequency spontaneous activity as measured by functional magnetic resonance imaging was observed between the medial parietal and medial prefrontal areas in the anesthetized monkey. However, there have been few attempts to demonstrate a default system that shows task-induced deactivation in nonhuman primates. We conducted a positron emission tomography study with [15O]H2O to demonstrate a default mode of brain activity in the awake monkey sitting on a primate chair. Macaque monkeys showed higher level of regional blood flow in these medial brain areas as well as lateral and orbital prefrontal areas during rest compared with that under a working memory task, suggesting the existence of internal thought processes in the monkey. However, during rest in the monkey, the highest level of blood flow relative to that in other brain regions was observed not in the default system but in the dorsal striatum, suggesting that regions with the highest cerebral blood flow during rest may differ depending on the resting condition and/or species.

Functional significance of delay-period activity of primate prefrontal neurons in relation to spatial working memory and reward/omission-of-reward expectancy

The lateral prefrontal cortex (LPFC) is important in cognitive control. During the delay period of a working memory (WM) task, primate LPFC neurons show sustained activity that is related to retaining task-relevant cognitive information in WM. However, it has not yet been determined whether LPFC delay neurons are concerned exclusively with the cognitive control of WM task performance. Recent studies have indicated that LPFC neurons also show reward and/or omission-of-reward expectancy-related delay activity, while the functional relationship between WM-related and reward/omission-of-reward expectancy-related delay activity remains unclear. To clarify the functional significance of LPFC delay-period activity for WM task performance, and particularly the functional relationship between these two types of activity, we examined individual delay neurons in the primate LPFC during spatial WM (delayed response) and non-WM (reward–no-reward delayed reaction) tasks. We found significant interactions between these two types of delay activity. The majority of the reward expectancy-related neurons and the minority of the omission-of-reward expectancy-related neurons were involved in spatial WM processes. Spatial WM-related neurons were more likely to be involved in reward expectancy than in omission-of-reward expectancy. In addition, LPFC delay neurons observed during the delayed response task were not concerned exclusively with the cognitive control of task performance; some were related to reward/omission-of-reward expectancy but not to WM, and many showed more memory-related activity for preferred rewards than for less-desirable rewards. Since employing a more preferred reward induced better task performance in the monkeys, as well as enhanced WM-related neuronal activity in the LPFC, the principal function of the LPFC appears to be the integration of cognitive and motivational operations in guiding the organism to obtain a reward more effectively.

Differential changes in glutamate concentration in the primate prefrontal cortex during spatial delayed alternation and sensory-guided tasks

Glutamate is a major neurotransmitter in the mammalian brain and glutamatergic neurotransmission in the frontal cortex is indicated to play important roles in cognitive operations. We previously examined changes in extracellular dopamine in the primate frontal cortex in cognitive tasks, and in this paper we extend this to glutamate. We employed, as cognitive tasks, a delayed alternation task where the animal must retain information in working memory, and a sensory-guided task in which there is no working memory requirement but there may be more sensory processing requirements. Using the in vivo microdialysis method, we examined changes in extracellular glutamate concentration in the dorsolateral, arcuate, orbitofrontal, and premotor areas of the primate frontal cortex. Compared to basal rest levels, we observed significant increases in glutamate concentration in dorsolateral and arcuate areas of the prefrontal cortex during the sensory-guided task, but did not find significant changes in any of the frontal areas examined during the delayed alternation task. When glutamate concentration was compared between the delayed alternation and sensory-guided tasks, difference was observed only in the dorsolateral prefrontal cortex, especially in the ventral lip area of the principal sulcus. The results indicate the importance of glutamate in processing sensory information but not in retaining information in working memory in the primate dorsolateral and arcuate prefrontal cortex. We also compared the concentration of glutamate and dopamine in the tasks. We found a double dissociation in the concentration of glutamate and dopamine in the dorsolateral area: there was an increase in glutamate but no change in dopamine during the sensory-guided task, whereas there was an increase in dopamine but no change in glutamate during the delayed alternation task. It is thus suggested that in the primate dorsolateral prefrontal cortex, increased glutamate tone without dopamine increase facilitates sensory-guided task performance, while increased dopamine tone without glutamate increase is beneficial for working memory task performance.

Tolerances of responses to visual patterns in neurons of the posterior inferotemporal cortex in the macaque against changing stimulus size and orientation, and deleting patterns

Neuronal activities were recorded in areas TEO and TE of the inferotemporal cortex in four hemispheres of two monkeys during the performance of a visual pattern discrimination task. Tolerances of responses to patterns against changing stimulus size and orientation, and deleting patterns halves were investigated and compared between TEO and TE neurons. Of 311 neurons tested, 80 (26%) responded to one or more patterns out of four standard patterns. Of these 80 neurons, 50 (63%) were recorded in area TEO and 30 (38%) in area TE. Neurons responsive to patterns were recorded in both areas TEO and TE, however degrees of tolerance of responses were different between TEO and TE neurons. Tolerances of TEO neurons were moderate and degrees of tolerance varied from neuron to neuron. Responses to particular patterns were dependent on stimulus size, stimulus orientation, and/or completeness of patterns. By contrast, tolerances of TE neurons were generally strong. Responses to particular patterns were not affected by changing stimulus size, changing stimulus orientation nor deleting patterns halves. These results suggest that area TEO rather than area TE is involved in detecting and processing particular visual shapes.

Higher dopamine release induced by less rather than more preferred reward during a working memory task in the primate prefrontal cortex.

An optimal level of dopamine (DA) in the mammalian prefrontal cortex (PFC) is critical for higher cognitive control of behavior. Too much or too little DA in the PFC induces impairment in working memory (WM) task performance. PFC DA is also concerned with motivation. When reward is anticipated and/or delivered, an increase in PFC DA release is observed. In the primate, more preferred reward induces enhanced WM-related neuronal activity in the dorsolateral PFC (DLPFC). We hypothesized that there would be more DA release in the primate DLPFC when more preferred, as compared with less preferred, reward is delivered during a WM task. Contrary to our hypothesis, we found higher DA release in the DLPFC when less rather than more preferred reward was used during a WM task, while unpredictable free reward delivery induced an increase in DLPFC DA release irrespective of the difference in the incentive value of the reward. Behaviorally, the monkey was more motivated with preferred than with less preferred reward, although it performed the task almost without error irrespective of the difference in the reward. Considering that mild stress induces an increase in DA release in the mammalian PFC, performing a WM task for less preferred reward could have been mildly stressful, and this mild stress may have induced more DLPFC DA release in the present study. The higher DA release in the DLPFC with less preferred reward may be beneficial for monkeys to cope with mildly stressful and unfavorable situations to achieve proficient WM task performance.

Representation of foveal visual fields in the ventral bank of the superior temporal sulcus in the posterior inferotemporal cortex of the macaque monkey

Using anesthetized and immobilized monkeys, this study investigated the representation of the visual field in the superior temporal sulcus in the posterior inferotemporal cortex. Of 1043 neurons in the posterior inferotemporal cortex including the sulcus and the gyrus, and surrounding areas that were tested, 540 (52%) responded to visual stimuli and their receptive fields were mapped. In the ventral bank of the superior temporal sulcus at the level corresponding to the posterior middle temporal sulcus, the foveal visual fields, which were dominant, were represented ventrally and the parafoveal visual fields dorsally. The upper and lower visual fields were represented intermingledly and no segregation between the representation of the upper and lower visual fields was seen. In the lateral convexity of the gyrus, the foveal visual fields were represented dorsally and the peripheral visual fields ventrally with the foveal visual fields being predominant. The upper visual fields were represented posteriorly, however locations and sizes of the representation of the upper and lower visual fields varied between hemispheres. The receptive field sizes of neurons in the sulcus were almost the same as those in the gyrus, and these receptive field sizes were intermediate between those of anterior inferotemporal neurons and V4 neurons. These findings suggest that the cortex in the sulcus in the posterior inferotemporal cortex is involved in the central vision, similarly to the cortex in the gyrus.

Release of neurotransmitters in the monkey frontal cortex is related to level of attention

Abstract Attention is reported to be maintained by monoamines, acetylcholine and amino acids systems. Changes in the releases of these neurotransmitters during the three stages comprising quiet wake (QW) and two arousal states (AW), which are activated from different sources, were investigated. Norepinephrine releases during AW were significantly higher than that during QW. Conversely, the levels of acetylcholine and serotonin that were released did not change significantly among these three stages. The interesting observation was the dissociation of the increase between glutamate and dopamine releases in the two AW states. These results indicate that attention level is related to the amount of norepinephrine release, and that attention quality is related to the interaction between dopamine and glutamate releases.

Responsiveness of neurons in the posterior inferotemporal cortex to visual patterns in the macaque monkey

Using anesthetized and immobilized monkeys, responses of neurons in the posterior inferotemporal cortex to visual patterns were examined. Response properties were compared between the sulcus and the gyrus, extending between the anterior tip of the posterior middle temporal sulcus and the inferior occipital sulcus. Of 682 neurons tested, 37% in the sulcus (134/365) and 36% in the gyrus (113/317) responded to one or more patterns. The preference of neurons for patterns varied from neuron to neuron; some neurons responded selectively to one particular pattern, whereas others responded to two or more patterns. To evaluate response properties of neurons, two indices were calculated (the pattern preference index and the pattern selectivity index). The distributions of these indices in the sulcus did not differ significantly from those of the gyrus. Furthermore, the relationship between the pattern preference index and the pattern selectivity index for each neuron was almost the same in these two portions; most neurons responding to a small number of patterns showed inhibitory or weak responses to the worst pattern. In both portions, most neurons had receptive fields with small eccentricities and receptive field sizes were almost the same. These results suggest that the cortex in the sulcus in the posterior inferotemporal cortex is involved in the detection of features of visual patterns, similarly to the cortex in the gyrus.