Abstract

To define the cortical areas that subserve spatial working memory in a nonhuman primate, we measured regional cerebral blood flow (rCBF) with [15O]H2O and positron emission tomography while monkeys performed a visually guided saccade (VGS) task and an oculomotor delayed-response (ODR) task. Both Statistical Parametric Mapping and regions of interest-based analyses revealed an increase of rCBF in the area surrounding the principal sulcus (PS), the superior convexity, the anterior bank of the arcuate sulcus (AS), the lateral orbitofrontal cortex (lOFC), the frontal pole (FP), the anterior cingulate cortex (ACC), the lateral bank of the intraparietal sulcus (lIPS) and the prestriate cortex. In the prefrontal cortex (PS, superior convexity, AS, lOFC and FP), rCBF values correlated positively with ODR task performance scores. From the hippocampus, rCBF values correlated negatively with ODR task performance. From the AS, superior convexity, lOFC, FP, ACC and lIPS, rCBF values of the PS correlated positively with rCBF values and negatively with hippocampus rCBF values. These results suggest that neural circuitry in the prefrontal cortex directly contributes the spatial working memory processes and that, in spatial working memory processes, the posterior parietal cortex and hippocampus have a different role to the prefrontal cortex.

Introduction

Working memory refers to the transient storage and processing of information (Baddeley, 1992). It is well known that the primate dorsolateral prefrontal cortex, especially the area surrounding the principal sulcus, participates in spatial working memory processes (Funahashi and Kubota, 1994; Goldman-Rakic, 1996; Petrides, 1996; Fuster, 1997). The firing rates of neurons in the dorsolateral prefrontal cortex increase tonically while monkeys actively retain memory of particular spatial locations (Kubota et al., 1974; Niki, 1974; Niki and Watanabe, 1976a; Kojima and Goldman-Rakic, 1982; Kubota and Funahashi, 1982; Funahashi et al., 1989, 1993b, 1997; Rainer et al., 1998; Sawaguchi and Yamane, 1999; Constantinidis et al., 2001). These results indicate that the dorsolateral prefrontal cortex takes part in spatial working memory, especially in the temporary active storage of spatial information (Funahashi and Kubota, 1994; Goldman-Rakic, 1996; Petrides, 1996; Fuster, 1997).

Recently, it was suggested that a network extending across several cortical areas, including the dorsolateral prefrontal cortex and posterior parietal cortex, takes part in spatial working memory. The dorsolateral prefrontal cortex interconnects with the posterior parietal cortex (Schwartz and Goldman-Rakic, 1984; Andersen et al., 1985; Barbas and Mesulam, 1985; Cavada and Goldman-Rakic, 1989; Andersen et al., 1990). During performance of spatial working memory tasks, posterior parietal cortex neurons were discovered mirroring dorsolateral prefrontal cortex (Chafee and Goldman-Rakic, 1998). A metabolic mapping study demonstrated an increase in glucose utilization in both the dorsolateral prefrontal and posterior parietal cortex during monkey performance of spatial working memory tasks (Friedman and Goldman-Rakic, 1994). These results indicate that when spatial working memory is required there is coactivation in both the dorsolateral prefrontal and posterior parietal cortex. In addition, the dorsolateral prefrontal and posterior parietal cortex commonly send efferents to the supplementary motor and premotor cortices, frontal eye field, orbitofrontal cortex, anterior and posterior cingulate cortices, parahippocampal gyrus and superior temporal cortex (Selemon and Goldman-Rakic, 1988). These cortical areas could constitute part of an anatomical network that mediates the spatial working memory (Goldman-Rakic, 1988). Human neuroimaging studies showed activation during the spatial working memory tasks in numerous cortical areas, including the dorsolateral prefrontal cortex, frontal eye field (Anderson et al., 1994; O‚Sullivan et al., 1995; Sweeney et al., 1996; Owen et al., 1996a, 1998), supplementary motor area (Anderson et al., 1994; O‚Sullivan et al., 1995; Goldberg et al., 1996; Smith et al., 1996; Sweeney et al., 1996), premotor cortex (Baker et al., 1996a; Smith et al., 1996), anterior cingulate cortex (Anderson et al., 1994; O‚Sullivan et al., 1995; Baker et al., 1996a; Belger et al., 1998; Owen et al., 1998), posterior parietal cortex (Anderson et al., 1994; O‚Sullivan et al., 1995; Baker et al., 1996a; Goldberg et al., 1996; Smith et al., 1996; Sweeney et al., 1996; Belger et al., 1998; Owen et al., 1996a, 1998) and occipital cortex (Anderson et al., 1994; O‚Sullivan et al., 1995; Goldberg et al., 1996; Owen et al., 1996a). These results provide supporting evidence that the neuronal links between these cortical areas constitute part of a working memory system.

To delineate the network that participates in spatial working memory processes, it is necessary to identify the cortical areas that subserve the spatial working memory and clarify the reciprocal relations between those areas. Recently, positron emission tomography (PET) techniques have enabled the measurement of monkey’s regional cerebral blood flow (rCBF) during performance of behavioral tasks (Onoe et al., 2001; Tanaka et al., 2001). Such PET studies using monkeys have several advantages. Repetition of measurements with the same animal avoids the need for complex normalization across subjects; direct comparison is possible with abundantly available anatomical and neurophysiological data; and, if new brain areas are discovered to be involved during the testing, we can analyze the neuronal mechanisms of those areas in detail using neuroanatomy, neurophysiology, neuropharmacology and other methods. In this study, to define the cortical areas that subserve spatial working memory in primates and correlations between those areas, we measured rCBF with H215O PET while monkeys performed oculomotor task required spatial working memory. Some of the data have been previously reported in abstract form (Inoue et al., 1999, 2001).

Materials and Methods

Subjects

Experiments were performed using two male rhesus monkeys (Macaca mulatta, monkey G, 4.3 kg; monkey N, 6.2 kg). They were housed individually in home cages. In their home cages, they were deprived of water, but could obtain their daily requirement of water as a reward during training and scan sessions. Experiments were conducted according to the Guide for the Care and Use of Laboratory Animals by the National Institute of Health, Bethesda, MD and the Guide for the Care and Use of Laboratory Primates by the Primate Research Institute, Kyoto University.

To fix the head during PET scans, a head-restraining device was attached to the monkey’s skull. Surgery was performed under aseptic conditions. The monkey was first given ketamine (50 mg) i.m. and then an intravenous injection of pentobarbital sodium (20 mg/kg). After partially exposing the skull, stainless steel or acrylic screws were used to attach firmly head-restraining device to the skull. These screws and the head-restraining device were then fixed with dental acrylic resin. The monkeys were given systemic antibiotics for a week after surgery and, for at least one week after surgery, free access to water and chow was allowed.

Apparatus

Figure 1A diagrammatically represents the experimental apparatus. The monkey sat in a primate chair in a dark room and during scan sessions, the head was immobilized by the head-restraining device. A PET scanner, SHR-7700 (Hamamatsu Photonics) (Watanabe et al., 1997), was used for all PET studies. This scanner has a 3.2 mm full-width half-maximum value and is able to image 31 × 3.6 mm-pitch slice at a resolution of 1.2 mm per pixel. Tilted 15° from the horizontal plane, the positioning of the PET scanner with the gantry arranged around the upper head area allowed the monkey to look downward at a 15-inch CRT monitor (PC-KM153R, NEC), placed 57 cm away from the eye plane. A computer (PC-9801FA, NEC) presented the fixation point and target stimuli on a CRT monitor. The monkey’s horizontal and vertical eye coordinates were sampled at 60 Hz by an infrared-camera monitoring system (X-Y Tracker C3162, Hamamatsu Photonics). The eye coordinates were fed to a computer (PC-9801BX2, NEC) via an A/D converter to evaluate maintenance of a fixation and whether saccades were properly executed. In addition, along with task events, eye positions during task performance were recorded onto magnetic tape by a data recorder (PC-208A, Sony-Magnescale).

Behavioral Tasks

To investigate the activation sites during the spatial working memory task, we measured the rCBF while monkeys performed two kinds of oculomotor tasks; an oculomotor delayed-response (ODR) task and a visually guided saccade (VGS) task.

Oculomotor Delayed-response Task (Fig. 1B)

After a short ITI (inter trial interval), a small red spot (0.1° diameter) was presented as a fixation point at the center of the CRT monitor. The monkey was required to look at the fixation point and maintain fixation. After the monkey maintained a fixation for 1 s, a red circle (0.5° diameter) was presented as a target cue for 100 ms (cue period). The target cue was randomly presented at one of eight predetermined positions (Fig. 1D). Their eccentricity was 5° from the fixation point. The monkey was required to maintain gaze at the fixation point during the cue period and the subsequent 2 s delay period. At the end of the delay period, the fixation point was extinguished and the monkey had to transfer gaze to the position where the target cue had been presented. If a saccade occurred within 500 ms, indicating that the monkey had succeeded in this task, a drop of water was presented as a reward. Proper saccade was defined as an eye movement that terminated within a 3.8° × 3.8° square zone around the correct cue position.

Visually Guided Saccade Task (Fig. 1C)

After a short ITI, a fixation point appeared at the center of the CRT monitor. The monkey was required to gaze at the fixation point and maintain fixation. When the monkey had maintained a fixation for 3.1 s, the fixation point was extinguished and a target cue was presented at one of the eight predetermined positions (Fig. 1D). After the target cue was presented, saccadic activity indicating transfer of fixation to the target cue had to occur within 500 ms. A correct saccade was defined as an eye movement that ended within a 3.8° × 3.8° square zone around the correct cue position.

PET Imaging

Before PET scanning, a venous cannula was placed in a sural vein. All drugs were administered through this cannula by automatic injector. When radioactivity had reached the brain, ∼10 s after the start of injection of a bolus of 15O-labeled H2O (2.5 GBq in 3.5 ml saline), a 100 s PET scan was obtained to record the distribution of H215O in the brain (Fig. 1E). To ensure adequate time for H215O decay, doses of tracer were administered every 15 min. During this interval, the monkey performed the task scheduled for the next PET scan. On experimental days, PET scans were performed 15–21 times. At the end of experimental day, a 30 min transmission scan was obtained to evaluate relative attenuation factors for image reconstruction.

Image Analysis

To yield transaxial images, all emission scans were reconstructed with a Hanning 4.5 mm filter. After normalizing the set of images taken each experimental day, the rCBF images (monkey G, 25 scans for both tasks; monkey N, 21 scans for both tasks) were statistically analyzed. For this analysis we used data from the 50 s period commencing 10 s after the 100 s emission scans had started (Fig. 1E). From the start of injection to the end of PET scan (110 s), the monkeys performed 23–28 trials in the ODR and VGS tasks. If the monkey failed to perform at least 17 trials correctly, data for that particular scan was excluded from final analysis.

Two methods were used to analyze the PET data. One was the data process using the Statistical Parametric Mapping (SPM96) software provided by the Wellcome Department of Cognitive Neurology, London, UK. The image data were smoothed using 2 mm Gaussian filter. Then, for task–task (ODR–VGS) comparison for each subject, z-scores were calculated on a pixel-by-pixel basis. Each pixel exceeding a z-score of 1.96 or greater, was superimposed on a relevant magnetic resonance (MR) image of the same subject. MR images were obtained by MRT-50A/II (Toshiba) at an intensity of 0.5 T (resolution: 0.586 mm × 0.586 mm × 3.0 mm per pixel). MR images were adjusted with regional cerebral glucose metabolism (rCMRGlu) images. To obtain rCMRGlu images, a bolus of [18F]2-fluoro-2-deoxy-d-glucose (40 MBq/kg in 5 ml saline) was delivered through a venous cannula placed in a sural vein and a 60 min emission scan was performed. These adjusted MR images were used with the rCMRGlu images to adjust the rCBF images. Coronal images and sagittal images of the results were made using a 3D BrainStation (Loats Associates Inc.). Based on the transaxial, coronal and sagittal images of the subtraction images (ODR–VGS) of each monkey, we listed the regions in both monkeys that were, according statistical evaluation, activated.

The second method was region of interest (ROI) based analysis. The MRI images were used to draw an individualized ROI template for each monkey. As shown in Figure 2, ROIs were drawn for a variety of regions involving the dorsolateral prefrontal cortex and its cortical connection (Barbas and Mesulam, 1985; Barbas and Pandya, 1989). We selected 15 regions: the area surrounding the principal sulcus (PS), the superior and inferior convexity, the anterior bank of the arcuate sulcus (AS), the lateral and medial orbitofrontal cortex (lOFC and mOFC), the frontal pole (FP) and the anterior cingulate cortex (ACC) in the frontal cortex; the lateral and medial bank of the intraparietal sulcus (lIPS and mIPS), area 7 and the posterior cingulate cortex (PCC) in the parietal cortex; the area surrounding the superior temporal sulcus (STS) in the temporal cortex; the prestriate cortex in the occipital cortex; and the hippocampus. These individualized ROI templates were then applied to the PET scan data of each subject and the rCBF values for each ROI were determined. The rCBF value was normalized to a relative value, derived from adjusting the global mean activity to 100. The normalized value under the VGS task was subtracted from the normalized value of the ODR task and the resulting difference was examined by paired t-test. Significance was set at P < 0.05.

Histological Analysis

After the PET scans were completed, the monkey was deeply anesthetized with pentobarbital sodium (35 mg/kg). The brain was first perfused with saline and then with 10% formalin solution. After finishing the perfusion, the brain was removed from the skull and cut transaxially into 100 µm serial sections. These sections were stained with cresyl violet. We matched the brain areas on the MRI images using the locations of the sulcus and cytoarchitecture of these sections.

Results

SPM Analysis

Using SPM, to compare the rCBF data during the ODR task with data during the VGS task, we made subtraction images (ODR–VGS) for each monkey and identified the regions (Table 1) that activated significantly during the ODR task in both monkeys. We detected activation in the PS (Fig. 3D) and other areas in the frontal cortex of both monkeys. Activity was located in the superior convexity (Fig. 3C), AS (Fig. 3E), FP (Fig. 3A), lOFC (Fig. 3F) and ACC (Fig. 3B). In the parietal cortex of both monkeys, a lateral and medial bank of the intraparietal sulcus (IPS) was activated. Significant activation was also detected in the PCC. In the temporal cortex, we found activation in the fundus or superior bank of superior temporal sulcus. In the occipital cortex, significant activation was apparent in the prestriate cortex.

ROI-based Analysis

In the ROI-based analysis, we compared the normalized rCBF values during the ODR and VGS tasks. Figure 4 shows the normalized rCBF values for all individual trials in the PS in both hemispheres of both monkeys. In the left hemisphere of monkey G, 16 scans (64%) during the ODR task showed greater activation than during the VGS task. The mean of the difference of the normalized rCBF values during the ODR task and the VGS task was 1.51. In remaining three hemispheres, 16 (64%, monkey G, right), 14 (67%, monkey N, left) and 13 scans (62%, monkey G, right) during the ODR task showed greater activation during the VGS task and the mean of the difference of the normalized rCBF values during the ODR task and the VGS task was 1.22 (monkey G, right), 2.97 (monkey N, left) and 2.61 (monkey N, right). The rCBF values during the ODR task and the VGS task were significantly different (df = 91, t = 3.44, P < 0.0001). In the PS, AS, superior convexity, lOFC, FP, ACC, lIPS, prestriate cortex and hippocampus, the rCBF values during the ODR task were significantly (P < 0.05) greater than those during the VGS task (Table 2). In other areas, such as the inferior convexity, mOFC, mIPS, area 7, PCC and STS, the differences were not significant (P > 0.05).

To analyze the relation between rCBF values recorded during the ODR task and the correct performance level of the task, we selected nine regions in which the rCBF significantly increased during the ODR task: the PS, AS, superior convexity, lOFC, FP, ACC, lIPS, prestriate cortex and hippocampus. Because event-related functional MRI studies have revealed that increase in rCBF are delayed by 3–6 s (McCarthy et al., 1996), we evaluated data gathered for 50 s (beginning 5 s after the start of PET scanning). In this analysis, we added scan data that were excluded from the SPM and ROI-based analysis because of low correct performance. No differences in values for the left and right hemisphere were found, so we combined data for the left and right hemisphere. After calculating the correlation coefficient between the rCBF values for each ROI template area and correct performance percentages for the ODR task, we found that in the PS, AS and FP, the rCBF values correlated positively (P < 0.05) with the correct performance percentages for the ODR task in both monkeys (Fig. 5A,B,E). In the superior convexity and lOFC, the rCBF values of monkey G correlated positively (P < 0.05) and the rCBF values of monkey N correlated positively but not significantly (P > 0.05) with correct performance percentage (Fig. 5C,D). In the ACC, the rCBF values of the monkey N negatively correlated (P < 0.05) with correct performance percentage and we could not find significant correlation (P > 0.05) between the rCBF values of the monkey G and the correct performance percentage (Fig. 5F). By contrast, the rCBF values of both monkeys in the hippocampus negatively correlated (P < 0.05) with correct performance percentages for the ODR task (Fig. 5H). In the lIPS and prestriate cortex, we could find no significant correlation (P > 0.05) between the rCBF value and correct performance percentages for the ODR task (Fig. 5G,I). In the VGS task, the monkeys made no trial errors, it was not possible to test the correlation of the rCBF values and correct performance percentage.

Research has shown that brain regions paired by significant correlations in normalized regional metabolic rates are functionally associated and that the strength of association is related to how well the correlations match (Horwitz, 1990, 1991). To investigate the relationship between regions in this experiment, we calculated the correlation of rCBF values for each ROI template. Figure 6 presents scatter diagrams showing the matching of rCBF values in the PS and other regions. In the PS, rCBF values correlated positively values in the AS, superior convexity, lOFC, FP, ACC and lIPS (Fig. 6A–F), and correlated negatively with values in the hippocampus (Fig. 6G). There was not correlation between rCBF values in the prestriate cortex (Fig. 6H). Figure 7 maps the correlation coefficients for paired cortical regions. The rCBF values of the prestriate cortex correlated negatively with those of the lIPS. The rCBF values of the lIPS correlated positively with those of the PS, AS, superior convexity and lOFC. The rCBF values of the prefrontal cortex (PS, AS, superior convexity, FP and lOFC) correlated well with each other. The rCBF values of the hippocampus correlated negatively with those of the frontal and parietal cortices.

Discussion

Functional neuroimaging with monkeys, unlike single-cell electrophysiology, enable the simultaneous investigation of multiple brain regions. Although lacking precision down to the neuronal level, functional neuroimaging provides an effective method for studying neural networks that are extensively distributed in the brain, including the network that appears to subserve the spatial working memory. Furthermore, neuroimaging data from monkeys is open to direct comparison with previous neurophysiological findings from monkeys.

In the present study, we found different activation patterns between the prefrontal and the posterior parietal cortices during the performance of spatial working memory tasks and were able to detect co-activation in areas in the prefrontal cortex. In the prefrontal cortex, the PS, AS, superior convexity, lOFC and FP showed greater activity during the ODR task and results from these areas were positively correlated. The rCBF values of the PS, AS, superior convexity, lOFC and FP correlated positively with the percentage of correct performance. On the other hand, activity in parietal areas did not correlate with how well the monkeys performed the ODR task. Although the hippocampus showed greater activity during the ODR task, the rCBF values correlated negatively with how well the ODR task was performed. These results suggest that neural circuitry in the prefrontal cortex (the PS, AS, superior convexity, lOFC and FP) directly contributes to spatial working memory processes and that the posterior parietal cortex and hippocampus, which are the sites of delay-period activity during delayed-response tasks, play a different role in spatial working memory processes.

SPM analysis and ROI-based analysis

Our study involved two broad approaches to data analysis, SPM analysis and ROI-based analysis. In the ROI-based analysis, the activation of a given structure is determined for each individual. When the activation is restricted to a region, this analysis is potentially powerful. However, if a region includes both activated and unactivated sites, significant activation is not found by ROI-based analysis. On the other hand, SPM analysis involves a pixel-by-pixel search. In this study, we used a smaller filter (2 mm) than is normally used in human PET studies. Thus, this analysis should be more than usually sensitive to small activation. In the both SPM and ROI-based analysis, significant activation occurred in the PS, AS superior convexity, lOFC, FP, ACC, lIPS and prestriate cortex. These areas likely include many activated sites and a few unactivated sites. However, there is some asymmetry of activated sites withinIn these areas in each monkey. In neurophysiological investigations of the association cortices, it is not unusual to find patchy hot spots of task-related neurons. Typical task-related neurons are clustered, and those clusters, rather than being continuously or systematically arranged as in the primary sensory areas, are sparsely distributed within each cortical area. This type of clustering has recently been confirmed in the prefrontal cortex using metabolic mapping of glucose utilization (Friedman and Goldman-Rakic, 1994; Sybirska et al., 2001) and the temporal association cortex using the optical imaging techniques (Wang et al., 1998). Similarly, activated patches in the association cortex are likely to be randomly and sparsely distributed. If this is the case, our finding that the activated sites by the SPM analysis are not symmetrical within each animal is not surprising. If the hot spot of activation site is located between slices, significant activation could not be found by the SPM analysis. This is probably reason of some asymmetry of activated sites within each monkey by the SPM analysis. Meanwhile, in the mIPS, PCC and STS, whereas SPM analysis revealed significant activation and ROI-based analysis did not, it is reasonable to consider that in these areas most sites were not activated but small spots were significantly activated. Since, if a region includes both activated and unactivated sites, significant activation is not found by ROI-based analysis but by SPM analysis significant activation could be found. On the other hand, in the hippocampus, most pixels in certain regions reveal slight activated and this showed up in ROI-based analysis and not in SPM analysis.

In both SPM and ROI-based analyses, because it is difficult to directly compare the level of activation on single task, we focused on the differences in rCBF values during the ODR task and the VGS task. These subtraction techniques are limited in that they only allow demonstration of the relative weighting given to working memory and sensory-guided processes in each area. Our present findings show that the lIPS and prefrontal area were more activated during the ODR task than during the VGS task. This suggests that these parietal and prefrontal areas are involved in working memory processes. Activity in the prefrontal areas, however, correlated with ODR task performance and activity in the lIPS did not correlated with ODR task performance. This, in turn, suggests that there are some differences between the prefrontal and parietal areas in the functional role of the working memory processes. It also suggests that the subtraction technique is inadequate for thorough investigation of brain regions involved in working memory processes and that other forms of analysis, for example, correlations between local activity and task performance, are required.

Dorsolateral Prefrontal Cortex

There is a body of evidence that the primate dorsolateral prefrontal cortex (surrounding the principal sulcus and superior convexity) plays an important role during the performance of delayed-response tasks. Many dorsolateral prefrontal neurons whose firing rates increase tonically during the delay period have directional selectivity (Kubota et al., 1974; Niki, 1974; Niki and Watanabe, 1976a; Kojima and Goldman-Rakic, 1982; Kubota and Funahashi, 1982; Funahashi et al., 1989, 1993b, 1997; Rainer et al., 1998; Sawaguchi and Yamane, 1999; Constantinidis et al., 2001). The occurrence of delay-period activity depends more on visual cue location rather than on direction of movement (Niki and Watanabe, 1976a; Funahashi et al., 1993b; Constantinidis et al., 2001; Takeda and Funahashi, 2002). These directional delay-period activities have been considered to be evidence of temporary storage of the spatial information required for the performance of spatial working memory tasks (Goldman-Rakic, 1987; Funahashi and Kubota, 1994; Fuster, 1997). Our detection of dorsolateral prefrontal activation during the ODR task in the both the SPM and ROI-based analysis may be further corrobor-ation of this tonic activation during the delay period in the ODR task.

In addition, rCBF values from the PS and superior convexity correlated positively with correct performance of the ODR task. This correlation may relate to the frequency of reward, because neuronal activities related to reward take place in the dorsolateral prefrontal cortex (Fuster et al., 1982; Watanabe, 1990). Even so, although the scores for correct task performance were higher during the VGS task than during the ODR task, the rCBF values during the VGS task were lower than during the ODR task. This suggests that changes in rCBF in these areas during the ODR task are related more to tonic neuronal activation than to reward. Neurophysiological data which indicate that delay-period activity in the dorsolateral prefrontal cortex decreases in erroneous trials (Fuster, 1973; Niki and Watanabe, 1976a; Funahashi et al., 1989, 1997) also supports conclusion. Our results provide further evidence that the dorsolateral prefrontal cortex participates in spatial working memory, especially the process of temporarily storing information.

Frontal Eye Field

Our application of both SPM and the ROI-based analysis demonstrated AS activation during the ODR task, with the rCBF values in the AS showing positive correlation with successful task performance scores. The AS includes the frontal eye field (FEF), which exists in the anterior bank of the arcuate sulcus. In the FEF, the visual, movement and visuomovement activities have been recorded and the results provide evidence that the FEF plays an important role in generating the eye movement (Bruce and Goldberg, 1985). In addition, some FEF neurons show tonic activation during the delay period of delayed-response tasks (Funahashi et al., 1989; Chafee and Goldman-Rakic, 1998; Ferrera et al., 1999). The activation of the AS might reflect a tonic neuronal activity during the delay period in the FEF.

Orbitofrontal Cortex

Both the SPM and the ROI-based analyses revealed greater rCBF in the lOFC during the ODR task. In addition, the rCBF values of lOFC correlated positively with how successfully the ODR task was performed.

Discovery that the OFC connects the amygdala (Aggleton et al., 1980; Barbas and De Olmos, 1990) and hypothalamus (Johnson, 1968) has provided evidence that the OFC plays an important role in motivational and emotional behavior (Rolls, 1999). Indeed, the presence in the OFC of neurons related to reward behavior (Rosenkilde et al., 1981; Thorpe et al., 1983; Rolls et al., 1990; Tremblay and Schultz, 1999) suggests the possibility that the increase in rCBF during the ODR task was related to receiving a reward. As described in the section on the dorsolateral prefrontal cortex, however, the increase in rCBF during the ODR task was likely due to factors other than reward.

Other studies have provided evidence that the lOFC participates in spatial working memory processes. For example, the lOFC connects with the posterior parietal cortex and the dorsolateral prefrontal cortex (Schwartz and Goldman-Rakic, 1984; Selemon and Goldman-Rakic, 1988; Cavada and Goldman-Rakic, 1989). Lesion of the lOFC also induces slight impairment of spatial delayed-response performance (Oscar-Berman, 1975). In the OFC, some neurons are activated during the delay period of spatial delayed-response tasks (Rosenkilde et al., 1981; Tremblay and Schultz, 1999). Recently, in a functional imaging study, activation in the human ventrolateral frontal cortex was detected when subjects performed spatial working memory tasks (Owen et al., 1996b). Our findings of increased rCBF in the lOFC during the ODR task, along with positive correlation between rCBF values and performance level during the ODR task, also suggest that the lOFC contributes to spatial working memory processes.

Frontal Pole

Studies have yet to clearly reveal the functional role of the primate FP. Human neuroimaging studies have offered evidence that the FP is involved with cognitive function, for example, planning of behavior and working memory (Baker et al., 1996b; Owen et al., 1996b; Koechlin et al., 1999, 2000). The present results, showing an increase in rCBF in the FP during the ODR task and a positive correlation between rCBF in FP with ODR task performance level, also suggest that this area participates in the performance of spatial working memory tasks. Further neurophysiological study is needed, however, to elucidate the functional role of this area in spatial working memory processes.

Anterior Cingulate Cortex

In the present study, although the results from the SPM and the ROI-based analyses indicated an increase in rCBF in the ACC during the ODR task, this change did not correlate with how successfully the ODR task was performed.

The ACC is often linked with emotion. For example, in this region, some neuronal activity has been found to be elicited by visual stimuli associated with reward or aversive stimuli and activation has depended on the significance of the visual stimuli (Nishijo et al., 1997; Koyama et al., 1998). In our present study, however, we found no correlation of ACC rCBF values with task performance results, which corresponds with the number of rewards received. This seems to suggest reward was not associated with greater rCBF in the ACC during the ODR task.

Ablation of the ACC has been reported to cause slight impairment of spatial delayed-response task performance (Meunier et al., 1997). Meanwhile, other research has found some neurons in the ACC that show tonic activation during the delay period of spatial delayed-response tasks (Niki and Watanabe, 1976b). These results, along with our current finding of increased rCBF in the ACC during the ODR task, suggests that the ACC plays some kind of role during spatial working memory tasks. We did not find, however, any significant correlation between rCBF values in the ACC and successful performance scores for the ODR task. This suggests that increased rCBF in the ACC does not simply reflect the tonic increase in spike rate during the delay period of the ODR task. In human event-related fMRI studies, activation of the ACC has been observed during the response period (McCarthy et al., 1996; Belger et al., 1998; Carter et al., 1998; MacDonald et al., 2000). These results and our current findings suggest that the ACC contributes to task education, for example, in processes concerned with performance monitoring, error detection or decision-making.

Posterior Parietal Cortex

In the posterior parietal cortex, especially the lateral bank of intraparietal sulcus (LIP), some neurons have been found to activate during the delay period of memory-guided saccade tasks (Barash et al., 1991a,b; Bracewell et al., 1996; Mazzoni et al., 1996; Snyder et al., 1997; Chafee and Goldman-Rakic, 1998). Monkey metabolic mapping investigation has revealed increased glucose utilization in the inferior parietal lobe during spatial working memory tasks (Friedman and Goldman-Rakic, 1994). In the present study, we found activation in the lIPS during the ODR task. Unlike the prefrontal cortex results, however, changes in rCBF in the lIPS did not significantly correlate with successful performance of the ODR task. These results suggest that during spatial working memory processes the posterior parietal cortex plays a role different to that of the prefrontal cortex. Evidence from other studies also points to a similar conclusion. Primate metabolic mapping has indicated that an enhancement of local cerebral glucose utilization is associated in the dorsolateral prefrontal cortex with memory task difficulty and in the posterior parietal cortex with sensory-motor task parameters (Friedman and Goldman-Rakic, 1994). Inactivation or ablation of the dorsolateral prefrontal cortex severely degrades successful performance of spatial working memory tasks (Bauer and Fuster, 1976; Funahashi et al., 1993a; Levy and Goldman-Rakic, 1999; Sawaguchi and Iba, 2001). On the other hand, inactivation of the posterior parietal cortex causes no or only slight degradation of performance (Bauer and Fuster, 1976; Li et al., 1999). Thus, the activation of the parietal cortex might reflect sensorimotor transformation or preparation for movement rather than actual spatial working memory processes.

SPM analysis also revealed activation in the mIPS during the ODR task. On the other hand, ROI-based analysis did not show any increase in rCBF values in the mIPS during the ODR task. This result suggests that the activation in the mIPS revealed by SPM analysis results from the spreading of the activation of the lIPS.

Prestriate Cortex

During ODR task performance, we found activation in the prestriate cortex. In area V3A, visual response and prestimulus activity are greater in memory-guided saccade tasks than in fixation tasks (Nakamura and Colby, 2000). This greater neuronal activity may induce increased rCBF during ODR tasks. Constant rCBF values from this area, uncorrelated with the successful performance of ODR task, also point to this conclusion.

Hippocampus

There is strong support for the hypothesis that the hippocampus is responsible for spatial memory. In the primate, following hippocampectomy, degradation of spatial delayed-response task performance has been observed (Zola-Morgan and Squire, 1985). Neurophysiological studies have revealed that many hippocampal neurons are activated during the delay period of spatial delayed-response tasks (Watanabe and Niki, 1985; Colombo et al., 1998). There is also evidence that glucose metabolism in the primate hippocampus increases during spatial delayed-response tasks (Friedman and Goldman-Rakic, 1988; Sybirska et al., 2000). In the current study, ROI-based analysis revealed that hippocampal rCBF was greater during the ODR task than during the VGS task. This finding is consistent with previous neuropsychological, neurophysiological and neuroimaging results and supports the hypothesis that the hippocampus plays a role in spatial working memory processes. On the other hand, rCBF changes in the hippocampus correlated negatively with successful task performance scores. This finding, along with the negative correlation of rCBF in the prefrontal cortex and the hippocampus, suggests that, in spatial working memory processes, the functional role of the hippocampus and the prefrontal cortex is distinctly different. One possible functional role in the hippocampus is error detection. In the hippocampus, a number of neurons respond during error trials in spatial delayed response tasks (Watanabe and Niki, 1985). Furthermore, monkeys with middle hippocampus damage have impaired ability to acquire matching-to-sample if training is conducted without error correction (Jackson, 1984). These results and current findings suggest that the hippocampus contributes to error detection.

Neural Circuitry Subserving Spatial Working Memory

We have already emphasized that the simultaneous observation of the activities of the multiple brain areas is one of the benefits of the PET studies. To use this advantage, we correlated activity data from different areas. Our findings show positive correlations between the posterior parietal cortex and the prefrontal cortices and between the prefrontal cortices and negative correlations between the prestriate cortex and posterior parietal cortex and between hippocampus and prefrontal and parietal areas.

These correlations correspond with anatomical findings which show that the spatial information of visual stimuli passes from the prestriate cortex to the posterior parietal cortex (Ungerleider and Mishkin, 1982). Because the posterior parietal cortex is connected with the dorsolateral prefrontal cortex (Schwartz and Goldman-Rakic, 1984; Andersen et al., 1985; Barbas and Mesulam, 1985; Cavada and Goldman-Rakic, 1989), FEF (Barbas and Mesulam, 1981; Cavada and Goldman-Rakic, 1989) and lOFC (Cavada and Goldman-Rakic, 1989), the spatial information can be considered to be sent from the posterior parietal cortex to the dorsolateral prefrontal cortex, FEF and lOFC. The strong correlation of rCBF values from the PS, AS, superior convexity, lOFC and FP suggests that these areas are functionally associated, with functional linkage based on anatomical connections (Barbas and Pandya, 1989). In these areas, rCBF changes correlated with the successful ODR task performance score. These results confirm that these prefrontal cortical regions play an important role to subserve spatial working memory in primates.

Neurophysiological and anatomical studies have offered evidence that the neural network in the dorsolateral prefrontal cortex and posterior parietal cortex plays an important role in the spatial working memory (Cavada and Goldman-Rakic, 1989; Wilson et al., 1993; Chafee and Goldman-Rakic, 1998). In the current study, however, in the lOFC and FP as well as the dorsolateral prefrontal cortex, we found activation related to spatial working memory processes, especially temporary storage. Anatomical data have provided evidence that a neural network that is distributed these areas contributes to the recognition or storage of spatial information (Pandya and Yeterian, 1984; Selemon and Goldman-Rakic, 1988). Both our current results and previous anatomical and neurophysiological results suggest that neural circuitry extending beyond the dorsolateral prefrontal cortex to the AS, lOFC and FP participates in the spatial working memory processes. We were unable, however, to directly measure the functional role of these areas in spatial working memory processes, because PET techniques provide inadequate temporal resolution, typically in the order of tens of seconds. With further research, we hope to clarify the functional role and interaction of these areas in the spatial working memory processes.

Comparison with Human Neuroimaging Data

Many human neuroimaging studies have been carried out to elucidate the brain regions that subserve the spatial working memory processes (Cabeza and Nyberg, 2000). A few comparative studies have tried to determine the brain regions that activate during ODR and VGS tasks, fixation tasks, or when at rest (Anderson et al., 1994; O‚Sullivan et al., 1995; Sweeney et al., 1996). Most studies have revealed activation of the dorsolateral prefrontal cortex during ODR tasks (O‚Sullivan et al., 1995; Sweeney et al., 1996). In addition to the dorsolateral prefrontal cortex, the FEF and supplementary motor area are activated during spatial working memory tasks (Anderson et al., 1994; O‚Sullivan et al., 1995; Sweeney et al., 1996). During ODR task performance there is activation in parietal regions, particularly in areas 7 and 40 (Anderson et al., 1994; O‚Sullivan et al., 1995; Sweeney et al., 1996). ODR tasks are also associated with activation in the anterior cingulate (Anderson et al., 1994; O‚Sullivan et al., 1995) and occipital regions (Anderson et al., 1994; O‚Sullivan et al., 1995).

In this PET study of the monkey, during the ODR task, we found activation in the dorsolateral prefrontal cortex, AS, lOFC, FP, ACC, lIPS and prestriate cortex. These results closely correspond with results from human neuroimaging studies. The similarity of results confirms that in monkeys and humans the same cortical regions are involved in spatial working memory.

This work was supported by in part by the Strategic Promotion System for Brain Science of the Ministry of Education, Culture, Sports, Science and Technology, the Human Frontier Science Program, Grant-in-Aid for Creative Scientific Research of Japan Society for the Promotion of Science.

Figure 1. (A) Diagram of PET scanner and apparatus. (B) Temporal sequence of task events in the ODR (oculomotor delayed-response) tasks and horizontal and vertical eye traces. F, fixation period (1 s); C, cue period (0.1 s); D, delay period (2 s); R, response period. (C) VGS (visually guided saccade) task: F, fixation period (3.1 s); R, response period. (D) Visual stimulus positions of the ODR and VGS task; FP, fixation point. (E) Temporal sequence of PET scan.

Figure 1. (A) Diagram of PET scanner and apparatus. (B) Temporal sequence of task events in the ODR (oculomotor delayed-response) tasks and horizontal and vertical eye traces. F, fixation period (1 s); C, cue period (0.1 s); D, delay period (2 s); R, response period. (C) VGS (visually guided saccade) task: F, fixation period (3.1 s); R, response period. (D) Visual stimulus positions of the ODR and VGS task; FP, fixation point. (E) Temporal sequence of PET scan.

Figure 2. ROI templates drawn on monkey G’s MRI scan prior to application to PET scan. 1, lateral bank of the intraparietal sulcus; 2, medial bank of the intraparietal sulcus; 3, posterior cingulate cortex; 4, area 7; 5, prestriate cortex; 6, superior convexity; 7, anterior bank of the arcuate sulcus; 8, the area surrounding the principal sulcus; 9, anterior cingulate cortex; 10, inferior convexity; 11, frontal pole; 12, lateral orbitofrontal cortex; 13, medial orbitofrontal cortex; 14, hippocampus; 15, superior temporal sulcus.

Figure 2. ROI templates drawn on monkey G’s MRI scan prior to application to PET scan. 1, lateral bank of the intraparietal sulcus; 2, medial bank of the intraparietal sulcus; 3, posterior cingulate cortex; 4, area 7; 5, prestriate cortex; 6, superior convexity; 7, anterior bank of the arcuate sulcus; 8, the area surrounding the principal sulcus; 9, anterior cingulate cortex; 10, inferior convexity; 11, frontal pole; 12, lateral orbitofrontal cortex; 13, medial orbitofrontal cortex; 14, hippocampus; 15, superior temporal sulcus.

Figure 3. Activation sites in monkey G where blood flow increased significantly (P < 0.05) more during the ODR task than the VGS task. Color-coded z-score maps were overlaid on the MR coronal and transaxial images. (A) Frontal pole; (B) anterior cingulate cortex; (C) superior convexity; (D) principal sulcus; (E) arcuate sulcus; and (F) lateral orbitofrontal cortex.

Figure 3. Activation sites in monkey G where blood flow increased significantly (P < 0.05) more during the ODR task than the VGS task. Color-coded z-score maps were overlaid on the MR coronal and transaxial images. (A) Frontal pole; (B) anterior cingulate cortex; (C) superior convexity; (D) principal sulcus; (E) arcuate sulcus; and (F) lateral orbitofrontal cortex.

Figure 4. Normalized rCBF values during ODR and VGS task in the area surrounding the principal sulcus (PS). Normalized rCBF values during ODR were higher than during VGS.

Figure 4. Normalized rCBF values during ODR and VGS task in the area surrounding the principal sulcus (PS). Normalized rCBF values during ODR were higher than during VGS.

Figure 5. Scatter diagrams showing the correlations during the ODR task between the correct performance score (%) and normalized rCBF values in the regions of interest. (A) PS; (B) AS; (C) superior convexity; (D) lOFC; (E) FP; (F) ACC; (G) lIPS; (H) hippocampus; and (I) prestriate cortex of monkey G and monkey N. Normalized rCBF values from the PS, AS, superior convexity, lOFC and FP correlated positively and values from the hippocampus correlated negatively, with ODR task performance.

Figure 5. Scatter diagrams showing the correlations during the ODR task between the correct performance score (%) and normalized rCBF values in the regions of interest. (A) PS; (B) AS; (C) superior convexity; (D) lOFC; (E) FP; (F) ACC; (G) lIPS; (H) hippocampus; and (I) prestriate cortex of monkey G and monkey N. Normalized rCBF values from the PS, AS, superior convexity, lOFC and FP correlated positively and values from the hippocampus correlated negatively, with ODR task performance.

Figure 6. Scatter diagrams showing correlations between normalized rCBF values in the PS and in other regions of interest. (A) AS; (B) superior convexity; (C) lOFC; (D) FP; (E) ACC; (F) lIPS; (G) hippocampus; and (H) prestriate cortex. Normalized rCBF values from the PS correlated positively with values from the AS, superior convexity, lOFC, FP, ACC and IPS and negatively with values from the hippocampus.

Figure 6. Scatter diagrams showing correlations between normalized rCBF values in the PS and in other regions of interest. (A) AS; (B) superior convexity; (C) lOFC; (D) FP; (E) ACC; (F) lIPS; (G) hippocampus; and (H) prestriate cortex. Normalized rCBF values from the PS correlated positively with values from the AS, superior convexity, lOFC, FP, ACC and IPS and negatively with values from the hippocampus.

Figure 7. Interaction between regions of interest during the ODR task. The strength of correlation is proportional to the width of the lines. Values for width gradient are shown below the figure. Solid lines show positive and dashed lines negative, correlations.

Figure 7. Interaction between regions of interest during the ODR task. The strength of correlation is proportional to the width of the lines. Values for width gradient are shown below the figure. Solid lines show positive and dashed lines negative, correlations.

Table 1


 Sites of activation, determined by SPM, during the ODR task

 Hemisphere Monkey G  Monkey N 
  A–P L–R z score  A–P L–R z score 
Frontal           
 PS A28 R15.5 +22 3.43  – – – – 
 – – – –  A25.5 L13 +18 2.06 
Sup. convexity A30.5 R11 +26 2.63  A30 R14 +27 2.19 
 A30.5 L7 +26 2.13  – – – – 
 AS A27 R17.5 +18 2.31  – – – – 
 A27 L15 +14 2.44  A24 L15.5 +17 2.00 
 OFC A25 R15 +11 2.13  A24 R16.5 +10 2.38 
 A25 L13 +13 2.75  A21.5 L16.5  +8 2.31 
 FP A39 L1 +18 2.19  A36.5 L0.5 +24 3.19 
 ACC A32.5 L3.5 +23 2.06  A31 L1 +24 2.13 
Parietal           
 IPS P2 R6 +25 2.25  – – – – 
 A3 L8.5 +25 2.81  P4 L6 +18 2.19 
 PCC – – – –  A1.5 R2.5 +18 2.06 
 A2 L2.5 +25.5 2.19  – – – – 
Temporal           
 STS A13 R14  +4 2.13  – – – – 
 A13 L16.5  +4 2.75  A13.5 L19   0 2.63 
Occipital           
 Prestriate cortex P14 R5 +14.5 3.56  P12.5 R13 +11 2.56 
 P12.5 L7 +10.5 2.06  P14 L8.5 +10 2.31 
 Hemisphere Monkey G  Monkey N 
  A–P L–R z score  A–P L–R z score 
Frontal           
 PS A28 R15.5 +22 3.43  – – – – 
 – – – –  A25.5 L13 +18 2.06 
Sup. convexity A30.5 R11 +26 2.63  A30 R14 +27 2.19 
 A30.5 L7 +26 2.13  – – – – 
 AS A27 R17.5 +18 2.31  – – – – 
 A27 L15 +14 2.44  A24 L15.5 +17 2.00 
 OFC A25 R15 +11 2.13  A24 R16.5 +10 2.38 
 A25 L13 +13 2.75  A21.5 L16.5  +8 2.31 
 FP A39 L1 +18 2.19  A36.5 L0.5 +24 3.19 
 ACC A32.5 L3.5 +23 2.06  A31 L1 +24 2.13 
Parietal           
 IPS P2 R6 +25 2.25  – – – – 
 A3 L8.5 +25 2.81  P4 L6 +18 2.19 
 PCC – – – –  A1.5 R2.5 +18 2.06 
 A2 L2.5 +25.5 2.19  – – – – 
Temporal           
 STS A13 R14  +4 2.13  – – – – 
 A13 L16.5  +4 2.75  A13.5 L19   0 2.63 
Occipital           
 Prestriate cortex P14 R5 +14.5 3.56  P12.5 R13 +11 2.56 
 P12.5 L7 +10.5 2.06  P14 L8.5 +10 2.31 
Table 2


 Differences of rCBF in ROI between ODR task and VGS task

 Monkey G  Monkey N  t P 
     
PS  1.51  1.22   2.97  2.61  3.44   0.0009 
AS  1.29  1.25   1.16  0.50  2.07   0.04 
Sup. convexity  1.53  1.31   1.49  1.25  2.41   0.02 
Inf. convexity  0.63  0.21   0.65 –1.80  0.08   0.94 
lOFC  2.47  1.97   2.47  3.88  4.10 <0.0001 
mOFC  0.97 –0.32   0.60  0.10  0.82   0.41 
FP  2.14  1.48   6.89  2.21  4.44 <0.0001 
ACC  2.03  0.48   2.14  4.72  6.07 <0.0001 
area7  1.18  0.78   1.33  1.12  1.4   0.17 
lIPS  1.47  0.52   1.11  0.94  3.12   0.003 
mIPS  0.85  0.17  –0.14 –0.68  0.21   0.84 
PCC –1.08 –1.25   1.03 –3.28  1.48   0.14 
STS  0.53  0.64  –0.61 –1.33  0.28   0.78 
Hippocampus  1.31  0.70   1.74  2.33  2.41   0.001 
Prestriate cortex  2.27  2.12   1.86  2.61  4.25 <0.0001 
 Monkey G  Monkey N  t P 
     
PS  1.51  1.22   2.97  2.61  3.44   0.0009 
AS  1.29  1.25   1.16  0.50  2.07   0.04 
Sup. convexity  1.53  1.31   1.49  1.25  2.41   0.02 
Inf. convexity  0.63  0.21   0.65 –1.80  0.08   0.94 
lOFC  2.47  1.97   2.47  3.88  4.10 <0.0001 
mOFC  0.97 –0.32   0.60  0.10  0.82   0.41 
FP  2.14  1.48   6.89  2.21  4.44 <0.0001 
ACC  2.03  0.48   2.14  4.72  6.07 <0.0001 
area7  1.18  0.78   1.33  1.12  1.4   0.17 
lIPS  1.47  0.52   1.11  0.94  3.12   0.003 
mIPS  0.85  0.17  –0.14 –0.68  0.21   0.84 
PCC –1.08 –1.25   1.03 –3.28  1.48   0.14 
STS  0.53  0.64  –0.61 –1.33  0.28   0.78 
Hippocampus  1.31  0.70   1.74  2.33  2.41   0.001 
Prestriate cortex  2.27  2.12   1.86  2.61  4.25 <0.0001 

References

Aggleton JP, Burton MJ, Passingham RE (
1980
) Cortical and subcortical afferents to the amygdala of the rhesus monkey (Macaca mulatta).
Brain Res
 
190
:
347–
368.
Andersen RA, Asanuma C, Cowan WM (
1985
) Callosal and prefrontal associational projecting cell populations in area 7A of the macaque monkey: a study using retrograde transported fluorescent dyes.
J Comp Neurol
 
232
:
443–
455.
Andersen RA, Asanuma C, Essick G, Siegel RM (
1990
) Corticocortical connections of anatomically and physiologically defined subdivisions within the inferior parietal lobule.
J Comp Neurol
 
296
:
65–
113.
Anderson TJ, Jenkins IH, Brooks DJ, Hawken MB, Frackowiak RS (
1994
) Cortical control of saccades and fixation in man a PET study.
Brain
 
117
:
1073–
1084.
Baddeley A (
1992
) Working memory.
Science
 
255
:
556–
559.
Baker SC, Frith CD, Frackowiak RSJ, Dolan RJ (
1996
) Active representation of shape and spatial location in man.
Cereb Cortex
 
6
:
612–
619.
Baker SC, Rogers RD, Owen AM, Frith CD, Dolan RJ, Frackowiak RSJ, Robbins TW (
1996
) Neural systems engaged by planning: a PET study of the Tower of London task.
Neuropsychologia
 
34
:
515–
526.
Barash S, Bracewell RM, Fogassi L, Gnadt JW, Andersen RA (
1991
) Saccade-related activity in the lateral intraparietal area. I. Temporal properties; comparison with area 7a.
J Neurophysiol
 
66
:
1095–
1108.
Barash S, Bracewell RM, Fogassi L, Gnadt JW, Andersen RA (
1991
) Saccade-related activity in the lateral intraparietal area. II. Spatial properties.
J Neurophysiol
 
66
:
1109–
1124.
Barbas H, De Olmos J (
1990
) Projection from amygdala to basoventral and mediodorsal prefrontal regions in the rhesus monkey.
J Comp Neurol
 
300
:
549–
571.
Barbas H, Mesulam MM (
1981
) Organization of afferent input to subdivisions of area 8 in the rhesus monkey.
J Comp Neurol
 
200
:
407–
431.
Barbas H, Mesulam MM (
1985
) Cortical afferent input to the principalis region of the rhesus monkey.
Neuroscience
 
15
:
619–
637.
Barbas H, Pandya DN (
1989
) Architecture and intrinsic connections of the prefrontal cortex in the rhesus monkey.
J Comp Neurol
 
286
:
353–
375.
Bauer RH, Fuster JM (
1976
) Delayed-matching and delayed-response deficit from cooling dorsolateral prefrontal cortex in monkeys.
J Comp Physiol Psychol
 
90
:
293–
302.
Belger A, Puce A, Krystal JH, Gore JC, Goldman-Rakic PS, McCarthy G (
1998
) Dissociation of mnemonic and perceptual processes during spatial and nonspatial working memory using fMRI.
Hum Brain Mapp
 
6
:
14–
32.
Bracewell RM, Mazzoni P, Barash S, Andersen RA (
1996
) Motor intention activity in the macaque’s lateral intraparietal area. II. Changes of motor plan.
J Neurophysiol
 
76
:
1457–
1464.
Bruce CJ, Goldberg ME (
1985
) Primate frontal eye fields. I. Single neurons discharging before saccades.
J Neurophysiol
 
53
:
603–
635.
Cabeza R, Nyberg L (
2000
) Imaging cognition. II. An empirical review of 275 PET and fMRI studies.
J Cogn Neurosci
 
12
:
1–
47.
Carter CS, Braver TS, Barch DM, Botvinik MW, Noll D, Cohen JD (
1998
) Anterior cingulate cortex, error detection and the online monitoring of performance.
Science
 
280
:
747–
749.
Cavada C, Goldman-Rakic PS (
1989
) Posterior parietal cortex in rhesus monkey. II., Evidence for segregated corticocortical networks linking sensory and limbic areas with the frontal lobe.
J Comp Neurol
 
287
:
422–
445.
Chafee MV, Goldman-Rakic PS (
1998
) Matching patterns of activity in primate prefrontal area 8a and parietal area 7ip neurons during a spatial working memory task.
J Neurophysiol
 
79
:
2919–
2940.
Colombo M, Fernandez T, Nakamura K, Gross CG (
1998
) Functional differentiation along the anterior–posterior axis of the hippocampus in monkeys.
J Neurophysiol
 
80
:
1002–
1005.
Constantinidis C, Franowicz MN, Goldman-Rakic PS (
2001
) The sensory nature of mnemonic representation in the primate prefrontal cortex.
Nat Neurosci
 
4
:
311–
316.
Ferrera VP, Cohen JL, Lee BB (
1999
) Activity of prefrontal neurons during location and color delayed matching tasks.
Neuroreport
 
10
:
1315–
1322.
Friedman HR, Goldman-Rakic PS (
1988
) Activation of the hippocampus and dentate gyrus by working-memory: a 2-deoxyglucose study of behaving rhesus monkeys.
J Neurosci
 
8
:
4693–
4706.
Friedman HR, Goldman-Rakic PS (
1994
) Coactivation of prefrontal cortex and inferior parietal cortex in working memory tasks revealed by 2DG functional mapping in the rhesus monkey.
J Neurosci
 
14
:
2775–
2788.
Funahashi S, Bruce CJ, Goldman-Rakic PS (
1989
) Mnemonic coding of visual space in the monkey’s dorsolateral prefrontal cortex.
J Neurophysiol
 
61
:
331–
349.
Funahashi S, Bruce CJ, Goldman-Rakic PS (
1993
) Dorsolateral prefrontal lesions and oculomotor delayed-response performance: evidence for mnemonic ‘scotomas’.
J Neurosci
 
13
:
1479–
1497.
Funahashi S, Chafee MV, Goldman-Rakic PS (
1993
) Prefrontal neuronal activity in rhesus monkeys performing a delayed anti-saccade task.
Nature
 
365
:
753–
756.
Funahashi S, Inoue M, Kubota K (
1997
) Delay-period activity in the primate prefrontal cortex encoding multiple spatial positions and their order of presentation.
Behav Brain Res
 
84
:
203–
223.
Funahashi S, Kubota K (
1994
) Working memory and prefrontal cortex.
Neurosci Res
 
21
:
1–
11.
Fuster JM (
1973
) Unit activity in prefrontal cortex during delayed-response performance: neuronal correlates of transient memory.
J Neurophysiol
 
36
:
61–
78.
Fuster JM (
1997
) The prefrontal cortex. Philadelphia, PA: Lippincott-Raven.
Fuster JM, Bauer, RH, Jervey JP (
1982
) Cellular discharge in the dorsolateral prefrontal cortex of the monkey in cognitive task.
Exp Neurol
 
77
:
679–
694.
Goldberg TE, Berman KF, Randolph C, Gold JM, Weinberger DR (
1996
) Isolating the mnemonic component in spatial delayed response: a controlled PET 15O-labeled water regional cerebral blood flow study in normal humans.
Neuroimage
 
3
:
69–
78.
Goldman-Rakic PS (
1987
) Circuitry of primate prefrontal cortex and regulation of behavior by representational memory. In: Handbook of physiology, the nervous system, higher functions of the brain (Mountcastle VB, Plum F, Geiger SR, eds), vol. 5, pp.
373–
417. Bethesda, MD: American Physiological Society.
Goldman-Rakic PS (
1988
) Topography of cognition: parallel distributed networks in primate association cortex.
Annu Rev Neurosci
 
11
:
137–
156.
Goldman-Rakic PS (
1996
) The prefrontal landscape: implications of functional architecture for understanding human mentation and the central executive.
Phil Trans R Soc Lond B
 
351
:
1445–
1453.
Horwitz B (
1990
) Simulating functional interactions in the brain: a model for examining correlations between regional cerebral metabolic rates.
Int J Biomed Comput
 
26
:
149–
170.
Horwitz B (
1991
) Functional interactions in the brain: use of correlations between regional metabolic rates.
J Cereb Blood Flow Metab
 
11
:
114–
120.
Inoue M, Mikami A, Ando I, Tsukada H (
1999
) Functional mapping of the macaque brain region related to spatial working memory.
Soc Neurosci Abstr
 
25
:884.
Inoue M, Mikami A, Ando I, Tsukada H (
2001
) Activity of the primate prefrontal cortex is correlated with performance level of the spatial working memory task.
Soc Neurosci Abstr
 
27
:852.17.
Jackson WJ (
1984
) Regional hippocampal lesions alter matching by monkeys: an anorexiant effect.
Physiol Behav
 
32
:
593–
601.
Johnson TN (
1968
) Projection from behaviorally-defined sectors of the prefrontal cortex to the basal ganglia, septum and diencephalon of the monkey.
Exp Neurol
 
21
:
20–
34.
Koechlin E, Basso G, Pietrini P, Penzer S, Grafman J (
1999
) The role of the anterior prefrontal cortex in human cognition.
Nature
 
399
:
148–
151.
Koechlin E, Corrado G, Pietrini P, Grafman J (
2000
) Dissociating the role of the medial and lateral anterior prefrontal cortex in human planning.
Proc Natl Acad Sci USA
 
97
:
7651–
7656.
Kojima S, Goldman-Rakic PS (
1982
) Delay-related activity of prefrontal neurons in rhesus monkeys performing delayed response.
Brain Res
 
248
:
43–
49.
Koyama T, Tanaka YZ, Mikami A (
1998
) Nociceptive neurons in the macaque anterior cingulate activate during anticipation of pain.
Neuroreport
 
9
:
2663–
2667.
Kubota K, Funahashi S (
1982
) Direction-specific activities of dorsolateral prefrontal and motor cortex pyramidal tract neurons during visual tracking.
J Neurophysiol
 
47
:
362–
376.
Kubota K, Iwamoto T, Suzuki H (
1974
) Visuokinetic activities of primate prefrontal neurons during delayed-response performance.
J Neurophysiol
 
37
:
1197–
1212.
Levy R, Goldman-Rakic PS (
1999
) Association of storage and processing functions in the dorsolateral prefrontal cortex of the nonhuman primate.
J Neurosci
 
19
:
5149–
5158.
Li CR, Mazzoni P, Andersen RA (
1999
) Effect of reversible inactivation of macaque lateral intraparietal area on visual and memory saccade.
J Neurophysiol
 
81
:
1827–
1838.
MacDonald AW III, Cohen JD, Stenger VA, Carter CS (
2000
) Dissociating the role of the dorsolateral prefrontal and anterior cingulate cortex in cognitive control.
Science
 
288
:
1835–
1838.
McCarthy G, Puce A, Constable RT, Krystal JH, Gore JC, Goldman-Rakic PS (
1996
) Activation of human prefrontal cortex during spatial and nonspatial working memory tasks measured by functional MRI.
Cereb Cortex
 
6
:
600–
611.
Mazzoni P, Bracewell RM, Barash S, Andersen RA (
1996
) Motor intention activity in the macaque’s lateral intraparietal area. I. Dissociation of motor plan from sensory memory.
J Neurophysiol
 
76
:
1439–
1456.
Meunier M, Bachevalier J, Mishkin M (
1997
) Effects of orbital frontal and anterior cingulate lesions on object and spatial memory in rhesus monkeys.
Neuropsychologia
 
35
:
999–
1015.
Nakamura K, Colby CL (
2000
) Visual, saccade-related and cognitive activation of single neurons in monkey extrastriate area V3A.
J Neurophysiol
 
84
:
677–
692.
Niki H (
1974
) Differential activity of prefrontal units during right and left delayed response trials.
Brain Res
 
70
:
346–
349.
Niki H, Watanabe M (
1976
) Prefrontal unit activity and delayed response: relation to cue location versus direction of response.
Brain Res
 
105
:
79–
88.
Niki H, Watanabe M (
1976
) Cingulate unit activity and delayed response.
Brain Res
 
110
:
381–
386.
Nishijo H, Yamamoto Y, Ono T, Uwano T, Yamashita J, Yamashima T (
1997
) Single neuron responses in the monkey anterior cingulate cortex during visual discrimination.
Neurosci Lett
 
227
:
79–
82.
Onoe H, Komori M, Onoe K, Takechi H, Tsukada H, Watanabe Y (
2001
) Cortical networks recruited for time perception: a monkey positron emission tomography (PET) study.
Neuroimage
 
13
:
37–
45.
Oscar-Berman M (
1975
) The effects of dorsolateral-frontal and ventrolateral-orbitofrontal lesions on spatial discrimination learning and delayed response in two modalities.
Neuropsychologia
 
13
:
237–
246.
O‘Sullivan EP, Jenkins IH, Henderson L, Kennard C, Brooks DJ (
1995
) The functional anatomy of remembered saccades: a PET study.
Neuroreport
 
6
:
2141–
2144.
Owen AM, Doyon J, Petrides M, Evans AC (
1996
) Planning and spatial working memory: a positron emission tomography study in humans.
Eur J Neurosci
 
8
:
353–
364.
Owen AM, Evans AC, Petrides M (
1996
) Evidence for a two-stage model of spatial working memory processing within the lateral frontal cortex: a positron emission tomography study.
Cereb Cortex
 
6
:
31–
38.
Owen AM, Stern CE, Look RB, Tracy I, Rosen BR, Petrides M (
1998
) Functional organization of spatial and nonspatial working memory processing within the human lateral frontal cortex.
Proc Natl Acad Sci USA
 
95
:
7721–
7726.
Pandya DN, Yeterian EH (
1984
) Proposed neural circuitry for spatial memory in the primate brain.
Neuropsychologia
 
22
:
109–
122.
Petrides M (
1996
) Specialized systems for the processing of mnemonic information within the primate frontal cortex.
Phil Trans R Soc Lond B
 
351
:
1455–
1462.
Rainer G, Asaad WF, Miller EK (
1998
) Memory fields of neurons in the primate prefrontal cortex.
Proc Natl Acad Sci USA
 
95
:
15008–
15013.
Rolls ET (
1999
) The brain and emotion. Oxford: Oxford University Press.
Rolls ET, Yaxley S, Sienkiewicz ZJ (
1990
) Gustatory responses of single neurons in the caudolateral orbitofrontal cortex of the macaque monkey.
J Neurophysiol
 
64
:
1055–
1066.
Rosenkilde CE, Bauer RH, Fuster JM (
1981
) Single cell activity in ventral prefrontal cortex of behaving monkey.
Brain Res
 
209
:
375–
394.
Sawaguchi T, Iba M (
2001
) Prefrontal cortical representation of visuospatial working memory in monkeys examined by local inactivation with mucimol.
J Neurophysiol
 
86
:
2041–
2053.
Sawaguchi T, Yamane I (
1999
) Properties of delay-period neuronal activity in the monkey dorsolateral prefrontal cortex during a spatial delayed matching-to-sample task.
J Neurophysiol
 
82
:
2070–
2080.
Schwartz ML, Goldman-Rakic PS (
1984
) Callosal and intrahemispheric connectivity of the prefrontal association cortex in rhesus monkey: relation between intraparietal and principal sulcal cortex.
J Comp Neurol
 
226
:
403–
420.
Selemon LD, Goldman-Rakic PS (
1988
) Common cortical and subcortical targets of the dorsolateral prefrontal and posterior parietal cortices in the rhesus monkey: evidence for distributed neural network subserving spatially guided behavior.
J Neurosci
 
8
:
4049–
4068.
Smith EE, Jonides J, Koeppe RA (
1996
) Dissociating verbal and spatial working memory using PET.
Cereb Cortex
 
6
:
11–
20.
Snyder LH, Batista AP, Andersen RA (
1997
) Coding of intention in the posterior parietal cortex.
Nature
 
386
:
167–
170.
Sweeney JA, Mintun MA, Kwee S, Wiseman MB, Brown DL, Rosenberg DR, Carl JR (
1996
) Positron emission tomography study of voluntary saccadic eye movements and spatial working memory.
J Neurophysiol
 
75
:
454–
468.
Sybirska E, Davachi L, Goldman-Rakic PS (
2000
) Prominence of direct entorhinal-CA1 pathway activation in sensorimotor and cognitive tasks revealed by 2-DG functional mapping in nonhuman primate.
J Neurosci
 
20
:
5827–
5834.
Sybirska E, Xiong W, Goldman-Rakic PS (
2001
) Focal activation of prefrontal cortex revealed by 2-deoxyglucose imaging of rhesus monkey performing cognitive task.
Soc Neurosci Abstr
 
27
:852.4.
Takeda K, Funahashi S (
2002
) Prefrontal task-related activity representing visual cue location or saccadic direction in spatial working memory task.
J Neurophysiol
 
87
:
567–
588.
Tanaka H, Onoe H, Tsukada H, Fujita I (
2001
) Attentional modulation of neural activity in the macaque inferior temporal cortex during global and local processing.
Neurosci Res
 
39
:
69–
472.
Thorpe SJ, Rolls ET, Maddison S (
1983
) The orbitofrontal cortex: neuronal activity in the behaving monkey.
Exp Brain Res
 
49
:
93–
115.
Tremblay L, Schultz E (
1999
) Relative reward preference in primate orbitofrontal cortex.
Nature
 
398
:
704–
708.
Ungerleider LG, Mishkin M (
1982
) Two cortical visual systems. In: Analysis of visual behavior (Ingle DJ, Goodale MA, Mansfield RJW, eds), pp.
549–
586. Cambridge, MA: MIT Press.
Wang G, Tanifuji M, Tanaka K (
1998
) Functional architecture in monkey inferotemporal cortex revealed by in vivo optical imaging.
Neurosci Res
 
32
:
33–
46.
Watanabe M (
1990
) Prefrontal unit activity during associative learning in the monkey.
Exp Brain Res
 
80
:
296–
309.
Watanabe T, Niki H (
1985
) Hippocampal unit activity and delayed response in the monkey.
Brain Res
 
325
:
241–
254.
Watanabe M, Okada H, Shimizu K, Omura T, Yoshikawa E, Kosugi T, Mori S, Yamashita T (
1997
) A high resolution animal PET scanner using compact PS-PMT detectors.
IEEE Trans Nucl Sci
 
44
:
1277–
1282.
Wilson FAW, O’Scalaidhe SP, Goldman-Rakic PS (
1993
) Dissociation of object and spatial processing in primate prefrontal cortex.
Science
 
260
:
1955–
1958.
Zola-Morgan S, Squire LR (
1985
) Medial temporal lesions in monkeys impair memory on a variety of tasks sensitive to human amnesia.
Behav Neurosci
 
99
:
22–
34.