Single unit activity in a border area between the dorsal prefrontal and premotor regions in the visually conditioned motor task of monkeys

Supplementary motor area structure and function: Review and hypotheses

Though its existence has been known for well over 30 years, only recently has the supplementary motor area (SMA) and its role in the cortical organization of movement come to be examined in detail by neuroscientists. Evidence from a wide variety of investigational perspectives is reviewed in an attempt to synthesize a conceptual framework for understanding SMA function. It is suggested that the SMA has an important role to play in the intentional process whereby internal context influences the elaboration of action. It may be viewed as phylogenetically older motor cortex, derived from anterior cingulate periarchicortical limbic cortex, which, as a key part of a medial premotor system, is crucial in the “programming” and fluent execution of extended action sequences which are “projectional” in that they rely on model-based prediction. This medial system can be distinguished from a lateral premotor system postulated to have evolved over phylogeny from a different neural source. An anatomico-physiologic model of the medial premotor system is proposed which embodies the principles of cyclicity and reentrance in the process of selecting those neural components to become active in conjunction with the performance of a particular action. The postulated dynamic action of this model in the microgenesis of a discrete action is outlined. It is concluded that although there is a great deal to be learned about the SMA, a convergence of current evidence can be identified. Such evidence suggests that the SMA plays an important role in the development of the intention-to-act and the specification and elaboration of action through its mediation between medial limbic cortex and primary motor cortex.

Activity in ventral and dorsal premotor cortex in response to predictable force-pulse perturbations in a precision grip task

This study compared the responses of ventral and dorsal premotor cortex (PMv and PMd) neurons to predictable force-pulse perturbations applied during a precision grip. Three monkeys were trained to grasp an unseen instrumented object between the thumb and index finger and to lift and hold it stationary within a position window for 2-2.5 s. The grip and load forces and the object displacement were measured on each trial. Single-unit activity was recorded from the hand regions in the PMv and PMd. In some conditions a predictable perturbation was applied to the object after 1,500 ms of static holding, whereas in other conditions different random combinations of perturbed and unperturbed trials were given. In the perturbed conditions, some were randomly and intermittently presented with a warning flash, whereas some were unsignaled. The activities of 198 cells were modulated during the task performance. Of these cells, 151 were located in the PMv, and 47 were located in the PMd. Although both PMv and PMd neurons had similar discharge patterns, more PMd neurons (84 vs. 43%) showed early pregrip activity. Forty of 106 PMv and 10/30 PMd cells responded to the perturbation with reflexlike triggered reactions. The latency of this response was always <100 ms with a mean of about 55 ms in both the PMv and the PMd. In contrast, 106 PMv and 30 PMd cells tested with the perturbations, only 9 and 10%, respectively, showed significant but nonspecific adaptations to the perturbation. The warning stimulus did not increase the occurrence of specific responses to the perturbation even though 21 of 42 cells related to the grip task also responded to moving visual stimuli. The responses were retinal and frequently involved limited portions of both foveal and peripheral visual fields. When tested with a 75 x 5.5-cm dark bar on a light background, these cells were sensitive to the direction of movement. In summary, the periarcuate premotor area activity to related to predictable force-pulse perturbations seems to reflect a general increase in excitability in contrast to a more specific anticipatory activity such as recorded in the cerebellum. In spite of the strong cerebello-thalamo-cortical projections, the results of the present study suggest that the cortical premotor areas are not involved in the elaboration of adaptive internal models of hand-object dynamics.

Radioactive 2-DG incorporation patterns in the mesial frontal cortex of task-performing monkeys

The pattern of radioactive 2-deoxy-D-glucose (2-DG) uptake in the rostral mesial cortex was investigated in seven hemispheres of four task-performing monkeys (a delayed-response task performed with a forelimb). A two-dimensional 2-DG map was constructed from frontal sections. Blob-like 2-DG incorporation sites (2-DG active sites) were observed in single frontal sections, e.g., in the anterior cingulate gyrus (CiG) and supplementary and primary motor cortices in the mesial surface, and around the superior precentral sulcus in the premotor area. Blob-like 2-DG incorporation sites were also observed in the medial part of the dorsal frontal cortex near the midline. However, most of these blob-like 2-DG active sites were revealed in fact not to be blobs. They formed rostrocaudally continuous streaks when they were constructed in a two-dimensional map. Streaks fused with one another in some areas, and gave off branches in other areas. These 2-DG uptake patterns were similar between the paired left and right hemispheres of three brains. It is highly probable that these 2-DG active streaks (or blobs) reflected neuronal activity related to somatomotor and/or eye movements, because the 2-DG-labeled areas included motor, premotor, supplementary motor, and possibly part of the supplementary eye fields. It is also probable that this 2-DG incorporation was related to cognitive or memory functions, because neuronal activity related to performance of a delayed-response was reported in the rostral mesial cortex and in the CiG.

Unimanual motor learning impaired by frontomedial and insular lesions in man

We correlated impaired unimanual motor learning with the lesion site in 53 patients with chronic lesions predominantly of the frontal lobe. The lesions were assessed using computed tomography (CT), then transferred to standard templates of nine slices parallel to the canthomeatal plane and digitized with a raster matrix of 3 mm by 3 mm width. The learning task was to track a moving target on a computer screen with a dot guided by the preferred hand, while the horizontal coupling between hand movement and screen was inverted. The mean tracking error was recorded over eight successive trials of 80s duration. If the mean error of the last three trials was not lower than that of the first three trials, impaired motor learning was assumed. We correlated performance and lesion with a contingency table analysis for each raster element. Impaired motor learning was associated with a lesion within the supplementary motor area and adjacent anterior cingulate, and within the anterior insular region. Our results indicate that these regions are critical for motor learning and functional plasticity in man. Our data support activation patterns obtained with positron emission tomography.

Effects of attention on visuomotor activity in the premotor and prefrontal cortex of a primate

We examined neuronal activity in the primate premotor (PM) and prefrontal (PF) areas during a demanding spatial matching task. On each behavioral trial, a rhesus monkey moved its forelimb when a visual stimulus, called the "prime stimulus," reappeared at a previously cued location. Because it triggered a movement, the part of space cued by the prime stimulus had to be either remembered or attended during the time between prime stimulus presentations. Between the first and second appearances of the prime stimulus, behaviorally irrelevant visual stimuli could appear at one or several locations other than that of the prime stimulus. We could thereby examine the activity that followed a stimulus when it was attended versus when it was irrelevant and presumably unattended. We found that visuospatial attention affected neuronal activity in both the motor and "nonmotor" parts of the frontal cortex. The magnitude of attention effects exceeded that previously reported--a finding that probably resulted from the intensive attentional demands of the present task.

Cortical connections of the inferior arcuate sulcus cortex in the macaque brain

Injections of the retrograde/anterograde tracers Wheat Germ Agglutinin-Horseradish peroxidase (WGA-HRP) into the cortex along the banks of the inferior limb of the arcuate sulcus in the cortex of 4 macaque monkeys (Macaca fascicularis) were used to investigate its cortico-cortical connections. All injections produced transported label within the sulcus principalis, the ventral lateral prefrontal cortex, the anterior cingulate sulcus and the dorsal insular cortex. The distribution of label within each of these areas differed slightly depending on the injection site. Injections along the caudal bank of the inferior arcuate sulcus label premotor, supplementary motor, and precentral motor areas but produce relatively sparse prefrontal labeling. Posteriorly label is transported to the inferior parietal cortex and the dorsal opercular bank of the Sylvian fissure. Injections along the rostral bank of the sulcus do not label motor areas but produce labeling in dorsal, lateral and orbital prefrontal areas, and in cortex along the ventral bank of the superior branch of the arcuate sulcus. Posteriorly label is transported to cortical areas in the superior temporal gyrus including the dorsal bank of the superior temporal sulcus. The more dorsal rostral bank injection produced both superior temporal and some sparse inferior parietal labeling and the more ventral rostral bank injection produced extensive superior temporal labeling but no parietal labeling. No labeling was ever seen in cortex ventral to the fundus of the superior temporal sulcus. Although other auditory recipient prefrontal areas have been reported, this is the first demonstration of a region chiefly devoted to auditory connections within the ventral frontal cortex. Its adjacency to areas associated with vocal muscle movement, and its connections to midline cortical areas associated with vocal functions in both primates and humans may provide important clues to the organization of Broca's language area.

Attention-related unit activity in the frontal association cortex during a go/no-go visual discrimination task

Unit activity related to a go/no-go visual discrimination task was studied in four rhesus monkeys. We recorded 272 task-related cells from frontal cortex in a region extending from the midprincipal sulcus to the central sulcus, and medially to the cingulate sulcus. Units located in anterior regions (dorsolateral prefrontal and anterior cingulate cortex) were typically related to both go and no-go trials (designated type II units) and showed similar ("symmetrical") activity in both kinds of trials; some of them also showed prestimulus ("anticipatory") activity. Such units were present but less common in posterior regions (postarcuate and precentral and posterior cingulate cortex). Units in these posterior regions were active predominantly in go trials (designated type I units). Also found posteriorly were "asymmetrical" type II cells whose activity was greater in go trials and occurred later in the trial, around the behavioral response. The anterior symmetrical and anticipatory type II units in the frontal association cortex were similar to such units described earlier in the brain stem reticular formation and may have similar functions in supporting focused and preparatory attention. On the other hand, asymmetrical type II units in the posterior frontal regions may have a role in the initiation of visually guided motor behavior.

Effects of premotor cortex cooling upon visually initiated hand movements in the monkey

The premotor cortex was temporarily impaired with a local cooling method and effects of the cooling upon visually initiated hand movement and upon cortical field potentials associated with the movement were examined in the monkey. Bilateral cooling of the premotor cortex (the dorsolateral part of presumed area 6) disorganized the well-trained reaction-time movement, in which a lever was lifted within duration of the light stimulus (about 0.5 s) delivered at random time intervals. Accordingly the monkey showed nearly self-paced, random movements with little regard to the light stimulus during the cooling. To obtain a certain number of appropriate reaction-time movements, about twice as many as trials in normal conditions were required during the premotor cooling. Unilateral (contralateral to the moving hand) cooling of the premotor cortex produced similar but weak effects. No appreciable paresis was observed by cooling. Such effects of cooling the premotor cortex faded in successive experimental days of several weeks or even in successive cooling sessions on the same day. This was in contrast with effects of motor cortex cooling which were reported to be almost unfaded on repetition. It is suggested that compensatory actions are quickly elicited in some other structures even on temporal impairment of the premotor cortex and that the actions are gradually accumulated on repetitive cooling with days and weeks.

The relationship of corpus callosum connections to electrical stimulation maps of motor, supplementary motor, and the frontal eye fields in owl monkeys

Microstimulation and anatomical techniques were combined to reveal the organization and interhemispheric connections of motor cortex in owl monkeys. Movements of body parts were elicited with low levels of electrical stimulation delivered with microelectrodes over a large region of precentral cortex. Movements were produced from three physiologically defined cortical regions. The largest region, the primary motor field, M-I, occupied a 4-6-mm strip of cortex immediately rostral to area 3a. M-I represented body movements from tail to mouth in a grossly somatotopic mediolateral cortical sequence. Specific movements were usually represented at more than one location, and often at as many as six or seven separate locations within M-I. Although movements related to adjoining joints typically were elicited from adjacent cortical sites, movements of nonadjacent joints also were produced by stimulation of adjacent sites. Thus, both sites producing wrist movements and sites producing shoulder movements were found next to sites producing digit movements. Movements of digits of the forepaw were evoked at several locations including a location rostral to or within cortex representing the face. Overall, the somatotopic organization did not completely correspond to previous concepts of M-I in that it was neither a single topographic representation, nor two serial or mirror symmetric representations, nor a "nesting about joints" representation. Instead, M-I is more adequately described as a mosaic of regions, each representing movements of a restricted part of the body, with multiple representations of movements that tend to be somatotopically related. A second pattern of representation of body movements, the supplementary motor area (SMA), adjoined the rostromedial border of M-I. SMA represented the body from tail to face in a caudorostral cortical sequence, with the most rostral portion related to eye movements. Movements elicited by near-threshold levels of current were often restricted to a single muscle or joint, as in M-I, and the same movement was sometimes multiply represented. Typically, more intense stimulating currents were required for evoking movements in SMA than in M-I. A third motor region, the frontal eye field (FEF), bordered the representation of eyelids and face in M-I. Eye movements elicited from this cortex consisted of rapid horizontal and downward deviation of gaze into the contralateral visual hemifield.

Set-related neuronal activity in the premotor cortex of rhesus monkeys: effects of changes in motor set

Previous studies have suggested that the premotor cortex plays a role in motor preparation. We have tested this hypothesis in macaque monkeys by examining neuronal activity during an enforced, 1.5-3.0 s delay period between the presentation of an instruction for movement and the onset of that movement. Two targets for movement were available to the monkey, one on the left and one on the right. Illumination of one of the targets served as the instruction for a forelimb movement. It is known that there are cells in the premotor cortex that have directionally specific, sustained activity increases or decreases following such instructions. If the premotor cortex is involved in the preparation for movement in a particular direction, then changing the target from one to the opposite side during the delay period should lead to a pronounced change in sustained neuronal activity. Further, removing the instruction, while still requiring movement to the target, should have little or no sustained effect. Seventy cells showed the predicted activity patterns, thus supporting the view that the premotor cortex plays a role in motor preparation.

Cortical afferents and efferents of monkey postarcuate area: an anatomical and electrophysiological study

A study has been made of the corticocortical efferent and afferent connections of the posterior bank of the arcuate sulcus in the macaque monkey. The distribution of efferent projections to the primary motor cortex (MI) was studied by injecting three different fluorescent retrograde tracers into separate regions of MI. The resultant labeling showed a discrete and topographically organized projection: neurons lying below the inferior limb of the arcuate sulcus project into the MI face area, while neurons located in the posterior bank of the inferior limb of the arcuate sulcus and in the arcuate spur region project into the MI hand area. These findings were confirmed electrophysiologically by demonstrating that postarcuate neurons could only be activated antidromically by stimulation within restricted regions of MI. HRP injections within postarcuate cortex indicated that afferents to this region arise from a number of cortical areas. However, the largest numbers of labeled neurons were found in the posterior parietal cortex (area 7b; PF) and in the secondary somatosensory region (SII). Neurons in both 7b (PF) and SII could be antidromically activated by postarcuate stimulation. It was further shown that stimulation of area 7b (PF) gives rise to short-latency synaptic responses in postarcuate neurons, including some neurons with identified projections to MI. The results are discussed in relation to the possible function of the postarcuate region of the premotor cortex in the sensory guidance of movement.

Motor aspects of cue-related neuronal activity in premotor cortex of the rhesus monkey

Neurons in the premotor cortex of rhesus monkeys were studied under two conditions: (1) visuospatial cues were given to guide the amplitude, direction, and onset time of forearm movements or (2) physically identical visual cues were given when reward was contingent on withholding movement. Neurons with sustained activity following the cues were preferentially active when the cues triggered a movement. Thus, activity of certain neurons in this cortical field is linked to motor set, i.e. intention to make a movement in response to the cue, rather than the visual cue per se.

Neuronal activity in the cortical supplementary motor area related with distal and proximal forelimb movements

Monkeys were trained to perform two different motor acts, one involving muscle activity in distal forelimb muscles and the other in proximal forelimb and shoulder girdle muscles. After confirming spatial and temporal dissociation of muscle activity in the two motor acts, single unit activity in the supplementary motor area (SMA) was recorded. SMA neurons related with the distal and proximal forelimb movements were found to be arranged rostrocaudally with a considerable overlap. In the overlapping region, neurons related with the distal movement were located more deeply.

Cortical projection to hand-arm motor area from post-arcuate area in macaque monkeys: a histological study of retrograde transport of horseradish peroxidase

In four macaque monkeys horseradish peroxidase (HRP) was injected into physiologically defined hand-arm motor area. Ipsilaterally, HRP labeled neurons were found in both upper and lower limbs of the posterior bank of the arcuate sulcus and in an area surrounding the arcuate spur. Contralaterally, labeled neurons were found in the same areas, though less dense in concentration. Labeled neurons were found mostly in layer III of the cortex.