Recently, several studies showed that neurons in ventral premotor cortex F5 and inferior parietal areas AIP and PFG respond to the observation of biological movements of both humans and monkeys. These studies analyzed neuronal responses to the observation of videos in different conditions.
One example is shown in Fig. 4b, where a neuron that responded to the basic condition of MGI discharged during both video epochs but decreased its response during obscuration of the first one.
Premotor cortex
The premotor cortex is intimately involved in selecting a movement from the repertoire of possible movements. It is sensitive to the inferred intention of a movement, and this sensitivity distinguishes it from the primary motor cortex, which is sensitive to the actual motor performance of a movement. It is also important for predicting movement errors. For example, if the monkey observes another person move incorrectly, the premotor cortex is more active than when the monkey observes itself moving correctly. This suggests that the premotor cortex is sensitive to behavioral context, as well as internal and external cues.
Studies of premotor cortical activity have shown that monkeys prepare to make a movement in anticipation of an instruction, but only if the movement is expected to be correct. In this preparatory stage, the lateral premotor cortex (PMd) and medial premotor cortex (PMc) are both activated. The activation of PMc is based on the anticipated direction and speed of a movement, while that of PMd is based on the behavioral context of the intended movement. The behavioral context of a movement is judged by comparing the intended outcome to the target, and deciding whether the target is achievable in the available time.
In one study, researchers measured neural activity during an enforced 1.5-3.0 s delay between the presentation of an instruction for a movement and its execution. They found that PMd neurons signaled the decision to make a movement by increasing their firing rate in preparation for the motion. PMc neurons, on the other hand, did not change their firing rates in the same way. The researchers speculate that these changes indicate that the decision is made by comparing the kinematics of the intended movement with its behavioral context.
In a related experiment, they recorded the activity of individual PMv neurons while 2 monkeys performed a visual discrimination task. The monkeys had to decide whether the orientation of a current stimulus was toward the right or left of a memorized trace of a previously shown stimulus. The results showed that the PMv neurons encoded the monkeys’ decisions, regardless of whether the stimuli were recently shown or retrieved from long-term memory.
Primary motor cortex
The primary motor cortex contains a map of the different body parts that can be moved. Neurons in this region are arranged in an area called the motor homunculus, which is shaped like a little person’s head and feet and includes regions that represent the leg, the arm, the hand, and other body parts. There are also patches of neurons that correspond to particular muscles. Each patch receives proprioceptive input from a specific muscle or group of synergistic muscles and sends its output to the appropriate group of muscles via a multisynaptic pathway in the brainstem and spinal cord.
The cytoarchitectural organization of the primary motor cortex is generally similar across catarrhine species, including Old World monkeys, great apes, and humans. However, some differences are apparent. For example, in a region of the motor cortex that serves the axial and proximal muscles, a greater percentage of neurons are enriched for NPNFP and PV than in Old World monkeys (Figure 17). These changes might contribute to the more dexterous control of orofacial muscles that is exhibited by great apes and humans.
Some neurons in the primary motor cortex encode the direction of a movement relative to an external frame of reference. In one study, a monkey was trained to prepare a certain movement to the right or to the left depending on an instruction stimulus, and to delay the movement until the “Move” signal is given. Neurons in the premotor cortex encoding preparation for movements fired more strongly during movements that are up and to the right than for other directions.
There is also evidence that the premotor cortex contains neurons that are selective for particular patterns of muscle activation associated with movements. In one study, Kakei and Hoffman recorded from neurons in the premotor cortex of a monkey as it moved its arms to produce various stereotyped positions. They found that some neurons fired more frequently when the monkey drew its arms to the side or toward the mouth than when it moved them closer to its face.
Other neurons fired more frequently when the monkey drew the arms to the front of its body, as though the monkey were bringing food to its mouth. These neurons were referred to as mirror neurons, and it is thought that they are involved in action observation and imitation.
Postmotor cortex
Postmotor cortex neurons can be activated by a movement in the monkey’s body or by observing another’s movement. Neurons in this area are also active during the few seconds of delay or preparation before the monkey performs a movement that is instructed by a sensory cue. They are less active during the actual movement itself. This suggests that the postmotor cortex is involved in preparing to make a movement and selecting a motor plan to execute it.
This information is incorporated into a neural representation of the movement and stored in rostral PMv. Neurons in this region are then able to compare the current stimulus (S2) with a trace of S1 that has been memorized during the delay period, decide on the direction of the difference between the 2 (sign(S2-S1)), and communicate this decision by making an eye movement toward one of several targets. The emergence of this behavioural decision can be tracked by recording the activity of neurons in rostral PMv using a CD task. These tasks involve a monkey being presented with two different objects. The monkey has to decide which target it should move toward. The researchers recorded the activity of neurons in the premotor cortex and primary motor cortex of a monkey while they were performing these tasks. The neurons in the premotor cortex were more active than those in the primary motor cortex. This suggested that the premotor cortex is more important for planning or preparing to make a movement than it is for executing the movement itself.
Recordings of single unit activity from the premotor cortex were made with tungsten microelectrodes in the posterior bank of the ventral arm of the sulcus arcuatus in 4 hemispheres of 2 monkeys. The recordings were performed during the CD and FDIR tasks described above. The authors found that the neurons in the premotor cortex were active in response to a visual instruction cue and remained active during the few seconds of delay or preparation for a movement. The authors concluded that these neurons were part of a frontal hierarchy that stores motor memory, and they suggest that this is phyletic memory, a fund of knowledge about fundamental movements of the species, that is, those movements required to survive.
Visual cortex
Visual cortical areas are characterized by a distinct cytoarchitecture and layer thickness. These characteristics allow cytological identification of neurons within the cortical sheet and are the basis for Korbinian Brodmann’s first maps of visual cortex (Brodmann, 1909). Moreover, they exhibit an early areal distribution independent of thalamocortical inputs. Furthermore, a number of cellular markers have been used to demarcate these modules. One such marker is the antibody clone Cat-301, which detects chondroitin sulfate proteoglycans. These molecules are present in the extracellular matrix and are enriched around synapses. They are also present in the proximal dendrites and cell body of neurons. The expression of Cat-301 is restricted to layers 3 and 5 in most visual association cortex areas, allowing for the delimitation of visual cortical areal boundaries.
Another cellular marker is the calcium-binding protein Cadherin-D28k, or Cb. This protein is expressed by GABAergic interneurons in the neocortex. Its early expression during corticogenesis, coupled with a later upregulation during arealisation in layers 4-6, makes it ideal for identifying visual areas in non-primate species. A recent study showed that the Cb expression profile in ferret visual cortex reflects a dynamic area-specific and layer-specific pattern, with cadherin20 and protocadherin10 being selectively expressed in V1, while cadherin8 and -11 are restricted to V2 (Krishna et al., 2009).
fMRI is a powerful technique that allows researchers to quickly identify functionally specialized brain regions. However, its physiological basis remains largely unknown. The underlying signaling mechanism is not well understood, and the relationship between neuronal activity and BOLD signal strength is not fully understood. In order to improve the sensitivity of fMRI, it is necessary to combine it with single-cell recording techniques.
To overcome motion artifact in fMRI studies, we developed an apparatus to hold the monkey’s head in a fixed position during the experiment. It consisted of a Plexiglas cylinder that was attached to a head post implanted on the monkey’s skull. The headpost was connected to a radiofrequency coil, which was scaled down from the birdcage coil used in human fMRI. A slit was cut into the plastic of the RF coil, which allowed for the headpost to be lowered through.