Analysis of patterns of natural activity of neurones in the precentral gyrus of conscious monkeys

Pathway-specific use-dependent dynamics of excitatory synaptic transmission in rat intracortical circuits

Information processing in neuronal networks is determined by the use-dependent dynamics of synaptic transmission. Here we characterize the dynamic properties of excitatory synaptic transmission in two major intracortical pathways that target the output neurons of the neocortex, by recording unitary EPSPs from layer 5 pyramidal neurons evoked in response to action potential trains of increasing complexity in presynaptic layer 2/3 or layer 5 pyramidal neurons. We find that layer 2/3 to layer 5 synaptic transmission is dominated by frequency-dependent depression when generated at fixed frequencies of > 10 Hz. Synaptic depression evolved on a spike-by-spike basis in response to action potential trains that possessed a broad range of interspike intervals, but a low mean frequency (10 Hz). Layer 2/3 to layer 2/3 and layer 2/3 to layer 5 synapses were incapable of sustained release during prolonged, complex trains of presynaptic action potential firing (mean frequency, 48 Hz). By contrast, layer 5 to layer 5 synapses operated effectively across a wide range of frequencies, exhibiting increased efficacy at frequencies > 10 Hz. Furthermore, layer 5 to layer 5 synapses sustained release throughout the duration of prolonged, complex spike trains. The use-dependent properties of synaptic transmission could be modulated by pharmacologically changing the probability of release and by induction of long-term depression. The dynamic properties of intracortical excitatory synapses are therefore pathway-specific. We suggest that the synaptic output of layer 5 pyramidal neurons is ideally suited to control the neocortical network across a wide range of frequencies and for sustained periods of time, a behaviour that helps to explain the pivotal role played by layer 5 neurons in the genesis of periods of network 'up' states and epileptiform activity in the neocortex.

The influence of changes in discharge frequency of corticospinal neurones on hand muscles in the monkey

1. The possibility that the discharge pattern of monkey corticomotoneuronal cells influences the degree to which they facilitate their target hand muscles was tested by compiling spike-triggered averages of EMG recorded from these muscles. 2. Records were made from area 4 corticomotoneuronal cells in three conscious macaque monkeys while they performed a precision grip between index finger and thumb. Simultaneous EMG recordings were made from up to six different intrinsic hand muscles. Twenty cells which produced clear post-spike facilitation of one or more muscles were selected for further analysis. 3. Spikes recorded from these cells were grouped according to the occurrence of a previous spike in the periods 0-10 ms, 10-20 ms, and so on up to 60-70 ms before the trigger spike. The post-spike period in which no additional spikes were allowed to fall was kept at either 12.5 or 25 ms. 4. Spikes selected in this way produced a transient facilitation of their target muscle EMG activity. The peak amplitude of this facilitation was normalized as a percentage of modulation of the background EMG level. The background level was determined from a period in the average to which the cell could not have contributed, because of the post-trigger spike interval. We verified that the percentage of modulation was not influenced by the overall level of EMG activity, since, for a given interval, the modulation was the same whether the relevant spikes were selected during periods of high- or low-level EMG activity. 5. The relative amplitude of the post-spike facilitation (i.e. the percentage of modulation) showed marked variation with interspike interval. A full analysis was completed for seventeen neurones. Spikes with the shortest intervals (less than 10 ms) usually produced the strongest effects, and evidence is presented that this was due to temporal summation and facilitation at the corticomotoneuronal synapse. Mid-range intervals (10-40 ms) were generally far less effective, although they constituted the highest proportion of cell activity. 6. A striking finding was the strong facilitation generated by the longer interspike intervals (40-70 ms). Although the absolute size of this post-spike effect was much smaller than that of the shortest intervals, its percentage of modulation was similar. It is suggested that this enhanced facilitation results from a combination of lower frequency discharge among the active motoneurones, and increased synchrony in the corticomotoneuronal input to them. 7. All of the above results were confirmed by examining cross-correlations between single corticomotoneuronal cells and single motor units in their target muscle.(ABSTRACT TRUNCATED AT 400 WORDS)

The monkey globus pallidus: neuronal discharge properties in relation to movement

1. Recordings were made of the natural discharges of 388 pallidal neurones in awake, free-to-move monkeys in order to describe the discharge properties of such neurones in relation to normal movement performance. 2. Of the 388 neurones, 156 discharged only in association with one direction of movement of the forelimb about a specific joint. If that movement was not taking place the neurone would not discharge. 3. All joints and directions of movement for the upper limb were represented by clusters of cells within the pallidal population. 4. Twenty-nine per cent of neurones co-varied with movement of both contralateral and ipsilateral limb for the same direction of movement about a given joint; distal movements were represented with similar frequency to proximal movements in this group. 5. Afferent information provided by natural stimulation of peripheral receptors did not directly influence either the discharging or non-discharging pallidal neurones. 6. Movement related neurones were regionally organized and were found in the posterior part of the pallidum.

Preparation for movement in the cat. II. Unit activity in the basal ganglia and thalamus

Of the movement-related units in the globus pallidus and entopeduncular nucleus 30--40% show early (more than 500 msec) onsets of their movement-related activity preceding self-initiated 'elbow'-flexing movements in cats. The medial pericruciate motor cortex and the VL-VA thalamic nuclei display similar distributions of onset times, in contrast to the lateral cruciate cortex where 97% of neurons change their activity much later. The possible significance of the early activity is discussed in relation to the notion of 'response set'. It is suggested that these data support the concept that the basal ganglia participate in the enabling and sequencing of movements rather than in directly causing them to occur.

Preparation for movement in the cat. I. Unit activity in the cerebral cortex

Single unit activity was recorded from the 'motor' cortex of cats during performance of a forelimb flexion movement. Two classes of cortical neurons were defined with respect to the onset of electrographic activity associated with this movement. 'Early' unit, first showing changes in firing rates more than 0.5 sec prior to the movement, were found almost exclusively in the medial precruciate cortex. The lateral cortex appears to be made up almost uniformly of 'late' unit, that is, neurons whose activities are more closely related with the actual movement. The medial cortex, on the other hand, contains both 'early' and 'late' units and thus may have the additional function of participating in the neural system which is the substrate for response set.

Afferent input to movement-related precentral neurones in conscious monkeys

Monkeys were trained to perform a stereotyped movement task, and to accept passive manipulation and natural stimulation of the limbs while remaining relaxed and quiet. All training, both of movement and for relaxation, was with food rewards. The effects of natural stimuli on 257 precentral neurones showing consistent modulations in discharge frequency during the performance of the movement task were investigated. Most precentral neurones had small, stable input zones located on the contralateral arm: 197 facilitatory and 17 inhibitory responses were obtained, while the remaining 43 cells were unaffected by the natural stimuli used. The most common natural stimulus capable of influencing precentral neurones was joint movement: 152 cells responded to joint movement, including 98 which only responded to movement at a single joint. Joint movement rather than joint position was the effective stimulus and none of these cells was influenced by palpation of muscles acting at the joint. The next most common natural stimulus capable of influencing precentral neurones was muscle palpation: 35 cells responded to a tap applied to a localized portion of the muscle belly, including 26 cells which also responded to movement of the joint at which the muscle acted. The direction of joint movement which influenced the cell was usually such as to stretch the muscle containing the receptors for the effective afferent input set up by tapping. The natural stimulus which influenced the smallest number of precentral neurones was tactile stimulation of the skin: 27 cells had cutaneous receptive fields, most of which were small ( < 5 cm 2 ) and confined to the hand. Included in the total sample were 51 pyramidal tract neurones. The behaviour of these was found to be similar to the unidentified neurones examined in the same animals with respect to their afferent input. However, there was a tendency for pyramidal tract neurones to be in receipt of a more convergent input than unidentified neurones in their vicinity. The majority of neurones recorded in close proximity to one another (within 500 μm or less) usually received their afferent input from the same peripheral region, but a significant proportion of such cells received inputs from different and remote peripheral zones. Hence the afferent input to the precentral motor cortex is not organized to provide independent and spatially segregated projections from particular peripheral sites only to limited and localized radial aggregations of neurones.

Relationship between the activity of precentral neurones during active and passive movements in conscious monkeys

Recordings have been made from neurones in the precentral cortex of conscious monkeys carrying out a stereotyped movement task for food rewards. The activity of each movement related neurone was investigated while the monkey performed a wide variety of movements in order to collect the food reward placed in a number of different positions. Of 362 task-related neurones, 176 neurones showed definite modulation of their discharge frequency which could be related to the performance of specific voluntary movements about one joint, or at a number of associated digit joints. These neurones could therefore be classified to have dis­charges associated with particular voluntary movements of shoulder, elbow, wrist, hand or fingers. The responses of the same neurones were then examined with natural stimulation of the skin, joints and muscles of the arm while the animal sat relaxed as it had been trained to do. 149 of the 176 neurones responded to afferent input from particular peripheral territories. 130 of the 176 neurones had afferent inputs from zones which were anatomically closely related to the joint involved in the specific active movement with which the neurone’s natural discharges were clearly associated. An analysis of the anatomical distribution in the motor cortex of the cells exhibiting given input and output associations revealed that not all the members of a restricted local population of neurones shared the same peripheral territory. When they did, the direction of movement of that territory which was associated with discharge of the cell could be the same or opposite under active and passive conditions. Moreover, not all the neurones with a particular association with active movement of a joint and an input from that same region were aggregated in the one restricted local area of the motor cortex; some members of the group sharing these input/output characteristics could be situated up to 5 mm from the main aggregation. An attempt is made to distinguish between the properties of receiving neurones and output neurones of the precentral gyrus.

Timing of the responses in the motor cortex of monkeys to an unexpected disturbance of finger position

Monkeys were trained with food rewards to hold the wrist and fingers of their right hand in a flexed posture and maintain force with the finger tips against an isometric lever for a number of seconds. Once the animal had learned to produce a reliable performance of the task an assembly was attached to the skull through which microelectrodes could be introduced into the precentral cortex to record the activity of single neurones. Neurones whose activity was correlated with the force of finger flexion were studied; some of these could be identified as pyramidal tract neurones by their response to electrical stimulation in the medullary pyramids. While the monkey was flexing against it, the lever was sometimes suddenly released so that the fingers flexed without resistance. This unexpected disturbance was often followed by a change in the discharge of precentral neurones, although the monkey had not been trained to respond to the release in any particular way. On release of the lever the discharge of a given cortical neurone might either increase or decrease, and the direction of this change could not be predicted from the behaviour of the neurone during the isometric task. The most common response was an increase in cortical cell firing in neurones whose natural discharge was associated with the active development of force. The discharge of pyramidal tract neurones changed 25-50 msec after the sudden unexpected peripheral disturbance. Earlier changes were seen in some other neurones situated within the precentral gyrus and in the anterior bank of the central sulcus.

Relationship of neuronal discharges in the precentral gyrus of monkeys to the performance of arm movements

Recordings have been made from 162 pyramidal tract neurones which discharged bursts of nerve impulses in characteristic temporal association with performances of a stereotyped motor task by monkeys. Clinical evaluation of the relationship between discharges of the neurones and free movement led to the view that each cell's firing was associated with a characteristic aspect of movement performance and the contraction of a particular group of muscles. Quantitative evaluation of these relationships led to the conclusion that coding of the recruitment of motor units to the movement task could have been conferred by the number of pyramidal tract neurones discharging to those motoneurone targets. A ramp of "recruitment" of pyramidal tract neurones preceded the development of a ramp of force by about 100 msec. This general conclusion was supported by the observations made in a single animal in which orderly discharge of precentral neurones in relation to a sterotyped movement performance was clearly evident.

The effect of a preceding stimulus on temporal facilitation at corticomotoneuronal synapses

1. Intracellular recordings were made of minimal corticomotoneuronal e.p.s.p.s in lumbar motoneurones of anaesthetized monkeys. For intervals of 2 msec and greater between paired cortical shocks, the average time course of facilitation of the second e.p.s.p. with respect to the first could be fitted closely by a negative exponential with a time constant of 10 msec.2. In the same motoneurones, ;triplets' of corticomotoneuronal e.p.s.p.s were generated by delivering three identical stimuli to the motor cortex. Considering the triplet as a conditioning e.p.s.p. followed by a test pair, the facilitation of the third e.p.s.p. with respect to the second was measured for various combinations of test and conditioning intervals. In each case the amplitude of the third e.p.s.p. was also compared with that of the first (conditioning) e.p.s.p.3. The effect of a brief conditioning interval was to reduce considerably the facilitation of the third e.p.s.p. with respect to the second at all test intervals from 2 to 50 msec. Combinations of brief conditioning intervals (e.g. 2 or 5 msec) and long test intervals (e.g. 20 or 50 msec) caused the third e.p.s.p. to be smaller than the second. As the conditioning interval lengthened, facilitation in the test pair increased towards the unconditioned values at all test intervals.4. Facilitation of the third e.p.s.p. with respect to the first could be described approximately as the linear addition of two facilitation components, one due to the conditioning input and one due to the first stimulus of the test pair. Each component followed the same negative exponential time course as found for an isolated pair of e.p.s.p.s and each of the first two inputs contributed to the facilitation of the third e.p.s.p. as if the other of these two inputs had not occurred.