Action selection and refinement in subcortical loops through basal ganglia and cerebellum

Subcortical loops through the basal ganglia and the cerebellum form computationally powerful distributed processing modules (DPMs). This paper relates the computational features of a DPM's loop through the basal ganglia to experimental results for two kinds of natural action selection. First, functional imaging during a serial order recall task was used to study human brain activity during the selection of sequential actions from working memory. Second, microelectrode recordings from monkeys trained in a step-tracking task were used to study the natural selection of corrective submovements. Our DPM-based model assisted in the interpretation of puzzling data from both of these experiments. We come to posit that the many loops through the basal ganglia each regulate the embodiment of pattern formation in a given area of cerebral cortex. This operation serves to instantiate different kinds of action (or thought) mediated by different areas of cerebral cortex. We then use our findings to formulate a model of the aetiology of schizophrenia.

Movement-related discharge in the cerebellar nuclei persists after local injections of GABA(A) antagonists

Limb movement-related neurons in the cerebellar nuclei (CN) typically produce bursts of discharge in association with movement. Consequently, given the inhibitory nature of the Purkinje cell (PC) projection to CN, it is puzzling that only a minority of movement-related PCs pause; the majority burst. Some of the movement-related CN activity may be the result of excitation from collaterals of mossy and climbing fiber projections to the cerebellar cortex. The only other input to CN is diffuse and neuromodulatory, from locus ceruleus and raphe nuclei. To investigate the role of the excitatory mossy fiber input, single units in CN were recorded in macaque monkeys during the performance of reaching and manipulation tasks, before and after blocking the PC input with local microinjections of GABA(A) antagonists (bicuculline or SR95531). After these injections, the movement-related modulation of CN discharge was greater and began earlier, compared with the modulation in the preinjection group of neurons. These observations indicate that an important excitatory drive is provided by extracerebellar inputs to CN, most likely from collaterals of mossy fibers. PCs may serve primarily to regulate this activity, by either pausing or bursting as necessary.

Features of motor performance that drive adaptation in rapid hand movements

In order to explore how subjects correct for errors in movement and adapt their motor programs, we studied rapid hand movements. Subjects grasped a grooved knob and made brisk turning movements to various targets, similar to tuning a radio dial. A motor attached to the knob shaft was configured to apply a destabilizing negative viscous perturbation. Following a period of practice with no perturbations, the negative viscosity was engaged, which caused a large change in overall kinematics: the peak velocity increased, the movement amplitude was too large, and discrete corrective submovements were generated to bring the pointer back onto the target. After about an hour and nearly 1000 trials, subjects learned to move accurately in the new dynamic environment, returning their overall kinematics near to previous levels. Measures of performance included the endpoint error of the primary movement (the initial movement segment), the frequency and amplitude of corrective submovements, task success rate, mean squared jerk, and deviation from a "normal" angular velocity temporal profile. Both the amplitude and frequency of corrective submovements decreased progressively during adaptation as the subjects made fewer target overshoot errors. These results are consistent with motor learning schemes in which adaptation of the motor controller is driven by an attempt to reduce the endpoint error of the primary movement. While there have been many theories regarding what is being optimized in motor control, in general, biologically plausible mechanisms for implementing these schemes have not been described. A biologically plausible optimization criterion is the minimization of the occurrence and amplitude of corrective submovements, since the latter have been proposed as realistic climbing fiber training signals for adaptive changes in the cerebellum. We postulate that the other criteria that have been proposed are instead secondary to an increased accuracy of the primary movement and a corresponding decrease in the occurrence and amplitude of corrective submovements.

The role of the cerebellum in modulating voluntary limb movement commands

We recorded the activity of cerebellar Purkinje cells (PCs), primary motor cortical (M1) neurons, and limb EMG signals while monkeys executed a sequential reaching and button pressing task. PC simple spike discharge generally correlated well with the activity of one or more forelimb muscles. Surprisingly, given the inhibitory projection of PCs, only about one quarter of the correlations were negative. The largest group of neurons burst during movement and were positively correlated with EMG signals, while another significant group burst and were negatively correlated. Among the PCs that paused during movement most were negatively correlated with EMG. The strength of these various correlations was somewhat weaker, on average, than equivalent correlations between M1 neurons and EMG signals. On the other hand, there were no significant differences in the timing of the onset of movement related discharge among these groups of PCs, or between the PCs and M1 neurons. PC discharge was modulated largely in phase, or directly out of phase, with muscle activity. The nearly synchronous activation of PCs and muscles yielded positive correlations, despite the fact that the synaptic effect of the PC discharge is inhibitory. The apparent function of this inhibition is to restrain activity in the limb premotor network, shaping it into a spatiotemporal pattern that is appropriate for controlling the many muscles that participate in this task. The observed timing suggests that the cerebellar cortex learns to modulate PC discharge predictively. Through the cerebellar nucleus, this PC signal is combined with an underlying cerebral cortical signal. In this manner the cerebellum refines the descending command as compared with the relatively crude version generated when the cerebellum is damaged.

The use of overlapping submovements in the control of rapid hand movements

Rapid targeted movements are subject to special control considerations, since there may be inadequate time available for either visual or somatosensory feedback to be effective. In our experiments, subjects rapidly rotated a knob to align a pointer to one of several targets. We recognized three different types of movement segments: the primary movement, and two types of submovement, which frequently followed. The submovements were initiated either before or after the end of the primary movement. The former, or "overlapping" type of submovement altered the kinematics of the overall movement and was consequently difficult to detect. We used a direct, objective test of movement regularity to detect overlapping submovements, namely, examining the number of jerk and snap zero crossings during the second half of a movement. Any overlapping submovements were parsed from the overall movement by subtracting the velocity profile of the primary movement. The velocity profiles of the extracted submovements had near-symmetric bell shapes, similar to the shapes of both pure primary movements and nonoverlapping submovements. This suggests that the same neural control mechanisms may be responsible for producing all three types of movement segments. Overlapping submovements corrected for errors in the amplitude of the primary movement. Furthermore, they may account for the previously observed, speed-dependent asymmetry of the velocity profile. We used a nonlinear model of the musculoskeletal system to explain most of the kinematic features of these rapid hand movements, including how discrete submovements are superimposed on a primary movement. Finally, we present a plausible scheme for how the central nervous system may generate the commands to control these rapid hand movements.

Cerebellar input to magnocellular neurons in the red nucleus of the mouse: synaptic analysis in horizontal brain slices incorporating cerebello-rubral pathways

We studied the synaptic input from the nucleus interpositus of the cerebellum to the magnocellular division of the red nucleus (RNm) in the mouse using combined electrophysiological and neuroanatomical methods. Whole-cell patch-clamp recordings were made from brain slices (125-150 microm) cut in a horizontal plane oriented to pass through both red nucleus and nucleus interpositus. Large cells that were visually selected and patched were injected with Lucifer Yellow and identified as RNm neurons. Using anterograde tracing from nucleus interpositus in vitro, we examined the course of interposito-rubral axons which are dispersed in the superior cerebellar peduncle. In vitro monosynaptic responses in RNm were elicited by an electrode array placed contralaterally in this pathway but near the midline. Mixed excitatory post-synaptic potentials (EPSPs)/inhibitory post-synaptic potentials (IPSPs) were observed in 48 RNm neurons. Excitatory components of the evoked potentials were studied after blocking inhibitory components with picrotoxin (100 microM) and strychnine (5 microM). All RNm neurons examined continued to show monosynaptic EPSPs after non-N-methyl-D-aspartate (NMDA) glutamate receptor components were blocked with 10 microM 6,7-dinitroquinoxaline-2,3-dione or 5 microM 2,3-dihydro-6-nitro-7-sulfamoyl-benzo(f)-quinoxaline (NBQX; n=12). The residual potentials were identified as NMDA receptor components since they (i) were blocked by the addition of the NMDA receptor antagonist, D,L-2-amino-5-phosphonovaleric acid (APV), (ii) were voltage-dependent, and (iii) were enhanced by Mg(2+) removal. Inhibitory components of the evoked potentials were studied after blocking excitatory components with NBQX and APV. Under these conditions, all RNm neurons studied continued to show IPSPs. Blockade of GABA(A) receptors reduced but did not eliminate the IPSPs. These were eliminated when GABA(A) receptor blockade was combined with strychnine to eliminate glycine components of the IPSPs. Thus, IPSPs evoked by midline stimulation of the superior cerebellar peduncle, while blocking alpha-amino-3-hydroxy-5-methylisoxazole-4-propionic acid (AMPA) and NMDA receptors, raise the possibility of direct inhibitory inputs to RNm from the cerebellum. In summary we propose that the special properties of the NMDA receptor components are considered important for the generation of RNm motor commands: their slow time course will contribute a steady driving force for sustained discharge and their voltage dependency will facilitate abrupt transitions from a resting state of quiescence to an active state of intense motor command generation.

Context dependency in the globus pallidus internal segment during targeted arm movements

Extracellular discharges from single neurons in the internal segment of the globus pallidus (GPi) were recorded and analyzed for rate changes associated with visually guided forearm rotations to four different targets. We sought to examine how GPi neurons contribute to movement preparation and execution. Unit discharge from 108 GPi neurons recorded in 35 electrode penetrations was aligned to the time of various behavioral events, including the onset of cued and return movements. In total, 39 of 108 GPi neurons (36%) were task-modulated, demonstrating statistically significant changes in discharge rate at various times between the presentation of visual cues and movement generation. Most often, strong modulation in discharge rate occurred selectively during either the cued (n = 32) or return (n = 2) phases of the task, although a few neurons (n = 5) were well-modulated during both movement phases. Of the 34 neurons that were modulated exclusively during cued or return movements, 50% (n = 17) were modulated similarly in association with movements to any target. The remaining 17 neurons exhibited considerable diversity in their discharge properties associated with movements to each target. Cued phases of behavior were always rewarded if executed correctly, whereas return phases were never rewarded. Overall, these data reveal that many GPi neurons discharged in a context-dependent manner, being modulated during cued, rewarded movements, but not during similar self-paced, unrewarded movements. When considered in the light of other observations, the context-dependence we have observed seems likely to be influenced by the animal's expectation of reward.

Functional connectivity between cerebellum and primary motor cortex in the awake monkey

Simultaneous single neuron and local field potential (LFP) recordings were made in arm-related areas of the cerebellar nuclei (CN) and primary motor cortex (M1) of two monkeys during a reaching and button pressing task. Microstimulation of focal sites in CN caused short latency (median = 3.0 ms) increases in discharge in 25% of 210 M1 neurons. Suppressive effects were less common (13%) and observed at longer latencies (median = 9.9 ms). Stimulation in CN also caused reciprocal facilitation and suppression in averages of antagonist muscle electromyograms (EMGs). The latency of these effects was approximately 8-11 ms. In contrast to the selectivity of unit and EMG effects, stimulation-evoked changes in LFP occurred over a broad range of sites. There were no significant short-latency effects detected in cross-correlation histograms between single neurons in CN and M1. However, CN spike-triggered averages of M1 LFPs were observed in a few cases (10% of 126 cases). In one-half of these, there were effects both before and after the CN spikes, which may reflect causal effects from M1 to CN, as well as from CN to M1. Overall, these results demonstrate a spatially specific, short latency, primarily excitatory pathway from CN to M1. The relatively rare effects at the single neuron level may have resulted from the difficulty in achieving optimal alignment between cerebellar and cerebral sites because of the specificity of these connections.

Kinematic properties of rapid hand movements in a knob turning task

In order to understand how the central nervous system controls the kinematics of rapid finger and hand movements, we studied the motions of subjects turning a knob to light-emitting diode targets, similar to tuning a radio dial. On many trials, subjects turned the knob with a single, smooth, and regular motion as revealed by the angular position and velocity trajectories, but on others, subjects produced irregularities in the kinematics. Like many past studies, we interpreted these irregularities as discrete corrective submovements. Unlike other studies, we used a direct, objective algorithm to identify overlapping submovements, detecting appreciable inflections in the acceleration traces by examining zero crossings in their derivatives, jerk and snap. The movements without overlapping submovements on average had a near symmetric, bell-shaped velocity profile that was independent of speed, and which matched the theoretical minimum jerk velocity very closely. We proposed three plausible mechanisms for altering the shape of movement kinematics, and implemented a mass-spring model with nonlinear damping to explore the possibilities. Although there was relatively little variability in the shape and symmetry of movements across trials, there was a fair amount of variability in their amplitude. We show that subjects attempted to eliminate the need for corrective submovements by making more accurate primary movements with practice, but that the variability inherent in rapid movements dictated the need for corrective submovements. Subjects used corrective submovements to improve final endpoint accuracy while reducing endpoint variability, resulting in higher task success rates.

Emergence of symmetric, modular, and reciprocal connections in recurrent networks with Hebbian learning

While learning and development are well characterized in feedforward networks, these features are more difficult to analyze in recurrent networks due to the increased complexity of dual dynamics - the rapid dynamics arising from activation states and the slow dynamics arising from learning or developmental plasticity. We present analytical and numerical results that consider dual dynamics in a recurrent network undergoing Hebbian learning with either constant weight decay or weight normalization. Starting from initially random connections, the recurrent network develops symmetric or near-symmetric connections through Hebbian learning. Reciprocity and modularity arise naturally through correlations in the activation states. Additionally, weight normalization may be better than constant weight decay for the development of multiple attractor states that allow a diverse representation of the inputs. These results suggest a natural mechanism by which synaptic plasticity in recurrent networks such as cortical and brainstem premotor circuits could enhance neural computation and the generation of motor programs.

A cerebellar model of timing and prediction in the control of reaching

A simplified model of the cerebellum was developed to explore its potential for adaptive, predictive control based on delayed feedback information. An abstract representation of a single Purkinje cell with multistable properties was interfaced, using a formalized premotor network, with a simulated single degree-of-freedom limb. The limb actuator was a nonlinear spring-mass system based on the nonlinear velocity dependence of the stretch reflex. By including realistic mossy fiber signals, as well as realistic conduction delays in afferent and efferent pathways, the model allowed the investigation of timing and predictive processes relevant to cerebellar involvement in the control of movement. The model regulates movement by learning to react in an anticipatory fashion to sensory feedback. Learning depends on training information generated from corrective movements and uses a temporally asymmetric form of plasticity for the parallel fiber synapses on Purkinje cells.

Relationship between modulation of the cerebellorubrospinal system in the in vitro turtle brain and changes in motor behavior in rats: effects of novel sigma ligands

Saturation and competition binding studies showed that the turtle brain contains sigma sites labeled by both [3H]di-o-tolylguanidine (DTG) and [3H](+)-pentazocine. There was a significant correlation between the IC50 values of sigma ligands for [3H]DTG sites in the turtle vs. rat brain, suggesting that the sites are comparable in the two species. In contrast, [3H](+)-pentazocine, which primarily labels sigma1 sites in the rodent brain, labels a heterogeneity of sites in the turtle brain. In extracellular recordings from the in vitro turtle brainstem, some sigma ligands enhanced the burst responses of red nucleus (RN) neurons (DTG, haloperidol, BD1031, BD1052, BD1069) while other sigma ligands decreased the burst responses (BD1047, BD1063). Control compounds (turtle Ringer vehicle control, opiate antagonist naloxone, atypical neuroleptic sulpiride) had no significant effects on the RN burst responses recorded from the in vitro turtle brain. The ED50s of the ligands for altering the burst responses in RN neurons from the turtle brain were correlated with their IC50s for turtle brain sites labeled with [3H]DTG, but not [3H](+)-pentazocine; this pattern is identical to that previously reported in rats, where there is a correlation between the potencies of sigma ligands for producing dystonic postures after microinjection into the rat RN and their binding to rat brain sites labeled with [3H]DTG, but not [3H](+)-pentazocine. When the novel sigma ligands were microinjected into the rat RN, dystonic postures were produced by ligands that increased the burst duration of RN neurons in the turtle brain. Novel sigma ligands that reduced the burst responses in the in vitro turtle brain have previously been reported to have no effects on their own when microinjected into the rat RN, but to block the dystonic postures produced by other sigma ligands. Taken together, the data suggest that the opposite effects of the novel ligands in the turtle electrophysiological studies represent the actions of agonists vs. antagonists, and that the directionality of the effects has predictive value for the expected motor effects of the drugs.

Model of cortical-basal ganglionic processing: encoding the serial order of sensory events

Several lines of evidence suggest that the prefrontal (PF) cortex and basal ganglia are important in cognitive aspects of serial order in behavior. We present a modular neural network model of these areas that encodes the serial order of events into spatial patterns of PF activity. The model is based on the topographically specific circuits linking the PF with the basal ganglia. Each module traces a pathway from the PF, through the basal ganglia and thalamus, and back to the PF. The complete model consists of an array of modules interacting through recurrent corticostriatal projections and collateral inhibition between striatal spiny units. The model's architecture positions spiny units for the classification of cortical contexts and events and provides bistable cortical-thalamic loops for sustaining a representation of these contextual events in working memory activations. The model was tested with a simulated version of a delayed-sequencing task. In single-unit studies, the task begins with the presentation of a sequence of target lights. After a short delay, the monkey must touch the targets in the order in which they were presented. When instantiated with randomly distributed corticostriatal weights, the model produces different patterns of PF activation in response to different target sequences. These patterns represent an unambiguous and spatially distributed encoding of the sequence. Parameter studies of these random networks were used to compare the computational consequences of collateral and feed-forward inhibition within the striatum. In addition, we studied the receptive fields of 20,640 model units and uncovered an interesting set of cue-, rank- and sequence-related responses that qualitatively resemble responses reported in single unit studies of the PF. The majority of units respond to more than one sequence of stimuli. A method for analyzing serial receptive fields is presented and utilized for comparing the model units to single-unit data.

Organization of ascending pathways to the forelimb area of the dorsal accessory olive in the cat

The purpose of these experiments was to define the topography of cuneate and spinal projections to the forelimb representation in the rostral dorsal accessory olive (rDAO). We were interested in determining whether the spinal and cuneate inputs constitute a homogeneous afferent source, and whether there is evidence that they serve different functional roles. We were also interested in determining whether the somatotopy of rDAO is the result of a point-to-point projection from its afferent sources, or whether the projection suggests a reorganization of afferents at the olive. Single unit recording was used to identify specific regions of rDAO, and the topography of inputs to the identified regions was determined by using wheat germ agglutinin-horseradish peroxidase (WGA-HRP) as a tracer. The results from retrograde tracing were confirmed by using WGA-HRP as an anterograde tracer from input sources. The cuneate and spinal neurons providing input to rDAO constitute two distinct neural populations. One consists of cells in the caudal cuneate nucleus and lamina VI of the rostral two cervical segments, the other consists of cells in the rostral cuneate nucleus. The cells in the caudal cuneate nucleus and the rostral cervical segments are large, multipolar neurons that form a single column of rDAO input cells. The column of cells projects to the contralateral rDAO in a topographic fashion with rostral regions of the column projecting to rostral rDAO, which contains cells that respond to somatosensory stimulation of the contralateral shoulder, trunk, and proximal forelimb. Caudal regions of the column project to caudal rDAO, which contains cells that respond to stimulation of the distal forelimb. Despite this topography, there is a large degree of overlap in the terminations from neighboring regions of the input column, indicating that a major reorganization occurs at the rDAO. The projection from the rostral cuneate nucleus arises from small neurons that project bilaterally to rDAO, and the input from the rostral cuneate nucleus lacks a clear topography. We propose that input from the cell column is responsible for the somatosensory sensitivity of rDAO neurons, whereas input from rostral cuneate is most likely modulatory, probably inhibitory, in nature.

Organization of reaching and grasping movements in the primate cerebellar nuclei as revealed by focal muscimol inactivations

Two monkeys were trained to point to targets and to retrieve fruit bits from a Kluver board, bottles, and tubes. Once proficient in the tasks, the macaques underwent aseptic surgical implantation of a recording chamber over the cerebellar nuclei on the side of their preferred hand. After recovery from surgery, a series of mapping penetrations were completed to identify task-related areas within the cerebellar nuclei. Muscimol (4- 16 microgram; 1-2 microgram/microliter) was pressure injected at different sites within the forelimb zone, and the resultant deficits were observed as the monkeys performed the behavioral tasks. Quantitative measures of task performance were supplemented by direct observation of live and videotaped performance. The locations of nuclear inactivation sites were reconstructed from marking lesions and tracks visible in histological sections. Injections placed in the cerebellar interpositus nucleus and adjacent regions of dentate caused a variety of deficits in forelimb function. A prominent anteroposterior specialization was apparent within the forelimb zone of this intermediate nuclear region. Injections into the anterior interpositus nucleus and adjacent dentate impaired preshaping of the hand and the manipulation of objects, whereas injections placed more posteriorly in posterior interpositus nucleus and adjacent dentate produced deficits in the aiming of reach and the stability of the arm. During anterior injections, the monkeys failed to adequately extend their fingers in preparation for target contact, as documented for >85% of the reaches in the pointing task of monkey J. Up to 38% of the fruit bits it attempted to retrieve from the Kluver board were dropped. In comparison, during posterior inactivations, 15% were dropped and during control conditions 3% were dropped. The monkeys made significantly greater pointing errors during posterior inactivations (11 times for monkey J and 4 times for monkey C) than during anterior inactivations (8 times for monkey J and 2 times for monkey C). We refer to the region producing hand deficits as the anterior hand zone and the region producing reaching deficits as the posterior reach zone. These results are discussed in relation to the problem of achieving spatiotemporal coordination in the large population of nuclear cells that participate in any given movement. The results do not favor the hypothesis that coordination is achieved through a selection of Purkinje cells along beams of parallel fibers. Instead, it is proposed that distal and proximal musculature is coordinated by the adaptive influences of climbing fiber input to Purkinje cells. We envision a relatively nonspecific recruitment of anterior and posterior nuclear cells due to positive feedback in the limb premotor network, which then is shaped into an appropriate spatiotemporal pattern of discharge through the inhibitory input from Purkinje cells.

Cerebellar guidance of premotor network development and sensorimotor learning

Single unit and imaging studies have shown that the cerebellum is especially active during the acquisition phase of certain motor and cognitive tasks. These data are consistent with the hypothesis that particular sensorimotor procedures are acquired and stored in the cerebellar cortex and that this knowledge can then be exported to the cerebral cortex and premotor networks for more efficient execution. In this article we present a model to illustrate how the cerebellar cortex might guide the development of cortical-cerebellar network connections and how a similar mechanism operating in the adult could mediate the exportation of sensorimotor knowledge from the cerebellum to the motor cortex. The model consists of a three-layered recurrent network representing the cerebello-thalamocortical-ponto-cerebellar limb premotor network. The cerebellar cortex is not explicitly modeled. Our simulations show that Hebbian learning combined with weight normalization allows the emergence of reciprocal and modular structure in the limb premotor network. Reciprocal connections allow activity to reverberate around specific loops. Modularity organizes the connections into specific channels. Furthermore, we show that cerebellar learning can be exported to motor cortex through these modular and reciprocal premotor circuits. In particular, we simulate developmental alignment of visuomotor relations and their realignment as a consequence of prism exposure. The exportation of sensorimotor knowledge from the cerebellum to the motor cortex may allow faster and more efficient execution of learned motor responses.

Network models of the basal ganglia

Over the past year, a number of conceptual and mathematical models of the basal ganglia and their interactions with other areas of the brain have appeared in the literature. Even though the models each differ in significant ways, several computational principles, such as convergence, recurrence and competition, appear to have emerged as common themes of information processing in the basal ganglia. Simulation studies of these models have provoked new types of questions at the many levels of inquiry linking biophysics to behavior.

Prediction of complex two-dimensional trajectories by a cerebellar model of smooth pursuit eye movement

A neural network model based on the anatomy and physiology of the cerebellum is presented that can generate both simple and complex predictive pursuit, while also responding in a feedback mode to visual perturbations from an ongoing trajectory. The model allows the prediction of complex movements by adding two features that are not present in other pursuit models: an array of inputs distributed over a range of physiologically justified delays, and a novel, biologically plausible learning rule that generated changes in synaptic strengths in response to retinal slip errors that arrive after long delays. To directly test the model, its output was compared with the behavior of monkeys tracking the same trajectories. There was a close correspondence between model and monkey performance. Complex target trajectories were created by summing two or three sinusoidal components of different frequencies along horizontal and/or vertical axes. Both the model and the monkeys were able to track these complex sum-of-sines trajectories with small phase delays that averaged 8 and 20 ms in magnitude, respectively. Both the model and the monkeys showed a consistent relationship between the high- and low-frequency components of pursuit: high-frequency components were tracked with small phase lags, whereas low-frequency components were tracked with phase leads. The model was also trained to track targets moving along a circular trajectory with infrequent right-angle perturbations that moved the target along a circle meridian. Before the perturbation, the model tracked the target with very small phase differences that averaged 5 ms. After the perturbation, the model overshot the target while continuing along the expected nonperturbed circular trajectory for 80 ms, before it moved toward the new perturbed trajectory. Monkeys showed similar behaviors with an average phase difference of 3 ms during circular pursuit, followed by a perturbation response after 90 ms. In both cases, the delays required to process visual information were much longer than delays associated with nonperturbed circular and sum-of-sines pursuit. This suggests that both the model and the eye make short-term predictions about future events to compensate for visual feedback delays in receiving information about the direction of a target moving along a changing trajectory. In addition, both the eye and the model can adjust to abrupt changes in target direction on the basis of visual feedback, but do so after significant processing delays.

Somatosensory and movement-related properties of red nucleus: a single unit study in the turtle

Extracellular recordings were performed from turtle red nucleus neurons to examine their responsiveness to peripheral somatic stimulation and to study differences between rubral sensory and movement-related responses. In pentobarbital sodium-anesthetized or decerebrate turtles, red nucleus neurons could be divided into two categories based on their response characteristics. The first group, which included 87% of neurons studied, had low spontaneous rates of activity and responded with excitation to electrical stimulation of the spinal cord or the cerebellum, or during active movement of the contralateral limbs. Neurons in this category were likely to be rubrospinal cells. The remaining 13% of cells studied had higher rates of spontaneous discharge and were inhibited by electrical stimulation or during active movement. These cells might be rubral GABAergic interneurons. Single red nucleus neurons responded with excitation and/or inhibition to somatosensory stimulation. Unlike the motor fields, which were restricted to a single contralateral limb, red nucleus sensory receptive fields were wide and often bilaterally distributed. Rubral responsiveness to sensory stimulation was found to be significantly diminished during active limb movements, thereby suggesting that sensory inputs to the red nucleus are not used for the on-line modification of motor commands. Inactivation of the cerebellar cortex enhanced the sensory responsiveness of rubral neurons and expanded the size of red nucleus receptive fields. These results suggest that the red nucleus receives substantial sensory input, and that the cerebellar cortex can modify the flow of sensory information to the red nucleus.

Motor co-ordinates in primate red nucleus: preferential relation to muscle activation versus kinematic variables

1. Magnocellular red nucleus (RNm) neurones (n = 158) were recorded from two macaque monkeys during a tracking task using one of six single-degree-of-freedom manipulanda. This task allowed us to study discrete movements about most of the joints of the arm. Single-unit, kinematic and electromyographic (EMG) signals from ten to twenty muscles of the upper limb were collected for approximately 2 min while the monkey used a given manipulandum. Movements about different joints were studied by switching among manipulanda. 2. Cross-correlation functions were calculated between RNm discharge rate and the kinematic variables, position and velocity, and between RNm and each of the EMG signals. Statistically significant cross-correlation peaks were found in 24% of the position correlations, 22% of the velocity correlations and 32% of the EMG correlations. The highest correlations were for EMG, reaching above 0.60. The peak correlation provided an effective means of identifying neurones with strong functional relations to one or more movements and/or muscles. These could then be analysed in detail, on a trial-by-trial basis. 3. The similarity between the dynamics of EMG and velocity signals of many highly practised movements makes it difficult to determine which might be the more likely target of RNm control. Therefore, we sought exceptions to this pattern, in order to distinguish between these two possible modes of control. For example, at the end of a movement, muscles occasionally remained active as velocity approached zero. Small corrective movements were often accompanied by a disproportionately large EMG. During these periods, RNm activity usually followed the time course of one or more of the EMG signals as opposed to the velocity signal. In the majority of cases, RNm responses were bidirectional, less frequently unidirectional and rarely reciprocal. These patterns were similar to the patterns of muscle activity. They did not resemble the velocity signals unless the latter were passed through a rectifier. 4. The results support the hypothesis that the red nucleus generates motor commands in a muscle-based co-ordinate system. Covariation between RNm discharge and velocity may result indirectly from correlations between muscle activation and movement. We discuss how the cerebellar cortex might convert the distributed representation of target position, known to be present in the posterior parietal cortex, directly into dynamic, muscle-based commands in the rubro-cortico-cerebellar limb premotor network.

Bistability in cerebellar Purkinje cell dendrites modelled with high-threshold calcium and delayed-rectifier potassium channels

Phase-plane analysis of the ionic currents underlying dendritic plateau potentials was carried out to study the nonlinear dynamics and steady-state transfer properties of the dendritic tree in cerebellar Purkinje cells. The results of an analysis of the P-type calcium and delayed rectifier potassium channel system are presented in this study. These channels constitute a simple system that can support bistability and plateau potentials. By requiring both the steady-state current-voltage curve and nullclines to mimic basic plateau potential properties, we obtained well-defined ranges of specific conductance that can support bistability. Hysteresis was found to be surprisingly prevalent in this simple ion-channel system. Using the steady-state current voltage relationship, we derive concise, algebraic expressions for the voltage and current thresholds of state transitions as functions of specific conductance. The significance of bistability in this ion-channel system is discussed with respect to the generation of plateau potentials in Purkinje cells dendrites and the role of the cerebellum in motor control.

In vitro classical conditioning of abducens nerve discharge in turtles

In vitro classical conditioning of abducens nerve activity was performed using an isolated turtle brainstem-cerebellum preparation by direct stimulation of the cranial nerves. Using a delayed training procedure, the in vitro preparation was presented with paired stimuli consisting of a 1 sec train stimulus applied to the auditory nerve (CS), which immediately preceded a single shock US applied to the trigeminal nerve. Conditioned and unconditioned responses were recorded in the ipsilateral abducens nerve. Acquisition exhibited a positive slope of conditioned responding in 60% of the preparations. Application of unpaired stimuli consisting of CS-alone, alternate CS and US, or backward conditioning failed to result in conditioning, or resulted in extinction of CRs. Latencies of CR onset were timed such that they occurred midway through the CS. Activity-dependent uptake of the dye sulforhodamine was used to examine the spatial distribution of neurons labeled during conditioning. These data showed label in the cerebellum and red nucleus during conditioning whereas these regions failed to label during unconditioned responses. Furthermore, the principal abducens nucleus labeled heavily during conditioning. These findings suggest the feasibility of examining classical conditioning in a vertebrate in vitro brainstem-cerebellum preparation. It is postulated that the abducens nerve CR represents a behavioral correlate of a blink-related eye movement. Multiple sites of conditioning are hypothesized, including the cerebellorubral circuitry and brainstem pathways that activate the principal abducens nucleus.

Three-dimensional reconstruction of the rubrocerebellar premotor network of the turtle

Neuroanatomical studies have demonstrated that the organization of the reptilian rubrocerebellar limb premotor network is similar to that of mammals. This network is composed of prominent recurrent connections among the red nucleus, lateral cerebellar nucleus and lateral reticular nucleus. In this paper the rubrocerebellar system of the turtle was three-dimensionally reconstructed to permit detailed examination of its anatomical organization. Each nucleus and its major efferent pathway was imaged and reconstructed from separate anatomical cases. Section images were used to draw tissue boundaries, mark cell positions and locate axonal trajectories. For each nucleus, drawings of section images containing labeled cells were stacked in the rostrocaudal direction using anatomical landmarks, and a graphic model of the surface was constructed using the method of triangulation. An ellipsoid of equal concentration was computed for each nucleus to ascertain their three-dimensional boundaries and location within the brainstem. To examine the entire rubrocerebellar network, a template of the turtle brainstem and cerebellum was constructed. The component nuclei of the rubrocerebellar network and their axonal projections were then spatially warped onto the template reconstruction on a section by section basis. The final three-dimensional reconstruction of the turtle rubrocerebellar limb premotor network could be rotated in space, allowing proper visualization of the anatomical details of this system. Furthermore, we were able to mathematically section through the reconstruction to obtain brainstem slices with differing orientations and thickness.

Distributed modular architectures linking basal ganglia, cerebellum, and cerebral cortex: their role in planning and controlling action

The motor system includes structures distributed widely through the CNS, and in this feature article we present a scheme for how they might cooperate in the control of action. Distributed modules, which constitute the basic building blocks of our model, include recurrent loops connecting distant brain structures, as well as local circuitry that modulates loop activity. We consider interconnections among the basal ganglia, cerebellum, and cerebral cortex and the specialized properties of certain cell types within each of those structures, namely, striatal spiny neurons, cerebellar Purkinje cells, and neocortical pyramidal cells. In our model, striatal spiny neurons of the basal ganglia function in contextual pattern recognition under the training influence of reinforcement signals transmitted in dopamine fibers. Cerebellar Purkinje cells also function in pattern recognition, in their case to select and execute actions through training supervised by climbing fibers, which signal discoordination. Neocortical pyramidal cells perform collective computations learned through a local training mechanism and also function as information stores for other modular operations. We discuss how distributed modules might function in a parallel, cooperative manner to plan, modulate, and execute action.

Synaptic linkages between red nucleus cells and limb muscles during a multi-joint motor task

The magnocellular red nucleus (RNm) becomes highly active when a monkey reaches to grasp an object. However, the only spike-triggered averaging studies of the RNm to date have been restricted to a simple wrist tracking paradigm and electromyographic (EMG) measurements of muscles of the forearm. We have now measured EMG signals from a large number of muscles throughout the shoulder, arm, forearm, and hand during a variety of tasks, including unconstrained reaching and grasping movements. Relations between these EMG signals and single-unit activity were assessed by on-line spike-triggered averaging and revealed significant post-spike effects among muscles of the shoulder and proximal arm, as well as intrinsic hand muscles. Although there remained a strong bias toward the extensor muscles of the forearm, as has been shown earlier, these results reinforce the importance of the RNm in the control of coordinated, whole-limb reaching movements.

Anatomical organization of the limb premotor network in the turtle (Chrysemys picta) revealed by in vitro transport of biocytin and neurobiotin

The in vitro turtle brainstem-cerebellum preparation has been a valuable tool in the study of central motor programs. In the present study, we investigate the anatomical organization of the turtle rubrocerebellar limb premotor network and its sensory connections in vitro by combining the rapid anterograde and retrograde transport of neurobiotin and biocytin with the extended viability of the isolated turtle brainstem-cerebellum. These compounds retrogradely labeled soma, dendrites, and axons, and orthogradely labeled axons and, to a lesser extent, terminals. The chelonian red nucleus receives a dense input from the contralateral lateral cerebellar nucleus and projects heavily to the contralateral spinal cord. Rubral axons sparsely innervate the lateral cerebellar nucleus and project heavily to the lateral reticular nucleus. Lateral reticular axons heavily innervate the lateral cerebellar nucleus before terminating in the pars lateralis of the cerebellar cortex as mossy fibers. These prominent, recurrent loops among the lateral cerebellar nucleus, red nucleus, and lateral reticular nucleus constitute the turtle rubrocerebellar limb premotor network. Sensory inputs to the red nucleus originate in the contralateral dorsal column nuclei, the principal trigeminal nucleus, and the spinothalamic system. These sites project bilaterally to the lateral reticular nucleus. The lateral cerebellar nucleus receives a contralateral input from the dorsal column nuclei. The red nucleus projects sparsely to the dorsal column nuclei. The red nucleus also receives an ipsilateral descending projection from the suprapeduncular nucleus, located in the diencephalon, and an ascending input from the rostral rhombencephalic reticular formation. An ipsilateral descending pathway originating in the red nucleus is likely to be the rubro-olivary tract.

Model of topographic map development guided by a transiently expressed repulsion molecule

The projection from the retina develops into a precise map of the visual world on the surface of the tectum. The search for molecular position cues that mediate map formation has recently yielded a tectal molecule that exerts a repulsion to fibers from the entire temporal half retina. This molecule appears not to function in the generally accepted gradient manner but instead provides only binary position information, and it is only expressed transiently during early development. Here we describe modeling results that compare the efficacy of binary versus graded position cues in topographic map formation; the model also includes an activity dependent process. We find that binary repulsion is more efficient than graded chemoaffinity in the rapid establishment of map polarity, and transient expression of either cue provides sufficient guidance for precise map formation.

Correlation of primate red nucleus discharge with muscle activity during free-form arm movements

1. We recorded from 239 neurons located in the magnocellular division of the red nucleus of four alert macaque monkeys. At the same time, we recorded electromyographic (EMG) signals from as many as twenty electrodes chronically implanted on muscles of the shoulder, arm, forearm and hand. We recorded EMG signals for periods ranging from several months to a year. 2. The monkeys were trained to perform three free-form food retrieval tasks, each of which activated all of the recorded muscles and most of the neurons. The 'prehension' task required simply that the monkey grasp a piece of food from a fixed point in space. The 'barrier' task required the monkey to reach around a small barrier to obtain the food, and the 'Kluver' task required that food be removed from small holes. During the prehension task, we found approximately equal numbers of neurons that were strongly active while the hand was being moved toward the target (70% of units), and while the food was being grasped (60%). Relatively few units were active as the hand was returned to the mouth (15%). 3. Data files of 1-2 min duration were collected while the monkey performed a single behavioural task. Whenever possible, we recorded files for all three tasks from each neuron. For each file we calculated long time-span analog cross-correlations (+/- 1.28 s) between instantaneous neuronal firing rate and each of the full-wave rectified, low-pass filtered EMG signals. We used the peak correlation and the time of the peak as two summary measures of the functional relation between modulation of neuronal activity and EMG. 4. The magnitude of the strongest correlations was between 0.4 and 0.5 (normalized to a perfect correlation of +/- 1.0). Distal muscles were the most frequently correlated, and extensors were more frequently correlated than flexors. For all monkeys, the lags for well correlated muscles were distributed broadly about a uni-modal value near 0 ms. Eighty five per cent of the correlations larger than or equal to 0.25 had peaks between -150 and 200 ms. 5. The activity of each neuron was represented in a muscle co-ordinate system by an n-dimensional 'functional linkage vector', each element of which was the peak correlation with one of n muscles. The vector for any given neuron points in a particular direction in muscle space, depending on the similarity between the activity of the neuron and the activity of each muscle.(ABSTRACT TRUNCATED AT 400 WORDS)

Intrinsic and synaptic properties of turtle red nucleus neurons in vitro

Burst discharges in the red nucleus are correlated with discrete limb movements. Intracellular recordings from red nucleus neurons in the in vitro turtle brainstem-cerebellum was performed to elucidate mechanisms underlying these bursts. Depolarizing intracellular current injection failed to demonstrate endogenous membrane currents that might produce burst discharges, and neurons did not exhibit significant spike frequency adaptation, which is a characteristic of synaptically driven bursts. Responses of red nucleus neurons to synaptic input demonstrated a late, slow depolarizing synaptic potential (slow EPSP) having a latency of 9-12 ms, and a maximal duration of 600 ms. it is concluded that neither intrinsic membrane responses, nor the duration of the slow EPSP, can fully account for the behavior of red nucleus neurons during burst discharge. We hypothesize that activity in the red nucleus is driven by a gradual recruitment of NMDA receptors, and lpr by polysynaptic excitatory pathways.

Receptive fields of single cells from the face zone of the cat rostral dorsal accessory olive

Natural stimulation was used to map the receptive fields of single cells recorded from the rostral medial portion of the dorsal accessory olive (rDAO) and the subjacent principal olive (PO) of the barbiturate anesthetized cat. Previous reports indicated a somatotopic mapping of the entire contralateral body within the rDAO which included a small face zone and a larger zone with a very precise map for the limbs. While concentrating on the face zone of the rDAO we confirmed the previously reported somatotopy (face: rostral and medial; forelimb: caudal and medial; hindlimb; caudal and lateral; and trunk: rostal and lateral) and found a somatotopy within, and adjacent to, the face zone. At the border between rDAO regions representing forelimb and face, cells with forelimb fields were found to lie dorsally to cells with facial fields. Within the rDAO face region, cells with large facial fields lie dorsally to cells with small facial fields. In both cases, the more ventral cells lie in the ventral lamella of the PO, which suggests a functional as well as physical continuity between rDAO and the ventral lamella of the PO. We therefore conclude that the face zone in the rDAO and the face zone in the PO form one continuous and complete map of the face with an orderly progression of receptive fields. Furthermore, we have found that stimulation of the red nucleus can inhibit rDAO cells with facial receptive fields just as it does cells with receptive fields from the rest of the body.

Anatomy of the turtle cerebellorubral circuit studied in vitro using neurobiotin and biocytin

We have combined the rapid anterograde and retrograde transport of neurobiotin and biocytin with the extended viability of the isolated turtle brainstem-cerebellum to conduct in vitro studies of the chelonian cerebellorubral circuit. Tracers were pressure injected in 15-25 nl quantities and the optimal transport time was 16 h. Tissue sections were incubated with avidin-biotin-HRP complex and reacted with DAB. Retrogradely labeled soma, dendrites and axons, and anterogradely labeled axons and to a lesser extent terminals were visible with both tracers. Red nucleus injections resulted in dense retrograde label in the contralateral lateral cerebellar nucleus and a heavily labeled contralateral rubrospinal tract. Cerebellar nucleus injections revealed light retrograde and dense terminal label in the contralateral red nucleus, together with retrograde label in a cell cluster in the ipsilateral ventrolateral medullary reticular formation, an area we identify as the lateral reticular nucleus. Injections into this medullary region resulted in heavy mossy fiber input to the ipsilateral cerebellum and moderate retrograde label in the contralateral red nucleus. These results identify prominent recurrent projections between the lateral cerebellar nucleus, red nucleus and lateral reticular nucleus, in addition to revealing other features of the cerebellorubral circuit.

Distributed Representation of Limb Motor Programs in Arrays of Adjustable Pattern Generators

This paper describes the current state of our exploration of how motor program concepts may be related to neural mechanisms. We have proposed a model of sensorimotor networks with architectures inspired by the anatomy and physiology of the cerebellum and its interconnections with the red nucleus and the motor cortex. We proposed the concept of rubrocerebellar and corticocerebellar information processing modules that function as adjustable pattern generators (APGs) capable of the storage, recall, and execution of motor programs. The APG array model described in this paper extends the single APG model of Houk et al. (1990) to an array of APGs whose collective activity controls movement of a simple two degree-of-freedom simulated limb. Our objective was to examine the APG array theory in a simple computational framework with a plausible relationship to anatomy and physiology. Results of simulation experiments show that the APG array model is capable of learning how to control movement of the simulated limb by adjusting distributed motor programs. Although the model is based on many simplifying assumptions, and the simulated motor control task is much simpler than an actual reaching task, these results suggest that the APG array model may provide a useful step toward a more comprehensive understanding of how neural mechanisms may generate motor programs.

Movement-related inputs to intermediate cerebellum of the monkey

1. The primary goal of this study was to characterize the information about single-joint forelimb movements supplied to intermediate cerebellar cortex by mossy fibers. Discharge of mossy fibers and Golgi cells was studied while monkeys operated six devices that required movements about specific joints. Additional control experiments in anesthetized cats and monkeys established criteria for identification of mossy fibers and Golgi cells. 2. The control experiments demonstrate that mossy fibers can be distinguished from Purkinje and Golgi cells by the waveshapes of their action potentials. Asynaptic activation from the inferior cerebellar peduncle, in combination with histological localization of recording sites in granular layer or subcortical white matter, verified that mossy fibers produce a variety of waveshapes that are characterized by brief initial phases and relatively small amplitudes. The same waveshapes were observed for the mossy fiber recordings from awake monkeys, and many identified mossy fibers had sensory properties similar to those found in the awake animals. From these combined criteria, we conclude that the recordings in the awake animals were from mossy fibers. Golgi cells, recorded exclusively in the granular layer of cerebellar cortex, were characterized by action potentials of longer duration and larger amplitude as compared with mossy fibers, and none were asynaptically activated from the inferior cerebellar peduncle. 3. Units were isolated while the monkeys made free-form and tracking movements. We studied movement-related discharge of 80 mossy fibers and 12 Golgi cells. Mossy fibers showed high modulations during use of at least one of the six manipulanda and had clear preferences for movement about a specific joint, although they often showed consistent but weaker firing during movement about a neighboring joint. Separation of movements by more than one joint produced a large reduction in discharge: shoulder units never fired well to movements of the finger, and finger units never fired well to movement of the shoulder. 4. The tracking task required maintenance of fixed limb positions (a static phase) as well as movements between these positions (a dynamic phase). Of 80 mossy fibers, 18% had purely tonic discharge patterns, 63% were phasic-tonic, and 20% were purely phasic. Discharge patterns were reciprocal (45%), bidirectional (42%), or unidirectional (13%). 5. Eighty percent of the mossy fibers exhibited tonic discharge that was significantly (P < 0.01) correlated with joint angle (r = 0.65 +/- 0.19, mean +/- SD), and about one third had phasic components that were significantly correlated with movement velocity.(ABSTRACT TRUNCATED AT 400 WORDS)

Distributed motor commands in the limb premotor network

Neuroanatomical studies have demonstrated extensive interconnections between the motor cortex, red nucleus and cerebellum, forming a premotor network for controlling limb movement. Single-unit studies indicate that command signals for limb movements are distributed broadly throughout this network. Cellular studies have demonstrated multiple recurrent loops in this network, and the presence of excitatory and inhibitory amino acid neurotransmitters. A recent model suggests that movement commands are initiated by sensory inputs to these loops, and that positive feedback, regulated by inhibition from cerebellar Purkinje cells, distributes commands throughout the limb premotor network. This model offers a new framework for exploring relationships between basic neural mechanisms and concepts of motor performance that derive from experimental psychology.

Output organization of intermediate cerebellum of the monkey

1. The goal of this study was to investigate the motor organization of monkey nucleus interpositus (NI) and neighboring regions of the lateral nucleus (NL) by correlating discharge of single neurons with active movements. Neurons were surveyed during free-form movements as well as during operation of six devices that required movement about specific forelimb joints. The paradigm allowed us to test the hypothesis that discharge of individual cells relates to movements about individual joints. 2. One hundred sixty-two isolated nuclear neurons from two monkeys were studied. Eighty-three percent showed large increases in discharge (an average of 3 times resting rate for forelimb neurons) during movement of one body part, either forelimb, hindlimb, mouth/face, or eyes. 3. Anterior interpositus contains neurons related to hindlimb movement in anterior regions and neurons related to forelimb movement in posterior regions. A mouth/face-related area exists in the dorsal-posterior regions and is continuous with a mouth/face area in the dorsal regions of NL. Posterior interpositus (NIP) showed no clear separation between forelimb and hindlimb neurons: forelimb neurons were encountered throughout the nucleus, and hindlimb neurons were encountered in the medial-anterior two thirds. A distinct eye movement area exists in lateral, posterior, and ventral regions of NIP. This area borders regions of NL that also contain eye movement-related neurons. 4. Forelimb interpositus neurons discharged strongly during reach and grasp; discharge rates were recorded for 41 neurons during a stereotyped reach and the average depth of modulation was 149 imp/s. Nineteen neurons that modulated during device tracking were also tested during reaching, and the depth of modulation was much greater during reaching. 5. Fifty-nine forelimb neurons were tested with device tracking. Twenty-seven (46%) produced no audible modulation, regardless of the joint being exercised. The remaining 32 neurons modulated during movement on at least one device (mean depth of modulation = 84 imp/s). Comparison of discharge during use of different devices revealed no strong evidence for device-specific discharge. 6. Discharge modulations during device tracking were phasic, preceded movement, and, for a small number of cells, showed consistent parametric relations to duration, amplitude, and velocity of movement. 7. Despite a clear somatotopy within NI and NL, there is no finer mapping based on active movements about individual joints within forelimb regions. Discharge modulation depends on movements involving the whole limb. Progress in understanding the function of intermediate cerebellum depends on determining the variables required to elicit consistent and high modulation of neural discharge.(ABSTRACT TRUNCATED AT 400 WORDS)

Sulforhodamine labeling of neural circuits engaged in motor pattern generation in the in vitro turtle brainstem-cerebellum

A fluorescent molecular probe was used in combination with a novel in vitro preparation to study spatial patterns of neural activity associated with motor pattern generation. The in vitro brainstem-cerebellum preparation takes advantage of the turtle's unusual resistance to anoxia to preserve the entire neural network that connects the cerebellum, red nucleus, and reticular formation. This preparation was bathed in a 0.01% solution of sulforhodamine while it was activated unilaterally by electrical stimulation of the dorsal quadrant of the spinal cord for 1 hr. Sulforhodamine is a small, sulfonated, highly charged fluorescent molecule that is taken up by endocytosis. To examine its distribution in the cerebellum and brainstem, coronal sections were prepared and viewed under epifluorescence illumination. Distinctive spatial patterns of labeling were associated with unilateral electrical stimulation of the in vitro network, suggesting that dye uptake was activity dependent. Blockade of uptake with altered magnesium and calcium concentrations indicated that single spike discharge evoked ortho- or antidromically was insufficient to induce dye uptake. Instead, sulforhodamine staining correlated with the presence of burst discharge that was recorded extracellularly from the red nucleus. Blockade of burst discharge with excitatory amino acid receptor antagonists prevented dye uptake in the red nucleus, the lateral cerebellar nucleus, and other structures that are known to be interconnected by recurrent anatomical pathways. These results suggest that sulforhodamine is internalized by intensely active neurons. The spatial distributions of label support the hypothesis that burst discharges in the turtle red nucleus are mediated by excitatory amino acid neurotransmitters and sustained by recurrent excitation in cerebellorubral synaptic pathways. Positive feedback in these recurrent pathways may provide an important driving force for the generation of motor programs that control limb movements.

Spatial overlap of rubrospinal and corticospinal terminals with input to the inferior olive

Somatosensory responses of cells in the dorsal accessory olive are suppressed following stimulation of the magnocellular red nucleus. Since the magnocellular red nucleus of the cat does not project directly to the dorsal accessory olive, the present experiments were designed to identify indirect pathways that might mediate suppression of olivary responsiveness. Wheat germ agglutinin-horseradish peroxidase was used to compare the location of magnocellular red nucleus terminals with the locations of cells providing input to the rostral dorsal accessory olive. Cells projecting to forelimb rostral dorsal accessory olive can be divided into two main groups: one group comprises a column of large cells located in the ventral caudal cuneate nucleus extending into lamina VI of C1 and C2, and a second group comprises smaller cells located in the ventral rostral cuneate nucleus. Terminations of fibers originating in the magnocellular red nucleus were found to target both groups of cells projecting to the dorsal accessory olive. Therefore, it is possible that the responsiveness of olivary cells is influenced via these terminations. Stimulation of sensorimotor cortex has also been shown to inhibit olivary responsiveness. Terminations from sensorimotor cortex target the same regions of cells that project to the dorsal accessory olive as those of the magnocellular red nucleus, and a similar, perhaps identical, anatomical substrate may serve to modulate olivary sensitivity by the two descending systems.

Evidence for GABAergic interneurons in the red nucleus of the painted turtle

Immunocytochemical and electrophysiological evidence supporting the presence of GABAergic interneurons in the turtle red nucleus is presented. Injections of HRP into the spinal cord produced labeling of large neurons in the contralateral red nucleus. The peroxidase-antiperoxidase (PAP) method revealed smaller cells immunoreactive to an antibody against glutamate decarboxylase (GAD), the synthetic enzyme for the inhibitory neurotransmitter GABA, that were interspersed among larger immunonegative neurons. Similar small neurons were densely immunostained by antibodies to GABA-glutaraldehyde conjugates obtained from different sources and applied according to pre-embedding and postembedding protocols. Rubrospinal neurons retrogradely labeled with HRP measured 16 and 27 microns in mean minor and major cell body diameters, while GABA-like immunopositive neurons situated within the red nucleus measured 7 and 13 microns. There was very little overlap in soma size between the two cell populations. Therefore, we suggest that the GAD- and GABA-positive neurons may be local inhibitory interneurons. This notion is further supported by observations of pre-embedding immunostaining for GAD and postembedding immunostaining for GABA showing that the turtle red nucleus is amply innervated by immunoreactive axon terminals. These puncta are closely apposed to cell bodies and dendrites of both immunonegative large neurons and immunopositive small neurons. Moreover, immunogold staining at the electron microscopic level demonstrated that GABA-like immunoreactive axon terminals with pleomorphic synaptic vesicles formed symmetric synapses with cell bodies and dendrites of the two types of red nucleus cells. These ultrastructural features are commonly assumed to indicate inhibitory synapses. A moderately labeled bouton with round vesicles and asymmetric synapses was also observed. In addition, the two types of red nucleus neurons received asymmetric axosomatic and axodendritic synapses with GABA-negative boutons provided with round vesicles, features usually associated with excitatory functions. To obtain electrophysiological evidence for inhibition, intracellular recordings from red nucleus neurons were conducted using an in vitro brainstem-cerebellum preparation from the turtle. Small, spontaneous IPSPs were recorded from 7 out of 14 red nucleus cells studied. These morphological and physiological results provide strong support for concluding that the turtle red nucleus, like its mammalian counterpart, contains GABAergic inhibitory interneurons. While we have not identified the main source of input to these interneurons, in view of the scarce development of the reptilian cerebral cortex, this input is unlikely to come from the motor cortex as it does in mammals.(ABSTRACT TRUNCATED AT 400 WORDS)

Red nucleus: role in motor control

Experimental reports in the past year have provided a better understanding of the motor functions of excitatory and inhibitory neurotransmitters in the red nucleus, and of the sensorimotor properties of single rubral neurons. These data fit well within the framework of a neural network model of the rubrocerebellar system.

Role of excitatory amino acids in mediating burst discharge of red nucleus neurons in the in vitro turtle brain stem-cerebellum

1. Bursts of discharge have been recorded in the red nucleus in several species and are thought to represent the expression of motor commands. A cerebellorubral circuit comprised of recurrent connections among the cerebellum, red nucleus, and reticular formation was postulated to function as a positive feedback loop that generates these motor commands and transmits them to the spinal cord via the rubrospinal pathway. We have used an in vitro preparation from the turtle that leaves the circuitry connecting the cerebellum, brain stem, and spinal cord intact to study the role of excitatory amino acid neurotransmitters and recurrent excitation in mediating the generation of burst discharges in the red nucleus. 2. Burst discharges were recorded extracellularly from single cells in the red nucleus in response to single pulse or brief train stimulation of the contralateral spinal cord or brief train stimuli applied to the ipsilateral cerebellar cortex. The firing characteristics and pharmacologic sensitivities of the bursts were independent of the type of stimulus used. The bursts had long durations ranging from 2 to 17 s and showed spike frequency adaptation. 3. Transection of the cerebellar peduncle, which eliminates inhibition impinging onto the cerebellorubral circuit, greatly enhanced the spontaneous activity and burst discharges recorded in the contralateral red nucleus. Furthermore, bath application of a solution containing elevated levels of calcium and magnesium blocked the expression of burst discharges even though synaptic activation of the neurons was not blocked. 4. The possibility that excitatory amino acid receptors mediate burst responses in the red nucleus was investigated in light of the antagonistic effects of elevated magnesium ions on bursting. Bath application of 100 microns DL-2-amino-5-phosphonovaleric acid (APV), a specific N-methyl-D-aspartate (NMDA) receptor antagonist; [10 microM 6-cyano-7-nitroquinoxaline-2,3-dione (CNQX)], a specific non-NMDA receptor antagonist; or 100 microM, DL-2-amino-4-phosphonobutyric acid (AP4), an agonist of a fourth class of excitatory amino acid receptor, blocked burst activity in the red nucleus. With a multibarreled pipette for simultaneous ejection of drug and recording, iontophoresis of APV or CNQX into the red nucleus blocked bursting whereas AP4 failed to show a significant effect. These data suggest that red nucleus neurons have both NMDA and non-NMDA receptors. The site of action of the AP4-sensitive receptor appears to be elsewhere in the cerebellorubral circuit. 5. Iontophoretic application of excitatory amino acid receptor agonists NMDA and quisqualate (Q) induced excitation of red nucleus neurons.(ABSTRACT TRUNCATED AT 400 WORDS)

Inhibition of sensory responses of cat inferior olive neurons produced by stimulation of red nucleus

1. The sensory responsiveness of cells in the inferior olive is known to be suppressed during certain phases of active movement. These experiments were designed to test the possibility that activity in the rubrospinal pathway contributes to this suppression. We recorded from cells sensitive to light touch located in one of the divisions of the inferior olive, the rostral dorsal accessory olive (rDAO), in cats anesthetized with pentobarbitol sodium. Responsiveness to peripheral stimuli was tested during and after trains of conditioning stimuli delivered to the rubrospinal pathway. 2. All 44 cells in our sample of rDAO neurons showed an inhibition of responsiveness to peripheral stimuli after conditioning stimulation of the rubrospinal pathway. Typical conditioning trains consisted of 0.2-ms pulses at 200 Hz for 100 ms. The mean current required for a reduction in firing probability to 0.5 was 31 microA. Slight increases in intensity often completely inhibited responses to peripheral stimuli. 3. Inhibition of responsiveness showed a delayed time course. Peak inhibition occurred approximately 50 ms after the last pulse in the conditioning train. In many cases there was no demonstrable inhibition during the conditioning train. Increases of train frequency, train duration, or stimulus intensity produced stronger and broader periods of olivary inhibition. 4. The lowest threshold points for eliciting rDAO inhibition coincided with either the magnocellular red nucleus (RNm) or the rubrospinal tract (RST). Stimulation at RST sites produced inhibition of responses in the contralateral but not in the ipsilateral rDAO. Transection of the RST in the upper brain stem blocked the inhibition produced by red-nucleus stimulation without altering the inhibition produced by tract stimulation caudal to the transection. The inhibitory effects thus appear to be caused by activation of the rubrospinal pathway. 5. The inhibitory timing observed in this study may be appropriate for explaining the suppression of olivary responsiveness to contact that has been observed in awake animals. Bursts of movement-related, red nucleus discharge often cease approximately 50 ms before the end of movement. This timing would allow peak inhibition to develop at approximately the time of contact with an object at the end of a goal-directed limb movement.

An in vitro preparation for studying motor pattern generation in the cerebellorubrospinal circuit of the turtle

In vivo studies in mammals have suggested that the cerebellorubrospinal circuit functions as a recurrent excitatory loop that generates motor commands and transmits them to the spinal cord via the rubrospinal pathway. Here we describe an in vitro preparation from the turtle exhibiting functional synaptic connections between the cerebellum, brainstem and upper spinal cord that is suitable for detailed analysis of this circuit. Electrical stimulation of the spinal cord was used to activate the cerebellorubrospinal circuit while activity was sampled with extracellular recordings from single cells in the red nucleus. Single units responded to stimulation with short and long latency synaptic responses, in addition to antidromic activation. Some cells showed bursts of activity lasting several hundred milliseconds suggesting the presence of recurrent excitation. Interruption of Purkinje cell inhibitory input impinging on the cerebellorubrospinal loop prolonged bursting and enhanced spontaneous activity. This preparation should facilitate the examination of the role of the cerebellorubrospinal circuit in motor pattern generation.

Activity of primate magnocellular red nucleus related to hand and finger movements

Single cells were recorded in the magnocellular red nucleus (RNm) of two cynomolgus monkeys using tungsten microelectrodes. The first monkey was trained to press finger switches and to operate a push-pull device. Comparison of responses while operating the two devices demonstrated a strong distal bias. The finger device elicited large modulations in discharge (greater than or equal to 50 impulses/s) in 75% of the sampled neurons. Most cells fired optimally during thumb switch operation, but also fired vigorously in association with other switch operations. The left motor cortex was removed from the second monkey 18 months prior to microelectrode recording. Cells in the cortically denervated RNm discharged vigorously in association with grouped finger movements that opened and closed the affected right hand. These results coupled with our previous findings suggest that the RNm is preferentially linked to distal limb muscles, and the primary role of the forelimb zone may be to control coordinated hand function including grouped movements of the fingers.

Control strategies in physiological systems

In this paper, written for a general audience, I review and contrast various strategies that the body uses to control homeostasis and movement. Messages, signals, communication channels, and control systems are dealt with from both a cellular and an integrative perspective. The major global control strategies are feedback, feedforward, and adaptive control, and examples of each are presented to highlight advantageous and disadvantageous features. Many physiological systems use these three strategies in combination.

Selective projections from the cat red nucleus to digit motor neurons

Classical studies of the cat rubrospinal tract describe dense terminations in spinal laminae V-VII and an absence of any significant projection to lamina IX. In contrast, our recent studies, utilizing the anterograde transport of wheat germ agglutinin conjugated with horseradish peroxidase, have demonstrated a consistent and circumscribed area of label in lamina IX at caudal cervical segments. The present study was undertaken to determine the distribution of rubrospinal terminals among motor neurons in lamina IX as well as to identify the likely target muscles of those motor neurons located near rubrospinal terminals. We injected wheat germ agglutinin-horseradish peroxidase into the red nucleus and unconjugated horseradish peroxidase into selected forearm muscles of the same side of the body. The locations of rubrospinal terminals showing anterograde label on one side of the spinal cord could then be compared with the locations of motor neurons showing retrograde label on the opposite side of the cord. The results demonstrated a clear focus of rubrospinal terminals in the lateral and dorsal portions of the ventral horn beginning at C8 and extending through rostral T1. No other segments of the spinal cord showed a focus of rubrospinal terminations in lamina IX. Retrogradely labeled motor neurons from the muscle injections showed that the rubrospinal terminals overlap extensively with motor neuronal pools supplying distal forearm muscles. Several lines of evidence indicate that the terminals are from rubrospinal fibers and are not due to transneuronal transport.

Correlation and spectral analysis of relations between single unit discharge and muscle activities

Correlation and spectral analysis was used to study functional linkages between single-cell discharge in the magnocellular red nucleus and the electromyographic activity of several limb muscles. Long sequences of unit discharge and EMG activity were recorded while feline subjects performed a food retrieval task. Unit discharge and muscle activity were patterned in bursts that corresponded to different phases of the task. There was sufficient variability in the parameters of these bursts to regard the signals as pseudorandom variables, thus facilitating a correlation analysis. Cross-correlation functions computed between unit discharge and each muscle EMG served to characterize the strength and reliability of linkages between a single unit and various limb muscles. Auto-correlation and auto-spectral density functions provided summary measures of the temporal and frequency characteristics of the signals. Power in the signals was concentrated in a behaviorally relevant range (0.2-8 Hz). Coherence functions showed peaks that indicated which frequency components were well correlated with unit discharge. Two-sided impulse responses served to characterize the dynamic properties of the linkages. These methods are shown to be valuable in characterizing noncausal, as well as causal, linkages through multisynaptic pathways in the nervous system.