| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Journal of Neurophysiology Vol. 87 No. 3 March 2002, pp. 1473-1487
Copyright ©2002 by the American Physiological Society
Department of Kinesiology, University of Wisconsin, Madison, Wisconsin 53706-1532
 |
ABSTRACT |
|---|
 TOP
 ABSTRACT
 INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Van Kan, Peter L. E. and Martha L. McCurdy. Contribution of Primate Magnocellular Red Nucleus to Timing of Hand Preshaping During Reaching to Grasp. J. Neurophysiol. 87: 1473-1487, 2002. Magnocellular red nucleus (RNm) is involved in controlling goal-directed limb movements such as reaching to grasp. We tested two hypotheses related to RNm's role in controlling reach-to-grasp movements. One hypothesis is that forelimb RNm neurons are grasp specific, and the other is that they specify the timing of metacarpi-phalangeal (MCP) extension to preshape the hand during the appropriate phase of the reach. We recorded single-unit discharge while monkeys performed two behavioral tasks that elicited similar reaches but differed in grasp. One task consisted of a reach with a precision grasp that elicited independent use of thumb and forefinger; the other included a whole-hand grasp that elicited concerted use of the four fingers. Most RNm neurons tested were engaged strongly during both the whole-hand and precision tasks, and the magnitude of discharge modulation did not differ between tasks. Thus most RNm neurons are not grasp specific but, instead, may contribute to behavioral features common to the two tasks. Two methods were used to investigate relations between single-unit discharge and kinematic data from the same individual trials of the whole-hand and precision tasks for a subset of forelimb RNm neurons. One method focused on correlations between parameters of RNm discharge and the duration, amplitude, and velocity of rotation of forelimb joints for each of the tasks. The second method compared between-task differences in times of peak neuronal discharge to between-task differences in times of rotations of forelimb joints. Parameters of reach-related RNm discharge were more frequently correlated with parameters of MCP extension than with parameters of rotation of wrist, elbow, and shoulder joints. Analyses of temporal relations between discharge and kinematic data during both the whole-hand and precision tasks indicate that discharge was time locked most frequently to MCP extension and, to a lesser extent, elbow extension during both tasks. We conclude that RNm may command muscle synergies that provide a basic preshape of the hand at the appropriate phase of limb transport. In addition, the timing of RNm's contribution to hand preshaping varies with the behavioral requirements of the task.
| |
INTRODUCTION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
Reaching to grasp is crucial
for our ability to interact effectively with our environment. Motor
commands that preshape the hand while moving it to the object to be
grasped require extensive processing of visuomotor information. This
processing may occur in functional modules that are coupled but are
controlled in parallel: one is limb transport and another is hand
preshaping (for review: Jeannerod 1988
). Transport and
preshaping of the hand interact. Perturbing object location, for
example, affects transport of the hand as well as kinematics and timing
of hand preshaping (Paulignan et al. 1991b). The
correction in transport requires additional time, and, correspondingly,
preshaping of the hand occurs later in the reach. In addition,
perturbing object size requires additional time to achieve the
appropriate preshaping of the hand, and, correspondingly, transport of
the hand is slowed (Paulignan et al. 1991a). Thus the
success of reach-to-grasp movements depends on appropriate spatiotemporal coordination between distal and proximal forelimb muscles.
An unresolved issue in the neural control of reaching to grasp is
whether preshaping and transport of the hand are controlled by a single
neural system or by separate systems. The organization of descending
motor pathways into medial and lateral systems (Kuypers 1982) may appear more compatible with the latter possibility
because the medial system preferentially targets proximal muscles, and the lateral system preferentially targets distal muscles. Behavioral studies indicate, however, that control through medial and lateral systems is not restricted to proximal and distal muscles, respectively. Transection of both corticospinal and rubrospinal pathways, for example, causes permanent deficits in independent hand and finger movements, but leaves intact control of distal muscles within the
context of whole-body movements, such as grasping while climbing (Lawrence and Kuypers 1968b). Thus, although medial and
lateral systems are functionally distinct and their functions depend on differential innervation of proximal and distal musculature, the relative contribution of a given pathway to control of distal versus
proximal parts of the limb, and the functional overlap between
pathways, remain unknown.
The rubrospinal tract is a component of the lateral system and provides
direct access to spinal circuitry used in the production of movement of
contralateral body parts in monkeys, cats, and rats (for review:
Keifer and Houk 1994; Massion 1967). Rubrospinal fibers
originate from magnocellular red nucleus (RNm), which receives major
input from intermediate cerebellum via nucleus interpositus (NI). RNm
neurons target interneurons at all spinal levels but terminate
selectively among motoneurons in
C8-T1 that innervate digit
muscles in cats (McCurdy et al. 1987) and monkeys
(Holstege 1987). Single-unit recording studies of RNm
and NI support the hypothesis that cerebellar output via RNm is
specialized for controlling aspects of hand use. Testing movement
specificity of monkey RNm neurons showed that devices that elicited
rotation of metacarpi-phalangeal (MCP) joints in isolation or in
combination with rotations about the wrist were more effective in
eliciting strong discharge than devices that elicited isolated wrist,
elbow, or shoulder rotations (Gibson et al. 1985a). RNm
discharge may command movements or muscles of distal forelimb joints
(Gibson et al. 1985a,b; Mewes and Cheney
1994; Miller and Houk 1995). Whereas only some
RNm and NI neurons discharge strongly during movements restricted to
individual distal or individual proximal joints (Gibson et al.
1985a; Van Kan et al. 1993), most forelimb RNm
and NI neurons discharge strongly during reaching to grasp
(Gibson et al. 1985a, 1996; Miller
and Houk 1995; Van Kan et al. 1993,
1994). In addition, the same RNm and NI neurons that
were strongly activated during reaching to grasp were only weakly
activated during similar reaches without a grasp (Van Kan and
McCurdy 2001; Van Kan et al. 1994). Furthermore,
reach-to-grasp-related discharge of NI neurons did not depend on
variations in trajectory, amplitude, and direction of the reach but,
instead, was contingent on grasp (Gibson et al. 1996;
Van Kan et al. 1994). The combined results emphasize the
importance of the discharge of RNm and NI neurons to hand use.
Recently, we have reported that one aspect of hand use that is
important to the discharge of some RNm neurons is MCP extension (Van Kan and McCurdy 2001). These findings suggest
functional specialization: cerebellar output via RNm is strongly
activated during coordinated reach-to-grasp movements but may relate
specifically to preshaping rather than transport of the hand. The
present study addressed this question by testing the same RNm neurons
during performance of reaching tasks that differed in grasp. One task required concerted use of the four fingers, and the other required a
precision grip. MCP extension was associated with hand preshaping during both tasks and occurred during a later phase of the precision than whole-hand task. The resulting temporal dissociation between MCP
extension and movements of proximal joints provided an opportunity to
evaluate coupling between discharge of a given RNm neuron and preshaping versus transport of the hand. In addition, testing the same
neurons during performance of both tasks provided information related
to RNm's role in controlling grouped versus individuated finger
movements. Brief reports of some of the results have been published
(Van Kan and McCurdy 1998; Van Kan et al.
2000).
| |
METHODS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
We studied neurons in the RNm of two monkeys (Macaca
mulatta) during reaching to grasp. All animal care and
experimental procedures complied with the National Institutes of Health
Policy on Humane Care and Use of Laboratory Animals, conformed to the
National Institutes of Health Guide for the Care and Use of Laboratory Animals, and were approved by the Institutional Animal Care and Use
Committee. Methods and experimental procedures have been described in
detail previously (Van Kan and McCurdy 2001).
Behavioral paradigm
The monkeys were trained to perform two reach-to-grasp tasks that involved the proximal forelimb similarly but differed in grasp. The animals reached with their right forelimb for cereal rewards presented at four target locations in front of them. Two target locations (left and right) were at shoulder height at angles of 31° to the left and 28° to the right of the parasagittal plane through the shoulder. The other two (upper and lower) were within the sagittal plane through the shoulder at angles of 56° above and 5° below the horizontal plane through the shoulder. Reaches started from a common location near the waist where the monkeys held onto a handle while receiving water reward. During an intertrial interval of variable duration (3-5 s), a mechanical arm (Scorbot ER-III, Eshed Robotec) moved the target assembly in a pseudorandom order to one of the four target locations. Illumination of a light-emitting diode (LED) at the target served as cue to move. Simultaneously with LED onset, a computer-controlled air cylinder dispensed a single piece of cereal (Froot Loop, Kellogg).
WHOLE-HAND TASK. In the whole-hand task, the cereal was presented in a beaker (32 mm ID). The beaker was tilted at a 45° angle toward the animal. The animal released the handle, reached, and inserted four fingers into the beaker to retrieve the cereal (Fig. 1, A and C).
|
PRECISION TASK. In the precision task, the cereal was presented in the center of a horizontally oriented slot (25 × 25 × 6 mm). Fractionated use of the forefinger and thumb were required to retrieve the cereal (Fig. 1, B and D).
Surgical preparation
Following completion of behavioral training, the monkeys were prepared for chronic recording. Under surgical anesthesia, a recording cylinder was fastened above a craniotomy. On completion of collection of neuronal data, 21 forelimb muscles were implanted percutaneously with fine bipolar electromyographic (EMG) electrodes: EDC, ED2,3, and ED4,5, extensor digitorum communis, two and three, and four and five, respectively; APL and EPL, abductor and extensor pollicus longus, respectively; FDS and FDP, flexor digitorum superficialis and profundus; PL, palmaris longus; ECR and ECU, extensor carpi radialis and ulnaris; FCR and FCU, flexor carpi radialis and ulnaris; TRI, triceps; BIC, biceps; BR, brachioradialis; spDLT, spinodeltoid; clDLT, cleidodeltoid; acDLT, acromion deltoid; TM, teres major; LAT, latissimus dorsi; PEC, pectoralis.
Data collection
Discharge of individual RNm neurons was recorded in daily sessions of 2- to 3-h duration. EMG activity was recorded during performance of the same tasks by the same animals but in sessions separate from those in which neuronal discharge was recorded. Reach onset and reach end were defined as the times of making and breaking contact with the handle and beaker or slot, respectively. Times of occurrence of action potentials were recorded with 100-µs precision. Output signals of the contact sensors, and rectified, integrated EMG signals were digitized at 167 Hz (CED 1401plus, Cambridge Electronic Design). RNm discharge, behavioral marker signals, and the animals' movements were recorded by a custom video display system that enabled us to record a histogram of discharge rate and behavioral marker signals in the same video image as the animal's limb.
Data analysis
All analyses were performed on data for reaches to the upper target. Amplitude and timing of task-related modulations in discharge rate of RNm neurons and EMG activity of forelimb muscles were quantified during the period from 0.25 s before reach onset to 0.25 s following reach end because most neurons and muscles attained their largest task-related modulations in activity during this period. Task-related modulations in neuronal discharge rate and EMG activity during individual trials were quantified by calculating the average discharge rate or EMG activity over a 100-ms window that was moved, 6 ms at a time, between the start and end times of the analysis interval. Mean peak discharge modulation was based on the 100-ms window with the highest discharge rate and was defined by subtracting from this rate the average rate during a 0.5-s interval starting 1.0 s prior to movement cue onset during which the animal sat quietly with the hand at the waist. Individual trial measurements of mean peak discharge modulation and time of peak modulation for each neuron and muscle tested during performance of the whole-hand and precision tasks were evaluated by student t-statistics (P < 0.05).
Task performance was analyzed kinematically from videotaped images of
the moving limb. Joint angles were calculated from the x-y
coordinates of the head of the humerus, the rotation point of the
elbow, the proximal end of the carpals, the proximal phalanges, and the
proximal interphalanges. Relations between RNm discharge and kinematic
variables of the MCP, wrist, elbow, and shoulder joints were evaluated
with correlation and regression analyses similar to those used by
Gibson et al. (1985b). These analyses, based on
individual trial records of the cumulative sum (cusum) and joint
angles, provide comparisons of burst onset latency versus movement
onset latency, burst offset latency versus movement offset latency,
number of spikes in the burst versus movement amplitude, and frequency
within bursts versus movement velocity. A detailed description is
provided in Van Kan and McCurdy (2001).
Between-task differences in timing of MCP extension and wrist flexion near reach end, and of elbow extension and shoulder flexion, were determined as follows. Plots of joint angle measurements versus time for individual trials of task performance were fitted with, and were well described by, polynomial functions using automated curve fitting software (TableCurve 2D, SPSS). The coefficient of determination (R2), a measure of the goodness-of-fit, and the order of the equations fitted were similar for the whole-hand (n = 47) and precision trials (n = 46) analyzed. Figure 2 shows time plots of discharge rate (A), cusum (B), and simultaneously recorded joint angles (C-F) for an individual trial of the whole-hand task and an individual trial of the precision task. The times of peak discharge (A) and the times of peak velocity of joint rotation (C-F) are indicated by vertical lines. The times of peak velocity were used as measures of times of joint rotation for quantitative analysis. Between-task differences in times of peak RNm discharge, in times of joint rotations, and in times of peak EMG activity of forelimb muscles were evaluated by Student's t-statistics.
|
| |
RESULTS |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
RNm discharge during reaching with different types of grasp
This report is based on 67 of 157 forelimb RNm neurons for which
complete data sets were obtained during performance of the whole-hand
and precision tasks. Data from the same neurons during the whole-hand
task have been included in a previous report (Van Kan and
McCurdy 2001). Most neurons discharged strongly during performance of both tasks (Figs. 3,
A-D, I, and J, and 8, A
and F). Mean peak discharge modulations for the neuronal
population tested did not differ between the whole-hand versus
precision tasks [131 ± 52 (SD) imp/s vs. 128 ± 53 imp/s,
respectively; Fig. 3H]. The majority (66%, 44/67) of
neurons had discharge modulations that were >100 imp/s for both tasks;
only 15% (10/67) of neurons had discharge modulations that were <100
imp/s for both tasks (Fig. 3H). Few neurons (13%, 9/67) had
discharge modulations that differed by 50 imp/s or more between the two
tasks (Fig. 3, E-G).
|
The times of peak discharge of the 67 forelimb RNm neurons tested were distributed throughout the period of limb transport for both the whole-hand and precision tasks (Fig. 4A, between the dotted lines). Most neurons (73% whole-hand, 64% precision) attained peak discharge within the interval extending from 30 to 100% of the period of limb transport, which corresponded to the second half of the reach trajectory as is illustrated in Fig. 3, I and J. Between-task comparisons showed that, on average, the population of forelimb RNm neurons tested attained peak discharge later during the precision than whole-hand task. The between-task differences in mean times of peak discharge relative to reach onset were significant (Fig. 4A; precision, 0.28 ± 0.11 vs. whole-hand, 0.20 ± 0.11 s, Student's t statistics, P < 0.05). Each of the data points in Fig. 4A compares the mean time of peak discharge of an individual neuron during the precision versus whole-hand task. Seventy-two percent (48/67) of data points lie below the diagonal, indicating that the neurons attained peak discharge during a later phase of the precision than whole-hand task.
|
In summary, the magnitude and timing of discharge of the forelimb RNm neurons tested during the whole-hand and precision tasks are consistent with the hypothesis that most of the neurons contribute to behavioral features that are common to the two tasks but occur at different phases of each of the tasks.
Relations between RNm discharge and joint rotations during reaching to grasp
We used two methods to examine relations between neuronal
discharge and kinematic variables during individual trials of the whole-hand and precision tasks. One method focused on correlations between parameters of RNm discharge and the duration, amplitude, and
velocity of rotation of MCP, wrist, elbow, and shoulder joints for each
of the tasks. The second method compared between-task differences in
timing of peak neuronal discharge to between-task differences in timing
of rotations of MCP, wrist, elbow, and shoulder joints. Both methods
were applied to single-unit discharge and kinematic data from the same
individual trials of the whole-hand and/or precision tasks for a subset
of 12 (18%) of the 67 forelimb RNm neurons tested. Selection of the 12 neurons for analysis was based on 3 criteria: 1) the neurons
were tested during both the whole-hand and precision tasks,
2) they had mean peak discharge modulations >50 imp/s for
each of the tasks, and 3) sufficient kinematic data recorded
simultaneously with neuronal discharge. The latter criterion was most
often the limiting factor. In addition, data for the free-reach task
were available for 6 of the 12 neurons analyzed (Van Kan and
McCurdy 2001). Neurons were not selected for analysis based on
their movement-related discharge characteristics. Ten of the 12 neurons
were analyzed during both the whole-hand and precision tasks. The
remaining two neurons were analyzed during the whole-hand task only
because one neuron did not discharge consistently during the precision
task and the other neuron lacked a discreet peak in discharge during
the precision task.
Relations between RNm discharge and movement parameters
Parameters of reach-related RNm discharge were strongly and
frequently correlated with parameters of MCP extension. Figures 2,
A-C, and 5, A-C,
show examples of the detailed time course of MCP extension and
associated modulations in discharge rate for three RNm neurons during
individual trials of the whole-hand and precision tasks (see also Figs.
11 and 12 in Van Kan and McCurdy 2001). Correlations
between parameters of discharge and MCP extension for the same three
neurons are illustrated in Fig. 6. Burst
onset latency was strongly correlated with the onset latency of MCP extension (Fig. 6, A-C), and burst offset latency was
strongly correlated with the offset latency of MCP extension.
Regression lines typically had slopes near 1.0 and fell below the line
of simultaneity, indicating that burst onset led movement onset and burst offset led movement offset (Table 2). Furthermore, burst duration, discharge amplitude, and discharge frequency were strongly correlated with the duration (Fig. 6, D-F), amplitude (Fig.
6, G-I), and velocity of MCP extension (Fig. 6,
J-L), respectively. Relations between parameters of
discharge and joint rotations for the 12 neurons analyzed are
summarized quantitatively in Table 1.
Criteria for inclusion in Table 1 were 1) at least one of the three linear correlation coefficients (for duration, amplitude, or
velocity) was significant, and 2) the average linear
correlation coefficient was >0.5. Relations between discharge of one
neuron (B046) and elbow extension during the whole-hand task
were excluded from Table 1 because burst onset consistently led
movement onset by an overly long interval (>0.35 s). In summary,
discharge of 10 of 12 neurons analyzed was related to MCP extension
during both the whole-hand and precision tasks.
|
|
|
Relations between parameters of neuronal discharge and movements of the wrist, elbow, and shoulder joints were observed less frequently than relations for MCP extension. Using inclusion in Table 1 as criterion of relatedness, relations for MCP extension were observed for 63% of entries in Table 1 (22/35 rows), whereas only 9% (3/35 rows) of entries were observed for wrist flexion, 26% (9/35 rows) for elbow extension, and 3% (1/35 rows) for shoulder flexion. Relations between discharge and wrist rotations were infrequent because parameters of wrist flexion were typically weakly correlated with parameters of neuronal discharge during the whole-hand task. In addition, bursts of discharge did not consistently correspond to wrist flexion or extension during the precision task. Relations between discharge and elbow and shoulder rotations were infrequent because the timing of discharge relative to the timing of joint rotations was highly variable. For example, burst onset preceded or coincided with onset of shoulder flexion during trials of the whole-hand tasks (Fig. 5F, thin vertical lines) but discharge of the same neuron substantially lagged onset of shoulder flexion during trials of the precision task (Fig. 5F, thick vertical lines). In addition, multiple discharge bursts within a single trial were typically not each associated with either elbow extension or shoulder flexion (Fig. 2, E and F). Furthermore, the paucity of parametric correlations between RNm discharge and movements of wrist, elbow, and shoulder joints is not easily accounted for by differences in the ranges of movement parameters because the duration, amplitude, and velocity of joint rotations each varied over a greater than fivefold range. Although relations between parameters of discharge and elbow extension were observed less frequently than relations involving MCP extension, it is noteworthy that the strongest relations involving elbow extension were as robust as the strongest relations involving MCP extension (Table 2).
|
The strongest evidence for relatedness of neuronal discharge and
rotation of a given joint is that strong correlations are present
regardless of task. If discharge is causally related to rotation about
a given joint or combination of joints, then comparing correlations for
a given neuron across the whole-hand, precision and, when available,
free-reach tasks should identify the joint or combination of joints
most strongly influenced by the discharge (Van Kan and McCurdy
2001). The present results show that discharge of 10 of 12 neurons analyzed was related to MCP extension for both the whole-hand
and precision tasks and, when available, for the free-reach task. In
contrast, only 1 of 12 neurons was related to wrist flexion for both
the whole-hand and precision tasks, only 3 of 12 neurons were related
to elbow extension for both tasks, and none of the neurons were related
to shoulder flexion for both tasks. The neuron related to wrist flexion
during both the whole-hand and precision tasks and one of the three
neurons related to elbow extension for both tasks were related solely to MCP extension when tested during the free-reach task (Van Kan and McCurdy 2001), suggesting that for these two neurons
discharge was not causally related to wrist flexion and/or elbow
extension. In conclusion, based on consistency of relations between
neuronal discharge and kinematics across the free-reach (Van Kan
and McCurdy 2001), whole-hand, and precision tasks, 10 of 12 neurons analyzed were related to MCP extension, and 2 of 12 neurons
were related to elbow extension and MCP extension.
Relations between RNm discharge and joint rotations were also evaluated by comparing the times of peak discharge to the times of joint rotations. The measure of timing of joint rotation used is illustrated in Fig. 2. The scatter plots in Fig. 7, A-C, compare the times of peak RNm discharge to the times of MCP extension, elbow extension, and shoulder flexion, respectively, for the 10 neurons analyzed during both the whole-hand and precision tasks. Each data point represents data from an individual trial, and each plot includes a total of 45 whole-hand trials and 45 precision trials combined from all 10 neurons analyzed. The times of peak discharge were more strongly correlated to the times of MCP extension than to the times of elbow extension or shoulder flexion (r = 0.78 vs. 0.52 and 0.16, respectively).
|
Between-task differences in times of peak RNm discharge and joint rotations
The observation that most RNm neurons attained peak discharge during a later phase of the precision than whole-hand task (Fig. 4A) provided another opportunity to evaluate temporal relations between discharge of a given RNm neuron and rotations of MCP, elbow, and shoulder joints during reaching to grasp. Although, on average, the population of 67 forelimb RNm neurons attained peak discharge later during the precision than whole-hand task when aligned on reach onset (Fig. 4A), the mean times of peak discharge during the two tasks did not differ when aligned on reach end (Fig. 4B, P > 0.05). We compared the times of peak neuronal discharge to times of joint rotations under the two alignment conditions to determine which joints mimicked the pattern of neuronal discharge given each alignment condition.
The times of peak neuronal discharge were compared with the times of joint rotations for the 10 RNm neurons analyzed during both the whole-hand and precision tasks. Sets of individual trial records of discharge from one neuron and associated joint rotations are shown in Fig. 8 for two conditions of alignment: aligned on reach onset (Fig. 8, A-E) and aligned on reach end (Fig. 8, F-J). RNm discharge as well as MCP extension were significantly (P < 0.05) delayed during the precision as compared with the whole-hand task when records were aligned on reach onset (Fig. 8, A and B, black vs. green lines and tic marks), but between-task differences in times of peak discharge as well as MCP extension were not significant when aligned on reach end (Fig. 8, F and G). Elbow extension was also delayed during the precision task when records were aligned on reach onset (Fig. 8D), but, unlike RNm discharge, elbow extension occurred consistently earlier during the precision than whole-hand task when records were aligned on reach end (Fig. 8I). Between-task differences in the times of shoulder flexion were minimal when records were aligned on reach onset (Fig. 8E), but shoulder flexion occurred consistently earlier during the precision than whole-hand task when records were aligned on reach end (Fig. 8J), which differs from RNm discharge. Thus under both alignment conditions, the between-task differences in times of peak neuronal discharge corresponded best to between-task differences in times of MCP extension and, to a lesser extent, elbow extension.
|
A similar result was observed when data from all 10 neurons was analyzed. Figure 9 shows data aligned on reach onset (Fig. 9, A-D) and aligned on reach end (Fig. 9, F-I). Each data point plots the time of peak discharge (Fig. 9, A and F) or the time of joint rotation (Fig. 9, B-D and G-I) from a whole-hand trial versus a precision trial, and each plot includes 45 pairs of a total of 47 whole-hand and 46 precision trials combined from all 10 neurons. Values of t refer to the results of paired Student's t statistics for the comparisons in each of the plots. The times of peak RNm discharge for precision and whole-hand trials varied throughout the period of limb transport, as was true for the neuronal population (Fig. 9A vs. Fig. 4A). The wide range of times of peak discharge corresponds better to the wide range of times of MCP extension (Fig. 9, A vs. B) than to the narrow ranges of times of elbow extension and shoulder flexion (Fig. 9, A vs. C and D, respectively). The four groups of measurements of between-task differences in timing (for peak RNm discharge and rotations of MCP, elbow, and shoulder joints) differed significantly when records were aligned on reach onset (Fig. 9E, Kruskal-Wallis 1-way ANOVA on ranks, H = 57.618, DF = 3, P < 0.001) as well as when records were aligned on reach end (Fig. 9J, Kruskal-Wallis 1-way ANOVA on ranks, H = 56.762, DF = 3, P < 0.001). All-pairwise multiple comparison procedures (Dunn's method) indicated that between-task differences in times of peak RNm discharge differed from between-task differences in times of elbow and shoulder rotations but did not differ from between-task differences in times of MCP extension, regardless of alignment condition. In summary, the times of peak discharge of the 10 forelimb RNm neurons analyzed during both the whole-hand and precision tasks fit better with the times of MCP extension than with the times of elbow extension and shoulder flexion.
|
Activity of forelimb muscles during reaching with different types of grasp
The interval within which most RNm neurons attained peak discharge (Fig. 4A) included MCP extension (Fig. 1C, frames 7-13, and D, frames 9-17) and rotations of the wrist, elbow, and shoulder. Many distal as well as proximal forelimb muscles were activated strongly during performance of both the whole-hand and precision tasks (Fig. 10). The between-task differences in times of peak discharge of RNm neurons agreed better with the between-task differences in times of peak EMG activity of distal extensor muscles than with those of proximal muscles. Figure 11 shows scatter plots of individual trial measurements of times of peak EMG activity for a few representative muscles during the whole-hand versus precision tasks (Fig. 11, A-D and F-I). Histograms summarize the between-task differences in times of peak EMG activity (Fig. 11, E and J). Peak EMG activity of EDC was significantly (P < 0.05) delayed during the precision as compared with the whole-hand task when records were aligned on reach onset (Fig. 11, A and E) but between-task differences in times of peak EMG activity were not significant when aligned on reach end (Fig. 11, F and J). This result is consistent with the pattern of between-task differences in times of peak discharge of most RNm neurons (Figs. 4, A and B, and 9, A and F). Peak EMG activity of ECR also occurred significantly (P < 0.05) later during the precision as compared with the whole-hand task when records were aligned on reach onset (Fig. 11, B and E) but occurred significantly earlier during the precision than whole-hand task when the same data were aligned on reach end (Fig. 11, G and J). Peak EMG activity of proximal muscles, such as triceps and spinodeltoid, was simultaneous with or occurred earlier during the precision than whole-hand task regardless of alignment condition (Fig. 11, C-E and H-I), which does not agree with the between-task differences in times of peak discharge of most RNm neurons. In conclusion, the observed patterns of muscle activation during the whole-hand and precision tasks are consistent with a role for RNm in producing MCP extension at the appropriate phase of limb transport.
|
|
| |
DISCUSSION |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
One aspect of hand use that is strongly related to the
discharge of at least some forelimb RNm neurons during reaching is extension of the MCP joints (Van Kan and McCurdy 2001).
The results of the present analyses extend these observations in two
ways. First, testing the same RNm neurons during reaches with different types of grasp revealed that, for most neurons, the large discharge modulations during reaching to grasp do not depend on the specific type
of grasp. Second, the dissociation of MCP extension from movements of
the elbow and shoulder joints between the tasks provided an opportunity
to evaluate relations between discharge of individual RNm neurons and
preshaping versus transport of the hand during reaching to grasp.
Kinematic analyses on a subset of neurons indicate that discharge was
time locked most frequently to MCP extension during both tasks. Because
MCP extension occurred during a later phase of the precision than
whole-hand task, the result implies that the timing of RNm's
contribution to hand preshaping relative to the phase of limb transport
varies with the behavioral requirements of the task. The results
support the hypothesis that cerebellar output via RNm is important for
appropriately timed MCP extension during coordinated whole-limb movements.
The relations between neuronal discharge and MCP extension that
we observed for a subset of RNm neurons are consistent with previous
analyses of RNm discharge during a task that required a monkey to close
individual finger switches while the forearm was supported (Houk
et al. 1988; Kennedy 1987). Although the
monkey's instructions were to operate individual switches using the
thumb, index, middle, and ring fingers in a fixed sequence, only a
small proportion (17%) of RNm neurons tested showed parametric
relations between discharge and movement of an individual digit
(Kennedy 1987). Discharge of most neurons may have been
related to grouped finger movements that positioned the hand and
fingers above the switches rather than movements of individual fingers
in isolation (Houk et al. 1988). The present result that
most RNm neurons discharged well during both the whole-hand and
precision tasks are consistent with Houk et al.'s conclusion. The
combined results support a role for RNm in grouped finger extension
during various behaviors.
In monkeys, hand use depends on corticospinal and rubrospinal
neurons, the main components of the lateral system of descending motor
pathways in this species (Kuypers 1982). Control of
relatively independent finger movements critically depends on motor
cortex (for review: Porter and Lemon 1993). Grouped
finger movements are preserved in monkeys following bilateral
pyramidotomy, but permanent loss of hand use results from additional
lesions of the RNm or rubrospinal tract (Lawrence and Kuypers
1968a,b), indicating that preserved hand use following damage
to the corticospinal system derives largely from the rubrospinal
system. Currently available data support a role for rubrospinal neurons
in the control of muscle synergies that produce grouped finger
extension on which corticospinal neurons, particularly
corticomotoneuronal neurons, superimpose control for individuated
finger movements (cf. model 3 in Schieber
1990).
Kinematic analyses of subsets of RNm neurons during various
reaching tasks in the present and our previous study (Van Kan and McCurdy 2001) combined revealed that relations between RNm discharge and proximal joint rotations are coupled to relations involving MCP extension, although analysis of a larger sample of
neurons might have included neurons with proximal preference. The
results are in good agreement with studies of movement specificity of
RNm neurons using multiple device testing (Gibson et al.
1985a,b). Gibson and his colleagues demonstrated that discharge
of RNm neurons was preferentially related to use of distal devices;
little evidence was obtained for relations between RNm discharge and
isolated proximal joint rotations. Our findings are also consistent
with behavioral observations following bilateral pyramidotomy in
monkeys (Lawrence and Kuypers 1968a). The monkeys
extended their digits while reaching for food but had difficulties
opening the hand at the mouth to feed themselves, and this deficit
persisted for the entire survival period. Because the rubrospinal
system is the remaining intact lateral system available for controlling independent limb movements, the observations suggest that in monkeys RNm may be more important for grouped digit extension that is part of a
coordinated whole-limb movement rather than grouped digit extension in isolation.
RNm neurons may contribute to muscle synergies that include digit
extensor and/or combinations of other forelimb muscles
(Belhaj-Saif et al. 1998; Miller et al.
1993; Sinkjaer et al. 1995). Grouped finger
extension to preshape the hand during reaching to grasp typically is
combined with activation of muscle synergies that include muscles of
the entire forelimb. It is noteworthy that even isolated MCP extension
elicited by a metacarpal device, which included a support for the
forearm, is accompanied by activation of wrist extensor (ECU), elbow
extensor (TRI), and shoulder flexor (anterior deltoid) muscles (Fig. 1 in Miller and Houk 1995), and that RNm discharge was
related to EMG activity of all of these muscles during use of the
metacarpal device (Fig. 2 in Miller and Houk 1995).
These observations suggest that even relatively pure MCP extension
movements may be produced by synergies of digit extensor muscles
combined with various other forelimb muscles.
The results reviewed in the preceding text raise an important
question: what is the functional significance of RNm's contribution to
proximal forelimb muscles? One possibility is that RNm's influences on
proximal muscles may provide the stability of the limb that is required
for MCP extension to be accurate and appropriately timed with respect
to the phase of limb transport. Interestingly, behavioral studies
following red nucleus lesions in rats have provided evidence that the
red nucleus in this species may provide "the tonus or supporting
frame work" that stabilizes the limb during digit extension and
grasping (Whishaw and Gorny 1996).
Timing of RNm's contribution to hand preshaping during reaching to grasp
The rubrospinal system may command muscle synergies that
preshape the hand at the appropriate phase of limb transport so that preshaping will be coordinated with transport of the hand. The idea
that the timing of discharge of red nucleus neurons may be important
for coordinated reach-to-grasp movements has been advanced by Jarratt
and Hyland on the basis of single-unit recording studies of red nucleus
neurons in rats during a reach-to-grasp task (Jarratt and Hyland
1999). The times of peak discharge of RNm neurons in the
monkey, like the onset times of discharge of rat red nucleus neurons,
were spread throughout the period of limb transport, and
reach-to-grasp-related discharge was well synchronized with reach end
in both species. The results from our between-task comparisons of a
subset of RNm neurons extend these observations by showing that
neuronal discharge was time locked to MCP extension and, to a lesser
extent, elbow extension. Because, in the monkey, the rubrospinal tract
is an important pathway by which cerebellar output influences spinal
circuitry (Keifer and Houk 1994), our findings implicate
cerebellar output in timing digit extension relative to limb transport.
The hypothesis that cerebellar output is important for appropriately
timed MCP extension has received considerable attention from recent
psychophysical studies of people with cerebellar deficits (Rand
et al. 2000; Timmann et al. 1999; Zackowski et al. 1999).
In conclusion, a given RNm neuron may influence muscles in a weighted, neuron-specific fashion according to that neuron's unique combination of spinal terminations, and each neuron may be activated strongly when the muscles that are influenced by its spinal projections are involved in the production of the movement. It is interesting to speculate that the ensemble of muscle synergies commanded by populations of RNm neurons may be the neural basis for producing appropriately timed MCP extension. The present result that the times of peak discharge of RNm neurons during reaching to grasp are widely distributed throughout the interval of limb transport indicates that a given neuron contributes maximally to its muscle synergy at a specific phase of task performance. Perhaps RNm's contribution to appropriately timed MCP extension involves activating a large number of RNm neurons in a specific temporal sequence.
| |
ACKNOWLEDGMENTS |
|---|
We thank M. Hickey for development of data acquisition software and C. Boyce, D. Jacobson, and J. Ruhland for assistance in data collection and analysis.
This work was supported by National Institute of Neurological Disorders and Stroke Grant NS-24042.
| |
FOOTNOTES |
|---|
Address for reprint requests: P.L.E. Van Kan, Dept. of Kinesiology, Rm. 3195, Medical Sciences Center, 1300 University Ave., Madison, WI 53706-1532 (E-mail: vankan{at}education.wisc.edu).
Received 17 January 2001; accepted in final form 24 October 2001.
| |
REFERENCES |
|---|
|
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
|
|---|
This article has been cited by other articles:
|
|
 |
|
  N. Krouchev, J. F. Kalaska, and T. Drew Sequential Activation of Muscle Synergies During Locomotion in the Intact Cat as Revealed by Cluster Analysis and Direct Decomposition J Neurophysiol, October 1, 2006; 96(4): 1991 - 2010. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
|
) and precision (
) trials. The curves
drawn through the data points represent polynomial functions fitted to
the data points (whole-hand trial, thin lines; precision trial, thick
lines). The coefficient of determination
(R2) and the order of the equation
fitted to the whole-hand and precision data, respectively, are
C, R2 = 0.891, 0.956, order = 13, 14; D,
R2 = 0.975, 0.881, order = 15, 12;
E, R2 = 0.949, 0.988, order = 11, 13; F,
R2 = 0.998, 0.999, order = 12, 12. The time of peak velocity was taken as the measure of timing of joint
rotation. These times are indicated by vertical lines (whole-hand
trial, thin lines; precision trial, thick lines).
