| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Journal of Neurophysiology Vol. 80 No. 5 November 1998, pp. 2475-2494
Copyright ©1998 by the American Physiological Society
Department of Physiology and Biophysics and Regional Primate Research Center, University of Washington, Seattle, Washington 98195
 |
ABSTRACT |
|---|
 Abstract
 Introduction
Methods
Results
Discussion
References
|
|---|
Perlmutter, Steve I., Marc A. Maier, and Eberhard E. Fetz. Activity of spinal interneurons and their effects on forearm muscles during voluntary wrist movements in the monkey. J. Neurophysiol. 80: 2475-2494, 1998. We studied the activity of 577 neurons in the C6-T1 spinal cord of three awake macaque monkeys while they generated visually guided, isometric flexion/extension torques about the wrist. Spike-triggered averaging of electromyographic activity (EMG) identified the units' correlational linkages with 
12 forearm muscles. One hundred interneurons produced changes in the level of average postspike EMG with onset latencies consistent with mono- or oligosynaptic connections to motoneurons; these were classified as premotor interneurons (PreM-INs). Most PreM-INs (82%) produced postspike facilitations in forearm muscles. Earlier spike-related features, often beginning before the trigger spike, were seen in spike-triggered averages from 72 neurons. Postspike effects were present in one muscle for 64% of the PreM-INs. Neurons with divergent linkages to larger "muscle fields" usually generated postspike effects in synergistic muscles. Fifty-eight percent of the PreM-INs had postspike effects in flexor muscles only and 29% in extensor muscles only. Postspike effects were distributed relatively evenly among the primary flexor and extensor muscles studied. The mean percent change in EMG level from baseline and the mean onset latencies for postspike facilitations and postspike suppressions were similar. PreM-INs exhibited a variety of response patterns during the generation of isometric wrist torque. The response patterns and output effects of 24% of the PreM-INs were consistent with a strict reciprocal organization of flexor and extensor muscle control. For another 60% of the PreM-INs, there was a congruent relation between activity and output effects for only one direction of torque production. These neurons were active for both flexion and extension torques, including 37 neurons that exhibited bidirectional increases in discharge rate. The relatively small number of postspike suppressions observed suggests that inhibitory interneurons were silent when their target muscles were recruited. Compared with premotor neurons in the motor cortex, the red nucleus and the C8-T1 dorsal root ganglia, spinal PreM-INs affected flexor muscles in greater proportions and had smaller muscle fields. The magnitudes of postspike facilitations were similar in all premotor populations. Bidirectional activity, common for PreM-INs, was rare for corticomotoneuronal and premotor dorsal root ganglion cells, which discharge only for torques in their preferred direction.
Supraspinal control of voluntary limb movements is mediated in large part by descending projections to spinal interneurons. Anatomic studies have demonstrated substantially denser projections from cortical and brain stem motor areas to intermediate spinal laminae than to motor nuclei (Armand et al. 1997 Spinal interneurons were studied in three male macaque monkeys (2 Macaca nemestrina and 1 M. mulatta). The experiments were approved by the Animal Care Committee at the University of Washington. All procedures conformed to the National Institutes of Health "Guide for the Care and Use of Laboratory Animals."
Behavioral paradigm
The monkeys were trained to generate isometric, ramp-and-hold flexion and extension torques about the wrist. Torque controlled the position of a cursor on a video screen in front of the animal. To obtain a fruit juice or applesauce reward, the monkeys held the cursor for 1.5-2 s within a target window that specified a given flexion or extension torque
Surgical implants
All surgeries were performed with the use of aseptic techniques with the animals under 1-1.5% halothane or isoflurane anesthesia after training was completed. Atropine sulfate was administered preoperatively; antibiotics (cefazolin, 25 mg/kg) and analgesics (ketoprofen, 5 mg/kg) were given postoperatively. Head stabilization lugs were cemented to the exposed skull with dental acrylic, which was anchored to the bone with screws.
Recording procedure
During recording sessions, the head and vertebral implants were secured to the primate chair with nylon screws. The implants were held firmly, but both restraints were somewhat flexible, allowing the animal to make small postural adjustments. Monkey B received occasional intramuscular injections of diazepam (2-4 mg) to eliminate excessive movements that jeopardized stable neuronal recordings. These procedures were well tolerated: all monkeys moved voluntarily from their home cages and seated themselves in the primate chair for each day's session.
Spike-triggered averaging of electromyographic activity
Action potentials of individual spinal units were monitored and pulses generated by a time-amplitude window discriminator (Bak Electronics). Electromyographic activity (EMG) was recorded differentially and band-pass filtered (30 Hz to 3 kHz). The action potentials of isolated neurons were used to trigger averages of EMG to identify functional connections to muscles (Fetz and Cheney 1980
Cross-talk between muscle recordings
Electrical cross-talk between electrodes in two muscles can produce significant features in the STA of one muscle that are actually due to spike-related effects in the other muscle (Fetz and Cheney 1980 Response averages
To characterize task-related modulation, we averaged the activity of isolated units over many trials of ramp-and-hold torque production. Histograms of neuronal activity and analog averages of rectified EMG and torque (8-ms bins) were compiled separately for flexion and extension trials, aligned on the onset of the torque ramps. Torque in the flexion direction is represented in all figures as a positive deflection; extension torques are shown as negative. Response patterns were assigned to describe the profile of averaged neural activity (Maier et al. 1998 Microstimulation
Microstimuli were frequently delivered through the recording electrode to determine output effects evoked at localized sites within the cord. Single-pulse stimuli (200-µs biphasic pulses) of Histological procedures
Near the conclusion of each experiment, we made small electrolytic lesions at several sites within the spinal cord by passing DC current (10-30 µA for 1-2 min) through the recording electrode. Animals were euthanized with pentobarbital sodium (50 mg/kg iv) and perfused with 10% Formalin or 4% paraformaldehyde. The spinal cord was cut into 75-µm sections, which were stained alternately with cresyl violet and Perls iron reaction for the detection of iron deposits made with Elgiloy electrodes.
Neurons were recorded in the C6-T1 spinal segments of three monkeys. Cells exhibiting spontaneous activity while the monkey sat quietly with no visual targets displayed were encountered at all depths of the gray matter. Firing rates were generally higher for dorsal horn cells and ranged widely for cells in the intermediate layers. Spontaneous firing rates ranged from <1 to >100 spikes/s. A few neurons exhibited activity modulated in phase with the monkey's respiratory or cardiac cycles.
Spike-related effects on muscles
For 134 neurons, STAs revealed significant changes, above or below baseline fluctuations, of spike-related EMG from forearm muscles. Firing rate was modulated during the task for 129 of the neurons and unmodulated for 5. The onset latencies of the STA features were distributed broadly, ranging from 34 ms before to 24 ms after the trigger spike. Postspike effects were defined as sustained changes of average EMG level with onset latencies consistent with mono- or oligosynaptic connections to motoneurons. Early spike-related effects with shorter latencies were attributed to other mechanisms (see DISCUSSION).
Postspike effects
One-hundred neurons were classified as PreM-INs on the basis of appropriately timed postspike facilitation or suppression of EMG in at least one muscle. Eighteen of these PreM-INs met the stringent criterion for classification as last-order interneurons. The population of PreM-INs produced 161 postspike effects, 20 of which had latencies of <4.5 ms. Neurons classified as last-order interneurons had properties similar to those of other PreM-INs, and the populations are combined for most of the subsequent description of results. Last-order and other PreM-INs are described separately only when their properties differed significantly from each other.
Distribution of postspike effects in muscles
Postspike effects were present in only one of the coactivated muscles for 64 of the 100 PreM-INs (e.g., Figs. 6 and 8). The other PreM-INs had larger "muscle fields": 22 had postspike effects in 2 muscles (e.g., Fig. 5) and 14 produced effects in
Parameters of postspike effects
Postspike effects had a wide distribution of onset times, with a mean latency of 7.4 ± 3.5 (SD) ms (Fig. 11A). Facilitatory and suppressive effects had similar onsets for neurons classified as last-order interneurons, but postspike facilitations tended to have earlier onsets than postspike suppressions for other PreM-INs (Table 1). Seventy-nine percent of all postspike effects peaked within 3.5 ms of onset and had durations of <8.0 ms (Fig. 11, E and F). Some effects lasted >10 ms, and in a few cases, the STAs showed two clear peaks in the averaged EMG. Postspike features with long durations may reflect PreM-IN connections to fast and slow motor units or the combined effects of mono- and oligosynaptic connections to motoneurons. Neurons classified as last-order interneurons exhibited postspike effects with longer durations in flexor muscles (but not in extensor muscles) and longer rise times than other PreM-INs. There were no other significant differences in the timing of postspike effects in flexor and extensor muscles (Table 1).
Response averages and relation to output effects
PreM-INs exhibited various firing patterns during the generation of isometric flexion/extension torques, including tonic (Figs. 5 and 7), phasic (Fig. 6), phasic-tonic (Fig. 10), and decrementing (Fig. 8) responses. Neuronal firing properties are discussed in detail in the companion paper (Maier et al. 1998
Location of PreM-INs
Recordings were made primarily in C8-T1 in monkey B, in caudal C7-C8 in monkey R, and in C6-C7 in monkey W. The approximate position of PreM-INs in the transverse plane of the spinal cord was estimated for monkeys B and W (Fig. 13), in which iron deposits were recovered in histological sections. Each cell's mediolateral position was determined by the relation of the recording track to the lesion tracks. We could not reliably estimate the dorsoventral position of neurons on the basis of these lesions because we did not have a consistent dorsoventral reference from track to track. Instead, we estimated the depth of the neuron by mapping its position relative to the depth of the first and last clearly recorded somatic activity within the track onto the outline of the gray matter at the appropriate mediolateral coordinate.
Spike-related effects with early onsets
TIMING.
The STAs of 72 neurons exhibited spike-related effects with onset latencies of <3.5 ms. A total of 138 early effects were seen.
PARAMETERS AND DISTRIBUTION.
Early spike-related effects had, on average, slower rise times, longer durations (Fig. 11, E and F), and larger magnitudes than postspike effects (P < 0.01; Table 1). The rise times of early-posttrigger effects (mean = 4.1 ms) tended to be shorter than those of synchrony effects (mean = 6.5 ms) but longer than those of postspike effects (2.1 ms).
Our goals in this study were to provide the first direct information on two functionally significant properties of spinal interneurons: their effects on muscle activation and their activity patterns during normal voluntary movements. In this paper, we focused on the first objective by characterizing the effects of individual PreM-INs on the firing probability of forearm muscles. In the companion paper, we addressed the second issue in more detail. Here, we compare our results on the postspike effects of spinal interneurons with previous data from supraspinal and afferent premotor neurons obtained under similar conditions and with data from spinal interneurons obtained in anesthetized or decerebrate animals. Then we discuss the implications of these results for the control of motoneuron activity during voluntary wrist movements. Before addressing these issues, we consider the synaptic linkages that produce the postspike effects.
Synaptic pathways mediating postspike effects
By our estimate, the shortest possible latency of STA features that could be mediated by monosynaptic connections of spinal interneurons to fast conducting motor units is 3.5 ms. This estimate is based on the latency of stimulus-evoked effects (Fig. 4) and is consistent with available data on the timing of synaptic transmission and spike initiation in motoneurons (Gustaffson and Jankowska 1976 Distribution of postspike effects in muscles
The majority of PreM-INs had postspike effects in only one target muscle. This was somewhat surprising in light of previous data from spinal interneurons and other premotor neurons.
Activity patterns of Prem-INs
The discharge patterns of spinal interneurons during isometric torque production at the wrist share many features with those of motoneurons of primary wrist flexor and extensor muscles (Maier et al. 1998 Early spike-related effects in EMG
Several mechanisms can account for STA features with onset times that are too early to be mediated by monosynaptic connections from the trigger neuron to motoneurons (Fig. 18A), as shown schematically in Fig. 18, B-E.
SYNCHRONY EFFECTS.
The majority of early effects began before and ended after the occurrence of the triggering action potential. Such effects are similar to the zero-delay peaks found in cross-correlograms of the activity of neighboring neurons in many areas of the CNS (Fetz et al. 1991 OTHER EARLY EFFECTS.
Early STA features that do not straddle the trigger time probably are not due to the synchronous firing of the recorded cell and premotor neurons, as this would require highly unlikely relations between the firing of these neurons. For example, effects with onsets earlier than 10 ms before the trigger time would require common inputs that reliably drive neurons through three or more synapses.
Control of motoneuron activity
The firing pattern and output effects of 24 PreM-INs were well matched to the alternating activation of forearm muscle motoneurons during the task: they facilitated primary flexor or extensor muscles, were active when their target muscles were active, and were silent for torques in the opposite direction. However, 71 PreM-INs with postspike effects in primary wrist movers were active for both directions of torque and/or had postspike effects that were not simply correlated with the cell's firing pattern during the task (Fig. 12). Responses often matched target muscle activity for torques in one direction, but for opposite torques these neurons either facilitated muscles that were silent or inhibited muscles that were actively recruited.
INTRODUCTION
Abstract
Introduction
Methods
Results
Discussion
References
; Bortoff and Strick 1993; Cheema et al. 1984; Dum and Strick 1996; Kuypers 1981; Ralston and Ralston 1985; Robinson et al. 1987). Spinal interneurons appear to integrate convergent supraspinal and peripheral afferent information and distribute appropriate signals to motoneurons to activate muscles in coordinated patterns (Baldissera et al. 1981; Jankowska 1992).
), virtually all physiological investigations have been performed in anesthetized or reduced preparations. Of the few descriptions of interneuron properties in awake animals, most concern sensory responses (Bromberg and Fetz 1977; Cleland and Hoffer 1986; Collins et al. 1990; Courtney and Fetz 1973; Soja et al. 1996; Sorkin et al. 1988; Wall et al. 1967). Other studies have reported the activity of some spinal interneurons during fictive locomotion and scratching (Berkowitz and Stein 1994; Jordan 1991; Terakado and Yamaguchi 1990) or inferred the activity of interneurons during normal movements from analyses of muscle responses, limb kinematics, or force (e.g., Nielsen and Kagamihara 1992; Pierrot-Deseilligny 1989; Shefner et al. 1992; Stein and Kearney 1995).
; Hongo et al. 1989; Jankowska and Roberts 1972; Rudomin et al. 1987; review in Cheney et al. 1991). Consequently, despite the crucial role of spinal interneurons in the control of voluntary movements, many basic questions concerning their function during normal behaviors remain unanswered.
; Courtney and Fetz 1973; Flament et al. 1992). This paper focuses on the output properties of spinal premotor interneurons (PreM-INs), identified with spike-triggered averages of muscle activity (Fetz and Cheney 1980). The companion paper (Maier et al. 1998) describes the responses of PreM-INs and other spinal neurons during movement and their relation to force produced at the wrist.
METHODS
Abstract
Introduction
Methods
Results
Discussion
References
0.2 N-m. For animals B and R, the target windows alternated back and forth between fixed flexion and extension levels. These animals produced ramp-and-hold torques in each direction starting from a maintained torque in the opposite direction (Fig. 1A). Monkey W began each trial by positioning the cursor within a center target window, corresponding to zero torque, for 1-4 s. Then a target window appeared at one of six positions randomly selected from three flexion and three extension levels. Monkey W produced ramp-and-hold torques in each direction starting from rest (Fig. 1B).
View larger version (45K):
[in a new window]
FIG. 1.
Behavioral paradigms and discharge patterns of 2 spinal interneurons during the alternating (A) and the center-out (B) tasks. Traces show simultaneously recorded action potentials of the interneuron, isometric torque trajectory, and electromyogram (EMG) of 1 flexor and 1 extensor muscle.
View larger version (45K):
[in a new window]
FIG. 2.
Schematic diagrams of the surgical procedure for implanting a stainless steel chamber over a unilateral laminectomy of the C5-T1 vertebrae and of the system for advancing metal recording electrodes through the chamber into the spinal cord.
(Fig. 2).
). Spike-triggered averages (STAs) of full-wave rectified and occasionally unrectified EMGs, digitized at 4 kHz, included 
20 ms before the trigger time. Only interneurons with STAs compiled from >2,000 triggers were included in the data set; previous experience has demonstrated that such STAs show reproducible effects (Fetz and Cheney 1980; Fetz et al. 1989; Flament et al. 1992; Mewes and Cheney 1991). STAs from putative motoneurons revealed consistent features in averages from less spikes, often with only a few dozen triggers (Maier et al. 1998).
Facilitatory effects were characterized by the mean and peak percent increase in EMG (MPI and PPI) and suppressive effects by the mean and peak percent decrease (MPD and PPD).
View larger version (35K):
[in a new window]
FIG. 3.
Methods for quantifying 3 types of spike-related effects in spike-triggered averages (STAs) of rectified EMG. A: postspike facilitation (PSF). B: synchrony suppression (SS). C: PSF superimposed on synchrony facilitation (SF). Top: action potential that triggered the STA in A. Significant effects had sustained increases or decreases in average EMG above or below the baseline mean ± 2 SD.
; Lemon et al. 1986). This cross-talk confounds estimates of the distribution of effects of individual interneurons to different muscles and was identified by cross-correlating muscle activity.
,b; DeLuca and Mambrito 1987). We chose the following criterion to distinguish synchronous firing of motoneurons from electrical cross-talk. All spike-related effects in muscles for which the amplitude ratio was >15% were excluded as redundant. Neurons for which STAs were compiled with three or more independent, coactive muscles are included in the data set (average was 4.6 independent, coactive muscles for both flexors and extensors).
).
50 µA were used to trigger averages of EMG. Stimulus-triggered averages were analyzed in the same manner as STAs for isolated units; however, stimulus-triggered averaging yielded reproducible poststimulus features with far fewer triggers, often <10.
RESULTS
Abstract
Introduction
Methods
Results
Discussion
References
) and are not considered further here.
30 µA is shown in Fig. 4B.
View larger version (32K):
[in a new window]
FIG. 4.
Determination of the latency criterion for postspike effects based on output effects elicited with microstimulation in the spinal cord. A: example of poststimulus effects in forearm muscles. B: histogram of onset latencies of effects elicited from 117 stimulation sites with intensities of 30 µA. C: comparison of onset latencies of the postspike effect of a putative motoneuron (4.0 ms) and the poststimulus effect elicited with 5-µA stimulation at the recording site (4.5 ms). Averages compiled with unrectified EMG show an apparent single motor unit potential. Records are 50 ms long. Difference in postspike and poststimulus latencies attributed to stimulus utilization time.
found that cat lumbar motoneurons typically were activated 0.2 ms after an extracellular stimulus, and we adopted this value as an estimate of stimulus utilization time for cervical motoneurons in the macaque.
), and 0.2 ms between the onset of the excitatory postsynaptic potential (EPSP) and spike initiation in the motoneuron (Cope et al. 1987; Gustaffson and Jankowska 1976). Thus we estimated that the synaptic effect of a last-order interneuron on motoneurons could produce postspike effects in STAs of EMG with latencies of 3.5 ms.
3.5 ms. We use the term "premotor" to indicate a functional linkage to motoneurons rather than a direct anatomic connection. The majority of postspike effects probably are produced by monosynaptic connections to motoneurons (Fetz and Cheney 1980). However, the range of conduction times for motor axons makes it impossible to determine with certainty if our longer latency postspike effects are mediated by mono- or oligosynaptic connections. A last-order interneuron with monosynaptic connections to slowly conducting motoneurons could produce postspike effects with onset latencies equal to, or longer than, those produced by a PreM-IN with disynaptic connections to very fast conducting motoneurons.
1 ms to the earliest possible latency for postspike effects on EMG (0.5-ms conduction along the PreM-IN axon, 0.3-ms synaptic delay, 0.2-ms EPSP-spike initiation in the interposed interneuron). Thus postspike effects with onset latencies of 3.5 and <4.5 ms were certainly due to monosynaptic connections to motoneurons. PreM-INs that produced a postspike effect in at least one muscle with a latency of <4.5 ms were classified as last-order interneurons. Although longer latency postspike effects could be mediated by oligosynaptic connections, the majority are likely to be monosynaptic (see DISCUSSION).
View larger version (28K):
[in a new window]
FIG. 5.
STAs and response averages of a facilitatory premotor interneuron (PreM-IN) located in lamina VII of caudal C8. Data were collected during the alternating task. Left: triggering action potential above the STAs of EMG in 3 independently recorded coactive muscles. *Muscles with significant postspike effects. STAs are plotted with the variances in baseline activity normalized to the STA of the muscle with the largest postspike effect (PT). Right: average activity of the neuron during flexion torques above the average torque trajectory (positive torque in flexion, negative torque in extension) and average responses of the coactive muscles. Neuron exhibited tonic firing during static torques, at a higher rate for flexion. Respective data for PT and flexor carpi ulnaris (FCU): mean percent increase (MPI), 5.0 and 3.9%; peak percent increase (PPI), 9.9 and 5.9%; onset latency, 3.6 and 3.5 ms; peak latency, 8.1 and 5.9 ms; duration of postspike facilitation, 13.0 and 9.3 ms.
View larger version (33K):
[in a new window]
FIG. 6.
STAs and response averages of an inhibitory PreM-IN in rostral T1; format as in Fig. 5. Cell fired phasically during extension torques and produced a postspike suppression in the extensor digitorum-4,5 (ED-4,5) muscle. mean percent decrease (MPD), 
8.5%; peak percent decrease (PPD), 12.0%; onset latency, 9.3 ms; peak latency, 16.8 ms; duration of postspike suppression, 13.5 ms. Note that on average postspike facilitations and suppressions had similar rise times (Table 1), despite the difference between the effect in this figure and those in Fig. 5.
). In FDS and FCU, these effects were superimposed on early facilitations with onset latencies of 5.0 and 0.6 ms, respectively (
). The STAs of the neuron in Fig. 8 showed large early facilitations in the four independently recorded flexor muscles and a postspike facilitation in the FDP muscle. Early features usually had substantially larger MPIs than superimposed postspike effects, as in Fig. 8.
View larger version (24K):
[in a new window]
FIG. 7.
STAs and response average of a PreM-IN in lamina VII of the C8-T1 border; format as in Fig. 5. Postspike facilitations ( ) were superimposed on synchrony facilitations () in the flexor digitorum superficialis (FDS) and FCU muscles. "Peak" immediately preceding the postspike facilitation in FCU was not present in repeated STAs compiled with independent triggers and was regarded as baseline noise. Neuron was activated tonically during flexion torques and was silent during extension. MPI and PPI of postspike effects are 5.4 and 10.9% in FDS, 3.4 and 5.5% in FCU, and 7.6 and 14.2% in flexor digitorum profundus (FDP). MPI and PPI of synchrony effects are 5.4 and 7.4% in FDS and 5.4 and 8.7% in FCU.
View larger version (29K):
[in a new window]
FIG. 8.
STAs and response average of a PreM-IN located near the motor nuclei in T1; format as in Fig. 5. STA shows a postspike facilitation ( ) superimposed on a synchrony facilitation () in FDP and pure synchrony facilitations in the other 4 independently recorded flexor muscles. Neuron was activated for flexion torques with a decrementing response pattern. MPI and PPI of postspike effect are 4.7 and 7.7% in FDP. MPI and PPI of synchrony effects are 14.0 and 21.4% in FDP, 18.7 and 33.0% in FCU, 12.3 and 20.8% in flexor carpi radialis (FCR), and 6.5 and 9.3% in FDS.
View larger version (19K):
[in a new window]
FIG. 9.
Relative distributions of postspike effects to wrist flexor and extensor muscles for 98 PreM-INs. 
, neurons producing postspike facilitations; 
, neurons producing postspike suppressions. Neurons with reciprocal postspike effects (
) included 2 that facilitated flexors and suppressed extensors and 2 that facilitated extensors and suppressed flexors. Two neurons with postspike effects only in brachioradialis muscle are not included. Percentages of total number of premotor neurons are listed and depicted as relative area of associated balloon.
3 muscles. Most PreM-INs with multiple target muscles (25/36) produced postspike effects only in coactive muscles (i.e., the 5 primary flexor muscles plus PT during flexion torques or the 5 primary extensor muscles plus APL during extension torques). The other PreM-INs cofacilitated (Fig. 10), cosuppressed, or produced reciprocal effects in flexor and extensor muscles. A larger proportion of neurons classified as last-order interneurons (5/18, 28%) had postspike effects in both flexor and extensor muscles than other PreM-INs (6/82, 7%).
View larger version (34K):
[in a new window]
FIG. 10.
STAs and response averages of a lamina VII neuron in C8 that cofacilitated extensor and flexor muscles; format as in Fig. 5. Neuron had tonic background firing for maintained torques, with discharge rate higher for flexion than extension torques, and a phasic increase in rate at the onset of torques in either direction. Postspike facilitation was produced in extensor digitorum communis (EDC), abductor pollicis longus (APL), extensor carpi radialis (ECR), and FDP. First peak in ECR is a posttrigger early effect (latency 2.6 ms). MPI and PPI of effects are 8.7 and 14.0% in EDC, 2.5 and 3.9% in APL, 3.5 and 5.5% in ECR, and 4.7 and 9.0% in FDP.
View this table:
TABLE 1.
Parameters of output effects
View larger version (37K):
[in a new window]
FIG. 11.
Timing of spike-related effects in forearm muscles. Histograms show onset latency (A and B) and peak latency (C and D) for postspike (left) and early effects (right). 
, facilitatory effects; 
, suppressive effects. Histogram outlines show duration (E) and rise time (F) of postspike (
) and early () effects. 
, means (milliseconds) for all effects.
4.2%) were similar, as were the peak percent changes (Table 1). There was no significant difference between the magnitudes of postspike effects in flexor and extensor muscles (Table 1) or among effects in the three monkeys.
), but a brief description is relevant here to compare response patterns with the output effects of PreM-INs. Two features of each neuron's firing properties are considered: the cell's "preferred" direction and its activity during torques in the opposite direction.
View larger version (26K):
[in a new window]
FIG. 12.
Relation between response patterns and postspike effects of PreM-INs. Neuron's agonist muscles are those active in the direction of torque for which the neuron has the largest increase in activity (not defined for neurons with unmodulated activity, 50 entries not aligned with agonist or antagonist columns). Fac, facilitatory effects; Sup, suppressive effects; dashed lines indicate connections that are not detectable by spike-triggered averaging. Two PreM-INs with postspike effects in brachioradialis (BR) are not included; 2 PreM-INs with postspike effects only in PT during trials in which PT was bidirectionally active are also not included. Square, circles, and triangles indicate increasing incongruity between response pattern and postspike effects (see text for details).
View larger version (19K):
[in a new window]
FIG. 13.
Histological reconstruction of recording sites for neurons in monkeys B and W. Sites are collapsed onto 1 frontal section of the gray matter in each monkey. 
, PreM-INs; , putative motoneurons; 
, neurons with spike-related early effects in STAs (and no postspike effects). See text for details of reconstruction method. One motoneuron appears to be located in lamina VI or VII, which probably reflects a limitation of this procedure.
, 59%, 81 effects for 45 neurons) straddled the trigger time, having onsets before and offsets after the trigger (Figs. 7, 8, and 15, A and B). Early-posttrigger effects (, 19%, 27 effects for 24 neurons) had onsets that occurred after the trigger but were too early (<3.5 ms) to be classified as postspike effects. Pretrigger effects (, 22%, 30 effects for 22 neurons) began and ended entirely before the occurrence of the triggering action potential. A few large, consistent pretrigger effects had onset latencies of less than 20 ms (Fig. 15, E and F).
View larger version (18K):
[in a new window]
FIG. 14.
Onset times plotted against offset times for all early effects. Effects that straddle the trigger time are classified as synchrony effects ( ); effects entirely after the trigger are classified as early-posttrigger effects (); effects entirely before the trigger are classified as pretrigger effects ().
View larger version (27K):
[in a new window]
FIG. 15.
Examples of different types of early effects from 6 different spinal interneurons. A and B: facilitatory and suppressive synchrony effects, respectively, associated with neurons that had no observable postspike effect in any muscle. C and D: multiphasic effects; insets: autocorrelogram of trigger cell. E and F: entirely pretrigger facilitation and suppression of EMG, respectively.
; McKiernan and Cheney 1993). We recompiled STAs with biphasic effects straddling the trigger time using only triggers preceded by a long interspike interval (30-50 ms). If the initial suppression of EMG was related to a trough in the neuron's autocorrelogram (Fig. 15, insets), this procedure would increase the duration of the suppressive phase. However, there was no significant change in the shape or duration of the effects.
View larger version (26K):
[in a new window]
FIG. 16.
Relative distributions of early effects to flexor and extensor muscles; format as in Fig. 9.
DISCUSSION
Abstract
Introduction
Methods
Results
Discussion
References
; Jankowska and Roberts 1972), motor axon conduction velocities in monkeys (90 m/s) (Carp 1992; Eccles et al. 1968; Gilliatt et al. 1976), and the estimated length of forearm motor axons (20-35 cm) in the three monkeys of this study.
) and longer conduction distances (C7-T1 motoneurons extend over ~2.5 cm in macaques) (Jenny and Inukai 1983) can account for delays of several milliseconds. Even more importantly, the size of the affected motoneurons determines the conduction time along the motor axon, which is a significant component of the onset latency of postspike effects. Many large, fast-conducting motoneurons may not be activated at the low force levels used in these experiments. Differences in conduction velocities of slow- and fast-twitch motoneurons (40-90 m/s in monkeys) (Carp 1992; Eccles et al. 1968; Gilliatt et al. 1976) can contribute differences in onset delays of ~5 ms (for motor axons 35 cm long) for postspike effects produced by PreM-INs projecting selectively to small versus large motoneurons (Garnett and Stephens 1980; Heckman and Binder 1988). Thus postspike effects that begin 10 ms after the trigger time still could be mediated by monosynaptic connections to motoneurons (Kasser and Cheney 1985).
) and rubral (Mewes and Cheney 1991) neurons and excitatory linkages from thalamic neurons (Mewes and Cheney 1990) to forearm muscles. These effects are at least disynaptically mediated. Short onset latencies can result from disynaptic pathways with short conduction distances to fast-conducting motoneurons. We estimated that postspike effects in EMG with latencies of 4.5 ms could be mediated by disynaptic pathways.
; Czarkowska et al. 1981; Jankowska and Roberts 1972) and cervical (Hongo et al. 1989) cord of cats. This may represent a difference in the organization of primate cervical and feline lumbar spinal circuits or may be due to our sampling of different species of interneurons. Alternatively, interneurons with widespread axonal branching may produce postspike effects in only one muscle for several possible reasons. Some anatomic connections may produce postsynaptic potentials that are too small to produce postspike effects that rise above the background EMG level (Cope et al. 1987) or are in motor units that are not sampled by the EMG electrodes. It is also possible that presynaptic inhibition of interneuron terminals (Aggelopoulos et al. 1997) during normal behavior (Hultborn et al. 1987) decreases the effectiveness of the synapses from specific axonal branches. Finally the interneurons could have branches to large motoneurons that were inactive for the task used in this study.
; Kasser and Cheney 1985), red nucleus (RM cells) (Mewes and Cheney 1991), and C8-T1 dorsal root ganglia (DRG cells) (Flament et al. 1992) have more divergent facilitatory outputs than PreM-INs (Fig. 17). Excitatory spinal PreM-INs may have more focused effects on motor pools than supraspinal cells and segmental afferents and thus may exert more fractionated control of muscles than do CM, RM, and DRG cells.
View larger version (28K):
[in a new window]
FIG. 17.
Relative distributions of postspike effects to flexor and extensor muscles for PreM-INs (same as Fig. 9) and premotor populations previously studied in the motor cortex, red nucleus, and dorsal root ganglia under similar behavioral conditions. Percentages of total number of premotor neurons are listed and depicted as relative area of associated balloon. Relative size of average muscle fields for postspike facilitations and postspike suppressions (considered separately, even when produced by individual neurons with reciprocal output effects) as a percentage of the number of independent, coactive muscles are shown (right) for each population.
; Kasser and Cheney 1985; Mewes and Cheney 1991). Comparing the latency of the postspike effects with poststimulus effects elicited by single microstimuli in the cortex or red nucleus (Cheney and Fetz 1985; Mewes and Cheney 1991) suggests that the short-latency STA features of CM and RM cells did include synchrony effects. Consequently, the muscle field sizes of CM and RM cells may have been overestimated by inclusion of muscles for which STAs exhibited purely synchrony effects. Synchrony and postspike effects were differentiated for DRG afferents (Flament et al. 1992), so their muscle fields and those of PreM-INs are directly comparable.
; Kasser and Cheney 1985) and DRG (Flament et al. 1992) cells generated postspike effects equally in flexor and extensor muscles, and RM cells generated effects more often in extensor muscles than in flexors (Belhaj-Saif et al. 1998; Mewes and Cheney 1991). In addition, the magnitude of postspike effects produced by CM cells was greater in extensor muscles than in flexors (Fetz and Cheney 1980). It is possible that the preferential distribution of postspike effects to flexor muscles reflects a sampling bias of our interneuron recordings. For example, forearm flexor muscle motoneurons are located more rostrally than extensor motoneurons in the cervical enlargement (Jenny and Inukai 1983). However, the rostrocaudal distribution of PreM-INs with effects on flexor versus extensor muscles showed no clear difference.
; Cangiano and Lutzemberger 1972; Clough et al. 1968; Wilson and Peterson 1981). The preferential linkages of PreM-INs to small numbers of forearm flexors may reflect a greater involvement of spinal processing in the control of individual muscles or small groups of muscles during diverse grasping movements of the hand and less influence over more stereotyped hand opening and wrist extension movements.
; Sherrington 1910), and many PreM-INs with facilitatory postspike effects in flexor muscles may be interneurons mediating flexion reflexes. The widespread flexion reflexes elicited by activation of diverse afferent pathways in spinalized animals suggest that spinal interneurons exert more control on flexor than extensor motoneurons. However, it is clear that the interneurons that mediate flexion reflexes are used for a variety of other movements under different behavioral conditions (Lundberg et al. 1987; McCrea 1992). Because the monkeys in our study produced extension torques as often as flexion torques, it appears that interneurons in the primate cervical spinal cord are organized preferentially to influence the activity of flexor motoneurons.
; Kasser and Cheney 1985) and DRG (Flament et al. 1992) neurons and only slightly larger than those of RM cells (Mewes and Cheney 1991). On the other hand, postspike suppressions produced by spinal PreM-INs had a shorter mean latency and a larger mean amplitude than those produced by CM (Kasser and Cheney 1985) and RM cells (Mewes and Cheney 1991). This suggests that suppressive PreM-INs were probably last-order inhibitory interneurons.
). However, motoneurons discharge only when the monkey produces torque in one direction. In contrast, 74% of PreM-INs (71% of neurons classified as last-order interneurons) maintained at least some background activity during torques in both directions and thus are not organized into simple, reciprocally firing pathways.
; Mewes and Cheney 1994), uncommon for DRG cells and motor cortical neurons during single joint movements (Crutcher and Alexander 1990; Evarts 1968; Flament et al. 1992; Murphy et al. 1982; Thach 1978), and virtually absent for CM cells projecting to forearm muscles (Cheney and Fetz 1980). In their study on hand movements, Gibson et al. (1985) found that magnocellular red nucleus neurons had little resting discharge when monkeys were not making active muscle contractions. Similarly Mewes and Cheney (1994) reported that the background activity of RM cells decreased substantially when the monkeys were not actively engaged in the task. We occasionally observed that spinal interneurons with activity during zero or nonpreferred torques had significantly decreased discharge rates during "time outs" when the target displays were turned off and the monkey was distracted from the task. In the flexion/extension task, the activity of PreM-INs for zero and nonpreferred torques and of RM cells for nonpreferred torques (Mewes and Cheney 1994) may represent a generalized increased excitability related to the generation of movement. Apparently this increased excitability was not evoked by the behavioral paradigms of Gibson et al. and is absent in CM and DRG neurons.
View larger version (26K):
[in a new window]
FIG. 18.
Schematic diagrams of circuits that can account for spike-related effects reported in this paper. A: connectivity of a PreM-IN. B: motor axon collateral activates neuron C-IN, resulting in an early-posttrigger effect in STA triggered from C-IN. C: common input to a PreM-IN and neuron S-IN results in synchrony effect in STA triggered from S-IN. D: neuron G-IN is driven by input from a Golgi tendon organ (GTO) that is activated synchronously with contraction of a motor unit, resulting in pretrigger effect in STA triggered from G-IN. In A-D, the neuron triggering the STA is shown as a filled circle in the circuit diagram; the timing of activity is shown (right of each circuit diagram). Events for each neuron show the earliest possible occurrence of action potentials, and thick horizontal lines indicate other possible occurrence times, representing the width of the central peak in the cross-correlogram between the triggering neuron [or the motoneuron (MN), in D] and the other cells in the circuit. Cross-correlogram is shown above the horizontal line for MN in A. Diagonal lines indicate synaptic delays that result in probabilistic increases in firing of postsynaptic cell. STAs of rectified EMG are shown in bottom row. In A-C, MU Hist depicts the histogram of motoneuron firing in relation to the trigger time. Longer duration of the spike-related feature in the STA as compared with MU Hist reflects the width of the muscle unit action potential and the recruitment of other motoneurons by the trigger cell (Palmer and Fetz 1985 ). In D, twitch tension shows the time course of the increase in force produced by activation of the motoneuron. Duration of the spike-related feature in the STA reflects the width of the muscle-unit action potential and the histogram of firing of G-IN relative to the motoneuron. E: highly interconnected spinal premotor network, as suggested by present data. This diagram is schematic: all pathways to motoneurons and connections between interneurons are not necessarily present in the spinal cord. STAs of numbered neurons shown to right: 1, postspike effect; 2, synchrony effect; 3, postspike effect superimposed on synchrony effect.
; Singer 1993). These central peaks are evidence of common synaptic input that increases above chance the probability that the two neurons will discharge synchronously (Kirkwood and Sears 1978). Analogously, peaks (or troughs) that straddle the trigger time in STAs of EMG may reflect common input to the recorded cell and to another neuron with an excitatory (or inhibitory) synaptic connection to motoneurons (Fig. 18C) (Flament et al. 1992).
for cortical neurons; Prut et al. 1996 for spinal interneurons). Such interactions among spinal cells would produce output effects dispersed in time by more than the conduction delay of motor axons.
). In addition, two analytic tests performed on the data (selective triggering with spikes preceded by long intervals and deconvolution of STAs with the autocorrelogram of the triggering cell) failed to show any significant influence of autocorrelogram features on multiphasic synchrony effects. This suggests that these complex STA features reflect the activity in a population of premotor neurons that only weakly entrains the trigger cell.
; Singer 1993; Welsh et al. 1995) and between motor units of the same and different muscles (Bremner et al. 1991a; Datta and Stephens 1990; DeLuca and Mambrito 1987; Sears and Stagg 1976). Motor-unit synchronization in human hand muscles is associated with coherent firing at a frequency of 16-32 Hz, suggesting that common inputs fire rhythmically (Farmer et al. 1993). This is consistent with reports of periodic features in the activity of corticospinal pathways (Baker et al. 1997; Conway et al. 1995; Murthy and Fetz 1996).
; Jankowska 1992), the similarity between synchrony effects in STAs triggered from spinal interneurons and DRG neurons (Flament et al. 1992), and the similarity between the frequencies of oscillatory features in STAs triggered from interneurons and motor cortex (Baker et al. 1997; Murthy and Fetz 1996) suggest common afferent and descending inputs to spinal interneurons. In conclusion, synchrony features in STAs of EMG probably represent the collective activity of premotor circuits that affect the triggering cell. Our data suggest a model of spinal premotor circuitry in which a network of interconnected interneurons receives common supraspinal and feedback signals that tend to produce synchronized, occasionally rhythmic, firing of interneurons and motoneurons (Fig. 18E).
). Early-posttrigger effects were not accompanied by later postspike suppressions of the target muscles, but STAs from Renshaw cells would not necessarily reveal the recurrent inhibition. The greater efficacy of the cell's excitatory input from motoneurons relative to the efficacy of its inhibitory output to motoneurons (see Baldissera et al. 1981), could produce a STA with an early effect that masks the recurrent inhibition.
; Humphrey and Reed 1983). Because most PreM-INs facilitated muscle activity, it appears that these shared pathways are primarily excitatory.
) counter the inappropriate excitation produced by facilitatory PreM-INs during movements in which the target muscles are not recruited. Alternatively, for incongruent postspike effects that are mediated by disynaptic connections, inappropriate effects would be avoided if the intervening interneuron is inactive.
for example, inhibitory PreM-INs that are silent when the muscles they suppress are activewill not be detected with this technique. This reasoning would suggest that many inhibitory PreM-INs in the spinal cord are active in a strict alternating pattern during voluntary flexion and extension of the wrist. These were not detected by STAs because they fired only when their target muscles were suppressed. Ia inhibitory interneurons probably were included in this grouponly one PreM-IN (shaded circle in Fig. 12, "Sup antag" column and "high-low" row) had properties expected of this neuron type (Jankowska 1992). "Unidentified interneurons" with strictly alternating activity patterns, described in the companion paper [Maier et al. (1998), in Table 2, the 84 U-INs (20% of U-INs) with "no activity" in the nonpreferred direction], probably include inhibitory PreM-INs. This conclusion suggests a difference in the organization of excitatory and inhibitory inputs to cervical motoneurons. Excitatory effects seem to be distributed to all wrist muscles during wrist movements, whereas specific inhibitory effects sharpen the reciprocal activation of flexors and extensors.
demonstrated that 28% of CM cells produce reciprocal postspike suppressions in antagonists of their facilitated target muscles. They activated CM cells with glutamate iontophoresis during movements in their nonpreferred direction, when they were normally silent. Among RM cells, 29% also had reciprocal effects (Mewes and Cheney 1991). Although not tested with glutamate activation, most DRG cells, which were likely Ia afferents (Flament et al. 1992), also would be expected to exert reciprocal inhibition through the Ia inhibitory interneuron (Baldissera et al. 1981). In contrast, only 4% of the spinal PreM-INs had reciprocal effects on flexor and extensor muscles. This is probably an underestimate: some of the 24 unidirectionally active PreM-INs (Fig. 12) might have exhibited reciprocal effects if tested with glutamate activation. Nonetheless, the reciprocal control of flexor and extensor muscles seems to be determined largely by supraspinal inputs during wrist movements.
).
| |
ACKNOWLEDGEMENTS |
|---|
We thank E. Jankowska for valuable comments on the manuscript, J. Garlid and L. Shupe for expert technical assistance, K. Elias for editorial assistance, and M. Domenowske for the illustration in Fig. 2.
This study was supported by National Institutes of Health Grants RO1-NS-12542, F32-NS-9189, and RR-00166; American Paralysis Association Grants PB1-9402-1 and PBR2-9502; and a Swiss National Science Foundation fellowship (M. A. Maier).
| |
FOOTNOTES |
|---|
Present address of M. A. Maier: Sobell Dept. of Neurophysiology, Institute of Neurology, Queen Square, London WC1N 3BG, UK.
Address for reprint requests: S. I. Perlmutter, Dept. of Physiology and Biophysics, Box 357330, University of Washington, Seattle, WA 98195-7330.
Received 2 December 1997; accepted in final form 28 July 1998.
| |
REFERENCES |
|---|
|
Abstract
Introduction
Methods
Results
Discussion
References
|
|---|
This article has been cited by other articles:
|
|
 |
|
  K. Seki, S. I. Perlmutter, and E. E. Fetz Task-Dependent Modulation of Primary Afferent Depolarization in Cervical Spinal Cord of Monkeys Performing an Instructed Delay Task J Neurophysiol, July 1, 2009; 102(1): 85 - 99. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Takei and K. Seki Spinomuscular Coherence in Monkeys Performing a Precision Grip Task J Neurophysiol, April 1, 2008; 99(4): 2012 - 2020. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Yanai, N. Adamit, R. Harel, Z. Israel, and Y. Prut Connected Corticospinal Sites Show Enhanced Tuning Similarity at the Onset of Voluntary Action J. Neurosci., November 7, 2007; 27(45): 12349 - 12357. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Maieron, G. D. Iannetti, J. Bodurka, I. Tracey, P. A. Bandettini, and C. A. Porro Functional Responses in the Human Spinal Cord during Willed Motor Actions: Evidence for Side- and Rate-Dependent Activity J. Neurosci., April 11, 2007; 27(15): 4182 - 4190. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. H. Schieber and G. Rivlis Partial Reconstruction of Muscle Activity From a Pruned Network of Diverse Motor Cortex Neurons J Neurophysiol, January 1, 2007; 97(1): 70 - 82. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. T. Moritz, T. H. Lucas, S. I. Perlmutter, and E. E. Fetz Forelimb Movements and Muscle Responses Evoked by Microstimulation of Cervical Spinal Cord in Sedated Monkeys J Neurophysiol, January 1, 2007; 97(1): 110 - 120. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Jackson, S. N. Baker, and E. E. Fetz Tests for presynaptic modulation of corticospinal terminals from peripheral afferents and pyramidal tract in the macaque J. Physiol., May 15, 2006; 573(1): 107 - 120. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. H. Schieber and G. Rivlis A Spectrum From Pure Post-Spike Effects to Synchrony Effects in Spike-Triggered Averages of Electromyographic Activity During Skilled Finger Movements J Neurophysiol, November 1, 2005; 94(5): 3325 - 3341. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. M. Baekey, K. F. Morris, S. C. Nuding, L. S. Segers, B. G. Lindsey, and R. Shannon Ventrolateral medullary respiratory network participation in the expiration reflex in the cat J Appl Physiol, June 1, 2004; 96(6): 2057 - 2072. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. B. Nielsen Sensorimotor integration at spinal level as a basis for muscle coordination during voluntary movement in humans J Appl Physiol, May 1, 2004; 96(5): 1961 - 1967. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Prut and S. I. Perlmutter Firing Properties of Spinal Interneurons during Voluntary Movement. I. State-Dependent Regularity of Firing J. Neurosci., October 22, 2003; 23(29): 9600 - 9610. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Prut and S. I. Perlmutter Firing Properties of Spinal Interneurons during Voluntary Movement. II. Interactions between Spinal Neurons J. Neurosci., October 22, 2003; 23(29): 9611 - 9619. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. H. Schieber Training and Synchrony in the Motor System J. Neurosci., July 1, 2002; 22(13): 5277 - 5281. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. J. De Luca and Z. Erim Common Drive in Motor Units of a Synergistic Muscle Pair J Neurophysiol, April 1, 2002; 87(4): 2200 - 2204. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M.A. Maier, J. Armand, P.A. Kirkwood, H.-W. Yang, J.N. Davis, and R.N. Lemon Differences in the Corticospinal Projection from Primary Motor Cortex and Supplementary Motor Area to Macaque Upper Limb Motoneurons: An Anatomical and Electrophysiological Study Cereb Cortex, March 1, 2002; 12(3): 281 - 296. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J M Kilner, S N Baker, and R N Lemon A novel algorithm to remove electrical cross-talk between surface EMG recordings and its application to the measurement of short-term synchronisation in humans J. Physiol., February 1, 2002; 538(3): 919 - 930. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Olivier, S. N. Baker, K. Nakajima, T. Brochier, and R. N. Lemon Investigation Into Non-Monosynaptic Corticospinal Excitation of Macaque Upper Limb Single Motor Units J Neurophysiol, October 1, 2001; 86(4): 1573 - 1586. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. M Baekey, K. F Morris, C. Gestreau, Z. Li, B. G Lindsey, and R. Shannon Medullary respiratory neurones and control of laryngeal motoneurones during fictive eupnoea and cough in the cat J. Physiol., July 15, 2001; 534(2): 565 - 581. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E Jankowska Spinal interneuronal systems: identification, multifunctional character and reconfigurations in mammals J. Physiol., May 15, 2001; 533(1): 31 - 40. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Hager-Ross and M. H. Schieber Quantifying the Independence of Human Finger Movements: Comparisons of Digits, Hands, and Movement Frequencies J. Neurosci., November 15, 2000; 20(22): 8542 - 8550. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Nakajima, M. A. Maier, P. A. Kirkwood, and R. N. Lemon Striking Differences in Transmission of Corticospinal Excitation to Upper Limb Motoneurons in Two Primate Species J Neurophysiol, August 1, 2000; 84(2): 698 - 709. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R Shannon, D M Baekey, K F Morris, Z Li, and B G Lindsey Functional connectivity among ventrolateral medullary respiratory neurones and responses during fictive cough in the cat J. Physiol., May 15, 2000; 525(1): 207 - 224. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. A. Maier, S. I. Perlmutter, and E. E. Fetz Response Patterns and Force Relations of Monkey Spinal Interneurons During Active Wrist Movement J Neurophysiol, November 1, 1998; 80(5): 2495 - 2513. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
