J Neurophysiol 90: 2090 –2097, 2003.
First published May 28, 2003; 10.1152/jn.00200.2003.
Sustained Muscle Contractions Maintained by Autonomous Neuronal Activity
Within the Human Spinal Cord
Daichi Nozaki,1 Noritaka Kawashima,1 Yu Aramaki,2 Masami Akai,1 Kimitaka Nakazawa,1 Yasoichi Nakajima,2
and Hideo Yano1
1
Departments of Motor Dysfunction and 2Sensory and Communication Disorders, Research Institute of National Rehabilitation Center for
the Disabled, 4 –1 Namiki, Tokorozawa 359-8555, Japan
Submitted 3 March 2003; accepted in final form 27 May 2003
Conventional belief has maintained that the spinal motoneuron is a passive neuronal integrator (Henneman and Mendell
1981); namely, it cannot fire without neuronal inputs. However, recent studies using reduced animal preparations have
challenged this notion (Bennett et al. 1998; Crone et al. 1988;
Hounsgaard et al. 1984, 1988; Schwindt and Crill 1980a,b).
Transient current injection or excitatory synaptic inputs can
trigger sustained motoneuron activity that outlasts input cessation. Most importantly, it comes from intrinsic properties that
generate sustained depolarization of the motoneuron (plateau
potential) (Hultborn 1999).
However, the emergence of plateau potentials in awake
animals or humans is difficult to demonstrate because the
membrane voltage of the motoneuron cannot be recorded intracellularly. Nevertheless, indirect evidence that plateau potentials contribute to motoneuron firing has been provided by
investigations of the motor-unit firing pattern in animals that
were awake (Bennett et al. 2001; Eken 1998; Gorrassini et al.
1999) and in humans (Gorassini et al. 1998, 2002; Kiehn and
Eken 1997). A paired motor-unit method adopted in these
studies, where the firing frequency of the lower threshold
motor unit is used as a measure of common synaptic inputs, has
ingeniously shown that motoneuron firing could be partly
maintained by intrinsically mediated depolarization.
If the plateau potential actually contributes to motoneuron
firing, as implied in previous studies, self-sustained motoneuron activity can also be observed in humans. Preliminary
evidence can be found in the work of Lang and Vallbo (1967).
They described low intensity stimulation of the tibial nerve that
occasionally induced sustained muscle activity, although they
did not emphasize this point. Other evidence was obtained
from patients who suffered from cramping (Baldissera et al.
1994). These authors revealed that electrical stimulation of Ia
afferents could trigger sustained muscle contractions; furthermore, they ascribed this mechanism to the plateau potentials of
human spinal motoneurons. More recently, it has been shown
that direct electrical stimulation of the muscle belly in normal
human subjects can induce sustained, involuntary muscle contractions that outlast the stimulus period (Collins et al. 2001).
However, several counter arguments should be answered to
reach the conclusion that these muscle contractions come from
an intrinsic mechanism of the motoneurons or interneurons.
For example, although Collins et al. (2001) described that the
sustained muscle contractions could be induced even for sleeping subjects, it was not evaluated to what extent this muscle
contraction was involuntary. Also, it remains unclear to what
extent these contractions were maintained by reverberating
activity within closed neuronal circuits.
The percutaneous application of an electrical pulse to the
tibial nerve is well known to reflexively activate human triceps
Address for reprint requests and other correspondence: Daichi Nozaki,
Department of Motor Dysfunction, Research Institute NRCD, 4 –1 Namiki
Tokorozawa, Saitama 359-8555, Japan, (E-mail: [email protected]).
The costs of publication of this article were defrayed in part by the payment
of page charges. The article must therefore be hereby marked ‘‘advertisement’’
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
INTRODUCTION
2090
0022-3077/03 $5.00 Copyright © 2003 The American Physiological Society
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 18, 2017
Nozaki, Daichi, Noritaka Kawashima, Yu Aramaki, Masami Akai,
Kimitaka Nakazawa, Yasoichi Nakajima, and Hideo Yano. Sustained
muscle contractions maintained by autonomous neuronal activity within
the human spinal cord. J Neurophysiol 90: 2090 –2097, 2003. First
published May 28, 2003; 10.1152/jn.00200.2003. It is well known that
muscle contraction can be easily evoked in the human soleus muscle
by applying single-pulse electrical stimulation to the tibial nerve at the
popliteal fossa. We herein reveal the unexpected phenomenon of
muscle contractions that can be observed when train stimulation is
used instead. We found, in 11 human subjects, that transient electrical
train stimulation (1-ms pulses, 50 Hz, 2 s) was able to induce sustained muscle contractions in the soleus muscle that outlasted the
stimulation period for greater than 1 min. Subjects were unaware of
their own muscle activity, suggesting that this is an involuntary
muscle contraction. In fact, the excitability of the primary motor
cortex (M1) with the sustained muscle contractions evaluated by
transcranial magnetic stimulation was lower than the excitability with
voluntary muscle contractions even when both muscle contraction
levels were matched. This finding indicates that M1 was less involved
in maintaining the muscle contractions. Furthermore, the muscle
contractions did not come from spontaneous activity of muscle fibers
or from reverberating activity within closed neuronal circuits involving motoneurons. These conclusions were made based on the respective evidence: 1) the electromyographic activity was inhibited by
stimulation of the common peroneal nerve that has inhibitory connections to the soleus motoneuron pool and 2) it was not abolished after
stopping the reverberation (if any) for approximately 100 ms by
inducing the silent period that followed an H-reflex. These findings
indicate that the sustained muscle contractions induced in this study
are most likely to be maintained by autonomous activity of motoneurons and/or interneurons within the human spinal cord.
AUTONOMOUS NEURONAL ACTIVITY WITHIN HUMAN SPINAL CORD
surae muscles (H-reflex) (Schieppati 1987). In the present
study, first, we demonstrate that train stimulation, instead of
single-pulse stimulation, induces sustained muscle contractions
that outlast the stimulation period, as reported in previous
human (Collins et al. 2001, 2002) and animal experiments
(Crone et al. 1988). We then provide evidence that this muscle
contraction comes from autonomous neuronal activity within
the spinal cord (possibly plateau potentials in motoneurons
and/or interneurons). Specifically, we have focused mainly on
the degree of “autonomy” of muscle contractions with various
types of electrical stimulation and transcranial magnetic stimulation of the primary motor cortex.
Parts of the present study have been presented in abstract
form (Nozaki et al. 2002).
METHODS
In total, 11 male subjects (22–39 yr old) participated in this study.
The subjects provided informed consent. This study was performed
according to the Declaration of Helsinki and was approved by the
ethics committee of our research institute. Participants were placed on
a bed on their left side with the inside of their right leg also placed on
the bed. Their hip and knee joints were flexed approximately 60°. The
right foot was not fixed because foot fixation with an ankle brace or
footplate allowed subjects to be aware of muscle contractions through
sensation of the reaction force from the sole of the foot. The surface
electromyography (EMG) signal was recorded from the right soleus
(Sol), the medial (MG) and lateral (LG) heads of the gastrocnemius;
and the tibialis anterior (TA) with use of a bipolar electrode. The
EMG signal was amplified (Amplifier; Nihon-Kohden, AB-651-J)
with band-pass filtering between 15 Hz and 3 kHz and digitized at 5
kHz. The tibial nerve was percutaneously stimulated by applying
rectangular electrical pulses of 1 ms duration to the popliteal fossa
(Fig. 1) with a constant voltage stimulator (DPS-1300D, Dia Medical
System). Typically, 50 Hz train stimulation for 2 s of 1.2 ⫻ H-reflex
threshold (this was below the direct motor response threshold) was
used to elicit sustained muscle contractions in the triceps surae muscles (Fig. 1). This stimulus intensity did not evoke any apparent ankle
joint movement (i.e., the induced muscle contraction was isometric)
because the friction between the inside of the foot and bed surface was
sufficient to prevent movement. The subjects were instructed to ignore
the electrical stimulation as far as possible and to remain relaxed. In
addition, the same experiment was conducted for two subjects after
placement of a topical anesthetic on the skin surrounding the stimulation electrode by subcutaneous injection of 3 ml lidocaine (10
mg/ml). We then determined whether cutaneous sensation was responsible for involuntary sustained muscle contractions. In the following parts of this paper, we tentatively refer to this muscle contraction as “the self-sustained muscle contraction,” even though this has
not been clarified at this stage. At the end of experiment, the subjects
conducted isometric maximal voluntary plantarflexion and dorsiflexion in the same posture described above and the EMG signals were
recorded.
Conditioning electrical stimulation
Six subjects participated in experiments 1 and 2. In experiment 1,
the common peroneal nerve (CPN) was stimulated during the selfsustained muscle contraction by using a bipolar surface electrode at
the level of the caput fibulae (Fig. 1, arrow 1) (Tanaka 1983). The
stimulus intensity was set to 1.05, 1.2, and 1.5 ⫻ motor threshold
(MT) of TA. Similarly, in experiment 2, the tibial nerve was stimulated to elicit an H-reflex in the Sol by applying an electrical pulse via
the same electrodes as those used to induce the self-sustained muscle
contraction (Fig. 1, arrow 2). The stimulus intensity was set to 1.0 ⫻
motor threshold of the Sol. In both cases, the interstimulus interval
was 6 s. Ten responses were elicited in one trial.
Transcranial magnetic stimulation
FIG. 1. Method inducing sustained muscle contractions in the human triceps surae muscles and experimental setting. Sustained involuntary muscle
contractions were induced by transient electrical stimulation to the tibial
afferent nerve (50 Hz, 2 s). To investigate whether the induced muscle
contractions originated from the plateau potentials, we provided a conditioning stimulus to the common peroneal nerve (CPN) in experiment 1
(arrow 1), to the tibial nerve (TN) in experiment 2 (arrow 2), and the
primary motor cortex using a transcranial magnetic stimulator in experiment 3 (arrow 3).
J Neurophysiol • VOL
In experiment 3, the primary motor cortex (M1) for the ankle joint
muscles of the right leg was stimulated using Magstim 200 (The
Magstim Company, UK) for 10 subjects (Fig. 1, arrow 3). A double
cone coil was placed over and around the vertex with its long axis
parallel to the frontal plane. The coil was firmly fixed to the subject’s
head using a custom-built helmet. During the experiment, subjects
were asked to close their eyes. The stimulus intensity was set to the
threshold level of the Sol in the resting condition (from 55 to 90% of
maximal stimulator output). The interstimulus interval was random
(ⱖ7 s apart) and a total of five responses were evoked in one trial.
First, the motor evoked potential (MEP) was evoked during the
self-sustained muscle contraction while background muscle activity
(BGA) of the Sol was continuously monitored on a realtime basis. The
BGA was calculated from the root mean square value (RMS) of the
EMG during the 100 ms just before the onset of stimulation. After
that, the MEP was evoked during voluntary plantarflexion, where the
subjects were requested to adjust the Sol BGA level to the last
sustained muscle contraction. This cycle was repeated six times per
subject. The magnitude of the MEP was evaluated by the area of the
EMG response.
To investigate the excitability of the Sol motoneuron pool, the Sol
H-reflex was evoked using the same method from experiment 2. The
experimental procedure was the same as the TMS experiment described above. The stimulus intensity was adjusted to evoke a constant
motor response (M-response) in the Sol (7 ⫾ 2% of maximal Mresponse). The magnitude of the Sol H-reflex and M-response was
evaluated by the peak-to-peak amplitude of the EMG response.
90 • OCTOBER 2003 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 18, 2017
Induction of sustained muscle contraction
2091
2092
NOZAKI ET AL.
Additional experiments
RESULTS
We found, in all 11 subjects, that electrical train stimulation
to the tibial afferent nerve can induce continuous muscle activity in the triceps surae muscles that outlasts the stimulation
period (Fig. 2A). The average muscle activity levels for the 60 s
recording induced by 1.2 ⫻ H-reflex threshold stimulus intensity were 9.6 ⫾ 1.9, 4.1 ⫾ 1.9, and 4.4 ⫾ 3.1% (means ⫾ SE
across all subjects) of maximal voluntary contraction on the
EMG basis for Sol, MG, and LG, respectively. Interestingly,
subjects were unaware of their muscle activities unless they
were informed. (However, when the ankle joint was fixed by an
ankle brace, subjects could feel muscle contractions by haptic
sensation. see METHODS.) It should be noted that this sustained
muscle contraction did not originate from involuntary tensing
of muscles secondary to stimulation pain. None of the subjects
felt pain from the stimulation. In addition, the sustained muscle
contractions were still observed after the application of a
topical anesthetic to the skin surrounding the stimulation electrode.
In previous studies using direct electrical stimulation to the
muscle belly (Collins et al. 2001, 2002), large reflex-like force
increments during the stimulation period have been reported.
In contrast to studies with stimulus contaminations to EMG
signals, our method allowed a full investigation of the EMG
response during the stimulation period (Fig. 2B). After a large
reflex response was induced by the first stimulus, several
subsequent responses were considerably depressed. However,
the response gradually recovered with repeated stimulations.
Moreover, the size of response never reached that of the first
H-reflex. The averaged size of the last two H-reflexes was
19.5 ⫾ 2.3% of the first H-reflex (means ⫾ SE across all
subjects). This gradual reflex increment was always observed
in all 11 subjects prior to subsequent sustained muscle contractions.
Figure 3A shows the effect of an electrical pulse stimulation
FIG. 2. Sustained muscle contractions induced by electrical train stimulation. A: typical trace of electromyographic
signals recorded from the triceps surae muscles and tibialis
anterior. Even after stimulation was turned off, electromyographic (EMG) activity of triceps surae muscles was maintained. B: EMG activity of the soleus during the electrical
stimulation period (same trace as in A). Enlarged signals are
shown in the bottom. After the reflexive muscle contraction
induced by the first electrical stimulation, several stimulations failed to evoke the reflex (bottom left). However, the
amplitude of the reflex gradually recovered (bottom right).
Such gradual recovery in reflexes was always observed prior
to the occurrence of sustained muscle contractions.
J Neurophysiol • VOL
90 • OCTOBER 2003 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 18, 2017
Experiment 3 was designed to clarify the extent of motor cortex
association with the sustained muscle contraction by comparing the
MEP and the H-reflex between both muscle contraction modes. However, two problems remained with this strategy. First, since the Hreflex was much larger than the MEP (see Fig. 4A and C), it is
possible that the large H-reflex does not have sufficient sensitivity to
detect the difference in contraction modes (Crone et al. 1990). Second,
since the MEP contains considerable polysynaptic components, it
would reflect not only the motor cortex excitability but also the
interneuron excitability (Nielsen et al. 1999). We conducted the
following two additional experiments to overcome these problems.
The first additional experiment consisted of matching the size of the
H-reflex with the MEP by using a very weak stimulus intensity for six
subjects. The experimental procedure was the same as the main
H-reflex experiment; however, the H-reflex size was quantified by the
area of EMG response, as done in the TMS experiment. In this case,
we were unable to monitor the M-response to guarantee the constancy
of the stimulus intensity, so we readjusted the stimulus intensity every
cycle (unless the stimulus intensity was adjusted, the drift was sometimes observed in H-reflex magnitude because of the change of
impedance between the skin and electrode). We compared the size of
the H-reflex size between voluntary and self-sustained muscle contractions every cycle.
In the second additional experiment, we compared the amount of
the earliest H-reflex facilitation conditioned by the TMS (Nielsen et
al. 1993) between self-sustained and voluntary muscle contractions.
This method is suitable for examining the excitability of the motor
cortex because the earliest H-reflex facilitation is probably caused by
activation of direct monosynaptic projections from the motor cortex to
spinal motoneurons (Nielsen and Petersen 1995). Five subjects participated in this experiment. First, the test H-reflexes were conditioned
by TMS applied at various conditioning-test intervals from ⫺10 to 3
ms with 1 ms resolution (negative value indicates that the TMS was
applied after tibial nerve stimulation) while the subjects conducted
voluntary plantarflexion. The amplitude of unconditioned H-reflex
was set to approximately 15% of maximal M-response throughout this
experiment. The earliest conditioning-test time interval to facilitate
the H-reflex was obtained for each subject. The H-reflex for each
muscle contraction mode then was conditioned by the TMS applied at
this interval or at 1 ms earlier than this interval (in the latter case, the
size of the H-reflex was almost equivalent to that of the unconditioned
H-reflex). These two intervals were alternated randomly and ⱖ20
H-reflexes were obtained for each muscle contraction mode and each
time interval. The amount of H-reflex facilitation was calculated for
each muscle contraction mode by taking the ratio of H-reflexes
induced at these two interstimulus intervals. The same ratio was
calculated for the Sol BGA to examine whether the H-reflex facilitation was due to the larger Sol BGA.
The repeated measures ANOVA with replication was used to
compare the size of MEP, H-reflex, M-response, and BGA between
the two muscle contraction modes. In the second additional experiment, a Student’s t-test was used for each subject to test whether the
H-reflex facilitation was significant. The probability P ⬍ 0.05 was
accepted as a significant level.
AUTONOMOUS NEURONAL ACTIVITY WITHIN HUMAN SPINAL CORD
2093
FIG. 3. Soleus (Sol) EMG response to conditioning electrical stimulation. A: ensemble averaged responses (over 10
responses) of the rectified Sol EMG to the CPN stimulation in
a subject. The rectified EMG was depressed during approximately 30 –50 ms after stimulation (arrows). The inhibitory
effect grew larger as stronger stimuli were used. B: electrical
train stimulation to CPN (50 Hz, 2 s, 1.3 ⫻ motor threshold of
TA) often terminated the sustained muscle contractions. C:
ensemble averaged response (over 10 responses) of the soleus
rectified EMG of the TN stimulation (1.0 ⫻ MT) in a subject.
Although the rectified EMG was suppressed over 100 ms after
an H-reflex, it recovered after this silent period. Similar results
as shown in A–C were observed for all subjects.
the train CPN stimulation. This is because the train tibial
nerve stimulation failed to stop the sustained muscle contractions (data not shown), although it also induced the
H-reflex and the subsequent silent period (see Fig. 3C) as
the CPN stimulation did.
The effect of an electrical pulse stimulation of the tibial
nerve (Fig. 1, arrow 2) during sustained muscle contractions is
shown in Fig. 3C. The stimulus intensity was set to 1.0 ⫻ MT,
which can induce nearly a maximal H-reflex in the Sol. This
reflexive muscle contraction was followed by the silent period
of approximately 100 ms (Fig. 3C, gray bar). However, the Sol
EMG level recovered to the level prior to the application of
conditioning stimulation.
To quantify the influence of descending commands from the
M1, the size of the MEP induced by TMS to M1 (Fig. 4A) was
evaluated. Figure 4B indicates that the Sol MEP was significantly smaller (P ⬍ 0.005) for the sustained muscle contractions than for voluntary plantarflexion. The Sol BGA was not
different between the two muscle contraction modes. Furthermore, the significant interactions (P ⬍ 0.005) between subjects
FIG. 4. Excitability of the primary motor cortex during
muscle contractions. A: typical trace of ensemble averaged
Sol motor evoked potential (MEP) (over 30 responses)
during self-sustained muscle contractions induced by our
method (black) and during voluntary muscle contractions
(gray) in a subject. B: while the Sol background electromyographic activity (BGA) was at the same level for both
contraction modes (n.s., not significant), the Sol MEP was
larger for voluntary plantarflexion (***P ⬍ 0.005). Similarly, the MEP was larger for voluntary contraction for the
TA (*P ⬍ 0.05), the MG (**P ⬍ 0.01), and the LG (*P ⬍
0.05). Although there were no significant BGA differences
between the two muscle contraction modes for these muscles, significant interactions were observed between subjects and muscle contraction modes (###P ⬍ 0.005). Data
represent the means ⫾ SE for all subjects. C: typical trace
of ensemble averaged Sol EMG response (H-reflex) to tibial
nerve stimulation (over 30 responses) during self-sustained
muscle contractions (black) and during voluntary muscle
contractions (gray) in a subject. In this trace, the BGA was
obscure because of the large size of the H-reflex and the
attenuation of the EMG by calculating the ensemble average. D: H-reflex test indicates that the size of the H-reflex
was not different between the two contraction modes. The
BGA level and the size of the M-response were also not
different. However, there was an interaction between subjects and muscle contraction modes (###P ⬍ 0.005). Data
represent the means ⫾ SE for all subjects.
J Neurophysiol • VOL
90 • OCTOBER 2003 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 18, 2017
of the CPN (Fig. 1, arrow 1) on the sustained muscle contractions. This conditioning electrical stimulation reduced the rectified Sol EMG with latency 30 to 40 ms (Fig. 3A, arrows),
indicating this EMG reduction might be caused by disynaptic
Ia reciprocal inhibition. The degree of reduction in Sol EMG
was larger when larger stimulations were used (Fig. 3A). Furthermore, electrical train stimulation of the CPN larger than
1.2 ⫻ MT (50 Hz ⫻ 2 s) reduced (13 of 29 trials in 4 subjects)
or often terminated (13 trials) the sustained muscle contractions (Fig. 3B), as previously reported in animal experiments
(Crone et al. 1988). When the intensity of stimulation was
1.05 ⫻ MT, it had no effect. As shown in Fig. 3A, the possible
disynaptic Ia reciprocal inhibition was followed by excitation
and subsequent inhibition. These might partly come from the
synchronized firing of motor units. In addition, when the stimulation intensity was strong, it is possible that the strong
dorsiflexion evoked by the CPN stimulation induced the stretch
reflex response in the soleus muscle and the subsequent silent
period. However, it was unlikely that these factors contributed
to the termination of the sustained muscle contractions by
2094
NOZAKI ET AL.
and muscle contraction modes existed for the BGA of other
muscles (i.e., some subjects had larger BGA for the voluntary
plantarflexion). Hence, these subjects had to activate MG, LG,
and even TA (i.e., cocontraction) to voluntarily adjust the Sol
FIG. 6. Effects of conditioning TMS on the size of the test Sol H-reflex. A: typical example of the effect of TMS in a subject.
In this subject, the earliest facilitation of the H-reflex was observed at the conditioning-test interval of –3 ms. Namely, when the
conditioning-test interval was – 4 ms, the size of H-reflex was equivalent to that of the control H-reflex. The facilitatory effect was
significant for voluntary muscle contraction (*P ⬍ 0.05), but it was not for self-sustained muscle contraction (n.s., not significant).
Data represent the means ⫾ SE. B: difference in facilitatory effect between self-sustained (SS) and voluntary muscle contractions
(Vol) in all 5 subjects. The size of facilitated H-reflex was expressed as the percentage of that conditioned at ⬃1 ms earlier of the
effective interval. The H-reflex facilitation was significant (*P ⬍ 0.05) in 4 of 5 subjects only for voluntary muscle contraction.
In 1 subject, the t value did not reach the significance level (P ⫽ 0.08). C: BGA level at the effective conditioning-test interval was
expressed as the percentage of that at 1 ms earlier than this interval. Note that the BGA level did not explain the results in B.
J Neurophysiol • VOL
90 • OCTOBER 2003 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 18, 2017
FIG. 5. Result of Sol H-reflex additional test when weak stimulation was
used. A: in this experiment, the weak stimulus intensity was used so that the
size of H-reflex was comparable to that of MEP. When the Sol BGA was
matched, there was no difference of H-reflex size between both contraction
modes (n.s., not significant). Data represent the means ⫾ SE for 6 subjects. B:
result of TMS experiment for 6 subjects participating in this additional experiment. The result was the same as shown in Fig. 4B (***P ⬍ 0.005). Data
represent the means ⫾ SE for these 6 subjects.
BGA level to that during sustained muscle contractions. The
MEP of these muscles was also significantly larger (P ⬍ 0.01
for MG, P ⬍ 0.05 for LG and TA) for voluntary contractions
(Fig. 4B). However, whether this result reflects some substantial difference between muscle contraction modes is unclear
from our experiment, because BGA levels were not matched
for both contraction modes.
It is possible that the Sol MEP size only reflects the excitability of the spinal motoneuron pool rather than the M1.
Therefore we also examined the excitability of the spinal
motoneuron pool using the H-reflex method (Fig. 4C). The
result was different from the TMS experiment (Fig. 4D). Specifically, the Sol H-reflex amplitude was at the same level for
both contraction modes (the BGA level and the size of the
M-response were not different). Furthermore, the significant
interaction between subjects and muscle contraction modes
(P ⬍ 0.005) indicates that the Sol H-reflex of some subjects
was smaller for the voluntary plantarflexion. Figure 5A indicates the result of additional H-reflex experiments for six
subjects. Even when the magnitude of H-reflex was adjusted to
that of TMS, the larger MEP for voluntary muscle contractions
(Fig. 5B) cannot be explained by the difference in the excitability of the spinal motoneuron pool.
In the second additional experiment, the earliest facilitation
of the H-reflex was observed at the conditioning-test interval of
⫺3 to –1 ms. The typical facilitation effect is shown in Fig. 6A.
In this subject, the earliest facilitation was observed at the
conditioning-test interval of –3 ms (TMS intensity was 35% of
maximal stimulator output). The amount of the facilitation was
statistically significant for the voluntary muscle contraction
(P ⬍ 0.05), but it was not for the sustained muscle contraction.
Figure 6B indicates that the amount of the earliest H-reflex
facilitation for all five subjects. In four of five subjects, the
amount of the facilitation was significant only for the voluntary
AUTONOMOUS NEURONAL ACTIVITY WITHIN HUMAN SPINAL CORD
muscle contraction (in 1 subject, P ⫽ 0.08). The difference in
the BGA level could not explain such a facilitation effect in
voluntary muscle contraction (Fig. 6C).
Thus these results indicate that M1 activity itself is lower
during the self-sustained muscle contraction than for the ordinary voluntary muscle contractions.
DISCUSSION
J Neurophysiol • VOL
reinforce this hypothesis, several possible counter arguments
should be answered. 1) Is the sustained muscle contraction
really associated with motoneuron activity? 2) Isn’t the sustained muscle contraction generated by reverberating activity
within the closed neuronal circuit? 3) Is it true that spinal
motoneurons do not receive volitional descending commands
from the motor cortex? We conducted three experiments to
examine each possibility.
The first matter to be clarified is whether motoneurons are
really firing during the sustained muscle contractions. For
some pathological situations, like the syndrome of continuous
muscle-fiber activity (Isaacs 1961), muscle fibers can be active
without motoneuron firing. To investigate this, we provided
conditioning electrical pulse stimulation to the common peroneal nerve, which is known to have an inhibitory connection to
the Sol motoneuronal pool (Fig. 1, arrow 1) (Crone et al. 1988;
Tanaka 1983). If the observed muscle contractions were not
associated with motoneuronal activity, this stimulation would
not affect the Sol EMG signal. However, the conditioning
stimulation reduced the rectified Sol EMG via the reciprocal
inhibitory pathway (Fig. 3A), indicating that the self-sustained
muscle contractions are certainly supported by spinal motoneuronal activity.
The second possibility is that reverberations of neuronal
signals within closed neuronal circuits maintained the selfsustained muscle activity. To investigate this, the silent period
of 100 ms was induced by providing electrical pulse stimulation to the tibial nerve during self-sustained muscle contractions (Fig. 1, arrow 2). During the silent period, the firing of
spinal motoneurons was actually stopped by several mechanisms, such as the pause of Ia afferent discharges during a
twitch contraction (Merton 1951), the Renshaw inhibition
(Anastasijević and Vučo 1980), and so on. Hence, it may stop
reverberations of neuronal signal(s) within the closed neuronal
loop, if the motoneurons are involved in that loop. Therefore,
if the sustained muscle contraction was due to its reverberation,
it should have also terminated after the silent period. However,
this single pulse stimulation was unable to terminate the sustained muscle contractions (Fig. 3C), indicating that at least
they are not supported by reverberating activity within closed
neuronal loops involving the motoneurons. In contrast, when
the possible closed neuronal loops do not involve the motoneurons, the situation is more complicated and it is very difficult to
draw a decisive conclusion (see following text).
Whether the volitional descending commands contribute to
the motoneuron activity is the third and most critical matter.
The following facts still cannot guarantee that descending
commands were not associated with the self-sustained muscle
contraction: 1) subjects maintained relaxation and 2) they did
not recognize their muscle activity. Likewise, although Collins
et al. (2001) have described that the self-sustained muscle
contraction can be induced even for sleeping subjects, the
involvement of M1 in the muscle contraction should be examined quantitatively. To quantify the influence of descending
commands from M1, we evaluated the size of the MEP when
TMS was applied to M1. If the observed self-sustained muscle
contractions were supported by autonomous neuronal activity
within the spinal cord, they should not have been accompanied
with M1 activity. On the other hand, M1 activity should have
always been involved during voluntary plantarflexion (Johannsen et al. 2001). Our hypothesis was that the MEP should
90 • OCTOBER 2003 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 18, 2017
We can induce the self-sustained muscle contractions in
human triceps surae muscles by applying transient electrical
stimulation to tibial nerves. These self-sustained muscle contractions were induced by repetitive Ia afferent inputs to the
motoneuron pools. This is evidenced by the fact that we used
a stimulus intensity that was below the motor threshold. Also,
blocked sensation of the skin around the stimulus electrode did
not affect it. This is similar to the phenomenon reported in
previous studies that used direct electrical stimulation of the
muscle belly (Collins et al. 2001, 2002). In those studies, large
reflex-like force increments during the stimulation period were
considered evidence of plateau potentials in human spinal
motoneurons. This is because it closely resembled the firing
frequency acceleration of the motoneuron during the neuronal
input period at the initiation of plateau potentials (Hounsgaard
et al. 1984, 1988; Hultborn 1999). Since stimulus contamination to EMG signals is much smaller in our method, the EMG
activity during the stimulation period was fully investigated
(Fig. 2B). When double stimulation procedure is used, it is well
known that the second H-reflex is almost depressed if the
interval is within several tens of milliseconds (Kagamihara et
al. 1998; Magladery et al. 1952). In accordance with this
well-known behavior, several subsequent responses were considerably depressed (Fig. 2B) after a large reflex response was
induced by the first stimulus. However, the response gradually
recovered with repeated stimulation (Fig. 2B), which is similar
to the frequency acceleration phenomenon (Hounsgaard et al.
1984, 1988). Although gradual increases in muscular responses
have been reported in a study on the tonic vibration reflex (De
Gail et al. 1966; Matthews 1966; Stuart et al. 1986), this type
of H-reflex behavior during repetitive electrical stimulations
has not been reported yet. Hence, some novel factors have to
counter the well-known depression effect during this recovery
period. One of the explanations for this phenomenon is the
depolarization of the membrane voltage of the motoneuron
through plateau potential emergence.
The gradual reflex increment (Fig. 2B), as well as the force
increment reported in previous studies (Collins et al. 2001,
2002), supports the hypothesis that induced muscle activity
comes from plateau potentials in human spinal motoneurons.
However, as Granit et al. (1957) originally considered, the
posttetanic potentiation by repetitive stimulation (Hirst et al.
1981) can contribute to the maintained increase in motoneuronal excitability. Furthermore, it is possible that the temporal
summation of excitatory postsynaptic potential evoked by repetitive stimulation could have affected our results. Therefore
we have to turn our attention to the sustained muscle contraction rather than to the EMG behavior during the stimulation
period.
We hypothesize that the sustained muscle contraction induced here is evidence for the emergence of plateau potentials
in human spinal motoneurons or in interneurons at least. To
2095
2096
NOZAKI ET AL.
J Neurophysiol • VOL
et al. 2001), muscle contractions induced by electrical stimulation to the afferent nerve has several advantages over conventional functional electrical stimulation. It can recruit the
nonfatigable motor unit at first and it also prevents synchronous muscle fiber contractions. Furthermore, this type of stimulation can be used as a useful and efficient method in preventing muscle atrophy during inactivity of elderly people or
during the microgravity conditions of astronauts.
We thank C. Capaday and K. Yamanaka for helpful discussions.
DISCLOSURES
This work was supported by the Japanese Ministry of Health, Labor and
Welfare; the Sasakawa Scientific Research Grant from The Japan Science
Society; the Meiji Life Foundation of Health and Welfare; and the Japanese
Ministry of Education, Culture, Sports, Science and Technology.
REFERENCES
Anastasijević R and Vučo J. Renshaw cell discharge at the beginning of
muscular contraction and its relation to the silent period. Exp Neurol 69:
589 –598, 1980.
Baldissera F, Cavallari P, and Dworzak F. Motor neuron ‘bistability’: a
pathogenetic mechanism for cramps and myokymia. Brain 117: 929 –939,
1994.
Bennett DJ, Hultborn H, Fedirchuk B, and Gorassini M. Synaptic activation of plateaus in hindlimb motoneurons of decerebrate cats. J Neurophysiol 80: 2023–2037, 1998.
Bennett DJ, Li Y, Harvey PJ, and Gorrassini M. Evidence for plateau
potentials in tail motoneurons of awake chronic spinal rats with spasticity.
J Neurophysiol 86: 1972–1982, 2001.
Collins DF, Burke D, and Gandevia SC. Large involuntary forces consistent
with plateau-like behavior of human motoneurons. J Neurosci 21: 4059 –
4065, 2001.
Collins DF, Burke D, and Gavdevia SC. Sustained contractions produced by
plateau-like behaviour in human motoneurones. J Physiol 538: 289 –301,
2002.
Crone C, Hultborn H, Kiehn O, Mazieres L, and Wingström H. Maintained changes in motoneuronal excitability by short-lasting synaptic inputs
in the decerebrate cat. J Physiol 405: 321–343, 1988.
Crone C, Hultborn H, Maziéres L, Morin C, Nielsen J, and PierrotDeseilligny E. Sensitivity of monosynaptic test reflexes to facilitation and
inhibition as a function of the test reflex size: a study in man and the cat. Exp
Brain Res 81: 35– 45, 1990.
De Gail P, Lance JW, and Neilson PD. Differential effects on tonic and
phasic reflex mechanisms produced by vibration of muscles in man. J Neurol Neurosug Psychiatry 29: 1–11, 1966.
Eken T. Spontaneous electromyographic activity in adult rat soleus muscle.
J Neurophysiol 80: 365–376, 1998.
Gorassini M, Bennett DJ, Kiehn O, Eken T, and Hultborn H. Activation
patterns of hindlimb motor units in the awake rat and their relation to
motoneuron intrinsic properties. J Neurophysiol 82: 709 –717, 1999.
Gorassini M, Yang JF, Siu M, and Bennett DJ. Intrinsic activation of human
motoneurons: possible contribution to motor unit excitation. J Neurophysiol
87: 1850 –1858, 2002.
Gorassini MA, Bennett DJ, and Yang JF. Self-sustained firing of human
motor units. Neurosci Lett 247: 13–16, 1998.
Granit R, Phillips CG, Skoglund S, and Steg G. Differentiation of tonic
from phasic alpha ventral horn cells by stretch, pinna and crossed extensor
reflexes. J Neurophysiol 20: 470 – 481, 1957.
Hallett M. Transcranial magnetic stimulation and the human brain. Nature
406: 147–150, 2000.
Hirst GDS, Redman SJ, and Wong K. Post-tetanic potentiation and facilitation of synaptic potentials evoked in cat spinal motoneurones. J Physiol
321: 97–109, 1981.
Henneman E and Mendell LM. Functional organization of motoneuron pool
and its inputs. In: Handbook of Physiology. The Nervous System. Motor
Control. Bethesda, MD: Am. Physiol. Soc., 1981, sect. 1, vol. II, chapt. 11,
p. 423–507.
90 • OCTOBER 2003 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 18, 2017
be smaller during the self-sustained muscle contractions than
during voluntary plantarflexion due to the lower excitability of
M1 (Hallett 2000) (assuming equal muscle activity levels). The
results shown in Figs. 4, B and D, 5, and 6 indicate that this
hypothesis might be correct. Although the result that both the
MEP and the H-reflex facilitation by the conditioning TMS
was smaller for the self-sustained muscle contractions does not
completely exclude the possibility that M1 is not involved in
supporting them, the smaller activity of M1 should be compensated by other factors, such as autonomous activity of the
spinal motoneuron. Furthermore, the result that some subjects
could not voluntarily adjust all muscle activity levels simultaneously to those during the sustained muscle contractions (Fig.
4B) indicates that the self-sustained muscle contractions are
apparently different from the voluntary muscle contractions.
Our results reinforce the view that self-sustained muscle
contractions induced by our method are primarily maintained
by autonomous neuronal activity within the human spinal cord.
Considering the similarity of this study to the animal study by
Crone et al. (1988), this autonomous neuronal activity is most
likely to be explained by the plateau potentials in spinal motoneurons. However, alternative explanations are still possible
for the experimental results. For example, our experimental
method was unable to exclude the contribution of the closed
neuronal circuits whose circuit times were longer than the
silent period. However, considering that most of the EMG
response to stretch in the Sol lies within approximately 100 ms
from stretch onset (Schieppati and Nardone 1997), the contribution of such circuits is small. Another possibility is that
neuronal systems, except M1, provide the sustained synaptic
excitation to the spinal motoneurons. Their candidate is the
plateau potentials in spinal interneurons (Morisset and Nagy
1999) or the reverberating activity within closed neuronal
circuits that do not involve the spinal motoneurons as their
constituents. These mechanisms can also account for our experimental results without any contradiction. It should be noted
that, even in these cases, the present results demonstrate the
novel phenomenon that the plateau potentials in interneurons
or the sustained activity of neuronal network can be triggered
in humans by applying electrical train stimulation to an Ia
afferent nerve.
From the conventional viewpoint of motor control, the spinal motoneuron can fire only by receiving excitatory inputs
from the supraspinal systems or the peripheral neuronal receptors. However, the present results suggest that the human spinal
cord has the ability to generate autonomous neuronal activity
maintaining muscle contraction. This finding forces reconsideration of this common belief. In the future, this autonomous
activity of the spinal neuronal system should be taken into
consideration when investigating the control of movements
(Bennett et al. 1998; Collins et al. 2002; Gorassini et al. 2002;
Hultborn 1999). For example, under the existence of the autonomous activity mode, the motor cortex does not have to
continue to send commands for motoneurons to fire continuously. It only has to send a signal of “switch on” or “switch
off” to control them. This may help reduce the burden on
supraspinal systems for tasks such as locomotion or standing
(Eken 1998). Furthermore, a recent study (Collins et al. 2002)
has shown the possibility that the strength of plateaus can be
controlled by volitional commands. From the viewpoint of
clinical applications, as pointed out in a previous study (Collins
AUTONOMOUS NEURONAL ACTIVITY WITHIN HUMAN SPINAL CORD
J Neurophysiol • VOL
Morisset V and Nagy F. Ionic basis for plateau potentials in deep dorsal horn
neurons of the rat spinal cord. J Neurosci 19: 7309 –7316, 1999.
Nielsen J, Morita H, Baumgarten J, Petersen N, and Christensen LOD. On
the comparability of H-refelexes and MEPs. Electroenchephalogr Clin
Neurophysiol Suppl 51: 93–101, 1999.
Nielsen J and Petersen N. Changes in the effect of magnetic brain stimulation
accompanying voluntary dynamic contraction in man. J Physiol 484: 777–
789, 1995.
Nielsen J, Petersen N, Deuschl G, and Ballegaard M. Task-related changes
in the effect of magnetic brain stimulation on spinal neurones in man.
J Physiol 471: 223–243, 1993.
Nozaki D, Kawashima N, Aramaki Y, Nakajima Y, Nakazawa K, and
Akai M. Autonomous activity of human spinal motoneurons: sustained
muscle contractions with less involvement of motor cortex. Soc Neurosci
Abstr 2002.
Schieppati M. The Hoffmann reflex: a means of assessing spinal reflex excitability and its descending control in man. Prog Neurobiol 28: 345–376, 1987.
Schieppati M and Nardone A. Medium-latency stretch reflexes of foot and
leg muscles analysed by cooling the lower limb in standing humans.
J Physiol 503: 691– 698, 1997.
Schwindt P and Crill W. Role of a persistent inward current in motoneuron
bursting during spinal seizures. J Neurophysiol 43: 1296 –1318, 1980a.
Schwindt PC and Crill WE. Properties of a persistent inward current in
normal and TEA-injected motoneurons. J Neurophysiol 43: 1700 –1724,
1980b.
Stuart GJ, Rymer WZ, and Schotland JL. Characteristics of reflex excitation in close synergist muscles evoked by muscle vibration. Exp Brain Res
65: 127–134, 1986.
Tanaka R. Reciprocal Ia inhibitory pathway in normal man and in patients
with motor disorders. Adv Neurol 39: 433– 441, 1983.
90 • OCTOBER 2003 •
www.jn.org
Downloaded from http://jn.physiology.org/ by 10.220.33.2 on June 18, 2017
Hounsgaard J, Hultborn H, Jespersen B, and Kiehn O. Intrinsic membrane
properties causing a bistable behaviour of ␣-motoneuron. Exp Brain Res 55:
391–394, 1984.
Hounsgaard J, Hultborn H, Jespersen B, and Kiehn O. Bistability of
␣-motoneurones in the decerebrate cat and in the acute spinal cat after
intravenous 5-hydroxytryptophan. J Physiol 405: 345–367, 1988.
Hultborn H. Plateau potentials and their role in regulating motoneuronal
firing. Prog Brain Res 123: 39 – 48, 1999.
Isaacs H. A syndrome of continuous muscle-fiber activity. J Neurol Neurosurg
Psychiatry 24: 319 –325, 1961.
Johannsen P, Christensen LOD, Sinkær T, and Nielsen JB. Cerebral
functional anatomy of voluntary contractions of ankle muscles in man.
J Physiol 535: 397– 406, 2001.
Kagamihara Y, Hayashi A, Okuma Y, Nagaoka M, Nakajima Y, and
Tanaka R. Reassessment of H-reflex recovery curve using the double
stimulation procedure. Muscle Nerve 21: 352–360, 1998.
Kiehn O and Eken T. Prolonged firing in motor units: evidence of plateau
potentials in human motoneurons. J Neurophysiol 78: 3061–3068, 1997.
Lang AH and Vallbo ÅB. Motoneuron activation by low intensity tetanic
stimulation of muscle afferents in man. Exp Neurol 18: 383–391, 1967.
Magladery JW, Teasdall RD, Park AM, and Languth HW. Electrophysiological studies of reflex activity in patients with lesions of the nervous
systems. I. A comparison of spinal motoneuron excitability following afferent nerve volleys in normal persons and patients with upper motor neurone
lesions. Bull Johns Hopkins Hosp 91: 219 –244, 1952.
Matthews PBC. The reflex excitation of the soleus muscle of the decerebrate
cat caused by vibration applied to its tendon. J Physiol 184: 450 – 472, 1966.
Merton PA. The silent period in a muscle of the human hand. J Physiol 114:
183–198, 1951.
2097