SEARCH

SEARCH BY CITATION

Keywords:

  • corpus callosum;
  • human;
  • motor cortex;
  • transcranial magnetic stimulation

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References
  9. Supporting Information

The corpus callosum is essential for neural communication between the left and right hemispheres. Although spatiotemporal coordination of bimanual movements is mediated by the activity of the transcallosal circuit, it remains to be addressed how transcallosal neural activity is involved in the dynamic control of bimanual force execution in human. To address this issue, we investigated transcallosal inhibition (TCI) elicited by single-pulse transcranial magnetic stimulation (TMS) in association with the coordination condition of bimanual force regulation. During a visually-guided bimanual force tracking task, both thumbs were abducted either in-phase (symmetric condition) or 180° out-of-phase (asymmetric condition). TMS was applied to the left primary motor cortex to elicit the disturbance of ipsilateral left force tracking due to TCI. The tracking accuracy was equivalent between the two conditions, but the synchrony of the left and right tracking trajectories was higher in the symmetric condition than in the asymmetric condition. The magnitude of force disturbance and TCI were larger during the symmetric condition than during the asymmetric condition. Right unimanual force tracking influenced neither the force disturbance nor TCI during tonic left thumb abduction. Additionally, these TMS-induced ipsilateral motor disturbances only appeared when the TMS intensity was strong enough to excite the transcallosal circuit, irrespective of whether the crossed corticospinal tract was activated. These findings support the hypotheses that interhemispheric interactions between the motor cortices play an important role in modulating bimanual force coordination tasks, and that TCI is finely tuned depending on the coordination condition of bimanual force regulation.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References
  9. Supporting Information

In electrophysiological studies, interhemispheric neural interactions between motor cortices have been well investigated in association with unimanual actions (Ferbert et al., 1992; Perez & Cohen, 2008), showing that transcallosal inhibition (TCI) is modulated inversely between the left and right motor cortices. In this situation, TCI toward the motor cortex innervating the active hand decreases (Murase et al., 2004; Liuzzi et al., 2010), whereas TCI toward the contralateral motor cortex increases (Mochizuki et al., 2004; Hinder et al., 2010). That is, TCI subserves the lateralized excitation of the motor cortex to generate an isolated unimanual action (Mayston et al., 1999). However, little is known about how TCI underlies motor organization during bimanual action. As bimanual actions require the co-activation of the bilateral motor cortices, the transcallosal inhibitory circuit may be organized differently compared with that observed during unimanual actions. Recently, a few studies investigated TCI with respect to bimanual actions (Yedimenko & Perez, 2010; Liuzzi et al., 2011). However, these studies were conducted either in the pre-movement phase or during static muscle contraction; hence, it remains to be addressed how the transcallosal inhibitory circuit is engaged in dynamic bimanual control during an ongoing action. As the static and dynamic contractions showed different activation patterns of corticomotoneuronal neurons (Cheney & Fetz, 1980), the transcallosal circuit might also exhibit different activity during dynamic force control.

During bimanual motor control, there is a characteristic behavioral constraint according to the spatiotemporal congruency of the left and right actions (Swinnen, 2002). In general, a simultaneous action using both sets of homologous muscle groups is more stable than that of non-homologous ones. Furthermore, even during a symmetric action, it is difficult to produce different magnitudes of muscle forces simultaneously (Steglich et al., 1999; Hu & Newell, 2011). Interestingly, patients with a lesion of the corpus callosum (CC) are likely to be freed from such bimanual constraints (Diedrichsen et al., 2003), indicating that bimanual isometric force control is also mediated by interhemispheric neural interactions via the transcallosal circuit.

Given these neurophysiological and behavioral backgrounds, we hypothesized that TCI is finely tuned for performing dynamic regulation of bimanual forces with different coordination strategies for different tasks. To test this hypothesis, we addressed the following questions: first, whether TCI differs between the symmetric and asymmetric bimanual force regulations, and second, whether TCI modulation during bimanual force regulation is different from that during unimanual action. In the present study, TCI was assessed by examining the effect of single-pulse transcranial magnetic stimulation (TMS) applied to the left primary motor cortex (M1) on the muscle activity of the ipsilateral hand. Suprathreshold TMS over the M1 disrupts motor activity in the muscles of the ipsilateral hand via TCI (Ferbert et al., 1992). Supporting this notion, some lesion studies demonstrated that such disruption disappeared in patients with a complete callosal lesion (Meyer et al., 1995), but is preserved in those with a subcortical vascular lesion (Boroojerdi et al., 1996).

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References
  9. Supporting Information

Participants

Eleven healthy male volunteers, 22–35 years old, participated in this study (six participated in all of the experiments, four participated only in the main experiment, and one participated only in the control experiments). All participants gave informed consent for the experimental procedure, which was approved by the local ethics committee at Chiba University, Faculty of Education, and was in accordance with the guidelines established in the Declaration of Helsinki. All participants were right-handed, as confirmed by the Edinburgh handedness inventory (Oldfield, 1971), and had no musical training.

Stimulation

To induce transcallosal motor interference on the activity of the left muscle, TMS was delivered to the left M1 using a magnetic stimulator (Magstim 200; Magstim Co., UK) with a figure-of-eight-shaped coil (each diameter 70 mm). The coil was located at a hot-spot where weak stimulation elicited the largest motor response in the right abductor pollicis brevis (APB), and was held tangentially over the scalp and rotated clockwise at 45°. The induced current in the cortex was set to run in the posterior–anterior direction. The stimulus intensity was set at 1.5 times the resting motor threshold (RMT). This intensity was quite strong because we aimed to induce an observable perturbation in the abduction force of the left thumb. However, we confirmed that TCI tested during isometric contraction was not saturated at this intensity (Supporting Information Fig. S1). The RMT was defined as the minimum stimulus intensity that produced a > 50 μV motor evoked potential (MEP) at the right APB in at least 5 out of 10 consecutive trials.

Motor task

Main experiment

The participants sat comfortably on a reclining chair with both shoulders and elbow angles semi-flexed throughout the experiment. Their left and right hands were separately placed on wooden boards with their palms downward. Each hand was strapped at the metacarpophalangeal joints of four fingers and the wrists. The thumbs were extended approximately 40° and the thumb cushion was in contact with a horizontal metal plate (Fig. 1A). The contact area was confined to 20 × 20 mm and was covered with a rubber sheet.

image

Figure 1. (A) Experimental setup. Left and right thumb abduction forces were displayed on the left and right halves, respectively, of an oscilloscope screen. The target line, displayed on the right half of the screen, moved up and down with a frequency of 0.1 Hz. TMS was given at a 10- or 15-s interval once the target line passed the 6 N level. Bimanual visually guided force tracking in the symmetric (B) and asymmetric (C) conditions. Black and gray indicate the left and right sides, respectively. During the symmetric condition, bilateral force feedback was displayed symmetrically. During the asymmetric condition, right side feedback was displayed in the opposite direction. The dotted arrows inside the oscilloscope schema represent the direction of force line movement when the abduction force was applied. Analysis of transcallosal interference during the left force incremental (D) and decremental (E) phases. From top to bottom: left force, rectified APB EMG traces, subtracted electromyography, and cumulative sum of the mean (CUSUM). The black and gray lines represent the 20-times averaged traces with and without TMS, respectively. The dotted slope on the force line was estimated by linear fitting in the 200-ms pre-stimulus baseline. The vertical arrows indicate the amount of force disturbance. TCI was depicted in the gray shaded area on the EMG traces and estimated by the amplitude of CUSUM.

Download figure to PowerPoint

The force regulation task was constructed on the basis of our previous experimental design (Kida et al., 2004). The participants were instructed to perform bimanual thumb abductions under visuomotor tracking. The target line moved sinusoidally up and down at 0.1 Hz on the right half of a dual-beam oscilloscope screen (VC-9; Nihon Kohden, Japan) positioned in front of the participants at a distance of 60 cm (Fig. 1A). The range of the target line displacement on the oscilloscope was 8 cm in height, which corresponded to a force range from 1 to 11 N (with a resolution of approximately 0.02 N). Left and right abduction forces were displayed as horizontal lines on the left and right half of the oscilloscope, respectively. In the symmetric condition, bilateral forces were displayed in the same manner; when the participant pushed the plates with both thumbs, both lines moved from bottom to top (Fig. 1B). Under this condition, the participants tracked the target line with bilateral thumb abduction forces in a symmetrical manner. In contrast, in the asymmetric condition, the right force line was displayed upside down by using an inverse function switch (Fig. 1C). Under this condition, the direction of force line movement in relation to the direction of the right force was opposite to that of the left force; therefore, to track the target line with both the left and right lines, the participants had to perform bilateral thumb abduction in an asymmetrical manner. One tracking sequence continued for 120 s and was repeated four times with a resting interval of 5 min. The order of the symmetric and asymmetric conditions was counterbalanced across participants. In each tracking condition, TMS was delivered for a total of 40 times when the target line passed the 6 N level. The TMS trigger was randomized across the incremental and decremental phases in the left thumb abduction force (i.e. the interstimulus interval was either 10 or 15 s). A practice tracking session without TMS was conducted three times prior to beginning each tracking condition.

Control experiment 1

To clarify whether TMS-induced force disturbance and TCI were modulated in association with unimanual force regulation, we designed the first control experiment in which the participants were instructed to keep the left side force constant at 6 N and to track the target line with only the right side force. The left side force and electromyographic (EMG) activity were averaged separately with reference to the TMS trigger during the right side tracking phase. The effect of TMS on the left tonic force and EMG activity were compared between the force incremental and decremental phases of the right side force.

Control experiment 2

The second control experiment was designed to investigate whether excitation of the crossed corticospinal tract (CST) was always accompanied by excitation of the transcallosal pathway. To this end, the participants also performed unimanual tracking on the left side in addition to the two bimanual conditions (symmetric and asymmetric). TMS was initially delivered at an intensity of 1.5 times the RMT during the unimanual condition (i.e. the right thumb was relaxed). During the second and third trials, one of the bimanual conditions was conducted in a counterbalanced order across the participants. The TMS intensity during the bimanual conditions was adjusted so that the size of the MEP in the right APB was equivalent to that during the unimanual condition (approximately 0.8 × RMT). By comparing the results from the unimanual and bimanual conditions, we evaluated the magnitude of the transcallosal effect elicited by different stimulus intensities under almost equivalent excitabilities of the crossed CST.

Data acquisition and analysis

Behavioral data

Bilateral thumb abductor forces were measured using strain gauges (type KFWS; Kyowa Dengyo Co., Ltd, Japan) attached to the metal pressure plates. The force signal was amplified (DC 5 kHz, gain × 106), displayed concurrently on an oscilloscope for visual feedback, and stored on a computer with a sampling rate of 1 kHz using a CED 1401 A/D converter (Cambridge Electrical Design, UK). The stored force signal was low-pass filtered (Butterworth filter, two-order, 30-Hz cut-off) for offline analysis. To evaluate the tracking performance, the tracking accuracy and tracking synchrony were calculated in an 8-s pre-stimulus period. This period enabled us to evaluate the tracking performance in the absence of the tracking disturbance induced by TMS. Tracking accuracy was assessed by computing the root mean square amplitude of the deviation of the force line from the target line. To estimate the tracking synchrony, a cross-correlogram was constructed from the rates of bilateral force line displacements. The maximum correlation coefficient indicated the degree of synchrony between left and right force line displacements.

To evaluate the magnitude of the tracking disturbance due to TCI, the left abduction force was averaged over 20 TMS triggers in each tracking phase. In an averaged trace, a linear regression line was estimated from a 200-ms pre-stimulus period as the baseline (Fig. 1D and E). The first peak deflection from the baseline, within 200 ms after TMS, was measured as the tracking disturbance. In control experiment 2, as the weak TMS intensity did not elicit an observable disturbance, the tracking disturbance was measured at the point that was identical to that of the unimanual tracking condition. Moreover, to estimate the tracking disturbance in the right force, the peak amplitude of the TMS-induced twitch response was measured (Tazoe et al., 2009).

Electromyography

The EMG signals were recorded from the bilateral APBs. A pair of surface Ag–AgCl electrodes (8 mm in diameter) was positioned 15 mm apart over the muscle belly. The EMG signals were amplified with a bandwidth of 16–3000 Hz, and sampled at a rate of 5 kHz using a CED 1401 A/D converter. In offline analysis, the left side electromyography was rectified and averaged over 20 TMS triggers in each tracking phase, and was subtracted from the control EMG trace obtained at the respective tracking phase without TMS to detect pure EMG suppression (Sakamoto et al., 2006; Fig. 1D and E). The subtracted EMG trace was then transformed to a cumulative sum of the mean trace that was constructed by the consecutive accumulation of the value at each time point, subtracted from the mean value of the 200-ms pre-stimulus baseline (King et al., 2006). The onset and offset of TCI were defined as the point at which at least 10 ms of continuous inhibition started and the first point at which the inhibition retuned to baseline, respectively. The magnitude of TCI was defined as the onset-to-offset amplitude in the cumulative sum of the mean trace. In the bimanual conditions with weak TMS intensity, as TCI was not obvious, the amplitude from the highest point to the lowest point was measured from 30–60 ms after TMS. For the MEP in the right APB, the peak-to-peak amplitude of the unrectified, averaged trace was measured.

Statistical analysis

For statistical comparisons of tracking performance, a two-factor anova with repeated measures was performed with hand (left, right) and tracking condition (symmetric, asymmetric) as factors. Tracking disturbance and TCI were tested using two-factor repeated-measures anova with tracking condition (symmetric, asymmetric) and tracking phase (incremental, decremental) as factors. The significant values reported for the F-values are those obtained after Greenhouse–Geisser correction, when appropriate, and the correlation coefficient epsilon is given only when the degrees of freedom were adjusted. For post-hoc measurements, a pairwise t-test with Bonferroni's correction and Student's t-test were performed. Significant differences were recognized at < 0.05 in all analyses.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References
  9. Supporting Information

Tracking performance

The maximum left and right thumb abduction forces were 45.3 ± 12.7 N (mean ± SD) and 47.9 ± 20.2 N, respectively. Thus, the tracking range corresponded to approximately 2–20% of the participants' maximal effort. Figure 2A shows typical examples of the tracking performance of one participant. Although the tracking error fluctuated slightly according to the tracking cycle (Fig. 2A, third group of traces), the deviation was not significantly different, irrespective of hand or tracking condition (hand, F1,9 = 4.0, = 0.076; tracking condition, F1,9 = 0.000, = 0.985; interaction, F1,9 = 0.019, = 0.895; Fig. 2B). By contrast, the peak correlation coefficient for the rate of force line displacement on the left and right sides showed a slight difference across tracking conditions (Fig. 2C), being significantly higher during the symmetric condition than during the asymmetric condition (< 0.05). These results indicate that, although left–right synchrony was slightly less well correlated during the asymmetric condition compared with during the symmetric condition, tracking accuracy was retained, irrespective of tracking condition.

image

Figure 2. (A) Typical recordings of tracking performance during the symmetric (left) and asymmetric (right) conditions. From top to bottom: trajectories of the force line, rate of force line displacement, and tracking error. Tracking error is the deviation of the force line from the target line. Note, as all traces indicate the line trajectories on the screen, both conditions are shown in a symmetric fashion. The vertical dotted lines indicate the time points of TMS. (B) Average (+SEM) tracking error of the 10 participants. The black and white columns represent the symmetric and asymmetric conditions, respectively. (C) The correlation between the left and right sides for the rate of force line displacement. *< 0.05.

Download figure to PowerPoint

Tracking disturbance and transcallosal inhibition during bimanual tracking

For TMS, the RMT of the right APB was 46.6 ± 7.0% of the maximal stimulator output, i.e. the intensity of the test stimulus was 70.0 ± 10.5% of the maximal stimulator output. The thumb abduction forces at the TMS trigger were constant, irrespective of hand (F1,9 = 0.024, = 0.879) or tracking condition (F1,9 = 0.058, = 0.816), but not for tracking phase (F1,9 = 103.472, < 0.001). TMS to the left M1 induced a marked tracking disturbance that appeared at approximately 100 ms post-stimulation (Fig. 3A, top traces), and this disturbance exhibited significant differences according to the tracking condition (F1,9 = 12.704, < 0.01) and phase (F1,9 = 522.789, < 0.001), but there was no interaction (F1,9 = 0.286, = 0.605). During the force incremental phase, the magnitude of the tracking disturbance was greater during the symmetric condition than during the asymmetric condition (= 2.581, < 0.05; Fig. 3B), but it did not reach significance during the force decremental phase (= 1.557, = 0.153). The rate of force change of the pre-stimulus baseline was not significantly different, irrespective of the tracking condition (F1,9 = 0.245, = 0.632). The disturbance of right thumb tracking due to the twitch response was not significantly different across tracking conditions (main effect, F1,9 = 0.755, = 0.407; interaction with phase, F1,9 = 0.106, = 0.751).

image

Figure 3. Effect of TMS on bimanual force tracking. (A) Typical recordings of left force and EMG traces in a single participant. Each trace is an average of 20 TMS applications. Average (+SEM) of tracking disturbance (B) and TCI (C) in 10 participants. The black and white columns indicate the symmetric and asymmetric conditions, respectively. *< 0.05. CUSUM, cumulative sum of the mean.

Download figure to PowerPoint

We next examined TCI following these stimulations (Fig. 3A, second and third traces). In line with the behavioral measurements, the magnitude of TCI was greater during the symmetric condition than during the asymmetric condition, irrespective of the tracking phase (F1,9 = 8.211, < 0.05; incremental phase, = 2.393, < 0.05; decremental phase, = 2.410, < 0.05; Fig. 3C). The duration of TCI shortened slightly in the asymmetric condition (F1,9 = 12.540, < 0.01) because of the slight prolongation of TCI onset (F1,9 = 8.085, < 0.05; Table 1). The background EMG activity for the 200-ms pre-stimulus baseline did not differ across the tracking conditions (main effect, F1,9 = 1.129, = 0.316; interaction with phase, F1,9 = 1.114, = 0.319; Table 1). The amplitude of the MEP in the right APB was not significantly different, irrespective of the tracking condition (F1,9 = 0.470, = 0.510) or phase (F1,9 = 0.007, = 0.933; Table 1).

Table 1. TMS-induced response of thenar muscle electromyography
 HandIncremental phaseDecremental phase
SymmetricAsymmetricSymmetricAsymmetric
  1. Values represent averages of 10 participants (± SD). *< 0.05, comparison across symmetric and asymmetric conditions.

Baseline electromyography (μV)L147.1 ± 36.2142.1 ± 34.6128.9 ± 25.0116.4 ± 24.9
R165.4 ± 65.2155.6 ± 40.1120.4 ± 33.1107.8 ± 39.6
TCI latency (ms)L32.3 ± 4.334.7 ± 7.631.6 ± 5.435.1 ± 5.0*
TCI duration (ms)L33.4 ± 5.128.6 ± 8.0*34.5 ± 2.430.0 ± 7.1*
MEP amplitude (mV)R4.39 ± 2.94.71 ± 3.94.37 ± 3.04.71 ± 4.3

Effect of contralateral unimanual tracking

To clarify whether the observed effects arising from TMS were due to bimanual motor organization, we examined to what extent the right tracking phase affected force disturbance and TCI during tonic abduction of the left thumb (Fig. 4A). Neither the disturbance of left tonic abduction nor TCI differed with respect to the phase of right side tracking (force disturbance, = 0.754; TCI cumulative sum of the mean, = 0.299, Fig. 4C and E). These findings indicate that simultaneous force regulation with the bilateral thumbs is essential for modulating force disturbance and TCI.

image

Figure 4. (A) Schematic illustration and typical example of the unilateral tracking task. The participants tracked the right target line with their right force while the left force was isometrically held at 6 N. The vertical arrows indicate the time of TMS application during the right force incremental (black) and decremental (gray) phases. Effect of TMS on the left force (B) and APB EMG activity (D). Each trace is an average of over 20 TMS applications. (C and E) Average (+SEM) of seven participants. The black and white columns indicate the incremental and decremental phases, respectively, of the applied right force. CUSUM, cumulative sum of the mean.

Download figure to PowerPoint

Involvement of the crossed corticospinal tract

To determine whether the modulation of TCI on the left APB was associated with excitation of the crossed CST of the right APB, we further examined the relationship between TCI and the activity in the crossed CST. To this end, the participants performed the task using both unimanual tracking and bimanual tracking (Fig. 5A). Moreover, force disturbance and TCI in all three tracking conditions were compared in a situation under which almost equal MEPs were obtained in the right APB ('Materials and methods'). TMS intensity under the bimanual conditions was 83.0 ± 3.6% RMT (range 70–100%). The size of the MEPs was not significantly different across the tracking conditions (incremental phase, F2,12 = 1.259, = 0.319; decremental phase, F2,12 = 0.587, = 0.571; Fig. 5D and G). Nevertheless, there were marked differences in both force disturbance (F2,12 = 90.05, < 0.001; Fig. 5E) and TCI (F1.09,6.55 = 35.08, ε = 0.546, < 0.001; Fig. 5F). Although force disturbance and TCI were observed clearly in the unimanual condition, they were virtually obscured during both of the bimanual conditions. Force disturbance and TCI in the unimanual condition were significantly greater than in both bimanual conditions (force disturbance, all < 0.001; TCI, all < 0.001). However, there was no difference between the bimanual symmetric and asymmetric conditions (force disturbance, both phases, > 0.05; TCI, both phases, > 0.05). These findings suggest that the task-related modulation of TCI was dissociated from the excitability of the crossed CST.

image

Figure 5. (A) Schematic illustrations of unimanual left tracking (left), bimanual symmetric tracking (center), and bimanual asymmetric tracking (right). The intensity of TMS in each tracking condition is shown under the respective illustration. Typical recordings of the left force (B), TCI of the left APB (C), and MEP of the right APB (D) in a single participant. (E–G) Respective averages (+SEM) of seven participants. The black, white, and gray columns indicate the unimanual, bimanual symmetric, and bimanual asymmetric conditions, respectively. ***< 0.001.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References
  9. Supporting Information

Although it has long been demonstrated that bimanual motor performance is mediated by the function of the CC (Preilowski, 1972; Franz et al., 1996; Eliassen et al., 1999, 2000; Stephan et al., 1999; Serrien et al., 2001; Kennerley et al., 2002; Diedrichsen et al., 2003; Johansen-Berg et al., 2007; Muetzel et al., 2008), little is known about the neural activity of the transcallosal circuit during bimanual motor actions (Soteropoulos & Baker, 2007). Recently, Yedimenko & Perez (2010) demonstrated that interhemispheric inhibition, as assessed by paired-pulse TMS, is modulated according to the direction of static forces of bilateral index fingers. Our experiment further expands this notion to the dynamic regulation of bimanual forces.

In the present study, we demonstrated that TCI between the motor cortices was modulated according to the condition of bimanual force regulation. TCI was greater when bimanual force regulation was performed in a symmetrical manner compared with when it was performed in an asymmetrical manner. In line with this, the perturbation of force tracking performance induced by TMS over the ipsilateral M1 was greater during the symmetric condition than during the asymmetric condition. Therefore, the transient disruption of right M1 activity due to TCI could mainly account for the modulation of the left tracking disturbance.

Furthermore, our findings could be a manifestation of the specific neural organization of the transcallosal inhibitory circuit for bimanual force control. Although TCI showed a different magnitude depending on whether TMS was applied during the left force incremental phase or decremental phase, the magnitude of TCI was generally larger during the symmetric condition than during the asymmetric condition, irrespective of the tracking phase. In addition, TCI of tonic muscle contraction was not modulated by unimanual force regulation of the right thumb (Fig. 4). These findings demonstrated that simultaneous force regulation with different coordination conditions accounts for the observed modulation of TCI, but unilateral force regulation was insufficient to induce such modulation.

Transcallosal inhibition during bimanual force control

The most important finding in the present study was that TCI during the symmetric condition, which required synchronous bilateral force regulation of the thumb, was greater than during the asymmetric condition. However, this finding may not be in line with the accepted role of TCI between the motor cortices. During a unimanual action, one of the most important functions of TCI is to prevent unwanted motor activity of the muscles contralateral to the acting hand (Mayston et al., 1999; Duque et al., 2007; Hübers et al., 2008; Giovannelli et al., 2009). Accordingly, this consideration might lead us to predict that TCI is weaker during symmetric muscle contractions than during asymmetric muscle contractions (Meister et al., 2010). Nevertheless, our results contradicted this prediction, suggesting that the transcallosal neural circuit has a specific strategy for bimanual motor actions, which is different from that for unimanual motor actions. This hypothesis is also supported by the fact that unimanual force regulation with the contralateral thumb was unable to induce the observed modulation of TCI.

The most plausible explanation for our results may be the characteristics of the present task in which bilateral homonymous muscles (i.e. APBs) acted as the prime movers in the symmetric and asymmetric conditions. Even while a muscle force is gradually released, the M1 is likely to play an important role in the regulation of an isometric force (Toma et al., 1999; Spraker et al., 2009). Therefore, it might not be an appropriate strategy for the isometric force regulation task to simply suppress the activity of the contralateral M1.

As another possibility, visual information might be involved in our findings. Visual feedback from an action has been demonstrated to have a prominent effect on the stability of bimanual coordination (Byblow et al., 1999; Mechsner et al., 2001). In the present study, the required movement of the force line was identical between the symmetric and asymmetric conditions to perform force regulation with as equal accuracy as possible ('Materials and methods'). Accordingly, the mapping rule for transforming the direction of force to the direction of the line movement on the oscilloscope was quite different across the symmetric and asymmetric conditions. The congruency of the visual feedback and the actual behavior has a severe impact on the excitability of cortical motor circuits (Johansson et al., 2006). Furthermore, the interhemispheric neural interactions seem to be influenced by the action direction in the extrinsic coordinated frame. The magnitude of interhemispheric interactions changes according to whether the direction of a side of action is egocentrically congruent to that of the contralateral tested side (Duque et al., 2005; Yedimenko & Perez, 2010). Therefore, if the external framework of a hand action is involved in the neural processing of visual information, the mechanism of visuomotor transformation might influence the excitability of the transcallosal circuits.

Using static contraction of bilateral index finger muscles, Yedimenko & Perez (2010) recently demonstrated that interhemispheric inhibition was larger when both the left and right index finger forces are directed toward the body midline compared with when left and right forces are directed in the same direction with respect to an allocentric coordinated frame. This result is in agreement with our findings that interhemispheric inhibitory interactions changed according to the direction of the left and right forces. However, care should be taken to interpret the symmetry of the force directions. According to the allocentric coordinated frame (i.e. parallel movements are recognized as symmetrical), the previous finding is compatible with ours (Yedimenko & Perez, 2010). However, the modulation of interhemispheric inhibition contradicts our findings of TCI modulation in terms of an egocentric coordinated frame (i.e. mirror-directed movements are recognized as symmetrical). Further study is needed to clarify this effect. There are several discrepancies in the methodology for examining interhemispheric interactions, including tested muscle (thumb vs. index finger), contraction manner (static and dynamic), TMS techniques (single-pulse vs. paired-pulse), directions of forces and cursors (up–down vs. left–right), and the contribution of antagonistic muscles. In our study, bilateral thumb abductions required almost the same amount of effort. However, the magnitude of left and right contractions was different in the previous experiment (Yedimenko & Perez, 2010). Thus, it remains unclear whether those different parameters account for the discrepant findings regarding interhemispheric interactions.

Involvement of the uncrossed corticospinal tract

Animal experiments demonstrated that some neurons in M1 have uncrossed motor pathways to the ipsilateral limb muscles (Edgley et al., 2004; Lacroix et al., 2004; Jankowska & Stecina, 2007; Brus-Ramer et al., 2009; Yoshino-Saito et al., 2010). In line with these findings, human experiments demonstrated that an MEP can be elicited at the muscle ipsilateral to the M1 where TMS was applied (Wasserman et al., 1994; Ziemann et al., 1999; Kagerer et al., 2003). The close latency between the ipsilateral and contralateral MEPs indicated that TMS was able to excite the ipsilateral muscle without going through a transcallosal circuit. Although we cannot completely exclude the possible involvement of such uncrossed motor pathways or other subcortical mechanisms, we argue that the ipsilateral motor response obtained in the present study resulted from the transcallosal motor circuit.

First, studies conducted on callosotomy patients demonstrated that the CC is essential for producing an inhibitory response in ipsilateral hand muscles, with a latency of approximately 30 ms (Meyer et al., 1995, 1998). The latency in the present study was almost identical to that in these lesion studies. Second, to generate an ipsilateral MEP consistently requires a relatively high stimulus intensity and strong activation of the muscle at which the ipsilateral MEP is evoked (Wasserman et al., 1994; Ziemann et al., 1999), and the probability of obtaining an ipsilateral MEP is muscle-dependent. For intrinsic hand muscles, an ipsilateral MEP was frequently observed in the first dorsal interosseous, but not in the APB, even at high TMS intensity and muscle activation (Ziemann et al., 1999; Jung & Ziemann, 2006). Indeed, we did not observe any ipsilateral MEP components in any of the participants (Figs 3-5). Third, our control experiment demonstrated that the magnitude of the ipsilateral inhibitory response was independent of the excitation of the crossed CST, suggesting that this inhibition was derived from supraspinal sources. Regarding this notion, it was reported that the cortical neurons of the CST were partly independent of those of the CC (Catsman-Berrevoets et al., 1980; Lang et al., 2004; Lee et al., 2007). These findings imply that the absence of an ipsilateral inhibitory response with weak TMS reflected the failure of CC neurons to be excited, even though crossed CST neurons were excited. This notion is consistent with previous findings demonstrating that the threshold to induce TCI was higher than the RMT for contralateral MEPs (Ferbert et al., 1992; Trompetto et al., 2004).

Functional significance

The stability of bimanual cyclic movement with different coordination conditions has been expressed by dynamic pattern theory, such as the Haken–Kelso–Bunz model (Haken et al., 1985; Schöner & Kelso, 1988). Based on this model, the phase shift between left and right cycles critically affects the stability of bimanual action. However, the bistable characteristic can be observed at low frequency; the bimanual action is stable at both in-phase and 180° out-of-phase. In the present study, the participants performed the symmetric and asymmetric force tracking tasks with almost equivalent accuracy, although synchrony of the left–right tracking trajectory was slightly lower during the asymmetric condition. This suggests that performance degradation due to bimanual constraint in the asymmetric force coordination was relatively low and was compensated for by the strategy of bimanual regulation, which was different from that in the symmetric condition. On the basis of this context, it may be that the observed modulation of TCI was due to an aspect of neural organization necessary for implementing a motor strategy to evade the constraints imposed on bimanual actions.

As previous studies demonstrated, a lack of transcallosal communication leads to either deterioration (Serrien et al., 2001; Kennerley et al., 2002) or improvement (Franz et al., 1996; Eliassen et al., 1999; Diedrichsen et al., 2003) in bimanual task performance according to the respective requirement for spatiotemporal coordination. That is, the functional importance of transcallosal neural communication depends on whether the coordination of left- and right-sided movements is required. In support of this, we recently observed that TCI modulation was influenced by the coordination requirement of left and right hands during a bimanual task (T. Tazoe, S. Sasada & T. Komiyama, unpublished observation). Following this line, as our experiment was not designed to manipulate the required coordination between the symmetric and asymmetric conditions, two different interpretations may be possible for our findings of TCI modulation. One is that, during the asymmetric condition, the inhibitory effect between the motor cortices decreased, uncovering the excitatory interhemispheric neural communication. The CC is reported to have both excitatory and inhibitory transcallosal circuits (Asanuma & Okuda, 1962; Ugawa et al., 1993; Hanajima et al., 2001; Bäumer et al., 2006). Recent research also indicated a possibility that inhibition from the contralateral motor cortex was decreased to acquire tight interhemispheric neural coupling during coordinated, asynchronous, bimanual movements (Duque et al., 2010). Therefore, it is likely that degraded TCI may be beneficial for improving spatiotemporal bimanual coordination, even if the bimanual action is carried out asymmetrically.

Another interpretation is that the decrease in TCI is a manifestation of the general suppression of the absolute impact of transcallosal interference. It was proposed that the gain control of excitatory and inhibitory transcallosal discharges countervails the neural interference between the motor cortices (Rokni et al., 2003). This allows each motor cortex to work independently without any interference from the contralateral cortex. Regarding this notion, callosotomy patients reportedly acquire a high degree of independence for movements on each side during bimanual movements at the expense of their ability to coordinate bimanual movements (Eliassen et al., 1999). Thus, when movements on each side have their own respective task goals, it should be beneficial that the movements on each side are organized individually and that they do not interfere with each other. Recent behavioral studies reported that such motor organization was implemented depending on the task requirement (Diedrichsen et al., 2004; Diedrichsen, 2007; Mutha & Sainburg, 2009). Given these reports, our findings might provide a good perspective of the CC circuit as a key structure influencing task-dependent bimanual interactions, even though the observed modulation of TCI did not demonstrate directly the extent of interhemispheric connectivity.

Methodological consideration

Although we demonstrated that the symmetric condition exhibited larger TCI than the asymmetric condition, it could be claimed that the transcallosal inhibitory circuit was occluded during asymmetric condition. A relatively high intensity of TMS is required to elicit TCI. Therefore, if the transcallosal circuit is highly activated during the asymmetric condition, such high TMS intensity might produce some effects that give rise to the underestimation of TCI. We cannot completely rule out this possibility from a physiological point of view, even though we confirmed that TCI was further increased as TMS intensity was > 150% RMT during static muscle contraction (Supporting Information Fig. S1).

In addition, we need to consider the data-processing methods for both force and EMG averaging. The present study adopted a signal averaging approach to increase the signal-to-noise ratio of the TMS-induced response on the ongoing EMG activity. It is true that the temporal profile of averaged force trace may not be representative of any single trial. However, it is also true that a single trial was not enough to properly detect TCI onset and offset. To obtain reliable data, we averaged more than 20 signals for all experiments, which improved our ability to assess TCI. This merit of this averaging approach was also applicable to analyzing tracking force. Force trajectory in a single trial was jagged because of the slowness of force tracking speed, and the averaging approach allowed us to extract pure force trajectory and the disturbance produced by TMS. However, unconscious corrective response in successive trials may contain the disturbed response of tracking force. In the redundant system of motor control, the data averaging approach may obscure components observed in a single trial.

Abbreviations
APB

abductor pollicis brevis

CC

corpus callosum

CST

corticospinal tract

EMG

electromyographic

M1

primary motor cortex

MEP

motor evoked potential

RMT

resting motor threshold

TCI

transcallosal inhibition

TMS

transcranial magnetic stmulation

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References
  9. Supporting Information

Supporting Information

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgement
  8. References
  9. Supporting Information

Please note: As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer-reviewed and may be re-organized for online delivery, but are not copy-edited or typeset by Wiley-Blackwell. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
ejn12026-sup-0001-FigS1.tifimage/tif222KFig. S1. TCI during static thumb abduction.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.