Retention of motor learning can be enhanced or degraded by subsequent performance of a different task. Neurophysiologically this may reflect interference in synaptic plasticity by ongoing neural activity in the brain. Here we demonstrate that N-methyl-D-aspartate (NMDA) dependent aftereffects of repetitive transcranial magnetic stimulation (rTMS) also are subject to interference effects, suggesting that it may be possible to investigate these basic mechanisms in the intact human brain. We measured the motor-evoked potential (MEP) amplitude and short-interval intracortical inhibition (SICI) in the first dorsal interosseous (FDI) muscle after continuous or intermittent theta burst (cTBS/iTBS) forms of rTMS. In resting subjects, cTBS depressed MEPs and reduced SICI for about 20 min, whereas iTBS had the opposite effect. However, if subjects contracted the FDI during TBS, then effects on the MEP were abolished, although effects of cTBS on SICI remained. Contraction immediately after TBS enhanced the facilitatory effect of iTBS and reversed the usual inhibitory effect of cTBS into facilitation. Contraction 10 min after cTBS (iTBS not tested) had only a transient (3–4 min) effect on MEPs. These interactions with behavior may relate to mechanisms of interference between learning paradigms in human and be similar to effects on synaptic long-term potentiation/depression described in animal experiments.
It has long been known that long-term potentiation (LTP) and long-term depression (LTD) induced by electrical stimulation of nervous pathways in animal preparations can be reversed by patterns of stimulation that on their own have no long-term effects on synaptic transmission. These effects are known as depotentiation and dedepression, respectively. For example, LTP can be reversed in hippocampus by a low-frequency train of stimulation at 1–2 Hz (Chen et al. 2001; Huang et al. 2001), particularly if the latter is applied within minutes of the end of the induction of LTP. More recently, there have been a number of reports that LTP can also be reversed by natural patterns of activation in conscious animals (Xu et al. 1998; Manahan-Vaughan and Braunewell 1999; Zhou et al. 2003). Thus, electrically elicited LTP in hippocampus was reversed in rats when they entered a new environment. Zhou and Poo (2004) suggested that this type of effect could enhance the emergence of connectivity based on patterned inputs and reduce the error in activity-dependent circuit refinement resulting from random inputs and activities.
Similar processes may explain interference between tasks in experiments involving motor learning. Retroactive interaction is well recognized in all theories of memory consolidation (Wixted 2004; Krakauer and Shadmehr 2006). Such interactions are often negative, as illustrated by the fact that learning of a single motor skill is often impaired if it is followed by training in a second task (Lechner et al. 1999; Goedert and Willingham 2002). However, retroactive interactions may also be positive in that some tasks facilitate prior learning (Bock et al. 2001); indeed, sleep benefit (Walker et al. 2002) can be viewed as another example of how later behaviors affect consolidation of previously learned skills.
The behavior of conscious humans is easy to modify by prior instruction, and hence in many ways, humans are a highly suitable model in which to investigate the influence of physiological activity on synaptic plasticity. However, this depends on the existence of an optimal tool to produce reliable effects on synaptic plasticity in human subjects. To date, various forms of repetitive transcranial magnetic stimulation (rTMS) have been proposed to produce LTP- and LTD-like effects in human motor cortex (Ziemann 2004); hence, these may be a useful practical way of investigating interactions with physiological activity. In our previous experiments (Huang et al. 2005), we developed a number of novel paradigms of rTMS, which we have called theta burst stimulation (TBS), based on theta burst pattern, which can alter motor cortical excitability quickly and efficiently. When stimulation is delivered to the “motor hot spot” for the first dorsal interosseous (FDI) muscle, 20 s of continuous TBS (cTBS) suppresses motor cortical excitability for 20 min, whereas intermittent TBS (iTBS) produces facilitatory effect on motor system that outlasts the conditioning by 20 min. Together with the NMDA dependency of TBS (Huang et al. 2007), we believe the aftereffects of TBS, like those of other forms of rTMS (Ziemann 2004), are likely to reflect types of synaptic plasticity seen in animal preparations. In this study, we explored how the aftereffects of TBS were affected by voluntary muscle contraction applied during TBS or after termination of TBS. The results are consistent with the idea that physiological activity in the brain interacts with synaptic plasticity–like processes in a way very similar to that in animal experiments. We propose that the effects are caused by voluntary activation of neuronal networks in the motor cortex in ways that depotentiate or dedepress the changes in synaptic effectiveness evoked by TBS. If so, this may be a useful paradigm to study the interaction between ongoing physiological activity and learning in the intact human brain.
Materials and Methods
All subjects for these experiments were right-handed healthy volunteers between 23 and 43 years of age (mean age: 30.9 ± 6.8 years) and gave their informed consent for the experiments. The project protocol was approved by the Joint Ethics Committee of the National Hospital for Neurology and Neurosurgery and the Institutional Review Board of the Chang Gung Memorial Hospital.
Stimulation and Recording
For all experiments, subjects were seated in a comfortable chair. Electromyographies were recorded using Ag–AgCl electrodes from the right FDI (the dominant hand in all subjects). EMG activity was sampled at 5000 Hz and recorded with a gain of 1000 and 5000 and filtered with a band-pass filter (3 Hz–2 kHz) through a Digitimer D360 amplifier (Digitimer Ltd, Welwyn Garden City, Herts, UK).
Magnetic stimulation was given using hand-held figure-eight coils with an outer winding diameter of 70 mm (Magstim Co., Whitland, Dyfed, UK). Single and paired pulses were delivered by Magstim 200 machines that produce monophasic pulses, whereas rTMS was delivered using a Magstim rapid stimulator connected to 4 booster modules that produce biphasic pulses. Stimulation was delivered over the motor hand area with the coil placed tangentially to the scalp with the handle pointing posteriorly. The motor hand area was defined as the location on the scalp where magnetic stimulation produced the largest motor-evoked potential (MEP) from the contralateral FDI when the subject was relaxed (the motor hot spot). The stimulation intensity was defined in relation to the active motor threshold (AMT) of the subject. The AMT was defined for each Magstim machine separately as the minimum intensity of single-pulse stimulation required to produce an MEP of greater than 200 μV on more than 5 out of 10 trials from the contralateral FDI while the subject was maintaining a voluntary contraction of about 20% of maximum in the FDI. Raw EMG signal on the screen was provided as visual feedback to the subject to help maintain a constant muscle contraction of the correct force.
One-minute trains of constant current electrical nerve stimuli were applied (Digitimer) to the ulnar nerve in the right wrist in order to mimic volitional contraction of the FDI muscle.
Single-pulse TMS was used to evaluate the effect of TBS on MEPs elicited in the contralateral FDI. The intensity of these “probe” TMS pulses was set at an intensity that produced MEPs of 1 mV in control conditions, unless described separately. MEP size and short-interval intracortical inhibition (SICI) were always assessed with the subject relaxed.
The TBS Paradigms
The basic TBS pattern was a burst containing 3 pulses of 50 Hz magnetic stimulation given every 200 ms (i.e., at 5 Hz). Two different stimulation paradigms were used in this experiment: 1) iTBS: the basic pattern given in a short train lasting 2 s (i.e., 10 bursts), repeated every 10 s for 20 cycles. 2) cTBS: the basic pattern given in a continuous train lasting 20 s (i.e., 100 bursts). Because we investigated interactions with behaviors at different times after TBS, we chose iTBS with 600 pulses and cTBS with 300 pulses because their aftereffects are of very similar duration (∼20 min). TBS was delivered to the motor hand area at an intensity of 80% AMT.
Baseline MEP recording was performed using 30 pulses delivered every 4.5–5.5 s. TBS was then given with the subject relaxed (TBS-relaxed), and MEP size was assessed using single pulses of TMS delivered in trains of 12 pulses given every 4.5–5.5 s every 1 min for 6 min and then every 2 min until 23 min after the end of TBS. Nine subjects (6 men, 3 women; mean age 31 ± 8 years) participated in the TBS experiment.
Contraction during TBS Conditioning (cTBS-Contract or iTBS-Contract)
TBS was given with the subject performing a voluntary contraction of the FDI contralateral to the site of stimulation by squeezing a 3-cm block between the thumb and index finger. This voluntary contraction was maintained throughout the period of TBS at about 10% of maximal force with visual feedback provided to the subject to encourage a constant force of contraction throughout stimulation. MEP size was tested on all 9 subjects in cTBS-contract, 7 of them (5 men, 2 women; mean age 32 ± 5 years) in iTBS-contract, and SICIs were tested on 7 subjects (4 men, 3 women; mean age 27 ± 3 years) in the motor hand area before and after TBS-contract. The technique of SICI measures the influence of a subthreshold “conditioning” pulse of TMS given over the hand motor area on a subsequent suprathreshold “test” pulse given over the same area. It is believed that SICI is produced by activity of γ-aminobutyric acidAergic interneurones in the cortex (Ziemann et al. 1996, 1998; Chen et al. 1998; Hanajima et al. 1998). We assessed SICI using the double-pulse method described by Kujirai et al. (1993). SICI was evaluated at an interstimulus interval (ISI) of 2 ms using a conditioning intensity of 80% AMT. We chose an ISI of 2 ms so that we could compare the present data with that reported in our previous paper in totally relaxed muscle (Huang et al. 2005). We adjusted the intensity of the test stimuli while assessing SICI after TBS to maintain the amplitude of test MEPs at approximately 1 mV. MEP size and SICI were assessed in separate sessions. Two blocks of baseline MEP or SICI were recorded. Each block was recorded with 10 trials of each condition every 4.5–5.5 s randomly intermixed with controls. After iTBS, MEP size and SICI were recorded every 4 min until 20 min following conditioning. After cTBS, MEP size and SICI were recorded every 5 min until 25 min after conditioning. We chose slightly shorter intervals for testing after iTBS in order to capture the initial peak and decline in the MEP curve that was evident in our previous paper (Huang et al. 2005) and can be seen in the present Figure 3A.
Contraction Immediately after TBS (cTBSc0 or iTBSc0)
In this part, subjects were asked to perform a voluntary contraction (10% maximum) of the FDI muscle contralateral to the site of stimulation for 1 min with visual feedback provided to the subject immediately after TBS-relaxed. Baseline MEP recording was performed using 30 pulses delivered every 4.5–5.5 s. TBS was then given and followed by the 1-min contraction. MEP size was assessed using single pulses of TMS delivered in trains of 12 pulses given every 4.5–5.5 s every 1 min for 5 min and then every 2 min until 23 min after the end of TBS. Same 9 subjects participated in the cTBS experiment, and only 7 of them (5 men, 2 women; mean age 32 ± 6 years) participated in the iTBS experiment. MEP size was always assessed with the subject relaxed.
We also did a control study using sham cTBS followed by the 1-min contraction in 7 of the 9 subjects (5 men, 2 women; mean age 31 ± 7 years). The protocol is exactly the same as cTBSc0, except cTBS was replaced by a sham stimulation using a coil tilted 90 degrees away from the scalp.
In addition, we assessed SICI using a similar method to that described above before and after cTBSc0 on 7 subjects (5 men, 2 women; mean age 26 ± 5 years). We adjusted the intensity of the test stimuli after conditioning to maintain the amplitude of test MEPs at approximately 1 mV. Two blocks of baseline SICI were recorded. SICI was recorded every 5 min until 25 min following conditioning.
Contraction at 10 Min after the End of cTBS (cTBSc10 Only)
The time course of changes in MEP size elicited from the contralateral FDI was measured using a similar method to that described in TBS paradigms on the same 9 subjects. The only difference is that the subjects were requested to perform a voluntary contraction of the FDI (10% maximum with visual feedback) contralateral to the site of stimulation at 10 min after the end of cTBS-relaxed for 1 min (cTBSc10). Other than the minute when subjects were asked to activate the FDI muscle, subjects were relaxed for the MEP assessment.
cTBS Followed by 1 Min Peripheral Stimulation Mimicking the Contraction
The protocol of this experiment is exactly the same as the one described in the session of “contraction immediately after TBS,” except that the voluntary contraction was replaced by a 1-min electrical stimulation of the ulnar nerve at wrist in the conditioned hand. The electrical stimuli had a duration of 500 μs, and the intensity used was that which produced a contraction of about 10% of maximal force in FDI muscle. This was achieved by stimulating at 17 Hz, which produced a comfortable fused contraction of the FDI muscle. Seven of the 9 subjects (5 men, 2 women; mean age 31 ± 7 years) participated in this part.
Contraction of the Other Muscles Immediately after cTBS
In this session, the protocol was exactly the same as the one described in “contraction immediately after TBS” paradigm, except that voluntary contraction of the FDI was replaced by voluntary contraction of the abductor digiti minimi (ADM) muscle in the same hand as the targeted FDI (cTBSc0-ADM) or the biceps muscle ipsilateral to the site of stimulation by flexing the elbow (cTBSc0-biceps). Surface EMG of the FDI was continuously monitored to ensure that the FDI was relaxed while the other muscle was being activated. Five subjects (2 men, 3 women; mean age 30 ± 6 years) not included in experiments above were recruited. The experiment of cTBSc0 was also repeated in these subjects for comparison.
Data were analyzed using SPSS for Windows version 11.0. Repeated measures analysis of variance (ANOVA) was used to compare variables before and after TBS. Statistics for the data were performed on absolute amplitude values, whereas the averaged baseline MEP amplitudes were normalized to 1 in the graphs to make the graphs easy read. Because we normalized baseline MEP amplitudes after they were averaged, error bars of baseline MEP were presented in the figures. A P < 0.05 was considered statistically significant.
Application of TBS at Rest Compared with during Voluntary Contraction
Figure 1 compares the after effects of applying TBS at rest (TBS-relaxed) with TBS given during a static voluntary contraction of FDI (TBS-contract). The upper panel shows the effects of iTBS and the lower panel the data from cTBS. At rest, iTBS facilitated MEPs, as described in our previous study (Huang et al. 2005); cTBS suppressed MEPs. However, if TBS was applied during contraction, then neither iTBS nor cTBS had any effect on MEP amplitude. This was confirmed in separate 2-factor ANOVAs for iTBS and cTBS. There was a significant time × contraction interaction for both the iTBS (F5,30 = 2.92, P = 0.029) and cTBS data (F5,40 = 5.91, P = 0.000), indicating that time had a different effect on MEPs according to the state of muscle contraction when TBS was applied.
Separate 1-factor ANOVAs showed that MEPs were enhanced by iTBS-relaxed for up to 20 min (F15,120 = 2.79, P = 0.001), whereas they were suppressed by cTBS-relaxed for 20 min (F15,120 = 4.35, P = 0.000). No significant effects were noted on MEP size at any time point if TBS was applied during contraction (cTBS: F5,30 = 1.38, P = 0.261; iTBS: F5,40 = 0.959, P = 0.454).
Figure 2A shows the data for SICI following iTBS performed during voluntary contraction. There was no significant change in SICI (F5,30 = 1.78, P = 0.231) following iTBS performed during voluntary contraction. Figure 2B shows that cTBS, given with the subjects actively contracting, reduced the amount of SICI for the following 20 min or more (1-factor ANOVA on the time course: F5,30 = 5.83, P = 0.001). Post hoc paired t-tests with Bonferroni correction showed that SICI was significantly reduced at 10–15, 15–20, and 20–25 min after the end of conditioning.
Because 7 of the subjects participated in the present experiments as in the previous paper (Huang et al. 2005), we got the chance to test the test–retest reliability of TBS-relaxed. Postconditioning data were normalized to baseline MEP and grouped into 3 phases (phase 1: 0–7 min, phase 2: 8–15 min, and phase 3: 16–23 min) to test the correlation between 2 sets of data. In iTBS-relaxed, they correlated well at phase 1 (Pearson correlation = 0.86) and phase 2 (0.80) but less correlated at phase 3 (0.75). In cTBS-relaxed, highly significant correlation between 2 experiments was found at all 3 phases. (Pearson correlation = 0.94, 0.94, and 0.82, respectively)
Contraction Immediately after TBS (iTBSc0 and cTBSc0)
Figure 3A shows the results of iTBS-relaxed and iTBS followed immediately by a 1-min contraction. It appears as if the aftereffect of iTBS was stronger when it was followed by contraction. This was confirmed in the ANOVA: there was a significant interaction between iTBS-relaxed and iTBSc0 (F14,84 = 2.45, P = 0.006).
Figure 3B compares the amplitude of MEPs elicited before and after cTBS-relaxed, cTBSc0, or contraction after sham cTBS. Surprisingly cTBSc0 reversed the effect of cTBS from inhibition to facilitation, whereas sham TBS followed by contraction had no significant effect on MEPs (1-factor ANOVA: F14,84 = 0.40, P = 0.971). Separate 2-factor ANOVAs showed that there was a significant interaction between cTBS-relaxed and cTBSc0 and also between cTBSc0 and contraction only (F14,98 = 3.60, P = 0.000; F14,84 = 1.96, P = 0.031, respectively). A 1-way ANOVA on the data from cTBSc0 showed that the facilitatory effect on MEPs was highly significant (F14,98 = 2.47, P = 0.005) and lasted for around 15 min.
In contrast to cTBS-relaxed that we previously reported (Huang et al. 2005) (2-factor ANOVA, time × contraction interaction [F5,30 = 2.40, P = 0.048]), cTBSc0 had no effect on SICI (1-factor ANOVA on the time course: F5,30 = 0.47, P = 0.796) (Fig. 4).
Contraction 10 Min after the End of cTBS (cTBSc10)
Figure 3C compares the effect of cTBSc10 with cTBS-relaxed. A 2-way ANOVA showed there was a significant interaction between the main factors of time and contraction (F15,120 = 2.64, P = 0.002). Post hoc paired t-tests with Bonferroni correction on normalized data showed that it was due to the first minute after contraction.
cTBS Followed Immediately by 1 Min Peripheral Stimulation Mimicking Voluntary Contraction
When cTBSc0 was replaced by cTBS-relaxed followed by 1-min stimulation of the ulnar nerve, the effect of cTBS was not reversed. MEPs were significantly suppressed for 20 min (F14,84 = 1.94, P = 0.033) (Fig. 3D). A 2-factor ANOVA comparing cTBS-relaxed with cTBS-relaxed plus ulnar nerve stimulation showed no significant time × group interaction (F14,84 = 0.60, P = 0.855) or a main effect of group (F1,6 = 0.86, P = 0.390).
Contraction of the Other Muscles Immediately after cTBS
When cTBSc0 was replaced by cTBS-relaxed followed by a voluntary contraction of the biceps muscle ipsilateral to the stimulation (cTBSc0-biceps), the effect of cTBS was not reversed (Fig. 5A). A 2-factor ANOVA comparing cTBS-relaxed with cTBS-relaxed plus a voluntary contraction of the biceps muscle ipsilateral to the stimulation showed no significant time × group interaction (F14,56 = 1.27, P = 0.356) or a main effect of group (F1,4 = 0.60, P = 0.480). Figure 5A also shows how, in contrast, the usual inhibition that is seen after cTBS-relaxed is reversed by voluntary contraction of the FDI muscle (cTBSc0) in the same subjects.
Similarly, there was no reversal of the cTBS effect when the activated muscle was replaced by the ADM muscle in the same hand as the targeted FDI (cTBSc0-ADM). Indeed, there was a tendency to facilitate responses in ADM (Fig. 5B).
The present results show that 1-min voluntary contraction of the target muscle (FDI) during or after TBS has a profound influence on the size and direction of the aftereffects produced by TBS. Contraction during TBS abolishes all effects on the MEP, although effects of cTBS on SICI remain. Contraction immediately after TBS enhances the facilitatory effect of iTBS and reverses the inhibitory effect of cTBS from suppression to facilitation. Contraction 10 min after cTBS (iTBS was not tested) reduced MEP suppression, but only for 3–4 min.
Previous papers (Di Lazzaro et al. 2005; Huang et al. 2005) have provided strong evidence that the aftereffects of TBS are due to lasting action on the excitability of circuits in the motor cortex. Thus, the present results suggest that muscle contraction interacts in some way with the cortical circuits responsible for the aftereffects. This interaction seems likely to involve volitional motor output to the contracting muscle because there was no effect if the muscle was made to contract by electrical stimulation of peripheral nerve. In addition, the fact that contraction of the biceps or ADM muscles failed to change the usual pattern of aftereffects suggests that attentional factors (Stefan et al. 2004) are also unlikely to have been responsible.
Contraction during TBS Conditioning
Effects on MEPs
We would like to put forward 2 hypotheses to account for the effects on MEPs. First, it may be that, like LTP and LTD at central synapses in animal experiments (Artola et al. 1990; Randic et al. 1993; Ngezahayo et al. 2000), our aftereffects are sensitive to the physiological state of neurons stimulated during TBS. Thus, contraction may have changed the membrane potential or Ca2+ concentration of postsynaptic neurons, and these affected the response to the conditioning protocol. A second possibility is that contraction activates the same set of synaptic connections as those stimulated by TBS. This could cause a “busy line” effect, in which the extra activation evoked by TBS was negligible.
Whatever the mechanism, the data may help explain why induction of synaptic plasticity in the neocortex of freely moving animals is more difficult than induction in brain slices or anesthetized animals (Trepel and Racine 1998; Froc et al. 2000). It may be that part of this difficulty is due to unregulated amounts of synaptic activation in conscious animals. If so, it may turn out that the human cortex is a much better model for exploring this type of effect, given the ease with which we can control behavior by simple instruction.
Effects on SICI
iTBS enhances SICI and cTBS depresses SICI if applied in subjects at rest (Huang et al. 2005). In the present experiments, iTBS-contract had no effect on SICI whereas cTBS-contract continued to reduce SICI. Because the same subjects participated in the present experiments as in the previous paper and because we measured SICI at the same times after TBS in both cases, we were able to compare the results directly. When cTBS was applied at rest, SICI was maximally reduced at 5–10 min, whereas if cTBS was applied during contraction, then the maximal effect was at 15–20 min. This was confirmed by a 2-factor ANOVA that showed there was a significant interaction between the main factors of time and contraction (F5,30 = 2.77, P = 0.036).
Why did contraction during cTBS abolish the effect on the MEP but not on SICI? One possibility relates to the fact that although contraction of FDI increases excitability of MEP pathways, it also reduces excitability in the SICI pathway to FDI (Ridding et al. 1995). In analogy with the explanation above for MEPs, we might imagine that the membrane potential of SICI neurons was either unaffected or hyperpolarized during FDI contraction, whereas that of the MEP pathway was depolarized. This differential change in their membrane potential could then account for the difference in the effect of contraction on SICI and MEP. A comparable result has been seen using a session of 5 Hz rTMS to condition SICI. When given in the relaxed state, 5 Hz rTMS caused a lasting reduction in SICI (Di Lazzaro et al. 2002), which was enhanced if the rTMS was applied during contraction of the target muscle (Fujiwara and Rothwell 2004).
The differential effects of voluntary contraction on the responses to TBS in MEPs and SICI underline the point that the effects of any rTMS protocol depend on the physiological state of the neuronal populations at the time the stimulus is applied. This is evident in concepts such as “homeostatic” plasticity (Siebner et al. 2004; Ziemann et al. 2004) where the previous history of neural activity determines the ease of producing LTP/LTD-like synaptic effects. It also explains why not all rTMS protocols are affected by ongoing muscle activity: protocols such as 1 Hz rTMS that typically use much higher stimulus intensities and hence activate much larger populations of neurones are sometimes unaffected by voluntary contraction (Siebner, Auer, et al. 1999; Siebner, Tormos, et al. 1999; Ragert et al. 2003; Hummel et al. 2005)
Contraction Immediately after TBS
Effects on MEP
Contraction of the FDI immediately after cTBS reversed the usual effect of cTBS from suppression to facilitation and enhanced the effect of iTBS. The results were not due to contraction alone because 1 min contraction in the absence of TBS had no lasting effect on MEPs. This effect was specific to the targeted muscle: voluntary contraction of the ADM in the same hand or of the biceps muscle ipsilateral to cTBS did not change the response of the FDI to cTBS. Because the effects were absent after a contraction produced by electrical stimulation of the ulnar nerve, we conclude that voluntary activation of the motor cortex led to activity that interfered with the inhibitory aftereffects of TBS.
We have argued that TBS leads to a mixture of inhibitory and excitatory aftereffects with the latter being predominant after iTBS, whereas inhibition predominates after cTBS (Huang et al. 2005). Our hypothesis is that voluntary activity interferes with the build up of the inhibitory processes allowing concurrent excitatory effects to become evident and facilitate MEPs. Similarly, interference with any inhibitory effects that are recruited after iTBS will remove some of the underlying competition with excitation and may enhance the facilitation of MEPs. The scenario would be similar to previous studies in animals that have shown that physiological activity after induction of LTP/LTD can reduce or abolish changes in synaptic plasticity. Interestingly, as in the present experiments, the effect is greatest the nearer it is to the end of the induction period (Chen et al. 2001). The fact that physiological activity may enhance or suppress prior changes in synaptic efficiency may also explain why a period of learning can either improve or interfere with prior learning.
Why are only inhibitory aftereffects influenced by voluntary activity, whereas the excitatory effects are unchanged or even enhanced? One possibility is that they relate to the existence of discrete states of synaptic plasticity that have been described at both hippocampal and thalamocortical synapses (Montgomery and Madison 2004). “Silent” synapses represent a state in which only LTP is possible. Furthermore, induction of potentiation at such synapses is resistant to change for a period of 30 min or so before the synapses enter the “active” state. We suggest that part of the LTP-like effects of TBS may be caused by recruitment of such “silent” synapses and that these are resistant to modification by physiological activity for a short period after induction. In contrast, we suggest that induction of LTD-like effects depends on deactivation of synapses in the active and/or the “potentiated” state, both of which are modifiable at all times.
Another possibility is that the balance between the LTP- and LTD-like effects after cTBS and iTBS is controlled by Ca2+ levels in the postsynaptic cell. Activating the target muscle voluntarily may cause a sudden large influx of calcium into postsynaptic cells that desensitizes the inosital triphosphate (InsP3) receptor (Bezprozvanny et al. 1991). Nishiyama et al. (2000) showed that when InsP3 receptor–dependent Ca2+ release from internal stores is blocked, an LTD protocol is reversed to produce LTP. On contrary, the facilitatory effect is not affected by the blockage of InsP3-induced Ca2+ release. Thus, contraction could block the induction of LTD-like effects while preserving the LTP-like consequences of TBS.
Finally, we note that “homeostatic” effects are sometimes observed between 2 sets of conditioning protocols (Siebner et al. 2004; Ziemann et al. 2004; Stefan et al. 2006). Thus, it is possible that cTBS enhanced the usually short-lasting facilitation caused by voluntary contraction and that this was responsible for reversing the apparent effect of cTBS from inhibition to facilitation. However, because there was no lasting effect of the same contraction at 10 min after cTBS, we think this explanation is unlikely.
Effects on SICI
In contrast to the results on MEPs, the effect of cTBS on SICI was not changed by contraction immediately after TBS. The most likely explanation for the difference is that during contraction, circuits involved in MEPs are likely to be excited, whereas those involved in SICI onto the FDI muscle are likely to be suppressed (see above). The difference in activation presumably affects the response to contraction, abolishing the effect of cTBS on SICI while it reverses its effect on MEPs.
Contraction at 10 Min after the End of cTBS
In contrast to the effect immediately after cTBS, contraction after 10 min, at a time when the inhibitory aftereffect had built up toward maximum, had no permanent effect on MEP amplitude. There was only a short-lasting (3–4 min) reduction of suppression that might be related to postexercise facilitation (Samii et al. 1996). The absence of such changes is consistent with animal studies, which show that reversal of LTP can only happen when a depotentiating event is applied within a short time window after induction of LTP (Xu et al. 1998; Staubli and Scafidi 1999; Chen et al. 2001). However, in those experiments, the time window for an effect was closer to 1 h than 10 min. Perhaps if we had applied cTBS for longer, to produce a more lasting suppression of MEPs, then the time window for reversal by physiological activity may have been longer. Alternatively, it may be that if we had extended the duration of our contraction from 1 to 5 min, then a permanent effect could have been revealed even in the present experiments.
In conclusion, we have demonstrated that voluntary muscle contraction during or after TBS can interact with the aftereffects of TBS on MEPs and SICI. We suggest that this is because the ongoing neural activity in the brain interacts with synaptic plasticity–like processes in the conscious human brain.
The work was funded by the Taiwan National Science Council (NSC 94-2314-B-182A-057), Chang Gung Memorial Hospital (CMRPG34001), and UK-Taiwan Joint Project Grant. Conflict of Interest: None declared.