Multi-session delivery of synchronous rTMS and sensory stimulation induces long-term plasticity

Multi-session delivery of synchronous rTMS and sensory stimulation induces long-term plasticity

Combining training or sensory stimulation with non-invasive brain stimulation has shown to improve performance in healthy subjects and improve brain function in patients after brain injury. However, the plasticity mechanisms and the optimal parameters to induce long-term and sustainable enhanced performance remain unknown.

This work was designed to identify the protocols of which combining sensory stimulation with repetitive transcranial magnetic stimulation (rTMS) will facilitate the greatest changes in fMRI activation maps in the rat's primary somatosensory cortex (S1).

Several protocols of combining forepaw electrical stimulation with rTMS were tested, including a single stimulation session compared to multiple, daily stimulation sessions, as well as synchronous and asynchronous delivery of both modalities. High-resolution fMRI was used to determine how pairing sensory stimulation with rTMS induced short and long-term plasticity in the rat S1.

All groups that received a single session of rTMS showed short-term increases in S1 activity, but these increases did not last three days after the session. The group that received a stimulation protocol of 10 Hz forepaw stimulation that was delivered simultaneously with 10 Hz rTMS for five consecutive days demonstrated the greatest increases in the extent of the evoked fMRI responses compared to groups that received other stimulation protocols.

Our results provide direct indication that pairing peripheral stimulation with rTMS induces long-term plasticity, and this phenomenon appears to follow a time-dependent plasticity mechanism. These results will be important to lead the design of new training and rehabilitation paradigms and training towards achieving maximal performance in healthy subjects.

Throughout history humans have been pursuing new regimes to augment and maximize motor and cognitive performance. Intense physiological training is known to increase endurance and enhance motor performance in athletes; purposeful physical therapy is instrumental to acquire and rebuild sensorimotor abilities in patients with impaired brain function; and cognitive skills training via traditional learning methods, and more recently by virtual reality and gaming-based methods have shown to increase mental endurance, maximize academic abilities [] and improve brain function in stroke patients [].

Advances in non-invasive brain stimulation technologies had opened a new frontier in achieving motor and cognitive functions in levels and speed that exceed traditional training methods. Repetitive transcranial magnetic stimulation (rTMS) is known to increase neural activity, and its application over a period of time has been shown to induce long-term and sustainable effects in healthy [] and in disease conditions, in human [] and in animal models []. These approaches may be particularly valuable to patients who may be unable to fully participate in a traditional training routine due to disability.

New paradigm in human performance now seeks to capitalize on benefits achieved from both traditional training and non-invasive brain stimulation technologies to reach peak performance. Pairing of peripheral and central nervous system stimulations has shown to improve endurance and athletic performance in healthy individuals [], improve motor functions in stroke patients [], and increase the cognitive processing speed in adults []. These changes are believed to occur through associative, Hebbian-like plasticity mechanisms. Indeed, new evidence using optical imaging in an animal model shows that a visual stimulation delivered during TMS can change cortical maps []. Nevertheless, the exact mechanisms and the optimal parameters to induce long-term and sustainable enhanced performance remain unknown; If indeed the mechanism is time-dependent plasticity, then it is likely that the exact timing of which the tactile, sensory or cognitive stimulation is presented during the brain stimulation protocol will determine the effectivity of this approach. Elucidating the plasticity mechanism associated with these protocols would greatly impact performance of healthy individuals and their adaptation in clinical practice.

This work was designed to identify the protocols of which combining sensory stimulation with rTMS will facilitate the greatest changes in activation maps in the rat's primary sensory cortex (S1). We quantified the spatial functional MRI (fMRI) activity and the expression of molecular markers associated with plasticity. The results demonstrate that rTMS significantly increases short- and long-term plasticity, and that synchronous delivery of the peripheral stimulation and rTMS have led to significant and sustainable increase in S1 performance.

All animal procedures were conducted in accordance with the NIH Guide for the Care and Use of Laboratory Animals and approved by the Michigan State University Animal Care and Use Committee. Forty-two adult Sprague-Dawley rats (14 males and 28 females, 250 g) were provided with food and water ad libitum.

Rats were anesthetized with 1.5% isoflurane followed by an initial s. c. injection of dexmedetomidine (0.1 mg/kg) which is known to preserve neurovascular coupling []. Then the isoflurane was discontinued and dexmedetomidine (0.1 mg/kg/h) was delivered SC. Rats were imaged in a 7 T/16 cm aperture bore small-animal scanner (Bruker BioSpin). A 72-mm quadrature volume coil and a H receive-only 2 × 2 rat brain surface array coil (RF ARR 300 1H R. BR. 2 × 2 RO AD) were used to transmit and receive magnetic resonance signals, respectively. An MRI oximeter (Starr Life Sciences, Pennsylvania, USA) was used to measure the respiration rate, heart rate, and partial pressure of oxygen saturation throughout the experiment. For fMRI, Free Induction Decay -echo-planner images (FID-EPI) were used with a resolution of 150 × 150 × 1000 μm. Five coronal slices covering the somatosensory cortex were acquired with TR/TE 1000/16.5 ms, FOV 3.5 cm, Flip angle 75°, matrix size 128 × 128 and slice thickness 1.0 mm. A T2-weighted TurboRARE sequence was used to acquire high-resolution anatomical images with TR/TE 3000/33 ms, FOV 3.5 cm and matrix size 256 × 256. Two needle electrodes were inserted into the left forepaw to deliver 3 mA for two 20-s (when it was delivered outside the MRI with or without the rTMS) or two 40-s (when it was delivered during fMRI) tactile-electrical stimulation. In the first scanning day, the stimulation electrodes remained at the same location for the group of the short-term experiments but were removed at the end of the scanning before animals recovered from anesthesia and returned to their housing. For the group of the long-term experiments, the stimulation electrodes were inserted into the forepaw everyday only for the duration of the stimulation period. Inside the scanner, rats were wrapped with circulating water jacket whose temperature was precisely controlled at 37C by a thermostat. Respiration rate, heart rate and partial pressure of oxygen were continuously monitored throughout all measurements (Starr Life Sciences).

The rTMS system (Magstim, Rapid2) was equipped with a figure eight, 25 mm custom rodent coil that was placed directly over the center of the head at bregma 0. This coil design has been shown to induce focal stimulation in rats []. rTMS was delivered with the following parameters: 20 s cycles, 20 s interval, and 2 periods (total of 400 pulses per day). This stimulation frequency has been found to have long-term effects in rats []. The power level was set to 30% of the instrument's maximum output and all the rats received the exact same rTMS protocol. The rTMS and the peripheral stimulation were manually turned at the exact time for synchronous or asynchronous stimuli. Rats were randomly assigned into seven groups for short-term (ST) and long-term (LT) plasticity studies: ST Group 1 received 10 Hz rTMS stimulation (n = 6); ST Group 2 received asynchronous 10 Hz rTMS stimulation with 3 Hz forepaw stimulation (n = 6); ST Group 3 received 3 Hz forepaw stimulation (n = 6). For long-term studies, rats received stimulation for five continuous days. They did not receive rTMS on the fMRI scanning days (i.e., Day 1 and Day 7). LT Group 1 received 10 Hz forepaw stimulation synchronized with 10 Hz rTMS stimulation (n = 6); LT Group 2 received only 10 Hz rTMS stimulation (n = 6); LT Group 3 received asynchronous 10 Hz rTMS stimulation with 3 Hz forepaw stimulation (n = 6); and LT Group 4 received only 10 Hz forepaw stimulation (n = 6).

Analysis: Functional images were processed with SPM fMRat software (SPM, University College London, UK). For each subject, the functional images were realigned to T2-weighted high-resolution images. In addition, head motion correction was done in three translational and three rotational directions (X,Y,Z). For each subject, EPI images were re-oriented, averaged and smoothed with full width half maximum (FWHM) = 1.25 mm in the coronal direction spatial in order to reduce randomly generated noise. Finally, a fMRI block design was used, and activation maps were obtained using the general linear model. For each individual, the Z-score statistics was cluster-size threshold for an effective significance of P  0.05. Statistics was conducted with a threshold of Z > 4.58. For group analysis, the anatomical images from the rats brain atlas [] were used for reference frames. All the images were coregistered and normalized to this template using SPM software. Every Z-score map was clustered into a new mask. The overlap between the masks is shown as the t-value.

Histology: Rats were perfused with 0.1 M phosphate buffer saline solution (PBS) in pH 7.4 followed by ice cold 4% paraformaldehyde solution and the brains were removed and immersed in sucrose solution. Brains were sliced on a cryostat to obtain 30 μm thick sections. Sections were incubated overnight with primary antibodies to detect Ca/calmodulin-dependent protein kinase II (CamKII) (anti-CaMKII rabbit, Abcam #ab52476); activity-regulated cytoskeleton-associated protein (Arc) (anti-Arc rabbit antibody, SYSY #156003). After three washes with PBS, sections were incubated for 3 h at room temperature with secondary antibodies for CaMKII (Alexa Fluor 647, Jackson #711605152) and Arc (Alexa Fluor 488, Abcam #ab150073), processed with ProLong Gold antifade reagent with DAPI (Thermo Fischer Scientific 2078923). Confocal images were acquired using the Nikon A1-Rsi Confocal Laser Scanning Microscope (Nikon Instruments, Inc., Tokyo, Japan) configured on a Nikon Eclipse Ti inverted microscope. Images were collected using either a Nikon 10x Plan Apo (NA 0.45) objective or a Nikon 20x Plan Apo VC (NA 0.75) objective. Image acquisition was performed using Nikon NIS-Elements AR software (version 5.20.02). DAPI fluorescence was excited using a 402 nm diode laser, and blue fluorescence emission was detected through a 450/50 nm band pass emission filter. Green fluorescence was excited using a 488 nm diode laser, and fluorescence emission was detected through a 525/50 nm band pass emission filter. Far red fluorescence was excited using a 647 nm diode laser, and fluorescence emission was detected through a 700/75 nm band pass emission filter. For each data set, a confocal series through the thickness of the tissue section was collected. For the 20× objective, confocal images were collected in 2 μm increments through an average thickness of 10 μm. For each confocal series, a Maximum Intensity Projection image was generated, representing the brightness intensity pixels through the Z-depth. ImageJ was used for cell counting and analysis. The number of cells were counted for an ROI of 1024 × 25 μm. All the histology procedures, imaging and cell counting had been performed by investigators blinded to the experimental condition.

Illustration demonstrating the experimental design for short-term plasticity study. Rats received a single session of rTMS, rTMS combined with forepaw stimulation, or only forepaw stimulation. fMRI was conducted within minutes after the stimulation.

Evoked fMRI responses to forepaw stimulation were measured before and after a single-session stimulation protocol. Representative BOLD z-score activation maps corresponding to p  F = 6.6). shows incidents maps of the fMRI responses in the center of S1(bregma 0) demonstrating the consistent distribution of the activated pixels for each condition and within each group. A representative time course and the peak fMRI values before and after rTMS application is shown in . The time course of the activated voxels was averaged and demonstrate a reproducible fMRI temporal responses.

Immunohistology for plasticity markers CaMKII and Arc. Following the last fMRI measurement, rats were perfused and processed for immunohistology. High-magnification images of neurons immunostained for CaMKII (red) and Arc (green) are shown in top panel (100X, scale bar = 10 μm). Microscopy images demonstrated increased fluorescent in neurons located in right S1 in groups that received both rTMS and left forepaw stimulation (Scale bar = 50 μm). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

After the final fMRI measurement, rats were perfused and immunohistology was performed to measure cellular markers associated with plasticity. Thirty μm thick brain sections were stained for CaMKII, a gene known to be involved in long-term potentiation (LTP) and Arc, an immediate-early gene known to play a role in synaptic plasticity []. Both the right and the left S1, contralateral and ipsilateral to the limb stimulation, respectively, were imaged. The number of CaMKII-positive and Arc-positive cells were counted in each region and averaged for n = 5 in each of the four groups. shows representative immunohistology results as well as the quantitative measurements. It is manifested that all three groups that received rTMS over 7 days exhibit increased expression of both CaMKII and Arc in S1 contralateral to stimulation, but only LT Group 1 and LT Group 3 that received both rTMS and forepaw stimulation showed differential expression in both plasticity markers between right and left S1. The group that received only the forepaw stimulation (LT Group 4) did not show any difference between right and left S1 groups (LT Group 1, 10 Hz Forepaw stim + 10 Hz rTMS, CamKII/L 614.6 ± 48.6, CamKII/R 736.8 ± 63.8, Arc/L 464.0 ± 61.2, Arc/R 582.8 ± 79.0, F = 8.4 > F = 7.7; LT Group 2, 10 Hz rTMS, CamKII/L 495.0 ± 16.0, CamKII/R 656.6 ± 20.3, Arc/L 447.4 ± 61.2, Arc/R 501.8 ± 61.4, F = 27.3 > F = 7.7; LT Group 3, 10 Hz rTMS+ 3 Hz Forepaw, CamKII/L 507.0 ± 46.1, CamKII R 660.2 ± 63.9, Arc/L 496.0 ± 73.0, Arc/R 679.0 ± 65.1, F = 10.7 > F = 7.7, two-way ANOVA analysis without replication). These results demonstrate that the pairing of rTMS with peripheral stimulation induce plasticity that can be detected in the cellular and the network levels.

[29] The neural code between neocortical pyramidal neurons depends on neurotransmitter release probability. [31] Acute changes in short-term plasticity at synapses with elevated levels of neuronal calcium sensor-1. [5] Immediate effects of repetitive magnetic stimulation on single cortical pyramidal neurons. [32] The TMS motor map does not change following a single session of mirror training either with or without motor imagery. [33] Non-invasive brain stimulation of motor cortex induces embodiment when integrated with virtual reality feedback. Short term plasticity is important for neurons to produce appropriate responses to acute changes in activity []. Short term plasticity lasts for milliseconds to minutes and is known to work through mechanisms of depression due to vesicle depletion or facilitation due to elevated calcium levels []. The results demonstrate that a single rTMS application immediately increased neural activity in S1 as was evident by the fMRI results. These results are consistent to what have been previously demonstrated in human and animal models []. However, this short-term increase failed to lead to long-term changes; Three days after rTMS application, the extent of the fMRI responses was identical to the pre-rTMS stimulation ones. It is plausible that the one-time rTMS application was not long enough to induce long-term plasticity, which is also supported by human studies []. Another interesting observation was that rTMS alone was more effective than combining 10 Hz rTMS with 3Hz sensory stimulation. This is also supported by the observations in the long-term plasticity experiments and suggests that the temporal coherence between the stimulation is crucial to determine brain responses. Thus, it is possible that because the stimulations were not synchronized, they decreased the evoked responses as was also observed in human studies []. Thus, the main implication is that for sustainable and long-term changes in cortical function, multiple rTMS sessions are required.

[35] Spike timing-dependent synaptic depression in the in vivo barrel cortex of the rat. [36] Receptive-field modification in rat visual cortex induced by paired visual stimulation and single-cell spiking. [37] Synaptic modifications in cultured hippocampal neurons: dependence on spike timing, synaptic strength, and postsynaptic cell type. [5] Immediate effects of repetitive magnetic stimulation on single cortical pyramidal neurons. [26] Long term plasticity is fundamental for learning, memory and recovering function after injury, and has been shown to last for minutes and days. There are several forms of long-term plasticity that induce rapid and long-lasting changes including LTP, long-term depression (LTD), and Hebbian synaptic plasticity []. A critical factor for these changes is the temporal sequence and interval between the pre- and post-synaptic spikes, known as spike timing-dependent-plasticity (STDP) []. A large amount of work showed that repetitive post-synaptic spiking within a time window of 5–20 msec after synaptic activation leads to LTP, and when the post-synaptic neuron is activated before the pre-synaptic neurons within a time window of 20 msec it leads to LTD []. In our experiments we could not determine the exact timing of the synaptic activity, but we attempted to control a constant synchronous or asynchronous delivery of the stimulation modules. Previously, when we used patch-clamp recordings in brain slices we found on average 3 msec delay between TMS pulse to neural activity [], and when we used extracellular recordings in vivo we found 8 msec delay between peripheral forepaw sensory stimulation to cortical activity []. Taking into consideration the milliseconds delay that are bound to occur using both modalities, we can assume a 3–10 msec time window of pre-synaptic activation. This time window agrees with known temporal sequence of spike timing that induces long-term plasticity. It will be interesting to further identify the cellular mechanisms associated with the combination of rTMS and another modality using direct measurements of neural activity via electrophysiology and optical imaging methods. The results indicate that a protocol consisting of daily rTMS stimulation is effective in inducing long-term changes in cortical function. Notably, the greatest and most significant change in fMRI responses was evoked when the rTMS was delivered at the same frequency as the sensory stimulation. This suggests, that rTMS combined with an additional stimulation considerably augment brain response, and that this long-term effect may be possibly modulated via a time-dependent mechanism such as STDP. The latter is also supported by the observation that in both the short-term and long-term studies, when the rTMS was combined with sensory stimulation, but each stimulation was delivered asynchronously (10 Hz rTMS and 3 Hz forepaw sensory stimulation) then the changes in fMRI responses were less than when both modes of stimulation were delivered synchronously. In the future it will interesting to test additional experimental designs where the two stimulation modalities are delivered with different phase shifts to further understand the cellular plasticity mechanism. Subsequently, it is plausible that delivering rTMS and another mode of stimulation in an asynchronous manner, may diminish effectivity.

[39] Prolonged peripheral nerve stimulation induces persistent changes in excitability of human motor cortex. 41 Transcranial magnetic stimulation for the prediction and enhancement of rehabilitation treatment effects. 42 Continuous theta burst stimulation over the contralesional sensory and motor cortex enhances motor learning post-stroke. 43 Hand function improvement with low-frequency repetitive transcranial magnetic stimulation of the unaffected hemisphere in a severe case of stroke. 44 Inhibition of the unaffected motor cortex by 1 Hz repetitive transcranical magnetic stimulation enhances motor performance and training effect of the paretic hand in patients with chronic stroke. [13] Transcranial magnetic stimulation and environmental enrichment enhances cortical excitability and functional outcomes after traumatic brain injury. Previous reports have showed that peripheral electrical stimulation by itself can induce cortical plasticity [], especially when the electrical stimulation was delivered for over 60 min []. An interesting result of our study was that sensory stimulation by itself did not lead to short-term and long-term plasticity. This may be due to the considerably short length of the stimulation itself which occurred over the course of less than 5 min. This also builds on a great amount of evidence suggesting that learning and memory is best achieved with multimodal forms of stimulation and experiences. For example, non-invasive brain stimulation paired with current stimulation increased LTP in brain slices [], and TMS combined with visual stimuli led to remodeling of maps in cat's primary visual cortex []. The combination of TMS with another modality of stimulation has also shown to improve post-stroke function in humans [] and in animal models of injury [].

The nature of this study required the rodents to be lightly anesthetized which limits the amplitude of the neural signals and by that the breadth of the information that could be discovered in conscious, behaving animals. However, the results indicate that even in the anesthetized animal the temporal application of the stimulation is critical in determining the long-term responses. Another limitation of the study is that the behavioral output is challenging to evaluate. The increased activity in S1 may have translated to increased dexterity or perhaps increased abilities in discrimination tasks. It would be impactful to evaluate these behavioral changes in humans where tests to evaluate sensory and motor performance are well established. It is conceivable that rTMS combined with another form of stimulation that is applied in a conscious, participating individual, may lead to even greater changes in brain function. This might be especially significant if the individual receives tactile feedback on top of the rTMS and sensory stimulation for individuals engaged in rehabilitative therapy, or visual and auditory feedback if the rTMS is delivered to enhance mental and cognitive learning.

[46] Repetitive transcranial magnetic stimulation for upper extremity motor recovery: does it help?. [47] Functional MRI impulse response for BOLD and CBV contrast in rat somatosensory cortex. 48 Enhancement of motor coordination by applying high frequency repetitive TMS on the sensory cortex. 49 50 Modulation of proprioceptive integration in the motor cortex shapes human motor learning. [27] Functional MRI detection of bilateral cortical reorganization in the rodent brain following peripheral nerve deafferentation. 52 Study of the spatial correlation between neuronal activity and BOLD fMRI responses evoked by sensory and channelrhodopsin-2 stimulation in the rat somatosensory cortex. 53 Catheter confocal fluorescence imaging and functional magnetic resonance imaging of local and systems level recovery in the regenerating rodent sciatic nerve. 55 Interhemispheric plasticity protects the deafferented somatosensory cortex from functional takeover after nerve injury. 56 Interhemispheric neuroplasticity following limb deafferentation detected by resting-state functional connectivity magnetic resonance imaging (fcMRI) and functional magnetic resonance imaging (fMRI). 57 Correlation between brain reorganization, ischemic damage, and neurologic status after transient focal cerebral ischemia in rats: a functional magnetic resonance imaging study. The results demonstrate that longitudinal rTMS application led to significant increases in the extent of the fMRI responses, even beyond S1 cortical representation, suggesting possible remapping of this region. This plasticity may translate to faster processing speed, increased perception and sensitivity, and faster motor responses. Remodeling of sensory maps after rTMS application was also documented in the cat's visual cortex [] and is believed to be linked to behavioral gains [], and is also well documented after stroke where it is considered beneficial for recovery []. The basis of the fMRI blood-oxygenation-level-dependent (BOLD) contrast is neurovascular coupling. This is a complex integration of neural activity with a substantial contribution from large draining veins along the pial surface towards midline which may affect the fMRI BOLD signal in the peripheral regions of S1 []. Thus, it is possible that due to the anatomy of the rat's cortex, the extent of the fMRI signal may exceed the neural activity that takes place in S1 and the regions adjacent to it. It is also plausible that rTMS effect the neural activity in the primary motor cortex itself. Indeed, human studies suggest that sensory stimulation, either via noninvasive brain stimulation or by peripheral stimulation induce increased neural activity and priming of the motor cortex []. Future studies using non-BOLD fMRI contrasts and electrophysiology could determine with greater spatial resolution neuroplasticity mapping. Nevertheless, our study builds on a growing amount of work showing that preclinical fMRI is becoming an instrumental tool in basic and translational neuroscience to non-invasively detect changes associated with neural activity in health and disease [].

[58] The roles of protein kinases in learning and memory. [59] 63 Arc/Arg3.1 is essential for the consolidation of synaptic plasticity and memories. 64 New views of Arc, a master regulator of synaptic plasticity. 65 Somatic Arc protein expression in hippocampal granule cells is increased in response to environmental change but independent of task-specific learning. 66 Environment-specific expression of the immediate-early gene Arc in hippocampal neuronal ensembles. 67 Delayed wave of c-Fos expression in the dorsal hippocampus involved specifically in persistence of long-term memory storage. 68 Sustained transcription of the immediate early gene Arc in the dentate gyrus after spatial exploration. We used two immunohistological markers that are known to be correlated to long-term plasticity. CamKII is essential for LTP induction [] and the associated structural plasticity of dendritic spines []. This protein is known to have a long profile of expression, and thus its detection after rats were perfused even a day after the last fMRI session supports that the observed fMRI changes were accompanied by cellular, and structural modifications. On the other hand, Arc is a member of the Immediate early gene family, and has been used as a robust marker for neuronal activity [] due to the crucial role it plays in the consolidation of new memories []. Previous studies have shown that upregulation of Arc during exploration of a novel environment in the hippocampus remains elevated for over 8 h []. C-Fos, another member of the immediate early gene family is also known to be detected many hours after the stimulation seized []. Perhaps, one of the reasons Arc expression is maintained several hours post-stimulation is the time frame needed for the protein to be degraded. The cellular mechanism by which Arc was found to be correlated to the CamKII levels and the fMRI results may not be entirely clear, but the results suggest that the upregulation of Arc correlated to long-term plasticity induced by the stimulation protocols that were delivered.

Images Powered by Shutterstock