Suppression of motor cortical excitability in anesthetized rats by low frequency repetitive transcranial magnetic stimulation

Abstract

Repetitive transcranial magnetic stimulation (rTMS) is a widely-used method for modulating cortical excitability in humans, by mechanisms thought to involve use-dependent synaptic plasticity. For example, when low frequency rTMS (LF rTMS) is applied over the motor cortex, in humans, it predictably leads to a suppression of the motor evoked potential (MEP), presumably reflecting long-term depression (LTD) -like mechanisms. Yet how closely such rTMS effects actually match LTD is unknown. We therefore sought to (1) reproduce cortico-spinal depression by LF rTMS in rats, (2) establish a reliable animal model for rTMS effects that may enable mechanistic studies, and (3) test whether LTD-like properties are evident in the rat LF rTMS setup. Lateralized MEPs were obtained from anesthetized Long-Evans rats. To test frequency-dependence of LF rTMS, rats underwent rTMS at one of three frequencies, 0.25, 0.5, or 1 Hz. We next tested the dependence of rTMS effects on N-methyl-D-aspartate glutamate receptor (NMDAR), by application of two NMDAR antagonists. We find that 1 Hz rTMS preferentially depresses unilateral MEP in rats, and that this LTD-like effect is blocked by NMDAR antagonists. These are the first electrophysiological data showing depression of cortical excitability following LF rTMS in rats, and the first to demonstrate dependence of this form of cortical plasticity on the NMDAR. We also note that our report is the first to show that the capacity for LTD-type cortical suppression by rTMS is present under barbiturate anesthesia, suggesting that future neuromodulatory rTMS applications under anesthesia may be considered.

Figure 1

(A) EMG-rTMS rat setup. Photograph shows rat in stereotaxic frame. Monopolar stainless steel needle electrodes are placed into the brachioradialis of each forelimb and between the third and fourth digit in the footpad (arrows). A ground electrode is inserted in the tail. A 40 mm figure-of-eight coil is fixed to a micromanipulator arm and positioned over the left or right hemisphere. (B) Demonstration of contralateral activation by TMS in rat. Representative brachioradialis MEPs (average of ten consecutive sweeps) are shown. Note, at motor threshold, lateralized TMS elicits isolated MEP in the contralateral forelimb.

Figure 2

(A) Representative recruitment curves. Averaged integrated voltage (uV*s) for the baseline recruitment acquisition from all animals is compared to post-stimulation values at both the 8 and 35-minutes timepoints after rTMS. By 2-way ANOVA, there is a significant reduction in the integrated MEP size following 1 Hz stimulation at both 8-minutes [stimulation effect: F(1,162) = 294.3, p<0.0001; MO effect: F(8,162) = 127.5, p<0.0001; interaction effect: F(8,162) = 25.86, p<0.0001] and the 35-minutes [stimulation effect: F(1,162) = 355.6, p<0.0001; MO effect: F(8,162) = 118.2, p<0.0001; interaction effect: F(8,162) = 27.23, p<0.0001] timepoints. Significance is confirmed by post-hoc t-test from 75–95% MO. Error bars indicate SD. ***p<0.001. (B) MEP changes following 1 Hz rTMS. All values are normalized to sham rTMS control. The difference between active and sham rTMS recruitment curves is significant by 2-way ANOVA at the 8-minute [stimulation effect: F(1,90) = 28.35, p<0.0001] and 35-minutes [stimulation effect: F(1,90) = 30.29, p<0.0001] timepoints. At individual stimulator intensities, using post-hoc t-test, there is a significant suppression at 75, 80, 85 and 90% MO (p<0.01, p<0.001, p<0.01, p<0.05, respectively) at the 8-minutes timepoint, and significant suppression at 75, 80 and 85% MO (p<0.05, p<0.01, p<0.01) at the 35-minutes timepoint. Error bars indicate SD. *p<0.05; **p<0.01; ***p<0.001. (C) Representative MEPs following 1 Hz active or sham rTMS. Ten consecutive MEP sweeps at 55% MO, 100% MT, and 95% MO from one animal per condition, active or sham rTMS, are shown. Baseline traces are followed by MEPs generated at the same MO at the 8-minutes and 35-minutes timepoints.

Figure 3. MEP change as a function of rTMS frequency.

All conditions are shown at 80% MO. All ratios of average MEP integrated voltage relative to pre-rTMS baseline were log transformed. Compared to its baseline, 1 Hz active rTMS is the only condition where the MEP size was suppressed at both the 8-minues (p<0.01) and 35-minutes (p<0.05) timepoints. Error bars indicate SD. *p<0.05, **p<0.01.

Figure 4

(A) Average motor threshold per treatment group. Percent machine output needed to achieve motor thresholds across sham 1(74.2%MO ±7.3%MO, 75.8%MO ±4.9%MO, 74.2%MO ±3.8%MO, 70%MO ±4.5%MO respectively). No significant difference between groups was detected. Error bars indicate SD. (B) Effects of NMDAR antagonists on MEP suppression by 1 Hz rTMS. Integrated MEP voltage (uV*s) is measured at 80% MO. Statistics per treatment pair were obtained by comparison with the corresponding sham condition. For the 1 Hz rTMS group, there is a 48.2% (p<0.01) and 53% (p<0.05) reduction at the 8-minutes and 35-minutes timepoints after rTMS, respectively. There is no significant change from pre-rTMS baseline after pretreatment with MK801 or AP5 conditions. Error bars indicate SD.*p<0.05, **p<0.01.