NMDA-receptor agonist reveals LTP-like properties of 10-Hz rTMS in the human motor cortex

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

rTMS was administered at the time of peak drug bioavailability, which is 2 hours after oral ingestion (Fig. 1A). Neuronavigated determination of primary motor cortex (M1) hotspot and resting motor threshold (rMT) was previously described [7]. Paired-pulse measures were obtained before and after rTMS, or approximately 1 hour, and 3 hours after drug disbursement, respectively [7]. Paired pulses were separated by inter-stimulus interval (1) of 3 ms for ICI, and 15 ms for ICF [7]. The conditioning stimulus (CS) was subthreshold intensity (80% rMT), and the testing stimulus (TS) was at a predetermined rMT for 1 mV (~120% rMT). We collected single-and paired-pulse MEPs with Magstim 200 2 and BiStim capacitors (Magstim, UK) in bins of 20 pulses, jittered 4e7 sec apart. We analyzed MEPs with Spike2 software (Cambridge Electronic Devices, UK). We amplified and filtered the raw signal with CED 1902 and 1401 microprocessors (Cambridge Electronic Devices, UK).
We administered 10-Hz rTMS over the left M1 with MagPro R30 with figure-8 B65 cooled coil (MagVenture, Denmark) because the MagStim device could not deliver rTMS. We delivered 300 pulses over 20 minutes with a 1.5 sec-on/58.5 sec-off duty cycle at 80% rMT (rMT obtained separately for MagVenture system), as previously described [7].
We calculated the average MEP peak-to-peak amplitude for each bin to create a paired-pulse/single-pulse ratio (MEP PP /MEP SP ) to yield ICF and ICI measures in respective protocols, then calculated change in ICF and ICI associated with rTMS (i.e., D ¼ post-rTMS value minus pre-rTMS value) to generate DICF and DICI. Wilcoxon signed-rank tests compared DICF and DICI between DCS and PBO conditions within subjects (e.g., DICF DCS vs. DICF PBO ) to assess within-subject relative change. Two-sided P value < 0.05 was considered statistically significant.
Mean DICF DCS was À0.48, compared to 0.36 for DICF PBO (Fig. 1B, At face value, these results were paradoxical. In order to better understand them, we examined the raw MEP values before ratios were calculated ( Fig. 1C and E, MEP values in legends). The PBO condition produced modest facilitation before rTMS, which then became robust (steeper slope) after rTMS (Fig. 1C). This is consistent with the hypothesis that 'excitatory' rTMS would enhance glutamatergic tone in an LTP-like manner. DCS, by contrast, produced robust facilitation (Mean (PBO) ICF: 1.69 vs. 1.25) before rTMS; consistent with a glutamatergic (NMDA receptor) agonist effect. We attribute the modest ICF and lack of facilitation after 10 Hz in the PBO group to the inherent MEP variability from a single time point (consider [2]), which can be mitigated through same-subject comparisons enabled by a crossover design. Intriguigingly, after rTMS, which creates a high baseline MEP SP (Mean ± SE, 1.55 ± 0.28), the subsequent facilitation is quite modest, suggesting a saturation or 'occlusion' effect -a classic finding in LTP studies [9]. Interestingly, observations of numerous human MEP studies suggest there may be a physiologic upper-limit near 75% above baseline.
In contrast to ICF, ICI was mitigated by rTMS alone (i.e., PBO; Fig. 1E), again consistent with the expected effect following an 'excitatory' stimulation protocol. DCS, on the other hand, expectedly blunted ICI before rTMS, but unexpectedly, produced the strongest inhibition after rTMS. We speculate that this effect may be the result of concurrent homeostatic depression superimposed on a short-term plasticity ICI protocol [10]. Animal studies are needed to verify this hypothesis, but we find it remarkable that these results occur in conjunction with the 'occlusion' observed with ICF. Moreover, all MEP SP data and PBO data are consistent with what we might expect from 'excitatory' rTMS and NMDA receptor agonist activity.
These results suggest that 10-Hz rTMS produces an excitatory effect through glutamatergic activity; and that NMDA receptor partial agonist DCS in combination with 10-Hz rTMS may produce occlusion of ICF, and homeostatic depression unmasked by ICI. In summary, these results suggest LTP-like mechanisms underlying 10-Hz rTMS in the human motor cortex; though our small sample size warrants caution to avoid type I error. Replication is warranted.

Declaration of competing interest
The authors declare no conflict of interests.