Abstracts

Rationale: Cathodal transcranial direct current stimulation (tDCS), a method for focal noninvasive cortical stimulation, leads to long-term depression (DCS-LTD) of cortical excitability that is mediated by mGluR5 signaling (Ann Neurol. 2016; 80(2):233-46). tDCS is undergoing active testing in epilepsy, but its clinical efficacy is modest, at best. We thus tested whether inhomogenous cortical excitability modulation may contribute to the incomplete cathodal tDCS antiepileptic effect, and whether a novel drug-device pair improves DCS efficacy. Methods: Direct current stimulation (DCS) was delivered through Ag/AgCl electrodes to isolated mouse primary motor cortical (M1) slices, to the dorsal scalp of anesthetized mice, or to isolated human cortical slices derived from epilepsy surgery. %change in field excitatory postsynaptic potential (fEPSP) slopes recorded by microelectrode array spanned across all cortical layers, 1 hour after isolated in vitro slice DCS (400µA, 25min), was plotted as an interpolated 2D map. Scalp electroencephalographic (EEG) signals were acquired from urethane (2g/kg) anesthetized mice with two thin Ag/AgCl Teflon coated EEG subdermal wire electrodes, with a reference contact positioned over the dorsal snout at midline, and one active contact over the parietal region. Mice received PTZ (300 mg/kg, i.p.) followed by sham, tDCS (1mA, 25min), or tDCS+drug [NMDAR blocker memantine or mGluR5 positive allosteric modulator (PAM) CDPPB] treatment, and the latency to EEG epileptic activity was compared between all groups. Results: DCS-LTD of excitatory synaptic strength was reliably induced in superficial cortical layers, while a long-term potentiation (DCS-LTP) was observed in deep cortical layers below the stimulating site in M1 slices. Given that DCS-LTD depends on the activation of mGluR5-mTOR pathway, we also found non-uniform changes of phospho-S6 ribosomal protein (pS6) expression after cathodal tDCS across M1 layers: pS6 increased 1 hour after cathodal tDCS in the cortical layer 2/3 and 5, but not in the deeper layers in M1 slices. This heterogeneous pattern of cathodal DCS effects was also detected in human neocortical slices. Given the distinct molecular pathways of DCS-LTP and DCS-LTD, cathodal DCS produced LTD throughout the cortical thickness after either blocking NMDA-type glutamate receptors, or facilitating mGluR5-type glutamate receptors. Scalp EEG data showed that applying cathodal tDCS had an increase trend (1.3 ± 0.2 fold of sham, n=7) in the latency to first PTZ-induced EEG spike as compared to shams (p>0.05). However, pretreatment of mGluR5 PAM CDPPB (20mg/kg, s.c.) significantly prolonged the latency to first PTZ spike (2.5 ± 0.5 fold of sham, n=6, p<0.05) after cathodal tDCS. Conclusions: Non-uniform DCS effects across the stimulated cortical volume may explain the weak tDCS effect in clinical trials. We suggest a novel clinical use of pharmacological metabotropic glutamate receptor modulation to enhance cathodal tDCS antiepileptic efficacy. Funding: Translational Research Program, Boston Children's Hospital;Epilepsy Therapy Project