Polarizing cerebellar neurons with transcranial Direct Current Stimulation.

In this issue, Parazzini and colleagues (Parazzini et al., 2014), present computational models of brain current flow using transcranial Direct Current Stimulation (tDCS) designed for cerebellum stimulation. Using MRI-derived models of three subjects, they predict that a montage with one 5 ! 9 cm sponge electrode placed roughly over the cerebellum area and an extra-cephalic electrode on the right arm, results in brain current flow that is largely restricted to the cerebellum – though evidently current must exit the brain through deep brain structures. This prediction of relatively targeted stimulation is notable in contrast to imaging studies and modeling predictions using other tDCS sponge-montages that suggest current influences large areas of brain between electrodes (Datta et al., 2009). The possibility of targeted neuromodulation using conventional sponge-tDCS is thus compelling and the cerebellum is a promising target for a range of indications (Ferrucci et al., 2008, 2013; Galea et al., 2009; Hamada et al., 2012; Boehringer et al., 2013; Ferrucci and Priori, 2013). Like most modeling studies of tDCS, Parazzini et al. (2014) rely on the ‘‘quasi-uniform’’ assumption (Bikson et al., 2013) to suggest electric field (or current density) intensity in any given region indicates the relative neuromodulation of that region. None-the-less, an open question is given that both cortical architecture and cell morphology of the cerebellum differ from cortical targets investigated in a majority of tDCS studies, for example motor regions (Nitsche and Paulus, 2000; Lang et al., 2004). Will the cerebellum respond distinctly to an applied electric field? Indeed, the direction of brain current flow relative to cellular morphology is believed to modulate the direction of induced excitability changes (Bikson et al., 2004; Rahman et al., 2013) while cell morphology is further assumed to determine sensitivity to electric fields (Radman et al., 2009). We predicted brain current flow using a stand-alone simulation package (HDExplore Version 2.3, Soterix Medical Inc., New York, NY) considering varied deployments of High-Definition (1 cm diameter) electrodes (Fig. 1). Though in each of the four montages selected the region of current flow is restricted to the cerebellum, the direction of current flow across the cerebellum varies. Inward, outward, lateralized left, or lateralized right direction of flow can be achieved by selecting electrode position and polarity. In each case the current flow direction through the cerebellum seems broadly uniform (i.e. local current flow seems largely unaffected by cortical folding and CSF). While highdefinition tDCS allows significantly enhanced flexibility in optimization of current flow (Dmochowski et al., 2011), conventional sponge-tDCS is simpler to apply and Parazzini et al. (2014) demonstrate particularly compelling performance for cerebellum stimulation. Finite element models of transcranial electrical stimulation thus show diffuse current flow throughout the cerebellum with the direction of current flow guided by a specific montage. How current flow modulates cerebellar activities depends on the physiological and morphological characteristics of neurons in the cerebellum. In vivo and in vitro animal models of transcranial electrical stimulation show that subthreshold DC fields can change membrane excitability, spike timing, and firing rates (Terzuolo and Bullock, 1956; Purpura and McMurtry, 1965; Fröhlich and McCormick, 2010; Reato et al., 2010). These effects are specific for individual neurons with their longitudinal axis oriented along the direction of current flow. Simplistically, current flow along the somato–dendritic axis of a neuron will hyperpolarize membrane compartments proximal to the current source (near anode) and depolarize distal membrane compartments (nearer the cathode; Fig. 2A) (Rahman et al., 2013). Therefore, current flow parallel to the somato–dendritic axis of a cerebellar Purkinje cell can be used to hyperpolarize or depolarize both somatic and dendritic compartments. In a seminal series of papers, Chan, Nicholson and colleagues (Chan and Nicholson, 1986; Chan et al., 1988; Lopez et al., 1991) used isolated turtle cerebellum to quantify Purkinje and stellate cell polarization under DC and low-frequency sinusoidal electric fields. These studies identified morphological determinants of neuron sensitivity to externally applied electric fields. Purkinje cells polarized linearly along the somato–dendritic axis (pia proximal to the current source; Fig. 2B) (Chan et al., 1988) and their results support the concept that current flow that depolarizes the soma leads to excitation while current flow that hyperpolarizes the soma leads to inhibition of firing (Chan and Nicholson, 1986). Pharmacological blockade of active conductances and synaptic transmission revealed passive membrane polarization by the field (thus depending simply on the resistive and capacitive properties of the cell membrane) underlying active voltage-dependent responses. Chan et al. (1988) also report Purkinje cells will polarize by 0.2 mV per 1 V/m applied electric field, which is within the 0.03–0.49 mV per V/m range reported for layer V/VI pyramidal cells (Radman et al., 2009). Further evidence for the morphological dependence of cerebellar cells to DC stimulation effects is given by the variety of stellate cells in the cerebellum. While the geometry and orientation of the dendrites of stellate cells of the lower molecular layer (distal to the