The targeted modulation of brain activity via controlled magnetic or electric stimulation is a valuable research tool in neuroscience, as well as an emerging clinical intervention in neurological and psychiatric diseases [1].

Within the group of non-invasive brain stimulation techniques (NIBS), tDCS has developed into a frequently used approach in neuroscience. tDCS is applied with surface electrodes and with low intensity electrical currents [2]. Dependent on the stimulation protocol, the intervention induces acute, but also prolonged, shifts of cortical excitability ^[3][4][5][3,4,5].

Effects of tDCS are supposed to be accomplished by a subthreshold modulation of neuronal membrane resting potentials that alters cortical excitability, and can subsequently induce more enduring effects related to neuroplasticity [3]. With respect to the after-effects of this intervention, tDCS over the motor cortex of healthy humans alters cortical excitability for a couple of minutes up to hours ^[4][6][4,6] after intervention. Furthermore, it was shown that tDCS effects depend on the current flow direction in relation to the targeted neuron populations or brain regions ^[3][7][8][3,7,8]. For the primary motor cortex model in humans, stimulation with the anode placed over the motor cortex target and a cathodal return electrode positioned over the contralateral supraorbital area increased motor evoked potentials (MEPs) (long-term potentiation (LTP)-like mechanism) and reversed electrode positioning reduced MEPs (long-term depression (LTD)-like mechanism) ^[3][4][5][3,4,5]. However, with other return electrode positions, no clear physiological effects emerged [3].

A considerable number of studies in humans have explored the physiological foundation of the effects of tDCS. Pharmacological studies have shown that the acute effects of short-lasting tDCS are not synaptically driven, but depend on membrane polarization, because the effects of depolarizing anodal tDCS were prevented by voltage-gated ion channel blockers, but not by glutamatergic N-methyl-D-aspartate (NMDA) receptor block [9]. In contrast, neuroplastic after-effects were prevented by NMDA receptor block, but increased by NMDA receptor enhancement ^[9][10][9,10]. This suggests that plasticity induced by tDCS depends on the glutamatergic system and involves calcium-dependent mechanisms. Moreover, Tropomyosin receptor kinase B (TrkB), the main receptor of brain derived neurotrophic factor (BDNF), was found to be involved in LPT-like effects of DCS [7]. Furthermore, tDCS reduced γ-amino-butyric acid (GABA) activity independent from stimulation polarity [11], which might have a gating effect on tDCS-induced plasticity. Early in vivo animal studies moreover suggest that after-effects of DCS, which last for more than 3 h and thus resemble late phase plasticity, require protein synthesis [12]. Beyond these regional effects of tDCS, it was shown that this intervention has an impact on cortical networks in terms of functional connectivity alterations and topological functional organization [13].

Based on respective physiological effects, tDCS has a relevant impact on psychological processes including perception, executive functions, and learning and memory, amongst others [14]. Some heterogeneities of these effects have been described between studies which underline the importance of improving mechanistic understanding of this intervention to optimize effects.

In the clinical domain, tDCS emerged as a valuable non-invasive brain stimulation tool to ameliorate symptoms in diseases accompanied by pathological alterations of cortical activity and plasticity such as depression, schizophrenia, pain syndromes, epilepsy, and in rehabilitation, amongst others ^[15][16][15,16]. However, heterogeneities of results are observed between studies, which might be caused at least partially by intervention protocol differences, and thus stress the need to understand basic tDCS effects better to improve the efficacy of interventions.

Although various mechanistic studies are available in humans ^[17][18][19][17,18,19] and also in animal models ^[7][20][21][7,20,21], the exact molecular mechanisms underlying the neuromodulatory effects of tDCS are yet not fully understood. Therefore, gathering more direct evidence using sophisticated neurobiological techniques such as cell-based assays (in vitro), brain slices (ex vivo), or in vivo animal models are required to supplement existing knowledge. Pelletier and Cicchetti summarized relevant research in 2015 [22]. They described cellular effects related to electrotaxis, metabolism, differentiation, cell orientation, as well as galvanotropism ^{[23][24][25][26][27][28][29][30][31][32][33][34][35][36]}[23,24,25,26,27,28,29,30,31,32,33,34,35,36], to name a few. However, all these morphological and structural changes have only been observed after either long-term stimulation, e.g., for 12 h [37], or with high electric field intensities of, e.g., 55.5 A/m² [38], which differ from those applied in healthy humans and patients. In human clinical studies, stimulation duration is typically around 20 min and electrical field strength is ≤1 V/m with a maximum current density of about 0.28 A/m^2 [39][40][39,40]. Nevertheless, Pelletier and Cicchetti also summarized animal data investigating DCS protocols more closely comparable to those applied in humans. These showed effects on brain physiology by different mechanisms, including the dependency on direction and impact of DC-induced membrane polarization on the relative orientation of the cells within the electric field, the involvement of modulation of presynaptic compartments, and modulation of action potential generation in efferent neurons ^[8][41][8,41]. Beyond these neuronal effects, DC stimulation has an impact on various other physiological processes.

DC electric fields can influence inflammatory processes (both anti- and pro-inflammatory [42]), structural neuroplasticity (neurite outgrowth [43]), neurogenesis (in case of directed neural stem cell migration towards a lesion or damaged location [30]), and angiogenesis [44].

2. Orientation of the Electric Field in Relation to Neuronal Population Morphology/Alignment, and Location

Kronberg et al. (2017) used rat hippocampal slices and applied current strengths of 100–200 µA for 45 s, and 3, 15, and 30 min. The electric field was oriented parallel to the somato-dendritic axis of cornu ammonis (CA1) pyramidal neurons. Trains of 900 electrical pulses at varying frequencies (0.5, 1, 5, and 20 Hz) were applied to generate plasticity before DCS, and field excitatory postsynaptic potentials (fEPSPs) were monitored at the dendritic level. Cathodal DCS enhanced LTP in apical dendrites and anodal DCS enhanced LTP in basal dendrites. Interestingly, both anodal and cathodal DCS reduced LTD in apical dendrites and DCS had no effect on weakly active synapses during NMDA block ^[45][47]. In a subsequent study, the same group showed that anodal DCS applied during theta burst stimulation (TBS) for some seconds enhanced Hebbian LTP [21].

Rahman et al. applied 10–150 µA currents for 3–5 s in rat motor cortex slices. Electrical fields were oriented orthodromic to L2/3 to induce fEPSP alterations. The postsynaptic membrane polarization during DCS and ongoing presynaptic activity induced by a train of presynaptic inputs delivered with constant or Poisson-distributed stimuli resulted in sustained and cumulative enhancement of fEPSPs ^[46][48].

Chakraborty et al. (2018) stimulated mouse coronal prefrontal cortical slices for 1 min with current strengths between 34.8 and 58.3 µA parallel and orthogonal to the dendrito-axonic axis of layer-V pyramidal neurons. They used recordings of membrane polarization (V/m) as readout. Chakraborty et al. (2018) showed that suprathreshold stimulation (important for, e.g., deep brain stimulation) induces action potentials at axon terminals, whereas subthreshold stimulation (important for DCS) modulates synaptic efficacy through axon terminal polarization. Moreover, only orientation of the electrical field parallel to the dendrite-axonal axis of the neurons induced these effects ^[47][49].

Recent experiments added the feature of cell localization as an important aspect affecting DCS effects. In contrast to previous studies, they used mouse and human cortical slices stimulated with 400 µA for 25 min. Slices were oriented orthogonal to the pia and parallel to the vertical inter-layer primary motor cortex (M1) projections with the cathode proximal to the cortical pia surface and the anode beneath the subcortical white matter. Their experiments focused on cortical excitability in human and mouse slices with cathodal stimulation. DCS generated LTD in superficial cortical layers, and LTP-like plasticity in deep cortical layers ^[48][52].

3. Acute, Prolonged, and Chronic DCS

Beyond application of DCS for some seconds, which has been discussed above and induces acute effects on neuronal membranes but no plasticity, DCS can also be applied for prolonged (minutes, effect on plasticity), or even chronic time courses (hours to days, effects on cell migration and neuronal orientation) [22].

Some articles already discussed in the previous section applied prolonged stimulation, such as Kronberg et al. (2017) and Sun et al. (2020) ^[45][48][47,52], and other contributions with prolonged stimulation protocols will be discussed in the next section.

Reato et al. (2015) for instance focused on the after-effects of prolonged DCS on gamma power and multi-unit activity (MUA). They used rat hippocampal slices and stimulated these for 10 min at a field strength varying from −20 up to + 20 mV/mm parallel to CA3 pyramidal neurons after induction of gamma oscillations by carbachol. They defined positive fields as anodal (0 to 20 mV/mm) and negative fields as cathodal (−20 to 0 mV/mm). Electrodes were oriented parallel to CA3 pyramidal cells. Their results showed altered gamma power and multi-unit activity (MUA) in a polarity-specific manner after 10 min of DCS: −20 and −10 mV/mm led to an acute downregulation of MUA and gamma power while +10 and + 20 mV/mm upregulated both. These results persisted for 10 min after stimulation ^[49][53]. Latchoumane et al. (2018) applied repeated prolonged DCS on embryonic stem cell (ESC)-derived neurons and glial cells after L-Glutamate-induced impairment of neuronal maturation. They treated mouse ESCs with currents of 10 µA over 5 days, 15 min per day. The intervention enhanced neuronal excitability as well as network synchronization. They furthermore demonstrated an upregulation of the NMDA receptor subunit NR2A, and Ras-related protein RAB3A in mouse Hb9 ESC-derived neurons and glial cells [50].

Zhao et al. (2015) applied chronic DCS in a stem cell model. They stimulated mouse neuronal precursor cells (NPCs) for 90 min with a current strength of 0.25 nA. The DC electric field enhanced mobility and caused cultured NPC migration to the cathode. A calcium-dependent mechanism was explored by adding the calcium chelator Ethylene glycol-bis(2-aminoethylether)-N,N,N’,N’-tetraacetic acid (EGTA) into the medium during DCS. The results showed that EGTA significantly decreased cell migration during DCS ^[51][54].

4. Molecular Changes—Plasticity and Neuromodulation

It has been suggested that the after-effects of tDCS are related to molecular mechanisms which play a vital role in activity-dependent synaptic plasticity. The molecular changes after DCS regarding plasticity and neuromodulation have been explored in the following contributions.

Ranieri et al. (2012) applied DCS with 200–250 µA for 20 min in rat hippocampal slices, with the electrical field oriented parallel to the somato-dendritic axis of CA1 pyramidal cells. They demonstrated a modulatory effect of DCS on LTP induced by a standard high frequency stimulation (HFS) protocol consisting of four trains of 50 stimuli at 100 Hz (500 ms each) repeated every 20 s ^[52][55]. Specifically, anodal DCS increased, while cathodal DCS decreased LTP at CA3-CA1 synapses in rat hippocampus. Furthermore, an induction of early genes such as c-fos and Zif268, which are relevant for structural plasticity ^[53][56], was observed following DCS.

Chang et al. (2015) stimulated mouse thalamocingulate brain slices with the electrical field oriented parallel as well as perpendicular to the direction of axodendritic fibers to investigate suppressive effects of DCS on the anterior cingulate cortex (ACC). DCS for 15 min at 400 µA induced LTD. Excitatory postsynaptic currents (EPSCs) were monitored via MEA and patch clamp recordings. DCS significantly decreased epileptic EPSCs, which were generated by 4-aminopyridine treatment prior to DCS. Furthermore, the NMDA receptor antagonist D-1-2-amino-5-phosphonopentanoic acid (APV) totally abolished this DCS effect [20].

Cortical excitability alterations due to cathodal DCS have also been explored in another study. mGluR5-mTOR signaling was identified as a novel pathway by which tDCS modulates cortical excitability. Mouse and human coronal slices oriented parallel or orthogonal to M1 fibers (layer V to II/III projections) were exposed for 10 or 20 min to current strengths of 300 or 400 µA. DCS induced LTD in both human and mouse cortices. These effects were abolished by an mGluR5-negative allosteric modulator but stabilized by a mGluR5-positive allosteric modulator [18].