Basic and functional effects of transcranial Electrical Stimulation (tES)-An introduction.

Non-invasive brain stimulation (NIBS) has been gaining increased popularity in human neuroscience research during the last years. Among the emerging NIBS tools is transcranial electrical stimulation (tES), whose main modalities are transcranial direct, and alternating current stimulation (tDCS, tACS). In tES, a small current (usually less than 3mA) is delivered through the scalp. Depending on its shape, density, and duration, the applied current induces acute or long-lasting effects on excitability and activity of cerebral regions, and brain networks. tES is increasingly applied in different domains to (a) explore human brain physiology with regard to plasticity, and brain oscillations, (b) explore the impact of brain physiology on cognitive processes, and (c) treat clinical symptoms in neurological and psychiatric diseases. In this review, we give a broad overview of the main mechanisms and applications of these brain stimulation tools.

direction of plasticity appears to depend on the stimulation intensity, where low intensities (0.4 mA) result in diminution of motor cortex excitability, and higher intensities (1 mA) result in enhanced motor cortical excitability. These results may be due to differences in the sensitivity of excitatory and inhibitory synapses to different stimulation intensities (Moliadze et al., 2012). Increases in cortical excitability are also observed following 1, 2, and 5 kHz Preprint submitted to Neurosci Biobehav Rev Finally published in Neurosci Biobehav Rev 85: 81-92 (2018) https://doi.org/10. 1016/j.neubiorev.2017.06.015 13 tACS over the primary motor cortex (1 mA, 10 min, with reference electrode over the contralateral orbit) which might be explained by mechanisms similar to tDCS, i.e., alteration in the neuronal membrane potential (Chaieb et al., 2011). To summarize, depending on stimulation frequency, intensity, and duration, tACS may non-linearly modify cortical excitability, both during and after intervention.

Remote effects of tES
tES results in not only regional effects as described so far, but also in widespread, networklevel changes across the brain which can be monitored with fMRI and EEG. Recent evidence suggests that tDCS affects cortical regions not only beneath the electrodes, but also other cortical and subcortical structures (Keeser et al., 2011;Polanía et al., 2012b). Moreover, effects of tDCS on remote regions that are functionally connected to the stimulated area can be in the same (Antal et al., 2011) or opposite (Stagg et al., 2009) direction compared to the area underneath the electrodes.
A tDCS-fMRI and also a tDCS-EEG study showed enhanced motor network activation, including premotor areas and posterior parietal areas by anodal stimulation of the left primary motor cortex (Polania et al., 2011;Polania et al., 2012b). Moreover, the combined tDCS-fMRI study showed enhanced functional connectivity between the left motor cortex and the ipsilateral thalamus, and caudate nucleus by anodal and reduced connectivity between the left motor cortex and the contralateral putamen by cathodal stimulation of the left motor cortex (Polanía et al., 2012b). Another tDCS-fMRI study showed that bilateral stimulation of the sensorimotor cortices results in extensive changes in functional connectivity, particularly in primary and secondary motor as well as prefrontal cortex (Sehm et al., 2012). In another study by Polanía and co-workers (Polania et al., 2011), EEG signals were recorded from 62 channels to analyze tDCS effects on cortical network function. EEG data were recorded Preprint submitted to Neurosci Biobehav Rev Finally published in Neurosci Biobehav Rev 85: 81-92 (2018) https://doi.org/10. 1016/j.neubiorev.2017.06.015 14 before and after application of anodal tDCS over the left M1 both during performance of voluntary hand movements (finger tapping) and during rest. Results showed that tDCS is not only able to affect resting networks, but also has a boosting effect on motor-task related cortico-cortical functional circuits, especially in the gamma frequency range (Polania et al., 2011).
Likewise, tACS also induces changes in remote oscillatory activity and long-distance, areato-area interactions. In a study by Polanía and co-workers, in-phase and anti-phase 6 Hz tACS over the left prefrontal and parietal cortices were employed to induce theta synchronization and desynchronization between these regions. Respective stimulation protocols had improving/deteriorating effects on performance of a working memory (WM) task. This effect was interpreted as evidence for the causal relevance of theta phase-coupling between prefrontal and parietal areas for WM performance in healthy humans (Polania et al., 2012a). In another study, bihemispheric anti-phase tACS was applied over occipital-parietal areas in the gamma frequency band (40 Hz) while participants fixated on the center of a display presenting a stroboscopic alternative motion (SAM) pattern and pressing response buttons in the case of change in the perceived direction of horizontal or vertical motion (Struber et al., 2014). Increased interhemispheric gamma band coherence during perceived horizontal compared to vertical motion of the SAM display was previously shown in correlative EEG studies (Rose and Buchel, 2005). Elevated interhemispheric coherence (phase synchronization), and thereby increased perception of horizontal compared to vertical motion by tACS, demonstrated the causal role of gamma oscillations for bistable perception (Struber et al., 2014). In a study on rats, Ozen and co-workers applied tES with a sinusoid waveform (0.8, 1.25 or 1.7 Hz) and performed extracellular and intracellular recordings from neocortical and hippocampal neurons. Entrainment of neuronal activity by tES was observed substrates of learning and memory formation (Liebetanz et al., 2002;Rioult-Pedotti et al., 2000). Therefore, tES may be a promising tool for modulating the activity of presumed taskrelevant brain areas and thereby identifying their causal significance for motor learning and adaptation.
To investigate the contribution of motor and frontal cortices in implicit motor learning, Nitsche and colleagues combined tDCS with the serial reaction time task (SRTT) (Nitsche et al., 2003c). In the SRTT, subjects learn to perform a finger movement sequence. Functional imaging and TMS studies suggest an involvement of the primary motor cortex, supplementary motor area, the prefrontal cortex, and the rostral inferior parietal cortex in this task (Honda et al., 1998b;Pascual-Leone et al., 1994). To explore the causal relevance of these areas for the learning process, 1 mA anodal or cathodal tDCS was applied over either the primary motor (M1), premotor, lateral prefrontal, or medial prefrontal cortices during SRTT performance (online tDCS). Whereas stimulation of the remaining cortices had no effect, anodal tDCS of M1 improved motor learning. These observations suggest a critical role of neuroplasticity in the motor learning process, and also the involvement of the primary motor cortex in the acquisition and early consolidation phases of motor learning (Nitsche et al., 2003c). Further tDCS-SRTT studies revealed a critical role of timing of stimulation for the effects of tDCS on motor learning. In contrast to online tDCS, offline stimulation before SRTT performance of either anodal or cathodal tDCS over M1 did not modulate performance in this task (Kuo et al., 2008). These differences might be due to a more unspecific priming effect of tDCS in this case, which will -in contrast to online tDCS -not be focused on taskrelated activated neurons. Previous literature has suggested that the premotor cortex is primarily involved in consolidation of SRTT-related learning, which takes place during rapid eye movement (REM) sleep (Maquet et al., 2000). Given that consolidation also involves Preprint submitted to Neurosci Biobehav Rev Finally published in Neurosci Biobehav Rev 85: 81-92 (2018) plasticity of contributing neuronal connections, and that online boosting of plasticity is superior to offline intervention, it could be hypothesized that stimulation of this area during REM sleep might improve sequence consolidation. In one study, healthy participants learned the SRTT in the evening and stimulation was applied over premotor cortex during REM sleep. In two control experiments, tDCS was delivered during performance of a SRTT-like task, but without repeating sequences (to control for the specificity of stimulation effects on motor learning consolidation), and while subjects were awake (to control for time-dependent but not sleep-dependent consolidation). Premotor anodal tDCS during sleep enhanced consolidation compared to sham, while no improvement was observed in the other conditions. These results suggest an involvement of the premotor cortex in REM sleepassociated consolidation of procedural memory (Nitsche et al., 2010). Further evidence of this involvement was found in a study by Stagg and colleagues (Stagg et al., 2011). Here, sequence learning was investigated by a similar sequential finger press task combined with 1 mA anodal/cathodal tDCS over M1 (10 min, with the return electrode over the contralateral supraorbital ridge). Their results show that stimulation applied during motor practice modulated learning rates in a polarity-specific manner: anodal tDCS increased, while cathodal stimulation decreased the rate of motor sequence learning. The results further showed the dependency of effects on the relative timing of stimulation and motor task performance as anodal or cathodal tDCS applied prior to the motor task slowed learning In addition to tDCS, other modes of tES have also been explored for the modulation of motor functions. Specifically, modulating brain oscillations in a frequency-specific manner by tACS may be relevant for investigating the neurophysiological underpinnings of motor processes.
Motor learning is associated with changes in oscillatory activity and synchronization at alpha within and between brain regions in the motor network. The amplitude of beta activity in contralateral motor areas has been shown to decrease during motor performance and increase after movement cessation (Pfurtscheller, 1981). The causal role of these brain oscillations for motor performance can be explored by their modulation with tACS (Antal and Paulus,

Functional effects of tES on emotional and cognitive processing system in healthy humans
The prefrontal cortex (PFC) provides the neuronal basis for many high level cognitive functions. For example, the dorsolateral prefrontal cortex (DLPFC) is involved in mood and emotional processes (Dolcos et al., 2004;Grimm et al., 2006;Weigand et al., 2013). Positive emotional stimuli and a happier mood are associated with higher activity in the left DLPFC (Habel et al., 2005;Herrington et al., 2005;Sergerie et al., 2005), while higher activity in the right cortex is associated with a negative affect (Belyi, 1987;Perini, 1986;Robinson and Lipsey, 1984). Neuroimaging and electrophysiological data have further identified the left DLPFC as a core region in emotional processing, specifically in the down-regulation of negative emotional conditions (Davidson et al., 2000). As such, tES has been employed in various emotion-related studies as a potentially useful tool to regulate mood and emotional processes through alteration of prefrontal activity and excitability.
A number of studies have shown that tDCS application over the DLPFC does not modulate mood in healthy subjects (Morgan et al., 2014;Motohashi et al., 2013;Plazier et al., 2012), but may suppress negative feelings and affect when subjects are exposed to negative stimuli (Rego et al., 2015) or frustration (Plewnia et al., 2015). Nitsche and colleagues  applied tDCS over the left DLPFC to evaluate its effect on subjective emotional state and emotional state-related information processing in healthy Preprint submitted to Neurosci Biobehav Rev Finally published in Neurosci Biobehav Rev 85: 81-92 (2018) https://doi.org/10. 1016/j.neubiorev.2017.06.015 28 In a proof-of-principle pioneer study, anodal tDCS was applied over the left DLPFC in 10 patients with major depression for five consecutive days (1 mA, 20 min, 35 cm 2 electrodes with the return electrode positioned over the contralateral supraorbital area). Evaluation by the Hamilton Depression Rating Scale (HDRS) and the Beck Depression Inventory (BDI) scores revealed four responders in the active group versus no responders in the sham group.
Improvement was suggested to be associated with enhanced excitability of left DLPFC, which is pathologically hypoactive in major depression (Fregni et al., 2006a). In a subsequent double-blind study in 40 patients, the number of sessions was extended to 10 days and stimulation intensity was increased to 2 mA (Boggio et al., 2008). Anodal tDCS of the occipital cortex and sham tDCS were performed as active and placebo control conditions, respectively. Here, only prefrontal tDCS significantly reduced depression scores evaluated by HDRS and BDI, which persisted for at least 30 days after the last stimulation session. investigate the safety and efficacy of tDCS combined with sertraline hydrochloride in 120 patients with MDD. Serotonin has a notable influence on neuroplasticity and affects learning and memory formation in animals and humans. Previous studies showed that acute application of a selective serotonin reuptake inhibitor (SSRI) enhances and prolongs anodal tDCS-induced LTP-like plasticity in healthy humans (Nitsche et al., 2009). Moreover, chronic application of a SSRI in healthy humans extends the duration of LTP-like plasticity induced by anodal tDCS for more than 24 hours after intervention (Kuo et al., 2016a). In this study by Brunoni and co-workers, participants were divided into four groups and each group received real or sham tDCS combined with either a real SSRI (sertraline) or placebo medication. In all groups, the anode was placed over the left DLPFC and the cathode over the right DLPFC. Stimulation was performed for 10 consecutive working days. The combination of tDCS with sertraline was found to be significantly superior compared to all other intervention groups. A potential mechanism underlying the more efficacious response achieved by combined therapy might be the modulation of key nodes of the mood-relevant cortico-limbic network, which are disrupted during MDD, i.e., synergistic regulation of DLPFC activity by tDCS and limbic system activity by SSRI (Brunoni et al., 2013a;D'Ostilio and Garraux, 2016).
In contrast to the above-mentioned results, some studies have shown no superior effects of active tDCS compared to sham stimulation (Loo et al., 2010;Palm et al., 2012), despite using the same parameters as studies with favourable results (Boggio et al., 2008;Fregni et al., 2006b). Potential explanations for this apparent inconsistency may be the role of pharmacological history, as patients in the sham group were on antidepressant medication (Loo et al., 2010), inter-study differences in the severity of symptoms, and/or a small sample size resulting in insufficient statistical power.