Batteries and the brain

Abstract

As a result of the work of Galvani, Volta, Purkyne, Hitzig and others over 150 years ago, we know that weak electrical currents can exert potent effects on neural tissue in animals and humans. They can polarise neuronal membranes and modulate the firing of neurones as in Galvanic stimulation of the vestibular system (Goldberg et al. 1984). Development of transcranial direct current stimulation (tDCS) as a research tool and as a potential therapy in neurology that is both simple and safe has focused attention on evidence of how it modulates brain activity. Delivery of direct current (DC) between two large electrodes placed on the head produces long-lasting changes in the CNS that greatly outlast the stimulus. These have variably been shown to alter motor functions (e.g. skill learning, recovery of function following stroke) and also sensory functions. Changes in cognitive function and mood have also attracted attention. Critical issues are what mechanisms underlie these changes and how might they be harnessed reliably. It is usually assumed that tDCS acts on the cerebral cortex, and plastic changes in cortical networks are frequently invoked to explain its long-lasting after-effects. In this issue, Bolzoni et al. (2013) tackle this assumption for motor function and, by extrapolation, for many other functions. In studies in anaesthetised cats, they found that tDCS also produces long-lasting changes in the ‘excitability’ of descending motor pathways that originate in the brainstem brought about by tDCS. This was established using well-understood methods to evoke responses in descending motor pathways which were recorded as compound action potentials from the spinal cord en route to spinal motor nuclei. An important feature is that electrical pulses delivered via in-dwelling extracellular electrodes can be adjusted to activate descending motor pathways trans-synaptically (via activation of terminals presynaptic to the cells with descending axons) and/or directly at the soma/initial segment or descending axon (which usually requires higher stimulus intensities). In each case, the trans-synaptic and directly activated responses can be unambiguously differentiated. All descending paths which were tested (including rubrospinal, reticulospinal and vestibulospinal paths) showed increased trans-synaptic activation following tDCS. The increased responsiveness developed gradually during application of the DC stimulus and greatly outlasted it, a time course observed in human studies. Thus, it is possible that some of the motor effects described in humans include effects mediated ‘remotely’ by brainstem descending motor pathways and that the effects are not exclusively mediated via corticospinal or corticofugal pathways. Changes ascribed to tDCS in other systems may therefore also be mediated via non-cortical paths. Regarding the mechanisms, these robust increases in indirect activation of rubrospinal and reticulospinal neurons following tDCS (rubrospinal probably via cerebellar efferents and reticulospinal from intrinsic reticular or corticoreticular terminals) were also from pathways that are not commonly associated with plasticity, and occurred during anaesthesia. In addition, tDCS sometimes increased output from directly activated axons, including reticulospinal and corticospinal tracts, axons of which were activated in the medulla some distance from sites of synaptic input. Thus, some effects of tDCS may be exerted on axons and therefore be unrelated to synaptic mechanisms, and therefore unrelated to long-term plastic changes. There is ample evidence from animal studies that the precise orientation of the dendritic fields to the current path will influence the mode and magnitude of neuronal excitation (e.g. Hern et al. 1962; Kabakov et al. 2012) and that tDCS can also alter presynaptic function (e.g. Marquez-Ruiz et al. 2012). An issue that needs to be resolved is to what extent the changes described in this study might also operate in humans. Observations in such different species must involve some differences: for example the smaller skull of the cat precludes the use of similarly sized electrodes to those used in the human studies and the different proportional brain anatomy must influence the current paths. The current densities in the study in the cat were probably at least twice those used in human studies. Not surprisingly there were also differences in the time course of effects. A remarkable feature of the work is that depressions in excitability were not reported, whereas the polarity and cranial localization of electrodes used to deliver the tDCS appear critical to its effectiveness in humans. To what extent there are common cellular features underlying the effects of DC stimulation at cortical, subcortical and spinal sites remains to be resolved.