The associative brain at work: Evidence from paired associative stimulation studies in humans

A. Suppa, A. Quartarone, H. Siebner, R. Chen, V. Di Lazzaro, P. Del Giudice, Walter Paulus, J.C. Rothwell, U. Ziemann, J. Classen
2017 Clinical Neurophysiology  
SECTION 1: PHYSIOLOGY OF PAS PAS represents a non-invasive brain stimulation protocol by which bidirectional changes of neuronal excitability can be induced. In the original protocol one route of stimulation is via somatosensory afferents and the other route via TMS of the motor cortex, but pairings with events converging at other sites of the nervous system and inductions with other stimulation modalities have also generated comparable physiological effects. Repetitive pairings of TMS of the
more » ... imary motor cortex (M1) conjointly with an afferent input to M1 (such as somatosensory information by peripheral nerve stimulation, e.g. median nerve stimulation ( MNS), result in changes of the amplitude of motor evoked potentials (MEP) (Stefan et al., 2000, for review see Müller-Dahlhaus, 2010; Carson and Kennedy, 2013). The direction of changes of MEP amplitudes critically depends on the interval between MNS and TMS. This form of timing dependent plasticity in conscious humans has similarities with STDP as revealed in a variety of model systems (Caporale and Dan, 2008), ranging from cultured neurons (Bi and Poo, 1998) and cortical slice preparations (Magee and Johnston, 1997, Markram et al., 1997) to intact animals (Zhang et al., 1998). These similarities include rapid induction, lasting duration and specificity to the stimulated representation, although the latter property cannot by proven at a microscopic scale using non-invasive stimulation methods. Several lines of evidence suggested that the site of action of PAS-induced plasticity is at the level of the cortex (Müller-Dahlhaus et al., 2010; Carson and Kennedy, 2013). Interstimulus intervals (ISIs) in the order of 20-25 ms led to a lasting enhancement of MEP amplitudes, whereas ISIs of around 10ms (PAS10) result in depression (Wolters et al., 2003; Weise et al., 2013). This has been suggested to reflect the sequence of events induced at the level of the M1 (Wolters et al., 2003, 2005). In the original version of PAS, the interval between MNS and TMS was 25ms (PAS25). The first component (N20) of the median nerve somatosensory-evoked potential (MN-SEP) typically arrives in the primary somatosensory cortex (S1) at around 20 ms (Allison et al., 1991) . Taking into induced oppositely directed changes. Together these findings indicate that the notion of a singular afferent route to the M1 driving PAS-related plasticity is probably wrong; it also suggests that simple models of the effects of TDCS and TBS on motor cortex may not be valid when translated to the different anatomy of the cerebellum. The data point to cerebellar contribution to the canonical PAS response, yet its nature remains to be further clarified. According to an influential theory, long-term potentiation/depression (LTP/LTD) of glutamatergic synapses is induced by the magnitude, and perhaps the speed, of Ca 2+ surge in the postsynaptic cell (Lisman, 1989) . There are at least three mechanisms by which postsynaptic [Ca 2+ ] is increased: Upon near-synchronous stimulation, Ca 2+ ions may enter the postsynaptic cell through (i) N-Methyl-D-Aspartate (NMDA)-receptors or (ii) by voltage-dependent Ca 2+ -channels. Ca 2+ may also increase (iii) as a consequence of release from intracellular Ca 2+ -stores. An important component of the canonical STDP is the participation of back-propagating action potentials which travel from the neuron's soma or from the proximal dendrite back into the dendritic tree (Feldman, 2012). Back-propagation provides an important dendritic signal which interacts with incoming signals at the dendritic tree. PAS-induced plasticity clearly represents a system level response toward a stimulation protocol whose components each engage neuronal ensembles, not single neuronal cells or even compartments of neurons. Although a macroscopic response, relatively simple models that are built on principles of STDP and Ca 2+ -dependent plasticity at the level of glutamatergic synapses successfully capture several of its essential properties (Fung and Robinson, 2013) . While PAS-induced plasticity shares certain properties with STDP of glutamatergic synapses important differences remain and similarity must not be taken as identity and caution must be applied when comparing system-level findings with cellular mechanisms (Müller-Dahlhaus et al., 2010; Carson and Kennedy, 2013) . Excitability enhancing PAS did not alter γ-aminobutyric acid (GABA)A-ergic inhibition in M1 (Stefan et al., 2002), as indexed by short-interval paired-pulse TMS (Kujirai et al., 1993) . Thus, LTP-like PAS effects do not arise from long-term disinhibition of the M1at least as far as testable
doi:10.1016/j.clinph.2017.08.003 pmid:28938144 fatcat:6onuazbuvjd2nmw6pfo2ey3jxu