The ability to non-invasively modulate synaptic plasticity within the human brain, with the presumption that that might aid learning and improve outcome in a variety of neurological and psychiatric conditions has been a major research goal for many years. Understanding the mechanisms underpinning this non-invasive brain stimulation (NIBS) is an important goal, not only as a scientific pursuit, but also so its likely efficacy can be improved and potential side-effects minimised.
Understanding the physiological effects of NIBS is difficult, as the detailed experimental approaches available to colleagues working with cellular or slice preparations or even whole animal models are impossible in healthy human subjects. And yet, if we are truly to achieve the ambition of safely and effectively modulating activity in the human brain, it is vital to continue with these studies, interpreting the results cautiously and constantly testing our assumptions. In an elegant series of experiments in this issue of The Journal of Physiology, McNickle & Carson (2015) do just that.
A sensible approach to the rational development of NIBS in humans is to start by mimicking stimulation protocols used in animal preparations, where we have a detailed cellular understanding of the resulting synaptic plasticity. A number of different NIBS protocols have been developed in this way in recent years, including paired associative stimulation (PAS), the subject of McNickle and Carson's study.
In 2000, Stefan and colleagues took inspiration directly from protocols used routinely in animal models to induce Hebbian plasticity to develop PAS. They showed that repeatedly pairing a discrete median nerve stimulus with a single transcranial magnetic stimulation (TMS) pulse applied to the abductor pollicus brevis muscle representation within the primary motor cortex led to an increase in cortical excitability (Stefan et al. 2000). The precise timing of these two stimuli appears to be critical – ulnar nerve stimulation applied 25 ms before the TMS pulse (a timing that results in the afferent input arriving in the cortex a few milliseconds before the TMS pulse) results in an increase in synaptic strength, whereas if the interstimulus interval is decreased, so that the afferent input arrives in the cortex a few milliseconds after the TMS pulse, then synaptic depression results (Wolters et al. 2003). The effects of this stimulation protocol clearly have much in common with spike timing-dependent plasticity (STDP), and therefore it has been widely held, that PAS equates to STDP in humans (Müller-Dahlhaus et al. 2010).
If STDP is the sole result of PAS, however, two conclusions can be inevitably drawn – that the precise timing of the two stimuli is critical, and that resulting synaptic effects will be seen only within specific ana-tomical pathways. However, as McNickle and Carson state as the starting point of their paper, it is difficult to reconcile this need for temporally specific stimuli with the known effects of a single TMS pulse. A TMS pulse will induce complex neuronal responses that are not clearly localised either within specific anatomical pathways or with a clear temporal pattern. Likewise, a single peripheral nerve stimulus will lead to a series of events that are both anatomically and temporally widespread.
McNickle and Carson test the assumption that temporally discrete stimuli are necessary for PAS by using short bursts of high frequency transcranial alternating current stimulation (tACS) as their cortical stimulation paired with brief trains of peripheral afferent stimulation, repeated for approximately 30 min.
As the name suggests, tACS utilises an alternating electrical current, with a peak-to-peak amplitude of 1–3 mA, passed through the brain via two electrodes placed on the scalp. Here, frequencies from 80 Hz to 250 Hz were used. Neither peripheral nerve stimulation alone nor tACS alone led to any increases in cortical excitability. Crucially, however, pairing the two stimuli led to significant increases in cortical excitability, which outlasted the stimulation by at least 30 min. The authors demonstrated that these effects are highly replicable across subjects. Most importantly, the authors clearly demonstrate that the PAS effect seen was not dependent on temporally discrete peripheral or cortical stimuli.
What does this study tell us? Primarily, it strongly suggests that the LTP-like effects seen after PAS can expressed in the absence of tightly-defined temporal stimuli. This suggests that they are probably due to a multiplicity of cellular pathways, including, but not limited to, STDP. While it does not formally eliminate the possibility that the effects of the traditional PAS are due to STDP alone, it makes it much less likely. This finding is of prime importance if we are to fully understand the effects of PAS, but also acts as a clear warning that directly translating stimulation protocols from animal models to humans is not trivial.
This paper also raises a number of interesting questions. In particular, the effects of tACS on the brain are not yet well understood and one important but as yet unresolved question is whether the application of tACS is able to entrain ongoing oscillatory activity in the brain. In this paper, increased cortical excitability was evident after PAS with three different tACS frequencies, raising the intriguing possibility that the stimulation entrains lower-frequency endogenous oscillations, particularly in the high gamma (60–90 Hz) range. Whether the PAS approach used here would have been more effective if the peripheral stimulation had been timed to the peaks of the tACS is something that warrants exploration.
As the authors themselves highlight, the parameters within which this novel form of PAS is effective have yet to be fully defined in terms of the frequency of the tACS, the peak-to-peak current, and the length of the stimuli used. However, the implication that cellular mechanisms in addition to STDP underlie this form of plasticity is important, and should inform our understanding of these stimulation approaches in the future.
Additional information
Competing interests
None declared.
References
- McNickle E. Carson RG. Paired associative transcranial alternating current stimulation increases the excitability of corticospinal projections in humans. J Physiol. 2015;593:1649–1666. doi: 10.1113/jphysiol.2014.280453. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Müller-Dahlhaus F, Ziemann U. Classen J. Plasticity resembling spike-timing dependent synaptic plasticity: the evidence in human cortex. Front Synaptic Neurosci. 2010;2:34–34. doi: 10.3389/fnsyn.2010.00034. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Stefan K, Kunesch E, Cohen L, Benecke R. Classen J. Induction of plasticity in the human motor cortex by paired associative stimulation. Brain. 2000;123:572–584. doi: 10.1093/brain/123.3.572. [DOI] [PubMed] [Google Scholar]
- Wolters A, Sandbrink F, Schlottmann A, Kunesch E, Stefan K, Cohen LG, Benecke R. Classen J. A temporally asymmetric Hebbian rule governing plasticity in the human motor cortex. J Neurophysiol. 2003;89:2339–2345. doi: 10.1152/jn.00900.2002. [DOI] [PubMed] [Google Scholar]