Reply from Vincenzo Di Lazzaro, Fabio Pilato, Michele Dileone, Pietro A. Tonali and Ulf Ziemann

Abstract

We thank Pompeia and colleagues for their insightful comments on our recent paper (Di Lazzaro et al. 2005b) in which we demonstrate that measures of transcranial magnetic stimulation (TMS) provide an opportunity to segregate physiologically relevant differences of benzodiazepine action in the intact human brain. By using paired pulse TMS protocols, or by coupling of peripheral nerve stimulation with TMS of the contralateral motor cortex, it is possible to recruit specific neuronal circuits of the human brain and to evaluate in vivo the effects of drugs on several neurotransmitter systems (Ziemann, 2004). The short latency inhibitory effect produced by peripheral nerve stimulation on the excitability of the contralateral motor cortex is known as short latency afferent inhibition (SAI). SAI tests an inhibitory circuit in motor cortex (Tokimura et al. 2000) that is controlled by central cholinergic activity: SAI is decreased by the muscarinic receptor antagonist scopolamine in normal subjects (Di Lazzaro et al. 2000b), is significantly reduced in Alzheimer's disease patients and, in these patients, can be increased by acetylcholinesterase inhibitors (Di Lazzaro et al. 2005a). A particular paired pulse TMS protocol (Kujirai et al. 1993) elicits short latency intracortical inhibition (SICI). SICI tests an inhibitory circuit in motor cortex in which neurotransmission through the GABAA receptor is involved (Ziemann et al. 1996; Di Lazzaro et al. 2000a; Ilic et al. 2002) while, in contrast to SAI, this measure is not modified by scopolamine (Di Lazzaro et al. 2000b), and decreased by an acetylcholinesterase inhibitor (Korchounov et al. 2005). Thus, evaluation of the effects of benzodiazepines on SAI and SICI provides a novel and fascinating opportunity to evaluate their effects on distinct inhibitory circuits in human motor cortex. In our recent study (Di Lazzaro et al. 2005b), we provide evidence, for the first time, for a dissociation of lorazepam and a ‘classical benzodiazepine’ (diazepam) on these measures. While both drugs enhance SICI, lorazepam decreases SAI and diazepam increases it. This dissociation may contribute to our understanding of why these two benzodiazepines impair memory function differently. What makes lorazepam different from other benzodiazepines remains unknown though Pompeia and colleagues suggest in their letter that a possible explanation for the atypical effects of lorazepam might be that it has a unique binding profile, possibly to as yet uncharacterized benzodiazepine receptors. This idea bears much similarity with our as yet unproven proposal that lorazepam and diazepam differ in their affinity to subtypes of the GABAA receptor bearing different alpha-units (Di Lazzaro et al. 2005b). If we conceive SAI as a GABAergic cortical inhibition positively controlled by acetylcholine, then the dissociation between lorazepam and diazepam on SAI may be caused by the GABAA receptor that mediates inhibition in the GABAergic motor cortical SAI circuit, or the GABAA receptor that controls release of acetylcholine at brainstem or intracortical levels (Giorgetti et al. 2000). According to the first scenario, which we speculated upon in our recent paper (Di Lazzaro et al. 2005b), lorazepam and diazepam both reduce acetylcholine release to a similar extent, therefore exerting a similar depression on SAI, but diazepam is significantly more effective than lorazepam in enhancing inhibition in the GABAergic motor cortical SAI circuit. This could explain the observed net effect of increased SAI under diazepam but decreased SAI under lorazepam (Di Lazzaro et al. 2005b). The second scenario, which is equally possible, is that lorazepam decreases acetylcholine release more effectively than diazepam while both drugs have similar effects on inhibition in the GABAergic SAI circuit, again resulting in the same dissociating net effects on SAI as above. If true, this would indicate that measures of TMS can be applied to probe cortical inhibition at the level of different subtypes of GABAA receptors. This would be a major advance, not only in comparison to psychophysical studies and less specific electrophysiological studies (Itil et al. 1989; Pompeia et al. 2003), as quoted by Pompeia and colleagues in their letter to the editor, but also in comparison to all previous TMS drug studies (Ziemann, 2004). The significance of this research lies in the fact that cortical GABAergic interneurones are highly diverse and operate with a corresponding diversity of GABAA receptor subtypes in controlling developmental plasticity and behaviour (Fagiolini et al. 2004; Mohler et al. 2004). We agree with Pompeia and colleagues that much more needs to be done, including dose–response curves and testing of other atypical benzodiazepines, to corroborate the view that TMS can distinguish non-invasively and at the systems level of human cerebral cortex between neuronal circuits bearing different GABAA receptor subtypes. However, it appears that a first step towards this important goal has been achieved.