Neuropsychologia 50 (2012) 2026–2031 Contents lists available at SciVerse ScienceDirect Neuropsychologia journal homepage: www.elsevier.com/locate/neuropsychologia Facilitation of speech repetition accuracy by theta burst stimulation of the left posterior inferior frontal gyrus Julia Restle a,b,1, Takenobu Murakami a,1, Ulf Ziemann a,n a Motor Cortex Group, Department of Neurology, Goethe University, Schleusenweg 2-16, Frankfurt am Main D-60528, Germany b Department of Physical Education, Gutenberg University, Saarstrasse 21, Mainz D-55122, Germany article info abstract Article history: The posterior part of the inferior frontal gyrus (pIFG) in the left hemisphere is thought to form part of Received 29 December 2011 the putative human mirror neuron system and is assigned a key role in mapping sensory perception Received in revised form onto motor action. Accordingly, the pIFG is involved in motor imitation of the observed actions of 3 April 2012 others but it is not known to what extent speech repetition of auditory-presented sentences is also a Accepted 1 May 2012 function of the pIFG. Here we applied fMRI-guided facilitating intermittent theta burst transcranial Available online 11 May 2012 magnetic stimulation (iTBS), or depressant continuous TBS (cTBS), or intermediate TBS (imTBS) over the Keywords: left pIFG of healthy subjects and compared speech repetition accuracy of foreign Japanese sentences Speech repetition before and after TBS. We found that repetition accuracy improved after iTBS and, to a lesser extent, after Sensorimotor integration of speech imTBS, but remained unchanged after cTBS. In a control experiment, iTBS was applied over the left Mirror neuron system middle occipital gyrus (MOG), a region not involved in sensorimotor processing of auditory-presented Dorsal auditory stream Transcranial magnetic stimulation speech. Repetition accuracy remained unchanged after iTBS of MOG. We argue that the stimulation Posterior inferior frontal gyrus type and stimulation site specific facilitating effect of iTBS over left pIFG on speech repetition accuracy Human indicates a causal role of the human left-hemispheric pIFG in the translation of phonological perception to motor articulatory output for repetition of speech. This effect may prove useful in rehabilitation strategies that combine repetitive speech training with iTBS of the left pIFG in speech disorders, such as aphasia after cerebral stroke. & 2012 Elsevier Ltd. All rights reserved. 1. Introduction matter of debate, area F5 of the macaque brain is thought to be partially homologous to pIFG in the human brain (Petrides & Acquisition of language skills in humans involves social interac- Pandya, 1994; Petrides, Tomaiuolo, Yeterian, & Pandya, 2012). This tions in earliest infancy, and speech repetition plays a fundamental proposition is supported by several neurophysiological and neuroi- role by mapping auditory and sensory input onto matching motor maging studies, which provided evidence for the existence of a output (Hickok & Poeppel, 2004, 2007; Wernicke, 1874). At the level putative mirror neuron system in the human pIFG (Fadiga, Craighero, of neuronal networks in the brain, speech repetition occurs within & D’Ausilio, 2009; Fadiga, Craighero, & Olivier, 2005; Rizzolatti & the auditory dorsal stream, which is constituted by a temporo- Craighero, 2004). parieto-frontal network including the posterior inferior frontal gyrus Transcranial magnetic stimulation (TMS) represents a useful tool (pIFG) (Hickok, Houde, & Rong, 2011; Hickok & Poeppel, 2004, for investigating the mechanisms of sensorimotor integration in the 2007). It has been hypothesized that, in the frontal areas, so called human brain. One technique of investigation is single or paired mirror neurons contribute to this network (Corballis, 2010; pulseTMSoftheprimarymotorcortex (M1) during observing or Rizzolatti & Arbib, 1998). Mirror neurons have first been character- listening to speech while motor evoked potentials (MEP) are ized in macaque area F5 by discharging when the monkey performs recorded as an index of excitability of the stimulated motor actions and perceives the actions of others (Ferrari, Gallese, representation (Hallett, 2007). Increases in MEP amplitudes in hand Rizzolatti, & Fogassi, 2003; Gallese, Fadiga, Fogassi, & Rizzolatti, muscles have been demonstrated during observation of hand and 1996; Rizzolatti, Fadiga, Gallese, & Fogassi, 1996). Although still a arm movements performed by another individual (Baldissera, Cavallari, Craighero, & Fadiga, 2001; Clark, Tremblay, & Ste-Marie, 2004; Fadiga, Fogassi, Pavesi, & Rizzolatti, 1995; Strafella & Paus, n Corresponding author. Present address: Department of Neurology, Hertie 2000). In addition, MEP amplitudes of facial or articulatory muscles Institute of Clinical Brain Research, Eberhard-Karls University, Hoppe-Seyler-Str. involved in speech production were facilitated during observation of 3, Tubingen,¨ D-72076, Germany. Tel.: þ49 69 6301 5739; fax: þ49 69 6301 4498. E-mail address: [email protected] (U. Ziemann). speech related lip movements (Murakami, Restle, & Ziemann, 2011; 1 These authors contributed equally to this work. Sundara, Namasivayam, & Chen, 2001; Watkins, Strafella, & Paus, 0028-3932/$ - see front matter & 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.neuropsychologia.2012.05.001 J. Restle et al. / Neuropsychologia 50 (2012) 2026–2031 2027 2003), and even when just listening to speech (Fadiga, Craighero, hypothesized that significant modulating effects of speech repetition Buccino, & Rizzolatti, 2002; Murakami et al., 2011; Watkins et al., accuracy would occur with TBS of pIFG but not MOG, since MOG is 2003). In an experiment that combined TMS of the face representa- involved in visual processing but is not part of the auditory speech tion of the left M1 with positron emission tomography, increases in processing network. MEP amplitude in the facial orbicularis oris muscle correlated To the best of our knowledge this is the first study that directly with metabolic activity in the left pIFG during listening to examines the importance of the left pIFG for speech repetition speech (Watkins & Paus, 2004). These findings supported the notion accuracy by using rTMS. We provide here direct evidence for a that the left pIFG facilitates excitability of the face and articulatory modulating role of the left pIFG in speech repetition. representations of the left M1 during speech perception, and there- fore, operates at the interface of speech perception and articulation. Another technique of investigation is to use TMS in the so 2. Material and Methods called ‘virtual lesion’ mode (Ziemann, 2010). By applying repeti- 2.1. Participants tive TMS (rTMS) over a certain brain area, the activity of the stimulated network is disrupted or facilitated, which should Eighteen healthy right-handed volunteers (mean age7S.D., 26.074.7 years) result, respectively, in decline or improvement of behavioral participated in this study. All subjects were German native speakers and performance if this network is relevant for the performed task. completely naı¨ve to Japanese language. Their educational level was homogenous Two previous virtual lesion studies provided evidence for the as all of them were university students. Handedness was assessed by the importance of the pIFG in sensorimotor integration processing: Edinburgh Handedness Inventory (Oldfield, 1971) with a mean laterality score of 95.178.6%. The study was approved by the Ethics Committee of medical faculty rTMS of pIFG resulted in reduced accuracy of hitting the correct of the Goethe University Frankfurt, and conducted in accordance with the latest target in a finger movement imitation task (Heiser, Iacoboni, version of the Declaration of Helsinki. All participants gave their written informed Maeda, Marcus, & Mazziotta, 2003) and delayed reaction time in consent prior to participation. an automatic finger movement imitation task (Catmur, Walsh, & Heyes, 2009). 2.2. Speech repetition task Here we present two experiments to clarify the role of the left pIFG in sensorimotor integration of speech processing. In the Subjects sat on a comfortable chair while listening to short Japanese poem main experiment we applied theta burst stimulation (TBS) sentences (IROHAKARUTA) presented via dual loudspeakers. The Japanese sen- tences were recorded from a native Japanese male speaker, using a video camera. (Huang, Edwards, Rounis, Bhatia, & Rothwell, 2005), a particular Each sentence consisted of 8 up to 13 syllables and was presented 4 times at form of rTMS, to the left pIFG, and evaluated the effects of three random intertrial intervals of 10–13 s before moving to the next sentence. Forty TBS protocols (cTBS: continuous TBS, iTBS: intermittent TBS, eight different sentences were prepared and divided into 6 different blocks of imTBS: intermediate TBS) on accuracy of speech repetition of a 8 sentences. For each experimental session one block (i.e., consisting of 4 repeats foreign language (Japanese). The task ensured specific testing of  8 sentences¼32 trials) was randomly selected. Subjects were instructed to repeat every sentence immediately after presentation (Fig. 1A, B). Repetition sensorimotor integration of language along the dorsal auditory accuracy was recorded by a digital video camera at every second and fourth stream because the foreign language provided phonological repetition of a given sentence (i.e., 16 trials were evaluated per block of trials). For material