Primary motor cortex and phonological recoding: A TMS-EMG study

Abstract Since the 1960s, evidence from healthy participants and brain-damaged patients, neuroimaging and non-invasive brain stimulation studies has specified the neurofunctional architecture of the short-term memory (STM) system supporting the temporary retention of a limited amount of verbal material. Auditory-verbal, later termed Phonological (Ph), STM, or the Phonological Loop, comprises two sub-components: i) the main storage system, the Phonological Short-Term Store (PhSTS), to which auditory verbal stimuli have direct access, and where phonologically coded information is retained for a few seconds; ii) a Rehearsal Process (REH), which actively maintains the trace held in the PhSTS, preventing its decay, and conveys visual verbal material to the PhSTS, after the process of Phonological Recoding (PhREC, or Grapheme-to-Phoneme Conversion) has taken place. PhREC converts visuo-verbal, graphemic, representations into phonological ones. The neural correlates of PhSTM include two discrete regions in the left hemisphere: the temporo-parietal junction (PhSTS) and the inferior frontal gyrus in the premotor cortex (REH). The neural basis of PhREC has been much less investigated. A few single case studies of patients made anarthric by focal or degenerative cortical damage, who show a pattern of impairment indicative of a deficit of PhREC, sparing the REH process, suggest that the primary motor cortex (M1) might be involved. To test this hypothesis in healthy participants with a neurophysiological approach, we measured the corticospinal excitability of M1, by means of Transcranial Magnetic Stimulation (TMS)-induced Motor Evoked Potentials (MEPs), during the execution of phonological judgements on auditorily vs. visually presented words (Experiment #1). Crucially, these phonological tasks involve REH, while PhREC is required only with visual presentation. Results show MEPs with larger amplitude when stimuli are presented visually. Task difficulty does not account for this difference and the result is specific for linguistic stimuli, namely: visual and auditory stimuli that cannot be verbalized lead to different behavioral and neurophysiological patterns (Experiment #2). The increase of corticospinal excitability when words are presented visually can be then interpreted as an indication of the involvement of M1 in PhREC. The present findings elucidate the neural correlates of PhREC, suggesting an involvement of the peripheral motor system in its activity.