Mapping the Dynamic Network Interactions Underpinning Cognition: A cTBS-fMRI Study of the Flexible Adaptive Neural System for Semantics

Cerebral Cortex, August 2016;26: 3580–3590 doi:10.1093/cercor/bhw149 Advance Access Publication Date: 30 May 2016 Original Article ORIGINAL ARTICLE Mapping the Dynamic Network Interactions Underpinning Cognition: A cTBS-fMRI Study of the Flexible Adaptive Neural System for Semantics Neuroscience and Aphasia Research Unit (NARU), School of Psychological Sciences, University of Manchester, Manchester, UK Address correspondence to Matthew A. Lambon Ralph, Neuroscience and Aphasia Research Unit (NARU), School of Psychological Sciences (Zochonis Building), University of Manchester, Brunswick Street, Manchester M13 9PL, UK. Email: ; JeYoung Jung, Neuroscience and Aphasia Research Unit (NARU), School of Psychological Sciences (Zochonis Building), University of Manchester, Brunswick Street, Manchester M13 9PL, UK. Email: Abstract Higher cognitive function reflects the interaction of a network of multiple brain regions. Previous investigations have plotted out these networks using functional or structural connectivity approaches. While these map the topography of the regions involved, they do not explore the key aspect of this neuroscience principle—namely that the regions interact in a dynamic fashion. Here, we achieved this aim with respect to semantic memory. Although converging evidence implicates the anterior temporal lobes (ATLs), bilaterally, as a crucial component in semantic representation, the underlying neural interplay between the ATLs remains unclear. By combining continuous theta-burst stimulation (cTBS) with functional magnetic resonance imaging (fMRI), we perturbed the left ventrolateral ATL (vATL) and investigated acute changes in neural activity and effective connectivity of the semantic system. cTBS resulted in decreased activity at the target region and compensatory, increased activity at the contralateral vATL. In addition, there were task-specific increases in effective connectivity between the vATLs, reflecting an increased facilitatory intrinsic connectivity from the right to left vATL. Our results suggest that semantic representation is founded on a flexible, adaptive bilateral neural system and reveals an adaptive plasticity-based mechanism that might support functional recovery after unilateral damage in neurological patients. Key words: anterior temporal lobe, bilateral system, cTBS, fMRI, semantic representation Introduction Human higher cognitive function is not localized to single brain region but, rather, reflects the interaction of a network of multiple brain regions that act in concert to achieve flexible cognitive behaviors. Previous studies have demonstrated these brain networks using functional or structural connectivity approaches (Honey et al. 2009; van den Heuvel et al. 2009). While these studies provide the topology of brain networks, they pretermitted a key aspect of this neuroscience principle: how brain regions interact in a dynamic fashion to achieve cognitive function. Here, we explored this issue targeting a higher cognitive function, semantic memory, by employing a combination of transcranial magnetic stimulation (TMS) and fMRI. Semantic memory refers to our collective knowledge about words, pictures, objects, people, emotions, etc. The neural basis of semantic memory reflects a large-scale network of distributed, interconnected brain regions (Patterson et al. 2007; Binder et al. 2009). Accumulating, convergent evidence indicates that among © The Author 2016. Published by Oxford University Press. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. JeYoung Jung and Matthew A. Lambon Ralph A Flexible Adaptive Bilateral Neural System For Human Semantics | 3581 hemisphere (Lee et al. 2003; O’Shea et al. 2007). The researchers suggested that the upregulation in nonstimulated hemisphere reflected adaptive short-term compensatory plasticity that may underlie functional recovery after stroke. Accordingly, we expected that cTBS over the left vATL would suppress semantically related neural activity at the target region and induce upregulation in the homologous right vATL as well as, potentially, in other parts of the semantic network. To test whether the upregulation of the right vATL reflected adaptive short-term plasticity, we used dynamic causal modeling (DCM) of our fMRI data. We hypothesized that if the upregulation of the right vATL is linked with an acute compensatory reorganization, we should expect increased facilitatory input from the unlesioned to the lesioned hemisphere to improve task performance. Materials and Methods Participants Thirty-five healthy individuals participated in this study, with 10 participants in the behavioral experiment (4 males, mean age, 22.8 ± 2.9 years, range from 20 to 28 years) and 25 in the fMRI experiment (7 males, mean age, 21.9 ± 3.7 years, range from 19 to 34 years). Data from 2 participants in the fMRI experiment were discarded because of excessive head movements (over a voxel). All participants were native English speakers and right handed (Oldfield 1971). Written informed consent was obtained from all participants. The experiment was approved by the local ethics committee. Behavioral Experimental Design and Procedure All participants performed a synonym judgment task and a number judgment task as a control task. The synonym judgment task was adapted from the previous offline-rTMS and fMRI studies targeting ATL (Pobric et al. 2007, 2010a, 2010b, 2010c; Binney et al. 2010; Visser et al. 2012). The task required participants to select which of 2 words was more related in meaning to a probe word. In each trial, 3 words were presented on the screen, a probe (e.g., handy) on the top, the target word (e.g., functional), and the unrelated word (e.g., receptive) at the bottom. The 3 words in each trial were matched for imageability (P = 0.46), word frequency (P = 0.99), and word length (P = 0.62). The number judgment task was designed to match the synonym judgment task for general difficulty level, which was also adapted from previous studies. Participants were asked to select 3 digit numbers (e.g., 325 and 367) that was closer to the probe number (e.g., 358) in numerical value. Participants received TMS stimulation at the left vATL or control site on different days (counterbalanced across participants). The experiment consisted of 2 sessions, the baseline (no-TMS) session and the after TMS session. The baseline session was conducted before or 1 h after the TMS session, and the order of sessions was counterbalanced across participants to control TMS and order effects. Each session contained both tasks, each of which had 60 trials. The trial started with 500 ms fixation, then the stimuli were presented until response or 3000 ms. The order of trials was randomized. E-prime software (...truncated)