Dopaminergic modulation of cortical motor network lateralization

Introduction Unilateral movements are primarily processed in contralateral cortical and subcortical areas and additionally in ipsilateral cerebellum, leading to an asymmetric pattern of neural activation. Decrease of lateralization is characteristic of aging (Naccarato et al., 2006; Wu et al., 2005), and disease, for example, in unilateral brain lesions or stroke (Carr et al., 1993; Rehme et al., 2011) and Parkinson's disease (PD; Wu et al., 2015). The explanation for imbalanced lateralization in drug-naive PD is an adaptive compensation, compatible with the finding that PD-associated deficient input from cortico-subcortical circuits is compensated by reduced cortical inhibition and increased cortical facilitation (Blesa et al., 2017). Here, we investigated the effect of dopamine depletion and substitution on cortical motor lateralization, with the hypothesis that lateralization decreases in advanced PD and that administration of levodopa, at least to a certain extent, reinstates lateralization. Methods We used fMRI to study motor activation in advanced PD patients and in healthy controls (HC) during unilateral upper and lower limb movements. Ten right-handed, left side symptom-dominant PD patients were tested in pseudo-randomized order after intake of their usual dopaminergic medication – 'ON' state – and after withdrawal of medication – 'OFF' state. Eighteen right-handed age-matched HC participated in a single session. We quantified activation lateralization using the average laterality index (AveLI; Matsuo et al., 2012) in three cortical motor regions of interest (ROIs): primary motor cortex (M1), supplementary motor area (SMA) and premotor cortex (PMC), during the four movement conditions. We compared AveLI between group pairs (PD OFF vs. HC, PD ON vs. HC, PD OFF vs. PD ON) within each ROI and movement condition. We estimated the effective connectivity between ROIs using bilinear dynamic causal modeling (DCM; Friston et al., 2003) and developed a measure to quantify the lateralization of the resulting connectivity networks to compare between groups. By constructing a group level parametric empirical Bayes (PEB) model (Friston et al., 2016) over all the subjects and conducting a search over nested models, we compared DCM parameter estimates between groups, thus providing the potential link between changes in motor lateralization and connectivity. Results In line with our predictions, motor activation lateralization as estimated with the AveLI showed a trend towards decrease in the PD OFF group compared to HC, in all three ROIs during left hand movement and in M1 during left foot movement (Fig. 1). Between-group differences were observed solely in conditions corresponding to movement of the more affected body side. Contrary to our hypothesis, dopamine substitution did not reinstate lateralization – in fact, AveLI in the PD ON group closely resembled that of the PD OFF group. Connectivity lateralization of input-specific modulation (DCM.B) networks was significantly lower in all conditions in the PD group as compared to HC. While on the body side more affected by PD, differences were found for both PD OFF and PD ON, input-specific modulation related to the less affected side was more altered in PD ON. PEB analysis revealed qualitatively more between-group differences in input-specific modulation on the more affected PD side and included many interhemispheric connections (Fig. 2). Conclusions Decreased lateralization is not only present in drug-naïve PD patients (Wu et al., 2015) but also in dopa-treated patients. Acute dopamine modulation does not alter lateralization. Decreased lateralization is evident in both fMRI activation amplitudes (as estimated with AveLI) and effective connectivity (as demonstrated through the DCM analysis).