Loss of action-related function and connectivity in the blind extrastriate body area

Abstract

The Extrastriate Body Area (EBA) participates in the visual perception and motor actions of body parts. We recently showed that EBA's perceptual function develops independently of visual experience, responding to stimuli with body-part information in a supramodal fashion. However, it is still unclear if the EBA similarly maintains its action-related function. Here, we used fMRI to study motor-evoked responses and connectivity patterns in the congenitally blind brain. We found that, unlike the case of perception, EBA does not develop an action-related response without visual experience. In addition, we show that congenital blindness alters EBA's connectivity profile in a counterintuitive way-functional connectivity with sensorimotor cortices dramatically decreases, whereas connectivity with perception-related visual occipital cortices remains high. To the best of our knowledge, we show for the first time that action-related functions and connectivity in the visual cortex could be contingent on visuomotor experience. We further discuss the role of the EBA within the context of visuomotor control and predictive coding theory.

Keywords: body representation; congenital blindness; extrastriate body area; neuroimaging; plasticity; resting-state fMRI; visuomotor interactions.

FIGURE 1

Motor-evoked responses to body part movements (whole-brain and EBA localizer). (A) Whole-brain statistical parametric maps of hand movements, most commonly associated with EBA activity, plotted on both cortical hemispheres after cortex-based alignment [random effects General Linear Model (GLM)]. Hand movements recruit EBA in sighted individuals but not in blind individuals (see Supplementary Figure 5 for other body parts). EBA, Extrastriate Body Area; CS, central sulcus. (B) Regions of Interest (ROI) analysis of EBA response to various body parts. The region is significantly active in the sighted and not in the blind. A between-group analysis (random effects GLM, n = 17) reveals significant differences in the response to movements of the right foot, right hand, left foot, and left hand (for results in separate hemispheres, see Supplementary Figure 6).

FIGURE 2

M1 responses are comparable between blind and sighted individuals [random effects General Linear Model (GLM)]. There were no significant group differences in M1 activity (left hemisphere) for all unilateral and face movements (Supplementary Table 5). Regions of Interest (ROI) analysis of right M1 produced similar results (Supplementary Figure 7). The sighted and blind groups show similar somatotopic organization in S1 and M1. We created the maps by contrasting each condition with all others. We then plotted and colored all statistically significant vertices (p < 0.05, corrected). The results on the left hemisphere represent the response to right hand, right foot, and face movements. We found similar results in the right hemisphere for contralateral effectors (Supplementary Figure 5). Error bars are for the standard error. Large asterisks indicate between-group significance, small asterisks indicate within-group significance against the baseline. NS, non-significant.

FIGURE 3

Whole-brain connectivity of Extrastriate Body Area (EBA) seed in blind and sighted individuals [random effects General Linear Model (GLM)]. (A) There is a high functional connection between the EBA [independent localizer from Human Connectome Project (HCP) atlas] and visual regions in sighted and blind individuals. However, EBA connectivity to sensorimotor areas exhibits a strikingly different picture. EBA has widespread connectivity to sensorimotor cortices in the sighted. Yet, there is a disconnection between EBA and sensorimotor regions in the congenitally blind. CS, central sulcus. (B) Single-subject connectivity overlap in sighted and blind groups. We created the probabilistic maps from single-subject EBA connectivity results. Connectivity to visual areas was consistent across blind and sighted participants. In addition, there was a connectivity overlap in the sighted around S1, M1, and associative sensorimotor cortices. (C) Group differences of EBA connectivity. The sighted had higher connectivity around sensorimotor regions and weaker connectivity around the Precuneus and lateral Parietal cortex.

FIGURE 4

There is a functional disconnection between the blind Extrastriate Body Area (EBA) and sensorimotor cortices. (A) Regions of Interest (ROIs) from the Human Connectome Project (HCP) atlas (Glasser et al., 2016). A full cytoarchitectural breakdown of each area is in Supplementary Table 3. (B) Congenital blindness alters EBA’s connectivity profile. Seed-to-seed analysis of connectivity strength from the EBA to sensorimotor and visual seeds. In both groups, resting-state connectivity to visual cortices was significantly higher than the baseline. Connectivity to sensorimotor regions was significantly higher in the sighted compared to the blind group (random effects GLM, n = 29). We found similar results when analyzing the connectivity from EBA to the right and left hemispheres separately (Supplementary Figure 8). Error bars are for the standard error. Large asterisks indicate between-group significance, small asterisks indicate within-group significance against the baseline. NS, non-significant.