A thalamocortical pathway for fast rerouting of tactile information to occipital cortex in congenital blindness

Abstract

In congenitally blind individuals, the occipital cortex responds to various nonvisual inputs. Some animal studies raise the possibility that a subcortical pathway allows fast re-routing of tactile information to the occipital cortex, but this has not been shown in humans. Here we show using magnetoencephalography (MEG) that tactile stimulation produces occipital cortex activations, starting as early as 35 ms in congenitally blind individuals, but not in blindfolded sighted controls. Given our measured thalamic response latencies of 20 ms and a mean estimated lateral geniculate nucleus to primary visual cortex transfer time of 15 ms, we claim that this early occipital response is mediated by a direct thalamo-cortical pathway. We also observed stronger directed connectivity in the alpha band range from posterior thalamus to occipital cortex in congenitally blind participants. Our results strongly suggest the contribution of a fast thalamo-cortical pathway in the cross-modal activation of the occipital cortex in congenitally blind humans.

Fig. 1

Neurophysiological responses following left index finger stimulation. a The source traces of MEG data were estimated across the cortical surface and extracted from regions of interest in thalamus, S1 and V1 in congenitally blind (CB) and sighted control (SC) subjects. The electrocutaneous finger stimulation occurred at 0 ms. Shading indicates the standard error of the mean (SEM) for each group. The significant S1 response in CB and SC and the early V1 response in CB only are indicated by black arrows. b Cortical activations between 35 and 50 ms following left index finger stimulation were stronger in CB vs. SC subjects over contralateral parieto-occipital regions (p < 0.05 uncorrected, permutation test with 5000 randomizations, t-test unequal variance). Medial view of the right hemisphere

Fig. 2

Linear regression model of ROI contributions to MEG data. a Temporal variations of the linear mixing model coefficients shown in each region of interest, following electrocutaneous stimulation of the index finger (0 ms). The data from left and right index finger stimulation were pooled for greater statistical power. The linear coefficient time series are shown for contralateral posterior thalamus (top, with early peak marked with an arrow), primary somatosensory (S1, middle), and visual (V1, bottom) ROIs in CB and SC subjects. Shading indicates standard error of the mean estimates. Gray-shaded boxes highlight segments of significant differences between the two groups (p ≤ 0.05), with their borders colored according to the group showing stronger V1 contribution (two-tailed independent t test, p ≤ 0.05—not corrected for multiple comparison due to a limited number of participants). The thalamus model coefficients are shown only over the first 50 ms after stimulus onset to emphasize the early peak around 20 ms (black arrow). b Average model coefficients in V1 within the [35, 50]-ms time window showed greater V1 contribution in CB vs. SC. Mixed model ANOVA with hemisphere as within-subject variable and group as between-subject variable: no main effect of hemisphere or interactions between hemisphere and group, but significant effect of group (F(1) = 15.6, p = 0.001) reflecting stronger activity in V1 for CB subjects

Fig. 3

Directed functional connectivity measured with phase-transfer entropy (PTE). a CB subjects showed stronger directed connectivity in the alpha frequency band (8–12 Hz) from thalamus (Th) to V1 (p = 0.019) and from S1 to V1 (p = 0.05; two-tailed Wilcoxon rank-sum test, Bonferroni corrected). b SC subjects showed stronger PTE connectivity in the beta-band (12–30 Hz) directed from S1 (p = 0.027) and V1 (p = 0.021) to thalamus (two-tailed Wilcoxon rank-sum test, Bonferroni corrected)

Fig. 4

Proposed scheme showing tactile information flow to the occipital cortex. Tactile information from large myelinated (Aβ) afferents is sent to the VPL and then rerouted to the LGN, from which it is further relayed to the primary visual cortex via the optic radiations. This de novo pathway (in blue) allows fast transmission of tactile information to the occipital cortex in congenitally blind subjects, as witnessed by short occipital response times around 35 ms. Tactile information can also be rerouted via a strengthening of an existing cortico-cortical pathway (in red). Tactile information is relayed from the VPL to SI, and from there further to the occipital cortex via the posterior parietal cortex. Because it involves more synapses, this is a much slower pathway, with response times around 65 ms. Abbreviations: LGN lateral geniculate nucleus, OR optic radiations, PO parietal–occipital area, S1/S2 primary/secondary somatosensory cortex, VIP ventral intraparietal area, VPL ventroposterolateral thalamic nucleus, V1 primary visual cortex, V2/V3 visual areas V2 and V3