Recovery of sensorimotor function after experimental stroke correlates with restoration of resting-state interhemispheric functional connectivity

Abstract

Despite the success of functional imaging to map changes in brain activation patterns after stroke, spatiotemporal dynamics of cerebral reorganization in correlation with behavioral recovery remain incompletely characterized. Here, we applied resting-state functional magnetic resonance imaging (rs-fMRI) together with behavioral testing to longitudinally assess functional connectivity within neuronal networks, in relation to changes in associated function after unilateral stroke in rats. Our specific goals were (1) to identify temporal alterations in functional connectivity within the bilateral cortical sensorimotor system and (2) to elucidate the relationship between those alterations and changes in sensorimotor function. Our study revealed considerable loss of functional connectivity between ipsilesional and contralesional primary sensorimotor cortex regions, alongside significant sensorimotor function deficits in the first days after stroke. The interhemispheric functional connectivity restored in the following weeks, but remained significantly reduced up to 10 weeks after stroke in animals with lesions that comprised subcortical and cortical tissue, whereas transcallosal neuroanatomical connections were preserved. Intrahemispheric functional connectivity between primary somatosensory and motor cortex areas was preserved in the lesion border zone and moderately enhanced contralesionally. The temporal pattern of changes in functional connectivity between bilateral primary motor and somatosensory cortices correlated significantly with the evolution of sensorimotor function scores. Our study (1) demonstrates that poststroke loss and recovery of sensorimotor function is associated with acute deterioration and subsequent retrieval of interhemispheric functional connectivity within the sensorimotor system and (2) underscores the potential of rs-fMRI to assess spatiotemporal characteristics of functional brain reorganization that may underlie behavioral recovery after brain injury.

Figure 1.

Lesion incidence maps. A, B, Color-coded local incidence of T₂-based lesion, from 20 to 100% for group I (n = 5) (A) and from 11 to 100% for group II (n = 9) (B), overlaid on consecutive coronal rat brain slices from a T₂-weighted template. Lesion areas with clear T₂ prolongation were manually outlined on T₂ maps at 3 d after stroke. The ROIs M1 (primary motor cortex) (dark blue), S1fl (forelimb region of the primary somatosensory cortex) (light blue), and V1 (primary visual cortex) (purple) are depicted on the template. The green areas in group II are voxels with lesioned tissue that overlap with S1fl and V1 regions (B) and were excluded from the functional connectivity analysis by previous lesion segmentation.

Figure 2.

Time course of sensorimotor functions. A, B, Mean (±SD) SPS (in points from 0 to −20) (A) and time difference of adhesive removal from left affected forelimb (in seconds) (B) at different time points after tMCA-O for groups I (dotted lines) and II (solid lines). *p < 0.05 versus pre; ^#p < 0.05 versus group I.

Figure 3.

Functional connectivity with right S1fl. A, B, Mean functional connectivity maps of groups I (A) and II (B), calculated from a seed in the structurally intact right, ipsilesional S1fl before and at different time points after tMCA-O. Maps display Fisher-transformed correlation coefficients (z′) ranging from 0.1 to 0.8 and −0.1 to −0.8 for positive and negative correlations, respectively, overlaid on consecutive coronal rat brain slices from a T₂-weighted template.

Figure 4.

Intrahemispheric connectivity between S1fl and M1. A, B, Mean (±SD) intrahemispheric functional connectivity (z′) between S1fl and M1, in the right ipsilesional (A) and left contralesional hemisphere (B), for groups I (dotted lines) and II (solid lines), before and at different time points after tMCA-O. Linear mixed-model analysis demonstrated a significantly increased functional connectivity between S1fl and M1 in the contralesional hemisphere in group II compared with group I (p < 0.05) (B).

Figure 5.

Interhemispheric connectivity between bilateral brain regions. A–C, Mean (±SD) interhemispheric functional connectivity (z′) between right and left S1fl (A), M1 (B), and V1 (C), for groups I (dotted lines) and II (solid lines), before and at different time points after tMCA-O. *p < 0.05 versus pre; ^#p < 0.05 versus group I.

Figure 6.

Neuroanatomical connectivity with contralesional M1 at chronic poststroke stage. A, ΔR₁ map of a coronal brain slice from a group II rat at 74 d after tMCA-O. Manganese-induced R₁ increase is evident in the contralesional sensorimotor cortex, contralesional caudate–putamen, corpus callosum, and ipsilesional S1fl and M1 at 2 d after MnCl₂ injection in the left, contralesional M1. ΔR₁ was negligible inside the lesion. B, ΔR₁ [in seconds⁻¹; mean (±SD)] in CC and right, ipsilesional M1 and S1fl, calculated from the difference between tissue R₁ before and 2 d after manganese injection in contralesional M1 in rats from group I (white bars) and II (black bars) at 72 d after stroke. C, Coronal brain section from a group II rat at 85 d after tMCA-O, immunohistochemically stained for BDA and PHA-L at 13 d after iontophoretic injections in left, contralesional M1. Staining of the anterograde tracers is clearly visible in the contralesional sensorimotor cortex, transcallosal tracts, and the ipsilesional M1 and S1fl. Remaining tissue within the lesion was detached during tissue preparation.