Fig 1.
Unilateral forelimb restraint.
A. Schematic showing a neonate mouse (P1) with unilateral constraint of the right forelimb against the chest using a surgical bandage tape. B. Experimental timeline. C. Experimental design. We traced separately the motor cortex (MCX) contralateral to the constrained right forelimb (termed constrained MCX; blue) and the MCX contralateral to the unconstrained left forelimb (termed unconstrained MCX; red).
Fig 2.
Topographic distribution of labeled CST axon terminals at C7/8.
Averaged heat maps for mice (n = 5) in each group and condition (A-D). EphA4 conditional knockout controls (B) show extensive ipsilateral misprojections. Constraint of one forelimb caused a significant decrease in the density of CST projections from constrained MCX (C) shifting the distribution toward a pattern similar to WT group (A). CST projections from unconstrained MCX show a bilateral increase (D). Contours enclosing the region of highest density of labeling (the border between yellow and green on the heat maps) are shown for each condition. E. Mediolateral distributions of mean axon density from unconstrained MCX (red) and constrained MCX (blue) within gray matter. The y-axis plots the regional density of CST projections. Note the substantial ipsilateral projection from the unconstrained MCX that is as dense as the contralateral projection from the constrained MCX. Light shading plots ±SEM. The arrow indicates the midline. Inset compares unconstrained distribution (red line) with EphA4 conditional knockout control (black), scaled in amplitude to match that of the unconstrained distribution. Constrained distribution (blue line) with EphA4 conditional knockout control, scaled in amplitude to match that of the constrained distribution. The gray line is the distribution of WT controls. Control data replotted from [8]. F. Bar graphs plot the average laterality index (measured as ratio of ipsilateral gray matter labeling divided by contralateral labeling for MCX on each side). Data show a robust bilateral projection in EphA4 conditional knockout control mice with experience (black) due to the abundant ipsilateral CST misprojections (one-way ANOVA, p<0.0001, F3, 71 = 30.18; Bonferroni post-test: p<0.05). The unconstrained MCX reveals no significant changes (red), the constrained MCX shows a robust decrease in the aberrant ipsilateral terminations leading to a more contralateral organization (blue) similar to WT group (gray) (Bonferroni post-test: p>0.05). Constr. MCX.: Constrained motor cortex, unconstr. MCX: unconstrained motor cortex, Ctrl.: control. The color bar represents the axon length in micrometers within in each region of interest. Calibration (A) for heatmaps: 250 μm.
Fig 3.
Forelimb constraint causes dual changes in CS axon terminals morphology.
A-C confocal projection stacked image (30 optical slices). CST axon morphology in EphA4 conditional KO control (A), constrained MCX (B), and unconstrained MCX (C); ipsilateral (left) and contralateral (right) CST. D. Bar graphs plot CST axon branching within ROIs (inset; average of 4–5 mice/group, 4 sections/animal). Contralateral CST axon morphology on the constrained side (blue) of the spinal cord (right) revealed an overall significant difference between the groups (one-way ANOVA, p<0.0001, F2,51 = 37.70). Bonferroni posthoc testing revealed a significant 50% reduction in the mean number of contralateral CST axon branch points per μm originating from constrained MCX (blue) and controls (dark gray, p<0.05). Importantly, the contralateral projections of the unconstrained MCX (red) showed a significant 1.8 times increase in the mean axonal branch points relative to EphA4 control conditional knockout mice (dark gray). There was no difference in the ipsilateral branches for the unconstrained M1 (light red) relative to EphA4 control conditional knockout mice (light gray) (one-Way ANOVA, p<0.0001, F2,51 = 11.71; p>0.05, Bonferroni posthoc) but ipsilateral branching from the constrained MCX (light blue) was significantly less that either in controls (p<0.05, Bonferoni posthoc) or from the unconstrained side (p<0.05, Bonferoni posthoc). The inset shows the ROIs (95μm x95μm) we analyzed CS axon terminals morphology. Calibration: scale bar: 50μm.
Fig 4.
Forelimb constraint reduces mirror sites in EphA4 conditional knockout mice.
A, B. Color maps plot the occurrence of evoked mirror movements at each MCX site. The percent of mirror sites is represented according to a color scale, from the lowest (blue) to the highest (red). The constrained MCX (A) shows a propensity of sites where mirror movements were not evoked (blue) and no sites were mirror movements were evoked most frequently (red). The unconstrained MCX (B) also showed a substantial reduction in mirror sites and a paucity of mirror movement sites. C. Bar graphs plot the average (n = 8 mice; 25 MCX sites within each of 16 hemispheres) of the percentage of sites from which the microstimulation evoked a mirror response. There was a 79% decrease in mirror sites from the constrained MCX (blue) and a 55% reduction in the unconstrained MCX (red) compared with EphA4 conditional knockout controls (p = 0.0004 Mann-Whitney test). The inset shows no mirror sites found in the WT at the threshold (all ‘‘blue” sites), whereas nearly all sites in EphA4 conditional knockout controls evoked mirror movements (mostly “red” sites); reanalyzed from data in [8].
Fig 5.
Forelimb constraint reduces bilateral reaching movements in EphA4 conditional knockout mice.
A. Stacked bar graphs plot the average forelimb use (n = 10–13 mice/group) during exploratory reaching behavior. Wild type mice show a small incidence of simultaneous use of both forelimbs in this task (black, left bar) In contrast, EphA4 conditional knockout control mice use mirror reaching movements nearly 80% of the time (black, middle bar). Unilateral forelimb constraint caused approximately a 50% reduction in mirror movements while reaching the cylinder wall (black right bar). The overall difference between the groups was highly significant (one-Way ANOVA, p<0.0001, F2.31 = 57.70). Bonferroni post-hoc testing revealed a significant difference between EphA4 constrained and conditional knockout control mice (p<0.05), indicating significantly more independent forelimb use during reaching. However, there still was increased mirroring compared with wild type mice (Bonferroni posthoc: p<0.05,). B. Bar graphs plot the normalized mean value of independent forelimb use. We found no significant change in right-left forelimb use in the EphA4 constrained mice (Fig 5B) (p = 0.44 paired t-test) and EphA4 conditional knockout control mice (p = 0.37 paired t-test). Thus, the reduction in mirror reaching movements is due to improved independent limb use not a failure to use one limb.
Fig 6.
Forelimb constraint causes changes in motor skill.
A. Adult mice make more errors while walking on a grid floor, with the limb bandaged during their infancy (right forelimb in this study). The bar graph shows the percent of forelimb slips was significantly higher in the constrained limb than in the unconstrained limb (p = 0.003, paired t-test). No difference was found between the two sides in the control EphA4 conditional knockout group (p > 0.05, paired t-test). Like foot placement accuracy impairment, the grasping time as shown in B was significantly shorter for the constrained than the unconstrained limb (p = 0.004, paired t-test). We found no inter-limb difference in the control EphA4 conditional knockout mice (p = 0.78, paired t-test).
Fig 7.
Effect of unilateral forelimb constraint in wild type mice.
A, B, Heat maps for WT mice (average of 5 mice) show a significant decrease in density of CST labeling from constrained MCX (A), one-way ANOVA, p<0.0009, F2, 49 = 8.20; Bonferroni post-test: p<0.05). Although the CST projections from unconstrained MCX (B) show a slight expansion, there were no significant changes relative to WT controls (Bonferroni post-hoc: p>0.05). Lines mark contours at the yellow-green boundary. Black line is WT control contour (from Fig 2A) for comparison with the wild type constrained/unconstrained MCX (gray lines). Constrained WT show both a forelimb placement (C) and grasp capability (D) impairments (p = 0.0078, paired t-test, p<0.0001 paired t-test, respectively). Calibration (A) for heat maps: 250 μm.