Figures and data
Behavioral task.
A, drawings of the reaching task setup. Monkeys were placed in a neck and left arm restraint (dark grey). Handle and cup (light grey) positions were fixed to the table; here the trial has the handle on the animal’s right, and the cup on the left. Holding the handle for one second unlocked the cup, and the food reward was retrieved. A camera positioned above the monkey recorded high speed video footage used for kinematic analysis. B, twelve possible handle and cup arrangements for assessing both elbow extension and flexion reach performance, shown with the same perspective as the camera’s field of view.
Example kinematics.
A, reach trajectories plotted from DeepLabCut predictions of hand location converted to physical distance (mm). Results are from Monkey D following a New M1 and S1 Brodmann’s Area 3a/b lesion. Plots shown are taken from an example baseline session (left four plots) and early post-lesion session (right four plots). Each plot represents a different handle location (start of reach, black square), with three corresponding cup locations (end of reach; black, red or green circles). Repeats of the same reaching movement are overlaid in the same color. Dotted lines represent mean trajectories. B, hand speed plotted for all trials for one movement direction, shown from same baseline and post-lesion sessions as in 2A. The left most plot for each session represents hand speed plotted 0.1 s before to 0.25 s after handle release (time 0 s). Green dots are the maximum speeds per trial; red traces are the mean speed averaged over all trial repeats. The right most plots display overlaid speed traces, interpolated to a standardized length of 50 samples from when the hand first exceeds (reach start) and then returns below (reach end) 0.1 m/s.
Visual representation of results from Table 1.
Regions marked with an asterisk are the lesion targets for that animal. Overall values were calculated using an average of damage across all cortical regions in each monkey. RNm results have been expressed as 100% or 0% as we do not have numerical values for this lesion type. Bars representing damage in New M1 (orange) and S1 Area 3 (green) for Monkey B are presented with dotted lines; these represent the more medial values calculated from 3.5-14.5 mm from midline in this animal due to considerable damage noted to both banks of the sulcus in this monkey only.
Summary of regions damaged from lesions in each monkey.
Data were obtained from post-mortem histological analysis. Rows separate monkeys, ordered from greatest overall damage to cortical layer V cells (Monkey Ca) to least (Monkey Zd). Columns separate cortical regions; numbers represent the percentage of layer V cells in that cortical region that were affected by the lesion. No numerical data were available for RNm lesion verification – the analysis was qualitative. * denotes percentage damage calculated from a more medial extent (3.5-14.5 mm from midline), where there was substantial damage to both anterior and posterior banks of the sulcus for Monkey B, not seen in the other animals.
RNm lesion locations confirmed with histological staining for Monkeys Cm and Ca.
50 µm-thick midbrain sections from both monkeys were stained with cresyl violet to visualize Nissl bodies. Additional counterstaining in alternate sections was conducted for Monkey Ca using Nuclear Fast Red and Perl’s Prussian Blue, allowing for visualization of iron deposits left behind by the lesion electrode. Sections were photographed under bright field illumination. Each column shows an individual lesion site at its largest extent (three in Monkey Ca; two lesions in Monkey Cm). High resolution images of each lesion site are shown next to the corresponding whole midbrain section. AQ = aqueduct; BCX = decussation of the brachium conjunctivum; CG = central gray; CST = corticospinal tract; MM = mammillary bodies; PC = posterior commissure; RNm = magnocellular red nucleus; RNp = parvocellular red nucleus; SN = substantia nigra; SNc = substantia nigra pars compacta; IIIN = oculomotor nerve; III = oculomotor nucleus; 3V = third ventricle
Trajectory variability per week averaged over all reaching directions.
Each graph contains average Mahalanobis distance squared values (AMD2) per week for each monkey. Higher AMD2 scores represent poorer reaching quality. Coloured bars above each plot show the percentage of each cortical region lesioned. Plots have been arranged in an order of largest degree of New M1 damage (orange bar) at the top, to the smallest at the bottom. Vertical lines show lesion times (RNm or cortical); horizontal lines indicate baseline mean ± standard error; circles are median values from all trials performed that week. Standard error of the median was calculated by bootstrapping with 1000 iterations (error bars). AMD2 ranges on the y axis are the same across all plots to allow easier visual comparison. Filled circles are significantly different from baseline (p < 0.05), identified using the two-sample t test and corrected for multiple comparisons using the Benjamini-Hochberg procedure with a false discovery rate of 5%.
Maximum reaching speed per week averaged over all reaching directions.
Each graph contains average maximum speed per week from an individual monkey. The coloured bars above each plot show the percentage of each cortical region lesioned. Plots have been arranged in an order from largest degree of Posterior Old M1 damage (purple bar) at the top, to the smallest at the bottom. Vertical lines show lesion times (RNm or cortical); horizontal lines indicate baseline mean ± standard error; circles represent mean weekly values, error bars show standard error of the mean. Filled circles are significantly different from baseline (p < 0.05), identified using the two-sample t test and corrected for multiple comparisons using the Benjamini-Hochberg procedure with a false discovery rate of 5%.
Maximum trajectory variability per week compared for elbow flexion vs extension reaching directions.
Similar display as Fig. 8, but results shown are the difference between flexion and extension values. The y axis range is comparable across all plots between monkeys to allow easier visual comparison. Note that the axis scale has been inverted, so that worse performance in extension is shown by upwards deflection.
Maximum reaching speed per week compared for elbow flexion vs extension reaching directions.
Similar display as Fig. 9, but results shown are the difference between flexion and extension values to appreciate any differences. The y axis range is comparable across all plots between monkeys to allow easier visual comparison. Worse performance in extension is shown by upwards deflection.
