Arm Control and its Recovery after Selective Lesions of Sensorimotor Cortex and the Red Nucleus: A Kinematic Study in Non-Human Primates

Cortical and subcortical lesions to the motor system, as often occur with stroke, typically lead to transitions through a stereotyped upper limb recovery sequence. After initial weakness and loss of dexterity, spasticity and fixed muscle coactivation patterns (synergies) appear. Early work suggested that different features arise from distinct primary motor cortex (M1) subdivisions. Here we investigated this with modern methods, using ischemic lesions of various cortical areas and electrocoagulation lesions of magnocellular red nucleus (RNm) in rhesus monkeys. Nine animals were trained on a reach and grasp task; hand kinematics were assessed with markerless tracking. The proportion of damaged cortical layer V cells in each cortical area was quantified, and corresponding kinematic effects evaluated. Reaching speed showed greater and more persistent reductions with larger lesions to the posterior part of M1 on the gyrus (Posterior Old M1 in Stricks terminology). Initial increases in trajectory variability were more consistent with greater damage within the central sulcus (New M1); these partially recovered. Lesions involving Anterior Old M1 (Area 4s in Hines terminology) had no additive negative effects. An extensive cortical lesion, which combined New and Old M1 with pre-motor and somatosensory cortex damage did not produce a worse or more persistent deficit than lesions limited to M1, suggesting that loss of arm control arose mainly from damage to descending pathways rather than cortico-cortical interactions. Lesions of RNm led to long-lasting slowing of reach, but no increase in variability. Subsequent cortical lesions to Old M1 led to more severe effects, and worse recovery, than without the preceding RNm lesion. This suggests an important neural compensatory role for the rubrospinal tract following cortical damage in monkey, which is not available in humans where the rubrospinal tract is vestigial. None of the lesions investigated led to overt abnormal synergies. The results are consistent with known differences in descending connections from each area: New M1 has fast cortico-motoneuronal output, known to be important for fine motor control (here assessed by trajectory variability); Old M1 has cortico-reticular connections able to activate the reticulospinal tract, important for generating the high forces needed for fast movements.