Somatosensory cortical map changes following digit amputation in adult monkeys

Abstract

The cortical representations of the hand in area 3b in adult owl monkeys were defined with use of microelectrode mapping techniques 2–8 months after surgical amputation of digit 3, or of both digits 2 and 3. Digital nerves were tied to prevent their regeneration within the amputation stump. Successive maps were derived in several monkeys to determine the nature of changes in map organization in the same individuals over time. In all monkeys studied, the representations of adjacent digits and palmar surfaces expanded topographically to occupy most or all of the cortical territories formerly representing the amputated digit(s). With the expansion of the representations of these surrounding skin surfaces (1) there were severalfold increases in their magnification and (2) roughly corresponding decreases in receptive field areas. Thus, with increases in magnification, surrounding skin surfaces were represented in correspondingly finer grain, implying that the rule relating receptive field overlap to separation in distance across the cortex (see Sur et al., '80) was dynamically maintained as receptive fields progressively decreased in size. These studies also revealed that: (1) the discontinuities between the representations of the digits underwent significant translocations (usually by hundreds of microns) after amputation, and sharp new discontinuous boundaries formed where usually separated, expanded digital representations (e.g., of digits 1 and 4) approached each other in the reorganizing map, implying that these map discontinuities are normally dynamically maintained. (2) Changes in receptive field sizes with expansion of representations of surrounding skin surfaces into the deprived cortical zone had a spatial distribution and time course similar to changes in sensory acuity on the stumps of human amputees. This suggests that experience‐dependent map changes result in changes in sensory capabilities. (3) The major topographic changes were limited to a cortical zone 500–700 μm on either side of the initial boundaries of the representation of the amputated digits. More distant regions did not appear to reorganize (i.e., were not occupied by inputs from surrounding skin surfaces) even many months after amputation. (4) The representations of some skin surfaces moved in entirety to locations within the former territories of representation of amputated digits in every monkey studied. In man, no mislocation errors or perceptual distortions result from stimulation of surfaces surrounding a digital amputation. This constitutes further evidence that any given skin surface can be represented by many alternative functional maps at different times of life in these cortical fields (Merzenich et al., '83b). These studies further demonstrate that basic features of somatosensory cortical maps (receptive field sizes, cortical sites of representation of given skin surfaces, representational discontinuities, and probably submodality column boundaries) are dynamically maintained. They suggest that cortical skin surface maps are alterable by experience in adults, and that experience‐dependent map changes reflect and possibly account for concomitant changes in tactual abilities. Finally, these results bear implications for mechanisms underlying these cortical map dynamics.