RAPID COMMUNICATION Emergence of Radial Nerve Dominance in Median Nerve Cortex After Median Nerve Transection in an Adult Squirrel Monkey

Schroeder, C. E., S. Seto, and P. E. Garraghty. Emergence of action potential patterns, as indexed by laminar profiles of radial nerve dominance in median nerve cortex after median nerve current source density (CSD) and multi-unit activity, samtransection in an adult squirrel monkey. J. Neurophysiol. 77: 522– pled using linear array multi-electrodes. Throughout the gla526, 1997. Throughout the glabrous representation in Area 3b, brous representation in Area 3b, electrical stimulation of the electrical stimulation of the dominant (median or ulnar) input prodominant (median or ulnar) input produced robust, shortduces robust, short-latency excitation, evident as a net extracellular latency excitation, evident as a large current sink accompa‘‘sink’’ in the Lamina 4 current source density (CSD) accompanied nied by action potentials in Lamina 4. Stimulation of the by action potentials. Stimulation of the collocated nondominant collocated nondominant (radial nerve) input produced a sub(radial nerve) input produces a subtle short-latency response in tle short-latency response in the Lamina 4 CSD, unaccompathe Lamina 4 CSD unaccompanied by action potentials and followed by a clear excitatory response 12–15 ms later. Laminar nied by action potentials, and followed by a clear excitatory response profiles for both inputs have a ‘‘feedforward’’ pattern, response 12–15 ms later. Laminar response profiles for both with initial activation in Lamina 4, followed by extragranular lamidominant and nondominant inputs had a ‘‘feedforward’’ patnae. Such corepresentation of nondominant radial nerve inputs with tern, with initial activation in Lamina 4, followed by extrathe dominant (median or ulnar nerve) inputs in the glabrous hand granular laminae. surface representation provides a likely mechanism for reorganizaThe normal corepresentation of nondominant dorsum tion after median nerve section in adult primates. To investigate hand (radial) inputs with the dominant (median or ulnar) this, we conducted repeated recordings using an implanted linear inputs in the glabrous hand surface representation would multi-electrode array straddling the cortical laminae at a site in seem to provide a clear vehicle for the biased patterns of ‘‘median nerve cortex’’ ( i.e., at a site with a cutaneous receptive reorganization found after peripheral nerve section (Garfield on the volar surface of D2 and thus with its dominant afferent input conveyed by the median nerve) in an adult squirrel monkey. raghty and Muja 1996; Garraghty et al. 1994; Merzenich et We characterized the baseline responses to median, radial, and al. 1983a,b) . The present experiment tested the prediction ulnar nerve stimulation. We then cut the median nerve and semithat in median nerve cortex, after removal of the dominant chronically monitored radial nerve, ulnar nerve and median nerve input, the nondominant, radial nerve input would acquire the (proximal stump) evoked responses. The radial nerve response in physiologic characteristics of the dominant input. median nerve cortex changed progressively during the weeks after median nerve transection, ultimately assuming the characteristics M E T H O D S of the dominant nerve profile. During this time, median, and ulnar nerve profiles displayed little or no change. Using previously described methods (Schroeder et al. 1995), one squirrel monkey (Saimiri sciureus) weighing 0.6 kg was implanted with a plastic cranial pedestal to provide for head restraint I N T R O D U C T I O N and stabilization of the multi-electrode during repeated measurements. To approximate the conditions of micro-electrode mapping The glabrous hand surface representation in somatosenexperiments, during recording, the monkey was anesthetized with sory cortical Area 3b of primates appears to be divisible a mixture of xylazine (10 mg/kg) and ketamine (30 mg/kg). into median and ulnar nerve ‘‘territories’’ based on which Depth of anesthesia was monitored using palpebral and pad withperipheral nerve provides the dominant or obvious cutaneous drawal reflexes. Body temperature was kept within normal limits input (Wall et al. 1993). Recently, Garraghty et al. (1994) using a water heating blanket. Laminar CSD and multi-unit activity profiles were obtained by and Schroeder et al. (1995) suggested that, in addition to recording with a linear array multicontact electrode, positioned so its dominant input, the cortex within the glabrous surface as to straddle the cortex from the pial surface to the white matter. representation also receives cryptic or nondominant inputs Tactile stimulation was used to define the cutaneous representation from the dorsal hand surface, conveyed by the radial nerve. at the recording site and electrical stimulation then was used to In the latter report, the suggestion was based on the direct quantify the strength, laminar distribution, and temporal pattern of observation that in locations dominated by median or ulnar median, ulnar, and radial nerve inputs. Electrical stimulation was nerve inputs from the glabrous hand surfaces, either cutaprovided by 100 ms, constant-current, square-wave pulses applied neous stimulation of the hand dorsum or electrical stimulawith bipolar electrodes to the skin of the forearm at points detertion of the radial nerve could produce a ‘‘nondominant’’ mined to permit isolated stimulation of individual nerves. Stimulapattern of response. Dominant and nondominant responses tion sites and appropriate current levels (2–3 mA) were determined from preliminary studies of nerve stimulation concurrent were characterized by their local postsynaptic potential and