AN ELECTROPHYSIOLOGICAL STUDY OF DORSAL HORN NEURONS IN THE SPINAL-CORD OF RATS WITH AN EXPERIMENTAL PERIPHERAL NEUROPATHY

1. Extracellular single-unit recordings have been made from 295 dorsal horn neurons in the lumbar enlargement of rat spinal cord; 191 neurons in 20 rats with an experimental peripheral neuropathy, and 104 in 1 0 sham-operated rats. Recordings were made 9-11 days after inducing the neuropathy by tying four loose ligatures around the sciatic nerve in the nerve-injured rats or performing a sham procedure in the sham-operated rats. 2. A survey of the general properties of all neurons encountered was made in the 1 0 sham-operated rats ( 104 neurons) and compared with those seen in 17 of the nerve-injured animals ( 1 80 neurons). The vast majority (87%; 156/180) of neurons recorded in the nerve-injured animals showed abnormal characteristics; these included responses to very gentle mechanical stimulation of the nerve-injury site and to manipulations that resulted in movement of this site such as extension of the leg and probing of the skin and muscle of the thigh (53%), absence of detectable peripheral receptive fields (RFs; 56%), and very high spontaneous activity (7%). In the sham-operated rats none of the neurons recorded could be activated by gentle mechanical stimulation of the sciatic nerve, and only 6% had no detectable peripheral RF. 3. In the nerve-injured animals, 31% (55/180) of cells had both a peripheral RF, and a response to gentle mechanical stimulation of the nerve-injury site. All cells of this type tested (n = 5) showed very prolonged responses (up to 10 min long) to 15 s pinch stimuli applied to the RF and to 15 s gentle tapping of the injury site. The majority of cells in this group were excited by noxious stimuli (71%; 39/55) and had C-fiber inputs (60%; 33/55). 4. The mean threshold temperatures for evoking responses to heat stimuli in cells tested in nerve-injured rats and in sham-operated animals were not different. However, there was a group of neurons in the nerve-injured rats that had low thresholds, failed to encode stimulus intensity, and did not have a C-fiber input. 5. There were significantly fewer neurons excited by low-intensity stimulation of the skin in the nerve-injured (24%; 43/180) than in the sham-operated rats (71%; 74/104). Measurements of mechanical threshold with von Frey hairs showed that, although the mean threshold did not change, none of the cells tested in the nerve-injured animals had thresholds < 12 mN, whereas the lowest threshold recorded in the sham-operated animals was 0.2 mN. 6. We conclude that, whereas some of the abnormalities seen in the properties of dorsal horn neurons in the nerve-injured rats are likely to be deficits produced by the substantial damage to the nerve, such as the loss of peripheral RFs and loss of low-threshold inputs, others, such as the response to stimulation of the injury site, abnormally prolonged responses to stimuli applied to RFs, and increased spontaneous activity are likely to be due to central changes and to the presence of a novel source of input at the site of nerve injury. We propose that input from the nerve-injury site accounts for the pain-related guarding behavior seen in the nerve injured rats and may summate with ''normal'' input to produce the lowered thresholds to external stimuli seen in these animals. This input may also serve to maintain a central abnormality. The prominent afterdischarges seen in some cells are likely to contribute to the abnormally prolonged and intense responses of the nerve-injured animals to even mild noxious stimuli. We conclude that dorsal horn mechanisms make a considerable contribution to the abnormal behavior of rats with an experimental peripheral neuropathy and are likely to contribute to abnormal sensations in patients with neuropathic pain.