CHANGES IN THE RESPONSE STATES OF PRIMATE SPINOTHALAMIC TRACT CELLS CAUSED BY MECHANICAL DAMAGE OF THE SKIN OR ACTIVATION OF DESCENDING CONTROLS

1. The responses of a population of 318 spinothalamic tract (STT) cells to mechanical stimulation of the skin were recorded in anesthetized macaque monkeys by several teams of investigators. The responses were subjected to k-means cluster analysis, a multivariate statistical procedure. 2. For an analysis that pertained to the responsiveness of the neurons, the mean responses to four standard mechanical stimuli (Brush, Pressure, Pinch, and Squeeze) were used. Although no true clusters were found, the cells could be partitioned into four groups (called clusters a, b, c, and d) that responded progressively more vigorously to the stimuli. 3. For an analysis that pertained to the selectivity of the cells for various stimulus intensities, from innocuous to highly noxious, the data were normalized by taking the ratio of the mean response evoked by each stimulus to the sum of the responses and multiplying by 100. This procedure does not have a bias toward selection of any particular number of clusters and resulted in three clusters of STT cells. 4. Cluster 1 STT cells responded best to Brush. Cluster 2 cells responded weakly to Brush and Pressure and maximally to Pinch. Cluster 3 cells responded weakly to Brush, Pressure, and Pinch and maximally to Squeeze. 5. The response states of STT cells with respect to mechanical stimulation of the skin can be defined by their cluster assignments on the basis of the responsiveness (clusters a-d) and selectivity (clusters 1-3) of the cells. The response states of newly recorded STT cells can be determined by discriminant analysis from the nearest centroids of the two types of clusters in the reference population of STT cells. 6. No consistent changes in response state were detected when a second series of mechanical stimuli was applied 1 cm from the site stimulated initially or when the stimulus series was alternately repeated at the initial site and at progressively more proximal sites. However, when the stimulus series was applied five times to the initial site, the response state of five of eight cells tested showed a change. Although a change in response state required repetitive damage, even a single stimulus series increased background activity and responses to Brush at undamaged sites. 7. The background activity and responses to Brush and Pressure of all five STT cells recorded in the superficial laminae increased after repeated testing. The background activity of five STT cells recorded in the nucleus proprius also increased, but the responses of only three of the cells to Brush and Pressure increased. However, one cell that showed no change belonged to cluster 1, and the other was tested with a minimally damaging pattern of stimulation. 8. The response states of STT cells could also be changed by descending control systems. Stimulation of the SI cortex or cerebral peduncle preferentially inhibited the responses to Brush and therefore shifted the selectivity of the cells tested. Stimulation in the periaqueductal gray produced a preferential decrease in responses to compressive noxious stimuli and decreased responsiveness. 9. The cluster analyses used in this study for classifying STT cells provide an improvement over those used in previous work from our laboratory. Furthermore, the two types of clusters, reflecting the responsiveness and selectivity of STT cells, describe the response states of these cells with respect to mechanical stimulation of the skin and provide a convenient baseline against which to assess changes due to pathological or physiological manipulations.