Correlation between activity of Purkinje cells and its modification by natural peripheral stimuli

This study was designed to evaluate the hypothesis that correlations exist between the simple spike activity of neighboring Purkinje cells. In addition to characterizing these correlations using cross-correlation techniques, the experiments examine the modifications of these temporal correlations produced by natural peripheral stimuli. These studies were performed in unanesthetized cats decerebrated at the mid-brain level. The activity of two to four Purkinje cells were recorded simultaneously on the surface of the cerebellar cortex in the anterior lobe using multiple microelectrodes. The electrodes were aligned relative to the direction of parallel fibers by directly determining the distribution of these fibers using a surface-stimulating electrode. Natural proprioceptive and cutaneous inputs consisted of stretch of the gastrocnemius-soleus muscle or sinusoidal deformation of cutaneous structures of the forelimb, respectively. Cross-correlation of the simple spike discharge of Purkinje cells was calculated during periods of spontaneous activity as well as during the application of proprioceptive and cutaneous stimuli. Short-duration positive correlations with short lag times occurred only between those Purkinje cells separated by less than 100 μm or recorded on the same microelectrode. Correlations of this type were present between some cells during spontaneous activity, while for other pairs of cells they were found only during the application of natural stimuli. Long-duration positive and/or negative correlations were found for approximately one-quarter of the Purkinje cell pairs separated by distances up to 750 μm. There was no obvious relationship between the presence of positive or negative correlations and the spatial relationship of the Purkinje cells relative to the direction of parallel fibers. The application of natural peripheral stimuli often augmented the long-duration positive correlation occurring between the simple spike discharge of Purkinje cells. For many pairs of Purkinje cells these changes in the correlation were independent of the changes in their average firing rate as assessed by poststimulus time histograms or cycle histograms. These findings demonstrate that, in addition to short-latency, short-duration positive correlations in the simple spike discharge of Purkinje cells separated by short distances, long-duration correlations exist between the discharge of Purkinje cells separated by considerably greater distances. Equally important, these correlations could be accentuated by natural peripheral and cutaneous stimuli independent of changes these stimuli evoked in the average firing rate. Due to the pattern of convergence of the corticonuclear projection, it is hypothesized that the changes in these correlations resulting from peripheral stimuli may provide one of the mechanisms by which the cerebellar cortex can modify the output of cerebellar nuclear neurons.