Morphological classification of the rat lateral cerebellar nuclear neurons by principal component analysis.

The deep cerebellar nuclei (DCN) constitute the major structures by which the cerebellum forwards its output to the rest of the brain. Although the connectivity of the DCN has been well studied, little is known about the interface-the neurons' soma and dendrites-between the DCN's inputs and outputs. We therefore decided to analyze the neurons' somatic and dendritic morphology by applying a multivariate approach (principal component analysis; PCA), in order to define morphological groups possibly related to distinct positions in the nuclear microcircuitry. The PCA was based on intracellularly stained neurons from the rat's lateral DCN and on 19 parameters that described the neurons' morphology. The PCA yielded two principal components that accounted for 46% of the variance. The first component, correlated with soma size, separated the majority of neurons (type I) from a population of small neurons (type II). The second component showed negative correlation with larger cells with more numerous primary dendrites and a more multipolar appearance (type Ia) and positive correlation with smaller neurons with asymmetric dendritic fields and tufted dendrites (type Ib). The preponderance of small somata in our type Ib neurons suggests that these neurons probably correspond to the inferior olive projection neurons. In summary, our results are in agreement with previous classifications, which distinguished projection neurons (type I) from local neurons (type II); furthermore, our results point to a hitherto undescribed dendritic morphological difference in the projection neurons. The latter may be important for understanding the phylogenetic changes seen in the mammalian lateral cerebellar nucleus. Copyright 2002 Wiley-Liss, Inc.