Reevaluation of synaptic relationships of cerebellar terminals in the ventral lateral nucleus of the rhesus monkey thalamus based on serial section analysis and three-dimensional reconstruction

Abstract

Terminals of cerebellar afferents (CB) to different regions of the ventral lateral nucleus (VL) of the rhesus monkey thalamus were labeled with wheat germ agglutinin-horseradish peroxidase following injections into the dentate nucleus. Synaptic relationships of 17 CB with projection neuron dendrites (PNd) and local circuit neuron dendrites (LCNd) were analyzed in serial ultrathin sections from dorsal and ventral VL regions, which are known to differ cytoarchitecturally and functionally. Three terminals were reconstructed using three-dimensional (3D) computer image analysis techniques to obtain volumetric and planar measurements. CB in the ventral VL were often flat and elongated with synaptic vesicles arranged in clusters. Each CB was engaged with one PNd and one to four LCNd. A single bouton formed 8-50 synaptic contacts, with those on PNd outnumbering the ones on LCNd 4.1:1. Only some CB in the ventral VL were engaged in complex synaptic arrangements such as triads and serial synapses. Most CB in the dorsal VL displayed a roundish shape and numerous uniformly distributed synaptic vesicles. They formed 5-25 synaptic contacts with a 3:1 ratio of contacts on PNd compared with those on LCNd. CB in the dorsal VL participated in a variety of complex synaptic arrangements. Two types of triads were found: classic with CB, PNd and LCNd, and unconventional with CB and two LCNd. CB were also involved in serial synapses with two LCNd or LCNd and another PNd, and serial sequential synapses with two LCNd and a PNd. Three glomerulus-like structures were encountered in the dorsal VL. 3D reconstruction and volumetric measurements revealed that synaptic contacts formed by CB on PNd had varying shapes and sizes (0.022-0.274 microns2). Synapses formed on LCNd were larger (0.09-0.407 microns2). The total area of all active zones of a single CB on LCNd was either equal to or about 40% smaller than that of synapses on PNd. The entire active zone area comprised 1-1.6% of the total CB surface area and did not seem to correlate with the volume. Synaptic contacts formed by associated LCNd on PNd in complex arrangements were usually small (0.021-0.044 micron2). The results suggest that: synapses formed by CB on PNd and LCNd, and synapses formed by LCNd on PNd may differ in strength; a variety of different circuits participate in the processing of cerebellar afferent information in the primate VL; and these circuits differ in functionally different VL subdivisions.

Organization of projections from the anterior pole of the nucleus reticularis thalami (NRT) to subdivisions of the motor thalamus: Light and electron microscopic studies in the Rhesus monkey

“…In the monkey, motor-related nuclei the dendrodendritic synapses are usually a part of complex synaptic arrangements positioned to modulate the input from subcortical afferents (Mason et al, 1996;Ilinsky et al, 1997;Kultas-Ilinsky et al, 1997). The dendrites of local circuit neurons also form synapses on each other resulting in rather complex local circuitry, especially in the VL .…”

Serial electron microscopic reconstruction of axon terminals on physiologically identified thalamocortical neurons in the cat ventral lateral nucleus

“…Similar attempts to assess terminal distribution have been performed on four TCNs (Wilson et al, 1984), two TCNs (Raczkowski et al, 1988), and one LCN (Hamos et al, 1985) in the cat LGN. Although previous studies implied that afferents to TCNs are spatially segregated in the VL, this implication was derived from observations of short and limited segments of dendrites (Rinvik and Grofová, 1974a;Kultas-Ilinsky et al, 1985;Kultas-Ilinsky and Ilinsky, 1991;Sawyer et al, 1994;Mason et al, 1996;Sato et al, 1996). Precise quantitative information about the distribution of different types of synapses on single identified VL neurons was previously lacking.…”

“…That study focused on cerebellar terminals, and we quantitatively analyzed short segments of TCN dendrites that were contacted by cerebellar terminals by using serial ultrathin sections. However, in addition to cerebellar terminals, various other terminals are concentrated on TCNs (Rinvik and Grofová , 1974a;Harding and Powell, 1977;KultasIlinsky et al, 1985;Kultas-Ilinsky and Ilinsky, 1991;Mason et al, 1996;Sato et al, 1996). Therefore, information about the distribution of the synapses formed by various afferents should aid in revealing how motor information is processed in the VL.…”

“…However, there has been no systematic attempt to determine quantitatively the distribution of particular types of synapses on the surface of a single identified cell in the VL. Previous studies have involved isolated electron photomicrographs or a short series of electron photomicrographs of isolated dendritic segments (Kultas-Ilinsky and Ilinsky, 1991;Mason et al, 1996;Sato et al, 1996). Therefore, these previous studies do not provide quantitative information regarding how terminals of different origins are distributed, or how excitatory and inhibitory terminals are distributed on the somata and different portions of dendrites of single TCNs of the VL.…”

“…There are two types of synaptic boutons with round vesicles in the VL of the cat and monkey. One type, which is larger, originates from the cerebellar nuclei (Harding, 1973b;Rinvik and Grofová, 1974b;Kultas-Ilinsky and Ilinsky, 1991;Mason et al, 1996;Sato et al, 1996), and the other smaller type originates from the motor area of the cerebral cortex (Harding, 1973b;Grofová and Rinvik, 1974;KultasIlinsky and Ilinsky, 1991). Synaptic boutons with flattened vesicles have been classified into three types in the cat motor thalamus (Rinvik and Grofová , 1974a).…”