Spino-olivary projections in the rat are anatomically separate from postsynaptic dorsal column projections

Abstract

The gracile nucleus (GN) and lateral part of rostral dorsal accessory olive (rDAO) are important relays for indirect, postsynaptic dorsal column, and direct ascending pathways, respectively, that terminate as climbing fibers in the "hindlimb-receiving" parts of the C1 and C3 zones in the cerebellar cortex. While the spinal cells of origin of that project to GN and rDAO are from largely separate territories in the spinal cord, previous studies have indicated that there could be an area of overlap between these two populations in the medial dorsal horn. Given the access of these two ascending tracts to sensory (thalamic) versus sensorimotor (precerebellar) pathways, the present study therefore addresses the important question of whether or not individual neurons have the potential to contribute axons to both ascending pathways. A double-fluorescent tracer strategy was used in rats (red Retrobeads and Fluoro-Ruby or green Retrobeads and Fluoro-Emerald) to map the spatial distribution of cells of origin of the two projections in the lumbar spinal cord. The two pathways were found to receive input from almost entirely separate territories within the lumbar cord (levels L3-L5). GN predominantly receives input from lamina IV, while rDAO receives its input from three cell populations: medial laminae V-VI, lateral lamina V, and medial laminae VII-VIII. Cells that had axons that branched to supply both GN and rDAO represented only about 1% of either single-labeled cell population. Overall, the findings therefore suggest functional independence of the two ascending pathways.

Keywords: cerebellum; gracile; inferior olive; tracer.

Figure 1

Photomicrographs of anterograde and retrograde labeling. A: An example of anterogradely labeled terminal fibers (arrow) in contralateral rostral dorsal accessory olive (DAO) as a result of injection of bidirectional tracer into the GN (case CFP78). PO, principal olive. B: Low-power photomicrograph of retrogradely labeled neurons (arrows) in lumbar segment L3 arising from an injection into the GN (case CFP104). Approximate outline of spinal cord gray matter is shown by dotted line. C: High-power photomicrograph of retrogradely labeled cells (arrows) in the ipsilateral L4 dorsal root ganglion, following injection of retrograde tracer into GN (case CFP78). D: High-power photomicrographs of anterogradely labeled climbing fibers in contralateral cerebellar cortex arising from injection of bidirectional tracer into rDAO (case CFP87). Arrowhead shows climbing fiber in molecular layer (ml) with arrow pointing to climbing fiber stem axon in granule cell layer (gl). E: Standard sagittal outline of the cerebellar cortex 1.4 mm lateral from the midline with hatching to indicate location of anterogradely labeled climbing fibers after an injection into DAO in case CFP83. II–V, cerebellar lobules II–V; CI, crus I; CII, crus II; Cop, copula pyramidis; PML, paramedian lobule; Sim, lobulus simplex. F: Low-power photomicrograph of retrogradely labeled spinal neurons (arrow) in lumbar segment L3 following injection of retrograde tracer in DAO in case CFP104. Scale bars = 200 μm in A,C; 500 μm in B,F; 50 μm in D.

Figure 2

Distribution of spinal cord neurons following injections into rDAO and GN. Schematic transverse representations of injection sites in rDAO (A), and GN (B) in case CFP89. Values for each level indicate approximate AP coordinates from Bregma (Paxinos and Watson, 2005). ECN, external cuneate nucleus; DAO, dorsal accessory olive; GN, gracile nucleus; MAO, medial accessory olive; MCN, main cuneate nucleus; PO, principal olive. C: Standard transverse outlines of the lumbar spinal cord depicting the pattern of retrograde cell labeling in segmental levels L3, L4, and L5. Magenta dots denote cells labeled from the GN injection site. These are concentrated in a mediolateral band within lamina IV (arrow); green dots denote cell labeling arising from the DAO injection site. Cell labeling from rDAO form three populations: a cluster in the medial aspect of lamina V in L3 (filled arrowhead); a cluster in lateral lamina V and adjacent white matter in L3 (unfilled arrow); and a cluster in the ventromedial aspect of lamina VII/VIII in L3, L4, and L5 (open arrowhead). Each dot represents an individual retrogradely labeled cell. Spinal laminae (I–X) according to Molander et al. (1984). LSN, lateral spinal nucleus; IML, intermediolateral cell column. Scale bars = 500 μm.

Figure 3

Overall distributions of spinal neurons that project to the GN and rDAO. A,B: Schematic transverse outlines of GN and DAO to show location of all injections sites considered for detailed analysis (n = 13). Dark shading shows core injection site area of all cases; light shading shows maximum extent of injection sites. C: Pooled data from all double bidirectional tracer experiments to illustrate the distributions of neurons retrogradely labeled from injection sites centered on GN (magenta) and rDAO (green) in lumbar segments L3, L4, and L5. Each dot indicates a retrogradely labeled cell. Three populations of cells were retrogradely labeled from the injection site in rDAO: i) Filled arrowhead indicates location of cells in the medial part of lamina V in segment L3; ii) arrow indicates location of cells in lateral part of lamina V and adjacent white matter in segment L3; and iii) open arrowhead indicates the cell population in the medial part of laminae VII and VIII in L3, L4, and L5. D: Same segments showing data for each case depicted as outlines of “clusters” of neurons. Magenta outlines show regions projecting to GN, green outlines show regions projecting to rDAO. See Materials and Methods for further details. Arrow indicates main region of overlap of magenta and green outlines. Scale bar = 500 μm.

Figure 4

Quantitative analysis of lamina distribution of L3–L5 spinal projection neurons. A: Histogram plots mean cell counts per spinal laminae and spinal nuclei for segmental levels L3–L5 arising from injections into GN. B: Histogram plots mean cell counts per spinal laminae and spinal nuclei for segmental levels L3–L5 arising from injections into rDAO. Counts are Abercrombie-corrected (see Materials and Methods for further details). Data are plotted as mean ± SEM, n = 13 double tracer experiments. Statistical comparisons (one-way ANOVA and Tukey’s multiple comparisons test) were made between laminae at each segmental level, ***P < 0.001. LSN, lateral spinal nucleus.

Figure 5

Lack of double-labeled neurons following retrograde tracer injections into GN and DAO. A–C: High-power photomicrographs showing an example double-labeled cell in case CFP111 (A, field viewed for green fluorescence; B, same field viewed for red fluorescence; C, both views combined). D: Injection sites in GN and rDAO. E: Upper panel, distribution of magenta and green single-labeled cells in spinal cord segmental levels L3–L5. Each dot represents an individual labeled cell. Lower panel, distribution of double-labeled cells (blue crosses). Same abbreviations as in Fig. 2. Scale bar = 50 μm in C (applies to A,B).