Topography of the complete corticopontine projection: from experiments to principal Maps

Abstract

The mammalian brain is characterized by orderly spatial distribution of its cellular components, commonly referred to as topographical organization. The topography of cortical and subcortical maps is thought to represent functional or computational properties. In the present investigation, we have studied map transformations and organizing principles in the projections from the cerebral cortex to the pontine nuclei, with emphasis on the mapping of the cortex as a whole onto the pontine nuclei. Following single or multiple axonal tracer injections into different cortical regions, three-dimensional (3-D) distributions of anterogradely labeled axons in the pontine nuclei were mapped. All 3-D reconstructed data sets were normalized to a standardized local coordinate system for the pontine nuclei and uploaded in a database application (FACCS, Functional Anatomy of the Cerebro-Cerebellar System, available via The Rodent Brain Workbench, http://www.rbwb.org). The database application allowed flexible use of the data in novel combinations, and use of a previously published data sets. Visualization of different combinations of data was used to explore alternative principles of organization. As a result of these analyses, a principal map of the topography of corticopontine projections was developed. This map followed the organization of early spatiotemporal gradients present in the cerebral cortex and the pontine nuclei. With the principal map for corticopontine projections, a fairly accurate prediction of pontine target area can be made for any site of origin in the cerebral cortex. The map and the underlying shared data sets represent a basis for modeling of topographical organization and structure-function relationships in this system.

Keywords: 3-D reconstruction; axonal tracing; brain map; cerebellum; development; neuroinformatics; pontine nuclei.

Figure 1

Distribution of triple tracer injections and anterogradely labeled fibers. A–C, schematic drawings of coronal sections through the center of Fluoro-Ruby, FR (A), biotinylated dextran amine, BDA (B), and Fluoro-Emerald, FE (C) injection sites in animal R413. D, Schematic drawing of the cerebral cortex viewed from dorsal, showing the position and size of the injection sites in A (red), B (blue), and C (green) in relation to the Bregma-related stereotaxic coordinate system (grid, in millimeter units) of Paxinos and Watson (1982), as well as the cytoarchitectonic parcellation of the cerebral cortex (gray lines) of Zilles and Wree (1995). E–G, Corresponding photomicrographs of a transverse section through the pontine nuclei (at level indicated in J) showing the pontine distribution of FR (E), BDA (F), and FE (G) labeled fibers. (H), Overlay of the images in E-G. (I), Computerized plot of the section shown in E–H. Labeled axons are semiquantitatively represented by dots, corresponding to the observed density of labeling. Lines indicate the outer boundaries of the ventral surface of the brain stem, the pontine nuclei, and the descending peduncle. The plexuses of labeled axons are largely segregated, and distributed with an inside-out shift in pontine location. (J) Computerized 3-D reconstruction of the pontine nuclei with labeled axons represented as color coded dots. The outer boundaries of the pontine nuclei are shown as transparent surfaces, and the boundaries of the peduncle as solid surfaces. A bounding box, representing the standardized coordinate system for the pontine nuclei, has been orientated along the long axis of the brain stem and fitted to the external boundaries of the pontine nuclei. The arrow indicates the rostrocaudal level of the section shown in I. (K) Photograph of the rat brain stem in view from ventral, with the pontine nuclei coordinate system superimposed. Scale bars, 500 μm (A–C), and 50 μm (E–I).

Figure 2

Regional organization of corticopontine projections. (A) The location and size of 31 tracer injection sites is mapped onto a diagram of the right cerebral surface. Major brain regions are color coded (blue, frontal; red, parietal; green, temporal; yellow, occipital). Presentation otherwise as in Figure 1. Injection sites indicated in gray did not give rise to pontine projections. (B) The pontine nuclei coordinate system (view from ventral), with solid curved lines indicating the outer boundaries of the pontine nuclei. Dots, representing labeled corticopontine axons from all cases investigated, are color coded in correspondence with A. (C) Selected individual experiments showing representative distributions of corticopontine projections from each region. The distribution of labeling is shown as dot maps and as cluster maps, with solid surfaces representing the external boundaries of densely labeled axonal clusters. The experiments clearly show that cerebral regional topography is maintained in the pontine nuclei. Projections arising from the occipital cortex (animal R412, BDA injection) are distributed laterally in the ipsilateral pontine nuclei. Projections from the temporal cortex (animal R411, FE injection) are shifted internally, whereas projections from parietal cortex (R407, FR injection) are distributed in central and caudal regions, and projections from frontal cortex (R410) are located medially and rostrally. Contralateral projections are sparse, but appear as mirror image of the (medially located) ipsilateral projections. The individual experiments show the characteristic clustered appearance of corticopontine projections. Restricted cortical tracer injections typically give rise to two to three distinct spherical or elongated axonal clusters. Bars, 1 mm.

Figure 3

3-D distribution of pontine clusters. Computer-generated stereo pairs showing the 3-D shape and relationship of pontine nuclei projections arising from occipital (yellow), temporal (green), parietal (red), and frontal (red) cortical regions (same cases as shown in Figure 2) in standard angles of view. The viewer must cross the eye axis to let the pair of images merge into a 3-D image. The cluster maps clearly show that the topographically organized corticopontine clusters are largely segregated but have close neighboring relationships at multiple locations.

Figure 4

Organization of corticopontine projections according to cortical gradients. Selected combinations of cerebral injection sites reveal concentric topographical organization in the pontine nuclei. (A-D) Cortical neurogenesis (A, reproduced from Smart, , with permission) and regionalization follows graded gene expression patterns (B, C, reproduced from O'Leary and Nakagawa, , with permission) that shift from anterolateral to posteriomedial. D, Interpretation of cerebal maturation gradients mapped onto the standard dorsal view drawing of the cerebral cortex. (E–G) Three different combinations of injection sites revealing a concentric, inside-out arrangement of terminal fields in the pontine nuclei. Presentation as in Figure 2. The selected three injection sites are located along the cerebral neurogenetic gradient, colored from red to black to blue, from the anterolateral cortex toward occipital (A, animal R113, red; D46/BDA, black; D46/FR, blue), (B, R113, red; R407/FR, black; R121/FR, blue) and frontal (C, R113, red; R406/FE, black; R410, blue), all giving rise to concentrically organized patterns with gradually more external location of the black and blue dots surrounding the centrally located red dots. (H–J), Three other combinations of injection sites, oriented from frontal and medial toward occipital and lateral, located approximately within the same neurogenetic zone (H, R406/FR, red; R406/FE, blue; R406/FB, blue; R406/F-G, green; I, R409, red; R121/BDA, black; R121/FR, blue; R412/FE, green; J, R412/FR, black, R403, red). Together, the different axonal clusters form circular volumes, increasing in diameter as the rows of injection sites are shifted from anterolateral toward dorsomedial, along the cerebral neurogenetic gradient.

Figure 5

Map transformation principles in the rat corticopontine system. Injection sites and corresponding dot populations representing labeled fibers are color coded according to graded patterns in the cerebral cortex. (A, C) When all experimental data are colored following the putative orientation of cortical neurogenesis, from anterolateral (dark gray) toward medial and posterior (white), a pattern of concentric organization is revealed (E), with black dots located in the center, and gradually brighter colored dots distributed more externally. (B, D) When the same data are colored following the cortical frontal to occipital axis (white to red), a general medial to lateral organization is seen in the pontine nuclei (F), with frontally originating dots primarily located in the medial half of the pontine nuclei, and darker colored dots gradually distributed toward lateral. (G, H) Principal map of corresponding locations in the cerebral cortex (viewed from dorsal) and pontine nuclei (viewed from ventral). In G, the gray (A) and red (B) graded cortical patterns are combined. Gridlines are alphabetically named from frontal to occipital, and labeled with roman numbers from anterolateral to dorsomedial. (H) Graphical model of corresponding topographical distributions (inside-out, black to white; medial to lateral, white to red) in the pontine nuclei. The two coordinate grids allow prediction of the overall pontine distribution of axonal plexuses arising from any location in the cerebral cortex.

Figure 6

Correspondence between predicted and observed distribution of corticopontine projections. (A) four injection site positions (same data as shown in Figures 3 and 4) are indicated on a dorsal view of the cerebral cortex together with a positional grid of the cerebral surface (explained in Figure 5). Grid fields containing the injection site center are correspondingly colored. (B) Predicted distributions of projections from the four injection sites in (A) shown in our principal graphical representation (explained in Figure 5). Prediction of overall pontine distribution is made by coloring pontine grid fields corresponding to the cortical grid fields in (A). (C) Overlay of predicted (transparent colored fields) and observed (colored dots) pontine distributions shows a high degree of correspondence.