We studied the distribution of corticopontine cells in the monkey cerebral cortex. Horseradish peroxidase (HRP) was injected into the brainstem of monkeys in an attempt to fill the pontine nuclei on one or both sides. In control animals we injected the medullary pyramids or varied the route, size, or location of pontine injections. All retrograde filled corticopontine neurons were layer V pyramidal cells. Corticopontine cells were distributed within a largely continuous area of cortex which extended from the cingulate cortex medially to the sylvian fissure laterally; from the superior temporal fissure caudally to the medial part of the frontal granular cortex rostrally. Areas 4 and 6 of Brodmann (1905) contained the highest density of filled cells. In the primary visual cortex, area 17, there were a few labelled cells restricted to the rostral portion of the upper bank of the calcarine fissure, in a region representing the lower periphery of the visual field. The results are discussed in relation to the possible functions of the corticopontine system, especially the role of the extrastriate visual areas in visually guided movement.

The cerebellum plays an important role in the visual guidance of movement. In order to understand the anatomical basis of visuomotor control, we studied the projection of pontine visual cells onto the cerebellar cortex of monkeys. Wheat germ agglutinin horseradish peroxidase was injected into the dorsolateral pons two monkeys. Retrogradely labelled cells were mapped in the cerebral cortex and superior colliculus, and orthogradely labelled fibers in the cerebellar cortex. The largest number of retrogradely labelled cells in the cerebral cortex was in a group of medial extrastriate visual areas. The major cerebellar target of these dorsolateral pontine cells is the dorsal paraflocculus. There is a weaker projection to the uvula, paramedian lobe, and Crus II, and a sparse but definite projection to the ventral paraflocculus. There are virtually no projections to the flocculus. There are sparse ipsilateral pontocerebellar projections to these same regions of cerebellar cortex. In nine monkeys, we made small injections of the tracer into the cerebellar cortex and studied the location of retrogradely filled cells in the pontine nuclei and inferior olive. Injections into the dorsal paraflocculus or rostral folia of the uvula retrogradely labelled large numbers of cells in the dorsolateral region of the contralateral pontine nuclei. Labelled cells were found ipsilaterally, but in reduced numbers. Injections outside of these areas in ventral paraflocculus or paramedian lobule labelled far fewer cells in this region of the pons. We conclude that the principal source of cerebral cortical visual information arises from a medial group of extrastriate visual areas and is relayed through cells in the dorsolateral pontine nuclei. The principal target of pontine visual cells is the dorsal paraflocculus.

The distribution of cortical cells projecting to the pontine nuclei in rats was studied by making large injections of horseradish peroxidase that filled the basilar pons and measuring the density of labelled cells in each cortical area. All retrogradely labelled cells were layer V pyramidal cells. The highest densities of labelled cells were observed in the motor areas. The lowest densities were in temporal association cortex and perirhinal cortex. Visual cortical areas, including the primary visual cortex, provided a major source of pontine projections. The distribution of corticopontine cells within the primary visual cortex was studied in more detail. In all cases the highest density of labelled cells was observed in the region of cortex that represents the nasal visual field. Control injections into brainstem regions adjacent to the pontine nuclei produced a much lower absolute density of retrogradely labelled cortical cells and the distribution of those cells was different from that observed following pontine injections. We conclude that every area of the rat's cerebral cortex projects to the pontine nuclei and that there are consistent variations in the density of the projections both between and within areas.

There are two great subcortical circuits that relay sensory information to motor structures in the mammalian brain. One pathway relays via the pontine nuclei and cerebellum, and the other relays by way of the basal ganglia.We studied the cells of origin of these two major pathways from the posteromedial barrel subfield ofrats, a distinct region ofthe somatosensory cortex that contains the sensory representation of the large whiskers. We itjected tracer substances into the caudate putamen or the pontine nuclei and charted the location of retrogradely filled cortical cells. In preliminary studies, we used double-labeling techniques to determine whether the cells of origin of these two pathways send axon collaterals to other subcortical targets. Lamina V of the rat posteromedial barrel subfield contains two distinct populations of subcortically projecting neurons, which are organized into distinct sublaminae. Corticopontine cells are located exclusively in sublamina Vb, the deeper of two sublaminae revealed by cytochrome oxidase staining. Corticostriate cells are located almost exclusively in the more superficial sublamina Va. Experiments using double-labeling fluorescent tracers demonstrate that about one-quarter of the corticopontine cells send a collateral branch to the superior colliculus. Other studies have shown that cells in Vb are activated at very short latency after vibrissal stimulation; hence, they would seem to be an appropriate relay for the rapid transmission of sensory information to the cerebellum for use in sensory guidance of movement.Sensory areas of the cerebral cortex are connected to motor areas by way of corticocortical circuits and two prominent subcortically directed pathways. One subcortical pathway goes to the basal ganglia, and the other goes to the cerebellum via the pontine nuclei (1-4). The efferent terminals from the basal ganglia and cerebellum remain largely distinct in the thalamus and their thalamic targets project to nonoverlapping regions of the motor cortex (5). Do the basal ganglia and the cerebellum receive inputs from the same or a different population of sensory cortical cells?The great majority of the cells in the sensory cortex that project to these two motor systems are located in lamina V of the cerebral cortex (2, 6-8). Lamina V may be divided into two sublaminae: a superficial sublamina Va and a deep sublamina Vb, which differ in a number of important respects. In mice, lamina Vb receives direct input from the somatosensory relay nucleus in the thalamus; Va does not (9). Stimulation of the appropriate whisker in rats activates cells in sublamina Vb at a latency that is as short as that for activating cells in lamina IV, the principal target of incoming thalamic fibers (10). In contrast, cells in Va respond at much longer latency to such stimulation. Cells in Va stain palely when tested for the presence of the metabolic enzymes cytochrome oxidase or succinic dehydrogenase (11). Cells in sublamina Vb stain far more densely, suggesting that they are capable ofh...

No abstract