The aim of this study was to map the viscerotopic representation of the upper alimentary tract in the sensory ganglia of the IXth and Xth cranial nerves and in the subnuclei of the solitary and spinal trigeminal tracts. Therefore, in 172 rats 0.5-65 microliters of horseradish peroxidase (HRP), wheat germ agglutinin-HRP, or cholera toxin-HRP were injected into the trunks and major branches of the IXth and Xth cranial nerves as well as into the musculature and mucosa of different levels of the upper alimentary and respiratory tracts. The results demonstrate that the sensory ganglia of the IXth and Xth nerves form a fused ganglionic mass with continuous bridges of cells connecting the proximal and distal portions of the ganglionic complex. Ganglionic perikarya were labeled in crude, overlapping topographical patterns after injections of tracers into nerves and different parts of the upper alimentary tract. After injections into the soft palate, pharynx, esophagus, and stomach, anterograde labeling was differentially distributed in distinct subnuclei in the nucleus of the tractus solitarius (NTS). Palatal and pharyngeal injections resulted primarily in labeling of the interstitial and intermediate subnuclei of the NTS and in the paratrigeminal islands (PTI) and spinal trigeminal complex. Esophageal and stomach wall injections resulted in labeling primarily of the subnucleus centralis and subnucleus gelatinosus, respectively. The distribution of upper alimentary tract vagal-glossopharyngeal afferents in the medulla oblongata has two primary groups of components, i.e., a viscerotopic distribution in the NTS involved in ingestive and respiratory reflexes and a distribution coextensive with fluoride-resistant acid-phosphatase-positive regions of the PTI and spinal trigeminal nucleus presumably involved in visceral reflexes mediated by nociceptive or chemosensitive C fibers.

The motor innervation for palatal, pharyngeal, laryngeal, and esophageal muscles originates within the nucleus ambiguus. Although the viscerotopic organization of the upper alimentary tract in the nucleus ambiguus has been extensively studied, little information concerning the dendritic arborization of nucleus ambiguus motoneurons is available. The neural tracer cholera toxin-horseradish peroxidase, which is particularly effective at revealing dendrites of retrogradely labeled neurons, was used to determine the dendritic architecture and organization of nucleus ambiguus motoneurons. In 72 rats, cholera toxin-horseradish peroxidase in volumes of 1.0-18 microliters was directly applied under pressure to the musculature of various sites along the upper alimentary tract. Motoneurons innervating the soft palate, pharynx, cricothyroid muscle, and cervical esophagus were all found to have extensive dendrites that extended into the adjacent reticular formation with a distinct pattern for each muscle group. In contrast, the dendrites of motoneurons innervating the thoracic and subdiaphragmatic esophagus were confined to the compact formation of the nucleus ambiguus. Dendritic bundling within the confines of the nucleus ambiguus was prominent following injection of tracer into the soft palate, pharynx, and esophagus. The bundles were primarily oriented in a rostrocaudal direction. These data suggest that the extensive extranuclear dendritic arborization of motoneurons innervating the soft palate, pharynx, larynx, and cervical esophagus provide a wide ranging target for multiple central afferents that may be involved in the differential control of muscles that participate in multiple complex motor functions. The lack of extensive extranuclear dendrites of motoneurons innervating the distal esophagus suggest that they receive focused central afferents. The prominent bundling of dendrites within the nucleus ambiguus may provide for synchronization of motoneurons innervating a specific muscle and perhaps for synchronization of motoneurons innervating different muscles acting in sequence.