Ultrastructure of primary afferent terminals and synapses in the rat nucleus of the solitary tract: comparison among the greater superficial petrosal, chorda tympani, and glossopharyngeal nerves

The greater superficial petrosal (GSP), chorda tympani (CT), and glossopharyngeal (IX) nerves terminate in overlapping patterns in the brainstem in the rat nucleus of the solitary tract (NTS). There is one region, in particular, that receives overlapping inputs from all three nerves and is especially plastic during normal and experimentally altered development. To provide the requisite data necessary ultimately to delineate the circuitry in this region, we characterized the morphology of the synaptic inputs provided by the GSP, CT, and IX nerves through transmission electron microscopy. Although all three nerves had features characteristic of excitatory nerve terminals, ultrastructural analysis revealed dimorphic morphologies differentiating IX terminals from GSP and CT terminals. IX terminals had a larger area than GSP and CT terminals, and more synapses were associated with IX terminals compared with GSP and CT terminals. Additionally, IX terminals formed synapses most often with spines, as opposed to GSP and CT terminals, which formed synapses more often with dendrites. IX terminals also exhibited morphological features often associated with synaptic plasticity more often than was seen for GSP and CT terminals. These normative data form the basis for future studies of developmentally and environmentally induced plasticity in the rodent brainstem.

Fine structural survey of the intermediate subnucleus of the nucleus tractus solitarii and its glossopharyngeal afferent terminals

The intermediate subnucleus of the nucleus tractus solitarii (imNTS) receives somatosensory inputs from the soft palate and pharynx, and projects onto the nucleus ambiguus, thus serving as a relay nucleus for swallowing. The ultrastructure and synaptology of the rat imNTS, and its glossopharyngeal afferent terminals, have been examined with cholera toxin-conjugated horseradish peroxidase (CT-HRP) as an anterograde tracer. The imNTS contained oval or ellipsoid-shaped, small to medium-sized neurons (18.2 x 11.4 microm) with little cytoplasm, few cell organelles and an irregularly shaped nucleus. The cytoplasm often contained one or two nucleolus-like stigmoid bodies. The average number of axosomatic terminals was 1.8 per profile. About 83% of them contained round vesicles and formed asymmetric synaptic contacts (Gray's type I), while about 17% contained pleomorphic vesicles and formed symmetric synaptic contacts (Gray's type II). The neuropil contained small or large axodendritic terminals, and about 92% of them were Gray's type I. When CT-HRP was injected into the nodose ganglion, many labeled terminals were found in the imNTS. All anterogradely labeled terminals contacted dendrites but not somata. The labeled terminals were usually large (2.69+/-0.09 mum) and exclusively of Gray's type I. They often contacted more than two dendrites, were covered with glial processes, and formed synaptic glomeruli. A small unlabeled terminal occasionally made an asymmetric synaptic contact with a large labeled terminal. The large glossopharyngeal afferent terminals and the neurons containing stigmoid bodies characterized the imNTS neurons that received pharyngeal afferents.

Electron microscopic observations of glossal circumvallate papillae in dysgeusic patients

A number of reports have been published describing the ultrastructure of normal taste buds but few have discussed their pathology, particularly in patients with gustatory abnormalities. We therefore conducted an electron microscopic study of biopsy specimens of glossal circumvallate papillae from patients with dysgeusia. We found depletion of microvilli and dense substance in the taste pores, vacuolation and hyaline degeneration of the taste bud cells and condensation of nuclear material and cytoplasm in cells of indeterminate histological type. In one patient with idiopathic dysgeusia, disruption of subsurface cisternae contiguous with the nerve endings was evident in Type II cells. In another patient, who had undergone radiotherapy, the number of cored vesicles in Type III cells was reduced.

Synaptology of the direct projections from the nucleus of the solitary tract to pharyngeal motoneurons in the nucleus ambiguus of the rat

During the pharyngeal phase of the swallowing reflex, the nucleus of the solitary tract (NTS) receives peripheral inputs from the pharynx by means of the glossopharyngeal ganglion and is the location of premotor neurons for the pharyngeal (PH) motoneurons. The semicompact formation of the nucleus ambiguus (AmS) is composed of small and medium-sized neurons that do not project to the pharynx, and large PH motoneurons. We investigated whether the neurons in the NTS projected directly to the PH motoneurons or to the other kinds of neurons in the AmS by using the electron microscope. When wheat germ agglutinin-conjugated horseradish peroxidase (WGA-HRP) was injected into the NTS after cholera toxin subunit B-conjugated HRP (CT-HRP) injections into the pharyngeal muscles of male Sprague-Dawley rats, many nerve terminals anterogradely labeled with WGA-HRP were found to contact PH motoneurons retrogradely labeled with CT-HRP. Most of the labeled axodendritic terminals (63%) contained pleomorphic vesicles with symmetric synaptic contacts (Gray's type II), and the remaining ones contained round vesicles with asymmetric synaptic contacts (Gray's type I). About 14% of the axosomatic terminals on PH motoneuron in a sectional plane were anterogradely labeled, and about 70% of the labeled axosomatic terminals were Gray's type II. Observations of serial ultrathin sections revealed that both the small and the medium-sized neurons received only a few labeled axosomatic terminals that were exclusively Gray's type I. These results indicate that the NTS neurons may send mainly inhibitory as well as a few excitatory inputs directly to the PH motoneurons in the AmS.

Unilateral innervation of guinea pig vallate taste buds as determined by glossopharyngeal neurectomy and HRP neural tracing

The innervation pattern by primary afferent nerve fibres and the neurotrophic effect on taste cells were investigated in the guinea pig vallate taste bud by means of glossopharyngeal neurectomy and horseradish peroxidase (HRP) or wheat germ agglutinin-horseradish peroxidase (WGA-HRP) tracing. In the glossopharyngeal neurectomy study, taste buds in the vallate papillae of adult guinea pigs were denervated by unilateral resection of the right glossopharyngeal nerve. Denervated animals were killed on days 1, 3 and 5 and weeks 1-9, 12 and 24 postneurectomy. The results showed that, on the denervated side, the taste buds decreased significantly in number during the 1st 2 wk, and disappeared completely by wk 3; no mature taste buds were present even 24 wk after neurectomy. This suggests that the vallate taste buds disappear in the absence of the glossopharyngeal nerve. In the neural tracing study, HRP or WGA-HRP was injected into the proximal end of the right glossopharyngeal nerve, near the jugular foramen. After a survival time of 24 h, the vallate papillae were sectioned and examined by light and electron microscopy. Light microscopy revealed that the HRP or WGA-HRP-labelled fibres innervated the vallate taste buds of the injected side. Most of the taste cells in the buds were labelled with HRP or WGA-HRP reaction products from the basal to the apical region. At the ultrastructural level, the reaction products were confined to the cytoplasm of the labelled cells, which were identified as type I, II and III cells, but not basal cells. Labelled intragemmal nerve profiles were seen among the taste cells. No synapse formation was seen with nerve profiles abutting on type I and II cells, whereas on certain type III cells, typical synapses were formed. We conclude that both the right and left vallate papilla in the guinea pig are unilaterally innervated by the glossopharyngeal nerve without cross-innervation.

The sites of origin and termination of afferent and efferent components in the lingual and pharyngeal branches of the glossopharyngeal nerve in the Japanese monkey (Macaca fuscata)

Afferent and efferent components in the lingual and pharyngeal branches of the glossopharyngeal nerve (Li and Ph) of the Japanese monkey (Macaca fuscata) were examined. After injecting wheat germ agglutinin-conjugated horseradish peroxidase (WGA-HRP) unilaterally into the central cut end of the Li and Ph, or into the stylopharyngeal muscle, labeled neuronal cell bodies and terminal labeling were observed in the medulla oblongata, peripheral ganglia of the glossopharyngeal and vagus nerves, and cervical ganglia of the sympathetic trunk. The following conclusions were deduced from the results. The Li contains efferent fibers from the inferior salivatory nucleus, and superior cervical ganglion. The afferent fibers in the Li are composed mainly of peripheral processes of ganglion neurons in the superior and petrous ganglia of the glossopharyngeal nerve, and additionally of those of ganglion neurons in the jugular ganglion of the vagus nerve. The afferent fibers in the Li terminate mainly in the lateral division of the nucleus of the solitary tract, and additionally in the dorsal aspect of the lateral marginal region of the interpolar spinal trigeminal nucleus. The Ph is mainly composed of efferent fibers from the ambiguous nucleus and superior cervical ganglion; only a small number of afferent fibers from the sensory ganglia of the glossopharyngeal and vagus nerves are contained in the Ph. Stylopharyngeal motoneurons are distributed in the retrofacial part of the ambiguous nucleus.

Distribution and synaptology of glossopharyngeal afferent nerve terminals in the nucleus of the solitary tract of the hamster

The distribution and synaptology of the afferent fibers of the glossopharyngeal nerve (IXN) in the hamster were studied by using horseradish peroxidase (HRP) histochemistry visualized with light and electron microscopy. Crystals of HRP were applied to the trunk of IXN in the vicinity of the petrosal ganglion. The densest IXN afferent label was distributed within the nucleus of the solitary tract (nst), just caudal to but overlapping with the area of termination of the facial nerve. Labeled IXN fibers extended rostrally to the principal trigeminal nucleus and caudally to the cervical spinal cord. There was significant labeling within the spinal trigeminal complex; the area postrema and the medullary reticular formation contained some labeled fibers. Ultrastructurally, the synaptic arrangements of anterogradely labeled IXN fibers were examined in the nst. Quantitative measures were taken of the area, maximum diameter, perimeter, and vesicles of labeled endings and the length of their synaptic junctions with dendritic processes. These endings were compared to comparable endings in control material and to published descriptions of VIIth nerve afferent terminals in the hamster nst. The synaptic relations of IXN afferent endings were predominantly with dendritic spines and shafts. The majority (86.6%) of IXN afferent endings were with dendritic processes that were not in apparent contact with other, unlabeled processes. Only 13.4% of IXN synaptic relationships were with dendritic processes that were also contacted by unlabeled vesicle-containing processes. This is in contrast to 31.2% of facial nerve afferent endings in the nst which make synaptic contact with such processes. There were more direct synaptic contacts between facial endings and unlabeled vesicle-containing processes (26.1%) than between IXN endings and unlabeled vesicle-containing processes (1.3%). Thus, unlike the glomerular-like endings of the gustatory fibers of the VIIth nerve, less complex relations appeared to characterize IXN synapses in the nst. These differences were related to the differential physiology of gustatory fibers in the VIIth nerve and IXN.

Ultrastructure of calcitonin gene-related peptide-immunoreactive, unmyelinated afferents to the cat carotid body: a case of volume transmission

To relate the ultrastructure of unmyelinated afferents to the cat carotid body with the known electrophysiological properties of cat chemosensory C-fibers, we took advantage of the fact that the calcitonin gene-related peptide is exclusively present in a population of sparsely branched afferents to the carotid body. They have a morphology identical to the afferents originating from carotid sinus nerve unmyelinated axons. Immunoreactive axons were stained using pre-embedding protocols and horseradish peroxidase-labeled secondary antibody. Labeling was present only in unmyelinated axons and boutons distributed in the interstitial and parenchymal tissue. The varicosities had an average diameter of 0.7 micron, and contained both small, clear vesicles and larger dense-core vesicles. No labeled axons were ever seen to contact glomus cells, but could be observed as close as 0.2 micron to a glomus cell, always with an interposed glial process. With a very sensitive protocol, that used tungstate-stabilized tetramethylbenzidine as the chromogen, amorphous deposits of reaction product were often detected in the extracellular space around a labeled bouton. We interpret these findings as indicating that the reciprocal chemical transmission between the oxygen-sensitive glomus cells and the unmyelinated afferents takes place through non-synaptic transmission, via the rather large extracellular space of the carotid body. In addition, the larger distances between glomus cells and unmyelinated afferents could explain the lowered sensitivity and sluggishness of chemosensory C-fibers, compared to the A-fibers.

Convergence of afferents from the SLN and GPN in cat medullary swallowing neurons

We demonstrated the convergence of information from the pharyngeal and laryngeal mucosa, transmitted by the glossopharyngeal nerve (GPN) and superior laryngeal nerve (SLN), in the nucleus of the tractus solitarius (NTS). First, the distribution of terminals of the GPN and SLN in the NTS was examined by an HPR tracing technique in cats, and the synapse formation of these neurons with NTS neurons was demonstrated by electron microscopy. The HRP-labeled SLN and GPN terminals were localized in a small area of the interstitial subnucleus of the NTS, slightly rostral to the obex, forming synapses with NTS neurons. Next, using extracellular recording in anesthetized cats, we determined whether or not swallowing-related neurons in the medulla oblongata receive peripheral inputs. Convergence of peripheral sensory inputs from the SLN and GPN was observed in more than 80% of the NTS cells. These results suggest that the NTS is not only a sensory-relay nucleus but also integrates information necessary for eliciting protective reflexes of the upper airway, such as swallowing.

Cellular relations in mouse circumvallate taste buds

The fine structure of the taste buds of circumvallate papillae of two strains of mice was studied by electron microscopy. Mice anesthetized with ketamine were perfused through the heart with a double aldehyde mixture in cacodylate buffer and the tissues embedded in Epon. Semi-serial sections were employed. The morphology and relationships of cell types are consistent with the majority of descriptions of mammalian taste buds served by the ninth cranial nerve. Cells of type II are particularly well documented, as the stages in their origin, maturation and degeneration could be followed. Significant differences, however, relate to cell type I. These cells contain large dense-cored granules, contrasted with the more irregular and somewhat larger dark granules of the type I cells in the rabbit. These granules do not produce a dense homogenous product for the pore, as seen in the rabbit. Rather the pore substance consists of small, empty vesicles in a diffuse dark matrix. These granules are only moderately larger than the dense-cored vesicles of the type III cells. All features of the type III cell were demonstrated, although no instance of a complete cell was seen in any section. No significant differences were noted between the two strains of mice. Intimate proximity of a nerve to a cell nucleolus, suggestive of a trophic pathway, is illustrated.

Synaptic connections of central carotid sinus afferents in the nucleus of the tractus solitarius of the rat. I. An electron microscopic study

A transganglionic transport technique was used to study the synaptic connections of the central carotid sinus afferents in the nucleus of the tractus solitarius of the rat by electron microscopy. The caudal part of the nucleus was profusely innervated. Labelled fibres extended to the contralateral nucleus, and to the ipsilateral dorsal motor nucleus of the vagus nerve, nucleus ambiguus, spinal nucleus of the trigeminal nerve and the area postrema. The labelled terminals were densely packed with clear, predominantly spherical vesicles about 50 nm in diameter and a few often swollen mitochondria. The terminals synapsed on dendrites of various calibres, spindle- or pear-shaped somal profiles with short axes lesser than 8 microns, and axon terminals. In axo-axonal synapses, most labelled terminals appeared to be presynaptic. Frequently, profiles of labelled terminals were in direct apposition with one another. The latter may represent the morphological substrate of the interaction between baro- and chemoreceptor inputs in the nucleus of the tractus solitarius and warrants further study. The present results indicate that in addition to direct inputs, the carotid sinus afferents are able to influence second-order neurons in the nucleus of the tractus solitarius indirectly through presynaptic modulation.

[Early development and cell differentiation of the parasympathetic vagus-glossopharyngeal nucleus in cattle. Light and electron microscopic studies]

The early development, differentiation of the cell and cell migration of the nucleus parasympathicus nervi vagi et glossopharyngei were examined by light microscope in 32 bovine embryos with crown-rump-lengths (CRL) ranging from 1 cm to 53 cm. During this period the nucleus is being enlarged 6 to 7 times and the size of the cell increases to 35-40 microns. The ultrastructure during the differentiation of the cell is shown electron microscopically in embryos with CRL of 2.5 cm and 3.6 cm. Several layers of matrix cells arise from the neuroepithelium of the neural tube by mitosis. They migrate in the shape of dark nucleated cells into the parasympathetic cell column. With advancing age of the embryos the number of cells with light nucleus increases. They represent the presumptive neurons. In embryos of 2.5 cm CRL their cytoplasm surrounds the nucleus on three sides in the shape of a narrow rim while on the fourth side it is enlarged into an outgrowing process. In this process a smaller number of organelles and their preliminary stages appears. Their number is significantly increased in embryos of 3.6 cm CRL and they can be seen throughout the growing process. In the following stages of maturity cytological development proceeds. In embryos of 53 cm CRL topographical and cytological data are comparable to those in adult animals.

The glossopharyngeal nerve of the axolotl labeled with carbocyanine dye (diI)

Fluorescent carbocyanine dye (diI) was used to label the glossopharyngeal (IX) nerve in the fixed preparation of the Mexican salamander, axolotl. When the cell bodies were viewed with a confocal laser scanning microscope and Nomarski optics, the cytoplasm was brightly fluorescent, but not the cell nucleus. The cell bodies which send peripheral axons in the two branches of the IX nerve were mainly distributed in the rostral part of the combined glossopharyngeal-vagus ganglion, but a few cells were also distributed in the middle and caudal parts. This may indicate a relatively undifferentiated organization of the IX nerve in the ganglion.

The onset and rate of myelination in six peripheral and autonomic nerves of the rat

A light and electron microscopic study was carried out of the numbers of myelinated fibres in 6 nerves of the rat for 7 age groups from birth to 73 weeks. The hypoglossal nerve and the mandibular branch of the facial nerve had short and early myelination periods, essentially complete by the second week. The glossopharyngeal nerve and the sympathetic rami communicantes myelinated late and over a protracted period. Myelination of the rami communicantes continued up to 20 weeks, followed by a marked loss of fibres in the 73 week animals. Intercostal and saphenous nerves had intermediary patterns. There was evidence of subpopulations myelinating at different times. Measurements of myelin sheath thickness showed variations of relative sheath thickness with age, between nerves and for subpopulations of nerves. Late myelination corresponded to relatively thin myelin sheaths. Statistical two-stage-density cluster analysis by computer was used for analysing complex fibre populations. The developmental changes of three subpopulations of the intercostal nerve are documented. Nerves also differed in their rates of axon growth. The increment in axon calibre was small and late for sympathetic fibres. Intercostal and facial nerve fibres had rapid axon growth with different growth rates for subpopulations.

Direct contacts between glossopharyngeal afferent terminals and hypoglossal motoneurons revealed by double labeling with cobaltic-lysine and horseradish peroxidase in the Japanese toad

Glossopharyngeal (IX) afferents and hypoglossal (XII) motoneurons of the Japanese toad were simultaneously labeled with cobaltic-lysine and horseradish peroxidase, respectively. Some of the terminal branches of the IX afferents had direct contacts with the dorsal dendrites, the lateral dendrites and the somata of the XII motoneurons, but not with the medial dendrites. Such direct contacts mainly occurred in the rostral region of the dorsomedial XII nucleus, where tongue-retractor motoneurons predominate, but not in the caudal region nor in the ventrolateral XII nucleus.

Ultrastructural studies of the nucleus ambiguus in cat and monkey following injection of HRP into the vagus nerve

The fine structure of retrogradely labelled neurons in the nucleus ambiguus (NA) has been examined in cat and monkey (Macaca fascicularis) following injections of horseradish peroxidase (HRP) into the vagus nerve. Many retrogradely labelled neurons were observed in the NA although unlabelled neurons were also present within the boundaries of the NA as identified by the distribution of retrogradely labelled cells. In both species a wide range of sizes of labelled neurons (20-60 microns in long axis) was observed from rostral to caudal levels of the NA. Large labelled neurons were generally oval or spindle-shaped while smaller neurons were oval in cross-section. Unlabelled neurons observed among labelled NA neurons tended to be smaller on average than the labelled neurons and ranged in size from 15 to 30 microns in long axis. The unlabelled neurons typically had nuclei which were more invaginated than those of the labelled neurons. Quantitative analyses of the synaptic organization in the NA revealed high proportions of terminals containing round vesicles and making asymmetrical or symmetrical contact with the somata in both monkey and cat. Terminals containing pleomorphic vesicles and making symmetrical contact with somata were slightly less numerous. Retrogradely labelled neurons exhibited a positive correlation between the size of neuron and density of synapses on the surface. There tended to be a greater synaptic density on monkey NA neurons than on cat NA neurons of comparable size.

Autoradiographic localization of opiate receptors in rat brain. I. Spinal cord and lower medulla

The localization of opiate receptors in the spinal cord and lower medulla has been elucidated by the autoradiographic identification of stereospecific [3H]diprenorphine (a potent opiate antagonist) binding sites. The opiate receptors were higly localized to: layers I (marginal cell zones) and II (substantia gelatinosa) of the dorsal horn of the spinal cord; the substantia gelationsa of the spinal trigeminal nucleus; components of the vagal system, including the vagus nerve, nucleus tractus solitarius, nucleus commissuralis, nucleus intercalatus, nucleus ambiguus and nucleus originis dorsalis vagus; the area postrema. Examination of [3H]etorphine (a potent opiate agonist) binding sites showed the same distribution. We conclude that, in these brain regions, opiate receptors are (1) highly associated with areas receiving small, afferent primary fibers, (2) strategically placed to modulate noxious stimuli as well as explain some visceral side effects of opiate administration.

An ultrastructural study of cat petrosal ganglia: a search for autonomic ganglion cells

It is not known whether cranial and spinal ganglia contain autonomic (motor) neurons in addition to sensory cells. Early light microscopic studies indicate the possibility of synaptic input to some nerve cells via perisomatic axon terminals and dendrites. It has also been suggested that visceromotor relay-type cells in the petrosal ganglion might be involved in the efferent control of the carotid body. The present ultrastructural analysis of 902 neurons in 17 levels from two cat petrosal ganglia provides no evidence for synapses on the cell somata or on their axonal extensions. Processes resembling dendrites were lacking although small surface projections typical of sensory cells were common. The petrosal ganglion thus appears to consist entirely of afferent unipolar neurons.