Regeneration of taste buds after reinnervation of a denervated tongue papilla by a normally nongustatory nerve

Taste buds degenerate and disappear after transection of their sensory nerve supply, and they differentiate anew from epithelial cells (e.g., lingual) following regeneration of sensory but not motor or autonomic axons. A controversy exists as to whether only gustatory sensory nerves can cause buds to reform or whether any sensory nerve can perform this function. This issue arose because the results of cross-innervation studies revealed a specificity whereas grafting data demonstrated a nonspecificity. A retest of specificity in the cross-reinnervation situation was performed by reinnervating the denervated vallate papilla of adult rat tongue with a sensory branch of the vagus nerve that is not normally gustatory. It was found that taste buds disappeared and remained lost from acutely and chronically denervated papilla. However, some buds were found 90–100 days after reinnervation by the normally nongustatory vagus nerve branch. Transection of the regenerated vagus nerve resulted in the loss of innervation and the degeneration of taste buds from reinnervated papilla indicating that this nerve had supported buds. These results show that a normally nongustatory nerve can induce the formation of taste buds after its axons grow into appropriate tissue. It appears that the ability to support taste buds is a nonspecific, rather than a specific, property of sensory nerve.     

Ultrastructure of mouse foliate taste buds: synaptic and nonsynaptic interactions between taste cells and nerve fibers

High voltage electron microscopy and conventional transmission electron microscopy were used to examine the ultrastructure of foliate taste buds of mice. Computer-assisted, three-dimensional reconstructions from serial sections were used to visualize regions of interaction between taste cells and nerve fibers. Based on criteria previously established for murine vallate taste buds (Kinnamon et al., '85), foliate taste cells were classified as dark, light, or intermediate depending on their cytoplasmic content and the characteristics of their nuclei. Cells of foliate taste buds display a continuous range of morphologies, from "typical" dark cells to "typical" light cells. Cells of dark, intermediate, and light morphologies all make afferent synapses onto nerve processes, suggesting that cells of all 3 types are sensory in function. Synapses between taste cells and nerve processes may be either macular or fingerlike in shape. No efferent synapses were found. In addition to conventional synapses, taste cells exhibit 2 other types of specializations at sites of apposition with nerve fibers: subsurface cisternae and atypical mitochondria. Subsurface cisternae are narrow sacs of endoplasmic reticulum that are closely apposed to the inner leaflet of the taste cell membrane. Possible functions of subsurface cisternae include synthesis of synaptic membrane components, modification of the electrical or adhesive properties of the taste cell membrane, and exchange of trophic factors with nerve processes. Atypical mitochondria are usually much larger than typical taste cell mitochondria, and their cristae often display a swollen, twisted configuration. These mitochondria are closely apposed to the inside of the taste cell membrane adjacent to nerve fibers. Atypical mitochondria may be providing unusual amounts of energy for metabolic reactions in their vicinities or participating in calcium buffering in the taste cell. Within taste cells, presynaptic specializations, subsurface cisternae, and mitochondria are often clustered together to form "synaptic ensembles." We hypothesize that the functions served by the subsurface cisternae and mitochondria, as well as synaptic transmission, may be important in interactions between taste cells and nerve fibers.     

Synapsin I-like immunoreactivity in nerve fibers associated with lingual taste buds of the rat

Immunoreactivity to synapsin I, a neuronal phosphoprotein, was localized in free-floating tissue sections prepared from lingual tissue of rats. Many nerve fibers within the tissue exhibited clear immunoreactivity including motor endplates on striated muscle, autonomic fibers innervating blood vessels or glands, and sensory fibers innervating muscles or the lingual epithelium including taste buds. Numerous immunoreactive fibers occurred within each taste bud, with fewer, fine fibers being dispersed in the epithelium between taste buds. The majority of the intragemmal immunoreactive fibers extended throughout the taste buds most of the distance outward from the basal lamina toward the surface of the epithelium. Fine, perigemmal fibers reached nearly to the epithelial surface. Ultrastructural analysis of the immunoreactive sensory fibers revealed that synapsin I-immunoreactivity occurred diffusely throughout the cytoplasm, and heavily in association with microvesicles. The synaptic vesicles at the taste receptor cell-to-afferent fiber synapse were, however, not immunoreactive for synapsin I, although these vesicles fall into the size class shown to be immunoreactive in other systems. This absence of synapsin I may be a common property of vesicles in axonless short receptor cells.     

Shrinkage of ipsilateral taste buds and hyperplasia of contralateral taste buds following chorda tympani nerve transection

The morphological changes that occur in the taste buds after denervation are not well understood in rats, especially in the contralateral tongue epithelium. In this study, we investigated the time course of morphological changes in the taste buds following unilateral nerve transection. The role of the trigeminal component of the lingual nerve in maintaining the structural integrity of the taste buds was also examined. Twenty-four Sprague-Dawley rats were randomly divided into three groups: control, unilateral chorda tympani nerve transection and unilateral chorda tympani nerve transection + lingual nerve transection. Rats were allowed up to 42 days of recovery before being euthanized. The taste buds were visualized using a cytokeratin 8 antibody. Taste bud counts, volumes and taste receptor cell numbers were quantified and compared among groups. No significant difference was detected between the chorda tympani nerve transection and chorda tympani nerve transection + lingual nerve transection groups. Taste bud counts, volumes and taste receptor cell numbers on the ipsilateral side all decreased significantly compared with control. On the contralateral side, the number of taste buds remained unchanged over time, but they were larger, and taste receptor cells were more numerous postoperatively. There was no evidence for a role of the trigeminal branch of the lingual nerve in maintaining the structural integrity of the anterior taste buds.     

Regeneration of taste buds in tongue grafts after reinnervation by neurons in transplanted lumbar sensory ganglia

Taste buds disappear after denervation and reappear after nerve regeneration. Sensory neurons are responsible since reinnervation by motor or autonomic fibers of peripheral nerve fail to induce bud regeneration. However, we do not know whether some neurons in all sensory ganglia can support buds or whether gustatory (i.e., taste bud inducing) neurons are localized to specific cranial ganglia. The present study was therefore pefrormed to determine whether neurons in transplanted spinal ganglia could support taste buds similarly to those in transplanted cranial ganglia. Grafts of lumbar or vagal nodose ganglia were combined with grafts of tongue's vallate papillae in the anterior chamber of rats' eyes and the papillae examined for taste buds 35 days later. Neurons were present in all transplanted ganglia, and all papillae reinnervated by them contained regenerated taste buds. Nerve fibers could be traced from the transplanted ganglia to the epithelium of the tongue grafts which bore the regenerated taste buds. Papillae transplanted without ganglia lacked buds. These findings indicate that some neurons in all sensory ganglia can induce taste bud formation. The present results could occur if gustatory neurons are intrinsically present in all sensory ganglia, but an alternative interpretation is that the tongue grafts transformed some neurons into gustatory neurons and, hence, that neuronal plasticity is involved.     

Neural induction of taste buds

Bilateral innervation allows more than 80% of the 610 vallate taste buds to survive removal of one IXth nerve in adult rats. Removal of both IXth nerves in neonatal or adult rats results in the absence of taste buds. In studying development, we found that removing or crushing one IXth nerve in three-day-old neonates profoundly decreased the number of vallate taste buds that subsequently developed. Specifically, after removal of one IXth nerve at 3 days, only 228 taste buds formed, compared with 496 taste buds that one nerve would maintain in adults. Thus, during normal development, the right and left IXth nerves interact synergistically, as at least 150 more taste buds develop than predicted by the sum of the independent action of each IXth nerve. This suggests that vallate taste buds are induced by the IXth nerve. A second example of synergism, representing evidence for the neural induction of taste buds, came from experiments in which we crushed the left IXth nerve 3 days after birth and found that these regenerated IXth nerve axons induced 4 times as many taste buds in the presence of the normal right IXth nerve (118 taste buds) as in its early absence (30 taste buds). We conclude that taste buds are neurally induced and that axons of the IXth nerve interact synergistically in inducing them, rather than competing for targets. We propose that in development innervated progenitor cells form stem cells which lead to taste bud cells.     

Calretinin immunoreactivity in taste buds and afferent fibers of the grey mullet Chelon labrosus

The presence of the calcium-binding protein calretinin in taste buds of a teleost, the thick-lipped grey mullet, was investigated using immunohistochemical techniques. Taste bud sensory cells had calretinin immunoreactivity. The nerve fiber plexus innervating taste buds, the ganglia and the viscerosensory roots projecting to the vagal lobe, also showed calretinin immunoreactivity. These results demonstrate for the first time the occurrence of calretinin in the taste buds and the taste afferent system of a teleost.     

The role of substance P and calcitonin gene-related peptide containing nerve fibers in maintaining fungiform taste buds in the rat after a chronic chorda tympani nerve injury.

Taste buds in the anterior part of the tongue of adult rats were denervated by unilateral resection of the chorda tympani nerve in the middle ear. Three months later one group of animals was perfused and their tongues were processed for demonstration of substance P (SP) and calcitonin gene-related peptide (CGRP) immunoreactivity. Fungiform taste buds found on the denervated side showed increased numbers of intragemmal SP- and CGRP-immunoreactive (IR) fibers compared to the normal side. Compared to the normal side, the number of taste buds appeared to be fewer on the denervated side. Moreover, taste buds on this side seemed to be only partially restored. Another group of animals was given the neurotoxin capsaicin which causes a depletion of SP and CGRP from sensory axons. The animals were perfused 2 or 3 weeks after the capsaicin treatment, and their tongues prepared for SP and CGRP immunohistochemistry or for histological examination of taste buds. Very few SP- and CGRP-IR fibers were present in capsaicin-treated animals. In these animals almost all fungiform taste buds and papillae on the chorda tympani-injured side disappeared. In contrast, normal numbers of taste buds were still present on the contralateral side where the chorda tympani innervation remained intact. It is conceivable that taste buds on the chorda tympani-innervated part of the tongue, deprived of the normal chorda tympani-innervation, can regenerate and become reinnervated by SP- and CGRP-containing fibers, and that these are essential for partially restoring and maintaining the structure of the denervated taste buds and the fungiform papillae.     

Immunoelectron-microscopic study on the fine structure of substance-P-containing fibers in the taste buds of the rat

The fine structure of substance-P-like immunoreactive [SPI] fibers in the taste buds of the circumvallate papillae of the rat tongue was investigated by means of electron microscopy using the unlabeled antibody-enzyme method. Outside the epithelium, SPI and non-SPI fibers are surrounded by the cytoplasm of Schwann cells. When the SPI fibers enter the epithelium, they immediately lose this cytoplasmic sheath and begin to traverse the taste buds. Though passing through the taste buds, no profiles suggesting clear synaptic contact between SPI fibers and underlying cells are identified. SPI terminals are filled with small synaptic vesicles and contain a few mitochondria. No SPI-positive structures are found in nerve endings that make synaptic contact with type III cells, the gustatory receptor cells.     

Claudin-based permeability barriers in taste buds

Tight junctions operate as semipermeable barriers in epithelial tissue, separating the apical from the basolateral sides of the cells. Membrane proteins of the claudin family represent the major tight junction constituents, and some reinforce permeability barriers, whereas others create pores based on solute size and ion selectivity. To outline paracellular permeability pathways in gustatory tissue, all claudins expressed in mouse taste buds and in human fungiform papillae have been characterized. Twelve claudins are expressed in murine taste-papillae-enriched tissue, and five of those are expressed in human fungiform papillae. A subset of the claudins expressed in mouse papillae is uniquely found in taste buds. By immunohistochemistry, claudin 4 has been found in mouse taste epithelium, with high abundance around the taste pore. Claudin 6 is explicitly detected inside the pore, claudin 7 was found at the basolateral side of taste cells, and claudin 8 was found around the pore. With the ion permeability features of the different claudins, a highly specific permeability pattern for paracellular diffusion is apparent, which indicates a peripheral mechanism for taste coding.     

Subtype-dependent postnatal development of taste receptor cells in mouse fungiform taste buds

Taste buds contain two types of taste receptor cells, inositol 1,4,5-triphosphate receptor type 3-immunoreactive cells (type II cells) and synaptosomal-associating protein-25-immunoreactive cells (type III cells). We investigated their postnatal development in mouse fungiform taste buds immunohistochemically and electrophysiologically. The cell density, i.e. the number of cells per taste bud divided by the maximal area of the horizontal cross-section of the taste bud, of type II cells increased by postnatal day (PD)49, where as that of type III cells was unchanged throughout the postnatal observation period and was equal to that of the adult cells at PD1. The immunoreactivity of taste bud cell subtypes was the same as that of their respective subtypes in adult mice throughout the postnatal observation period. Almost all type II cells were immunoreactive to gustducin at PD1, and then the ratio of gustducin-immunoreactive type II cells to all type II cells decreased to a saturation level, ～60% of all type II cells, by PD15. Type II and III cells generated voltage-gated currents similar to their respective adult cells even at PD3. These results show that infant taste receptor cells are as excitable as those of adults and propagate in a subtype-dependent manner. The relationship between the ratio of each taste receptor cell subtype to all cells and taste nerve responses are discussed.     

Persistence of taste buds in denervated fungiform papillae

Taste buds in hamster fungiform papillae persist in an atrophic state for as long as 330 days after chorda tympani denervation or 50 days after combined chorda tympani-lingual nerve resection. Although taste bud structure depends on innervation, there is no absolute neural requirement for taste bud survival.     

Cellular expression of alpha-gustducin and the A blood group antigen in rat fungiform taste buds cross-reinnervated by the IXth nerve.

Although taste buds are trophically dependent on their innervation, cross-reinnervation experiments have shown that their gustatory sensitivities are determined by the local epithelium. Both the gustatory G-protein, alpha-gustducin, and the cell-surface carbohydrate, the A blood group antigen, are expressed by significantly fewer fungiform than vallate taste cells in the rat. In these experiments, one side of the anterior portion of the tongue was cross-reinnervated by the IXth nerve in order to determine whether the molecular expression of taste bud cells is determined by the epithelium from which they arise or by the nerve on which they are trophically dependent. The proximal portion of the IXth nerve was anastomosed to the distal portion of the chorda tympani (CT) nerve using fibrin glue (IX-CT rats). Control animals had the CT cut and reanastomosed using the same technique (CT-CT rats), or had the CT avulsed from the bulla and resected to prevent regeneration (CTX rats). The animals survived for 12 weeks postoperatively, and the tongues were removed, stained with methylene blue, and the fungiform taste pores counted on both sides. Tissue from the anterior 5 mm of the tongue was cut into 50-microm sections, which were incubated with antibodies against alpha-gustducin and the human blood group A antigen. In both CT-CT and IX-CT rats, there was regeneration of fungiform taste buds, although in both groups there were significantly fewer taste buds on the operated side of the tongue. The normal vallate papilla had a mean of 8.37 alpha-gustducin-expressing cells and 5.22 A-expressing cells per taste bud, whereas the fungiform papillae contained 3.06 and 0.23 cells per taste bud, respectively. In both CT-CT and IX-CT rats there was a normal number of cells expressing alpha-gustducin or the A antigen in regenerated taste buds; in the CTX animals there was a significant decrease in the expression of these markers. These results demonstrate that the molecular phenotype of taste bud cells is determined by the local epithelium from which they arise and not by properties of the innervating nerve.