Early development of fungiform papillae on the fetal rat tongue was examined: (1) to determine whether morphogenesis of the taste bud-bearing fungiform papillae is induced by nerve and (2) to study the growth pattern of the two sensory nerves that innervate the papilla. The papillae first appear on the 15th day of gestation (E15; E1 is the day when the dam is sperm positive) in rows parallel to the midline sulcus. There appears to be a medial-lateral and an anterior-posterior gradient in the sequence of papilla differentiation. The epithelium of the early papilla resembles a multilayered placode topped by a flattened surface periderm. Close examination of the peridermal cells at the apex of the papillae reveals that the cells have fewer surface microvilli and their cytoplasm is more electron opaque than that of similar cells in interpapillary regions. The basal cells in the placode-like epithelium differ from those in interpapillary regions in that they are postmitotic and have more mitochondria. At later stages, the papilla acquires a mesenchymal core and nerves grow into the core. Results from organ culture experiments of tongue fragments taken from E14 fetuses indicate that morphogenesis of fungiform papillae is initiated in the absence of sensory nerve influence, but the nerve exerts a trophic effect on their maintenance. The two sensory nerves of the tongue, the chorda tympani and the lingual branch of the trigeminal nerve, enter the tongue mesenchyme at E14 and grow toward the epithelium. By E15 the chorda tympani branches have reached the developing fungiform papillae, by E16 many have entered the papilla, and by E17 they have penetrated the epithelium at the papilla apex. Their fibers are associated exclusively with the cells at the papilla apex, where the taste bud will develop. The trigeminal nerve ramifies beneath the surface of the entire epithelium by E15. Later, it, too, sends branches into fungiform papillae; these ascend along the trunk of the chorda tympani and at E17 terminate in the connective tissue core around the chorda tympani field. The results are compatible with the notion that the tongue epithelium exerts a general tropic effect on growing axons of both sensory nerves, and the epithelial cells of the fungiform papilla apex exert a similar effect to which only the chorda tympani axons are responsive.

The initiation of the morphogenesis of gustatory papillae is independent of innervation. To address the question of whether taste bud formation is associated with gustatory papilla morphogenesis, we examined developing tongues in mouse embryos from embryonic day 11 to birth. Despite the smooth morphological appearance of the lingual dorsal surface at 13 days of gestation, we observed embryonic taste bud primordia as discrete collections of cytokeratin 8-positive and elongated cells in epithelial placodes in the anterior tongue. In subsequent stages until birth, cytokeratin 8 continues to be expressed in embryonic taste buds distributed in punctuate patterns at regular intervals along rows that are symmetrically located on both sides of the median sulcus in the dorsal anterior developing tongue. Embryonic taste buds were observed in the developing circumvallate papillae from 15.5 days of gestation until birth. The dorsal epithelium of the anterior tongue is not innervated when embryonic taste buds first occur. The increased numbers of embryonic taste buds in developing fungiform papillae until birth are not correlated with the neural invasion of the epithelium. Thus, taste buds occur prenatally more likely independently of the innervation.

The rat tongue has an extensive, complex innervation from four cranial nerves. However, the precise developmental time course and spatial routes of these nerves into the embryonic tongue are not known, although this knowledge is crucial for studying mechanisms that regulate development and innervation of the lingual taste organs, gustatory papillae and resident taste buds. We determined the initial spatial course of nerves in the developing tongue and papillae, and tested the hypothesis that sensory nerves first innervate the tongue homogeneously and then retract to more densely innervate papillae and taste buds. Antibodies to GAP-43 and neurofilaments were used to label nerve fibers in rat embryo heads from gestational day 11 through 16 (E11-E16). Serial sagittal sections were traced and reconstructed to follow paths of each nerve. In E11 rat, geniculate, trigeminal and petrosal ganglia were labeled and fibers left the ganglia and extended toward respective branchial arches. At E13 when the developing tongue is still a set of tissue swellings, the combined chorda/lingual, hypoglossal and petrosal nerves approached the lingual swellings from separate positions. Only the chorda/lingual entered the tongue base at this stage. At E14 and E15, the well-developed tongue was innervated by all four cranial nerves. However, the nerves maintained distinctive entry points and relatively restricted mesenchymal territories within the tongue, and did not follow one another in common early pathways. Furthermore, the chorda/lingual and glossopharyngeal nerves did not set up an obvious prepattern for gustatory papilla development, but rather seemed attracted to developing papillae which became very densely innervated compared to surrounding epithelium at El5. To effect this dense papilla innervation, sensory nerves did not first innervate the tongue in a homogeneous manner with subsequent retraction and/or extensive redirection of fibers into the taste organs. Results contribute to a set of working principles for development of tongue innervation. Points of entry and initial neural pathways are restricted from time of tongue formation through morphogenesis, suggesting distinctive lingual territories for each nerve. Thus, sensory and motor nerves distribute independently of each other, and sensory innervation to anterior and posterior tongue remains discrete. For taste organ innervation, gustatory papillae are not induced by a prepatterned nerve distribution. In fact, papillae might attract dense sensory innervation because neither chorda/lingual nor glossopharyngeal nerve grows homogeneously to the lingual epithelium and then redistributes to individual papillae.

Early afferent innervation and differentiation of sensory vestibular cells were studied in mouse embryos from gestation day (GD) 13 to 16. Afferent neurites were found as early as GD 13 in the epithelium when there were no clearly differentiated sensory cells. By GD 14 the earliest sensory cells which exhibited short hair bundles at their luminal pole were then contacted by afferent endings at their basal part. On GD 15 nerve endings establishing specialized synaptic contacts, characterized by asymmetrical membrane densities and synaptic bodies, were observed. At this stage, microtubules contacting the presynaptic membranes, as well as coated vesicles were found. On GD 16 the hair cells were multi-afferented and numerous synaptic bodies were found. These results showing a concomitance between the hair cell differentiation and the establishment of nerve contacts are discussed with particular respect to nerve-hair cell interactions during sensory differentiation. This study does not point to a primary induction of vestibular hair cell differentiation by nerve endings, but it is consistent with the possibility that the ingrowth of nerve fibers is one of many factors that influence the differentiation of receptor cells. With respect to synapse formation, it is assumed that the location of synaptic bodies at presynaptic densities is determined by the arrival of afferent nerve endings.