Anne Lawton

BDNF is required for the normal development of taste neurons in vivo.

THE vallate gustatory epithelium of neonatal trkB null mutant mice (−/−) lacked innervation. This prompted the evaluation of null mutant mice corresponding to the three neurotrophin ligands for tyrosine kinase receptor B (TrkB): brain-derived neurotrophic factor (BDNF), neurotrophin (NT)3, NT4. The vallate gustatory epithelium of nt3−/− mice and of nt4−/− mice appeared normal. Only bdnf−/− mice had a vallate papilla that was stunted, sparsely innervated, and lacked up to 98% of its taste buds. All three defects persisted. For example, the vallate papilla of 12-day-old bdnf−/− mice remained markedly less well innervated than the vallate of 7-day-old or newborn bdnf+/+ mice. The foliate taste papillae of neonatal bdnf−/− mice had similar defects. We conclude that the normal development of taste neurons requires BDNF.

THE MORPHOGENESIS OF MOUSE VALLATE GUSTATORY EPITHELIUM AND TASTE BUDS REQUIRES BDNF-DEPENDENT TASTE NEURONS

The developmental absence of brain-derived neurotrophic factor (BDNF) in null mutant mice caused three interrelated defects in the vallate gustatory papilla: sparse innervation, a reduction in the area of the gustatory epithelium, and fewer taste buds. On postnatal day 7, the stunted vallate papilla of bdnf null mutant mice was 30% narrower, the trench walls 35% reduced in area, and the taste buds 75% less abundant compared with wild-type controls. Quantitative assessment of innervation density was carried out to determine if the small trench walls and shortage of taste buds could be secondary consequences of the depletion of gustatory neurons. The diminished gustatory innervation was linearly associated with a reduced trench wall area (r=+0.94) and fewer taste buds (r=+0.96). Residual taste buds were smaller than normal and were innervated by a few surviving taste neurons. We conclude that BDNF-dependent taste neurons contribute to the morphogenesis of lingual gustatory epithelia and are necessary for both prenatal and postnatal mammalian taste bud formation. The gustatory system provides a conspicuous example of impaired sense organ morphogenesis that is secondary to sensory neuron depletion by neurotrophin gene null mutation.

Neural control of ectopic filiform spines in adult tongue.

The tongue surface directly above a fungiform taste bud is flat, thinly keratinized, and free of filiform spines. We examined fungiform papillae in serial sections of rat and gerbil tongues after unilateral transection of the chorda-lingual nerve had caused many fungiform taste buds to degenerate. Such empty fungiform papillae often formed a solitary keratinized outgrowth that closely resembled the spine of an ordinary filiform papilla. By six months an ectopic spine was found on 61% of empty fungiform papillae, but never on fungiform papillae that contained a taste bud. Experimental innervation of the tongue reduced the incidence of ectopic filiform spines in proportion to the cross-sectional area of the trigeminal nerve branches tested (the mylohyoid nerve, the lingual nerve, lingual + mylohyoid or lingual + auriculotemporal nerves). The chorda tympani nerve was 60 times more effective than trigeminal nerves in preventing ectopic filiform spines. We suggest that positive and negative trophic actions are normal characteristics of taste axons, for they promote the formation of taste buds and prevent the expression of ectopic filiform spines. By preventing the outgrowth of ectopic spines on fungiform papillae, taste axons maintain a thinly keratinized apical surface that can be breached by the taste receptor cells.

Glycoconjugates and keratin 18 define subsets of taste cells.

SummarySections of neonatal, normal adult and denervated adult rat tongue were examined with lectin histochemistry. Attention was focused upon intragemmal cells (cells within the taste bud) and the surrounding perigemmal cells. Informative staining patterns were observed with four of 12 lectins: Ulex europaeus (UEA-I), Bauhinia purpurea (BPA), Helix pomatia (HPA) and Lotus tetragonolobus (LTA) agglutinins. In normal adult tongues, BPA bound to those lingual epithelial cells lacking contact with the basal lamina. After they formed, vallate taste buds were laterally surrounded by distinctive BPA-positive cells. HPA reacted selectively with 28% and LTA with 23% of the intragemmal cells in vallate/foliate taste buds. In double-stained taste buds there was, a statistically significant overlap of LTA-positive cells and keratin 18-positive cells. The overlap between HPA binding and keratin 18 was more marked: double-stained cells comprized 67% of all stained cells. During taste bud development in neonates keratin 18 synthesis preceded HPA binding. In contrast, during the replacement of adult taste cells, keratin 18 synthesis and HPA binding were generally concurrent. Keratin 18 and HPA probably identify the same subset of older taste receptor cells. HPA may bind to glycoconjugates on the surface of keratin 18-positive cells. In denervated adult tongue the loss of all UEA-I-positive or BPA-positive perigemmal cells suggests that perigemmal as well as intragemmal cells are nerve-dependent.

Keratin Polypeptides and Taste Buds

Subsets of 20 soft keratin polypeptides have been identified in animal tissues by two-dimensional gel electrophoresis [1–6]. Differences among sets of keratins detected by immunocytochemistry distinguish types of epithelia or even specific regions of an epithelium [1,7–10]. From morphologic considerations we hypothesized that taste buds were islets of simple epithelium embedded in stratified squamous epithelium. If this hypothesis is correct, the keratins in taste buds ought to be distinctive from those of surrounding cells. We examined several gustatory epithelia with antikeratin antibodies to compare the immunoreactivity of fusiform, perigemmal, and basal gustatory cells and to identify useful cell markers. The present report concludes that antibodies against keratins 8 and 19 may be used as general differentiation markers for taste receptor cells.