Coevolutionary patterning of teeth and taste buds

Abstract

Teeth and taste buds are iteratively patterned structures that line the oro-pharynx of vertebrates. Biologists do not fully understand how teeth and taste buds develop from undifferentiated epithelium or how variation in organ density is regulated. These organs are typically studied independently because of their separate anatomical location in mammals: teeth on the jaw margin and taste buds on the tongue. However, in many aquatic animals like bony fishes, teeth and taste buds are colocalized one next to the other. Using genetic mapping in cichlid fishes, we identified shared loci controlling a positive correlation between tooth and taste bud densities. Genome intervals contained candidate genes expressed in tooth and taste bud fields. sfrp5 and bmper, notable for roles in Wingless (Wnt) and bone morphogenetic protein (BMP) signaling, were differentially expressed across cichlid species with divergent tooth and taste bud density, and were expressed in the development of both organs in mice. Synexpression analysis and chemical manipulation of Wnt, BMP, and Hedgehog (Hh) pathways suggest that a common cichlid oral lamina is competent to form teeth or taste buds. Wnt signaling couples tooth and taste bud density and BMP and Hh mediate distinct organ identity. Synthesizing data from fish and mouse, we suggest that the Wnt-BMP-Hh regulatory hierarchy that configures teeth and taste buds on mammalian jaws and tongues may be an evolutionary remnant inherited from ancestors wherein these organs were copatterned from common epithelium.

Keywords: bipotency; placode patterning; plasticity; quantitative trait loci; tooth/taste bud development.

Fig. 1.

QTL profile for significant tooth (red) and taste bud (blue) genetic effects, with chromosome position plotted against LOD score. Best-scoring SNP markers from MQM models were located in cichlid genomes and all annotated genes 1 Mb on either side were identified. (Lower) Candidate genes for tooth and taste bud density are indicated along expanded 2-Mb portion of the x axis approximately positioned from the center of the peak for tooth (red) and taste bud (blue). Note shared QTL for tooth and taste bud density on chromosomes 17 and 20.

Fig. 2.

Expression of pitx2, (A and D), sox2 (C and F), and shh (B and E) at early stages of tooth and taste bud copatterning. Sagittal section at initial stages of jaw formation, early 5 dpf (20×; scale bar, 20 µm) and late 5 dpf (40×; scale bar, 40 µm). Black arrows show shared first arch lamina and white arrowhead shows reduced pitx2 in posterior pharynx (A) compared with shh (B) and sox2 (C). Rostral is to the left of page, ventral to the bottom; 18-µm-thick sections.

Fig. 3.

Expression of candidates Bmper and Sfrp5 in mouse teeth and tongues. Bmper and Sfrp5 expression, as shown in frontal section of bud stage molar teeth at E12.5 (A and D) and cap stage teeth at E15.5 (A′ and D′) (20×). (Scale bar, 200 µm.) Gene expression in tongue as observed in sagittal (B, B′, E, and E′) and frontal (C, C′, F, and F′) sections at E12.5 and E15.5, respectively (10×) (Scale bar, 400 µm.)

Fig. 4.

Effects of chemical treatment on tooth and taste bud density. calb2 ISH was used to score taste bud density (A–D) and cleared and stained jaws were used to score tooth density (A′–D′). The midline of cleared and stained fish is marked by a white line; teeth in each half of the dentary are marked by white numerals. Dorsal views of dentaries, labial to bottom of page. (Scale bars, 100 µm.) Box plots summarize statistical analysis of chemical treatments vs. control (DMSO), for taste buds (no. of taste buds per 10 μm²) and teeth (no. of teeth per 100 μm²). All treatments are significantly different from control, with P < 0.0001; n = numbers of animals used.

Fig. 5.

ISH of genes in cichlid dentary following 24-h treatment with LDN or CYC initiated at 6 dpf and immediate euthanasia. Dorsal views, labial to bottom of page. (Scale bar, 100 µm.)

Fig. 6.

Dorsal views of dentary after whole-mount IHC of taste bud marker Calb2 (green) and nerve marker acetylated Tubulin (red) following LDN or CYC treatment. Three-dimensional rendering of 150-µm optical sections overlaid to bright-field image at 10×. White arrows indicate ectopic taste bud. Labial to bottom of page. (Scale bar, 100 µm.)

Fig. 7.

A model of evolutionary conserved patterning networks for oral organs. Humans possess taste buds on the tongue and teeth on the dental arch, whereas these organs are copatterned in cichlids and other fishes. Inferred roles of Wnt-BMP-Hh interactions and effects on tooth and taste bud patterning in cichlids represented before placode condensation, during placode formation, and during induction of mesenchyme to both organs, from proximal to distal of horizontal plane. Genomic candidates are highlighted in red text. Preplacode stage represents the model of placode recruitment from a bipotent epithelium. These interactions are consistent with reports in the mouse for taste bud (9, 11, 70) and tooth (8, 21, 38, 47, 49) patterning networks, when studied independently. Based on cichlid expression and treatment data, Wnts drive the formation of taste placodes proximally and tooth placodes distally, whereas BMPs and Hhs are inhibitory of taste bud differentiation and permissive for tooth germs. Anti-BMPs, such as osr2, may reduce BMP activity in the taste field to promote taste bud formation and sfrp5, expressed in both fields, may repress Wnt signaling. At the point of mesenchyme induction, pitx2 is expressed in tooth placodes and the underlying mesenchyme expresses bmp2/4. Simlarly, taste buds express calb2 and anti-BMPs osr2 and bmper are expressed in the lingual mesenchyme.