Involvement of Wnt, Eda and Shh at defined stages of sweat gland development

Abstract

To maintain body temperature, sweat glands develop from embryonic ectoderm by a poorly defined mechanism. We demonstrate a temporal cascade of regulation during mouse sweat gland formation. Sweat gland induction failed completely when canonical Wnt signaling was blocked in skin epithelium, and was accompanied by sharp downregulation of downstream Wnt, Eda and Shh pathway genes. The Wnt antagonist Dkk4 appeared to inhibit this induction: Dkk4 was sharply downregulated in β-catenin-ablated mice, indicating that it is induced by Wnt/β-catenin; however, its overexpression repressed Wnt target genes and significantly reduced gland numbers. Eda signaling succeeded Wnt. Wnt signaling was still active and nascent sweat gland pre-germs were still seen in Eda-null mice, but the pre-germs failed to develop further and the downstream Shh pathway was not activated. When Wnt and Eda were intact but Shh was ablated, germ induction and subsequent duct formation occurred normally, but the final stage of secretory coil formation failed. Thus, sweat gland development shows a relay of regulatory steps initiated by Wnt/β-catenin - itself modulated by Dkk4 - with subsequent participation of Eda and Shh pathways.

Keywords: Ectodermal dysplasia; Exocrine gland; Hair follicle; Heatstroke; Hyperhidrosis; Mouse; Skin appendage.

Fig. 1.

Ablation of β-catenin from skin epidermis results in blocking of sweat gland induction. (A) β-catenin and Lef1 are highly expressed in sweat gland germs in wild type (arrows), but undetectable in β-Cat cKO embryos. Scale bar: 20 µm. (B) Developmental histology. Pre-germs/early germs can be occasionally observed at E16.5, germs/advanced germs at E17.5 and early coiling at around P0 (arrows in wild type). No pre-germ/germ formation in β-Cat cKO embryos. Right panels show the absence of hair follicle development in the mutant mice. Scale bars: 25 µm. (C) Cell proliferation and cell death status in β-Cat cKO footpads. Ki67-positive cells are scattered in the basal layers of wild-type and β-Cat cKO embryos (upper panels). Caspase 3 is not present in either wild-type or β-Cat cKO embryos (lower panels). Arrows indicate pre-germs. Caspase 3 is occasionally found in cells close to the epidermal ducts in wild-type adult mice (an arrow in right panel). Scale bars: 25 µm.

Fig. 2.

Genes regulated by Wnt/β-catenin during sweat gland development. (A) Expression profiling was carried out with whole footpad skin from wild-type and β-Cat cKO embryos at E15.5, E16.5 and E17.5. (B) The number of genes significantly affected in β-Cat cKO embryos. Forty genes were affected at all three developmental time points. (C) qRT-PCR assays confirmed significant downregulation of Edar, Shh, Dkk4, Wnt10b and Fzd10 in β-Cat cKO footpads. ***P≤0.05 for Fzd10 and ***P≤0.001 for others. Data are mean±s.e.m. (D) Edar protein (arrow) was localized to the membrane of sweat gland germ cells in wild-type, but not in the mutant, embryos. Dashed lines demarcate the epidermal-dermal junction. Scale bar: 20 µm.

Fig. 3.

Sweat gland numbers were significantly reduced in Dkk4 transgenic mice. (A) In situ hybridization revealed selective expression of Dkk4 in sweat gland germs in wild-type embryos, and high expression in the basal layer of skin epidermis and sweat gland germs in the transgenic mice. Arrows indicate sweat gland germs. Scale bar: 50 µm. (B) Sweat test showed decreased numbers of sweating spots in the transgenic mice. (C) A reduction of ∼40% of sweating spots was seen in the transgenic mice. *P≤0.01. Data are mean±s.e.m. (D) The sizes of sweat gland clusters (outlined) in the transgenic mice were about half of wild-type controls in adult (P75) and developing (P8) mice. Secretory regions and ducts opened normally in the transgenic mice (P75). Scale bar: 100 µm. (E) Formed sweat gland germs at E16.5 (arrows) and E17.5 (arrowheads) normally expressed Lef1 and Edar in the transgenic embryos. Scale bar: 20 µm. (F) qRT-PCR assays show moderate but significant downregulation of several Wnt target genes in transgenic embryos. * and **P≤0.005 for Edar and Shh; P≤0.05 for the others. Data are mean±s.e.m.

Fig. 4.

Abortive sweat gland pre-germ formation in Tabby. (A) A nascent pre-germ like structure formed in Tabby at E17.5 (arrows in Ta). Scale bar: 25 µm. (B) Lef1 staining shows developing sweat glands in wild-type and nascent pre-germs in Tabby (arrows). Lef1 is found in pre-germs, germs and cells in the tip of growing ducts in wild type, and in pre-germs in Tabby (arrows). Lef1 is undetectable in a sweat duct ready to coil in wild type (right panel). Scale bar: 25 µm. (C) Ki67 is found at high levels in advanced germs in wild type and at lower levels in pre-germ cells in Tabby (arrows). Caspase 3 is not present in either wild-type or Ta footpads. Scale bar: 20 µm. (D) qRT-PCR assays show sharp downregulation of Eda and Shh in Tabby. Dkk4 was significantly downregulated only at E17.5. ***P≤0.002. Data are mean±s.e.m. (E) Dkk4 transgenic Tabby mice lack sweat glands, but form pre-germ like structures, as seen in Tabby. Arrows indicate pre-germs and germs of sweat glands for each genotype. Scale bar: 25 µm.

Fig. 5.

Ablation of Shh from skin epidermis results in rudimentary secretory region formation. (A) Hair follicles were arrested at the germ stage in the mutant mice (arrows). Scale bar: 100 µm. (B) Secretory regions started to coil, but further progression was interrupted in Shh cKO mice (P5 and P8). Arrows indicate sweat ducts with normal spacing in the mutant mice. Scale bar: 100 µm. (C) Gland germs were induced as normal, and Lef1 and Edar were expressed as normal in the mutant mice. Scale bar: 20 µm. (D) Krt14 and Krt8 staining shows normal sweat ducts, but rudimentary secretory regions (circled by broken lines) formed in Shh cKO mice. Scale bar: 20 µm.

Fig. 6.

Schematic representation of sweat gland development. Wnt/β-catenin regulates sweat gland induction through Eda/Edar and Dkk4, and possibly also through Wnt10b, Fzd10 and Epgn (see text). Dkk4 suppresses gland induction through a negative-feedback effect on Wnt/β-catenin. The Bmp pathway and Engrailed may regulate Wnt/β-catenin upstream, and Eda/Edar may regulate Dkk4 at late developmental stages. Eda and Edar are required for sweat duct formation downstream of Wnt/β-catenin; Wnt10b, Fzd10, Tbx3 and Mmp9 may also be involved. Shh is required for secretory region formation, and Foxa1 is likely involved in the completion of sweat gland formation.