The specification of feather and scale protein synthesis in epidermal-dermal recombinations

Keratin proteins synthesized by dorsal or tarsometatarsal embryonic chick epidermis in heterotopic and heterospecific epidermal-dermal recombinants were analyzed by polyacrylamide gel electrophoresis and were compared to those produced by normal nondissociated dorsal and tarsometatarsal embryonic skin, as well as to those produced by control homotopic recombinants. Recombinant skins were grafted on the chick chorioallantoic membrane and grown for 8 or 11 days. Recombinants comprising dorsal feather-forming dermis formed feathers, irrespective of the origin of the epidermis. The electrophoretic band patterns of the keratins extracted from these feathers were of typical feather type. Conversely recombinants comprising tarsometatarsal scale-forming dermis formed scales, irrespective of the origin of the epidermis. The band patterns of the keratins extracted from the epidermis of these scales were of typical scale type. Heterospecific recombinants comprising chick dorsal feather-forming epidermis and mouse plantar dermis gave rise to six footpads arranged in a typical mouse pattern. In these recombinants, the chick epidermis produced keratins, the band pattern of which was of typical chick scale type. These results demonstrate that the dermis not only induces the formation of cutaneous appendages in confirmity with its regional origin, but also triggers off in the epidermis the biosynthesis of either of two different keratin types, in accordance with the regional type (feather, scale, or pad) of cutaneous appendages induced. The possible relationship between region-specific morphogenesis and cytodifferentiation is discussed in comparison with results obtained in other kinds of epithelial-mesenchymal interactions.     

Induction of epidermal mucous metaplasia by culture of recombinants of undifferentiated epidermis and retinol-treated dermis in a chemically defined medium

Epidermal mucous metaplasia of cultured skin is known to be induced by excess retinol. Studies were made on whether retinol affects primarily the epidermis or the dermis during retinol-induced epidermal mucous metaplasia of 13-day-old chick embryonic skin in culture. When recombinants of 13-day-old normal epidermis and retinol-treated dermis were cultured for 7 days in chemically defined medium in the absence of retinol, hormones, and serum, they showed altered epidermal differentiation toward secretory epithelium (mucous metaplasia). Thus retinol acted primarily on dermal cells.     

Acceleration of Retinol-Induced Epidermal Mucous Metaplasia by Stimulating the Dermal Adenylate Cyclase-cAMP System in Chick Embryonic Skin: Appearance of cAMP-Dependent Phosphorylated Proteins in Dermis of Retinol-Pretreated Skin after 2 h-Treatment with cAMP

Epidermal mucous metaplasia of cultured 13-day-old chick embryonic tarsometatarsal skin can be induced by culture in medium containing excess retinol (20 μM) for only 8–24 h and then in a chemically defined medium with Bt₂cAMP (0.2–2 mM) and without retinoids or serum for 2 days. In this work, stimulation of the adenylate cyclase-cAMP system in retinol-pretreated skin by forskolin, pertussis toxin, cholera toxin or AIF₄^– was found to accelerate the synthesis of epidermal sulfated glycoprotein (mucin). In skin induced toward mucous metaplasia by retinol, treatment with forskolin for 1 day increased the cAMP content 10-fold in the dermis but only 2-fold in the epidermis over the control levels. The cAMP level of Bt₂cAMP (0.2 mM)-treated skin was 18 times higher in the dermis but rather lower in the epidermis than untreated skin. These results suggest the importance of an adenylate cyclase-cAMP system in the dermis of skin in stimulating mucous metaplasia induced by retinoids. In fact, cAMP-dependent protein phosphorylation was seen only in the dermis of retinol-pretreated skin after 2 h-treatment with cAMP. As no transfer of cAMP from the dermis to the epidermis of forskolin-treated skin was detected, there may be no gap junctional communication between the epidermis and the dermis, while the basement membrane becomes discontinuous during mucous metaplasia.     

Use of retinoic acid for the analysis of dermal-epidermal interactions in the tarsometatarsal skin of the chick embryo

Feet of chicks are normally covered with scales. Injection of retinoic acid into the amniotic cavity of 10-day chick embryos causes the formation of feathers on the foot scales. To elucidate whether retinoic acid affects primarily the epidermis or the dermis, heterotypic dermal-epidermal recombinants of tarsometatarsal skin were tested as to their morphogenetic capacity, when grafted to the chick chorioallantoic membrane. Recombinants involving treated epidermis and untreated dermis formed feathered scales, while the reverse recombinants of untreated epidermis and treated dermis led to the formation of scales only. Likewise the association of treated tarsometatarsal dermis with untreated epidermis from a non-appendage-forming region (the midventral apterium) resulted in the formation of scales only. These results show that retinoic acid affects primarily the epidermis. Further insight into the mechanism of dermal-epidermal interaction was gained by heterotopic recombinations of early (8.5- and 10-day) untreated tarsometatarsal dermis with epidermis from the midventral apterium. These recombinants formed scales, proving that tarsometatarsal dermis is endowed with scale-forming properties as early as 8.5 days of incubation. Finally, it is concluded that retinoic acid acts on the chick foot epidermal cells by temporarily inhibiting their scale placode-forming properties, allowing their latent feather placode-forming properties to be expressed.     

Neutral morphogenetic activity of epithelium in heterologous tissue recombinations

To assess the existence of specific and nonspecific epithelial instructions for mesenchymal cell differentiation we compared homospecific and heterospecific mouse and quail tissue recombinations. In heterospecific recombinants between trypsin-dissociated mouse molar mesenchyme and quail epithelia neither odontoblasts nor chondrocytes differentiated. Cartilage appeared if the quail epithelium was contaminated with homologous limb mesenchyme and odontoblasts differentiated if the mouse dental epithelium was contaminated with dental papilla cells.     

Neutral morphogenetic activity of epithelium in heterologous tissue recombinations

Abstract. To assess the existence of specific and nonspecific epithelial instructions for mesenchymal cell differentiation we compared homospecific and heterospecific mouse and quail tissue recombinations. In heterospecific recombinants between trypsin-dissociated mouse molar mesenchyme and quail epithelia neither odontoblasts nor chondrocytes differentiated. Cartilage appeared if the quail epithelium was contaminated with homologous limb mesenchyme and odontoblasts differentiated if the mouse dental epithelium was contaminated with dental papilla cells.     

Effects of retionoids on invasion of organ cultures of chick chorioallantoic membrane by adenovirus transformed cells

Summary Invasion of chick chorioallantoic membrane (CAM) organ cultures by rat 3Y1 cells transformed by the highly oncogenic human adenovirus type 12 (3Y1/12-10 cells) was inhibited by several retinoids tested. The anti-invasive activity of the retinoids was dependent on retinoid concentration and continuous (4d) exposure of the CAM. The 50% retinoid dose (dose effective in achieving a response in half of the organ cultures) that inhibited invasion was 0.85 μg/ml of retinol palmitate, 0.39 μg/ml of retinoic acid, or 0.16 μg/ml of retinol acetate. This dose was of the same order of magnitude as that which induced CAM differentiation, and was three-to fourfold less than the dose that caused cytotoxic damage of CAM. In addition, the retinoids inhibited 3Y1/12-10 cell growth by approximately 40% at levels over 10-fold higher than those needed for anti-invasion activity. The findings suggest that the anti-invasive activity of retinoids was at least partly due to direct induction of cell differentiation of the CAM host tissue. This work was supported by National Cancer Institute Grant CA 13231 and by University of Akron Grant RG 832.     

beta-Catenin has sequential roles in the survival and specification of ventral dermis

The dermis promotes the development and maintains the functional components of skin, such as hair follicles, sweat glands, nerves and blood vessels. The dermis is also crucial for wound healing and homeostasis of the skin. The dermis originates from the somites, the lateral plate mesoderm and the cranial neural crest. Despite the importance of the dermis in the structural and functional integrity of the skin, genetic analysis of dermal development in different parts of the embryo is incomplete. The signaling requirements for ventral dermal cell development have not been established in either the chick or the mammalian embryo. We have shown previously that Wnt signaling specifies the dorsal dermis from the somites. In this study, we demonstrate that Wnt/beta-catenin signaling is necessary for the survival of early ventral dermal progenitors. In addition, we show that, at later stages, Wnt/beta-catenin signaling is sufficient for ventral dermal cell specification. Consistent with the different origins of dorsal and ventral dermal cells, our results demonstrate both conserved and divergent roles of beta-catenin/Wnt signaling in dermal development.     

The use of retinoids as probes for analyzing morphogenesis of glands from epithelial tissues

Summary Thirty-five years ago Honor Fell and Edward Mellanby were studying effects of high doses of vitamin A on skeletal development in chick embryos when they noticed that a piece of epidermis, accidentally included in an organ culture, had undergone mucous metaplasia. Further studies by Fell and others eventually led to an understanding of the important role of vitamin A in modulating epithelia in vivo. Fifteen years later another organ culture experiment showed me that excess vitamin A could also initiate the morphogenesis of branching and mucus-secreting glands from developing vibrissa follicles in upper lip skin of embryonic mice. Since then our group has shown that induction of this novel structure by naturally occurring retinoids resembles a normal embryonic induction in that it is stage-dependent, time-dependent, and irreversible. Tissue separation and recombination studies showed that isolated upper lip epidermis can form these glands when combined with retinoid-treated upper lip dermis. Untreated mouse epidermis can form similar glands after combination with chick dermis containing higher retinoid levels. The hamster cheek pouch, normally devoid of glandular structures, can also form mucous glands when treated with a retinoid, either in vivo or in vitro. Recombination studies in organ culture have now shown that mesenchyme exposed to retinoid is essential for gland morphogenesis from pouch epithelium. Evidences is accumulating that retinoic acid may even be the active morphogen in some normally developing systems.     

Developmental changes in endogenous retinoids during pregnancy and embryogenesis in the mouse.

Vitamin A and its analogs (retinoids) have acquired particular significance in embryonic development since the discovery that retinoic acid (RA) possesses properties of an endogenous morphogen and that embryonic tissues contain specific nuclear receptors for RA. Since the mammalian embryo does not synthesize RA de novo but rather must acquire it directly or in a precursor form from the maternal circulation, we sought to establish the relationship between levels of RA, retinol, and retinyl esters in the maternal system and their acquisition by the embryo, particularly during organogenesis in the mouse. Results indicate profound changes in maternal vitamin A levels during pregnancy in the mouse. These changes were characterized by a large, transient decrease in plasma retinol levels coincident with the period of organogenesis (e.g. gestational Days 9-14), and an apparent increase in mobilization from hepatic stores to the conceptus. During organogenesis, the embryo exhibited a steady increase in retinol levels with little increase in retinyl esters and virtually no change in RA. Analysis of retinoid accumulation patterns in the embryonic liver indicate that functional onset of vitamin A storage occurs by mid-organogenesis. In contrast, placental levels of these retinoids remained unchanged throughout organogenesis. Analysis of the conceptus as a developmental unit revealed that during early organogenesis the majority of retinoids are contained in the placenta (8-fold more than in the embryo). However, by mid-organogenesis the retinoid content of the embryo exceeds that of the placenta. Together, these results provide evidence that pregnancy in the mouse is accompanied by pronounced alterations in maternal retinoid homeostasis that occur coincident with the period of high embryonic sensitivity to exogenous retinoids.     

Induction of Mucous Metaplasia in Chick Embryonic Skin by Retinol-Pretreated Embryonic Chick or Quail Dermal Fibroblasts through Cell-Cell Interaction: Correlation of a Transient Increase in Retinoic Acid Receptor β mRNA in Retinol-Treated Dermal Fibroblasts with Their Competence to Induce Epidermal Mucous Metaplasia

Epidermal mucous metaplasia of 13-day-old chick embryonic tarsometatarsal skin can be induced by culture in medium containing 20 μM retinol for only 8 hr and then in a chemically defined medium without retinol for 2 days. Retinol primarily affects the dermal cells, which then transform the epithelial cells into mucus-secreting cells. In this study, we developed a system using a combination of retinol-pretreated chick or quail dermal fibroblasts and chick skin, and showed that retinol-pretreated quail embryonic dermal fibroblasts invaded the dermis of chick embryonic skin to beneath the epidermal basal cells within 1 day of culture and induced metaplasia, suggesting that epidermal mucous metaplasia of the skin was induced by the direct interaction of retinol-pretreated dermal fibroblasts with the epidermal cells or by low diffusible paracrine factor produced by the fibroblasts.
Increase in retinoic acid receptor β (RARβ) mRNA in dermal fibroblasts was observed after 8 hr-treatment with retinol which preceded morphological changes induced by retinol and this increase was correlated with the competence of the dermal fibroblasts to induce epidermal mucous metaplasia. Thus some gene product(s) controlled by RARβ in dermal fibroblasts may be an essential signal for induction of epidermal mucous metaplasia.     

UVA/B exposure promotes the biosynthesis of dehydroretinol in cultured human keratinocytes

Retinol and its metabolites modulate epithelial differentiation and serve as cellular UV sensors through changes in retinoid status. Of note is the dehydroretinol family which may serve functions distinct from parental retinol. This study focuses on the metabolism of this family and its potential participation in the response of normal epidermal human keratinocytes to UV irradiation. There were three findings. First, keratinocytes contain two pools of dehydroretinyl esters, one of which is shielded from UVB-, but not from UVA-induced decomposition. Second, using a novel in vitro assay we demonstrated that both UVA and UVB promote dehydroretinol biosynthesis in keratinocytes, but only UVB exposure promotes retinoid ester accretion by enhancing the activity of at least one acyl transferase. Finally, dehydroretinol sufficiency reduces UVA/B driven apoptosis more effectively than retinol sufficiency. This may in part be due to differences in the expression of Fas ligand, which we found to be upregulated by retinoic acid, but not dehydroretinoic acid. These observations implicate a role of dehydroretinol and its metabolites in UVA/B adaptation. Thus, the keratinocyte response to UV is jointly shaped by both the retinoids and dehydroretinoids.     

Molecular mechanisms controlling dorsal dermis generation from the somitic dermomyotome

The initiation of the development of skin appendages (hair/feathers/scales) requires a signal from the competent dense dermis to the epidermis (Dhouailly, 1977). It is therefore essential to understand how to make a competent dermis. In recent years, a few studies have focused on the development of the dorsal dermis from the somitic dermomyotome. Our first aim in this review is to attempt to reconcile the available data on the origin of the dorsal dermis and summarize the present knowledge on the molecular mechanisms implicated in dermal lineage induction. Secondly, we open the discussion on the formation of a loose pre-dermal mesenchyme and more importantly of a dense dermis capable of participating in appendage development. To go further we draw a comparison between the chick and mouse systems to gain a new insight into how to initiate appendage morphogenesis and regulate the extent of hair/feather fields.     

Endogenous retinoids in the hair follicle and sebaceous gland

Vitamin A and its derivatives (retinoids) are critically important in the development and maintenance of multiple epithelial tissues, including skin, hair, and sebaceous glands, as shown by the detrimental effects of either vitamin A deficiency or toxicity. Thus, precise levels of retinoic acid (RA, active metabolite) are needed. These precise levels of RA are achieved by regulating several steps in the conversion of dietary vitamin A (retinol) to RA and RA catabolism. This review discusses the localization of RA synthesis to specific sites within the hair follicle and sebaceous gland, including their stem cells, during both homeostasis and disease states. It also discusses what is known about the specific roles of RA within the hair follicle and sebaceous gland. This article is part of a Special Issue entitled: Retinoid and Lipid Metabolism.     

Differences in the histogenesis and keratin expression of avian extraembryonic ectoderm and endoderm recombined with dermis

The responses of the chorionic ectoderm and allantoic endoderm (from 8-day chick embryos) to dermal induction were compared through tissue recombinants grafted onto the chorioallantoic membrane. The chorionic epithelium formed the appropriate epidermis with a fully developed stratum corneum in response to both spur and scutate scale dermises. Analysis of these recombinant epidermal tissues by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) demonstrated that tissue-specific expression of the alpha (alpha) and beta (beta) keratin polypeptides occurred. In addition, indirect immunofluorescence studies with antisera to alpha or beta keratins showed that the beta stratum, which characterizes the epidermis of spurs and scutate scales, was formed, and the alpha keratins were distributed as in the normal epidermal tissues. In contrast, although the allantoic endoderm became stratified in association with either spur or scutate scale dermis, a stratum corneum with a beta stratum did not develop. SDS-PAGE analysis demonstrated that while the characteristic beta keratins of scutate scales and spur were not detected, most of the alpha keratins normally elaborated by these structures were present, suggesting that even without histogenesis of a stratum corneum the expression of alpha keratins of endoderm could be regulated in a tissue-specific manner by dermis. This study also demonstrated that there are differences in the abilities of the chorionic and allantoic epithelia to respond to the same dermal cues, which may reflect earlier restrictions in their developmental potentials.     

Tgm2/Gh, Gbx1 and TGF-beta are involved in retinoic acid-induced transdifferentiation from epidermis to mucosal epithelium

We previously demonstrated that retinoic acid (RA) induces epidermis to transdifferentiate to mucosal epithelium with goblet cells in chick embryonic cultured skin. To characterize the molecular mechanism of this transdifferentiation process, we used rat embryonic cultured skin and immunohistochemistry to confirm that RA-induced epidermal transdifferentiation accompanies the expression of markers of esophagus epithelium. Because Gbx1, TG2/Gh (transglutaminase2) and TGF-beta2 are reported individually to be induced by RA in cultures of chick embryonic skin, mouse epidermal cells and human hair follicles respectively, here, we investigated whether cooperative interplay of Gbx1, TG2/Gh and TGF-beta2 is required for the transdifferentiation of epidermal cells to mucosal cells. We have shown that expression of Gbx1, TG2/Gh and TGF-beta proteins were all upregulated in RA-induced transdifferentiated skin and that the former two were expressed in the epidermis, while TGF-beta was expressed in the dermis. Inhibitors of the TGF-beta signal pathway partially inhibited transdifferentiation. Overexpression of both hTG2/Gh and mGbx1 together in the epidermis by electroporation resulted in cuboidal cells in the upper cell layers of the epidermis without keratinized layers, although epidermal keratinization was observed in skin by overexpression of either of them. Labeling DNA with BrdU indicated that RA directly transdifferentiated transient amplifying epidermal cells, not stem cells, to mucosal cells. This study showed that coexpression of TG/2 and Gbx1 in the epidermis was required for esophagus-like mucosal transdifferentiation, and that increase in TGF-beta2 expression by RA in the dermis was essential to induce transdifferentiation through epithelial-mesenchymal interaction.     

Enzymes and binding proteins affecting retinoic acid concentrations

Free retinoids suffer promiscuous metabolism in vitro. Diverse enzymes are expressed in several subcellular fractions that are capable of converting free retinol (retinol not sequestered with specific binding proteins) into retinal or retinoic acid. If this were to occur in vivo, regulating the temporal-spatial concentrations of functionally-active retinoids, such as RA (retinoic acid), would be enigmatic. In vivo, however, retinoids occur bound to high-affinity, high-specificity binding proteins, including cellular retinol-binding protein, type I (CRBP) and cellular retinoic acid-binding protein, type I (CRABP). These binding proteins, members of the superfamily of lipid binding proteins, are expressed in concentrations that exceed those of their ligands. Considerable data favor a model pathway of RA biosynthesis and metabolism consisting of enzymes that recognize CRBP (apo and holo) and holo-CRABP as substrates and/or affecters of activity. This would restrict retinoid access to enzymes that recognize the appropriate binding protein, imparting specificity to RA homeostasis; preventing, e.g. opportunistic RA synthesis by alcohol dehydrogenases with broad substrate tolerances. An NADP-dependent microsomal retinol dehydrogenase (RDH) catalyzes the first reaction in this pathway. RDH recognizes CRBP as substrate by the dual criteria of enzyme kinetics and chemical crosslinking. A cDNA of RDH has been cloned, expressed and characterized as a short-chain alchol dehydrogenase. Retinal generated in microsomes from holo-CRBP by RDH supports cytosolic RA synthesis by an NAD-dependent retinal dehydrogenase (RalDH). RalDH has been purified, characterized with respect to substrate specificity, and its cDNA has been cloned. CRABP is also important to modulating the steady-state concentrations of RA, through sequestering RA and facilitating its metabolism, because the complex CRABP/RA acts as a low K_m substrate.     

Exogenous retinoic acid decreases in vivo and in vitro proliferative activity during the early migratory stage of neural crest cells

We have previously demonstrated that directional migration of neural crest cells (NCC) is associated with a high cell density, resulting from an active cell proliferation. It is also known that treatment with retinoic acid (RA) causes a dose-dependent inhibition of proliferation of some cell types, and that administration of RA during the early stages of embryonic development, induces cranio-facial abnormal patterns corresponding to NCC derivatives. In view of these findings, it was of interest to determine if exogenous RA is a potential modulator of the mitotic rate of NCC, and to explore the hypothesis of an inhibitory effect exerted by RA on the proliferative behaviour of NCC in vivo and in vitro. Homogenates of RA-treated chick embryos showed a low [³H]dT incorporation, indicating a generalized diminution of DNA synthesis. The labelling index (LI=number of labelled cells/total number of cells) revealed that NCC from RA-treated and control embryos had higher values of [³H]dT incorporation than neural tube cells (P < 0.0001). Autoradiographs of RA-treated chick embryos showed a significantly lower [³H]dT incorporation in NCC at the prosencephalic and mesencephalic levels, as well as in the neural tube cells at the prosencephalic, mesencephalic and rhombencephalic levels, than in control chick embryos (P < 0.0001). NCC cultures treated with 1 or 10 μm RA had a significantly lower LI than in cultures treated with 0.1 μm RA or control cultures (P < 0.04). In chick embryos, the mitotic index of NCC was 0.026 for RA-treated and 0.033 for controls, while the duration of the cell cycle was significantly longer in the NCC of RA-treated embryos (～ 40 h) than in controls (～ 25 h). The length of the cell cycle phases of NCC was similar in both experimental conditions, except for G₁ phase, which was significantly longer in the RA-treated group than in controls. These results show that RA blocks DNA synthesis and lengthens the proliferative behaviour of NCC both in early chick embryos and in vitro, effects that could modify the morphogenetic patterns of NCC distribution through a decreased cell population.