Ectoderm-Targeted Overexpression of the Glucocorticoid Receptor Induces Hypohidrotic Ectodermal Dysplasia
http://www.100md.com
内分泌学杂志 2005年第6期
Department of Animal Pathology, Veterinary Faculty (J.L.C., A.B., H.L.), University of Santiago de Compostela, E-27002 Lugo; Instituto de Biomedicina de Valencia-Consejo Superior de Investigaciones Cientifica (IBV-CSIC) (E.D., H.L., P.P.), E-46010 Valencia; Project on Cell and Molecular Biology and Gene Therapy (J.M.P., J.L.J.), Centro de Investigaciones Energeticas Medioambientales y Tecnologicas, E-28040 Madrid; and Fundación Valenciana de Investigaciones Biomédicas (FVIB) (E.D., H.L., P.P.), 46013 Valencia, Spain
Address all correspondence and requests for reprints to: Paloma Pérez, Laboratory of Animal Models, Genomics and Pharmacoproteomics Program, Fundación Valenciana de Investigaciones Biomédicas, Avenida Autopista del Saler 16, Camino de las Moreras, E-46013 Valencia, Spain. E-mail: pperez@ochoa.fib.es.
Abstract
Hypohidrotic ectodermal dysplasia is a human syndrome defined by maldevelopment of one or more ectodermal-derived tissues, including the epidermis and cutaneous appendices, teeth, and exocrine glands. The molecular bases of this pathology converge in a dysfunction of the transcription factor nuclear factor of the -enhancer in B cells (NF-B), which is essential to epithelial homeostasis and development. A number of mouse models bearing disruptions in NF-B signaling have been reported to manifest defects in ectodermal derivatives. In ectoderm-targeted transgenic mice overexpressing the glucocorticoid receptor (GR) [keratin 5 (K5)-GR mice], the NF-B activity is greatly decreased due to functional antagonism between GR and NF-B. Here, we report that K5-GR mice exhibit multiple epithelial defects in hair follicle, tooth, and palate development. Additionally, these mice lack Meibomian glands and display underdeveloped sweat and preputial glands. These phenotypic features appear to be mediated specifically by ligand-activated GR because the synthetic analog dexamethasone induced similar defects in epithelial morphogenesis, including odontogenesis, in wild-type mice. We have focused on tooth development in K5-GR mice and found that an inhibitor of steroid synthesis partially reversed the abnormal phenotype. Immunostaining revealed reduced expression of the inhibitor of B kinase subunits, IKK and IKK, and diminished p65 protein levels in K5-GR embryonic tooth, resulting in a significantly reduced B-binding activity. Remarkably, altered NF-B activity elicited by GR overexpression correlated with a dramatic decrease in the protein levels of Np63 in tooth epithelia without affecting Akt, BMP4, or Foxo3a. Given that many of the 170 clinically distinct ectodermal dysplasia syndromes still remain without cognate genes, deciphering the molecular mechanisms of this mouse model with epithelial NF-B and p63 dysfunction may provide important clues to understanding the basis of other ectodermal dysplasia syndromes.
Introduction
GLUCOCORTICOIDS (GCs) ARE a vital class of steroid hormones that mediate profound and diverse physiological effects in vertebrate development, metabolism, neurobiology, and programmed cell death (1). Natural GCs and their synthetic analogs function through the GC receptor (GR), a member of the nuclear hormone receptor superfamily of ligand-activated transcription factors. Many morphogenetic processes depend on the precise spatiotemporal expression pattern of GR and its ligand; birth defects are reported after maternal exposure to GCs. However, the precise mechanisms underlying GC teratogenic effects are not fully understood.
Hypohidrotic ectodermal dysplasias (HEDs) are malformation syndromes in humans and mice characterized by severe defects in hair formation (hypotrichosis, partial or total alopecia), abnormal or absent teeth, and hypoplastic or aplastic sweat glands (2, 3). These organs derive from the embryonic ectoderm and thus share similar early morphogenesis (4). It has been shown that, among other common molecular mechanisms, the nuclear factor-B (NF-B) family of transcription factors is required for the normal development of ectodermal-derived tissues (5, 6). NF-B comprises five different proteins in mammals (p50-NF-B1, p52-NF-B2, p65/RelA, c-Rel, and RelB) that can form hetero- or homodimers that are sequestered by cytoplasmic inhibitor of B (IB) proteins. Various cellular stimuli including proinflammatory cytokines, bacterial and viral products, and mitogens activate an IB kinase (IKK) complex to phosphorylate the IBs, thereby triggering their proteasomal degradation and the subsequent release of NF-B (7).
IKK is formed by the catalytic subunits IKK and IKK?, the essential regulatory subunit IKK (or NF-B essential modulator), and the recently cloned regulatory protein ELKS [derived from the relative abundance of its constitutive amino acid glutamic acid (E), leucine (L), lysine (K), and serine (S)] (7, 8, 9). IKK is the bottleneck common to many activation pathways that lead to the nuclear translocation of NF-B. Interestingly, the lack of NF-B essential modulator function causes the disease incontinentia pigmenti (IP) in humans and mice (5, 10, 11). IP is a rare X-linked dominant genodermatosis causing lethality in males and a complex dermatological disease in females that is characterized by abnormal skin pigmentation and is often associated with developmental anomalies of hair, teeth, and eyes.
There are several mouse models for studying ectodermal dysplasia, and all are characterized by defective NF-B signaling. Mice expressing the superrepressor molecule IB display defects in early morphogenesis of hair follicles, exocrine glands, and teeth (12). The mouse mutants tabby, downless, and crinkled exhibit identical HED phenotypes, and they bear mutations in ectodysplasin-A (Eda), a TNF family member ligand, its receptor Edar, and the Edar-associated death domain adapter protein Edaradd, respectively (3). There are two functional isoforms of Eda, Eda-A1 and Eda-A2, which specifically bind to Edar and the related TNF receptor family member X-linked ectodermal dysplasia receptor, respectively. Similar to many other TNF receptor members, downstream responses of Edar and Xedar are mediated by NF-B (3).
We have previously shown that transgenic mice constitutively overexpressing the GR under the control of the keratin 5 (K5) promoter (K5-GR mice) exhibit profound alterations in skin development (13) and severe ocular lesions (14). The use of the K5 promoter drives transgene expression to epithelial cells, closely resembling the spatiotemporal pattern of endogenous K5 expression and, thus, targeting tissues that are derived from the embryonic ectoderm (13). Because GR interferes with NF-B signaling in different cell types including epithelial cells and given that NF-B function is impaired in all reported mouse models of ectodermal dysplasia (ED), we aimed to study the consequences of GR overexpression in the development of ectodermal derivatives.
Here, we report that K5-GR mice exhibit abnormal epithelial morphogenesis of the hair, teeth, and exocrine glands, thus recapitulating the triad of signs in the HED syndrome. Adult mice displayed abnormal hair follicle morphogenesis and defective odontogenesis characterized by microdontia or absence of molar teeth. Orofacial alterations included epithelial palate hypoplasia and cleft palate in some individuals and aplasia/hypoplasia of palatine and tongue salivary glands. K5-GR mice lacked Meibomian glands and exhibited underdeveloped sweat and preputial glands. The observed phenotypic malformations, including defective odontogenesis, appear to be GR-specific because the GC analog dexamethasone (Dex) induced defects in wild-type (wt) mice similar to those caused by GR transgene expression. Moreover, treatment with the GC synthesis inhibitor metyrapone reversed abnormal tooth development in K5-GR embryos. Our studies suggest that reduced expression of IKK, IKK, and p65 during epithelia development and reduced B-binding activity may underlie the ED phenotype of K5-GR transgenic mice. Our data indicate that Np63 expression is dramatically reduced in tooth epithelia and other oral epithelia of K5-GR 18.5-d post conception (dpc) embryos as compared with wt littermates. However, neither Akt activity nor bone morphogenetic protein-4 (BMP4) and forkhead box class O-3a (Foxo3a) signaling seemed to be relevant for the defective odontogenesis found in K5-GR mice, although they seem to play a role in the development of other oral epithelia, thus suggesting that tissue-specific mechanisms are relevant for tooth development. Given that many of the 170 clinically distinct ED syndromes still remain without cognate genes, deciphering the molecular mechanisms of this mouse model with epithelial NF-B and p63 dysfunction may provide important clues to understand the basis of other ED syndromes.
Materials and Methods
Animal handling and treatments
K5-GR hemizygous mice of line 285 (B6D2 mixed genetic background) were previously described (13). Transgenic mice for these studies were generated by mating heterozygous K5-GR mice with wt mice, thus obtaining 50% of transgenic and 50% of nontransgenic progeny. Because littermates are the only appropriate control for direct comparison with the K5-GR transgenic mice, in all experiments, nontransgenic littermates were used as controls.
All the protocols used regarding animal experimentation were approved by the institutional animal care and use committee, in accordance with accepted standards of humane animal care and in compliance with international guidelines.
For experiments evaluating the effects of the Dex on tooth development, wt female mice (B6D2/F2) were ip injected every other day with Dex (Sigma Chemical Co., St. Louis, MO) (1 μg/pregnant dam) or saline starting at 12.5 dpc (the morning of the day that the vaginal plug was seen was considered as 0.5 dpc) until delivery. To evaluate the effect of GR in the absence of endogenous ligand, the GC synthesis inhibitor metyrapone (BIOMOL Research Laboratories, Inc., Plymouth Meeting, PA) was administered in drinking water to pregnant dams (500 μg/ml) (15).
The total number of control and K5-GR transgenic mice examined for histopathological analysis was 178. Of these, 36 were 18.5-dpc embryos, 32 were newborn (postnatal d 0, P0) without treatment, 20 were P0 resulting from Dex treatment, 22 were P0 resulting from metyrapone treatment, and 68 were adult mice.
Antibodies
The primary antibodies used included rabbit polyclonal antibodies directed against GR (sc-1004; Santa Cruz Biotechnology, Inc., Santa Cruz, CA), p65 (sc-372), K5 (Babco, Berkeley, CA), IKK (sc-7182; Santa Cruz Biotechnology, Inc.) and IKK (sc-8330, Santa Cruz Biotechnology, Inc.), and mouse monoclonal antibody directed against IKK? (10AG2, IMGENEX, San Diego, CA). p-Akt (Ser-473) was obtained from Cell Signaling Technology Inc. (Beverly, MA). DNp63 (4A4), BMP4, and Foxo3a were obtained from Santa Cruz Biotechnology, Inc. Mouse monoclonal antirat GR-specific antibody was kindly provided by Sam Okret (Karolinska Institute, Novum, Huddinge, Sweden). The secondary biotin-conjugated antirabbit or antimouse antibodies were from Jackson ImmunoResearch Laboratories, Inc. (West Grove, PA).
Histological and immunohistochemical analysis
Embryos were obtained by cesarean derivation at 18.5 dpc. Tissues from adults, whole embryos, or newborn mice were formalin or ethanol fixed (for histology and immunostaining, respectively) and embedded in paraffin. Consecutive 4-μm-thick sections of the corresponding tissues were obtained and routinely stained with hematoxylin/eosin. For immunostaining, paraffin sections were blocked with 1% BSA and incubated with the primary antibody for at least 1 h. Slides were washed three times with PBS, then incubated with conjugated secondary antibodies for 1 h; the reaction was visualized with the Avidin-Biotin-Complex kit from Dako (Vectastain Elite; Vector Laboratories, Inc., Burlingame, CA) using 3,3-diaminobenzidine in PBS as chromogenic substrate for peroxidase. Control immunostaining using the secondary antibody in the absence of the primary antibody was routinely performed. Slides were mounted and analyzed by light microscopy, and microphotographs were taken at the indicated magnification.
Quantitative immunohistochemistry
Chromogen quantification was performed calculating the cumulative signal strength, or so-called energy (intensity), of the digital file representing any portion or an image as previously described (15). By means of an algorithm, we have determined the absolute amount of antibody-specific chromogen per pixel for any cellular region. In brief, digital images were acquired at 1000x using an Olympus E-20p digital camera attached to an Olympus BX61 microscope system (Olympus, Melville, NY). Chromogen abundance was quantified by calculating the cumulative signal strength within 10 randomly selected areas (25 x 25 pixels). The same first upper molar was used in all the assessments by placing the windows over ameloblast or preameloblast cell layer in both wt and transgenic mice. Thus, chromogen/pixel can be quantified by subtracting the energy of the control slide (i.e. not exposed to primary antibody) from that in the homologous regions of the experimental slide (i.e. slide exposed to primary antibody) for each antibody used. Statistical analysis was performed using SPSS for Windows version 11.5 (SPSS Inc., Chicago, IL). Values were reported as arbitrary energy units per pixel. The paired Student’s t test was used to compare the data for the wt and transgenic mice, with the level of significance set at 5% (P 0.05).
Morphometric studies
Quantitation of alopecia in adult transgenic mice was performed by measurements of the number of hair follicles per millimeter on 4-μm paraffin sections stained with hematoxylin-eosin. To count the number of hair follicles, samples were obtained from the back skin of mice at 5 wk of age during the anagen growing phase of the second postnatal hair cycle (n = 7, four transgenics, three wt) and at 20 wk of age, during the resting telogen phase (n = 8, four transgenics, four wt). Five representative skin sections from each mouse were analyzed. The number of hair follicles per millimeter was determined by counting with an ocular micrometer; only those hair follicles with at least one third of their length in the section were counted (magnification x100, 30 fields per time point). Statistical analyses were performed with the two-tailed unpaired Student’s t test using the statistical program Systat 11.0 (SPSS Inc.).
Immunoblotting
Whole-cell protein extracts were prepared as previously described (7), separated by SDS-PAGE, and transferred to Hybond membrane (Amersham Biosciences Inc., Piscataway, NJ). Membranes were stained with Ponceau S (Sigma) to verify equal protein loading and transfer. Bands were visualized using Amersham ECL reagent and Hyperfilm (Amersham Biosciences Inc.). Experiments were performed by duplicate, and the chemiluminescence of indicated bands was quantitated using a phosphor imager (Bio-Rad) and Image Gauge program (Fuji Photo Film USA, Inc., Edison, NJ). The phosphor imager values for nontransgenic samples (GR protein levels in wt mouse tooth epithelium) were arbitrarily set as 1, and transgenic protein levels were expressed as relative to control values.
EMSAs
Tooth epithelium from the upper mandible was dissected from five nontransgenic and 10 K5-GR 18.5-dpc transgenic embryos and pooled. Whole-cell extracts were prepared, protein concentration was determined, and 7 μg was used for each lane. EMSA was performed by incubating whole-cell extracts with a labeled oligonucleotide corresponding to a palindromic B site as previously described (7). The sequence of the B oligonucleotide coding strand was 5'-GATCCAACGGCAGGGGAATTCCCCTCTCCTTA-3'. The composition of the retarded complexes was determined by supershift experiments, as reported (7). Experiments were performed by duplicate and quantitated using a phosphor imager. The phosphor imager values for control samples (basal B binding in wt mouse tooth epithelium) were arbitrarily set as 1, and transgenic B binding was expressed as relative to control values.
Results
We aimed to study the consequences of GR overexpression in the morphogenesis of organs that develop as appendages of embryonic ectoderm, including hair, teeth, and exocrine glands. We have used K5-GR transgenic mice as a model system because the K5 promoter drives transgene expression to epithelial cells, thus targeting tissues that are derived from the embryonic ectoderm (13). Adult K5-GR transgenic mice exhibited abnormal hair follicle morphogenesis and diffuse alopecia along the whole coat (Fig. 1). Histological examination of these mice showed many atrophic hair follicles (Fig. 1B, arrowhead) and orphan sebaceous glands (Fig. 1B, arrows) compared with wt mouse littermates (Fig. 1A). The severity of alopecia was assessed by quantitation of hair follicles, which revealed an average decrease of 50% in adult transgenic mice compared with wt mice (Fig. 1C, graphic bars). With age, some transgenic mice developed severe alopecia throughout the face (Fig. 1D). However, K5-GR mice exhibited the four hair types, monotrich, awl, auchene, and zigzag, similar to wt mice (data not shown). Histological sections of the whisker pad and face of these mice showed atrophy of vibrissae (Fig. 1F, arrowhead) and of pelage hair follicles, which were all replaced by hypertrophic sebaceous glands (Fig. 1F, arrows). Some transgenic mice developed moderate pigmentary incontinence that could be easily detected as heavily pigmented areas of the skin in tail (Fig. 1, G and H) and foot pads (Fig. 1, I and J).
FIG. 1. Abnormal hair follicle morphogenesis in K5-GR transgenic mice. A and B, Hematoxylin/eosin-stained sections of back skin demonstrate reduced number of hair follicles in K5-GR skin (B) compared with wt mice (A). Note atrophic hair follicle (arrowhead) and orphan sebaceous glands (arrows) in transgenic skin (B). C, Histomorphometric analysis of skin sections: measurements of number of hair follicles per millimeter in 5- and 20-wk-old wt (light bars) and transgenic (tg) K5-GR (dark bars) mice. *, P < 0.05, statistically significant. D, Severe patchy alopecia observed in elder K5-GR mice. E and F, Sections of whisker pad in elder wt (E) and K5-GR (F) mice. Absence of hair follicles (arrows) and severe atrophy of the vibrissae (arrowhead) in K5-GR mice (F). G and H, Tip of tail. I and J, Footpads. Note pigmentary incontinence in tail and footpads of K5-GR mice (H and J) in comparison with the wt littermates (G and I). Scale bars in A and B, 100 μm; E and F, 200 μm; G to J, 500 μm.
We next analyzed whether ectoderm-targeted overexpression of GR elicited malformation of other ectodermal-derived tissues. K5-GR adult mice displayed alterations in the oral cavity (Fig. 2; data not shown), with 62% of mice (n = 45) presenting oligodontia and/or microdontia of the molar teeth, with the absence of the third and, sometimes, second molar, representing the most common abnormalities (Fig. 2, A and B). In addition, we note that minor salivary glands of the tongue in transgenic mice were underdeveloped (serous glands) or absent (mucous glands) in contrast with the wt littermates (Fig. 2, C and D).
FIG. 2. Oral phenotype in K5-GR adult mice. A and B, Frontal view of the upper jaw in wt (A) and K5-GR (B) adult mice. Note bilateral absence of the second and third molars in the transgenic (B). C and D, Hematoxylin-eosin-stained sections of the tongue in wt (C) and K5-GR mice (D). The transgenic tongue shows a decreased amount of serous glands and lack of mucous glands. Scale bars in A and B, 500 μm; C and D, 200 μm.
Because K5-GR mice exhibited defects in hair follicle and tooth formation, we evaluated whether GR overexpression caused morphogenetic defects in exocrine glands, including sweat, preputial, and Meibomian glands (Figs. 3 and 4). Previous reports demonstrated that GR is endogenously expressed in the sweat glands (16), and we found underdeveloped sweat glands in K5-GR mice (Fig. 3). The number of sweat glands in transgenic foot pads was dramatically reduced compared with wt mice (Fig. 3, A and B). Ductal epithelial cells and myoepithelial cells around the sweat glands acini were targeted by the transgene because both were positive for K5 expression (Fig. 3, C, C', D, and D'). In the transgenics, the scanty sweat gland acini present in the foot pads showed disorganization of secretory cells and underdevelopment of myoepithelial cells (Fig. 3, C' and D').
FIG. 3. Sweat and preputial glands are underdeveloped in K5-GR mice. A to D', Sections of foot pads in wt (A, C, and C') and K5-GR (B, D, and D') mice. A and B, Hematoxylin-eosin-stained section to show severe decrease of eccrine sweat glands (arrowheads) and ducts (arrows) in the dermis of the foot pad in the transgenic in comparison with the wt. C and D', Expression of K5 in the basal layer of epidermis and ducts (C and D) as well as in myoepithelial cells around the secretory acini (C' and D'). Note disorganization of epithelial secretory cells and decrease of myoepithelial cells in the scanty sweat glands present in the transgenic (D'). E to I, Preputial glands; in K5-GR mice these glands are approximately half the size of glands in the wt (E). Immunostaining using an anti-K5 antibody shows K5 expression in myoepithelial cells around secretory acini of wt (F) and K5-GR mice (G). Note delayed differentiation of reserve cells to sebocytes (arrows) and interstitial sclerosis (asterisks) in the transgenic. Immunostaining using an anti-GR antibody demonstrates that GR is cytoplasmic in the wt (H) and both cytoplasmic and nuclear in transgenic mice (I). Scale bars in A, B, F, and G, 200 μm; C and D, 100 μm; C' and D', 25 μm; H and I, 50 μm.
FIG. 4. Meibomian glands are absent in K5-GR mice. A to D, Sections of the lower eyelids in wt (A and C) and K5-GR mice (B and D) to show K5 expression in myoepithelial cells of Meibomian glands located between the conjunctival epithelium (c) and the orbicular muscle (m, muscle extension is marked by a dotted line). Note absence of Meibomian glands in the transgenic eyelid. Scale bars in A and B, 200 μm; C and D, 100 μm.
Preputial glands are identical in appearance and function to sebaceous glands that open into the preputial space, and GR overexpression correlated with underdevelopment of these glands in adult transgenic males compared with control age-matched males (Fig. 3E). Ductal epithelial cells and myoepithelial cells around the secretory acini were positive to K5 (Fig. 3, F and G), and, thus, confirmed transgene expression. K5-GR mice showed not only underdeveloped glands but also delayed maturation of sebocytes (Fig. 3, F and G, arrows) and sclerosis of interstitial conjunctive tissue (Fig. 3G, asterisks). Altered differentiation of these glands correlated with a strong expression of the GR transgene in the nuclei of the myoepithelial and secretory cells (Fig. 3, H and I) that produced a disorganized acinar pattern in comparison with the control littermates (Fig. 3, C and D).
Adult transgenic eyelids lacked the tarsal plate containing Meibomian glands that could be seen in the wt between the conjunctival epithelium and the orbicular muscle (Fig. 4, A and B, arrows). wt littermates showed a discontinuous pattern of K5 expression that specifically marked the myoepithelial cells around the Meibomian acini (Fig. 4C), whereas immunostaining of transgenic mouse eye showed no K5 expression in the region underlying conjunctival epithelium, due to the absence of glands (Fig. 4D).
Consistent with the observed maldevelopment of ectodermal epithelia in K5-GR adults, both palatogenesis and odontogenesis were dramatically altered in transgenic newborns (Figs. 5 and 6). Histopathological analysis of newborn transgenic mice revealed hypoplasia of palate epithelium, occasionally consisting of a single layer of cells, as compared with the well-differentiated, stratified squamous epithelium in the controls (Fig. 5, A and B, arrows). Both the dorsal tongue and the palate are targets for transgene expression. Immunostaining with an anti-GR antibody demonstrated cytoplasmic localization of endogenous GR in wt mice (Fig. 5C), whereas in the transgenics, altered differentiation of the oral epithelia correlated with a strong expression of the GR transgene in the nuclei of the basal layers (Fig. 5D, arrows). It is known that administration of GC hormones during pregnancy induces cleft palate, an inadequate fusion of the palatal shelves, in the offspring (17). We have detected cleft palate in 13.3% of transgenic newborns (n = 32); the cleft palate should be completely closed at this developmental stage, as occurs in control mice (Fig. 5, E and F; data not shown).
FIG. 5. Impaired palatogenesis in newborn K5-GR mice. A and B, Hematoxylin-eosin section of palate epithelium of wt (A) and K5-GR (B) newborn mice to show underdevelopment and vacuolation of epithelial cells in the transgenic (arrows). C and D, Expression of GR in the epithelium of the palate and tongue in the wt (C) is cytoplasmic but in the transgenic (D) is both cytoplasmic and nuclear. E and F, Frontal view of the upper jaw in newborn wt (E) and K5-GR (F) mice to show cleft palate in the transgenic. Scale bars in A and B, 100 μm; C and D, 200 μm; E and F, 500 μm.
FIG. 6. Impaired odontogenesis in newborn K5-GR mice. A, C, G, and I, Hematoxylin-eosin transversal section of the mouth at the level of the first molars. B, B', D, D', H, H', J, and J', Higher magnification of the upper right first molar of the corresponding left picture. Note delayed differentiation of dental primordia in K5-GR newborn (C, D, and D') in comparison with the wt (A, B, and B'). E and F, GR expression in wt (E) and transgenic (F) newborn; impaired odontogenesis in the transgenic correlates with nuclear expression of GR in the internal epithelial cells that are poorly differentiated to ameloblasts. G, H, and H', in utero exposure of wt to Dex induced a delay in the differentiation of the dental primordia (compare H and H' with B and B'). I, J, and J', in utero exposure to metyrapone-induced reversion of the observed phenotype in the transgenics (compare I, J, and J' with C, D, and D'). Scale bars in A, C, G, and I, 0.5 mm; B, D to F, H, and J, 100 μm; B', D', H', and J', molar epithelium at x600 magnification.
In addition, odontogenesis was also profoundly altered in K5-GR newborn mice, with a marked delay in tooth formation as well as abnormal differentiation of the epithelial cells in tooth primordia (Fig. 6 and Table 1). We obtained head transversal sections at the level of the first molars to comparatively analyze tooth development in transgenic and control littermates. In newborn control mice, the molars appeared at the late bell stage of tooth development (Fig. 6A and Table 1), with the internal dental epithelial cells differentiated in columnar ameloblasts that secrete enamel matrix and the dental papilla cells differentiated into odontoblasts that give rise to predentine (Fig. 6, B and B'). In transgenic littermates, a marked delay in tooth development was noted because 50% of the molars showed either the early or late cap stage of tooth development (Fig. 6C; Table 1), where internal epithelial cells appeared cuboidal to slightly columnar with a central nucleus, and dental papilla cells were not yet differentiated (Fig. 6, D and D'; compare B' and D'). Altered differentiation of the teeth correlated with a strong expression of the GR transgene in the nuclei of the internal and external dental epithelial cells, as detected by an antibody specific for rat GR (Fig. 6, E and F). To further confirm that the reported defective odontogenesis was specifically elicited by GR overexpression, we performed experiments in which wt mice were intrauterinely exposed to Dex from 12.5 dpc to birth, thus mimicking the developmental onset of GR transgene expression (13). Remarkably, Dex induced a marked delay in the differentiation of internal epithelial cells and dental papilla cells to ameloblasts and odontoblasts, respectively, in newborn mice (Fig. 6, G, H, and H'; note in Table 1 that 100% of the teeth analyzed were at the early bell stage of development). In addition, when K5-GR mice were treated with the steroid synthesis inhibitor metyrapone, we observed reversion of the tooth phenotype, as demonstrated by almost normal differentiation and organization of odontoblasts (one or two rows of cuboidal cells) and preameloblasts (Fig. 6, I, J, and J'; note in Table 1 that 28.57 and 33.33% of the teeth analyzed were at the early and late stages, respectively, of molar development). Metyrapone treatment of nontransgenic mice did not affect molar development whereas Dex treatment of K5-GR mice only slightly altered effects of the transgene, probably reflecting that constitutively active GR in transgenic epithelia is already saturated and does not respond to further addition of exogenous ligand.
TABLE 1. First upper molar stage development found in the different groups of newborn mice
Given that NF-B signaling plays a crucial role in the normal development of ectodermal-derived organs (6), we next examined whether delayed differentiation of teeth correlated with defective expression of IKK/NF-B family members. We performed immunohistochemistry on first molar sections of 18.5-dpc embryos using antibodies against IKK, IKK?, IKK, and p65 (Fig. 7). IKK? expression was absent in teeth of both wt and transgenics at this developmental stage (data not shown). In wt embryos, the internal dental epithelial cells appeared completely differentiated to columnar ameloblasts that highly expressed IKK, p65, and IKK (Fig. 7, A, C, and E, respectively). Our previous results demonstrated that GR overexpression correlates with reduced levels of IKK and, thus, with diminished IKK activity and NF-B binding activity (18). Accordingly, we found a significant reduction in the expression of IKK in the medial region of the tooth primordial, coinciding with an undifferentiated ameloblast layer of K5-GR 18.5-dpc embryos (Fig. 7B, arrows). The same regions also displayed diminished expression of p65 and IKK (Fig. 7, D and F, arrows). These data have been quantitatively assessed by a previously reported method that allows calculating the cumulative signal strength of antibody-specific chromogen per pixel for any cellular region (15). Values are reported as arbitrary energy units per pixel, with the level of significance set at 5% (P 0.05). Quantitative immunostaining shows that IKK, p65, and IKK expression was reduced by 3.8-, 1.5-, and 3.4-fold, respectively, in the ameloblast layer of K5-GR 18.5-dpc embryos (Fig. 7G). To further assess the functional relevance of altered IKK/NF-B expression due to GR overexpression, we performed EMSAs by using whole-cell extracts obtained from tooth epithelia of K5-GR 18.5-dpc embryos and nontransgenic littermates (Fig. 7H). Immunoblotting using a specific anti-GR antibody showed that GR protein levels in transgenic tooth epithelia were 2.5- to 3.7-fold increased in transgenics compared with wt littermates. GR overexpression elicited a significant decrease of the B-binding activity composed of p50/p65 complexes (Fig. 7H; data not shown), indicating a defective NF-B function in tooth epithelium of K5-GR transgenic mice.
FIG. 7. Defective IKK/NK-B expression and activity during odontogenesis in K5-GR 18.5-dpc embryos. A to F, Sections of the lower first molar in wt (A, C, and E) and transgenic (Tg/TG; B, D, and F) to show reduced expression of IKK, p65, and IKK in the medial region of the dental primordia in the transgenics. The reduced expression of IKK, p65, and IKK correlated with a more severely delayed differentiation of internal epithelial cells to ameloblasts (arrows in B, D, and F, respectively). Scale bars in A to F, 100 μm. G, IKK, IKK and p65 chromogen quantity found in wt and K5-GR 18.5-dpc transgenic embryos first upper molar, expressed in terms of arbitrary energy units per pixel. Chromogen amount was determined using the entire ameloblasts or preameloblasts cytoplasmic region. *, Significant differences (P < 0.01). H, Upper panel, Immunoblotting of whole-cell extracts obtained from tooth epithelia of K5-GR 18.5-dpc transgenic embryos and nontransgenic littermates with an anti-GR antibody. Nonspecific band is shown as a loading control. Lower panel, EMSA showing p50/p65 NF-B complexes in tooth epithelia of K5-GR 18.5-dpc transgenic embryos and nontransgenic littermates. The assay was done by using the same extracts as in the upper panel and a consensus B oligonucleotide-labeled probe.
We aimed to elucidate further mechanisms by which GR altered signaling induces epithelial dysmorphogenesis. Because phenotypical alterations found in K5-GR mice resemble those described in ED patients, and given the reported involvement of p63 in these disorders (19), we studied the expression of Np63 in K5-GR transgenic mice. We found that Np63 expression in the ameloblast layer of K5-GR 18.5-dpc embryos was almost completely abolished as compared with wt littermates (Fig. 8, left). This dramatic decrease of Np63 expression was not exclusive to the tooth epithelia but also found in the epithelial mucosa as well as in the tongue and in palate (Fig. 8, left and right). We next investigated the possible mechanism responsible for the decreased Np63 expression and focused on Akt because this kinase induces Np63 (20) and is repressed by GR (18). As expected, immunostaining using a specific anti-Akt-P antibody showed reduced Akt phosphorylation in K5-GR oral mucosa (Fig. 9, upper panel). However, no significant phosphorylated Akt expression was observed in the ameloblast layer, suggesting that the interaction between GR and Akt reported in other epithelia is not responsible for the observed tooth defects in K5-GR embryos. In search of additional mechanisms underlying tooth epithelial defects, we examined BMP4 expression by immunostaining because Np63 is a transcriptional target of Bmp signaling in other model systems (21), and BMP4 signaling has been shown to modulate tooth development (22, 23). We found decreased BMP4 expression in oral epithelia, but it does not seem to affect the ameloblast layer (Fig. 9, middle panel), as it occurs with Akt expression. Finally, we also studied the expression of the transcription factor Foxo3a (Fig. 9, lower panel). Foxo3a is modulated by Akt but, interestingly, also by IKK through an Akt-independent manner (24, 25). We observed a mild similar Foxo3a expression in the ameloblast layer of both transgenic and wt littermates (Fig. 9, lower panel, insets), which reinforces the lack of Akt activity in this epithelium and indicates that Foxo3a is not responsible for the observed decrease expression of Np63. In view of these data, one might postulate that the effect of GR on Np63 expression is due to transcriptional repression. We are currently exploring whether the effect of GR on Np63 expression is mediated through transcriptional repression.
FIG. 8. GR-targeted overexpression dramatically reduces Np63 expression in tooth epithelia and other oral epithelia of K5-GR 18.5-dpc embryos. Sections of the lower first molar in control and transgenic littermates were immunostained with an antibody against Np63. A reduced expression of Np63 was found in tooth epithelia and in other oral epithelia, including palate and tongue.
FIG. 9. Phosphorylated Akt (Akt-P), BMP4, and Foxo3a are down-regulated in oral epithelia of K5-GR 18.5-dpc embryos but unchanged in tooth epithelia of K5-GR 18.5-dpc embryos. Sections of the lower first molar in control and transgenic littermates were immunostained with specific antibodies for Akt-P (ser473), BMP4, and Foxo3a.
Discussion
The human syndrome HED is the most common form of ED, with an incidence of 1 per 10,000 to 1 per 100,000 live births. Of the 170 EDs described to date, fewer than 30 have been explained at the molecular level by identification of the modifying gene (2). In the current report, we present a detailed study of the development of ectoderm-derived tissues in K5-GR transgenic mice including hair, palate, teeth, and exocrine glands. The use of the keratin 5 promoter to drive the expression of the GR transgene targets every tissue in the developing ectoderm (13); however, we consistently found that some ectoderm-derived tissues are more severely affected than others. This may reflect graded tissue sensitivity with the most affected tissues being the classical targets of GCs, and thus, most sensitive to an increase in GR signaling (26). Our data show that overexpression of GR elicited abnormal morphogenesis of hair follicle and tooth, consistent with previous results indicating the crucial role of GR in the development of these ectodermal derivatives. Classical reports have demonstrated that injection of Dex in mice induces retardation of hair growth, hair loss, and alopecia (27). In agreement with these observations, augmented GR levels caused reduced number of hair follicles and abnormal hair follicle morphogenesis along with localized IP in K5-GR adult mice (Fig. 1).
Tooth bud primordia have been reported to express high GR mRNA levels, thus implying a relevant role of GR in tooth formation (17). The fact that Dex treatment of wt mice mimicked the abnormal differentiation program of tooth epithelia observed in K5-GR mice (Fig. 6) argues in favor of a direct role of GR. Furthermore, inhibition of GC synthesis, through the use of metyrapone, was able to reverse the abnormal tooth morphogenesis in K5-GR mice (Fig. 6, compare C and D and I and J; Table 1). Additionally, overexpression of GR elicited striking abnormalities in the oral cavity, including cleft palate (Fig. 5), which agrees with the reported role of GR in the regulation of regional growth and differentiation during mouse palatogenesis. Interestingly, cleft palate is induced by pharmacological doses of GCs in humans and rodents (17); thus, our data strongly suggest a direct role for high levels of nuclear GR in secondary palate formation, which is an epithelium-governed process.
Surprisingly, ocular anomalies were one of the most consistent defects found in K5-GR mice (14). The absence of Meibomian glands is characteristic of some ED diseases (28, 29). In addition, corneal defects and other severe ocular anomalies have been described in some patients with ED (30), consistent with the effect of GR overexpression in ocular epithelia.
A number of transgenic mice have been reported to exhibit a phenotype resembling HED, thus underscoring the relevance of the targeted molecules in the development of ectodermal tissues. A common feature of all these mouse models, including our K5-GR mice, is defective NF-B signaling (13, 18). Constitutively blocked NF-B activity resulting from the overexpression of a superrepressor IB molecule in transgenic mice produced ectodermal maldevelopment (12), and mutations in serines S32, S36 of IB have been found in HED patients (31). Although GCs have been reported to transcriptionally increase IB in certain cell types (32), we did not detect up-regulation of IB in the skin of K5-GR mice (13). However, the NF-B activity was greatly impaired in these mice, at least partially, through reduced IKK protein levels (18). IKK-deficient mice exhibit malformation of ectodermal derivatives (10, 11); interestingly, mutations in the carboxy-terminal domain of IKK have been described in HED patients (33).
In an attempt to understand the molecular mechanisms by which overexpressed GR induced HED, we evaluated the expression pattern of some IKK/NF-B proteins by immunolocalization studies (Fig. 7). Our data showed that altered tooth differentiation in K5-GR mice correlated with a reduced expression of IKK and p65, thus indicating that GR interferes with the NF-B function at multiple levels. We also detected dysregulated levels of IKK in the teeth of K5-GR mice. Of all the reported mouse models for ED, we noticed striking similarities between the phenotype of K5-GR transgenic mice and that of IKK-deficient mice. Impaired ocular development (14) and cleft palate of K5-GR mice (Fig. 5) were not previously reported in most ED models. However, these features have been reported in IKK-deficient mice, which exhibit cleft palate and underdeveloped eyelids due to poor differentiation of the epithelium of the cornea and conjunctiva (34, 35). The precise mechanisms by which IKK exerts its developmental and morphogenetic functions are not yet known. Some developmental problems of IKK-deficient mice, including ocular defects, might be due to impairments in the canonical or alternate pathways activating NF-B (36, 37). However, in IKK-deficient mice, the abnormal epidermal morphogenesis is mediated through the production of a soluble factor that induces keratinocyte differentiation (37), whereas abnormal tooth development is mediated by Notch/Wnt/Shh signaling pathways (38). Neither of these pathways requires NF-B. Thus, we cannot exclude the possibility that the observed morphogenetic defects effecting epithelial tissues in K5-GR mice reflect contributions of pathways independent of NF-B function.
To identify additional defective developmental signals produced by dysregulated GR expression, we have explored additional signaling pathways that may be altered in K5-GR mouse model. Because p63 has been reported to be involved in ED disorders (19), specifically in the disease ectrodactyly ectodermal dysplasia-cleft palate, and given that some phenotypical alterations found in K5-GR mice, such as cleft palate, resemble those described in EEC patients, we studied the expression of Np63 in the K5-GR mice. Our studies indicate that Np63 expression in the ameloblast layer of K5-GR 18.5-dpc embryos was almost completely abolished as compared with wt littermates. This was not exclusive to the tooth epithelia but also found in the epithelial mucosa as well as in the tongue and in palate (Fig. 8). Despite Np63 and Foxo3a being regulated by the PI3K/Akt pathway (20, 24), we did not find that Akt activity or Foxo3a signaling seemed to be relevant for the defective odontogenesis found in K5-GR mice. Similarly, we did not detect altered BMP4 expression despite its reported function during tooth development (22). However, Akt, BMP, and Foxo3a signaling pathways seemed to play a role in the development of other oral epithelia, thus suggesting that tissue-specific mechanisms are relevant for tooth development. We are currently exploring additional hypothesis to explain a direct effect of GR on Np63 expression.
Acknowledgments
We thank A. Ramírez and D. Burks for critical reading of the manuscript. We acknowledge personnel at the Animal Facility of the Instituto de Biomedicina de Valencia for animal care.
References
Yudt MR, Cidlowski JA 2002 The glucocorticoid receptor: coding a diversity of proteins and responses through a single gene. Mol Endocrinol 16:1719–1726
Lamartine J 2003 Towards a new classification of ectodermal dysplasias. Clin Exp Dermatol 28:351–355
Mikkola ML, Thesleff I 2003 Ectodysplasin signaling in development. Cytokine Growth Factor Rev 14:211–224
Thesleff I, Vaahtokari A, Partanen AM 1995 Regulation of organogenesis. Common molecular mechanisms regulating the development of teeth and other organs. Int J Dev Biol 39:35–50
Aradhya S, Nelson DL 2001 NF-B signaling and human disease. Curr Opin Genet Dev 11:300–306
Smahi A, Courtois G, Rabia SH, Doffinger R, Bodemer C, Munnich A, Casanova JL, Israel A 2002 The NF-B signalling pathway in human diseases: from incontinentia pigmenti to ectodermal dysplasias and immune-deficiency syndromes. Hum Mol Genet 11:2371–2375
Karin M, Yamamoto Y, May Wang Q 2004 The IKK/NF-B system: a treasure trove for drug development. Nat Rev 3:17–26
Yamamoto Y, Gaynor RB 2004 IB kinases: key regulators of the NF-B pathway. Trends Biochem Sci 29:72–79
Ducut Sigala JL, Bottero V, Young DB, Shevchenko A, Mercurio F, Verma I M 2004 Activation of transcription factor NF-B requires ELKS, an IB kinase regulatory subunit. Science 304:1963–1967
Makris C, Godfrey VL, Kr?hn-Senftleben G, Takahashi T, Roberts JL, Schwarz T, Feng L, Johnson RS, Karin M 2000 Female mice heterozygous for IKK/NEMO deficiencies develop a dermopathy similar to the human X-linked disorder incontinentia pigmenti. Mol Cell 5:969–979
Schmidt-Supprian M, Bloch W, Courtois G, Addick K, Israel A, Rajewsky K, Pasparakis M 2000 NEMO/IKK-deficient mice model incontinentia pigmenti. Mol Cell 5:981–992
Schmidt-Ullrich R, Aebischer T, Hülsken J, Birchmeier W, Klemm U, Scheidereit C 2001 Requirement of NF-B/Rel for the development of hair follicles and other epidermal appendices. Development 128:3843–3853
Pérez P, Page A, Bravo A, del Río M, Gimenez-Conti I, Budunova I, Slaga TJ, Jorcano JL 2001 Altered skin development and impaired proliferative and inflammatory responses in transgenic mice overexpressing the glucocorticoid receptor. FASEB J 15:2030–2032
Cascallana JL, Bravo A, Page A, Budunova I, Slaga TJ, Jorcano JL, Pérez, P 2003 Disruption of eyelid and cornea development by targeted overexpression of the glucocorticoid receptor. Int J Dev Biol 47:59–64
Matkowskyj KA, Schonfeld D, Benya RV 2000 Quantitative immunohistochemistry by measuring cumulative signal strength using commercially available software Photoshop and MatLab. J Histochem Cytochem 48:303–312
Smith JT, Waddell BJ 2000 Increased fetal glucocorticoid exposure delays puberty onset in postnatal life. Endocrinology 141:2422–2428
Abbott BD, McNabb FMA, Lau C 1994 Glucocorticoid receptor expression during the development of the embryonic mouse secondary palate. Craniofac Genet Dev Biol 14:87–96
Leis H, Page A, Ramírez A, Bravo A, Segrelles C, Paramio J, Barettino D, Jorcano JL, Pérez P 2004 Glucocorticoid receptor counteracts tumorigenic activity of Akt in skin through interference with the phosphatidylinositol 3-kinase (PI3K) signaling pathway. Mol Endocrinol 18:303–311
Little NA, Jochemsen AG 2002 Molecules in focus: p63. Int J Biochem Cell Biol 34:6–9
Barbieri CE, Barton CE, Pietenpol JA 2003 Delta Np63 expression is regulated by the phosphoinositide 3-kinase pathway. J Biol Chem 278:51408–51414
Bakkers J, Hild M, Kramer C, Furutani-Seiki M, Hammerschmidt M 2002 Zebrafish DeltaNp63 is a direct target of Bmp signaling and encodes a transcription repressor blocking neural specification in the ventral ectoderm. Dev Cell 2:617–627
Tucker AS, Matthews KL, Sharpe PT 1998 Transformation of tooth type induced by inhibition of BMP signaling. Science 282:1136–1138
Ruiz S, Segrelles C, Bravo A, et al 2003 Abnormal epidermal differentiation and impaired epithelial-mesenchymal tissue interactions in mice lacking the retinoblastoma relatives p107 and p130. Development 130:2341–2353
Kops GJ, de Ruiter ND, De Vries-Smits AM, Powell DR, Bos JL, Burgering BM 1999 Direct control of the Forkhead transcription factor AFX by protein kinase B. Nature 398:630–634
Hu MC, Lee DF, Xia W, et al 2004 IB kinase promotes tumorigenesis through inhibition of forkhead FOXO3a. Cell 117:225–237
Bamberger CM, Schulte HM, Chrousos GP 1996 Molecular determinants of GR function and tissue sensitivity to glucocorticoids. Endocr Rev 17:245–261
Stenn KS, Paus R, Dutton T, Sarba B 1993 Glucocorticoid effect on hair growth initiation: a reconsideration. Skin Pharmacol 6:125–134
Mondino BJ, Bath PE, Foos RY, Apt L, Bajacich GM 1984 Absent meibomian glands in the ectrodactyly, ectodermal dysplasia, cleft lip-palate syndrome. Am J Ophthalmol 97:469–500
Bonnard E, Logan P, Eustace P 1996 Absent meibomian glands: a marker for EEC syndrome. Eye 10:355–361
Donahue JP, Shea CJ, Taravella MJ 1999 Hidrotic ectodermal dysplasia with corneal involvement. J AAPOS 3:372–375
Courtois G, Smahi A, Reichenbach J, D?ffinger R, Cancrini C, Bonnet M, Puel A, Chable-Bessia C, Yamaoka S, Feinberg J, Dupuis-Girod S, Bodemer C, Livadiotti S, Novelli F, Rossi P, Fischer A, Isra?l A, Munnich A, Le Deist F, Casanova JL 2003 A hypermorphic IB mutation is associated with autosomal dominant anhidrotic ectodermal dysplasia and T cell immunodeficiency. J Clin Invest 112:1108–1115
De Bosscher K, Vanden Berghe W, Haegeman G 2003 The interplay between the glucocorticoid receptor and nuclear factor-B or activator protein-1: molecular mechanisms for gene repression. Endocr Rev 24:488–522
Orange JS, Geha RS 2003 Finding NEMO: genetic disorders of NF-B activation. J Clin Invest 112:983–985
Li Q, Lu Q, Hwang JY, Buscher D, Lee KF, Izpisua-Belmonte JC, Verma IM 1999 IKK1-deficient mice exhibit abnormal development of skin and skeleton. Genes Dev 13:1322–1328
Yoshida K, Hu Y, Karin M 2000 IB Kinase is essential for development of the mammalian cornea and conjunctiva. Invest Ophtalmol Vis Sci 41:3665–3669
Senftleben U, Cao Y, Xiao G, Greten FR, Krahn G, Bonizzi G, Chen Y, Hu Y, Fong A, Sun S, Karin M 2001 Activation by IKK of a second, evolutionary conserved, NF-kB signaling pathway. Science 293:1495–1499
Hu Y, Baud V, Oga T, Kim K, Yoshida K, Karin M 2001 IKK controls formation of the epidermis independently of NF-B. Nature 410:710–714
Ohazama A, Hu Y, Schmidt-Ullrich R, Cao Y, Scheidereit C, Karin M, Sharpe PT 2004 A dual role for IKK in tooth development. Dev Cell 6:219–227(Jose Luis Cascallana, Ana)
Address all correspondence and requests for reprints to: Paloma Pérez, Laboratory of Animal Models, Genomics and Pharmacoproteomics Program, Fundación Valenciana de Investigaciones Biomédicas, Avenida Autopista del Saler 16, Camino de las Moreras, E-46013 Valencia, Spain. E-mail: pperez@ochoa.fib.es.
Abstract
Hypohidrotic ectodermal dysplasia is a human syndrome defined by maldevelopment of one or more ectodermal-derived tissues, including the epidermis and cutaneous appendices, teeth, and exocrine glands. The molecular bases of this pathology converge in a dysfunction of the transcription factor nuclear factor of the -enhancer in B cells (NF-B), which is essential to epithelial homeostasis and development. A number of mouse models bearing disruptions in NF-B signaling have been reported to manifest defects in ectodermal derivatives. In ectoderm-targeted transgenic mice overexpressing the glucocorticoid receptor (GR) [keratin 5 (K5)-GR mice], the NF-B activity is greatly decreased due to functional antagonism between GR and NF-B. Here, we report that K5-GR mice exhibit multiple epithelial defects in hair follicle, tooth, and palate development. Additionally, these mice lack Meibomian glands and display underdeveloped sweat and preputial glands. These phenotypic features appear to be mediated specifically by ligand-activated GR because the synthetic analog dexamethasone induced similar defects in epithelial morphogenesis, including odontogenesis, in wild-type mice. We have focused on tooth development in K5-GR mice and found that an inhibitor of steroid synthesis partially reversed the abnormal phenotype. Immunostaining revealed reduced expression of the inhibitor of B kinase subunits, IKK and IKK, and diminished p65 protein levels in K5-GR embryonic tooth, resulting in a significantly reduced B-binding activity. Remarkably, altered NF-B activity elicited by GR overexpression correlated with a dramatic decrease in the protein levels of Np63 in tooth epithelia without affecting Akt, BMP4, or Foxo3a. Given that many of the 170 clinically distinct ectodermal dysplasia syndromes still remain without cognate genes, deciphering the molecular mechanisms of this mouse model with epithelial NF-B and p63 dysfunction may provide important clues to understanding the basis of other ectodermal dysplasia syndromes.
Introduction
GLUCOCORTICOIDS (GCs) ARE a vital class of steroid hormones that mediate profound and diverse physiological effects in vertebrate development, metabolism, neurobiology, and programmed cell death (1). Natural GCs and their synthetic analogs function through the GC receptor (GR), a member of the nuclear hormone receptor superfamily of ligand-activated transcription factors. Many morphogenetic processes depend on the precise spatiotemporal expression pattern of GR and its ligand; birth defects are reported after maternal exposure to GCs. However, the precise mechanisms underlying GC teratogenic effects are not fully understood.
Hypohidrotic ectodermal dysplasias (HEDs) are malformation syndromes in humans and mice characterized by severe defects in hair formation (hypotrichosis, partial or total alopecia), abnormal or absent teeth, and hypoplastic or aplastic sweat glands (2, 3). These organs derive from the embryonic ectoderm and thus share similar early morphogenesis (4). It has been shown that, among other common molecular mechanisms, the nuclear factor-B (NF-B) family of transcription factors is required for the normal development of ectodermal-derived tissues (5, 6). NF-B comprises five different proteins in mammals (p50-NF-B1, p52-NF-B2, p65/RelA, c-Rel, and RelB) that can form hetero- or homodimers that are sequestered by cytoplasmic inhibitor of B (IB) proteins. Various cellular stimuli including proinflammatory cytokines, bacterial and viral products, and mitogens activate an IB kinase (IKK) complex to phosphorylate the IBs, thereby triggering their proteasomal degradation and the subsequent release of NF-B (7).
IKK is formed by the catalytic subunits IKK and IKK?, the essential regulatory subunit IKK (or NF-B essential modulator), and the recently cloned regulatory protein ELKS [derived from the relative abundance of its constitutive amino acid glutamic acid (E), leucine (L), lysine (K), and serine (S)] (7, 8, 9). IKK is the bottleneck common to many activation pathways that lead to the nuclear translocation of NF-B. Interestingly, the lack of NF-B essential modulator function causes the disease incontinentia pigmenti (IP) in humans and mice (5, 10, 11). IP is a rare X-linked dominant genodermatosis causing lethality in males and a complex dermatological disease in females that is characterized by abnormal skin pigmentation and is often associated with developmental anomalies of hair, teeth, and eyes.
There are several mouse models for studying ectodermal dysplasia, and all are characterized by defective NF-B signaling. Mice expressing the superrepressor molecule IB display defects in early morphogenesis of hair follicles, exocrine glands, and teeth (12). The mouse mutants tabby, downless, and crinkled exhibit identical HED phenotypes, and they bear mutations in ectodysplasin-A (Eda), a TNF family member ligand, its receptor Edar, and the Edar-associated death domain adapter protein Edaradd, respectively (3). There are two functional isoforms of Eda, Eda-A1 and Eda-A2, which specifically bind to Edar and the related TNF receptor family member X-linked ectodermal dysplasia receptor, respectively. Similar to many other TNF receptor members, downstream responses of Edar and Xedar are mediated by NF-B (3).
We have previously shown that transgenic mice constitutively overexpressing the GR under the control of the keratin 5 (K5) promoter (K5-GR mice) exhibit profound alterations in skin development (13) and severe ocular lesions (14). The use of the K5 promoter drives transgene expression to epithelial cells, closely resembling the spatiotemporal pattern of endogenous K5 expression and, thus, targeting tissues that are derived from the embryonic ectoderm (13). Because GR interferes with NF-B signaling in different cell types including epithelial cells and given that NF-B function is impaired in all reported mouse models of ectodermal dysplasia (ED), we aimed to study the consequences of GR overexpression in the development of ectodermal derivatives.
Here, we report that K5-GR mice exhibit abnormal epithelial morphogenesis of the hair, teeth, and exocrine glands, thus recapitulating the triad of signs in the HED syndrome. Adult mice displayed abnormal hair follicle morphogenesis and defective odontogenesis characterized by microdontia or absence of molar teeth. Orofacial alterations included epithelial palate hypoplasia and cleft palate in some individuals and aplasia/hypoplasia of palatine and tongue salivary glands. K5-GR mice lacked Meibomian glands and exhibited underdeveloped sweat and preputial glands. The observed phenotypic malformations, including defective odontogenesis, appear to be GR-specific because the GC analog dexamethasone (Dex) induced defects in wild-type (wt) mice similar to those caused by GR transgene expression. Moreover, treatment with the GC synthesis inhibitor metyrapone reversed abnormal tooth development in K5-GR embryos. Our studies suggest that reduced expression of IKK, IKK, and p65 during epithelia development and reduced B-binding activity may underlie the ED phenotype of K5-GR transgenic mice. Our data indicate that Np63 expression is dramatically reduced in tooth epithelia and other oral epithelia of K5-GR 18.5-d post conception (dpc) embryos as compared with wt littermates. However, neither Akt activity nor bone morphogenetic protein-4 (BMP4) and forkhead box class O-3a (Foxo3a) signaling seemed to be relevant for the defective odontogenesis found in K5-GR mice, although they seem to play a role in the development of other oral epithelia, thus suggesting that tissue-specific mechanisms are relevant for tooth development. Given that many of the 170 clinically distinct ED syndromes still remain without cognate genes, deciphering the molecular mechanisms of this mouse model with epithelial NF-B and p63 dysfunction may provide important clues to understand the basis of other ED syndromes.
Materials and Methods
Animal handling and treatments
K5-GR hemizygous mice of line 285 (B6D2 mixed genetic background) were previously described (13). Transgenic mice for these studies were generated by mating heterozygous K5-GR mice with wt mice, thus obtaining 50% of transgenic and 50% of nontransgenic progeny. Because littermates are the only appropriate control for direct comparison with the K5-GR transgenic mice, in all experiments, nontransgenic littermates were used as controls.
All the protocols used regarding animal experimentation were approved by the institutional animal care and use committee, in accordance with accepted standards of humane animal care and in compliance with international guidelines.
For experiments evaluating the effects of the Dex on tooth development, wt female mice (B6D2/F2) were ip injected every other day with Dex (Sigma Chemical Co., St. Louis, MO) (1 μg/pregnant dam) or saline starting at 12.5 dpc (the morning of the day that the vaginal plug was seen was considered as 0.5 dpc) until delivery. To evaluate the effect of GR in the absence of endogenous ligand, the GC synthesis inhibitor metyrapone (BIOMOL Research Laboratories, Inc., Plymouth Meeting, PA) was administered in drinking water to pregnant dams (500 μg/ml) (15).
The total number of control and K5-GR transgenic mice examined for histopathological analysis was 178. Of these, 36 were 18.5-dpc embryos, 32 were newborn (postnatal d 0, P0) without treatment, 20 were P0 resulting from Dex treatment, 22 were P0 resulting from metyrapone treatment, and 68 were adult mice.
Antibodies
The primary antibodies used included rabbit polyclonal antibodies directed against GR (sc-1004; Santa Cruz Biotechnology, Inc., Santa Cruz, CA), p65 (sc-372), K5 (Babco, Berkeley, CA), IKK (sc-7182; Santa Cruz Biotechnology, Inc.) and IKK (sc-8330, Santa Cruz Biotechnology, Inc.), and mouse monoclonal antibody directed against IKK? (10AG2, IMGENEX, San Diego, CA). p-Akt (Ser-473) was obtained from Cell Signaling Technology Inc. (Beverly, MA). DNp63 (4A4), BMP4, and Foxo3a were obtained from Santa Cruz Biotechnology, Inc. Mouse monoclonal antirat GR-specific antibody was kindly provided by Sam Okret (Karolinska Institute, Novum, Huddinge, Sweden). The secondary biotin-conjugated antirabbit or antimouse antibodies were from Jackson ImmunoResearch Laboratories, Inc. (West Grove, PA).
Histological and immunohistochemical analysis
Embryos were obtained by cesarean derivation at 18.5 dpc. Tissues from adults, whole embryos, or newborn mice were formalin or ethanol fixed (for histology and immunostaining, respectively) and embedded in paraffin. Consecutive 4-μm-thick sections of the corresponding tissues were obtained and routinely stained with hematoxylin/eosin. For immunostaining, paraffin sections were blocked with 1% BSA and incubated with the primary antibody for at least 1 h. Slides were washed three times with PBS, then incubated with conjugated secondary antibodies for 1 h; the reaction was visualized with the Avidin-Biotin-Complex kit from Dako (Vectastain Elite; Vector Laboratories, Inc., Burlingame, CA) using 3,3-diaminobenzidine in PBS as chromogenic substrate for peroxidase. Control immunostaining using the secondary antibody in the absence of the primary antibody was routinely performed. Slides were mounted and analyzed by light microscopy, and microphotographs were taken at the indicated magnification.
Quantitative immunohistochemistry
Chromogen quantification was performed calculating the cumulative signal strength, or so-called energy (intensity), of the digital file representing any portion or an image as previously described (15). By means of an algorithm, we have determined the absolute amount of antibody-specific chromogen per pixel for any cellular region. In brief, digital images were acquired at 1000x using an Olympus E-20p digital camera attached to an Olympus BX61 microscope system (Olympus, Melville, NY). Chromogen abundance was quantified by calculating the cumulative signal strength within 10 randomly selected areas (25 x 25 pixels). The same first upper molar was used in all the assessments by placing the windows over ameloblast or preameloblast cell layer in both wt and transgenic mice. Thus, chromogen/pixel can be quantified by subtracting the energy of the control slide (i.e. not exposed to primary antibody) from that in the homologous regions of the experimental slide (i.e. slide exposed to primary antibody) for each antibody used. Statistical analysis was performed using SPSS for Windows version 11.5 (SPSS Inc., Chicago, IL). Values were reported as arbitrary energy units per pixel. The paired Student’s t test was used to compare the data for the wt and transgenic mice, with the level of significance set at 5% (P 0.05).
Morphometric studies
Quantitation of alopecia in adult transgenic mice was performed by measurements of the number of hair follicles per millimeter on 4-μm paraffin sections stained with hematoxylin-eosin. To count the number of hair follicles, samples were obtained from the back skin of mice at 5 wk of age during the anagen growing phase of the second postnatal hair cycle (n = 7, four transgenics, three wt) and at 20 wk of age, during the resting telogen phase (n = 8, four transgenics, four wt). Five representative skin sections from each mouse were analyzed. The number of hair follicles per millimeter was determined by counting with an ocular micrometer; only those hair follicles with at least one third of their length in the section were counted (magnification x100, 30 fields per time point). Statistical analyses were performed with the two-tailed unpaired Student’s t test using the statistical program Systat 11.0 (SPSS Inc.).
Immunoblotting
Whole-cell protein extracts were prepared as previously described (7), separated by SDS-PAGE, and transferred to Hybond membrane (Amersham Biosciences Inc., Piscataway, NJ). Membranes were stained with Ponceau S (Sigma) to verify equal protein loading and transfer. Bands were visualized using Amersham ECL reagent and Hyperfilm (Amersham Biosciences Inc.). Experiments were performed by duplicate, and the chemiluminescence of indicated bands was quantitated using a phosphor imager (Bio-Rad) and Image Gauge program (Fuji Photo Film USA, Inc., Edison, NJ). The phosphor imager values for nontransgenic samples (GR protein levels in wt mouse tooth epithelium) were arbitrarily set as 1, and transgenic protein levels were expressed as relative to control values.
EMSAs
Tooth epithelium from the upper mandible was dissected from five nontransgenic and 10 K5-GR 18.5-dpc transgenic embryos and pooled. Whole-cell extracts were prepared, protein concentration was determined, and 7 μg was used for each lane. EMSA was performed by incubating whole-cell extracts with a labeled oligonucleotide corresponding to a palindromic B site as previously described (7). The sequence of the B oligonucleotide coding strand was 5'-GATCCAACGGCAGGGGAATTCCCCTCTCCTTA-3'. The composition of the retarded complexes was determined by supershift experiments, as reported (7). Experiments were performed by duplicate and quantitated using a phosphor imager. The phosphor imager values for control samples (basal B binding in wt mouse tooth epithelium) were arbitrarily set as 1, and transgenic B binding was expressed as relative to control values.
Results
We aimed to study the consequences of GR overexpression in the morphogenesis of organs that develop as appendages of embryonic ectoderm, including hair, teeth, and exocrine glands. We have used K5-GR transgenic mice as a model system because the K5 promoter drives transgene expression to epithelial cells, thus targeting tissues that are derived from the embryonic ectoderm (13). Adult K5-GR transgenic mice exhibited abnormal hair follicle morphogenesis and diffuse alopecia along the whole coat (Fig. 1). Histological examination of these mice showed many atrophic hair follicles (Fig. 1B, arrowhead) and orphan sebaceous glands (Fig. 1B, arrows) compared with wt mouse littermates (Fig. 1A). The severity of alopecia was assessed by quantitation of hair follicles, which revealed an average decrease of 50% in adult transgenic mice compared with wt mice (Fig. 1C, graphic bars). With age, some transgenic mice developed severe alopecia throughout the face (Fig. 1D). However, K5-GR mice exhibited the four hair types, monotrich, awl, auchene, and zigzag, similar to wt mice (data not shown). Histological sections of the whisker pad and face of these mice showed atrophy of vibrissae (Fig. 1F, arrowhead) and of pelage hair follicles, which were all replaced by hypertrophic sebaceous glands (Fig. 1F, arrows). Some transgenic mice developed moderate pigmentary incontinence that could be easily detected as heavily pigmented areas of the skin in tail (Fig. 1, G and H) and foot pads (Fig. 1, I and J).
FIG. 1. Abnormal hair follicle morphogenesis in K5-GR transgenic mice. A and B, Hematoxylin/eosin-stained sections of back skin demonstrate reduced number of hair follicles in K5-GR skin (B) compared with wt mice (A). Note atrophic hair follicle (arrowhead) and orphan sebaceous glands (arrows) in transgenic skin (B). C, Histomorphometric analysis of skin sections: measurements of number of hair follicles per millimeter in 5- and 20-wk-old wt (light bars) and transgenic (tg) K5-GR (dark bars) mice. *, P < 0.05, statistically significant. D, Severe patchy alopecia observed in elder K5-GR mice. E and F, Sections of whisker pad in elder wt (E) and K5-GR (F) mice. Absence of hair follicles (arrows) and severe atrophy of the vibrissae (arrowhead) in K5-GR mice (F). G and H, Tip of tail. I and J, Footpads. Note pigmentary incontinence in tail and footpads of K5-GR mice (H and J) in comparison with the wt littermates (G and I). Scale bars in A and B, 100 μm; E and F, 200 μm; G to J, 500 μm.
We next analyzed whether ectoderm-targeted overexpression of GR elicited malformation of other ectodermal-derived tissues. K5-GR adult mice displayed alterations in the oral cavity (Fig. 2; data not shown), with 62% of mice (n = 45) presenting oligodontia and/or microdontia of the molar teeth, with the absence of the third and, sometimes, second molar, representing the most common abnormalities (Fig. 2, A and B). In addition, we note that minor salivary glands of the tongue in transgenic mice were underdeveloped (serous glands) or absent (mucous glands) in contrast with the wt littermates (Fig. 2, C and D).
FIG. 2. Oral phenotype in K5-GR adult mice. A and B, Frontal view of the upper jaw in wt (A) and K5-GR (B) adult mice. Note bilateral absence of the second and third molars in the transgenic (B). C and D, Hematoxylin-eosin-stained sections of the tongue in wt (C) and K5-GR mice (D). The transgenic tongue shows a decreased amount of serous glands and lack of mucous glands. Scale bars in A and B, 500 μm; C and D, 200 μm.
Because K5-GR mice exhibited defects in hair follicle and tooth formation, we evaluated whether GR overexpression caused morphogenetic defects in exocrine glands, including sweat, preputial, and Meibomian glands (Figs. 3 and 4). Previous reports demonstrated that GR is endogenously expressed in the sweat glands (16), and we found underdeveloped sweat glands in K5-GR mice (Fig. 3). The number of sweat glands in transgenic foot pads was dramatically reduced compared with wt mice (Fig. 3, A and B). Ductal epithelial cells and myoepithelial cells around the sweat glands acini were targeted by the transgene because both were positive for K5 expression (Fig. 3, C, C', D, and D'). In the transgenics, the scanty sweat gland acini present in the foot pads showed disorganization of secretory cells and underdevelopment of myoepithelial cells (Fig. 3, C' and D').
FIG. 3. Sweat and preputial glands are underdeveloped in K5-GR mice. A to D', Sections of foot pads in wt (A, C, and C') and K5-GR (B, D, and D') mice. A and B, Hematoxylin-eosin-stained section to show severe decrease of eccrine sweat glands (arrowheads) and ducts (arrows) in the dermis of the foot pad in the transgenic in comparison with the wt. C and D', Expression of K5 in the basal layer of epidermis and ducts (C and D) as well as in myoepithelial cells around the secretory acini (C' and D'). Note disorganization of epithelial secretory cells and decrease of myoepithelial cells in the scanty sweat glands present in the transgenic (D'). E to I, Preputial glands; in K5-GR mice these glands are approximately half the size of glands in the wt (E). Immunostaining using an anti-K5 antibody shows K5 expression in myoepithelial cells around secretory acini of wt (F) and K5-GR mice (G). Note delayed differentiation of reserve cells to sebocytes (arrows) and interstitial sclerosis (asterisks) in the transgenic. Immunostaining using an anti-GR antibody demonstrates that GR is cytoplasmic in the wt (H) and both cytoplasmic and nuclear in transgenic mice (I). Scale bars in A, B, F, and G, 200 μm; C and D, 100 μm; C' and D', 25 μm; H and I, 50 μm.
FIG. 4. Meibomian glands are absent in K5-GR mice. A to D, Sections of the lower eyelids in wt (A and C) and K5-GR mice (B and D) to show K5 expression in myoepithelial cells of Meibomian glands located between the conjunctival epithelium (c) and the orbicular muscle (m, muscle extension is marked by a dotted line). Note absence of Meibomian glands in the transgenic eyelid. Scale bars in A and B, 200 μm; C and D, 100 μm.
Preputial glands are identical in appearance and function to sebaceous glands that open into the preputial space, and GR overexpression correlated with underdevelopment of these glands in adult transgenic males compared with control age-matched males (Fig. 3E). Ductal epithelial cells and myoepithelial cells around the secretory acini were positive to K5 (Fig. 3, F and G), and, thus, confirmed transgene expression. K5-GR mice showed not only underdeveloped glands but also delayed maturation of sebocytes (Fig. 3, F and G, arrows) and sclerosis of interstitial conjunctive tissue (Fig. 3G, asterisks). Altered differentiation of these glands correlated with a strong expression of the GR transgene in the nuclei of the myoepithelial and secretory cells (Fig. 3, H and I) that produced a disorganized acinar pattern in comparison with the control littermates (Fig. 3, C and D).
Adult transgenic eyelids lacked the tarsal plate containing Meibomian glands that could be seen in the wt between the conjunctival epithelium and the orbicular muscle (Fig. 4, A and B, arrows). wt littermates showed a discontinuous pattern of K5 expression that specifically marked the myoepithelial cells around the Meibomian acini (Fig. 4C), whereas immunostaining of transgenic mouse eye showed no K5 expression in the region underlying conjunctival epithelium, due to the absence of glands (Fig. 4D).
Consistent with the observed maldevelopment of ectodermal epithelia in K5-GR adults, both palatogenesis and odontogenesis were dramatically altered in transgenic newborns (Figs. 5 and 6). Histopathological analysis of newborn transgenic mice revealed hypoplasia of palate epithelium, occasionally consisting of a single layer of cells, as compared with the well-differentiated, stratified squamous epithelium in the controls (Fig. 5, A and B, arrows). Both the dorsal tongue and the palate are targets for transgene expression. Immunostaining with an anti-GR antibody demonstrated cytoplasmic localization of endogenous GR in wt mice (Fig. 5C), whereas in the transgenics, altered differentiation of the oral epithelia correlated with a strong expression of the GR transgene in the nuclei of the basal layers (Fig. 5D, arrows). It is known that administration of GC hormones during pregnancy induces cleft palate, an inadequate fusion of the palatal shelves, in the offspring (17). We have detected cleft palate in 13.3% of transgenic newborns (n = 32); the cleft palate should be completely closed at this developmental stage, as occurs in control mice (Fig. 5, E and F; data not shown).
FIG. 5. Impaired palatogenesis in newborn K5-GR mice. A and B, Hematoxylin-eosin section of palate epithelium of wt (A) and K5-GR (B) newborn mice to show underdevelopment and vacuolation of epithelial cells in the transgenic (arrows). C and D, Expression of GR in the epithelium of the palate and tongue in the wt (C) is cytoplasmic but in the transgenic (D) is both cytoplasmic and nuclear. E and F, Frontal view of the upper jaw in newborn wt (E) and K5-GR (F) mice to show cleft palate in the transgenic. Scale bars in A and B, 100 μm; C and D, 200 μm; E and F, 500 μm.
FIG. 6. Impaired odontogenesis in newborn K5-GR mice. A, C, G, and I, Hematoxylin-eosin transversal section of the mouth at the level of the first molars. B, B', D, D', H, H', J, and J', Higher magnification of the upper right first molar of the corresponding left picture. Note delayed differentiation of dental primordia in K5-GR newborn (C, D, and D') in comparison with the wt (A, B, and B'). E and F, GR expression in wt (E) and transgenic (F) newborn; impaired odontogenesis in the transgenic correlates with nuclear expression of GR in the internal epithelial cells that are poorly differentiated to ameloblasts. G, H, and H', in utero exposure of wt to Dex induced a delay in the differentiation of the dental primordia (compare H and H' with B and B'). I, J, and J', in utero exposure to metyrapone-induced reversion of the observed phenotype in the transgenics (compare I, J, and J' with C, D, and D'). Scale bars in A, C, G, and I, 0.5 mm; B, D to F, H, and J, 100 μm; B', D', H', and J', molar epithelium at x600 magnification.
In addition, odontogenesis was also profoundly altered in K5-GR newborn mice, with a marked delay in tooth formation as well as abnormal differentiation of the epithelial cells in tooth primordia (Fig. 6 and Table 1). We obtained head transversal sections at the level of the first molars to comparatively analyze tooth development in transgenic and control littermates. In newborn control mice, the molars appeared at the late bell stage of tooth development (Fig. 6A and Table 1), with the internal dental epithelial cells differentiated in columnar ameloblasts that secrete enamel matrix and the dental papilla cells differentiated into odontoblasts that give rise to predentine (Fig. 6, B and B'). In transgenic littermates, a marked delay in tooth development was noted because 50% of the molars showed either the early or late cap stage of tooth development (Fig. 6C; Table 1), where internal epithelial cells appeared cuboidal to slightly columnar with a central nucleus, and dental papilla cells were not yet differentiated (Fig. 6, D and D'; compare B' and D'). Altered differentiation of the teeth correlated with a strong expression of the GR transgene in the nuclei of the internal and external dental epithelial cells, as detected by an antibody specific for rat GR (Fig. 6, E and F). To further confirm that the reported defective odontogenesis was specifically elicited by GR overexpression, we performed experiments in which wt mice were intrauterinely exposed to Dex from 12.5 dpc to birth, thus mimicking the developmental onset of GR transgene expression (13). Remarkably, Dex induced a marked delay in the differentiation of internal epithelial cells and dental papilla cells to ameloblasts and odontoblasts, respectively, in newborn mice (Fig. 6, G, H, and H'; note in Table 1 that 100% of the teeth analyzed were at the early bell stage of development). In addition, when K5-GR mice were treated with the steroid synthesis inhibitor metyrapone, we observed reversion of the tooth phenotype, as demonstrated by almost normal differentiation and organization of odontoblasts (one or two rows of cuboidal cells) and preameloblasts (Fig. 6, I, J, and J'; note in Table 1 that 28.57 and 33.33% of the teeth analyzed were at the early and late stages, respectively, of molar development). Metyrapone treatment of nontransgenic mice did not affect molar development whereas Dex treatment of K5-GR mice only slightly altered effects of the transgene, probably reflecting that constitutively active GR in transgenic epithelia is already saturated and does not respond to further addition of exogenous ligand.
TABLE 1. First upper molar stage development found in the different groups of newborn mice
Given that NF-B signaling plays a crucial role in the normal development of ectodermal-derived organs (6), we next examined whether delayed differentiation of teeth correlated with defective expression of IKK/NF-B family members. We performed immunohistochemistry on first molar sections of 18.5-dpc embryos using antibodies against IKK, IKK?, IKK, and p65 (Fig. 7). IKK? expression was absent in teeth of both wt and transgenics at this developmental stage (data not shown). In wt embryos, the internal dental epithelial cells appeared completely differentiated to columnar ameloblasts that highly expressed IKK, p65, and IKK (Fig. 7, A, C, and E, respectively). Our previous results demonstrated that GR overexpression correlates with reduced levels of IKK and, thus, with diminished IKK activity and NF-B binding activity (18). Accordingly, we found a significant reduction in the expression of IKK in the medial region of the tooth primordial, coinciding with an undifferentiated ameloblast layer of K5-GR 18.5-dpc embryos (Fig. 7B, arrows). The same regions also displayed diminished expression of p65 and IKK (Fig. 7, D and F, arrows). These data have been quantitatively assessed by a previously reported method that allows calculating the cumulative signal strength of antibody-specific chromogen per pixel for any cellular region (15). Values are reported as arbitrary energy units per pixel, with the level of significance set at 5% (P 0.05). Quantitative immunostaining shows that IKK, p65, and IKK expression was reduced by 3.8-, 1.5-, and 3.4-fold, respectively, in the ameloblast layer of K5-GR 18.5-dpc embryos (Fig. 7G). To further assess the functional relevance of altered IKK/NF-B expression due to GR overexpression, we performed EMSAs by using whole-cell extracts obtained from tooth epithelia of K5-GR 18.5-dpc embryos and nontransgenic littermates (Fig. 7H). Immunoblotting using a specific anti-GR antibody showed that GR protein levels in transgenic tooth epithelia were 2.5- to 3.7-fold increased in transgenics compared with wt littermates. GR overexpression elicited a significant decrease of the B-binding activity composed of p50/p65 complexes (Fig. 7H; data not shown), indicating a defective NF-B function in tooth epithelium of K5-GR transgenic mice.
FIG. 7. Defective IKK/NK-B expression and activity during odontogenesis in K5-GR 18.5-dpc embryos. A to F, Sections of the lower first molar in wt (A, C, and E) and transgenic (Tg/TG; B, D, and F) to show reduced expression of IKK, p65, and IKK in the medial region of the dental primordia in the transgenics. The reduced expression of IKK, p65, and IKK correlated with a more severely delayed differentiation of internal epithelial cells to ameloblasts (arrows in B, D, and F, respectively). Scale bars in A to F, 100 μm. G, IKK, IKK and p65 chromogen quantity found in wt and K5-GR 18.5-dpc transgenic embryos first upper molar, expressed in terms of arbitrary energy units per pixel. Chromogen amount was determined using the entire ameloblasts or preameloblasts cytoplasmic region. *, Significant differences (P < 0.01). H, Upper panel, Immunoblotting of whole-cell extracts obtained from tooth epithelia of K5-GR 18.5-dpc transgenic embryos and nontransgenic littermates with an anti-GR antibody. Nonspecific band is shown as a loading control. Lower panel, EMSA showing p50/p65 NF-B complexes in tooth epithelia of K5-GR 18.5-dpc transgenic embryos and nontransgenic littermates. The assay was done by using the same extracts as in the upper panel and a consensus B oligonucleotide-labeled probe.
We aimed to elucidate further mechanisms by which GR altered signaling induces epithelial dysmorphogenesis. Because phenotypical alterations found in K5-GR mice resemble those described in ED patients, and given the reported involvement of p63 in these disorders (19), we studied the expression of Np63 in K5-GR transgenic mice. We found that Np63 expression in the ameloblast layer of K5-GR 18.5-dpc embryos was almost completely abolished as compared with wt littermates (Fig. 8, left). This dramatic decrease of Np63 expression was not exclusive to the tooth epithelia but also found in the epithelial mucosa as well as in the tongue and in palate (Fig. 8, left and right). We next investigated the possible mechanism responsible for the decreased Np63 expression and focused on Akt because this kinase induces Np63 (20) and is repressed by GR (18). As expected, immunostaining using a specific anti-Akt-P antibody showed reduced Akt phosphorylation in K5-GR oral mucosa (Fig. 9, upper panel). However, no significant phosphorylated Akt expression was observed in the ameloblast layer, suggesting that the interaction between GR and Akt reported in other epithelia is not responsible for the observed tooth defects in K5-GR embryos. In search of additional mechanisms underlying tooth epithelial defects, we examined BMP4 expression by immunostaining because Np63 is a transcriptional target of Bmp signaling in other model systems (21), and BMP4 signaling has been shown to modulate tooth development (22, 23). We found decreased BMP4 expression in oral epithelia, but it does not seem to affect the ameloblast layer (Fig. 9, middle panel), as it occurs with Akt expression. Finally, we also studied the expression of the transcription factor Foxo3a (Fig. 9, lower panel). Foxo3a is modulated by Akt but, interestingly, also by IKK through an Akt-independent manner (24, 25). We observed a mild similar Foxo3a expression in the ameloblast layer of both transgenic and wt littermates (Fig. 9, lower panel, insets), which reinforces the lack of Akt activity in this epithelium and indicates that Foxo3a is not responsible for the observed decrease expression of Np63. In view of these data, one might postulate that the effect of GR on Np63 expression is due to transcriptional repression. We are currently exploring whether the effect of GR on Np63 expression is mediated through transcriptional repression.
FIG. 8. GR-targeted overexpression dramatically reduces Np63 expression in tooth epithelia and other oral epithelia of K5-GR 18.5-dpc embryos. Sections of the lower first molar in control and transgenic littermates were immunostained with an antibody against Np63. A reduced expression of Np63 was found in tooth epithelia and in other oral epithelia, including palate and tongue.
FIG. 9. Phosphorylated Akt (Akt-P), BMP4, and Foxo3a are down-regulated in oral epithelia of K5-GR 18.5-dpc embryos but unchanged in tooth epithelia of K5-GR 18.5-dpc embryos. Sections of the lower first molar in control and transgenic littermates were immunostained with specific antibodies for Akt-P (ser473), BMP4, and Foxo3a.
Discussion
The human syndrome HED is the most common form of ED, with an incidence of 1 per 10,000 to 1 per 100,000 live births. Of the 170 EDs described to date, fewer than 30 have been explained at the molecular level by identification of the modifying gene (2). In the current report, we present a detailed study of the development of ectoderm-derived tissues in K5-GR transgenic mice including hair, palate, teeth, and exocrine glands. The use of the keratin 5 promoter to drive the expression of the GR transgene targets every tissue in the developing ectoderm (13); however, we consistently found that some ectoderm-derived tissues are more severely affected than others. This may reflect graded tissue sensitivity with the most affected tissues being the classical targets of GCs, and thus, most sensitive to an increase in GR signaling (26). Our data show that overexpression of GR elicited abnormal morphogenesis of hair follicle and tooth, consistent with previous results indicating the crucial role of GR in the development of these ectodermal derivatives. Classical reports have demonstrated that injection of Dex in mice induces retardation of hair growth, hair loss, and alopecia (27). In agreement with these observations, augmented GR levels caused reduced number of hair follicles and abnormal hair follicle morphogenesis along with localized IP in K5-GR adult mice (Fig. 1).
Tooth bud primordia have been reported to express high GR mRNA levels, thus implying a relevant role of GR in tooth formation (17). The fact that Dex treatment of wt mice mimicked the abnormal differentiation program of tooth epithelia observed in K5-GR mice (Fig. 6) argues in favor of a direct role of GR. Furthermore, inhibition of GC synthesis, through the use of metyrapone, was able to reverse the abnormal tooth morphogenesis in K5-GR mice (Fig. 6, compare C and D and I and J; Table 1). Additionally, overexpression of GR elicited striking abnormalities in the oral cavity, including cleft palate (Fig. 5), which agrees with the reported role of GR in the regulation of regional growth and differentiation during mouse palatogenesis. Interestingly, cleft palate is induced by pharmacological doses of GCs in humans and rodents (17); thus, our data strongly suggest a direct role for high levels of nuclear GR in secondary palate formation, which is an epithelium-governed process.
Surprisingly, ocular anomalies were one of the most consistent defects found in K5-GR mice (14). The absence of Meibomian glands is characteristic of some ED diseases (28, 29). In addition, corneal defects and other severe ocular anomalies have been described in some patients with ED (30), consistent with the effect of GR overexpression in ocular epithelia.
A number of transgenic mice have been reported to exhibit a phenotype resembling HED, thus underscoring the relevance of the targeted molecules in the development of ectodermal tissues. A common feature of all these mouse models, including our K5-GR mice, is defective NF-B signaling (13, 18). Constitutively blocked NF-B activity resulting from the overexpression of a superrepressor IB molecule in transgenic mice produced ectodermal maldevelopment (12), and mutations in serines S32, S36 of IB have been found in HED patients (31). Although GCs have been reported to transcriptionally increase IB in certain cell types (32), we did not detect up-regulation of IB in the skin of K5-GR mice (13). However, the NF-B activity was greatly impaired in these mice, at least partially, through reduced IKK protein levels (18). IKK-deficient mice exhibit malformation of ectodermal derivatives (10, 11); interestingly, mutations in the carboxy-terminal domain of IKK have been described in HED patients (33).
In an attempt to understand the molecular mechanisms by which overexpressed GR induced HED, we evaluated the expression pattern of some IKK/NF-B proteins by immunolocalization studies (Fig. 7). Our data showed that altered tooth differentiation in K5-GR mice correlated with a reduced expression of IKK and p65, thus indicating that GR interferes with the NF-B function at multiple levels. We also detected dysregulated levels of IKK in the teeth of K5-GR mice. Of all the reported mouse models for ED, we noticed striking similarities between the phenotype of K5-GR transgenic mice and that of IKK-deficient mice. Impaired ocular development (14) and cleft palate of K5-GR mice (Fig. 5) were not previously reported in most ED models. However, these features have been reported in IKK-deficient mice, which exhibit cleft palate and underdeveloped eyelids due to poor differentiation of the epithelium of the cornea and conjunctiva (34, 35). The precise mechanisms by which IKK exerts its developmental and morphogenetic functions are not yet known. Some developmental problems of IKK-deficient mice, including ocular defects, might be due to impairments in the canonical or alternate pathways activating NF-B (36, 37). However, in IKK-deficient mice, the abnormal epidermal morphogenesis is mediated through the production of a soluble factor that induces keratinocyte differentiation (37), whereas abnormal tooth development is mediated by Notch/Wnt/Shh signaling pathways (38). Neither of these pathways requires NF-B. Thus, we cannot exclude the possibility that the observed morphogenetic defects effecting epithelial tissues in K5-GR mice reflect contributions of pathways independent of NF-B function.
To identify additional defective developmental signals produced by dysregulated GR expression, we have explored additional signaling pathways that may be altered in K5-GR mouse model. Because p63 has been reported to be involved in ED disorders (19), specifically in the disease ectrodactyly ectodermal dysplasia-cleft palate, and given that some phenotypical alterations found in K5-GR mice, such as cleft palate, resemble those described in EEC patients, we studied the expression of Np63 in the K5-GR mice. Our studies indicate that Np63 expression in the ameloblast layer of K5-GR 18.5-dpc embryos was almost completely abolished as compared with wt littermates. This was not exclusive to the tooth epithelia but also found in the epithelial mucosa as well as in the tongue and in palate (Fig. 8). Despite Np63 and Foxo3a being regulated by the PI3K/Akt pathway (20, 24), we did not find that Akt activity or Foxo3a signaling seemed to be relevant for the defective odontogenesis found in K5-GR mice. Similarly, we did not detect altered BMP4 expression despite its reported function during tooth development (22). However, Akt, BMP, and Foxo3a signaling pathways seemed to play a role in the development of other oral epithelia, thus suggesting that tissue-specific mechanisms are relevant for tooth development. We are currently exploring additional hypothesis to explain a direct effect of GR on Np63 expression.
Acknowledgments
We thank A. Ramírez and D. Burks for critical reading of the manuscript. We acknowledge personnel at the Animal Facility of the Instituto de Biomedicina de Valencia for animal care.
References
Yudt MR, Cidlowski JA 2002 The glucocorticoid receptor: coding a diversity of proteins and responses through a single gene. Mol Endocrinol 16:1719–1726
Lamartine J 2003 Towards a new classification of ectodermal dysplasias. Clin Exp Dermatol 28:351–355
Mikkola ML, Thesleff I 2003 Ectodysplasin signaling in development. Cytokine Growth Factor Rev 14:211–224
Thesleff I, Vaahtokari A, Partanen AM 1995 Regulation of organogenesis. Common molecular mechanisms regulating the development of teeth and other organs. Int J Dev Biol 39:35–50
Aradhya S, Nelson DL 2001 NF-B signaling and human disease. Curr Opin Genet Dev 11:300–306
Smahi A, Courtois G, Rabia SH, Doffinger R, Bodemer C, Munnich A, Casanova JL, Israel A 2002 The NF-B signalling pathway in human diseases: from incontinentia pigmenti to ectodermal dysplasias and immune-deficiency syndromes. Hum Mol Genet 11:2371–2375
Karin M, Yamamoto Y, May Wang Q 2004 The IKK/NF-B system: a treasure trove for drug development. Nat Rev 3:17–26
Yamamoto Y, Gaynor RB 2004 IB kinases: key regulators of the NF-B pathway. Trends Biochem Sci 29:72–79
Ducut Sigala JL, Bottero V, Young DB, Shevchenko A, Mercurio F, Verma I M 2004 Activation of transcription factor NF-B requires ELKS, an IB kinase regulatory subunit. Science 304:1963–1967
Makris C, Godfrey VL, Kr?hn-Senftleben G, Takahashi T, Roberts JL, Schwarz T, Feng L, Johnson RS, Karin M 2000 Female mice heterozygous for IKK/NEMO deficiencies develop a dermopathy similar to the human X-linked disorder incontinentia pigmenti. Mol Cell 5:969–979
Schmidt-Supprian M, Bloch W, Courtois G, Addick K, Israel A, Rajewsky K, Pasparakis M 2000 NEMO/IKK-deficient mice model incontinentia pigmenti. Mol Cell 5:981–992
Schmidt-Ullrich R, Aebischer T, Hülsken J, Birchmeier W, Klemm U, Scheidereit C 2001 Requirement of NF-B/Rel for the development of hair follicles and other epidermal appendices. Development 128:3843–3853
Pérez P, Page A, Bravo A, del Río M, Gimenez-Conti I, Budunova I, Slaga TJ, Jorcano JL 2001 Altered skin development and impaired proliferative and inflammatory responses in transgenic mice overexpressing the glucocorticoid receptor. FASEB J 15:2030–2032
Cascallana JL, Bravo A, Page A, Budunova I, Slaga TJ, Jorcano JL, Pérez, P 2003 Disruption of eyelid and cornea development by targeted overexpression of the glucocorticoid receptor. Int J Dev Biol 47:59–64
Matkowskyj KA, Schonfeld D, Benya RV 2000 Quantitative immunohistochemistry by measuring cumulative signal strength using commercially available software Photoshop and MatLab. J Histochem Cytochem 48:303–312
Smith JT, Waddell BJ 2000 Increased fetal glucocorticoid exposure delays puberty onset in postnatal life. Endocrinology 141:2422–2428
Abbott BD, McNabb FMA, Lau C 1994 Glucocorticoid receptor expression during the development of the embryonic mouse secondary palate. Craniofac Genet Dev Biol 14:87–96
Leis H, Page A, Ramírez A, Bravo A, Segrelles C, Paramio J, Barettino D, Jorcano JL, Pérez P 2004 Glucocorticoid receptor counteracts tumorigenic activity of Akt in skin through interference with the phosphatidylinositol 3-kinase (PI3K) signaling pathway. Mol Endocrinol 18:303–311
Little NA, Jochemsen AG 2002 Molecules in focus: p63. Int J Biochem Cell Biol 34:6–9
Barbieri CE, Barton CE, Pietenpol JA 2003 Delta Np63 expression is regulated by the phosphoinositide 3-kinase pathway. J Biol Chem 278:51408–51414
Bakkers J, Hild M, Kramer C, Furutani-Seiki M, Hammerschmidt M 2002 Zebrafish DeltaNp63 is a direct target of Bmp signaling and encodes a transcription repressor blocking neural specification in the ventral ectoderm. Dev Cell 2:617–627
Tucker AS, Matthews KL, Sharpe PT 1998 Transformation of tooth type induced by inhibition of BMP signaling. Science 282:1136–1138
Ruiz S, Segrelles C, Bravo A, et al 2003 Abnormal epidermal differentiation and impaired epithelial-mesenchymal tissue interactions in mice lacking the retinoblastoma relatives p107 and p130. Development 130:2341–2353
Kops GJ, de Ruiter ND, De Vries-Smits AM, Powell DR, Bos JL, Burgering BM 1999 Direct control of the Forkhead transcription factor AFX by protein kinase B. Nature 398:630–634
Hu MC, Lee DF, Xia W, et al 2004 IB kinase promotes tumorigenesis through inhibition of forkhead FOXO3a. Cell 117:225–237
Bamberger CM, Schulte HM, Chrousos GP 1996 Molecular determinants of GR function and tissue sensitivity to glucocorticoids. Endocr Rev 17:245–261
Stenn KS, Paus R, Dutton T, Sarba B 1993 Glucocorticoid effect on hair growth initiation: a reconsideration. Skin Pharmacol 6:125–134
Mondino BJ, Bath PE, Foos RY, Apt L, Bajacich GM 1984 Absent meibomian glands in the ectrodactyly, ectodermal dysplasia, cleft lip-palate syndrome. Am J Ophthalmol 97:469–500
Bonnard E, Logan P, Eustace P 1996 Absent meibomian glands: a marker for EEC syndrome. Eye 10:355–361
Donahue JP, Shea CJ, Taravella MJ 1999 Hidrotic ectodermal dysplasia with corneal involvement. J AAPOS 3:372–375
Courtois G, Smahi A, Reichenbach J, D?ffinger R, Cancrini C, Bonnet M, Puel A, Chable-Bessia C, Yamaoka S, Feinberg J, Dupuis-Girod S, Bodemer C, Livadiotti S, Novelli F, Rossi P, Fischer A, Isra?l A, Munnich A, Le Deist F, Casanova JL 2003 A hypermorphic IB mutation is associated with autosomal dominant anhidrotic ectodermal dysplasia and T cell immunodeficiency. J Clin Invest 112:1108–1115
De Bosscher K, Vanden Berghe W, Haegeman G 2003 The interplay between the glucocorticoid receptor and nuclear factor-B or activator protein-1: molecular mechanisms for gene repression. Endocr Rev 24:488–522
Orange JS, Geha RS 2003 Finding NEMO: genetic disorders of NF-B activation. J Clin Invest 112:983–985
Li Q, Lu Q, Hwang JY, Buscher D, Lee KF, Izpisua-Belmonte JC, Verma IM 1999 IKK1-deficient mice exhibit abnormal development of skin and skeleton. Genes Dev 13:1322–1328
Yoshida K, Hu Y, Karin M 2000 IB Kinase is essential for development of the mammalian cornea and conjunctiva. Invest Ophtalmol Vis Sci 41:3665–3669
Senftleben U, Cao Y, Xiao G, Greten FR, Krahn G, Bonizzi G, Chen Y, Hu Y, Fong A, Sun S, Karin M 2001 Activation by IKK of a second, evolutionary conserved, NF-kB signaling pathway. Science 293:1495–1499
Hu Y, Baud V, Oga T, Kim K, Yoshida K, Karin M 2001 IKK controls formation of the epidermis independently of NF-B. Nature 410:710–714
Ohazama A, Hu Y, Schmidt-Ullrich R, Cao Y, Scheidereit C, Karin M, Sharpe PT 2004 A dual role for IKK in tooth development. Dev Cell 6:219–227(Jose Luis Cascallana, Ana)