Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Dev Biol. Author manuscript; available in PMC 2009 Jun 15.
Published in final edited form as:
PMCID: PMC2696204

The Edar subfamily in feather placode formation


A subgroup of the TNF receptor family, composed of Edar, Troy and Xedar, are implicated in the development of ectodermal appendages, such as hair follicles, teeth and sweat glands. We have isolated chicken orthologues of these three receptors and analysed their roles in early feather development. Conservation of protein sequences between mammalian and avian proteins is variable, with avian Edar showing the greatest degree of sequence identity. cXedar differs from its mammalian orthologue in that it contains an intracellular death domain. All three receptors are expressed during early feather morphogenesis and dominant negative forms of each receptor impair the epithelial contribution to feather bud morphogenesis, while the dermal contribution appears unaffected. Hyperactivation of each receptor leads to more widespread assumption of placode fate, though in different regions of the skin. Receptor signaling converges on NF-κB, and inhibiting this transcription factor alters feather bud number and size in a stage specific manner. Our findings illustrate the roles of these three receptors during avian skin morphogenesis and also suggest that activators of feather placode fate undergo mutual regulation to reach a decision on skin appendage location and size.

Keywords: TNFR, placode, feather, skin, evolution, death domain, Edar, Troy, Xedar


Skin appendages, such as hairs, feathers, scales and glands, are class defining features in vertebrates. While the skin itself forms a barrier, its appendages play a wide variety of roles, from defence and display to insulation and aerodynamics. During development these cutaneous appendages are laid out in a periodic pattern in the embryonic ectoderm, the cells of which must all choose between appendage and surface keratinocyte fates. The cells that will become appendages first condense to form dense patches, called placodes. After placode specification, cell proliferation generates downgrowths in mammals to produce hair follicles (Hardy, 1992), or outgrowths in birds to produce a feather bud (Lin et al., 2006). The process of appendage formation in all vertebrate classes relies on a series of reciprocal interactions between the epidermis and its underlying dermis (Sengel, 1990; Hardy, 1992; Fuchs et al., 2001; Millar, 2002).

In mouse, hair follicles are generated across the entire surface of the embryo in a series of temporally defined pulses late in gestation, the later forming follicles filling in the gaps that open up between older follicles as the skin grows. In contrast, in avian skin several tracts are formed first; following which a defined morphogenetic wave moves across specific tracts, leaving a very regular array of placodes in its wake (Jiang et al., 2004). Despite the similarities between early hair and feather follicle morphogenesis, they appear to be convergently evolved structures (Wu et al., 2004).

The formation of feather buds takes place in hierarchical levels (Chang et al., 2004). In the first level, feather fields (which later become feather tracts) form from presumptive dermis and ectoderm. In the second level, periodic patterning takes place and the originally homogeneous feather field breaks into individual feather buds and interbud regions. In the subsequent levels, feather buds undergo morphogenetic events to form an anterior-posterior axis and branches (Yu et al., 2004; Lin et al., 2006).

The expression pattern of genes involved in periodic pattern formation can be categorized into two distinct modes, ‘restrictive’ and ‘de novo’, reflecting the successive stages of their functions (Jiang et al., 2004). Molecules with ‘restrictive’ expression are involved in negotiating placode position. They are initially expressed homogenously at a moderate level. As appendage locations are specified, these genes become restricted to or up regulated in the placodes and down regulated in the surrounding regions, or vice versa. β-catenin is an example and is considered to be required for establishing the competence of feather field. Molecules with ‘de novo’ expression, such as Shh, appear directly in the placode once its position has been defined. They serve to regulate bud growth, shaping, axis determination and outgrowth morphogenesis (Jiang et al., 1999; Widelitz et al., 2000). The molecules involved in pattern formation fall into two functional categories; activators and inhibitors of placode fate. Whether direct or indirect, interactions between these inhibitors and activators are responsible for breaking the symmetry of the early skin into the hexagonal feather pattern that emerges (Jiang et al., 2004).

The tumour necrosis factor receptor (TNFR) family is expanded in the vertebrate lineage, with mammalian genomes containing about 30 members (Locksley et al., 2001), compared to a single gene in Drosophila (Kanda et al., 2002). This expansion appears to be correlated with acquisition of roles in vertebrate evolutionary novelties, such as the adaptive immune system, bone, mammary gland, and skin appendages (Locksley et al., 2001). The TNFR family can be subdivided in two ways. Several subfamilies are defined by sequence similarities in the receptors’ extracellular ligand binding domains, with each subfamily being a product of gene duplication and divergence (Locksley et al., 2001). Alternatively, TNFR family members can be allocated to one of two functional classes according to their mode of signaling, which depends on whether they contain an intracellular death domain or not. Eight receptors (Edar, p75 NGFR, TNFR1, Fas, DR3, DR4, DR5, DR6), which are scattered among the subfamilies, contain a C-terminal death domain which is used to recruit cytoplasmic death domain adaptor proteins (Wajant, 2003). These adaptors in turn recruit members of the Traf family to transduce signals, commonly resulting in activation of the transcription factor NF–κB, and sometimes initiating apoptosis. A majority of TNFRs do not contain a death domain and initiate signaling by recruiting Trafs directly to their cytoplasmic tails (Inoue et al., 2000).

Ectodysplasin (Eda) is a member of the TNF family of ligands and was initially implicated in appendage development by the cloning of a gene underlying hypohidrotic ectodermal dysplasia (HED) in mouse and human (Kere et al., 1996; Thesleff and Mikkola, 2002). HED is characterised by agenesis or malformation of ectoderm-derived appendages, such as teeth, sweat glands and hair follicles, while the skin itself develops normally. Positional cloning identified a receptor for Eda, a member of the TNFR superfamily called Edar (Headon and Overbeek, 1999), and a cytoplasmic transducer of Edar signals called Edaradd (Headon et al., 2001; Yan et al., 2002). Like the ligand, Eda, both receptor and adaptor are mutated in mouse and human HED. Mice with genetic lesions that affect this signaling pathway display a phenotype in which primary hair follicles, normally developing between embryonic day 14 (E14) and E16, are entirely absent, while the later developing secondary hair follicles are almost normal (Headon and Overbeek, 1999; Laurikkala et al., 2002). Therefore, the Eda pathway is required specifically for initiation of primary hair follicles, while whiskers and secondary follicles are minimally affected by its absence. A reciprocal phenotype is caused by mutation of the transcription factor Lef-1 or the BMP inhibitor Noggin, which are required to initiate development of secondary, but not primary, hair follicles (van Genderen et al., 1994; Botchkarev et al., 2002; Plikus et al., 2004). Thus two genetically distinct pathways are utilised to activate hair follicle development at different stages of mouse development. Interestingly, Lef-1, Noggin and the Eda pathway components are all expressed in both primary and secondary follicle placodes, and so their expression characteristics do not indicate their functional roles in a given follicle subtype. Though the evolutionary relationships between ectodermal appendages in different vertebrate classes are unclear, a conserved role for Edar signaling in their development is indicated by the finding that its mutation underlies a spontaneous fish mutant that lacks scales (Kondo et al., 2001).

Following identification of Edar, two novel TNFRs, Troy and Xedar, were cloned and found to be expressed in the embryonic epidermis and appendages (Kojima et al., 2000; Yan et al., 2000). The extracellular domains of Edar, Troy and Xedar mark them out as a distinct subfamily within the TNFRs, though their intracellular domains are unrelated to one another (Locksley et al., 2001). Xedar binds to a specific splice variant of Eda, EdaA2, which differs by two amino acids from the Edar binding variant, EdaA1 (Yan et al., 2000). No TNF ligand has been identified for Troy (Bossen et al., 2006). All three receptors employ Trafs, which ultimately leads to activation of NF-κB (Kojima et al., 2000; Yan et al., 2000). However, while Edar contains a death domain used to recruit Edaradd for signal transduction, this domain is entirely absent from the mammalian Troy and Xedar proteins. Recent studies of null mutations in Xedar and Troy have reported an absence of gross skin or appendage phenotypes (Newton et al., 2004; Shao et al., 2005).

Previous work has described the expression of cEda, cEdar and cEdaradd in the forming feather field and showed that ectopic activation of cβ-catenin was sufficient to induce cEdar expression (Houghton et al., 2005). Here we describe the effects of suppression and activation of Edar and its related receptors in developing chicken skin in vitro and in vivo, and the effects of suppression of NF-κB activity. This work defines functional roles for signaling from these receptors, and for NF-κB, in feather development. Our findings also suggest a mutual feedback regulation among activators in the periodic patterning process. It is perhaps the summation of these activities that specify the placode and inter-placode fates, rather than the linear cβ –catenin - cEdar axis proposed in previous studies (Houghton et al., 2005).

Materials and Methods

Isolation of chicken cDNAs

Chicken EST databases (Boardman et al., 2002) were queried using human Eda, Edar, Edaradd, Troy and Xedar amino acid sequences. Available sequences were used to design oligonucleotides for 5′ and 3′ RACE to generate cDNAs covering a full length ORF. RACE ready cDNA was generated from E8 chicken skin using the Generacer kit (Invitrogen). When the first round of RACE did not give discrete products a nested reaction was performed. Oligonucleotides used were:




Sequences for cEdar were not represented in the EST database therefore a partial sequence within the death domain between amino acids 372 and 387 was obtained by degenerate PCR. This sequence was then extended by 5′ and 3′ nested RACE reactions. cEdar5′S: 5′-GTGGTGAAAACGTGGCGTCACCTT-3′; cEdar5′SN: 5′-CTTTGGGCTGAAGAGGGACGAGAT-3′; cEdar3′AS: 5′-ATCTCGTCCCTCTTCAGCCCAAAG-3′; cEdar3′ASN: 5′-CTCTCCGCAAGGTGACGCCACGTT-3′

Novel sequences have been submitted to GenBank (cTroy-long DQ360499, cTroy-short DQ360500, cXedar DQ360501)


Epidermis was separated from dermis using dispase and RNA isolated from each tissue using TRI reagent (Sigma). cDNA was synthesised using random primers (Invitrogen) and AMV reverse transcriptase (Roche). The cEda RT-PCR was run on a 4% agarose gel to distinguish the A1 and A2 forms, which differ in size by 6 nucleotides. For quantitative RT-PCR cDNA was generated from 3 individual embryos for each age examined. Quantification of 18S rRNA was used to normalize Eda expression levels. For each cDNA sample three 20 μl reactions containing SYBR green (Molecular Probes) were run on an Opticon 2 thermal cycler (MJ Research) under the following conditions: 95°C, 10′ (X1) 95°C, 15″; 59°C, 30″; 72°C, 30″ (X40). The following primers were used for amplification:















Whole mount in situ hybridisation

Whole mount in situ hybridisations were carried out on White Leghorn or scaleless mutant (University of Connecticut, Storrs, Connecticut) embryos using digoxigenin-labelled riboprobes as described (Riddle et al., 1993), with a Proteinase K treatment of 5 μg/ml for 5 minutes. The templates used to generate riboprobes were: cβ-catenin: codons 1–127; cEda: 80 nt of ORF and 333 nt of 3′UTR; cEdar: entire ORF and 43 nt of 5′ and 3′UTRs; cTroy: codons 17–257; cXedar: 202 nt of 5′UTR and codons 1–337. To identify feather placodes embryos were stained with 0.02% ethidium bromide as described (Eames and Schneider, 2005). For detection of viral proteins after in situ hybridization stained embryos were processed for cryosectioning by snap-freezing tissue in OCT (Sakura Tissue-Tek) in liquid nitrogen. Immunostaining was performed on 16 μm cryosections using the anti-gag antibody AMV3C2 (D. Boettiger, Developmental Studies Hybridoma Bank, University of Iowa) at a 1:750 dilution followed by antibody detection using the Vectastain ABC kit (Vector) and SigmaFAST 3,3′-Diaminobenzidine.

Western blotting

E14 wild type, EdaTa/Ta and Edaraddcr/cr mouse, and E8 chicken dorsal, skins were dissected, homogenized in TRI reagent and proteins isolated. Samples were run on an 18% SDS-PAGE gel and transferred to a nitrocellulose membrane. Rabbit anti-Eda antibody (AL166, produced by immunisation using mouse Eda amino acids 245–391, Schneider et al., 2001) was used at 1/1000, anti-mouse HRP secondary at 1/10000 (Amersham) and mouse monoclonal anti-βActin-horseradish peroxidase at 1/25000 (AC-15, Sigma). Signal was detected using SuperSignal West Pico Chemiluminescent Substrate (Pierce).

Generation of proviral constructs

Dominant negative receptors were generated by replacing intracellular signaling domains with enhanced green fluorescent protein (GFP, amplified from pEGFP, Clontech). DN-cEdar consisted of amino acids 1–292, DN-cTroy amino acids 1–200, and DN-cXedar amino acids 1–167.

Constitutively active receptors were generated by fusing a cDNA encoding the six transmembrane domains of the Epstein Barr viral LMP1 protein (codons 1–188) to the receptors’ intracellular domains (cEdar codons 213–448, cTroy codons 193–409, cXedar codons 161–504). The LMP1 cDNA was amplified from a pSG5:LMP1 template (from Bill Sugden, University of Wisconsin).

Fused cDNA PCR products were cloned into the pCR8/GW/TOPO Gateway entry vector (Invitrogen) and sequenced. An LR recombination reaction was performed to transfer the cDNAs to a Gateway compatible RCASBP (A) vector (Loftus et al., 2001).

Preparation of RCAS virus

DF-1 cells (from C. Tickle, University of Dundee) were cultured in DMEM/10% FBS and transfected with proviral plasmid DNA using Lipofectamine 2000 (Invitrogen). When infected cells reached confluence the medium was replaced with a minimal volume of DMEM/1%FBS, which was collected after 24 hours and filtered through a 0.45 μm filter. The filtrate was used directly for in ovo injections.

NF-κB reporter assay

RCAS infected DF-1 cells were cultured for 7 days and then cotransfected with a 10:1 molar ratio of the pNF-κBluc reporter plasmid (Clontech) and the pRL-TK (Renilla luciferase) plasmid (Promega). Luciferase activity was measured 18 hours after transfection using the Dual Luciferase Reporter Assay System (Promega).

Viral transduction of reconstituted skin explants

Reconstitution assays were performed as described (Jiang et al., 1999). Dorsal skins from stage 31 white leghorn embryos were dissected in HBSS and incubated in 2X calcium/magnesium free medium with 0.25% EDTA. Epithelium and mesenchyme were separated and the mesenchyme was incubated in 0.1% collagenase/trypsin for 10 mins at 37°C to prepare a single cell suspension. Cells were pelleted by centrifugation and resuspended in virus containing medium for one hour. Reconstituted explants were made by plating mesenchymal cells at high density on tissue culture inserts and then placing the intact epithelium back on top of the mesenchyme. The reconstituted explants were cultured in DMEM/10% FCS and incubated at 37°C in 5% CO2.

In ovo injection

Conditioned media from RCAS infected cultures were used to inject stage 20 embyos in ovo. Approximately 5 μl of viral suspension was delivered between the amniotic membrane and the dorsal ectoderm of the embryos between the wing and limb buds. Embryos were harvested at E8, fixed in 4% PFA in PBS overnight at 4°C, then processed for in situ hybridisation. Numbers of embryos analysed were: CA-cEdar (n=35), CA-cTroy (n=31), CA-cXedar (n=14), DN-cEdar (n=16; 7 with bud suppression phenotype = 44%), DN-cTroy (n=15; 6 with bud suppression phenotype = 40%), DN-cXedar (n=19; 5 with bud suppression phenotype = 26%), GFP control (n=12) and LMP1 control (n=12). The presence of virus was monitored by detection of GFP positive foci or by antibody staining to detect the viral gag protein.

NF-κB inhibition

Dorsal skin was dissected from E7 embryos, placed on a culture insert and treated with 1 μl/ml DMSO (control) or 100 μM NF-κB inhibitor BAY 11–7082 (Sigma Aldrich) dissolved in DMSO. Explants were photographed and processed for sectioning and in situ hybridization.


Cloning and expression of the chicken Edar subfamily

We used the sequences of human Eda, Edar, Troy and Xedar to query chicken EST sequence databases and verified full length ORFs by 5′ and 3′ RACE. This approach identified a single orthologue for each component of the mammalian Edar signaling pathway and its related receptors (Fig. 1A). The predicted extracellular domains of Edar, Troy and Xedar were all well conserved between chicken and human (Edar 84%, Troy 79%, Xedar 65%), while the degree of conservation of their intracellular domains was much more variable (Edar 85%, Troy 48%, Xedar 14%). We identified two forms of cTroy which differ in the length of their intracellular domains (217 versus 66 cytoplasmic amino acids), comparable to the long and short Troy isoforms described in mammalian genomes (223 versus 21 cytoplasmic amino acids) (Kojima et al., 2000). Troy-short may represent an endogenous dominant negative form or a receptor with a distinct signaling output.

Fig. 1
The chicken Edar subfamily and its expression during skin development. (A) Schematic of cEdar, cXedar, cTroy-long and cTroy-short proteins. Sequence similarities (% amino acid identity) between domains are indicated. The 13 C-terminal amino acids of cTroy-short ...

As in mammals, the extracellular domains of cTroy and cXedar are more closely related to one another than either is to cEdar. Also, cEdar and cTroy, and cTroy and cXedar, have no detectable sequence similarity in their intracellular domains. In contrast to the mammalian proteins, however, cXedar contains within its last exon a region that is similar to the death domain of cEdar. Conservation of key amino acid residues between this region of cXedar and the death domains human EDAR and p75 NGFR, as well as its location in the protein, confirm that this is a death domain (Fig. 1B). Thus between avian and mammalian lineages the Edar sequence and domain structure is highly conserved, the Troy protein moderately so, and Xedar is highly diverged.

We examined the expression pattern of each of these genes during early feather development. RT-PCR at incubation day 8 (E8) detected cEdar and cTroy specifically in the epidermis, while cXedar was present in both epidermis and dermis (Fig. 1C). Whole mount in situ hybridisation revealed that cEdar and cTroy have similar expression characteristics. Both genes are expressed in the primary row of the spinal tract at E6 (Fig. 1D,E) and at E7 are expressed in the placodes and the morphogenetic wave (Fig. 1F,G). Thus cEdar and cTroy both undergo the restrictive mode of expression, from homogenous staining in the morphogenetic wave to upregulation in the placodes. In the humeral tracts at E7 the two receptors are expressed in placodes (Fig. 1H,I), but cEda mRNA is detected in the interplacode region (Fig. 1J), as previously reported (Houghton et al., 2005). In the scaleless mutant fowl, which almost completely lacks the morphological and molecular indications of feather development, cEdar, cTroy and cEda are all expressed diffusely (Fig. 1K–M), as previously described for other genes expressed in the restrictive mode (Widelitz et al., 2000). In E8 humeral tracts cEdar and cTroy, but not cXedar, were expressed in a punctate expression pattern (Fig. 1N–P), but by E10 all three receptors are upregulated in the outgrowing short buds (Fig. 1Q–S). cEdar and cTroy are also expressed in foot scales (Fig. 1T,U), with cEda transcript in a reciprocal pattern in the interscale region (Fig. 1V).

The sequence conservation and expression characteristics of Edar, Troy and Xedar between mammals and birds suggest that they may play similar, yet distinct, roles in skin development across amniote lineages.

Eda splice variants and protein solubilisation

Eda is produced in two functionally distinct splice forms; the EdaA1 protein specifically activates Edar, while EdaA2, which lacks 2 amino acids present in the A1 form, activates Xedar (Yan et al., 2000). We performed RT-PCR using oligos adjacent to this alternative splice site to estimate the relative abundance of each isoform during development. Both isoforms are expressed in epidermis and dermis, with the EdaA1:EdaA2 ratio essentially constant through development (Fig. 2A). Thus in this system alternative splicing of Eda appears to be a constitutive event, with no evidence of temporal or spatial regulation. Quantitative RT-PCR detected higher levels of total Eda transcript in the dermis than in the epidermis at E7 and E10 (Fig. 2B). Eda is synthesised as a transmembrane protein, but cell culture studies have demonstrated that it can be released into the medium by furin mediated cleavage at an extracellular site (Schneider et al., 2001). To determine whether this cleavage occurs in vivo, we first validated an anti-Eda polyclonal antibody by detecting Eda protein in extracts of wild type and Edaraddcr/cr embryonic mouse skin. EdaTa/Ta samples were used as a negative control to confirm the specificity of the antibody (Fig. 2C) as the genetic lesion in this line removes the Eda promoter and first exon (Srivastava et al., 1997). Immunodetection of proteins from developing chicken and mouse skin indicated that Eda is present in both species in the cleaved soluble form (Fig. 2D). As the cEda transcript is detected in the interbud domain, with cEdar expressed in the placodes (Fig. 1H,J), this solubilisation of cEda would allow stimulation of cEdar signaling within the placode.

Fig. 2
Analysis of the Eda isoforms in feather tract development and detection of solubilised Eda protein in vivo. (A) Detection of cEdaA1 and cEdaA2 in embryonic epidermis and dermis by RT-PCR. The ratio of A1:A2 shows little variation through feather development ...

Reagents to manipulate receptor signaling

To examine the functions of cEdar, cTroy and cXedar in feather bud development we engineered dominant negative (DN) and constitutively active (CA) forms of each protein (Fig. 3A). To suppress receptor signaling we replaced intracellular signaling domains with GFP. These truncated proteins should sequester wild type receptors into nonfunctional complexes, analogous to the mode of action of the spontaneous EdarSleek dominant allele in mouse (Headon and Overbeek, 1999). GFP tagging allows visualisation of sites of viral infection and transgene expression.

Fig. 3
Design and evaluation of reagents to manipulate receptor signaling. (A) Constitutively active (CA) cEdar, cTroy and cXedar were generated by fusing their intracellular domains (ICD) to the transmembrane domains of LMP1. Dominant negative (DN) receptors ...

Activation of receptor signaling could be achieved by overexpressing the native receptors, though with this approach signaling may be influenced by ligand availability. Therefore, we engineered chimeric proteins consisting of the six transmembrane domains of latent membrane protein 1 (LMP1) fused to the receptors’ intracellular domains (Fig. 3A). LMP1 is an Epstein–Barr virus encoded protein that self-aggregates in the cell membrane and mimics a constitutively active TNFR (Hatzivassiliou et al., 1998). Thus the chimeric receptors are predicted to display ligand independent signaling.

We tested the NF-κB activation capabilities of the fusion proteins by an NF-κB reporter assay in chicken DF-1 cells. Fusion of receptor intracellular domains to LMP1 activated NF-κB in the case of cEdar and cTroy, while the cXedar fusion protein induced only weak, though statistically significant, activation. Dominant negative receptors displayed no NF-κB activation above the basal level for GFP virus infected cells, confirming ablation of key signaling domains (Fig. 3B).

Modulating receptor signaling in reconstituted skin

We first tested the effects of the cEdar and cTroy on feather bud formation in an ex vivo assay. Dissociation of embryonic skin removes placode positional information and allows manipulation of periodic patterning from its initial stages. Also, it allows efficient viral infection in culture. Upon reaggregation, the skin goes on to produce an array of outgrowing feather buds (Jiang et al., 1999). Transduction of reconstituted cultures with RCAS carrying DN-cEdar (n=12) or DN-cTroy (n=8) inhibited feather bud outgrowth when compared to the effects of control virus at day 2 (Fig. 4A–E). In the transduced explants placode formation was initiated and dermal condensations were visible (Fig. 4E, arrowheads), however, bud outgrowth and development to the short bud stage was not sustained. After 5 days in culture there were no short buds formed while in control explants short buds had elongated to the long bud stage (data not shown). These data suggest therefore that the earliest placode specification stage appears to occur with suppressed cEdar or cTroy signaling, however, placodes fail to progress beyond this stage and after 5 days in culture retain only dermal placode components (Fig. 4E).

Fig. 4
Manipulation of signaling in skin reconstitution assays. Feather buds after two days in culture. (A) Reconstituted skin transduced with control RCAS virus. Explants transduced with (B) DN-cEdar or (C) DN-cTroy exhibit an overall suppression of bud outgrowth, ...

Introduction of CA-cEdar during this patterning process yielded a more tightly packed feather array (n=12, Fig. 4F–H). CA-cEdar expressing cultures contained both small (Fig. 4G, arrowheads) and large buds. The larger buds appeared after 24 hr in culture, similar to the control. The smaller buds emerged after an additional 24 hr, representing a second wave of bud formation (Fig. 4F, G). Quantification showed that the bud density increases, but the average size decreases (Fig. 4H). Thus in this in vitro model, inhibition of receptor signaling suppresses bud morphogenesis, but does not abolish the initial dermal condensation, while increased receptor activity leads to an increase in bud density.

In vivo suppression of cEdar subfamily signaling

We infected embryos in ovo with RCAS expressing dominant negative receptors. Phenotypes of embryos with GFP positive patches were assessed by detection of cβ–catenin at incubation day 8. cβ-catenin undergoes the restrictive mode of expression and so can be used to visualise both placode fate and the location of the morphogenetic wave (Widelitz et al., 2000). Expression of GFP alone in control embryos did not perturb placode development, while DN-cEdar, DN-cTroy and DN-cXedar all suppressed placode identity, resulting in weak or occasionally absent cβ–catenin expressing foci in tracts (Fig. 5B,D,G). The buds with weakened expression that were detectable retained their spatial fidelity. The phenotypes correlated with GFP positive regions and thus transgene expression (Fig. 5A,C,F), though in some cases suppression of bud development was observed beyond the detectable GFP fluorescence. This may reflect the ability of the receptor extracellular domains to sequester freely diffusing ligand and so act in a non-cell autonomous manner by restricting ligand availability beyond infected regions. The morphological appearance of feather placodes can be visualised by staining with ethidium bromide (Eames and Schneider, 2005). Staining of an embryo infected with DN-cXedar (Fig. 5C,D,E) shows that within the infected tract (Fig. 5D,H), the lateral spread of feather placodes as the tract expands has been arrested, with fewer rows of identifiable placodes compared to the contralateral non-infected tract (Fig. 5E,I). These results agree with the in vitro reconstitution studies, indicating that these receptors reinforce placode identity and promote morphogenesis, but they do not appear to be individually required for spatial patterning or initial fate determination.

Fig. 5
Suppression of receptor signaling in vivo. (A,C,F) GFP fluorescence and (B,D,G) corresponding cβ-catenin expression in (A,B) DN-cEdar, (C,D) DN-cXedar and (F,G) DN-cTroy infected embryos. cβ-catenin expression is suppressed in regions ...

Activated receptors promote placode fate in distinct regions in vivo

To determine the effects of ectopic receptor activation, RCAS vectors were used to deliver constitutively active receptor constructs to developing embryos. Introduction of CA-cEdar induced patches of ectopic cβ–catenin expression immediately peripheral to the expanding tracts (Fig. 6A,A′). We did not observe ectopic cβ–catenin more distant from the tract in the apteric region. This phenotype is very similar to that reported for ectopic cEdar expression induced by activated β–catenin, except that ectopic placodes induced by β-catenin are not limited to the region immediately peripheral to the tract (Houghton et al., 2005). In contrast to CA-cEdar’s effects, both CA-cTroy and CA-cXedar generated fusions between buds within the tracts (Fig. 6B,C,C′), but we did not observe ectopic β–catenin within or outside the morphogenetic wave. Exclusion of cEda expression from these fusions (Fig. 6D), indicates that they have assumed full placode fate and are not cells with a combination of bud and interbud identities. To confirm that the ectopic or fused placode phenotypes observed correlated with foci of infection, embryos post in situ hybridization were cryosectioned, and immunostained for viral gag protein. This demonstrated that within phenotypically abnormal regions viral protein staining was present in both the epithelial and mesenchymal layers of the skin (Fig. 6E,H). In regions of the same embryo where placode phenotype was normal, staining for the viral gag protein was absent (Fig. 6F,I).

Fig. 6
Activation of receptor signaling in vivo. (A) CA-cEdar induces ectopic cβ-catenin expressing foci adjacent to the expanding feather tracts. (A′) Enlargement of the boxed area in A. The ectopic foci (arrowheads) are generally circular and ...

Suppression of NF-κB in feather bud development

Edar, Troy and Xedar activate the transcription factor NF–κB (Kojima et al., 2000; Yan et al., 2000), and repression of NF-κB function in the skin of transgenic mice phenocopies Edar ablation (Schmidt-Ullrich et al., 2001). To directly test a requirement for NF-κB function in feather development, we treated skin cultures with BAY 11–7082, an inhibitor of IκB phosphorylation (Pierce et al., 1997). Over the 3 day culture period tract expansion in the control explants resembled the events that occur in vivo (Fig. 7A,C,E). In inhibitor treated explants, the total bud area in the whole explant was reduced. This is manifested as a change in bud density or size, depending on the location within the explant (Fig. 7B,D,F). This finding can be explained by the fact that the feather buds in the midline form before those in lateral rows, thus reflecting stage dependent responses. Placode size towards the midline of the explants was similar to that of the control (n=12, Fig. 7D,F, arrowhead). These early feather buds had initiated development at the time when inhibitors were added. Some buds appear to escape the inhibition, while some recede over a 2 day period. More laterally, feather placodes were not established when the inhibitors were added. Here the number of buds was reduced but their size was increased (Fig. 7D,F, arrow, K). This difference is apparent on a scatter plot of bud size relative to location in the skin (midline versus lateral, Fig. 7K). We further examined these buds by in situ hybridisation with cShh, a known marker of the placode epidermis. In control explants the expression of cShh is restricted to the distal epithelium of the long buds (Fig. 7G,I). In the presence of inhibitor, the enlarged buds which form also express cShh. In these enlarged buds, cShh is properly expressed in the distal region of the bud. (Fig. 7H,J). Thus although the sizing of the feather buds may be irregular, marker gene expression appears to be normal. In the inhibitor treated explants, the total bud area of the explant is reduced compared to controls, largely due to a decrease in bud density (Fig. 7K).

Fig. 7
NF-κB inhibition during feather development. E7 dorsal skin was cultured in the absence or presence of the NF-κB inhibitor, BAY 11–7082. In control explants (A,C,E) long buds begin to form near the midline of the explant (arrow ...


We have examined the sequence, expression and functional characteristics of three related receptors of the TNFR superfamily in feather placode development. While the sequence of cEdar is very similar to its human orthologue, cTroy is moderately, and cXedar is highly, divergent from their mammalian counterparts. Our finding of a death domain in cXedar was unexpected, representing to our knowledge the first example of a signaling TNFR which contains a death domain in one lineage but lacks it in another. This finding implies that the ancestral vertebrate Xedar contained a death domain which has been lost in mammals and suggests that Xedar signaling mechanisms are distinct in different vertebrates. The expression patterns of the three receptors are similar to their characteristics in mammalian skin development. In mouse, Edar and Troy are first expressed throughout the epidermis and then become restricted to the hair placodes (Pispa et al., 2003), and Xedar is upregulated slightly later during appendage morphogenesis (Yan et al., 2000). cEda is first expressed all over the epidermis, and then becomes restricted to the interbud region in a pattern reciprocal to that of cEdar, in an expression pattern resembling that observed in the mouse interfollicular epidermis (Houghton et al., 2005). Thus, with the exception of cXedar, these molecules are expressed in the restrictive mode during skin appendage formation, suggesting a role in the micropatterning process of forming feather buds within tracts (Lin et al., 2006).

While manipulation of receptor and NF-κB signaling activities yielded diverse phenotypes (Fig. 8), some general principles have emerged. We propose that, though not required for the initial assumption of placodal fate, the level of receptor activity modulates both the density and the size of feather buds. Elevated receptor activity leads to a greater fraction of the skin assuming a bud fate, while suppressed signaling has the opposite effect and destabilises placodes.

Fig. 8
Model for roles of receptor and NF-κB signaling at different stages of feather development. (A) Modulation of receptor/NF-κB activity during periodic patterning leads to the alteration of placode density. This is represented in the lateral ...

The phenotypes caused by expressing dominant negative receptors are not consistent with their being essential for the periodic patterning process. Rather, they appear to regulate epithelial morphogenesis. Our finding that dominant negative cTroy and cXedar suppress feather development is in apparent contrast to their null mutations in mouse, which are reported to cause no gross abnormalities (Shao et al., 2005; Newton et al., 2004). There are several potential explanations for this difference. The first is that the Troy and Xedar mutant mice have subtle defects in hair follicle formation. Given the distinct subtypes of follicles in the mouse coat, it is possible that a significant alteration in the number of any one type would not appear abnormal at a gross level. A second explanation is that Troy and Xedar functions differ between mouse and chicken, an argument perhaps supported by the divergence observed in their protein sequences compared to the high degree of conservation of Edar. The presence of two forms of Troy, a long form capable of signaling and a putative dominant negative short form, suggests a third explanation. While the mouse null mutation ablated both active and antagonistic Troy forms, our dominant negative approach exclusively enhanced the negative Troy function. If Troy functionally interacts with other receptors or signaling pathways, then selective enhancement of negative functions might have more severe consequences than loss of both positive and negative inputs. The essentially identical phenotypes induced by each mutated receptor construct are consistent with such a transdominant negative effect. This effect may be caused by heteromeric Eda ligand complexes recruiting the different receptors into a single signaling complex. Such heteromeric EdaA1-EdaA2 complexes have not been reported, but the presence of a collagen-like domain towards the N-terminus of both Eda variants (Srivastava et al., 1997) suggests that these proteins would be likely to multimerise when coexpressed.

cXedar expression is not detectable by in situ hybridization until E10, after the initiation of feather bud morphogenesis. This finding could indicate that the endogenous receptor normally works with cTroy and cEdar to regulate bud shape and growth, but not the early stages of placode formation. That endogenous cXedar might not be as important in placode formation is perhaps supported by the lower efficiency of bud suppression caused by DN-cXedar (26%), compared to the equivalent cEdar (44%) and cTroy (40%) constructs. The phenotypic effects may represent dominant negative inhibition at a time when cXedar signaling may not normally be active, therefore the effects observed may be indicative of ectopic interaction with cEdar and cTroy.

Elevated signaling activity was achieved using LMP1-receptor fusions. Enhanced signaling acting in the morphogenetic zone, where placodes are emerging, led to an increase in bud number. This effect is seen as ectopic bud formation in the morphogenetic wave using CA-cEdar in vivo, and increased bud density in the reconstitution assay in vitro. If this enhanced activity occurred after placode locations were established it instead converted some interbud domains to bud by increasing bud size, leading to placode fusions as seen using CA-cTroy and CA-cXedar.

In vivo, cEdar generated ectopic buds in the morphogenetic wave, ahead of agreed placode locations, while cTroy and cXedar promoted placode fusion within tracts. The reason for this difference is unclear, but may reflect when and where these receptors act relative to the inhibitors of placode fate. If cEdar acts upstream of the inhibitors, and its signaling is susceptible to their action, then this would explain the lack of intratract fusions observed. That Edar itself is susceptible to the inhibitory effects of BMP signaling has been shown in mouse and chick (Houghton et al., 2005, Mou et al., 2006), though whether its cytoplasmic signal transduction apparatus is also inhibited by these factors remains to be tested. Conversely, cTroy and cXedar may act downstream of the inhibitors, evading lateral inhibitory signals emanating from the placodes and allowing fusions to form within the tracts. It is also possible that signaling pathways other than NF-κB are differentially activated by these three receptors, and that this divergence in downstream signal transduction is the cause of the spatially distinct phenotypes caused by the activated receptors. The weak NF-κB activation induced by CA-cXedar in vitro might support the notion that other signaling pathways are employed by these receptors.

Overexpression of EdaA2 in mouse has been reported to have no effect on skin or appendage development (Newton et al., 2004), suggesting that restricted receptor expression, rather than ligand availability, is the primary mode of regulating of this pathway. Stimulation of Edar signaling by application of recombinant EdaA1 to embryonic mouse skin cultures has been reported to cause enlarged, fused hair placodes (Mustonen et al., 2004). This phenotype is similar to that we find caused by activation of cTroy and cXedar, but not by cEdar. Though similar in ultimate appearance, the mechanisms of hair and feather placode fusion may be distinct. Hair placode fusion is a result of the joining of independently generated placodes, each of which is approximately the same age. Since feathers are generated as the morphogenetic wave sweeps laterally across the skin, fusions in this system are between older and younger buds, and are most likely caused by a failure of separation rather than the union of two initially distinct placodes.

Suppression of pathway activity was achieved by expressing dominant negative receptors or by pharmacological inhibiton of NF-κB activity. When suppression by dominant negative receptors took place at the morphogenetic wave stage, in the reconsititution system, there was an overall suppression of bud formation, although the stable dermal condensation of the feather primordium was visible. The dermal papilla condensates appeared periodically patterned, though a morphologically distinct epithelial placode did not appear and bud outgrowth did not occur. This finding in chicken skin is in apparent contrast to the situation in the Edar pathway mouse mutants, where some effort at epithelial placode formation seems to occur without formation of a distinct dermal papilla (Schmidt-Ullrich et al., 2006). This may indicate a difference in the relative contributions of dermal and epidermal components to early appendage formation in the different species, with the dermis being more clearly predominant in feather development.

NF-κB has been identified as a promoter of placodal fate in mouse (Schmidt-Ullrich et al., 2006). Pharmacological inhibition of NF-κB in explant cultures allows an analysis of its roles in the formation of the feather primordia. We were able to determine the phenotypic effects of blockade of all signals activating this transcription factor as an addition to dominant negative inhibition of individual receptor signaling. When NF-κB inhibition occurs at a stage when feather primordia are already formed (the central rows of the feather explant), suppression of NF-κB leads to some of the midline buds receding over a two day period. When NF-κB inhibition acts at a stage when periodic patterning is still ongoing (in the lateral regions of the explant), there is an overall reduction in the number of feather placodes. These “reset buds” are larger than normal buds possibly due to the increase in interbud spacing and lack of inhibitory signals from the surrounding placodes.

Phenotypic effects resulting from the blockade of NF-κB activity are more pronounced than those generated by expression of individual dominant negative receptors. That inhibition of all NF-κB activating inputs has this stronger effect suggests that there are other factors that activate NF-κB during feather morphogenesis, or else that the three receptors analysed here have redundant functions relative to one another. Taken together, these data from both in vivo and in vitro models suggest that the Edar subfamily-NF-κB axis plays an important role in setting the number of buds formed during the initial pattern forming stage, and regulates bud size during early feather bud morphogenesis.

Of the genes known to regulate appendage development, β –catenin is the most intensively studied and has produced the most dramatic phenotypes when manipulated in developing skin. Expression of activated β –catenin in embryonic chicken skin causes ectopic placodogenesis and feather growth (Noramly et al., 1999; Houghton et al., 2005). The ectopic placodes express cEdar (Houghton et al., 2005), a point that has been used to suggest that β-catenin lies upstream of Edar in a linear developmental pathway. Our finding that cEdar can upregulate β-catenin expression indicates that this linear model for placode initiation may not be tenable, or at least that evidence obtained from gain of function experiments is insufficient to draw these conclusions. Rather, genes expressed in the restrictive mode prior to placode formation appear to be mutually reinforcing, a phenomenon that would help achieve threshold levels of activation required to assume a placode fate. β-catenin has a greater range of placode inducing ability, being able to produce buds in essentially any location in the skin, while we observed cEdar induced buds only in the morphogenetic wave bordering the developing tracts. This spatial characteristic may simply reflect the distribution of signaling co-factors, with full cEdar signal transduction requiring co-expression of cEdaradd, which is restricted to the buds and the morphogenetic wave (Houghton et al., 2005).

Data from mammals, fish and birds indicates that Edar’s sequence and function are conserved across the vertebrate classes (Kondo et al., 2001; Houghton et al., 2005). Thus a highly constrained Edar perhaps lies at the core of an ancestral skin appendage development programme. Troy and Xedar are more extensively diverged, indicating that they may be more peripheral to this programme and thus more malleable evolutionary agents. Importantly the retention of the death domain in Xedar in the avian genome will be addressed by examination of the function of the domain both biochemically and in vivo. Future study of the Edar subfamily will dissect their interactions, aiming to illustrate the uses to which duplicated genes may be put in order to generate the complex organ forms we see today.


We thank Keith Brennan, Heithem El-Hodiri, David Garrod, Stacie Loftus, Karen More, Pascal Schneider, Bill Sugden, Cheryll Tickle and Randall B. Widelitz. This work was supported by funding from the BBSRC, UK (DJH) and NIAMS AR 042177 and 047364 from NIH, USA (CMC).


Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.


  • Boardman PE, Sanz-Ezquerro J, Overton IM, Burt DW, Bosch E, Fong WT, Tickle C, Brown WR, Wilson SA, Hubbard SJ. A comprehensive collection of chicken cDNAs. Curr Biol. 2002;12:1965–1969. [PubMed]
  • Bossen C, Ingold K, Tardivel A, Bodmer JL, Gaide O, Hertig S, Ambrose C, Tschopp J, Schneider P. Interactions of tumor necrosis factor TNF and TNF receptor family members in the mouse and human. J Biol Chem. 2006;281:13964–13971. [PubMed]
  • Botchkarev VA, Botchkareva NV, Sharov AA, Funa K, Huber O, Gilchrest BA. Modulation of BMP signaling by noggin is required for induction of the secondary (nontylotrich) hair follicles. J Invest Dermatol. 2002;118:3–10. [PubMed]
  • Chang CH, Jiang TX, Lin CM, Burrus LW, Chuong CM, Widelitz R. Distinct Wnt members regulate the hierarchical morphogenesis of skin regions spinal tract and individual feathers. Mech Dev. 2004;121:157–171. [PMC free article] [PubMed]
  • Eames BF, Schneider RA. Quail-duck chimeras reveal spatiotemporal plasticity in molecular and histogenic programs of cranial feather development. Development. 2005;132:1499–1509. [PMC free article] [PubMed]
  • Fuchs E, Merrill BJ, Jamora C, DasGupta R. At the roots of a never-ending cycle. Dev Cell. 2001;1:13–25. [PubMed]
  • Hardy MH. The secret life of the hair follicle. Trends Genet. 1992;8:55–61. [PubMed]
  • Hatzivassiliou E, Miller WE, Raab-Traub N, Kieff E, Mosialos G. A fusion of the EBV latent membrane protein-1 LMP1 transmembrane domains to the CD40 cytoplasmic domain is similar to LMP1 in constitutive activation of epidermal growth factor receptor expression, nuclear factor-kappa B, and stress-activated protein kinase. J Immunol. 1998;160:1116–1121. [PubMed]
  • Headon DJ, Overbeek PA. Involvement of a novel Tnf receptor homologue in hair follicle induction. Nat Genet. 1999;22:370–374. [PubMed]
  • Headon DJ, Emmal SA, Ferguson BM, Tucker AS, Justice MJ, Sharpe PT, Zonana J, Overbeek PA. Gene defect in ectodermal dysplasia implicates a death domain adapter in development. Nature. 2001;414:913–916. [PubMed]
  • Houghton L, Lindon C, Morgan BA. The ectodysplasin pathway in feather tract development. Development. 2005;132:863–872. [PubMed]
  • Inoue J, Ishida T, Tsukamoto N, Kobayashi N, Naito A, Azuma S, Yamamoto T. Tumor necrosis factor receptor-associated factor (TRAF) family: adapter proteins that mediate cytokine signaling. Exp Cell Res. 2000;254:14–24. [PubMed]
  • Jiang TX, Jung HS, Widelitz RB, Chuong CM. Self-organization of periodic patterns by dissociated feather mesenchymal cells and the regulation of size, number and spacing of primordia. Development. 1999;126:4997–5009. [PubMed]
  • Jiang TX, Widelitz RB, Shen WM, Will P, Wu DY, Lin CM, Jung HS, Chuong CM. Integument pattern formation involves genetic and epigenetic controls: feather arrays simulated by digital hormone models. Int J Dev Biol. 2004;48:117–135. [PMC free article] [PubMed]
  • Kanda H, Igaki T, Kanuka H, Yagi T, Miura M. Wengen, a member of the Drosophila tumor necrosis factor receptor superfamily, is required for Eiger signaling. J Biol Chem. 2002;277:28372–28375. [PubMed]
  • Kere J, Srivastava AK, Montonen O, Zonana J, Thomas N, Ferguson B, Munoz F, Morgan D, Clarke A, Baybayan P, Chen EY, Ezer S, Saarialho-Kere U, de la Chapelle A, Schlessinger D. X-linked anhidrotic hypohidrotic ectodermal dysplasia is caused by mutation in a novel transmembrane protein. Nat Genet. 1996;13:409–416. [PubMed]
  • Kojima T, Morikawa Y, Copeland NG, Gilbert DJ, Jenkins NA, Senba E, Kitamura T. TROY, a newly identified member of the tumor necrosis factor receptor superfamily, exhibits a homology with Edar and is expressed in embryonic skin and hair follicles. J Biol Chem. 2000;275:20742–20747. [PubMed]
  • Kondo S, Kuwahara Y, Kondo M, Naruse K, Mitani H, Wakamatsu Y, Ozato K, Asakawa S, Shimizu N, Shima A. The medaka rs-3 locus required for scale development encodes ectodysplasin-A receptor. Curr Biol. 2001;11:1202– 1206. [PubMed]
  • Laurikkala J, Pispa J, Jung HS, Nieminen P, Mikkola M, Wang X, Saarialho-Kere U, Galceran J, Grosschedl R, Thesleff I. Regulation of hair follicle development by the TNF signal ectodysplasin and its receptor Edar. Development. 2002;129:2541–2553. [PubMed]
  • Lin CM, Jiang TX, Widelitz RB, Chuong CM. Molecular signaling in feather morphogenesis. Curr Opin Cell Biol. 2006;18:730–741. [PMC free article] [PubMed]
  • Locksley RM, Killeen N, Lenardo MJ. The TNF and TNF receptor superfamilies: integrating mammalian biology. Cell. 2001;104:487–501. [PubMed]
  • Loftus SK, Larson DM, Watkins-Chow D, Church DM, Pavan WJ. Generation of RCAS vectors useful for functional genomic analyses. DNA Res. 2001;8:221–226. [PubMed]
  • Millar SE. Molecular mechanisms regulating hair follicle development. J Invest Dermatol. 2002;118:216–225. [PubMed]
  • Mou C, Jackson B, Schneider P, Overbeek PA, Headon DJ. Generation of the primary hair follicle pattern. Proc Natl Acad Sci USA. 2006;103:9075–9080. [PMC free article] [PubMed]
  • Mustonen T, Ilmonen M, Pummila M, Kangas AT, Laurikkala J, Jaatinen R, Pispa J, Gaide O, Schneider P, Thesleff I, Mikkola ML. Ectodysplasin A1 promotes placodal cell fate during early morphogenesis of ectodermal appendages. Development. 2004;131:4907–4919. [PubMed]
  • Newton K, French DM, Yan M, Frantz GD, Dixit VM. Myodegeneration in EDA-A2 transgenic mice is prevented by XEDAR deficiency. Mol Cell Biol. 2004;24:1608–1613. [PMC free article] [PubMed]
  • Noramly S, Freeman A, Morgan BA. beta-catenin signaling can initiate feather bud development. Development. 1999;126:3509–3521. [PubMed]
  • Pierce JW, Schoenleber R, Jesmok G, Best J, Moore SA, Collins T, Gerritsen ME. Novel inhibitors of cytokine-induced IkappaBalpha phosphorylation and endothelial cell adhesion molecule expression show anti-inflammatory effects in vivo. J Biol Chem. 1997;272:21096–21103. [PubMed]
  • Pispa J, Mikkola ML, Mustonen T, Thesleff I. Ectodysplasin, Edar and TNFRSF19 are expressed in complementary and overlapping patterns during mouse embryogenesis. Gene Expr Patterns. 2003;3:675–679. [PubMed]
  • Plikus M, Wang WP, Liu J, Wang X, Jiang TX, Chuong CM. Morpho-regulation of ectodermal organs: integument pathology and phenotypic variations in K14-Noggin engineered mice through modulation of bone morphogenic protein pathway. Am J Pathol. 2004;164:1099–1114. [PMC free article] [PubMed]
  • Riddle RD, Johnson RL, Laufer E, Tabin C. Sonic hedgehog mediates the polarizing activity of the ZPA. Cell. 1993;75:1401–1416. [PubMed]
  • Schmidt-Ullrich R, Aebischer T, Hulsken J, Birchmeier W, Klemm U, Scheidereit C. Requirement of NF-kappaB/Rel for the development of hair follicles and other epidermal appendices. Development. 2001;128:3843–3853. [PubMed]
  • Schmidt-Ullrich R, Tobin DJ, Lenhard D, Schneider P, Paus R, Scheidereit C. NF-kappaB transmits Eda A1/EdaR signalling to activate Shh and cyclin D1 expression, and controls post-initiation hair placode down growth. Development. 2006;133:1045–1057. [PubMed]
  • Schneider P, Street SL, Gaide O, Hertig S, Tardivel A, Tschopp J, Runkel L, Alevizopoulos K, Ferguson BM, Zonana J. Mutations leading to X-linked hypohidrotic ectodermal dysplasia affect three major functional domains in the tumor necrosis factor family member ectodysplasin-A. J Biol Chem. 2001;276:18819–18827. [PubMed]
  • Sengel P. Pattern formation in skin development. Int J Dev Biol. 1990;34:33–50. [PubMed]
  • Shao Z, Browning JL, Lee X, Scott ML, Shulga-Morskaya S, Allaire N, Thill G, Levesque M, Sah D, McCoy JM, Murray B, Jung V, Pepinsky RB, Mi S. TAJ/TROY, an orphan TNF receptor family member, binds Nogo- 66 receptor 1 and regulates axonal regeneration. Neuron. 2005;45:353–359. [PubMed]
  • Srivastava AK, Pispa J, Hartung AJ, Du Y, Ezer S, Jenks T, Shimada T, Pekkanen M, Mikkola ML, Ko MS, Thesleff I, Kere J, Schlessinger D. The Tabby phenotype is caused by mutation in a mouse homologue of the EDA gene that reveals novel mouse and human exons and encodes a protein (ectodysplasin-A) with collagenous domains. Proc Natl Acad Sci USA. 1997;94:13069–13074. [PMC free article] [PubMed]
  • Thesleff I, Mikkola ML. Death receptor signaling giving life to ectodermal organs. Sci STKE. 2002;2002:E22. [PubMed]
  • van Genderen C, Okamura RM, Farinas I, Quo RG, Parslow TG, Bruhn L, Grosschedl R. Development of several organs that require inductive epithelial-mesenchymal interactions is impaired in LEF-1-deficient mice. Genes Dev. 1994;8:2691–2703. [PubMed]
  • Wajant H. Death receptors. Essays Biochem. 2003;39:53–71. [PubMed]
  • Widelitz RB, Jiang TX, Lu J, Chuong CM. beta-catenin in epithelial morphogenesis: conversion of part of avian foot scales into feather buds with a mutated beta-catenin. Dev Biol. 2000;219:98–114. [PubMed]
  • Wu P, Hou L, Plikus M, Hughes M, Scehnet J, Suksaweang S, Widelitz R, Jiang TX, Chuong CM. Evo-Devo of amniote integuments and appendages. Int J Dev Biol. 2004;48:249–270. [PMC free article] [PubMed]
  • Yan M, Wang LC, Hymowitz SG, Schilbach S, Lee J, Goddard A, de Vos AM, Gao WQ, Dixit VM. Two-amino acid molecular switch in an epithelial morphogen that regulates binding to two distinct receptors. Science. 2000;290:523– 527. [PubMed]
  • Yan M, Zhang Z, Brady JR, Schilbach S, Fairbrother WJ, Dixit VM. Identification of a novel death domain-containing adaptor molecule for ectodysplasin-A receptor that is mutated in crinkled mice. Curr Biol. 2002;12:409–413. [PubMed]
  • Yu M, Yue Z, Wu P, Wu DY, Mayer JA, Medina M, Widelitz RB, Jiang TX, Chuong CM. The biology of feather follicles. Int J Dev Biol. 2004;48:181–191. [PMC free article] [PubMed]
PubReader format: click here to try


Save items

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Gene
    Gene records that cite the current articles. Citations in Gene are added manually by NCBI or imported from outside public resources.
  • Gene (nucleotide)
    Gene (nucleotide)
    Records in Gene identified from shared sequence and PMC links.
  • GEO Profiles
    GEO Profiles
    Gene Expression Omnibus (GEO) Profiles of molecular abundance data. The current articles are references on the Gene record associated with the GEO profile.
  • HomoloGene
    HomoloGene clusters of homologous genes and sequences that cite the current articles. These are references on the Gene and sequence records in the HomoloGene entry.
  • MedGen
    Related information in MedGen
  • Nucleotide
    Primary database (GenBank) nucleotide records reported in the current articles as well as Reference Sequences (RefSeqs) that include the articles as references.
  • Protein
    Protein translation features of primary database (GenBank) nucleotide records reported in the current articles as well as Reference Sequences (RefSeqs) that include the articles as references.
  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem chemical substance records that cite the current articles. These references are taken from those provided on submitted PubChem chemical substance records.

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...