Human Molecular Genetics, 2001, Vol. 10, No. 9 953-962
© 2001 Oxford University Press
Ectodysplasin is released by proteolytic shedding and binds to the EDAR protein
1Department of Medical Genetics, Haartman Institute, 2Finnish Genome Center and 3Institute of Biotechnology, Viikki Biocenter, 00014 University of Helsinki, Helsinki, Finland and 4Department of Dermatology, Helsinki University Central Hospital, Helsinki, Finland
Received 3 January 2001; Revised and Accepted 2 March 2001.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Anhidrotic ectodermal dysplasia (EDA) is an X-linked disorder characterized by abnormal development of ectoderm and its appendices. The EDA gene encodes different isoforms of ectodysplasin, a transmembrane protein. The two longest isoforms, ectodysplasin-A1 and -A2, which differ by an insertion of two amino acids, are trimeric type II membrane proteins with an extracellular portion containing a short collagenous domain and a TNF ligand motif in the C-terminal region. We show that ectodysplasin is released from cells to the culture medium. Deletion constructs were used to localize the cleavage site and show that the putative recognition sequence of a furin-like enzyme is needed for the cleavage. Some EDA patients have missense mutations affecting this recognition sequence, suggesting that cleavage has biological significance in vivo. EDAR, a recently cloned member of the TNFR family and the product of the downless gene, is able to co-precipitate ectodysplasin, confirming that they form a ligand–receptor pair. In situ hybridization and immunostaining studies show that ectodysplasin and EDAR are expressed in adjacent or partially overlapping layers in the developing human skin. We conclude that as a soluble ligand, ectodysplasin is able to interact with EDAR and mediate signals needed for the development of ectodermal appendages.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
The normal development of ectodermal appendages, including teeth, sweat glands and hair follicles, requires an intact anhidrotic ectodermal dysplasia (EDA) gene or its orthologue in many mammalian species. The human EDA gene (EDI) is located in the X chromosome, causing anhidrotic (or hypohidrotic) ectodermal dysplasia in males when mutated (1). X-linked EDA is the most common among more than 150 clinically distinct syndromes characterized by variably abnormal development of epidermis and its appendages (2). The orthologous X-linked murine gene is called Tabby (Ta) (3,4).
We have found that the EDA gene undergoes complex alternative splicing producing at least seven different transcripts (5). All transcripts include exon 1, which encodes a short intracellular domain, a transmembrane domain and 73 amino acids of extracellular protein sequence. The two longest isoforms, ectodysplasin-A1 and -A2, which differ only by an insertion of two amino acids (EDA-A1, 391 amino acids; EDA-A2, 389 amino acids), are trimeric type II membrane proteins with an extracellular portion containing a collagenous domain and a TNF-ligand motif in the C-terminal region (5–7).
Phenotypically highly similar but genetically distinct conditions of autosomal ectodermal dysplasia exist in both human and mouse. Recently, the gene mutated in the mouse downless (dl) strain was cloned and mutations in the corresponding human gene, DL or EDAR, were shown to cause autosomal recessive and dominant forms of ectodermal dysplasia (8,9). The EDAR gene encodes a novel protein which belongs to the TNF receptor family. The phenotypic similarity of the X-linked and autosomal forms of ectodermal dysplasias suggested that the ectodysplasin and EDAR proteins might function as a ligand–receptor pair in a common signalling pathway which participates in the development of ectodermal appendages. However, murine in situ data suggested different expression patterns for ectodysplasin and EDAR (10,11) and their possible interaction has also been questioned (10). Recently another receptor candidate, called XEDAR, was reported to bind specifically the 389 amino acid ectodysplasin-A2 isoform but not the 391 amino acid A1 isoform (12).
We aimed to study the suggested ligand–receptor relationship of the ectodysplasin and EDAR proteins and show here for the first time that ectodysplasin is released from cells to the cell culture medium, suggesting that it can act as a diffusible ligand. Using deletion variants of ectodysplasin we localized the cleavage site and showed that a putative recognition sequence of a furin-like enzyme is responsible for the cleavage. Furthermore, we showed that ectodysplasin-A1 and the EDAR protein co-precipitate, confirming that they can form a ligand–receptor pair. Finally, our in situ hybridization and immunostaining studies revealed that ectodysplasin and EDAR have different, but neighbouring expression patterns in the developing human skin, and that ectodysplasin is detected at sites of EDAR expression. We conclude that as a diffusible ligand, ectodysplasin can interact with EDAR and thus mediate signals necessary for ectodermal development.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Ectodysplasin is released from cells by proteolytic shedding
To study whether ectodysplasin is shed from the plasma membrane to cell culture medium, 293A and COS-1 cells were transiently transfected with the human and murine full-length EDA cDNAs (EDA-A1, EDA-A2 or Ta-A). After transfection, culture media were changed and cell lysate and medium samples were prepared for SDS–PAGE at different time points (4–48 h). Samples were analysed by western blotting using a novel polyclonal antibody against the 18 amino acid peptide sequence SKHTTFFGAIRLGEAPAS from the C-terminus of ectodysplasin-A, which is conserved between human and mouse. In the cell lysates of transfectants, a 45–50 kDa product was observed (Fig. 1A), corresponding to the full-length ectodysplasin-A polypeptide in reducing conditions (6,7). In medium samples of EDA-A and Ta-A transfectants, the antibody revealed a band of
35–40 kDa (Fig. 1B). The proteolytic processing was observed in both cell lines used, COS-1 and 293A, and also in transfected primary keratinocytes (not shown). The cleaved form was present already after 4 h collection in non-concentrated medium samples of COS-1 and 293A cells (data not shown). No difference was observed in the cleavage between ectodysplasin-A1 and -A2; both forms were cleaved equally (data not shown). The medium of mock-transfected cells was negative for ectodysplasin-A or other products. Antibodies against the N-terminus of ectodysplasin or the shortest isoform (EDA-O, 135 amino acids) did not bind to any proteins in the medium samples (data not shown).
|
In order to study the specificity of cleavage and localize the cleavage site, we prepared deletion constructs. We made four deletion constructs truncating the spacer region (Sp) of ectodysplasin-A between the transmembrane and collagenous domains (Fig. 2). The dSp1 construct has the largest deletion, lacking 100 amino acids (residues 79–179), dSp2 lacks 40 amino acids (residues 79–118) and dSp3 lacks 23 amino acids (residues 119–141). As ectodysplasin is known to have two overlapping putative cleavage sites (R-X-K/R-R) for furin-like proteases, residues 152–159 (RRVRRNKR) were deleted in the dFur construct.
|
When cells were transiently transfected with the dSp1 plasmid, followed by western analysis with C-terminal antibody (Fig. 1), the soluble form of ectodysplasin was absent from the medium sample (Fig. 1B). In the cell lysate, a 35–40 kDa band was observed (Fig. 1A) which agrees well with the calculated molecular weight for the dSp1 variant. To more specifically localize the shedding region, we carried out transfections with the dSp2, dSp3 and dFur deletion constructs. When transfectants were assayed on immunoblots, the soluble form was observed in the medium samples of dSp2 and dSp3, whereas the medium sample of dFur was negative (Fig. 1B), suggesting that a furin-like proteolytic enzyme is responsible for cleavage of ectodysplasin. In the cell lysate samples of dSp2, dSp3 and dFur, bands corresponding to proteins of the expected size were seen, indicating that the proteins were translated and stable (Fig. 1A). All the products, also the cleaved form, were seen as two or three closely migrating bands, as has been reported earlier (6).
The size observed for the cleaved product on westerns was larger than calculated. We estimated that the cleaved form is 35–40 kDa, whereas the calculated molecular weight for the peptide from furin site to C-terminus is 25 kDa. This is most likely because of glycosylation; there are two N-glycosylation sites on the ectodomain of ectodysplasin (and none in the N-terminal region) and the collagenous domain may also be glycosylated. Thus, extracellular glycosylation accounts for 10–15 kDa, which agrees well with previous data (5–7). There appeared to be some experimental variation also in the relative amounts of protein produced by the different constructs.
In order to study whether mutations in the first putative furin-like recognition site have an effect on cleavage, a new expression construct, called Mut, with Arg153His and Arg156His mutations was prepared. When the non-concentrated medium sample of Mut transfectants was assayed on immunoblots with the C-terminal antibody, no protein product could be observed (Fig. 1B) or only a faint band was present after longer exposures. This small amount of cleaved form observed with Mut construct can be explained with non-complete destruction of the furin-like recognition site, as there are two overlapping sites for furin-like proteases. The band observed in the lysate sample of Mut was as strong as bands in other lysates of ectodysplasin constructs (Fig. 1A), suggesting that Arg mutations disturb cleavage but not stability of protein. When 10-fold concentrated medium samples of dSp1 and dFur were immunoblotted, dSp1 was negative and only faint bands were observed in dFur sample (data not shown) which may also be unspecific degradation products. As the dFur clearly inhibits most cleavage we consider it unlikely that another major cleavage site would exist.
Shed ectodomain forms trimers
To study whether the deletion variants of ectodysplasin are able to form trimers like the native protein (6,7), we analysed the transfectants on SDS–PAGE under non-reducing conditions. On immunoblots the C-terminal ectodysplasin antibody revealed three different forms in lysate samples (Fig. 3A). The smallest bands of 40–50 kDa (35 kDa in dSp1) correspond to monomers. The sharp band of 90 kDa (70 kDa in dSp1) and the larger diffuse bands 120–130 kDa (100 kDa in dSp1) correspond to the expected dimeric and trimeric sizes of different ectodysplasin deletion variants. The result agrees well with our previous data of trimer formation of ectodysplasin (6).
|
Because the trimeric form seemed unstable in the normal SDS–PAGE conditions, we used milder conditions as described by Tanaka et al. (13) to detect the trimeric Fas ligand, to study whether the shed form of ectodysplasin exists as a trimer. The medium sample of ectodysplasin-A was concentrated using the Centricon concentrator and changed into PBS for chemical crosslinking with difluorodinitrobenzene (DFDNB) or disuccinimidyl suberate (DSS). Media samples with or without crosslinking were mixed in 0.5% SDS sample buffer and run on a gel at +4°C (Fig. 3B). On immunoblots without crosslinking, the antibody against ectodysplasin recognized a monomeric cleaved form of 35–40 kDa and a faint larger band corresponding to the expected trimeric size of
100 kDa. After crosslinking, the putative trimeric form was stronger and a band of 70–80 kDa, corresponding to dimeric size, was also observed. Additional larger bands are probably due to hyperlinking. Similar results were obtained with both crosslinker chemicals (data not shown). Our results showed that the shed form of ectodysplasin is likely to exist as a trimer like the uncleaved forms. Finally, we used cell surface biotinylation to examine whether the deletion forms of ectodysplasin were transported to the cell surface. COS-1 cells transfected with the full-length or deletion variants of ectodysplasin were biotinylated on ice, followed by precipitation of biotinylated proteins by streptavidin and western blotting (Fig. 3C). The results showed that all ectodysplasin variants were present on the cell surface, further supporting the idea that defective cleavage explained the lack of secreted protein. As a control, cells were transfected with a pCMV5-HCR plasmid encoding the intracellular HCR protein (14) (K. Asumalahti, S. Suomela, T. Laitinen, O. Elomaa, U. Saarialho-Kere and J. Kere, unpublished data). As expected, this protein was not detected in the surface protein fraction precipitated by streptavidin-agarose, though it was detectable in the total cell lysate with the anti-HCR antibodies (Fig. 3C).
Co-precipitation of ectodysplasin with the EDAR protein
The entire coding region or the putative extracellular domain of human EDAR cDNA (9) was cloned to the pCDNA3-V5-His vector for expression in eukaryotic cells. Total lysates of transiently transfected 293A and COS-1 cells and partially purified protein were analysed by western blotting using two polyclonal antibodies, one against the 20-mer peptide sequence (CEGFFRATVLTPGDMENDAE) from the N-terminal region and the other against the last 16 amino acids (EWAGVVPPASQPHAAS) from the C-terminus of the EDAR protein. In cells transfected with the V5-His-tagged full-length EDAR, both antibodies recognized a band of 70–75 kDa (Fig. 4A). The calculated size for the EDAR polypeptide is 49 kDa and for the V5-His tags is 5 kDa, suggesting post-translational glycosylation accounting for 20–30 kDa. Also the antibody against the V5 tag revealed the same polypeptide of 70–75 kDa (data not shown). In mock-transfected cells (Fig. 4A) antibodies did not recognize any proteins. Stainings with preimmune sera were negative (data not shown).
|
In order to study whether ectodysplasin and the EDAR protein are able to form a ligand–receptor pair, we carried out immunoprecipitations. For the precipitations, the ectodomain of EDAR was expressed as a secreted V5-His-fusion protein (EDARec-V5-His), which was partially purified from the culture medium with a Ni-affinity column. The purified product was verified by immunoblotting with the antibody against the N-terminal region of EDAR (Fig. 4A). On the gel, EDARec runs as a band of 35–40 kDa. The calculated size for it is 21 kDa and for V5-His tag is 5 kDa. The 10–15 kDa difference in size is likely to be due to the glycosylation in the N-glycosylation site of the extracellular region of EDAR (9). As expected, the antibody against the C-terminus did not recognize EDARec-V5-His. For binding studies, cells transiently transfected with the full-length EDA-A1 cDNA were lysed and labelled with biotin. The cell lysate was incubated with the EDARec-V5-His fusion protein followed by immunoprecipitation with anti-V5 antibody and protein A-agarose. The precipitates were resolved by SDS–PAGE and analysed by western blotting and staining with peroxidase-conjugated streptavidin (SPOD). The EDARec-V5-His protein precipitated specifically a single polypeptide of
50 kDa, corresponding to ectodysplasin-A (Fig. 4B). On western blot the antibody against the C-terminus of ectodysplasin-A revealed the same band as streptavidin (Fig. 4B). The anti-V5 antibody and protein A sepharose alone were not able to precipitate ectodysplasin-A1 (Fig. 4B). The coprecipitation of ectodysplasin-A1 with the EDAR protein confirms that they are able to form a ligand–receptor pair as suggested (12).
Localization of ectodysplasin and EDAR by in situ hybridization and immunohistochemistry
We have shown previously that ectodysplasin is expressed in the developing epithelium and also in epithelial cells in adult tissues (15). In order to compare the expression patterns of ectodysplasin and EDAR in the developing human skin, in situ hybridization and immunostaining studies were performed using parallel sections of paraffin-embedded tissues. In the skin, at week 20 of gestation, both genes were expressed in epidermis, whereas the dermis and smooth muscle were negative (Fig. 5). Both immunohistochemistry (data not shown) and in situ hybridizations showed that ectodysplasin was expressed throughout the epidermis as previously reported (15) whereas the EDAR was focally restricted to the periderm (Fig. 5). Similar peridermal expression was observed in immunostainings with antibodies against the C-terminus of EDAR. Interestingly, EDAR was seen in association of thickenings of epithelia, which are the first signs for formation of epidermal appendages such as hair follicles. By immunostaining, ectodysplasin was shown to be present in epithelial cells and in the hair matrix of hair follicles at gestational week 20. This agrees well with our previous in situ data (1). By in situ hybridization, EDAR mRNA was found in the same area of hair follicle as ectodysplasin (Fig. 5). The expression of ectodysplasin and EDAR was also observed in sebaceous glands. The antibody used to detect ectodysplasin was against the C-terminus of the protein (EDA-A) and thus also able to recognize the cleaved form. Different expression patterns of ectodysplasin and EDAR were also observed. At 10 weeks of gestation, ectodysplasin was expressed in the epithelium of jaws but tooth buds were negative, whereas EDAR was expressed both in the epithelium of jaws and also in tooth buds (data not shown). The first signs of EDAR expression were detected at week 7 in the epidermis of fingers (data not shown). Sense RNA probes used as controls in in situ hybridizations did not show any specific signal (data not shown).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Even though the cloning of the EDA (ED1) gene disclosed a previously unknown player in epithelial morphogenesis, the mechanisms of action of the EDA gene or its product ectodysplasin have remained unclear. Its recognition as a trimeric membrane protein with a TNF-like domain (6) and the cloning of two putative receptors, EDAR and XEDAR, suggested a novel signalling pathway. EDAR was identified through the positional cloning of the mouse mutant gene downless (dl) (8) and the subsequent finding that its human orthologue is mutated in some patients with autosomal dominant and recessive ectodermal dysplasias (9). Unexpectedly, a second receptor with a TNFR motif, called XEDAR, was cloned from the human X chromosome and two specific ligand–receptor pairs were recognized (12). Of the two ectodysplasin-A isoforms that differ only by two amino acids (5), A1 was shown to bind to the EDAR, whereas A2 bound to XEDAR (12). The specificity of ectodysplasin-A1 binding to EDAR in contrast to a number of other TNFR homologues was also shown by Tucker et al. (10). However, their in situ data suggested that ectodysplasin and EDAR were not expressed in overlapping domains in developing tooth (10,11), further suggesting that the ectodysplasin–EDAR binding may not be physiologically relevant in vivo. Thus the failure to detect a soluble form of ectodysplasin, and indeed doubts about its existence, have obscured the view. In this report, we show that both isoforms of ectodysplasin (EDA-A1 and EDA-A2) are specifically cleaved from transfected cells. We also show that EDAR protein is able to co-precipitate ectodysplasin-A1 isoform, suggesting that they form a ligand–receptor pair.
Proteolytic shedding of ectodysplasin
The members of the TNF-ligand family are type II membrane proteins. Most of them are involved in host defence and immune regulation, inducing cell death, survival, proliferation and differentiation (16,17). They can act either locally through direct cell-to-cell contacts or as soluble proteins capable of diffusing to more distant targets. Soluble forms are released from the cell surface by a proteolytic cleavage process known as ectodomain shedding. Ectodysplasin has a putative recognition site (R-X-K/R-R) for proprotein convertase proteases, such as furin, but no evidence of ectodysplasin shedding from the plasma membrane has been available. In order to clarify this issue, we made a novel antibody against the C-terminus of ectodysplasin-A and used it to show that ectodysplasin could be released into the medium from transfected cells. Furthermore we showed that the cleaved form is likely to exist as a trimer. Using deletion and mutation variants of ectodysplasin, we found that the furin-like recognition site is essential for cleavage.
Furin is the mammalian prototype of a family of serine proteases called subtilisin-like proprotein convertases (SPCs) (18). They play an important role in the processing of hormones, growth factors and adhesion molecules as well as a number of other protein precursors (18–20). To date at least seven members of this family have been identified. Three members of this family, PC1/3, PC2 and PC4, are expressed exclusively in endocrine tissues or in testis and thus are not relevant protease candidates to cleave ectodysplasin. In contrast furin, PACE4, PC5/6 and PC7 are expressed more ubiquitously. The substrate specificity of the SPC family members is not yet well characterized as they all have similar sequence specificity. However, this problem has been overcome to some extent by the use of cell lines that are unable to produce functional furin; a human colon carcinoma cell line, LoVo, or a mutant Chinese hamster ovary line, RPE.40 (18–20). A furin-like cleavage site is found in the TNF ligands BAFF, APRIL and Tweak, and the shedding of BAFF at this site has been demonstrated (21,22). Recently, the ectodomain cleavage at a furin-like site has been shown also in the transmembrane collagen types XIII (23,24) and XVII (25). Both collagens have been postulated to have a role as cell adhesion proteins, but the biological meaning of their soluble forms is still unclear.
Because of the low expression level of ectodysplasin (1,15) and technical limitations, neither we nor others have been able to detect even the unprocessed protein in tissue extracts or primary keratinocytes by western blotting or immunoprecipitation. Thus, all cleavage experiments must be done with transfected cells. However, we believe that the cleavage has biological significance in vivo, because in at least seven EDA patients, one of two adjacent arginine residues (residues 155 and 156) at the putative furin-like recognition site has been changed into histidine or cysteine by a missense mutation in exon three (5,26,27). In the present study we showed that Arg-His mutations (residues 153 and 156) at the furin-like recognition site greatly disturb cleavage. It is known that in a number of proteins, such as proalbumin, insulin pro-receptor, pro-factor IX and fibrinogen A 
-chain, mutations at the furin cleavage site are responsible for a variety of genetic disorders (18).
In the present study we showed that the shed ectodysplasin exists as a trimer like other soluble TNF-ligands. When our study was in progress, others showed that at least ectodomains of ectodysplasin expressed in transfected cells are able to interact with the membrane-anchored receptors EDAR and XEDAR (10,12). Furthermore, Yan et al. (12) used a soluble form of ectodysplasin (amino acids 179–391), which is only slightly shorter than the cleaved form (amino acids 160–391). Thus, most likely shedding produces an active ligand able to interact with its receptor. The proportion of cleaved ectodysplasin seems to be significant since it can be detected in the medium already after 4 h without any metabolic labelling and immunoprecipitation enrichment. On the other hand, we found by surface labelling and immunofluorescence stainings (data not shown) that in transfected cells there is still a lot of membrane-bound ectodysplasin left. We found the cleaved form also from transfected keratinocytes, suggesting that these epithelial cells have proteases required for shedding.
Expression of ectodysplasin and EDAR
We report here for the first time the expression of ectodysplasin and EDAR in the developing human skin. According to our data, these two proteins are expressed in partially overlapping or adjacent tissues in skin and are thus able to interact with each other. In the developing skin at week 20 of gestation, ectodysplasin is expressed throughout the epidermis whereas EDAR is more restricted to the periderm. Moreover, EDAR is found in association with placodes, thickenings of epithelia where the formation of epidermal appendages begins. In the developing mouse skin, at E14/E15 before the appearance of placodes, expression becomes focally upregulated (8,12), as observed in human skin, and remains high in the follicle epithelial cells that retain contact with the dermal condensation at the base of epidermal downgrowth (8). Ectodysplasin-A1 and -A2 isoforms may have distinctive temporal and spatial expression patterns in the developing murine skin, suggesting that these two forms may have distinct roles in the developing skin (12). The expression of XEDAR has not yet been studied extensively. It will be interesting to study whether EDAR and XEDAR have clear differences in their expression profiles.
Function of ectodysplasin in ectodermal morphogenesis
TNF signalling has been implicated previously in host defence and immunity, but the identification of the ED1 (EDA) and EDAR genes (Ta and dl in mice) defective in syndromes affecting ectodermal development implied another role. Based on results cumulated over the past few years, we suggest a model for ectodysplasin action (Fig. 6). Ectodysplasin and EDAR form the first TNF–TNFR pair known to regulate embryonic morphogenesis (11). The expression of EDAR is induced by activinßA, whereas Wnt6 induces ectodysplasin expression (11). The downstream signalling of ectodysplasin has also been studied. Like the classical death domain-containing TNF receptors, EDAR is capable of activating the nuclear factor-
B (NF-B) pathways (12,28) (P. Koppinen, J. Pispa, J. Laurikkala, I. Thesleff and M. L. Mikkola, manuscript in preparation). Our observations (6) (K. Pulkkinen, O. Elomaa and J. Kere, unpublished data) and those of Koppinen et al. suggest that ectodysplasin-EDAR signalling is not likely to induce apoptosis. Thus, we propose that ectodysplasin as a soluble ligand mediates a positive signal for cell survival, growth or differentiation.
|
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
Production of antibodies
To produce polyclonal antibodies against the EDAR and ectodysplasin proteins, synthetic peptides were prepared and used to immunize rabbits (Genosys). These consisted of an 18 amino acid peptide SKHTTFFGAIRLGEAPAS (residues 374–391) from the C-terminus of ectodysplasin-A (5), a 16 amino acid peptide EWAGVVPPASQPHAAS (residues 433–448) from the C-terminus of EDAR and a 20 amino acid peptide CEGFFRATVLTPGDMENDAE (residues 93–112) from the N-terminal region of EDAR (9). Antisera were purified with the peptide antigen bound to matrix using standard procedures. The affinity columns for the purification of EDAR antisera were prepared by Genosys and for the ectodysplasin-A antiserum according to the manufacturer’s instructions (Epoxy-activated Sepharose 6B; Pharmacia Biotech). Antibodies against the N-terminus of ectodysplasin (A17) and the 135 amino acid EDA-O protein have been described previously (6,29).
Preparation of expression constructs
The constructs coding the full-length forms of human ectodysplasin-A, pCMV-EDA-A1 and pCMV-EDA-A2, and its murine homologue Tabby, pCI-neo-Ta, have been described previously (5,7). To prepare the deletion constructs dSp1, dSp2, dSp3 and dFur of ectodysplasin-A, two PCR fragments were first amplified from each side of the deletion using the full-length EDA-A2 cDNA as template (5). The dSp1 construct lacks residues 79–179 (nucleotides 477–779), dSp2 lacks residues 79–118 (nucleotides 477–596), dSp3 lacks residues 119–141 (nucleotides 597–665) and dFur lacks residues 152–159, RRVRRNKR (nucleotides 696–719). The 5' portions of mutant variants were amplified using the sense primer E1Aext (nucleotides 210–231) containing the initiation codon and deletion-specific antisense primers. The 3' portions were amplified using deletion-specific sense primers complementary to antisense primers and a primer called ECR (nucleotides 1426–1444) from the 3' side of the stop codon. The primary PCR products were mixed and used as a template in secondary PCRs with the E1Aext and ECR primers. The Mut construct of ectodysplasin-A2 encoding arginine
histidine mutations (residues 153 and 156) was prepared by PCR using the same method as applied for deletion constructs mutation engineered into the primers (for residue 153, nucleotides 699–701 and for residue 156, nucleotides 708–710, CGC–CAC). The PCR fragments containing deletions were directly cloned into the pcDNA3-V5-His vector using the TOPO TA cloning kit (Invitrogen). All constructs used in the study were confirmed by DNA sequencing.
Human EDAR cDNA was amplified with PCR from fetal liver cDNA (Clontech) using primers derived from the EDAR cDNA sequence (9). The full-length coding region of EDAR (nucleotides 413–1800) and the region encoding the putative ectodomain, EDARec (nucleotides 413–999) were cloned into the pcDNA3-V5-His vector (TOPO TA cloning kit, Invitrogen) which encodes the C-terminal V5-6 x His epitope tags.
Cell culture and transfections
293A and COS-1 cells were grown in DMEM-GLUTAMAX medium (Gibco Laboratories) supplemented with 10% fetal calf serum (FCS), 100 IU/ml penicillin and 100 µg/ml streptomycin. Cells were transiently transfected using the lipofection method with the FuGENE 6 reagent (Boehringer Mannheim) according to the manufacturer’s suggestions. Transfectants were used after 16–48 h for western blotting, immunoprecipitation or immunofluorescence. For shedding assays, the culture media of transfectants were replaced with fresh serum free media or media containing 2% FCS. At various time points (4–48 h), media samples were harvested and prepared for SDS–PAGE.
Preparation of samples for SDS–PAGE and western blotting
Transfected cells were washed with PBS and lysed into Laemmli loading buffer. For reducing conditions, a loading buffer with 5% ß-mercaptoethanol was used and samples were boiled for 3 min. Lysates were run on 10% or 8% SDS–PAGE and electroblotted. Media samples of transfectants were harvested on ice, the protease inhibitors aprotinin (1 µg/ml), leupeptin (1 µg/ml), phenylmethysulfonyl fluoride (100 µg/ml), and 1,10-phenanthroline (1 mM) (Sigma) were added, and the samples were centrifuged at +4°C. Supernatants were mixed with Laemmli loading buffer and analysed on SDS–PAGE followed by western blotting. For trimerization assays medium samples were concentrated and changed into PBS buffer with Sentricon-10 concentrator (Millipore). For crosslinking, samples in PBS were incubated with 0.25 mM DSS (Pierce) or with 0.35 mM DFDNB (Sigma) for 30 min at +4°C, after which the reaction was stopped by adding (1/10 volume) 1 M Tris–HCl pH 7.5. The samples were mixed with Laemmli loading buffer with lower SDS concentration (0.5%) and run at +4°C. In immunostainings, the anti-EDA-C-terminal serum was diluted 1:5000 and the affinity purified antibodies anti-EDAR-C-terminal and anti-EDAR-N-terminal were both used at the concentration of 1 µg/ml. Peroxidase-conjugated anti-rabbit IgG (Sigma) was used as the secondary antibody. For biotinylated samples, peroxidase-conjugated streptavidin (Boehringer Mannheim) was applied. The proteins were visualized using the enhanced chemiluminescence method. Staining with preimmune sera was used as a negative control.
Cell surface biotinylation and streptavidin precipitation
For biotinylation, transfected cells were rinsed with ice-cold PBS and cooled on this buffer on ice. Cell surface proteins were labelled for 30 min on ice with sulfo-N-hydroxysuccinimide-biotin using the cellular labelling and immunoprecipitation kit (Roche Diagnostics). For streptavidin precipitation, cells were lysed into a lysis buffer (50 mM sodium borate pH 8, 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 100 µg/ml phenylmethysulfonyl fluoride, 1 µg/ml aprotinin and 1 µg/ml leupeptin) followed by centrifugation. The cell lysate was incubated with streptavidin-agarose (Sigma) for 1 h at +4°C, after which the agarose was centrifuged. Streptavidin precipitants were washed three times with the lysis buffer and twice with 10 mM Tris pH 7.5 and 0.5% NP-40 buffer. Precipitants were eluted by heating in SDS loading buffer for 5 min. Total lysates and avidin-precipitated proteins were analysed by SDS–PAGE, followed by immunoblotting with anti-EDA-antibodies. The HCR expression construct (in pCMV5 vector) and anti-HCR antibodies used as a control for surface labelling will be described separately by K. Asumalahti, S. Suomela, T. Laitinen, O. Elomaa, U. Saarialho-Kere and J. Kere (unpublished data).
Protein co-precipitations
The ectodomain of EDAR (EDARec) was expressed as a secreted V5-His fusion protein in transiently transfected cells. The protein was partially purified from cell culture media using the Xpress system protein purification protocol with ProBond resin (Invitrogen). The trapped proteins were eluted from the resin with 200–500 mM imidazole. The high concentration fractions were collected and applied to a PD-10 column (Pharmacia Biotech) eluted in PBS to remove imidazole. Samples were concentrated with Sentricon-10 concentrators (Millipore) and the protein concentration was estimated by measuring the absorbance at 280 nm. Purified proteins were verified by SDS–PAGE and western blotting. For precipitation studies, cells (3 x 106) transfected with the full-length ectodysplasin-A1 cDNA were lysed into lysis buffer (50 mM sodium borate pH 8, 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate, 100 µg/ml phenylmethysulfonyl fluoride, 1 µg/ml aprotinin and 1 µg/ml leupeptin) and biotinylated using the cellular labelling and immunoprecipitation kit (Roche Diagnostics). Biotinylated lysate (1 ml) was centrifuged and 50 µl or 100 µl was used in receptor precipitation with 0.5 µg of EDARec-V5-His. Binding volumes were increased to 0.3 ml using PBS with bovine serum albumin (0.25%) and Tween (0.01%). The binding reaction proceeded overnight at +4°C, after which the reaction volume was adjusted to 0.6 ml with dilution buffer (50 mM Tris pH 7.5, 150 mM NaCl and 0.1% Nonidet P-40) for immunoprecipitation. Complexes were captured with the anti-V5-antibody (1 µg for 1 h) followed by protein A-agarose (50 µl for 1 h) precipitation. In the control precipitation, anti-V5-antibody and protein A-agarose alone were incubated with the biotinylated ectodysplasin-A1. Precipitates were washed extensively five times using buffers recommended by the manufacturer (Roche Diagnostics) after which they were resolved on SDS–PAGE and electroblotted. Biotinylated proteins were detected with peroxidase-conjugated streptavidin and enhanced chemiluminescence.
In situ hybridization and immunohistochemistry
In situ hybridization was performed on formalin-fixed, paraffin-embedded specimens. In situ hybridization and immunohistochemical studies on fetal samples were approved by the ethics committees of the Departments of Medical Genetics and Dermatology, University of Helsinki. To prepare cRNA probes, PCR fragments of 502 bp (nucleotides 449–950) and 553 bp (nucleotides 1226–1779) were amplified from the human EDAR cDNA (9) with primers containing the T7 and Sp6 promoter sequences at opposite ends for in vitro transcription. Preparation of the 549 bp ectodysplasin-specific probe (nucleotides 214–762 of EDA-O cDNA) has been described earlier (1). Probes were labelled with (-33P)-UTP (Amersham) using the Riboprobe in vitro transcription systems (Promega). A sense probe was used as a control for non-specific hybridization in all experiments.
Deparaffinized sections were processed for immunohistochemistry according to standard procedures recommended by Vector Laboratories. Affinity purified antibodies were applied for 1 h at room temperature at the following concentrations: anti-EDA-C-terminal at 20 µg/ml; anti-EDAR-N-terminal and anti-EDAR-C-terminal both at 10 µg/ml. Pre-immune sera or normal rabbit IgG were used as controls. Bound antibodies were detected using the avidin-biotin complex method (Vectastain Elite ABC kit; Vector Laboratories) with 3,3'-diaminobenzidine tetrahydrochloride as the chromogenic substrate. Sections were counterstained with hematoxylin.
|
ACKNOWLEDGEMENTS |
|---|
We thank Irma Thesleff and members of her group for valuable discussions and sharing data and material in the course of the study. We thank Riitta Herva for providing tissue samples and Seija Hartikainen and Alli Tallqvist for excellent technical assistance. This work was supported by the Sigrid Jusélius Foundation, Academy of Finland and Helsinki University Central Hospital Research Funds.
|
FOOTNOTES |
|---|
+ These authors contributed equally to this work
To whom correspondence should be addressed at: Finnish Genome Center, PO Box 21 (Tukhlmankatu 2), 00014 University of Helsinki, Helsinki, Finland; Tel: +358 9 1912 6538; Fax: +358 9 1912 6789; Email: juha.kere@helsinki.fi
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSREFERENCES |
|---|
1 Kere, J., Srivastava, A.K., Montonen, O., Zonana, J., Thomas, N., Ferguson, B., Munoz, F., Morgan, D., Clarke, A., Baybayan, P. et al. (1996) X-linked anhidrotic (hypohidrotic) ectodermal dysplasia is caused by mutation in a novel transmembrane protein. Nature Genet., 13, 409–416.[Web of Science][Medline]
2 Pinheiro, M. and Freire-Maia, N. (1994) Ectodermal dysplasias: a clinical classification and a causal review. Am. J. Med. Genet., 53, 153–162.[Web of Science][Medline]
3 Ferguson, B.M., Brockdorff, N., Formstone, E., Ngyuen, T., Kronmiller, J.E. and Zonana, J. (1997) Cloning of Tabby, the murine homolog of the human EDA gene: evidence for a membrane-associated protein with a short collagenous domain. Hum. Mol. Genet., 6, 1589–1594.
4 Srivastava, A.K., Pispa, J., Hartung, A.J., Du, Y., Ezer, S., Jenks, T., Shimada, T., Pekkanen, M., Mikkola, M.L., Ko, M.S. et al. (1997) 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, 94, 13069–13074.
5 Bayes, M., Hartung, A.J., Ezer, S., Pispa, J., Thesleff, I., Srivastava, A.K. and Kere, J. (1998) The anhidrotic ectodermal dysplasia gene (EDA) undergoes alternative splicing and encodes ectodysplasin-A with deletion mutations in collagenous repeats. Hum. Mol. Genet., 7, 1661–1669.
6 Ezer, S., Bayes, M., Elomaa, O., Schlessinger, D. and Kere, J. (1999) Ectodysplasin is a collagenous trimeric type II membrane protein with a tumor necrosis factor-like domain and co-localizes with cytoskeletal structures at lateral and apical surfaces of cells. Hum. Mol. Genet., 8, 2079–2086.
7 Mikkola, M.L., Pispa, J., Pekkanen, M., Paulin, L., Nieminen, P., Kere, J. and Thesleff, I. (1999) Ectodysplasin, a protein required for epithelial morphogenesis, is a novel TNF homologue and promotes cell-matrix adhesion. Mech. Dev., 88, 133–146.[Web of Science][Medline]
8 Headon, D.J. and Overbeek, P.A. (1999) Involvement of a novel Tnf receptor homologue in hair follicle induction. Nature Genet., 22, 370–374.[Web of Science][Medline]
9 Monreal, A.W., Ferguson, B.M., Headon, D.J., Street, S.L., Overbeek, P.A. and Zonana, J. (1999) Mutations in the human homologue of mouse dl cause autosomal recessive and dominant hypohidrotic ectodermal dysplasia. Nature Genet., 22, 366–369.[Web of Science][Medline]
10 Tucker, A.S., Headon, D.J., Schneider, P., Ferguson, B.M., Overbeek, P., Tschopp, J. and Sharpe, P.T. (2000) Edar/Eda interactions regulate enamel knot formation in tooth morphogenesis. Development, 127, 4691–4700.[Abstract]
11 Laurikkala, J., Mikkola, M., Mustonen, T., Åberg, T., Koppinen, P., Pispa, J., Nieminen, P., Galceran, R. and Thesleff, I. (2001) TNF signaling via ligand-receptor pair ectodysplasin and edar controls the function of epithelial signaling centers and is regulated by Wnt and activin during organogenesis. Dev. Biol., 229, 443–455.[Web of Science][Medline]
12 Yan, M., Wang, L.C., Hymowitz, S.G., Schilbach, S., Lee, J., Goddard, A., de Vos, A.M., Gao, W.Q. and Dixit, V.M. (2000) Two-amino acid molecular switch in an epithelial morphogen that regulates binding to two distinct receptors. Science, 290, 523–527.
13 Tanaka, M., Suda, T., Takahashi, T. and Nagata, S. (1995) Expression of the functional soluble form of human fas ligand in activated lymphocytes. EMBO J., 14, 1129–1135.[Web of Science][Medline]
14 Asumalahti, K., Laitinen, T., Itkonen-Vatjus, R., Lokki, M.L., Suomela, S., Snellman, E., Saarialho-Kere, U. and Kere, J. (2000) A candidate gene for psoriasis near HLA-C, HCR (Pg8), is highly polymorphic with a disease-associated susceptibility allele. Hum. Mol. Genet., 9, 1533–1542.
15 Montonen, O., Ezer, S., Saarialho-Kere, U.K., Herva, R., Karjalainen-Lindsberg, M.L., Kaitila, I., Schlessinger, D., Srivastava, A.K., Thesleff, I. and Kere, J. (1998) The gene defective in anhidrotic ectodermal dysplasia is expressed in the developing epithelium, neuroectoderm, thymus, and bone. J. Histochem. Cytochem., 46, 281–289.
16 Smith, C.A., Farrah, T. and Goodwin, R.G. (1994) The TNF receptor superfamily of cellular and viral proteins: activation, costimulation, and death. Cell, 76, 959–962.[Web of Science][Medline]
17 Wallach, D., Kovalenko, A.V., Varfolomeev, E.E. and Boldin, M.P. (1998) Death-inducing functions of ligands of the tumor necrosis factor family: a Sanhedrin verdict. Curr. Opin. Immunol., 10, 279–288.[Web of Science][Medline]
18 Nakayama, K. (1997) Furin: a mammalian subtilisin/Kex2p-like endoprotease involved in processing of a wide variety of precursor proteins. Biochem. J., 327, 625–635.
19 Steiner, D.F. (1998) The proprotein convertases. Curr. Opin. Chem. Biol., 2, 31–39.[Web of Science][Medline]
20 Bassi, D.E., Mahloogi, H. and Klein-Szanto, A.J. (2000) The proprotein convertases furin and PACE4 play a significant role in tumor progression. Mol. Carcinog., 28, 63–69.[Web of Science][Medline]
21 Chicheportiche, Y., Bourdon, P.R., Xu, H., Hsu, Y.M., Scott, H., Hession, C., Garcia, I. and Browning, J.L. (1997) TWEAK, a new secreted ligand in the tumor necrosis factor family that weakly induces apoptosis. J. Biol. Chem., 272, 32401–32410.
22 Schneider, P., MacKay, F., Steiner, V., Hofmann, K., Bodmer, J.L., Holler, N., Ambrose, C., Lawton, P., Bixler, S., Acha-Orbea, H. et al. (1999) BAFF, a novel ligand of the tumor necrosis factor family, stimulates B cell growth. J. Exp. Med., 189, 1747–1756.
23 Snellman, A., Tu, H., Väisänen, T., Kvist, A.P., Huhtala, P. and Pihlajaniemi, T. (2000) A short sequence in the N-terminal region is required for the trimerization of type XIII collagen and is conserved in other collagenous transmembrane proteins. EMBO J., 19, 5051–5059.[Web of Science][Medline]
24 Peltonen, S., Hentula, M., Hägg, P., Ylä-Outinen, H., Tuukkanen, J., Lakkakorpi, J., Rehn, M., Pihlajaniemi, T. and Peltonen, J. (1999) A novel component of epidermal cell-matrix and cell-cell contacts: transmembrane protein type XIII collagen. J. Invest. Dermatol., 113, 635–642.[Web of Science][Medline]
25 Schäcke, H., Schumann, H., Hammami-Hauasli, N., Raghunath, M. and Bruckner-Tuderman, L. (1998) Two forms of collagen XVII in keratinocytes. A full-length transmembrane protein and a soluble ectodomain. J. Biol. Chem., 273, 25937–25943.
26 Monreal, A.W., Zonana, J. and Ferguson, B. (1998) Identification of a new splice form of the EDA1 gene permits detection of nearly all X-linked hypohidrotic ectodermal dysplasia mutations. [Published erratum appears in Am. J. Hum. Genet. (1998) 63, 1253–1255.] Am. J. Hum. Genet., 63, 380–389.
27 Pääkkönen, K., Kere, J., Cambiaghi, S., Novelli, G., Penttinen, M. and Srivastava, A.K. (2001) The mutation spectrum of the EDA gene in X-linked anhidrotic ectodermal dysplasia. Hum. Mutat., in press.
28 Kumar, A., Eby, M.T., Sinha, S., Jasmin, A. and Chaudhary, P.M. (2000) Ectodermal dysplasia receptor activates the nuclear factor kappa B, c- Jun N-terminal kinase and cell death pathways and binds to ectodysplasmin A. J. Biol. Chem., 276, 2668–2677.
29 Ezer, S., Schlessinger, D., Srivastava, A. and Kere, J. (1997) Anhidrotic ectodermal dysplasia (EDA) protein expressed in MCF-7 cells associates with cell membrane and induces rounding. Hum. Mol. Genet., 6, 1581–1587.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  G. Veit, E. P. Zimina, C.-W. Franzke, S. Kutsch, U. Siebolds, M. K. Gordon, L. Bruckner-Tuderman, and M. Koch Shedding of Collagen XXIII Is Mediated by Furin and Depends on the Plasma Membrane Microenvironment J. Biol. Chem., September 14, 2007; 282(37): 27424 - 27435. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. P. Zimina, A. Fritsch, B. Schermer, A. Yu. Bakulina, M. Bashkurov, T. Benzing, and L. Bruckner-Tuderman Extracellular Phosphorylation of Collagen XVII by Ecto-Casein Kinase 2 Inhibits Ectodomain Shedding J. Biol. Chem., August 3, 2007; 282(31): 22737 - 22746. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C.-W. Franzke, P. Bruckner, and L. Bruckner-Tuderman Collagenous Transmembrane Proteins: Recent Insights into Biology and Pathology J. Biol. Chem., February 11, 2005; 280(6): 4005 - 4008. [Full Text] [PDF]
|
|
|
|
 |
|
 T. Mustonen, M. Ilmonen, M. Pummila, A. T. Kangas, J. Laurikkala, R. Jaatinen, J. Pispa, O. Gaide, P. Schneider, I. Thesleff, et al. Ectodysplasin A1 promotes placodal cell fate during early morphogenesis of ectodermal appendages Development, October 15, 2004; 131(20): 4907 - 4919. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Rouse, E. Siegfried, W. Breer, and G. Nahass Hair and Sweat Glands in Families With Hypohidrotic Ectodermal Dysplasia: Further Characterization Arch Dermatol, July 1, 2004; 140(7): 850 - 855. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C.-W. Franzke, K. Tasanen, L. Borradori, V. Huotari, and L. Bruckner-Tuderman Shedding of Collagen XVII/BP180: STRUCTURAL MOTIFS INFLUENCE CLEAVAGE FROM CELL SURFACE J. Biol. Chem., June 4, 2004; 279(23): 24521 - 24529. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Mayer, G. Boileau, and M. Bendayan Sorting of Furin in Polarized Epithelial and Endothelial Cells: Expression Beyond the Golgi Apparatus J. Histochem. Cytochem., May 1, 2004; 52(5): 567 - 580. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Newton, D. M. French, M. Yan, G. D. Frantz, and V. M. Dixit Myodegeneration in EDA-A2 Transgenic Mice Is Prevented by XEDAR Deficiency Mol. Cell. Biol., February 15, 2004; 24(4): 1608 - 1613. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C.-Y. Cui, M. Durmowicz, C. Ottolenghi, T. Hashimoto, B. Griggs, A. K. Srivastava, and D. Schlessinger Inducible mEDA-A1 transgene mediates sebaceous gland hyperplasia and differential formation of two types of mouse hair follicles Hum. Mol. Genet., November 15, 2003; 12(22): 2931 - 2940. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Laurikkala, J. Pispa, H.-S. Jung, P. Nieminen, M. Mikkola, X. Wang, U. Saarialho-Kere, J. Galceran, R. Grosschedl, and I. Thesleff Regulation of hair follicle development by the TNF signal ectodysplasin and its receptor Edar Development, March 7, 2003; 129(10): 2541 - 2553. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C.-Y. Cui, M. Durmowicz, T. S. Tanaka, A. J. Hartung, T. Tezuka, K. Hashimoto, M. S.H. Ko, A. K. Srivastava, and D. Schlessinger EDA targets revealed by skin gene expression profiles of wild-type, Tabby and Tabby EDA-A1 transgenic mice Hum. Mol. Genet., July 15, 2002; 11(15): 1763 - 1773. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. Thesleff and M. L. Mikkola Death Receptor Signaling Giving Life to Ectodermal Organs Sci. Signal., May 7, 2002; 2002(131): pe22 - pe22. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. K. Srivastava, M. C. Durmowicz, A. J. Hartung, J. Hudson, L. V. Ouzts, D. M. Donovan, C.-Y. Cui, and D. Schlessinger Ectodysplasin-A1 is sufficient to rescue both hair growth and sweat glands in Tabby mice Hum. Mol. Genet., December 1, 2001; 10(26): 2973 - 2981. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||