Logo of embojLink to Publisher's site
EMBO J. 2008 Mar 19; 27(6): 886–897.
Published online 2008 Feb 21. doi:  10.1038/emboj.2008.24
PMCID: PMC2274933

Perturbed desmosomal cadherin expression in grainy head-like 1-null mice


In Drosophila, the grainy head (grh) gene plays a range of key developmental roles through the regulation of members of the cadherin gene family. We now report that mice lacking the grh homologue grainy head-like 1 (Grhl1) exhibit hair and skin phenotypes consistent with a reduction in expression of the genes encoding the desmosomal cadherin, desmoglein 1 (Dsg1). Grhl1-null mice show an initial delay in coat growth, and older mice exhibit hair loss as a result of poor anchoring of the hair shaft in the follicle. The mice also develop palmoplantar keratoderma, analogous to humans with DSG1 mutations. Sequence analysis, DNA binding, and chromatin immunoprecipitation experiments demonstrate that the human and mouse Dsg1 promoters are direct targets of GRHL1. Ultrastructural analysis reveals reduced numbers of abnormal desmosomes in the interfollicular epidermis. These findings establish GRHL1 as an important regulator of the Dsg1 genes in the context of hair anchorage and epidermal differentiation, and suggest that cadherin family genes are key targets of the grainy head-like genes across 700 million years of evolution.

Keywords: cadherins, desmoglein, epidermis, grainy head, hair


Although the grh gene in Drosophila has been linked to a range of developmental events, its predominant expression in the developing cuticle underpins the best-characterized of its mutant phenotypes, which manifest as defects in cuticle integrity and wound repair (Bray and Kafatos, 1991; Mace et al, 2005). In the developing wing, grh mutants exhibit multiple hairs of abnormal polarity, analogous to the defects observed in mutants of the frizzled-dependent PCP pathway (Lee and Adler, 2004). It has been postulated that a cadherin family member, starry night (stan) (also known as flamingo (fmi)), is a direct target of GRH. Grh mutant clones in the wing also exhibited a delay in hair morphogenesis, which was not mimicked in the stan/fmi mutants (Lee and Adler, 2004). Another member of the cadherin gene family directly regulated by grh is E-cadherin/shotgun (Almeida and Bray, 2005). This regulation is critical in post-embryonic neuroblasts, where grh expression is essential for these cells to adopt the stem cell programme appropriate for their position within the central nervous system (Almeida and Bray, 2005).

Our interest has focused on the roles of the three mammalian homologues of grh, Grhl1–3 in the developing epidermis (Wilanowski et al, 2002; Ting et al, 2003b). All three genes are highly expressed in the developing mouse skin, with differing times of initial and maximal expression (Auden et al, 2006). We have previously shown that Grhl3 is essential for closure of the neural tube (Ting et al, 2003a), and for establishing and maintaining the epidermal barrier in mice (Jane et al, 2005; Ting et al, 2005). This latter function is mediated, in part, through the regulation of a protein/lipid cross-linking enzyme, analogous to the role of grh in regulating the Dopa decarboxylase gene in the fly cuticle. This finding emphasizes the degree of functional conservation between members of this gene family. We now report a further insight into these functional homologies in our generation and study of mice deficient for Grhl1. These animals display a delay in the initial appearance of their coat, coupled with defective hair anchoring to the follicle. They also exhibit palmoplantar keratoderma (PPK) with marked thickening of the epidermal layer of the paws. These phenotypes are unified by the demonstration of diminished expression of the desmosomal cadherin genes coding for desmoglein 1 (Dsg1), which we show are direct targets of GRHL1.


Grhl1-null mice exhibit an abnormal hair coat

Mice carrying a targeted Grhl1 allele were generated as described in the Supplementary data. Genotyping of offspring from Grhl1+/– intercrosses at weaning showed that Grhl1–/– mice were represented in normal Mendelian proportions. Of 218 pups tested, 53 were +/+ (24%), 120 were +/− (55%), and 45 were −/− (21%). Although the same in size at birth, at weaning, the Grhl1–/– pups were smaller than their littermates, and 2 months later these mice displayed no significant difference in size or weight (Supplementary Figure 2). Over 30% of the Grhl1–/– pups were devoid of hair at weaning, and the remainder show a marked paucity of hair growth (Figure 1A, upper panel). Growth of whiskers, nails, and teeth were normal in these animals. Later in life, recovery in hair growth was observed; however, all Grhl1-null animals exhibited a sparser coat throughout life compared with their wild type littermates. An example of most pronounced hair regrowth is shown in Figure 1A (lower panel). In addition, we noted that many Grhl1–/– mice displayed severe regional hair loss that appeared unrelated to the hair cycle (Figure 1B, upper panel). This aspect of the phenotype was related to grooming by cage-mates, and mice housed in isolation showed sparse hair coats without regional hypotrichosis (Figure 1B, lower panel). This suggested that hair anchoring may be defective in the Grhl1–/– animals. Examination of hair types revealed no differences in frequency or appearance between the wild-type and mutant animals (data not shown). In all other aspects, the Grhl1–/– mice were healthy, with normal fertility and longevity. The formation of the skin barrier and wound healing were unaffected in the Grhl1–/– animals (Supplementary Figure 3; data not shown).

Figure 1
Phenotype of Grhl1-deficient mice. (A) A representative Grhl1 mutant (–/–) and wild-type (+/+) littermate photographed at weaning (upper panel). Mice from a separate litter at 37 days (lower panel). (B) An example of overgrooming ...

Grhl1 is expressed in the developing epidermis and hair follicles

During embryogenesis, Grhl1 expression was first detectable by in situ hybridization at very low levels in the surface ectoderm at embryonic day 10.5 (E10.5) (Auden et al, 2006). By E15.5, Grhl1 was readily detectable in the skin and in the developing hair germs (Figure 2A). At later time points during development and in the adult, expression of Grhl1 in the skin was confined to the suprabasal layers of the epidermis and to the hair follicles (Figure 2B). In the whiskers at E17.5, Grhl1 expression was observed throughout the length of the follicle in the inner root sheath (IRS), but with sparing of the dermal papilla (Figure 2C). These expression patterns were confirmed in mice carrying a lacZ reporter gene fused in frame to the first codon of exon 1 of the Grhl1 gene (Supplementary Figure 1). In E14.5 embryos, β-galactosidase activity was observed in the skin and hair placodes (Figure 2E). In the adult, expression was identified in the suprabasal layers of the epidermis (Figure 2D), IRS of the hair follicle, but not in the dermal papilla (Figure 2F–G). Significant expression of Grhl1 was also noted during development in other tissues lined by squamous epithelium, including the oral cavity and anal canal. Expression was also observed in several other organs, including lung, kidney, and stomach (Auden et al, 2006).

Figure 2
Expression of Grhl1 in embryonic and adult skin and hair follicles. (AC) Embryonic skin sections probed with Grhl1-specific antisense riboprobe. Left: dark field photograph; right: light field photograph. (A) At E15.5, the Grhl1 transcript is ...

Grhl1-null mice show defective hair anchoring

To determine the underlying pathology in the Grhl1–/– animals, we examined dorsal skin histology in mice at 0, 7, 14, and 21 days (Supplementary Figure 4; data not shown). No abnormalities in the dermal or epidermal architecture were observed at these time points. In particular, the structure and density of hair follicles was identical in both Grhl1–/– and wild-type littermates. Scanning electron microscopy of Grhl1 mutant hair shafts did not detect any structural differences compared with wild-type hairs (Supplementary Figure 5). Skin sections from areas without obvious hair loss in older mice (3–7 months) were unremarkable, with the exception of sections taken from the palmar or plantar surfaces of the fore and hind paw, respectively (see below). Sections from the hypotrichotic regions of adult mice revealed areas containing empty cysts (Figure 3A). These were reminiscent of the empty dilated telogen hair follicles described in mice carrying a targeted deletion of the desmosomal gene, Dsg3 (Koch et al, 1998).

Figure 3
Analysis of the epidermis and hair in adult Grhl1−/− mice. (A) Histology of the skin from a hypotrichotic area of a 4-month-old Grhl1-null mouse revealing cysts (arrows) in telogen follicles. (B) Histology of horizontal cross-sections ...

We repeated the histological examination on skin from adult Grhl1–/– mice fixed in 70% ethanol. Processing in this way has been shown to display clefts between the IRS and outer root sheath (ORS) in Dsg-deficient mice that were not apparent following formalin fixing (Hanakawa et al, 2004). The formation of these clefts is due to shrinkage of the tissue caused by dehydration from the ethanol fixing. As shown in Figure 3B, cleft formation was evident between the IRS and ORS in Grhl1–/– mice, but not in wild-type littermate controls. We also examined hair anchorage using a tape-stripping test, in which adhesive tape was applied to the dorsal region of the trunk and peeled off in the direction of hair growth (Figure 3C; Koch et al, 1998). Despite the thinner hair covering of the Grhl1–/– mice, substantially more hair was lost from the mutant animals than their wild-type littermate controls. Hair numbers adherent to the tape per low power microscopic field varied between 400 and 500 in the mutant strain compared with 100–200 in the wild-type animals.

The rate of hair regrowth following depilation was identical in both wild-type and mutant mice (Supplementary Figure 6). Examination of depilated hairs removed during anagen revealed that the shafts remained intact, and had not broken at any point along the shaft. Toluidine blue staining of these hairs from wild-type and mutant mice uncovered a marked difference (Figure 3D). Adherent to the proximal tip of the wild-type hair was an intensely stained ball of tissue that consisted of the IRS, companion layer and ORS (Hanakawa et al, 2004). In contrast, the mutant hair displayed naked anagen hair bulbs lacking any of the usual tissue layers. This finding is reminiscent of the appearance of plucked anagen hairs in human patients with loose anagen hair syndrome (Baden et al, 1992). Consistent with this, we observed marked differences in the appearance of depilated follicles between wild-type and mutant mice, with retention of the IRS in the Grhl1-null animals (Figure 3E).

Defining the GRHL1 consensus DNA-binding site

The molecular events underlying the defects observed in the Grhl1-null hair follicles presumably reflected altered expression of target genes of this transcription factor. To date, the consensus DNA-binding sequence of this factor remained unknown. To address this, we initially performed cyclic amplification and selection of targets (CASTing) (Wright et al, 1991). For this purpose, we used cellular extracts from a human placental choriocarcinoma cell line JEG-3 that we had previously shown to contain abundant GRHL1 (see Materials and methods). We had successfully employed this strategy to define the consensus DNA-binding site for GRHL3 (Ting et al, 2005). After six cycles, the population of selected binding sites was cloned and individual isolates sequenced and aligned (Supplementary Figure 7). The defined GRHL1 DNA-binding consensus sequence (AACCGGTT) was identical to that defined for GRHL3, and also matched the consensus sequence for Drosophila GRH DNA binding, which we had previously identified by alignment of multiple GRH-responsive gene regulatory regions (Wilanowski et al, 2002; Ting et al, 2005). Of note, the first of the two cytosines and the second of the guanines were invariant in both GRHL1 and GRHL3 CASTing assays.

Reduced expression of the Dsg1 genes in Grhl1-deficient mice

Our histological analyses, as well as analyses of others, suggested that defects in hair anchoring underpinned the Grhl1-deficient phenotype. Previous studies have implicated members of the desmosomal cadherin family in this process (Koch et al, 1998; Hanakawa et al, 2004), and this, coupled with the links between grh and cadherin family members in Drosophila prompted us to examine the potential role of these genes in the Grhl1-deficient mice.

The desmosomal cadherins are adhesion molecules that are critical components of the intercellular junctions in the epidermis, the desmosomes (Garrod et al, 2002; Yin and Green, 2004). The family consists of the desmocollins (Dsc) and Dsgs, each of which has multiple subtypes in mice and humans (Dsg1–4 and Dsc1–3) that are expressed in a differentiation-dependent pattern (Garrod et al, 2002; Yin and Green, 2004). In addition, mice have three Dsg1 isoforms (Dsg1α, Dsg1β, and Dsg1γ) (Kljuic and Christiano, 2003; Pulkkinen et al, 2003; Whittock, 2003), which exhibit varying expression patterns in the differentiating cell layers in the epidermis (Brennan et al, 2004). In the hair follicle, all three Dsg1 isoforms are present, but are subject to significant hair cycle-dependent changes (Brennan et al, 2004). To determine whether the desmosomal cadherins were potential GRHL1 target genes, we surveyed the regulatory regions of these genes for potential GRHL1-binding sites. As shown in Figure 4A, the three mouse Dsg1 genes and the human DSG1 gene all contained a putative GRHL1-binding motif, with a conserved first cytosine and second guanine. The putative GRHL1-binding site in the human DSG1 promoter was located 12–34 bp upstream of the four reported transcription start sites (Adams et al, 1998). Although the locations of the transcription start sites in the mouse Dsg1 genes have not been reported, there is significant homology between the three mouse genes and the human gene in the promoter regions, suggesting that GRHL1-binding sites also reside in the proximal Dsg1 promoters in the mouse. No GRHL1 consensus-binding sites were found in the human or mouse Dsg2, Dsg3, or Dsg4 genes, or in any of the Dsc genes from these species (data not shown).

Figure 4
Mouse and human Dsg1 genes are direct targets of GRHL1 regulation. (A) Alignment of the promoter regions of Dsg1 genes from mouse and human to the consensus DNA-binding sequence for GRHL1. Conserved bases are bolded and the invariant cytosine and guanine ...

Our earlier RNA and reporter gene data (Figure 2) had demonstrated that Grhl1 was expressed in the IRS of the hair follicle, a region also known to express the Dsg1 genes (Brennan et al, 2004; Hanakawa et al, 2004). To confirm this at a protein level, and to visualize the coexpression of GRHL1 and DSG1, we performed immunohistochemistry on single sections of epidermis from wild-type mice using antisera to both proteins and the ImmPRESS multiple antigen labelling system (see Material and methods). GRHL1 and DSG1 were coexpressed in the IRS (Figure 4B) and the sebaceous glands (data not shown).

To ascertain whether the level or distribution pattern of DSG1 protein was altered in the Grhl1−/− animals, we performed immunohistochemistry on wild-type and mutant skin with antisera to DSG1 (Figure 4C). A marked reduction in protein expression was observed in both the hair follicles and interfollicular epidermis of the Grhl1 mutants compared with the controls. We then performed northern analysis using RNA from skin with a probe that would hybridize to all three mouse Dsg1 genes. We also examined the expression of Dsg3, which lacks a GRHL1-binding site in its regulatory regions, in these samples. As shown in Figure 4D, expression of Dsg1 was specifically reduced in the mutant strain. Quantitation of the signals by densitometry revealed that Dsg1 levels were threefold lower in the Grhl1-deficient skin. In contrast, the expression of Dsg3 was not altered in the Grhl1-deficient skin. To confirm our northern analysis, and to investigate whether changes in expression of other cell adhesion genes may contribute to the observed phenotype, we studied the expression of Dsg1, Dsg2, Dsg3, E-cadherin (Cdh1), P-cadherin (Cdh3), desmoplakin (Dsp), and desmocollin 1 (Dsc1) in the Grhl1-null mice by Q-RT–PCR (Figure 4E). Dsg1 expression was significantly reduced in the Grhl1−/− epidermis. No significant differences in expression levels of the other genes were observed in the mutants compared with their wild-type littermates, with the exception of Dsg2, which was upregulated twofold in the mutant epidermis. The upregulation of Dsg2 is unlikely to contribute significantly to the desmosomal cadherin protein pool, as the levels of DSG2 protein in adult skin are extremely low in normal adult skin (Mahoney et al, 2006). We also performed a microarray (Illumina) experiment, using epidermal-derived RNA from three wild-type and three mutant mice. We did not find any significant changes in the expression of cell adhesion genes in the Grhl1-knockout mice, consistent with the reduced sensitivity of this methodology as compared to Q-RT–PCR. A number of genes were found to be differentially expressed, but none of these have been linked to hair anchoring (data not shown).

To confirm the Dsg1 genes as direct targets of GRHL1, we performed electrophoretic mobility shift assays (EMSA) (Figure 4F). The putative GRHL1-binding motif (GACTGGTT) is perfectly conserved, together with 6 bp upstream and 12 bp downstream flanking sequences, in three of the Dsg1 promoters: mouse Dsg1α, mouse Dsg1γ, and human DSG1 (Figure 4A). In the mouse Dsg1β promoter, this motif is slightly different (AACTGGTT), although the flanking sequences are still conserved. Consequently, we designed two probes for the EMSA, one derived from the human DSG1 promoter, and the other based on the mouse Dsg1β promoter. Using nuclear extract derived from the JEG-3 cell line, we demonstrated several DNA/protein complexes with the mouse probe (lane 1). The upper band (arrowed) was specifically supershifted with the addition of anti-GRHL1 antisera (lane 2). This band was specific, as it was competed off with excess unlabelled mouse Dsg1β probe (lane 3). It was also competed off with excess unlabelled human DSG1 probe (lane 4). The human DSG1 probe revealed a less complex array of protein/DNA complexes, but the GRHL1 complex was retained (lane 5) and specifically supershifted with anti-GRHL1 antisera (lane 6). The GRHL1 band was also competed away with excess unlabelled, human DSG1 (lane 7) and mouse Dsg1β probe (lane 8), indicating that GRHL1 binds specifically to the Dsg1 promoters in mice and humans.

To confirm this binding in an in vivo setting, we performed chromatin immunoprecipitation (ChIP) using the human keratinocyte cell line, HaCaT (Figure 4G). We initially determined by RT–PCR that both GRHL1 and DSG1 are expressed in this cell line (data not shown). Crosslinked protein/chromatin complexes were immunoprecipitated with two different anti-GRHL1 antisera (lanes 4 and 5), and after reversal of the crosslinking the precipitates were subjected to PCR with primers flanking the GRHL1-binding site in the DSG1 promoter (upper panel). The promoter from the MYOD gene (which is not expressed in these cells) was used as a negative control (lower panel) (Sawado et al, 2001). Pre-immune sera from one of the GRHL1 immunized rabbits (lane 3) and no sera (lane 2) served as additional controls. Occupancy of the DSG1 promoter by GRHL1 was evident in this cell line. The input chromatin for each experiment is shown in lane 1.

We also performed reporter gene assays with the individual mouse and human Dsg1 promoters in the setting of enforced Grhl1 expression (Figure 4H). The human embryonic kidney cell line 293T was cotransfected with the individual human and mouse Dsg1 promoters linked to the firefly luciferase gene and a mammalian expression vector carrying the mouse Grhl1 gene. An empty vector was transfected with the reporter constructs as a control, and transfection efficiency was established with a plasmid expressing Renilla luciferase. All four promoters were trans-activated to varying extents by Grhl1 in this context, consistent with the reduction in Dsg1 expression observed in the Grhl1-deficient mice. We then examined the effect of mutating the nearly optimal GRHL1-binding site in the Dsg1 promoter and found that loss of this site in isolation did not ablate GRHL1-dependent activation (data not shown). We reexamined the region of the promoter that we identified by ChIP to contain the GRHL1-binding site(s) and, eventually, identified two other sites matching the binding sequences previously identified in our CASTing assay, which were also conserved in the human DSG1 promoter. This suggests, as with many other transcription factors, that multiple binding sites are present in key regulatory regions. To investigate the specificity of the response to GRHL1, we examined whether the Dsg1 promoter was also responsive to GRHL3 in transfection experiments (Supplementary Figure 8). Despite the fact that GRHL1 and GRHL3 have identical in vitro-defined DNA-binding consensus sequences, GRHL3 was unable to activate the Dsg1 promoter.

Grhl1-null mice exhibit palmoplantar keratoderma

In humans, mutations in the DSG1 gene result in the dominantly inherited disorder striate PPK (SPPK), typified by hyperkeratosis at sites of pressure and abrasion, the palms, and soles (Rickman et al, 1999; Hunt et al, 2001). Histological examination of the palmopalmar surface of the paws of the Grhl1-null mice revealed marked thickening of the stratum corneum, consistent with PPK (Figure 5A). Analysis of expression of different keratins in humans with DSG1-SPPK revealed a reduction of K5 and K14, with weak staining of K10 remaining in the lower suprabasal layers (Wan et al, 2004). Studies in the Grhl1-mutant mice also revealed a marked reduction in K14 expression and a less dramatic reduction in K5 expression compared with wild-type controls (Figure 5B and C). Expression of K10 was also altered, with weak staining confined to the upper spinous and granular layers (Figure 5D). However, the levels of involucrin, a differentiation marker, were decreased (Figure 5E), which was inconsistent with DSG1-SPPK patient data (Wan et al, 2004).

Figure 5
Grhl1-null mice exhibit features of PPK. (A) Histology of the palmar surface of the forepaw of a Grhl1-null (–/–) and wild-type (+/+) littermate at 7 months of age. (BE) Immunohistochemistry (C, D) or immunofluorescence ...

Ultrastructural studies of the interfollicular epidermis unveiled more abnormalities in the Grhl1-null mice (Figure 5F). Fewer desmosomes were identifiable (Table I), and these appeared to be shorter and less well organized (Figure 5F). We then examined desmosomal number in cultured wild-type and mutant epidermis following treatment with EGTA, which has no effect on normal epidermis, but induces dissociation in the setting of desmosomal instability (Garrod et al, 2005). Consistent with our other findings, EGTA treatment led to a further reduction in desmosomal number (Table I).

Table 1
Comparison of desmosome numbers in Grhl1+/+ and Grhl1−/− epidermis untreated or treated with EGTA


Many cellular functions require stringent control of attachment to, and detachment from, other cells or biological surfaces. An example of dynamic adhesive interactions is anchoring of the hair shaft to the hair follicle, in which the desmosomal cadherins have a pivotal role. The studies reported here implicate the transcription factor Grhl1 in the regulation of hair anchoring and reveal a direct link between this factor and the human and mouse Dsg1 genes. Expression of Grhl1 in the mouse hair follicle was observed in the IRS of the follicle, but not in the dermal papilla cells, mirroring the pattern of expression of the Dsg1 genes (Brennan et al, 2004). Mice deficient in Grhl1 showed a delay in the appearance of their coat, sporadic regional hair loss, and excess hair loss with tape stripping and abnormal clefting between the IRS and ORS, which are the features of loss of Dsg1 expression (Hanakawa et al, 2004). Northern blotting and Q-RT–PCR confirmed that the levels of Dsg1 were reduced in the mutant mice compared with wild-type littermate controls. Sequence analysis, DNA binding, and ChIP experiments demonstrated that the mouse and human Dsg1 promoters were direct targets of GRHL1, and transfection assays confirmed the ability of GRHL1 to activate these regulatory elements. These findings suggest that GRHL1 is an important regulator of the Dsg1 genes in the context of mammalian hair anchorage.

Studies of mice carrying mutations in the desmosomal cadherin genes have illustrated the role of this family in the growth and anchoring of the hair follicle (Garrod et al, 2002). Mice carrying a targeted deletion of the Dsg3 gene show normal follicular neogenesis, and then develop defective hair anchoring in the quiescent (telogen) phase of the growth cycle (Koch et al, 1998). These animals also exhibit a skin phenotype that resembles pemphigus vulgaris (Koch et al, 1997). Mice deficient in Dsc1 and Dsg4 also exhibit skin and hair phenotypes, although defective hair anchorage is not a feature of these mutants (Chidgey et al, 2001; Kljuic et al, 2003). Mutations in the desmoplakin and plakoglobin genes manifest with PPK, hair abnormalities, and cardiomyopathy (McGrath, 2005). The hair phenotype in these patients is described as ‘woolly', and as such does not resemble the anchorage defects seen in the Grhl1-null mice. Histological examination of the heart from the Grhl1 mutant mice revealed no evidence of cardiomyopathy (data not shown). The presence of three Dsg1 genes in mice has presumably precluded their inactivation by standard gene targeting. Instead, inactivation of two of the DSG1 isoforms (DSG1α and DSG1β) by a serine protease produced by Staphylococcus aureus, exfoliative toxin A (ETA) in the setting of Dsg3 deficiency has been examined (Hanakawa et al, 2004). These mice exhibited a specific loss of anagen hair with tape stripping and separation between the ORS and IRS of the hair follicle, a histological feature we also observed in the Grhl1-deficient mice. In contrast, ETA-induced loss of these DSG1 isoforms in the setting of wild-type levels of DSG3 did not lead to defective hair anchoring or histological abnormalities, suggesting that DSG3 can compensate for loss of DSG1α and DSG1β. Our data suggest that DSG1γ, the ETA-resistant isoform of DSG1 (Brennan et al, 2004), also has a key role in this compensation, as reduced total Dsg1 expression in Grhl1-deficient mice in the setting of wild-type levels of Dsg3 leads to severe hair anchoring defects. Dsg1γ is highly expressed in the IRS of the lower hair follicle and in the sebaceous glands, analogous to Grhl1, and the Dsg1γ promoter contains a GRHL1 consensus site. Computational analyses suggest that the very low hydrophilic potential of the extracellular anchoring domain of DSG1γ may provide it with a stronger affinity for the cell membrane than the other DSG1 isoforms, underpinning its potential importance in hair shaft anchoring (Brennan et al, 2004).

The phenotype induced by the Grhl1-mediated reduction in total Dsg1 levels is unlike the ETA-treated Dsg3-null mice in that it manifests as poor hair anchorage at all stages of the hair cycle, not just in anagen. This was evident in our tape-stripping experiments, in which hair was readily removed irrespective of the growth cycle stage, and in the histological identification of empty dilated telogen hair follicles, reminiscent of the mice carrying a targeted deletion of the Dsg3 gene. We suspect that the delay in initial hair appearance was contributed to by maternal grooming, as hair growth improved at the time of weaning, while hair loss in young pups was variable between litters, but consistent within the same litter. Also commensurate with this, older mice showed sporadic regional hair loss, which was not observed when animals were housed in isolation. Although it is possible that the phenotypic variability observed was also influenced by modifier loci, these have not been described in the context of other depilation defects.

In addition to hair anchorage defects, the Grhl1-null mice exhibited keratoderma of the palmar surface of the front-paws and the plantar surface of the hind-paws accompanied by disordered expression of keratins 5, 10, and 14, a phenotype mimicking the appearance of SPKK in patients with DSG1 mutations. The expression of involucrin was decreased, which is in contrast to patients with SPPK, indicating that subtle species differences exist in this condition. Hair disorders have also not been reported in these patients, and we postulate that this may be due to a dosage effect, as DSG1 mutations in humans result in haploinsufficiency, with proteins levels predicted to be at 50% of normal. The lack of skin or hair phenotypes in the Grhl1+/– animals is not surprising, as these mice presumably exhibit much higher levels of Dsg1 expression than homozygous Grhl1−/− mutants.

Accumulating evidence suggests that the specific distribution pattern of desmosomal proteins is not merely a result of differentiation, but may in fact drive tissue morphogenesis and function (Green and Gaudry, 2000). Despite this, little is known about the spatio-temporal transcriptional regulation of desmosomal components. C/EBP transcription factors have been shown to regulate the expression of desmocollin genes, but not Dsgs (Smith et al, 2004). A region 4.2 kb 5′ of the transcription start site in the human DSG1 promoter is sufficient to confer expression in the epidermis and hair follicles (Adams et al, 1998). This region encompasses the GRHL1-binding site. In addition, protein kinase C (PKC) has been shown to positively regulate DSG1, with little effects on the other DSG genes (Denning et al, 1998). Although we are yet to establish whether PKC signalling acts through Grhl1, this pathway has been linked to the Grhl family in the context of neural tube closure, as demonstrated in the curly tail mice, which carry a mutation in the Grhl3 gene (Ting et al, 2003a; Cogram et al, 2004). Our data suggest that hair anchorage is sensitive to the absolute levels of Dsg expression. Consistent with this, a two- to threefold elevation in Dsg1 and Dsg3 levels in the plucked (pk) mutant mouse leads to impaired hair growth with retention of IRS in the follicle (Luo et al, 2005), and altered patterns of expression of individual Dsgs perturb epidermal differentiation (Elias et al, 2001; Merritt et al, 2002). These observations highlight the need for precise regulation of Dsg expression in the hair follicle.

Disruptions of DSG1 protein function can also result in cell–cell dis-adhesion and blistering in the superficial epidermis (Anhalt et al, 1982; Udey and Stanley, 1999; Amagai et al, 2000; Hanakawa et al, 2002). The Grhl1-deficient mice did not exhibit a blistering phenotype, although we observed desmosomal defects in the interfollicular epidermis, which had fewer desmosomes than controls. These desmosomes were shorter, less-organized, and weakly adhesive, as demonstrated by their sensitivity to EGTA treatment (Garrod et al, 2005). However, this desmosomal phenotype was not sufficiently severe to cause skin blistering. Consistent with this, the Nikolsky test for epidermal fragility was negative (data not shown) (Polifka and Krusinski, 1980). We conclude that there were sufficient desmosomes present in the Grhl1−/− epidermis to prevent skin blistering.

Our studies expand the functional comparisons between Drosophila grh and the mammalian Grhl genes that were initially examined in the context of formation and repair of the epidermal barrier (Ting et al, 2005). The consensus DNA-binding sequence identified for GRHL1 is identical to that defined for Drosophila GRH (Wilanowski et al, 2002). In addition, the identification of cadherin genes as direct transcriptional targets of GRHL1 extends the links between the grh and cadherin gene families from Drosophila to mammals. The cadherin family consists of more than 100 members (including the protocadherins) involved in many diverse cellular and developmental events (Nollet et al, 2000). Among the known targets of Drosophila, GRH are two cadherin superfamily members, stan/fmi and E-cadherin/shotgun (Lee and Adler, 2004; Almeida and Bray, 2005). The links between the two gene families have recently been extended in another model organism. A study of genome-wide predictions of genetic interactions in C. elegans has identified a number of cadherin family members as prime targets for significant functional interactions with Grh-1, the worm homologue of grh (Zhong and Sternberg, 2006). These included cdh-6, a homologue of Drosophila stan, and hmp-2, a β-catenin. We have recently identified the mammalian homologue of E-cadherin as a target of another member of the Grh-like family, Grhl2 (JMC and SMJ, unpublished). Our ongoing studies will examine their functional relevance.

The ablation of Grhl1 expression does not completely inactivate the Dsg1 genes, suggesting that there are other factors involved in their regulation. Also, the Grhl1 gene is expressed in many tissues that do not express Dsg1, and the apparent lack of phenotype associated with these tissues may be due to compensation from other Grhl family members. To this end, we have generated mice carrying mutations at both Grhl1 and Grhl3 loci. Mice null for both Grhl1 and Grhl3 died at birth, with neural tube defects that were identical to those observed in Grhl3-null embryos. Grhl1−/−Grhl3+/− animals did not show any differences in phenotype to that observed with loss of Grhl1 alone. It is unclear whether the residual Grhl3 allele is sufficient to partially compensate for the loss of Grhl1, or compensation is provided by another factor. Our transfection data suggest that GRHL3 is incapable of activating the Dsg1 promoter, suggesting that it may have no role in this context. The development of a conditional Grhl3 allele will allow this question to be addressed further.

Materials and methods

Microscopy, digital photography, and image processing

Gross images were obtained using a digital camera (Fuji FinePix S2 Pro with 60 mm Micro Nikon lens) or a dissection microscope (model SMZ-U, Nikon) equipped with a camera (AxioCam, Zeiss) driven by AxioVision (Zeiss). Microscopy images were obtained using a microscope (Optiphot-2, Nikon) and the same camera.


This assay was carried out according to the previously published protocol, using the same nuclear extract and antibodies (Wilanowski et al, 2002). Two oligonucleotide probes were used (sense strand only given): human Dsg1 promoter 5′-GGTGGGGATCCAGACTGGTTATACGTACCTTC-3′, mouse Dsg1β promoter 5′-GGGTGGAGATCCAACTGGTTATACGTACCTTC-3′.

Luciferase reporter gene activity assay

Luciferase activity assays employed the dual-luciferase reporter assay system (Promega). Promoter elements from the mouse and human Dsg1 genes were PCR amplified and cloned into KpnI-XhoI cloning sites of the pGL3-Basic vector. The PCR primers introduced the required KpnI and XhoI restriction sites at the ends of the amplification products. The primer sequences are included in the Supplementary data. For expression of Grhl1, the full-length mouse Grhl1 cDNA (Wilanowski et al, 2002) was cloned into the pCAGGS vector. 293T cells were transiently cotransfected with the approximately 1 kb fragments of proximal promoters of the mouse and human Dsg1 promoters linked to the firefly luciferase reporter gene, and either the empty pCAGGS vector or the pCAGGS vector containing the Grhl1 cDNA. Cotransfection with the pTK-RL plasmid expressing Renilla luciferase served as the control of transfection efficiency. The results were standardized using Renilla luciferase assay. Each experiment was undertaken in triplicate.

Other experimental procedures

RT–PCR, Q-RT–PCR, histology, in situ hybridization, β-galactosidase staining, and electron microscopy were carried out as described (Wilanowski et al, 2002; Ting et al, 2003a, 2003b; Wan et al, 2004; Majewski et al, 2006). Primer sequences are included in the Supplementary data. Desmosomal counts were obtained using two different methodologies. In the first, images were taken randomly, with the only criterion being that the cornified layer appeared in the image, and all images were captured at 25 K. In the second, images were taken side-by-side all at 25 K and were rotated so that the cornified layer lay parallel to one of the long axes. Twenty to 24 images were taken per section.

The CASTing assay was performed as described previously (Ting et al, 2005), except nuclear extract from JEG-3 cells, and anti-GRHL1 antibody no. 67 was used (Wilanowski et al, 2002). Northern blots were performed as described previously (Ting et al, 2005). The Dsg1-specific probe encompassed nt 1808–2405 (GenBank entry AK036986), a region common to all three Dsg1 isoforms. The Dsg3-specific probe encompassed nt 1162–1822 (GenBank entry NM_030596). Signal strengths were quantified using PhosphorImager, driven by ImageQuant 5.0 (Molecular Dynamics).

The ChIP experiment followed the published protocol (Forsberg et al, 2000).

Primers specific for the human DSG1 promoter were:


Primers specific for the MYOD promoter were used as negative control (Zhao et al, 2004).

Immunohistochemistry on single sections of wild-type mouse skin was performed using the ImmPRESS multiple antigen labelling system as per the manufacturer's instructions (Vector Laboratories, Burligame, CA). Polyclonal antisera to mouse DSG1 was added first and developed with NovaRed (red colour). After re-blocking with horse serum, polyclonal antisera to mouse GRHL1 was added and developed with DAB-Ni (gray/black colour).

Supplementary Material

Supplementary Figure 1

Supplementary Figure 2

Supplementary Figure 3

Supplementary Figure 4

Supplementary Figure 5

Supplementary Figure 6

Supplementary Figure 7

Supplementary Figure 8

Supplementary Information


We thank members of the Jane lab, Walt Holleran, and Peter Elias for helpful discussions; S Hunjan and Ozgene Inc. (Perth, Australia) for technical assistance; and John Stanley for the kind gift of DSG1 and DSG3 antisera. Animal support was provided by staff from the Walter & Eliza Hall Institute. SMJ is a Principal Research Fellow of the Australian National Health and Medical Research Council (NHMRC). This work was supported by Grants from the Australian NHMRC and The March of Dimes Foundation.


  • Adams MJ, Reichel MB, King IA, Marsden MD, Greenwood MD, Thirlwell H, Arnemann J, Buxton RS, Ali RR (1998) Characterization of the regulatory regions in the human desmoglein genes encoding the pemphigus foliaceous and pemphigus vulgaris antigens. Biochem J 329: 165–174 [PMC free article] [PubMed]
  • Almeida MS, Bray SJ (2005) Regulation of post-embryonic neuroblasts by Drosophila Grainyhead. Mech Dev 122: 1282–1293 [PubMed]
  • Amagai M, Matsuyoshi N, Wang ZH, Andl C, Stanley JR (2000) Toxin in bullous impetigo and staphylococcal scalded-skin syndrome targets desmoglein 1. Nat Med 6: 1275–1277 [PubMed]
  • Anhalt GJ, Labib RS, Voorhees JJ, Beals TF, Diaz LA (1982) Induction of pemphigus in neonatal mice by passive transfer of IgG from patients with the disease. N Engl J Med 306: 1189–1196 [PubMed]
  • Auden A, Caddy J, Wilanowski T, Ting SB, Cunningham JM, Jane SM (2006) Spatial and temporal expression of the Grainyhead-like transcription factor family during murine development. Gene Expr Patterns 6: 964–970 [PubMed]
  • Baden HP, Kvedar JC, Magro CM (1992) Loose anagen hair as a cause of hereditary hair loss in children. Arch Dermatol 128: 1349–1353 [PubMed]
  • Bray SJ, Kafatos FC (1991) Developmental function of Elf-1: an essential transcription factor during embryogenesis in Drosophila. Genes Dev 5: 1672–1683 [PubMed]
  • Brennan D, Hu Y, Kljuic A, Choi Y, Joubeh S, Bashkin M, Wahl J, Fertala A, Pulkkinen L, Uitto J, Christiano AM, Panteleyev A, Mahoney MG (2004) Differential structural properties and expression patterns suggest functional significance for multiple mouse desmoglein 1 isoforms. Differentiation 72: 434–449 [PubMed]
  • Chidgey M, Brakebusch C, Gustafsson E, Cruchley A, Hail C, Kirk S, Merritt A, North A, Tselepis C, Hewitt J, Byrne C, Fassler R, Garrod D (2001) Mice lacking desmocollin 1 show epidermal fragility accompanied by barrier defects and abnormal differentiation. J Cell Biol 155: 821–832 [PMC free article] [PubMed]
  • Cogram P, Hynes A, Dunlevy LP, Greene ND, Copp AJ (2004) Specific isoforms of protein kinase C are essential for prevention of folate-resistant neural tube defects by inositol. Hum Mol Genet 13: 7–14 [PubMed]
  • Denning MF, Guy SG, Ellerbroek SM, Norvell SM, Kowalczyk AP, Green KJ (1998) The expression of desmoglein isoforms in cultured human keratinocytes is regulated by calcium, serum, and protein kinase C. Exp Cell Res 239: 50–59 [PubMed]
  • Elias PM, Matsuyoshi N, Wu H, Lin C, Wang ZH, Brown BE, Stanley JR (2001) Desmoglein isoform distribution affects stratum corneum structure and function. J Cell Biol 153: 243–249 [PMC free article] [PubMed]
  • Forsberg EC, Downs KM, Bresnick EH (2000) Direct interaction of NF-E2 with hypersensitive site 2 of the beta-globin locus control region in living cells. Blood 96: 334–339 [PubMed]
  • Garrod DR, Berika MY, Bardsley WF, Holmes D, Tabernero L (2005) Hyper-adhesion in desmosomes: its regulation in wound healing and possible relationship to cadherin crystal structure. J Cell Sci 118: 5743–5754 [PubMed]
  • Garrod DR, Merritt AJ, Nie Z (2002) Desmosomal cadherins. Curr Opin Cell Biol 14: 537–545 [PubMed]
  • Green KJ, Gaudry CA (2000) Are desmosomes more than tethers for intermediate filaments? Nat Rev Mol Cell Biol 1: 208–216. [PubMed]
  • Hanakawa Y, Li H, Lin C, Stanley JR, Cotsarelis G (2004) Desmogleins 1 and 3 in the companion layer anchor mouse anagen hair to the follicle. J Invest Dermatol 123: 817–822 [PubMed]
  • Hanakawa Y, Matsuyoshi N, Stanley JR (2002) Expression of desmoglein 1 compensates for genetic loss of desmoglein 3 in keratinocyte adhesion. J Invest Dermatol 119: 27–31 [PubMed]
  • Hunt DM, Rickman L, Whittock NV, Eady RA, Simrak D, Dopping-Hepenstal PJ, Stevens HP, Armstrong DK, Hennies HC, Kuster W, Hughes AE, Arnemann J, Leigh IM, McGrath JA, Kelsell DP, Buxton RS (2001) Spectrum of dominant mutations in the desmosomal cadherin desmoglein 1, causing the skin disease striate palmoplantar keratoderma. Eur J Hum Genet 9: 197–203 [PubMed]
  • Jane SM, Ting SB, Cunningham JM (2005) Epidermal impermeable barriers in mouse and fly. Curr Opin Genet Dev 15: 447–453 [PubMed]
  • Kljuic A, Bazzi H, Sundberg JP, Martinez-Mir A, O'Shaughnessy R, Mahoney MG, Levy M, Montagutelli X, Ahmad W, Aita VM, Gordon D, Uitto J, Whiting D, Ott J, Fischer S, Gilliam TC, Jahoda CA, Morris RJ, Panteleyev AA, Nguyen VT et al. (2003) Desmoglein 4 in hair follicle differentiation and epidermal adhesion: evidence from inherited hypotrichosis and acquired pemphigus vulgaris. Cell 113: 249–260 [PubMed]
  • Kljuic A, Christiano AM (2003) A novel mouse desmosomal cadherin family member, desmoglein 1 gamma. Exp Dermatol 12: 20–29 [PubMed]
  • Koch PJ, Mahoney MG, Cotsarelis G, Rothenberger K, Lavker RM, Stanley JR (1998) Desmoglein 3 anchors telogen hair in the follicle. J Cell Sci 111: 2529–2537 [PubMed]
  • Koch PJ, Mahoney MG, Ishikawa H, Pulkkinen L, Uitto J, Shultz L, Murphy GF, Whitaker-Menezes D, Stanley JR (1997) Targeted disruption of the pemphigus vulgaris antigen (desmoglein 3) gene in mice causes loss of keratinocyte cell adhesion with a phenotype similar to pemphigus vulgaris. J Cell Biol 137: 1091–1102 [PMC free article] [PubMed]
  • Lee H, Adler PN (2004) The grainy head transcription factor is essential for the function of the frizzled pathway in the Drosophila wing. Mech Dev 121: 37–49 [PubMed]
  • Luo J, Zhang L, Stenn K, Prouty S, Parimoo S (2005) Desmoglein genes are up-regulated in the pk mutant mouse. Biochem Biophys Res Commun 327: 64–69 [PubMed]
  • Mace KA, Pearson JC, McGinnis W (2005) An epidermal barrier wound repair pathway in Drosophila is mediated by grainy head. Science 308: 381–385 [PubMed]
  • Mahoney MG, Hu Y, Brennan D, Bazzi H, Christiano AM, Wahl JK (2006) Delineation of diversified desmoglein distribution in stratified squamous epithelia: implications in diseases. Exp Dermatol 15: 101–109 [PubMed]
  • Majewski IJ, Metcalf D, Mielke LA, Krebs DL, Ellis S, Carpinelli MR, Mifsud S, Di Rago L, Corbin J, Nicola NA, Hilton DJ, Alexander WS (2006) A mutation in the translation initiation codon of Gata-1 disrupts megakaryocyte maturation and causes thrombocytopenia. Proc Natl Acad Sci USA 103: 14146–14151 [PMC free article] [PubMed]
  • McGrath JA (2005) Inherited disorders of desmosomes. Australas J Dermatol 46: 221–229 [PubMed]
  • Merritt AJ, Berika MY, Zhai W, Kirk SE, Ji B, Hardman MJ, Garrod DR (2002) Suprabasal desmoglein 3 expression in the epidermis of transgenic mice results in hyperproliferation and abnormal differentiation. Mol Cell Biol 22: 5846–5858 [PMC free article] [PubMed]
  • Nollet F, Kools P, van Roy F (2000) Phylogenetic analysis of the cadherin superfamily allows identification of six major subfamilies besides several solitary members. J Mol Biol 299: 551–572 [PubMed]
  • Polifka M, Krusinski PA (1980) The Nikolsky sign. Cutis 26: 521–525, 526 [PubMed]
  • Pulkkinen L, Choi YW, Kljuic A, Uitto J, Mahoney MG (2003) Novel member of the mouse desmoglein gene family: Dsg1-beta. Exp Dermatol 12: 11–19 [PubMed]
  • Rickman L, Simrak D, Stevens HP, Hunt DM, King IA, Bryant SP, Eady RA, Leigh IM, Arnemann J, Magee AI, Kelsell DP, Buxton RS (1999) N-terminal deletion in a desmosomal cadherin causes the autosomal dominant skin disease striate palmoplantar keratoderma. Hum Mol Genet 8: 971–976 [PubMed]
  • Sawado T, Igarashi K, Groudine M (2001) Activation of beta -major globin gene transcription is associated with recruitment of NF-E2 to the beta-globin LCR and gene promoter. Proc Natl Acad Sci USA 21: 21 [PMC free article] [PubMed]
  • Smith C, Zhu K, Merritt A, Picton R, Youngs D, Garrod D, Chidgey M (2004) Regulation of desmocollin gene expression in the epidermis: CCAAT/enhancer-binding proteins modulate early and late events in keratinocyte differentiation. Biochem J 380: 757–765 [PMC free article] [PubMed]
  • Ting SB, Caddy J, Hislop N, Wilanowski T, Auden A, Zhao LL, Ellis S, Kaur P, Uchida Y, Holleran WM, Elias PM, Cunningham JM, Jane SM (2005) A homolog of Drosophila grainy head is essential for epidermal integrity in mice. Science 308: 411–413 [PubMed]
  • Ting SB, Wilanowski T, Auden A, Hall M, Voss AK, Thomas T, Parekh V, Cunningham JM, Jane SM (2003a) Inositol- and folate-resistant neural tube defects in mice lacking the epithelial-specific factor Grhl-3. Nat Med 9: 1513–1519 [PubMed]
  • Ting SB, Wilanowski T, Cerruti L, Zhao LL, Cunningham JM, Jane SM (2003b) The identification and characterization of human Sister-of-Mammalian Grainyhead (SOM) expands the grainyhead-like family of developmental transcription factors. Biochem J 370: 953–962 [PMC free article] [PubMed]
  • Udey MC, Stanley JR (1999) Pemphigus—diseases of antidesmosomal autoimmunity. JAMA 282: 572–576 [PubMed]
  • Wan H, Dopping-Hepenstal PJ, Gratian MJ, Stone MG, Zhu G, Purkis PE, South AP, Keane F, Armstrong DK, Buxton RS, McGrath JA, Eady RA (2004) Striate palmoplantar keratoderma arising from desmoplakin and desmoglein 1 mutations is associated with contrasting perturbations of desmosomes and the keratin filament network. Br J Dermatol 150: 878–891 [PubMed]
  • Whittock NV (2003) Genomic sequence analysis of the mouse desmoglein cluster reveals evidence for six distinct genes: characterization of mouse DSG4, DSG5, and DSG6. J Invest Dermatol 120: 970–980 [PubMed]
  • Wilanowski T, Tuckfield A, Cerruti L, O'Connell S, Saint R, Parekh V, Tao J, Cunningham JM, Jane SM (2002) A highly conserved novel family of mammalian developmental transcription factors related to Drosophila grainyhead. Mech Dev 114: 37–50 [PubMed]
  • Wright WE, Binder M, Funk W (1991) Cyclic amplification and selection of targets (CASTing) for the myogenin consensus binding site. Mol Cell Biol 11: 4104–4110 [PMC free article] [PubMed]
  • Yin T, Green KJ (2004) Regulation of desmosome assembly and adhesion. Semin Cell Dev Biol 15: 665–677 [PubMed]
  • Zhao Q, Cumming H, Cerruti L, Cunningham JM, Jane SM (2004) Site-specific acetylation of the fetal globin activator NF-E4 prevents its ubiquitination and regulates its interaction with the histone deacetylase, HDAC1. J Biol Chem 279: 41477–41486 [PubMed]
  • Zhong W, Sternberg PW (2006) Genome-wide prediction of C. elegans genetic interactions. Science 311: 1481–1484 [PubMed]

Articles from The EMBO Journal are provided here courtesy of The European Molecular Biology Organization
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Gene
    Gene links
  • Gene (nucleotide)
    Gene (nucleotide)
    Records in Gene identified from shared sequence links
  • GEO Profiles
    GEO Profiles
    Related GEO records
  • HomoloGene
    HomoloGene links
  • MedGen
    Related information in MedGen
  • Nucleotide
    Published Nucleotide sequences
  • Pathways + GO
    Pathways + GO
    Pathways, annotations and biological systems (BioSystems) that cite the current article.
  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem Substance links
  • Taxonomy
    Related taxonomy entry
  • Taxonomy Tree
    Taxonomy Tree

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...