Lgr4 (Leucine-rich repeat containing G-protein coupled receptor 4) is one of the genes identified as novel GPCRs (Hsu et al.,1998; Loh et al.,2000), designated Lgr4 ∼ Lgr8, from an expressed sequence tag (EST) database with high homology to glycoprotein hormone receptors including FSHR (Baker et al.,2003), LH/CGR (Huhtaniemi et al.,2002; Zhang et al.,2001), and TSHR (Postiglione et al.,2002).
As Lgr4 showed high homology with FSHR, LHR, and THSR, this receptor was initially thought to be involved in the reproductive systems of animals. Mazerbourg et al. reported the generation of Lgr4 gene-interrupted mice using a gene-trapped ES cell line (Leighton et al.,2001), in which the expression of Lgr4 was severely attenuated by the insertion of the β-geo gene into the first intron (Mazerbourg et al.,2004). In that report, they only described the neonatal lethality of the null mice, but not the cause. On the other hand, our Lgr4−/− mice, completely lacking exon18, which encodes the whole transmembrane domain of Lgr4, also show embryonic/neonatal lethality with an additional phenotype of hypoplastic kidneys and an increased concentration of plasma creatinine (Kato et al.,2006). Mendive et al. (2006) as well as Hoshii et al. (2007) suggested that Lgr4 was essential for postnatal development of the male reproductive tract on a DBA or CD1 genetic background. Recently, we reported that Lgr4 was a critical regulator of keratinocyte motility in the epidermal tissue of eyelids using our Lgr4-deficient mice, which showed the eye-open at birth (EOB) phenotype, generally considered as a typical phenotype with reduced keratinocyte motility (Kato et al.,2007). In that report, we additionally generated keratinocyte-specific Lgr4-deficient mice (Lgr4K5 KO), which spared the embryonic/neonatal lethality, and almost all of them showed the EOB phenotype as well (Kato et al.,2007).
Lgr4K5 KO mice could survive to adulthood, and interestingly, all of them showed sparse head hair and focal alopecia behind their ears. As well as the phenotypes observed in the epidermis of Lgr4K5 KO mice, in which gene deletion was restricted to epithelial cells, Lgr4−/− mice also showed similar abnormalities in hair follicles, and we further analyzed this defect in Lgr4K5 KO and/or Lgr4−/− mice. These phenotypes are similar to those observed in tabby (Eda) and downless (Edar) mice, which are pathophysiological models of hypohidrotic (anhidrotic) ectodermal dysplasia (HED) in human. HED is a congenital disorder of ectodermal differentiation in which afflicted individuals have no sweat glands. They also have sparse head hair and abnormal teeth (Srivastava et al.,1997; Headon and Overbeek,1999; Monreal et al.,1999; Naito et al.,2002; Mikkola and Thesleff,2003; Hammerschmidt and Schlake,2007). In this report, we first describe the novel function of Lgr4 in the control of hair follicle development.
Abnormal Phenotypes in Hair of Postnatal Lgr4K5 KO Mice
Previously, we reported Lgr4K5 KO mice, which were spared the embryonic/neonatal lethality, and the body size and the weight of several organs including the kidney and liver of the neonatal mice were all normal (Kato et al.,2007). The number of Lgr4K5 KO mice that could survive until the weaning stage was the same as that of the control mice (Table 1). Lgr4K5 KO mice showed sparse head hair (Fig. 1A) and focal alopecia behind their ears (Fig. 1B). This phenotype was similar to those observed in tabby (Eda) and downless (Edar) mice (Srivastava et al.,1997,2001; Headon and Overbeek,1999; Naito et al.,2002; Mikkola and Thesleff,2003; Hammerschmidt and Schlake,2007), and by postnatal day (P) 7 (Fig. 1C), the histology clearly showed the abnormal hair follicle structure (Fig. 1D). These results strongly suggested that Lgr4 deletion caused abnormal hair follicle formation.
We confirmed the expression of Lgr4 in the epidermis of embryonic day (E) 16.5 embryos and P0 mice by reverse transcriptase-polymerase chain reaction (RT-PCR) analysis (Fig. 2A). Primer sets were described previously (Kato et al.,2007). To analyze the localization of Lgr4 expression in skin tissues, in situ hybridization was performed, and its expression was detected in the hair follicles and epidermis of P0 mice (Fig. 2B). It was also reported that, in adult mice, the expression of Lgr4 appeared in hairs and vibrissae at the beginning of the hair shaft region (Van Schoore et al.,2005).
Lgr4 Depletion Causes a Reduction in Hair Follicle Morphogenesis in Embryo
We next analyzed the hair follicle development in Lgr4−/− mice and found a clear difference between Lgr4−/− and Lgr4+/+ mice. Lgr4−/− mice had reduced numbers of hair follicle plugs at E16.5 and clearly showed a reduction in the number of hair follicles at P0 (Fig. 3A–D). Lgr4−/− mice showed a significant decrease (P < 0.05) in the number of hair follicles compared with age-matched Lgr4+/+ mice, and the number of hair follicles seen in the Lgr4−/− mice was only 25% of that seen in the Lgr4+/+ mice (Fig. 3G). A similar phenotype was also shown in Lgr4K5 KO mice (Fig. 3E,F). The number of hair follicles of Lgr4K5 KO mice was significantly reduced (Fig. 3H). We further analyzed to determine whether there was a defect in primary hair follicles in the Lgr4−/− mice. Because the identification of placodes was reported using the alkaline phosphatase activity (Paus et al.,1999; Ito et al.,2007), we applied this procedure to detect primary hair placodes in the embryos at E14.5, and we visualized them by whole-mount alkaline phosphatase staining (Fig. 3I,J). This result showed the defect in primary hair placode formation. These results strongly suggested that the deletion of Lgr4 in the epidermis caused abnormal hair follicle morphogenesis.
Effects of Lgr4 Deletion on the Expression of Several Genes Related to Hair Follicle Development
Many signaling molecules and their inhibitors including Wnts, sonic hedgehog (Shh), fibroblast growth factors (Fgfs), transforming growth factor-βs (Tgf-βs) superfamily, Ectodysplasin-A (Eda), and its receptor (Edar) are expressed in the placodes or by the underlying condensed mesenchyme, and they are reported to be necessary for hair follicle formation (Millar,2002; Pispa and Thesleff,2003; Schmidt-Ullrich and Paus,2005). We measured the expression levels of these genes in the mRNA prepared from Lgr4+/+ and Lgr4−/− mice epidermis (P0), and the transcripts of Edar, Lef1 and Shh were significantly reduced in the whole epidermis from Lgr4−/− mice (Fig. 4).
Lgr4 Deletion Affects the Induction of Hair Follicles
To further analyze the role of Lgr4 in hair follicle morphogenesis, we observed the distribution of shh expression and the level of Bmps signal. The downstream signal of Eda/Edar is transmitted through the Shh signaling pathway, which is essential for controlling the ingrowth and morphogenesis of the hair follicle (St-Jacques et al.,1998; Chiang et al.,1999; Millar,2002; Pummila et al.,2007). The immunohistochemical analysis showed higher expression of shh not only in hair follicles but also in the placodes (asterisks) of Lgr4+/+ mice, but these were not detected in the Lgr4−/− mice (Fig. 5A). The suppression of bone morphogenetic proteins (BMPs) signal is required for hair follicle induction (Botchkarev et al.,1999; Jamora et al.,2003). We next examined the level of the phosphorylated/active form of Smad1/5/8 using P-smad1/5/8 antibody, which are key molecules mediating the downstream effects of BMPs. In the resultant immunohistochemical analysis, we detected higher phosphorylation of Smad1/5/8 in Lgr4−/− mice epidermis (Fig. 5B).
Deletion of Lgr4 causes abnormal development of kidney, eyelid and epididymis (Kato et al.,2006,2007; Mendive et al.,2006; Hoshii et al.,2007). In this study, we suggest a novel function of Lgr4 in its important role in hair follicle development by using Lgr4 deficient mice. The number of hair follicles at the head and the expression of Edar, Lef1 and Shh, which were hair follicle morphogenesis-related genes, showed significant reductions in the epidermis in Lgr4−/− mice. Shh, reported to be required for mature hair follicle morphogenesis, acts downstream of Eda/Edar (St-Jacques et al.,1998; Chiang et al.,1999; Pummila et al.,2007). Although the intensity of immunostaining using Shh antibody for placodes prepared from P0 Lgr4−/− mice was not different from that when using P0 Lgr4+/+ mice, Lgr4−/− mice showed fewer placodes than Lgr4+/+ mice. In addition, the suppression of bone morphogenic proteins (BMPs) signal is required for hair follicle induction (Botchkarev et al.,1999; Jamora et al.,2003). The increased numbers of cells showing higher phosphorylation of smad1/5/8 suggested that placode induction was suppressed in the Lgr4−/− mice, and we suspected that there was the interaction between the Bmp signal and Lgr4 signal in the regulation of placode formation. Our results strongly suggested a relationship between Lgr4 and hair placode morphogenesis, and the reduced expression of shh, Lef1, and Edar in the Lgr4−/− mice might reflect the lack of hair placode induction. We then suspected that Lgr4 was required for keratinocytes to respond to inductive signals of hair follicle morphogenesis. Our observation in keratinocytes with reduced cell migration would help to elucidate the mechanisms of abnormal hair follicle morphogenesis in Lgr4 deficient mice. We are planning further investigation of the molecular function of Lgr4 in keratinocytes to elucidate the common mechanisms causing reduced keratinocyte migration and abnormal hair follicle morphogenesis.
Moreover, BMP, Shh, and Wnt signals were reported to be essential for kidney morphogenesis (Dressler,2002; Yu et al.,2004; Gill and Rosenblum,2006), and our observation of reduced expression and accumulated phosphorylation of these genes' products might be an important clue for explaining the mechanism causing the renal hypoplasia observed in Lgr4−/− mice.
The Lgr4K5 KO mice showed a similar phenotype as those observed in tabby (Eda) and downless (Edar) mice, which were models of HED. The study of Lgr4 along with its cognate ligands might reveal useful applications for the regeneration of hair follicles and for the treatment of human genetic diseases such as HED.
The generation of the Lgr4fx/fx, Lgr4−/− and Lgr4K5 KO mice has been described previously (Kato et al.,2006,2007). The care and use of mice in this study were approved by the Institution Animal Care and Use Committee of Tohoku University.
Histology, Immunostaining, and In Situ Hybridization
For the skin histology, the skin of the mice was fixed in 4% formaldehyde. After dehydration, the specimens were embedded in paraffin, and the paraffin blocks were sectioned at 5 μm thickness and stained with H&E (hematoxylin–eosin). In the immunohistochemical analysis, paraffin-embedded sections of skin were analyzed with the following antibodies: rabbit anti-P-smad1/5/8 (Cell Signaling Technology, Tokyo, Japan) and rabbit anti-shh (Cell Signaling Technology). These antibodies were diluted to 1/500 with 1% BSA (bovine serum albumin) in TBS (Tris-buffered saline) and applied for overnight at 4°C. Nuclei were counterstained with hematoxylin. Nonradioactive in situ hybridization was performed as described previously (Kawamata et al.,2007). The paraffin blocks were sectioned at 5 μm thickness, and the samples were prehybridized in hybridization buffer (50% formamide, 5× standard saline citrate (SSC), 1 mg/ml yeast tRNA, 2% Blocking Powder [Roche], 0.1% Triton X-100, 0.5% CHAPS, 5 mM EDTA (ethylenediaminetetraacetic acid), 50 μg/ml heparin) without RNA probe at 50°C for 1 hr, and then hybridized with a RNA probe of 1 μg/ml at 50°C for 16 hr. Hybridization was detected using an anti-digoxigenin Fab (Roche) coupled to alkaline phosphatase using NTB/BCIP solution (Roche).
Alkaline Phosphatase Staining
For whole-mount alkaline phosphatase staining, embryos were fixed with 4% paraformaldehyde, then stained at 37°C in alkaline phosphatase staining buffer (0.1 mg/ml Naphthol AS-MX phosphate, 0.6 mg/ml Fast Blue BB Salt, 5 μl/ml N,N-dimethlyformamide, 2 mM MgCl2, 0.1 mM Tris-HCl).
The number of hair follicles per millimeter of epidermal length was calculated using paraffin sections of Lgr4−/− or Lgr4K5 KO mice at E16.5 and P0, and was compared with that of age-matched Lgr4+/+ or Lgr4K5 ctrl mice (each group n = 4 or 5). Each sample was cut and every ninth point was analyzed and quantified by Student's t-test. A P value < 0.05 was considered significant.
The expression levels of Wnt5a, β-catenin, Lef1, Shh, Eda, and Edar were measured. The epidermis was separated from the dermis with 0.8 U/ml dispase (Roche), and total RNA was extracted. cDNA was then synthesized from 2 μg of the total RNA in 40 μl of the reaction mixture according to standard procedures. For quantitative PCR, 1 μl of the cDNA was added to 20 μl of the reaction mixture containing 10 μl DyNAmo SYBR Green qPCR kit (FINNZYMES) and 1 μl of 12.5 μM primers (forward and reverse). For each sample, a parallel reaction was set up with acidic ribosomal phosphoprotein PO (arbp) primers for the endogenous control. The primer sequences used were as follows:
Arbp S, 5′-ATAACCCTGAAGTGCTCGACAT-3′; Arbp AS, 5′-GGGAAGGTGTACTCAGTCTCCA-3′; Wnt5a S, 5′-AATCCACGCTAAGGGTTCCT-3′; Wnt5a AS, 5′-CCGCGCTATCATACTTCTCC-3′; β-catenin S, 5′-TGACCAGTTCCCTCTTCAGG-3′; β-catenin AS, 5′-ATGCTCCATCATAGGGTCCA-3′Lef1 S, 5′-TGAGTGCACGCTAAAGGAGA-3′; Lef1 AS, 5′-GCTGTCATTCTGGGACCTGT-3′; Shh S, 5′-GCAGGTTTCGACTGGGTCTA-3′; Shh AS, 5′-GAAGGTGAGGAAGT- CGCTGT-3′; Eda S, 5′-GTGGACGGCACCTACTTCAT-3′;Eda AS, 5′-CA- TCTTCACGGCGATTTTCT-3′Edar S, 5′-TGTGTATGCCAACGTGTGTG-3′; and Edar AS, 5′-CCCAATCTCATCCCTCTTCA-3′; The reactions were run in a DNA Engine Opticon System (MJ Research, Japan). Each reaction was performed in duplicate.
All experimental data are expressed as mean SEM. Statistical comparisons for all the physiological and laboratory data were made among the genotype groups using analysis of variance followed by Student's t-test for individual comparisons. P values of < 0.05 were considered significant.
We thank Dr. J Takeda for providing us the K5 Cre TG mice.