Advertisement

The Role of Smads in Skin Development

  • Philip Owens
    Affiliations
    Department of Otolaryngology, Oregon Health Sciences University, Portland Oregon, USA

    Department of Cell and Developmental Biology, Oregon Health Sciences University, Portland Oregon, USA

    Department of Dermatology, Oregon Health Sciences University, Portland Oregon, USA
    Search for articles by this author
  • Gangwen Han
    Affiliations
    Department of Otolaryngology, Oregon Health Sciences University, Portland Oregon, USA

    Department of Cell and Developmental Biology, Oregon Health Sciences University, Portland Oregon, USA

    Department of Dermatology, Oregon Health Sciences University, Portland Oregon, USA
    Search for articles by this author
  • Allen G. Li
    Affiliations
    Department of Otolaryngology, Oregon Health Sciences University, Portland Oregon, USA

    Department of Cell and Developmental Biology, Oregon Health Sciences University, Portland Oregon, USA

    Department of Dermatology, Oregon Health Sciences University, Portland Oregon, USA
    Search for articles by this author
  • Xiao-Jing Wang
    Correspondence
    VAMC Building 103, Room F-221, 3710 SW US Veterans Hospital Road, Mail code R&D46, Portland, Oregon 97239, USA
    Affiliations
    Department of Otolaryngology, Oregon Health Sciences University, Portland Oregon, USA

    Department of Cell and Developmental Biology, Oregon Health Sciences University, Portland Oregon, USA

    Department of Dermatology, Oregon Health Sciences University, Portland Oregon, USA
    Search for articles by this author
      Smads are a group of signaling mediators and antagonists of the transforming growth factor-β (TGF-β) superfamily, responding but not limited to signaling from TGF-β, Activin, and bone morphogenetic proteins (BMPs). As all of these three signaling pathways play important roles in skin development, we have been actively pursuing studies assessing the role of Smads in skin development. Our studies revealed that Smad-4 affects hair follicle differentiation primarily by mediating BMP signaling. Smad-7 significantly affects hair follicle development and differentiation by blocking the TGFβ/Activin/BMP pathway and by inhibiting WNT/β-catenin signaling via ubiquitin-mediated β-catenin degradation. In contrast, other Smads may have redundant or dispensable functions in skin development. Here, we review the work that shows the emergence of Smad functions in skin development via traditional and novel signaling pathways.

      Abbreviations:

      ALK
      activin-like kinase
      BMP
      bone morphogenetic protein
      FGF
      fibroblast growth factor
      IRS
      inner root sheath
      R-Smad
      receptor-specific Smad
      Co-Smad
      common partner Smad
      I-Smad
      inhibitory Smad
      SHH
      sonic Hedgehog
      TGF-β
      transforming growth factor-β

      Introduction

      Smads are intracellular proteins that shuttle from the cytoplasm to the nucleus. There are three classes of Smad proteins termed receptor-specific Smads (R-Smad – 1–3, 5, and 8), inhibitory Smads (I-Smad – 6 and 7), and the common partner or mediator Smad (Smad-4) (
      • ten Dijke P.
      • Hill C.S.
      New insights into TGF-beta-Smad signalling.
      ;
      • Massague J.
      • Seoane J.
      • Wotton D.
      Smad transcription factors.
      ). R-Smads are phosphorylated by the serine/threonine kinase domain of the transforming growth factor-β (TGF-β) family receptors, once these receptors are activated by their ligands (Figure 1). Phosphorylated R-Smads then bind with Smad-4 and enter the nucleus, where the Smad complex transcriptionally regulates gene expression by binding sequence-specific elements within the promoters of target genes termed “Smad binding elements.” The Smad complex at the Smad binding elements requires recruits co-factors to elicit a myriad of transcriptional responses including activation or repression. Among R-Smads, Smad-2 and -3 are regulated by TGF-β and Activin, whereas Smad-1, -5, and -8 are primarily activated by bone morphogenetic proteins (BMPs). The I-Smads block the ability of their respective receptors to phosphorylate R-Smads. Smad-6 preferentially antagonizes the signaling of Smad-1, -5, and -8, whereas Smad-7 preferentially antagonizes the signaling of Smad-2 and -3.
      Figure thumbnail gr1
      Figure 1Smad signaling. Smads are intracellular signaling molecules that are phosphorylated by the serine/threonine kinase receptors of transforming growth factor-β (TGF-β)/Activin or bone morphogenetic protein (BMP). These receptors are activated (phosphorylated) upon ligand binding to their extracellular domains. Once phosphorylated, receptor-specific Smads (R-Smads) then partner with Smad-4 where they form a hetero–trimeric complex and shuttle into the nucleus. Upon entry to the nucleus, Smads form diverse associations with transcription co-factors and bind site-specific Smad binding elements (SBE). Smad-2 and -3 (Smad-2/3) are signaled from TGF-β/Activin ligands, whereas Smad-1, -5, and -8 (Smad-1/5/8) are activated from BMP ligands. For each R-Smad group, there lies an inhibitory Smad (I-Smad): Smad-6 preferentially inhibits Smad-1/5/8 and Smad-7 preferentially inhibits Smad-2/3.
      The mammalian epidermis and its appendages develop from a simple ectoderm to a stratified epithelium during embryogenesis. In mice, epidermal development begins at embryonic day 9.5 (E9.5). Upon birth, the epidermis is fully differentiated, expressing differentiation markers for the terminally differentiated epidermis (
      • Fuchs E.
      • Raghavan S.
      Getting under the skin of epidermal morphogenesis.
      ). Epidermal appendages consist of hair follicles, sebaceous/sweat glands, mammary glands, teeth, nails/claws, as well as parts of the external genitalia structure. Of these appendages, the best studied and understood is the hair follicle. The primary murine hair follicle, which gives rise to the guard hair, begins to develop at E14.5. Secondary hair follicles (awl and zigzag) begin to develop approximately 2 days later and make up the majority of hair follicles in mouse skin. When hair follicles develop, epidermal keratinocytes receive instructive signals from the underlying mesenchyme (dermis), and then aggregate and form a condensation (placode). The placodes then become associated with underlying dermal condensates and give rise to the hair germ. As hair follicles move downwards, the dermal condensates are engulfed by the broadened bottom of downgrowing hair follicles (hair pegs), forming dermal papillae. The instructive signals for hair follicle morphogenesis are from but not limited to the pathways of WNT/β-catenin, Hedgehog, fibroblast growth factor (FGF) and BMP. Once the hair follicles have been specified, they grow until approximately two weeks postnatally and begin their cycling. Hair cycling begins with the catagen or regressing phase. Once they are physically separated from the dermal papillae, the hair follicles enter into the telogen or resting phase. The transition from telogen to the anagen or regrowth phase utilizes the molecular mechanisms involved in hair follicle induction during embryonic hair development. Sebaceous glands develop in the first week after birth and alter their size proportionally with the associated hair follicles during hair cycling. Underneath the sebaceous glands, the size of the ‘bulge’ region remains persistent throughout the hair cycle, which has been widely characterized as a region of multipotent stem cells
      • Blanpain C.
      • Fuchs E.
      Epidermal stem cells of the skin.
      ;
      • Moore K.A.
      • Lemischka I.R.
      Stem cells and their niches.
      ). Delicate machinery that regulates stem cell fate and progenitor differentiation is thus required for proper skin development during embryonic stages and for the maintenance of tissue homeostasis. Recent studies have shown that expression of Smads and their target genes are enriched in the epidermal stem cell population (
      • Morris R.J.
      A perspective on keratinocyte stem cells as targets for skin carcinogenesis.
      ;
      • Tumbar T.
      • Guasch G.
      • Greco V.
      • Blanpain C.
      • Lowry W.E.
      • Rendl M.
      • et al.
      Defining the epithelial stem cell niche in skin.
      ).

      The contribution of the TGF-β superfamily to epidermal development

      TGF-β and Activin ligands activate Smad-2 and -3 via the type I receptors activin-like kinase (Alk)-1, -4, -5, and -7. Among the three TGF-β isoforms, TGF-β2 is both required and sufficient to induce hair follicles in mice (
      • Foitzik K.
      • Paus R.
      • Doetschman T.
      • Dotto G.P.
      The TGF-beta2 isoform is both a required and sufficient inducer of murine hair follicle morphogenesis.
      ), via a mechanism of RAS/mitogen-activated protein kinase activation and subsequent transcription of Snail (
      • Jamora C.
      • Lee P.
      • Kocieniewski P.
      • Azhar M.
      • Hosokawa R.
      • Chai Y.
      • et al.
      A signaling pathway involving TGF-beta2 and snail in hair follicle morphogenesis.
      ). Although neither TGF-β1 nor TGF-β3 is required for hair follicle development, TGF-β1 is a potent inducer of catagen (
      • Foitzik K.
      • Lindner G.
      • Mueller-Roever S.
      • Maurer M.
      • Botchkareva N.
      • Botchkarev V.
      • et al.
      Control of murine hair follicle regression (catagen) by TGF-beta1 in vivo.
      ), and TGF-β3 has been shown to regulate keratinocyte migration during wound healing (
      • Bandyopadhyay B.
      • Fan J.
      • Guan S.
      • Li Y.
      • Chen M.
      • Woodley D.T.
      • et al.
      A “traffic control” role for TGFbeta3: orchestrating dermal and epidermal cell motility during wound healing.
      ). In addition to TGF-β ligands, Activin exerts an effect on skin development. Mice lacking activin-βA lack vibrissae and vibrissae follicles (
      • Matzuk M.M.
      • Kumar T.R.
      • Vassalli A.
      • Bickenbach J.R.
      • Roop D.R.
      • Jaenisch R.
      • et al.
      Functional analysis of activins during mammalian development.
      ). Germline deletion of follistatin, an Activin-binding protein and antagonist, results in a hyperkeratotic epidermis and abnormal vibrissae that appear thin and inappropriately orientated (
      • Matzuk M.M.
      • Lu N.
      • Vogel H.
      • Sellheyer K.
      • Roop D.R.
      • Bradley A.
      Multiple defects and perinatal death in mice deficient in follistatin.
      ). Correspondingly, follistatin knockout mice and Activin-βA transgenic mice show a delay of hair follicle morphogenesis. Treatment of wild-type embryonic skin explants with follistatin protein stimulated hair follicle development. This effect can be attenuated by Activin A. Activin-βA transgenic mice demonstrate a failure of catagen entry (
      • Nakamura M.
      • Matzuk M.M.
      • Gerstmayer B.
      • Bosio A.
      • Lauster R.
      • Miyachi Y.
      • et al.
      Control of pelage hair follicle development and cycling by complex interactions between follistatin and activin.
      ). These findings highlight the important signaling and crosstalk from mesenchymal cells of the dermal papillae to differentiated keratinocytes of the hair follicle. Presumably, these ligands and antagonists coordinate their function via Smad activation.
      BMPs comprise a large group of soluble proteins that activate Smad-1, -5, and -8 via the type I receptors ALK-2, -3, and -6. It has been difficult to study the requirement for BMP signaling in epidermal development because different BMP ligands and receptors are expressed in different epithelial and stromal compartments or in cells of different lineages within the same compartment. Advances in understanding BMP functions in the skin have mainly come from keratinocyte-specific BMP transgenic mouse models (
      • Botchkarev V.A.
      • Sharov A.A.
      BMP signaling in the control of skin development and hair follicle growth.
      ). A recent study shows that BMP-2 induces transcription of canonical WNT/β-catenin family ligands and receptors to regulate cell fate in human keratinocytes (
      • Yang L.
      • Yamasaki K.
      • Shirakata Y.
      • Dai X.
      • Tokumaru S.
      • Yahata Y.
      • et al.
      Bone morphogenetic protein-2 modulates Wnt and frizzled expression and enhances the canonical pathway of Wnt signaling in normal keratinocytes.
      ). Many BMP ligands and their antagonists do not determine specific cell fate but help to regulate the proper timing and behavior of a specific cell lineage. For instance, administration of Noggin protein into telogen mouse skin revealed that Noggin is required for postnatal hair follicle cycling (
      • Botchkarev V.A.
      • Botchkareva N.V.
      • Nakamura M.
      • Huber O.
      • Funa K.
      • Lauster R.
      • et al.
      Noggin is required for induction of the hair follicle growth phase in postnatal skin.
      ). The mechanism of Noggin-induced hair follicle re-growth is associated with attenuating the inhibitory effect of BMP-4 on Sonic Hedgehog (Shh) expression. A recent study demonstrates that mice overexpressing Noggin using a keratin 5 promoter increased cell proliferation via affecting genes controlling cell-cycle progression (
      • Sharov A.A.
      • Sharova T.Y.
      • Mardaryev A.N.
      • di Vignano A.T.
      • Atoyan R.
      • Weiner L.
      • et al.
      Bone morphogenetic protein signaling regulates the size of hair follicles and modulates the expression of cell cycle-associated genes.
      ). In contrast, when primary keratinocyte explants were treated with BMPs, they became quiescent and underwent growth arrest (
      • Sharov A.A.
      • Sharova T.Y.
      • Mardaryev A.N.
      • di Vignano A.T.
      • Atoyan R.
      • Weiner L.
      • et al.
      Bone morphogenetic protein signaling regulates the size of hair follicles and modulates the expression of cell cycle-associated genes.
      ). A study from Elaine Fuchs' laboratory provided further evidence for the involvement of BMP in skin development by showing that BMP-6 is expressed in the dermal papillae and the stem cell niche (
      • Fuchs E.
      • Tumbar T.
      • Guasch G.
      Socializing with the neighbors: stem cells and their niche.
      ). This finding suggests that BMP-6 is a potent inducer of matrix cells in the developing hair follicle. The BMP antagonist Cripto was also upregulated in isolated and enriched epidermal stem cells, the function of which remains to be determined (
      • Fuchs E.
      • Tumbar T.
      • Guasch G.
      Socializing with the neighbors: stem cells and their niche.
      ). Several studies have provided evidence for the requirement of ALK-3 (bone morphogenetic protein receptor 1A) in postnatal hair follicle maintenance. In these studies, ALK-3 was conditionally deleted in keratinocytes, which resulted in the collapse of the hair follicles postnatally (
      • Kobielak K.
      • Pasolli H.A.
      • Alonso L.
      • Polak L.
      • Fuchs E.
      Defining BMP functions in the hair follicle by conditional ablation of BMP receptor IA.
      ;
      • Andl T.
      • Ahn K.
      • Kairo A.
      • Chu E.Y.
      • Wine-Lee L.
      • Reddy S.T.
      • et al.
      Epithelial Bmpr1a regulates differentiation and proliferation in postnatal hair follicles and is essential for tooth development.
      ;
      • Ming Kwan K.
      • Li A.G.
      • Wang X.J.
      • Wurst W.
      • Behringer R.R.
      Essential roles of BMPR-IA signaling in differentiation and growth of hair follicles and in skin tumorigenesis.
      ;
      • Yuhki M.
      • Yamada M.
      • Kawano M.
      • Iwasato T.
      • Itohara S.
      • Yoshida H.
      • et al.
      BMPR1A signaling is necessary for hair follicle cycling and hair shaft differentiation in mice.
      ). These reports demonstrated that ALK-3 is required for the formation of the differentiated matrix cells of the hair follicle. More recently, a study has shown that deletion of ALK-3 in the hair follicle lead to an increased population of hair follicle stem cells, which fail to utilize β-catenin to specify hair follicle lineages. In contrast, activation of Akt and loss of PTEN (phosphatase and Tensin homolog) were found in these stem cells (
      • Zhang J.
      • He X.C.
      • Tong W.G.
      • Johnson T.
      • Wiedemann L.M.
      • Mishina Y.
      • et al.
      Bone morphogenetic protein signaling inhibits hair follicle anagen induction by restricting epithelial stem/progenitor cell activation and expansion.
      ). This report suggested that ALK-3 functions to define the progeny of the hair follicle stem cell niche and regulate a discrete balance of committed and undifferentiated cells (
      • Zhang J.
      • He X.C.
      • Tong W.G.
      • Johnson T.
      • Wiedemann L.M.
      • Mishina Y.
      • et al.
      Bone morphogenetic protein signaling inhibits hair follicle anagen induction by restricting epithelial stem/progenitor cell activation and expansion.
      ). It remains to be determined if the effects of these BMP ligands and receptors require spatio-temporal Smad activation.

      R-Smads: little effect on skin development

      Even though TGF-β superfamily members and receptors exert profound effects on skin development and there is abundant expression of Smads 1–5 in the epidermis and hair follicles (
      • He W.
      • Cao T.
      • Smith D.A.
      • Myers T.E.
      • Wang X.J.
      Smads mediate signaling of the TGFbeta superfamily in normal keratinocytes but are lost during skin chemical carcinogenesis.
      ), studies have revealed little effect of individual R-Smads on skin development and differentiation. Germline Smad-2 knockout mice die at the embryonic stage before hair follicle development due to failure of germ layer specification and primitive streak formation (
      • Waldrip W.R.
      • Bikoff E.K.
      • Hoodless P.A.
      • Wrana J.L.
      • Robertson E.J.
      Smad2 signaling in extraembryonic tissues determines anterior–posterior polarity of the early mouse embryo.
      ). Transgenic mice overexpressing Smad-2 using the keratin 14 promoter exhibit delayed hair growth, underdeveloped ears, and shortened tails. Further analysis suggests that these phenotypes correlated with enhanced signaling from TGF-β and Activin (
      • Ito Y.
      • Sarkar P.
      • Mi Q.
      • Wu N.
      • Bringas Jr, P.
      • Liu Y.
      • et al.
      Overexpression of Smad2 reveals its concerted action with Smad4 in regulating TGF-beta-mediated epidermal homeostasis.
      ). It remains to be determined if there are any pathological conditions in which Smad-2 is overexpressed in the skin. At the physiological level, Smad-2 alone appears to have little effect on skin development, as our unpublished data reveal that keratinocyte-specific Smad-2 deletion did not result in obvious abnormalities in skin development. Similarly, Smad-3 knockout mice have no apparent phenotype related to aberrant epidermal development. With respect to Smad-1, -5, and -8, very little is known about the individual roles of these Smads in the epidermis. Antibody staining, which recognizes all three phosphorylated form, thereby serving as a readout of BMP activity, revealed abundant phospho-Smad-1/5/8 staining in both interfollicular epidermis and hair follicles (
      • Han G.
      • Li A.G.
      • Liang Y.Y.
      • Owens P.
      • He W.
      • Lu S.
      • et al.
      Smad7-induced beta-catenin degradation alters epidermal appendage development.
      ). Some links for the involvement of these Smads in skin development and differentiation has been suggested by studies of Smad co-factors and target genes. For instance, Smad-1 and -4 form a transcriptional complex on the promoter of the Dlx3 transcription factor (
      • Park G.T.
      • Morasso M.I.
      Bone morphogenetic protein-2 (BMP-2) transactivates Dlx3 through Smad1 and Smad4: alternative mode for Dlx3 induction in mouse keratinocytes.
      ). Dlx3 is a homeodomain transcription factor that helps pattern and specify cell fate in keratinocytes and its expression is induced by BMP-2 (
      • Park G.T.
      • Morasso M.I.
      Bone morphogenetic protein-2 (BMP-2) transactivates Dlx3 through Smad1 and Smad4: alternative mode for Dlx3 induction in mouse keratinocytes.
      ). Another potential link to Smad-1/5/8 may be RUNX3, a transcription factor often demonstrated to partner with Smad-1/5/8 during osteoblast formation. Runx3 was found to be a critical determinant of hair shape as well as nail and gland formation (
      • Raveh E.
      • Cohen S.
      • Levanon D.
      • Groner Y.
      • Gat U.
      Runx3 is involved in hair shape determination.
      ). However, direct evidence for BMP-specific Smads in skin development is lacking. Our unpublished data revealed that keratinocyte-specific deletion of either Smad-1 or -5 did not affect skin development and differentiation. To date, there is no report for Smad-8 function in the epidermis. Because of the structural similarity and the lack of antibodies specific for individual BMP-specific Smads, it is difficult to determine if they have redundant functions in the skin.

      Smad-4: the bridge between R-Smads

      The lack of skin phenotypes in R-Smad knockout keratinocytes suggests that individual Smads may compensate for each other's loss during skin development and differentiation. Indeed, when Smad-4 is deleted in keratinocytes, which results in abrogation of most Smad signaling, hair follicles collapse (
      • Yang L.
      • Mao C.
      • Teng Y.
      • Li W.
      • Zhang J.
      • Cheng X.
      • et al.
      Targeted disruption of Smad4 in mouse epidermis results in failure of hair follicle cycling and formation of skin tumors.
      ;
      • Qiao W.
      • Li A.G.
      • Owens P.
      • Xu X.
      • Wang X.J.
      • Deng C.X.
      Hair follicle defects and squamous cell carcinoma formation in Smad4 conditional knockout mouse skin.
      ). These mice lost postnatal hair follicles before the first catagen entry. Spontaneous squamous cell carcinomas were also identified. Interestingly, the phenotypes of Smad4/ hair follicles were very similar to those in epidermal-specific ALK-3 knockout mice (
      • Kobielak K.
      • Pasolli H.A.
      • Alonso L.
      • Polak L.
      • Fuchs E.
      Defining BMP functions in the hair follicle by conditional ablation of BMP receptor IA.
      ;
      • Andl T.
      • Ahn K.
      • Kairo A.
      • Chu E.Y.
      • Wine-Lee L.
      • Reddy S.T.
      • et al.
      Epithelial Bmpr1a regulates differentiation and proliferation in postnatal hair follicles and is essential for tooth development.
      ;
      • Ming Kwan K.
      • Li A.G.
      • Wang X.J.
      • Wurst W.
      • Behringer R.R.
      Essential roles of BMPR-IA signaling in differentiation and growth of hair follicles and in skin tumorigenesis.
      ;
      • Yuhki M.
      • Yamada M.
      • Kawano M.
      • Iwasato T.
      • Itohara S.
      • Yoshida H.
      • et al.
      BMPR1A signaling is necessary for hair follicle cycling and hair shaft differentiation in mice.
      ), which illustrates a possible dependence of Smad-4 on BMP signaling but not on Activin or TGF-β signaling (Figure 1). Smad-4 has been identified to partner with Lef1 at the Msx2 promoter (
      • Hussein S.M.
      • Duff E.K.
      • Sirard C.
      Smad4 and beta-catenin co-activators functionally interact with lymphoid-enhancing factor to regulate graded expression of Msx2.
      ), representing a significant link to the WNT/β-catenin signaling cascade, which is critical for hair development and differentiation. Therefore, Smad-4 may cooperate with WNT/β-catenin signaling at the transcriptional level. Another transcriptional partner of Smads shown in T cells is GATA-3 (
      • Blokzijl A.
      • Ten Dijke P.
      • Ibanez C.F.
      Physical and functional interaction between GATA-3 and Smad3 allows TGF-beta regulation of GATA target genes.
      ). In the skin, GATA-3 is expressed specifically in the inner root sheath (IRS) (
      • Kaufman C.K.
      • Zhou P.
      • Pasolli H.A.
      • Rendl M.
      • Bolotin D.
      • Lim K.C.
      • et al.
      GATA-3: an unexpected regulator of cell lineage determination in skin.
      ). Mice with a GATA-3 gene deletion exhibited a collapsed hair follicle phenotype (
      • Kaufman C.K.
      • Zhou P.
      • Pasolli H.A.
      • Rendl M.
      • Bolotin D.
      • Lim K.C.
      • et al.
      GATA-3: an unexpected regulator of cell lineage determination in skin.
      ) similar to the keratinocyte-specific Smad-4 knockout hair follicles. It remains to be determined if GATA-3 and Smads form a functional transcription complex that specifies the differentiated IRS of the hair follicle. Another interesting Smad partner may be FOXO factors, which, like Smads, are proteins that shuttle between the nucleus and cytoplasm and are found to be expressed in human keratinocytes (
      • Gomis R.R.
      • Alarcon C.
      • He W.
      • Wang Q.
      • Seoane J.
      • Lash A.
      • et al.
      A FoxO–Smad synexpression group in human keratinocytes.
      ). FOXO transcription factors belong to the winged helix domain of transcription factors, which is a large family that also includes FoxN1, a transcription factor responsible for the nude mouse phenotype (
      • Mecklenburg L.
      • Nakamura M.
      • Sundberg J.P.
      • Paus R.
      The nude mouse skin phenotype: the role of Foxn1 in hair follicle development and cycling.
      ). It remains to be determined if FoxN1 has a direct interaction with Smads, similar to what has been demonstrated with FOXO family members.
      Smad-4 has been considered necessary for Smad signaling because it was widely required for many Smad-associated processes. This notion has been challenged by the observation that when Smad-4 is inhibited in HaCaT keratinocytes, R-Smad transcriptional activity still occurred (
      • Levy L.
      • Hill C.S.
      Smad4 dependency defines two classes of transforming growth factor {beta} (TGF-{beta}) target genes and distinguishes TGF-{beta}-induced epithelial–mesenchymal transition from its antiproliferative and migratory responses.
      ). Further evidence from the hematopoietic system has shown that Smad-4 loss results in new stoichiometric accumulations of R-Smads with the TIF1-γ transcription factor (
      • He W.
      • Dorn D.C.
      • Erdjument-Bromage H.
      • Tempst P.
      • Moore M.A.
      • Massague J.
      Hematopoiesis controlled by distinct TIF1gamma and Smad4 branches of the TGFbeta pathway.
      ). TIF1-γ is exclusively nuclear and does not aid in the shuttling of Smad-2/3 into the nucleus. This study shows that even when Smad-4 is removed, phosphorylated R-Smad complexes may still bind their target genes. For the above reasons, it is not surprising that epidermal and hair follicle development is largely intact when Smad-4 is deleted in keratinocytes. However, regardless of the initial normal embryonic skin development, homeostasis in Smad4-deleted keratinocytes apparently is lost, which eventually results in skin tumor formation.

      I-Smads: restricting and refining R-Smads

      Within the two I-Smads, abundant Smad-6 mRNA was detected in the developing mouse skin at E15 (
      • Flanders K.C.
      • Kim E.S.
      • Roberts A.B.
      Immunohistochemical expression of Smads 1–6 in the 15-day gestation mouse embryo: signaling by BMPs and TGF-betas.
      ). This report also identified that the vibrissae follicles lack Smad-6 expression. There have been reports of transient expression or loss of expression of Smad-6 in early mouse development and in pathological conditions. For instance, Smad-6 and -7 are lost in keloids, which are benign tumors in the skin that result from overaccumulation of extracellular matrix proteins as a result of un-inhibited TGF-β signaling (
      • Yu H.
      • Bock O.
      • Bayat A.
      • Ferguson M.W.
      • Mrowietz U.
      Decreased expression of inhibitory SMAD6 and SMAD7 in keloid scarring.
      ). Further studies on Smad-6 functioning in skin pathology remain to be done. Almost little or no expression of Smad-6 is detected in the adult skin (
      • He W.
      • Li A.G.
      • Wang D.
      • Han S.
      • Zheng B.
      • Goumans M.J.
      • et al.
      Overexpression of Smad7 results in severe pathological alterations in multiple epithelial tissues.
      ). However, the role of Smad-6 in the skin is basically unknown. A thorough examination of Smad-6 expression patterns during different developmental stages and physiological/pathological states of the skin will serve as an initial step toward identifying which, if any, role of Smad-6 in the skin.
      To date, Smad-7 has shown the most effects among all Smads on skin development. Smad-7 is expressed at a very low level in normal keratinocytes (
      • He W.
      • Cao T.
      • Smith D.A.
      • Myers T.E.
      • Wang X.J.
      Smads mediate signaling of the TGFbeta superfamily in normal keratinocytes but are lost during skin chemical carcinogenesis.
      ), but is often overexpressed under pathological conditions, for example, in intrinsically aged and photoaged human skin (
      • He W.
      • Li A.G.
      • Wang D.
      • Han S.
      • Zheng B.
      • Goumans M.J.
      • et al.
      Overexpression of Smad7 results in severe pathological alterations in multiple epithelial tissues.
      ;
      • Quan T.
      • He T.
      • Kang S.
      • Voorhees J.J.
      • Fisher G.J.
      Ultraviolet irradiation alters transforming growth factor beta/smad pathway in human skin in vivo.
      ) and during skin carcinogenesis (
      • He W.
      • Cao T.
      • Smith D.A.
      • Myers T.E.
      • Wang X.J.
      Smads mediate signaling of the TGFbeta superfamily in normal keratinocytes but are lost during skin chemical carcinogenesis.
      ). To access the overall role of Smad signaling in skin development, we generated Smad-7 transgenic mice, which overexpress Smad-7 under the control of a keratin 5 promoter (K5.Smad-7) to levels that are sufficient to block Smad signaling from TGF-β/Activin and BMP (
      • He W.
      • Li A.G.
      • Wang D.
      • Han S.
      • Zheng B.
      • Goumans M.J.
      • et al.
      Overexpression of Smad7 results in severe pathological alterations in multiple epithelial tissues.
      ). These transgenic mice exhibit multiple developmental defects in the stratified epithelia, including decreased hair follicle size. They die perinatally due to epithelial hyperkeratosis in the upper digestive tract and severe thymic atrophy (
      • He W.
      • Li A.G.
      • Wang D.
      • Han S.
      • Zheng B.
      • Goumans M.J.
      • et al.
      Overexpression of Smad7 results in severe pathological alterations in multiple epithelial tissues.
      ). Since the phenotype severity correlates with the degree of Smad signaling inhibition, these data suggest the importance of Smad signaling in skin development, which would not necessarily be revealed in individual Smad knockout skin if the Smads have functional redundancy. However, one K5.Smad-7 transgenic line, which expressed a low level of Smad-7 transgene, and therefore did not block Smad signaling, still exhibited a hair loss phenotype (
      • Han G.
      • Li A.G.
      • Liang Y.Y.
      • Owens P.
      • He W.
      • Lu S.
      • et al.
      Smad7-induced beta-catenin degradation alters epidermal appendage development.
      ).
      To understand further the mechanism, we generated Smad-7 transgenic mice, in which the Smad-7 transgene can be induced in keratinocytes at different developmental stages and at different levels in keratinocytes including epidermal stem cells. This inducible transgenic system consists of a transactivator line (GLp65) (
      • Cao T.
      • He W.
      • Roop D.R.
      • Wang X.J.
      K14-GLp65 transactivator induces transgene expression in embryonic epidermis.
      ;
      • Lu S.L.
      • Reh D.
      • Li A.G.
      • Woods J.
      • Corless C.L.
      • Kulesz-Martin M.
      • et al.
      Overexpression of transforming growth factor beta1 in head and neck epithelia results in inflammation, angiogenesis, and epithelial hyperproliferation.
      ) and a target line (tata.Smad-7). The keratin 5 vector was used to target the GLp65 transactivator to the basal layer of the epidermis and the outer root sheath of the hair follicle, including epidermal stem cells (
      • Arin M.J.
      • Longley M.A.
      • Wang X.J.
      • Roop D.R.
      Focal activation of a mutant allele defines the role of stem cells in mosaic skin disorders.
      ). The target transgene (tata.Smad-7) consists of a GAL4 UAS enhancer upstream of a tata minimal promoter (
      • Cao T.
      • He W.
      • Roop D.R.
      • Wang X.J.
      K14-GLp65 transactivator induces transgene expression in embryonic epidermis.
      ;
      • Lu S.L.
      • Reh D.
      • Li A.G.
      • Woods J.
      • Corless C.L.
      • Kulesz-Martin M.
      • et al.
      Overexpression of transforming growth factor beta1 in head and neck epithelia results in inflammation, angiogenesis, and epithelial hyperproliferation.
      ) and cDNA encoding mouse Smad-7 with a Flag tag (
      • He W.
      • Li A.G.
      • Wang D.
      • Han S.
      • Zheng B.
      • Goumans M.J.
      • et al.
      Overexpression of Smad7 results in severe pathological alterations in multiple epithelial tissues.
      ). RU486 is used to regulate transgene expression in bigenic mice (GLp65/tata.Smad-7). The ability to control Smad-7 expression at a pathologically relevant level and in an acute or a sustained manner furthered our analysis of the underlying molecular mechanisms. This model has also helped us to determine the role of Smad-7 overexpression at stages critical for stem cell fate decision and differentiation as well as in postnatal hair cycling. We found that Smad-7 transgene induction resulted in a significant delay in embryonic hair follicle development and a complete blockade of hair follicle differentiation (Figure 2). When transgenic Smad-7 expression was induced in a sustained manner beginning either on E10.5 or on E14.5, the effect on bigenic mouse skin from E16.5 throughout P1 was essentially identical. In both cases, histological examination showed that hair follicle morphogenesis was delayed in Smad-7 transgenic skin in comparison with normal skin. On E16.5, hair follicles in normal embryos are in developmental stage 2 or 3, in which follicles appear to be in the form of a hair germ or peg, respectively. In contrast, only scattered hair follicle placodes were observed in Smad-7 transgenic embryos. On E18.5, certain follicles in control skin have entered stage 4, characterized by the beginning of IRS formation and engulfment of dermal papilla within the hair bulb. In contrast, Smad-7 E18.5 hair follicles were in stages 1 and 2. In P1 normal skins, more than 50% of the hair follicles have passed stage 4, and some have entered stage 5, evidenced by the presence of melanin in the precortex and an elongated IRS. However, the majority of the hair follicles in P1 Smad-7 transgenic skin were in stages 2 and 3, equivalent in appearance to the E16.5 normal embryonic hair follicles. After birth, if Smad-7 transgene induction was continuously maintained, hair follicle differentiation was abrogated. At P10, normal skin hair follicles have each formed a well-differentiated hair shaft. However, Smad-7 transgenic follicles were much shorter and disoriented with prominent sebaceous glands and clustered melanin. Hair follicle differentiation markers, the AE15 antibody stained IRS cells and medulla cells of the hair shaft and the AE13 antibody stained upper cortical and cuticle cells, are absence in Smad-7 transgenic hair follicles that appeared disorganized and lacked any obvious IRS or hair shaft. In contrast, sebaceous gland development was significantly accelerated in Smad-7 transgenic skin (Figure 3). Sebocytes are visible on P1 Smad-7 transgenic skin, but not on control skin. Again, when the Smad-7 transgene is expressed at levels high enough to inhibit Smad signaling, these skin phenotypes are exacerbated and epidermal differentiation is also perturbed. However, even when Smad-7 is expressed at a lower level that does not affect Smad signaling, these skin abnormalities still occur.
      Figure thumbnail gr2
      Figure 2Smad-7 induction perturbed hair follicle development and differentiation. (a) Hematoxylin and eosin staining of embryonic E16.5, E18.5, and P1 skin that was exposed to daily RU486 treatment in utero beginning on E14.5. Arrows point to representative hair follicles in different developmental stages (S). Note that the epidermis of E18.5 and P1 transgenic skin also exhibited mild hyperplasia. The bar in the first panel represents 50 μm for all sections in (a). The AE15 antibody stained the inner root sheath (IRS) cells and medulla cells of the hair shaft and the AE13 antibody stained upper cortical and cuticle cells. (b) Histology and immunostaining using AE13 and AE15 antibodies on P10 control and Smad-7 transgenic skins. In hematoxylin and eosin sections (left), arrows in Smad-7 transgenic skin point to the sebaceous glands. In immunostaining panels (middle and right), arrows in control skin point out examples of normal staining patterns. Note that arrows in Smad-7 transgenic skin point out aberrant staining patterns. Dark staining in Smad-7 transgenic follicles represents irregular melanin deposits. The bar in the first panel represents 50 μm for all sections in (b). This is supplemental in Han et al., Dev Cell, 11:301–312, 2006.
      Figure thumbnail gr3
      Figure 3Smad-7 induction accelerates sebaceous gland morphogenesis. Oil-red-O staining was used to detect sebocytes. Hematoxylin was used as a counter stain. Note that sebaceous glands, hair canal, and subcutaneous fat tissue are positively stained by Oil-red-O. Arrows denote representative sebocytes or sebaceous glands. The bar in the first panel represents 100 μm for all sections.
      Further analysis revealed that independent of its role in anti-Smad signaling, Smad7-bound β-catenin and induced β-catenin degradation by recruiting an E3 ligase, Smurf2 to the Smad-7/β-catenin complex. Consequently, WNT/β-catenin signaling was suppressed in Smad-7 transgenic hair follicles. Co-expression of Smurf2 and Smad-7 transgenes exacerbated Smad7-induced abnormalities in hair follicles and sebaceous glands. Conversely, when endogenous Smad-7 was knocked down, keratinocytes exhibited increased β-catenin protein and enhanced WNT signaling. These data suggest that even at a low, physiological level, endogenous Smad-7 participates in β-catenin turnover (Figure 4). This mechanism also explains the phenotype difference between Smad4/ skin and Smad-7 transgenic skin, which cannot be explained by the difference in the degree of abrogating Smad signaling. Although Smad-4 also potentially affects WNT/β-catenin signaling at the transcriptional level, the direct effect of Smad-7 on β-catenin degradation appears to be more potent in affecting WNT/β-catenin signaling. This is probably why Smad-7 transgenic, but not Smad4 null skin, exhibits delayed hair follicle development.
      Figure thumbnail gr4
      Figure 4Antagonizing Smad signaling and Wnt signaling by Smad-7. (a) Smad-7 inhibits phosphorylation and receptor Smad complex assembly to restrict and terminate signaling from transforming growth factor β (TGFβ)/activin or bone morphogenetic protein (BMP). If WNT (wingless related) signaling is unaffected by Smad-7, WNT ligands bind cognate frizzled receptors, resulting in releasing of β-catenin (β-cat) from the membrane and sent to the nucleus to induce transcription of targets with LEF/TCF (lymphoid enhancement factor/T-cell factor) transcriptional partners. (b) Smad-7 recruits the ubiquitin E3 ligase Smurf2 to β-catenin in the cytoplasm to induce β-catenin degradation. β-Catenin is degraded in the 26S proteasome. This activity of Smad-7 inhibits WNT signaling by preventing β-catenin entering into the nucleus.

      Conclusions and Future Perspectives

      To date, the most significant advances in our understanding of Smad functions in the skin have come from mouse models with spatio-temporal expression/ablation of individual Smad genes. Among them, only Smad-4 loss or Smad-7 overexpression in keratinocytes results in abnormalities of epidermal and hair follicle development and/or differentiation (Figure 5). In the future, thorough examination of Smad expression patterns at specific developmental stages and pathological conditions of the skin (e.g., wound healing, cancer, and various skin diseases) will provide valuable lessons for Smad function in the epidermis. More complex genetic approaches such as knocking out more than one Smad gene at a time in keratinocytes and combination of these knockouts with transgenics overexpressing TGF-β family ligands will further elucidate the overlapping and compensatory functions of individual Smads in the skin. The recent finding that Smad-7 functions in the WNT/β-catenin pathway demonstrates that Smad proteins are not simply mediators/antagonists of TGF-β signaling. It remains to be determined how other pathways are integrated by Smads and under what physiological/pathological conditions these interactions occur.
      Figure thumbnail gr5
      Figure 5Smad signaling in skin development. Smads are highly enriched in the hair follicle bulge stem cell niche. Sebocytes hyperproliferation occur when Smad-4 is absent or Smad-7 is overexpressed. Smad-4 loss or Smad-7 increase also results in epidermal hyperproliferation. Terminal differentiation of the epidermis is blocked by Smad-7 overexpression.

      Conflict of Interest

      The authors state no conflict of interest.

      ACKNOWLEDGMENTS

      P.O. is the recipient of Tartar Trust and supported by an NIH training grant. The original work in our laboratory was supported by NIH Grant nos. CA87849, AR47898, and GM70966 to X.J.W.

      REFERENCES

        • Andl T.
        • Ahn K.
        • Kairo A.
        • Chu E.Y.
        • Wine-Lee L.
        • Reddy S.T.
        • et al.
        Epithelial Bmpr1a regulates differentiation and proliferation in postnatal hair follicles and is essential for tooth development.
        Development (Cambridge, England). 2004; 131: 2257-2268
        • Arin M.J.
        • Longley M.A.
        • Wang X.J.
        • Roop D.R.
        Focal activation of a mutant allele defines the role of stem cells in mosaic skin disorders.
        J Cell Biol. 2001; 152: 645-649
        • Bandyopadhyay B.
        • Fan J.
        • Guan S.
        • Li Y.
        • Chen M.
        • Woodley D.T.
        • et al.
        A “traffic control” role for TGFbeta3: orchestrating dermal and epidermal cell motility during wound healing.
        J Cell Biol. 2006; 172: 1093-1105
        • Blanpain C.
        • Fuchs E.
        Epidermal stem cells of the skin.
        Annu Rev Cell Dev Biol. 2006; 22: 339-373
        • Blokzijl A.
        • Ten Dijke P.
        • Ibanez C.F.
        Physical and functional interaction between GATA-3 and Smad3 allows TGF-beta regulation of GATA target genes.
        Curr Biol. 2002; 12: 35-45
        • Botchkarev V.A.
        • Botchkareva N.V.
        • Nakamura M.
        • Huber O.
        • Funa K.
        • Lauster R.
        • et al.
        Noggin is required for induction of the hair follicle growth phase in postnatal skin.
        FASEB J. 2001; 15: 2205-2214
        • Botchkarev V.A.
        • Sharov A.A.
        BMP signaling in the control of skin development and hair follicle growth.
        Differentiation. 2004; 72: 512-526
        • Cao T.
        • He W.
        • Roop D.R.
        • Wang X.J.
        K14-GLp65 transactivator induces transgene expression in embryonic epidermis.
        Genesis. 2002; 32: 189-190
        • Flanders K.C.
        • Kim E.S.
        • Roberts A.B.
        Immunohistochemical expression of Smads 1–6 in the 15-day gestation mouse embryo: signaling by BMPs and TGF-betas.
        Dev Dyn. 2001; 220: 141-154
        • Foitzik K.
        • Lindner G.
        • Mueller-Roever S.
        • Maurer M.
        • Botchkareva N.
        • Botchkarev V.
        • et al.
        Control of murine hair follicle regression (catagen) by TGF-beta1 in vivo.
        FASEB J. 2000; 14: 752-760
        • Foitzik K.
        • Paus R.
        • Doetschman T.
        • Dotto G.P.
        The TGF-beta2 isoform is both a required and sufficient inducer of murine hair follicle morphogenesis.
        Dev Biol. 1999; 212: 278-289
        • Fuchs E.
        • Raghavan S.
        Getting under the skin of epidermal morphogenesis.
        Nat Rev. 2002; 3: 199-209
        • Fuchs E.
        • Tumbar T.
        • Guasch G.
        Socializing with the neighbors: stem cells and their niche.
        Cell. 2004; 116: 769-778
        • Gomis R.R.
        • Alarcon C.
        • He W.
        • Wang Q.
        • Seoane J.
        • Lash A.
        • et al.
        A FoxO–Smad synexpression group in human keratinocytes.
        Proc Natl Acad Sci USA. 2006; 103: 12747-12752
        • Han G.
        • Li A.G.
        • Liang Y.Y.
        • Owens P.
        • He W.
        • Lu S.
        • et al.
        Smad7-induced beta-catenin degradation alters epidermal appendage development.
        Dev Cell. 2006; 11: 301-312
        • He W.
        • Cao T.
        • Smith D.A.
        • Myers T.E.
        • Wang X.J.
        Smads mediate signaling of the TGFbeta superfamily in normal keratinocytes but are lost during skin chemical carcinogenesis.
        Oncogene. 2001; 20: 471-483
        • He W.
        • Dorn D.C.
        • Erdjument-Bromage H.
        • Tempst P.
        • Moore M.A.
        • Massague J.
        Hematopoiesis controlled by distinct TIF1gamma and Smad4 branches of the TGFbeta pathway.
        Cell. 2006; 125: 929-941
        • He W.
        • Li A.G.
        • Wang D.
        • Han S.
        • Zheng B.
        • Goumans M.J.
        • et al.
        Overexpression of Smad7 results in severe pathological alterations in multiple epithelial tissues.
        EMBO J. 2002; 21: 2580-2590
        • Hussein S.M.
        • Duff E.K.
        • Sirard C.
        Smad4 and beta-catenin co-activators functionally interact with lymphoid-enhancing factor to regulate graded expression of Msx2.
        J Biol Chem. 2003; 278: 48805-48814
        • Ito Y.
        • Sarkar P.
        • Mi Q.
        • Wu N.
        • Bringas Jr, P.
        • Liu Y.
        • et al.
        Overexpression of Smad2 reveals its concerted action with Smad4 in regulating TGF-beta-mediated epidermal homeostasis.
        Dev Biol. 2001; 236: 181-194
        • Jamora C.
        • Lee P.
        • Kocieniewski P.
        • Azhar M.
        • Hosokawa R.
        • Chai Y.
        • et al.
        A signaling pathway involving TGF-beta2 and snail in hair follicle morphogenesis.
        PLoS Biol. 2005; 3: e11
        • Kaufman C.K.
        • Zhou P.
        • Pasolli H.A.
        • Rendl M.
        • Bolotin D.
        • Lim K.C.
        • et al.
        GATA-3: an unexpected regulator of cell lineage determination in skin.
        Genes Dev. 2003; 17: 2108-2122
        • Kobielak K.
        • Pasolli H.A.
        • Alonso L.
        • Polak L.
        • Fuchs E.
        Defining BMP functions in the hair follicle by conditional ablation of BMP receptor IA.
        J Cell Biol. 2003; 163: 609-623
        • Levy L.
        • Hill C.S.
        Smad4 dependency defines two classes of transforming growth factor {beta} (TGF-{beta}) target genes and distinguishes TGF-{beta}-induced epithelial–mesenchymal transition from its antiproliferative and migratory responses.
        Mol Cell Biol. 2005; 25: 8108-8125
        • Lu S.L.
        • Reh D.
        • Li A.G.
        • Woods J.
        • Corless C.L.
        • Kulesz-Martin M.
        • et al.
        Overexpression of transforming growth factor beta1 in head and neck epithelia results in inflammation, angiogenesis, and epithelial hyperproliferation.
        Cancer Res. 2004; 64: 4405-4410
        • Massague J.
        • Seoane J.
        • Wotton D.
        Smad transcription factors.
        Genes Dev. 2005; 19: 2783-2810
        • Matzuk M.M.
        • Kumar T.R.
        • Vassalli A.
        • Bickenbach J.R.
        • Roop D.R.
        • Jaenisch R.
        • et al.
        Functional analysis of activins during mammalian development.
        Nature. 1995; 374: 354-356
        • Matzuk M.M.
        • Lu N.
        • Vogel H.
        • Sellheyer K.
        • Roop D.R.
        • Bradley A.
        Multiple defects and perinatal death in mice deficient in follistatin.
        Nature. 1995; 374: 360-363
        • Mecklenburg L.
        • Nakamura M.
        • Sundberg J.P.
        • Paus R.
        The nude mouse skin phenotype: the role of Foxn1 in hair follicle development and cycling.
        Exp Mol Pathol. 2001; 71: 171-178
        • Ming Kwan K.
        • Li A.G.
        • Wang X.J.
        • Wurst W.
        • Behringer R.R.
        Essential roles of BMPR-IA signaling in differentiation and growth of hair follicles and in skin tumorigenesis.
        Genesis. 2004; 39: 10-25
        • Moore K.A.
        • Lemischka I.R.
        Stem cells and their niches.
        Science. 2006; 311: 1880-1885
        • Morris R.J.
        A perspective on keratinocyte stem cells as targets for skin carcinogenesis.
        Differentiation. 2004; 72: 381-386
        • Nakamura M.
        • Matzuk M.M.
        • Gerstmayer B.
        • Bosio A.
        • Lauster R.
        • Miyachi Y.
        • et al.
        Control of pelage hair follicle development and cycling by complex interactions between follistatin and activin.
        FASEB J. 2003; 17: 497-499
        • Park G.T.
        • Morasso M.I.
        Bone morphogenetic protein-2 (BMP-2) transactivates Dlx3 through Smad1 and Smad4: alternative mode for Dlx3 induction in mouse keratinocytes.
        Nucleic Acids Res. 2002; 30: 515-522
        • Qiao W.
        • Li A.G.
        • Owens P.
        • Xu X.
        • Wang X.J.
        • Deng C.X.
        Hair follicle defects and squamous cell carcinoma formation in Smad4 conditional knockout mouse skin.
        Oncogene. 2006; 25: 207-217
        • Quan T.
        • He T.
        • Kang S.
        • Voorhees J.J.
        • Fisher G.J.
        Ultraviolet irradiation alters transforming growth factor beta/smad pathway in human skin in vivo.
        J Invest Dermatol. 2002; 119: 499-506
        • Raveh E.
        • Cohen S.
        • Levanon D.
        • Groner Y.
        • Gat U.
        Runx3 is involved in hair shape determination.
        Dev Dyn. 2005; 233: 1478-1487
        • Sharov A.A.
        • Sharova T.Y.
        • Mardaryev A.N.
        • di Vignano A.T.
        • Atoyan R.
        • Weiner L.
        • et al.
        Bone morphogenetic protein signaling regulates the size of hair follicles and modulates the expression of cell cycle-associated genes.
        Proc Natl Acad Sci USA. 2006; 103: 18166-18171
        • ten Dijke P.
        • Hill C.S.
        New insights into TGF-beta-Smad signalling.
        Trends Biochem Sci. 2004; 29: 265-273
        • Tumbar T.
        • Guasch G.
        • Greco V.
        • Blanpain C.
        • Lowry W.E.
        • Rendl M.
        • et al.
        Defining the epithelial stem cell niche in skin.
        Science. 2004; 303: 359-363
        • Waldrip W.R.
        • Bikoff E.K.
        • Hoodless P.A.
        • Wrana J.L.
        • Robertson E.J.
        Smad2 signaling in extraembryonic tissues determines anterior–posterior polarity of the early mouse embryo.
        Cell. 1998; 92: 797-808
        • Yang L.
        • Mao C.
        • Teng Y.
        • Li W.
        • Zhang J.
        • Cheng X.
        • et al.
        Targeted disruption of Smad4 in mouse epidermis results in failure of hair follicle cycling and formation of skin tumors.
        Cancer Res. 2005; 65: 8671-8678
        • Yang L.
        • Yamasaki K.
        • Shirakata Y.
        • Dai X.
        • Tokumaru S.
        • Yahata Y.
        • et al.
        Bone morphogenetic protein-2 modulates Wnt and frizzled expression and enhances the canonical pathway of Wnt signaling in normal keratinocytes.
        J Dermatol Sci. 2006; 42: 111-119
        • Yu H.
        • Bock O.
        • Bayat A.
        • Ferguson M.W.
        • Mrowietz U.
        Decreased expression of inhibitory SMAD6 and SMAD7 in keloid scarring.
        J Plast Reconstr Aesthet Surg. 2006; 593: 221-229
        • Yuhki M.
        • Yamada M.
        • Kawano M.
        • Iwasato T.
        • Itohara S.
        • Yoshida H.
        • et al.
        BMPR1A signaling is necessary for hair follicle cycling and hair shaft differentiation in mice.
        Development. 2004; 131: 1825-1833
        • Zhang J.
        • He X.C.
        • Tong W.G.
        • Johnson T.
        • Wiedemann L.M.
        • Mishina Y.
        • et al.
        Bone morphogenetic protein signaling inhibits hair follicle anagen induction by restricting epithelial stem/progenitor cell activation and expansion.
        Stem Cells. 2006; 24: 2826-2839