By continuing to browse this site you agree to us using cookies as described in About Cookies
Wiley Online Library is migrating to a new platform powered by Atypon, the leading provider of scholarly publishing platforms. The new Wiley Online Library will be migrated over the weekend of March 17 & 18. You should not experience any issues or loss of access during this time. For more information, please visit our migration page: http://www.wileyactual.com/WOLMigration/
Mammalian epidermis is maintained throughout adult life by stem cells, which self-renew and produce progeny that differentiate along the lineages of the interfollicular epidermis, sebaceous glands, and hair follicles (Niemann and Watt, 2002; Watt et al., 2006). The hair follicles undergo well-defined cycles of growth (anagen), regression (catagen), and rest (telogen; Alonso and Fuchs, 2006). During anagen, the base of the follicle, known as the bulb, is the major site of proliferation and differentiation. The widest part of the hair bulb is defined as the “critical line” or “line of Auber”: cells proximal to the line rapidly proliferate, while distal cells initiate differentiation along the lineages of the inner root sheath and hair shaft (Auber, 1952). In catagen, a subset of cells undergoes apoptosis, the hair follicle shrinks in length and retracts toward the skin surface. Thereafter, the follicle enters the resting, telogen, state (Muller-Rover et al., 2001; Alonso and Fuchs, 2006). Throughout the hair cycle, the bulb remains in contact with a specialized group of mesenchymal cells called the dermal papilla and reciprocal signalling between the dermal papilla and bulb is known to regulate the hair cycle (Jahoda and Reynolds, 1996; Oliver and Jahoda, 1988).
Two of the signalling pathways that control hair follicle growth and differentiation are the Notch and Wnt pathways. Inhibition of Wnt signalling prevents formation of hair follicles in the embryo, whereas in the adult Wnt inhibition results in conversion of follicles into cysts of interfollicular epidermis (Watt et al., 2006). Conversely, activation of the Wnt effector, β-catenin, is sufficient to induce ectopic hair follicles in adult epidermis (Watt et al., 2006). Notch activation is not required for hair follicle formation in the embryo, but blocking Notch signalling in postnatal epidermis results in conversion of hair follicles into interfollicular epidermal cysts (Yamamoto et al., 2003; Pan et al., 2004; Vauclair et al., 2005; Estrach et al., 2006). In contrast to β-catenin activation, activation of Notch is not sufficient to induce anagen or ectopic hair follicle formation. Nevertheless, Notch can promote differentiation of ectopic hair follicles induced by β-catenin (Estrach et al., 2006). We have recently shown that Notch signalling is activated by β-catenin and that the Notch ligand Jagged1 is a direct transcriptional target of β-catenin/Lef/Tcf complexes (Estrach et al., 2006).
Whereas the sites of expression of Notch receptors and their ligands have been documented in postnatal mammalian epidermis (Powell et al., 1998; Favier et al., 2000; Lowell et al., 2000; Nickoloff et al., 2002), less is known about Notch target genes (Lin et al., 2000). We found that several of these genes, including members of the Hes and Hey (HERP, Hesr) families (Iso et al., 2003), are up-regulated by β-catenin activation in the epidermis. The Hes family of transcriptional repressors comprises seven members; of these, Hes1, 5, and 7 are directly regulated by Notch (Bessho et al., 2001; Ohtsuka et al., 2001). The Hey family, also transcriptional repressors, consists of three members: Hey1, Hey2, and HeyL (Iso et al., 2003). To examine the association of the Notch and Wnt pathways during normal hair cycling and ectopic hair follicle formation, we have documented the expression of Notch target genes and Jagged1 in mouse skin and examined the effects of β-catenin activation.
RESULTS AND DISCUSSION
We first performed in situ hybridization of wild-type adult mouse back skin (Figs. 1, 2) using 35S-labeled RNA probes. We examined the mRNA expression patterns of the Notch ligand Jagged1 and of the Notch target genes Hes1, Hes5, Hey2, HeyL, and Hey1 (Iso et al., 2003). Skin sections were hybridized with a β-actin probe as a positive controls (Figs. 1B,2B). In telogen skin (Fig. 1A), only Jagged1 and Hes1 were detected (Fig. 1C–F, and data not shown). Jagged1 and Hes1 mRNAs were found at low levels in the hair follicle (Fig. 1C,D) and the interfollicular epidermis (Fig. 1E,F). We were able to confirm weak expression of Jagged1 and Hes1 by immunolabeling skin sections (data not shown). We could not reach any conclusions about sebaceous gland expression because of high levels of nonspecific signal (data not shown).
Next, we examined back skin of adult mice in which the hair follicles were in anagen (Fig. 2A,B). Jagged1 and Hes1 were expressed in the interfollicular epidermis, but no other Notch pathway genes were detected there (data not shown). In anagen hair follicles, Jagged1 and Hes1 mRNAs were expressed at high levels in cells in the bulb matrix (cells proximal to the line of Auber) and precortex (bulb cells distal to the line of Auber; Fig. 2C,D). Although Jagged1 mRNA was restricted to the bulb, Hes1 mRNA was detected more widely in the cuticle, cortex, and medulla cells of the hair shaft, and in a few outer root sheath cells in the distal portion of the follicle (Fig. 2C,D, and data not shown).
The other Notch pathway genes were primarily expressed in unique layers in the hair follicle or in the surrounding mesenchyme. Hey2 was expressed in dermal papilla cells (Fig. 2E), and HeyL transcripts were found in the inner root sheath and hair shaft layers, with some expression in the precortex (Fig. 2F; see also Lin et al., 2000). Hey1 transcripts were undetectable (data not shown). Cells expressing Hes5 transcripts were found in the differentiating hair shaft medulla and in small, asymmetrical patches in the bulb near the line of Auber (Fig. 2G); these latter cells may correspond to melanoblasts, which are known to express Hes5 (Moriyama et al., 2006). Hes5-positive cells at the line of Auber were present in fewer than 1% of anagen follicles, suggesting transient expression (Fig. 2G).
We next examined how the Notch pathway is affected by activation of β-catenin in adult mouse epidermis. We used K14ΔNβ-cateninER transgenic mice in which a stabilized, N-terminally truncated form of β-catenin was activated by topical application of the drug 4-hydroxy-tamoxifen (4OHT; Lo Celso et al., 2004). Microarray analysis has shown that Jagged1, Hes5, Hey2, and HeyL are all up-regulated at least 2-fold in transgenic skin treated with 4OHT for 7 days, whereas Hes1 is up-regulated 1.5-fold (Estrach et al., 2006). In contrast, the levels of Notch pathway genes are not affected by 4OHT treatment of wild-type skin (Estrach et al., 2006).
We began by analyzing expression of Notch signaling components in back skin of the high transgene copy number line D4, treated with 4OHT for 7 days (Fig. 3A,B). One of the earliest effects of β-catenin activation is to induce anagen of existing follicles (Lo Celso et al., 2004), and expression of Jagged1, Hes1, HeyL, and Hey2 was indistinguishable in transgenic and wild-type anagen follicles (Fig. 3, and data not shown). In contrast, Hes5 was not detected in β-catenin–induced anagen back skin follicles (Fig. 3E).
Activation of the β-catenin transgene causes formation of ectopic follicles, which appear as epithelial outgrowths from existing follicles, sebaceous glands, and interfollicular epidermis (Fig. 3A,A′ asterisks; Fig. 4A,A′). Although with time some of these outgrowths form a mature hair shaft, at early time points they can be distinguished from existing follicles both by their location and because they lack distinct inner root sheath and hair shaft layers (Silva-Vargas et al., 2005; Estrach et al., 2006). β-Catenin–induced follicles in both back (Fig. 3) and paw (Fig. 4) skin expressed high levels of Jagged1, Hes1, and HeyL (Figs. 3B–D,H, 4B–D,H). In paw, but not back skin, there was also patchy expression of Hes5 (Figs. 3E, 4E) and expression of Hey2 in the dermal papillae associated with new follicles (Figs. 3G, 4G, and data not shown).
Hey1 mRNA transcripts were detected in the dermal papillae of existing paw follicles and in mesenchymal cells surrounding new follicles, in a pattern that was similar to Hey2 (Fig. 4F). In back skin, Hey1 was additionally detected in the bulb of existing follicles (Fig. 3F). The expression pattern of Hey1 was not unexpected, because a previous microarray study has reported that Hey1 mRNA is expressed in dermal papilla cells (Rendl et al., 2005) and a second study has reported expression of Hey1 protein in anagen hair bulbs (Blanpain et al., 2006). Our failure to detect Hey1 mRNA in wild-type back skin suggests that expression is up-regulated by β-catenin.
We conclude that the patterns of expression of Notch target genes are similar in normal anagen hair follicles and β-catenin–induced ectopic follicles. The minor differences in gene expression in ectopic follicles in back and paw skin may reflect differences in follicle size, the ectopic follicles being larger in paw skin.
To examine the kinetics of β-catenin–induced expression of Notch pathway genes, we treated back skin from a low copy line (D2) of K14ΔNβcateninER transgenic mice with 4OHT for 7, 14, or 28 days. The β-catenin–induced phenotype develops more slowly in this line than in the D4 line of transgenic mice (Silva-Vargas et al., 2005). After 7 days, Jagged1 mRNA was found in all sites of ectopic β-catenin activity: cells of the outer root sheath, the periphery of the sebaceous gland, and the basal layer of the interfollicular epidermis (Fig. 5A, and data not shown). In contrast, expression of Hes1 (Fig. 5B) was similar to that observed in wild-type telogen skin (Fig. 1D,F), and Hes5, Hey2, and HeyL were undetectable (Fig. 5C, and data not shown).
By 14 days of 4OHT treatment, Jagged1 and Hes1 were strongly expressed, both in ectopic and existing follicles (Fig. 5D,E). Expression of Hes5 and Hey2 was similar to wild-type anagen follicles (Fig. 5F, and data not shown), while HeyL could also be detected in ectopic follicles (Fig. 5J). After 28 days of 4OHT treatment, all of the genes examined were detected in both existing and new ectopic hair follicles (Fig. 5G–I,K), in a pattern similar to 4OHT-treated D4 transgenic paw skin (Fig. 4).
To examine the effects of β-catenin activation on total protein levels of Notch pathway components, we performed Western blotting of protein extracts from total skin of 8-week-old wild-type and K14ΔNβ-cateninER mice treated with 4OHT for 15 days and from 34-day-old wild-type mice (Fig. 5L,M, and data not shown). The hair follicles in 8-week-old wild-type skin were in telogen at the time of extraction, whereas β-catenin had induced anagen in 8-week-old transgenic skin. Hair follicles in 34-day-old wild-type skin were in anagen.
Hes1, Hes5, Hey2, and Jagged1 protein levels were lower in wild-type telogen skin than in wild-type anagen skin and 4OHT-treated K14ΔNβ-cateninER skin. The levels of each protein were similar in transgenic and wild-type anagen skin. Western blotting with an antibody specific for the intracellular domain of Notch1 showed that there was a similar increase in cleaved, active Notch1 in both wild-type anagen and transgenic skin. Equal loading of protein lysates was confirmed by staining gels that had not been subjected to electroblotting (Fig. 5M).
In conclusion, our data show that expression of Jagged1 and Notch target genes of the Hes and Hey families is regulated during the normal hair cycle and on overexpression of activated β-catenin. The increase in gene transcripts during anagen is consistent with the observation that Notch activity is necessary for the growth phase of the hair cycle (Pan et al., 2004; Vauclair et al., 2005; Estrach et al., 2006). It is interesting that Hes1, Hes5, and HeyL are expressed in the precortex of the hair follicle bulb, because this is a site of nuclear β-catenin and Lef1 expression and is also where lineage commitment occurs (DasGupta and Fuchs, 1999; Niemann and Watt, 2002). Up-regulation of Jagged1 is consistent with its role in anagen entry and the finding that it is a direct β-catenin target gene (Estrach et al., 2006).
Although there was some overlap in expression of Hes1 and HeyL in the hair cuticle, and Hes1 and Hes5 in the hair shaft medulla, most Notch targets were expressed in unique locations. Embryonic deletion of individual Notch receptors or ligands produces tissue specific effects, with minimal compensation between family members (Kopan and Weintraub, 1993; Swiatek et al., 1994; Jiang et al., 1998; Xue et al., 1999). Multiple Notch receptors and ligands are expressed in the hair follicle, and it is possible that unique ligand and receptor combinations activate individual target genes. In support of this hypothesis, there is evidence that Hey2 transcription is dependent on Dll1 activation of the Notch pathway (Kokubo et al., 1999; Iso et al., 2001) and that Hey2 is found in the hair follicle dermal papilla, a region that expresses Dll1 (Powell et al., 1998). It will be of interest to discover what pathways, in addition to β-catenin, intersect with Notch signaling to regulate hair follicle induction and growth.
The K14ΔNβ-cateninER mice (D2 and D4 lines) have been described previously (Lo Celso et al., 2004; Silva-Vargas et al., 2005). K14ΔNβ-cateninER and wild-type mice were treated topically with 1 mg of 4OHT three times per week for up to 28 days.
In Situ Hybridization
In situ hybridization was performed as previously described (Poulsom et al., 1998). DNA plasmids used to make 35S-labeled antisense probes for Hey1, Hey2, and HeyL (gift of M. Gessler), Hes1 and Hes5 (gift of R. Kageyama), and Jagged1 (gift of J. Lewis), have been described previously (Leimeister et al., 1999; Ohtsuka et al., 2001). A β-actin antisense probe was used as a positive control (Lo Celso et al., 2004). Slides were photographed using bright- and darkfield illumination on a Nikon Ellipse Microscope.
Back skin was collected from 8-week-old K14ΔNβ-cateninER (D2 line) and wild-type mice treated with 4OHT for 15 days and from 34-day-old wild-type mice. Postcollection tissue was frozen immediately in liquid nitrogen. Samples were allowed to thaw on ice before addition of RIPA buffer (50 mM Tris pH 7.4, 1% NP-40, 0.25% sodium deoxycholate, 150 mM NaCl, and 1 mM ethylenediaminetetraacetic acid) with protease inhibitor cocktail (Roche, 11836170001). Lysates were homogenized, incubated on ice for 20 min, and then centrifuged to pellet insoluble material. Supernatants were collected and protein concentrations were determined using a BCA assay (Pierce). Lysates were subjected to polyacrylamide gel electrophoresis on 4–12% Novex Tris-Glycine acrylamide gels (Invitrogen). Gels were stained with Brilliant Blue G (Sigma) according to the manufacturer's instructions or transferred to polyvinylidene difluoride membrane (Amersham) and blocked with 10% milk powder in phosphate buffered saline plus 0.1% Tween-20 (PBT). Blots were incubated overnight with antibodies to the following proteins diluted in the blocking solution: Jagged1 1:100 (C-20, Santa Cruz Biotechnology), Hes1 1:100 (gift of T. Sudo, Toray Industries), Hes5 1:500 (Chemicon), Hey2 1:500 (Chemicon), and active Notch1 1:200 (Val1744, Cell Signaling Technologies). Blots were washed in PBT and incubated with anti-rabbit (1:5,000, Amersham) or anti-goat (1:5,000, Dako) secondary antibodies. Blots were visualized by using ECL Western Blotting Detection Reagents (Amersham).
We thank the CR-UK London Research Institute In Situ Hybridisation Laboratory for expert technical assistance. We also thank everyone who provided reagents and advice, in particular Angela Mowbray, Manfred Gessler, Julian Lewis, and Ryoichiro Kageyama. C.A.A. was supported by the National Institutes of Health under a Ruth L. Kirschstein National Research Service Award.