Kobielak et al. 10.1073/pnas.0703004104. |
Fig. 5. The hair follicle stem cells (SCs) and their lineage. Schematics depict the mature hair follicle (Left) and the lineages that arise upon transition from the resting (telogen) to growing (anagen) phase of the hair cycle (Right). Growth is accompanied by SC activation, self-renewal, and conversion to proliferating progeny that subsequently progress to differentiate along distinct lineages to produce the hair shaft, and the ORS and IRS.
Fig. 6. Postnatal ablation of Bmpr1a affects quiescence within the SC niche in adult hair follicles. (A) Status of active canonical BMP signaling in WT follicles, as depicted by anti-P-Smad1/5/8 immunofluorescence (Upper), anti-BMPR1A immunofluorescence (Lower Insets), and Bmpr1a in situ hybridization (ISH) (Lower) with digoxygenin cRNA probes specific for exon 2 of BMPR1A. Analyses were performed over the period encompassing the first telogen (P19) to anagen (P20 and P24) transition. Note lack of labeling in anti-P-Smad1/5/8 immunofluorescence in new germ (hg). Note that Bmpr1a ISH correlates well with anti-P-Smad1 staining at anagen P24, suggesting that its pattern of expression may be more faithful than immunofluorescence. (B) H&E-stained sections of skins from Bmpr1a(fl/fl)/K14-CreTM and Bmpr1a(fl/fl) mice before TM treatment began at P44. Initiation of TM treatment coincided with the onset of the second telogen phase of the hair cycle. (C) Genotyping of DNAs isolated from Bmpr1a CONTM and cKOTM skins. PCR analysis confirmed deletion of exon2 of Bmpr1a (Bottom) only after TM treatment and only in mice carrying the double-floxed allele (Top) and K14-CreTM (Middle). (D) Lack of anti-P-Smad-1/5/8 immunofluorescence shows that the Bmpr1a gene was efficiently targeted within 16d of topical TM treatment. (E) ISH with a cRNA probe specific for the deleted exon 2. Note the absence of hybridization in Bmpr1a cKOTM skin at P59. (F) Prior to TM treatment of mice, the CD34 monoclonal antibody marks the bulge SCs of the P44 resting telogen-phase follicles in both Bmpr1a(fl/fl)/K14-CreTM and Bmpr1a(fl/fl) mice. Note that there is autofluorescence of the hair shaft and background fluorescence within the dermis. The anti-CD34 staining of the epithelium, however, faithfully recapitulates the natural expression and specificity of CD34. (G and H) Increased BrdU incorporation to bulge and to the cells migrating out of the niche in Bmpr1a cKOTM mice after receiving a 4-h pulse at P59 and P77, respectively. In contrast, no anti-BrdU antibody labeling was detected in the bulge of CONTM animals at P59, and only a few cells were marked in the hair germ at the onset of anagen at P77. Broken lines denote dermal-epidermal boundaries, and arrows denote labeling. Bu, bulge; hg, hair germ; Mx, matrix.
Fig. 7. Expansion of the proliferating populations of hair follicles after inhibition of BMP signaling postnatally. cKOTM and CONTM mice were targeted for Bmpr1a ablation at P44 and analyzed just after the first morphological changes were observed at P59. At this stage, the CONTM follicles were still in telogen, but the cKOTM follicles had all precociously entered anagen and had formed an early hair germ. Frozen skin sections (10 mm) were subjected to indirect immunofluorescence and ISH. (A) Anti-Sox9 staining was detected in the cKOTM and CONTM bulge, and anti-Sox9 immunofluorescence was very strong in the early hair germ of the cKOTM follicles at P59. (B) Anit-Lhx2 staining was detected in the cKOTM and CONTM bulge but was consistently slightly stronger in the CONTM bulge. Anti-Lhx2 immunofluorescence was very strong in the early hair germ of the cKOTM follicles at P59. Note that intense Lhx2 staining overlaps with up-regulated P-cad staining in the cKOTM germ. (C) ISH of Sox4 shows expression in the hair germ and the bulge of the cKOTM, but not CONTM, P59 follicle. (D) Shh ISH shows expression in the distal portion of the hair germ close to the DP of P59 cKOTM follicles. No expression in the niche was detected. Broken lines mark dermo-epithelial borders, and brackets denote bulge (Bu). hg, hair germ.
Fig. 8. Regeneration of epidermal wounds by hair follicle cells following targeted removal of BMP receptor 1A expression. Two different experiments were conducted to examine this issue. (A) Bmpr1a (fl/fl)/K14-Cre and control mice were bred on the background of K14-H2BGFP. At birth, epidermis was removed enzymatically by dispase treatment (Upper), discarded, and then the dermis harboring BMPR1A-deficient or WT hair follicles was grafted to the backs of nude mice. Reepithelialization was judged by the number of GFP-positive cells (arrowheads) in the BMPR1A-deficient and WT hair follicle regenerated epidermis. (B) Bmpr1a cKOTM mice were mated to Rosa26-fl-stop-fl-lacZ mice; Bmpr1a ablation was induced during the extended second telogen phase of adult mice, and marked by permanent β-galactosidase expression. Following Bmpr1a ablation, skins were scraped to remove the epidermis, and wound repair was monitored for the periods indicated by X-gal staining to detect β-galactosidase activity and counterstain with eosin. Signs of reepithelialization were judged by the presence of β-gal-expressing cells within the regenerated epidermis (arrows). Epi, epidermis; Der, Dermis; HF, hair follicle.
Fig. 9. Precocious and sustained activation of Lef1/β-catenin and P-PTEN/PI3K/AKT signaling in hair follicles lacking BMPR1A. cKOTM and CONTM mice were analyzed at P77 after TM treatment. Frozen skin sections (10 mm) were subjected to indirect immunofluorescence or immunohistochemistry by using the antibodies indicated in the Lower Right of each frame. Color coding is according to the fluorescent tags of the secondary antibodies. (A) Immunofluorescence shows up-regulated nuclear Lef1 staining in the SC niche and cells migrating out of the niche (arrowheads) in cKOTM P77 follicles. In CONTM follicles, only a few cells in the new hair germ show nuclear Lef1 labeling. (B) Nuclear b-catenin staining is markedly increased in P77 cKOTM follicles in both the SC niche and in the cell masses migrating down out of the niche. In the control CONTM anagen follicle, nuclear b-catenin staining is only seen in the hair germ at P77. (C) Immunofluorescence reveals that anti-P-GSK3b staining is enhanced in the P77 cKOTM niche and in the proliferating progeny emanating from the bulge. In control CONTM follicles, anti-P-GSK3b staining is seen within the P77 bulge and is increased at the telogen-to-anagen transition (see Fig. 3C for comparison with telogen bulge). (D) Anti-P-PTEN immunohistochemistry shows that P-PTEN is normally only seen in the emerging hair germs (hg) of WT follicles at the telogen-to-anagen transition (P21 and P23, first transition; and P77, second transition). In the P77 cKOTM bulge and emerging cell mass, P-PTEN is up-regulated. (E) As judged by immunofluorescence, 14-3-3z was detected in the P77 cKOTM SC niche but was up-regulated in the proliferating progeny migrating out of the niche. In the control CONTM skin, 14-3-3z was also detected in the bulge and elevated in the hair germ at the start of the new anagen. Lam5, laminin5; β-cat, β-catenin; β4, β4-integrin; Bu, bulge; hg, hair germs. Arrowheads denote aberrant expression.
Fig. 10. Mice expressing a dominant active BMPR1A transgene display precocious hair follicle differentiation. dTg mice for the K14-Teton doxycycline-regulatable transcription factor (rtTA), and for a doxycycline-regulatable enhancer driving the expression of BMPR1A-CA, an HA-epitope tagged, constitutively active form of the BMPR1A, were engineered as described in the main text. (A) PCR analysis of DNAs isolated from dTgDox and CONDox mice confirmed the presence of transgenes. (B) Elevated anti-P-Smad-1/5/8 immunofluorescence in dTgDox skin after doxycycline induction. (C) Absence of anti-CD34 staining in the SCs of the dTgDox niche at P47. Note that a cyst emerges from the dTgDox hair follicle and that the hair shaft displays autofluorescence, as do hair-shaft cells, present within the P47 follicle. (D) Monoclonal antibody AE13 staining marks the hair-specific keratins. In CONDox follicles, it is present within the cortical cells that form the hair shaft; in dTgDox skin, AE13 marks some cells within the cyst. (E) Filaggrin staining (Fil) marks the granular layer of epidermis. This marker is excluded from the cysts of dTgDox. (F-H) Immunofluorescence and ISH show highly reduced levels of Sox9, Sox4 and Shh, respectively, in dTgDox ,whereas in CONDox, expression was observed in the expected pattern. Mx, matrix; Bu, bulge.
SI Materials and Methods
Histochemistry, Immunofluorescence, and in Situ Hybridization (ISH).
For all analyses, skins were taken from the middorsal side of the animals. Histology, immunofluorescence, and ISH were performed as described [Kobielak K, Pasolli HA, Alonso L, Polak L, Fuchs E (2003) J Cell Biol 163:609-623]. The following primary antibodies were used at the indicated concentrations: AE13 (mouse, 1:50; H. Sun, New York University School of Medicine, New York, NY), AE15 (mouse, 1:10; H. Sun), K5 (rabbit, 1:600; Fuchs laboratory, The Rockefeller University), GATA-3 (mouse, 1:100; Santa Cruz Biotechnology), BMPRIA (rabbit, 1:100; P. ten Dijke, Ludwig Institute for Cancer Research, Uppsala, Sweden), Lef-1 (rabbit, 1:100; Fuchs laboratory), a6-integrin (rat, 1:100; BD Pharmingen), CD34 (rat, 1:50; BD Pharmingen), BrdU (rat, 1:100; Abcam), b4-integrin/CD104 (rat, 1:100; BD Pharmingen), P-Smad-1/5/8 (rabbit, 1:100; Cell Signaling Technology), Sox9 (rabbit, 1:100; Santa Cruz Biotechnology), Lhx2 (rabbit, 1:2,500; T. Jessell, Columbia University, New York, NY), P-cadherin (rat, 1:100; M. Takeichi, RIKEN, Kobe, Japan), K10 (rabbit, 1:500; Covance), Filaggrin (rabbit, 1:500; Covance), HA (rat, 1:100; Roche Applied Science), and 14-3-3z (rabbit, 1:100; Cell Signaling Technology). Relevant FITC- or Texas Red-conjugated donkey or goat secondary antibodies (1:300; Jackson ImmunoResearch Laboratories) were used for detection of primary antibodies.
For immunohistochemistry, back skins were fixed in 4% paraformaldehyde, and tissues were dehydrated and embedded in paraffin. Unmasking of b-catenin (mouse, 1:50; Sigma-Aldrich) and P-PTEN (rabbit, 1:100; Cell Signaling Technology) epitopes was accomplished with 2100 Retriever (Prestige Medical) in 10 mM citrate buffer (pH 6.0) for 10 min. Nuclei were labeled by using 4'6'-diamidino-2-phenylindole (DAPI). Imaging was performed with Zeiss Axioskop microscope equipped with Zeiss Axiocam digital cameras. Probes specific for Shh, Sox4, and exon 2 deleted in Bmpr1a cKOTM were used for ISH.
Immunoblots.
The following antibodies were dissolved in 5% BSA in 1´ TBS-T and incubated overnight at 4°C: GSK3b (rabbit, 1:1,000; Cell Signaling Technology), P-GSK3b (rabbit, 1:1,000; Cell Signaling Technology), P-PTEN (rabbit, 1:1,000; Cell Signaling Technology), P-AKT (rabbit, 1:1,000; Cell Signaling Technology), AKT (rabbit, 1:1,000; Cell Signaling Technology), and b-actin (mouse 1:1,000; Sigma). Immunoblots were performed with standard protocols.
FACS Analysis.
Purification of bulge cells and total epidermal keratinocytes from adult mouse back skins was described previously [Blanpain C, Lowry WE, Geoghegan A, Polak L, Fuchs E (2004) Cell 118:635-648]. For hair-follicle SC analysis, we used the following primary antibodies: anti-a6 integrin (CD49f) directly coupled to PE (BD Pharmingen) and anti-CD34 coupled to biotin (BD Pharmingen). BrdU detection was performed with a BD Pharmingen BrdU flow kit.