Portal de Periódicos da CAPES

Patterns of Proliferation and Apoptosis during Murine Hair Follicle Morphogenesis

2001; Elsevier BV; Volume: 116; Issue: 6 Linguagem: Inglês

10.1046/j.0022-202x.2001.01368.x

1523-1747

Markus Magerl, Desmond J. Tobin, Sven Müller‐Röver, Evelin Hagen, G Lindner, Ian A. McKay, Ralf Paus,

Cancer and Skin Lesions

In this study, we have correlated cutaneous apoptosis and proliferation in neonatal mice during hair follicle morphogenesis. We have applied a novel triple- staining technique that uses Ki67 immunoreactivity as a marker of proliferation as well as TUNEL and Hoechst 33342 staining as apoptosis markers. We have also assessed the immunoreactivity of interleukin-1β-converting enzyme, caspase 1, a key enzyme in the execution of apoptosis, and of P-cadherin, which has been suggested as a key adhesion receptor in segregating proliferating keratinocytes. The TUNEL data were systematically compared with high resolution light microscopy and transmission electron microscopy data. Virtually all keratinocytes of the developing hair bud were strongly Ki67+, suggesting that the hair bud is not an epidermal invagination but primarily the product of localized keratinocyte proliferation. As hair follicle development advanced, three distinct foci of proliferation became apparent: the distal outer root sheath around the hair canal, the mid outer root sheath, and the proximal hair matrix. Of these proliferating hair follicle keratinocytes only defined subsets expressed P-cadherin. TUNEL+ cells in the hair follicle were not found before stage 5 of murine hair follicle morphogenesis. During the early stages of hair follicle development, interleukin-1β-converting enzyme immunoreactivity was present on all keratinocytes, but virtually disappeared from the proximal hair follicle epithelium later on. High resolution light microscopy/transmission electron microscopy revealed scattered and clustered apoptotic keratinocytes in all epithelial hair follicle compartments throughout hair follicle development, including its earliest stages. This highlights striking differences in the demarcation of apoptotic hair follicle keratinocytes between the TUNEL technique and high resolution light microscopy/transmission electron microscopy and suggests a role for apoptosis in sculpting the hair follicle even during early hair follicle development. In this study, we have correlated cutaneous apoptosis and proliferation in neonatal mice during hair follicle morphogenesis. We have applied a novel triple- staining technique that uses Ki67 immunoreactivity as a marker of proliferation as well as TUNEL and Hoechst 33342 staining as apoptosis markers. We have also assessed the immunoreactivity of interleukin-1β-converting enzyme, caspase 1, a key enzyme in the execution of apoptosis, and of P-cadherin, which has been suggested as a key adhesion receptor in segregating proliferating keratinocytes. The TUNEL data were systematically compared with high resolution light microscopy and transmission electron microscopy data. Virtually all keratinocytes of the developing hair bud were strongly Ki67+, suggesting that the hair bud is not an epidermal invagination but primarily the product of localized keratinocyte proliferation. As hair follicle development advanced, three distinct foci of proliferation became apparent: the distal outer root sheath around the hair canal, the mid outer root sheath, and the proximal hair matrix. Of these proliferating hair follicle keratinocytes only defined subsets expressed P-cadherin. TUNEL+ cells in the hair follicle were not found before stage 5 of murine hair follicle morphogenesis. During the early stages of hair follicle development, interleukin-1β-converting enzyme immunoreactivity was present on all keratinocytes, but virtually disappeared from the proximal hair follicle epithelium later on. High resolution light microscopy/transmission electron microscopy revealed scattered and clustered apoptotic keratinocytes in all epithelial hair follicle compartments throughout hair follicle development, including its earliest stages. This highlights striking differences in the demarcation of apoptotic hair follicle keratinocytes between the TUNEL technique and high resolution light microscopy/transmission electron microscopy and suggests a role for apoptosis in sculpting the hair follicle even during early hair follicle development. dermal papilla hair follicle hair matrix high resolution light microscopy interleukin-1β-converting enzyme, caspase 1 immunoreactivity keratinocyte TdT-mediated dUTP-digoxigenin nick end labeling Hair follicle (HF) pattern formation is a complex process of organogenesis that is tightly regulated by a well-balanced interplay of cell proliferation, polarization, differentiation, and programmed cell death (Glücksmann, 1951Glücksmann A. Cell deaths in normal vertebrate ontogeny.Biol-Rev. 1951; 26: 59Crossref PubMed Scopus (864) Google Scholar;Saunders, 1966Saunders Jr, Jw Death in embryonic systems.Science. 1966; 154: 604-612Crossref PubMed Scopus (698) Google Scholar;Paus et al., 1993Paus R. Rosenbach T. Haas N. Czarnetzki B.M. Patterns of cell death: the significance of apoptosis for dermatology.Exp Dermatol. 1993; 2: 3-11Crossref PubMed Scopus (77) Google Scholar). Previously,Polakowska et al., 1994Polakowska R.R. Piacentini M. Bartlett R. Goldsmith L.A. Haake A.R. Apoptosis in human skin development: morphogenesis, periderm, and stem cells.Dev Dyn. 1994; 199: 176-188Crossref PubMed Scopus (245) Google Scholar reported prominent apoptosis in the epithelium of the developing human HF. In that study, however, the relative contribution of apoptosis to HF morphogenesis, and its coordination with epithelial cell proliferation, adhesion molecule expression, and caspase expression were not systematically assessed. Thus, no comprehensive apoptosis study is yet available on murine hair follicle development, a very instructive research model for general studies on morphogenesis (Gat et al., 1998Gat U. DasGupta R. Degenstein L. Fuchs E. De novo hair follicle morphogenesis and hair tumors in mice expressing a truncated beta-catenin in skin.Cell. 1998; 95: 605-614Abstract Full Text Full Text PDF PubMed Scopus (933) Google Scholar;Oro and Scott, 1998Oro A.E. Scott M.P. Splitting hairs: dissecting roles of signaling systems in epidermal development.Cell. 1998; 95: 575-578Abstract Full Text Full Text PDF PubMed Scopus (138) Google Scholar;Philpott and Paus, 1998Philpott M. Paus R. Principles of hair follicle morphogenesis.in: Chuong C.M. Molecuar Basis of Epithelial Appendage Morphogenesis. Landes Bioscience Publication, Austin, TX1998: 75-103Google Scholar;St Jacques et al., 1998St Jacques B. Dassule H.R. Karavanova I. et al.Sonic hedgehog signaling is essential for hair development.Curr Biol. 1998; 8: 1058-1068Abstract Full Text Full Text PDF PubMed Google Scholar;Botchkarev et al., 1999Botchkarev V. Botchkareva N. Roth W. et al.Noggin is a mesenchymally derived stimulator of hair-follicle induction.Nature Cell Biol. 1999; 3: 158-164Google Scholar). One key question in hair biology is the long-standing controversy whether HF morphogenesis is primarily a cell-migration-driven invagination of the epidermis or whether HF development rather reflects a highly restricted and stringently regulated burst of keratinocyte (KC) proliferation and segregation, which leads to hair bud formation (Danneel, 1931Danneel R. Die Entwicklung der Haare bei der Ratte.Z Morph Ökol Tiere. 1931; 20: 733-754Crossref Scopus (1) Google Scholar;Fleischhauer, 1953Fleischhauer K. Über die Entstehung der Haaranordnung und das Zustandekommen räumlicher Beziehungen zwischen Haaren und Schweissdrüsen.Z Zellforsch U Mikroskop Anat. 1953; 38: 328-355Crossref Scopus (4) Google Scholar;Pinkus et al., 1958Pinkus H. Embryology of hair.in: Montagna W. Ellis R.A. The Biology of the Hair Growth. Academic Press, New York1958Crossref Google Scholar;Müller et al., 1991Müller M. Jasmin J.R. Monteil R.A. Loubiere R. Embryology of the hair follicle.Early Hum Dev. 1991; 26: 159-166Abstract Full Text PDF PubMed Scopus (19) Google Scholar;Paus et al., 1999Paus R. Müller-Röver S. van der Veen C. et al.A comprehensive guide for the recognition and classification of distinct stages of hair follicle morphogenesis.J Invest Dermat. 1999; 113: 523-532Crossref PubMed Scopus (421) Google Scholar). Therefore, this study aimed at characterizing and covisualizing the patterns of proliferation and apoptosis during murine HF morphogenesis using the C57BL/6J mouse model (Paus et al., 1999Paus R. Müller-Röver S. van der Veen C. et al.A comprehensive guide for the recognition and classification of distinct stages of hair follicle morphogenesis.J Invest Dermat. 1999; 113: 523-532Crossref PubMed Scopus (421) Google Scholar). Apoptotic (i.e., TUNEL+) and proliferating (i.e., Ki67+) cells were covisualized by employing a novel immunofluorescent triple-staining protocol that utilizes the DNA dye Hoechst 33342 as a nuclear counterstain. Furthermore, we have studied the coexpression patterns of Ki67 with P-cadherin, which has been suggested to be a key adhesion receptor for the segregation of proliferating KC in the epidermis and in the HF (Hirai et al., 1989Hirai Y. Nose A. Kobayashi S. Takeichi M. Expression and role of E- and P-cadherin adhesion molecules in embryonic histogenesis. II. Skin morphogenesis.Development. 1989; 105: 271-277PubMed Google Scholar;Shimoyama et al., 1989Shimoyama Y. Hirohashi S. Hirano S. Noguchi M. Shimosato Y. Takeichi M. Abe O. Cadherin cell-adhesion molecules in human epithelial tissues and carcinomas.Cancer Res. 1989; 49: 2128-2133PubMed Google Scholar). In addition, the codistribution of TUNEL+ cells with interleukin-1β-converting enzyme (caspase-1, ICE) immunoreactivity (IR) was assessed in order to correlate apoptosis with the follicular distribution of this key caspase in the death execution machinery (Darmon and Bleackley, 1996Darmon A.J. Bleackley R.C. An interleukin-1beta converting enzyme-like protease is a key component of Fas-mediated apoptosis.J Biol Chem. 1996; 271: 21699-21702Crossref PubMed Scopus (32) Google Scholar;Zhivotovsky et al., 1996Zhivotovsky B. Burgess D.H. Orrenius S. Proteases in apoptosis.Experientia. 1996; 52: 968-978Crossref PubMed Scopus (56) Google Scholar;Trump et al., 1997Trump B.F. Berezesky I.K. Chang S.H. Phelps P.C. The pathways of cell death: oncosis, apoptosis, and necrosis.Toxicol Pathol. 1997; 25: 82-88Crossref PubMed Scopus (305) Google Scholar;Allen et al., 1998Allen R.T. Cluck M.W. Agrawal D.K. Mechanisms controlling cellular suicide: role of Bcl-2 and caspases.Cellular Mol Life Sci May. 1998; 54: 427-445Crossref PubMed Scopus (164) Google Scholar). C57BL/6 mice of both sexes (Charles River, Sulzfeld, Germany) were housed in community cages under 12 h light periods, and were fed water and mouse chow ad libitum. All hair cycle stages of neonatal HF morphogenesis (stages 0–8) were studied as described previously (Paus et al., 1999Paus R. Müller-Röver S. van der Veen C. et al.A comprehensive guide for the recognition and classification of distinct stages of hair follicle morphogenesis.J Invest Dermat. 1999; 113: 523-532Crossref PubMed Scopus (421) Google Scholar). The TUNEL technique (TdT-mediated dUTP-digoxigenin nick end labeling, ApopTag, Oncor) was combined with the nuclear stain Hoechst 33342 and an antibody against ICE or Ki67 as previously described (Lindner et al., 1997Lindner G. Botchkarev V.A. Botchkareva N.V. Ling G. van der Veen C. Paus R. Analysis of apoptosis during hair follicle regression (catagen).Am J Pathol. 1997; 151: 1601-1617PubMed Google Scholar;Botchkarev et al., 1998Botchkarev V. Botchkareva N. Albers K. van der Veen C. Lewin G. Paus R. Neutrophin-3 involvement in the regulation of hair follicle morphogenesis.J Invest Dermatol. 1998; 111: 279-285https://doi.org/10.1046/j.1523-1747.1998.00277.xCrossref PubMed Scopus (48) Google Scholar). Eight micron cryostat sections of C57BL/6 back skin were freshly prepared and fixed in formalin (10%) for 10 min at room temperature and postfixed in ethanol/acetic acid (2:1) for 5 min at -20°C. The sections were then covered with equibrilation buffer for 5 min at room temperature, followed by incubation with terminal deoxynucleotidyl transferase (TdT) solution for 1 h at 37°C. This reaction was terminated with stop/wash buffer for 30 min at 37°C. After blocking with normal goat serum [10% in phosphate-buffered saline (PBS), 20 min), the sections were incubated with the primary antibody (ICE or Ki67, 1:100 in PBS, overnight, 4°C). Digoxigenin-dUTP labeled DNA compounds were detected by antidigoxigenin fluorescein isothiocyanate conjugated F(ab)2 fragments (30 min at room temperature). Thereafter, the tissue was covered with the secondary antibody solution (goat anti-rabbit rhodamine, 1:200 in PBS, 30 min, room temperature). Tissue sections were counterstained for 45 min at room temperature with Hoechst 33342 dye (10 µg per ml in PBS). Each staining phase was interspersed by a washing step (three times) in PBS. The antibody used for ICE (Santa Cruz Biotechnology, Santa Cruz, CA) detects the p10 subunit of the murine ICE-heterodimer and recognizes both the inactive and the active form of ICE (Bhat et al., 1996Bhat R.V. DiRocco R. Marcy V.R. et al.Increased expression of IL-1beta converting enzyme in hippocampus after ischemia: selective localization in microglia.J Neurosci. 1996; 16: 4146-4154PubMed Google Scholar;Shimizu et al., 1996Shimizu S. Eguchi Y. Kamiike W. Matsuda H. Tsujimoto Y. Bcl-2 expression prevents activation of the ICE protease cascade.Oncogene. 1996; 12: 2251-2257PubMed Google Scholar). The antibody employed to detect the proliferation-associated antigen Ki67 (Dianova, Hamburg, Germany) is a monoclonal antibody to a 345–395 kDa protein complex (Gerdes et al., 1991Gerdes J. Li L. Schlueter C. et al.Immunobiochemical and molecular biologic characterization of the cell proliferation-associated nuclear antigen that is defined by monoclonal antibody Ki-67.Am J Pathol. 1991; 138: 867-873PubMed Google Scholar) expressed primarily during the S, G2, and M phases of the cell cycle (Sawhney and Hall, 1992Sawhney N. Hall P.A. Ki67-structure, function, and new antibodies.J Pathol. 1992; 168: 161-162Crossref PubMed Scopus (158) Google Scholar;Schwarting, 1993Schwarting R. Little missed markers and Ki-67.Lab Invest. 1993; 68: 597-599PubMed Google Scholar;Gerlach et al., 1997Gerlach C. Golding M. Larue L. Alison M.R. Gerdes J. Ki-67 immunoexpression is a robust marker of proliferative cells in the rat.Lab Invest. 1997; 77: 697-698PubMed Google Scholar). Hoechst 33342 (2′-[4-ethoxyphenyl]-5-[4-methyl-1-piperazinyl]-2,5′-bi-1h-benzimidazole; Sigma Chemical, Diesenhofen, Germany) is a nonintercalating (Sailer et al., 1997Sailer B.L. Nastasi A.J. Valdez J.G. Steinkamp J.A. Crissman H.A. Differential effects of deuterium oxide on the fluorescence lifetimes and intensities of dyes with different modes of binding to DNA.J Histochem Cytochem. 1997; 45: 165-175Crossref PubMed Scopus (35) Google Scholar) DNA-reactive fluorochrome with preferential binding to A-T regions (Crissman and Steinkamp, 1993Crissman H.A. Steinkamp J.A. Cell cycle-related changes in chromatin structure detected by flow cytometry using multiple DNA fluorochromes.Eur J Histochem. 1993; 37: 129-138PubMed Google Scholar). It serves as both a nuclear counterstain and a morphologic indicator of apoptotic cells (Hardin et al., 1992Hardin J.A. Sherr D.H. DeMaria M. Lopez P.A. A simple fluorescence method for surface antigen phenotyping of lymphocytes undergoing DNA fragmentation.J Immunol Meth. 1992; 154: 99-107Crossref PubMed Scopus (58) Google Scholar). For the P-cadherin/Ki67/Hoechst 33342 triple staining, we used a slightly modified version of the protocol established for TUNEL/Ki67/Hoechst 33342 triple staining (Lindner et al., 1997Lindner G. Botchkarev V.A. Botchkareva N.V. Ling G. van der Veen C. Paus R. Analysis of apoptosis during hair follicle regression (catagen).Am J Pathol. 1997; 151: 1601-1617PubMed Google Scholar). The rat monoclonal antimouse P-cadherin antibody PCD-1 (Nose and Takeichi, 1986Nose A. Takeichi M. A novel cadherin cell adhesion molecule: its expression patterns associated with implantation and organogenesis of mouse embryos.J Cell Biol. 1986; 103: 2649-2658Crossref PubMed Scopus (345) Google Scholar) was applied in a dilution of 1:50. The IR patterns were evaluated with a fluorescence microscope (Zeiss, Oberkochen, Germany), using the appropriate filters. Photo micrographs were generated with the help of a digital image analysis system (ISIS, Metasystems, Belmont, MA). Representative tissue samples were obtained from two sites on the backs of 15 mice (three mice each at 0, 1, 3, 5 and 8 d post partum) for a total of 60 blocks from 30 mice. The tissues were fixed in Karnovsky's fixative (Karnovsky, 1965Karnovsky M. Formaldehyde-glutaraldehyde fixative of high osmolarity for use in electron microscopy.Histopathol. 1965; 27: 137-138Google Scholar), postfixed in 2% osmium tetroxide, uranyl acetate, and phenylenediamine, and embedded in resin as previously described (Tobin et al., 1990Tobin D.J. Fenton D.A. Kendall M.D. Ultrastructural observations on the hair bulb melanocytes and melanosomes in acute alopecia areata.J Invest Dermatol. 1990; 94: 803-807Abstract Full Text PDF PubMed Google Scholar;Tobin et al., 1991Tobin D.J. Fenton D.A. Kendall M.D. Cell degeneration in alopecia areata. An ultrastructural study.Am J Dermatopathol. 1991; 13: 248-256Crossref PubMed Scopus (55) Google Scholar). Semi-thin and ultra-thin sections were cut with a Reichart-Jung microtome; the former were stained with the metachromatic stain, toluidine blue/borax, examined by oil-immersion light microscopy, and photographed (Leitz, Wetzlar, Germany). The latter were stained with uranyl acetate and lead citrate and examined and photographed using a Jeol 100CX electron microscope (Jeol, Tokyo, Japan). Sixty HF at each time point were examined by light microscopy from each mouse (total of 300 HF) and 30 HF from each by TEM (total of 150 HF). This study provides strong evidence that the formation of the new HF is based on a tightly regulated and strictly coordinated interplay of apoptosis and cell proliferation, which occurs in a focal rather than a random pattern. Cell death, remarkably, was a significant feature of HF morphogenesis even in its earliest stages. Ki67+ cells were rare in the undifferentiated basal layer of the epidermis (Figure 1a). In contrast, virtually all HF KC in morphogenesis stages 1 and 2 were strongly and homogeneously Ki67+ (Figure 1b, c), indicating that they had entered into the cell cycle (Sawhney and Hall, 1992Sawhney N. Hall P.A. Ki67-structure, function, and new antibodies.J Pathol. 1992; 168: 161-162Crossref PubMed Scopus (158) Google Scholar;Schwarting, 1993Schwarting R. Little missed markers and Ki-67.Lab Invest. 1993; 68: 597-599PubMed Google Scholar;Gerlach et al., 1997Gerlach C. Golding M. Larue L. Alison M.R. Gerdes J. Ki-67 immunoexpression is a robust marker of proliferative cells in the rat.Lab Invest. 1997; 77: 697-698PubMed Google Scholar). This homogeneous Ki67 IR of early hair buds (Figure 1b–d,Figure 2) provides very suggestive evidence for the old concept that hair bud formation is the consequence of massive KC proliferation in a well-defined region of the epidermis, rather than resulting primarily from epithelial cell migration from the epidermis – so-called epidermal "invagination" (Danneel, 1931Danneel R. Die Entwicklung der Haare bei der Ratte.Z Morph Ökol Tiere. 1931; 20: 733-754Crossref Scopus (1) Google Scholar;Fleischhauer, 1953Fleischhauer K. Über die Entstehung der Haaranordnung und das Zustandekommen räumlicher Beziehungen zwischen Haaren und Schweissdrüsen.Z Zellforsch U Mikroskop Anat. 1953; 38: 328-355Crossref Scopus (4) Google Scholar;Pinkus et al., 1958Pinkus H. Embryology of hair.in: Montagna W. Ellis R.A. The Biology of the Hair Growth. Academic Press, New York1958Crossref Google Scholar).Figure 2Distribution of TUNEL+, Ki67+, and P-cadherin+ cells during murine HF morphogenesis. Schematic representation of TUNEL labeling (black) and Ki67 (gray) and P-cadherin (black membrane) IR during all stages of HF development. irs, inner root sheath; ors, outer root sheath.View Large Image Figure ViewerDownload (PPT) A more heterogeneous distribution of Ki67+ cells was a feature of stage 3 HF (Figure 1d), which became increasingly more restricted in maturing HFs to three distinct foci of proliferation (Figure 1e–j, Figure 2): the distal outer root sheath (ORS) around the hair canal (Figure 1i), the mid ORS (Figure 1i,Figure 4l), and the proximal hair matrix (HM) (Figure 1j). The number of Ki67+ cells in the proximal ORS declined during stage 5 (Figure 1f). When the HF reached its largest extension in stage 8, the hair bulb KC showed maximal Ki67 IR (Figure 1j), particularly below Auber's line (Bullough et al., 1958Bullough W.S. Laurence E.B. The mitotic activity of the follicle.in: Montagna W. Ellis R.A. The Biology of Hair Growth. Academic Press, New York1958Crossref Google Scholar;Ramirez et al., 1997Ramirez R.D. Wright W.E. Shay J.W. Taylor R.S. Telomerase activity concentrates in the mitotically active segments of human hair follicles.J Invest Dermatol. 1997; 108: 113-117Crossref PubMed Scopus (152) Google Scholar). Whereas TRITC labeling did not reveal any Ki67 IR on dermal papilla (DP) fibroblasts at any time during neonatal skin and HF development, the more sensitive avidin-biotin complex labeling technique (Müller-Röver et al., 1998Müller-Röver S. Peters E.J.M. Botchkarev V.A. Panteleyev A. Paus R. Distinct patterns of NCAM expression are associated with defined stages of murine hair follicle morphogenesis and regression.J Histochem Cytochem. 1998; 46: 1401-1410Crossref PubMed Scopus (47) Google Scholar) revealed Ki67+ DP fibroblasts located in the pre-DP and DP throughout all stages of HF morphogenesis (Figure 3b,d,f,h,j). Ki67+ fibroblasts were also located in the connective tissue sheath (CTS) during stages 3 and 4 (Figure 3f,h). These findings were confirmed by HRLM (Figure 3a,c,e,g,i). P-cadherin expression has been suggested by some as a useful marker for discriminating proliferating and nonproliferating epidermal and follicular KC (Hirai et al., 1989Hirai Y. Nose A. Kobayashi S. Takeichi M. Expression and role of E- and P-cadherin adhesion molecules in embryonic histogenesis. II. Skin morphogenesis.Development. 1989; 105: 271-277PubMed Google Scholar;Shimoyama et al., 1989Shimoyama Y. Hirohashi S. Hirano S. Noguchi M. Shimosato Y. Takeichi M. Abe O. Cadherin cell-adhesion molecules in human epithelial tissues and carcinomas.Cancer Res. 1989; 49: 2128-2133PubMed Google Scholar;Fujita et al., 1992Fujita M. Furukawa F. Fujii K. Horiguchi Y. Takeichi M. Imamura S. Expression of cadherin cell adhesion molecules during human skin development: morphogenesis of epidermis, hair follicles and eccrine sweat ducts.Arch Dermatol Res. 1992; 284: 159-166Crossref PubMed Scopus (58) Google Scholar;Müller-Röver et al., 1999Müller-Röver S. Tokura Y. Welker P. Furukawa F. Wakita H. Takigawa M. Paus R. E- and P-cadherin expression during murine hair follicle morphogenesis and cycling.Exp Dermatol. 1999; 4: 237-246Google Scholar). Ki67 and P-cadherin IR were very commonly co-localized in the same cells during HF morphogenesis stages 0–2, where homogeneous staining occurred in all follicular KC (Figure 1w). This co-expression pattern became restricted to the innermost portion of the HM, however, and to the mid ORS during stages 3–6 (Figure 1x,y). In contrast to human skin development (Fujita et al., 1992Fujita M. Furukawa F. Fujii K. Horiguchi Y. Takeichi M. Imamura S. Expression of cadherin cell adhesion molecules during human skin development: morphogenesis of epidermis, hair follicles and eccrine sweat ducts.Arch Dermatol Res. 1992; 284: 159-166Crossref PubMed Scopus (58) Google Scholar), we show here that KC in the outermost portion of the murine HM display isolated Ki67 IR but no P-cadherin IR (Figure 1x–z, Figure 2), whereas KC in the precortical murine HM display isolated P-cadherin IR and no Ki67 IR (Figure 1x–z, Figure 2). This suggests that follicular P-cadherin expression is much less tightly coupled to the proliferative epithelial cell compartment than previously proposed (Hirai et al., 1989Hirai Y. Nose A. Kobayashi S. Takeichi M. Expression and role of E- and P-cadherin adhesion molecules in embryonic histogenesis. II. Skin morphogenesis.Development. 1989; 105: 271-277PubMed Google Scholar;Shimoyama et al., 1989Shimoyama Y. Hirohashi S. Hirano S. Noguchi M. Shimosato Y. Takeichi M. Abe O. Cadherin cell-adhesion molecules in human epithelial tissues and carcinomas.Cancer Res. 1989; 49: 2128-2133PubMed Google Scholar). Throughout HF morphogenesis, scattered TUNEL+ cells were observed in the epidermis, most commonly in the stratum granulosum (Figure 1a,c,k,l,m,s). In contrast, intrafollicular TUNEL+ cells were never observed during the first stages of HF development (stages 1–4) (Figure 1a–e,k–o) and appeared first during stage 5. Interestingly, they were frequently seen just in and above the newly generated inner root sheath (IRS) (Figure 1f,p). During stage 6, several strongly TUNEL+ cells with an elongated, flattened appearance became visible in the IRS at the level of the tip of the emerging hair shaft (Figure 1g,q). Additional TUNEL+ cells appeared in the mid IRS as well as in the distal and infundibular ORS during stages 7 and 8 (Figure 1h,i,r,s). Clustering of TUNEL+ cells, as is characteristic of hair follicle regression (catagen) (Lindner et al., 1997Lindner G. Botchkarev V.A. Botchkareva N.V. Ling G. van der Veen C. Paus R. Analysis of apoptosis during hair follicle regression (catagen).Am J Pathol. 1997; 151: 1601-1617PubMed Google Scholar), occurred only late in HF development. During stages 7 and 8 of HF morphogenesis, TUNEL+ cells were concentrated within three distinct apoptosis "hot spots", namely the distal and mid IRS and the infundibular ORS, similar to human HF development (Polakowska et al., 1994Polakowska R.R. Piacentini M. Bartlett R. Goldsmith L.A. Haake A.R. Apoptosis in human skin development: morphogenesis, periderm, and stem cells.Dev Dyn. 1994; 199: 176-188Crossref PubMed Scopus (245) Google Scholar). The distribution pattern of TUNEL+ cells was correlated with the expression of ICE (caspase 1), a key enzyme in the death execution machinery of most cells (Zhivotovsky et al., 1996Zhivotovsky B. Burgess D.H. Orrenius S. Proteases in apoptosis.Experientia. 1996; 52: 968-978Crossref PubMed Scopus (56) Google Scholar;Porter et al., 1997Porter A.G. Ng P. Janicke R.U. Death substrates come alive.Bioessays. 1997; 19: 501-507Crossref PubMed Scopus (164) Google Scholar;Allen et al., 1998Allen R.T. Cluck M.W. Agrawal D.K. Mechanisms controlling cellular suicide: role of Bcl-2 and caspases.Cellular Mol Life Sci May. 1998; 54: 427-445Crossref PubMed Scopus (164) Google Scholar). During HF regression (catagen) in adolescent mice, we had previously demonstrated that all TUNEL+ cells are located in regions of the HF epithelium that are also strongly ICE+ (Lindner et al., 1997Lindner G. Botchkarev V.A. Botchkareva N.V. Ling G. van der Veen C. Paus R. Analysis of apoptosis during hair follicle regression (catagen).Am J Pathol. 1997; 151: 1601-1617PubMed Google Scholar). Whereas epidermal KC were ICE+ throughout HF morphogenesis, all KC in the HF were positive only until the hair bud had developed (Figure 1k–n). Thereafter, ICE IR disappeared from the proximal hair bulb during the formation of the IRS (Figure 1o,p), which itself became strongly ICE+ during stage 6 (Figure 1q). The bulge (Figure 1q,r), the infundibulum, and the mid to distal ORS were ICE+ during stages 6–8 (Figure 1r–v). DP or connective tissue sheath fibroblasts never displayed any ICE IR during HF morphogenesis. In stage 6, prominent ICE IR was observed in precursor sebocytes (Figure 1q). There was a gradual reduction in ICE IR in the IRS during stages 7 and 8 (Figure 1s). The very prominent ICE IR in the early stages of HF morphogenesis (Figure 1k–n) and the gradual yet marked decline in ICE IR in most of the middle and proximal part of the developing HF epithelium (Figure 1o–s) raises the possibility that this caspase also has enzymatic functions that are unrelated to apoptosis control. Morphologic evidence of apoptosis was obtained using HRLM for stages 1–8 HF in 0.5-1 µm resin sections, and was subsequently confirmed using transmission electron microscopy. Scattered apoptotic KC or clusters of apoptotic KC were detected in all follicular compartments throughout HF development (Figure 4a–k, m–o), notably also during the earliest stages of HF development (Figure 4a, b, d) when no TUNEL+ cells were found.Figure 4HRLM and TEM analysis of proliferation and apoptosis during hair follicle morphogenesis. Figures display longitudinal sections of developing hair follicles. (a) HRLM. Stage 1 HF. Note the presence of three apoptotic cells (arrowheads) located above columnar basal KC. (b) HRLM. Stage 2 HF. Note the presence of a single apoptotic cell (arrowhead) in the developing HF with classic apoptotic features including (i) rounding up of the cell and shrinkage from surrounding healthy cells, (ii) the rounding up and condensation of the nucleus, and (iii) the hyperchromatism of the cytoplasm. This cell has yet to be engulfed by surrounding cells. e, epidermis; d, dermis; v, vessel. (c) TEM. Stage 2 HF. Note KC mitosis in the developing HF (encircled area). (d) HRLM. Stage 3 HF (dotted line). Note the presence of three apoptotic cells (arrowheads) in the matrix just above the DP in close proximity to mitotic cell (encircled region). (e) HRLM. Note the formation of the IRS. Some of these differentiating IRS cells have a passing resemblance to pyknotic cells (arrowheads). Compare them with TUNEL+ cells in Fig 1(f,g) and 1(p,q). (f) TEM. Stage 5 HF. At the tip of the developing inner root sheath (irs) note nuclear remnants in the keratinizing IRS. Compare them with TUNEL+ cells in Figure 1(f,g) and 1(p,q). ors, outer root sheath. (g) HRLM. Stage 6 HF. Note the presence of three apoptotic cells (arrowheads) in the precortex of the hair bulb. Note also the upper apoptotic cell with classic apoptotic features including contraction of the cell and nuclear condensation. dp, dermal papilla. (h) TEM. Stage 6 HF. Tip of the developing hair cone. Note presence of three apoptotic KC (arrowheads) adjacent to the outer root sheath (ors). (i) TEM. Apoptotic KC (arrowheads) of Figure 1(h) in higher magnification. Note classic apoptotic features: contraction of the cell and nuclear condensation (left), vacuolization (right). Note also the retention of organelle integrity (right, at the bottom). (j) TEM. Stage 6 HF. Tip of hair cone passing by the sebocytes of developing sebaceous gland. Arrowhead: apoptotic cell; ors, outer root sheath. (k) TEM. Stage 6 HF. Higher magnification of the apoptotic cell (arrowhead) in Figure 1(j). Note cell shrinkage and vacuolization. (l) TEM. Stages 6–7 HF. Mitotic KC in the middle outer root sheath (ors). (m) TEM. Stage 7 HF. Hair cone including apoptotic KC (arrowheads). Note nuclear emargination, cell body shrinkage, and cytoplasmic vacuolization. Ors, outer root sheath. (n) TEM. Stage 8 HF. Note the presence of degenerating cell in the middle outer root sheath (ors) close to the inner root sheath (irs). There appears to be a breakdown of the nuclear material into condensed spherical structures (arrowhead) without typical apoptosis-associated changes in cell shape. he, Henle's layer; hu, Huxley's layer; irs, inner root sheath; ors, outer root sheath. (o) TEM. Longitudinal section of distal HF at level of the sebaceous gland. Note phagocytosis of apoptotic bodies (arrowhead) in the outer root sheath (ors). The apoptotic body consists of a highly condensed and electron dense nuclear fraction in addition to condensed cytoplasm. Note also copious amounts of glycogen in ORS KC. Lobules of the sebaceous gland (sg) appear in the upper part.View Large Image Figure ViewerDownload (PPT) Apoptosis was first convincingly detectable by HRLM/TEM during stages 1–2 (Figure 4a,b). Here, scattered cells (Figure 4b) or small clusters of two to four cells, invariably located in the suprabasal layer (Figure 4a), were observed to contain portions of intensely staining cytoplasm, usually with a smooth contour and with small, densely basophilic (pyknotic) nuclei. There was considerable cell shrinkage associated with cytoplasmic contraction and loss of cell-cell junctions (leaving a halo around the cell). Interestingly, visible apoptosis occurred in the primitive hair germ (stage 2) (Figure 4b,c), even before the formation of a recognizable DP. With advancing morphogenesis, stage 3–4 HF (Figure 4d, dotted line) exhibited clusters of apoptotic cells in the developing matrix just above the – now partially enclosed – DP. Cells undergoing apoptosis were observed in direct proximity to cells undergoing mitosis (Figure 4d, encircled area). Yet, until stage 4 (Figure 1a–e,k–o), no intrafollicular TUNEL+ cells were found, whereas HRLM already revealed clusters of apoptotic KC in the HM just above the DP (Figure 4a,d,e). Some intrafollicular TUNEL+ KC displayed distinct morphologic features not typical for apoptotic cells, such as elongated nuclei and extended forms (Figure 1g–i,q,r), instead of the shrunken and rounded phenotype (Figure 1i,l,r) characteristic of TUNEL+ apoptotic cells (Hossain et al., 1997Hossain M.M. Nakayama H. Takashima A. Goto N. Doi K. 5-Azacytidine (5Az) induces apoptosis in PC12 cells: a model for 5Az-induced apoptosis in developing neuronal cells.Histol Histopathol. 1997; 12: 439-445PubMed Google Scholar;Trump et al., 1997Trump B.F. Berezesky I.K. Chang S.H. Phelps P.C. The pathways of cell death: oncosis, apoptosis, and necrosis.Toxicol Pathol. 1997; 25: 82-88Crossref PubMed Scopus (305) Google Scholar). This suggests that these TUNEL+ cells are terminally differentiating, rather than truly apoptotic KC. Cell proliferation in the developing HF epithelium was common in stage 2–8 HF. When colocalized with apoptosis, proliferation tended to occur distally to these dying cells, particularly in the mid ORS or precortex region of the matrix. During stages 4–5 of HF morphogenesis, the developing IRS assumed a cone-shaped structure (Figure 4e,f), consisting of terminally differentiating KC that were highly condensed and contained pyknotic nuclei. These cells, however, did not show classic apoptotic morphology. KC in the precortex region of the HM, however, exhibited classic apoptotic features including nuclear fragmentation and cytoplasmic condensation (Figure 4g), whereas pyknotic IRS cells were highly extended in the hair growth direction and exhibited dissolution of cellular contents. Classic apoptosis was also regularly observed adjacent to the emerging hair cone during stages 6 and 7 (Figure 4h–k,m). Apoptotic cells were also seen in other regions of the developing HF, including mid HF close to the IRS (Figure 4n) and upper HF, at (Figure 4o) and below (Figure 4h,i) the level of the sebaceous gland. Ultrastructural analysis of these cells confirmed a diagnosis of apoptosis (Wyllie et al., 1981Wyllie A.H. Beattie G.J. Hargreaves A.D. Chromatin changes in apoptosis.Histochem J. 1981; 13: 681-692Crossref PubMed Scopus (142) Google Scholar;Cotter et al., 1990Cotter T.G. Lennon S.V. Glynn J.G. Martin S.J. Cell death via apoptosis and its relationship to growth, development and differentiation of both tumour and normal cells.Anticancer Res. 1990; 10: 1153-1159PubMed Google Scholar). Consistent with the findings ofBullough et al., 1958Bullough W.S. Laurence E.B. The mitotic activity of the follicle.in: Montagna W. Ellis R.A. The Biology of Hair Growth. Academic Press, New York1958Crossref Google Scholar and our Ki67 data, mitotic cells were occasionally found in the middle ORS (Figure 4l). These differences between TUNEL staining and HRLM/TEM underscore the importance of using additional in situ techniques for apoptosis detection such as HRLM/TEM or confocal laser scanning microscopy (Cervos Navarro and Schubert, 1996Cervos Navarro J. Schubert T.E. Pitfalls in the evaluation of apoptosis using TUNEL.Brain Pathol. 1996; 6: 347-348PubMed Google Scholar;Kimura et al., 1997Kimura K. Sasano H. Shimosegawa T. et al.Ultrastructural and confocal laser scanning microscopic examination of TUNEL-positive cells.J Pathol. 1997; 181: 235-242Crossref PubMed Scopus (33) Google Scholar) in order to validate TUNEL data. Our findings demonstrate that TUNEL staining, at least in murine HF epithelium, is neither highly sensitive (false-negative results during earliest stages of HF development), nor highly specific (false-positive results during later HF morphogenesis). In summary, methodologically, this study reports a novel triple staining for correlating apoptosis and proliferation in epithelial cell compartments and highlights that TUNEL data on HF apoptosis must be interpreted in conjunction with HRLM/TEM data to avoid false-negative and false-positive in situ end labeling results. Furthermore, this study further questions whether P-cadherin specifically discriminates proliferating from nonproliferating KC as suggested previously (Hirai et al., 1989Hirai Y. Nose A. Kobayashi S. Takeichi M. Expression and role of E- and P-cadherin adhesion molecules in embryonic histogenesis. II. Skin morphogenesis.Development. 1989; 105: 271-277PubMed Google Scholar;Shimoyama et al., 1989Shimoyama Y. Hirohashi S. Hirano S. Noguchi M. Shimosato Y. Takeichi M. Abe O. Cadherin cell-adhesion molecules in human epithelial tissues and carcinomas.Cancer Res. 1989; 49: 2128-2133PubMed Google Scholar). Our study also reports the first profile of caspase 1 expression during HF development. This profile raises the possibility that this key death execution enzyme has functional properties in hair biology that extend beyond apoptosis. In addition, we show that developing murine HF are characterized by well-defined, developmentally controlled foci of proliferation and apoptosis. Finally, we present evidence that early HF development is not primarily a migratory phenomenon of epidermal invagination but a proliferation-driven event including apoptotic processes to control the effects of over-proliferation and for fine sculpturing of the architecture of the follicle epithelium. The excellent technical assistance of Ruth Pliet is most gratefully acknowledged. This study was supported in part by grants from the EU (Brite EURam Project No. BE 97-4301) to I.A.M. and R.P. and Deutsche Forschungsgemeinschaft to R.P. (Pa 345/8-2).