| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
OPEN ACCESS ARTICLE
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TISSUE-SPECIFIC STEM CELLS |
aMassachusetts General Hospital, Center for Regenerative Medicine, Boston, Massachusetts, USA;
bCR-UK Cambridge Research Institute, Cambridge, United Kingdom;
cBarts and The London Queen Mary's School of Medicine and Dentistry, Institute of Cell and Molecular Science, Centre for Cutaneous Research, London, United Kingdom;
dWellcome Trust Centre for Stem Cell Research, Tennis Court Road, Cambridge, United Kingdom;
eUniversity of Massachusetts Cancer Center Tissue Bank, Departments of Cancer Biology and Pathology, University of Massachusetts Medical School, Worcester, Massachusetts, USA;
fDepartments of Dermatology, Venereology, Allergology and Immunology, Dessau Medical Center, Dessau, Germany
Key Words. Epidermis • Sebaceous glands • Cell differentiation • Myc oncogene • β-catenin
Correspondence: Fiona M. Watt, MA, D.Phil., CR-UK Cambridge Research Institute, La Ka Shing Centre, Robinson Way, Cambridge CB1 ORE, United Kingdom. Telephone: +44 1223 404400; Fax: +44 1223 404573; e-mail: fiona.watt{at}cancer.org.uk
Received August 10, 2007;
accepted for publication February 12, 2008.
First published online in STEM CELLS EXPRESS February 28, 2008.
 |
ABSTRACT |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionConclusionDisclosure of Potential...AcknowledgmentsReferences |
|---|
Disclosure of potential conflicts of interest is found at the end of this article.
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionDisclosure of Potential...AcknowledgmentsReferences |
|---|
There is good evidence for interdependence of the sebaceous gland and hair follicle: in situations in which one organ collapses the other is often also lost [3]. In addition, the sebaceous gland can be regenerated by the reservoir of stem cells in the hair follicle bulge [4–7]. Nevertheless, the interdependence is not obligatory. Sebaceous glands are present in some mouse mutants that lack hair follicles [8, 9], and sebaceous glands can be induced in footpad epidermis, an anatomic region normally devoid of hair follicles and sebocytes [10]. Retroviral lineage marking has provided strong evidence that the sebaceous gland can arise and be maintained independently of the hair follicle bulge [11], and the transcription factor Blimp1 has recently been described as a specific marker of sebaceous gland progenitor cells [12]. The different observations can be reconciled if there is a stem cell compartment that normally maintains the sebaceous gland, but can be replenished, following injury or Blimp1 deletion, by stem cells from the bulge [12, 13].
There is evidence from a variety of mouse models that c-Myc (Myc) and β-catenin exert opposing effects on sebaceous gland differentiation. Activation of Myc favors differentiation along the lineages of the interfollicular epidermis and sebaceous gland and results in the appearance of groups of differentiated sebocytes within the interfollicular epidermis [14–17]. Blimp1 is reported to negatively regulate Myc, which would be consistent with a role for Myc in promoting SG differentiation [12]. In contrast, activation of β-catenin induces de novo hair follicle morphogenesis and regression of sebaceous glands [18–21]. Disruption of β-catenin signaling by overexpression of dominant negative forms of Lef1 (
NLef1) suppresses hair follicle differentiation, stimulating sebocyte and interfollicular epidermal differentiation [22, 23]. NLef1 also triggers formation of sebaceous tumors [23]. The different effects of Myc and β-catenin on sebaceous gland differentiation are surprising, because Myc is reported to be a β-catenin target gene [24] and acts downstream of β-catenin in intestinal epithelium [25].
In the present report, we set out to investigate the effects of Myc and β-catenin in an in vitro model of sebocyte differentiation, the SZ95 line of immortalized human sebocytes [26]. In the course of these studies we have obtained evidence for the existence of a common progenitor of sebaceous gland and interfollicular epidermis, providing a model for human epidermal lineage selection in vitro.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionDisclosure of Potential...AcknowledgmentsReferences |
|---|
For the clonogenicity assay equal numbers (100–1000) of viable keratinocytes and SZ95 cells were plated per 60 mm dish in triplicate on a 3T3-J2 feeder layer and cultured in keratinocyte medium [28] for 2 weeks prior to fixation.
De-epidermized dermis (DED) was prepared from adult breast skin as described previously [30]. 105 SZ95 cells were resuspended in 20 µl complete Sebomed culture medium, seeded onto the denuded epithelial surface of each DED and cultured at the air-medium interface on sterile tissue culture inserts (Becton-Dickinson [BD Bioscience], Franklin Lakes, NJ, http://www.bd.com) in 6-well plates for 2–3 weeks.
The Seb-E6E7 line of immortalized sebaceous gland cells was generated from adult human facial skin collected with IRB approval following a facelift procedure. After treating the skin with Dispase (Sigma, Gillingham, Dorset, U.K., http://www.sigmaaldrich.com) overnight at 4°C, sebaceous glands were micro-dissected under a dissecting microscope. The glands were digested with 0.05% trypsin-EDTA (Gibco, Paisley, U.K., http://www.invitrogen.com) for 10 minutes and then with a mixture of trypsin-EDTA and Versene (Gibco, 1.33:1) for an additional 20 minutes at room temperature. Disaggregated cells were collected by centrifugation for 5 minutes at 800 rpm and plated on mitomycin C (1.5 µg/ml DMEM for 2 hours) treated 3T3-J2 cells in keratinocyte medium (KCM), as described previously [31]. Three weeks later the cells were immortalized by transduction with a retroviral vector containing HPV16/E6E7 genes (LXSN-16E6E7) packaged in PA317 cells (kindly provided by James Rheinwald, Harvard Medical School, Boston, MA), as described previously [32]. Sebocytes were transduced by co-culture with mitomycin C-treated packaging cells in the presence of 3T3-J2 cells. Six days later the PA317 and 3T3-J2 cells were replaced with mitomycin-treated 3T3-J2 NHP cells (neomycin, hygromycin, puromycin resistant 3T3-J2 cells, kindly provided by James Rheinwald, Harvard Medical School, Boston, MA). Infected cells were selected in 0.2 mg/ml of G418 (Gibco) for 6 days, then re-plated at clonal density. One stably transduced clone of cells was isolated and expanded to give rise to the Seb-E6E7 line.
Retroviral Infection of Keratinocytes and SZ95 Cells
SZ95 cells and primary human keratinocytes were transduced with retroviral vectors using the AM12 packaging line as a feeder layer, essentially as described previously [33]. The following vectors were used: pBabePuro (empty vector), pBabePuroMycER [28], pBabePuroNβ-catenin [34], pBabePuroInvolucrin-GFP reporter [35], and pBabePuroIhh. To improve the efficiency of the infection, polybrene was added to a final concentration of 2.5 µg/ml for the first 2 days of co-culture. After 2–7 days, the AM12 cells were removed by treatment with EDTA and infected SZ95 cells were selected in puromycin (1 µg/ml for the first day then 0.6 µg/ml for an additional week). Cells transduced with MycER were incubated with 200 nM 4-hydroxy-Tamoxifen (4OHT) for 2 or 3 days to induce Myc activation. No differences in proliferation or differentiation were observed between uninfected or empty vector infected cells treated with 4OHT and untreated MycER infected cells.
Time-Lapse Video Microscopy
For time-lapse video microscopy, 1–2 x 105 SZ95 cells were plated on a 35 mm dish, kept humidified at 37°C in 5% CO2, and videotaped for up to 48 hours. Frames were taken every 4 minutes using Olympus IMT1 or IMT2 inverted microscopes (Olympus, Philadelphia, http://www.olympusamerica.com) driven by Broadcast Animation Controllers (BAC 900) and fitted with monochrome CCD cameras and video recorders (Sony M370 CE and PVW-2800P, Sony, Tokyo, http://www.sony.net). Recordings were digitized, and the sequence of all frames was run on a PC. Motility was measured using a cell tracking extension (CR-UK) written for IPLab (Signal Analytics Inc. IP Lab, Rockville, MD, http://www.scanalytics.com), and speed was calculated using a program written in Mathematica by Daniel Zicha (CR-UK).
SZ95 Xenografts
To assess the differentiation potential of human SZ95 sebocytes in vivo, we modified a previously described skin reconstitution protocol [6, 35]. Five x 105 fibroblasts, freshly isolated from the skin of 1–3-day old Balb/C mice, and 5 x 105 SZ95 cells were suspended in 0.1 ml phosphate buffered saline and injected subcutaneously into adult nude mice. After four weeks, cysts were collected and embedded in optimal cutting temperature (OCT) compound for preparation of frozen sections.
Generation and Treatment of Double Transgenic Mice
All mouse husbandry and experimental procedures were conducted in compliance with the protocols established by the Cancer Research UK animal ethics committee under the terms of a UK Home Office license. Heterozygous K14Nβ-cateninER mice (line D4) [20] were crossed with heterozygous K14MycER mice [14] to generate double-transgenic mice and littermate controls. At the start of every experiment mice were 6–8 weeks old and, therefore, in the telogen (resting) phase of the hair cycle.
To activate the MycER and Nβ-cateninER transgenes, 1.5 mg 4OHT (Sigma, H6278) diluted in 200ul acetone was applied to tail skin using a paintbrush every second day. In some experiments, cyclopamine (Biomol, Exeter, U.K., http://www.biomol.com) (50 µM in ethanol) was topically applied to tail skin daily. When mice were treated with both drugs, 1 mg 4OHT was applied 30 minutes after cyclopamine every second day. To label cells that were going through S-phase of the cell cycle, 100 mg/kg of sterile-filtered bromodeoxyuridine (BrdU) (Sigma) was injected into the intraperitoneal cavity 1 hour before harvesting of skin.
Antibodies
The following primary antibodies were used at the dilutions indicated: actin (AC-40, mouse monoclonal, Sigma, 1:1,000); Blimp1 (rabbit polyclonal, 1:100; kind gift of Reuben Tooze; [12]); cornifin (SQ37C, rabbit polyclonal, 1:500 [36]); c-Myc (N-262, rabbit polyclonal, Santa Cruz Biotechnology, Inc., Santa Cruz, CA, http://www.scbt.com; 1:100); EMA/MUC1 (E29, mouse monoclonal, Dako, Carpinteria, CA, http://www.dako.com; 1:100); fatty acid synthase (18341, rabbit polyclonal (IBL, Tokyo, http://www.ibl-japan.co.jp/en/service/service.html), 1:50); involucrin (SY5, mouse monoclonal, 1:100 for immunofluorescence and immunohistochemistry, 1:2,000 for western blotting, [37]); keratin seven (K7) (LP1K mouse monoclonal, 1:100 for immunofluorescence, 1:1,000 for western blotting [38]); keratin 10 (PRB-159P, rabbit polyclonal, [Covance, Princeton, NJ, http://www.covance.com] 1:000); keratin 14 (PRB-155P, rabbit polyclonal, Covance, 1:1,000); keratin 17 (rabbit polyclonal, 1:1000; kind gift from Pierre Coulombe [39]); ER (MC-20, rabbit polyclonal, Santa Cruz, 1:500); tubulin (SAP4G5, mouse monoclonal, Sigma, 1:10,000) and BrdU (OBT0030, [Transduction Laboratories, BD Biosciences] 1:100). Species-specific secondary antibodies were conjugated to AlexaFlour 488 or 594 (Invitrogen, 1:1,000) or HRP (BD Biosciences, 1:2,000). Nuclei were counterstained with DAPI (Invitrogen; 1:10,000).
Histology and Immunolabeling
Human scalp tissue (obtained as surgical waste with appropriate ethical approval), SZ95 DED cultures and mouse dorsal and tail skin were either frozen, unfixed, in OCT compound (BDH, Poole, U.K., http://uk.vwr.com) for cryosectioning or fixed in 10% neutral buffered formalin overnight for paraffin embedding. Cultured cells on coverslips were fixed and permeabilized essentially as described previously [28]. Whole mount mouse tail samples were prepared and stained as described previously [16].
For immunofluorescence staining, frozen sections and cells on coverslips were blocked for one hour in 10% FCS in PBS, then incubated with primary antibody for 1 hour (room temperature) or overnight (4°C), washed and incubated with the appropriate secondary antibody for 30 minutes. All antibodies were diluted in PBS containing 10% FCS. BrdU labeled cells were detected in paraffin sections as described previously [20]. Nile Red (Sigma) was used to detect lipid droplets in frozen sections and cultured cells. A 1,000x stock solution (10 mg/ml) was prepared in methanol and diluted in PBS for use. Sections and cells were incubated with Nile Red for 15 minutes. Fluorescence images were acquired with a Zeiss Photomicroscope three epifluorescence microscope or a Zeiss LSM 510 confocal microscope (Zeiss, Welwyn Garden City, U.K., http://www.zeiss.com). Photoshop CS2 (Adobe, San Jose, CA, http://www.adobe.com) was used to adjust brightness, contrast and picture size.
To quantitate BrdU incorporation in tissue sections, the average number of BrdU-positive cells per 25x microscopic field (± SEM) was determined for a minimum of three experimental mice. For quantitation of differentiating SZ95 cells by immunofluorescence staining, the number of cornifin, involucrin or Nile Red positive cells was scored in five random 25x microscopic fields per coverslip. At least three coverslips per experiment were examined and each experiment was performed in triplicate. The two-tailed t-test was used to assess the statistical significance of the data.
Western Blotting
Transgene expression was assessed by western blotting with an anti ER antibody as described previously [20]. For analysis of K7 and involucrin expression, cells were lysed in RIPA buffer (50 mM Tris-HCl pH 7.4, 1% NP40, 0.25% sodium deoxycholate, 150 mM NaCl, 1 mM EGTA) supplemented with protease and phosphatase inhibitor cocktails (Roche, Basel, Switzerland, http://www.roche.com). Cells were incubated in RIPA buffer for 10–30 min at 4°C then centrifuged at 14,000 rpm for 10 min at 4°C on a bench top centrifuge. Cleared lysates were assayed for protein content and then immunoblotted using the same method as for ER detection [20].
Chromatin Immunoprecipitation
Primary human keratinocytes were infected with the inducible constructs MycER and Myc106ER or pBabe empty vector as a control [28]. After incubation with 4OHT for 24 hours, the cells were harvested and lysates containing protein/DNA complexes were prepared using a ChIP Assay Kit (Upstate, Millipore, Watford, U.K., http://www.millipore.com), according to the manufacturer's instructions with slight modifications.
For each immunoprecipitation, 2 ml of diluted lysate was precleared with 60 µl blocked protein A beads (Upstate) for 2 hours at 4°C with agitation. Samples were immunoprecipitated overnight at 4°C with rabbit anti c-Myc (4 µg N262, Santa Cruz) or an isotype matched antibody as a negative control. Immune complexes were recovered with 60 µl blocked protein A beads (4 hours at 4°C). Beads were washed and eluted, and crosslinks were reversed according to the manufacturer's instructions. Eluted material was phenol/chloroform-extracted and ethanol-precipitated. DNA was resuspended in 20 µl dH2O. PCR was performed with 2 µl of DNA in a final volume of 25 µl. Amplified constructs were visualized on 1.4% agarose gels. Primer sequences for the Indian hedgehog (Ihh) promoter region were: forward 5' GAG GTG GGA AAA GGA ACC TGC CC and reverse 5' TCA GGG AGC CAG CTC CAC CAT CC.
|
RESULTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionDisclosure of Potential...AcknowledgmentsReferences |
|---|
|
Blimp1 has been described as a marker of progenitor cells in the sebaceous glands of mouse skin [12]. In agreement with a recent report [44], we found that in human skin Blimp1 was expressed in the interfollicular epidermis, predominantly in the suprabasal layers, including the involucrin positive upper spinous layers (Fig. 1E). Blimp1 was also detected in scattered cells of the outer (involucrin negative) and inner (involucrin positive) root sheath of hair follicles (Fig. 1E) and in terminally differentiating cells of the sebaceous gland (Fig. 1I). We could not, therefore, use Blimp1 to discriminate between the interfollicular epidermis (IFE) and sebocyte lineages in adult human epidermis.
Expression of K7 and Involucrin by Immortalized Human Sebocytes
SZ95 cells are derived by immortalization of human facial sebocytes with SV40 large T antigen, and retain many typical characteristics of primary sebocytes [26] and when they undergo terminal differentiation they round up and accumulate lipids [26, 29]. In preconfluent cultures the majority of SZ95 cells express K7, whereas there are very few K7 positive cells in cultured keratinocytes from human interfollicular epidermis (Fig. 2A–2C).
|
In order to investigate the possibility that SZ95 cells expressed involucrin as a consequence of immortalization by SV40, we generated a new line of human sebocytes immortalized by introduction of HPV16 E6 and E7 genes [32]. These cells, Seb-E6E7, are a clonal line derived from sebaceous glands of adult human face skin and have been in culture for more than 30 passages. We compared them with vp, an HPV16 immortalized line of IFE keratinocytes derived from neonatal foreskin [27]. Seb-E6E7 cells, like SZ95 cells, expressed both K7 and involucrin (Fig. 2C, 2D). The level of involucrin expression was comparable in Seb-E6E7 and vp (Fig. 2D); however, Seb-E6E7 expressed higher levels of K7 (Fig. 2C). The latter observation is consistent with the fact that vp, unlike Seb-E6E7, do not undergo sebocyte terminal differentiation in culture (data not shown).
We conclude that, whereas SZ95 and SebE6-E7 cells are derived from sebaceous gland and undergo sebocyte differentiation, they also contain cells with the capacity to express involucrin, a marker of terminally differentiated cells in the IFE and inner root sheath of the HF.
SZ95 Cells Are Bipotential Progenitors of Sebaceous Gland and Interfollicular Epidermis
The differentiation capacity of SZ95 cells was further examined by culturing them on a human dermal substrate (de-epidermized dermis; DED) at the air-liquid interface for 2 weeks, conditions that promote interfollicular epidermal differentiation [30]. SZ95 cells formed a partially stratified epithelium (Fig. 2H). Cells in the basal layer expressed K7 (Fig. 2I, 2K, 2M), while cells above the basal layer consisted of patches of Nile Red positive (Fig. 2I) or involucrin and cornifin positive cells (Fig. 2J, 2L). Cells that were Nile Red positive tended not to express involucrin or cornifin (Fig. 2J), consistent with the absence of involucrin expression by terminally differentiated sebocytes (Fig. 1J, 1K). The suprabasal cells, whether accumulating lipids or interfollicular epidermal differentiation markers, had reduced levels of K7 compared to the basal cells (Fig. 2I, 2M).
To investigate whether individual sebocytes had bipotential differentiation capacity, and if so, at what frequency, we seeded SZ95 cells at a range of clonal densities on a feeder layer of J2–3T3 cells. 14 days later we stained the dishes for involucrin or, as a negative control, CD8 (Fig. 2N, 2O). SZ95 cells had a colony forming efficiency of approximately 33%, which was higher than the 10% colony forming efficiency of human interfollicular epidermal keratinocytes in these experiments. In addition, SZ95 clones tended to be smaller and more homogeneous than those formed by primary interfollicular epidermal keratinocytes (Fig. 2N, 2O and data not shown). We scored a minimum of 500 clones in triplicate experiments and found that every SZ95 clone contained involucrin positive cells. Whereas in clones of IFE keratinocytes the involucrin positive cells accumulate as suprabasal layers in the center of the colony [45], in SZ95 cells the involucrin positive cells tended to lie as scattered cells at the clone periphery (Fig. 2O).
Since involucrin and cornifin are markers both of IFE and HF inner root sheath, we investigated whether SZ95 cells could form IFE and HF in xenografts. We injected SZ95 cells into nude mice in combination with neonatal mouse dermal fibroblasts (Fig. 3), conditions that support differentiation of primary IFE keratinocytes into cysts containing IFE, SG and HF [6]. The cysts that formed contained both differentiated sebocytes and cornified cells, which could be distinguished by hematoxylin and eosin staining (Fig. 3A, 3B). Consistent with this interpretation, granular cells were identified adjacent to the cornified layers (Fig. 3B), and large cells with accumulated lipid (stained with Nile Red) were detected in other regions of the xenografts (Fig. 3C). Some cells within the xenografts expressed K7, indicative of commitment to the sebaceous gland lineage (Fig. 3E), whereas others expressed cornifin (Fig. 3D). Examination of sections double labeled with antibodies to K7 and cornifin confirmed that these markers were not co-expressed (Fig. 3D–3F). We saw no evidence for hair follicle formation in the SZ95 xenografts. We conclude that SZ95 cells are capable of undergoing terminal differentiation into interfollicular epidermal keratinocytes or into mature sebocytes, although we cannot rule out the possibility that, with an appropriate inductive stimulus, they could also form some or all of the HF lineages.
|
|
To investigate whether Myc overexpression had any effect on lineage selection, SZ95 cells were transduced with a retrovirus encoding MycER [28, 33]. Following two days of 4OHT treatment to activate Myc, cells transduced with MycER showed an increase in the number and size of cytoplasmic lipid droplets, whereas cells that were not exposed to 4OHT or were transduced with the empty retroviral vector (EV) did not (Fig. 5A, 5B and data not shown).
|
In order to investigate the consequences of activating β-catenin, SZ95 cells were transduced with a retroviral vector that expressed stabilized, N-terminally truncated, β-catenin (Nβ-catenin) [34]. Expression of both involucrin (Fig. 5E, 5F) and cornifin (Fig. 5G, 5H) increased in Nβ-catenin expressing SZ95 cells. In β-catenin transduced cultures involucrin and cornifin bright cells were more numerous, appeared to be either suprabasal or stretching along the edges of the colony, and showed an elongated morphology (Fig. 5F, 5H; compare Fig. 2O).
To quantitate these observations five different microscopic fields per condition were analyzed using the same confocal detection settings for EV, MycER, and Nβ-catenin transduced cells. Cells with a staining intensity higher than a set threshold were scored as positive. As shown in Figure 5I–5K, Myc activation led to an increase in cells with accumulated lipid droplets, and to a decrease in the number of cells expressing involucrin or cornifin. Nβ-catenin expression stimulated expression of both interfollicular epidermal differentiation markers (Fig. 5J, 5K), without affecting the proportion of cells that accumulated lipid droplets (Fig. 5I).
The differential effects of Myc and β-catenin on sebocyte differentiation were confirmed by monitoring the behavior of EV, Nβ-catenin and 4OHT treated MycER SZ95 cells using time-lapse videomicroscopy for 48 hours (Fig. 5L–5P). Cells were seeded at low density (approximately 50 cells/mm2), to ensure that the starting population consisted mainly of single cells, and then monitored for 48h. MycER cells were most efficient at forming sebaceous gland-like colonies (identified as containing cells that accumulated large lipid vacuoles and subsequently burst), whereasNβ-catenin cells were the least efficient (Fig. 5L–5N). Nβ-catenin cells also showed less tendency to form cohesive clusters than MycER and EV cells (Fig. 5L–5N). When the speed of individual cells was determined, MycER cells were found to move significantly more slowly than EV cells (Fig. 5O), consistent with the finding that Myc reduces motility of interfollicular epidermal keratinocytes [33]. Nβ-catenin cells had a similar average speed to EV transduced cells, but the range was greater, suggesting that β-catenin can stimulate SZ95 cell motility (Fig. 5P).
Antagonism Between β-Catenin and Myc in Regulating Sebocyte Differentiation In Vivo
The effects of Myc and β-catenin on differentiation of SZ95 cells are consistent with our previous observations of the consequences of activating each protein in transgenic mouse epidermis. Activation of Myc by 4OHT treatment of K14MycER transgenic mice results in enlargement of the sebaceous glands and ectopic sebocyte differentiation in the hair follicles and interfollicular epidermis [14, 16]. Conversely, prolonged activation of β-catenin, by 4OHT treatment of K14Nβ-cateninER transgenics, leads to loss of sebaceous glands and formation of ectopic hair follicles [20, 21].
To determine the epidermal response to simultaneous activation of Myc and β-catenin, we crossed K14MycER and K14Nβ-cateninER transgenic mice and applied 4OHT topically to the skin. As previously reported, activation of MycER alone resulted in an increase in thickness of the IFE and enlargement of sebaceous glands (Fig. 6A, 6B), whereas β-catenin activation led to induction of ectopic follicles from IFE and SG (Fig. 6C). Coexpression and relative abundance of the transgenes were confirmed by Western blotting protein lysates of total skin (Fig. 6D).
|
Nβ-cateninER transgenics, but the cells at the periphery of the sebaceous gland were small and proliferative, consistent with the observation that this is a common site of ectopic hair follicle formation (Fig. 6K) [21]. In double transgenics the sebaceous gland phenotype was intermediate (Fig. 6L), with regions of larger, more differentiated sebocytes (arrow) interspersed with small, proliferative cells (arrowhead). Thus, β-catenin-mediated induction of ectopic hair follicles in interfollicular epidermis and sebaceous glands was inhibited by overexpression of Myc, whereas β-catenin blocked Myc induced sebocyte differentiation. Activation of Myc or β-catenin individually results in increased epidermal proliferation, evaluated either by Ki67 labeling or incorporation of a 1-hour pulse of BrdU [16, 20] (Fig. 6M–6O, 6Q). Whereas Myc activation primarily increases proliferation in the interfollicular epidermis (Fig. 6N; [14, 16]), β-catenin induces proliferation locally at sites of ectopic follicle formation (Fig. 6O) [21]. Quantitation of the number of BrdU positive cells per microscopic field in tail whole mounts revealed that the increase in proliferation induced by β-catenin was greater than that induced by Myc (Fig. 6Q). In double transgenics proliferation was increased relative to wild-type epidermis but the total number of BrdU positive cells was intermediate between Myc and β-catenin single transgenics (Fig. 6P, 6Q). Our results demonstrate that, when co-expressed in the basal layer of mouse epidermis, Myc and β-catenin have mutually antagonistic effects on both lineage specification and proliferation.
Indian Hedgehog Is a Myc Target Gene but Is Not Sufficient for Myc Induced Lineage Selection
One of the pathways activated downstream of β-catenin and required for proliferation during ectopic hair follicle formation is Shh [18, 20, 21]. We have previously reported that Ihh is upregulated in sebaceous glands, sebaceous tumors and differentiated SZ95 cells and that cyclopamine inhibits the growth of SZ95 cells in culture [29, 46]. We screened the human Ihh gene for Myc-binding sites and found two conserved E-box elements located in the first intron, approximately 1,000 bp apart (gccCACGTGtct and ctcCACGTGagg). We performed chromatin immunoprecipitation on primary human keratinocytes transduced with MycER or a mutant that lacks the Myc transactivation domain (Myc106ER; [28]). We were able to amplify the Ihh promoter PCR fragment in infected keratinocytes when precipitated with a Myc antibody but not when precipitated with an isotype matched antibody (Lef1) or with the secondary antibody only (Fig. 7A). We conclude that Ihh is a Myc target gene. However, in SZ95 cells transduced with a retroviral vector encoding Ihh we did not see any effect on the proportion of cells expressing either interfollicular epidermal markers (cornifin, involucrin) or lipid droplets (assessed by Nile Red staining) (Fig. 7B). This suggests that specification of the sebaceous gland lineage by Myc is not mediated by induction of Ihh.
|
|
DISCUSSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionDisclosure of Potential...AcknowledgmentsReferences |
|---|
Of the different populations of epidermal stem cells, that of the sebaceous gland is least well-characterized, although recent evidence suggests the existence of unipotent Blimp1 positive cells that can be replenished by stem cells in the hair follicle bulge [12]. We found that in human epidermis Blimp1 was not selectively expressed in SG progenitor cells, but was also expressed by terminally differentiating cells in the IFE, SG, and HF, in good agreement with a recent study [44]. Thus Blimp1 is not a specific marker of SG stem or progenitor cells in human skin. While Blimp1 binds to and negatively regulates the Myc promoter [12, 44], there was no correlation between Blimp1 and Myc levels in individual human SZ95 cells, suggesting that additional factors regulate levels of Myc protein in sebocytes.
Our most striking finding was that clonogenic SZ95 cells not only generated cells that underwent SG differentiation but also cells that expressed involucrin and cornifin, differentiation markers expressed in the IFE and HF inner root sheath. On injection into nude mice SZ95 cells underwent SG and IFE differentiation but did not form HF, leading us to conclude that they are bipotent rather than multipotent. Nevertheless, we cannot exclude the possibility that in response to additional stimuli SZ95 cells may be able to give rise to cells of the HF lineages. Indeed, given the physical proximity of the involucrin positive cells of the inner root sheath to mature sebocytes in the sebaceous gland (Fig. 1K), it is tempting to speculate that this is indeed the case.
Evidence that sebaceous gland cells not only differentiate into mature sebocytes but also differentiate into involucrin positive cells was also obtained in a second human sebocyte line, Seb-E6E7, which were immortalized with HPV16 E6 and E7. This strongly suggests that the differentiation properties of SZ95 cells are not an indirect consequence of immortalization with SV40.
The bipotential differentiation capacity of SZ95 and SebE6-E7 cells provides an experimental model with which to study the factors that regulate lineage selection in human epidermis at the single cell level. While more work needs to be done to establish conclusively that such cells exist in normal skin, the finding that Myc activation results in the appearance of clusters of sebocytes in mouse IFE [16] is certainly consistent with their existence, as is the observation that hedgehog signaling can trigger sebaceous gland differentiation in IFE [10]. The location of involucrin positive cells at the periphery of SZ95 clones raises the alternative possibilities that differentiating sebocytes and IFE keratinocytes differ in their adhesive properties [49] or that there is a community effect whereby differentiating sebocytes promote the same differentiation pathway in their neighbors [50].
Retroviral transduction of SZ95 cells with activated Myc resulted in increased lipid accumulation while activated β-catenin stimulated expression of involucrin and cornifin. In vivo, Myc stimulates differentiation of both the IFE and SG lineages ([14, 15, 33, 51, 52], while prolonged activation of β-catenin results in loss of SG and ectopic HF formation in all epidermal sites including the SG [20, 21]. We observed that involucrin-positive SZ95 cells tended to have lower endogenous Myc levels than cells that underwent sebocyte differentiation (Fig. 4), raising the possibility that different levels of Myc promote different lineages. Since SZ95 cells did not form HF in nude mice (Fig. 3), β-catenin may promote IFE differentiation because HF lineage selection is blocked in SZ95 cells.
In vivo, there was mutual antagonism between Myc and β-catenin in cells coexpressing both transgenes. Myc blocked ectopic HF formation while β-catenin reduced accumulation of terminally differentiated sebocytes. In intestinal epithelium Myc mediates proliferation in response to Wnt activation [25]. However, in the epidermis the populations that proliferate in response to Myc and β-catenin are different. Whereas β-catenin induces local proliferation at sites of ectopic HF formation [21], Myc induced proliferation is predominantly in the IFE [14–16] (Fig. 6).
Shh mediates β-catenin induced proliferation during anagen [18] and inhibition with cyclopamine blocks ectopic HF formation [21]. Ihh mRNA is upregulated in SZ95 cells in response to N-terminal Lef1 mutations that block β-catenin signaling [46] and is also a direct Myc target gene (Fig. 7; D. Roop, personal communication). Ihh has been detected by immunohistochemical staining of mouse SG and sebaceous tumors [29]. The reduction in Myc induced proliferation by cyclopamine would be consistent with Myc mediated upregulation of Ihh, although we have been unable to detect Ihh in normal mouse epidermis by in situ hybridization (data not shown). Cyclopamine has previously been reported to not only reduce SZ95 proliferation, but also to stimulate SG differentiation [23]. However, overexpression of Ihh did not affect the proportion of SZ95 cells that underwent SG or IFE differentiation. From these studies we propose that Myc dependent upregulation of Ihh contributes to Myc induced proliferation, but not lineage selection.
While we have yet to define the mechanisms by which Myc and β-catenin exert different effects on lineage selection by SZ95 cells, it is interesting to note that they had different effects on SZ95 cell cohesion and motility. Cells transduced with Myc tended to cluster and showed reduced motility, while cells transduced with β-catenin tended to remain as single cells. In keratinocytes and SZ95 cells β-catenin does not increase average cell speed (Fig. 5) [34], but nevertheless some SZ95 cells showed increased motility when overexpressing β-catenin. These observations raise the possibility that Myc inhibits β-catenin induced hair follicle formation by inhibiting the cell movements associated with hair growth [5].
|
CONCLUSION |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionDisclosure of Potential...AcknowledgmentsReferences |
|---|
|
DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionDisclosure of Potential...AcknowledgmentsReferences |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionDisclosure of Potential...AcknowledgmentsReferences |
|---|
|
REFERENCES |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionConclusionDisclosure of Potential...AcknowledgmentsReferences |
|---|
and β in squamous differentiating epithelial tissues and several skin lesions. J Invest Dermatol 1997;108:200–204.[CrossRef][Medline]
