If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Cutaneous Biology Research Center, Massachusetts General Hospital and Harvard Medical School, Charlestown, Massachusetts, USADepartment of Dermatology, Tohoku University Graduate School of Medicine, Sendai, Japan
Cutaneous Biology Research Center, Massachusetts General Hospital and Harvard Medical School, Charlestown, Massachusetts, USAInstitute of Biochemistry, Lausanne University, Epalinges, Switzerland
We present here a dynamic model of functional equilibrium between keratinocyte stem cells, transit amplifying populations and cells that are reversibly versus irreversibly committed to differentiation. According to this model, the size of keratinocyte stem cell populations can be controlled at multiple levels, including relative late steps in the sequence of events leading to terminal differentiation and by the influences of a heterogeneous extra-cellular environment. We discuss how work in our laboratory, on the interconnection between the cyclin/CDK inhibitor p21WAF1/Cip1 and the Notch1 signaling pathways, provides strong support to this dynamic model of stem cell versus committed and/or differentiated keratinocyte populations.
Emerging evidence indicates that the balance between epithelial cell growth and differentiation involves a complex interplay between classical cell regulatory pathways and development-related signals. Keratinocytes provide an ideal experimental system to dissect the extra and intracellular signaling pathways involved in growth/differentiation control of epithelial cells, under normal conditions and during tumor development (
). Within the proliferative compartment of the epidermis, at least two kinds of keratinocyte populations are thought to exist: multipotent “stem cells”, slow cycling but with an indefinite self renewal potential and capable of generating all other types of growing and differentiating keratinocytes; and “transient amplifying cells”, actively proliferating but capable of a limited number of cell divisions and already committed towards differentiation (
An attractive model, depicted in Figure 1, is that there is a dynamic equilibrium between keratinocyte stem cells, transit amplifying populations and cells that are reversibly versus irreversibly committed to differentiation. According to this model, the size of stem cell populations is likely to be influenced by the number of surrounding transit amplifying cells that are in turn in equilibrium with cells that have withdrawn reversibly versus irreversibly from the cell cycle. The rate of conversion from one cell population to the next is likely to be further controlled by the surrounding cellular environment, i.e. the stroma and associated blood supply. Such cellular environment is likely to be non-homogeneously distributed, but organized along discontinuous signaling centers that are more (niches) or less permissive for stem cell populations. It is therefore likely that the number of keratinocyte stem cells can be controlled through mechanisms acting at multiple levels, including (A) relative late steps in the sequence of events leading to terminal differentiation; (B) influences of the surrounding cellular and extracellular environment. Here we will discuss how work in our laboratory, on the interconnection between the cyclin/CDK inhibitor p21WAF1/Cip1 and the Notch1 signaling pathway, which provides strong support to the dynamic model illustrated above.
Figure 1Model of a dynamic equilibrium between keratinocyte populations with different stem cell potential versus commitment to differentiation.
). This protein belongs to the Cip/Kip family of CKI (p21WAF1/Cip1, p27Kip1, p57Kip2), which share significant sequence homology in their amino-terminal portions and recognize a broad, but not identical range of cyclin/CDK targets (
). The amino terminal domain of p21, like the corresponding domains of p27 or p57, is both necessary and sufficient to inhibit cyclin/CDK activity in vitro and in vivo. The unique carboxy-terminal domain of p21 associates with the proliferating nuclear antigen (PCNA), a subunit of DNA polymerase δ and can inhibit DNA replication directly, without affecting DNA repair (
), besides associating with cyclin/CDKs and PCNA, p21 participates in a number of other specific protein–protein interactions. Some of these interactions still bear on cell cycle control, whereas others point to separate functions of p21, as a modulator of apoptosis or transcription.
In keratinocytes, like in other cell types, increased p21 expression plays an important role in cell cycle withdrawal which is associated with differentiation (
). p21, however, is not absolutely required for this process, as other cell cycle regulatory events are associated with the differentiation process that can compensate for lack of p21 function (
). In contrast to the transition between growth and growth arrest of these “transit amplifying” populations, we have found that p21 is likely to play an essential role in restricting the number of keratinocytes with high growth/differentiation potential (putative stem cells). Several lines of evidence support this conclusion.
Like in other organ systems, keratinocyte stem cells are only a small fraction of cells in the epidermis, with no unequivocal biochemical markers, so that their analysis is ultimately dependent on functional rather than biochemical assays. Stem cells should have two fundamental properties: (1) an indefinite life span, and (2) the capability to give rise to all other types of partially committed and differentiated cells. The existence of keratinocyte subpopulations with increased growth potential, capable of generating rapidly expanding colonies in culture has been well established (
). Primary keratinocytes derived from p21-/- mice undergo growth in arrest in response to increased extracellular calcium to a same extent as wild type cells (
). When tested in a growth clonogenicity assays, however, p21-/- keratinocyte cultures contain a much higher fraction of keratinocytes with clonogenic properties than their wild type counterpart (0.1% of total proliferating cells, vs <0.005%) (
Keratinocytes in the interfollicular epidermis undergo a vertical differentiation program along a single lineage. By contrast, keratinocyte differentiation in hair follicles proceeds along a minimum of six distinct lineages (
). To take into account this complex program of differentiation, we developed an in vivo skin/hair reconstitution assay with genetically labeled cells. The interfollicular epidermis reconstituted in these assays was composed of well-demarcated columnar units likely to originate from single progenitor cells. By contrast, reconstituted hair follicles were composed of cells of multiple origin, consistent with distinct progenitor cells for the shaft, inner (IRS) and outer root sheath (ORS), respectively (
), in our combined in vitro/in vivo assay with cultured wild-type keratinocytes, very few if any multipotent precursor cells capable to generate on their own entire hair follicles could be detected (
). By contrast, the presence of such multipotent progenitors forming entire hair follicles could be readily demonstrated by similar assays with primary keratinocytes derived from p21-/- mice (
Figure 2Hair follicle reconstitution potential of p21+/+versus p21-/- primary keratinocytes. Summary of the results obtained by skin/hair reconstitution assays with primary keratinocytes genetically labeled by infection with a retrovirus transducing the alkaline phosphatase (AP) marker gene (
). Experiments were performed with primary keratinocytes derived from either p21+/+ or p21-/- mice of the same genetic background. In each case, mixed populations of approximately 10% infected and 90% uninfected keratinocytes were grafted together with hair-inducing dermal papilla cells. Mice were euthanized 6 weeks after grafting (with reconstituted hair follicles having undergone three hair cycles), and hair follicles in cross section were counted for numbers that consisted of homogeneously AP-positive cells (consistent with being derived from single progenitors) versus numbers of follicles composed of both AP-positive and AP-negative populations (likely to be derived from multiple progenitors).
Exposure of primary keratinocytes to increased extracellular calcium concentrations induces a differentiation program similar to that occurring in vivo, including establishment of adherens junctions and desmosomes, induction of several enzymatic and structural markers of the upper epidermal layers and exit from the cell cycle (
). Importantly, establishment of this differentiation program can be reversed at early times (within 24 h) by simple removal of the calcium pro-differentiation stimulus, and only at later times (2–3 d and beyond) differentiation becomes irreversible. p21 appears also to be a critical determinant of the irreversibility of differentiation, as a significant fraction of primary keratinocytes from p21-/- mice (15%–30%), but not wild-type cells, can resume proliferation even after 3–6 d of exposure to elevated extracellular calcium (
Thus, although not essential for exit of the vast majority of “transit amplifying” keratinocytes from the cell cycle, p21 plays a key role in restricting the number of keratinocyte stem cell populations (as determined by clonogenicity and skin/hair reconstitution assays) as well as, further downstream, determining the irreversibility of differentiation. Thus, p21 may exert these functions by acting at two distinct levels (increasing k1/k2 as well as k5 in Figure 1). On the basis of our dynamic model, however, the same balance of keratinocyte stem cells versus committed populations could be achieved if p21 were to act at a single level (i.e. by simply increasing k5 in Figure 1).
Notch1, a direct upstream regulator of p21 expression, plays a key role in the coordinate control of keratinocyte self renewal and differentiation
The Notch gene family encodes evolutionarily conserved transmembrane receptors with a key role in cell-fate determination and differentiation (
). Notch activation is triggered by interactions with ligands of the Delta and Serrate/Jagged families, and results in release of the intracellular region of the receptor by proteolytic cleavage and nuclear translocation (
). The activated Notch intracellular region binds to a ubiquitous DNA-binding protein of the CSL family (Drosophila Suppressor of hairless (Su(H)) or its mammalian homolog RBP-Jκ/CBF1), converting it from a repressor into an activator of transcription (
). Notch activation results in the induction of a program of gene expression that can either suppress or promote differentiation in a cell type- and context-dependent manner (
Most work has focused on the ability of Notch to direct divergent pathways of differentiation among initially equipotent cells (Figure 3a). One-way Notch and ligands such as Delta act to promote divergent differentiation is termed lateral inhibition, a mechanism first described and characterized in the developing neuroectoderm of Drosophila (
). Notch and its ligands, however, can also be co-expressed in the same cells undergoing coordinate differentiation to a single fate (Figure 3b). Examples of such cells include those in developing mammalian muscle and skin, in which the Serrate-like ligands Jagged 1 and 2 are co-expressed with Notch (
). In this context, the Notch-dependent upregulation of Jagged1 expression has been proposed to function as a positive feedback mechanism, similar to that reported for Drosophila (
). The relative contribution of negative and positive feedback mechanisms for modulation of Notch signaling appears to be further controlled by ancillary proteins, such as members of the Fringe family that selectively inhibit Notch responsiveness to Serrate-like ligands (
Serrate-mediated activation of Notch is specifically blocked by the product of the gene fringe in the dorsal compartment of the Drosophila wing imaginal disc.
Figure 3Notch signaling in determination of cell fate and stem cell potential. As discussed in the text, there are two possible modes of action of Notch signaling, lateral inhibition (A) and positive reinforcement (B), that are determined by either a negative or positive feedback loop between expression of Notch (N) and its ligands, Delta (D) or Jagged (J).
A variety of data suggest that all these mechanisms for regulation of Notch activity operate in mammalian skin. A complex pattern of expression of Notch and its ligands is seen in hair follicles and feathers, consistent with their multiple keratinocyte cell lineages (
). A role for Notch in keratinocyte skin appendages formation and/or differentiation is supported by ectopic expression of Notch or its ligands, with Notch activation in one layer of the hair follicle modulating differentiation and/or lineage commitment in neighboring layers (
). In human interfollicular epidermis, localized expression of Delta in putative “stem cells” has been proposed to induce commitment of neighboring Notch1-expressing cells to a “transit amplifying” phenotype, a scenario reminiscent of lateral inhibition (
), we have found that in newborn mouse epidermis high expression of Jagged 1 and 2 coincide with Notch 1 and 2 in differentiating keratinocytes of the supra-basal layers, with both Jagged 1/2 and Notch1 expression being turned off in the terminally differentiated outermost layers (Rangarajan et al, 2001b). In mouse epidermis around birth, all three forms of Fringe are expressed (
). Of note, the spinous layer is the only region of the epidermis free of Fringe expression, thereby potentially creating a Notch signaling boundary between this layer and cells of the immediately underlying (basal) and overlying (granular) layers (
An essential role of Notch1 in epidermal border formation was demonstrated by the analysis of mice with a conditional keratinocyte-specific deletion of this gene. Loss of Notch1 results in an hyperplastic phenotype of the interfollicular epidermis, with inappropriate co-expression of keratins of the basal and upper epidermal layers, and a striking upregulation of integrin β1 and β4 expression in the supra-basal epidermal layers, where expression of these integrins is usually low or undetectable (
). Additionally, endogenous Notch1 activity is required for induction of p21 expression in differentiation. In fact, in mouse keratinocytes, the p21 gene is a direct transcriptional target of Notch1 activation. p21 expression is induced in these cells by activated Notch1 through RBP-Jκ-dependent transcription, and the RBP-Jκ protein binds directly to the endogenous p21 promoter (
). Importantly, Notch1 activity itself is induced in parallel with p21 expression in the transition between transit amplifying and differentiating cells (Figure 4a). In turn, increased Notch1 activity can also induce differentiation, through a mechanism likely to involve down-regulation of integrin α6/β4 and α3/β1 expression (
) (Figure 4b). Thus, Notch1 can function both upstream and in parallel with p21 in regulating the balance between keratinocyte stem cells and more committed populations. Like p21, Notch1 may exert this function acting at distinct multiple levels (increasing k1/k2 as well as k5 in Figure 1), but the same end result could be achieved if Notch1, like p21, were to act at a single level (i.e. only increasing k5).
Figure 4Notch1 signaling at the center of the keratinocyte differentiation program. As discussed in the text, Notch1 signaling functions as a direct upstream regulator of p21WAF1/Cip1 expression in mouse keratinocyte differentiation. Notch1 signaling itself is induced with keratinocyte differentiation (A) and, in turn, promotes the exit from the cell cycle and differentiation, through a mechanism, which includes down-modulation of integrin expression (B).
). Consistent with an increased size of stem cell populations, mice with a disruption of the p21 gene exhibit a substantially enhanced susceptibility to chemically induced skin carcinogenesis, with an increase numbers of tumors that are either malignant or benign, depending on protocol and genetic background (
). Strikingly, as for mice lacking the p21WAF1/Cip1 gene, mice with induced deletion of Notch1 exhibit a substantially increased susceptibility to chemically-induced skin carcinogenesis and, as with p21-/- cells, Notch1-/- keratinocytes infected with a retrovirus transducing the ras oncogene and injected subcutaneously into nude mice form aggressive squamous cell carcinomas, whereas Notch1+/+ cells do not (
Although p21 is one of the important downstream targets of Notch function in keratinocytes, Notch signaling is likely to affect other signaling pathways with a key regulatory function in this cell type. In fact, mice with an induced deletion of the Notch1 gene differ in other aspects of their phenotype from p21-/- animals. In particular, mice with the Notch1 deletion develop after 3 mo of age extensive hyperplasia and keratinization of the corneal epithelium that results in opaque plaque formation and blindness. Concurrently, they develop also spontaneous skin tumors in various parts of the body that are highly vascularized and consist of small proliferating basophilic cells similar to those in basal cell carcinomas (BCC) (
). In both hyperplastic epidermis and BCC-like tumors, Notch1 deficiency is associated with increased and sustained expression of Gli2, a downstream target of the sonic hedgehog (SHH) signaling pathway that has been causally linked to BCC tumor formation (
). Additionally, the Wnt/β-catenin pathway appears to be elevated in keratinocytes and tumors as a consequence of loss of Notch1 function, whereas β-catenin signaling can be suppressed by activated Notch1 expression (
Specific down-modulation of Notch1 signaling in cervical cancer cells is required for sustained HPV-E6/E7 expression and late steps of malignant transformation.
) in keratinocytes with elevated Notch1 activity. Importantly, inhibition of the latter pathway has been recently connected with a growth inhibitory function of Notch1 in an important type of human keratinocyte-derived tumors: HPV-induced cervical carcinomas (
Specific down-modulation of Notch1 signaling in cervical cancer cells is required for sustained HPV-E6/E7 expression and late steps of malignant transformation.
Thus, an exciting area of study for the future is the close interconnection among classical cell cycle and development-related pathways in control of keratinocyte stem cell potential, differentiation and tumorigenesis.
REFERENCES
Artavanis-Tsakonas S.
Rand M.D.
Lake R.J.
Notch signaling: Cell fate control and signal integration in development.
Serrate-mediated activation of Notch is specifically blocked by the product of the gene fringe in the dorsal compartment of the Drosophila wing imaginal disc.
Specific down-modulation of Notch1 signaling in cervical cancer cells is required for sustained HPV-E6/E7 expression and late steps of malignant transformation.
53015-7&id=gr1.jpg){kind=link}
53015-7&id=gr2.jpg){kind=link}
53015-7&id=gr3.jpg){kind=link}
53015-7&id=gr4.jpg){kind=link}