An effector of intercellular adhesion, beta-catenin also functions in Wnt signaling, associating with Lef-1/Tcf DNA-binding proteins to form a transcription factor. We report that this pathway operates in keratinocytes and that mice expressing a stabilized beta-catenin controlled by an epidermal promoter undergo a process resembling de novo hair morphogenesis. The new follicles formed sebaceous glands and dermal papilla, normally established only in embryogenesis. As in embryologically initiated hair germs, transgenic follicles induce Lef-1, but follicles are disoriented and defective in sonic hedgehog polarization. Additionally, proliferation continues unchecked, resulting in two types of tumors also found in humans. Our findings suggest that transient beta-catenin stabilization may be a key player in the long-sought epidermal signal leading to hair development and implicate aberrant beta-catenin activation in hair tumors.

Gene knockout technology has provided a powerful tool for functional analyses of genes expressed preferentially in a particular tissue. Given marked similarities between human and mouse skin, such studies with epidermally expressed genes have often provided valuable insights into human genetic skin disorders. Efficient silencing of a specified gene in a temporally regulated and epidermal-specific fashion could extend functional analyses to broadly expressed genes and increase the categories of human skin disorders to which parallels could be drawn. We have generated transgenic mice expressing Cre and a fusion protein between Cre recombinase and the tamoxifen responsive hormone-binding domain of the estrogen receptor (CreER tam ) under the control of the human keratin 14 (K14) promoter. This promoter is strongly active in dividing cells of epidermis and some other stratified squamous epithelia. With K14-Cre, transgenic embryos recombine genetically introduced loxP sequences efficiently and selectively in the genomes of keratinocytes that reside in embryonic day 14.5 skin, tongue, and esophagus. With K14-CreER tam , postnatal transgenic mice show no Cre activity until tamoxifen is administered. If orally administered, tamoxifen activates keratinocyte-specific CreER tam , allowing recombination of loxP sequences in epidermis, tongue, and esophagus. If topically administered, tamoxifen allows recombination in the area of skin where tamoxifen was applied. Finally, we show that epidermal cells harboring a Cre-dependent rearranged genome persist for many months after tamoxifen application, indicating that the epidermal stem cell population has been targeted efficiently. These tools now pave the way for testing the functional role of different somatic mutations that may exist in mosaic disorders of the skin, including squamous and basal cell carcinomas.

BPAG1 is the major antigenic determinant of autoimmune sera of bullous pemphigoid (BP) patients. It is made by stratified squamous epithelia, where it localizes to the inner surface of specialized integrin-mediated adherens junctions (hemidesmosomes). To explore the function of BPAG1 and its relation to BP, we targeted the removal of the BPAG1 gene in mice. Hemidesmosomes are otherwise normal, but they lack the inner plate and have no cytoskeleton attached. Though not affecting cell growth or substratum adhesion, this compromises mechanical integrity and influences migration. Unexpectedly, the mice also develop severe dystonia and sensory nerve degeneration typical of dystonia musculorum (dt/dt) mice. We show that in at least one other strain of dt/dt mice, BPAG1 gene is defective.

When surface epithelium was conditionally targeted for ablation of alpha-catenin, hair follicle development was blocked and epidermal morphogenesis was dramatically affected, with defects in adherens junction formation, intercellular adhesion, and epithelial polarity. Differentiation occurred, but epidermis displayed hyperproliferation, suprabasal mitoses, and multinucleated cells. In vitro, alpha-catenin null keratinocytes were poorly contact inhibited and grew rapidly. These differences were not dependent upon intercellular adhesion and were in marked contrast to keratinocytes conditionally null for another essential intercellular adhesion protein, desmoplakin (DP). KO keratinocytes exhibited sustained activation of the Ras-MAPK cascade due to aberrations in growth factor responses. Thus, remarkably, features of precancerous lesions often attributed to defects in cell cycle regulatory genes can be generated by compromising the function of alpha-catenin.

Keratinocyte growth factor (KGF), also known as fibroblast growth factor 7 (FGF7), is synthesized by skin fibroblasts. However, its mitogenic activity is on skin keratinocytes, where it is the most potent growth factor identified thus far. To explore KGF's function in vivo, we used embryonic stem cell technology to generate mice lacking KGF. Over time, their fur developed a matted appearance, very similar to that of the rough mouse, whose recessive mutation maps at or near the KGF locus on mouse chromosome 2. In contrast to the recently reported transforming growth factor-~ (TGF-~) and FGF5 knockouts, which showed defects in the follicle outer-root sheath and the hair growth cycle, respectively, the hair defect in the KGF knockout seemed to be restricted to the cells giving rise to the hair shaft. Thus, we have uncovered a third, and at least partially nonoverlapping, growth factor pathway involved in orchestrating hair follicle growth and/or differentiation. Surprisingly, the absence of KGF resulted in no abnormalities in epidermal growth or wound healing. This was true even when we engineered double knockout mice, null for both KGF and TGF-a, two factors that are increased dramatically in the normal wound-healing process. Whereas we found no evidence of compensatory changes at the mRNA level of wounded knockout mice, these data imply that the regulation of epidermal growth is complex and involves a number of growth stimulatory factors that go beyond what are thought to be the major paracrine and autocrine growth factors. We suggest that the redundancy in epidermal growth and wound healing is likely to stem from the vitality of these functions to the organism, a feature that is not a consideration for the hair follicle.

We have generated an epidermis-specific desmoplakin (DP) mouse knockout, and show that epidermal integrity requires DP; mechanical stresses to DP-null skin cause intercellular separations. The number of epidermal desmosomes in DP-null skin is similar to wild type (WT), but they lack keratin filaments, which compromise their function. DP-null keratinocytes have few desmosomes in vitro, and are unable to undergo actin reorganization and membrane sealing during epithelial sheet formation. Adherens junctions are also reduced. In vitro, DP transgene expression rescues these defects. DP is therefore required for assembly of functional desmosomes, maintaining cytoskeletal architecture and reinforcing membrane attachments essential for stable intercellular adhesion.

Abstract. Keratin 5 and keratin 14 have been touted as the hallmarks of the basal keratin networks of all stratified squamous epithelia. Absence of K14 gives rise to epidermolysis bullosa simplex, a human blistering skin disorder involving cytolysis in the basal layer of epidermis. To address the puzzling question of why this disease is primarily manifested in skin rather than other stratified squamous epithelia, we ablated the K14 gene in mice and examined various tissues expressing this gene. We show that a key factor is the presence of another keratin, K15, which was hitherto unappreciated as a basal cell component. We show that the levels of K15 relative to K14 vary dramatically among stratified squamous epithelial tissues, and with neonatal development. In the absence of K14, K15 makes a bona fide, but ultrastructurally distinct, keratin filament network with K5. In the epidermis of neonatal mutant mice, K15 levels are low and do not compensate for the loss of K14. In contrast, the esophagus is unaffected in the neonatal mutant mice, but does appear to be fragile in the adult. Parallel to this phenomenon is that esophageal K14 is expressed at extremely low levels in the neonate, but rises in postnatal development. Finally, despite previous conclusions that the formation of suprabasal keratin filaments might depend upon K5/K14, we find that a wide variety of suprabasal networks composed of different keratins can form in the absence of K14 in the basal layer.

Desmosomes first assemble in the E3.5 mouse trophectoderm, concomitant with establishment of epithelial polarity and appearance of a blastocoel cavity. Throughout development, they increase in size and number and are especially abundant in epidermis and heart muscle. Desmosomes mediate cell–cell adhesion through desmosomal cadherins, which differ from classical cadherins in their attachments to intermediate filaments (IFs), rather than actin filaments. Of the proteins implicated in making this IF connection, only desmoplakin (DP) is both exclusive to and ubiquitous among desmosomes. To explore its function and importance to tissue integrity, we ablated the desmoplakin gene. Homozygous −/− mutant embryos proceeded through implantation, but did not survive beyond E6.5. Mutant embryos proceeded through implantation, but did not survive beyond E6.5. Surprisingly, analysis of these embryos revealed a critical role for desmoplakin not only in anchoring IFs to desmosomes, but also in desmosome assembly and/or stabilization. This finding not only unveiled a new function for desmoplakin, but also provided the first opportunity to explore desmosome function during embryogenesis. While a blastocoel cavity formed and epithelial cell polarity was at least partially established in the DP (−/−) embryos, the paucity of desmosomal cell–cell junctions severely affected the modeling of tissue architecture and shaping of the early embryo.