Characterization of an inducible, epidermal‐specific knockout system: Differential expression of lacZ in different Cre reporter mouse strains

Abstract

To generate epidermal-specific knockout mice, vectors can be used to target expression of Cre recombinase to epidermal stem cells. Cre-mediated recombination is then used to delete gene sequences flanked by loxP sites (“floxed gene”). However, most keratin-based expression vectors (e.g., K14, which direct transgene expression to epidermal stem cells), are also expressed in other epithelial tissues. Furthermore, transgene expression can occur as early as embryonic day 9.5 (Bickenbach et al., 1995). Therefore, constitutive expression of Cre recombinase with these vectors may induce embryonic lethality as a result of deletion of the gene in other epithelial tissues. In addition, even if it was possible to restrict the deletion of the gene to the epidermis, neonatal lethality may still occur if loss of expression of this gene has a severe effect on development of the epidermis (e.g., the p63 gene; Mills et al., 1999). To circumvent this problem, we have developed an inducible, epidermal-specific knockout system. This system utilizes epidermal-specific expression of a Cre recombinase, which is fused with a truncated progesterone receptor (PR1). The CrePR1 fusion protein is bound and paradoxically activated by progesterone antagonists, such as RU486 or its analogue ZK98.734. The Cre recombinase activity in this fusion protein is sequestered in the cytoplasm until a progesterone antagonist induces translocation of the fusion protein to the nucleus (Kellendonk et al., 1996). When the CrePR1 transgenic mice are crossed with a floxed target mouse line to generate bigenic mice, an epidermal-specific knockout of the target gene can be induced by topical application of a progesterone antagonist. A keratin K14 vector was used to target CrePR1 expression to the epidermis (K14.CrePR1; Fig. 1). This vector targets transgene expression to the basal layer of the epidermis and hair follicles (Kucera et al., 1996), the location of epidermal stem cells. Because stem cells renew this epithelium throughout the life of the organism, once RU486-induced excision occurs in an epidermal stem cell, there should be a focal clone of keratinocytes that persists without the deleted gene for the remainder of the life of the mouse. To confirm that K14.CrePR1 could mediate the inducible excision of a loxP flanked gene, we initially mated K14.CrePR1 mice with CAG-CAT-Z Cre reporter mice (Sakai and Miyazaki, 1997). If Cre activity was induced in stem cells, then we anticipated that topical application of RU486 to the skin of bigenic mice would induce expression of b-galactosidase (b-gal) throughout the epidermis and in hair follicles. Surprisingly, the bigenic skin containing the CAGCAT-Z transgene showed that the b-gal expression was primarily in the differentiated layers of the epidermis, with focal expression in the basal cells, and no expression in hair follicles (Fig. 2A). To date, the b-gal activity has persisted for over 4 months. Because the rate of renewal in mouse epidermis is about 1 week, Cre-mediated excision must have occurred in epidermal stem cells. One explanation for these results is that the CAG promoter is not expressed in keratinocytes residing in the basal layer or in hair follicles. To test this hypothesis, we mated K14.CrePR1 mice with ROSA26 Cre reporter mice (Soriano, 1999), and observed b-gal activity throughout the epidermis and hair follicles (Fig. 2B). These results document the differential expression of the CAG and ROSA26 promoters in the epidermis and