In a first phase of efforts, we discovered that the Tabby mouse, which has many of the features observed in human EDA, is specifically mutated in the corresponding mouse gene. We demonstrated that the Wnt pathway directly regulates EDA transcription. In published work, we further found that provision of DNA encoding a variant of ectodysplasin (Eda-A1, the longest isoform) in embryonic Tabby mice restores hair follicles and sweat glands. By generating Tet-regulated conditional transgenic mice, we have dissected spatiotemporal actions of Eda-A1 during hair follicle development. We also have characterized eye phenotypes of Tabby mice including blindness and inflammation susceptibility, and they are also reversed by supplementation with the same Eda-A1 isoform. This study has provided the first animal model for ocular surface disease, and also further increased the interest in the possibility of manipulating the Eda pathway to combat dry eye. By large scale genome-wide expression profiling of samples from wild-type and Tabby mice ranging from embryos to adult and from hair follicles to sweat glands and primary keratinocytes, we identified numerous downstream target genes of Eda, including lymphotoxin-, Shh, Wnt10b and Dkk4 in hair follicles and Shh and FoxA1 in sweat glands. More recently, we have further focused on the function of Eda and Eda target genes identified by expression profiling in mutant mouse models. We demonstrated that target gene lymphotoxin-, an immune gene, is involved in hair shaft formation, but not hair follicle induction. We also found that Dkk4, a Wnt antagonist, discriminates an Eda-independent mechanism of secondary hair follicle formation. Notably, both pathways converge at the activation of downstream Shh. Based on these observation, we proposed that different subtypes of hair follicles are formed by variant molecular mechanisms. Conditional Shh transgenic mice and skin-specific Shh knockout mice in wild-type and Tabby backgrounds showed that Shh is required for elongation of Tabby hair shafts, but not for the induction of the primary hair follicles that are missing in Tabby mice. We have also studied and compared the control of Eda-independent skn appendage developmental pathways. A similar signaling pathway (TNF/NF-kB) is required for development of secondary lymphoid organs, but with very different downstream effectors. Skin appendage nails/claws is also independent of EDA, again regulated by a Wnt pathway early on, but working through Fzd6, the loss of Fzd6 distorted claw formation in mice, in line with the demonstrated damage of nail formation in patients lacking an active gene copy. We have now initiated projects for skin exocrine glands, again with Tabby mice as a model system. In sweat glands, FoxA1 was strikingly affected gene in Tabby during late developmental stages and adult stage. Skin-specific FoxA1 knockout mice showed striking anhidrosis, with abundant accumulation of glycoproteins in the lumens and ducts of otherwise complete sweat glands;and we further showed that FoxA1 functions in sweat glands by promoting transcription of an anion channel protein, Best2. Best2 knockout mice also showed severe hypohidrosis/anhidrosis, revealing a FoxA1-Best2 cascade as a fundamental genetic pathway in sweat glands, regulating sweat secretion. Because Best2 is a calcium activated bicarbonate channel, we inferred two alternative cascades for sweating: calcium K/Cl (FoxA1) additional monovalent ion transporter cascade and calcium (FoxA1) Best2 K/Cl ion transporter, with the latter likely playing the major role. We found that four K channels and two Cl channels are highly expressed in sweat glands. Sublocalizations suggest that one of them may be directly activated by calcium (the first cascade), with a second activated by FoxA1/Best2 (the second cascade). Concerning another exocrine skin appendage, we previously showed that Eda-ablated Tabby mice develop ocular surface disease, and EDA patients show extreme dry eye. We plan to examine dry eye etiology with meibomian glands as an entry point. In a first phase, we characterized their development. They are missing in Tabby mice, and Shh knockout mice, Dkk4 transgenic mice, and beta-catenin knockout mice all completely lack meibomian glands. Currently we are characterizing meibomian gland phenotypes in these mice more systematically by time-course histological and immunohistochemical analyses, and assessing the possible trophic role of Eda in preventing aging-related deterioriation of Meibomian gland function. In further initiatives moving toward skin appendage regeneration, incisive studies have led to the isolation and understanding of several types of stem cells jointly required for hair follicle or sweat gland formation and maintenance, but these require very complex protocols and difficult-to-obtain quantities of stem cells to satisfy requirements of regenerative medicine. We will attempt first steps in a long-term but more direct route, starting from embryonic stem cells (ES cells). ES cells have been differentiated into keratinocyte progenitors and full thickness skin epidermis in presence of retinoic acid and Bmp4, but with low efficiency. We have initiated an approach that could be more efficient starting from a master transcription factor that directs them toward the keratinocyte lineage.

National Institute of Health (NIH)
National Institute on Aging (NIA)
Investigator-Initiated Intramural Research Projects (ZIA)
Project #
Application #
Study Section
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
National Institute on Aging
Zip Code
Cui, Chang-Yi; Klar, Joakim; Georgii-Heming, Patrik et al. (2013) Frizzled6 deficiency disrupts the differentiation process of nail development. J Invest Dermatol 133:1990-7
Kunisada, Makoto; Cui, Chang-Yi; Piao, Yulan et al. (2009) Requirement for Shh and Fox family genes at different stages in sweat gland development. Hum Mol Genet 18:1769-78
Hashimoto, Tsuyoshi; Schlessinger, David; Cui, Chang-Yi (2008) Troy binding to lymphotoxin-alpha activates NF kappa B mediated transcription. Cell Cycle 7:106-11