The homeoprotein DLX3 and tumor suppressor p53 co-regulate cell cycle progression and squamous tumor growth.

Epidermal homeostasis depends on the coordinated control of keratinocyte cell cycle. Differentiation and the alteration of this balance can result in neoplastic development. Here we report on a novel DLX3-dependent network that constrains epidermal hyperplasia and squamous tumorigenesis. By integrating genetic and transcriptomic approaches, we demonstrate that DLX3 operates through a p53-regulated network. DLX3 and p53 physically interact on the p21 promoter to enhance p21 expression. Elevating DLX3 in keratinocytes produces a G1-S blockade associated with p53 signature transcriptional profiles. In contrast, DLX3 loss promotes a mitogenic phenotype associated with constitutive activation of ERK. DLX3 expression is lost in human skin cancers and is extinguished during progression of experimentally induced mouse squamous cell carcinoma (SCC). Reinstatement of DLX3 function is sufficient to attenuate the migration of SCC cells, leading to decreased wound closure. Our data establish the DLX3-p53 interplay as a major regulatory axis in epidermal differentiation and suggest that DLX3 is a modulator of skin carcinogenesis.

Negative response elements in keratin genes mediate transcriptional repression and the cross-talk among nuclear receptors.

Very little is known about the mechanisms responsible for the findings that binding of nuclear receptors (NR) to some promoter elements leads to transcriptional activation, whereas binding to others leads to repression. Case in point is the group of epidermal keratin genes and their DNA sequences responsible for repression by NR. Keratin response elements (KREs) interact with receptors for retinoic acid, thyroid hormone, and glucocorticoids. KREs, by their structure and sequence, direct the binding of retinoic acid and thyroid hormone as homodimers and glucocorticoids as monomers. Such specific DNA-receptor interactions are crucial for the repression signal of transcription. In this paper we have analyzed the interactions between the KREs and NR that lead to such repression. We have found that KREs are promoter-independent. They not only provide a docking platform for the receptors, but also play a key role in directing the receptors to bind into particular configurations and coordinating the interactions among different receptors. Both an intact KRE and an intact receptor DNA-binding domain are necessary for the regulation to occur, which emphasizes the importance of interaction between the DNA and NR for proper signaling. Furthermore, KREs allow simultaneous binding of multiple receptors, thus providing fine-tuning of transcriptional regulation. The DNA/DNA-binding domain interactions in keratin promoters exemplify tissue and gene specificity of hormone action.

Novel mechanism of steroid action in skin through glucocorticoid receptor monomers.

Glucocorticoids (GCs), important regulators of epidermal growth, differentiation, and homeostasis, are used extensively in the treatment of skin diseases. Using keratin gene expression as a paradigm of epidermal physiology and pathology, we have developed a model system to study the molecular mechanism of GCs action in skin. Here we describe a novel mechanism of suppression of transcription by the glucocorticoid receptor (GR) that represents an example of customizing a device for transcriptional regulation to target a specific group of genes within the target tissue, in our case, epidermis. We have shown that GCs repress the expression of the basal-cell-specific keratins K5 and K14 and disease-associated keratins K6, K16, and K17 but not the differentiation-specific keratins K3 and K10 or the simple epithelium-specific keratins K8, K18, and K19. We have identified the negative recognition elements (nGREs) in all five regulated keratin gene promoters. Detailed footprinting revealed that the function of nGREs is to instruct the GR to bind as four monomers. Furthermore, using cotransfection and antisense technology we have found that, unlike SRC-1 and GRIP-1, which are not involved in the GR complex that suppresses keratin genes, histone acetyltransferase and CBP are. In addition, we have found that GR, independently from GREs, blocks the induction of keratin gene expression by AP1. We conclude that GR suppresses keratin gene expression through two independent mechanisms: directly, through interactions of keratin nGREs with four GR monomers, as well as indirectly, by blocking the AP1 induction of keratin gene expression.

Specific organization of the negative response elements for retinoic acid and thyroid hormone receptors in keratin gene family.

Retinoic acid and thyroid hormone are important regulators of epidermal growth, differentiation, and homeostasis. Retinoic acid is extensively used in the treatment of many epidermal disorders ranging from wrinkles to skin cancers. Retinoic acid and thyroid hormone directly control the transcription of differentiation-specific genes including keratins. Their effect is mediated through nuclear receptors RAR and T3R. We have previously identified the response element in the K14 gene, K14RARE/TRE, to which these receptors bind, and found that it consists of a cluster of five half-sites with variable spacing and orientation. To determine whether this specific structure is found in other keratin genes, we have mapped and analyzed the RARE/TRE elements in three additional epidermal keratin genes: K5, K6, and K17. We used three different approaches to identify these elements: co-transfection of promoter deletion constructs, gel-shift assays, and site-specific mutagenesis. We localized the RARE/TRE elements relatively close to the TATA box in all three promoters. All three RARE/TRE elements have a similar structural organization: they consist of clusters of 3-6 half-sites with variable spacing and orientation. This means that the clustered structure of the RARE/TREs is a common characteristic for keratin genes. RARE and TRE in the K5 promoter are adjacent to each other whereas in the K17 promoter they overlap. All three keratin REs bind specifically both RAR and T3R in gel-shift assays. Interestingly, addition of ligand to the receptor changes the binding pattern ofthe T3R from homodimer to monomer, reflecting the change in regulation from induction to inhibition.