Role of the Retinoid X Receptor α (RXRα) and its heterodimeric Partners during Skin Carcinogenesis

Nuclear Receptors (NRs) are ligand-dependent transcription factors that play important roles during embryonic development and homoeostasis. Several NRs, including the retinoic acid receptors (RARα, βand γ), the retinoic X receptors (RXRα,β and γ), the vitamin D receptor (VDR), the peroxisome proliferator activated receptors (PPARα, β and γ) and the liver X receptors (LXRα and β), play key roles in the morphogenesis and homeostasis of the skin. RXRα is the predominant retinoid receptor in the epidermis and a obligatory heterodimeric partner for other non-steroidal NRs. Aberrant expression and function of several of these NRs have been found during skin carcinogenesis, both in mouse and human. Natural or synthetic ligands for RXRs, RARs, VDR and PPARs exhibit protective effects against tumor formation in human and mouse. To investigate the possiblerole of keratinocytic RXRα in skin caricnogenesis, we have subjected adult mice, in which RXRα was selectively ablated in keratinocytes (RXRαep-/- mice), to the chemical two-step tumorigenesis protocol. Upon DMBA/TPA treatment RXRαep-/- mice exhibited an increased tumor number, growth rate and a much higher frequency to malignant carcinoma, than control mice. DMBA/TPA-treated RXRαep-/- mice also develop melanocytic growths, which degenerate into melanoma. Furthermore, mice in which PPARγ was selectively ablated in keratinocytes (PPARγep-/-), and mice bearing a PPARα-null mutation are also more susceptible to DMBA/TPA-induced epithelial tumors, than control mice. However, PPARγep-/- and PPARα-null mice did not develop melanocytic growth or melanoma upon DMBA/TPA treatment. We also showed that the susceptibility to epithelial tumors induced by DMBA/TPA is not affected in mice bearing a null mutation of the VDR gene. However, VDR null mice developed a large number of MGs in response to DMBA/TPA treatment, but in contrast to the MGs from RXRαep-/- mice, these MGs did not progress to malignant melanoma. Thus, keratinocytic RXRα/PPARγ and RXRα/PPARα heterodimers may function as suppressor of epithelial tumors, whereas RXRα/VDR heterodimers may play a role in the formation of melanocytic growth. Further studies are required to identify the heterodimerc partner(s) of RXRα involved in the suppression of malignant melanoma. We also used Tg.AC mice as an alternative model to study the role of RXRα during skin carcinogenesis. Tg.AC transgenic mice carry an activated v-Ha-ras oncogene under the control of a partial ζ-globin promoter and are highly sensitive to tumor formation in response to TPA treatment. Surprisingly, Tg.AC/RXRαep-/-(c) mice, carrying the Ha-ras oncogene and in which the selective ablation of the RXRα alleles occurred in epidermal keratinocytes during fetal epidermal morphogenesis, developed skin papillomas in the absence of any promotion agent, some of which progressed to carcinoma. Moreover, in response to a single topical application of DMBA, RXRαep-/-(c) mice, develop epithelial tumors. Thus, ablation of RXRα in keratinocytes during epidermal morphogenesis, is sufficient to promote the formation of Ha-ras initiated skin tumors. All together our results demonstrate that keratinocytic RXRα is a tumor suppressor of epithelial carcinogenesis (most probably within a heterodimer with PPARα and PPARγ) and malignant melanoma, (most probably whitin a heterodimer with VDR and other NRs), via paracrine mechanisms