Retinoic Acid Receptor Function Research Articles

Anticoagulant Effects of Synthetic Retinoids Mediated Via Different Receptors on Human Leukemia and Umbilical Vein Endothelial Cells

AbstractWe recently found that retinoic acids (RAs) exert anticoagulant effects by upregulating thrombomodulin (TM) and downregulating tissue factor (TF ) expression in acute promyelocytic leukemia (APL) cells and monoblastic leukemia cells. Two classes of nuclear RA receptors, termed retinoic acid receptors (RARs) and retinoid X receptors, have been identified. Each receptor class consists of three subtypes. In the present study, we have used several synthetic retinoids to determine which receptor subtypes are involved in the regulation of TM and TF expression in NB4 APL cells, U937 monoblastic leukemia cells, and human umbilical vein endothelial cells (HUVECs). Am80, which has no binding affinity for RARγ, and Ch55, which does not bind to cytoplasmic retinoic acid binding protein (CRABP), upregulated TM and downregulated TF in NB4 and U937 cells, similar to all-trans RA (ATRA). A specific RARα antagonist, Ro41-5253, significantly suppressed the upregulation of TM by ATRA and Am80 in NB4 cells, U937 cells, and HUVECs. In contrast, only with preincubation with both RARα and RARβ antagonists was downregulation of TF by retinoids suppressed in NB4 cells. These findings indicate that the mechanism of transactivation and transrepression functions of RARs are distinct and also elucidate the major role of RARα in TM upregulation by retinoids in leukemic cells and HUVECs and the cooperation of RARα and RARβ in TF downregulation by retinoids. They also indicate that binding to CRABP is not required for the anticoagulant effect of retinoids and that synthetic retinoids will prove very useful in controlling distinct targets, the TM and TF genes, at the level of transcription, and will permit the development of retinoids with a new type of anticoagulant effect.

Anticoagulant Effects of Synthetic Retinoids Mediated Via Different Receptors on Human Leukemia and Umbilical Vein Endothelial Cells

We recently found that retinoic acids (RAs) exert anticoagulant effects by upregulating thrombomodulin (TM) and downregulating tissue factor (TF ) expression in acute promyelocytic leukemia (APL) cells and monoblastic leukemia cells. Two classes of nuclear RA receptors, termed retinoic acid receptors (RARs) and retinoid X receptors, have been identified. Each receptor class consists of three subtypes. In the present study, we have used several synthetic retinoids to determine which receptor subtypes are involved in the regulation of TM and TF expression in NB4 APL cells, U937 monoblastic leukemia cells, and human umbilical vein endothelial cells (HUVECs). Am80, which has no binding affinity for RARγ, and Ch55, which does not bind to cytoplasmic retinoic acid binding protein (CRABP), upregulated TM and downregulated TF in NB4 and U937 cells, similar to all-trans RA (ATRA). A specific RARα antagonist, Ro41-5253, significantly suppressed the upregulation of TM by ATRA and Am80 in NB4 cells, U937 cells, and HUVECs. In contrast, only with preincubation with both RARα and RARβ antagonists was downregulation of TF by retinoids suppressed in NB4 cells. These findings indicate that the mechanism of transactivation and transrepression functions of RARs are distinct and also elucidate the major role of RARα in TM upregulation by retinoids in leukemic cells and HUVECs and the cooperation of RARα and RARβ in TF downregulation by retinoids. They also indicate that binding to CRABP is not required for the anticoagulant effect of retinoids and that synthetic retinoids will prove very useful in controlling distinct targets, the TM and TF genes, at the level of transcription, and will permit the development of retinoids with a new type of anticoagulant effect.

8] Use of transgenic mice to eliminate retinoic acid receptor function in specific tissues

This chapter proposes a novel strategy to examine the physiological functions of retinoic acid (RA) in a tissue-specific manner. This novel strategy has revealed the absolute requirement of RA in normal skin development. In addition to the epidermis, there are several other interesting targets in which to express the dominant-negative receptor for RA (RAR), that is, lung, liver, intestine, bone, and testis. Thus, it would be possible to dissect the roles of RA systemically in these organs during embryogenesis and in adults. Because the structure and mode of actions of the thyroid hormone receptor (TR)/RAR subfamily are quite homologous, it is possible to construct a dominant-negative molecule of other members by the same procedure, which might be used to discover unknown physiologic functions of other lipophilic molecules. An application of this strategy to so-called orphan receptors would be quite difficult and might reveal the function of the orphan receptors before their ligands are known.

Identification and Functional Separation of Retinoic Acid Receptor Neutral Antagonists and Inverse Agonists

Inverse agonists are ligands that are capable of repressing basal receptor activity in the absence of an agonist. We have designed a series of C-1-substituted acetylenic retinoids that exhibit potent antagonism of retinoic acid receptor (RAR)-mediated transactivation. Comparison of these related retinoid antagonists for their ability to repress basal RAR transcriptional activity demonstrates that the identity of the C-1 substituent differentiates these ligands into two groups: RAR inverse agonists and neutral antagonists. We show that treatment of cultured human keratinocytes with a RAR inverse agonist, but not a RAR neutral antagonist, leads to the repression of the serum-induced differentiation marker MRP-8. While RAR-selective agonists also repress expression of MRP-8, cotreatment with a RAR inverse agonist and a RAR agonist results in a mutual repression of their individual inhibitory activities, indicating the distinct modes of action of these two disparate retinoids in modulating MRP-8 expression. Our data indicate that RARs, like beta2-adrenoreceptors, are sensitive to inverse agonists and that this new class of retinoids will provide insight into the molecular mechanisms of RAR function in skin and other responsive tissues.

Murine endodermal F9E cells, derived from the teratocarcinoma line F9, contain high basal levels of retinoic acid receptors (RARs and RXRs) but are not sensitive to the actions of retinoic acid

Retinoic acid (RA)-induced endodermal differentiation of F9 teratocarcinoma cells is accompanied by altered of many genes, including the retinoic acid receptor (RAR) α, β, and γ and retinoic x receptor (RXR) α and γ genes. In addition, RA enhances the binding of nuclear receptors to the RA responsive elements and increases the trans-activation of the RA-responsive genes in F9 cells. These data suggest that RA increases the overall function of RA receptors in F9 cells. However, the cell line F9E, established by long-term exposure of F9 cells to RA, is insensitive to RA. These cells expressed the genes for keratin 8 and tissue specific plasminogen activator (t-PA), which are characteristic of endoderm cells. Northern blot analysis showed that F9E cells expressed high basal levels of RAR and RXR mRNAs, but genes including the RARs and RXRs, which are normally regulated by RA in F9 cells, were no longer regulated in F9E cells. Further, F9E cells were not sensitive to the antiproliferative effect of RA. In these cells, high levels of RA receptors constitutively bound to RA responsive elements, but RA could neither increase the amount of RA receptor binding to the RA responsive elements, nor enhance the transcription of the RA target genes. These findings provide a possible explanation for insensitivity of endodermal cells to the actions of RA. Our data also indicate that the levels of endogenous RA receptors do not predict the degree of RA sensitivity, and that RA responsiveness might be a function of the cell's ability to regulate RA receptor genes.

Inhibition of Retinoic Acid Receptor Function and Retinoic Acid-regulated Gene Expression in Mouse Melanoma Cells by Calreticulin: A POTENTIAL PATHWAY FOR CYCLIC AMP REGULATION OF RETINOID ACTION

Calcium is a second messenger that controls a wide variety of cellular functions. Because of its multiple actions, there is a stringent requirement for calcium homeostasis, and this is achieved in part by a system of transport and storage proteins such as calreticulin located in the endoplasmic reticulum. Calreticulin is also found in the nucleus, suggesting that it may have a role in transcriptional regulation. It has been reported that calreticulin can inhibit steroid-regulated gene transcription by preventing receptor binding to DNA. Here we report that overexpression of the calreticulin gene in B16 mouse melanoma cells resulted in a decrease in retinoic acid (RA)-stimulated reporter gene expression. Gel shift analysis showed that purified calreticulin inhibited the binding of endogenous RAR to a beta-RA response element oligonucleotide, only if added prior to the addition of the oligonucleotide. Co-immunoprecipitation studies suggest a physical interaction between RAR and calreticulin. Transfection of the calreticulin gene into B16 cells inhibited the RA induction of protein kinase Calpha, a marker of RA-induced differentiation. We also found that cyclic AMP increased the expression of calreticulin. Cyclic AMP may act to antagonize RA action by both decreasing RAR expression (Y. Xiao, D. Desai, T. Quick, and R. M. Niles, J. Cell Physiol., in press) and stimulating calreticulin levels.

Nuclear retinoid receptor expression in normal human endometrium throughout the menstrual cycle.

Previous work has shown that retinoic acid receptors (RARs) and retinoid X receptors (RXRs) are expressed in human endometrial epithelial and stromal cells. These nuclear receptors mediate the biological effects of retinoic acid, a vitamin A derivative which may have an important, though poorly characterized role in the functional differentiation of secretory epithelia. The aim of this study was to find out whether the expression of RAR and RXR mRNA in endometrial epithelial and stromal cells varies in relation to the menstrual cycle. The expression of RARs and RXRs was investigated by Northern blotting and, for stromal cells, there were no differences in expression of RAR-alpha, RAR-beta, RAR-gamma and RXR-alpha between the proliferative and secretory phases of the menstrual cycle. Similarly, for epithelial tissue, there were no significant differences between the proliferative and secretory phases with respect to the expression of RAR-alpha, RAR-gamma and RXR-alpha. However, RAR-beta was expressed at a 1.7-fold higher level in epithelial samples from the proliferative phase compared to the secretory phase. Overall, the levels of expression of RAR-alpha, RAR-beta and RAR-gamma were 1.7- to 4-fold higher in stromal cells compared to epithelial cells whereas RXR-alpha was expressed at a similar level in both cell types. We have previously suggested that retinoic acid has a role in endometrial differentiation or function which may be reflected by cyclical changes in intracellular retinoic acid levels. These data indicate that RARs and RXRs are expressed at a similar level throughout the menstrual cycle, with the possible exception of RAR-beta, implying that any menstrual cycle-related function of RARs in controlled by ligand availability rather than by changes in expression of the receptors.

The consensus sequence of a major Alu subfamily contains a functional retinoic acid response element.

Alu repeats are interspersed repetitive DNA elements specific to primates that are present in 500,000 to 1 million copies. We show here that an Alu sequence encodes functional binding sites for retinoic acid receptors, which are members of the nuclear receptor family of transcription factors. The consensus sequences for the evolutionarily recent Alu subclasses contain three hexamer half sites, related to the consensus AGGTCA, arranged as direct repeats with a spacing of 2 bp, which is consistent with the binding specificities of retinoic acid receptors. An analysis was made of the DNA binding and transactivation potential of these sites from an Alu sequence that has been previously implicated in the regulation of the keratin K18 gene. These Alu double half sites are shown to bind bacterially synthesized retinoic acid receptors as assayed by electrophoretic mobility shift assays. These sites are further shown to function as a retinoic acid response element in transiently transfected CV-1 cells, increasing transcription of a reporter gene by a factor of approximately 35-fold. This transactivation requires cotransfection with vectors expressing retinoic acid receptors, as well as the presence of all-trans-retinoic acid, which is consistent with the known function of retinoic acid receptors as ligand-inducible transcription factors. The random insertion of potentially thousands of Alu repeats containing retinoic acid response elements throughout the primate genome is likely to have altered the expression of numerous genes, thereby contributing to evolutionary potential.

Separation of Transactivation and AP1 Antagonism Functions of Retinoic Acid Receptor α

Retinoic acid receptors (RARs) regulate gene expression either by directly binding to the RAR-responsive elements or by antagonizing the action of c-Jun/c-Fos (AP1). AP1 is involved in the expression of metalloproteases, cytokines and other factors which play critical roles in the turnover of extracellular matrix, inflammation and hyperproliferation in diseases such as psoriasis, rheumatoid arthritis and in tumor metastases. We demonstrate here that synthetic retinoids inhibit 12-O-tetradecanoylphorbol-14-acetate-induced transcription from the stromelysin AP1 motif through RAR alpha, -beta, and -gamma. Interestingly, these diaryl acetylenic retinoids, which are potent agonists only for RAR beta and RAR gamma, but not for RAR alpha, in transactivation assays, are able to inhibit AP1-dependent gene expression through RAR alpha. Thus these analogs can differentially affect the transactivation and AP1 antagonistic functions of RAR alpha. These results demonstrate that the transactivation and AP1 antagonistic functions are separable, and it should be possible to develop retinoids that are completely specific for AP1 antagonism through all RARs. Furthermore, using an RAR-selective ligand, we also demonstrate the separation of ligand binding and AP1 antagonism functions of RARs.

Reconstitution of thyroid hormone receptor and retinoic acid receptor function in the fission yeast Schizosaccharomyces pombe

Reconstitution of thyroid hormone receptor and retinoic acid receptor function in the fission yeast Schizosaccharomyces pombe

Reconstitution of thyroid hormone receptor and retinoic acid receptor function in the fission yeast Schizosaccharomyces pombe.

We report here a characterization of the thyroid hormone receptors (T3Rs), retinoic acid receptors (RARs), and retinoid X receptors (RXRs) by reconstituting their actions in the fission yeast Schizosaccharomyces pombe. S. pombe provide a well defined and readily manipulated genetic background devoid of known endogenous nuclear hormone receptors. All the receptors tested, when introduced exogenously into S. pombe, induced high levels of reporter gene activation in response to physiological concentrations of hormone ligand. In these properties, the S. pombe system exhibits significant advantages over the previously employed Saccharomyces cerevisiae system. Use of the S. pombe system permitted the elucidation of previously undescribed differences in the DNA sequence recognition properties of different isoforms of the RXR and RARs, and the identification of apparently novel forms of response element for RXRs and RARs. Intriguingly, the v-erb A allele of T3R, a transcriptional repressor in vertebrate cells, acts as a transcriptional activator both in S. cerevisiae and in the evolutionarily highly divergent S. pombe, underscoring the importance of cellular factors in the regulation of receptor transcriptional activity.

Alteration of a single amino acid residue in retinoic acid receptor causes dominant-negative phenotype.

Thyroid hormone receptor (TR) and retinoic acid receptor (RAR) are closely related not only in their structures but also in their modes of action. TR and RAR share many amino acids in the functionally important DNA binding domain, ligand binding domain, and heterodimeric interface. To function, both receptors dimerize with retinoid X receptor, recognize their cognate hormone response elements, and then respond to the ligand, leading to the activation of the target genes. Genetic analysis has revealed mutated TRs in thyroid hormone resistance syndrome, which displays autosomal dominant inheritance. Eventually, the mutated TRs show the dominant-negative phenotype on the wild-type TR. The mutations have been observed mostly in conserved amino acids between TR and RAR in the ligand binding domain and produce hormone-insensitive receptors. In this report, we demonstrate that the dominant-negative phenotype is transferable to RAR by a single amino acid substitution identified in the syndromes of thyroid hormone resistance. The mutated RAR can suppress the wild-type RAR function, especially at the physiological concentration of retinoic acid. Consistently, the mutated RAR is an absolute requirement of intact dimeric and DNA binding capacities for the dominant-negative phenotype, indicating the necessity of the maintenance of a machinery for correct recognition of the targets. Thus, the hormone-insensitive receptor may be interfering with the access of functional receptors to the hormone response elements. The dominant-negative RAR will serve as a molecular tool to elucidate physiological roles of RAR by blocking RAR-mediated signaling pathways.

V-erbA and citral reduce the teratogenic effects of all-trans retinoic acid and retinol, respectively, in Xenopus embryogenesis.

Treatment of late blastula/early gastrula stage Xenopus embryos with all-trans retinoic acid results in disruption of the primary body axis through effects on both mesoderm and neuroectoderm. This effect of retinoic acid, coupled with the known presence of retinoic acid in Xenopus embryos has led to the proposal that retinoic acid may be an endogenous morphogen providing positional information in early development. To further elucidate the role of retinoic acid in early Xenopus development, we have attempted to interfere with the retinoic acid signalling pathway both at the level of retinoic acid formation, by treatment with citral (3,7-dimethy-2,6-octadienal), and at the level of nuclear retinoic acid receptor function, by microinjection of v-erbA mRNA. The feasibility of this approach was demonstrated by the ability of citral treatment and v-erbA mRNA injection to reduce the teratogenic effects of exogenous retinol and retinoic acid, respectively, in early Xenopus development. Interestingly, v-erbA mRNA injection and citral treatment of gastrula stage embryos resulted in tadpoles with a similar set of developmental defects. The defects were chiefly found in tissues that received a contribution of cells from the neural crest, suggesting that at least a subset of neural crest cells may be sensitive to the endogenous level of retinoic acid. In accord with this proposal, it was found that the expression patterns of two early markers of cranial neural crest cells, Xtwi and XAP-2, were altered in embryos injected with v-erbA mRNA. These results indicate that structures in addition to the primary axis are regulated by retinoic acid signalling during early Xenopus development.

Reconstitution of retinoid X receptor function and combinatorial regulation of other nuclear hormone receptors in the yeast Saccharomyces cerevisiae.

The nuclear hormone receptor family of transcription factors regulates gene expression via a complex combinatorial network of interactions. Of particular interest is the ability of retinoid X receptors (RXRs) to form heterodimers with retinoic acid receptors (RARs) and thyroid hormone receptors (TRs), thereby modifying their activities. We report here that RXR, RAR, and TR function can be reconstituted in the yeast Saccharomyces cerevisiae and demonstrate that the combinatorial regulation seen in vertebrate cells can be reproduced in the yeast background. Using this system, we have shown that RARs respond to a wide variety of retinoid ligands but that RXRs are specific for the 9-cis isomer of retinoic acid. RXR enhanced the activity of RARs and TRs on a variety of hormone response elements without demonstrably altering their DNA specificity. Interestingly, the ability of RXR to potentiate gene activation by RARs and by TRs varied for different receptor isoforms.

ErbA: tumor suppressor turned oncogene?

Cancer has been commonly linked to aberrant proliferation and a failure of the transformed cells to differentiate. Activated proto-oncogenes are thought to provide continuous proliferation signals that enhance the growth of these cells. Conversely, cellular transformation may also be achieved by the inactivation of genes whose normal function is to constrain cell growth by either suppressing proliferation or inducing differentiation. Such an inactivation could result from dominant-negative mutations, leading to the expression of abnormal proteins that inhibit the function of their normal counterparts. A prototype example is the v-erbA oncogene of the avian erythroblastosis virus (AEV), which antagonizes the transcriptional regulatory function of the chicken c-ErbA/thyroid hormone receptors (c-ErbA/TR) and the structurally related retinoic acid receptors (RARs). The result of this inhibition is a loss of hormone responsiveness and hormone-induced differentiation. Here we have a parallel to the tumor suppressor gene where it is also a loss of function that induces the transformation process. In this way, the normal, hormone-activated c-ErbA/TRs and RARs act as growth suppressors because the resulting differentiated cells irreversibly lose proliferative potential. In this article, the properties of v-ErbA will be discussed in the context of c-ErbA/thyroid hormone and retinoic acid receptor function.

A tissue-specific enhancer confers Pit-1-dependent morphogen inducibility and autoregulation on the pit-1 gene.

Pit-1 is a tissue-specific POU domain factor obligatory for the appearance of three cell phenotypes in the anterior pituitary gland. Expression of the pit-1 gene requires the actions of a cell-specific 390-bp enhancer, located 10 kb 5' of the pit-1 transcription initiation site, within sequence that proves essential for effective pituitary targeting of transgene expression during murine development. The enhancer requires the concerted actions of a cell-specific cis-active element, Pit-1 autoregulatory sites, and atypical morphogen response elements. Pituitary ontogeny in the Pit-1-defective Snell dwarf mouse reveals that pit-1 autoregulation is not required for initial activation or continued expression during critical phases of Pit-1 target gene activation but, subsequently, is necessary for maintenance of pit-1 gene expression following birth. A potent 1,25-dihydroxyvitamin D3-responsive enhancer element defines a physiological site in which a single nucleotide alteration in the sequence of core binding motifs modulates the spacing rules for nuclear receptor response elements. Unexpectedly, the major retinoic acid response element is absolutely dependent on Pit-1 for retinoic acid receptor function. On this DNA element, Pit-1 appears to function as a coregulator of the retinoic acid receptor, suggesting an intriguing linkage between a cell-specific transcription factor and the actions of morphogen receptors that is likely to be prototypic of mechanisms by which other cell-specific transcription factors might confer morphogen receptor responsivity during mammalian organogenesis.

Function of retinoic acid receptor γ in the mouse

Null mutant mice for retinoic acid receptor γ2 (RARγ2) or all RARγ isoforms were generated. RARγ2 mutants appeared normal, whereas RARγ mutants exhibited growth deficiency, early lethality, and male sterility due to squamous metaplasia of the seminal vesicles and prostate. These defects were previously observed in vitamin A-deficient animals and could be prevented by RA administration, demonstrating that RARγ mediates some of the retinoid signal in vivo. Congenital defects included Harderian gland agenesis, tracheal cartilage malformations, and homeotic transformations along the rostral axial skeleton, establishing a direct link between RA and patterning of the axial skeleton. We also show that in utero RA-induced lumbosacral truncations are mediated by RARγ. The observed RARγ null phenotype suggests a high degree of functional redundancy among the RARs. The variable penetrance of some of the observed defects is discussed in light of this redundancy and stochastic variation of gene activity.

Functional inhibition of retinoic acid response by dominant negative retinoic acid receptor mutants.

The diverse effects of retinoids on the development, growth, and homeostasis of vertebrate organisms are mediated in part by three distinct isoforms of retinoic acid receptors (RARs). These proteins, which are structurally and functionally closely related to thyroid hormone receptors and the oncogene product v-ErbA, regulate patterns of gene expression in target tissues. One approach to study the distinct effects of retinoic acid in cells is to subvert this activity of endogenous receptors by expression of dominant negative receptor derivatives. We demonstrate here that RAR alpha, RAR beta, and RAR gamma can be converted into potent negative transcriptional regulators that block wild-type RAR function. Furthermore, these mutant RARs, but not the wild-type receptors, actively repress the basal transcription level of target promoters. When expressed in transgenic mice, the most potent of these inhibitory receptor mutants is apparently able to disturb developmental processes by inducing a cleft palate in transgenic offspring.

Retinoic acid resistance of estradiol-independent breast cancer cells coincides with diminished retinoic acid receptor function

Retinoic acid (RA) strongly inhibits proliferation of the estrogen (E 2)-dependent human breast cancer cell lines MCF7, T47D, and ZR75-1, but not the E 2-independent and E 2 receptor (ER)-negative lines MDA-MB231, MDA-MB468, BT20 and Hs578T. The specific sensitivity of the E 2-dependent cell lines seems not to be caused by an inhibitory effect of RA on ER functioning since RA inhibited the proliferative response not only to E 2 but also to insulin. Furthermore, endogenous RA receptors (RARs) hardly impaired transcriptional activation of an E 2 responsive element-tk-CAT reporter construct. RARα mRNA was highly expressed in the RA-responsive lines, but not in the unresponsive lines, except BT20. With the exception of Hs578T, also RARβ mRNA expression was low in the unresponsive lines. While in the dependent lines and Hs578T RA activated RA responsive element-dependent transcriptional activity, this response was very low in MDA-MB231, MDA-MB468, and BT20, suggesting that the RA resistance of these latter three ER-negative lines is due to underexpression of functional RARs. Our results suggest that the loss of functional RARs may be a frequent event, leading to RA unresponsiveness of ER-negative breast cancer cells. This implies that both the steroid and retinoid receptor status of breast tumors may be used to predict a successful treatment with retinoids.

Terminal differentiation in keratinocytes involves positive as well as negative regulation by retinoic acid receptors and retinoid X receptors at retinoid response elements.

Terminal differentiation of epidermal keratinocytes is inhibited by 1 microM retinoic acid, a concentration which induces differentiation in a number of cell types, including F9 teratocarcinoma cells. The molecular basis for these opposing retinoid responses is unknown, although retinoic acid receptors (RARs) and retinoid X receptors (RXRs) have been detected in both cell types. When F9 cells are stably transfected with a truncated RAR alpha lacking the E/F domain necessary for ligand binding and RAR/RXR dimerization, action at retinoid response elements is suppressed and cells produce a retinoic acid-resistant phenotype; i.e., they are blocked in differentiation (A. S. Espeseth, S. P. Murphy, and E. Linney, Genes Dev. 3:1647-1656, 1989). If retinoid receptors influence epidermal differentiation only in a negative fashion, then suppression of transactivation at retinoid response elements would be expected to enhance, rather than block, keratinocyte differentiation. In this study, we show that surprisingly, even though constitutive expression of an analogous truncated RAR gamma in keratinocytes specifically suppressed transactivation at retinoid response elements, keratinocytes were blocked, rather than enhanced, in their ability to undergo morphological and biochemical features of differentiation. These findings demonstrate a direct and hitherto unrecognized role for RARs and RXRs in positively as well as negatively regulating epidermal differentiation. Additionally, our studies extend those of Espeseth et al. (Genes Dev. 3:1647-1656, 1989), indicating a novel RAR function independent of the E/F domain.