The role of estrogen receptor-α and its activation function-1 for growth plate closure in female mice.

High estradiol levels in late puberty induce growth plate closure and thereby cessation of growth in humans. In mice, the growth plates do not fuse after sexual maturation, but old mice display reduced longitudinal bone growth and high-dose estradiol treatment induces growth plate closure. Estrogen receptor (ER)-α stimulates gene transcription via two activation functions (AFs), AF-1 and AF-2. To evaluate the role of ERα and its AF-1 for age-dependent reduction in longitudinal bone growth and growth plate closure, female mice with inactivation of ERα (ERα(-/-)) or ERαAF-1 (ERαAF-1(0)) were evaluated. Old (16- to 19-mo-old) female ERα(-/-) mice showed continued substantial longitudinal bone growth, resulting in longer bones (tibia: +8.3%, P < 0.01) associated with increased growth plate height (+18%, P < 0.05) compared with wild-type (WT) mice. In contrast, the longitudinal bone growth ceased in old ERαAF-1(0) mice (tibia: -4.9%, P < 0.01). Importantly, the proximal tibial growth plates were closed in all old ERαAF-1(0) mice while they were open in all WT mice. Growth plate closure was associated with a significantly altered balance between chondrocyte proliferation and apoptosis in the growth plate. In conclusion, old female ERα(-/-) mice display a prolonged and enhanced longitudinal bone growth associated with increased growth plate height, resembling the growth phenotype of patients with inactivating mutations in ERα or aromatase. In contrast, ERαAF-1 deletion results in a hyperactive ERα, altering the chondrocyte proliferation/apoptosis balance, leading to growth plate closure. This suggests that growth plate closure is induced by functions of ERα that do not require AF-1 and that ERαAF-1 opposes growth plate closure.

Roles of transactivating functions 1 and 2 of estrogen receptor-alpha in bone.

The bone-sparing effect of estrogen is primarily mediated via estrogen receptor-α (ERα), which stimulates target gene transcription through two activation functions (AFs), AF-1 in the N-terminal and AF-2 in the ligand binding domain. To evaluate the role of ERα AF-1 and ERα AF-2 for the effects of estrogen in bone in vivo, we analyzed mouse models lacking the entire ERα protein (ERα(-/-)), ERα AF-1 (ERαAF-1(0)), or ERα AF-2 (ERαAF-2(0)). Estradiol (E2) treatment increased the amount of both trabecular and cortical bone in ovariectomized (OVX) WT mice. Neither the trabecular nor the cortical bone responded to E2 treatment in OVX ERα(-/-) or OVX ERαAF-2(0) mice. OVX ERαAF-1(0) mice displayed a normal E2 response in cortical bone but no E2 response in trabecular bone. Although E2 treatment increased the uterine and liver weights and reduced the thymus weight in OVX WT mice, no effect was seen on these parameters in OVX ERα(-/-) or OVX ERαAF-2(0) mice. The effect of E2 in OVX ERαAF-1(0) mice was tissue-dependent, with no or weak E2 response on thymus and uterine weights but a normal response on liver weight. In conclusion, ERα AF-2 is required for the estrogenic effects on all parameters evaluated, whereas the role of ERα AF-1 is tissue-specific, with a crucial role in trabecular bone and uterus but not cortical bone. Selective ER modulators stimulating ERα with minimal activation of ERα AF-1 could retain beneficial actions in cortical bone, constituting 80% of the skeleton, while minimizing effects on reproductive organs.

The estrogen-responsive adrenomedullin and receptor-modifying protein 3 gene identified by DNA microarray analysis are directly regulated by estrogen receptor.

Recent studies have revealed that hundreds of genes in the uterus are activated by estrogen. Their expression profiles differ over time and doses and it is not clear whether all these genes are directly regulated by estrogen via the estrogen receptor. To select the genes that may be regulated by estrogen, we treated mice with several doses of estrogen and searched for those genes whose dose-response expression pattern mirrored the uterine growth pattern. Among those genes, we found that the dose-dependent expression of the adrenomedullin (ADM) gene correlated well with the uterotrophic effect of estrogen. ADM expression is induced early after estrogen administration and is restricted to the endometrial stroma. The spatiotemporal gene expression pattern of ADM was similar to that of receptor-modifying protein 3 (RAMP3). RAMP3 is known to modify calcitonin gene-related receptor (CRLR) so that it can then serve as an ADM receptor. Chromatin immunoprecipitation assays indicated that the estrogen receptor binds directly to the ADM promoter region and RAMP3 intron after estrogen administration. It was also shown that neither the ADM nor RAMP3 gene could be activated in estrogen receptor-alpha-null mouse. Although uterine ADM expression has been reported to occur in the myometrium, our observations indicate that estrogen-induced ADM is also expressed in the uterine stroma and that such variable, spatiotemporally regulated ADM expression contributes to a wider range of biological effects than previously expected.

Role of foxj1 and estrogen receptor alpha in ciliated epithelial cell differentiation of the neonatal oviduct.

Estrogen regulates proliferation and differentiation of epithelial cells in the mammalian oviduct, but pathways for cell-specific differentiation are not well understood. In the epithelial cells of the developing rat oviduct, we found estrogen receptor (ER) alpha is expressed at birth and persists in all cells through neonatal day (ND) 7 when ciliated cells appear. To determine a specific function of ER and foxj1, a transcription factor known to have fundamental roles in ciliogenesis in the lung, in differentiation of the ciliated epithelial cells, we treated newborn rats from ND 0 to 5 with estradiol-17beta (E2) with and without a selective ER antagonist. E2 enhanced the number of proliferating cells and accelerated the process of epithelial cell differentiation resulting in ciliogenesis by ND 5, and co-treatment with an ER antagonist inhibited these changes. Foxj1 was expressed only in the infundibulum and ampulla (INF/AMP). That expression preceded the appearance of cilia and was induced by E2. Cilia were absent in oviducts of foxj1-deficient mice, indicating that foxj1 plays a critical role in oviductal ciliogenesis. However, we found the presence of cilia in the ERalpha-deficient mouse oviduct. The widespread expression of ERalpha in oviductal epithelium, but restriction of cilia to the INF/AMP regions, and importantly, the presence of cilia in the ERalpha-deficient mice, suggested ER signaling is not essential for ciliated epithelial cell differentiation. These observations demonstrate that, although E2 stimulates the differentiation process of ciliated epithelial cells, foxj1 is directly required for epithelial cell ciliogenesis of the neonatal oviduct.

Role of fibroblast growth factor-2 isoforms in the effect of estradiol on endothelial cell migration and proliferation.

Both 17beta-estradiol (E2) and fibroblast growth factor-2 (FGF2) stimulate angiogenesis and endothelial cell migration and proliferation. The first goal of this study was to explore the potential link between this hormone and this growth factor. E2-stimulated angiogenesis in SC Matrigel plugs in Fgf2+/+ mice, but not in Fgf2-/- mice. Cell cultures from subcutaneous Matrigel plugs demonstrated that E2 increased both migration and proliferation in endothelial cells from Fgf2+/+ mice, but not from in Fgf2-/- mice. Several isoforms of fibroblast growth factor-2 (FGF2) are expressed: the low molecular weight 18-kDa protein (FGF2lmw) is secreted and activates tyrosine kinase receptors (FGFRs), whereas the high molecular weight (21 and 22 kDa) isoforms (FGF2hmw) remains intranuclear, but their role is mainly unknown. The second goal of this study was to explore the respective roles of FGF2 isoforms in the effects of E2. We thus generated mice deficient only in the FGF2lmw (Fgf2lmw-/-). E2 stimulated in vivo angiogenesis and in vitro migration in endothelial cells from Fgf2lmw-/- as it did in Fgf2+/+ mice. E2 increased FGF2hmw protein abundance in endothelial cell cultures from Fgf2+/+ and Fgf2lmw-/- mice. As shown using siRNA transfection, these effects were FGFR independent but involved FGF2-Interacting Factor, an intracellular FGF2hmw partner. This is the first report for a physiological role for the intracellular FGF2hmw found to mediate the effect of E2 on endothelial cell migration via an intracrine action.

Estradiol alters nitric oxide production in the mouse aorta through the alpha-, but not beta-, estrogen receptor.

Although estradiol (E(2)) has been recognized to exert several vasculoprotective effects in several species, its effects in mouse vasomotion are unknown, and consequently, so is the estrogen receptor subtype mediating these effects. We investigated the effect of E(2) (80 microg/kg/day for 15 days) on NO production in the thoracic aorta of ovariectomized C57Bl/6 mice compared with those given placebo. E(2) increased basal NO production. In contrast, the relaxation in response to ATP, to the calcium ionophore A23187, and to sodium nitroprusside was unaltered by E(2), whereas acetylcholine-elicited relaxation was decreased. The abundance of NO synthase I, II, and III immunoreactive proteins (using Western blot) in thoracic aorta homogenates was unchanged by E(2). To determine the estrogen receptor (ER) subtype involved in these effects, transgenic mice in which either the ERalpha or ERbeta has been disrupted were ovariectomized and treated, or not, with E(2). Basal NO production was increased and the sensitivity to acetylcholine decreased in ERbeta knockout mice in response to E(2), whereas this effect was abolished in ERalpha knockout mice. Finally, these effects of E(2) on vasomotion required long-term and/or in vivo exposure, as short-term incubation of aortic rings with 10 nmol/L E(2) in the isolated organ chamber did not elicit any vasoactive effects. In conclusion, this study demonstrates that ERalpha, but not ERbeta, mediates the beneficial effect of E(2) on basal NO production.

The AF-1 activation-function of ERalpha may be dispensable to mediate the effect of estradiol on endothelial NO production in mice.

Two isoforms of estrogen receptor (ER) have been described: ERalpha and ERbeta. The initial gene targeting of ERalpha, consisting in the introduction of a Neo cassette in exon 1 [alphaERKO, hereafter called ERalpha-Neo KO (knockout)], was reported in 1993. More recently, another mouse deficient in ERalpha because of the deletion of exon 2 (ERalphaKO, hereafter called ERalpha-delta2 KO) was generated. In ovariectomized ERalpha-wild-type mice, estradiol (E(2)) increases uterine weight and basal production of endothelial nitric oxide (NO). Both of these effects are abolished in ERalpha-delta2 KO mice. In contrast, we show here that both of these effects of E(2) are partially (uterine weight) or totally (endothelial NO production) preserved in ERalpha-Neo KO. We also confirm the presence of two ERalpha mRNA splice variants in uterus and aorta from ERalpha-Neo KO mice. One of them encodes a chimeric ERalpha protein (ERalpha55), partially deleted in the A/B domain, that was detected in both uterus and aorta by Western blot analysis. The other ERalpha mRNA splice variant codes for an isoform deleted for the A/B domain (ERalpha46), which was detected in uterus of ERalpha-Neo KO, and wild-type mice. This protein isoform was not detected in aorta. The identification of these two N-terminal modified isoforms in uterus, and at least one of them in aorta, probably explains the persistence of the E(2) effects in ERalpha-Neo KO mice. Furthermore, ERalpha-Neo KO mice may help in the elucidation of the specific functions of full-length ERalpha (ERalpha66) and ERalpha46, both shown to be physiologically generated in vivo.

Deletion of estrogen receptors reveals a regulatory role for estrogen receptors-beta in bone remodeling in females but not in males.

To determine the contributions of estrogen receptor (ER)alpha and ERbeta in bone growth and remodeling in male and female mice, we generated and analyzed full knockouts for each receptor, and a double ER knockout. Although suppression of the ligand to the ERs (i.e., estradiol) after menopause or gonadectomy in females led to a catastrophic increase in bone turnover and concomitant bone loss, deletion of one or both ERs failed to show such an effect. Complete deletion of ERalpha led to a decrease, not an increase, in bone turnover and an increase, not a decrease, in trabecular bone volume in both male and female animals. Deletion of ERbeta led to different responses in males, where bone was unaffected, and in females, where bone resorption was decreased and trabecular bone volume increased. In contrast, deletion of both ERs led to a profound decrease in trabecular bone volume in females, which was associated with a decrease, not an increase, in bone turnover. Finally, deletion of ERalpha, but not ERbeta, led to major changes in circulating levels of estradiol and/or testosterone, indirectly affecting bone remodeling and bone mass. Thus, only ERalpha was shown to regulate bone remodeling in males, whereas in females both receptor subtypes influenced this process and could, at least under basal knockout conditions, compensate for each other.

Increased anxiety and synaptic plasticity in estrogen receptor beta -deficient mice.

Estrogens are powerful modulators of neuronal physiology and in humans may affect a broad range of functions, including reproductive, emotional, and cognitive behaviors. We studied the contribution of estrogen receptors (ERs) in modulation of emotional processes and analyzed the effects of deleting ERalpha or ERbeta in mice. Behavior consistent with increased anxiety was observed principally in ERbeta mutant females and was associated with a reduced threshold for the induction of synaptic plasticity in the basolateral amygdala. Local increase of 5-hydroxytryptamine 1a receptor expression in medial amygdala may contribute to these changes. Our data show that, particularly in females, there is an important role for ERbeta-mediated estrogen signaling in the processing of emotional behavior.

Estradiol accelerates reendothelialization in mouse carotid artery through estrogen receptor-alpha but not estrogen receptor-beta.

BACKGROUND: The atheroprotective effect of 17beta-estradiol (E(2)) has been suggested in women and clearly demonstrated in animals through both an effect on lipid metabolism and a direct effect on the cells of the arterial wall. It has been shown, for example, that E(2) promotes endothelium-dependent relaxation and accelerates reendothelialization in rats. Similar studies have been undertaken in mice to appreciate the molecular mechanism of this process. METHODS AND RESULTS: We report here a model of electric carotid injury adapted from that described by Carmeliet et al (1997) that allows us to precisely evaluate the reendothelialization process. We demonstrate that E(2) accelerates endothelial regeneration in castrated female wild-type mice. In ovariectomized transgenic mice in which either the estrogen receptor (ER)-alpha or ERbeta gene has been disrupted, E(2) accelerated reendothelialization in female ERbeta knockout mice, whereas this effect was abolished in female ERalpha knockout mice. CONCLUSIONS: This study demonstrates that ERalpha but not ERbeta mediates the beneficial effect of E(2) on reendothelialization and potentially the prevention of atherosclerosis.

Effect of single and compound knockouts of estrogen receptors alpha (ERalpha) and beta (ERbeta) on mouse reproductive phenotypes.

The functions of estrogen receptors (ERs) in mouse ovary and genital tracts were investigated by generating null mutants for ERalpha (ERalphaKO), ERbeta (ERbetaKO) and both ERs (ERalphabetaKO). All ERalphaKO females are sterile, whereas ERbetaKO females are either infertile or exhibit variable degrees of subfertility. Mast cells present in adult ERalphaKO and ERalphabetaKO ovaries could participate in the generation of hemorrhagic cysts. Folliculogenesis proceeds normally up to the large antral stage in both ERalphaKO and ERbetaKO adults, whereas large antral follicles of ERalpha+/-ERbetaKO and ERalphabetaKO adults are markedly deficient in granulosa cells. Similarly, prematurely developed follicles found in prepubertal ERalphaKO ovaries appear normal, but their ERalphabetaKO counterparts display only few granulosa cell layers. Upon superovulation treatment, all prepubertal ERalphaKO females form numerous preovulatory follicles of which the vast majority do not ovulate. The same treatment fails to elicit the formation of preovulatory follicles in half of the ERbetaKO mice and in all ERalpha+/-/ERbetaKO mice. These and other results reveal a functional redundancy between ERalpha and ERbeta for ovarian folliculogenesis, and strongly suggest that (1) ERbeta plays an important role in mediating the stimulatory effects of estrogens on granulosa cell proliferation, (2) ERalpha is not required for follicle growth under wild type conditions, while it is indispensable for ovulation, and (3) ERalpha is also necessary for interstitial glandular cell development. Our data also indicate that ERbeta exerts some function in ERalphaKO uterus and vagina. ERalphabetaKO granulosa cells localized within degenerating follicles transform into cells displaying junctions that are unique to testicular Sertoli cells. From the distribution pattern of anti-Müllerian hormone (AMH) in ERalphabetaKO ovaries, it is unlikely that an elevated AMH level is the cause of Sertoli cell differentiation. Our results also show that cell proliferation in the prostate and urinary bladder of old ERbetaKO and ERalphabetaKO males is apparently normal.

Retinoic acid receptor beta 2 (RAR beta 2) null mutant mice appear normal.

Vertebrates are highly sensitive to both retinoic acid (RA) deficiency and excess. The RA signal is thought to be transduced by nuclear receptors (the RAR and RXR families) which activate the expression of target genes via cis-acting transcriptional enhancer elements. Each of the three RAR genes, RAR alpha, RAR beta, and RAR gamma, gives rise to several isoforms by differential usage of two promoters and alternative splicing. RAR beta 2 is the most abundant of the four RAR beta isoforms, and its transcription is spatially and temporally restricted in developing embryos, suggesting that it might perform specific functions. Furthermore, RAR beta 2 expression can be induced via a retinoic acid response element located in its promoter region. This RA effect is particularly interesting since under conditions of RA excess, RAR beta 2 promoter activity and transcript accumulation are induced in regions of developing embryos in which malformations subsequently appear, such as the craniofacial region, the hindbrain, and the limbs. These findings have led to the suggestion that the RAR beta 2 isoform might mediate some of the teratogenic effects of RA. In this study, we have eliminated RAR beta 2 expression by targeted gene disruption. RAR beta 2 null mutants exhibit an apparently normal phenotype, indicating that other RARs must compensate for RAR beta 2 sufficiently well to allow normal prenatal and postnatal development to proceed. By challenging RAR beta 2 null embryos with teratogenic doses of RA, we have also directly addressed the question of whether RAR beta 2 is required for mediating RA-induced malformations.

Retinoic acid signal transduction pathways.

Complexity in the retinoid signaling system arises from a combination of several forms of retinoic acid which possess differential activities, multiple cytoplasmic binding proteins and nuclear receptors that have distinct ligand specificities and functional properties, and the existence of polymorphic retinoic acid response elements. Additional diversity appears to be generated by heterodimeric interactions between the two classes of nuclear retinoic acid receptors and between retinoic acid receptors and some other members of the nuclear receptor superfamily. Thus, a complex array of combinatorial effects is beginning to emerge that may account for the pleiotropic effects of retinoids.

Retinoic acid receptors and binding proteins in human skin.

Nuclear retinoic acid receptors (RAR) are likely to mediate many of the pleiotypic cutaneous actions of retinoids by acting as ligand-dependent enhancer factors. The presence of nuclear RAR in skin was confirmed by identification of a 45-kDa nuclear RA binding activity by fast protein liquid chromatography (FPLC). Analysis of RNA extracted from skin specimens demonstrated expression of RAR-alpha and RAR-gamma transcripts, as well as expression of the homologous low-affinity receptor, RXR-alpha. Both isoforms of RAR-gamma RAR-gamma 1 and RAR-gamma 2 were detectable, with RAR-gamma 1 being the more strongly expressed. FPLC analysis also demonstrated a 15-kDa peak of specific RA binding activity, consistent with the presence of cellular retinoic acid binding protein (CRABP). Of the two known forms of CRABP, CRABP-II was much more strongly expressed than CRABP-I at the level of steady-state mRNA. CRABP-II was also expressed in keratinocytes and fibroblasts in vitro. CRABP-II was up-regulated by agents that induce keratinocyte differentiation, and inhibited by prolonged exposure to high concentrations of RA. In contrast, CRABP-II was consistently induced by RA in dermal, but not in lung fibroblasts. CRABP-I was expressed at low to undetectable levels under all these conditions. The presence of tissue-specific and differentiation-related regulation of CRABP-II suggests that it may be an important regulator of RA action in human skin.

Retinoic acid receptors as regulators of human epidermal keratinocyte differentiation.

To examine the role of nuclear retinoic acid (RA) receptors (RARs) in the regulation of squamous differentiation in normal human epidermal keratinocytes (NHEK), we analyzed binding activity, mRNA expression, and transcriptional activity of the endogenously expressed RARs. Specific RA-binding activity eluted from size-exclusion HPLC with an apparent mol wt of 50 kilodaltons and was predominantly (greater than 95%) associated with the NHEK nuclear cell fraction. This RAR-binding activity represented in part the expression of RAR alpha and RAR gamma genes, whose transcripts were expressed in similar abundance in undifferentiated NHEK. Differentiation resulted in lower mRNA expression of RAR alpha relative to the mRNA expression of RAR gamma. Treatment of NHEK cells with 10(-6) M RA did not induce expression of RAR beta mRNA. Similarly, three squamous cell carcinoma cell lines derived from human skin and oral cavity expressed RAR alpha and RAR gamma transcripts, but not RAR beta transcripts. Transfection of NHEK with chloramphenicol acetyltransferase (CAT) reporter plasmids indicated that the endogenously expressed RARs could activate transcription through the RAR beta response element in a concentration-dependent manner with doses of 10(-9) M RA and higher. CAT expression was not activated through TRE, a palindromic thyroid hormone response element with purported RA responsiveness. The competitive binding of benzoic acid derivatives of RA to RAR correlated with the ability of each analog to suppress mRNA expression of the squamous cell markers, involucrin, type I transglutaminase, and SQ37, and to activate transcription of the RAR beta response element-CAT reporter. These results demonstrate that the control of NHEK differentiation by RA is consistent with the interaction of the retinoid with RAR and the regulation of transcription by that ligand-receptor complex.

Differential regulation of retinoic acid receptors and binding proteins in human skin.

Many of the pleiotropic effects of retinoids are likely to be mediated by nuclear retinoic acid receptors (RAR) acting as ligand-dependent enhancer factors. However, in previous studies we have been unable to document altered RAR expression at the RNA level in response to retinoic acid (RA) treatment or in psoriatic lesions, conditions characterized by marked alterations in keratinocyte proliferation and differentiation, which are either caused by or responsive to RA. In an attempt to identify other potential regulators of RA responsiveness, we have used RNA blot hybridization to study the expression of the cellular retinoic acid binding proteins (CRABP) CRABP-I and CRABP-II, the RAR-gamma isoforms RAR-gamma 1 and RAR-gamma 2, and the low-affinity RAR homologue RXR in normal, RA-treated, and psoriatic human epidermis. CRABP-II is selectively and markedly induced by RA in adult human skin (J Biol Chem 266:17662-17666, 1991). However, in submerged, serum-free keratinocyte cultures, CRABP-II mRNA could not be induced by RA. Comparisons of intact human skin, submerged keratinocyte cultures, and human skin equivalent cultures indicated that induction of CRABP-II by RA requires epidermal stratification, dermal-epidermal interactions, or both. CRABP-II transcripts were also expressed in heat-separated human dermis at levels similar to those found in epidermal keratome biopsies, whereas CRABP-I transcripts were undetectable in dermal RNA. CRABP-II transcripts were markedly elevated in psoriatic lesions, as they were in RA-treated skin. In contrast, CRABP-I mRNA was undetectable and not increased in psoriatic lesions. Expression of RAR-gamma isoforms and RXR was not detectably altered in either psoriatic lesions or in RA-treated skin. Thus, altered expression of CRABP-II appears more likely to regulate the cutaneous actions of RA than does altered expression of CRABP-I, RXR, or RAR-gamma isoforms. From these and other results, a model for regulation of RA action involving sequestration of RA by CRABP-II is proposed.

Chromosomal assignment of retinoic acid receptor (RAR) genes in the human, mouse, and rat genomes.

The human genes encoding the alpha and beta forms of the retinoic acid receptor are known to be located on chromosomes 17 (band q21.1:RARA) and 3 (band p24:RARB). By in situ hybridization, we have now localized the gene for retinoic acid receptor gamma, RARG, on chromosome 12, band q13. We also mapped the three retinoic acid receptor genes in the mouse, by in situ hybridization, on chromosomes 11, band D (Rar-a); 14, band A (Rar-b); and 15, band F (Rar-g), respectively, and in the rat, using a panel of somatic cell hybrids that segregate rat chromosomes, on chromosomes 10 (RARA), 15 (RARB), and 7 (RARG), respectively. These assignments reveal a retention of tight linkage between RAR and HOX gene clusters. They also establish or confirm and extend the following homologies: (i) between human chromosome 17, mouse chromosome 11, and rat chromosome 10 (RARA); (ii) between human chromosome 3, mouse chromosome 14, and rat chromosome 15 (RARB); and (iii) between human chromosome 12, mouse chromosome 15, and rat chromosome 7 (RARG).

Retinoic acid-induced glandular metaplasia in mouse skin is linked to the dermal expression of retinoic acid receptor beta mRNA.

The distribution of transcripts of nuclear (RAR alpha, RAR beta, and RAR gamma) and cytosolic (CRABP) retinoic acid receptors was analyzed in 13.5-day mouse embryo upper-lip skin, cultured in vitro for 48 hr with or without added retinoic acid. The results show a significant up-regulation of the transcription of the RAR beta gene concomitant with the initiation of an alteration of hair vibrissae follicle development, leading, after transfer for 8 days to the chick embryo chorioallantoic membrane, to an exocrine-type gland morphogenesis.

Retinoic acid receptor gene expression in human skin.

Human skin exhibits a characteristic, pleiotypic response to topical retinoic acid. In attempting to understand this response at the molecular level, we have used fast protein liquid chromatography (FPLC) and RNA blot hybridization to characterize the expression of the nuclear retinoic acid receptor (RAR) alpha, beta, and gamma genes in adult human epidermis. Size exclusion FPLC of 0.6 M NaCl nuclear extracts prepared from keratome biopsies revealed two peaks of specific [3H] retinoic acid (RA) binding at Mr 45 and 18 kDa, in agreement with the expected sizes of RAR and cellular RA binding protein. Blot hybridization analysis of total RNA extracted from keratome biopsies revealed that RAR-gamma was the predominant RAR species expressed in human epidermis, as RAR-alpha transcripts were detectable only at low levels and RAR-beta transcripts were undetectable. RAR transcripts were not induced by topical treatment with 0.1% RA cream under occlusion for 4 h or 4 d. Moreover, there was no significant difference in RAR-gamma transcript levels in normal and psoriatic epidermis. RAR-gamma transcripts were constitutively expressed not only in cultured human keratinocytes, but also in human dermal and lung fibroblasts. RAR-beta was induced by RA in dermal fibroblasts, but not in keratinocytes. RA induced IL-1 beta transcripts in keratinocytes rapidly (2 to 4 h) and at low concentrations (3 x 10(-10) M), consistent with activation of the IL-1 beta gene via RAR. These results demonstrate constitutive expression of RAR-gamma in human epidermis, and suggest that RAR-gamma is a molecular target of RA action in adult human skin.

Multiple isoforms of the mouse retinoic acid receptor alpha are generated by alternative splicing and differential induction by retinoic acid.

Together with the previously described mouse retinoic acid receptor alpha-1 (mRAR-alpha 1, formerly mRAR-alpha 0), we have isolated and characterized here a total of seven mRAR-alpha cDNA isoforms (mRAR-alpha 1 to alpha 7). These isoforms are generated from mRAR-alpha primary transcript(s) of a single gene by alternative splicing of at least eight different exons with the exon which encodes the amino acid sequence of their common B region. All of these isoforms differ in their 5'-untranslated regions (5'-UTRs) and, in the case of mRAR-alpha 1 and alpha 2, also in the sequences encoding the N-terminal A region which is known to be important for differential trans-activation by other members of the nuclear receptor superfamily. In addition, the sequences encoding the open reading frames (ORFs) of mRAR-alpha 3 and alpha 4 cDNA isoforms remain open to their very 5' ends, which suggests that these two isoforms may also encode RAR-alpha s with unique A region amino acid sequences. The two predominant isoforms, mRAR-alpha 1 and alpha 2, were found to be differentially expressed in mouse adult and fetal tissues, as well as in P19 and F9 embryonal carcinoma (EC) cell lines. Interestingly, the expression of mRAR-alpha 2, in contrast to that of the mRAR-alpha 1 isoform, was induced by retinoic acid (RA) in EC cells, thus suggesting the presence of two promoters in the 5' region of the mRAR-alpha gene, which differ in their response to RA. The conservation between mouse and human RAR-alpha 1 and alpha 2 cDNA isoform sequences, as seen by cross-hybridization in Southern blots or by DNA sequence analysis, together with their differential patterns of expression, strongly suggests that they perform specific functions during embryogenesis and in the adult.