Search

We have engineered a mouse mutation that specifically deletes most of the RXR alpha N-terminal A/B region, which includes the activation function AF-1 and several phosphorylation sites. The homozygous mutants (RXR alpha af1(o)), as well as compound mutants that further lack RXR beta and RXR gamma, are viable and display a subset of the abnormalities previously described in RXR alpha-null mutants. In contrast, RXR alpha af1(o)/RAR(-/-)(alpha, beta or gamma) compound mutants die in utero and exhibit a large array of malformations that nearly recapitulate the full spectrum of the defects that characterize the fetal vitamin A-deficiency (VAD) syndrome. Altogether, these observations indicate that the RXR alpha AF-1 region A/B is functionally important, although less so than the ligand-dependent activation function AF-2, for efficiently transducing the retinoid signal through RAR/RXR alpha heterodimers during embryonic development. Moreover, it has a unique role in retinoic acid-dependent involution of the interdigital mesenchyme. During early placentogenesis, both the AF-1 and AF-2 activities of RXR alpha, beta and gamma appear to be dispensable, suggesting that RXRs act as silent heterodimeric partners in this process. However, AF-2 of RXR alpha, but not AF-1, is required for differentiation of labyrinthine trophoblast cells, a late step in the formation of the placental barrier.

Retinoic acid, one of the principle active metabolites of vitamin A (retinol), is believed to be essential for numerous developmental and physiological processes. Vitamin A deprivation (VAD) during development leads to numerous congenital defects. Previous studies of retinoic acid receptor (RAR) deficient mice failed to reveal any of these VAD-induced defects. This finding suggested that either the RARs are functionally redundant or that they are not critically required during development. In order to address these possibilities, we derived a number of RAR compound mutants. Unlike RAR single mutants, these compound null mutants died either in utero or shortly following birth. Histological analysis revealed essentially all of the defects characteristic of fetal VAD. A number of additional malformations, not described in previous VAD studies, were also observed. These included defects of the ocular and salivary glands and their ducts, the skeletal elements of the fore- and hindlimbs, and the cervical region of the axial skeleton. In addition, with the exception of derivatives forming within the first pharyngeal arch, most of the elements derived from mesectoderm emanating from cranial and hindbrain levels were affected. A number of these mutants also exhibited supernumerary cranial skeletal elements characteristics of the reptilian skull. A summary of the defects found in these RAR double mutants is presented.

We have disrupted the CRABPII gene using homologous recombination in embryonic stem cells, and shown that this disruption results in a null mutation. CRABPII null mutant mice are essentially indistinguishable from wild-type mice as judged by their normal development, fertility, life span and general behaviour, with the exception of a minor limb malformation. Moreover, CRABPI-/-/CRABPII-/- double mutant mice also appear to be essentially normal, and both CRABPII-/- single mutant and CRABPI-/-/CRABPII-/- double mutant embryos are not more sensitive than wild-type embryos to retinoic acid excess treatment in utero. Thus, CRABPI and CRABPII are dispensable both during mouse development and adult life. Our present results demonstrate that CRABPs are not critically involved in the retinoic acid signaling pathway, and that none of the functions previously proposed for CRABPs are important enough to account for their evolutionary conservation.

Compound null mutations of retinoic acid receptor (RAR) genes lead to lethality in utero or shortly after birth and to numerous developmental abnormalities. In the accompanying paper (Lohnes, D., Mark., M., Mendelsohn, C., Dollé, P., Dierich, A., Gorry, Ph., Gansmuller, A. and Chambon, P. (1994). Development 120, 2723-2748), we describe malformations of the head, vertebrae and limbs which, with the notable exception of the eye defects, were not observed in the offspring of vitamin A-deficient (VAD) dams. We report here abnormalities in the neck, trunk and abdominal regions of RAR double mutant mice, which include: (i) the entire respiratory tract, (ii) the heart, its outlow tract and the great vessels located near the heart, (iii) the thymus, thyroid and parathyroid glands, (iv) the diaphragm, (v) the genito-urinary system, and (vi) the lower digestive tract. A majority of these abnormalities recapitulate those observed in the fetal VAD syndrome described by Joseph Warkany's group more than fourty years ago [Wilson, J. G., Roth, C. B. and Warkany, J. (1953) Am. J. Anat., 92, 189-217; and refs therein]. Our results clearly demonstrate that RARs are essential for vertebrate ontogenesis and therefore that retinoic acid is the active retinoid, which is required at several stages of the development of numerous tissues and organs. We discuss several possibilities that may account for the apparent functional redundancy observed amongst retinoic acid receptors during embryogenesis.

A null mutation was generated in the mouse RXR alpha gene by targeted disruption. Growth deficiency occurred in heterozygote mice. Null mutants died in utero and displayed myocardial and ocular malformations. These malformations belong to the fetal vitamin A deficiency syndrome, supporting the idea that RXR alpha is involved in retinoid signaling in vivo. A phenotypic synergy was observed when the RXR alpha mutation was introduced into RAR alpha or RAR gamma mutant backgrounds: RXR alpha null mutants and RXR alpha +/-/RAR gamma-/- double mutants displayed similar ocular defects, which became more severe in RXR alpha-/-/RAR gamma+/- and RXR alpha-/-/RAR gamma-/- mutants. Furthermore, RXR alpha/RAR double mutants exhibited several malformations not seen in single mutants. This functional convergence strongly suggests that RXR alpha/RAR heterodimers mediate retinoid signaling in vivo.

That both deficiency and excess of vitamin A lead to a wide spectrum of congenital abnormalities has strongly implicated the active metabolite, retinoic acid (RA), in normal embryonic development. There are 3 families of RA receptors (RARs), RAR alpha, RAR beta and RAR gamma, each having at least two isoforms derived from primary transcripts initiated at two promoters P1 and P2 (reviewed in Leid et al., 1992) Transcripts encoding all 4 isoforms of RAR beta (RAR beta 1 to RAR beta 4) accumulate in embryonal carcinoma (EC) cells in the presence of RA. It has been previously shown that the RA modulation of RAR beta 2/beta 4 transcripts is achieved at the level of transcriptional initiation via a RA response element (RARE) present in the P2 RAR beta 2/beta 4 promoter. In contrast, the mechanism by which RA up-regulates RAR beta 1/beta 3 transcripts has not yet been elucidated. We describe here the isolation of the P1 RAR beta 1/beta 3 promoter and characterization of its activity in transgenic animals. We find that RAR beta 1/beta 3 promoter activity, which is apparently confined to the embryonic CNS, is not modified by RA treatment, unlike that of the RAR beta 2/beta 4 promoter (Mendelsohn et al., 1991). Nuclear run-on transcription analysis in EC cells supports the conclusion that RAR beta 1/beta 3 transcript initiation is not modulated by RA, and that the RA-induced accumulation of RAR beta 1/beta 3 transcripts occur via a RA-dependent release of a block in RNA chain elongation.

Retinoids, in particular all-trans retinoic acid (T-RA), are essential for normal development and homeostasis of vertebrates. Although many effects of retinoids, particularly with regard to teratogenicity, have been described in the literature, the mechanisms by which these simple signalling molecules work has only recently begun to be elucidated. We now recognize at least two classes of retinoid-binding proteins and two families of retinoid receptors. The ultimate interpretation of the retinoid signal within a given cell is probably the result of a complex series of interactions between these proteins, yet little is understood concerning the role each member of this signalling pathway plays. It is therefore imperative to dissect the molecular mechanisms which transduce the effects of these ligands, both in vivo and in isolated systems. One approach we are employing is gene targeting of retinoic acid receptors (RARs) and cellular retinoid-binding proteins to generate mice in which one or more of these genes has been functionally inactivated.