The long term objective of this project is to elucidate the role retinoids play in controlling limb pattern and to understand their mechanisms of action. The effects of retinoids are believed to be transduced by specific nuclear receptors. At least four classes of such receptors are known (RARalpha, RARbeta, RARgamma and RXR). The specific developmental function of each of these receptors will be tested using retrovirus-mediated gene transfer into developing chick limb buds. Four types of experiments will be performed: misexpression of wild type receptors, expression of dominant-negative variants of the receptors, expression of constitutively active variants of the receptors and expression of chimeric receptors whose ligand-binding domain has been replaced by a homologous domain of a steroid receptor. In the last case, the receptor will be activated by implantation in vivo of a steroid- carrying substrate without affecting the expression of endogenous retinoid receptors. Differentially spliced variants exist for each class of retinoid receptors. All those expressed during limb development will be isolated from cDNA libraries and their relative functions will be explored. Retinoic acid is known to exist in a graded distribution across the limb bud, but the shape of the gradient is only very roughly known. Moreover, other retinoics are known to be able to activate retinoid receptors. In order to elucidate the distribution of all retinoid receptor-activating molecules, an easily assayable gene (encoding beta-galactosidase) will be expressed from a retinoid-dependent promoter (a promoter containing a retinoic acid response element). This construct will be transduced into developing chick limbs with a replication-competent retroviral vector.
