Transcriptional regulation by triiodothyronine requires synergistic action of the thyroid receptor with another trans-acting factor.

Human placental lactogen B (hCS-B) promoter activity is strongly stimulated by triiodothyronine (T3) in pituitary GC cells through interaction between the thyroid receptor and a thyroid receptor-binding element (TBE) spanning coordinates -67 to -41. This TBE is adjacent to the binding site for pituitary factor GHF1 (-95 to -68) which seems necessary for T3 stimulation of hCS-B promoter activity (M. L. Voz, B. Peers, A. Belayew, and J. A. Martial, J. Biol. Chem. 266:13397-13404, 1991). We here demonstrate actual synergy between the thyroid receptor and GHF1. Indeed, in placental JEG-3 cells devoid of factor GHF1, hCS promoter activity is barely stimulated by T3, while a strong response is observed in pituitary GC cells. In the latter, furthermore, neither the TBE nor the GHF1-binding site alone is sufficient to render the thymidine kinase promoter responsive to T3, while in combination they promote strong T3 stimulation. Close proximity between these sites is required for optimal synergy: T3 stimulation globally decreases with increased spacing. Furthermore, synergy occurs not only with a GHF1-binding site but also with all other factor recognition sequences tested (Sp1, NF1, CP1, Oct1, and CACCC boxes) and even with two other copies of the TBE. Nor is it specific to hCS TBE, since the palindromic sequence TCAGGTCA TGACCTGA (TREpal) also exhibits cooperativity.     

Characterization of a single strong tissue-specific enhancer downstream from the three human genes encoding placental lactogen.

The human genes coding for growth hormone (hGH) and placental lactogen (choriosomatomammotropic hormone [hCS]) are clustered on chromosome 17 in the following order: 5' hGH-N hCS-L hCS-A hGH-V hCS-B 3'. So far, a single placenta-specific enhancer has been identified in the locus, 2 kb downstream from the hCS-B gene, and shown to comprise one in vitro binding site for a nuclear protein. We here provide evidence that the hCS-B enhancer is more complex: (i) protection against DNase I digestion in the 3' flanking region of the hCS-B gene reveals four binding sites (DF-1, DF-2, DF-3, and DF-4) for nuclear proteins from either placental or HeLa cells, and (ii) placenta-specific enhancer activity can be fully exerted in transient expression experiments by a 126-bp fragment comprising the DF-3 and DF-4 protein-binding sites. By dissecting this region, we show that enhancer activity is mediated by a synergy between DF-3 and DF-4. Competitions with various oligonucleotides in footprinting and gel retardation experiments indicate that the same protein or set of proteins, different in HeLa and placenta cell nuclei, interacts with sites DF-2, DF-3, and DF-4. We also studied the regions of the hCS-L and hCS-A genes which are highly similar to the hCS-B enhancer. Although they each present the same four protein-binding sites, they exhibit only minor enhancer activity.     

Retinoic acid activation and thyroid hormone repression of the human alcohol dehydrogenase gene ADH3.

Mammalian alcohol dehydrogenase (ADH) catalyzes the oxidation of retinol to retinaldehyde, the rate-limiting step in the synthesis of retinoic acid. There exists a family of ADH isozymes encoded by unique genes, and it is unclear which isozymes are most important for regulation of retinoic acid synthesis during differentiation or development. A region in the human ADH3 promoter from -328 to -272 base pairs was shown previously to function as a retinoic acid response element (RARE), prompting an hypothesis for a positive feedback mechanism controlling retinoic acid synthesis (Duester, G., Shean, M. L., McBride, M. S., and Stewart, M. J. (1991) Mol. Cell. Biol. 11, 1638-1646). The ADH3 RARE contains three direct AGGTCA repeats which constitute the critical nucleotides of RAREs present in other genes. We dissected the ADH3 RARE and determined that receptor binding as well as transactivation are dependent upon only the two downstream AGGTCA motifs separated by 5 base pairs, a structure noticed previously for a RARE in the promoter for the retinoic acid receptor beta (RAR beta) gene. ADH3 and RAR beta RAREs functioned similarly in transfection assays, suggesting that the feedback mechanisms controlling ADH3 and RAR beta utilize a common RARE. We also found that the normal functioning of the ADH3 RARE was abrogated by thyroid hormone receptor in the presence of thyroid hormone. A negative thyroid hormone response element in the human ADH3 promoter was found to colocalize with the RARE. Since ADH production in rat liver is known to be repressed by thyroid hormone, these findings suggest that human ADH production may also be subject to thyroid hormone repression and that the mechanism involves an interference with retinoic acid induction.     

The human M creatine kinase gene enhancer contains multiple functional interacting domains.

Cis-elements (-933 to -641) upstream of the human M creatine kinase gene cap site contain an enhancer that confers developmental and tissue-specific expression to the chloramphenicol acetyltransferase gene in C2C12 myogenic cells transfected in culture. Division of the enhancer at -770 into a 5' fragment that includes the MyoD binding sites (-933 to -770) and a 3' fragment that includes the MEF-2 binding site (-770 to -641) resulted in two subfragments that showed minimal activity but in combination interacted in a position- and orientation-independent fashion to enhance activity of the SV40 promoter in transient transfection experiments. A 5' enhancer construct (-877 to -832) including only one (the low affinity) MyoD binding site was active when present in multiple copies. In contrast, a 3' enhancer construct (-749 to -732) including the MEF-2 binding site was inactive even when present in multiple copies. However, if the 5' construct was extended to include the high-affinity MyoD binding site (-877 to -803) the 5' and 3' constructs interacted in a position- and orientation-independent fashion to activate the SV40 promoter. Thus, the human M creatine kinase enhancer comprises multiple functional interacting domains.     

Specific epidermal growth factor receptors on porcine and human thyroid membranes

Specific, high affinity, saturable receptors for epidermal growth factor (EGF) have been demonstrated both on porcine and on human thyroid membranes. The binding affinities of porcine (Ka 3.0 X 10(-9) M) and human thyroid EGF receptors (Ka 1.75 X 10(-9) M) are very similar. TSH does not inhibit the binding of 125I-EGF to either membrane. These results suggest the possibility that EGF may be involved in the regulation of human as well as porcine thyroid follicular cell growth and function.     

Functional activity and regulation of human beta 2-adrenergic receptors expressed in Xenopus oocytes

The recently cloned human beta-adrenergic cDNA and several mutated forms have been expressed in Xenopus laevis oocytes by injection of RNA made from the cDNA under the control of the bacteriophage SP6 promoter. The cDNA and gene of the beta 2-adrenergic receptor possess the unusual feature of having a second upstream ATG (-101 base pairs) and a 19-codon open reading frame 5' to the initiator methionine codon of the receptor (Kobilka, B. K., Dixon, R. A. F., Frielle, T., Dohlman, H. G., Bolanowski, M., Sigal, I. S., Yang-Feng, T. L., Francke, U., Caron, M. G., and Lefkowitz, R. J. (1987) Proc. Natl. Acad. Sci. U.S.A. 84, 46-50). RNA lacking this upstream AUG and open reading frame was translated approximately 10-fold more efficiently both in an in vitro rabbit reticulocyte system and in oocytes. Injected oocytes but not water injected controls expressed typical beta 2-adrenergic receptors as assessed by ligand binding (450 fmol/mg membrane protein) and catecholamine-stimulated adenylate cyclase (approximately 20 fold). Moreover, these receptors displayed typical agonist-induced homologous desensitization when oocytes were incubated with isoproterenol at room temperature for 3-24 h. Among a series of mutations, truncations of the membrane-anchored core of the receptor eliminated receptor binding and cyclase stimulating activity. In contrast, disruption of one of the cAMP-dependent protein kinase phosphorylation sites or removal of the serine/threonine-rich carboxyl terminus had little or no effect on these functions or on the extent of agonist-induced desensitization relative to that observed with native receptor. These studies validate the beta 2-adrenergic nature of the cloned human beta-adrenergic cDNA, document the utility of the Xenopus oocyte system for studying functional and regulatory properties of receptors coupled to adenylate cyclase, and suggest the possibility that elements in the 5' untranslated region of the beta 2-adrenergic receptor RNA may regulate its translation in vivo.