- Email: [email protected]
Triiodothyronine inhibits transcription from the human growth hormone promoter
261
Molecular and Cellular Endocrinology, 71 (1990) 261-267 Elsevier Scientific Publishers Ireland, Ltd.
MOLCEL 02314
Triiodothyronine inhibits transcription from the human growth hormone promoter A. Morin ‘, J. Louette
2, M.L.J. Voz 2, A. Tixier-Vidal
‘, A. Belayew
2 and J.A. Martial
2
’ UAO 411 I5 CNRS, Laboratoire de Neuroendocrinologie Cellulaire, Collsge de France, F-75231 Paris Cedex, France, and ’ Laboratoire de G&tie G&tique,
CJniversitJ de Li.?ge, Institut de Chimie B6, B-4000 Sart-Tilman,
Belgium
(Received 26 January 1990; accepted 19 April 1990)
Key words: CC cell; hGH-reporter gene; hGH-CAT;
Transient expression; Triiodothyronine dose-response
Three DNA constructs, the natural human growth hormone gene (hGH-hGH) its 500 bp promoter linked to the chloramphenicol acetyl transferase reporter gene (hGH-CAT), and its structural part linked to the herpes virus thymidine kinase promoter (TK-hGH) were introduced into rat pituitary GC cells by DEAE-dextran transfection. Transient expression was followed as a function of triiodothyronine (T3) concentration. The hGH-CAT expression was specifically inhibited by T3 following a typical dose-response curve while hGH-GH gene expression was not significantly modified. The transient expression of TK-hGH increased as a function of T3 concentration. These results indicate that T3 exerts two opposite effects on hGH gene expression. First, it down-regulates expression by acting on the promoter; second, it up-regulates expression by acting on the structural part of the gene. These action could be due to regulation of transcription and mRNA stabilization, respectively.
Introduction The thyroid hormone L-3,5,3’-triiodothyronine (T3) regulates the expression of several genes, mostly at the transcriptional level where its action is mediated by its binding to a specific nuclear receptor related to the c-erb-A oncogene (reviewed by Oppenheimer et al., 1987; Samuels et al., 1988). The expression of the rat growth hormone (rGH) gene is specifically stimulated by Tj (Samuels et al., 1976; Martial et al., 1977). This transcriptional regulation has been most extensively studied in
Address for correspondence: Joseph A. Martial, Laboratoire Central de G&e GCnttique, Institut de Chimie B6, B-4000 Sart-Tiian, Belgium. 0303-7207/90/$03.50
the rat pituitary GC cell line (reviews: Davis et al., 1988; Samuels et al., 1988). Various deletion mutants of the rGH promoter have been linked to reporter genes and the resulting fusion genes transfected into GC cells. T3 induction of the reporter genes was analyzed either in stable transformants or during transient expression. Functional thyroid response elements (TRE) defined in this way were compared to T3 receptor binding sites tentatively located in vitro on the promoter DNA. A close association was found between the regulatory elements involved in the tissue-specific expression and in the thyroid hormone regulation of the rGH gene (Ye et al., 1988) making it difficult to locate the TRE. However, synthetic oligonucleotides could be derived from the rGH
6 1990 Ekevier Scientific Publishers Ireland, Ltd.
262
promoter sequence (Glass et al., 1987; Koenig et al., 1987; Brent et al., 1989) that could transfer T3 regulation to the thymidine kinase promoter in transfected cells and bind the Ts receptor in vitro. The conclusions drawn about the T3 stimulation of the rat GH gene transcription could not be extended to the homologous human GH gene for which Cattini et al. (1986a) demonstrated an inhibition of expression at physiological T3 concentration. In this experiment, GC cells had been stably transformed by a 2.6 kb EcoRI genomic fragment comprising the hGH gene with 496 bp of 5’ flanking sequence. In a subsequent work Cattini and Eberhardt (1987) constructed a hGH-CAT fusion gene containing 470 bp of promoter sequences. This region contains all the sequences necessary for tissuespecific expression, allowing the analysis of its biological activity in pituitary cells (Cattini et al., 1986b; Bodner and Karin, 1987; Lefevre et al., 1987). When hGH-CAT transient expression was assayed after transfection in GC cells, it showed little or no response to 10e8 M Tj treatment (Cattini and Eberhardt, 1987; Brent et al., 1988). The latter results were surprising since some of us had previously shown by DNA-cellulose binding competition (Barlow et al., 1986) that the only region in the whole hGH gene where a specific interaction for the T,-receptor complex could be found in vitro was located between coordinates - 290 and - 129. It could thus be expected that any transcriptional regulation of hGH gene expression by Ts had to involve this promoter region. To resolve these ambiguous data, we have analyzed the transient expression of three distinct hGH constructs after transfection of GC cells: the hGH promoter linked to the CAT reporter gene (hGH-CAT); the natural hGH gene (hGH-hGH); the neutral TK promoter linked to the hGH structural gene (TK-hGH). When expression was monitored as a function of T3 concentration, typical dose-response curves were found corresponding to a maximal 2-fold inhibition of hGH-CAT gene expression at lop9 M Tj and a 2-fold stimulation of TK-hGH gene expression. The transient expression of the natural hGH-hGH gene in these GC cells was not significantly altered by increasing Tj concentrations.
Materials and methods Plasmid constructions A 2.6 kb EcoRI fragment comprising 496 bp of promoter region, the whole structural part of the previously cloned human growth hormone gene (hGH,; Eliard et al., 1985) and 528 bp past the poly A addition site was subcloned in the plasmid vector pBR322, yielding hGH-hGH. A BamHI restriction site is located at the origin of transcription. The hGH-CAT fusion gene was kindly provided by M. Karin. It was constructed by insertion of the 500 bp EcoRI/BamHI hGH promoter fragment and the CAT structural gene (Gorman et al., 1982) into a pUC polylinker. The TK-hGH fusion gene was made by Selden et al. (1986) by ligating a PuuII-BglII fragment comprizing 200 bp of TK promoter and 56 bp of sequences downstream from the transcription initiation site to the hGH structural part. It was purchased as part of an Allegro assay system (Nichols Institute, Los Angeles, CA, U.S.A.). The TK-CAT fusion gene with 105 bp of TK promoter was present in the pBLCAT2 vector (Luckow and Schutz, 1987) kindly provided by G. Schutz. Tissue culture Rat pituitary tumor GC cells were grown in Ham F12 medium supplemented with 15% horse serum and 2.5% fetal calf serum. Sera were depleted of steroid and thyroid hormones by successive treatments with charcoal-dextran (Laverrilre et al., 1986) and AG l-X8 resin (Samuels et al., 1979). DEAE-dextran transfection The transfection schedule was the following: on day 1, 1.5 x lo6 cells were plated per 35 mm dish (6-well plates, Nunc) in hormone-free medium. On day 4 the medium was changed. On day 5 the cells were transfected. We essentially used the DEAEdextran method of Kopchick and Stacey (1984), except that the transfected 3 pg of supercoiled plasmid DNA/plate, and left the DNA for only 30 min on the cells. The fresh medium added after transfection already contained L-triiodothyronine (T3) (Na salt, 98% pure, Sigma) when needed. 10e3 M T3 stock solution was made in 80%
263
ethanol/20% NaOH 0.1 M and diluted into culture medium. Medium was changed on day 6 and cells were harvested on day 7, i.e. 44 h after transfection. RNA extraction and analysis Total RNA was prepared by dissolving cells in a 5 M guanidinium thiocyanate solution and selectively precipitating RNA by addition of 4 M LiCl according to Cathala et al. (1983). Total RNA was quantified by a method (Caverriere et al., 1989) based on measures of the fluorescence emitted by RNA/ ethidium bromide complexes. Serial 3-fold dilutions starting from 16 to 3 pg of each RNA sample were spotted onto nitrocellulose and hybridized to an excess of 32P-labelled probes (see below) as described by Thomas (1980). After autoradiography, individual spots were cut and distributed into separate scintillation vials and their radioactivity measured. Potential DNA contamination of RNA samples was monitored by DNase RQl digestion. Probes were the cDNA inserts of the phGH cDNA (Martial et al., 1979) and prGH cDNA (Seeburg et al., 1977) plasmids that were labelled by nick translation (Maniatis et al., 1975) with [“P]dCTP or dATP (400 Ci/mmol; Amersham) to a specific activity of l-5 X lo8 cpm/pg. CA T assay Assays were performed essentially according to Gorman et al. (1982) with the following modifications: a preliminary 6 min heat inactivation of endogenous acetylases at 65” C and an overnight incubation of the cell extract in the presence of 4 mM acetyl coenzyme A (Sigma) and 0.25 FCi of [‘4C]chloramphenicol (40 Ci/mmol, Amersham). Chlor~phe~~l and its acetylated forms were extracted in ethyl acetate and separated by thinlayer chromatography (TLC). Following autoradiography the silica gel spots from the TLC plates containing either acetylated products or remaining substrate were scraped into separate scintillation vials and counted. Standardization All data involving transfected genes are given per dish, not taking into account the various degrees of cell proliferation induced by T3 concentra-
tion, as none of the plasmids used can replicate in these cells. However, endogenous rGH gene expression was corrected for cell proliferation using the amount of total RNA extracted as a measure of cell population (see Flug et al., 1987 for a detailed discussion). Computer work The sequence analysis software package was from the University of Wisconsin Genetics Computer Group (Devereux et al., 1984). It was run on a Digital VAX 11/780. Results T3 regulation of hGfi-CA T The hGH gene promoter defined as a 500 bp EcoRI-BamHI fragment has been linked to the chloramphenicol acetyl transferase (CAT; Gorman et al., 1982) reporter gene to yield the hGHCAT hybrid gene that can direct proper transcript initiation (Lefevre et al., 1987). In order to study the T3 regulation of hGH-CAT expression, this construct was introduced into rat pituitary GC cells by DEAE-dextran transfection. Cells were grown a further 44 h in culture medium supplemented with various amounts of T3. CAT activity was assayed in the cell homogenates according to Gorman et al. (1982) and plotted as a function of
-12
-11
-10
-9
-8
Log T3 concentration Fig. 1. Transient expression of hGH-CAT in transfected Gc cells as a function of Ts concentration in the culture medium. CAT enzymatic activity was measured according to Gorman et al. (1982) in cells transfected by the DEAE-dextran method. Cells were harvested 44 h after transfection (for details, see Materials and Methods). Each point determined at least in triplicate is drawn with its standard error.
264
T3 concentration (Fig. 1). A T3 dose-dependent inhibition of hGH-CAT transient expression was observed with the maximal effect (2-fold inhibition) reached at 10m9 M T3. 7; reguI~tion of the hGH gene
A 2.4 kb EcoRI genomic fragment comprising the hGH gene with 496 bp of 5’ flanking DNA and 528 bp of sequences 3’ to the poly A addition site was subcloned into the plasmid vector pBR322 yielding hGH-hGH. The transcripts of such a construct are correctly initiated (Cattini et al., 1986a). This ‘natural’ hGH-hGH gene was introduced into rat pituitary GC cells in culture by DEAE-dextran transfection and its transient expression was monitored as a function of the T3 concentration in the medium. Total RNA extracted 44 h after transfection was analyzed by dot-blot hybridization using 32Plabelled hGH cDNA as a probe. The amount of hGH mRNA found (Fig. 2) was not sig~fic~tly altered by increases in Ts concentration from lo-” to lo-’ M. The endogenous rat GH gene induction was followed in parallel in the same cells. A typical dose-response curve was obtained (Fig. 2), with a maximal (i-fold stimulation reached at physiological T3 concentrations (lop9 M), indicating a normal response of these GC cells to the hormone (Samuels and Shapiro, 1976; Martial et al., 1977). No cross-hybridization between the rat
I---+
Fig. 2. Transient expression of the hGH gene in transfected GC ceils as a function of Ts concentration in the culture medium. Cells were harvested 44 h after DRAGdextrau transfection and total RNA was extracted. Dot-blot hybridization to hGH cDNA or rat GH cDNA probes distinguishes expression from the exogenous or endogenous GH genes (for details, see Materials and Methods). Each point determined at least in triplicate is drawn with its standard error.
Fig. 3. Transient expression of TK-hGH in transfected GC cells as a function of Ts concentration in the culture medium. Cells were harvested 44 h after DRAE-dextran transfection to hGH cDNA or rat GH cDNA probes distinguishes expression from the exogenous or endogenous GH genes (for details, see Materials and Methods). Each point determined at least in triplicate is drawn with its standard error.
and human GH mRNAs/cDNAs took place in our experiments. Indeed, RNA extracted from mock-transfected rat pituitary cells only gave a background signal when probed with the human GH cDNA (data not shown). T3 regulation of TK-GH
Since 7; inhibition linked to the hGH promoter as seen in hGH-CAT could not be observed in the transient expression of the natural gene, we reasoned that its structural part could be responsible for a positive T3 response. To test this hypothesis, we used the pTK-hGH construct where the hGH gene structural part was fused to the herpes simplex thymidine kinase (TK) promoter (Selden et al., 1986). GC cells were transfected with this fusion gene by the DEAE-dextran method. The transient expression of pTK-hGH was similarly analyzed as a function of T3 concentration in the culture medium. RNA expressed from pTK-hGH accumulated in the cells (Fig. 3) when T3 concentration was raised, reaching a plateau (2-fold M. The endogenous induction) at 10-i’ to lo-’ rat GH mRNA induction was followed in parallel in the same cells (Fig. 3). A control pTK-CAT construct (pBLCAT2, Luckow and Schlitz, 1987) was assayed for transient expression in the presence or absence of 10e8 M T3 and showed a non-specific 1.2- to l.Cfold induction (data not shown) in agreement with published data (e.g., Glass et al., 1987).
Discussion In order to study the role of Ts on the expression of the hGH gene, we have assayed three constructs: the 500 bp hGH promoter linked to the CAT reporter gene (hGH-CAT), the natural hGH gene with the same 500 bp of promoter sequences (hGH-hGH) and the neutral thymidine kinase promoter linked to the hGH structural gene (TK-hGH). Transient expression of the three constructs was followed as a function of T3 concentration after introduction into GC cells. No significant effect was seen for the natural hGH gene expression while hGH-CAT and TK-hGH were respectively down- and up-regulated, to a similar extent. We conclude that T3 exerts two opposite effects on hGH gene expression. T3 effect through the hGH gene promoter We have shown in this paper that T3 down-regulates expression from the 500 bp hGH promoter under conditions that lead to a 6-fold induction of the rat gene expression. This T3 inhibition of transient hGH-CAT expression has been recently confirmed by electroporation of GC cells (Voz et al., in preparation). Previous data from other groups (Cattini and Eberhardt, 1987; Brent et al., 1988) showing a slight Ts inhibition or no effect on hGH promoter activity can be explained by different hormonal conditions. Indeed, a much lower T3 induction of the control rGH gene (1.5to 2-fold) is reported in these papers suggesting sub-optimal TX responses. Could the different T3 response observed with the hGH and rGH promoters be mediated by the same TRE? Some of us have previously shown that a human T3 receptor could specifically interact with a fragment located from - 290 to - 129 (Barlow et al., 1986) in the hGH promoter. Most of the many publications that have described positive regulatory elements involved in T3 regulation of rGH promoter have located them between coordinates - 208 and - 164 (Glass et al., 1987; Koenig et al., 1987; Wight et al., 1988; Ye et al., 1988; Brent et al., 1989). When comparing the rat and human GH promoters, and introducing gaps to maximize similarity (Fig. 4) the best conserved sequence within this region is found between nucleotides
hGH
rGH
TRE
related
SPl
GHFl I
Fig. 4. Sequence comparison of the human and rat GH proximal promoters. Sequences have been aligned and gaps have been introduced to maximize similarity using the ‘Gap’ program from the University of Wisconsin Genetics Computer Group. The boxes refer to cis elements involved in the transcriptional regulation of the human GH (upper line) and rat GH (lower line) genes. Two binding sites for the pituitary specific factor GHF-l/GCZ/GCl (Catanzaro et al., 1987; Lefevre et al., 1987; West et al., 1987; Ye et al., 1988) are shown as striped boxes on each promoter. The dotted boxes stand for SPI binding sites (Lefebre et al., 1987; Lemaigre et al., 1989). A sequence that is identical in both promoters, located in a region involved in Ts regulation of rat GH expression is outlined as a black box. Numbers refer to coordinates of the cis elements and gap boundaries, starting from the transcription initiation site taken as + 1. The introduced gap sizes are given in bp.
- 188 and - 173 in hGH and - 206 to - 191 in rGH. This location in hGH is 17 bp closer to a region extending from -106 to -148 in both promoters that can bind a pituitary specific factor named GHF-l/GC2 (Lefevre et al., 1987; West et al., 1987; Ye et al., 1988). In the absence of GHF-1, the distal part of this site can interact with the ubiquitous SPl factor (Ye et al., 1988; Lemaigre et al., 1989). The Tj receptor binding sequences might overlap with the GHF-1 in hGH site thus preventing activation, as was described for glucocorticoid inhibition of a CAMP-responsive enhancer (Akerblom et al., 1988). Alternatively, an altered spacing between the Ts receptor and GHF-1 binding sites in the hGH promoter could prevent the type of interactions found in the rGH gene where tissue-specific and
266
T:, response elements are tightly linked (Ye et al., 1988). Finally, the T3 receptor could interact with the GHF-1 protein bound to its specific DNA site and lower its ability to induce transcription as was shown for the ~uc~orticoid and estrogen receptors on rat prolactin gene expression (Adler et al., 1988). In order to distinguish between these possibilities, deletion mutants of the hGH promoter are presently being tested for their ability to confer T3 inhibition to a CAT reporter gene (Voz et al., in preparation). TX effect through hGH sequences located 3’ from the transcription initiation site As our results on TK-hGH show, Ts exerts a positive effect through the hGH structure gene or 3’ flanking sequences. These data, together with a report on the role of glucocorticoids in hGH mRNA stabilization (Paek and Axel, 1987), indicate that the use of the hGH reporter gene as initiated by Selden et al. (1986) should be reconsidered with caution when studying hormonal regulation. We do not believe this TX regulation is transcriptional since, using human IM9 cell T3 receptor, we could not find in vitro binding downstream from the hGH transc~ption i~tiation site (Barlow et al., 1986). These results have not been refuted by the recent data of Sap et al. (1990) showing that the third rat GH intron includes a TRE responding to the chicken c-erb-A protein. Indeed no sequence similar to that region is present in the smaller third intron of the hGH gene. We therefore favour a post-transc~ptional mechanism as was shown for Ts on liver malic enzyme (Dozin et al., 1986) and spot 14 (Narayan and Towle, 1985) gene expression. Tj effect on the 2.6 kb hGH gene We have not found a significant effect of T3 on the hGH-GH gene expression. This could result from the addition of the two opposite effects described above. Nothing is known about the Tj regulation of the natural hGH gene in normal human pituitary cells. However, the results presented here are in agreement with a work of Isaacs et al. (1987) on
dispersed GH-secreting human pituitary adenoma cells where no significant effect of T3 on hGH mRNA levels could be demonstrated. Our results disagree with data from Cattini et al. (1986a) on GC cells stably transformed by the hGH gene. This discrepancy is most probably due to experimental conditions which could affect to different extents the two opposite effects of TJ. It should also be noted that the TX effect can differ according to the GH, subclone used for transfection as was shown for rat prolactin expression (Forman et al., 1988). Another explanation for this discrepancy might be found in the heterogenous T3 receptor population (Samuels et al., 1977; Cattini et al., 1988; Wight et al., 1988; Hodin et al., 1989) present in GH, cells. Maximal inhibition would be observed as soon as T3 is added to the cell culture medium. Progressive disappearance of the depletable TX receptor population in GC cells would then lead to a release of transcription in~bition. A short incubation in the presence of Ts such as what can be done with stable hGH transformants (Cattini et al., 1986a) would favour detection of the inhibitory role of the hormone. A longer incubation as done routinely in transient expression (this study) would allow a balance between the stimulatory and inhibitory effects of ?;.
Acknowledgements We would like to thank G. Schutz and M. Karin for kindly providing pBLCAT2 and phGHCAT respectively. We acknowledge the help of M.J. Marchand, M.L. Scippo, E. Bellefroid and C, Lambert in an initial phase of this work, J.N. Laverriere and G. Rousseau for discussions, E. Rosenbaum for maintaining cell cultures and J. Lejeune for expert secretarial assistance. A.M. held a post-doctoral fellowship from the CNRS (France, Direction des Relations et de la Cooperation Internationale). J.A.L. and M.L.J.V. held IRSIA pre-doctoral fellowships (Belgium). This work was partly supported by a grant from the SPPS (Belgium, Sciences de la Vie, BI0/15) and the FNRS (Belgium, FRSM 3.4577.86).
267
References Ackerblom, I.E., Slater, E.P., Beato, M., Baxter, J.D. and Mellon, P.L. (1988) Science 241, 350-353. Adler, S., Waterman, M.L., He, X. and Rosenfeld, M.G. (1988) Cell 52, 685-695. Barlow, J.W., Voz, M.L.J., Eliard, P.H., Mathy-Hartert, M., De Nayer, P., Economidis, I.V., Belayew, A., Martial, J.A. and Rousseau, G. (1986) Proc. Natl. Acad. Sci. U.S.A. 83, 9021-9025. Bodner, M. and Karin, M. (1987) Cell SO, 267-275. Brent, G.A., Hamey, J.W., Moore, D.D. and Larsen, P.R. (1988) Mol. Endocrinol. 2, 792-798. Brent, G.A., Larsen, P.R., Hamey, J.W., Koenig, R.J. and Moore, D.D. (1989) J. Biol. Chem. 264, 178-182. Catanzaro, D.F., West, B.L., Baxter, J.D. and Reudelhuber, T.L. (1987) Mol. Endocrinol. I, 90-96. Cathala, G., Savouret, J.F., Mendez, B., West, B.L., Karin, M., Martial, J.A. and Baxter, J.D. (1983) DNA 2, 3299335. Cattini, P.A. and Eberhardt, N.L. (1987) Nucleic Acids Res. 15, 1297-1309. Cattini, P.A., Anderson, T.R., Baxter, J.D., Mellon, P. and Eberhardt, N.L. (1986a) J. Biol. Chem. 261, 13367-13372. Cattini, P.A., Peritz, L.N., Anderson, T.R., Baxter, J.D. and Eherhardt, N.L. (1986b) DNA 5, 503-509. Cattini, P.A., Kardami, E. and Eberhardt, N.L. (1988) Mol. Cell. Endocrinol. 56, 2633270. Davis, J.R.E., Belayew, A. and Sheppard, M.C. (1988) in Molecular Biology of Endocrinology (Baillieres Clinical Endocrinology and Metabolism), Vol. 2, pp. 797-834, W.B. Saunders, London. Devereux, J., Haeberli, P. and Smithies, 0. (1984) Nucleic Acids Res. 12, 387-395. Dozin, B., Magnuson, M.A. and Nikodem, V.M. (1986) J. Biol. Chem. 261, 10290-10292. Eliard, P.H., Marchand, M.J.. Rousseau, G.G., Formstecher, P.. Mathy-Hartert, M., Belayew, A. and Martial, J.A. (1985) DNA 4, 409-417. Flug, F., Copp, R.P., Casanova, J., Horowitz, Z.D., Janocko, L.. Plotnick, M. and Samuels, S.H.H. (1987) J. Biol. Chem. 262, 6373-6382. Forman, B.M., Chang-Ren, Y., Frederick, S., Casanova, J. and Samuels, H.H. (1988) Mol. Endocrinol. 2, 902-911. Glass. C.K., France, R., Weinberger, C., Albert, V., Evans, R.M. and Rosenfeld, M.G. (1987) Nature 329, 738-741. Gorman, CM., Moffat, L.F. and Howard, B.H. (1982) Mol. Cell. Biol. 2, 10441051. Hodin, R.A., Lazar, M.A., Wintman, B.I., Darling, D.S., Koenig, R.J., Larsen, P.R., Moore, D.D. and Chin, W.W. (1989) Science 244, 76-79. Isaacs, R.E., Findell, P.R., Mellon, P., Wilson, C.B. and Baxter, J.D. (1987) Mol. Endocrinol. 1, 569-576.
Koenig, R.J., Brent, G.A., Warner, R.L., Larsen, P.R. and Moore, D.D. (1987) Proc. Natl. Acad. Sci. U.S.A. 84, 5670-5674. Kopchick, J.J. and Stacey, D.W. (1984) Mol. Cell. Biol. 4, 240-246. Laverriere, J.N., Muller, M., Buisson, N., Tougard, C., TixierVidal, A., Martial, J.A. and Gourdji, D. (1986) Endocrinology 118, 198-206. Laverriere, J.N., Richard, J.L., Buisson, N., Martial, J.A., Tixier-Vidal, A. and Gourdji, D. (1989) Neuroendocrinology 50, 693-701. Lefevre, C., Imagawa, M., Dana, S., Grindlay, J., Bodner, M. and Karin, M. (1987) EMBO J. 6, 971-981. Lemaigre, F.P., Courtois, S.J., Lafontaine, D.A. and Rousseau, G.G. (1989) Eur. J. B&hem. 181, 555-561. Luckow, B. and Schutz, G. (1987) Nucleic Acids Res. 15, 5490. Maniatis, T., Jeffrey, A. and Kleid, D.G. (1975) Proc. Natl. Acad. Sci. U.S.A. 72, 1184-1188. Martial, J.A., Baxter, J.D., Goodman, H.M. and Seeburg, P.H. (1977) Proc. Natl. Acad. Sci. U.S.A. 74, 1816-1820. Martial, J.A., Hallewell, R.A., Baxter, J.D. and Goodman, H.M. (1979) Science 205, 602-607. Narayan, P. and Towle, H.C. (1985) Mol. Cell. Biol. 5, 26422646. Oppenheimer, J.H. (1987) in Molecular Basis of Thyroid Hormone Action (Oppenheimer, J.H. and Samuels, H.H., eds.), pp. l-35, Academic Press, New York. Paek, I. and Axel, R. (1987) Mol. Cell. Biol. 7, 1496-1507. Samuels, H.H. and Shapiro, L.E. (1976) Proc. Natl. Acad. Sci. U.S.A. 73, 3369-3373. Samuels, H.H., Stanley, F. and Shapiro, L.E. (1977) J. Biol. Chem. 252, 6052-6060. Samuels, H.H., Stanley, F. and Casanova, J. (1979) Endocrinology 105, 80-85. Samuels, H.H., Forman, B.M., Horowitz, Z.D. and Ye, Z.S. (1988) J. Clin. Invest. 81, 957-967. Sap, J., de Magi&s, L., Stunnenberg, H. and Vennstrom, B. (1990) EMBO J. 9, 887-896. Seeburg, P.H., Shine, J., Martial, J.A., Baxter, J.D. and Goodman, H.M. (1977) Nature 270, 486-494. Selden, R.F., Howie, K.B., Rowe, M.E., Goodman, H.M. and Moore, D.D. (1986) Mol. Cell. Biol. 6, 3173-3179. Thomas, P.S. (1980) Proc. Natl. Acad. Sci. U.S.A. 77, 52015205. West, B.L., Catanzaro, D.F., Mellon, S.H., Cattini, P.A., Baxter, J.D. and Reudelhuber, T.L. (1987) Mol. Cell. Biol. 7, 11931197. Wight, P.A., Crew, M.E. and Spindler, S.R. (1988) Mol. Endocrinol. 2, 536-542. Ye, Z.S., Forman, B.M., Aranda, A., Pascual, A., Park, H.Y., Casanova, J. and Samuels, H.H. (1988) J. Biol. Chem. 263, 7821-7829.
Molecular and Cellular Endocrinology, 71 (1990) 261-267 Elsevier Scientific Publishers Ireland, Ltd.
MOLCEL 02314
Triiodothyronine inhibits transcription from the human growth hormone promoter A. Morin ‘, J. Louette
2, M.L.J. Voz 2, A. Tixier-Vidal
‘, A. Belayew
2 and J.A. Martial
2
’ UAO 411 I5 CNRS, Laboratoire de Neuroendocrinologie Cellulaire, Collsge de France, F-75231 Paris Cedex, France, and ’ Laboratoire de G&tie G&tique,
CJniversitJ de Li.?ge, Institut de Chimie B6, B-4000 Sart-Tilman,
Belgium
(Received 26 January 1990; accepted 19 April 1990)
Key words: CC cell; hGH-reporter gene; hGH-CAT;
Transient expression; Triiodothyronine dose-response
Three DNA constructs, the natural human growth hormone gene (hGH-hGH) its 500 bp promoter linked to the chloramphenicol acetyl transferase reporter gene (hGH-CAT), and its structural part linked to the herpes virus thymidine kinase promoter (TK-hGH) were introduced into rat pituitary GC cells by DEAE-dextran transfection. Transient expression was followed as a function of triiodothyronine (T3) concentration. The hGH-CAT expression was specifically inhibited by T3 following a typical dose-response curve while hGH-GH gene expression was not significantly modified. The transient expression of TK-hGH increased as a function of T3 concentration. These results indicate that T3 exerts two opposite effects on hGH gene expression. First, it down-regulates expression by acting on the promoter; second, it up-regulates expression by acting on the structural part of the gene. These action could be due to regulation of transcription and mRNA stabilization, respectively.
Introduction The thyroid hormone L-3,5,3’-triiodothyronine (T3) regulates the expression of several genes, mostly at the transcriptional level where its action is mediated by its binding to a specific nuclear receptor related to the c-erb-A oncogene (reviewed by Oppenheimer et al., 1987; Samuels et al., 1988). The expression of the rat growth hormone (rGH) gene is specifically stimulated by Tj (Samuels et al., 1976; Martial et al., 1977). This transcriptional regulation has been most extensively studied in
Address for correspondence: Joseph A. Martial, Laboratoire Central de G&e GCnttique, Institut de Chimie B6, B-4000 Sart-Tiian, Belgium. 0303-7207/90/$03.50
the rat pituitary GC cell line (reviews: Davis et al., 1988; Samuels et al., 1988). Various deletion mutants of the rGH promoter have been linked to reporter genes and the resulting fusion genes transfected into GC cells. T3 induction of the reporter genes was analyzed either in stable transformants or during transient expression. Functional thyroid response elements (TRE) defined in this way were compared to T3 receptor binding sites tentatively located in vitro on the promoter DNA. A close association was found between the regulatory elements involved in the tissue-specific expression and in the thyroid hormone regulation of the rGH gene (Ye et al., 1988) making it difficult to locate the TRE. However, synthetic oligonucleotides could be derived from the rGH
6 1990 Ekevier Scientific Publishers Ireland, Ltd.
262
promoter sequence (Glass et al., 1987; Koenig et al., 1987; Brent et al., 1989) that could transfer T3 regulation to the thymidine kinase promoter in transfected cells and bind the Ts receptor in vitro. The conclusions drawn about the T3 stimulation of the rat GH gene transcription could not be extended to the homologous human GH gene for which Cattini et al. (1986a) demonstrated an inhibition of expression at physiological T3 concentration. In this experiment, GC cells had been stably transformed by a 2.6 kb EcoRI genomic fragment comprising the hGH gene with 496 bp of 5’ flanking sequence. In a subsequent work Cattini and Eberhardt (1987) constructed a hGH-CAT fusion gene containing 470 bp of promoter sequences. This region contains all the sequences necessary for tissuespecific expression, allowing the analysis of its biological activity in pituitary cells (Cattini et al., 1986b; Bodner and Karin, 1987; Lefevre et al., 1987). When hGH-CAT transient expression was assayed after transfection in GC cells, it showed little or no response to 10e8 M Tj treatment (Cattini and Eberhardt, 1987; Brent et al., 1988). The latter results were surprising since some of us had previously shown by DNA-cellulose binding competition (Barlow et al., 1986) that the only region in the whole hGH gene where a specific interaction for the T,-receptor complex could be found in vitro was located between coordinates - 290 and - 129. It could thus be expected that any transcriptional regulation of hGH gene expression by Ts had to involve this promoter region. To resolve these ambiguous data, we have analyzed the transient expression of three distinct hGH constructs after transfection of GC cells: the hGH promoter linked to the CAT reporter gene (hGH-CAT); the natural hGH gene (hGH-hGH); the neutral TK promoter linked to the hGH structural gene (TK-hGH). When expression was monitored as a function of T3 concentration, typical dose-response curves were found corresponding to a maximal 2-fold inhibition of hGH-CAT gene expression at lop9 M Tj and a 2-fold stimulation of TK-hGH gene expression. The transient expression of the natural hGH-hGH gene in these GC cells was not significantly altered by increasing Tj concentrations.
Materials and methods Plasmid constructions A 2.6 kb EcoRI fragment comprising 496 bp of promoter region, the whole structural part of the previously cloned human growth hormone gene (hGH,; Eliard et al., 1985) and 528 bp past the poly A addition site was subcloned in the plasmid vector pBR322, yielding hGH-hGH. A BamHI restriction site is located at the origin of transcription. The hGH-CAT fusion gene was kindly provided by M. Karin. It was constructed by insertion of the 500 bp EcoRI/BamHI hGH promoter fragment and the CAT structural gene (Gorman et al., 1982) into a pUC polylinker. The TK-hGH fusion gene was made by Selden et al. (1986) by ligating a PuuII-BglII fragment comprizing 200 bp of TK promoter and 56 bp of sequences downstream from the transcription initiation site to the hGH structural part. It was purchased as part of an Allegro assay system (Nichols Institute, Los Angeles, CA, U.S.A.). The TK-CAT fusion gene with 105 bp of TK promoter was present in the pBLCAT2 vector (Luckow and Schutz, 1987) kindly provided by G. Schutz. Tissue culture Rat pituitary tumor GC cells were grown in Ham F12 medium supplemented with 15% horse serum and 2.5% fetal calf serum. Sera were depleted of steroid and thyroid hormones by successive treatments with charcoal-dextran (Laverrilre et al., 1986) and AG l-X8 resin (Samuels et al., 1979). DEAE-dextran transfection The transfection schedule was the following: on day 1, 1.5 x lo6 cells were plated per 35 mm dish (6-well plates, Nunc) in hormone-free medium. On day 4 the medium was changed. On day 5 the cells were transfected. We essentially used the DEAEdextran method of Kopchick and Stacey (1984), except that the transfected 3 pg of supercoiled plasmid DNA/plate, and left the DNA for only 30 min on the cells. The fresh medium added after transfection already contained L-triiodothyronine (T3) (Na salt, 98% pure, Sigma) when needed. 10e3 M T3 stock solution was made in 80%
263
ethanol/20% NaOH 0.1 M and diluted into culture medium. Medium was changed on day 6 and cells were harvested on day 7, i.e. 44 h after transfection. RNA extraction and analysis Total RNA was prepared by dissolving cells in a 5 M guanidinium thiocyanate solution and selectively precipitating RNA by addition of 4 M LiCl according to Cathala et al. (1983). Total RNA was quantified by a method (Caverriere et al., 1989) based on measures of the fluorescence emitted by RNA/ ethidium bromide complexes. Serial 3-fold dilutions starting from 16 to 3 pg of each RNA sample were spotted onto nitrocellulose and hybridized to an excess of 32P-labelled probes (see below) as described by Thomas (1980). After autoradiography, individual spots were cut and distributed into separate scintillation vials and their radioactivity measured. Potential DNA contamination of RNA samples was monitored by DNase RQl digestion. Probes were the cDNA inserts of the phGH cDNA (Martial et al., 1979) and prGH cDNA (Seeburg et al., 1977) plasmids that were labelled by nick translation (Maniatis et al., 1975) with [“P]dCTP or dATP (400 Ci/mmol; Amersham) to a specific activity of l-5 X lo8 cpm/pg. CA T assay Assays were performed essentially according to Gorman et al. (1982) with the following modifications: a preliminary 6 min heat inactivation of endogenous acetylases at 65” C and an overnight incubation of the cell extract in the presence of 4 mM acetyl coenzyme A (Sigma) and 0.25 FCi of [‘4C]chloramphenicol (40 Ci/mmol, Amersham). Chlor~phe~~l and its acetylated forms were extracted in ethyl acetate and separated by thinlayer chromatography (TLC). Following autoradiography the silica gel spots from the TLC plates containing either acetylated products or remaining substrate were scraped into separate scintillation vials and counted. Standardization All data involving transfected genes are given per dish, not taking into account the various degrees of cell proliferation induced by T3 concentra-
tion, as none of the plasmids used can replicate in these cells. However, endogenous rGH gene expression was corrected for cell proliferation using the amount of total RNA extracted as a measure of cell population (see Flug et al., 1987 for a detailed discussion). Computer work The sequence analysis software package was from the University of Wisconsin Genetics Computer Group (Devereux et al., 1984). It was run on a Digital VAX 11/780. Results T3 regulation of hGfi-CA T The hGH gene promoter defined as a 500 bp EcoRI-BamHI fragment has been linked to the chloramphenicol acetyl transferase (CAT; Gorman et al., 1982) reporter gene to yield the hGHCAT hybrid gene that can direct proper transcript initiation (Lefevre et al., 1987). In order to study the T3 regulation of hGH-CAT expression, this construct was introduced into rat pituitary GC cells by DEAE-dextran transfection. Cells were grown a further 44 h in culture medium supplemented with various amounts of T3. CAT activity was assayed in the cell homogenates according to Gorman et al. (1982) and plotted as a function of
-12
-11
-10
-9
-8
Log T3 concentration Fig. 1. Transient expression of hGH-CAT in transfected Gc cells as a function of Ts concentration in the culture medium. CAT enzymatic activity was measured according to Gorman et al. (1982) in cells transfected by the DEAE-dextran method. Cells were harvested 44 h after transfection (for details, see Materials and Methods). Each point determined at least in triplicate is drawn with its standard error.
264
T3 concentration (Fig. 1). A T3 dose-dependent inhibition of hGH-CAT transient expression was observed with the maximal effect (2-fold inhibition) reached at 10m9 M T3. 7; reguI~tion of the hGH gene
A 2.4 kb EcoRI genomic fragment comprising the hGH gene with 496 bp of 5’ flanking DNA and 528 bp of sequences 3’ to the poly A addition site was subcloned into the plasmid vector pBR322 yielding hGH-hGH. The transcripts of such a construct are correctly initiated (Cattini et al., 1986a). This ‘natural’ hGH-hGH gene was introduced into rat pituitary GC cells in culture by DEAE-dextran transfection and its transient expression was monitored as a function of the T3 concentration in the medium. Total RNA extracted 44 h after transfection was analyzed by dot-blot hybridization using 32Plabelled hGH cDNA as a probe. The amount of hGH mRNA found (Fig. 2) was not sig~fic~tly altered by increases in Ts concentration from lo-” to lo-’ M. The endogenous rat GH gene induction was followed in parallel in the same cells. A typical dose-response curve was obtained (Fig. 2), with a maximal (i-fold stimulation reached at physiological T3 concentrations (lop9 M), indicating a normal response of these GC cells to the hormone (Samuels and Shapiro, 1976; Martial et al., 1977). No cross-hybridization between the rat
I---+
Fig. 2. Transient expression of the hGH gene in transfected GC ceils as a function of Ts concentration in the culture medium. Cells were harvested 44 h after DRAGdextrau transfection and total RNA was extracted. Dot-blot hybridization to hGH cDNA or rat GH cDNA probes distinguishes expression from the exogenous or endogenous GH genes (for details, see Materials and Methods). Each point determined at least in triplicate is drawn with its standard error.
Fig. 3. Transient expression of TK-hGH in transfected GC cells as a function of Ts concentration in the culture medium. Cells were harvested 44 h after DRAE-dextran transfection to hGH cDNA or rat GH cDNA probes distinguishes expression from the exogenous or endogenous GH genes (for details, see Materials and Methods). Each point determined at least in triplicate is drawn with its standard error.
and human GH mRNAs/cDNAs took place in our experiments. Indeed, RNA extracted from mock-transfected rat pituitary cells only gave a background signal when probed with the human GH cDNA (data not shown). T3 regulation of TK-GH
Since 7; inhibition linked to the hGH promoter as seen in hGH-CAT could not be observed in the transient expression of the natural gene, we reasoned that its structural part could be responsible for a positive T3 response. To test this hypothesis, we used the pTK-hGH construct where the hGH gene structural part was fused to the herpes simplex thymidine kinase (TK) promoter (Selden et al., 1986). GC cells were transfected with this fusion gene by the DEAE-dextran method. The transient expression of pTK-hGH was similarly analyzed as a function of T3 concentration in the culture medium. RNA expressed from pTK-hGH accumulated in the cells (Fig. 3) when T3 concentration was raised, reaching a plateau (2-fold M. The endogenous induction) at 10-i’ to lo-’ rat GH mRNA induction was followed in parallel in the same cells (Fig. 3). A control pTK-CAT construct (pBLCAT2, Luckow and Schlitz, 1987) was assayed for transient expression in the presence or absence of 10e8 M T3 and showed a non-specific 1.2- to l.Cfold induction (data not shown) in agreement with published data (e.g., Glass et al., 1987).
Discussion In order to study the role of Ts on the expression of the hGH gene, we have assayed three constructs: the 500 bp hGH promoter linked to the CAT reporter gene (hGH-CAT), the natural hGH gene with the same 500 bp of promoter sequences (hGH-hGH) and the neutral thymidine kinase promoter linked to the hGH structural gene (TK-hGH). Transient expression of the three constructs was followed as a function of T3 concentration after introduction into GC cells. No significant effect was seen for the natural hGH gene expression while hGH-CAT and TK-hGH were respectively down- and up-regulated, to a similar extent. We conclude that T3 exerts two opposite effects on hGH gene expression. T3 effect through the hGH gene promoter We have shown in this paper that T3 down-regulates expression from the 500 bp hGH promoter under conditions that lead to a 6-fold induction of the rat gene expression. This T3 inhibition of transient hGH-CAT expression has been recently confirmed by electroporation of GC cells (Voz et al., in preparation). Previous data from other groups (Cattini and Eberhardt, 1987; Brent et al., 1988) showing a slight Ts inhibition or no effect on hGH promoter activity can be explained by different hormonal conditions. Indeed, a much lower T3 induction of the control rGH gene (1.5to 2-fold) is reported in these papers suggesting sub-optimal TX responses. Could the different T3 response observed with the hGH and rGH promoters be mediated by the same TRE? Some of us have previously shown that a human T3 receptor could specifically interact with a fragment located from - 290 to - 129 (Barlow et al., 1986) in the hGH promoter. Most of the many publications that have described positive regulatory elements involved in T3 regulation of rGH promoter have located them between coordinates - 208 and - 164 (Glass et al., 1987; Koenig et al., 1987; Wight et al., 1988; Ye et al., 1988; Brent et al., 1989). When comparing the rat and human GH promoters, and introducing gaps to maximize similarity (Fig. 4) the best conserved sequence within this region is found between nucleotides
hGH
rGH
TRE
related
SPl
GHFl I
Fig. 4. Sequence comparison of the human and rat GH proximal promoters. Sequences have been aligned and gaps have been introduced to maximize similarity using the ‘Gap’ program from the University of Wisconsin Genetics Computer Group. The boxes refer to cis elements involved in the transcriptional regulation of the human GH (upper line) and rat GH (lower line) genes. Two binding sites for the pituitary specific factor GHF-l/GCZ/GCl (Catanzaro et al., 1987; Lefevre et al., 1987; West et al., 1987; Ye et al., 1988) are shown as striped boxes on each promoter. The dotted boxes stand for SPI binding sites (Lefebre et al., 1987; Lemaigre et al., 1989). A sequence that is identical in both promoters, located in a region involved in Ts regulation of rat GH expression is outlined as a black box. Numbers refer to coordinates of the cis elements and gap boundaries, starting from the transcription initiation site taken as + 1. The introduced gap sizes are given in bp.
- 188 and - 173 in hGH and - 206 to - 191 in rGH. This location in hGH is 17 bp closer to a region extending from -106 to -148 in both promoters that can bind a pituitary specific factor named GHF-l/GC2 (Lefevre et al., 1987; West et al., 1987; Ye et al., 1988). In the absence of GHF-1, the distal part of this site can interact with the ubiquitous SPl factor (Ye et al., 1988; Lemaigre et al., 1989). The Tj receptor binding sequences might overlap with the GHF-1 in hGH site thus preventing activation, as was described for glucocorticoid inhibition of a CAMP-responsive enhancer (Akerblom et al., 1988). Alternatively, an altered spacing between the Ts receptor and GHF-1 binding sites in the hGH promoter could prevent the type of interactions found in the rGH gene where tissue-specific and
266
T:, response elements are tightly linked (Ye et al., 1988). Finally, the T3 receptor could interact with the GHF-1 protein bound to its specific DNA site and lower its ability to induce transcription as was shown for the ~uc~orticoid and estrogen receptors on rat prolactin gene expression (Adler et al., 1988). In order to distinguish between these possibilities, deletion mutants of the hGH promoter are presently being tested for their ability to confer T3 inhibition to a CAT reporter gene (Voz et al., in preparation). TX effect through hGH sequences located 3’ from the transcription initiation site As our results on TK-hGH show, Ts exerts a positive effect through the hGH structure gene or 3’ flanking sequences. These data, together with a report on the role of glucocorticoids in hGH mRNA stabilization (Paek and Axel, 1987), indicate that the use of the hGH reporter gene as initiated by Selden et al. (1986) should be reconsidered with caution when studying hormonal regulation. We do not believe this TX regulation is transcriptional since, using human IM9 cell T3 receptor, we could not find in vitro binding downstream from the hGH transc~ption i~tiation site (Barlow et al., 1986). These results have not been refuted by the recent data of Sap et al. (1990) showing that the third rat GH intron includes a TRE responding to the chicken c-erb-A protein. Indeed no sequence similar to that region is present in the smaller third intron of the hGH gene. We therefore favour a post-transc~ptional mechanism as was shown for Ts on liver malic enzyme (Dozin et al., 1986) and spot 14 (Narayan and Towle, 1985) gene expression. Tj effect on the 2.6 kb hGH gene We have not found a significant effect of T3 on the hGH-GH gene expression. This could result from the addition of the two opposite effects described above. Nothing is known about the Tj regulation of the natural hGH gene in normal human pituitary cells. However, the results presented here are in agreement with a work of Isaacs et al. (1987) on
dispersed GH-secreting human pituitary adenoma cells where no significant effect of T3 on hGH mRNA levels could be demonstrated. Our results disagree with data from Cattini et al. (1986a) on GC cells stably transformed by the hGH gene. This discrepancy is most probably due to experimental conditions which could affect to different extents the two opposite effects of TJ. It should also be noted that the TX effect can differ according to the GH, subclone used for transfection as was shown for rat prolactin expression (Forman et al., 1988). Another explanation for this discrepancy might be found in the heterogenous T3 receptor population (Samuels et al., 1977; Cattini et al., 1988; Wight et al., 1988; Hodin et al., 1989) present in GH, cells. Maximal inhibition would be observed as soon as T3 is added to the cell culture medium. Progressive disappearance of the depletable TX receptor population in GC cells would then lead to a release of transcription in~bition. A short incubation in the presence of Ts such as what can be done with stable hGH transformants (Cattini et al., 1986a) would favour detection of the inhibitory role of the hormone. A longer incubation as done routinely in transient expression (this study) would allow a balance between the stimulatory and inhibitory effects of ?;.
Acknowledgements We would like to thank G. Schutz and M. Karin for kindly providing pBLCAT2 and phGHCAT respectively. We acknowledge the help of M.J. Marchand, M.L. Scippo, E. Bellefroid and C, Lambert in an initial phase of this work, J.N. Laverriere and G. Rousseau for discussions, E. Rosenbaum for maintaining cell cultures and J. Lejeune for expert secretarial assistance. A.M. held a post-doctoral fellowship from the CNRS (France, Direction des Relations et de la Cooperation Internationale). J.A.L. and M.L.J.V. held IRSIA pre-doctoral fellowships (Belgium). This work was partly supported by a grant from the SPPS (Belgium, Sciences de la Vie, BI0/15) and the FNRS (Belgium, FRSM 3.4577.86).
267
References Ackerblom, I.E., Slater, E.P., Beato, M., Baxter, J.D. and Mellon, P.L. (1988) Science 241, 350-353. Adler, S., Waterman, M.L., He, X. and Rosenfeld, M.G. (1988) Cell 52, 685-695. Barlow, J.W., Voz, M.L.J., Eliard, P.H., Mathy-Hartert, M., De Nayer, P., Economidis, I.V., Belayew, A., Martial, J.A. and Rousseau, G. (1986) Proc. Natl. Acad. Sci. U.S.A. 83, 9021-9025. Bodner, M. and Karin, M. (1987) Cell SO, 267-275. Brent, G.A., Hamey, J.W., Moore, D.D. and Larsen, P.R. (1988) Mol. Endocrinol. 2, 792-798. Brent, G.A., Larsen, P.R., Hamey, J.W., Koenig, R.J. and Moore, D.D. (1989) J. Biol. Chem. 264, 178-182. Catanzaro, D.F., West, B.L., Baxter, J.D. and Reudelhuber, T.L. (1987) Mol. Endocrinol. I, 90-96. Cathala, G., Savouret, J.F., Mendez, B., West, B.L., Karin, M., Martial, J.A. and Baxter, J.D. (1983) DNA 2, 3299335. Cattini, P.A. and Eberhardt, N.L. (1987) Nucleic Acids Res. 15, 1297-1309. Cattini, P.A., Anderson, T.R., Baxter, J.D., Mellon, P. and Eberhardt, N.L. (1986a) J. Biol. Chem. 261, 13367-13372. Cattini, P.A., Peritz, L.N., Anderson, T.R., Baxter, J.D. and Eherhardt, N.L. (1986b) DNA 5, 503-509. Cattini, P.A., Kardami, E. and Eberhardt, N.L. (1988) Mol. Cell. Endocrinol. 56, 2633270. Davis, J.R.E., Belayew, A. and Sheppard, M.C. (1988) in Molecular Biology of Endocrinology (Baillieres Clinical Endocrinology and Metabolism), Vol. 2, pp. 797-834, W.B. Saunders, London. Devereux, J., Haeberli, P. and Smithies, 0. (1984) Nucleic Acids Res. 12, 387-395. Dozin, B., Magnuson, M.A. and Nikodem, V.M. (1986) J. Biol. Chem. 261, 10290-10292. Eliard, P.H., Marchand, M.J.. Rousseau, G.G., Formstecher, P.. Mathy-Hartert, M., Belayew, A. and Martial, J.A. (1985) DNA 4, 409-417. Flug, F., Copp, R.P., Casanova, J., Horowitz, Z.D., Janocko, L.. Plotnick, M. and Samuels, S.H.H. (1987) J. Biol. Chem. 262, 6373-6382. Forman, B.M., Chang-Ren, Y., Frederick, S., Casanova, J. and Samuels, H.H. (1988) Mol. Endocrinol. 2, 902-911. Glass. C.K., France, R., Weinberger, C., Albert, V., Evans, R.M. and Rosenfeld, M.G. (1987) Nature 329, 738-741. Gorman, CM., Moffat, L.F. and Howard, B.H. (1982) Mol. Cell. Biol. 2, 10441051. Hodin, R.A., Lazar, M.A., Wintman, B.I., Darling, D.S., Koenig, R.J., Larsen, P.R., Moore, D.D. and Chin, W.W. (1989) Science 244, 76-79. Isaacs, R.E., Findell, P.R., Mellon, P., Wilson, C.B. and Baxter, J.D. (1987) Mol. Endocrinol. 1, 569-576.
Koenig, R.J., Brent, G.A., Warner, R.L., Larsen, P.R. and Moore, D.D. (1987) Proc. Natl. Acad. Sci. U.S.A. 84, 5670-5674. Kopchick, J.J. and Stacey, D.W. (1984) Mol. Cell. Biol. 4, 240-246. Laverriere, J.N., Muller, M., Buisson, N., Tougard, C., TixierVidal, A., Martial, J.A. and Gourdji, D. (1986) Endocrinology 118, 198-206. Laverriere, J.N., Richard, J.L., Buisson, N., Martial, J.A., Tixier-Vidal, A. and Gourdji, D. (1989) Neuroendocrinology 50, 693-701. Lefevre, C., Imagawa, M., Dana, S., Grindlay, J., Bodner, M. and Karin, M. (1987) EMBO J. 6, 971-981. Lemaigre, F.P., Courtois, S.J., Lafontaine, D.A. and Rousseau, G.G. (1989) Eur. J. B&hem. 181, 555-561. Luckow, B. and Schutz, G. (1987) Nucleic Acids Res. 15, 5490. Maniatis, T., Jeffrey, A. and Kleid, D.G. (1975) Proc. Natl. Acad. Sci. U.S.A. 72, 1184-1188. Martial, J.A., Baxter, J.D., Goodman, H.M. and Seeburg, P.H. (1977) Proc. Natl. Acad. Sci. U.S.A. 74, 1816-1820. Martial, J.A., Hallewell, R.A., Baxter, J.D. and Goodman, H.M. (1979) Science 205, 602-607. Narayan, P. and Towle, H.C. (1985) Mol. Cell. Biol. 5, 26422646. Oppenheimer, J.H. (1987) in Molecular Basis of Thyroid Hormone Action (Oppenheimer, J.H. and Samuels, H.H., eds.), pp. l-35, Academic Press, New York. Paek, I. and Axel, R. (1987) Mol. Cell. Biol. 7, 1496-1507. Samuels, H.H. and Shapiro, L.E. (1976) Proc. Natl. Acad. Sci. U.S.A. 73, 3369-3373. Samuels, H.H., Stanley, F. and Shapiro, L.E. (1977) J. Biol. Chem. 252, 6052-6060. Samuels, H.H., Stanley, F. and Casanova, J. (1979) Endocrinology 105, 80-85. Samuels, H.H., Forman, B.M., Horowitz, Z.D. and Ye, Z.S. (1988) J. Clin. Invest. 81, 957-967. Sap, J., de Magi&s, L., Stunnenberg, H. and Vennstrom, B. (1990) EMBO J. 9, 887-896. Seeburg, P.H., Shine, J., Martial, J.A., Baxter, J.D. and Goodman, H.M. (1977) Nature 270, 486-494. Selden, R.F., Howie, K.B., Rowe, M.E., Goodman, H.M. and Moore, D.D. (1986) Mol. Cell. Biol. 6, 3173-3179. Thomas, P.S. (1980) Proc. Natl. Acad. Sci. U.S.A. 77, 52015205. West, B.L., Catanzaro, D.F., Mellon, S.H., Cattini, P.A., Baxter, J.D. and Reudelhuber, T.L. (1987) Mol. Cell. Biol. 7, 11931197. Wight, P.A., Crew, M.E. and Spindler, S.R. (1988) Mol. Endocrinol. 2, 536-542. Ye, Z.S., Forman, B.M., Aranda, A., Pascual, A., Park, H.Y., Casanova, J. and Samuels, H.H. (1988) J. Biol. Chem. 263, 7821-7829.