THE JooxNnL OF BIOLOGICAL CHEMISTRY 0 1993 by The American Soclety for Biochemistry and Molecular Biology, Inc

Vol. 268, No. 31, Issue of November 5, pp. 23538-23543, 1993 Printed in U.S.A.

Repression of c-fos Gene Expression by Thyroid Hormone and Retinoic Acid Receptors*

(Received for publication, May 3, 1993, and in revised form, July 19, 1993)

Paloma Perez, Axel Schonthal$, and Ana ArandaO

28029 Madrid, Spain From the Znstituto de Znuestigaciones Biomedicas. Consejo Superior de Inuestigaciones Cientificas, Arturo Duperier 4,

Incubation with 1 nm triiodothyronine (T3) decreased cycloheximide-induced c-fos mRNAlevels and the mRNA response to the tumor promoter 12-0-tetradecanoyl- phorbol-13-acetate (TPA; 100 nm) or to forskolin (15 PM). T3 also reduced the abundance of nuclear proteins that bind to an AP-1 binding site and the levels of c-Fos pro- tein, as determined by Western blot. In transient trans- fection assays with a c-fos promoter reporter construct T3 decreased basal promoter activity by more than 50- 60% and strongly inhibited the TPA and forskolin-in- duced activity. T3 further suppressed basal promoter ac- tivity and led to a drastic decrease of the response to both stimulatory agents after co-transfection of the pro- moter with an expression vector for the receptor. A trun- cated receptor that lacks the DNA binding domain abol- ished the inhibitory effect of the endogenous and transfected T3 receptor. Retinoic acid (RA) had an effect similar to T3 reducing basal CAT activity and the re- sponse to TPA and forskolin by 40-50% in cells co-trans- fected with the c-fos promoter and an expression vector for the RA receptor. The inhibitory effect of the nuclear receptors on c-Fos is reciprocal, since overexpression of c-Fos repressed induction of the activity of the growth hormone promoter by T3 and RA. Co-transfection with an antisense c-fos vector relieved this inhibition and re- stored the response to both ligands. These results show the antagonism between the nuclear receptors and the membrane signal transduction pathways that converge on the c-fos oncogene.

Thyroid hormones and retinoids affect cell growth and dif- ferentiation by binding to nuclear receptors that function as ligand-inducible transcriptional enhancer factors (1). These re- ceptors stimulate transcription through binding to specific se- quences named hormone response elements, which are nor- mally located in the 5”flanking region of the regulated genes. The high degree of homology between the thyroid hormone and R A 1 receptors makes possible that both can regulate genes in common (2) by binding to similar or even the same (3-5) ele- ments.

* This work has been supported by Grant PM88-0007 from Direction General de Investigacion Cientifica y Tecnica, Grant C051/91 from CO- munidad de Madrid, and a grant from Fundacion Ramon Areces. The costs of publication of this article were defrayed in part by the payment

tisernent” in accordance with 18 U.S.C. Section 1734 solely to indicate of page charges. This article must therefore be hereby marked “aduer-

this fact.

fornia, Los Angeles, CA 90033-1054. $ Present address: Dept. of Microbiology, University of Southern Cali-

§ To whom correspondence should be addressed: Instituto de Inves-

Arturo Duperier 4, 28029 Madrid, Spain. Fax: 34-1-5854587. tigaciones BiomBdicas, Consejo Superior de Investigaciones Cientificas,

The abbreviations used are: RA, retinoic acid; F U R , RA receptor; TB, triiodothyronine; GH, growth hormone; CAT, chloramphenicol acetyl- transferase.

In addition to stimulating transcription, the nuclear recep- tors can also selectively inhibit the expression of certain genes. Negative regulation can involve negative hormone response elements (51, but it can also result from protein to protein interactions. In some cases repression of gene expression by T3 and RA receptors involves interaction of the receptors with the components of the AP-1 complex (6-10).

The AP-1 complex, which interacts with the control regions of genes containing TPA-responsive promoter elements (111, con- sists of two subunits encoded by the family of j u n a n d fos oncogenes. These nuclear oncogenes have important effects on cell proliferation and differentiation, and, therefore, regulation of its expression andlor activity is central to understanding their role in normal and neoplasic growth (see Refs. 12 and 13, and references therein).

Thyroid hormones have mitogenic effects on growth hormone (GH)-producing pituitary cells (14). Although RA has anti-pro- liferative effects in a variety of cell types often associated with expression of differentiated phenotype, we have observed that the retinoid increases proliferation of pituitary GH1 cells (15). Since expression of c-Fos is induced i n response to a variety of mitogenic signals, we have examined whether the effects of T3 and RA on pituitary cell growth could be related to changes in the expression of c-Fos. Our results indicate that despite their proliferative effects in pituitary cells, both ligands can tran- scriptionally decrease the expression of the c-fos oncogene. T3 and RA receptors repress the basal activity of the c-fos pro- moter and its response to TPA and cyclic AMP. This demon- strates that in pituitary cells the nuclear receptors antagonize the signals transduced by protein kinase A and protein kinase C, which converge in the c-fos oncogene. This inhibitory effect is reciprocal, and c-Fos blocks transcriptional induction of the GH promoter by the receptors.

EXPERIMENTAL PROCEDURES

Plasmids-The plasmid FC4 (16) contains 404 base pairs upstream of the c-fos start of transcription linked to chloramphenicol acetyltrans- ferase (CAT) gene. The plasmid TATA CAT without any c-fos sequences has been described ( 17). The expression vectors for RA and T, receptors contain the cDNA sequences of the human retinoic acid receptor M a (3) or chick c-erbAcr, a thyroid hormone receptor (18) under control of the Rous sarcoma virus promoter. A 5’ deletion mutant of ~ - e r b A ( M e t ~ ~ ~ - V a l ~ ~ ~ ) , which lacks the A/B domain and the DNA bind- ing domain, was obtained by cleaving at NcoI sites in the chick c-erbA c-DNA and cloning in the expression vector as described (19). Vectors for c-fos and antisense c-fos have been described previously (20). The re- porter plasmid -530GH-CAT contains a fragment of the rat GH gene (-530/+7) fused to CAT gene (21). We have shown previously that the GH promoter responds to T, and RA in pituitary cells (2).

Cell Culture and DNA Dansfection-GH1 cells were cultured in RPMI medium containing 10% horse serum and 2.5% fetal calf serum and were transfected by electroporation as described previously (2, 21). Ten pg of the reporter plasmids with or without the same amount of expression vectors of the receptors or c-fos were mixed with 20 million cells and exposed to a high voltage pulse (170-200 V, 960 microfarads)

23538

Repression of c-fos by Nuclear Receptors 23539

FIG. 1. Northern blot analysis of c-foe mRNA. GH1 cells were incubated for 48 h with 1 nM T, or 1 p~ RA in hormone-depleted medium and with 100 nM TPA (leppanel) for the last 30 or 90 min, or with 5 pglml cycloheximide for the last 6 h (right panel ). RNA was extracted, and Northern blot analysis was performed with a c-fos cDNA probe as described under "Experimental Procedures".

using a Bio-Rad electroporator with a capacitor extender. The total amount of transfected DNA was kept constant by addition of carrier DNA. The cells from each electroporation were split in different culture plates in medium containing 10% AGlx8 resin-charcoal stripped new- born calf serum (14). Treatments with RA, T,, andor TPA and forskolin were with the concentrations and at the times indicated in the text in this hormone-depleted medium. Each treatment was performed in du- plicate cultures that normally showed less than 5% variation in CAT activity, which was determined (2,211 by incubation of the cell extracts with ['4C1chloramphenicol. The unreacted and acetylated [14Clchloram- phenicol were separated by thin layer chromatography, identified by autoradiography, and quantitated. Each experiment was repeated a t least two to three times with similar relative differences in T3- and RA-regulated expression.

RNA Extraction and Hybridization-Total RNA was extracted from GH1 cultures with guanidinium thiocyanate. The RNA (15 pg) was run in 1% formaldehyde-agarose gels and transferred to nylon membranes (Nytran) for Northern blot analysis. The blots were hybridized with a 1.1-kilobase PstI fragment from pfos, a rat c-fos cDNA-bearing plasmid (22) labeled by random oligonucleotide priming. Hybridizations were at 42 "C with 50% formamide and the more stringent wash was a t 60 "C with 0.1 x SSPE, 1% SDS. The blots were quantified by densitometric scan of the c-fos mRNA band, and the data were corrected by the intensity of the ribosomal RNA bands.

Western Blot-50 pg of GH1 cell extracts were run in a 10% acryl- amide gel and transferred to an Immobilon polyvinylidene difluoride membrane. Nonspecific binding was blocked with 1% nonfat dried milk in TBS-T (Tris-buffered saline, 0.1% Tween 20) for 2 h a t mom tem- perature. The membrane was rinsed, incubated with a 1/1000 dilution of sheep c-Fos antibody (Cambridge Research Biochemicals, OA-11825) at 4 "C overnight. washed again, and incubated in 1/500 rabbit anti- sheep peroxidase-conjugated antibody (Cambridge Research Biochemi- cals, SA-10-435) for 2 h a t room temperature. Detection by enhanced chemiluminiscence (ECL, Amersham Corp.) was carried out following manufacturer's indications.

DNA Binding Assays-Gel retardation analysis were carried out with nuclear extracts (23) from GH1 cells. As probes we used an oligo- deoxynucleotide conforming the AP-1 binding site present in the c-fos promoter (5'-GGACAATC?GTCAGCAGG"T-3'). For the binding assays, the extracts were incubated on ice for 15 min in a buffer (20 mM Tris-HC1, pH 7.6,75 mM KCI, 1 mM dithiothreitol, 5 pg/ml bovine serum albumin, 13% glycerol) containing 3 pg of poly(dI-dC) and then for 20 min a t room temperature with approximately 70,000 cpm of double-stranded oligonucleotide end-labeled with [yJ2P1ATP using T4 polynucleotide kinase. For competition experiments, a n excess of unla- beled, double-stranded oligodeoxynucleotides was added to the binding reaction. When anti-Fos antibody was used, 1 pl of a monoclonal anti- body (1 pg/pl) against residues 359-378 of c-Fos (Oncogene Science), or 1 pg of purified IgG was incubated overnight with the reaction mixture prior to addition of the labeled probe. DNA-protein complexes were resolved on 5% polyacrylamide gels in 0.5% TBE buffer (0.089 M M a borate, 0.089 M boric acid, 2 mM EDTA).

c- la +

7- w - FIG. 2. Influence of T, and FtA on the c-foe mRNA responme to

forskolin. Northern blot analysis was performed with total RNA from GH1 cells incubated for 48 h with 1 nw T:, or 1 p~ RA and/or for the last 30 or 90 min with 15 p~ forskolin. The upper panel shows c-fos mRNA levels, and the lower panel the amount of ribosomal RNA applied to each lane.

RESULTS Regulation of c-fos mRNA a n d P r o t e i n i n GHI Cells-The

levels of c-fos mRNA were undetectable by Northern blot anal- ysis in control GH1 cells and in cells treated with TR or RA for different time periods between 30 min and 72 h. Since c-fos mRNA is superinduced in the absence of de novo protein syn- thesis due to prolonged transcription from the promoter and stabilization of the message (12), we examined c-fos mRNA levels in cells incubated with T3 or RA for 48 h and with 5 &ml cycloheximide for the last 6 h. Fig. 1 ( r i g h t p a n e l ) shows that cycloheximide strongly induced c-fos mRNA levels. This induc- tion was blunted to 40-500 of control levels in cells treated with 1 nM T3 and was not significantly reduced in cells treated with 1 J ~ M RA. The lefl panel in Fig. 1 illustrates the effect of Tn and RA on the response to the tumor promoter TPA, a well known inducer of c-fos gene expression. TPA (100 nM) induced a large increase of c-fos mRNA at 30 min, which almost re- turned to normal after 90 min. The response to TPA was mark- edly reduced in cells treated with T3 and slightly decreased in cells treated with RA. Quantification of the Northern blots showed that T3 decreased by approximately 5-fold the response to a 30-min incubation with TPA, whereas RA produced a less

23540 Repression of c-fos by Nuclear Receptors

than 2-fold decrease. Fig. 2 shows the effect of T3 and RA on the c-fos mRNA response to forskolin, which increases intracellular CAMP. The kinetics of c-fos mRNA induction were similar to that found with TPA, showing high c-fos mRNA levels after 30 min of incubation with forskolin and becoming undetectable after 90 min. T3 decreased to less than half the response to forskolin, and RA reduced mRNA levels to a similar extent.

Fig. 3 shows that, as studied by Western blot, the levels of c-Fos protein in response to TPA were also significantly de- pressed in cells incubated with T3. Treatment with RA had a minor effect and decreased the response to TPA by 20-3070. To analyze whether the decrease of c-Fos protein and mRNA caused by T3 was also reflected in a decrease in the abundance ofAP-1 complexes, gel retardation assays were performed with an oligonucleotide conforming to an AF"1 binding site. As

C TPA T3*1PA RA*TPA

FIG. 3. Expression of c-Fos in CH1 cells. Proteins (50 pg) from control cells ( C ) and cells incubated with 1 nM T3 or 1 p~ RA for 48 h in the presence or absence of 100 nM TPA for the last 60 min, were used for Western blot analysis with an anti-Fos antibody as described under "Experimental Procedures."

shown in Fig. 4 (left panel), nuclear extracts from GHl cells showed the presence of retarded bands with the AP-1 site, which were enhanced significantly in the cells treated with TPA. These complexes were competed efticiently by an excess of unlabeled AP-1 oligonucleotide, but they were unaffected by an excess of an unrelated oligonucleotide (not shown). In parallel with the decrease in the expression of the c-fos gene, the abun- dance of the retarded bands was significantly lower in the nuclear extracts from cells incubated with T3 than from control cells, whereas RA had little if any effect. Both ligands de- creased the response to TPA. Since different members of the fos and j u n families can form dimers that bind to AP-1 sites, the nuclear extracts were incubated with an anti-c-Fos antibody or with an equivalent amount of purified IgC (Fig. 4, right panel) to analyze the presence of c-Fos in the retarded bands. Incu- bation with the antibody produced a decrease in the intensity of the retarded bands formed with extracts from control and RA- treated cells and did not lower the already decreased intensity of the shift caused with extracts from T3-treated cells. Whereas in these groups the antibody did not cause the appearance of new protein-DNA complexes of slower mobility, a clear super- shift was apparent in extracts from cells incubated with TPA alone, RA + TPA, or T3 + TPA, although the intensity of the latter was decreased.

Influence of T.? and RA on the Activity of the c-fos promoter "To examine whether the decrease in c-Fos protein and mRNA levels following T3 treatment of GH1 cells is exerted a t a tran- scriptional level, we performed transient transfection assays with a fos-CAT fusion construct (16). Incubation with T3, but not with RA, produced a decrease of basal c-fos promoter ac- tivity. Time-course studies (Fig. 5) showed that the effect of T3 was negligible after 3 h but was observed a t 8 h, when a half- maximal decrease with respect to the values obtained after 48 h was found. Fig. 5 also shows that the inhibitory effect of T3 on the promoter was slower than the stimulation caused by 100 nM TPA and 15 p~ forskolin. The effect of both stimulators was barely detectable a t 3 h, but was already maximal between 8 and 24 h of treatment. After 8 h of incubation and at the concentrations used, T3 produced a 3040% decrease of CAT activity, forskolin produced a 4-5-fold increase, and TPA en- hanced this activity by 2-3-fold.

Fig. 6A shows the effect of a 48-h incubation with T3 and RA on basal c-fos promoter activity and on the promoter response

&os - + - + - + - + - + - + T3 - - + + - - " + + - - RA " " + + " " + +

TPA - - - - - - + + + + + + b w b -mm

73 - + " + - RA

TPA " - + + + " + " +

FIG. 4. Mobility shift assay of nuclear extracts with an AP-1 site. A 32P-labeled oligodeoxynucleotide conforming the AP-1 binding site of the c-fos promoter was used for gel retardation with 15 pg of nuclear extracts from control GH1 cells, from cells treated with 1 nM T, or 1 pu RA for 48 h, and from cells incubated with 100 nM TPA for the last 90 min. Panel A illustrates the influence of T,. RA. and/or TPA on the abundance of protein-DNA complexes. Panel B shows the influence of an anti-c-Fos antibody (a-FOS) on the gel retardation assay. In this case the extrahq were pretreated with either the antibody or an equivalent amount (1 pg) of purified I@ before incuhation with the labeled oligonucleotide. In both panels, lane 1 illustrates migration of the free probe, and lane 2 indicates competition of lane 3 with a 50-fold excess of unlabeled A P - I oligodeoxynucleotide. The arrows indicate the specific protein-DNA complexes formed.

Repression of c-fos by Nuclear Receptors 2354 1

to 100 nM TPA and 15 p~ forskolin. T3 reduced basal promoter activity by 5040% and strongly inhibited the TPA and forsko- lin-induced activity of the reporter gene. The negative regula- tion of T3 is specific for this promoter. This hormone does not cause a generalized gene repression, since i t did not affect activity of the minimal promoter pTATA CAT (not shown). To analyze the influence of overexpression of T3 receptors on the response of the c-fos promoter, the reporter plasmid was co- transfected with an expression vector encoding the a form of c-erbA, a thyroid hormone receptor. As shown in Fig. 6B, the basal activity of the c-fos promoter was further repressed in a ligand-dependent manner by co-transfection of c-erbAa. T3 caused a drastic decrease of the response to TPA and forskolin in cells overexpressing c-erbA.

As shown in Fig. 6 A , RA did not decrease but rather in- creased the activity of the c-fos promoter (100-170% of controls depending on the experiments) when acting through its endog- enous receptors. However, in cells transfected with an expres- sion vector for M a , the retinoid had a similar effect than T3 reducing basal CAT activity by 3040% (Fig. 6C). Overexpres- sion of the RAR also led to an inhibition on the response to forskolin and TPA in the presence of RA, although in both cases the effect of the retinoid was somewhat less marked than that produced by T3. The inhibition caused by the transfected RAR is specific, since overexpression of this receptor produces a sev- eralfold increase in the response of the GH promoter or other reporter constructs containing a positive response element to RA (data not shown).

A c-erbA Mutant Lacking the DNA Binding Domain Inhibits the Action of the T3 Receptor-A truncated c-erbA that is tran- scriptionally inactive acts in a dominant-negative fashion to

0 2 4 e 2 4 4

n m h 0 u n 1

FIG. 5. Regulation of the activity of the c-fos promoter in GHl celle. Electroporation was used for transient transfection of GH1 cells with 10 pg of the reporter plasmid FC4, which contains 404 base pairs of the c-fos promoter. After transfection, cells were grown in hormone- depleted medium for the times indicated in the presence of 1 nM T3, 1 p~ RA. 15 p~ forskolin (Fk), or 100 nM TPAand CAT activity determined as described under "Experimental Procedures." The data are mean * S.D of duplicate transfections.

block trans-activation by the T3 receptor (19). This deletion mutant, which lacks the DNA binding domain, abolished the inhibitory effect of T3 on the basal and stimulated expression of the c-fos promoter. The truncated receptor blocks the inhibition produced by the endogenous and transfected c-erbA. Fig. 7 shows the results obtained after co-transfection of 2.5 pg of the wild-type vector alone (panel A) or with 10 pg of the mutant receptor (panel B). Under these conditions, T3 did not inhibit the activity of the c-fos promoter or decrease the stimulation produced by TPA or forskolin. When equal amounts of both vectors (10 pg) were transfected, a partial response to T3 was found (not illustrated), thus showing a dose-dependent inhibi- tion of the effect of T3 by the truncated receptor.

Inhibition by c-Fos of the T.? and RA-mediated Stimulation of the Growth Hormone Promoter by c-Fos-T3 and RA increase transcription of the rat GH gene through a common hormone- responsive element located in the 5"flanking region of the gene (2, 24). To examine whether the inhibitory effect of the recep- tors on t h e m - 1 complex is reciprocal, we studied the influence of c-fos on the response of the growth hormone promoter to T3 and RA. Fig. 8A shows that T3 and RA increased by approxi- mately 3-4-fold the expression of a GH-CAT construct which contains 530 base pairs of flanking DNA. Fig. 88 illustrates the effect of co-transfection with a c-fos expression vector on the response of the GH promoter to T3 and RA. Overexpression of c-Fos totally repressed the ligand-dependent activation of the GH gene promoter by the endogenous T3 and RA receptors. When c J u n was transfected with c-Fos, an inhibition of the response was also found (not shown). As can be seen in Fig. 8C, co-transfection with an antisense c-fos vector (SOD did not de- crease the response of the GH promoter to T3 And RA. The antisense vector when used in combination with the c-fos vector relieved the inhibition caused by the oncoprotein and restored the response to both ligands (Fig. 80 ).

DISCUSSION

We report here a decrease in c-fos mRNA and protein levels following T3 treatment of pituitary GHl cells. This effect is exerted at a transcriptional level, since expression of a c-fos promoter-CAT fusion plasmid was also inhibited in the pres- ence of T3. The activity of the c-fos promoter was further re- pressed in a ligand-dependent manner by co-transfection of an expression plasmid encoding an a form of c-erbA, a t h p i d hormone receptor. In GH1 cells the endogenous RA receptors do not mediate a significant inhibition of the expression of the c-fos promoter, although we have shown that they increase transcription of the GH gene (2, 24). In fact, we observed in

CAT = t M Y m "1 FIG. 6. Iniluence of Ts and RA on the e x p d o n of the c-fos promoter. In A, GHl celle were transfected with 10 pg of the c-fo. promoter

plasmid FC4. In B and C, the plasmid was co-transfected with 10 pg of expression vectors for c-erbA or RAR. respectively. CAT activity was determined in cells incubated with medium alone (control cells, C), with 1 nM T, or 1 PM RA for 48 h, and/or with 100 nM TPA or 15 PU fornkolin ( F K ) for the last A h.

23542 Repression of c-fos by Nuclear Receptors

several experiments a slight enhancement in the activity of the c-fos promoter in cells treated with RA. However, after overex- pression of the RA receptor, the retinoid has an effect similar to T:% and down-regulates activity of the c-fos promoter. Interest- ingly, a similar type of regulation by RA h a s been described for the epidermal growth factor receptorlerbB gene. RA increases activity of the epidermal growth factor receptor promoter, while in the presence of transfected RAR the retinoid significantly inhibits this activity (25). These results suggest that receptor concentration and/or its interaction with other factors, which may be cell-specific, modulate the type of cellular response to RA. This is further supported by the description that the c-fos promoter is not repressed by RA in HeLa cells (9). In any case, the inhibition of the c-fos promoter after overexpression of the RA receptor is specific, since the positive response of t he GH promoter to RA is further enhanced under these conditions.

The f inding that the transcriptional response of c-fos to the tumor promoter TPA was strongly repressed in T,-treated cells, as well as in "treated cells after co-transfection with the RA receptor, suggests that the inhibitory response of the receptors could be mediated by AP-1 binding sites as recently demon- strated for the collagenase or stromelysin genes (6,7). We show here that the receptors not only repressed the response of the c-fos promoter to TPA, a signal acting through the protein ki-

A' - . . .

151 5 0 172 598 175 613 815 350 748

B) , .

Va- 256 208 232 797 812 760 833 9 2 6 9 1 0

FIG. 7. Influence of a truncated thyroid hormone receptor on the inhibition of the c-fos promoter by T,. The cells were co-trans- fected with 5 pg of the c-fos promoter construct and 2.5 pg of Rous sarcoma virus-c-erhA in the absence (panel A ) or presence (panel B ) 10 of pg of an expression vector for a truncated c-erbA. CAT activity was determined after incubation with 1 nM T, for 48 h and with 15 PM forskolin or 100 nM TPA for the last 8 h. The autoradiographs show the conversion of I 14Clchloramphenicol to i t s acetylated forms.

A) B)

IpSVtorl C ~ R A c T 3

0

m m e*

0 0

(fold induction) **' 3'7 CAT .etivHy

1 0.6

nase C pathway, but also inhibited the response to forskolin, which acts via a CAMP-dependent pathway.

The antagonism between the TR and RA receptors and c-Fos was mutual, since increasing endogenous A P - 1 complex by transfection with expression vectors for c-fos or c-fos plus c jun inhibited the response of the GH promoter to both ligands. This repression is not specific for the GH promoter. since other au- thors have shown that the response to both ligands of other constructs containing TR- or RA-responsive elements under control of viral promoters is also reduced in the presence of transfected c-fos (9, 10). We also show that the repressive effect of c-fos can be overcome by an antisense c-fos vector, thus show- ing that this antagonism depends on the concentration of the different factors implicated.

Negative regulation of gene expression by the ligands of the nuclear receptor superfamily by transcriptional interference with the A P - 1 complex h a s been observed in several instances. The main conclusion from this study is that the thyroid hor- mone receptor is able not only to inhibit AP-1 activity as re- cently shown (10) but also to effectively block c-Fos expression and, therefore, the formation of A P - 1 complexes. A decrease of c-fos should amplify the inhibitory effect of the receptors on AP-1 activity, which in turn will affect regulation of genes con- taining AP-1 sites, which are involved in important cell func- tions. The process of cell proliferation is under a very complex regulation in which the transcription factor A P - 1 seems to play an important role. Although expression of c-Fos is normally associated with the response to a variety of mitogenic signals and tumor promoters, i t can also be stimulated by differenti- ating agents that inhibit cell proliferation. This indicates the role of this transcriptional factor on the regulation not only of cell growth but also in differentiation processes. CHI cells con- tain comparatively high levels of T3 receptors, and these recep- tors mediated an important differentiated function of these cells, Le. the transcription of the GH gene. Different authors have emphasized that the antagonism of the nuclear receptors with the AP-1 complex is related with the anti-proliferative and anti-neoplasic effects of the ligands for the different receptors. However, in our cell model, TR significantly decreases the ex- pression of the c-fos gene despite significantly increasing cell proliferation. Therefore, a simplistic model in which A P - 1 in- creases proliferation and the nuclear receptors exert anti-pro- liferative effects via inhibition of c-foslcjun cannot be main- tained.

A deletion mutant of t h e TB receptor, which lacks the DNA binding domain and is transcriptionally inactive, has been shown to act in a dominant-negative fashion to block trans- activation by the native receptor (19). This mutant also abol- ished the repression of t h e c-fos promoter by T3. Interestingly, a recent observation from Desbois et al. (26) shows that v-erbA

C ) D) B I p s v l o r + P s v ~ l

RA c T 3 R A c n R A

* 0

0.6 1 3.4 3.3 1 4.4 4..0

FIG. 8. Influence of c-fm on the response of the growth hormone promoter to Ts end Rk CAT activity was determined in CHI cells transfected with 5 pg of -530GH-CAT and treated for 48 h with 1 nM T,, or 1 PM RA. In panel A, the cells were transfected with the reporter plasmid alone. In panel B, the reporter plasmid was co-transfected with 2 pg of an expression vector for c-fos IpSVfos). In panel C . the cells wrre transfected with 8 pg of an antisense c-fos expression vector (pSVsoQ in panel D, the cells received 2 pg of pSVfos + 8 pg of pSVsof expression vectors. The figure illustrates the conversion of [ ' T I chloramphenicol to its acetylated forms in the different transfections. and the quantification of the data expressed as the induction over the values obtained in the corresponding control untreated groups.

Repression of c-fos by Nuclear Receptors 23543

has an effect similar to that produced by the truncated c-erbA, since v-erbA abrogates the inactivation of an AP-1 element by thyroid hormone receptors. The carboxyl terminus of the T3 receptor and other members of the receptor superfamily con- tains regions necessary for homo- and heterodimerization. When expressed in excess, the truncated receptors could medi- ate the formation of inactive heterodimers with the native re- ceptors or titrate other proteins such as the retinoid X receptor, which is thought to be required for c-erbA activity (27-30). Depending on the relative affinity of the receptor for interac- tion with the mutant, other receptors, or even the components of the AP-1 complex, the mutant receptor can inhibit the effect of the native receptor. Although at present the domains of the receptors involved in this interaction have not been well de- fined, it has been shown that a truncated thyroid hormone receptor that lacks only 17 amino acids from the carboxyl ter- minus can no longer antagonize AP-1 activity (lo), whereas multiple functional domains appear to be required for optimal anti-A€'-1 activity of the RA receptor (9).

In conclusion, increasing evidence is being obtained on the extensive cross-talk between the nuclear receptors and the signal transduction pathways that modulate gene transcrip- tion in response to growth factors, hormones, tumor promot- ers, and several oncogenes. Regulation of c-fos gene expression by thyroid hormone and RA receptors provides a specific mechanism by which their ligands can modify the cellular re- sponse to a variety of external stimuli acting through two main signal transduction systems: protein kinase A and pro- tein kinase C.

Acknowledgments-We are grateful to Dr. R. Evans, H. Samuels, and I. Verma for providing plasmids and to A. Villa for expert technical assistance. We also gratefully acknowledge the help of T. Palomino with the gel retardation experiments.

REFERENCES 1. Evans, R. M. (1988) Science 240, 889-895 2. Bedo, G., Santisteban, P., and Aranda, A. (1989) Nature 339,231-234 3. Umesono, K., Guiguere, V., Glass, C. K., Rosenfeld, M. G., and Evans, R. M.

(1988) Nature 336, 262-265 Nature 355, 441446

4. Umesono, K., Murakami, K. K., Thompson, C. C., and Evans, R. M. (1991) Cell 65, 1255-1266

5. Naar, A. M., Boutin, J.-M., Lipkin, S. M., Yu, V. C., Holloway, J. M., Glass, C. K., and Rosenfeld, M. G. (1991) Cell 65, 1267-1279

6. Nicholson, R. C., Mader, S., Nagpal, S., h i d , M., Rocbette-Egly, C., and Cham- bon, P. (19901 EMBO J . 9,44434454

7. Lafyatis, R., Kim, S. J . , Angel, P., Roberts, A. B., Sporn, M. B., Karin, M., and

8. Schule, R., Rangarajan, P., Yang, N., Kliewer, S., Ransone, L. J., Bolado, J., Wilder, R. L. (1990) Mol. Endocrinol. 4, 973-980

Verma, I. M., and Evans, R. M. (1991) Proc. Natl. Acad. Sci. U. S . A. 88, 6092-6096

9. Yang-Yen, H.-F., Zhang, J.-K., Graupner, G., Tzukerman, M., Sakamoto, B.,

10. Zhang, X. K., Wills, K. N., Husmann, M., Hermann, T., and Pfhal, M. (1991) Karin, M., and Pfhal, M. (1991) New Biol. 3, 1206-1219

Mol. Cell. B i d . 11, 601G6025 11. Angel, P., Imagawa, M., Chiu, R., Stein, B., Imbra, R. J., Rahmsdorf, H. J. ,

12. Ransone, A. J., and Verma, I . M. (1990) Annu. Reu. Cell Biol. 6,539-577 Jonat, C., Herrlich, P., and Karin, M. (19871 Cell 49, 729-739

13. Herschman, H. R. (1991)Annu. Reu. Biochem. 60, 281-318

15. Perez, P., Sanchez-Pacheco, A., Bedo, G., Pascual, A,, and Aranda, A. (1990) in 14. Samuels, H. H., Stanley, F., and Casanova, J. (19791 Endocrinology 105,8045

16. Sassone-Corsi, P., and Verma, I . M. (1987) Nature 326, 507-510 17. Konig, H., Ponta, H., Rahmsdorf, U., Buscher, M., Schonthal, A., Rahmsdorf,

18. Forman, M. B., Yang, C., Stanley, F., Casanova, J., and Samuels, H. H. (1988) H. J., and Herrlich, P. (1989) EMBO J. 8,2559-2566

19. Forman, B. M., Yang, C., Au, M., Casanova, J., Ghysdael, J., and Samuels, H. Mol. Endocrinol. 2, 902-911

20. Schonthal, A., Buscher, M., Angel, P., Rahmsdorf, H. J., Ponta, H., Hattori, K., H. (1989) Mol. Endocrinol. 3, 1610-1626

21. Ye, Z., Forman, B. M., Aranda, A., Pascual, A,, Park, H., Casanova, J . , and Chiu, R., Karin, M., and Herrlich, P. (1989) Oncogene 4, 629-636

22. Curran, T., and Verma, I. M. (1984) J. Virol. 135, 218-228 Samuels, H. H. (1988) J. B i d . Chem. 263, 7821-7829

23. Andrews, N. C., and Faller, D. V. (1991) Nucleic Acids Res. 19, 2499 24. Garcia-Villalba, P., Au, M., Samuels, H. H., and Aranda, A. A. (19931 Biochem.

25. Hudson, L. G., Santon, J. B.. Glass, C. K., and Gill, G. N. (1990) Cell 62, Biophys. Res. Comrnun. 191, 580-586

26. Desbois, C., Aubert, D., Legrand, C., Pain, B., and Samarut, J. (1991) Cell 67, 1165-1175

27. Yu, V. C., Delsert, C., Andersen, B., Holloway, J . M., Devary, 0. V., Naar, A. M., 731-740

Kim, S. Y., Boutin, J., Glass, C. K., and Rosenfeld, M. G. (1991) Cell 67, 1251-1266

28. Leid, M., Kastner, P., Lyons, R., Nakshatri, H., Saunders, M.. Zacharewski, T., Chen, J., Staub, A., Gamier, J., Mader, S., and Chambon, P. (1992) Cell 68,

29. Kliewer, S. A., Umesono, K., Mangelsdorf, D. J., and Evans, R. M. (1992) 377-395

30. Zhang, X. K., Hoffmann, B., Tran, P. B.-V., Graupner, G., and Pfhal, M. (1992) Nature 355,446-449

20th FEBS Meeting, Abstracts, P-Th 050