Thyroid Hormone Regulates the Oxytocin Gene - The Journal of ...

Vol. 267. No. 6 , Issue of February 25, pp. 3771-3777,1992 Printed in U.S.A.

THEJOURNAL OF BIOLOGICAL CHEMISTRY 0 1992 by The American Society for Biochemiatry and Molecular Biology, Inc.

Thyroid HormoneRegulates the Oxytocin Gene* (Received for publication, October 3, 1991)

Roger A. H. Adan#, Joke J. Cox, Jorge P. van Kats, andJ. Peter H. Burbachg From the Rudolf Magnus Institute, Departmentof Pharmacology, Medical Faculty, Universityof Utrecht, Vondellaan6, 3521 GD Utrecht, The Netherlands

Oxytocin (OT)’ is a biologically active peptide with horEndocrinefactors involved inthetranscriptional regulation of the oxytocin (OT) gene were investigated monal actions related to the regulation of reproduction and in heterologous expression systems.Plasmids havinga with actions on the central nervous system. OT stimulates 5”flanking region of the rat OT gene (-363/+16) or milk ejection (Soloff, 1985) and prolactin release (Samson et the human OT gene (-382/+41) cloned in frontof the al., 1986), and induces uterus contraction (Fuchs, 1986). Horfireflyluciferasegenewereco-transfectedwith a n monally active OT is mainly produced in two hypothalamic expression vector forthe rat thyroid hormone receptor nuclei, the supraoptic nucleus and the paraventricular nu(Y in P19 embryonalcarcinoma (EC) cells. Thyroid cleus,which transport OT and its associated neurophysin hormone (T3) stimulated the activity of the rat and axonally to the posterior lobe of the pituitary gland, from human OT promoters about 10-fold. In MCF-7 breast where it canbe secreted into the circulation. The OT-expresstumor cells transfected with the humanOT promoter- ing neurons in the adult brain have high levels of OT mRNA. luciferase fusion gene,T3 stimulation through endog- OT gene expression is markedly up-regulated during develenous thyroid hormone receptors was about 5-fold. Co- opment in the first month of neonatal life (Almazan et al., transfection experiments in P19EC cells using 5’ dele- 1989; Van To1 et al., 1988a, 1990). The OT mRNA levels are tion mutants of the rat OT gene showed that thyroid increased during late pregnancy, during lactation, and upon hormone responsiveness was located in two regions, salt loading (Van To1 et al., 1987, 1988b; Zingg and Lefebvre, one located between nucleotides -195 and -172, the 1988). Hormones are thought t o play a role in the regulation other betweennucleotides -172 and-148. Each region of the response of hypothalamic OT neurons to altered physaccounted for about %fold T3 stimulation. Gel retar- iological and endocrine states, but it is not knownwhich dation analysis using extracts from HeLa cells over- hormones are involved. To understand the regulation of the producingthec-erbA/TRaprotein showedspecific OT system, endocrine factorsinvolved in transcriptionalregbinding to the -1721-148 element, while no binding ulation of the OTgene are being studied. One such hormone, occurred on the -1951-172 element. The -1721-148 estrogen, has been studied and theestrogen responsiveness of elementwhichcontains the imperfectestrogen re- the rat and human OTgenes has been revealed (Richard and sponse element, GGTGACCTTGACC, has inverted as Zingg, 1990; Burbach et al., 1990; Adan et al., 1991). The well as direct repeatsof the TGACC motif. Mutagenesis estrogen-response element (ERE) in the 5’-flanking region of of TGACC motifs separately reduced thyroid hormone these OT genes has the sequence GGTGACCTTGACC and responsiveness by about 50%. However, simultaneous is located at -168 to -155 from the transcription initiation mutation of two TGACC motifs abolished the respon- site of the rat OTgene. It has one mismatch with theperfect siveness to T3completely. There was no cooperativity ERE GGTCAnnnTGACC (Beato,1989). The human and rat between the activated thyroid hormone and estrogen 5’-flanking regions of the OT gene have a high degree of receptors in transfected MCF-7 cells nor in thyroid homology (Ivell and Richter, 1984; Sausville et al., 1985) (Fig. hormone receptor and estrogen receptor co-transfected 1). The ERE also contains a direct TGACCrepeat. The P19EC cells. Negative interactions between these two TGACC sequence is also present at -103 and -83, and as the receptors were observed and gel retardation assays inverted motif GGTCA at -187 in the rat OT upstream region showed interaction between the two receptors pro(Ivell and Richter, 1984). TGACC motifsare integral partsof teins. It was shown in an in vivo experiment that consensus sequences for thyroid hormone-response elements treatment of rats with thyroid hormone increased hy- (TREs) that have beendescribed such as the inverted sepothalamic OT mRNAlevels, the pituitaryOT content, quenceGGTCA (Day and Maurer, 1989; Sap et al., 1990), as well as OT levels inblood. The results reveal thyroidAGGT(C/A)A (Brent et al., 1989, 1991), and TCAGGTCA hormone as a physiological regulator of OT gene (Naar et al., 1991). Furthermore, the thyroidhormone recepexpression, which stimulatesOT promoter activity di- tor is known to bind an ERE, thereby inhibiting estrogenrectly through interaction with a thyroid hormonedependent transactivation (Glass et al., 1988).Recently, it response element in theOT gene. was shown that thyroid hormone receptors, in particular the a receptor, are expressed at high levels in thesupraoptic nucleus and at lower levels in the paraventricular nucleus of therathypothalamus(Bradley et al., 1989). This finding * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “uduertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ Supported by Netherlands Organization for Scientific Research Project 900-546-065. $ T o whom correspondence should be addressed. Tel.:31-30880521; Fax: 31-30-896034.

The abbreviations used are: OT, oxytocin; EC cells, embryocarcinoma cells; ERE, estrogen-response element; PTU, 6-n-propyl-2thiouracil; Tat 3,5,3’-~-triiodothyronine; T,, 3,5,3’,5’-tetraiodothyronine; TRE, thyroid hormone-response element; TSH, thyroid-stimulating hormone; Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid; RSV, Rous sarcoma virus; LTR,lateterminal repeat; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

3771

Expression Oxytocin Gene Hormone Thyroid and

3772

of pRSVGAL was included in the co-transfectionin all experiments. The next day the medium was changed and thecells received 100 nM 17P-estradiol and/or 100 nM thyroid hormone (T3, 3,5,3’-~-triiodothyronine)as used before inco-transfectionassays(Kumarand Chambon, 1988; Thompson and Evans,1989) with ethanol as vehicle. Controls received ethanol only. After 24 h the cells were harvested in 420 p1 of 100 nM potassium phosphate (pH 7.8), and luciferase activity was measured in 100 p1 of extract according to the protocol of De Wet et al. (1987), using a Lumac/Sm biocounter M2010A luminometer. Galactosidase activity was determined according to Sambrook et Ql. (1989). MCF-7 cells were cultured and transfected, according to the protocol of Seiler-Tuyns et al. (1986), as described previously (Burbach et al., 1990). Transfected MCF-7 cells were treated with 100 nM 17PMATERIALS ANDMETHODS estradiol and/or 100 nM thyroid hormone for 24 h. The cells were harvested and theluciferase activity measured as described above. Plasmids-pl9LUC is a derivative of pSVOAL-A 5’ and contains Gel Retardation Assays and Nuckar Entracts-Oligonucleotides a multiple cloning site in frontof the luciferase gene (Van Zonneveld used for gel retardationexperiments were 5”ACCTGAGGCGGTregion (Ivell and Richter, et al., 1988). The -363/+16 rat OT upstream GACCTTGACCCCAGCCCAG-3’ and its complementary strand 5‘1984) was cloned intopl9LUCas a HindIII-Sau3Afragmentto TCGACTGGGCTGGGGTCAAGGTCACCGCCTC-3’, inthetext construct pROLUC. p-l95ROLUC, p-l72ROLUC, and p-148ROLUC referred to as the -172/-148 element. A restrictionfragment are 5”deletion mutants of pROLUC (in Fig. 1 positions -195, -172, and -148 aremarkedwithasterisks), which were made by the (HindIII-BstEII) from p-195ROLUC (nucleotides -195 to -163) was also used for gel retardation assays. The complementary oligos with polymerase chain reaction. The region of the putative EREbetween 5”overhang and the restriction fragment were labeled by filling in -172 and -148 in p-172ROLUC was mutated using polymerase chain with Klenow DNA polymerase to a specific activity of approximately reaction. The oligos toobtainpGGTTROLUC,pCCTTROLUC, 50,000 cpm/ng of DNA.Labeled fragments were purified on 5% pCAROLUC, and pCAGGTTROLUC were, respectively: 5’polyacrylamide gel electrophoresis gels. For competition, the palinGGATCCCAAGCTTAGGCTTI’GACCTTGA, 5”ATCCCAAGCT- dromic T R E sequence (TREp) asdescribed by Umesono et al. (1991), TAGGCGGTGATTTTGAC~CCAGC, 5“GGTCATAAGCT5’-AAGCTTTCAGGTCATGACCTGAGAATTC, andits compleTAGGCGGTGAmTTGAACCCAGCCC, and 5”GGTCATAAGCT- mentary strand were used. TAGGCTEGACCTTGAZCCCAGCCC (mismatches with pNuclear extractsof HeLa cells infectedwith vaccinia viruscarrying 172ROLUC are underlined). The nucleotide sequence of all polymthe CEA-I1 cDNA encoding the chicken c-erbA/TR-a or its control erase chain reaction-generated constructs was confirmed bysequence (wild type nuclear extract) were obtained from Dr. H. Stunnenberg analysis. pHOLUC was constructed by cloning the -382/+41 up(EMBL, Heidelberg Germany). The preparation of these extracts is stream region of the human O T promoter (Sausville et al., 1985) as a described by Sap et al. (1990). Nuclear extracts of cells infected with BamHI-NaeI fragment in front of the luciferase gene in pl9LUC. bacculovirus carrying the estrogen receptor and its control were a pRSVLUC has the RSV-LTR in front of the luciferase gene (De Wet kind gift of Dr. M. G. Parker (Molecular Endocrinology Laboratory, et al., 19871, whereas pRSVGAL has the RSV-LTR in front of the London, UK). The preparation of these extracts described is byFawell LacZ gene. pHEO is a plasmid having the human estrogen receptor et al. (1990). coding region cloned in the eukaryoticexpression vector pSG5 (KuTypical binding reactions had a volume of 20 pl and contained 2 mar et al., 1986). pRSrTRa is an expression vector for the rat a - pg of poly [d(I-C)],1 pl of nuclear extract, 60mM KC1, 15 mM Hepes thyroidhormone receptor (Thompsonet d , 1987). pGEM4is a (pH 7.81, 1 mM dithiothreitol, 0.02% Nonidet P-40, and 0.2 ng of commercially available cloning vector(Promega Corp., Madison, WI). labeled DNA. This mixture was incubated on ice for 1 h followed by Transfection and Luciferase A s s Q ~ s - P ~EC ~ cells (McBurney et an incubation a t 25 “C for 30 min. The reaction mixture was loaded al., 1982) were cultured in Dulbecco’s modified Eagle’s medium with- on a 10% nondenaturing polyacrylamide gel electrophoresis gel in 0.5 out phenolred, and supplemented with steroid-free (charcoal treated) X Tris borate EDTA supplemented with 0.02% Nonidet P-40 and 7.5% fetal calf serum. The daybefore transfection the cells were run a t 10 V/cm a t 4 “C. The gel was then dried and exposed to x-ray plated to a 10% confluent density in 6-cm culture dishes. The cells film. were transfected overnight with thecalcium phosphate precipitation Treatment of Animak+“ale Wistar rats (150-180 g) were used in method (Vander Eb and Graham,1980). Briefly,5 pg of OT promoter all experiments. T o obtain hypothyroid animals, rats were given orally luciferase plasmid was co-transfected with 1 pg of pRSrTRa, pHEO, 2 mg of PTU in 1 mM NaOH/100 g body weight daily. Hyperthyroid or pGEM4 or combinations of these. As an internal control, 200 ng rats were given orally a daily doseof 3.5 pg of T d in 1 mM NaOH/100 g body weight. Control animalswere given orally a daily doseof 1 ml -217 of solvent (1 mM NaOH) per 100g body weight. All treatments were RAT TCCCCTTC TAGGCTGTGTCCCTTTTGAGCTCAGGTUT carried out orally via a stomach tube for a period of 3 weeks. At the IIII IIII IIII Ill HUM TCCCTTCCGCAAGGCACCTCACCTTCTGTGCCCAGACCAT end of the experiments animals were decapitated after ether anaes-213 -16R thesia and trunkblood was collectedin tubes containing200 pl of 0.1 RAT TAGCTGAGGCGGTGACCTTOCCCAGCCCAGACCCTGCA M EDTA, 0.15 M NaCl. Plasma was prepared by centrifugation (2500 Ill1 IIIIIIIIIIIIIIIII IIIIII IIIIII HUM TAGCCAACGCGGTGACCTTUCCCGGCCCAGGCCCTGCT rpmfor 10 min,4 “C) and stored at -20 “C for determination of ERE hormone levels. The whole brain was removed and the rostral partof PAT AATGAAGGGCCTGCTTCTAAACAGTGTGGAACAGTTTGACC the hypothalamus containing the PVN and SONwas dissected. The IIIIIII IIIII whole pituitary gland was collected. Tissues were frozen on dry ice HUM AATGAAGAGGAAAGCCCGTACGCACTCGGCC TGACC and stored at-70 “C. Determination of Hormones and Peptides-Plasma T,, Ts, and RAT AAAGAGACCTGGCTGTGACCAGTCATGCAGTCACCCTCT IIIIIIIII IIII IIIII T S H levels were determined using a radioimmunoassay. Plasma OT HUM CACGGCGACCCTCTGTGLCCCATACTACCAACCTCT levels were determined using a radioimmunoassay as previously described (Van de Heijninget al., 1991). PituitaryOT levels were PAT TAGACTGGGCCCCACCATGGCAGTGGACAAGGCATAAXLA determined by radioimmunoassay in acidic extracts preparedaccordII IIIII IIIIIIIIIII HUM TAAACAGAGCTCCACCGCCGCAATGCCC AGGCATAAXLA ing to Burbach andBin Liu (1989). TATA BOX Northern Blot Analysis of Hypothalamic RNA-RNA was isolated FIG. 1. Homologous elements in the 5‘-flanking region o f using the method of Wilkinson (1988). All RNA samples were denathe rat a n d h u m a nOT gene. The sequences upstream of the TATA tured withglyoxal and dimethylsulfoxide, run througha 1.4% agarose box are aligned. TGACC motifs often present in binding sitesfor the gel, and transferred to a nylon membrane as described before (Van thyroid hormone receptor are in bold. Asterisks indicate the 5’-endof To1 et al., 1989). DNA probes were labeled by random priming. The the used deletion mutants of the rat OT gene (see Fig. 5 and text). OT probe was a 219-base pair fragment having rat OT exon C (Van Only homology of four or moreserial basepairsis shown. The To1 and Burbach, 1989). A glyceraldehyde-3-phosphate dehydrogenidentified ERE is underlined (Adan et al., 1991). Sequences are from ase (GAPDH) probe was obtained by labeling of a 1269 cDNA (Fort et al., 1985).The GAPDH insertwas a kind gift of Dr. H. B. Nielander Ivell and Richter (1984) and Sausville et al. (1985). together with the sequence motifs inthe 5“flanking region of the OT gene raised the question whether thyroid hormone is involved in transcriptional regulation of the OT gene. Here it is reported that the h u m a n and rat OT 5”flanking regions significantly confer thyroid hormone responsiveness to a reporter gene (luciferase) in heterologous expression systems and a TRE is characterized in the rat OT gene.Thyroid hormone treatment of rats increases hypothalamic mRNA and OT peptide in uiuo, indicating that thyroid hormone is a sofar unrecognized hormonal mediator of OT gene regulation in the brain.

.

” _

Thyroid Hormone and Oxytocin Gene Expression

3773

(IMB, Utrecht, The Netherlands). Hybridization was performed as described by Van To1 and Burbach (1987). Signals on autoradiograms were quantified by laser densitometry using an UltrascanXL(LKB).GAPDHmRNA was used as a standard (Piechaczyk et al., 1984). OT mRNA levels were expressed as ratio to GAPDH mRNA. There were no differences in total yield of RNA or GAPDH mRNA content between the three groups. Data are presented as mean& standard error. Statistics-Statistical analysis was performed by one-way analysis of variance (ANOVA) at thep < 0.05 level.

T

RESULTS

Activation of the Rat and Human OT Gene Promoters by Thyroid Hormone-When the rat OT promoter-containing luciferase construct pROLUC (-363/+16) was co-transfected a expressionplasmid withthe thyroidhormonereceptor pRSrTRa in P19 EC cells, treatment of cells with T3caused a significant increase in luciferase activity.The average stimulation over five experiments was 10-fold (Fig. 2). A similar stimulation of luciferase activity was found using the human OT promoter-luciferase construct pHOLUC(-382/+41) (Fig. 2). ThusbothratandhumanOT genes showed marked responsiveness to thyroid hormone in this co-transfection assay. The human breast tumor cell line MCF-7 contains endogenous thyroid hormone receptors (Fukuda et al., 1988). T3 stimulated luciferase activity about 5-fold inMCF-7 cells transfected with pHOLUC only (Fig. 3). No response was obtained in nontransfected cells. The results indicate that endogenous levels of thyroid hormone receptors were also able t o enhance OT promoteractivity similarly as in receptor cotransfected cells. Identification of a TRE-To locate DNA elements of the rat OT upstream region that mediate the response to T3, 5’ deletions in pROLUC were made around the ERE at -168 to -155 (Fig. 1).These concerned deletion of the regions upstream of nucleotides -195,-172, and -148 (plasmids pl95ROLUC, p-l72ROLUC, and p-l48ROLUC, respectively). Co-transfection of each of these plasmidswith pRSrTRa in P19 EC cells showed no loss of responsiveness to thyroid hormone when the region upstream of -195 was

+T

T

T3

control

E

T3+E

FIG. 3. Stimulation of the human OT promoter by TSand 17P-estradiol through endogenous receptors. MCF-7 cells were transfected with pHOLUC and treated overnight with 100 nM T3 (+T) and/or 100 nM 0-estradiol (+E). Luciferaseactivity of T:%treated cells was expressed as fold induction as compared to controls that received only. The average value k standard deviation is shown.

FOLDINDUCTION

T ERE

pROLUC .1 -

GGTCA-

GGTGACCTTGACC

p- 195ROLUC -,.I

-GGTCA-

GGTGACCTTGACC

p- 172ROLUC “E

GGTGACCTTGACC

n LUC

2.9

p- 148ROLUC

-’*m

0.9

FIG. 4. Localization of thyroid hormone-responsive elements in the 5”flanking region of the rat OT gene. P19 EC cells were co-transfected with pROLUC (-363/+16), p-195ROLUC (-195/+16), p-172ROLUC (-172/+16), or p-148ROLUC (-148/+16) and a plasmid expressing the rat thyroid hormone 01 receptor and treated overnightwith 100 nM T:i (2’).The sequences of the ERE and the GGTCA motif are shown. Luciferase activity of T3-treated cells, corrected for transfection efficiency with galactosidase,was expressed as fold induction (LUC, luciferase gene).

deleted (Fig. 4). p-172ROLUC was on average 3-fold less responsive to TSstimulation than p-195ROLUC. On average over three separate experiments the TSresponsiveness of p172ROLUC accounted for a 3-fold stimulation of promoter activity. In p-l48ROLUC, TSresponsiveness was completely pHOLUC pHOLUC pROLUC pROLUC lost (Fig. 4). Thus, responsiveness was located in the regions FIG. 2. Stimulation of the rat and human OT promoters by spanning nucleotides -195 to -173 and -172 to -148. thyroid hormone. P19 ECcells were co-transfected with pROLUC To investigate next whether these two elements of the rat (a construct containing the-363/+16 upstream region of the rat OT OT gene actually bindtothethyroid hormonereceptor, gene in front of the luciferase gene) or pHOLUC (a construct with nuclear extracts from c-erbA/TRa vaccinia-infected HeLa the -382/+41 upstream region of the human OT gene in frontof the cells were used for gel retardation assays. Extracts from HeLa luciferase gene), togetherwithaplasmid that expressed therat thyroid hormone a receptor and treated overnight with 100 nM Ta cells infected with wild type vaccinia virus served as control. (+T). Luciferaseactivity was corrected for transfection efficiency Gel retardation analysisusing the labeled -172/-148 element with galactosidase activity and the effect of T, was expressed as fold revealed three retarded bandsof which two were also present induction ascompared to cells that only received vehicle. The average in control extracts(complexes A and B; Fig. 5 ) . A specifically fold induction f standard deviation is shown. The basal luciferase activity obtained with pROLUC and pHOLUC was similar. Data of retarded band (complex C, Fig. 5) was seen only in the lanes nuclear extract. This pROLUC are from five separate experimentswith a total n = 12, and containing the c-erbA/TRa enriched band was competed with a 50-fold molar excess of the TRE for pHOLUC with n = 3.

3774

Thyroid Oxytocin Hormone and 1 2 3 4 5 6 7 8

ZwE

-A

B

C

FIG.5. Binding of the thyroid hormone receptor binding to the -172/-148 element of the rat OT gene. The -172/-148 region of the rat OT gene was labeled and incubated with a nuclear extract from HeLa cells infected with a vaccinia virus expressing the c-erbA/TRa protein (lanes 5-8) or with a control nuclear extract from cells infected with wild type vaccinia virus (lanes 1-4). A competition with increasing amounts of TRE palindrome was performed 50-fold molar excess (lanes 1 and 8 ) , 5 M excess (lanes 2 and 7),and 0.5 M excess (lanes 3 and 6).

-172ROLUC AGGCbGTGACCTTGAC&C

-

2.9

GGTTROLUC

TT

1.4

F "

CCTTROLUC CAROLUC CAGGTTROLUC -148ROLUC

1.4

TT

TT

A

2.0

A

1.1 0.9

FIG. 6. Mutational analysis of the -172/-148 element of the rat OT gene. The TGACC motifs in the -172/-148 sequence were mutated asshown. Arrours indicate the position and orientation of the TGACC motifs. The thyroid hormone responsiveness of these constructs was determined as described in the legend to Fig. 4.

palindrome as described by Umesono et a1. (1991). No retardation was foundwith the labeled -195/-163 restriction fragment as a probe. Thus, element -172/-148 confers T3 responsiveness and binds the c-erbA/TRaprotein. To further delineate the relevance of TGACC sequence motifs within the element -172/-148 for the T3 responsiveness, mutations were made in p-172ROLUC (Fig. 6). Mutants in which the left-half palindrome GGTGA (nucleotides -168 to -163) was mutated into TTTGA (p-GGTTROLUC), the left TGACC motif (nucleotides -166 to -161) was converted into TGATT (pCCTTROLUC), or the right-half palindrome TGACC (nucleotides -160 to -155)was converted into TGAAC (pCAROLUC) all showed decreased response to T3. The decrease was approximately 50% for all mutants (Fig. 6). A mutant in which both half-palindromes were mutated simultaneously (pCAGGTTROLUC, where the mutations of pGGTTROLUC and pCAROLUC are combined), completely lost T3 responsiveness (Fig. 6). Thyroid and Estrogen Activation of the OT Gene Promoter Interfere-Since the element between nucleotides -172 and -148 was previously shown to mediate stimulation by estrogen as well (Burbach et al., 1990; Adan et al., 1991), it was of interest to investigate whether and how the activation of the rat OT promoter by T3 is related to the estrogen-induced activation. The human breast tumor cell line MCF-7, which contains endogenous estrogen as well as thyroidhormone receptors (Fukuda et al., 1988; De Launoit and Kiss, 1989), was transfected with pHOLUC and treated with TI and/or

Gene Expression 17P-estradiol. The stimulation of the human OT promoter activity by 17@-estradiolwas only slightly higher than by TI. There was no additive effect of simultaneous treatment with T3and 170-estradiol (Fig. 3). In subsequent experiments the expression plasmids for the humanestrogen receptor (pHEO) andthe expression plasmid for therat thyroid hormone receptor (Y (pRSrTRa) were co-transfected with the rat OT promoter constructsin P19EC cells, in the presence or absence of ligands. The stimulation of the rat OT promoter activity by 17P-estradiol was stronger than by T3: in over five experiments the estrogen stimulation was 191 & 8-fold and the T3 stimulation 9.6 k 2-fold. Simultaneous treatment of transfected cells with T3 and 17P-estradiol had no additive effect (Fig. 7A). Rather, thestimulation by 17P-estradiol was significantly suppressed when the unoccupied thyroid hormone receptor was present (Fig. 7A). The occupied thyroid hormone receptor caused a small decrease in estrogen stimulation. The activation of OT promoter activity by Ta was markedly suppressed by about 60% when the unoccupied estrogen receptor was present (Fig. 7B). The same experiments carried out with the promoter-less vector pl9LUC and the RSV-LTR-driven luciferase gene in pRSVLUC showed no effects of T3 or 17P-estradiol on luciferase activity per se nor any interactions. The possible interaction between the thyroid hormone receptor and the estrogen receptor was further analyzed by gel retardation using the labeled -172/-148 element as probe. The gel retardation assay revealed one specific retarded complex (complex C inFig. 8) which occurred only when a nuclear extract enriched inestrogen receptor was mixed with anuclear extract enriched in c-erbA/TRa protein prior to addition of the labeled -172 to -148 fragment. Thyroid Hormone Stimulates OT Gene Expression in Vivo-

--

A

+T

TF

TFI

IR+ER

FIG. 7. Interference of thyroid hormone and estrogen for the stimulation of the rat OT promoter. PI9 EC cells were cotransfected with pROLUC and plasmids that expressed the rat thyroid hormone a receptor (TR)and/or the human estrogen receptor ( E R ) , and treated overnight with 100 nM Ts (+TIand/or 100 nM 170-estradiol (+E).Panel A shows that the unoccupied thyroid hormone receptor suppressed the estrogen-induced activation. The same occurs to a lesser extent for the ligand-occupied thyroid receptor. The stimulation by estrogen in estrogenreceptor-transfected cells was 191-fold (average of five experiments). The average .t standard deviation isshown. Panel B shows that theunoccupied estrogen receptor suppressed the thyroid hormone stimulation of pROLUC. The stimulation by T:, in thyroid hormone receptor-transfected cells was 9.6fold (average of five experiments). Statistical significance was tested by ANOVA: p < 0.05.

Thyroid Hormone and Oxytocin Gene Expression

3775

Hypothalamio OT mRNA

*

31

k

e -D

I* 0 = 1

E

0

OT peptidelevel

FIG. 8. Interaction of c-erbA/TRa and estrogen receptorin binding to the -172/-148 element. The -172/-148 element of the rat OT gene was incubated with nuclear extracts enriched in cerbA/TRa protein ( l a n e 1 ), estrogen receptor ( l a n e 2), a combination of the two (lane 3) or a combination of the two control extracts ( l a n e 4 ) . Complex D is specific for c-erbA/TRa, complex B occurs in all lanes and complex A is broadened in extracts containing the estrogen receptor. Complex C is only seen in lane 3 and occurs only in the presence of both c-erbA/TRLu and estrogen receptor protein.

In order to evaluate the influence of thyroid hormone on hypothalamic expression of the OTgene, groups of rats ( n = 5 ) were treated orally with T4,PTU, or solvent for 3 weeks. T3, T , and TSH levels were measured to determine the thyroid status of the animals. The PTU treatment resulted in hypothyroidy: plasma levels of T3 (0.76 f 0.02 nmol/liter) and T4 (32.6 f 0.7 nmol/liter) were significantly lower than in the control group (1.11 f 0.05 and 71.6 f 6 nmol/liter, respectively) and plasma TSH levels were significantly higher (11.71 f 1.0 versus 1.14 f 0.1 ng/ml). The treatmentwith T4 had resulted in mildly hyperthyroid animals: plasma T4levels (118f 7 nmol/liter)were statistically significantincreased as compared to thecontrol group. The plasma levels of T3were 1.18 f 0.06 nmol/liter and of TSH, 0.89 f 0.1 ng/ml. The expression of the OTgene in thehypothalamo-neurohypophyseal system was analyzed in the three experimental groups at thelevel of hypothalamic mRNA, peptide stores of the pituitary gland, and circulating peptidelevels. OT mRNA levels of the hyperthyroid animalswere 1.6 times higher than the levels of control and hypothyroid animals (Fig. 9). The mean OT mRNA level of the hypothyroid group did not differ from that of control animals. Similarly, hyperthyroid rats had a significantly higher OT content in the neurointermediate lobe of the pituitary gland and a significantly higher plasma OT level than normal and hypothyroid rats. The levels of the PTU group did not differ from the control group. DISCUSSION

The datademonstrate that thyroid hormonestimulates OT gene expression through a direct influence of the T3-thyroid hormone receptor complex on OT gene promoter activity. It is further shown that the potential of the OT gene to be regulated by thyroid hormone is employed in vivo in hyperthyroid rats. Thus, thyroid hormone may be a physiological regulator of hypothalamic OT gene expression, involved in the signalling of peripheral information to OTneurons in the brain. The resultsonpromoteractivity were obtainedin two different heterologous expression systems: one in which the rat thyroid hormone receptor a was expressed via transient transfection of an eukaryotic expression vector, the other employing the endogenous thyroid hormone receptors in a

in NIL 0

OT peptldo b v d In blood

*

,,I-

CON PTU T4 FIG. 9. The effects of manipulation of the thyroid statuson OT expression in vivo in rats. Animals were chronically treated with PTU, T,, or vehicle ( C O N ) as described under “Materials and Methods.” Hypothalamic OT mRNA contents, OT peptide level in the neurointermediate lobe of the pituitary gland ( N I L ) , and OT levels in blood were determined. Statistically significant differences ( ‘ p < 0.05;ANOVA) were observed in the T,-treated rats as compared to control and PTU-treated rats.

human breast tumorcell line. In both systems there was low, but detectable expression of the luciferase gene when fused to the 5’-flanking regions of the rat andhuman OT genes in the absence of T3 and a marked stimulation by addition of T3. This effect was mediated specifically through the OT gene sequences since promoterless and RSV-LTR-fused luciferase gene constructs were unaffected by T S . The stimulation of OT promoter activity by T3 was thyroid hormone receptor-dependent, since the T3effect in P19 EC cells was only observed when the thyroid hormone receptor was expressed by cotransfection. Furthermore, endogenous thyroid hormone receptors in MCF-7 cells (Fukuda et al., 1988) could also mediate the T3 stimulation of the OT promoter, indicating that the T3 responsiveness of the OT gene can also be func-

3776

Thyroid Hormone

and Oxytocin Gene Expression

tional under normal physiological conditions of the cell. Thus far, thyroid hormone was not known to be involved in the regulation of OT systems. The present results provide the first evidence that the OT gene is a target for thyroid hormone actions. Treatment of rats with T.,, inducing a mild hyperthyroidy, increased OT mRNA level in the hypothalamus, and the OT peptide levels in pituitary and blood. The increase in hypothalamic mRNA content and pituitary peptide level indicate that gene expression and biosynthesis are enhanced by thyroid hormone, while the higher stimulated plasma level points to an increased biological availability of OT. There was no significant effect on any of these parameters in PTU-treated hypothyroid rats. This result suggests that there is no tonic influence of thyroid hormone under euthyroid conditions, but thatOT neurons respond to elevated thyroid hormone levels with increased gene expression and peptide synthesis. Other data indirectly supporting a physiological control of OT gene expression by thyroid hormone are: 1) the abundant presence of the thyroid hormone receptors mRNA in the hypothalamic nuclei expressing the OT gene (Bradley et al., 1989), and 2) the reduction in biosynthesis of neurosecretory material in thyroidectomized rats inthese nuclei (Talanti and Attila, 1972). We identify thyroid hormone as a hormone of peripheral origin responsible for the signaling of peripheral information to thebrain. Multiple sequence elements are involved in the thyroid hormone responsiveness of the rat OT gene.At least two regions can be indicated on thebasis of the 5"deletion studies, which are each responsible for a &fold stimulation of promoter activity by TI. One region is located between -195 and -173, the other between -172 and -148. In theformer region a single TGACC motif is present (nucleotides -184 to -179) but a DNA fragment having the -195/-173 region failed to bind to the c-erbA/TRaprotein. Thus, it may only function in conjunction with the -172/-148 element and involve accessory factors. Thelatter region containsthe motif GGTGACCTTGACC. Deletion of this region completely abolished T3 responsiveness of the OT promoter and this region binds to the c-erbA/TRa protein. Thus, the -172/ -148 element contains a functional TRE. TGACC motifs are part of several thyroidhormone-response elements (Glass et al., 1988; Brent et al., 1989, 1991; Sap et al., 1990). The sequence between -172 and -148 is a composite of three such motifs. It contains an inverted TGACC repeat (GGTGAcctTGACC) and a direct TGACC repeat (ggTGACCtTGACC).The former is analogous to the canonical ERE. It has been shown that thethyroid hormone receptor is able to bind a perfect ERE in a gel retardation assay, but fails to activate transcription (Glass et al., 1988). However, the -172/-148 element bindsthe thyroid hormone receptor and activates the OT promoter in response to Ts. Mutagenesis of each separate TGACC motif in the -172/ -148 element showed that each element contributed to the T3 responsiveness. The mutations concerned the Cs in the TGACC motifs since these reduce stimulation in other TREs (Brent et al., 1989; Umesono et al., 1991). However, with only one TGACC motif intact, no thyroidhormone responsiveness was retained. This result indicatesthat a single TGACC motif is not enough to confer thyroid hormone responsiveness. In addition, two single TGACC motifs are present downstream from the -172/-148 element (Fig. 1). Since -148 ROLUC was not sensitive to TJ, these two additional motifs did not confer T, responsiveness. The data show an interaction between thyroid hormone and estrogen activation of the OT promoter: the activation by both ligands is not additive, and theunoccupied estrogen-

receptor blunts thestimulation of the ligand-occupied thyroid hormone receptor and vice versa. Our observations are in line with previous reports concerning the action of thyroid hormone receptor a. It has been shown that the unoccupied thyroid hormone receptor a can bind a TRE (Dalman et al, 1990) and is able to repress transcriptional activity from promoters containing TREs (Hermannet al., 1991; Graupner et al., 1991). Furthermore, the thyroid hormone receptor binds a perfect ERE in a gel retardation assay (Glass et al., 1988). The unoccupied thyroid hormone receptor a has been shown to repress estrogen receptor stimulated expression from an imperfect ERE (Graupner et al., 1991). Different mechanisms of interaction have been proposed which can be related to our findings. First, it has been suggested that hindrance of accessibility can lead to repression (Graupner et al., 1991). This is in agreement with our data showing no additive effect of the two ligand-occupied receptors and inhibition of estrogen receptor-stimulated OT gene transcription by the unoccupied thyroid hormone receptor a. Surprisingly, the unoccupied estrogen receptor was able to repress thyroid receptor-stimulated transcriptional activity. As far as we know this has not been shown before, although it has been shown that under certain conditions an unoccupied estrogen receptor can bind DNA (Brown and Harp,1990). The two receptors irrespective of ligand binding may thus compete for the same DNA element,in our case the -172/-148 TRE. Second, proteinprotein interactionsmay explain the repression of unoccupied receptors. It could be that dimerization of estrogen and thyroid receptor occurs, to form a less active complex, similarly to what has been demonstrated for retinoic acid and thyroid hormone receptors (Glass et al., 1989). In a gel retardation assay using a mixture of extracts enriched in c-erbA/TRa protein and estrogen receptor protein, one extra retarded complex was seen, in addition to theseparate estrogen receptor-DNA and c-erbA/TRa-DNA specific retarded complexes. This suggests that when both receptors are present an additional interaction may take place with the -172/-148 fragment. In conclusion, the OT gene is a physiological target gene for thyroid hormone action. This action is mediated by at least two regions in the 5'-flanking sequences of the OTgene, one of which overlaps a functional ERE. Acknowledgments-We are grateful to Drs. R. hell and D. Richter for the gift of the cloned rat oxytocin gene, Dr. J. Battey for the human gene, Drs. S. Green and P. Charnbon for the estrogen receptor expression vector, and Drs. C. C. Thompson and R. M. Evans for the thyroid hormone receptor expression vector a. We thank Dr. W. M. Wiersinga for the measurement of plasma TSH, T, and T, levels. We thank Dr. H. Stunnenberg for the c-erbA/TRa nuclear extracts, Dr. M. G. Parker for the estrogen receptor nuclear extracts, and Dr. E. Schmidt for the TRE palindrome. REFERENCES Adan, R. A. H., Walther, N., Cox, J. J., hell, R., and Burhach, J. P. H. (1991) Biochem. Biophys. Res. Commun. 175,117-122 Almazan, G., Lefebvre, D. L., and Zingg, H. H. (1989) Deu. Brain Res. 45.69-75 Beato, M. (1989) Cell 5 6 , 335-344 Brent, G. A., Harney, J. W., Chen, Y., Warne, R. L., Moore, D. D., and Larsen, P. R. (1989) Mol. Endocrinol. 3 , 1996-2004 Brent, G. A,, Moore, D.D., and Larsen, P. R. (1991) Annu. Reu. Physiol. 5 3 , 17-35 Bradley, D. J., Young, S. W., 111, and Weinberger, C. (1989) Proc. Natl. Acad. Sci. U.S. A. 8 6 , 7250-7254 Brown, M.,and Harp, P. A. (1990) J . Bid. Chem. 2 6 5 , 11238-11243 Burbach, J. P. H., and Bin Liu (1989) Methods Enzymol. 1 6 8 , 385397 Burbach, J. P. H., Adan, R. A. H., Van Tol, H. H. M., Verbeeck, M. A. E., Axelson, J. F., Van Leeuwen, F. W., Beekman, J. M., and

Oxytocin Hormone Thyroid and AB, G. (1990) J. Neuroendocr. 2,633-639 Dalman, F. C., Koenig, R. J., Perdew, G. H., Massa, E., and Pratt, W. B. (1990) J. BWl. C h m . 2 6 5 , 1615-1618 Day, R. N., and Maurer, R. A. (1989) Mol. Endocrinol. 3 , 931-938 DeLeaunoit, Y., and Kiss, R. (1989) I n Vitro Cell Deu. Biol. 25,585591 De Wet J. R.,Wood,K.V., DeLuca, M., Helinski, D. R., and Subramani, S. (1987) Mol. Cell Biol. 7,725-737 Fawell, S. E., White, R., Hoare, S., Sydenham, M., Page, M., and Parker, M. G. (1990) Proc. Natl. Acad. Sci. U. S. A. 87,6882-6887 Fort, Ph., Mary, L., Piechaczyk, M., El Sabrouty, S., Dani, Ch., Jeanteur, Ph., and Blanchard, J. M. (1985) Nucleic Acids Res. 13, 1431-1442 Fuchs, A. R. (1986) in Oxytocin: Clinicaland Laboratory Studies (Amico, J. A., and Robinson, A.G., eds) pp. 207-256, Elsevier, Amsterdam Fukuda, T., Willingham, M. C., and Cheng, S. Y. (1988) Endocrinology 123,2646-2652 Glass, C. K., Holloway, J. M., Devary, 0. V., and Rosenfeld, M. G. (1988) Cell 5 4 , 313-323 Graupner, G., Wills, K. N., Tzukerman, M., Zhang, X., and Pfahl, M. (1989) Nature 340,653-656 Graupner, G., Zhang, X., Tzukerman, M., Wills, K., Hermann, T., and Pfahl, M. (1991) Mol. Endocrinol. 5 , 365-372 Hermann, T., Zhang, X., Tzukerman, M., Wills, K. N., Graupner, G., and Pfahl, M. (1991) Cell Regul. 2 , 565-574 Ivell, R., and Richter, D. (1984) Proc. Natl. Acad. Sci. U. S. A. 8 1 , 2006-2010 Kumar, V., and Chambon, P. (1988) Cell 5 5 , 145-156 Kumar, V., Green, S., Staub, A., and Chambon, P. (1986) EMBO J. 5,2231-2236 McBurney, M. W., Jones-Villeneuve, E. M. V., Edwards, M. K. S., and Anderson, P. J. (1982) Nature 2 9 9 , 165-167 Naar, A.M., Boutin, J., Lipkin, S. M., Yu, V. C., Holloway, J. M., Glass, C. K., and Rosenfeld, M. G. (1991) Cell 6 5 , 1267-1279 Piechaczyk, M., Blanchard, J. M., Marty, L., Dani, Ch., Panabieres, F., El Sabouty, S., Fort, Ph., and Jeanteur,Ph. (1984) Nucleic Acids Res. 12,6951-6963 Richard, S., and Zinh, H. H. (1990) J. Biol. Chem. 265,6098-6103

Expression Gene

3777

Sambrook, J., Fritsch, E. F., and Maniatus, T. (1989) Molecular Cloning: A Laboratory Manual 2nd Ed, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY Samson, W. K., Lumpkin, M. D., and McCann, S. M. (1986) Endocrinology 119,554-560 Saw J.. de Maeistris. L.. Stunnenbera. . . -. H... and Vennstrom.. B. (1990) EMBO J. 9,-887-896' Sausville. E.. Carnev. D.. and Battev. - , J. (1985) . . J. Biol. Chem. 260. 10236-10241 Seiler-Tuyns, A., Walker, P., Martinez, E., MBrillat, A.-M., Givel,F., and Wahli, W. (1986) Nucleic Acids Res. 14,8755-8770 Soloff, M. S. (1985) in Oxytocin: Clinicaland Laboratory Studies (Amico, J. A,, and Robins, A. G., eds) pp. 259-276, Elsevier, Amsterdam Talanti, S., and Attila, U.(1972) Life Sci. 11, 49-54 Thompson, C . C., and Evans, R. M. (1989) Proc. Natl. Acad. Sci. U. S. A. 86,3494-3498 Thompson, C. C., Weinberger, C., Lebo, R., and Evans, R. M. (1987) Science 237, 1610-1614 Umesono. K.. Murakami, K. K.. ThomDson. . C. C., and Evans, R. M. (1991) Ce11'65,1255-1266 Van de Heiinine. B. J. M.. Koekoek-van den Herik. I.. Rots. N. Y.. and Van-Wimersma Greidanus Tj. B. (1991) J. keuroendlocr. 3; 57-64 Van der Eb, A. J., and Graham, F. L. (1980) Methods Enzymol. 65, 826-839 Van Tol, H. H. M., and Burbach, J. P. H. (1987) Methods Enzymol. 168,398-413 Van Tol, H. H. M., Voorhuis, D. Th. A. M., and Burbach, J. P. H. (1987) Endocrinology 120, 71-76 Van Tol, H. H. M., Van den Buuse, M., De Jong, W., and Burbach, J. P. H. (1988a) Mol. Brain Res. 4 , 303-311 Van Tol, H. H. M., Bolwerk, E. L. M., Lui, B., and Burbach, J. P. H. (1988b) Endocrinology 122,945-951 Van Tol, H. H. M., Snijdewindt, F. G. M., Boer, G. J., and Burbach, J. P. H. (1990) Neurosci. Res. Commun. 7 , 25-33 Van Zonneveld, A.-J., Curriden, S. A., and Loskutoff, D. J. (1988) Proc. Natl. Acad. Sci. U. S. A. 8 5 , 5525-5529 Wilkinson, M. (1988) Nucleic Acids Res. 1 6 , 10933 Zingg, H. H., and Lefebvre, D. L. (1988) Mol. Brain Res. 4 , 1-6 -

I

I

'