Stimulation of luteinizing hormone beta gene promoter activity by the orphan nuclear receptor, steroidogenic factor-1.

The orphan nuclear receptor, steroidogenic factor-1 (SF-1), is expressed in the pituitary and in the gonadotrope precursor cell line, alphaT3-1, where it is believed to enhance expression of the common gonadotropin alpha-subunit gene through transactivation of the gonadotrope-specific element (GSE). Sequence analysis of the rat luteinizing hormone beta-subunit (LH beta) gene promoter revealed the presence of a consensus GSE at -127 to -119 (TGACCTTGT). We have demonstrated the ability of SF-1 to bind specifically to this putative GSE sequence by electrophoretic mobility shift assay, utilizing both alphaT3-1 nuclear extracts and in vitro translated SF-1. In addition, mutation of the putative LHbeta-GSE (TGAAATTGT) eliminated specific DNA binding. To examine the ability of SF-1 to enhance LHbeta promoter activity, CV-1 cells, which lack endogenous SF-1, were cotransfected with an SF-1-containing expression vector and an LHbeta-luciferase reporter construct. When cotransfected with -209/+5 of the LHbeta promoter, SF-1 increased luciferase activity by 56-fold. SF-1 responsiveness was markedly diminished with loss of the putative GSE region in deletion constructs and in the presence of a two base pair mutation, analogous to the mutation which eliminated DNA binding. Finally, the LHbeta-GSE was able to confer SF-1 responsiveness on a heterologous minimal growth hormone promoter, GH50 (57-fold). We conclude that SF-1 both binds to and transactivates the rat LHbeta promoter. These data suggest that SF-1 may participate in the expression of the LHbeta gene by the gonadotrope.

The orphan nuclear receptor, steroidogenic factor-1 (SF-1), is expressed in the pituitary and in the gonadotrope precursor cell line, ␣T3-1, where it is believed to enhance expression of the common gonadotropin ␣-subunit gene through transactivation of the gonadotropespecific element (GSE). Sequence analysis of the rat luteinizing hormone ␤-subunit (LH␤) gene promoter revealed the presence of a consensus GSE at ؊127 to ؊119 (TGACCTTGT). We have demonstrated the ability of SF-1 to bind specifically to this putative GSE sequence by electrophoretic mobility shift assay, utilizing both ␣T3-1 nuclear extracts and in vitro translated SF-1. In addition, mutation of the putative LH␤-GSE (TGAAATTGT) eliminated specific DNA binding. To examine the ability of SF-1 to enhance LH␤ promoter activity, CV-1 cells, which lack endogenous SF-1, were cotransfected with an SF-1-containing expression vector and an LH␤-luciferase reporter construct. When cotransfected with ؊209/؉5 of the LH␤ promoter, SF-1 increased luciferase activity by 56-fold. SF-1 responsiveness was markedly diminished with loss of the putative GSE region in deletion constructs and in the presence of a two base pair mutation, analogous to the mutation which eliminated DNA binding. Finally, the LH␤-GSE was able to confer SF-1 responsiveness on a heterologous minimal growth hormone promoter, GH50 (57fold). We conclude that SF-1 both binds to and transactivates the rat LH␤ promoter. These data suggest that SF-1 may participate in the expression of the LH␤ gene by the gonadotrope.
The pituitary gonadotropins, luteinizing hormone and follicle-stimulating hormone, are critical modulators of gamete maturation and gonadal steroidogenesis. These hormones are composed of a common ␣-subunit linked noncovalently to unique ␤-subunits which specify physiologic actions (1).
Several DNA regulatory elements have been defined for the ␣-subunit gene promoter. GnRH 1 -stimulated expression is believed to be mediated through a regulatory element located between positions Ϫ346 and Ϫ244 in the human ␣-subunit gene promoter, a region separate from those involved in basal and cAMP-stimulated expression (2). Activation of a cAMP response element (CRE) appears to be important for expression in both pituitary and placental cell types, while placentalspecific expression occurs through the activation of a trophoblast-specific element acting in concert with the CRE (3). Pituitary-specific expression of the ␣-subunit gene has been attributed to the presence of both a pituitary glycoprotein basal element and a gonadotrope-specific element (GSE) (4,5). The consensus GSE sequence (TGACCTTGT), defined in the common ␣-subunit by Mellon and colleagues, resembles a nuclear receptor binding half-site (4,6). Variations of this sequence, alternatively called the Ad4 response element, are also present in the promoter regions of multiple genes which play a role in steroidogenesis, sexual differentiation, and adult reproductive function (7). The GSE/Ad4 element has been shown to interact with the transcription factor, steroidogenic factor-1 (SF-1), in a number of genes, including the steroidogenic P450, the aromatase, and the Mü llerian inhibiting substance genes (8 -10). SF-1 is an orphan member of the nuclear hormone receptor superfamily. Best known for its selective expression in adrenal and gonadal cells, it has more recently been identified in the pituitary gland with localization to the gonadotrope (6,7).
In studies of the human ␣-subunit gene promoter, SF-1 has been shown to bind to the GSE region by electrophoretic gel mobility shift assay. Furthermore, reporter constructs which contain the ␣-subunit GSE site are expressed at higher levels in cell lines which contain endogenous SF-1 than in those cells which lack SF-1, consistent with a role for SF-1 in tissuespecific transcriptional activation of the ␣-subunit gene (4,6).
In contrast with the ␣-subunit, the cis-acting elements responsible for expression of either the LH␤or FSH␤-subunit mRNAs are poorly understood. Interestingly, transgenic mice null for the gene which encodes SF-1 not only express the ␣-subunit in low levels, but also fail to express the ␤-subunits, suggesting a functional role for SF-1 in LH␤ gene expression (7). As previous studies of the bovine LH␤ gene promoter have shown that the proximal 776 base pairs are sufficient to direct pituitary-specific expression in transgenic mice (11), we analyzed the corresponding region of the rat LH␤ promoter for the presence of an SF-1-binding site, or GSE.
The rat LH␤ gene promoter contains a consensus GSE at position Ϫ127 to Ϫ119 relative to the transcriptional start site (Fig. 1). Interestingly, this sequence is highly conserved across species among the LH␤ genes, suggesting physiologic significance (11,14). Inasmuch as the consensus GSE sequence is present in the LH␤ gene promoter, we wished to determine whether this putative GSE region has functional significance. We, therefore, investigated the ability of SF-1 to bind to and transactivate the rat LH␤ gene promoter.

MATERIALS AND METHODS
Oligonucleotides Used in Electrophoretic Mobility Shift Assay (EMSA)-The nucleotide sequence of the rat LH␤ gene promoter was based on Fig. 3 of Jameson et al. (12) with position Ϫ1 assigned to the nucleotide immediately 5Ј to the transcriptional start site. The LHSF oligonucleotide used in EMSA corresponds to bases Ϫ134 to Ϫ113 of the rat LH␤ gene (sense strand: 5Ј-TCCTTTCTGACCTTGTCTGTCT-3Ј). The LHSFM oligonucleotide sequence is identical to LHSF except for the conversion of the CC nucleotide pair at positions Ϫ124 and Ϫ123 to an AA pair (sense strand: 5Ј-TCCTTTCTGAAATTGTCTGTCT-3Ј). The Pit-1 oligonucleotide used in competition studies corresponds to Ϫ137 to Ϫ65 of the rat growth hormone promoter and contains two Pit-1/growth hormone factor-1 binding sites (sense strand: 5Ј-GGGAGGAGCTTCTA-AATTATCCATCAGCACAAGCTGTCAGTGGCTCCAGCCATGAATAA-ATGTATAGGGAAA-3Ј) (16). Except for the Pit-1 oligonucleotide, all oligonucleotides used for EMSA contained 5Ј-BamHI and 3Ј-BglII restriction sites in addition to the sequences listed above.
Sense and antisense oligonucleotides were annealed and end-labeled with [␥-32 P]ATP by T4 polynucleotide kinase and purified over a NICK column (Pharmacia Biotech Inc.).
Nuclear Extract and in Vitro Translated Proteins-The method of Andrews and Faller (17) was used to prepare crude nuclear extracts from a mouse gonadotrope-derived cell line (␣T3-1), a monkey kidney fibroblast cell line (CV-1), and a rat somatolactotrope cell line (GH 3 ). In vitro translated SF-1 protein was generated from a plasmid containing 2.1 kilobase pairs of the mouse SF-1 cDNA using the TNT coupled reticulocyte lysate system (Promega, Madison, WI) (18). The resultant product was determined to be of appropriate size by comparison with [ 35 S]methionine-labeled protein by sodium dodecyl sulfate-polyacrylamide gel electrophoresis.
Electrophoretic Mobility Shift Assays-Nuclear extract (5 g) or in vitro translated protein (1, 3, or 5 l) was incubated with 50,000 cpm of oligonucleotide probe in DNA-binding buffer (20 mM HEPES (pH 7.9), 60 mM KCl, 5 mM MgCl 2 , 10 mM phenylmethylsulfonyl fluoride, 10 mM dithiothreitol, 1 mg/ml bovine serum albumin, and 5% (v/v) glycerol) for 30 min on ice. For competition studies, excess unlabeled oligonucleotide was added 5 min prior to the addition of probe. Where indicated, antiserum (1 l) was added 30 min following the addition of probe, and incubation was continued for 2 h. Protein-DNA complexes were resolved on a 5% nondenaturing polyacrylamide gel in 0.5 ϫ Tris borate-EDTA buffer and subjected to autoradiography.
Plasmids Used in Transfection Studies-The largest LH␤ reporter construct used for these studies contained 794 base pairs of the 5Јflanking sequence of the rat LH␤ gene and the first 5 base pairs of the 5Ј-untranslated region fused to a luciferase reporter gene, pXP2 (19). Deletions in this construct were created by subcloning polymerase chain reaction products containing the LH␤ promoter sequences into the pXP2 vector using BamHI/HindIII sites which were introduced by the primers. The -209LH␤-MUT plasmid was created by introducing a two base pair mutation into the -209LH␤ construct using the transformer site-directed mutagenesis kit (Clontech Laboratories, Inc., Palo Alto, California). The selection primer was located in the pXP2 polylinker and the 3Ј-end of the LH␤ flanking sequence and converted a unique HindIII restriction site to a unique MluI site (sense strand: 5Ј-GGTAGGGAAGGTATCACGCGTGTCGACCCGGGTACC-3Ј).
The mutagenic primer spanned region Ϫ147 to Ϫ104 of the LH␤ promoter and eliminated a TthIII1 restriction site in addition to introducing the desired mutation (sense strand: 5Ј-GCTGGTCCCTGGCTTTTCT-GAAATTGTCTGTCTCGCCCCCAAAG-3Ј). To create GSE2-GH50, an oligonucleotide was designed which contained two copies of the putative GSE region as a tandem repeat flanked by BamHI/BglII restriction sites (sense strand: 5Ј-GATCCTTTTCTGACCTTGTCTGTCTCGC-CTCTGACCTTGTCTGTA-3Ј). This oligonucleotide was inserted upstream of the minimal growth hormone promoter, GH50, in the pXP1 luciferase reporter plasmid (19,20). All reporter constructs were confirmed by dideoxysequencing.
The SF-1 expression vector contained 2.1 kilobase pairs of the mouse SF-1 cDNA driven by cytomegalovirus promoter sequences (18). The Pit-1 expression vector was created by placing 915 base pairs of the rat Pit-1/growth hormone factor-1 cDNA sequence from pBluescript SK(Ϫ) (Stratagene, La Jolla, CA) into the pcDNAI vector (Invitrogen, San Diego, CA) using HindIII/NotI restriction enzyme sites (21).
Cell Culture and Assays-Monkey kidney fibroblast (CV-1) cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum. Cells growing in 3.5-cm tissue culture wells (Flow Laboratories, McLean, VA) were transfected with expression (0.1 g/ well) and reporter (1.65 g/well) plasmids using the calcium phosphate precipitation method (22). Control wells received the appropriate "empty" expression vector (0.1 g/well). Cotransfection with an RSV-␤galactosidase plasmid (1 g/well) allowed correction for differences in transfection efficiency between wells. The cells were harvested 48 h following transfection and the cell extracts analyzed for both luciferase (23) and ␤-galactosidase (24) activities. Luciferase activity was first normalized to the level of ␤-galactosidase activity. Results were then calculated as fold change relative to expression in the presence of the control empty expression vector. Data are shown as the mean Ϯ S.E. and represent a minimum of three independent experiments with each point run in triplicate in each experiment.
Sources of SF-1 and Pit-1 Antibodies and Plasmids-The rabbit SF-1 antiserum as well as the vectors containing the SF-1 cDNA were kindly provided by Dr. K. L. Parker (Duke University). The SF-1 antibody was generated against a glutathione S-transferase-SF-1 fusion protein (25). The Pit-1 antiserum, directed against amino acids 136 -150 of rat Pit-1/growth hormone factor-1, was provided by C. Bancroft (Mt. Sinai School of Medicine) (26). The Pit-1 cDNA in pBluescript SK(Ϫ) was provided by Dr. L. E. Theill (University of California, San Diego).

␣T3-1 Nuclear Extract Binds to the Putative GSE Region of the LH␤ Gene Promoter-
The ␣T3-1 cell line has previously been shown to express both SF-1 mRNA and protein (6, 7). Nuclear extracts from this cell line, presumed to contain SF-1, have been shown to bind to the ␣-subunit promoter GSE (4, 6). We therefore utilized EMSA to investigate whether these extracts were able to interact with the region of the rat LH␤ gene promoter that contains the putative GSE sequence (oligonucleotide LHSF). As shown in Fig. 2, the interaction of ␣T3-1 nuclear extracts with 32 P-labeled LHSF produced a specific protein-DNA complex as demonstrated by successful competition with unlabeled LHSF (lanes 1 and 2).
In order to confirm that the complex identified in Fig. 2 contained SF-1, we investigated the effect of a SF-1-specific antibody on the formation of the ␣T3-1 nuclear extract-LHSF  (4) is aligned with the putative GSE sites of the LH␤ gene in rat (12), cow (13), pig (14), and human (15). Also shown are regions of homology with the SF-1-binding sites in the 21-hydroxylase (8), aromatase (9), and Mullerian inhibiting substance (10) gene promoters. Variant nucleotides are underlined.

SF-1 Regulation of LH␤ Gene Expression
complex. This antibody has previously been shown to block the ability of SF-1 to bind to the promoter element of a number of genes, including the glycoprotein hormone ␣-subunit, aromatase, and 21-hydroxylase genes (6,9,25). Treatment with this SF-1-specific antiserum substantially decreased the intensity of the protein-DNA complex while the addition of an anti-Pit-1 antiserum, used as a negative control, had no effect (Fig.  2, lanes 3 and 4). This result confirms that the GSE of the LH␤ gene promoter is bound by SF-1, or an immunologically related protein, present in ␣T3-1 nuclear extracts.
Parallel EMSA was performed using the oligonucleotide LHSF as a probe in the presence of nuclear extracts from cell lines that do not contain SF-1. No specific protein-DNA interactions were detected with the use of nuclear extracts from either monkey kidney fibroblast cells (CV-1) or rat somatolactotrope cells (GH 3 ) (data not shown).
In Vitro Translated SF-1 Binds to the LH␤ Gene Promoter-Further confirmation that SF-1 binds to the LH␤-GSE promoter region was obtained by the use of in vitro translated SF-1 in EMSA. A binding reaction containing the labeled LHSF oligonucleotide and in vitro translated SF-1 resulted in a protein-DNA complex mobility similar to that obtained with the ␣T3-1 nuclear extracts (Fig. 2, lanes 5-10). Formation of this complex diminished in the presence of either excess unlabeled LHSF oligonucleotide or blocking antiserum directed against SF-1, but was unaffected by the Pit-1 antibody. Taken as a whole, these data clearly demonstrate that the putative GSE region of the LH␤ gene is recognized by SF-1, as either an endogenous (␣T3-1 nuclear extract) or an in vitro translated product.
Mutation of the Putative LH␤-GSE Site Eliminates Bind-ing-In order to localize further the SF-1 recognition site, a 2-base pair mutation was introduced in the wild-type LH␤ oligonucleotide sequence (LHSF) to form LHSFM. The choice of this mutation was based on the loss of DNA binding which resulted from analogous mutations in the Mü llerian inhibiting substance and glycoprotein ␣-subunit promoters (6, 10). EMSA was performed using ␣T3-1 nuclear extracts and either the wild-type LHSF (Fig. 3, lanes 1-4) or the mutant LHSFM (Fig.  3, lanes 5-8) as a probe. The intensity of the complex obtained with the LHSF probe was blunted by unlabeled wild-type LHSF, but not by the mutated sequence or by an unrelated oligonucleotide containing two binding sites for the pituitary transcription factor, Pit-1 (Fig. 3, lanes 2-4). As seen in lanes 5-8, the ␣T3-1 nuclear extract was not able to bind to LHSFM when used as a probe. These results establish that an intact LH␤-GSE sequence is required for binding by ␣T3-1 nuclear extract. SF-1 Specifically Increases LH␤ Promoter Activity-We next sought to determine the functional significance of the interaction between SF-1 and the LH␤ gene promoter sequences. In initial investigations utilizing the gonadotrope-derived ␣T3-1 cell line, LH␤ promoter-driven luciferase activity exceeded luciferase activity in the absence of cell extract (background activity) by less than 2-fold. At this level of expression, we were unable to evaluate reliably whether the presence or absence of the putative GSE sequence altered LH␤ promoter activity in response to the endogenous SF-1 present in this cell line. Furthermore, attempts to increase LH␤ gene expression through cotransfection with an SF-1 expression vector were unsuccessful.
These studies were therefore performed in the monkey kidney fibroblast cell line, CV-1, a cell line which has previously been shown to support SF-1-induced transactivation of the bovine P-450 CYP11B promoter. By Northern blot analysis, this cell line lacks the mRNA which encodes the SF-1 homolog, Ad4BP (27). As stated previously, we have also demonstrated that CV-1 nuclear extract fails to bind the LH␤-GSE region by  5-11). Competition with 500-fold molar excess of unlabeled LHSF is shown for both the nuclear extract and in vitro translated SF-1 (lanes 2 and 8). Incubation with antiserum specific to SF-1 (lanes 3 and 9) or Pit-1 (lanes 4 and 10) was also performed using both protein preparations. Note that 3 l of in vitro translated SF-1 were used in the cold competition and antibody studies (lanes 8 -10) and therefore band intensity should be compared against lane 6. Lane 11 contains a probe and the unprogrammed reticulocyte lysate used for in vitro translation. The arrowhead indicates the specific binding complex. A nonspecific band, indicated by the asterisk, is present in unprogrammed reticulocyte lysate.

SF-1 Regulation of LH␤ Gene Expression
EMSA, consistent with the absence of endogenous SF-1 (data not shown). Utilizing the CV-1 cell line, basal LH␤ gene promoter activity exceeded expression of the promoterless reporter plasmid (pXP2) by an average of 15-fold.
In Fig. 4A, CV-1 cells were cotransfected with region Ϫ209 to ϩ5 of the LH␤ gene promoter and cytomegalovirus-driven expression vectors containing either the SF-1 or Pit-1 cDNA. The presence of SF-1 markedly increased LH␤ promoter activity (56 Ϯ 5-fold). In contrast, the pituitary transcription factor Pit-1 did not alter luciferase levels, indicating the specificity of the SF-1 response.
Of importance, we have recently confirmed the ability of SF-1 to increase LH␤ promoter activity in the rat pituitaryderived somatolactotrope cell line, GH 3 . Utilizing transiently transfected GH 3 cells and conditions similar to those in CV-1 cells, SF-1 increased LH␤ promoter activity in the Ϫ209/ϩ5 construct by 15 Ϯ 1.5-fold.

SF-1 Stimulation of Rat LH␤ Gene Promoter Activity Is Dependent on the Presence of an Intact GSE-
In order to delineate the region in the LH␤ gene promoter responsible for providing SF-1 responsiveness, reporter constructs were generated which incorporated various deletions in the rat LH␤ gene promoter. Of note, no systematic changes in basal expression (i.e. in the absence of SF-1) were observed in these deletion constructs.
The evaluation of sequential 5Ј-deletion constructs revealed persistent SF-1 stimulation of LH␤ promoter activity with deletion to position Ϫ134, followed by an abrupt loss of the SF-1 response with deletion to position Ϫ82 (Fig. 4B). Based on these data, loss of LH␤ promoter sequences across the putative GSE region (positions Ϫ127 to Ϫ119) correlates with the loss of SF-1-stimulated promoter activity.
Further definition of the SF-1-responsive cis-acting element was obtained by the introduction of a two base pair mutation into the putative GSE site of the Ϫ209LH␤ luciferase reporter construct to form Ϫ209LH␤-MUT. This small change, analogous to the mutation which eliminated DNA-binding by nuclear extract (Fig. 3), substantially decreased the ability of SF-1 to increase promoter activity (6 Ϯ 1-fold versus 56 Ϯ 5-fold) (Fig. 4B). Thus, transactivation of the LH␤ promoter by SF-1 appears to be critically dependent on the presence of an intact GSE sequence.
LH␤ Promoter Sequences Confer SF-1 Responsiveness to a Heterologous Promoter-Having demonstrated that the putative GSE site is necessary for SF-1 responsiveness in the context of the LH␤ gene promoter, we next asked whether this sequence was sufficient to confer SF-1 responsiveness to a heterologous promoter. Two copies of the LH␤-GSE sequence were inserted upstream of the growth hormone minimal promoter, GH50. As seen in Fig. 4C, these sequences conferred a marked SF-1 response to this normally nonresponsive promoter (57 Ϯ 10-fold versus 1.3 Ϯ 0.2-fold).
Other investigators have shown previously that the human ␣-subunit promoter GSE (identical to the putative rat LH␤-GSE, see Fig. 1) increases thymidine kinase minimal promoter activity in SF-1 containing cell lines, but not in cell lines which lack SF-1 (6). However, interpretation of this study was limited by the possibility that additional cell-specific factors were contributing to the observed differences in transcriptional activity. As our results were obtained in a single cell line, stimulation of promoter activity in the presence of the GSE can be attributed solely to SF-1-induced effects.

DISCUSSION
Our results clearly demonstrate that SF-1 binds specifically to the putative GSE region of the rat LH␤ gene promoter and that, through this interaction, SF-1 substantially increases LH␤ gene promoter activity. Furthermore, we have shown that the introduction of a two base pair mutation within the LH␤-GSE sequence markedly blunts the ability of SF-1 to either bind to or transactivate this promoter.
It is of interest to note that the SF-1 response was not fully lost in the mutated construct, suggesting a role for additional nucleotides within the GSE in effecting SF-1-induced transactivation. While parallel functional assays have not been performed, it has been clearly demonstrated that DNA binding by SF-1 is severely blunted by mutation of nucleotide pairs at other positions within the GSE/Ad4 element (8). Alternatively, additional SF-1 binding sites may contribute to the regulation of LH␤ promoter activity. Sequence analysis of the rat LH␤ promoter identifies a number of regions which resemble the consensus GSE. It will be of interest to investigate the possible FIG. 4. An intact putative LH␤-GSE region confers SF-1 responsiveness to both the LH␤ promoter and a heterologous minimal promoter. CV-1 cells were transiently transfected with luciferase reporter constructs which contained various regions of the rat LH␤ gene promoter. Cells were cotransfected with plasmids encoding either SF-1 or Pit-1 and with an RSV-␤-galactosidase expression vector. Luciferase activity was normalized to ␤-galactosidase activity. Promoter activity was then calculated as fold change over expression in the presence of the appropriate control expression vector. Results are shown as the mean Ϯ S.E. of at least nine samples in three independent experiments. A, comparison of LH␤ promoter activity in response to SF-1 versus Pit-1. B, SF-1 stimulation of LH␤ promoter activity with loss of the intact GSE sequence by sequential 5Ј-deletion or mutagenesis. C, SF-1 responsiveness of the growth hormone minimal promoter (GH50) (20) or GH50 preceded by two copies of the putative LH␤-GSE region (GSE2-GH50). functional significance of these regions in future studies. SF-1 has been shown to act at multiple levels of the reproductive axis, including the hypothalamus, pituitary, and gonad (7). Within the pituitary, Mellon and colleagues have demonstrated SF-1-stimulated expression of the glycoprotein ␣-subunit gene (6). In conjunction with their results, our data regarding the regulation of LH␤ gene expression suggests that SF-1 may play a critical role in the coordinated expression of both subunit genes required for luteinizing hormone biosynthesis in the gonadotrope.
In the functional studies reported here, we utilized a heterologous system in which both an SF-1 expression vector and a reporter construct containing LH␤ promoter sequences were transiently transfected into a fibroblast cell line, CV-1. In preliminary studies, we have also observed SF-1-mediated increases in LH␤ promoter activity using a pituitary-derived cell line, the rat somatolactotrope cell line, GH 3 . Although these studies would ideally have been performed in a gonadotropederived cell line, currently available cell lines fail to express either endogenous or exogenous gonadotropin ␤-subunits at appreciable levels (28).
As gonadotrope-derived ␣T3-1 cells are known to contain SF-1 (7), the lack of LH␤ gene expression may seem inconsistent with a role for SF-1 in activation of the LH␤ promoter. However, a number of potential explanations are possible. For example, this cell line may be arrested at a stage of development in which both ␣-subunit (E12.5) and pituitary SF-1 (E13.5) expression have been established, but LH␤-subunit expression is absent (E16.5) (4, 7, 28). The ␣T3-1 cells may therefore be appropriately expressing an inhibitory factor(s) which is responsible for suppression of LH␤ promoter activity at this developmental stage. Alternatively, ␣T3-1 cells may lack the SF-1 ligand (currently postulated but as yet unidentified) or an additional cofactor required for LH␤ promoter activation. This explanation seems less likely, however, in view of the magnitude of the observed response in CV-1 cells, a cell line that does not express endogenous SF-1 nor any of the identified SF-1-regulated genes and would therefore be predicted to be less likely to contain the necessary cofactors.
The results presented here clearly demonstrate that the presence of the GSE element is sufficient to direct SF-1 responsiveness in the context of both the LH␤ gene promoter and the heterologous minimal growth hormone promoter, GH50. The question remains, however, as to whether SF-1 is required for LH␤ gene expression. Interestingly, in transgenic mice null for the Ftz-F1 gene which encodes SF-1, GnRH replacement was able to restore gonadotropin expression in four out of five animals (7,29). These results suggest that cells from the gonadotrope lineage are present in these animals and are capable of expressing the LH␤ gene despite the lack of SF-1. However, it is important to note that quantitatively normal levels of LH in the absence of SF-1 have yet to be shown. Furthermore, as is true for all gene "knockout" paradigms, this transgenic model system does not exclude possible redundancy in the pool of potential transactivating factors for this critical reproductive gene. Thus, while SF-1 may not be absolutely required for LH␤ gene expression, it may in fact be required for normal levels of expression in the intact animal. The magnitude of the SF-1directed increase in promoter activity observed in both the CV-1 and GH 3 cell lines strongly implies physiological significance.
Our results do not directly address whether SF-1, a member of the nuclear hormone receptor superfamily, functions as a monomer or whether it has the ability, and/or requirement, to undergo dimerization (18). Members of this family are best known for binding to pairs of recognition half-sites arranged as tandem or inverted repeats. More recently, an alternative mechanism for DNA interaction, monomer binding to a single 5Ј-extended half-site, has been described for both SF-1 and another orphan nuclear receptor, NGFI-B (30). The results presented here are at least consistent with the ability of SF-1 to bind to a single GSE element in the LH␤ gene promoter. By sequence analysis, the rat LH␤-GSE at positions Ϫ127 to Ϫ119 is present as a single site. Furthermore, as shown by EMSA, SF-1 is able to bind to an oligonucleotide probe in which the LH␤-GSE site is flanked by fewer than 8 additional base pairs on each side. SF-1 has also been shown to bind to similarly short GSE-containing promoter sequences from a variety of other genes, including the glycoprotein hormone ␣-subunit, aromatase, and Mullerian inhibiting substance genes (6,9,10).
The lack of a second DNA-response element in the region of the LH␤-GSE does not exclude potential SF-1 dimerization with a non-DNA-binding partner such has been shown to occur between the orphan nuclear receptor NurrI/NGFI-B and the 9-cis-retinoic acid receptor (31). Intriguingly, mutation of the DAX-1 gene in humans results in adrenal and gonadal hypoplasia similar to that observed in SF-1 deficient mice, implicating DAX-1 as a potential SF-1 dimerization partner (7,32).
The current studies clearly define a role for SF-1 in the regulation of basal expression of the LH␤ gene. Within the pituitary gland, SF-1 expression is restricted to the gonadotrope subpopulation and may therefore be an important modulator of cell-specific expression (7). SF-1 cannot, however, be the sole determinant of gonadotrope-specific expression as it is also expressed in non-LH-producing tissues such as the ventromedial hypothalamus, gonad, and adrenal gland (7,29). It will also be of interest to determine whether SF-1-induced increases in LH␤ promoter activity interact with GnRH-stimulated responses. The potential role of SF-1 in both tissuespecific and hormonally-mediated expression of the LH␤ gene awaits further exploration.