p300 Regulates the Synergy of Steroidogenic Factor-1 and Early Growth Response-1 in Activating Luteinizing Hormone-β Subunit Gene*

Tight regulation of luteinizing hormone-β subunit (LHβ) expression is critical for differentiation and maturation of mammalian sexual organs and reproductive function. Two transcription factors, steroidogenic factor-1 (SF-1) and early growth response-1 (Egr-1), play a central role in activating LHβ promoter, and the synergy between these two factors is essential in mediating gonadotropin-releasing hormone stimulation of LHβ promoter. Here we demonstrate that the transcriptional co-activator p300 regulates this synergy. Overexpression of p300 results in strong stimulation of LHβ promoter but only in the presence of both SF-1 and Egr-1, and not in the presence of other Egr proteins. Mutation of the binding sites for either SF-1 or Egr-1 completely abolishes the synergy between these two factors, as well as the influence of p300. Importantly, LHβ promoter is precipitated using p300 antibodies in a chromatin immunoprecipitation assay with LβT2 gonadotropes, and this effect is enhanced by gonadotropin-releasing hormone. The influence of p300 on LHβ promoter is potentiated by steroid receptor co-activator, as well as by E1A proteins, and attenuated by Smad nuclear interacting protein 1. Taken together, these results suggest that p300 is recruited to LHβ promoter where it coordinates the functional synergy between SF-1 and Egr-1.

Reproductive development and function in mammals depends on exquisite regulation of luteinizing hormone (LH) 1 release by anterior pituitary gonadotropes. Impaired fertility has been associated with underproduction, as well as overproduction, of LH (reviewed in Ref. 1). The secretion of LH is stimulated by pulsed release of the hypothalamic decapeptide gonadotropin-releasing hormone (GnRH), which also stimulates the synthesis and release of follicle-stimulating hormone. LH and follicle-stimulating hormone are heterodimeric glycoproteins that consist of a common ␣-subunit and a unique ␤-subunit (2). Binding of GnRH to its membrane receptor (3) induces a cascade of phosphorylation signals that lead to the activation of LH␤ promoter (reviewed in Ref. 4).
Several regulatory elements have been identified within 500 bp upstream of the transcription start site in LH␤ gene. However, a sequence located within 150 bp upstream of the transcription start site can drive the expression of a reporter gene in the gonadotrope cell line L␤T2 in response to GnRH treatment (5). This small region contains two tandem SF-1/Egr-1 elements separated by a Pitx1 binding element. Indeed, these factors have been clearly shown to regulate LH␤ gene transcription (5)(6)(7)(8)(9)(10). In vivo, both SF-1 and Egr-1 play an essential role in the regulation of LH␤ gene, as SF-1 knockout mice (11)(12)(13) or Egr-1 knockout mice (8,14) exhibit a dramatic LH␤ deficiency, resulting in abnormal development and infertility (15,16). These two factors are not only obligatory for activation of LH␤ promoter, but they also synergize with each other to activate LH␤ promoter in a GnRH-dependent manner (5,8,17). This synergistic interaction between SF-1 and Egr-1 has been observed using LH␤ promoter from different species, supporting the notion that this conserved mechanism is relevant for precise regulation of LH␤ gene expression (18,19). Although paramount to regulation of LH␤ gene, the mechanism of synergy between SF-1 and Egr-1 is not known (9,17). Notably, this synergy requires the binding of both SF-1 and Egr-1 to their respective LH␤ promoter elements (5). Direct interaction between SF-1 and Egr-1 has been detected in vitro using a pulldown experiment, but not using an electromobility shift assay (17).
The co-activator p300/CBP is known to regulate gene expression by bridging promoter-bound transcription factors with the basal transcriptional machinery (for review, see Ref. 20). Because SF-1 and Egr-1 can individually interact with the coactivator p300/CBP (21,22) we postulated that SF-1 and Egr-1, bound to their cognate response elements, form a docking platform for p300, which, in turn, would enhance the recruitment of additional co-factors leading to enhanced LH␤ transcription. We therefore tested the hypothesis that p300 influences the transcriptional activation of LH␤ promoter by SF-1 and Egr-1.
Cell Culture and Transfections-CV-1 and 293 cells were maintained in Dulbecco's modified Eagle's medium, with 10% fetal bovine serum and antibiotics at 37°C at 10 and 5% CO 2 , respectively. JEG3 cells were grown in minimum essential medium with 10% fetal bovine serum and antibiotics at 5% CO 2 . L␤T2 cells, obtained from Pamela Mellon (Uni-versity of California, San Diego, CA), were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and were cultured in a 5% CO 2 environment (32,33). GnRH (100 nM) was added to plated cells 24, 6, and 1 h before harvest. CV1, JEG3, and 293 cells were transfected using the modified calcium phosphate method described previously (34,35). L␤T2 cells were transfected using FuGENE 6 (Roche Applied Science), following the manufacturer's protocol. Empty expression plasmids were used to balance the total amount of transfected DNA, where necessary. Standard luciferase assays were performed 48 h after transfection as described previously (28), using a Lumistar 96-well plate reader (BMG, Durham, NC). Because we found that p300 influenced the expression of the control vector ␤-galactosidase, we normalized the results to transfection efficiency and cell viability by one of the following methods: (1) determination of total protein in transfected cells, and (2) Renilla luciferase activity, co-transfected using the pRL-null vector (Promega, Madison, WI) and determined by a dual-luciferase assay kit (Promega). Results were expressed as relative luciferase units (RLU). All experiments were performed in duplicate and repeated at least three times.
Chromatin Immunoprecipitation and PCR-Approximately 1 ϫ 10 7 L␤T2 cells were grown on 10-cm dishes and cross-linked with 1% formaldehyde at room temperature for 15 min. The reaction was stopped by the addition of glycine to a final concentration of 0.125 M. Cells were scraped in phosphate-buffered saline, collected by centrifugation, and incubated for 10 min on ice in 200 l of lysis buffer (50 mM Tris, pH 8.1, 10 mM EDTA, 1% SDS plus protease inhibitor mixture; Sigma). Cell lysates were then sonicated on ice to a mean length of 500 -1,000 bp and then centrifuged at 14,000 rpm. The chromatin solution was diluted in a buffer containing 0.01% SDS, 1.1% Triton X-100, 1.2 nM EDTA, 16.7 mM Tris, pH 8.1, 167 mM NaCl and then pre-cleared with 80 l of salmon sperm DNA/protein A-agarose slurry for 30 min at 4°C with agitation. Input samples were obtained at 0.1% (v/v), followed by immunoprecipitation with specific antibodies at 4°C overnight. The antibodies included anti-p300 (sc584), anti-Egr-1 (sc110), and anti-PPAR␥ (sc1984), all from Santa Cruz Biotechnology, Inc., and anti-SF-1 (06 -431) from Upstate Biotechnology. Immune complexes were collected with 60 l of salmon sperm DNA/protein A-agarose slurry and sequentially washed for 10 min each in low salt wash buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris, pH 8.1, 150 mM NaCl), high salt buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris, pH 8.1, 500 mM NaCl), LiCl wash buffer (0.25 M LiCl, 15 Nonidet P-40, 1% deoxycholate, 1 mM EDTA, 10 mM Tris, pH 8.1), and two washes with Tris/EDTA buffer. Precipitates were then extracted two times with 1% SDS, 0.1 M NaCO 3 . Eluates were pooled, and cross-linking was reversed by incubation at 65°C overnight. Samples were then digested with proteinase K (Roche Diagnostics) for 2 h at 45°C and purified by phenol-chloroform extraction followed by ethanol precipitation. The DNA was dissolved in 50 l of Tris/EDTA buffer, and 5 l was used for PCR, described below.
For standard PCR, immunoprecipitated DNA was used as a template in 25-l reactions using specific LH␤ promoter primers, which include forward primer 5Ј-TTCAGCGAGCAGCCTGCAGTGGCCTCCCCT and reverse primer 5Ј-CCACTAAGTAGTGGCTACAGGCTTGGGTAA. Band density was quantified using LabWorks 4.0 software (UVP, Upland, CA). Transcript level was determined by quantitative PCR using a 5700 Sequence Detector System (Applied Biosystems). Each 50 l of PCR included 3 l of reverse transcription products, forward primers (5Ј-GGTTACCCAAGCCTGTAGCCA) and reverse primers (5Ј-TGGCTTTATACCTGCGGGTT), and 25 l of SYBER Green Master mix (Applied Biosystems). Transcript level was determined by using Gene-Amp 5700 SDS software (Applied Biosystems).
Statistics-Each set of results was determined in duplicate and repeated at least three times. Because of the nature of transfection experiments, representative results are shown. For chromatin immunoprecipitation (ChIP) experiments, data presented are mean Ϯ S.D. from four independent experiments. Relevant paradigms were compared using analysis of variance (ANOVA), with post hoc Bonferroni t test (Primers of Biostatistics, v. 3.0). A value of p Ͻ 0.05 was determined significant.

P300 Strongly Stimulates the Activity of LH␤ Promoter-To
test the hypothesis that p300 influences the transcriptional activation of LH␤ promoter by SF-1 and Egr-1 we initially determined the effect of p300 on the activity of LH␤ promoter. For this purpose, we transiently transfected CV1 cells with expression plasmids for SF-1, Egr-1, and p300, along with an LH␤ reporter construct that spans nucleotides Ϫ156 to ϩ7 of rat LH␤ promoter upstream of luciferase. As shown in Fig. 1, in the absence of SF-1 and Egr-1 (lane 2) we observed minimal and non-significant activation of LH␤ promoter by p300. The influence of p300 on LH␤ promoter was pronounced in the presence of either SF-1 or Egr-1 (Fig. 1, lanes 5 and 8). Coexpression of SF-1 and Egr-1, along with p300, resulted in dramatic up-regulation of LH␤ promoter (Fig. 1, compare lane 11 with lanes 5 and 8), suggesting that p300 potentiates the synergy between SF-1 and Egr-1 in regulation of LH␤ promoter. As a control we tested the effect of steroid receptor co-activator-1 (SRC-1), a co-activator known to interact with and enhance the transcriptional activity of SF-1 (28,36). As expected, SRC-1 potentiated the interaction of SF-1 and Egr-1 on LH␤ promoter (Fig. 1, lane 12), but to a lesser extent than p300. To ensure that these results are not cell type-specific we repeated these experiments in JEG3 cells, as well as 293 cells, and recapitulated in these cell lines the strong p300-dependent potentiation of LH␤ promoter transactivation by SF-1/Egr-1 synergy (not shown). To ensure that the dramatic p300-dependent increase in LH␤ reporter activity is not because of significant alteration of SF-1 and Egr-1 expression we examined the level of these factors in transfected cells, using Western analysis. We found that the level of transfected SF-1, but not Egr-1, is weakly (3-5-fold) increased in the presence of overexpressed p300 (not shown). Moreover, transfection of 10-fold more SF-1 expression plasmid alone resulted in only a minor increase of the LH␤ promoter activity compared with the combined effect of SF-1/Egr-1 with p300, effectively ruling out altered SF-1/ Egr-1 expression level as an explanation for the striking effect of p300.
To obtain further support for the influence of p300 on SF-1/ Egr-1 synergy in the context of LH␤ promoter we used LH␤ promoter reporter plasmids that harbor mutated binding elements for Egr-1, SF-1, or both (5). As shown in Fig. 2, mutations in either SF-1 or Egr-1 elements resulted in a dramatic reduction in SF-1/Egr-1 synergy and markedly attenuated the influence of p300 on LH␤ promoter. These results demonstrate that p300 acts through interaction with DNA-bound SF-1 and Egr-1 in the context of LH␤ promoter. Whereas SF-1 and Egr-1 may directly interact (9,17), our results suggest that this direct interaction is not sufficient for promoter potentiation by p300, which requires tethering of SF-1 and Egr-1 to their promoter elements.
The murine gonadotrope cell line L␤T2 expresses LH␣ and LH␤ (32,37), as well as SF-1 and Egr-1, and up-regulates LH secretion by GnRH stimulation (5,32,33,38). We therefore sought to confirm the influence of p300 on LH␤ in these cells. We transiently transfected L␤T2 cells with p300, along with our LH␤-promoter reporter gene. As shown in Fig. 3, we found that overexpression of p300 in L␤T2 cells led to a significant up-regulation of basal LH␤ reporter. This effect was further potentiated by exposing L␤T2 cells to GnRH, thus providing further support to the influence of p300 on LH␤ activation by SF-1 and Egr-1.
LH␤ Promoter Occupancy in L␤T2 Cells-Having shown that p300 acts as a co-activator for SF-1/Egr-1 in transfected cells, we utilized ChIP to determine whether p300 is physically recruited to LH␤ promoter. We performed these experiments in L␤T2 cells, which produce LH␤ and in which p300 overexpression was shown to significantly enhance a LH␤ reporter. L␤T2 cells were treated with 1% formaldehyde to cross-link proteins to chromatin within the live cells, and complexes were captured using specific antibodies. To confirm the specificity of our immunoprecipitation we performed parallel experiments with a protein A-agarose mixture that either lacked the relevant antibody or had been bound to an irrelevant antibody (nonspecific IgG or anti-PPAR␥ antibody). As shown in Fig. 4A, the use of p300 antibody resulted in enrichment of the DNA genomic segment harboring LH␤ promoter, confirming the recruitment of this factor to the promoter. As positive control we used anti-SF-1 or -Egr-1 antibodies, both resulting in precipitation of the genomic LH␤ promoter fragment. In contrast, there was no precipitation using a control, nonspecific IgG (Fig. 4A) or anti-PPAR␥ antibody (not shown).
Because we have demonstrated previously that GnRH enhances the functional interaction of SF-1 and Egr-1, we assessed the influence of GnRH on the recruitment of p300 to LH␤ promoter. Using qualitative PCR in four independent ChIP experiments we consistently found an increase in PCR-amplified DNA from chromatin complexes precipitated with p300 in GnRH-stimulated L␤T2, compared with non-stimulated cells (Fig. 4B). Similar results were found with precipitation of Egr-1, but not SF-1. To quantify the changes in immunoprecipitated chromatin, we used real time quantitative PCR. As shown in Fig. 4C, a significant influence of GnRH was observed on precipitation of LH␤ chromatin fragment with p300 antibody. We also confirmed the qualitative PCR data, demonstrating a significant influence of GnRH on precipitation of LH␤ promoter fragment using anti Egr-1 antibodies. As expected, the effect of GnRH on SF-1 was not significant. Together, these results indicate that p300 is physically recruited to endogenous LH␤ promoter in L␤T2 cells and that this recruitment is enhanced following stimulation by GnRH.
p300 Potentiates the Effect of SRC/p160 Co-activators on LH␤ Promoter-SF-1 has been shown to interact with all three members of the p160 steroid receptor co-activator family of proteins, SRC-1 (28, 36), SRC-2/GRIP1/TIF2 (29, 39), and SRC-3/p/CIP/AIB1/ACTR/RAC3/TRAM-1 (30, 40 -44). These SRC proteins also enhance transactivation by additional nuclear receptors, where these SRC proteins likely enhance the recruitment of p300 (45)(46)(47)(48)(49)(50). Interestingly, all three SRC proteins have been detected in the pituitary (review in Ref. 51), with widespread expression of SRC-1 SRC-3 throughout the brain, but more restricted expression of SRC-2 to the pituitary (52). We therefore sought to determine whether overexpression of individual SRC/p160 proteins potentiates the effect of p300 on LH␤ stimulation. We transfected CV1 cells with expression plasmids for SF-1, Egr-1, and increasing amounts of each SRC expression plasmid, in the absence or presence of p300. As expected, transfection of SRC-1, SRC-2, or SRC-3 alone enhanced the transcription of LH␤ reporter gene in the absence of p300 (Fig. 5). Importantly, addition of p300 markedly amplified the effect of the three SRC proteins, particularly SRC-2. These results are consistent with a cooperative interaction between SRC/p160 and p300 and suggest that pituitary SRC proteins may enhance the recruitment of p300 to LH␤ promoter.
The Influence of p300 on SF-1 Interaction with Other Egr Proteins-In addition to Egr-1, other members of Egr family bind to promoter sites that are either identical or similar to the Egr-1 sites within the proximal region of LH␤ promoter (53). Moreover, we have demonstrated recently (5) that all Egr isoforms exhibit some degree of synergistic interaction with SF-1 in activation of LH␤ promoter. However we also showed that Egr-1-deficient mice selectively exhibit a very low level LH␤ expression, indicating that Egr family members are not redundant with Egr1 and cannot compensate for the absence of Egr-1 in regulation of LH␤ expression in pituitary gonadotropes. We therefore sought to determine whether p300 modulates the activity of Egr proteins in the context of LH␤ promoter. We transfected CV1 cells with an expression plasmid for each Egr protein in the absence or presence of SF-1, p300, or both and assessed their function in transactivation of LH␤-driven luciferase gene. Whereas p300 markedly potentiated the activity of Egr-1 (15-fold), its effect on other Egr members was weak (up to 2-fold) and statistically insignificant (Fig. 6). These data indicate that the influence of p300 on LH␤ promoter is unique to the interaction of SF-1 with Egr-1 and likely less relevant to LH␤ activation by other Egr proteins.
E1A Potentiates the Activity of LH␤ Gene-The adenoviral protein E1A interacts with p300 and has been shown to attenuate the co-activator functions of p300 (54 -56). To further demonstrate the involvement of p300 in activation of LH␤ promoter by SF-1/Egr-1 we sought to determine whether E1A represses LH␤ transactivation by SF-1 and Egr-1. Surprisingly, transfection of either 13S or 12S subunit of E1A into CV1 cells resulted in potent activation of LH␤ promoter (Fig. 7A). The direct interaction of p300, Egr-1, and SF-1 with the endogenous LH␤ promoter was determined by ChIP analysis in L␤T2 cells. Cells were treated with formaldehyde, washed, lysed, and sonicated. The DNA-protein complexes were immunoprecipitated with indicated antibodies and captured with protein A-agarose beads. The DNA was purified and amplified with primers flanking LH␤ promoter. Input DNA represents a fraction of the sonicated chromatin prior to immunoprecipitation. A, chromatin preparations from untreated cells were immunoprecipitated with p300, Egr-1, or SF-1 antibodies, and precipitated DNA was subjected to 30 cycles of PCR. Amplification products were visualized with ethidium bromide staining. Rabbit nonimmune IgG (or anti PPAR␥ antibodies; not shown) was included as a negative control. B, L␤T2 cells were stimulated with or without GnRH (10 Ϫ7 M), and immunoprecipitated DNA was amplified using PCR. Densitometric analysis of the bands is shown below the figure, demonstrating the influence of GnRH on precipitation of p300, Egr-1, or SF-1. C, chromatin immunoprecipitation was quantitated by real-time PCR, as described under "Experimental Procedures." The results depict mean Ϯ S.D., derived from four independent experiments, expressed as -fold from GnRH-stimulated cells over non-stimulated cells for each antibody used in immunoprecipitation. Control immunoprecipitation with IgG was arbitrarily defined as one. * denotes p Ͻ 0.05 compared with IgG control (ANOVA, with post hoc Bonferroni t test).

FIG. 5. Synergy of SRC proteins with p300 on LH␤ promoter.
CV-1 cells were transiently transfected with SF-1 (0.1 g) and Egr-1 (0.4 g) in the absence or presence of p300 (1.5 g), along with the LH␤-luciferase construct (0.5 g) and 0.5-1.5 g of SRC-1, SRC-2, and SRC-3 expression plasmids or an empty vector (labeled as control). The total amount of transfected DNA was kept uniform by adding empty expression plasmids. Luciferase activity was measured 48 h after transfection. Results (mean Ϯ S.D.) are expressed as RLU and represent three independent experiments, each performed in duplicate. The effect of p300 in the presence of each SRC construct (0.5 or 1.5 g) was significant (ANOVA; p Ͻ 0.01, with post hoc Bonferroni t test p Ͻ 0.05).
p300 Enhances the Synergy between SF-1 and Egr-1 Whereas either E1A 12S or E1A 13S up-regulated the activity of LH␤ promoter, the effect of E1A 13S was more potent (48versus 6-fold, respectively; see Fig. 7A). This effect of E1A subunits was abrogated by the disruption of SF-1 or Egr-1 binding sites within the LH␤ promoter (not shown). We repeated the experiments in 293 cells, which express the adenoviral E1A gene endogenously and exhibit high E1A activity (57). As expected, introduction of either E1A 12S or E1A 13S expression plasmids into 293 cells did not further enhance activation of LH␤ promoter (not shown). Together, these observations suggest that E1A proteins regulate the activity of LH␤ promoter. Because E1A interacts with other nuclear receptors, including thyroid receptor and retinoic acid receptor-␤ (58), we hypothesized that E1A stimulates LH␤ promoter by enhancement of SF-1 activity. Indeed, as shown in Fig. 7A, E1A modestly up-regulates (4-fold) the activation of LH␤ promoter by SF-1 even in the absence of Egr-1 or p300 (Fig. 7A). E1A proteins had a similar effect on Egr-1-stimulated LH␤ promoter, albeit to a lesser magnitude. We reproduced the effect of E1A on SF-1 using GAL4-SF-1 chimeric protein (35), co-transfected with a GAL4 reporter plasmid into CV-1 cells, in the presence or absence of co-expressed E1A subunits (not shown). Interestingly, the influence of E1A on LH␤ promoter was dramatically more potent (112-fold) in the presence of co-expressed p300 (Fig. 7A). Nevertheless, this influence was unchanged when we used p300del30, rendering p300 incapable of physical interaction with E1A. We next assessed the influence of increasing E1A 13S level on p300-stimulated SF-1/Egr-1 synergy. As shown in Fig. 7B, E1A 13S potentiated LH␤ activation by SF-1 alone, but not Egr-1 alone, with maximum effect using 0.1 g of transfected plasmid (lane 4). This potentiation was enhanced in the presence of co-transfected SF-1/Egr-1, with maximum effect reproduced using 0.1 g of transfected plasmid (lane 16). Importantly, the effect of E1A was markedly enhanced in the presence of co-transfected p300, and maximum stimulation was consistently observed with as little as 10 ng of E1A 13S plasmid (lane 20). Taken together, our data suggest that E1A stimulates activation of LH␤ promoter primarily via SF-1. This stimulation by E1A is dramatically enhanced in the presence of p300, even without direct interaction with E1A.
SNIP1 Inhibits Transactivation of LH␤ Promoter by SF-1 and Egr-1-It has been demonstrated recently (31,59) that SNIP1 suppresses p300/CBP-dependent transcriptional activation, likely by competing for binding of transcription factors requiring the C/H1 domain of CBP/p300. We therefore investigated whether SNIP1 might inhibit the transcriptional activation of LH␤ promoter by targeting p300 in our system. Transfection of increasing amount of SNIP1 expression vector in 293 cells resulted in a dramatic, concentration-dependent attenuation of LH␤ promoter activation by SF-1 and Egr-1 promoter (Fig. 8A). We observed similar results in CV1 cells, albeit to a lesser magnitude (not shown). Consistent with the hypothesis that SNIP1 sequestrates p300, overexpression of p300 in 293 cells reversed LH␤ promoter inhibition by SNIP1 (Fig. 8B), thereby providing further support to the influence of p300 on the transcriptional activity of LH␤ promoter.

DISCUSSION
The synergistic interaction of SF-1 and Egr-1 plays a critical role in regulation of LH␤ expression in gonadotropes (5,9,17,60). The mechanism of this interaction remained unclear. In this study we showed that the co-activator protein p300 is recruited to LH␤ promoter, where it potentiates the interaction of SF-1 and Egr-1. Importantly, disruption of either SF-1 or Egr-1 elements within LH␤ promoter, which abrogates SF-1/ Egr-1 synergy, attenuated p300-stimulated promoter activity. These data indicate that p300 cannot exert its strong transcriptional enhancement activity without binding of SF-1 and Egr-1 to LH␤ promoter. Previous studies using an in vitro pull-down approach have shown that SF-1 and Egr-1 physically interact and that this interaction could partly explain the synergy between these two factors (9,17). Interestingly, this interaction was not reproducible by Halvorson et al. using electrophoretic mobility shift assay (9,17), suggesting that the interaction is weak. Consistent with this notion, we could not demonstrate an interaction of p300 with SF-1 and Egr-1 using co-immunoprecipitation. A weak interaction of p300 with the SF-1/Egr-1 complex may not be surprising, as it would allow rapid on/off response to GnRH signals, necessary for fine-tuned regulation of LH␤ expression. In our experiments SF-1, Egr-1, and p300 are co-expressed and can potentially interact with each other. The finding that efficient transcriptional activation by p300 occurs only when both SF-1 and Egr-1 bind to their promoter elements indicates that physical interaction between SF-1 and Egr-1 is not sufficient, and SF-1/Egr-1 binding to their promoter cognate sites constitutes an obligatory step for the recruitment of p300 to LH␤ promoter and consequently efficient activation. These observations suggest a mechanism by which p300 acts as an adaptor molecule that links SF-1 and Egr-1 and thereby support the transcriptional synergy between these two proteins in the context of LH␤ promoter. Our findings using nonspecific cell lines were corroborated using L␤T2 cells, a gonadotrope cell line that expresses gonadotropins, including LH␤, and responds to GnRH (32). Transfection of p300 in L␤T2 cells resulted in the enhancement of LH␤ promoter. Furthermore, our ChIP analysis confirmed the recruitment of endogenous SF-1, Egr-1, and p300 to a genomic fragment harboring LH␤ promoter. Although these three transcription factors are associated with the promoter even in the absence of stimula-FIG. 6. The influence of p300 on SF-1 synergy with Egr proteins. CV-1 cells were transiently transfected with expression plasmids for SF-1 (0.1 g) and Egr-1, -2, -3, or -4 (0.4 g each) in the presence or absence of a plasmid for p300 (3 g), along with the LH␤-luciferase construct (0.5 g). The total amount of transfected DNA was kept uniform by addition of empty expression plasmids. Luciferase activity was measured 48 h after transfection. Results (mean Ϯ S.D.) are expressed as RLU and represent three independent experiments, each performed in duplicate. The effect of p300 on the interaction of SF-1 and Egr-1 was significant (ANOVA; p Ͻ 0.001, with post hoc Bonferroni t test p Ͻ 0.05).
p300 Enhances the Synergy between SF-1 and Egr-1 tion by GnRH, the association of p300 and Egr-1 with the endogenous LH␤ promoter is enhanced by GnRH. The stimulation by GnRH is consistent with the influence of GnRH on Egr-1 expression (5,9,17,60) and is of sufficient magnitude to up-regulate LH␤ production by L␤T2 cells (not shown).
The Egr family of proteins includes Egr-1 (NGFI-A, krox24, and zif-268) (61-63), Egr-2 (Krox20) (64,65) teins. Indeed, targeted ablation of each Egr protein results in a unique phenotype that cannot be compensated by the presence of other Egr proteins (8,14). Relevant to the present study, Egr-1-null mice exhibit profound deficiency of LH␤ in pituitary gonadotropes (8,14). Surprisingly, studies using L␤T2 cells and other cell lines have shown that all four members of the Egr family exhibit some degree of synergy with SF-1 in activating LH␤ promoter (5). Here we demonstrated that p300 overexpression results in marked potentiation of the synergy between SF-1 and Egr-1 and a weak effect on the synergy between SF-1 and Egr-4. The interaction between SF-1 and either Egr-2 or Egr-3 was not supported by p300. Although Egr-1 exhibits pleiotropic functions in diverse systems, our present data, as well as our previously published results, provide strong support to the unique role of Egr-1 in regulation of LH␤, because (1) Egr-1 deficiency results in profound reduction in expression of LH␤ and infertility (8,14), (2) GnRH uniquely stimulates the expression of Egr-1, but not other Egr proteins, in pituitary gonadotropes (5), and as shown in the present work (3), p300 specifically potentiates the interaction of Egr-1 with SF-1 in activation of LH␤ promoter.
We used additional approaches to buttress our conclusions regarding the role of p300 in regulation of LH␤ promoter by SF-1/Egr-1. In the first approach we examined the influence of SRC/p160 proteins on activation of LH␤ promoter by p300. As noted, SF-1 has been shown to interact with the different members of the SRC/p160 (28 -30, 36, 39 -44), and all three SRC proteins have been detected in the pituitary (review by Ref. 51), with specific expression of SRC-2 to the pituitary (52). In addition, several studies have shown that effective recruitment of p300 required the presence of SRC-1 (45,46,50). We found that co-expression of SRC/p160 proteins potentiates the co-activator effect of p300, suggesting that SRC proteins, most notably SRC-2, promote the interaction of p300 with SF-1 and Egr-1 on LH␤ promoter. In the second approach we assessed the influence of adenoviral E1A proteins on activation of LH␤ promoter by p300. E1A has been shown to abrogate the coactivator function of p300 in diverse cellular systems (reviewed in Ref. 70). We therefore reasoned that co-expression of E1A would counteract the action of p300 and consequently reduce the synergy between SF-1 and Egr-1. To our surprise we found that instead of a reduction in SF-1/Egr-1-mediated transactivation of LH␤ promoter we observed a marked increase of promoter activity. The stimulatory effect of E1A was seen even in the presence of SF-1 alone, albeit to a lesser extent. This is consistent with the original identification of E1A as a transcriptional activator (71,72) and with several observations that demonstrated the role of E1A as a co-activator for nuclear receptors (58,73). Importantly, we found that the stimulatory effect of E1A is concentration-dependent, with peak effect at 100 ng of transfected E1A 13S plasmid in the case of SF-1 or SF-1/Egr-1, followed by transcriptional repression. In contrast, the stimulatory influence of E1A 13S is markedly enhanced in the presence of p300, and peak effect is observed at a lower concentration (Ͻ10 ng; see Fig. 7B). The dual role of E1A as a transcriptional co-activator and a corepressor points to complex interaction of this protein with diverse basal and promoterspecific transcription factors. The mechanism underlying these divergent roles of E1A remains unknown. Although in our study we focused on the influence of E1A on SF-1/Egr-1, other transcription factors are recruited to the LH␤ promoter fragment employed in our studies. For example, a Ptx1 binding element that has been shown to be important for regulating LH␤ gene (9,10,74) is nested between the pairs of Egr-1 and SF-1 elements in LH␤ promoter. It is likely that physiological activation of LH␤ promoter by GnRH stimulation results in recruitment of p300-bound SF-1/Egr-1, along with additional LH␤ regulators, such as Ptx-1 and basal promoter factors. There is growing evidence indicating that transcriptional activation is a multi-step process, involving a complex array of different chromatin-modifying co-activators (75). We speculate that at low concentration E1A recruits additional co-factors to the promoter. These factors likely potentiate the effect of SF-1 and synergistically interact with p300 to further amplify the transcriptional response (Fig. 7B). Higher expression of E1A results in transcriptional interference, likely reflecting squelching of other transcription factors that are paramount to the activity of LH␤ promoter.
We attempted to suppress the activity of the endogenous p300 using a dominant-negative form of p300. The fragment of p300 1514 -1922 , which harbors the CH3 domain and additional peptide regions, has been reported to act as a dominant-negative peptide in diverse transcriptional pathways involving p300 (26, 76 -79). Although we observed a concentration-dependent reduction of the synergy between SF-1 and Egr-1 following transfection of p300 1514 -1922 into 293 cells, which express high E1A activity, we could not reproduce this result in CV-1 or JEG3 cells (not shown). The p300 1514 -1922 fragment contains regions that mediate binding of E1A. It is therefore likely that in 293 cells the expression of p300 1514 -1922 might target E1A and impede the formation of an active transcriptional complex between the endogenous E1A and SF-1.
In our final approach we used SNIP1 to sequester p300 away from LH␤ promoter and consequently inhibit its transcrip- FIG. 8. SNIP1 inhibits the SF-1/Egr-1-dependent activation of LH␤ promoter. A, 293 cells were transiently transfected with expression plasmids for SF-1 (0.1 g) and Egr-1 (0.4 g) in the presence of increasing concentrations of SNIP1, along with the LH␤-luciferase construct (0.5 g). The total amount of transfected DNA was kept uniform by addition of empty expression plasmids. Luciferase activity was measured 48 h after transfection. Results (mean Ϯ S.D.) are expressed as RLU and represent three independent experiments, each performed in duplicate. B, 293 cells were co-transfected 0.5 g of the reporter LH␤-luciferase with SF-1, Egr-1, and SNIP1 alone or with the indicated amount of p300 expression vectors. Luciferase activity was measured 48 h after transfection. Results (mean Ϯ S.D.) are expressed as RLU and represent three independent experiments, each performed in duplicate. * denotes p Ͻ 0.05 compared with control (ANOVA, with post hoc Bonferroni t test). tional activation by SF-1 and Egr-1. We did not transfect p300 in a first set of experiments. Thus, inhibition of LH␤ promoter by SNIP1 likely occurred through endogenous p300. In support of this hypothesis we showed that overexpression of p300 can overcome inhibition by SNIP1. These data buttress the role of p300 as an integrator in the transcriptional activation of LH␤ by SF-1 and Egr-1.
Our study provides molecular and biochemical evidence that supports the recruitment and the function of p300 in the activation of LH␤ promoter. Our data also point to the complex combinatorial interaction of transcription factors that regulate LH␤ promoter. This complexity is not surprising, considering the fact that exquisite regulation of LH␤ expression is critical for reproductive function of both sexes, and fine-tuned signals must integrate to achieve precise and accurate expression of this gonadotropin.