Conserved Amino Acids within CCAAT Enhancer-binding Proteins (C/EBPα and β) Regulate Phosphoenolpyruvate Carboxykinase (PEPCK) Gene Expression*

Thyroid hormone and cAMP stimulate transcription of the gene for phosphoenolpyruvate carboxykinase (PEPCK). CCAAT enhancer-binding proteins (C/EBPα and β) are involved in multiple aspects of the nutritional, developmental and hormonal regulation ofPEPCK gene expression. Previously, we have identified a thyroid hormone response element in the PEPCKpromoter and demonstrated that C/EBP proteins bound to the P3(I) site are participants in the induction of PEPCK gene expression by thyroid hormone and cAMP. Here, we identify several peptide regions within the transactivation domain of C/EBPα that enhance the ability of T3 to stimulate gene transcription. We also demonstrate that several conserved amino acids in the transactivation domain of C/EBPα and C/EBPβ are required for the stimulation of basal gene expression and identify amino acids within C/EBPβ that participate in the cAMP induction of the PEPCK gene. Finally, we show that the CREB-binding protein (CBP) enhanced the induction ofPEPCK gene transcription by thyroid hormone and that CBP is associated with the PEPCK gene in vivo. Our results indicate that both C/EBP proteins and CBP participate in the regulation of PEPCK gene transcription by thyroid hormone.

Thyroid hormone and cAMP stimulate transcription of the gene for phosphoenolpyruvate carboxykinase (PEPCK). CCAAT enhancer-binding proteins (C/EBP␣ and ␤) are involved in multiple aspects of the nutritional, developmental and hormonal regulation of PEPCK gene expression. Previously, we have identified a thyroid hormone response element in the PEPCK promoter and demonstrated that C/EBP proteins bound to the P3(I) site are participants in the induction of PEPCK gene expression by thyroid hormone and cAMP. Here, we identify several peptide regions within the transactivation domain of C/EBP␣ that enhance the ability of T 3 to stimulate gene transcription. We also demonstrate that several conserved amino acids in the transactivation domain of C/EBP␣ and C/EBP␤ are required for the stimulation of basal gene expression and identify amino acids within C/EBP␤ that participate in the cAMP induction of the PEPCK gene. Finally, we show that the CREB-binding protein (CBP) enhanced the induction of PEPCK gene transcription by thyroid hormone and that CBP is associated with the PEPCK gene in vivo. Our results indicate that both C/EBP proteins and CBP participate in the regulation of PEPCK gene transcription by thyroid hormone.
Transcription of the gene for phosphoenolpyruvate carboxykinase (PEPCK) 1 is stimulated in various nutritional and pathologic states such as high protein diets, fasting, hyperthy-roidism, and diabetes (1). Multiple hormones including glucagon (via cAMP), thyroid hormone (T 3 ), glucocorticoids, and retinoic acid increase PEPCK gene expression (2). The hormonal induction of the PEPCK gene by T 3 and glucocorticoids is mediated through the binding of nuclear receptors to weak hormone response elements in the PEPCK promoter (3,4). The full induction by these hormones requires accessory factors associated with the PEPCK promoter. Our studies and those of others have shown that CCAAT enhancer-binding proteins (C/ EBP) are accessory factors required for the stimulation by cAMP, T 3 , and glucocorticoids (5)(6)(7)(8).
We have defined two critical sites in the PEPCK promoter that are required for stimulation by T 3 (6). A thyroid hormone response element (TRE) (Ϫ330/Ϫ320) binds the thyroid hormone receptor (TR) as a heterodimer with the retinoid X receptor (RXR). In addition, a site called P3(I) (Ϫ250/Ϫ234) binds C/EBP␣ and C/EBP␤ (7). Both sites are required for T 3 to stimulate PEPCK gene expression. The induction of PEPCK transcription by cAMP involves multiple sites in the promoter including a cAMP response element (CRE) (Ϫ90/Ϫ82) and the P3(I) site (2,9). The PEPCK CRE can bind both CREB and C/EBP proteins with similar affinity (10). The P3(I) site is involved in both the T 3 and cAMP induction of the PEPCK gene. Glucocorticoids induce PEPCK gene expression through two weak glucocorticoid response elements, but multiple accessory factors are involved in the glucocorticoid induction of the PEPCK gene including C/EBP␤ bound to the CRE (8,11). Therefore, C/EBP proteins are centrally involved in regulating multiple hormone responses of PEPCK gene expression. C/EBP proteins also contribute to the tissue-specific expression and developmental regulation of the PEPCK gene (12).
The CCAAT enhancer-binding proteins consist of a family of bZIP proteins. The transactivation domain is contained in the amino terminus, while the DNA binding domain and leucine zipper are contained within the carboxyl-terminal region (13). C/EBP␣ and ␤ have been shown to have important roles in directing the expression of many genes encoding metabolic enzymes in the liver. In addition, C/EBP isoforms have prominent roles in adipocyte differentiation (14). However, these isoforms are not redundant. C/EBP␣ is a terminal differentiation factor that is associated with inhibition of cell division (15). C/EBP␣ stimulates expression of the PEPCK gene at birth. Both C/EBP␣ and C/EBP␤ knockout mice have impaired expression of the PEPCK gene (12,16).
The T 3 induction of gene transcription is mediated through the binding of the liganded TR to hormone response elements (17). The TR binds to TREs primarily as a heterodimer with RXR (18). T 3 is not required for DNA binding and in the absence of ligand the TR generally acts as a repressor of gene expression. Without T 3 , the TR is associated with nuclear corepressors such as NCoR and SMRT (17). When ligand is added, various coactivators may be recruited to the nuclear receptors including steroid receptor coactivator (SRC-1/NcoA-1), CREBbinding protein (CBP/p300), and thyroid receptor accessory proteins (TRAP/DRIP/PBP) (17,19). SRC-1 can interact with many liganded nuclear receptors and with orphan receptors such as HNF-4 and COUP-TF through a conserved LXXLL peptide motif (20). CBP is associated with SRC proteins and liganded nuclear receptors. CBP was initially described as a coactivator for CREB and enhancer of cAMP responsiveness (21). CBP is able to interact with a variety of proteins and therefore offers the potential for mediating the interactions between receptors and accessory factors (22). In these studies, we have defined specific regions within C/EBP that are involved in the T 3 induction of PEPCK transcription. In addition, we provide evidence that CBP can participate in the T 3 induction of PEPCK gene transcription.

Construction of CAT and Luciferase Vectors-
The ligation of the PEPCK promoter from Ϫ490 to ϩ73 to the CAT reporter gene (Ϫ490-PCAT) has been described (3). The introduction of the Gal4 binding site into the P3(I) site of the PEPCK promoter to create Ϫ490-P3G4-CAT was described previously (6). The Ϫ490 to ϩ73 region of the PEPCK promoter was ligated in front of the luciferase reporter gene by removing the PEPCK promoter fragment from Ϫ490-PCAT by digestion with KpnI and BglII and ligating into the polylinker of pGL3 basic (Promega). The Gal4 site was introduced into the TRE region of Ϫ330-PTRE/G4-CAT by PCR amplification with the 5Ј-primer, ccctctagatcggaggtactgtcctccgtctgac, containing the altered nucleotides and a 3Ј-primer, ttagatctcagagcgtctcgcc (ϩ73 to ϩ52), which includes the BglII site at ϩ73. The 5Ј-primer introduced an XbaI site. The amplified promoter fragment was digested with XbaI and BglII and ligated in front of the CAT reporter gene. The sequence was confirmed by sequence analysis at the St. Jude Center for Biotechnology (Memphis, TN).
Construction of Gal4-C/EBP Expression Vectors-The Gal4-C/EBP␤ vectors with alanine substitutions were constructed by two-step PCR amplification. The initial PCR reactions contained the forward primer, which contained an EcoRI site, and the first 18 nucleotides of the rat C/EBP␤ cDNA (tccgaattcatgcaccgcctgctggcctgggac) and a reverse primer with the alanine switches M27,28 (ggcagtcggggctcgtaggcggcgttggccacttccatg), M57,58,59 (gaagtcgatggcgcgcgcggccgcgccaatggccggctc), or M61,62 (ccaggtaggggctgaaggcggcggcgcgcgtgtgctcg). Additional PCR reactions contained forward primers with the alanine substitutions M27,28 (catggaagtggccaacgcccgcctacgagcccgactgcc), M57,58,59 (gagccggccgattggcgcggccgcgcgcgccatcgacttc), or M61,62 (cgagcacgagcgcgccgccgccttcagcccctacctgg) and reverse primers containing PstI sites and the nucleotides representing amino acids 108 -102 (gagctgcaggtaaccgtagtcggccggcttc). The PCR reactions were conducted using the PCR kit from CLONTECH and consisted of 20 cycles of 94°C for 30 s and 68°C for 2 min. All PCR reactions contained 10% GC melt buffer (CLONTECH) as this region of mouse C/EBP␤ is extremely GC rich. The mouse C/EBP␤ cDNA was the template (23). The PCR products were purified from agarose gels. The PCR products encompassing approximately amino acids 1-70 and 60 -108 were mixed along with the outside primers, and the PCR reactions were repeated. The appropriate 330-base pair DNA fragment was isolated from an agarose gel and subcloned into TOPO-TA vector (Invitrogen) as outlined by the manufacturer. The C/EBP␤ fragment was removed from TOPO-TA by digestion with EcoRI and PstI. This DNA fragment was ligated into the mammalian Gal4 DNA expression vector called pM (CLONTECH).
To create the Gal4-C/EBP␤-(1-100) and the Gal4-C/EBP␤-(1-100)-M86,87 vectors, PCR reactions were conducted with the forward primer encompassing nucleotides 1-18 and a reverse primer encoding either the wild-type sequence (cggctgcaggctcggcttggcgccgtagtcg) or containing mutations in the amino acids 86 and 87 (cggctgcaggctcggcttggcgccgtcgtcgtcggcgaagaggtcggagccgccgtcgtggtgcg). PCR conditions and subcloning into TOPO-TA were conducted as described above. Construction of the Gal4-C/EBP␣ vectors was described elsewhere (5,24). The sequence of all Gal4-C/EBP␤ vectors was confirmed by sequence analysis.
To construct the C/EBP␣ prey vectors for the mammalian two-hybrid assays, the C/EBP␣ fragment was isolated from the Gal4-C/EBP␣ vector by digestion with EcoRI and PstI. These C/EBP␣ fragments were ligated into the VP16 vector (CLONTECH).
Cell Transfections, Luciferase, and CAT Assays-HepG2 cells were transfected by calcium phosphate precipitation as described previously (6). CAT assays were conducted with [ 3 H]chloramphenicol and n-butyryl coenzyme A using the xylene phase extraction method (6). All transfections were performed in duplicate and repeated 3-6 times. Luciferase assays were conducted with the luciferin reagent as outlined by the manufacturer (Promega).
Chromatin Immunoprecipitation Assay-We used a modification of the technique described by Shang et al. (25). A 1% solution of formaldehyde prepared in buffer (0.1 M NaCl, 1 mM EDTA, 0.5 mM EGTA, 50 mM Hepes, pH 8.0) was added to hepatocytes for 5 min at 4°C to cross-link DNA and its associated proteins. Hepatocytes were prepared as we have described previously (26). The cross-linking reaction was stopped by the addition of glycine. Cross-linked cells were then recovered by centrifugation and washed three times with 5 ml of ice-cold phosphate-buffered saline. The cells were resuspended with 1 ml of buffer (20 mM Tris-HCl, pH 8.0, 2 mM EDTA, 150 mM NaCl) plus protease inhibitors (1 mM phenylmethylsulfonyl fluoride, 1 mM benzamidine, 0.1 g/ml aprotinin, 1 g/ml leupeptin, and 0.1 g/ml pepstatin) and sonicated at 4°C for 10 s at maximum setting. The sonication was repeated five times after 30 s intervals at 4°C.
The sonicated cells were centrifuged for 10 min at 4°C to remove cell debris, and each supernatant was diluted up to a final volume of 1.5 ml with binding buffer (20 mM Tris-HCl, pH 8.0, 2 mM EDTA, 150 mM NaCl, 1% Triton X-100) plus protease inhibitors. Each supernatant was precleared by adding 50 l of bovine serum albumin-blocked protein A-Sepharose (for rabbit antibodies) or protein G-Sepharose (for mouse antibodies). These mixtures were incubated overnight at 4°C with constant shaking, and precleared supernatant was recovered by centrifugation and finally transferred to prechilled microcentrifuge tubes. Immunoprecipitation was performed with specific antisera raised against C/EBP␤, C/EBP␣, TR␤1, and CBP (Santa Cruz Biotechnology). As a control, monoclonal anti-polyhistidine antibody (Sigma Chemical Co.) was used. The mixtures were incubated at 4°C for 1 h followed by isolation of antibody-protein-DNA complexes with 50 l of bovine serum albumin-blocked protein A-Sepharose or protein G-Sepharose for 1 h at 4°C. Immunoprecipitates were recovered by centrifugation, and the resins were washed sequentially three times for 3 min with wash buffer 1 (20 mM Tris-HCl, pH 8.0, 150 mM NaCl, 2 mM EDTA, 0.1% SDS, 1% Triton X-100); wash buffer 2 (20 mM Tris-HCl, pH 8.0, 500 mM NaCl, 2 mM EDTA, 0.1% SDS, 1% Triton X-100); and wash buffer 3 (10 mM Tris-HCl, pH 8.0, 1 mM EDTA, 0.25 M LiCl, 1% Igepal, 1% deoxycholate). Precipitates were then washed three times with TE buffer (10 mM Tris-HCl, pH 8.0, 1 mM EDTA) and extracted two times by incubation for 15 min at room temperature with 250 l of 1% SDS, 0.1 M NaHCO 3 . Eluates were pooled and 30 l of 5 M NaCl were added and heated at 65°C for 3 h to reverse the formaldehyde cross-linking. Proteinase K (Roche Molecular Biochemicals) was added, and the heating continued at 65°C for 3 h. DNA fragments were purified with pellet paint kit (Novagen). Precipitated DNA was washed with 70% ethanol, air-dried for 10 min, and resuspended in 100 l of sterile water. Bound DNA fragments were analyzed by PCR using AmpliTaq Gold Kit (Applied Biosystems Group), 2 M of each primer, and 5 l of immunoprecipitated DNA per reaction. Cycling parameters were: 1 cycle of 94°C for 9 min, 30 cycles of 94°C for 30 s, 63°C for 30 s, 72°C for 30 s, and 1 cycle at 72°C for 7 min. The primers used for amplification of promoter rat PEPCK were: forward TRE (Ϫ479/Ϫ459), 5Ј-cacgtctcagagctgaattcccttc-3Ј and reverse TRE (Ϫ290/Ϫ314), 5Ј-actataggctcttgccttaattgtc-3Ј. Amplified PCR products were electrophoresed through a 3% Nusieve/agarose gel in Tris acetate/EDTA buffer and visualized by ethidium bromide staining.

RESULTS
In previous studies, we found that mutation of the C/EBP binding site called P3(I) in the PEPCK promoter eliminated the T 3 induction of the PEPCK gene (6). A model of some of the key regulatory elements in the PEPCK promoter is shown in Fig.  1A. Our first experiments were designed to demonstrate that the P3(I) site could function as an enhancer of T 3 responsiveness out of the context of the PEPCK gene. One copy of an idealized thyroid hormone response element (DR4) was ligated in front of a minimal PEPCK promoter from Ϫ68 to ϩ73 that contains only a TATA box but no hormone response elements in order to create DR4X1 Ϫ68PLuc. A schematic of the luciferase vectors used in these studies is shown in Fig. 1B. Addition of three copies of the P3(I) element (DR4X1 P3WTX3 Ϫ68PLuc) greatly enhanced the T 3 induction from 0.3-to 2.4-fold (Table  I). We added three copies of the P3(I) site because there are three C/EBP binding sites in this region of the PEPCK gene including P3(I), P4(I), and P4(II) (2). A Gal4 site was ligated in front of the enhancerless SV40 promoter driving the luciferase reporter gene to generate Gal4X1 SV40-Luc. Cotransfection of this vector with Gal4-TR␤ allowed a strong 8.1-fold induction by T 3 . Addition of three P3(I) sites in the Gal4X1 P3WTX3 SV40-Luc increased this stimulation to 61-fold. This induction could be reduced either by cotransfection with a dominant negative C/EBP vector (A-C/EBP) or the introduction of mutations into the P3(I) site (P3Mut243 or P3Mut247) that eliminated the ability of C/EBP to bind the P3(I) site (27) (Table I).
The A-C/EBP vector will inhibit the binding of all C/EBP isoforms (27). These results indicate that C/EBP can enhance the T 3 induction through a TRE and that this enhancement does not require it in the context of the PEPCK promoter.
The next experiments were designed to identify domains within C/EBP␣ or C/EBP␤ that were involved in the enhancement of T 3 responsiveness. To test the ability of the TR and C/EBP to synergize, one copy of an idealized TRE and one copy of a Gal4 site were ligated in front of the Ϫ68PEPCK-Luc reporter gene. HepG2 cells were cotransfected with mammalian expression vectors for RSV-TR␤ and Gal4-C/EBP␣ as well as the DR4X1 Gal4X1 Ϫ68PLuc reporter gene. In the absence of C/EBP, a single TRE was unable to confer T 3 responsiveness to the minimal PEPCK promoter (Fig. 2). The Gal4-C/EBP␣-(6 -217) vector that contains the transactivation domain allowed a 4.1 Ϯ 0.2-fold induction by T 3 . Deletion of amino acids 175-217 diminished the T 3 induction. Likewise deletion of the first 50 amino acids in the Gal4-C/EBP␣-(50 -217) reduced the T 3 response. Three amino acids, tyrosine, phenylalanine, and leucine, at positions 67, 77, and 78 in C/EBP␣ had been found to be critical for the induction of basal expression (5). As is shown in Fig. 2, mutation of these amino acids in the vector called Gal4-C/EBP␣-TM reduced but did not eliminate the induction by T 3 . We had reported previously that these amino acids were not required for the cAMP induction of PEPCK gene expression (5).
Our next studies examined the possibility that TR␤ and C/EBP␣ could physically interact. To conduct these studies, we utilized several approaches including GST pull-downs and mammalian two-hybrid assays. We tested whether bacterially expressed GST-C/EBP␣ could pull-down 35 S-labeled TR␤ in GST pull-down assays. In these experiments, we found that GST-C/EBP␣ interacted with [ 35 S]TR␤ although only a small percentage of the input [ 35 S]TR␤ was retained by the GST-C/ EBP␣ (data not shown). Addition of T 3 or His-tagged RXR␣ did not strengthen the interaction. For the mammalian two-hybrid experiments, the Gal4X4 Ϫ68PLuc reporter gene was used. The Gal4-TR␤ was the bait and C/EBP␣-VP16 was the prey. In the absence of T 3 , coexpression of the C/EBP␣-VP16 did not increase luciferase expression (data not shown). Addition of T 3 caused a 26.5 Ϯ 5.8-fold increase in luciferase activity, while cotransfection of C/EBP␣-VP16 increased the T 3 induction to 37.5 Ϯ 12.0 (data not shown). Overall, our results suggested that while C/EBP␣ could interact with TR␤, these interactions were quite weak and most likely did not form the basis for the C/EBP enhancement of T 3 action.
Because both C/EBP␣ and C/EBP␤ are present in rat liver nuclei and can bind to the P3(I) element in the PEPCK promoter, we tested whether Gal4-C/EBP␤ could enhance the T 3 induction of the DR4X1 Gal4X1 Ϫ68PLuc (Fig. 2). Cotransfection with Gal4-C/EBP␤-(3-181) or Gal4-C/EBP␤-(1-100) increased the T 3 induction by 2-fold. However, cotransfection with a Gal4-CREB vector did not enhance the T 3 induction, indicating that this effect was mediated by C/EBP proteins (Fig. 2). These data indicate that regions within the first 100 amino acids of C/EBP␤ could increase the T 3 response. We examined the first 100 amino acids of the transactivation domains of the rat C/EBP␣ and mouse C/EBP␤ and found that several amino acid regions are conserved as is shown in Fig. 3. In particular, the phenylalanine and leucine amino acids are highly conserved. We introduced a mutation in the Gal4-C/ EBP␤-(1-100) expression vector in which the FL amino acids 86 and 87 were switched to alanine. Interestingly, mutation of amino acids 86 and 87 (C/EBP␤ M86,87), which are conserved between C/EBP␣ and C/EBP␤ did not decrease the induction by T 3 (Fig. 2). These results indicate that there is an additional domain within the first 100 amino acids of C/EBP␤ that participates in the T 3 induction of PEPCK transcription.
We tested whether the FL to alanine switch would affect the ability of C/EBP␤ to stimulate basal expression. The Gal4-C/ EBP␤ vectors were cotransfected with either Gal4X3 Ϫ68PLuc (Fig. 4A) or Ϫ490-P3G4-PLuc (Fig. 4B). Mutation of amino acids 86/87 decreased the stimulation of basal expression by C/EBP␤ from 15.2 Ϯ 3.7 to 2.4 Ϯ 0.3, indicating that these amino acids are critical in both the C/EBP␣ and C/EBP␤ isoforms in the stimulation of basal expression (Fig. 4A). We altered several additional amino acids in the transactivation domain of C/EBP␤. The amino acids 54 -70 of the mouse C/EBP␤ are partially conserved in the rat C/EBP␣ in amino acids 55-71 as well as a region of homology in the aminoterminal regions of these proteins (Fig. 3). Mutation of amino acids 56, 57, and 58 or 61 and 62 reduced the ability of Gal4-C/EBP␤ to stimulate basal transcription of a Gal4X3 Ϫ68PLuc reporter gene, but the expression of the Ϫ490-P3G4-PLuc was not reduced as compared with the Gal4-C/EBP␤-(1-108) (Fig.  4). The difference in the response to these Gal4-C/EBP␤ vectors indicates that promoter context as well as transactivation domains contribute to the ability of C/EBP␤ to stimulate transcription. The Ϫ68PLuc vector has a minimal promoter containing only a TATA box, while the Ϫ490 PEPCK promoter has a number of binding sites for other factors. Alteration of amino acids 27 and 28 in C/EBP␤ did not affect the ability of C/EBP␤ to stimulate basal expression.
Previously, we had demonstrated that the P3(I) site was required for the full induction of the PEPCK gene by cAMP (5,7). We tested the ability of these C/EBP␤ vectors to restore protein kinase A (PKA) responsiveness by cotransfecting Ϫ490-P3G4-PLuc with the Gal-C/EBP␤ vectors. The P3G4 vector has a Gal4 site substituted for the P3(I) site in the Ϫ490PLuc (Fig.  1B). Mutation of amino acids 61 and 62 and to a lesser extent 56, 57, and 58 in the Gal4-C/EBP␤ vectors decreased the PKA response. Amino acids 60 -72 of C/EBP␣ have been implicated in the cAMP induction of the PEPCK gene (28). These results suggest that the conserved amino acids between 55 and 71 in C/EBP␣ and C/EBP␤ may be important in the contribution of these proteins to cAMP responsiveness. C/EBP␤ M86,87 was as effective as the Gal4-C/EBP␤-(1-100) in mediating a PKA induction although the basal expression was greatly decreased (Fig. 5). The Gal4-C/EBP␤ was not able to mediate a cAMP induction out of the context of the PEPCK gene as overexpression of PKA did not increase the activity of Gal4X3 Ϫ68PLuc when cotransfected with Gal4-C/EBP␤ (data not shown).
Because we observed only a minimal physical interaction between C/EBP␣ and TR␤, we next tested whether overexpression of the coactivators SRC-1 or CBP could enhance the T 3 induction of the PEPCK gene. To conduct these experiments, we transfected Ϫ490-PCAT with RSV-TR␤ and mammalian expression vectors for SRC-1 and/or CBP. Overexpression of SRC-1 enhanced the basal expression of Ϫ490-PCAT 3-fold, and the reporter gene was stimulated an additional 4-fold by the addition of T 3 (Fig. 6). These results indicate that SRC-1 can interact with factors bound to the PEPCK promoter to enhance the basal expression of the gene. Overexpression of CBP did not elevate the basal expression of PEPCK-CAT. However, the T 3 response was increased from 4.9 Ϯ 0.7 to 7.1 Ϯ 0.7-fold. This stimulation was significant at a p value of 0.045 using a one-tailed Student's t test. Cotransfection of SRC-1 and CBP did not further increase the effect of T 3 . These results indicate that CBP can enhance the T 3 induction of the PEPCK gene. We tested whether tethering CBP next to a single TRE would restore T 3 responsiveness as had the Gal4-C/EBP␣-(6 -217) (Table II). Full-length CBP was ligated to Gal4 to create Gal4-CBP. Both the Gal4-C/EBP␣ and the Gal4-CBP stimulated the basal expression of the reporter gene (data not shown). The Gal4-CBP was able to restore T 3 responsiveness to this vector suggesting that recruitment of CBP to the PEPCK promoter would enhance the induction by T 3 .

FIG. 2. Identification of domains within C/EBP that enhance thyroid hormone responsiveness.
One copy of a T 3 responsive element (DR4X1) was ligated with one copy of a Gal4 site in front of Ϫ68 PEPCK-luciferase reporter gene. A model of the reporter vector is shown at the top of the figure. Transfections included 3 g of the luciferase reporter gene, 3 g of RSV-TR␤, and 0.5 g of the Gal4 expression vector into HepG2 cells. The numbers of the Gal4-C/EBP vectors indicate the amino acids of C/EBP. The Gal4-C/EBP␣-TM vector has alanines introduced at amino acids 67, 77, and 78. Cells were exposed to 100 nM T 3 in serum-free medium for 16 h and then harvested. The data are expressed as luciferase activity corrected for protein and transfection efficiency. All transfections were conducted in duplicate and repeated at least four times.

FIG. 3. Amino acid sequences of C/EBP␤ and C/EBPa.
The peptide sequences of rat C/EBP␣ from amino acids 55 to 85 and mouse C/EBP␤ between 54 and 94 are shown. The dark lines above the amino acid sequences indicate regions of high homology. The underlined alanines (A) are mutated from the wild-type sequence. These mutations were introduced into the Gal4-C/EBP␤ vectors.

TABLE I
The C/EBP binding site in the PEPCK promoter enhances the induction of transcription by thyroid hormone HepG2 cells were transiently transfected with the luciferase reporter genes. Each transfection contained 3 g of luciferase gene, 1 g of RSV-TR␤ or Gal4-TR␤, 0.5 g of the dominant negative C/EBP vector (A-C/EBP) and 0.5 g of TK-Renilla. Transfected cells were exposed to 100 nM T 3 for 16 h. The data are expressed as luciferase activity corrected for protein content and transfection efficiency. Given that CBP increased the T 3 responsiveness of the PEPCK gene, we examined whether we could observe physical interactions between CBP and C/EBP␣. To conduct these studies, we initially used a mammalian two-hybrid assay. Overexpression of Gal4-CBP strongly potentiated basal expression of the Gal4X4 Ϫ68PLuc reporter gene. Cotransfection with C/EBP␣-(6 -217)-VP16 increased the expression of the luciferase reporter gene 9.1 Ϯ 4.4-fold (Fig. 7). Our results indicated that CBP could interact with C/EBP␣. These experiments were repeated eight times as there was considerable variability in the extent of the induction although expression of the reporter gene was consistently induced by the C/EBP␣-VP16 vector. These results indicate that CBP can interact with C/EBP␣. However, we were not able to demonstrate interactions between various GST-CBP proteins and [ 35 S]C/EBP␣ in pulldown experiments. Our data suggest that the interactions between C/EBP and CBP are not strong and that CBP may need to interact with several factors for stable interaction with the PEPCK promoter. Because C/EBP proteins are involved in the cAMP induction of PEPCK transcription, we tested whether CBP could enhance the induction of PEPCK gene expression by overexpression of the catalytic subunit of PKA. Overexpression of CBP did not enhance the induction by PKA suggesting that CBP alone would not increase the cAMP response (data not shown).
Our final experiments examined whether the TR, C/EBP, and CBP were associated with the PEPCK promoter in vivo. To test this question, we utilized the chromatin immunoprecipitation (ChIP) assay. Freshly prepared hepatocytes were treated briefly with 1% formaldehyde to cross-link the proteins to DNA. Ϫ490 PEPCK-CAT vector was cotransfected with RSV-TR␤ and 3 g of mammalian expression vectors for either SRC-1 or CBP. The cotransfected expression vector is indicated on the left side. Following transfection, the cells were exposed to 100 nM T 3 for 40 h. The data are expressed as CAT activity corrected for protein and transfection efficiency. All experiments were repeated at least four times in duplicate.

TABLE II
Effect of CBP on the T3 induction of the PEPCK gene HepG2 cells were transfected with 3 g of DR4X1 Gal4X1-68PLuc, 1 g RSV-TR␤ and 100 ng of the Gal4 expression vector. Cells were exposed to 100 nM T 3 for 24 h. All transfections were conducted at least four times in duplicate. The data are expressed as luciferase activity corrected for protein content and transfection efficiency. Ϫ68PLuc was cotransfected with 10 ng of Gal4-CBP and 500 ng of C/EBP␣-VP16 vectors as described previously. The data are expressed as luciferase activity corrected for protein and transfection efficiency. All transfections were conducted in NIH-3T3 cells and were repeated eight times in duplicate.
The cross-linked proteins and DNA were immunoprecipitated with antibodies to either TR␤, C/EBP␣, C/EBP␤, or CBP. PCR primers were created that amplified regions around the TRE and the P3(I) site in the PEPCK promoter. Our experiments using immunoprecipitated DNA as a template for PCR reactions show that the TR␤ is associated with the PEPCK promoter (Fig. 8). In addition, antibodies to C/EBP␣, C/EBP␤, and CBP immunoprecipitated the PEPCK promoter. Given that the sheared chromatin DNA was up to 500 bp in length, these results do not demonstrate the specific binding of these proteins to any element in the gene. The PCR products may represent proteins bound to the CRE, and C/EBP proteins bind to the P3(II), P4(I), and P4(II) sites in the PEPCK promoter with high affinity. In addition, we found that C/EBP proteins and CBP were also associated with the PEPCK promoter in H4IIE rat hepatoma cells (data not shown). These data indicate that these proteins are associated with the PEPCK gene in vivo.

DISCUSSION
The PEPCK gene has been studied extensively as a model for the multihormonal regulation of gene expression (1,2,9). PEPCK transcription is regulated through the interactions of nuclear hormone receptors and accessory factors bound to the promoter. The involvement of multiple accessory factors allows for subtle modulation of PEPCK gene expression in response to dietary, developmental, and hormonal alterations. Our studies have focused on the role of C/EBP proteins in the cAMP and T 3 induction of PEPCK gene expression. Here, we show that C/EBP proteins participate in T 3 and cAMP responsiveness and that the coactivator CBP can enhance T 3 stimulation of the PEPCK gene.
C/EBP proteins have been identified as accessory factors in the thyroid hormone and cAMP induction of several genes. Recently, it was found that C/EBPs participated in the T 3 induction of the malic enzyme gene (29). In addition, it was reported that C/EBP␤ was involved in the ability of cAMP to induce the StAR and prolactin genes (30,31). In the kidney, C/EBP␤ rather than CREB mediates the cAMP induction of the PEPCK gene (32). Experiments from the laboratory of Richard Hanson (12) utilizing C/EBP␣ knockout mice have shown that C/EBP␣ is required for the induction of the PEPCK gene by cAMP. A TRE has been identified in the promoter of the C/EBP␣ gene, and it has been reported that T 3 can stimulate C/EBP␣ transcription (33). These observations raise the possibility that T 3 stimulates PEPCK gene expression by increasing C/EBP␣ levels. We were not able to observe an increase in C/EBP␣ or C/EBP␤ protein abundance in the livers of hyperthyroid as opposed to euthyroid rats (data not shown). Our data suggest that C/EBP proteins are directly involved in the T 3 induction and that T 3 does not stimulate the PEPCK gene by increasing C/EBP abundance.
There have been several studies that have identified regions of homology between C/EBP␣ isoforms from different species (34 -36). A recent report from MacDougald and coworkers (37) outlined four conserved regions in the transactivation domain of C/EBP␣, which were called CR1, 2, 3, and 4. The CR2 domain of C/EBP␣, which contains amino acids 55-108, has several conserved amino acids between C/EBP␣ and C/EBP␤ (Fig. 3). Our data highlight several points regarding the relevance of this CR2 region within C/EBP proteins in the regulation of PEPCK gene transcription. The phenylalanine/leucine amino acids are important for the stimulation of basal expression by both C/EBP␤ and C/EBP␣ because mutation of these two amino acids in this region eliminates the ability of C/EBP␤ to stimulate basal expression (Fig. 4). It was reported that the FL amino acids in C/EBP␣ contacted the TATA-binding protein, TBP, and were essential for stimulating basal expression (34). Our data suggested that the amino acids 61 and 62 were important for the participation of C/EBP␤ in the cAMP stimulation of PEPCK gene expression (Fig. 5). Deletional analysis of the C/EBP␣ transactivation domain identified a short peptide stretch from amino acids 55 to 65 involved in mediating cAMP responsiveness (27). This conserved stretch of amino acids between C/EBP␣ and ␤ is involved in the cAMP induction of the PEPCK gene.
Previous reports have suggested that CBP might have a role in regulating PEPCK gene expression through its ability to interact with CREB, NF-1, and various steroid receptors (38). In addition, overexpression of E1A reduced the ability of the catalytic subunit of PKA to stimulate PEPCK gene expression (39). Finally, overexpression of CBP stimulated basal expression of the PEPCK gene (38). Our data support the concept that CBP has a role in the hormonal regulation of PEPCK gene expression. However, we did not observe that overexpression of CBP enhances either the basal expression or the PKA induction of our PEPCK-Luc vectors in transient transfections. Such an observation does not rule out a role for CBP in the cAMP induction despite the fact that we did not observe any synergism between CBP and PKA in transient transfections. For example, the overexpression of the coactivator SRCAP, which interacts with CBP, greatly enhanced the PKA induction of the PEPCK gene (40).
Our data does suggest a role for CBP in the T 3 induction of PEPCK transcription. Our data are compatible with some of the previous observations regarding the interactions of CBP/ p300 with C/EBP␣. MacDougald and coworkers (37) reported that overexpression of p300 enhanced the ability of C/EBP␣ to stimulate leptin gene expression. The ability of C/EBP␣ to synergize with p300 was mediated through all the conserved motifs of C/EBP␣. In keeping with that observation, our C/EBP␣-VP16 vectors were able to interact with Gal4-CBP in our mammalian two-hybrid assays. Previous reports indicated that C/EBP␣ and CBP did not interact in GST pull-down assays (37). We also believe as demonstrated by our mammalian two hybrid experiments that the interactions between C/EBP␣ and CBP are not likely to be strong. CBP may be recruited to the PEPCK promoter through its interaction with multiple proteins including CREB, NF-1, C/EBP␣ and ␤ as well as others. CBP and C/EBP␣ have been reported to colocalize in the nucleus using fluorescently tagged proteins (37). C/EBP␤ proteins have been shown to interact with p300 through the amino terminus of C/EBP␤ and the CH3 domain of p300 (41). Previous studies have demonstrated that multiple regions in the amino terminus of C/EBP␤ are involved in the interaction with CBP (41). This region of C/EBP␤ contains the CR2 region, which is FIG. 8. TR␤, C/EBP, and CBP are associated with the PEPCK promoter in vivo. ChIP assays were performed using formaldehyde cross-linked hepatocytes and antibodies to the C/EBP␣, C/EBP␤, TR␤, and CBP as indicated above. Immunoprecipitation using antibody to the hexahistidine tag (His tag) was used as a control for these experiments. The cross-linking was conducted on freshly isolated hepatocytes that were in suspension. The PCR products were resolved on a 3% Nusieve agarose gel. The sequence of the primers around the TRE and the conditions for the ChIP assay are described under "Materials and Methods." highly conserved between both C/EBP isoforms. It is likely that this domain will be important for the interaction of CBP and C/EBP isoforms.
Several models have been developed in which the various nuclear coactivators are assembled to form a functional complex. For example, SRC-1/NCoA-1 can be recruited by many liganded nuclear receptors, and CBP can interact with SRC-1 (19,20). SRC-1 has been shown to interact with several receptors that are part of the PEPCK glucocorticoid response unit including HNF-4, glucocorticoid receptor, and COUP-TF (42). When tethered to the PEPCK promoter, SRC-1 can substitute for other members of the glucocorticoid response unit suggesting that SRC-1 is involved in the glucocorticoid induction (42). However, we did not observe any enhancement of the T 3 induction by cotransfection of an expression vector for SRC-1. SRC-1 increased basal expression of our PEPCK reporter genes indicating that SRC-1 was interacting with the PEPCK promoter. It has been shown in primary hepatocytes that glucocorticoids are required for the full induction of PEPCK transcription by glucagon (43). It is possible that SRC-1 and CBP are involved in the synergistic activation of PEPCK gene transcription by glucocorticoids and cAMP. In summary, our data have defined limited domains of the C/EBP␣ and C/EBP␤ transactivation domains that are involved in the stimulation of PEPCK gene transcription by cAMP and T 3 . In addition, we have determined that CBP can enhance T 3 induction of the PEPCK gene.