Identification of CCAAT/enhancer-binding protein alpha as a transactivator of the mouse amelogenin gene.

Amelogenin expression is ameloblast-specific and developmentally regulated at the temporal and spatial levels. In a previous transgenic mouse analysis, the expression pattern of the endogenous amelogenin gene was recapitulated by a reporter gene driven by a 2. 2-kilobase mouse amelogenin proximal promoter. To understand the molecular mechanisms underlying the spatiotemporal expression of the amelogenin gene during odontogenesis, the mouse amelogenin promoter was systematically analyzed in mouse ameloblast-like LS8 cells. Deletion analysis identified a minimal promoter (-70/+52) containing a CCAAT/enhancer-binding protein (C/EBP)-binding site upstream of the TATA box. In transient transfection assays, C/EBPalpha up-regulated the promoter activity in a dose-dependent manner. The C/EBP-binding site was necessary for both C/EBPalpha-mediated transactivation and basal promoter activity. Electrophoresis mobility shift assays demonstrated that C/EBPalpha bound to its cognate site in the amelogenin promoter and that the binding was specific. Endogenous C/EBPalpha was detected in LS8 cells, and overexpression of exogenous C/EBPalpha in LS8 cells was able to increase the expression level of the endogenous amelogenin protein. The activity of the amelogenin promoter in rat parotid Pa-4 cells and Madin-Darby canine kidney cells was minimal, ranging from 20 to 30% of the activity in ameloblast-like cells. Transient transfection experiments showed that C/EBPalpha transactivated the mouse amelogenin reporter gene in Pa-4 cells, but not in Madin-Darby canine kidney cells. Taken together, these data indicate that C/EBPalpha is a bona fide transcriptional activator of the mouse amelogenin gene in a cell type-specific manner.

One unique characteristic of tooth development is the formation of mineralizing extracellular matrices. Enamel, the only epithelially derived calcified tissue in vertebrates, is synthesized by ameloblasts. Amelogenins are essential to the proper regulation of enamel mineralization. Amelogenin expression is ameloblast-specific and developmentally regulated at the temporal and spatial levels (1)(2)(3)(4)(5)(6)(7)(8). A 2263-nucleotide proximal promoter element from the mouse X-chromosomal amelogenin gene has been demonstrated by transgenic mouse analysis to direct the expression of a reporter gene in a temporal and spatial pattern that is essentially identical to that of the en-dogenous amelogenin gene (5).
During organogenesis, a programmed differentiation of embryonic epithelial cells is often characterized by the highly regulated expression of tissue-specific genes activated in response to inductive interactions with embryonic mesenchyme. The biochemical identities of the inductive signals that direct mammalian epithelial determination and differentiation have been elusive; however, studies of the regulated expression of tissue-specific gene products in developing epithelia may facilitate the molecular dissection of these interactions. The regulated expression of the ameloblast-specific amelogenin gene in the developing mouse tooth organ is an excellent model for studying developmentally regulated gene expression.
We hypothesize that the regulated transcription of amelogenin by specific activator(s) and repressor(s) in ameloblast cell lineage results in the spatiotemporal expression of amelogenin required for proper enamel formation. Gene expression is regulated at several levels, including activation of gene structure, transcription initiation, termination of transcription, nuclear RNA processing, mRNA translation, and mRNA stability. Unlike several other developmentally regulated genes (9), the methylation pattern of CpG islands in the promoter region is not associated with the regulated transcription of the amelogenin gene (10). Extensive homologies exist in the promoter regions of the bovine, human, and murine X-chromosomal amelogenin genes. There is a 70% identity within the 300-base pair region upstream of the transcription initiation site, suggesting that transcriptional regulation is likely to play an important role in the spatiotemporal expression of amelogenin. For a TATA box-containing promoter, the preinitiation complex is assembled in a highly regulated and defined order (11). The assembly of the preinitiation complex on a core promoter is sufficient to initiate transcription at a minimal level. However, the rate of transcription can be increased by an activator or turned off by a repressor.
The CCAAT/enhancer-binding proteins (C/EBPs) 1 are a family of related basic region leucine zipper transcription factors involved in the regulation of various aspects of cellular differentiation and function in multiple tissues. Six different members of the family (C/EBP␣, -␤, -␥, -␦, -⑀, and -) have been isolated and characterized. The expression of C/EBPs is tissueand stage-specific during development. C/EBPs have been shown to play a key role in regulating cellular differentiation, terminal function, and response to inflammatory insults (12)(13)(14)(15)(16).
To investigate the role of C/EBP␣ in the regulation of amelogenin gene expression, the 2.2-kilobase mouse amelogenin promoter has been systematically analyzed in mouse ameloblast-like LS8 cells. Our experimental strategy includes the following four analyses. First, the potential of C/EBP␣ to serve as a transcriptional activator of the amelogenin gene is tested in LS8, Pa-4, and MDCK cells using cotransfection assays. Second, the putative C/EBP-binding site on the amelogenin promoter is resolved using deletion and mutation analysis. Third, the specific binding to the putative site by C/EBP␣ is examined by electrophoresis mobility shift assay (EMSA). Fourth, the effect of C/EBP␣ overexpression on the endogenous amelogenin gene in LS8 cells is determined by Western blot analysis.
Cell Culture-A mouse ameloblast cell line (LS8) established by immortalizing primary cultures of enamel organ epithelium with SV40 large T antigen was maintained in Dulbecco's modified Eagle's medium (Sigma) supplemented with fetal bovine serum (10%), penicillin (100 units/ml), and streptomycin (100 g/ml). The rat parotid epithelial cell line Pa-4, also known as the parotid C5 cell line (17), was provided by Dr. D. Ann (University of Southern California) and maintained as described (18). The MDCK cells were provided by Dr. C. Okamoto (University of Southern California) and maintained in Eagle's minimal essential medium supplemented with fetal bovine serum (5%), penicillin (100 units/ml), and streptomycin (100 g/ml).
Transient Transfection and Luciferase Assay-Test DNA (750 ng) was used for transient transfection of 2 ϫ 10 5 cells/well in 12-well plates. To monitor transfection efficiency, 75 ng of pCMV-lacZ/well was cotransfected as an internal control. Two hours before the addition of the DNA-liposome complex, cells was washed twice with Hanks' solution and subsequently cultured in serum-and antibiotic-free medium. The plasmid DNA was mixed with 50 l of medium in a 5-ml VWR culture tube. Five microliters of LipofectAMINE PLUS reagent (Life Technologies, Inc.) was added, mixed, and incubated at room temperature for 15 min. Two microliters of LipofectAMINE Reagent (Life Technologies, Inc.) was diluted into 50 l of medium in a second tube. The contents of these two tubes were combined, mixed, and incubated at room temperature for 15 min. The DNA-PLUS-LipofectAMINE complex was diluted with 0.5 ml of medium, mixed, and layered gently on top of the cells. The cells were incubated at 37°C and 5% CO 2 for 3 h. After removal of the DNA-PLUS-LipofectAMINE complex, cells were incubated in 1 ml of complemented medium for an additional 22 h and then subjected to luciferase assay with a Dual-Light kit (Tropix Inc.) according to the manufacturer's recommendation. Briefly, cells were washed twice with Hanks' solution and lysed in 100 l of lysis buffer (100 mM potassium phosphate (pH 7.8), 0.2% Triton X-100). Cell lysates were transferred into a 1.5-ml microcentrifuge tube and cleared by centrifugation at 21,000 ϫ g for 2 min. Luciferase activity was measured by mixing 10 l of extract with 25 l of Buffer A and 100 l of Buffer B sequentially. After a 45-min incubation at room temperature, ␤-galactosidase activity was measured by adding 100 l of Buffer C in a luminometer (Lumat, Berthold).
Mutagenesis-Site-directed mutagenesis was performed using the GeneEditor TM in vitro site-directed mutagenesis system (Promega) according to the manufacturer's instructions. The mutagenesis primer was synthesized at the Microchemical Core Facility (University of Southern California/Norris Comprehensive Cancer Center). To ascertain the mutations, the plasmids were analyzed by restriction mapping and DNA sequencing.
Preparation of Nuclear Extracts-LS8 cells (80% confluent) in a 100-mm Falcon dish were washed twice with cold phosphate-buffered saline (pH 7.4) and scraped off in 1 ml of lysis buffer (20 mM Hepes (pH 7.6), 20 % glycerol, 1.5 mM MgCl 2 0.2 mM EDTA, 0.1% Triton X-100, 10 mM NaCl). Cell lysates were Dounce-homogenized for 10 strokes with a type A pestle on ice, transferred to a 15-ml Falcon tube, and centrifuged at 2000 rpm for 10 min at 4°C. The pelleted nuclei were resuspended at 1 ϫ 10 7 nuclei/ml in nuclear extraction buffer (20 mM Hepes (pH 7.6), 20% glycerol, 1.5 mM MgCl 2 , 0.2 mM EDTA, 0.1% Triton X-100, and 500 mM NaCl) with a final concentration of 0.5 M NaCl and mixed on a rotator at 4°C for 1 h. Nuclear debris was pelleted by centrifugation at 10,000 rpm for 10 min at 4°C. Aliquots were frozen in liquid N 2 and stored at Ϫ80°C. The protein concentration was determined using a Bio-Rad protein assay kit with bovine serum albumin standards.
EMSA-Double-stranded oligonucleotide probes were generated by annealing an antisense strand to a 10-fold excess of sense strand and filling in with [␣-32 P]dATP (NEN Life Science Products) and Klenow (exo Ϫ ). The GelShift buffer kit (Stratagene) was used for the binding reaction, which was performed as recommended by the manufacturer. The reaction mixtures were resolved on a 6% nondenaturing polyacrylamide gel provided in the kit. The gel was dried, and the bands were visualized by autoradiography. The sequences of the oligonucleotides were as follows: wild-type antisense strand, 5Ј-GAACAGC CAAT-CAGGT TTCTGAATGAA-3Ј; wild-type sense strand, 5Ј-TTTTTCAT-TCAGAAACCTG ATTGGCT GTTC-3Ј; mutant antisense strand, 5Ј-GAACAGCCAATCTCTAGA CTGAATGAA-3Ј; and mutant sense strand, 5Ј-TTTTTCATTCAGTCTAGA GATTGG CTGTTC-3Ј (mutated sites are boldface).
Western Blot Analysis-Whole cell lysates were prepared from untreated LS8 cells, LS8 cells transfected with pcDNA3 (Invitrogen), and LS8 cells transfected with pcrC/EBP␣ (a C/EBP␣ expression vector in pcDNA3), respectively. Protein concentrations were determined by the Bio-Rad protein assay with a bovine serum albumin standard curve. Equal amounts of protein (10 -20 g) were subject to 12% SDS-polyacrylamide gel electrophoresis. The resolved proteins on the gel were electroblotted onto Immobilon-P membrane (Millipore Corp.), and the membrane was incubated with a primary antibody (anti-C/EBP␣, anti-C/EBP␤, or anti-amelogenin). Horseradish peroxidase-conjugated secondary antibody and the enhanced chemiluminescence (ECL) detection system (Amersham Pharmacia Biotech) were used to detect the bound antibodies.

RESULTS
The Ϫ70/Ϫ52 Region Is Essential to Amelogenin Promoter Activity-To identify the cis-element(s) required for the promoter activity of the mouse amelogenin gene, a series of 5Јdeletion reporter constructs were tested in LS8 cells with transient transfection assays. Consistent with the transgenic analysis (5), the 2263-base pair region from the mouse amelogenin promoter in the reporter construct p2207 was sufficient to direct the expression of the reporter gene in LS8 cells, a mouse ameloblast-like cell line. Sequential deletion of the region between Ϫ2207 and Ϫ71 in the mouse amelogenin promoter gave rise to a modest increase (Ͻ2-fold) in promoter activity. However, further deletion of a 19-nucleotide stretch (Ϫ70 to Ϫ52) resulted in a nearly complete ablation of reporter gene activity in LS8 cells, with p51 exhibiting a background level of activity similar to the promoterless construct pGL3-Basic ( Fig. 1). Therefore, the Ϫ70/ϩ51 region of the mouse amelogenin promoter included in the p70 construct functioned as a minimal promoter in LS8 cells, whereas the Ϫ70/Ϫ52 region was requisite to promoter activity.
In ameloblast-like LS8 cells, the Ϫ2207/Ϫ71 region appeared to have a limited effect on the promoter activity of the mouse amelogenin gene. To determine whether a cis-element(s) responsible for tissue specificity of the amelogenin promoter is located in the Ϫ2207/Ϫ52 region, the same set of reporter constructs was transiently transfected into non-ameloblast cells. Rat parotid Pa-4 and MDCK cells were used since they represent terminal differentiated oral and non-oral epithelial cells, respectively. The reporter gene activity of p2207 in Pa-4 and MDCK cells was 3-5-fold above that of the promoterless construct pGL3-Basic, in comparison with 15-fold observed in LS8 cells. Progressive deletion to Ϫ70 led only to a modest increase in reporter gene activity: Ͻ2-fold in Pa-4 cells and 2-3-fold in MDCK cells. However, the reporter gene activities of p51 were comparable in Pa-4, MDCK, and LS8 cells (Fig. 1). The transfection studies confirmed that the mouse amelogenin promoter is ameloblast-specific as demonstrated in a previous transgenic animal analysis (5). Given the fact that the corresponding amelogenin reporter constructs (p2207 versus p70) exhibited comparable promoter activities in Pa-4 and MDCK cells that were consistently between 20 and 30% of the activity in LS8 cells, the Ϫ2207/Ϫ71 region appeared to contribute little to the tissue specificity of the mouse amelogenin promoter. However, similar to that in LS8 cells, the Ϫ70/Ϫ52 region was necessary for promoter activity in both Pa-4 and MDCK cells.
C/EBP␣ Is a Transcriptional Activator of the Mouse Amelogenin Promoter-A putative C/EBP transcription factorbinding site was found in the Ϫ70/Ϫ52 region of the mouse amelogenin promoter. To test whether C/EBP␣ could function as a transcriptional activator of the mouse amelogenin promoter, a C/EBP␣ expression vector was cotransfected into LS8 cells with either the amelogenin promoter-reporter construct p2207 or p70. In response to increasing amounts of C/EBP␣, the reporter gene activity was increased up to 11-fold for p2207 ( Fig. 2A) and up to 27-fold for p70 (Fig. 2B). These data indicate that C/EBP␣ transactivates the mouse amelogenin promoter in a dose-dependent manner in LS8 cells.
To determine whether C/EBP␣ alone is sufficient to activate the mouse amelogenin promoter in non-ameloblast cells, similar cotransfection studies were performed in Pa-4 and MDCK cells. Cotransfection of C/EBP␣ had little effect on the promoter activity of both p2207 and p70 in MDCK cells. However, in Pa-4 cells, the reporter gene activity was increased up to 4-fold for p2207 ( Fig. 2A) and up to 6-fold for p70 (Fig. 2B) in response to increasing amounts of C/EBP␣. In the absence of exogenous C/EBP␣, the reporter gene activity of p2207 in Pa-4 cells was 30% of the basal level in LS8 cells; cotransfection of C/EBP␣ was able to increase the level to 140%. A similar result was obtained for p70. In the absence of exogenous C/EBP␣, the reporter gene activity of p70 in Pa-4 cells was 24% of the basal level in LS8 cells; cotransfection of C/EBP␣ was able to increase the level to 140%. Taken together, these data indicate that C/EBP␣ functions as a transcriptional activator of the mouse amelogenin gene in a cell type-specific manner.
The Putative C/EBP-binding Site Is Necessary for Amelogenin Promoter Activity-To understand the function of the putative C/EBP-binding site, a mutation was introduced into the core sequence of the C/EBP site in the context of p2207 (p2207mut) and p70 (p70mut) (Fig. 3A); the responsiveness to C/EBP␣ was then determined in LS8 cells with transient cotransfection assays. Mutation or deletion of the C/EBP site consistently abolished the basal promoter activity of these reporter constructs (p2207mut, p70mut, and p51). Furthermore, C/EBP␣-mediated transactivation of these constructs was reduced to the background level, similar to that of promoterless pGL3-Basic, whereas the reporter gene activity of the wild-type constructs was increased an order of magnitude by cotransfection with C/EBP␣ (Fig. 3B). Taken together, these data indicate that the putative C/EBP-binding site in the Ϫ70/Ϫ52 region of the mouse amelogenin promoter is required not only for C/EBP␣-mediated transactivation, but also for basal promoter activity in LS8 cells. Interestingly, the putative C/EBP-binding site identified in the Ϫ70/Ϫ52 region of the mouse amelogenin promoter is conserved between species, as shown in the alignment of murine, bovine, and human X-chromosomal amelogenin promoter nucleotide sequences (Fig. 3C).
C/EBP␣ Binds to the Mouse Amelogenin Promoter-The results described above provide functional evidence of a role for C/EBP␣ as a positive regulator of mouse amelogenin gene expression. To determine whether C/EBP␣ can bind to the mouse amelogenin promoter, a gel mobility shift assay (EMSA) was performed using the double-stranded mouse amelogenin C/EBP oligonucleotide, 5Ј-TTTTTC ATTCAGAAACCTGATT-GGCTGTTC-3Ј (C/EBP-binding site is in boldface). A protein-DNA complex was formed using nuclear extract prepared from LS8 cells as evidenced by the shifted band (Fig. 4A, lane 2). The intensity of the band was increased when nuclear extract prepared from C/EBP␣-transfected LS8 cells was used (Fig. 4A,  lane 3). In addition, an antibody specific to C/EBP␣ was able to supershift the protein-DNA complex (Fig. 4A, lane 4), whereas the antibody alone did not bind to the probe (lane 5). The addition of a 10-, 50-, or 100-fold molar excess of unlabeled C/EBP oligonucleotide inhibited the binding of C/EBP␣ to the labeled probe (Fig. 4B, lanes 3-5). However, no inhibition was observed (Fig. 4B, lanes 6 -8) with a molar excess of an oligonucleotide encoding a mutated C/EBP-binding site, 5Ј-TTTTTC ATTCAGtctaga GATTGGCTGTTC-3Ј (mutated site is in lowercase letters). These data demonstrate that C/EBP␣ is able to bind to the mouse amelogenin promoter and that the binding is specific.
As shown in Fig. 4D, the C/EBP consensus sequence within the mouse amelogenin promoter does not include a CCAAT box. Furthermore, a 32 P-labeled double-stranded oligonucleotide containing a mutated C/EBP-binding site (Fig. 4D, MUT) was unable to form a detectable protein-DNA complex, even though the CCAAT box remained intact (Fig. 4C, lanes 7-10). In contrast, the wild-type probe (Fig. 4D, WT) formed a protein-DNA complex in a dose-dependent manner (Fig. 4C, lanes 2-5). These data indicate that the C/EBP-binding site, independent of the CCAAT box, is required for the binding of C/EBP␣ to the mouse amelogenin promoter.
C/EBP␣ Up-regulates the Expression of the Amelogenin Protein in LS8 Cells-To ascertain that C/EBP␣ is a bona fide positive regulator of mouse amelogenin gene expression, two additional questions were asked. First, what is the expression status of C/EBP␣ in ameloblasts? Second, is C/EBP␣ able to regulate the expression of the endogenous amelogenin gene in the context of chromatin? The expression status of C/EBP␣ was analyzed in an ameloblast-like cell line (LS8) by Western blotting. An antibody specific to C/EBP␣ recognized a 42-kDa protein, which could be specifically competed off with a blocking peptide to the antibody (Fig. 5A), indicating that authentic C/EBP␣ is expressed in LS8 cells. Furthermore, the overexpression of exogenous C/EBP␣ in LS8 cells was achieved by transiently transfecting 2 g of C/EBP␣ expression vector. Exogenous C/EBP␣ was readily detected, whereas the endogenous protein required a 40 times longer exposure for detection (Fig. 5B, C/EBP alpha panel, compare lane 3 with lanes 1 and  2, respectively). The expression level of the endogenous amelogenin protein in LS8 cells was then determined. A Ͼ2fold increase in the amelogenin protein level was observed in LS8 cells overexpressing C/EBP␣ (Fig. 5B, histogram, third bar versus first bar), whereas the level of amelogenin expression in empty vector-transfected LS8 cells remained essentially the same as that in untransfected control LS8 cells (Fig. 5B, histogram, second bar versus first bar). With a typical 20 -30% transfection efficiency in our transient transfection experiments, the observed 2.5-fold overall increase (Fig. 5B, histogram, third bar) in the amelogenin protein level upon C/EBP␣ activation actually represented a 6 -9-fold increase in the cells that overexpressed exogenous C/EBP␣. Therefore, the existence of C/EBP␣ and its capability of activating the endogenous amelogenin gene in the ameloblast-like LS8 cells clearly indi-cate that C/EBP␣ is a genuine regulator of amelogenin gene expression.
C/EBP␤ Has a Marginal Effect on Amelogenin Gene Expression-As a first step to determine whether other C/EBP family members play a role in mouse amelogenin gene expression, C/EBP␤ was studied. The expression status of C/EBP␤ was analyzed in LS8 cells by Western blotting. An antibody specific to C/EBP␤ recognized a 32-kDa protein, which could be specifically competed off with a blocking peptide to the antibody (Fig.  6A), indicating that C/EBP␤ is expressed in LS8 cells. To test whether C/EBP␤ can function as a transcriptional activator of the mouse amelogenin promoter, a C/EBP␤ expression vector was cotransfected into LS8 cells with either amelogenin promoter-reporter construct p2207 or p70. C/EBP␤ was much less potent than C/EBP␣ in activating amelogenin reporter gene activity: 2-3-fold for C/EBP␤ versus 8 -15-fold for C/EBP␣ (Fig.  6B). The transfection data indicate that the capability of C/EBP␤ in transactivating the mouse amelogenin gene is very limited. To further determine whether endogenous C/EBP␤ in LS8 cells can bind to the amelogenin promoter, a gel mobility shift analysis was performed using the wild-type oligonucleotides (Fig. 4D) as a labeled probe. A protein-DNA complex was formed using nuclear extract prepared from LS8 cells as evidenced by the shifted band (Fig. 6C, lane 2). However, an antibody specific to C/EBP␤ was not able either to supershift or to disrupt the protein-DNA complex (Fig. 6C, lane 3), and the antibody alone was not able to bind to the probe (lane 4). This EMSA data argued that there was no detectable level of C/EBP␤ present in the protein-DNA complex. Taken together, the results indicate that C/EBP␤ has only a marginal effect, if any, on the transcriptional regulation of the mouse amelogenin gene.

DISCUSSION
The Mouse Amelogenin Promoter Proximal Regulatory Region-The proximal regulatory region of the mouse amelogenin promoter is the region immediately upstream of the TATA box. This region includes nucleotides Ϫ70 to Ϫ33 together with the TATA box and drives ϳ190% of the promoter activity in ameloblast-like LS8 cells relative to the 2.2-kilobase promoter tested in transgenic animals (5). The slight increase in promoter activity suggests that a silencer element(s) may be located in the region between Ϫ2207 and Ϫ70, which is consistent with a previous report in which an upstream fragment from the bovine amelogenin promoter decreased the activity of a heterol- ogous hamster sarcoma virus-thymidine kinase minimal promoter by ϳ50% in Chinese hamster ovary cells (19). In the mouse amelogenin promoter, the Ϫ70/Ϫ52 region contains the binding site for C/EBP. Deletion of the C/EBP-binding site results in the loss of promoter activity. Taking advantage of the ameloblast-like LS8 cells, we have identified the mouse amelogenin minimal promoter, consisting of the Ϫ70/ϩ51 region, in which the C/EBP-binding site-containing Ϫ70/Ϫ52 region is essential to promoter activity. Interestingly, this 19nucleotide stretch is highly conserved (15 out of 19 nucleotides) in the promoter regions of human, bovine, and murine X-chromosomal amelogenin genes (Fig. 3C). Most importantly, the core sequence (5Ј-GAAA-3Ј) for the C/EBP-binding site is identical in all three, suggesting that this region plays a basic regulatory role in amelogenesis.

C/EBP␣ Is an Important Regulator of Mouse Amelogenin
Gene Expression-We tested the hypothesis that C/EBP␣ is involved in the regulation of mouse amelogenin gene expression. Amelogenin expression is ameloblast-specific and developmentally regulated at the temporal and spatial levels. Nucleic acid hybridization experiments (1)(2)(3) demonstrated that the transcription of amelogenin is restricted to inner enamel epithelial cells that undergo terminal differentiation to the ameloblast phenotype, which is characterized by withdrawal from the cell cycle, polarization, and columnar morphology. Furthermore, the same promoter region as that in the p2207 construct is able to recapitulate the spatiotemporal expression pattern of the endogenous amelogenin gene in transgenic mouse analyses (5). In the present study, the promoter of the mouse amelogenin gene was systematically analyzed in ameloblast-like LS8 cells, rat parotid Pa-4 cells, and MDCK cells. The minimal promoter was identified, in which a 19-nucleotide stretch containing the C/EBP transcription factor-binding site was required for basal promoter activity. C/EBPs are composed of a family of basic region leucine zipper domain-containing transcription factors that are critical regulators of cell differentiation (12)(13)(14)(15)(16). C/EBP␣ has been demonstrated to mediate cell cycle arrest; cellular differentiation; and transcriptional regulation of tissue-specific genes in adipocytes, hepatocytes, keratinocytes, pneumocytes, and ovarian follicles (20 -34). In liver and adipose tissue, peak levels of C/EBP␣ mRNA are detected only in differentiated tissues (35,36). C/EBP␣ functions as a transcriptional activator in adipocytes, and the accumulation of C/EBP␣ late in preadipocyte differentiation is correlated with the expression of differentiation markers (37)(38)(39). Here, we demonstrate that C/EBP␣ functions as a transcriptional activator of the mouse amelogenin gene by the following evidence. 1) The overexpression of C/EBP␣ in mouse ameloblast-like LS8 cells can up-regulate the promoter activity of the mouse amelogenin gene in a dose-dependent manner. 2) The ability of C/EBP␣ to transactivate the mouse amelogenin gene is cell type-specific.
3) The C/EBPbinding site located in the Ϫ70/Ϫ52 region in the mouse amelogenin promoter is necessary for both C/EBP␣-mediated transactivation and basal promoter activity in LS8 cells. 4) C/EBP␣ is capable of specifically binding to the C/EBP site in the promoter. 5) C/EBP␣ is expressed in ameloblast-like LS8 cells, and overexpression of C/EBP␣ can increase the expression level of the endogenous amelogenin protein.
Consistent with the notion that a repressive element(s) is located in the Ϫ2207/Ϫ70 region of the mouse amelogenin promoter, the transactivation of the p2207 reporter construct by C/EBP␣ is less pronounced than that of p70. This finding suggests that the putative repressor(s) can interfere with the transactivating function of C/EBP␣, probably through proteinprotein interactions. The interference may be due to direct interactions with C/EBP␣ and/or interruption of the interactions between C/EBP␣ and basal transcriptional machinery. The mouse amelogenin promoter is cell type-specific. In nonameloblast cell lines, all reporter constructs tested so far have very low promoter activity, and deletion of the Ϫ2207/Ϫ71 region leads only to a modest increase in reporter gene activity. There are several possible explanations for this observation. 1) The cis-element(s) conferring tissue specificity is located in the Ϫ70/Ϫ52 region. 2) The Ϫ2207/Ϫ71 region lacks the cis-element(s) that binds to tissue-specific repressor(s). 3) The Ϫ2207/ Ϫ71 region contains a tissue-specific silencer(s); however, deletion of this region is not sufficient to increase promoter activity due to the lack of certain transcriptional activator(s) in non-ameloblast cells. Interestingly, the ability of C/EBP␣ to transactivate the mouse amelogenin gene is cell type-dependent. The presence of repressor(s) and/or the absence of coacti-

FIG. 3. Requirement of the C/EBP-binding site in the amelogenin promoter for C/EBP␣-mediated transactivation and basal promoter activity.
A, schematic representation of the proximal regulatory region of the amelogenin promoter in the reporter construct. The transcription initiation site is indicated by the arrow and designated as position ϩ1. The sequence of the Ϫ70/Ϫ33 region is shown, in which the 5Ј-boundary of the promoter in the p51 construct is indicated by Ϫ51. A C/EBP-binding site is located in the region between Ϫ70 and Ϫ51. The mutated nucleotides in the core sequence of the C/EBP-binding site are in lowercase. B, results of transfection experiments. The reporters p2207mut and p70mut are constructs with a mutated C/EBP-binding site in the context of p2207 and p70, respectively. Equal amounts of various reporter constructs were transiently transfected into LS8 cells with 250 ng of C/EBP␣ expression plasmid or empty vector pcDNA3. pCMV-lacZ was used as an internal control for transfection efficiency. The relative luciferase activity is the normalization of luciferase activity with ␤-galactosidase activity. The mean Ϯ S.D. from at least three independent experiments is represented. C, sequence alignment of the amelogenin promoter. The Ϫ70/Ϫ51 region and flanking sequence of the murine X-chromosomal amelogenin promoter is aligned with the corresponding regions of the bovine and human X-chromosomal amelogenin promoters. Dashes indicate identical nucleotides in all three genes. The C/EBP consensus sequence is identified in the alignment in boldface. WT, wild-type; MUT, mutated. vator(s) may account for the inability of C/EBP␣ to transactivate the mouse amelogenin promoter in MDCK cells.
The formation of a protein-DNA complex detected in LS8 nuclear extract (Fig. 4A, lane 2) most likely resulted from endogenous C/EBP␣ in LS8 cells. This notion is supported by the observation that exogenous C/EBP␣ was able to increase the intensity of the complex (Fig. 4A, lane 3). The protein-DNA complex in EMSA could not be completely supershifted by a C/EBP␣-specific antibody (Fig. 4A), which might be due to the following reasons. First, the epitope recognized by the antibody transferred to nitrocellulose; and immunoblotted with either an amelogenin-specific antibody (Amelogenin panel) or a C/EBP␣-specific antibody (C/EBP alpha panel). For better visualization of overexpressed exogenous C/EBP␣ in LS8 cells, the exposure time with the ECL system was optimized in the following manner. After immunoblotting with a C/EBP␣-specific antibody (C/EBP alpha panel), the membrane was cut in half. Exogenous C/EBP␣ in lane 3 was detected with a 15-s exposure, whereas the endogenous C/EBP␣ proteins in lanes 1 and 2 were detected with a 600-s exposure, a 40-fold difference. The expression level of the endogenous amelogenin protein was quantitated by densitometry, normalized with the loading control, and plotted. is partially blocked once C/EBP␣ binds to DNA as a dimer; therefore, the binding of antibody to the C/EBP␣ dimer is less efficient than that to the monomer, as seen in Western blot analysis. Second, the interaction between C/EBP␣ and its antibody is masked by other proteins in the EMSA complex such as transcription coactivators.
Although C/EBP␣ was originally identified as a CCAAT-box binding protein, the compiled consensus sequence for C/EBP ((C/T)TNNNGNAA(C/T)) clearly indicates that the CCAAT box is not an essential part of a given C/EBP-binding site. In the case of the mouse amelogenin promoter, a CCAAT box is located in the Ϫ58/Ϫ55 region on the sense strand, 4 nucleotides downstream of the core sequence of the C/EBP site (Fig. 4D). It is possible that proteins other than C/EBP␣ are also capable of binding to the 30-mer oligonucleotide probe used in EMSA. One of the candidates is NFY/CBF, which is a ubiquitous transcription factor composed of three subunits, A, B, and C (40). The conserved histone fold motif in both NFY-B and NFY-C subunits forms a dimer ("histone fold") and interacts with NFY-A (41). The CCAAT box motif specifically recognized by NFY/CBF is found in the promoter and enhancer regions of many eukaryotic genes (42). Polymerase chain reaction-mediated random binding site selection has been used (43) to establish the consensus sequence for NFY/CBF binding as TG ATTGG (T/C)(T/ C)(A/G). Changes in 1 or 2 nucleotides in the flanking sequences of the binding core (ATTGG) have been shown to only modestly decrease NFY/CBF binding (43).
The 32 P-labeled oligonucleotide used in our EMSA contains both the C/EBP site and the NFY/CBF site. However, two lines of evidence argue against the shifted band detected in our EMSA as being composed of a mixture of C/EBP␣-probe and NFY-probe complexes. First, no protein-DNA complex was detected in gel shift assay when we used a 32 P-labeled oligonucleotide probe in which the C/EBP site was mutated, although this mutation resulted in only the 5Ј-most nucleotide being changed in the NFY/CBF consensus binding site (Fig. 4, C and  D). Second, it is reasonable to assume that the binding of C/EBP␣ and NFY/CBF to the amelogenin promoter is mutually exclusive due to the proximity of their binding sites (4 base pairs apart). Given the fact that NFY/CBF is a ubiquitous transcription factor, it is unlikely that NFY/CBF plays a significant role in the spatiotemporal expression of the amelogenin gene during tooth development. The CCAAT box may be necessary for amelogenin promoter activity; however, in the absence of a functional C/EBP site, the CCAAT box alone is not sufficient to support reporter gene transcription in LS8 cells, as evidenced by the inactive mutated constructs p2207mut and p70mut (Fig. 3B).
A C/EBP␤-specific antibody is incapable of either supershifting or disrupting the protein-DNA complex (Fig. 6C). There are two possibilities. First, there is no C/EBP␤ protein in the complex. Second, the antibody fails to recognize C/EBP␤ protein in the complex. Given the fact that C/EBP␤ has a very modest transactivational effect on amelogenin reporter genes (Fig. 6B), C/EBP␤ is unlikely to play a direct role in mouse amelogenin gene expression.
Our results demonstrate that the C/EBP-binding site in the proximal region of the mouse amelogenin promoter is necessary not only for C/EBP␣-mediated transactivation, but also for basal promoter activity in ameloblast-like cells, and that overexpression of C/EBP␣ is sufficient to increase the protein level of the endogenous amelogenin gene. These findings suggest that C/EBP␣ plays a key role in the developmentally regulated expression of the amelogenin gene. Amelogenin, the major organic component of enamel matrix, has been demonstrated to play an important role in proper enamel mineralization. Amelogenins consist of ϳ90% of the enamel matrix proteins. Several mutations in the human X-chromosomal amelogenin gene have been identified in patients with the inherited enamel defect X-linked amelogenesis imperfecta (44 -46). Disruption of FIG. 6. Effect of C/EBP␤ on mouse amelogenin gene expression. A, equal amounts of LS8 cell lysate were electrophoresed, transferred to nitrocellulose, and immunoblotted with a C/EBP␤-specific antibody (sc-150, Santa Cruz Biotechnology) without (Ϫ) or with (ϩ) preincubation with blocking peptides (sc-150p, Santa Cruz Biotechnology). The arrow indicates C/EBP␤. The numbers indicate the molecular masses of marker proteins (in kilodaltons). B, shown are the results of transient cotransfection experiments. Equal amounts of reporter construct p2207 or p70 were transiently transfected into LS8 cells with 250 ng of empty vector pcDNA3 (vector bars) or C/EBP␣ (C/EBP alpha bars) or C/EBP␤ (C/EBP beta bars) expression plasmid. pCMV-lacZ was used as an internal control for transfection efficiency. The relative luciferase activity is the normalization of luciferase activity with ␤-galactosidase activity. The mean Ϯ S.D. from at least three independent experiments is represented, and the levels of p2207 and p70 in the absence of exogenous C/EBPs were set as 1. C, shown are the results from EMSA of the C/EBP-binding site. The EMSA study was performed essentially as described in the legend to Fig. 4. Only nuclear extracts prepared from untransfected LS8 cells (NE) were used. For supershift analysis, the extracts were preincubated with a C/EBP␤-specific antibody (Ab; sc-150, Santa Cruz Biological). Complexes were separated by electrophoresis and visualized by autoradiography. amelogenin synthesis during tooth development with either antisense oligonucleotides or ribozymes results in disorganized enamel (47,48). Understanding the developmentally regulated expression pattern of amelogenin, the major structural gene during tooth development, will facilitate the elucidation of the molecular events involved in enamel formation. C/EBP␣ is the first transcription factor identified to date to function as a positive regulator of amelogenin expression. Further delineating the relationship between C/EBP␣ and amelogenin expression will provide insights into the understanding of the mechanism underlying the strictly regulated, tissue-specific expression of the amelogenin gene. In the most simplified scenario, the expression profiles of C/EBP␣ and amelogenin are correlated with each other during ameloblast differentiation. However, it is more likely that C/EBP␣ works in concert with transcription repressors to regulate the spatiotemporal expression of the amelogenin gene during tooth development. Both derepression and activation are probably required for the observed stringent regulation of the amelogenin gene. Transcription repressors can function through their cognate sites on the amelogenin promoter and/or through disrupting the transactivational effect of C/EBP␣. The identification of C/EBP␣ as a potent transcriptional activator of amelogenin not only provides the first direct evidence that positive regulators are involved in amelogenin gene expression, but also facilitates the study of potential repressors acting through C/EBP␣.