Title: Cloning and Characterization of Rat Dentin Matrix Protein 1 (DMP1) Gene and Its 5′-Upstream Region
Abstract: Rat dentin matrix protein 1 (DMP1) is a highly acidic 58-kDa phosphoprotein, and DMP1 was the first gene to be cloned from the mineralized dentin matrix. It exists as a highly phosphorylated protein with a pI of 3 in the dentin matrix and, in that state, might have an important role in the mineralization process. The spatio-temporal distribution during development indicates that the expression of this gene is tightly regulated in the odontoblasts. It is now known that DMP1 is not unique to dentin but is present in other mineralized tissues like long bone, calvaria, and ameloblasts. To study the transcriptional regulation and the function of DMP1 in these tissues, a genomic clone with a functional promoter, introns, and exons was isolated. Sequence analysis showed that the ratDMP1 gene is comprised of six exons and five introns and spans ∼13 kilobases (kb). Exon 1 contains the 5′-untranslated sequences. Exon 2 encodes a total of 18 amino acids including the 16 amino acids of the signal sequence. Exons 3–5 encode 16, 11, and 15 amino acids, respectively. Exon 6 contains 1.3 kb of the coding sequence with the RGD domain, stop codon, and the 3′-untranslated region (1.1 kb). We have mapped two transcription start sites within the DMP1 promoter that are 280 and 321 base pairs, respectively, from the ATG start codon. The location of functional elements within the 5′-upstream DMP1 DNA fragment was determined by cloning it into a luciferase reporter gene. Transient transfection and luciferase assays revealed that the 3 kb fragment has the ability to drive the luciferase gene. However, this promoter activity was restricted to MC3T3-E1 cells (an osteoblast cell lineage). The promoter was silent in Chinese hamster ovary cells (an epithelial cell lineage), indicating the necessity of tissue-specific factors to drive the transcription. Rat dentin matrix protein 1 (DMP1) is a highly acidic 58-kDa phosphoprotein, and DMP1 was the first gene to be cloned from the mineralized dentin matrix. It exists as a highly phosphorylated protein with a pI of 3 in the dentin matrix and, in that state, might have an important role in the mineralization process. The spatio-temporal distribution during development indicates that the expression of this gene is tightly regulated in the odontoblasts. It is now known that DMP1 is not unique to dentin but is present in other mineralized tissues like long bone, calvaria, and ameloblasts. To study the transcriptional regulation and the function of DMP1 in these tissues, a genomic clone with a functional promoter, introns, and exons was isolated. Sequence analysis showed that the ratDMP1 gene is comprised of six exons and five introns and spans ∼13 kilobases (kb). Exon 1 contains the 5′-untranslated sequences. Exon 2 encodes a total of 18 amino acids including the 16 amino acids of the signal sequence. Exons 3–5 encode 16, 11, and 15 amino acids, respectively. Exon 6 contains 1.3 kb of the coding sequence with the RGD domain, stop codon, and the 3′-untranslated region (1.1 kb). We have mapped two transcription start sites within the DMP1 promoter that are 280 and 321 base pairs, respectively, from the ATG start codon. The location of functional elements within the 5′-upstream DMP1 DNA fragment was determined by cloning it into a luciferase reporter gene. Transient transfection and luciferase assays revealed that the 3 kb fragment has the ability to drive the luciferase gene. However, this promoter activity was restricted to MC3T3-E1 cells (an osteoblast cell lineage). The promoter was silent in Chinese hamster ovary cells (an epithelial cell lineage), indicating the necessity of tissue-specific factors to drive the transcription. rat dentin matrix protein 1 kilobase(s) base pair(s) Chinese hamster ovary polymerase chain reaction Tooth formation is regulated by temporally and spatially restricted reciprocal interactions between the epithelium and the mesenchyme, and this signal traverses back and forth until tooth development is complete (1.Thesleff I. Nieminen P. Curr. Opin. Cell Biol. 1996; 8: 844-850Crossref PubMed Scopus (173) Google Scholar, 2.Mina M. Kollar E.J. Arch. Oral Biol. 1987; 32: 123-127Crossref PubMed Scopus (408) Google Scholar, 3.Peters H. Balling R. Trends Genet. 1999; 15: 59-64Abstract Full Text Full Text PDF PubMed Scopus (294) Google Scholar). One type of cells resulting from this process is the odontoblasts, which are the principle cells that synthesize the dentin matrix. Mature odontoblasts are responsible not only for the secretion of the collagen-rich extracellular matrix but also for the secretion of noncollagenous proteins that are responsible for initiating the mineralization cascade. In dentinogenesis, a number of highly controlled extracellular and intracellular events are responsible for the formation of the well-defined mineralized tissue. The noncollagenous proteins produced by the single layer of odontoblasts may control these events (4.Butler W.T. Eur. J. Oral Sci. 1998; 106: 204-210Crossref PubMed Scopus (183) Google Scholar, 5.Veis A. Butler W.T. Chemistry and Biology of Mineralized Tissues. EBSCO Media, Birmingham, AL1985: 170Google Scholar, 6.Linde A. Goldberg M. Crit. Rev. Oral Biol. Med. 1993; 4: 679-728Crossref PubMed Scopus (360) Google Scholar). One of the noncollagenous proteins is dentin matrix protein 1 (DMP1)1 (7.George A. Sabsay B. Simonian P.A.L. Veis A. J. Biol. Chem. 1993; 268: 12624-12630Abstract Full Text PDF PubMed Google Scholar, 8.MacDougall M. Gu T.T. Luan X. Simmons D. Chen J. J. Bone Mineral Res. 1998; 13: 422-431Crossref PubMed Scopus (132) Google Scholar). Based on our initial data, DMP1 was thought to be dentin-specific, but later, its expression was seen in calvaria and long bone (9.George A. Gui J. Jenkins N.A. Gilbert D.J. Copeland N.G. Veis A. J. Histochem. Cytochem. 1994; 42: 1527-1531Crossref PubMed Scopus (80) Google Scholar, 10.D'Souza R.N. Cavender A. Sunavala S.G. Alvarez J. Ohshima T. Kulkarni A.B. MacDougall M. J. Bone Miner. Res. 1997; 12: 2040-2149Crossref PubMed Scopus (292) Google Scholar, 11.Kulkarni G.V. Jee S.W. Srinivasan R. Marks Jr., S. Veis A. George A. Goldberg M. Robinson C. Boskey A.L. Proceedings of the 6th International Conference on the Chemistry and Biology of Mineralized Tissues, November 1–6, 1998. American Academy of Orthopedic Surgeons, Chicago, IL1999Google Scholar). DMP1 is an acidic protein, rich in aspartic acid, glutamic acid, and serine residues. Fifty-two percent of these serines can be phosphorylated by casein kinase I- and II-like kinases. Upon phosphorylation, DMP1 could be involved in the mineralization process along with other noncollagenous phosphorylated proteins. DMP1 also has a RGD site in its cDNA (7.George A. Sabsay B. Simonian P.A.L. Veis A. J. Biol. Chem. 1993; 268: 12624-12630Abstract Full Text PDF PubMed Google Scholar). We have recently reported that the RGD sequence in DMP1 functions as a cell attachment domain (12.Kulkarni, G. V., Chen, B., Malone, J. P., Narayanan, A. S., and George, A. (1999) Arch. Oral Biol., in pressGoogle Scholar). Little is known about the mechanism regulating the ontogeny and restricted tissue-specific expression of DMP1 gene. The cDNA sequence for DMP1 has been determined in various species, implicating the evolutionary pathway of the DMP1 gene. Cloning of DMP1 in these species has revealed the conservation of the acidic residues and serines that are strategically positioned for phosphorylation (13.Toyosawa S. O'hUigin C. Klein J. J. Mol. Evol. 1999; 48: 160-167Crossref PubMed Scopus (25) Google Scholar, 14.Toyosawa S. O'hUigin C. Tichy H. Klein J. Gene (Amst.). 1999; 234: 307-314Crossref PubMed Scopus (21) Google Scholar, 15.Hirst K.L. Simmons D. Feng J. Aplin H. Dixon M.J. MacDougall M. Genomics. 1997; 15: 38-45Crossref Scopus (57) Google Scholar). To understand the tissue-specific mechanisms behind the DMP1transcription, we have attempted to characterize the promoter of theDMP1 gene. The regulatory factors that are necessary for the transcriptional control of the DMP1 gene may play a role in their tissue specificity. These factors may be cis-acting, such as promoters and enhancers, or trans-acting, such as various DNA binding factors. The differentiation of odontoblasts and secretion of the dentin matrix involve specific genes being expressed at the right time in a temporo-spatial manner (16.Bleicher F. Couble M.L. Farges J.C. Couble P. Magloire H. Matrix Biol. 1999; 18: 133-143Crossref PubMed Scopus (100) Google Scholar). To date, very little is known about the recognition elements and the DNA-binding proteins that regulate the transcription of this gene. To characterize the DMP1 gene and study the mechanisms of tissue-specific regulation, we have isolated a genomic clone containing all the introns as well as a ∼3-kb segment of the 5′-flanking region that includes a functional promoter. DMP1 promoter deletion constructs were also made, and their consequential effects on the expression of a reporter gene (luciferase) were analyzed. To address the question of tissue specificity, we have studied the promoter activity profile in transiently transfected MC3T3-E1, a murine calvaria-derived osteoblast cell line, and in an epithelial cell lineage, CHO. Our ultimate goal is to identify specific transcription factors within the odontoblasts/osteoblasts that are responsible for driving the expression of DMP1. Rat tail genomic DNA was isolated by following the procedure of Sambrook et al. (17.Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar). DNA (20 μg) was digested with different restriction enzymes, and the digested DNA fragments were separated on a 0.8% agarose gel and processed for Southern hybridization as described previously (18.George A. Srinivasan R. Thotakura S.R. Liu K. Veis A. Connect. Tissue Res. 1999; 40: 49-57Crossref PubMed Scopus (33) Google Scholar). Full-length ratDMP1 cDNA was used as a probe. Rat DMP1promoter was isolated by using the Rat Genomewalker kit fromCLONTECH. PCR reactions were carried with a gene-specific primer and an adapter primer. Gene-specific primer was made complementary to nucleotides 51–75, upstream of the translation codon ATG (5′-TGGGTCAAAAGCTCCGTCAGGTTC-3′) of the DMP1cDNA. The PCR amplicons were cloned in to TA vector and sequenced by automated sequencing. More gene-specific primers were made, and PCR was carried until 3 kb of 5′-flanking sequences were obtained. A PCR reaction was also carried with the rat genomic DNA in the 3′ direction with primers (forward and reverse) at nucleotides 361–378 (5′-TCGGCAGACACCACACAG-3′ and 5′-CATGAGGTAACACTTCAATGGCA-3′) to check whether there were any introns downstream of the nucleotide at position 378. PCR amplicons were cloned into TA vector and sequenced by automated sequencing. A synthetic oligonucleotide of 21-mer (5′-GTCAGGTTCTCCCAGAGG-3′) with a sequence complementary to nucleotides 83–64 (upstream of the ATG start codon) of the DMP1 published cDNA (7.George A. Sabsay B. Simonian P.A.L. Veis A. J. Biol. Chem. 1993; 268: 12624-12630Abstract Full Text PDF PubMed Google Scholar) was used for primer extension analysis to determine the transcriptional start site. Total RNA was isolated from MC3T3-E1 cells (which express DMP1) by using the Trizol reagent (Life Technologies, Inc.). Yeast tRNA was used as a negative control. Primer extension was carried according to Sambrook et al. (17.Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. 2nd ed. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar). Briefly, a γ-32P-end-labeled primer was annealed to 20 μg of total RNA at 55 °C for 18 h. The primer-RNA complex was precipitated and resuspended with diethyl pyrocarbonate-treated water. Extension of the annealed primer was carried out using Superscript II (Life Technologies, Inc.) at 42 °C for 90 min. The cDNA was precipitated in the presence of sodium acetate and analyzed on a 6% polyacrylamide/urea gel with a sequencing reaction as a ladder. The gel was dried onto a Whatman 3MM paper and exposed to Kodak x-ray film. A ∼3-kb fragment encompassing the upstream sequences of DMP1 was subcloned into a promoterless luciferase pGL3 luciferase basic vector (Promega). The pGL3-basic vector lacks eukaryotic promoter and enhancer sequences, and expression of luciferase activity in cells transfected with this plasmid depends solely on the putative regulatoryDMP1 sequences that were inserted upstream to the luciferase gene. Systematic nested deletions were made by PCR amplification using the primers listed in Table I. The PCR products were initially cloned into a TA vector (Invitrogen) and further subcloned into pGL3-basic vector (Promega).Table IList of primers used in experimentsName of the primerSequenceST1−2344GTTCCTGGGCTTGGA−2330ST2−1902GCATCTAGGGG−1891ST3−1408CGCAGTCAAACC−1397ST4−773CAAGCATCCTTACCC−759ST5−211CGCCTGGAAATAAGC−196ST0/BglIItctaga+228TCAAAAGCTCCG+217CP5′−288GAGATGAAATTCAAGGCC−271CP3′+38CAGAAGTGGGCCTTCTCT+21ST2-REV−1891CCCCATAGATGC−1902ST3-REV−1397AGGTTTGACTGCG−1408ST1 to ST5 primers were used in combination with ST0/BglIIfor the PCR-mediated deletions within the DMP1 promoter. ABg1II site was incorporated into theST0/BglII primer for easy cloning. CP5′ and CP3′ primers were used for the amplification of the core promoter. ST2-Rev and ST3-Rev were used in the amplification of enhancer fragments in combination with either ST1 or ST2, and denoted as K1, K2, and K3. Open table in a new tab ST1 to ST5 primers were used in combination with ST0/BglIIfor the PCR-mediated deletions within the DMP1 promoter. ABg1II site was incorporated into theST0/BglII primer for easy cloning. CP5′ and CP3′ primers were used for the amplification of the core promoter. ST2-Rev and ST3-Rev were used in the amplification of enhancer fragments in combination with either ST1 or ST2, and denoted as K1, K2, and K3. For the enhancer experiments, the PCR-amplified DNA fragments were cloned into TA vector as mentioned above and further subcloned into pGL3-promoter (SV40 promoter). PGL3-promoter vector has a SV40 promoter to drive the luciferase gene. When compared with the control, any effect on the luciferase activity will be due to the DNA fragment cloned upstream to the SV40 promoter. All the constructs were verified for orientation by partial sequencing or enzymatic digestion. CHO cells, an epithelial cell line, and MCT3T3 cells (mouse calvarial 3T3-like cells), an osteoblast cell line, were grown in F-12 and Dulbecco's modified Eagle's medium, respectively (Life Technologies, Inc.) supplemented with 10% fetal bovine serum. For transcriptional analysis, the various chimeric constructs were transfected into subconfluent cells (60–70% confluence) using the Superfect transfection reagent (Qiagen) as described by the manufacturer. The DNA-Superfectamine mixture was added to the cells, cells were incubated for 2 h, and the medium was replaced by regular medium. The cells were kept in culture for 24 h, and then the cells were lysed and clarified by centrifugation, and the luciferase assay was performed as described below. The amount of protein recovered after lysis was determined by the Bradford assay (Bio-Rad). The pRLSV-40 plasmid was co-transfected with all the transfections to normalize the variations in transfection efficiency. pRLSV-40 encodes Renilla luciferase, and its activity can be distinguished from that of the firefly luciferase encoded in pGL3 using the dual-luciferase assay system (Promega). All transfections were carried out in triplicates. Activities of the firefly luciferase and Renilla luciferase in a single sample were measured using the dual-luciferase reporter assay (DLRTM) system (Promega) according to manufacturer's instructions. In short, 50 μg of protein was dispensed into a 96-well plate and placed in a luminometer equipped with two dispensers. The first dispenser added the firefly luciferase-activating reagent, Luciferase Assay Reagent II, and luminescence was recorded. The second dispenser then added the Renilla luciferase-activating reagent, Stop and Glo Reagent, which simultaneously quenched firefly luminescence and activated the Renilla luciferase. The light emission was recorded using a Dynex luminometer. Variations in transfection efficiency were normalized by dividing the measurement for the firefly luciferase activity by that for the Renilla luciferase activity. To determine the copy number for the ratDMP1 gene, Southern analysis of the rat genomic DNA was performed with various restriction enzymes, and a short cDNA probe was used. A single band was detected in each lane, indicating thatDMP1 is a single copy gene (data not shown). A promoter walking technique was used to isolate 3-kb 5′-upstream sequences. All the introns for DMP1were also isolated from the rat genomic DNA with the Genomewalker technique. The genomic structure for DMP1 consists of six exons and five introns (Fig. 1). The first exon is ∼95-bp long and encodes most of the 5′-untranslated region. The second exon codes 19 bp of the 5′-untranslated region, the signal peptide and the first two N-terminal amino acids of the secreted protein. The third, fourth, and fifth exons code for 16, 11, and 15 amino acids, respectively. The sixth and largest exon codes for the rest of the protein. The exon-intron boundaries were determined and belongs to the class 0 type. The sequences at the boundaries conformed closely to the classical GT/AG rule. There are five introns, and the sizes of the introns 1–5 are 3791, 465, 2047, 162, and 1375 bp, respectively. Primer extension analysis demonstrated that a major band was observed at 280 bp relative to the ATG initiation code; therefore, this would be the transcription initiation site, and it is denoted as +1 in Fig. 2. However, a minor extension product at 321 bp upstream of ATG is also observed (Fig.3).Figure 3Complete nucleotide sequence of the 5′-upstream region of the DMP1 gene showing the position of various transcription factor binding sites. The consensus sequences for the binding factors are shown initalics. An arrow (↓; stronger) and anasterisk (weaker) indicate the transcription start sites. The core promoter region identified was indicated between −288 (start) and +38 bp. The enhancer element identified was between −1902 and −1397 bp.View Large Image Figure ViewerDownload Hi-res image Download (PPT) A 5′-upstream DNA fragment of ∼3 kb was isolated by promoter walk and cloned into the luciferase reporter vector to assay for transcriptional activity of this upstream fragment. Transient transfection assays reveal a severalfold increase in the luciferase activity compared with the pGL3-basic vector. Furthermore, in CHO cells, this construct failed in transcriptional activity (Fig. 4). Future reference of this 3-kb fragment will be as the DMP1 promoter. To determine whether the 5′-flanking region of the rat DMP1 region contains the regulatory elements utilized in a cell-type-specific manner, we made several chimeric DMP1-luciferase constructs. Serial deletions at the 5′-end were made in an attempt to identify various specific regions within the DMP1 promoter. Luciferase assays of these constructs (Fig. 5) revealed that an initial deletion of ∼500 bp (ST1) at the 5′-end of the DMP1 promoter increases its transcriptional activity to ∼11-fold when compared with the full-length DMP1 promoter (FL). Further deletion of another ∼500 bp (ST2) increases the transcriptional activity to ∼15-fold with respect to the full-length DMP1 promoter. With the deletion of ∼2 kb (ST4) at the 5′-end of the DMP1 promoter, a 2-fold transcriptional activity was still retained. Any further serial deletion (ST5) beyond this 2-kb region leads to the complete loss of promoter activity. Our initial serial deletions (a deletion of 500 bp at the 5′-end) revealed a sharp rise in the promoter activity. To identify the presence of any enhancer elements within this region, we have made PCR fragments between −2344 and −1902 bp (K1), −1902 and −1397 bp (K2), and −2344 bp and −1397 bp (K3). These DNA fragments were cloned into a pGL3-promoter vector (which had SV40 promoter). Any enhancer activity within these fragments will affect the luciferase activity through the SV40 promoter. Luciferase assay result of these constructs is shown in Fig.6. It is evident that the K1 construct has no enhancer activity, whereas K2 has increased luciferase activity (∼40-fold), indicating the presence of an enhancer element within this region. The K3 construct, on the other hand, has enhancer activity of about 20-fold with respect to the pGL3/SV40 transcriptional activity (Fig. 6). The enhancer activity of these fragments was restricted to MC3T3-E1 cells. Transfections in CHO cells did not have any significant enhancer activity. The minimal sequences within a promoter to drive the normal activity is called the core promoter. To identify the core promoter region, we have PCR-amplified a 300-bp fragment that spans +38 to −288 bp from the transcription start site. This fragment was cloned into the pGL3-basic vector to analyze its activity. Fig. 7 shows that the 300-bp fragment has complete promoter activity. In fact, this core promoter is almost 1-fold higher in activity than the full-length DMP1promoter. The core promoter identified was equally active in both MC3T3-E1 and CHO cell lines. DMP1, the first cloned dentin-related gene, has an overall composition between that of bone phosphoprotein and that of dentin phosphophoryn (7.George A. Sabsay B. Simonian P.A.L. Veis A. J. Biol. Chem. 1993; 268: 12624-12630Abstract Full Text PDF PubMed Google Scholar). The expression of DMP1 was found to be completely restricted to polarized odontoblasts and differentiated osteoblasts. To study the mechanisms regulating cell type- and differentiation-specific expression of the DMP1 gene, we have isolated and characterized the DMP1 gene along with its exon-intron boundaries and the upstream promoter region. This is the first report describing the genomic sequence and organization of the rat DMP1 gene. DMP1 gene is a single copy gene comprised of six exons and five introns. The fifth exon is present in an isoform ofDMP1 as shown in Fig. 1 and was first reported by MacDougallet al. (8.MacDougall M. Gu T.T. Luan X. Simmons D. Chen J. J. Bone Mineral Res. 1998; 13: 422-431Crossref PubMed Scopus (132) Google Scholar) in the mouse. This additional exon containing 15 amino acids was found to be spliced out from our original cDNA clone. The functional significance of this alternate splicing has not yet been determined. There are five introns of varying sizes, with the largest of them being the first intron. All of the introns are of the phase 0 type because they are flanked on either side by coding triplets. It is due to this type of arrangement that the presence of exon shuffling is noticed in one of the isoforms. The isolated rat DMP1 promoter is characterized by a TATA box and an inverted CCAAT box. In our experiments, we have shown that ratDMP1 is active in the correct orientation. An inverted CCAAT box is also seen in the promoter of bone sialoprotein and chicken osteopontin (19.Craig A.M. Denhardt D.T. Gene (Amst.). 1991; 100: 163-171Crossref PubMed Scopus (153) Google Scholar, 20.Rafidi K. Simkina I. Johnson E. Moore M.A. Gerstenfeld L.C. Gene (Amst.). 1994; 140: 163-169Crossref PubMed Scopus (24) Google Scholar). Analysis of the promoter sequences by the Matinspector program shows that DMP1 has binding sites for other transcription factors such as SP1, AP1, GATA, CREB, MSX 1, MYOD, AP4, c-MYC, ETS-1, and so forth and for serum-responsive elements such as ELK1, ISRE, and so forth. The most common binding sites are shown in Fig. 3. Primer extension studies using total RNA from the rat teeth/MC3T3-E1 cells have identified a major transcription start site at 280 bp upstream from the ATG site. An additional minor start site was also identified at 321 bp upstream of the ATG codon. The functional significance of the additional start site has yet to be analyzed. Similar patterns have been observed in other systems as well (20.Rafidi K. Simkina I. Johnson E. Moore M.A. Gerstenfeld L.C. Gene (Amst.). 1994; 140: 163-169Crossref PubMed Scopus (24) Google Scholar, 21.Banine F. Gangneux C. Lebreton J.P. Frebourg T. Salier J.P. Biochim. Biophys. Acta. 1998; 1398: 1-8Crossref PubMed Scopus (9) Google Scholar, 22.Fruscio M.D. Gilchrist C.A. Baker R.T. Gray D.A. Biochim. Biophys. Acta. 1998; 398: 9-17Crossref Scopus (11) Google Scholar, 23.Li Y.-P. Chen W. J. Bone Miner. Res. 1999; 14: 487-499Crossref PubMed Scopus (52) Google Scholar). To identify cell type-specific transcription of the DMP1gene, several deletions within the DMP1 promoter were made and analyzed for the ability to drive the luciferase gene. Luciferase assays demonstrated that the highest activity was found in the ST2 construct (with a deletion of 1 kb from the 5′-end of theDMP1 promoter). Construct ST4 (with a deletion of 2 kb from the 5′-end of the DMP1 promoter) showed a sharp drop in luciferase activity. However, the ST4 construct has a 1-fold higher activity than that of the full-length DMP1 promoter. Any further deletion of the ST4 construct showed a complete loss of promoter activity. In contrast, all the deletion constructs except ST4 showed no promoter activity in the CHO cell line. This indicates that the sequences upstream to ST4 in the DMP1 promoter construct probably contain tissue-specific elements that are necessary for the expression of DMP1 promoter in the cell lines tested. However, upon the deletion of these tissue-specific sequences, the ST4 construct was equally active in CHO cells. ST5 construct, on the other hand, failed to drive the luciferase gene in both MC3T3-E1 and CHO cell lines, possibly due to the loss of all transcription factor binding sites. Overall, these results suggest that there are tissue-specific regulatory elements within the 2 kb at the distal end of theDMP1 promoter. Furthermore, the above results also suggest that the 1-kb fragment at the 3′-end of the DMP1 promoter has the necessary sequences for a minimal promoter. Studies are being conducted to identify some of the tissue-specific factors that can control the DMP1 gene expression. The core/minimal promoter sequence resides between +38 and −288 bp with respect to transcription start site. The luciferase assay results show that this 300-bp fragment is sufficient to drive the reporter gene. This core promoter is equally active in CHO and MC3T3-E1 cells, indicating that the DMP1 promoter lost its tissue specificity. Data obtained from the serial deletions indicated the presence of an enhancer element within the DMP1 promoter. A 1-kb deletion at the 5′-end of the DMP1 promoter sharply increased its promoter activity by severalfold. This result prompted us to investigate the presence of any enhancer/repressor in and around this region. Luciferase assay showed the presence of an enhancer at −1902 to −1397 bp of the DMP1 promoter. Furthermore, transfection into CHO cells of the same constructs has no effect on the SV40 promoter activity. These results indicate that the enhancer is cell type-specific. In summary, our results have identified cell type-specific regions in the distal part of the DMP1 promoter. These findings will serve as a basis for future studies aimed at identifying the regulatory sequences and specific transcription factor(s) within odontoblasts or osteoblasts that are responsible for controlling the tissue-specific and developmental regulation of the DMP1 gene. We are also attempting to identify the tissue-specific enhancer binding factors responsible for the expression of DMP1 in mineralized matrix. We thank Dr. Vani Turumalla for help with primer extension assay.