Characterization of human bone morphogenetic protein (BMP)-4 and -7 gene promoters: activation of BMP promoters by Gli, a sonic hedgehog mediator1

Bone Vol. 29, No. 1 July 2001:54 – 61

Characterization of Human Bone Morphogenetic Protein (BMP)-4 and -7 Gene Promoters: Activation of BMP Promoters by Gli, a Sonic Hedgehog Mediator S. KAWAI* and T. SUGIURA Laboratory for Bone Research, Discovery Research Laboratories, Hoechst Marion Roussel, Ltd., Kawagoe, Saitama, Japan

actions of molecules involved in the skeletogenesis and developmental processes. (Bone 29:54 – 61; 2001) © 2001 by Elsevier Science Inc. All rights reserved.

Among the bone morphogenetic protein (BMP) family, which plays a crucial role not only in bone formation but also in development, BMP-2, -4, and -7 participate predominantly in various aspects. To undertake complex tasks, their expression is strictly controlled. In this study we isolated and analyzed the 5ⴕ-flanking regions of the human BMP-4 and -7 genes to elucidate the mechanism of their temporally and spatially specific expression. As for BMP-4 expression, a reverse transcription-polymerase chain reaction (RT-PCR) assay with specially designed sets of primers demonstrated that osteoblastic SaOS-2 and Hos cells expressed two types of transcripts comprising one of the 5ⴕ-untranslated first exons, whereas MG63 cells displayed only the transcript with the BMP-4 proximal first exon. Likewise, RT-PCR revealed that Hos and MG63 cells expressed BMP-7. Subsequent 5ⴕ-RACE confirmed an alternative usage of the BMP-4 first exons with clustered multiple transcription start sites in the distal exon and the sole start site in the proximal exon. The transcription start site of the BMP-7 gene was found to be far upstream (764 bp) of the initiation ATG codon. We constructed a series of deletion mutants of fusions between these BMP promoters and the luciferase gene and examined their activity by transient transfection into osteoblastic Hos and renal COS-7 cells. The degree of distal and proximal BMP-4 promoter activity was in accordance with the expression level of the corresponding transcripts. Both distal and proximal BMP-4 promoters possessed suppressor elements that are operative only in Hos cells. The positive and negative elements identified in the BMP-7 promoter were more remarkably effective in Hos cells. The activities of the respective BMP-4 promoters and BMP-7 promoter were all stimulated upon the cotransfection of a potential sonic hedgehog (SHH) mediator, Gli1 or Gli3 into COS-7 cells, providing direct evidence that the Gli proteins are capable of inducing the BMP expression. Our systems are helpful for assessment of the complicated inter-

Key Words: Bone morphogenetic proteins (BMPs); Promoter; Cloning; Osteoblasts; Gli; Sonic hedgehog. Introduction It is now recognized that expression of bone morphogenetic proteins (BMPs), constituting the transforming growth factor-␤ (TGF-␤) superfamily, is regulated precisely for accomplishing multiple functions in a tissue- and development-specific manner.6 To understand the mechanism of the elaborated control, it is fundamental to elucidate the promoter regions of the BMP genes. In this regard we have already investigated the promoters of the human BMP-2 and growth/differentiation factor (GDF)-5 genes and dissected their structures of organization, which are complicated and seemingly necessary for the accurate regulation of their expression.31,32 BMP-4, a representative member of the BMP family, is required for several different processes in early development, beginning with gastrulation and mesoderm formation.13 The deficiency in BMP-4 is lethal, highlighting its importance in embryogeneis.40 As for transcriptional control, BMP-4 expression is enhanced during fracture healing and development.8,18 Meanwhile, BMP-7 forms a subfamily with BMP-5, BMP-6, and BMP-8.6 BMP-7 governs development of several tissues during embryogenesis. In fact, mice in which the BMP-7 gene is disrupted die shortly after birth with poor kidney development, eye defects, and skeletal patterning defects.14 Both BMP-4 and BMP-7 are capable of inducing bone formation at ectopic sites in rodents.26,33 The findings that the BMP-4 and BMP-7 genes are expressed in overlapping regions,15 and that the activity of their heterodimer is superior to that of each homodimer,1 suggest that they may act as a heterodimer in vivo. Thus, BMP-4 and BMP-7 are multifunctional molecules and could work cooperatively, necessitating their expression in a strictly controlled manner. In this regard, our attention turned to the hedgehog pathway. Hedgehog proteins are a family of extracellular signaling proteins that regulate various aspects of embryonic development and skeletogenesis.16,17,38 The most extensively characterized hedgehog, the sonic hedgehog (SHH), is believed to induce BMPs, including BMP-4 and -7 in various mesenchymal sites.3,23,39 Indeed, ectopic expression of SHH results in bone formation.10 In flies, a transcription factor, Cubitus interrupts (Ci), is considered a transducer of

Address for correspondence and reprints: Dr. Takeyuki Sugiura, Discovery Research Laboratory, R&D Center, Daiichi Pharmaceutical Co., Ltd., 16-13, Kitakasai 1 chome, Edogawaku, Tokyo 134-8630. E-mail: [email protected] *Current address: Bone Diseases Group, Hoechst Marion Roussel, Romainville, France. Sequence data from this article have been submitted to the GenBank/ EMBL Data Libraries under the following accession numbers: human BMP-4 gene, AF210053; human BMP-7 gene, AF210054. © 2001 by Elsevier Science Inc. All rights reserved.

54

8756-3282/01/$20.00 PII S8756-3282(01)00470-7

Bone Vol. 29, No. 1 July 2001:54 – 61

hedgehog and regulates the transcription of decapentaplegic (dpp), a BMP family member of Drosophila.2 Accordingly, it can be hypothesized that Gli proteins, vertebrate homologs of Ci, would stimulate BMP expression, but this remains to be demonstrated.24 To extend our previous studies on BMP family gene promoters,31,32 the present investigation focuses on the promoter structures of other human BMP genes. We cloned and functionally characterized the 5⬘-flanking DNA sequence of the BMP-4 and -7 genes and found that Gli1 and Gli3 could enhance their promoter activity. To our knowledge, this is the first direct evidence that Gli proteins have the ability to induce BMP expression. Materials and Methods Cloning of 5⬘-Flanking Regions of BMP-4 and BMP-7 Genes We first sought a restriction enzyme suited to construct a ␭-phage library that accommodates a fragment of the target gene. After human placenta DNA (Clontech, Palo Alto, CA) was digested with various restriction enzymes, the digests were resolved by southern blot analysis with either BMP-4 or BMP-7 cDNA as a probe. It was revealed that the EcoRI treatment yielded a fragment of about 7 kb hybridized with the human BMP-4 probe. Likewise, a fragment of about 11 kb, encompassing the human BMP-7 gene, was produced by treatment of the placenta DNA with HindIII. We isolated the respective fragments from the agarose and cloned them into ␭DASH II (StrataGene, La Jolla, CA). Two genomic libraries were constructed by packaging the resultant vectors in vitro with Gigapack III XL Extract (StrataGene), followed by infection into XL1-Blue MRA (StrataGene). Individual clones from their size-fractioned libraries were screened by polymerase chain reaction (PCR) amplification7 using BMP-4 or BMP-7 specific primers to identify the genomic clones that contain such sequences. The oligonucleotide primers for PCR were designed based on cDNA sequences20,41 such as 5⬘-GGCAGAGGAGGAGGGAGGGAGGGAAGGAGC-3⬘ (forward) and 5⬘-CAGTAGCGGGCTCGCCAGCAGCAGCTCCTG-3⬘ (reverse) for the BMP-4 gene and 5⬘-GGGCGCAGCGGGGCCCGTCTGCAGCAAGTG-3⬘ (forward) and 5⬘-AGAGGATCTCGCGCTGCATCTCCCGCCGCT-3⬘ (reverse) for the BMP-7 gene. PCR conditions for both genes involved 35 cycles of 94°C for 1 min, 65°C for 1 min, and 72°C for 3 min. The expected sizes of the PCR products were 220 bp for the BMP-4 gene and 300 bp for the BMP-7 gene. After identification of the inserts by partial sequencing with an ALF autosequencer (Amersham Pharmacia Biotech, Uppsala, Sweden), the respective fragments (6.8 kb for BMP-4 and 10.8 kb for BMP-7) were subcloned into pUC18 and the entire sequences were determined with FITC-labeled specific primers.

S. Kawai and T. Sugiura Human BMP-4 and -7 promoters

55

that annealing temperatures for BMP-4 and BMP-7 transcripts were 55°C and 53°C, respectively. The products were resolved by electrophoresis on 1.5% agarose gels. 5⬘ Rapid Amplification cDNA Ends (RACE) The 5⬘-RACE reaction was performed with a commercial kit from Life Technologies (Rockville, MD) according to the manufacturers protocol. The gene specific primers for BMP-4 were 5⬘-TCAGGTATCAAACTAGCATGGCTCGCGCCT-3⬘, 5⬘CCTAGCAGGACTTGGCATAATAAAACGACC-3⬘, and 5⬘ATCAGCATTCGGTTACCAGGAATCATGGTG-3⬘ (451–480, 421– 450, and 391– 420 of sequence GeneBank accession no. M22490, respectively). All primer sequences exist in exon 4 of the human BMP-4 gene. Those for BMP-7 were 5⬘-AGAGGATCTCGCGCTGCATCTCCCGCCGCT-3⬘, 5⬘-CCTGGCTGCGGAGGCGCCGGTGGATGAAGC-3⬘, and 5⬘-TCGAGTGCACCTCGTTGTCCAGGCTGAAGT-3⬘ (245–274, 215–244, and 185–214 of sequence GeneBank accession no. M60316, respectively). Following cloning of PCR products into the pGEM-T vector (Promega, Madison, WI), independent colonies were taken randomly and subjected to sequencing analysis. Construction of Promoter-Reporter Expression Vectors The promoter fragments were ligated into pGL3-basic (pGL3B) luciferase reporter vector (Promega) that carries neither eukaryotic promoter nor enhancer. The restriction sites indicated in Figures 4 and 5 were utilized to construct serially deleted promoter-reporter vectors. Cell Culture, Transfection, and Luciferase Activity Assay Human osteosarcoma, Hos, MG63, and SaOS-2 cells, and simian kidney COS-7 cells, all obtained from the American Type Culture Collection, were routinely maintained in Dulbecco’s modified Eagle’s medium containing 10% fetal calf serum (FCS), 100 U/mL penicillin, and 50 ␮g/mL streptomycin. One day before transfection, cells were plated in 24 well culture plates at a density of 4 ⫻ 104 per well. Transient transfection of various promoter-luciferase reporter vectors (0.4 ␮g), along with 40 ng of an internal control, pcDNA3.1/His/LacZ, expressing ␤-galactosidase (Invitrogen, Carlsbad, CA), was done using LipofectAMINE⫹ (Life Technologies). About 72 h after transfection, luciferase activity was measured by the Luciferase Assay System (Promega) and normalized by ␤-galactosidase activity determined with the Galacto-Light Kit (Tropix, Inc., Bedford, MA). The experiments were done in triplicate and repeated at least twice.

Reverse Transcription-Polymerase Chain Reaction (RT-PCR) Poly(A)⫹ RNA was prepared and reverse transcribed with kits purchased from Amersham. Because it has been reported that the human BMP-4 gene contains two first exons (exons 1 and 2, see figure 1B), which are mutually excluded,37 two sets of primers were designed for BMP-4 transcripts: 5⬘-GAGTATCTAGCTTGTCTCCCC-3⬘ (forward sequence in exon 1); 5⬘-TGTTCACCGTTTTCTCGACTC-3⬘ (forward sequence in exon 2); and 5⬘-TCAGGTATCAAACTAGCATGG-3⬘ (reverse sequence in exon 3 used as a common primer). The primer set for the BMP-7 transcript was 5⬘-TGGAACATGACAAGGAATTCT-3⬘ (forward) and 5⬘-CATCCAGCGTCTCCACCGAGA-3⬘ (reverse). PCR conditions were the same as those for PCR screening except

Effects of Gli Proteins Coexpression on BMP Promoter Activity The human Gli vectors driven by SR␣ promoter were kind gifts from Dr. H. Sasaki of Osaka University.27 For cotransfection the cells were treated as described earlier with 0.2 ␮g of the BMP promoter-reporter vector (B4P1EX, B4P2XB, or B7XX), 40 ng of pcDNA3.1/His/LacZ, and different amounts of a Gli-expressing vector. The total amount of plasmid was adjusted to 0.4 ␮g by adding an empty pSR␣ vector. The following procedures were the same as reported earlier. The experiments were done in triplicate.

56

S. Kawai and T. Sugiura Human BMP-4 and -7 promoters

Bone Vol. 29, No. 1 July 2001:54 – 61

Assuming that there exist two kinds of human BMP-4 transcripts with respective first exons based on results of the mouse BMP-4 genes,4,11 we designed two sets of primers for RT-PCR. Electrophoresis patterns (Figure 3A) showed that Hos and SaOS-2 cells expressed the transcript spanning exon 1 and exon 3, whereas all cells tested expressed the transcript spanning exon 2 and exon 3, albeit at lower levels, corroborating the finding that the human BMP-4 gene expresses two types of transcripts composed of different first exons.37 All osteoblastic cells except for SaOS-2 cells showed a clear band for the BMP-7 transcript (Figure 3B). Although COS-7 cells were of monkey origin, a definite PCR band was detected at the expected size by the use of primers designed for the human BMP-7 sequence, presumably due to its high homology with the corresponding simian sequence. Determination of Transcription Initiation Sites

Figure 1. Structure of the human BMP genes. Boxes denote exons and lines noncoding sequences. Note that exons 1 and 3 are designated as exons 1 and 2 in Shore et al.29 ATG depicts the translation initiation codon. Restriction sites abbreviations: Bg, BglII; Bm, BamHI; E, EcoRI; H, HindIII; K, KpnI; N, NcoI; P, PstI; Sa, SacI; Sp, SphI; Xb, XbaI; Xh, XhoI. Some of the restriction sites were utilized for the construction of the deletion mutants in Figures 4 and 5. (A) BMP-4 gene. (B) Exon usage and alternative splicing patterns of the human BMP-4 gene are schematically represented, but not drawn to scale. (C) BMP-7 gene.

Results Cloning of 5⬘-Flanking Regions of Human BMP-4 and -7 Genes The human BMP-4 and BMP-7 genes were cloned by PCR screening the respective size-fractioned human genomic DNA libraries, as described in Materials and Methods. The overall identity of our BMP-4 sequence (6774 bp) with that from Shore et al.29 and van den Wijngaard et al.36 was 99.4% and 98.6%, respectively. The discrepancies could be ascribed to genetic polymorphic variations or possible sequencing errors in each clone. In any case, it was evident from sequence analysis that the cloned sequence originated from the BMP-4 gene. Alignment of the isolated BMP-7 sequence with the published cDNA sequence20 also assured that it was actually derived from the human BMP-7 gene. Comparison of the sequences demonstrated that the genomic DNA fragment comprised about 8.8 kb 5⬘-flanking region, 480 bp exon including the initiation ATG codon, and 1.6 kb intron. The exon-intron junctions were in accordance with the GT/AG rule.28 The restriction maps of the cloned sequences are indicated in Figure 1 and a part of the cloned sequences in Figure 2. Transcription of BMPs in Osteoblastic Cells Because human tissues were not available to us, we employed an established cell line for the transcription initiation study. To select the cell lines conferring the source of mRNA, human osteosarcoma cell lines, Hos, MG63, and SaOS-2, were examined by RT-PCR for BMP expression. Renal COS-7 cells were included as a positive control for the BMP-7 transcript as kidney expresses BMP-7 mRNA in abundance.21,22

Because the RT-PCR assay indicated that BMP-4 expression levels from both promoters were the most prominent in SaOS-2 cells, total RNA obtained from the cells was used for 5⬘-RACE. After cloning the PCR products, 20 independent colonies were randomly picked up and subjected to sequencing analysis. Thirteen clones contained the fragment linking exons 1 and 3, excluding any exon 2 sequence. The sequences of five clones were derived from exons 2 and 3. The remaining two clones were without insert. These results support the notion that the BMP-4 gene possesses two first exons, preceeded by distinct promoters, which are utilized in a mutually exclusive manner. Importantly, 5⬘-RACE indicated that in exon 1 transcription initiation occurred at various sites, confined to the region of about 300 bp (Figure 2A),29,36 whereas the sole site was determined as an initiation point in exon 2 (Figure 2B). In addition, in four clones of the exon 1/exon 3 transcript we also observed a fragment missing a 3⬘ portion, apparently resulting from alternative splicing (Figure 1B).36 Transcription start sites for the BMP-7 gene were mapped similarly by the 5⬘-RACE method using total RNA from MG63 cells appropriately expressing the gene. Out of six randomly selected colonies harboring the PCR products, four indicated the same position (⫹1) as a transcription start site (Figure 2C) and the other two clones defined different start sites (⫹600 and ⫺1131). As the last transcript makes a fairly long 5⬘ leading sequence (approximately 1.9 kb), it may be an artifact. It should be noted that the distal and proximal BMP-4 promoters and the BMP-7 promoter all lacked cononical TATA and CAAT consensus around their transcription initiation sites. Promoter Activity of 5⬘-Flanking Regions of BMP-4 and BMP-7 Genes To define the promoter region of the cloned genes, we constructed luciferase reporter vectors fused with the 5⬘-flanking sequences. We transiently transfected fusions into osteoblastic Hos and renal COS-7 cells and assessed the promoter activities, as represented by the luciferase activities. Hos cells were employed among the three osteoblastic cells because their transfection efficiency was the most favorable and they were the only cell line expressing all the BMP transcripts of interest (Figure 3A,B). Because it was suggested from the aforementioned results that BMP-4 expression is regulated by two promoters, promoter 1 (the 3.4 kb EcoRI-XhoI fragment) and promoter 2 (the 2.3 kb XhoI-BglII fragment) were ligated separately into a pGL3B luciferase reporter vector. As Figure 4A shows, although both BMP-4 promoters were effective in the osteoblastic as well as in

Bone Vol. 29, No. 1 July 2001:54 – 61

S. Kawai and T. Sugiura Human BMP-4 and -7 promoters

57

Figure 2. Partial sequence of the human BMP promoters. Lower case means intron. Numbers on the left (in bold) refer to the position relative to the transcription start sites (⫹1). Numbers under the start sites indicate frequency. Potential transcription factor binding motifs are boxed. They were assigned by TFSEARCH (Y. Akiyama, http://www.rwcp.or.jp/papia/). Matrix similarity scores were set at 90.0, except Sp1 at 80.0. (A) Basal region of BMP-4 promoter 1. The most upstream transcription start site is numbered as ⫹1. The alternatively spliced GT site is in italics. (B) Core region of BMP-4 promoter 2. (C) BMP-7 basal promoter region. The initiation ATG codon is underlined.

the kidney cells, the promoter 1 construct (B4P1EX) was more potent compared with the promoter 2 construct (B42PXB), which is consistent with the relative abundance of the respective BMP-4 transcripts in Hos cells (Figure 3A). The discernible activity of the BMP-4 promoters in the kidney cells was not surprising because BMP-4 is expressed in various tissues, including kidney.41

We further explored the positive and negative elements and the basal regions in the BMP-4 promoters by constructing stepwise deletion promoter-reporter plasmids and transfecting them into the two cell lines (Figure 4B). Removal of the sequence to position ⫺1917 from B4P1EX increased the activity in Hos, implicating the presence of a suppressor element. Sequential deletion to position ⫺1131 caused a gradual decrease in the

58

S. Kawai and T. Sugiura Human BMP-4 and -7 promoters

Bone Vol. 29, No. 1 July 2001:54 – 61

Figure 3. RT-PCR analysis of BMP expression. Arrows mark the position of the expected bands. DNA size marker was ␾x174-HaeIII digest (Takara Shuzo, Kyoto, Japan). (A) BMP-4 expression. Human BMP-4 cDNA was used as the template for a positive control. The predicted sizes of the PCR products specific for exon 1/exon 3 transcript and exon 2/exon 3 transcript were 370 and 250 bp, respectively. Note that the band intensities of the two transcripts in the control lane are comparable, permitting comparison of mRNA abundance between two BMP-4 transcripts. (B) BMP-7 expression. COS-7 cells were employed as a positive control. The expected size of the PCR product was 340 bp.

promoter activity but the B4P1NX construct still retained significant activity in the osteoblastic cells. In COS-7 cells, deletion of the sequence to ⫺1131 led to a stepwise decline in promoter activity. Again, basal activity was retained by the B4P1NX construct in this cell line. Thus, the ⫺297/⫹202 region contributed mainly to promoter 1 activity. Elimination of the ⫺2099/⫺1168 sequence from B4P2XB enhanced the activity in Hos cells, suggesting a negative regulatory element in this region. A more extensive deletion to position ⫺315 (B4P2SB) influenced the activity only slightly. In contrast, the promoter activity was not significantly affected by the removal of the sequence to position ⫺315 in COS-7 cells. In conclusion, the activity of promoter 2 was represented by the ⫺315/⫹191 region. The activity of the human BMP-7 promoter region was analyzed in a similar way by the transient transfection experiments. When the B7PXX construct containing the longest sequence of the BMP-7 5⬘-flanking region was transfected into the two cell lines, its activity was about fourfold (Hos) or tenfold (COS-7) greater than the promoterless pGL3B activity (Figure 5A). Northern blot analysis indicated that BMP-7 was progressively transcribed in the kidney tissue.21,22 The promoter region of the human BMP-7 gene was delineated in more detail by carrying out deletion mutants experiments (Figure 5). Removal of the ⫺4313/⫺1771 fragment from the parental construct enhanced the activity by about fourfold in Hos cells. Further deletion to position ⫺1277 (B7PPX) compromised the activity to some extent, but substantial activity was maintained by the minimal construct consisting of the ⫺491/⫹111 fragment (B7PKX). In COS-7 cells, BMP-7 promoter activity was fairly refractory to truncation, despite a similar tendency to that in Hos cells, at least until position ⫺491. In conclusion, we found suppressor elements in the ⫺3070/⫺1771 and ⫺1277/⫺491 sequences and an enhancer element in the ⫺1771/⫺1277 sequence. Basal activity was shown by the ⫺1277/⫺491 sequence. Potential transcription factor binding motifs within the core promoter regions of the respective BMP genes are shown in

Figure 4. Promoter activity of the human BMP-4 gene 5⬘ upstream region (A) and serial deletion analysis of the promoters (B). Activity of the human BMP-4 promoter constructs was assayed in Hos and COS-7 cells. Constructs are identified on the left of the diagram. The experiments were done in triplicate and repeated at least three times. The results are expressed as the mean ⫾ standard errors and indicated as fold inductions over the promoterless pGL3B vector (A) or as relative values to the activity of B4P1EX for the top panel and B4P2PB for the bottom panel set at 100% (B).

Figure 2. Note that all promoters contained multiple Sp1 sites, which is characteristic of the promoters without a TATA box.

Figure 5. Promoter activity of the human BMP-7 gene 5⬘ upstream region. Constructs are identified on the left of the diagram. Activity of the human BMP-7 promoter constructs was monitored in Hos and COS-7 cells. The results are expressed as the mean ⫾ standard errors and indicated as fold inductions relative to pGL3B.

Bone Vol. 29, No. 1 July 2001:54 – 61

S. Kawai and T. Sugiura Human BMP-4 and -7 promoters

59

dent manner. Surprisingly, no stimulation of promoter activity was found when the constructs were transfected into Hos cells (data not shown). Discussion

Figure 6. Effects of Gli protein coexpression on BMP-4 promoter 1 (A) and promoter 2 (B) and BMP-7 promoter (C) activity. COS-7 cells were transfected with 0.2 ␮g of the BMP promoter-reporter vector, the indicated amount (ng) of Gli expressing vector, and 40 ng of pcDNA3.1/ His/LacZ. Total amount of plasmid was adjusted to 0.4 ␮g by adding an empty pSR␣ vector. Data shown are representative of two experiments done in triplicate with similar results. Results are expressed as mean ⫾ standard error and indicated as fold inductions over the promoter activity without Gli contransfection.

Effects of Gli Expression on BMP Promoter Activities Because Gli proteins are mediators of SHH that have the potential to induce BMP expression, it was of interest to determine whether coexpression of Gli proteins could augment the activity of BMP promoters. As shown in Figure 6, the activity of all BMP promoters was enhanced upon the cotransfection of the Gli1- or Gli3-expressing vector in COS-7 cells in a dose-depen-

We have already investigated the promoter structures of the human BMP-2 and GDF-5 genes and extended the study to the BMP-4 and -7 genes in this report. Table 1 summarizes the noticeable features of the human BMP gene family promoters observed in this study or those that have appeared in the literature. It is clear from this table that, although we can classify BMP-2 and -4 and BMP-5, -6, and 7 into two respective subgroups, each member shows outstanding differences in the promoter features within the subgroups. In addition, GDF-5 is unique in its promoter structure, which may reflect its special role in chondrogenesis. It seems established that the mouse BMP-4 gene has the two nontranslated first exons that are transcribed in an exclusive manner.4,11 Nevertheless, the presence of the additional first exon in the human BMP-4 gene has been controversial because one group of investigators found an alternative use of two exons,36 similar to the corresponding murine gene,4,11 but the other group failed to detect this.29 Both groups employed the same cell line (U-2 OS), but obtained different results.29,36 In this study we have presented definitive data for the existence of the exon of interest (exon 2) in SaOS-2 and Hos cells by RT-PCR and 5⬘-RACE. Furthermore, the earlier experiment demonstrating alternative splicing by the 5⬘-RACE methods was performed with embryonic carcinoma cells,36 but here our 5⬘-RACE data supporting the alternative exon were obtained with osteoblastic SaOS-2 cells. Despite the more prevalent expression of transcript from the proximal promoter (promoter 2) in human osteoblastic cells (Figure 3A), its expression level was lower than that of the distal promoter (promoter 1) transcript. In fact, activities of promoter-reporter fusions in the transient transfection experiments (Figure 4A) mirrored the expression levels of the corresponding transcripts in Hos cells. One of the striking features of the basal region of promoter 1 in terms of the possible transcription factor binding sites shown in Figure 2A is that the combination of deltaE, ETS, and Sp1 occurs twice (⫺230, ⫺147, ⫺95 and 90, 83, 218, respectively). In contrast, no particular features other than multiple Sp1 sites were seen for the region around the start site of the proximal promoter (promoter 2). It is likely that expression from promoter 1 needs a set of transcription factors, thereby limiting the cell species competent for the expression.37 In turn, when a necessary combination is available, the transcriptional activity of promoter 1 may be augmented greatly by cooperation of the transcription factors. Conversely, prevalent expression of the promoter 2 transcript, albeit at a lower level, could be realized by the fewer numbers of necessary transcription factors. The significance of the individual transcription

Table 1. Comparison of the promoter organization between the human BMP family members

TIS TATA Leader sequence (bp) Remarks

BMP-231

BMP-4

BMP-525

Multiple No 147–493

Multiple No 37–328 Two exclusively utilized first exons

Two One TATA 588–698

BMP-635

BMP-7

GDF-532

GDF-105

Single No 218

Single No 764

Single Initiator 359 GT repeat

Multiple No 327–456

KEY: TIS, transcription initiation site. Multiple represents the number of the identified transcription start sites more than two. Leader sequence indicates a range from the shortest to the longest length when various initiation sites exist. The sequence around the GDF-5 transcription start site contains the initiator consensus30 instead of TATA. Data for BMP-4 and -7 genes are from this study. The BMP-7 gene might possess additional minor transcription initiation sites.

60

S. Kawai and T. Sugiura Human BMP-4 and -7 promoters

factors for the expression of each transcript awaits further investigation. In any case, all properties involved in the BMP-4 promoter structure should allow for the fine tuning of its expression. The BMP-7 gene contains one main transcription start site fairly far upstream (764 bp) of the initiation codon. The relatively long 5⬘-noncoding leader sequence is likely to be a characteristic of the BMP family of genes, ranging approximately 200 –700 bp in length (Table 1). This feature appears to be conserved more widely in the TGF-␤ gene family because the leader sequences of human TGF-␤1, -␤2, and -␤3 mRNA are 840, 1357, and 1110 bp in length, respectively.9,12,19 The long leader sequence is possibly helpful in facilitating translation.42 The deletion mutant experiments of the BMP-7 promoterreporter fusions indicated that the BMP-7 5⬘-flanking region bore both enhancers and silencers that were more prominently effective in the osteoblastic Hos cells than in the kidney COS-7 cells. This implies that the osteoblastic cells contained transcription factors binding to extended elements of the BMP-7 promoter other than those in COS-7 cells. Although BMP-7 transcription is observed in various tissues, including bone and kidney,21 it must be regulated accurately to achieve tissue and developmentspecific expression.15 The enhancers and silencers observed in Hos cells would be responsible for such precise control of the BMP-7 expression. Comparison between the core promoter region of BMP-7 and that of human BMP-2 (⫺372 to ⫺1 in Figure 2 of Sugiura31), which are colocalized frequently during development,15 has shown that both promoters contain common transcription factor binding motifs, such as multiple Sp1, MZF1, and GATA1 sites. These transcription factors might be the basis for the expression of BMP-2 and -7. Finally, it must be stressed that all BMP promoters (BMP-4 promoters 1 and 2 and BMP-7) responded to Gli protein coexpression in COS-7 cells. This result supports the idea that Gli proteins play a key role in inducing BMP molecules during development as a mediator of SHH function.38 Indeed, we found several sequences resembling Gli-binding motifs (TGGGTGGTC) in all the responsive BMP promoters (⫺2682, ⫺2583, ⫺2173, ⫺1768, ⫺1336 in BMP-4 promoter 1; ⫺1564, ⫺962, ⫺269, ⫺101 in BMP-4 promoter 2; and ⫺4300, ⫺4011, ⫺3732, ⫺3693, ⫺3576, ⫺3123, ⫺2779, ⫺1498, ⫺1384, ⫺1226, ⫺855, ⫺764, ⫺443, ⫺311, ⫺302, ⫺42 in BMP-7 promoter). However, surprisingly, none of the promoters was stimulated by the Gli proteins in Hos cells. As Gli proteins need cofactors such as CBF or Smad for BMP induction,2 such factors may not suffice for induction in Hos cells. Alternatively, the Gli proteins expressed in Hos cells might be localized in the cytoplasm rather than in the nucleus, thereby limiting the ability to induce BMP expression.2 Because Hos cells constitutively express all BMP transcripts (Figure 3A), and no induction of their expression by the Gli expression was observed, it is certain that the mechanism or transcription factors involved are different between constitutive and inducible expression of BMP proteins. We cannot exclude the possibility that no effect of Gli1 and Gli3 cotransfection on BMP promoters in Hos cells would be ascribed to their insufficient expression level. To test this, we examined the expression of Gli proteins by immunoblotting with respective anti-Gli proteins antibodies (Santa Cruz). We were unable to detect Gli proteins in Hos and COS-7 cells both before and after transfection (data not shown), presumably due to the low expression level of the proteins, low affinity of the antibodies, or both. Nevertheless, we believe that a considerable quantity of Gli proteins was expressed in Hos cells in view of the potent activity of SR␣ promoter in the Hos cells.34 In any case, we demonstrated effects of Gli proteins on BMP promoters that are important in the developmental process, and the systems exam-

Bone Vol. 29, No. 1 July 2001:54 – 61

ined herein would aid in clarifying the underlying mechanism of their interaction. Acknowledgments: The authors thank Dr. T. Matsuishi and Dr. R. Baron for advice and encouragement. We are also thankful to all the members of the BMP Promoter Project led by T.S.

References 1. Aono, A., Hazama, M., Notoya, K., Taketomi, S., Yamasaki, H., Tsukuda, R., Sasaki, S., and Fujisawa, Y. Potent ectopic bone-inducing activity of bone morphogenetic protein-4/7 heterodimer. Biochem Biophys Res Commun 210: 670 – 677; 1995. 2. Aza-Blanc, P. and Kornberg, T. B. Ci: A complex transducer of the hedgehog signal. Trends Genet 15:458 – 462; 1999. 3. Bitgood, M. J. and McMahon, A. P. Hedgehog and Bmp genes are coexpressed at many diverse sites of cell– cell interaction in the mouse embryo. Dev Biol 172:126 –138; 1995. 4. Feng, J. Q., Chen, D., Cooney, A. J., Tsai, M. J., Harris, M. A., Tsai, S. Y., Feng, M., Mundy, G. R., and Harris, S. E. The mouse bone morphogenetic protein-4 gene. Analysis of promoter utilization in fetal rat calvarial osteoblasts and regulation by COUP-TFI orphan receptor. J Biol Chem 270:28364 –28373; 1995. 5. Hino, J., Takao, M., Takeshita, N., Konno, Y., Nishizawa, T., Matsuo, H., and Kangawa, K. cDNA cloning and genomic structure of human bone morphogenetic protein-3B (BMP-3b). Biochem Biophys Res Commun 223:304 –310; 1996. 6. Hogan, B. L. Bone morphogenetic proteins: Multifunctional regulators of vertebrate development. Genes Dev 10:1580 –1594; 1996. 7. Israel, D. I. A PCR-based method for high stringency screening of DNA libraries. Nucl Acids Res 21:2627–2631; 1993. 8. Jones, C. M., Lyons, K. M., and Hogan, B. L. Involvement of bone morphogenetic protein-4 (BMP-4) and Vgr-1 in morphogenesis and neurogenesis in the mouse. Development 111:531–542; 1991. 9. Kim, S. J., Glick, A., Sporn, M. B., and Roberts, A. B. Characterization of the promoter region of the human transforming growth factor-beta 1 gene. J Biol Chem 264:402– 408; 1989. 10. Kinto, N., Iwamoto, M., Enomoto-Iwamoto, M., Noji, S., Ohuchi, H., Yoshioka, H., Kataoka, H., Wada, Y., Yuhao, G., Takahashi, H. E., Yoshiki, S., and Yamaguchi, A. Fibroblasts expressing sonic hedgehog induce osteoblast differentiation and ectopic bone formation. Fed Eur Biol Soc Lett 404:319 – 323; 1997. 11. Kurihara, T., Kitamura, K., Takaoka, K., and Nakazato, H. Murine bone morphogenetic protein-4 gene: Existence of multiple promoters and exons for the 5⬘-untranslated region. Biochem Biophys Res Commun 192:1049 –1056; 1993. 12. Lafyatis, R., Lechleider, R., Kim, S. J., Jakowlew, S., Roberts, A. B., and Sporn, M. Structural and functional characterization of the transforming growth factor beta 3 promoter. A cAMP-responsive element regulates basal and induced transcription. J Biol Chem 265:19128 –19136; 1990. 13. Lawson, K. A., Dunn, N. R., Roelen, B. A., Zeinstra, L. M., Davis, A. M., Wright, C. V., Korving, J. P., and Hogan, B. L. Bmp4 is required for the generation of primordial germ cells in the mouse embryo. Genes Dev 13:424 – 436; 1999. 14. Luo, G., Hofmannm, C., Bronckers, A. L., Sohocki, M., Bradley, A., and Karsenty, G. BMP-7 is an inducer of nephrogenesis, and is also required for eye development and skeletal patterning. Genes Dev 9:2808 –2820; 1995. 15. Lyons, K. M., Hogan, B. L., and Robertson, E. J. Colocalization of BMP 7 and BMP 2 RNAs suggests that these factors cooperatively mediate tissue interactions during murine development. Mech Dev 50:71– 83; 1995. 16. McMahon, A. P. More surprises in the hedgehog signaling pathway. Cell 100:185–188; 2000. 17. Mo, R., Freer, A. M., Zinyk, D. L., Crackower, M. A., Michaud, J., Heng, H. H., Chik, K. W., Shi, X. M., Tsui, L. C., Cheng, S. H., Joyner, A. L., and Hui, C. Specific and redundant functions of Gli2 and Gli3 zinc finger genes in skeletal patterning and development. Development. 124:113–123; 1997. 18. Nakase, T., Nomura, S., Yoshikawa, H., Hashimoto, J., Hirota, S., Kitamura, Y., Oikawa, S., Ono, K., and Takaoka, K. Transient and localized expression of bone morphogenetic protein 4 messenger RNA during fracture healing. J Bone Miner Res 9:651– 659; 1994. 19. Noma, T., Glick, A. B., Geiser, A. G., O’Reilly, M. A., Miller, J., Roberts,

Bone Vol. 29, No. 1 July 2001:54 – 61

20.

21.

22.

23.

24. 25.

26.

27.

28.

29.

30. 31. 32.

A. B., and Sporn, M. B. Molecular cloning and structure of the human transforming growth factor-beta 2 gene promoter. Growth Factors 4:247–255; 1991. Ozkaynak, E., Rueger, D. C., Drier, E. A., Corbett, C., Ridge, R. J., Sampath, T. K., and Oppermann, H. OP-1 cDNA encodes an osteogenic protein in the TGF-beta family. Eur Med Biol Org J 9:2085–2093; 1990. Ozkaynak, E., Schnegelsberg, P. N., Jin, D. F., Clifford, G. M., Warren, F. D., Drier, E. A., and Oppermann, H. Osteogenic protein-2. A new member of the transforming growth factor-beta superfamily expressed early in embryogenesis. J Biol Chem 267:25220 –25227; 1992. Ozkaynak, E., Schnegelsberg, P. N., and Oppermann, H. Murine osteogenic protein (OP-1): High levels of mRNA in kidney. Biochem Biophys Res Commun 179:116 –123; 1991. Roelink, H. Tripartite signaling of pattern: Interactions between hedgehogs, BMPs and Wnts in the control of vertebrate development. Curr Opin Neurobiol 6:33– 40; 1996. Ruiz i Altaba, A. Gli proteins and hedgehog signaling: Development and cancer. Trends Genet 15:418 – 425; 1999. Sakaue, M., Kitazawa, S., Nishida, K., Kitazawa, R., and Maeda, S. Molecular cloning and characterization of human bone morphogenic protein (BMP)-5 gene promoter. Biochem Biophys Res Commun 221:768 –772; 1996. Sampath, T. K., Maliakal, J. C., Hauschka, P. V., Jones, W. K., Sasak, H., Tucker, R. F., White, K. H., Coughlin, J. E., Tucker, M. M., Pang, R. H. L., Corbett, C., Oppermann, H., and Rueger, D. C. Recombinant human osteogenic protein-1 (hOP-1) induces new bone formation in vivo with a specific activity comparable with natural bovine osteogenic protein and stimulates osteoblast proliferation and differentiation in vitro. J. Biol Chem 267:20352–20362; 1992. Sasaki, H., Hui, C., Nakafuku, M., and Kondoh, H. A binding site for Gli proteins is essential for HNF-3 beta floor plate enhancer activity in transgenics and can respond to Shh in vitro. Development 24:1313–1322; 1997. Shapiro, M. B. and Senapathy, P. RNA splice junctions of different classes of eukaryotes: Sequence statistics and functional implications in gene expression. Nucl Acids Res 11:7155–7174; 1987. Shore, E. M., Xu, M., Shah, P. B., Janoff, H. B., Hahn, G. V., Deardorff, M. A., Sovinsky, L., Spinner, N. B., Zasloff, M. A., Wozney, J. M., and Kaplan, F. S. The human bone morphogenetic protein 4 (BMP-4) gene: Molecular structure and transcriptional regulation. Calcif Tissue Int 63:221–229; 1998. Smale, S. T. Transcription initiation from TATA-less promoters within eukaryotic protein-coding genes. Biochim Biophys Acta 1351:73–78; 1997. Sugiura, T. Cloning and functional characterization of the 5⬘-flanking region of the human bone morphogenetic protein-2 gene. Biochem J 338:433– 440; 1999. Sugiura, T., Ho¨tten, G., and Kawai, S. Minimal promoter components of the human growth/differentiation factor-5 gene. Biochem Biophys Res Commun 263:707–713; 1999.

S. Kawai and T. Sugiura Human BMP-4 and -7 promoters

61

33. Takaoka, K., Yoshikawa, H., Hashimoto, J., Ono, K., Matsui, M., and Nakazato, H. Transfilter bone induction by Chinese hamster ovary (CHO) cells transfected by DNA encoding bone morphogenetic protein-4. Clin Orthop 300:269 –273; 1994. 34. Takebe, Y., Seiki, M., Fujisawa, J., Hoy, P., Yokota, K., Arai, K., Yoshida, M., and Arai N. SR alpha promoter: An efficient and versatile mammalian cDNA expression system composed of the simian virus 40 early promoter and the R-U5 segment of human T-cell leukemia virus type 1 long terminal repeat. Mol Cell Biol 8:466 – 472; 1988. 35. Tamada, H., Kitazawa, R., Gohji, K., Kamidono, S., Maeda, S., and Kitazawa, S. Molecular cloning and analysis of the 5⬘-flanking region of the human bone morphogenetic protein-6 (BMP-6). Biochim Biophys Acta 1395:247–251; 1998. 36. van den Wijngaard, A., Pijpers, M. A., Joosten, P. H., Roelofs, J. M., van Zoelen, E. J., and Olijve, W. Functional characterization of two promoters in the human bone morphogenetic protein-4 gene. J Bone Miner Res 14:1432– 1441; 1999. 37. van den Wijngaard, A., van Kraay, M., van Zoelen, E. J., Olijve, W., and Boersma, C. J. Genomic organization of the human bone morphogenetic protein-4 gene: Molecular basis for multiple transcripts. Biochem Biophys Res Commun 219:789 –794; 1996. 38. Vortkamp, A., Pathi, S., Peretti, G. M., Caruso, E. M., Zaleske, D. J., and Tabin, C. J. Recapitulation of signals regulating embryonic bone formation during postnatal growth and in fracture repair. Mech Dev 71:65–76; 1998. 39. Wall, N. A. and Hogan, B. L. Expression of bone morphogenetic protein4(BMP-4), bone morphogenetic protein-7 (BMP-7), fibroblast growth factor-8 (FGF-8) and sonic hedgehog (SHH) during branchial arch development in the chick. Mech Dev 53:383–392; 1995. 40. Winnier, G., Blessing, M., Labosky, P. A., and Hogan, B. L. Bone morphogenetic protein-4 is required for mesoderm formation and patterning in the mouse. Genes Dev 9:2105–2116; 1995. 41. Wozney, J. M., Rosen, V., Celeste, A. J., Mitsock, L. M., Whitters, M. J., Kriz, R. W., Hewick, R. M., and Wang, E. A. Novel regulators of bone formation: Molecular clones and activities. Science 242:1528 –1534; 1988. 42. Yang, Y., Mumy, M., Romeo, D., and Wakefield, L. M. Identification of the start sites for the 1.9- and 1.4-kb rat transforming growth factor-beta1 transcripts and their effect on translational efficiency. Gene 219:81– 89; 1998.

Date Received: August 4, 2000 Date Revised: February 13, 2001 Date Accepted: February 14, 2001