Introduction

Bone morphogenetic proteins (BMPs) are multifunctional growth factors that are members of the transforming growth factor-β family [1]. They exhibit unique activity in bone matrix that is characterized by ectopic bone formation in muscle tissues in vivo [2]. BMPs inhibit the myogenic differentiation of myoblasts and induce osteogenic differentiation into osteoblastic lineage cells in vitro [3]. They physiologically regulate not only bone formation but also the development and regeneration of various tissues in vertebrates and lower animals [4].

BMP signaling is transduced by two different transmembrane serine/threonine kinase receptors—the type I and type II receptors [5, 6]. The BMP-bound type II receptor kinase phosphorylates the type I receptor. The activated BMP type I receptor kinase subsequently phosphorylates a serine–valine–serine motif at the C-termini of Smad1, Smad5 and Smad8, which are BMP receptor-regulated Smads (R-Smads). The phosphorylated R-Smads form heteromeric complexes with Smad4 and translocate into the nucleus to directly regulate the transcription of target genes.

The Id family members are negative regulators of myogenesis that have a common helix–loop–helix structure, and their genes have been shown to be typical early response genes of BMP signaling in various cell types [7–9]. A GC-rich 29-bp element in the 5′ enhancer region of the Id1 gene was identified as the BMP-responsive element (BRE), which is recognized by a complex of Smad1/5 and Smad4 in response to the activation of the BMP type I receptor [7]. Mad, a Drosophila homolog of mammalian Smad1, also binds to a GC-rich sequence in target genes [10, 11]. Additional GC-rich elements have been found in the target genes of BMPs [10–14].

In the present study, we analyzed functional BREs in the mouse Id2 and Id3 genes. We identified the conserved GGCGCC sequence in the BREs of the Id1-Id3 genes, which are recognized by Smad1 and Smad4 in response to BMP-4. Moreover, we found a novel BRE by searching the mouse genome for a region homologous to the Id1 BRE. This novel BRE regulates the transcription of a novel non-coding BMP-inducible transcript, termed BIT-1, that is mainly expressed in the testis, brain, kidney and cartilage in mice.

Materials and methods

Plasmid constructs

To construct the luciferase reporter plasmids, 3.1-, 5.7- and 0.8-kb fragments from the 5′ regions of the mouse Id2 and Id3 genes and BIT-1, respectively, were inserted between the KpnI and EcoRV sites of pGL4.14 (Promega, Madison, WI, USA) by a standard polymerase chain reaction (PCR) method using PrimeStar HS DNA polymerase (TaKaRa, Shiga, Japan) with mouse genomic DNA as a template. The primer sequences were as follows: 5′-atggtaccCTTCTGCCAAGGAAGAGTTC-3′ (mId2-5KpnI), 5′-atgatatcACCAGGCTCGGTTCAGAATG-3′ (mId2-3EcoRV), 5′-atggtaccGAGTGCTGAGACTGTTAAGTAT-3′ (mId3-5KpnI), 5′-atgatatcCCTAAAGCAAACAGTGC-3′ (mId3-3EcoRV), 5′-atggtaccGCACAAGTAGTTATGGAGAC-3′ (BIT-1-5KpnI) and 5′-atgatatcACAGAACCTCGGAGACTGAC-3′ (BIT-1-3EcoRV) (lowercase letters indicate the flanking sequence). Deletion mutants were generated by PCR using each specific upper primer and the mId2-3EcoRV and mId3-3EcoRV primers with Id2(3.1)-luc and Id3(5.7)-luc as templates, respectively. To generate Id2WT4F-luc, Id3WT4F-luc, Id2WT4R-luc, Id3WT4R-luc, Id2MUT4F-luc, Id3MUT4F-luc and the novel BREWT4F-luc, four tandem repeats of the specific wild-type or mutant DNA sequences of the 34-bp Id2 BRE, 35-bp Id3 BRE or 29-bp novel BRE were synthesized (TaKaRa). These repeat sequences were then subcloned into the HindIII recognition site of pGL4.26 (Promega). The sequences of the BREs are as follows: 5′-TTGCTATGGCAGCCGCCTGAGCGGCGCCGCGAGG-3′ (Id2WT), 5′-TTGCTATGGCAGCCGCCTGAGCTTTGCCGCGAGG-3′ (Id2MUT), 5′-ATGACGTCCCACCCTGGCGCCAGGCTGTCTGGGGC-3′ (Id3WT), 5′-ATGACGTCCCACCCTTTTGCCAGGCTGTCTGGGGC-3′ (Id3MUT) and 5′-CAGAGCGACCGCCCGGCCGGCGCCAGCAG-3′ (novel BRE) (the underlined regions indicate mutations). Id2-2867MUT-luc and Id3-3304MUT-luc were constructed via a standard PCR technique using the following primers: 5′-CAGCCGCCTGAGCTTTGCCGCGAGGACAAG-3′ (Id2-2867MUT-luc sense), 5′-CTTGTCCTCGCGGCAAAGCTCAGGCGGCTG-3′ (Id2-2867MUT-luc anti-sense), 5′-CGTCCCACCCTTTTGCCAGGCTGTCTGGGG-3′ (Id3-3303MUT-luc sense) and 5′-CCCCAGACAGCCTGGCAAAAGGGTGGGACG-3′ (Id3-3303MUT-luc anti-sense). Mouse Smad1 and Smad4 expression vectors have been described previously [7]. Mouse Rp58 (GenBank ID 30928) was obtained using standard reverse transcriptase (RT)-PCR techniques with PrimeStar HS DNA polymerase (TaKaRa) and cloned into the pcDEF3 expression vector [15]. The DNA sequence of each construct was confirmed by DNA sequencing using an ABI3500 Genetic Analyzer (Applied Biosystems, Foster City, CA, USA).

Cell culture, transfection and luciferase assay

C2C12 mouse myoblasts were maintained as previously described [3]. Primary osteoblasts were prepared from newborn mouse calvaria as described previously [16]. Cells in a 96-well plate were transfected with 200 ng/well of plasmid DNA using 0.5 μl/well of Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) as described previously [7]. The luciferase reporter assay was performed using the Dual-Glo Luciferase Assay system (Promega), luciferase reporter plasmids and phRL-SV40 (Promega) in the presence and absence of 10 ng/ml BMP-4 (R&D Systems, Minneapolis, MN, USA) with or without 100 nM LDN193189 (Stemgent, San Diego, CA, USA) for 24 h as previously described [7].

Electrophoretic mobility shift assay

Electrophoretic mobility shift assays (EMSAs) were performed as previously described [7, 17]. Oligonucleotides were labeled with the Biotin 3′ End DNA Labeling Kit (Thermo, Rockford, IL, USA). The DNA binding reaction was performed for 20 min at room temperature in 20 μl of binding buffer (10 mM Tris–HCl, pH 7.5; 50 mM KCl; 5 mM MgCl2; 1 mM dithiothreitol; 0.05 % Nonidet P-40; 2.5 % glycerol), 1 μg of poly(dI-dC), 100 fmol of biotin-labeled DNA and 10 μg of nuclear extract proteins. For supershift experiments, antibodies were added prior to the addition of the probe and incubated for 1 h at 4 °C. Antibodies against Smad1 (Invitrogen) and Smad4 (B-8; Santa Cruz Biotechnology, Santa Cruz, CA, USA) were used. For the competition experiments, the unlabeled double-stranded oligonucleotides were added to the mixtures at room temperature 30 min prior to the addition of the labeled probe. The sequences of the oligonucleotides used are as follows: 5′-CATGGCGACCGCCCGCGCGGCGCCAGCCT-3′ (Id1 wild-type BRE), 5′-CATGGCGACCGCCCGCGCTTTGCCAGCCT-3′ (Id1 mutant BRE), 5′-TTGCTATGGCAGCCGCCTGAGCGGCGCCGCGAGG-3′ (Id2 wild-type BRE), 5′-TTGCTATGGCAGCCGCCTGAGCTTTGCCGCGAGG-3′ (Id2 mutant BRE), 5′-ATGACGTCCCACCCTGGCGCCAGGCTGTCTGGGGC-3′ (Id3 wild-type BRE), 5′-ATGACGTCCCACCCTTTTGCCAGGCTGTCTGGGGC-3′ (Id3 mutant BRE) and 5′-CAGAGCGACCGCCCGGCCGGCGCCAGCAG-3′ (wild-type novel BRE) (the underlined regions indicate mutations). The samples were subjected to electrophoresis on a native 5 % acrylamide/0.5 × Tris-borate/EDTA gel and transferred to a nylon membrane. The biotin-labeled DNA was detected with the LightShift Chemiluminescent EMSA kit (Thermo).

Rapid amplification of cDNA ends

Rapid amplification of cDNA ends (3′ RACE) was performed using the 3′-Full RACE Core Set (TaKaRa). A library of adapter-ligated double-stranded cDNA was constructed from 1 μg of total RNA isolated from BMP-4-treated C2C12 cells according to the manufacturer’s instructions. To isolate the 3′-end of the BIT-1 cDNA, primer 1 (5′-GCTGGGGGGTCAGTCTCCGA-3′) and primer 2 (5′-CTTCAAGACTCCTCCTGAGG-3′) were designed from the sequence of the 3′-region of BIT-1. PCR reactions were performed according to the manufacturer’s instructions using AMV Reverse Transcriptase XL (TaKaRa). The PCR product was subcloned into T-Vector pMD20 (TaKaRa) and sequenced using an ABI3500 Genetic Analyzer (Applied Biosystems).

RT-PCR analysis

Total RNA was prepared using TRIzol Reagent (Invitrogen) and then reverse transcribed with SuperScript III reverse transcriptase (Invitrogen) according to the manufacturer’s instructions. PCR reactions were performed with GoTaq (Promega). The primer sets used are as follows: 5′-GGGTGTTGCTGGACAGTGCC-3′ (BIT-1 sense), 5′-CACTACCCTGAGCACCACCC-3′ (BIT-1 anti-sense), 5′-TCCTGCAGCATGTAATCGAC-3′ (mouse Id1 sense), 5′-GAGAGGGTGAGGCTCTGTTG-3′ (mouse Id1 anti-sense), 5′-CTCCAAGCTCAAGGAACTGG-3′ (mouse Id2 sense), 5′-CGCAACCCACACAGAACTTA-3′ (mouse Id2 anti-sense), 5′-CTCTTGGACGACATGAACCAC-3′ (mouse Id3 sense), 5′-AGTGAGCTCAGCTGTCTGGAT-3′ (mouse Id3 anti-sense), 5′-GAGAGGGAAATCGTGCGTGA-3′ (mouse β-actin sense) and 5′-ACATCTGCTGGAAGGTGGAC-3′ (mouse β-actin anti-sense).

Chromatin immunoprecipitation

Chromatin immunoprecipitation (ChIP) was performed with a ChIP assay kit (Cell Signaling, Beverly, MA, USA) according to the manufacturer’s instructions using antibodies against Smad1 (D59D7; Cell Signaling) and Smad4 (Cell Signaling). The purified DNA was analyzed by PCR using the following primers that detect sequences in the indicated promoter: 5′-TTCGCGCCCTAAGTCTGCAGGTG-3′ (Id1 promoter sense), 5′-AGGCCCGTGACGTCACCCATTC-3′ (Id1 promoter anti-sense), 5′-CCTTGCAGGCATTGATCAGCTG-3′ (Id2 promoter sense), 5′-GAGCCCCGGAGCAGACTCTCT-3′ (Id2 promoter anti-sense), 5′-GCTATTCACAAGCCGCTCGCCG-3′ (Id3 promoter sense), 5′-GCCCAGCTGTGGGTCCGGGTCC-3′ (Id3 promoter anti-sense), 5′-CCAAGCCTTGTTGCCTGTGAACG-3′ (BIT-1 promoter sense), and 5′-ACAGCTGATTCATTGGCTCGGC-3′ (BIT-1 promoter anti-sense).

Statistical analysis

Comparisons were made using an unpaired Student’s t test; the results are shown as the means ± SD. Statistical significance is indicated as *p < 0.05 and **p < 0.01.

Results

Identification of functional BREs in the mouse Id2 and Id3 genes

To examine the transcriptional mechanism of BMP early response genes, we analyzed the BREs in the mouse Id2 gene using a series of luciferase reporters harboring fragments of the 5′-flanking region of the mouse Id2 gene. For Id2(3.1)-luc, but not Id2(2.7)- or Id2(2.0)-luc, luciferase was induced in response to BMP-4 in C2C12 cells (Fig. 1a). The response to BMP-4 was dependent on the 34-bp region between positions −2867 and −2821 in the Id2 gene (5′-TTGCTATGGCAGCCGCCTGAGCGGCGCCGCGAGG-3′), which contains the GGCGCC sequence identified in the BRE of the Id1 gene (Fig. 1b) [7–9]. In both the forward and reverse orientations, the 34-bp region of the Id2 gene was sufficient to respond to BMP-4 (Fig. 1c). Moreover, mutation of GGCGCC to TTTGCC in the 34-bp region reduced the response to BMP-4 (Fig. 1d), suggesting that the 34-bp region functions as the BRE in the Id2 gene.

Fig. 1
figure 1

Identification of a BRE in the mouse Id2 gene. a and b Deletion analysis of the 5′ region of the mouse Id2 gene in luciferase reporter constructs. Deletion constructs were generated from Id2(3.1)-luc. c Characterization of the 34-bp region of the mouse Id2 gene as a BMP-responsive enhancer. Reporter constructs harboring four copies of the 34-bp element with a wild-type or mutant sequence in the forward (Id2WT4F-luc and Id2MUT4F-luc) or reverse orientation (Id2WT4R-luc) were generated. d The role of the GGCGCC sequence of the Id2 BRE in the BMP response. The GGCGCC sequence in the BRE was substituted with TTTGCC in Id2(2867)-luc. Luciferase expression was determined in C2C12 cells in the presence and absence of 10 ng/ml BMP-4. Data are reported as the means ± SD (n = 3)

Full size image

Using the same strategy, we analyzed a BRE within a 5.7-kb 5′ -flanking region of the mouse Id3 gene (Fig. 2). The response of the Id3 gene to BMP-4 was highly dependent on the 35-bp region between positions −3299 and −3265 (5′-ATGACGTCCCACCCTGGCGCCAGGCTGTCTGGGGC-3′), which also contains the GGCGCC sequence (Fig. 2a, b). Similar to the Id1 and Id2 genes, the GGCGCC sequence in the Id3 gene was both necessary and sufficient to mediate the response to BMP-4 (Fig. 2c, d), suggesting that this 35-bp region is the BRE in the Id3 gene.

Fig. 2
figure 2

Identification of a BRE in the mouse Id3 gene. a and b Deletion analysis of the 5′ region of the mouse Id3 gene in luciferase reporter constructs. Deletion constructs were generated from Id3(5.7)-luc. c Characterization of the 35-bp region of the mouse Id3 gene as a BMP-responsive enhancer. Reporter constructs containing four copies of the 35-bp element with a wild-type or mutant sequence in the forward (Id3WT4F-luc and Id3MUT4F-luc) or reverse orientation (Id3WT4R-luc) were constructed. d The role of the GGCGCC sequence of the Id3 BRE in the BMP response. The GGCGCC sequence in the BRE was substituted with TTTGCC in Id3(3304)-luc. Luciferase expression was determined in C2C12 cells in the presence and absence of 10 ng/ml BMP-4. Data are reported as the means ± SD (n = 3)

Full size image

Smad1 and Smad4 cooperatively induce the transcription of the Id2 and Id3 genes

Next, we analyzed the roles of Smad1 and Smad4 in the transcription of the Id2 and Id3 genes. Using EMSAs, we demonstrated that BMP-4 treatment induced the formation of a DNA–protein complex with both the Id2 and Id3 BREs (Fig. 3a, b). The DNA–protein complexes that formed with the BREs were supershifted upon the addition of antibodies against Smad1 and/or Smad4 (Fig. 3a, b). The DNA–protein complex that formed with the Id1 BRE was competed away by the addition of excess unlabeled DNA of not only the wild-type Id1 BRE but also the wild-type Id2 and Id3 BREs; however, the BREs with mutated GGCGCC sequences failed to compete (Fig. 3c). A ChIP assay demonstrated the binding of endogenous Smad1 and Smad4 to the BREs in the Id2 and Id3 genes in response to BMP-4 (Fig. 3d). The overexpression of Smad1 stimulated the expression of luciferase from both the Id2WT4F-luc reporter and the Id3WT4F-luc reporter (Fig. 3e, f). Co-transfection with Smad4 and Smad1 expression vectors enhanced luciferase expression in C2C12 cells (Fig. 3e, f).

Fig. 3
figure 3

Smad1 and Smad4 bind to BREs in the Id2 and Id3 genes. a and b The DNA–protein complexes formed with the Id2 and Id3 BREs contain Smad1 and Smad4. Nuclear extracts were prepared from C2C12 cells treated for 1 h with or without 100 ng/ml BMP-4. EMSAs were performed using the Id2 a or Id3 b BRE as a probe. Antibodies against Smad1 (S1) and Smad4 (S4) alone or in combination were added to the reaction mixtures. An arrow from left to right indicates BMP-induced complexes. S1, S4 and S1/S4 indicate supershifted bands with the indicated antibody. Asterisks indicate non-specific bands. c Sequence specificity of the BREs in the Id1, Id2 and Id3 genes. EMSAs were performed using nuclear extracts prepared from C2C12 cells treated with BMP-4 and the Id1 BRE as a probe. Excess unlabeled Id1, Id2 or Id3 BRE with a wild-type (WT) or mutant (Mut) GCCGCC sequence was added to the mixtures. An arrow from left to right indicates BMP-induced complexes. Asterisks indicate non-specific bands. d ChIP assays demonstrate that Smad1 and Smad4 bind to the Id2 and Id3 BREs in response to BMP-4 stimulation in C2C12 cells. e and f Smad1 and Smad4 activate the expression of Id2 and Id3 via BREs. Id2WT4F-luc e and Id3WT4F-luc f were co-transfected with Smad1 in the presence or absence of Smad4 in C2C12 cells. Data are reported as the means ± SD. (n = 3). **p < 0.01 compared with cells transfected with empty vector and Smad1, Smad1 alone or Smad4 with Smad1

Full size image

Effects of Rp58 on the transcription of Id1, Id2 and Id3

Recently, Rp58 (Zfp238) was identified as a transcriptional repressor of Id2 and Id3 during myogenesis [18]. We examined the effect of the relationship between the Smads and Rp58 on the transcription of BMP-responsive genes, including Id1, Id2 and Id3, using luciferase reporter constructs driven by each native promoter/enhancer region [Id1(2.1)-luc, Id2(3.1)-luc and Id3(5.7)-luc, respectively]. As reported previously, the overexpression of Rp58 suppressed BMP-4-induced luciferase expression from Id2(3.1)-luc and Id3(5.7)-luc, but Rp58 did not suppress expression from Id1(2.1)-luc (Fig. 4a–c). The consensus sequences for Rp58 were located at positions +174 and +16 of the Id2 and Id3 BREs, respectively, but there was no Rp58 site in proximity to the Id1 BRE (Fig. 4d). An unexpected stimulation, rather than inhibition, of the Id1(2.1)-luc luciferase reporter by Rp58 was also observed with empty pGL4.26 vector (2.0-fold, data not shown).

Fig. 4
figure 4

Rp58 suppresses the expression of Id2 and Id3 but not of Id1. a–c Rp58 suppresses Id2(3.1)-luc b and Id3(5.7)-luc c but not Id1(2.1)-luc a. C2C12 cells were transfected with one of the reporter constructs with or without Rp58, and luciferase expression was determined in the presence or absence of BMP-4. Data are reported as the mean ± SD (n = 3). d A scheme of the relationship between the BREs (solid squares) and the Rp58 sites (open circles) in the Id1(2.1)-luc, Id2(3.1)-luc and Id3(5.7)-luc reporter constructs. *p < 0.05; **p < 0.01

Full size image

Identification of a novel BMP-responsive gene containing the conserved BRE

Due to the high level of conservation of the GC-rich BRE sequences among the Id1, Id2 and Id3 genes, we performed a BLAST (http://blast.ncbi.nim.njh.gov/Blast.cgi) search on the mouse genome to identify novel genes that have potential BREs. Using the Id1 BRE as the query sequence, we found a highly conserved sequence (23 of 29 bases identical) on chromosome 13 at position 47,723,740-47,723,768, which showed 79.3, 62 and 45 % homology with the BREs of Id1, Id2 and Id3, respectively (Fig. 5a). We generated a luciferase reporter construct containing four tandem copies of this potential novel BRE upstream of the minimal promoter (Fig. 5b). This reporter was activated in response to BMP-4, and this activation was suppressed by a chemical inhibitor of BMP receptors (Fig. 5b). Using EMSAs and ChIP assays, we demonstrated that this novel BRE was recognized by Smad1 and Smad4 in response to BMP-4 (Fig. 5c, d). Moreover, the overexpression of Smad1 in the presence of Smad4 induced luciferase expression in C2C12 cells (Fig. 5e).

Fig. 5
figure 5

Identification of a novel BRE in the mouse genome. a Comparison of the novel potential BRE with the BREs of Id1, Id2 and Id3. Underlining indicates the novel potential BRE identified in the mouse genome on chromosome 13 at position 47,723,740–47,723,768. Nucleotides that are conserved between the novel potential BRE and the Id1 BRE are indicated by asterisks. b Functional analysis of the novel BRE. A reporter construct containing four copies of the novel BRE was transfected into C2C12 cells, and luciferase expression was determined in the presence or absence of 10 ng/ml BMP-4 with or without 100 nM LDN193189. Data are reported as the means ± SD (n = 3). c Smad1 and Smad4 bind to the novel BRE. EMSAs were performed using the novel BRE as a probe and nuclear extracts of C2C12 cells treated with BMP-4. Antibodies against Smad1 or Smad4 were added to the mixture. An arrow indicates BMP-induced complexes. S1, S4 and S1/S4 indicate supershifted bands with the indicated antibody. Asterisks indicate non-specific bands. d ChIP assays demonstrate that Smad1 and Smad4 proteins bind to the novel BRE and the Id1 BRE in response to BMP-4 stimulation of C2C12 cells. e Smad1 and Smad4 induce the transcriptional activity of BRE. C2C12 cells were co-transfected with the 4× novel BRE luciferase reporter and Smad1 in the presence or absence of Smad4. Data are reported as the means ± SD (n = 3). **p < 0.01 compared with cells transfected with empty vector and Smad1. *p < 0.05 compared with cells transfected with Smad1 alone or with Smad4 and Smad1

Full size image

A novel BMP-inducible transcript

Our findings suggest that the novel BRE regulates the transcription of an unidentified gene in response to BMP signaling. However, there were no genes that mapped within 10 kb of the novel BRE in the RefSeq database. We found that the novel BRE was located 255 bp upstream from a single cap analysis gene expression (CAGE) tag in the FANTOM database (TC ID T13F02D835DB; TU ID 83506 forward strand) (Fig. 6a, c). BMP-4 significantly stimulated luciferase expression from a reporter construct containing the genomic sequence from positions −805 to +49 relative to the CAGE tag, a region that includes the novel BRE, and this stimulation was suppressed by the BMP receptor inhibitor LDN-193189 (Fig. 6a). Furthermore, the corresponding endogenous transcript containing the CAGE tag, termed BIT-1, was induced in C2C12 cells within 30 min after treatment with BMP-4 (Fig. 6b). We determined the sequence of full-length BIT-1 from the CAGE tag in C2C12 cells cultured in the presence of BMP-4 using 3′ RACE (Fig. 6c). RT-PCR with a primer set that included the CAGE tag and the 3′ end sequence yielded the expected 1.0-kb single band, which was confirmed by direct DNA sequencing (Fig. 6d). The sequence data for BIT-1 have been deposited in the DDBJ/EMBL/GenBank databases under accession number AB679283. BIT-1 appears to be a non-coding RNA because the first open reading frame from the CAGE tag only encodes 6 amino acids and ATG codons do not correspond to the Kozak consensus sequence (Fig. 6c). Among the mouse tissues, BIT-1 was detected in the testis = brain > cartilage = kidney > and lung, in this order, and this expression pattern was quite different from that of other BMP-responsive genes, such as Id1, Id2 and Id3 (Fig. 6e). Treatment with BMP-4 induced the expression of BIT-1 mRNA and the luciferase reporter not only in C2C12 cells but also in primary osteoblasts (Supplemental Fig. 1). We did not detect any changes in the induction of ALP activity, Id1WT4F-luc expression or the inhibition of myogenesis by BMP signaling with a transient transfection of C2C12 cells with BIT-1 cDNA (data not shown).

Fig. 6
figure 6

The novel BRE regulates the transcription of BIT-1. a A luciferase reporter construct harboring the novel BRE was activated by BMP-4. A DNA fragment of mouse chromosome 13 position 47,723,190–47,724,044 containing the novel BRE (open arrow) and the CAGE tag (triangle) (−805 to +49) was cloned into the luciferase reporter vector pGL4.14 and transfected into C2C12 cells. Luciferase expression was determined in the presence or absence of BMP-4 with or without 100 nM LDN193189. Data are reported as the mean ± SD (n = 3). b RT-PCR analysis. C2C12 cells were treated with 100 ng/ml BMP-4 for the indicated time periods. The BIT-1, Id1 and actin mRNAs were amplified by RT-PCR. c Nucleotide sequence of the novel BRE and BIT-1. The novel BRE identified (open arrow), the CAGE tag (open box), and the 3′ end as determined by 3′ RACE (solid triangle) are indicated. The first nucleotide of the CAGE tag is denoted by +1. d Induction of BIT-1 by BMP-4 in C2C12 cells. The expected full-length BIT-1 was amplified by RT-PCR using primers that bind the CAGE tag and the 3′ end. e Expression of BIT-1, Id1, Id2 and Id3 in mouse tissues. RT-PCR analysis was performed using cDNA prepared from the following mouse tissues: C cartilage, Lu lung, S spleen, H heart, Bo bone, M muscle, T testis, K kidney, Li liver, Br brain, I intestine and St stomach. **p < 0.01 compared with control

Full size image

Discussion

In the present study, we analyzed the functional BREs in the mouse Id2 and Id3 genes, which are typical early response genes in BMP signaling. Both BREs were found in the 5′-flanking region of each gene; the Id2 BRE was located between positions −2854 and −2820, and the Id3 BRE was located between positions −3299 and −3625. Both mouse BREs were conserved at the corresponding positions in the human Id2 (97 %) and Id3 (100 %) genes (data not shown), suggesting that these regions are important for transcriptional regulation. The BREs in the Id1, Id2 and Id3 genes possess a highly conserved GGCGCC core sequence, and this sequence was recognized by a complex containing Smad1 and Smad4. We previously reported that mutation of the Id1 BRE from GGCGCC to TTTGCC destroys the response to BMP signaling by interrupting the binding of Smads to DNA [7].

Recently, other groups have also identified regions similar to the BREs in the mouse Id2 and human Id3 genes [19, 20]. ChIP-seq analysis using an antibody against Smad1/5 also identified two GC-rich Smad-binding elements in human endothelial cells, the GGCGCC sequence and a GGAGCC sequence [21]. Taken together, all of these findings indicate the importance of the GGCGCC sequence in the BMP response in mammalian cells. Interestingly, although two GGCGCC sequences were identified within a 60-bp region in the mouse Id2 gene, deleting only the region distal to the native promoter/enhancer abrogated the response to BMP-4. These results suggest that the GGCGCC sequence is essential, but not sufficient, for the BMP response. Additional sequences surrounding this core sequence may modify the binding affinity of Smads and/or other co-activators. Although a role for the CAGA (or complement GTCT) box in the BMP response has been suggested, our artificial BRE reporters, Id1WT4F-luc and Id2WT4F-luc, respond to BMP-4 although they lack the CAGA/GTCT box, suggesting that this element may not be essential for the response to BMP signaling. Additional studies are required to clarify the synergism between these sequences in BMP-regulated transcription.

We found a novel BMP-inducible transcript, BIT-1, by searching a genome database using the 29-bp Id1 BRE; the sequence was chosen because not only the core GGCGCC sequence but also the surrounding sequence may be required for the BMP response. The BIT-1 BRE exhibited the highest degree of homology with the Id1 BRE (79.3 %). BIT-1 seems to be one of the BMP-induced non-coding RNAs that include BMP/OP-resposive gene [22]. BIT-1 may have physiological roles in tissue development that are regulated by BMPs, as suggested by the fact that BIT-1 was found to be expressed in the target organs of BMPs, such as the testis, brain, kidney and cartilage. However, transient overexpression of the BIT-1 cDNA in C2C12 cells did not change the typical responses induced by BMPs, including alkaline phosphatase (ALP) activity, Id-luc expression or myogenesis (data not shown). Cell specificity may be involved in the functions of BIT-1. Gain-of-function and/or loss-of-function analyses in vivo will be required to elucidate the physiological roles of BIT-1. We have searched for novel BREs and BIT-1 in other species. Interestingly, whereas the novel BRE, including the GC-core sequence, is highly conserved not only in mouse but also in human, rat, orangutan, dog, horse and chicken, the conserved BIT-1 sequence is found in mouse and rat, but not in other species. This finding suggests that BIT-1 may function in rodents and that the conserved BRE in other species may regulate the expression of other transcripts in response to BMP signaling.

Recently, Rp58 has been identified as a transcriptional repressor of Id2 and Id3, which are negative regulators of the MyoD family of transcription factors [18]. The expression of Rp58 increased during myogenesis in C2C12 cells [18]. We confirmed the finding that Rp58 suppressed the transcription of the Id2 and Id3 genes via the consensus sites in the 5′-flanking region of each gene. We observed that these Rp58 sites were located close to the BREs in both the Id2 and Id3 genes. The luciferase reporters lacking the Rp58 sites were not suppressed by the overexpression of Rp58, suggesting that Rp58 inhibits the transcription of Id2 and Id3 induced by BMP signaling by binding to the target promoter/enhancer regions. The molecular mechanism of this inhibition will be elucidated by analyzing the components that are recruited to the consensus site by Rp58.

In conclusion, we identified functional BREs in BMP early response genes containing the core GGCGCC sequence. A novel BMP-inducible transcript, BIT-1, may play a role in BMP-regulated tissue development.