Estimating bloodstain age in the short term based on DNA fragment length using nanopore sequencer.

We used a nanopore sequencer to quantify DNA fragments > 10,000 bp in size and then evaluated their relationship with short-term bloodstain age. Moreover, DNA degradation was investigated after bloodstains were wetted once with water. Bloodstain samples on cotton gauze were stored at room temperature and low humidity for up to 6 months. Bloodstains stored for 1 day were wetted with nuclease-free water, allowed to dry, and stored at room temperature and low humidity for up to 1 week. The proportion of fragments > 20,000 bp in dry bloodstains tended to decrease over time, particularly for fragments > 50,000 bp in size. This trend was modeled using a power approximation curve, with the highest R2 value (0.6475) noted for fragments > 50,000 bp in size; lower values were recorded for shorter fragments. The proportion of longer fragments was significantly reduced in bloodstains that were dried after being wetted once, and there was significant difference in fragments > 50,000 bp between dry conditions and once-wetted. This result suggests that even temporary exposure to water causes significant DNA fragmentation, but not extensive degradation. Thus, bloodstains that appear fresh but have a low proportion of long DNA fragments may have been wetted previously. Our results indicate that evaluating the proportion of long DNA fragments yields information on both bloodstain age and the environment in which they were stored.

Simplified detection of the species of origin of antler velvets using single-stranded tag hybridization chromatographic printed-array strip.

In this study, we developed a convenient and easy-to-use origin identification method for antler velvets based on a simple DNA extraction technique and single-stranded tag hybridization chromatographic printed-array strip (STH-PAS). The primer sets used to detect Cervus elaphus, Rangifer tarandus, and 12S rRNA did not engage in non-specific reactions such as primer dimer formation. In both the triplex and singleplex assays, the sensitivity was < 1 ng DNA. Moreover, Cervus elaphus DNA could be detected in OTC crude drug products. Although the detection sensitivity resulting from the simplified extraction was slightly lower than that obtained with extraction by conventional methods, the amount of DNA was sufficient even from a small sample. The choice of a triplex or singleplex assay will depend on the purpose of the test. For example, if it is important to determine whether the antler velvet is derived from Cervus elaphus or Rangifer tarandus, a triplex assay is appropriate. If it is necessary to explore whether antler velvet from Cervus elaphus is included in an OTC crude drug product, a singleplex assay using the Cervus elaphus primer set is informative. If it is necessary to explore whether powdered antler velvet includes counterfeit products (from Rangifer tarandus), a singleplex assay employing the Rangifer tarandus primer is appropriate. The singleplex assay detects minor components even at a 1,000:1 ratio. Our study thus demonstrated the utility of a method combining simple DNA extraction with STH-PAS for efficient identification of the origin of antler velvets.

A method for determining the origin of crude drugs derived from animals using MinION, a compact next-generation sequencer.

We evaluated whether MinION, an inexpensive portable sequencer, can be used to identify the origin of crude drugs derived from animals. Standard and nonstandard crude drugs with different species of origin were examined. In addition, standards mixed with nonstandard samples were used. As a target gene, cytochrome c oxidase I was amplified and sequenced. The fast mode results had a slightly lower match ratio than high-accuracy mode, but the animals of origin were correctly determined by BLAST for all samples. For antler velvet derived from Rangifer tarandus, even when the sequences were aligned based on Cervus elaphus, the animal of origin was determined correctly. Minor contents could be detected from mixtures of two animals, if the mixtures contained at least 19:1 mtDNA when the coverage allele-fraction threshold was 0.05. By contrast, in fast mode, two sequences could not be separated due to the low accuracy of the base-calling for each read. For fieldwork, the species of origin of crude drugs could be identified with only simple DNA extraction and library preparation. Therefore, MinION appears to be a convenient tool for identifying the origins of crude drugs derived from animals.

A sudden infant death associated with atrial septal aneurysm and abnormality of the connecting site between the inferior vena cava and right atrium.

An atrial septal aneurysm (ASA) is a rare cardiac anomaly characterized by varicose bulgingofthe atrial septum (oval fossa) into the left or right atrium. Pathogenesis and clinical significance of ASA are controversial. We report an autopsy case of a huge undiagnosed ASA with abnormality of the connecting site between the inferior vena cava and the right atrial ostium in a 2-month-oldJapanesefemale who died suddenly and unexpectedly. She was born at 36 weeks 4 days (body weight 3,110 g). No abnormality was detected during pregnancy or delivery. The postnatal growth was normal with no cardiac problem detected at the 1-month checkup. The ASA bulged off in a mass to the left atrium (width, 0.8 cm; excursion ratio, 53%), reaching close to the inflow site of the right pulmonary vein, with dilation of the pulmonary vein. The connecting site between the inferior vena cava and the right atrium was atypically located 1.6 cm away from the atrioventricular groove. Although most cases of ASA in an infant resolve physiologically as the infant grows, the infant in the present case is thought to have had an exceptional pathological ASA, possibly causing supraventricular arrhythmia. The abnormality of the connecting site between the inferior vena cava and the right atrium might have affected the development and continuation of the ASA.

Diagnostic significance of the histopathology of bone marrow macrophages in forensic autopsies.

Forensic pathologists often encounter autopsies that require an assessment of antemortem general conditions (e.g., infection, metabolic disorders). To establish evaluation clues for such cases, we quantitatively examined macrophages and the general pathology of bone marrow in samples from 180 forensic autopsy cases of decedents with various conditions. Hematoxylin-eosin staining, Berlin blue staining, and immunostainings for CD163, CD138, and CD61 were performed. We determined the numbers per field (density) of total macrophages, swollen macrophages, macrophages with hemophagocytosis, and hemosiderin-laden macrophages. Each density was standardized by identifying its ratio to the total number of macrophages. The decedents' background data (cause of death, other pathological findings, postmortem interval, antemortem symptoms, and presence of resuscitation) were extracted. No correlations were found between the postmortem interval and the other decedent data, indicating that these data are not affected by postmortem changes. In the group in which inflammatory disease was the cause of death, there were significant elevations in the ratio of the swollen macrophage density to total macrophages. Significantly higher ratios of the density of swollen and hemophagocytic macrophages were observed in the group in which conditions with a prolonged agonal period were the cause of death. The group with a return of spontaneous circulation to resuscitation showed a significantly higher ratio of macrophage density with hemophagocytosis. This study provides the first statistical analysis focused on bone marrow histopathology in forensic autopsies. The results will be useful for elucidating causes of death and agonal-period conditions.

Estimating individual mtDNA haplotypes in mixed DNA samples by combining MinION and MiSeq.

We tried to estimate individual mtDNA haplotypes in mixed DNA samples by combining MinION and MiSeq. The BAM files produced by MiSeq were viewed using Integrative Genomics Viewer (IGV) to verify mixed bases. By sorting the reads according to base type for each mixed base, partial haplotypes were determined. Then, the BAM files produced by MinKNOW were viewed using IGV. To determine haplotypes with IGV, only mixed bases determined by MiSeq were used as target bases. By sorting the reads according to base type for each target base, each contributor's haplotype was estimated. In mixed samples from two contributors, even a haplotype with a minor contribution of 5% could be distinguished from the haplotype of the major contributor. In mixed samples of three contributors (mixture ratios of 1:1:1 and 4:2:1), each haplotype could also be distinguished. Sequences of C-stretches were determined very inaccurately in the MinION analysis. Although the analysis method was simple, each haplotype was correctly detected in all mixed samples with two or three contributors in various mixture ratios by combining MinION and MiSeq. This should be useful for identifying contributors to mixed samples.

Estimating included animal species in mixed crude drugs derived from animals using massively parallel sequencing.

We developed a method that can detect each animal species of origin for crude drugs derived from multiple animal species based on massively parallel sequencing analysis of mitochondrial genes. The crude drugs derived from animals investigated in this study were Cervi Parvum Cornu and Trogopterorum feces, which are derived from a mix of different animal species, two chopped cicada sloughs, and two commercial Kampo drugs. The mitochondrial 12S rRNA, 16S rRNA, and cytochrome oxidase subunit I gene regions were amplified and sequenced using MiSeq. The ratios of haplotype to total number of sequences reads were calculated after sequence extraction and trimming. Haplotypes that exceeded the threshold were defined as positive haplotypes, which were compared with all available sequences using BLAST. In the Cervi Parvum Cornu and Trogopterorum feces samples, the haplotype ratios corresponded roughly to the mixture ratios, although there was a slight difference from mixture ratios depending on the gene examined. This method could also roughly estimate the compositions of chopped cicada sloughs and Kampo drugs. This analysis, whereby the sequences of several genes are elucidated, is better for identifying the included animal species. This method should be useful for quality control of crude drugs and Kampo drugs.

Bloodstain examination and DNA typing from hand-washed bloodstains on clothes.

We investigated whether bloodstain examination and DNA typing can be performed on washed bloodstains on clothes. Blood was dropped onto T-shirts made from 100% cotton or 100% polyester. After drying, the T-shirts were hand-washed with handwashing soap, dishwashing detergent, laundry detergent, soap, or just water until the bloodstains could not be seen. After drying the T-shirts, DNA and RNA were extracted simultaneously from the bloodstained areas using commercial kits. RNA was reverse-transcribed to DNA, and then the detection of the mRNAs for HBB, ACTB, and 18S rRNA was examined. DNA was quantified via real-time PCR, and then STR typing was performed with a commercial kit. The luminol and leucomalachite green tests were used as preliminary bloodstain tests, and an immuno-chromatography kit was used to identify human bloodstains. DNA could be extracted from all washed bloodstains, but more DNA was extracted from cotton T-shirts than from polyester T-shirts. STR typing was successful for all bloodstains without issues such as PCR inhibition. In the human bloodstain identification test using mRNA, almost all bloodstains produced a Ct value for HBB and all bloodstains produced a Ct value for 18S rRNA, whereas few bloodstains produced a Ct value for ACTB. All bloodstains reacted positively to luminol, but some were negative for leucomalachite green. Most of the bloodstains did not react positively in the human bloodstain identification test using the immuno-chromatography kit. The results suggest that human bloodstain identification and DNA typing can still be performed after clothes with bloodstains are washed.

The origin identification method for crude drugs derived from arthropods and annelids using molecular biological techniques.

We evaluated whether the origins of crude drugs derived from arthropods and annelids could be identified using molecular biological techniques. DNA was extracted from 20 crude drugs prepared from different animals using a commercial kit with added phenol treatment. The target regions used to identify origin were the mitochondrial 16S ribosomal RNA (rRNA), 12S rRNA, and cytochrome oxidase subunit I (COI) gene regions. Extracted DNA was amplified by polymerase chain reaction, and then sequenced by the Sanger method. The aligned sequences were compared with all available sequences using BLAST to estimate the origins of the crude drugs. The origin of crude drugs used in this study could be estimated using this method. The COI region was the best for identifying origin among three regions examined, based on the success rate of PCR amplification and analysis. Moreover, the 12S rRNA region was also useful for origin identification, with the exception of the earthworm. However, the origin of some crude drugs could not be strictly identified due to matches to various species in all three regions. One likely cause was that the species of origin of a crude drug has not been registered in DNA databases. We found that even the same crude drug from the same pharmaceutical company had different origins by production lot or import source country. Therefore, this method is useful not only for DNA-based origin identification but also quality control of production lots.

Estimation of the number of contributors to mixed samples of DNA by mitochondrial DNA analyses using massively parallel sequencing.

We evaluated whether the number of contributors to mixed DNA samples can be estimated by analyzing the D-loop of mitochondrial DNA using massively parallel sequencing. The A- (positions 16,209-16,400) and B- (positions 30-284) amplicons in hypervariable regions 1 and 2, respectively, were sequenced using MiSeq with 2 × 251 cycles. Sequence extraction and trimming were performed using CLC Genomics Workbench 11 and the number of observed haplotypes was counted for each amplicon type using Microsoft Excel. The haplotype ratios were calculated by dividing the number of counted reads of the corresponding haplotype by the total number of sequence reads. Haplotypes that were over the threshold (5%) were defined as positive haplotypes. The number of larger positive haplotypes in either of the two amplicon types was defined as the number of contributors. Samples were collected from seven individuals. Seventeen mixed samples were prepared by mixing DNA from two to five contributors at various ratios. The number of contributors was correctly estimated from almost all of the mixed samples containing equal amounts of DNA from two to five people. In mixed samples of two or three people, the minor components were detected down to a ratio of 20:1 or 8:2:1. However, heteroplasmy, base deletions, and sharing of the same haplotypes caused incorrect estimations of the number of contributors. Although this method still has room for improvement, it may be useful for estimating the number of contributors in a mixed sample, as it does not rely on forensic mathematics.

Establishment of widely applicable DNA extraction methods to identify the origins of crude drugs derived from animals using molecular techniques.

We established widely applicable DNA extraction methods to identify the origins of crude drugs derived from animals. Twenty-one samples including 17 kinds of crude drug derived from animals were examined. DNA was extracted from most of the crude drugs by adjustment of the QIAamp® DNA Mini Kit. DNA extraction was performed successfully using phenol to remove impurities after applying a proteinase treatment. DNA extraction was performed successfully by decalcification treatment using ethylenediaminetetraacetic acid (EDTA), before applying the proteinase treatment for crude drugs having high calcium content, such as those from oyster shell and cuttlefish bone. DNA could not be extracted from sea-ear shell using the EDTA decalcification treatment, but was extracted successfully using a TBONE EX KIT. The mitochondrial 16S ribosomal RNA (rRNA) gene region was amplified, and Basic Local Alignment Search Tool (BLAST) analysis was performed after sequencing. Polymerase chain reaction (PCR) products of approximately 600 bp in length were obtained from all samples except donkey glue, one of the two seahorses, and longgu. Drug origins were determined in all samples by sequence analysis based on the BLAST results, and match rates were >97 %. Moreover, 16 samples had a match rate >99 %. Our DNA extraction methods were widely applicable to evaluation of many crude drugs derived from animals, and proved very useful for identifying the origins of such drugs.

Identification of canine saliva using mRNA-based assay.

A dog saliva analysis in addition to a bite-mark analysis may be important for evidence when a crime involves a dog bite. In this study, the utility of detecting canine saliva-specific mRNAs to identify canine saliva was evaluated. Canine saliva swabs (n = 20), urine swabs (n = 20), body surface swabs (n = 20), whole blood samples (n = 10), human saliva (n = 20), human skin surface swabs (n = 20), and human whole blood (n = 20) were tested. The saliva-specific genes encoding statherin (STATH), carbonic anhydrase VI (CA-VI), and dog allergens (Canf1 and Canf2) were analyzed as candidate genes. Moreover, glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as confirmation of canine mRNA extraction. STATH, CA-VI, Canf1, Canf2, and GAPDH mRNAs were detected in 19/20, 1/20, 11/20, 4/20, and 20/20 saliva samples, respectively. The STATH, CA-VI, Canf1, Canf2, and GAPDH mRNAs did not exhibit cross-reactivity with samples of human origin. This mRNA-based assay was also able to detect canine saliva in mock forensic samples. The results of this study indicated that the detection of STATH mRNA is useful for the identification of canine saliva, and GAPDH is a suitable marker for canine mRNA extraction.

Sudden death of a child from myocardial infarction due to arteritis of the left coronary trunk.

An eight-year-old Japanese boy developed abdominal pain, followed by convulsion and loss of consciousness. He was taken to an emergency room but could not be resuscitated. At autopsy, the left main coronary trunk (LMT) demonstrated an increase in caliber with severe luminal narrowing, and the left anterior descending branch (LAD) subsequent to the LMT showed severe stenosis. Microscopically, the intima of the LMT demonstrated severe fibrosis and infiltration of lymphocytes and histiocytes suggesting vasculitis, and the small lumen was occupied by a fresh thrombus. The LAD showed significant intimal thickening with strong lymphocytic inflammation at the edge of the thickening. The left ventricle showed widespread myocardial infarction in the recovery stage. There were no findings of atherosclerosis, vasculitis or fibrocellular changes in the ascending aorta or intravisceral arteries other than the LMT and the LAD under investigation. The increase in the caliber of the LMT and the limitation of arteritis to the LMT and the subsequent branch suggested Kawasaki disease (KD), but it was atypical that the patient had no clinical history consistent with KD. The present case showed no findings suggesting classical polyarteritis nodosa (cPAN) at the acute or scar stage in the other vessels being investigated, and cPAN in childhood is rare compared to KD. A nonspecific inflammatory reaction (single organ vasculitis, SOV) was also considered as a possible cause, but it is difficult to determine whether the cause of the coronary stenosis in the present case was SOV because the sampling of arteries was insufficient. If forensic pathologists make unusual findings suggesting vasculitis at autopsy, the collection of a sufficient number of vessels of various sizes is warranted.

Establishment of a novel model of chondrogenesis using murine embryonic stem cells carrying fibrodysplasia ossificans progressiva-associated mutant ALK2.

Fibrodysplasia ossificans progressiva (FOP) is a genetic disorder characterized by heterotopic endochondral ossification in soft tissue. A mutation in the bone morphogenetic protein (BMP) receptor ALK2, R206H, has been identified in patients with typical FOP. In the present study, we established murine embryonic stem (ES) cells that express wild-type human ALK2 or typical mutant human ALK2 [ALK2(R206H)] under the control of the Tet-Off system. Although wild-type ALK2 and mutant ALK2(R206H) were expressed in response to a withdrawal of doxycycline (Dox), BMP signaling was activated only in the mutant ALK2(R206H)-expressing cells without the addition of exogenous BMPs. The Dox-dependent induction of BMP signaling was blocked by a specific kinase inhibitor of the BMP receptor. The mutant ALK2(R206H)-carrying cells showed Dox-regulated chondrogenesis in vitro, which occurred in co-operation with transforming growth factor-ß1 (TGF-ß1). Overall, our ES cells are useful for studying the molecular mechanisms of heterotopic ossification in FOP in vitro and for developing novel inhibitors of chondrogenesis induced by mutant ALK2(R206H) associated with FOP.

Identification of a novel bone morphogenetic protein (BMP)-inducible transcript, BMP-inducible transcript-1, by utilizing the conserved BMP-responsive elements in the Id genes.

Bone morphogenetic proteins (BMPs) inhibit myogenesis and induce osteoblastic differentiation in myoblasts. They also induce the transcription of several common genes, such as Id1, Id2 and Id3, in various cell types. We have reported that a GC-rich element in the Id1 gene functions as a BMP-responsive element (BRE) that is regulated by Smads. In this study, we analyzed and identified BREs in the 5'-flanking regions of the mouse Id2 and Id3 genes. The core GGCGCC sequence was conserved among the BREs in the Id1, Id2 and Id3 genes and was essential for the response to BMP signaling via Smads. We found a novel BRE on mouse chromosome 13 at position 47,723,740-47,723,768 by searching for conserved sequences containing the Id1 BRE. This potential BRE was found in the 5'-flanking region of a novel gene that produces a non-coding transcript, termed BMP-inducible transcript-1 (BIT-1), and this element regulated the expression of this gene in response to BMP signaling. We found that BIT-1 is expressed in BMP target tissues such as the testis, brain, kidney and cartilage. These findings suggest that the transcriptional induction of the Ids, BIT-1 and additional novel genes containing the conserved BRE sequence may play an important role in the regulation of the differentiation and/or function of target cells in response to BMPs.

Identification and functional analysis of Zranb2 as a novel Smad-binding protein that suppresses BMP signaling.

Smads 1/5/8 transduce the major intracellular signaling of bone morphogenetic proteins (BMPs). In the present study, we analyzed Smad1-binding proteins in HEK293T cells using a proteomic technique and identified the protein, zinc-finger, RAN-binding domain-containing protein 2 (ZRANB2). Zranb2 interacted strongly with Smad1, Smad5, and Smad8 and weakly with Smad4. The overexpression of Zranb2 inhibited BMP activities in C2C12 myoblasts in vitro, and the injection of Zranb2 mRNA into zebrafish embryos induced weak dorsalization. Deletion analyses of Zranb2 indicated that the serine/arginine-rich (SR) domain and the glutamine-rich domain were required for the inhibition of BMP activity and the interaction with Smad1, respectively. Zranb2 was found to be localized in the nucleus; however, the SR domain-deleted mutant localized to the cytoplasm. The knockdown of endogenous Zranb2 in C2C12 cells enhanced BMP activity. Zranb2 suppressed Smad transcriptional activity without affecting Smad phosphorylation, nuclear localization, or DNA binding. Taken together, these findings suggested that Zranb2 is a novel BMP suppressor that forms a complex with Smads in the nucleus.

A novel mutation of ALK2, L196P, found in the most benign case of fibrodysplasia ossificans progressiva activates BMP-specific intracellular signaling equivalent to a typical mutation, R206H.

Fibrodysplasia ossificans progressiva (FOP) is a rare autosomal dominant congenital disorder characterized by progressive heterotopic ossification in muscle tissues. Constitutively activated mutants of a bone morphogenetic protein (BMP) receptor, ALK2, have been identified in patients with FOP. Recently, a novel ALK2 mutation, L196P, was found in the most benign case of FOP reported thus far. In the present study, we examined the biological activities of ALK2(L196P) in vitro. Over-expression of ALK2(L196P) induced BMP-specific activities, including the suppression of myogenesis, the induction of alkaline phosphatase activity, increased BMP-specific luciferase reporter activity, and increased phosphorylation of Smad1/5 but not Erk1/2 or p38. The activities of ALK2(L196P) were higher than those of ALK2(G356D), another mutant ALK2 allele found in patients with FOP and were equivalent to those of ALK2(R206H), a typical mutation found in patients with FOP. ALK2(L196P) was equally or more resistant to inhibitors in comparison to ALK2(R206H). These findings suggest that ALK2(L196P) is an activated BMP receptor equivalent to ALK2(R206H) and that ALK2(L196P) activity may be suppressed in vivo by a novel molecular mechanism in patients with this mutation.

Suppression of BMP-Smad signaling axis-induced osteoblastic differentiation by small C-terminal domain phosphatase 1, a Smad phosphatase.

Bone morphogenetic proteins (BMPs) induce osteoblastic differentiation in myogenic cells via the phosphorylation of Smads. Two types of Smad phosphatases--small C-terminal domain phosphatase 1 (SCP1) and protein phosphatase magnesium-dependent 1A--have been shown to inhibit BMP activity. Here, we report that SCP1 inhibits the osteoblastic differentiation induced by BMP-4, a constitutively active BMP receptor, and a constitutively active form of Smad1. The phosphatase activity of SCP1 was required for this suppression, and the knockdown of SCP1 in myoblasts stimulated the osteoblastic differentiation induced by BMP signaling. In contrast to protein phosphatase magnesium-dependent 1A, SCP1 did not reduce the protein levels of Smad1 and failed to suppress expression of the Id1, Id2, and Id3 genes. Runx2-induced osteoblastic differentiation was suppressed by SCP1 without affecting the transcriptional activity or phosphorylation levels of Runx2. Taken together, these findings suggest that SCP1 may inhibit the osteoblastic differentiation induced by the BMP-Smad axis via Runx2 by suppressing downstream effector(s).

Canonical Wnts and BMPs cooperatively induce osteoblastic differentiation through a GSK3beta-dependent and beta-catenin-independent mechanism.

Both BMPs and Wnts play important roles in the regulation of bone formation. We examined the molecular mechanism regulating cross-talk between BMPs and Wnts in the osteoblastic differentiation of C2C12 cells. Canonical Wnts (Wnt1 and Wnt3a) but not non-canonical Wnts (Wnt5a and Wnt11) synergistically stimulated ALP activity in the presence of BMP-4. Wnt3a and BMP-4 synergistically stimulated the expression of type I collagen and osteonectin. However, Wnt3a did not stimulate ALP activity that was induced by a constitutively active BMP receptor or Smad1. Noggin and Dkk-1 suppressed the synergistic effect of BMP-4 and Wnt3a, but Smad7 did not. Overexpression of beta-catenin did not affect BMP-4-induced ALP activity. By contrast, inhibition or stimulation of GSK3beta activity resulted in either stimulation or suppression of ALP activity, respectively, in the presence of BMP-4. Taken together, these findings suggest that BMPs and canonical Wnts may regulate osteoblastic differentiation, especially at the early stages, through a GSK3beta-dependent but beta-catenin-independent mechanism.

Dual roles of smad proteins in the conversion from myoblasts to osteoblastic cells by bone morphogenetic proteins.

Bone morphogenetic proteins (BMPs) induce ectopic bone formation in muscle tissue in vivo and convert myoblasts such that they differentiate into osteoblastic cells in vitro. We report here that constitutively active Smad1 induced osteoblastic differentiation of C2C12 myoblasts in cooperation with Smad4 or Runx2. In floxed Smad4 mice-derived cells, Smad4 ablation partially suppressed BMP-4-induced osteoblast differentiation. In contrast, the BMP-4-induced inhibition of myogenesis was lost by Smad4 ablation and restored by Smad4 overexpression. A nuclear zinc finger protein, E4F1, was identified as a possible component of the Smad4 complex that suppresses myogenic differentiation in response to BMP signaling. In the presence of Smad4, E4F1 stimulated the expression of Ids. Taken together, these findings suggest that the Smad signaling pathway may play a dual role in the BMP-induced conversion of myoblasts to osteoblastic cells.