Identification of Type I and Type II Serine/Threonine Kinase Receptors for Growth/Differentiation Factor-5*

Growth/differentiation factor-5 (GDF-5) is a member of the bone morphogenetic protein (BMP) family, which plays an important role in bone development in vivo . Mutations in the GDF-5 gene result in in mice and Hunter-Thompson type chondrodysplasia in human. BMPs transduce their effects through binding to two different types of serine/threonine kinase receptors, type I and type II. However, binding abilities appear to be different among the members of the BMP family. BMP-4 binds to two different type I receptors, BMP receptors type IA (BMPR-IA) and type IB (BMPR-IB), and a type II receptor, BMP receptor type II (BMPR-II).

BMPs were originally identified as proteins that induce ectopic bone and cartilage formation in vivo (7,8). In vitro studies have revealed that BMPs have various biological effects on different cell types, e.g. stimulation of proteoglycan synthesis in chondroblasts (9), synthesis of collagen and alkaline phosphatase during chondrogenic and osteogenic differentiation (9 -11), and induction of differentiation in neural cells (12,13). GDF-5 also stimulates chondrogenic phenotype expression in vitro and induces cartilage and bone formation in vivo (14).
BMPs are widely distributed not only in bone and cartilage but in other tissues; e.g. BMP-3, -4, -5, and -6 are found in lung and liver, and OP-1/BMP-7 is expressed in kidney (15). BMPs play important roles in the embryonal development. Null mutation in the BMP-4 gene leads to defects in mesoderm formation during the early embryonic stage (16). OP-1/BMP-7-deficient mice die shortly after birth because of poor kidney development and have eye defects and skeletal abnormalities (17,18). GDF-5/CDMP-1 is predominantly expressed in the precartilaginous mesenchymal condensation and the cartilaginous cores of the developing long bone (1,3). Mutations in the murine GDF-5 gene result in abnormal skeletal development, known as brachypodism (1). In a recent study, a mutation in the human GDF-5 gene was shown to be associated with a recessive human chondrodysplasia (Hunter-Thompson type) (19). The resulting phenotype of this disorder is similar to murine brachypodism.
Members of the TGF-␤ superfamily transduce their signals through the formation of heteromeric complexes of two different types of serine/threonine kinase receptors, i.e. type I recep-* This work was supported in part by grants-in-aid for scientific research from the Ministry of Education, Science and Culture of Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  tors of about 50 -55 kDa and type II receptors of more than 70 kDa (20 -22). A series of receptor serine/threonine kinases, termed activin receptor-like kinase (ALK)-1 to -6, was previously identified to constitute a type I receptor family, including a TGF-␤ type I receptor (T␤R-I/ALK-5), two activin type I receptors (ActR-I/ALK-2 and ActR-IB/ALK-4), and two BMP type I receptors (BMPR-IA/ALK-3 and BMPR-IB/ALK-6) (23)(24)(25)(26)(27)(28)(29). Type II receptors for activin (ActR-II and ActR-IIB) (30 -32), for TGF-␤ (T␤R-II) (33), and for BMPs (BMPR-II) (34 -37) have been identified in mammals. In the TGF-␤ and activin receptor systems, ligand binds first to its specific type II receptor, and the complex of ligand and type II receptor is then recognized by type I receptor. Upon formation of the heteromeric receptor complex, type I receptor is phosphorylated by type II receptor, and subsequent activation of the catalytic activity of type I receptor kinase is essential for signaling (38,39).
BMP-4 binds to BMPR-IA and BMPR-IB efficiently (27,28,40,41) in the presence of DAF-4, a type II receptor in Caenorhabditis elegans (42), whereas OP-1/BMP-7 binds to BMPR-IB and less efficiently to BMPR-IA (27). OP-1/BMP-7, but not BMP-4, can also bind to one of the activin type I receptors, ActR-I, in the presence of DAF-4 (27). In addition, OP-1/BMP-7 was recently shown to bind ActR-II and ActR-IIB and mediate certain activin-like effects through the ActR-II⅐ActR-I complex (43). We have recently shown that ALK-1 (also termed TGF-␤ superfamily receptor type I or R3) mediates certain signals after stimulation by OP-1/BMP-7. 2 Human BMPR-II was recently cloned, and it was shown that BMP-2, BMP-4, and OP-1/BMP-7 bound to BMPR-II and transduced signals in combination with certain type I receptors after forming heteromeric complexes (35)(36)(37). In contrast to the TGF-␤ and activin receptors, BMP type I and type II receptors bind ligands independently, but binding affinity is up-regulated in the presence of both receptor types. The ligand-receptor interactions of BMPs in mammals are remarkably similar to those observed with the DPP receptor system in Drosophila (44,45). However, the precise signaling mechanism of the BMP/DPP receptor systems remains unknown.
We investigated the biological effect of GDF-5 on osteoprogenitor-like cell lines and identified type I and type II receptors for GDF-5; BMPR-IB and BMPR-II, but not BMPR-IA, bound GDF-5 in ROB-C26 cells and other cell types. Moreover, we show here that GDF-5 transduces its signal through heteromeric complexes of BMPR-IB and various type II receptors.
Alkaline Phosphatase (ALP) Activity-For a histochemical analysis of ALP activity, cells were fixed for 10 min with 3.7% formaldehyde at room temperature. After washing with phosphate-buffered saline, the cells were incubated for 20 min with a mixture of 0.1 mg/ml of naphthol AS-MX phosphate (Sigma), 0.5% N,N-dimethylformamide, 2 mM MgCl 2 , and 0.6 mg/ml of fast blue BB salt (Sigma) in 0.1 M Tris-HCl, pH 8.5, at room temperature.
For a quantitative analysis of ALP activity, cells were washed with 20 mM Tris-HCl, pH 7.5, and 150 mM NaCl and extracted with lysis buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 1% Triton X-100). ALP activity was determined by an established technique using pnitrophenyl phosphate (Sigma) as a substrate (48). Protein concentration in each extract was measured by DC protein assay (Bio-Rad) using bovine serum albumin as a standard. Preparation of Polyclonal Antibodies-Antisera to the type I receptors were made against synthetic peptides corresponding to the intracellular juxtamembrane parts of the type I receptors (26,49). An antiserum against ActR-II (ARC-2), which detects only ActR-II, was generated against a peptide corresponding to the C-terminal tail of ActR-II (50). An antiserum against the intracellular part of ActR-II (mARII) (51), which cross-reacts with ActR-IIB, was a gift from K. Verschueren (University of Leuven, Belgium). Antisera against TGF-␤ receptor type II (DRL) and BMPR-II (SMN and NRR) were generated against peptides corresponding to the C-terminal tails of the receptors, as previously reported (26,36).
Transient Transfection of cDNAs-Transient transfection plasmids encoding the type I receptors were previously described (26,27). ActR-II cDNA was a gift from L. S. Mathews and W. W. Vale (Salk Institute, San Diego, CA). ActR-IIB1 cDNA, BMPR-II His construct and p3TP-Lux promoter-reporter construct were obtained from J. Massagué. For transient transfection, cDNAs for type I or type II receptors subcloned into pSV7d (52), pcDNA1, pcDNA3 (Invitrogen), or pCMV5 (53) expression vectors were used. These plasmids and p3TP-Lux promoter-reporter construct (1 g of each) were transfected into COS-1 or R mutant Mv1Lu cells by the transfection kit of eukaryotic cells (Tfx TM -50, Promega), following the manufacturer's protocol. One or two days after, the cells were used for affinity cross-linking and immunoprecipitation studies or transcriptional response assay.
Binding, Affinity Cross-linking, and Isolation of the Cross-linked Complexes-Recombinant human BMP-2 and GDF-5 3 were iodinated according to the chloramine-T method as described (26). Cells were incubated on ice for 2-3 h with 0.2-0.5 nM of 125 I-labeled ligands in the presence or absence of unlabeled ligands in a binding buffer (phosphatebuffered saline containing 0.9 mM CaCl 2 , 0.49 mM MgCl 2 , and 1 mg/ml bovine serum albumin). After incubation, the cells were washed with the binding buffer without bovine serum albumin, and cross-linking was done in the same buffer containing 0.27 mM of disuccinimidyl suberate (Pierce) and 1 mM of bis(sulfosuccinimidyl) suberate (Pierce) for 15 min on ice. The cells were washed once with a buffer containing 10 mM Tris-HCl, pH 7.4, 1 mM EDTA, 10% glycerol, and 0.3 mM phenylmethylsulfonyl fluoride (Sigma) and lysed for 20 min in lysis buffer (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 10 mM EDTA, 1% Triton X-100, 1% sodium deoxycholate) containing 1.5% Trasylol (Bayer) and 1 mM phenylmethylsulfonyl fluoride and clarified by centrifugation. Crosslinked materials were then incubated with antisera for 45 min at 4°C. Immune complexes were bound to protein A-Sepharose (Kabi-Pharmacia) for 30 min at 4°C, washed once with a buffer containing 20 mM Tris-HCl, pH 7.5, 500 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, 0.2% SDS, followed by one wash in distilled water. For isolation of BMPR-II His complexes, cells were extracted with a buffer containing 50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.5% Triton X-100, and protease inhibitors. Cell extracts were clarified by centrifugation and incubated with Ni 2ϩ -NTA-agarose (Qiagen) for 1 h at 4°C in the presence of 20 mM imidazole. Beads were rinsed briefly once with the same buffer. The immune complexes or complexes isolated by Ni 2ϩ -NTA-agarose were eluted by boiling for 3 min in SDS sample buffer (100 mM Tris-HCl, pH 8.8, 0.01% bromphenol blue, 36% glycerol, 4% SDS) containing 10 mM dithiothreitol and analyzed by SDS-7% polyacrylamide gel electrophoresis. The gels were fixed, dried, and subjected to the analysis using a Fuji BAS 2000 Bio-Imaging Analyzer (Fuji Photo Film).
Transcriptional Response Assay-R mutant Mv1Lu cells were cotransfected with p3TP-Lux promoter-reporter construct (25,54) with plasmids containing the type I or type II receptor cDNAs as described above. One day after transfection, cells were starved in Dulbecco's modified Eagle's medium containing 0.2% FBS for 6 h and then exposed to 300 ng/ml of GDF-5 for 24 h. Luciferase activity in the cell lysate was measured using the luciferase assay system (Toyo Ink) according to the manufacturer's protocol and a luminometer (AutoLumat LB953; EG&G Berthold).

RESULTS
ALP Activity-GDF-5 stimulated mesenchyme aggregation and chondrogenesis in rat limb bud cells in vitro and induced ectopic cartilage and bone formation in mice tissues of rodents in vivo (14). However, several osteoblastic cell lines, such as MC3T3-E1 cells, did not efficiently respond to GDF-5 as measured by ALP activity in contrast to the effects of other BMPs. To identify the signaling receptors for GDF-5, we first attempted to find cell lines that respond to GDF-5 using the enzyme histochemical study of ALP activity. In ROB-C26 rat osteoprogenitor-like cell line, GDF-5 increased the number of ALP-positive cells at 300 ng/ml (Fig. 1A). However, most cell lines, including MC3T3-E1 mouse osteoblastic cells, ROS17/2.8 rat osteosarcoma cells, and C2C12 mouse myoblastic cells, had no significant increase in ALP activity by the treatment with GDF-5 (data not shown). Fig. 1B shows the dose-dependent effect of GDF-5 on ROB-C26 cells treated for 6 days. ALP activity was induced by GDF-5 in a dose-dependent manner.
Identification of GDF-5 Receptors in Nontransfected Cell Lines-In order to investigate which serine/threonine kinase receptors act as type I and type II receptors for GDF-5, the ROB-C26 cells were tested for the binding of GDF-5. The cells were affinity-labeled using 125 I-GDF-5, and the cross-linked complexes were analyzed by immunoprecipitation using the antiserum against each of type I and type II receptors, followed by SDS-gel electrophoresis under reducing conditions. Crosslinked complexes of 80 -90 kDa could be immunoprecipitated by the antiserum to BMPR-IB ( Fig. 2A). A high molecular mass complex of 150 -200 kDa, which may represent a type II receptor complex, was co-immunoprecipitated by the BMPR-IB antiserum. The type II receptor complex could be immunoprecipitated by the BMPR-II antiserum, and co-immunoprecipitation of the type I receptor complex was also seen ( Fig. 2A). Immunoprecipitation by the BMPR-II antiserum was less efficient than that by the BMPR-IB antiserum, which may be due to a poor affinity of the BMPR-II antiserum (36). Weak bands could be seen after immunoprecipitation by the antisera against ALK-1, BMPR-IA, and T␤R-II ( Fig. 2A), but these were not reproducible.
Binding of 125 I-BMP-2 was also tested in the ROB-C26 cells (Fig. 2B). Although 125 I-BMP-2 is known to bind BMPR-IA and BMPR-IB as well as BMPR-II in other systems, binding to only BMPR-IB and BMPR-II could be seen in this cell type, suggesting that this cell line predominantly expresses BMPR-IB rather than BMPR-IA.
To identify the endogenous receptors for GDF-5 in other cell types, we tested some other cell lines for the binding of GDF-5. In the U1240 MG glioblastoma cell line and the Mv1Lu mink lung epithelial cell line, BMPR-IB could bind GDF-5 (Fig. 3). In contrast, 125 I-BMP-2-cross-linked complexes to Mv1Lu were immunoprecipitated by BMPR-IA (data not shown). In most cell types investigated, including ATDC5 chondroblastic cells, MC3T3-E1 mouse osteoblastic cells, ROS17/2.8 rat osteosarcoma cells, BEC bovine endothelial cells, and C2C12 mouse myoblastic cells, binding of GDF-5 was not clear (data not shown).
Binding of GDF-5 to Type I and Type II Receptors Expressed in COS-1 Cells-In order to further investigate the type I and type II receptors for GDF-5, binding was tested using COS-1 cells transfected with the cDNAs for serine/threonine kinase receptors. For the transfection of BMPR-II cDNA, a C-terminally truncated form of BMPR-II His , which encodes 530 amino acid residues with a hexahistidine tag in its C terminus and, therefore, forms about 100 kDa of cross-linked complexes (35), was used. The cross-linked complexes were precipitated using the specific antisera or Ni 2ϩ -NTA agarose beads.
When singly transfected, we could observe binding of GDF-5 only to BMPR-IB among six type I receptors. Among different type II receptors, ActR-II, ActR-IIB1, and BMPR-II His bound GDF-5 (Fig. 4). GDF-5 did not bind well to other serine/threo-

FIG. 1. The effects of GDF-5 on induction of ALP activity in ROB-C26 cells.
ROB-C26 cells were cultured in ␣-minimal essential medium containing 10% FBS for 6 days with (ϩ) or without (Ϫ) 500 ng/ml GDF-5. The cells were fixed and stained for ALP as described under "Experimental Procedures" and photographed by a phase contrast microscopy (A). B, dose-dependent induction of ALP activity by GDF-5. ROB-C26 cells in 24-well cell culture plates were treated with GDF-5 for 6 days. ALP activity was measured spectrophotometrically with p-nitrophenyl phosphate as a substrate. Enzyme specific activity was presented as nmol of p-nitrophenol produced/min/mg of protein.
When type I receptor cDNAs were co-transfected with the BMPR-II His , ActR-II, or ActR-IIB1 cDNA, GDF-5 bound different sets of type I receptors. In the presence of BMPR-II His , GDF-5 bound efficiently to BMPR-IB, but not to the other type I receptors (Fig. 5A). BMPR-IB as well as BMPR-II bands could be seen when the cross-linked complexes were immunoprecipitated by the BMPR-IB antiserum. When isolated with Ni 2ϩ -NTA agarose beads, co-precipitation of the BMPR-IB complex could also be observed (Fig. 5A). Binding to BMPR-IB was up-regulated in the presence of BMPR-II His , compared with its absence (data not shown). In the presence of ActR-II, GDF-5 bound most efficiently to BMPR-IB (Fig. 5B). Weak binding of GDF-5 was also seen to BMPR-IA in the presence of ActR-II (Fig. 5B). In the presence of ActR-IIB1, GDF-5 bound efficiently to BMPR-IB, but also to ActR-I and to BMPR-IA very weakly (Fig. 5C). Thus, the binding of GDF-5 is most efficient to BMPR-IB compared with the other type I receptors, and weak binding to BMPR-IA and ActR-I is observed in the presence of different type II receptors. When COS-1 cells were co-transfected with BMPR-IB and ActR-IIB1 cDNAs, the BMPR-IB complex could not be immunoprecipitated with the type II receptor antiserum (Fig. 5C), suggesting that BMPR-IB forms a heteromeric complex with BMPR-II and ActR-II upon GDF-5 binding but less efficiently with ActR-IIB1.
Signaling Activity in Response to GDF-5-We next investigated whether type I and type II receptors are capable of

FIG. 2. Binding of GDF-5 and BMP-2 in ROB-C26 cells.
ROB-C26 cells were affinity-labeled with 125 I-GDF-5 (A) and 125 I-BMP-2 (B), followed by cross-linking. The cross-linked complexes were immunoprecipitated with type I and type II receptor antisera. Antiserum to BMPR-II used in A is the NRR antiserum, which was raised against the C-terminal tail of BMPR-II (36). Two different antisera to BMPR-II (SMN and NRR; Ref. 36) were used in B. Binding of 125 I-GDF-5 and 125 I-BMP-2 in the presence or absence of unlabeled GDF-5 or BMP-2 was examined by immunoprecipitation using the BMPR-IB antiserum (C). Samples were subjected to SDS-gel electrophoresis, followed by an analysis using a Bio-Imaging Analyzer (BAS 2000; Fuji). Markers of molecular mass are indicated to the left.
signaling upon binding GDF-5 using a p3TP-Lux promoterreporter construct (54). R mutant Mv1Lu cells were transfected with type I and/or type II receptors, together with p3TP-Lux, and stimulated or not stimulated by GDF-5. Since transfection of empty pSV7d vector showed no luciferase response to GDF-5 (Fig. 6, first set of bars), these cells were used for the analysis of transfected GDF-5 receptor signaling activity. Cells transfected with ActR-II or BMPR-II alone did not respond to GDF-5 (Fig. 6). Cells transfected with BMPR-IA or BMPR-IB alone showed a very weak response to GDF-5, which may be ascribed to the presence of endogenous activin and BMP type II receptors in the R mutant cells (26,36). When p3TP-Lux was cotransfected with ActR-II and BMPR-IB, transcriptional activation was clearly observed after stimulation by GDF-5 (Fig. 6). BMPR-IA also showed a less but significant increase in luciferase activity in the presense of ActR-II. Similarly, co-transfection of BMPR-II and BMPR-IB mediated transcriptional activation; however, co-transfection of BMPR-II and BMPR-IA did not show significant transcriptional activation (Fig. 6). In the R mutant cells co-transfected with ActR-IIB1 and type I receptors (ActR-I, BMPR-IA, and BMPR-IB), the activation of transcription by GDF-5 was not detected (data not shown).

DISCUSSION
In the present paper, we first tried to find the cell lines that respond to GDF-5 in order to identify the signaling receptors for GDF-5. In contrast to BMP-2, BMP-4, and OP-1/BMP-7, GDF-5 did not efficiently induce the ALP activity in most cell lines, including MC3T3-E1, ROS17/2.8, and C2C12 cells. Only the osteoprogenitor-like cell line, ROB-C26, could respond to GDF-5 (Fig. 1, A and B). These data suggested to us that the bioactivity of GDF-5 can be observed in limited cell types, and the receptor-binding profile of GDF-5 may be different from those of other BMPs. We then investigated the receptors in the ROB-C26 cells using antibodies against known serine/threonine kinase receptors. Interestingly, GDF-5 bound to only BMPR-IB and BMPR-II in this cell line but not to the other serine/threonine kinase receptors, including BMPR-IA ( Fig.  2A). Since BMP-2 showed a similar binding profile in the ROB-C26 cells (Fig. 2B), this cell line may predominantly express BMPR-IB and BMPR-II compared with the other receptors.  4. Binding of GDF-5 in COS-1 cells singly transfected with type I and type II receptor cDNAs. COS-1 cells were transfected with cDNAs for type I and type II receptors. The cells were affinitylabeled using 125 I-GDF-5, followed by cross-linking. Cell lysates were precipitated with each of the corresponding antisera against type I and type II receptors or with Ni 2ϩ -NTA-agarose beads. Samples were subjected to SDS-gel electrophoresis, followed by an analysis using a Bio-Imaging Analyzer . Isolation A-I, B-IA, A-IB, T-I, B-IB, A-II, A-IIB, and T-II represent immunoprecipitation by ActR-I, BMPR-IA, ActR-IB, T␤R-I, BMPR-IB, ActR-II, ActR-IIB and T␤R-II antisera, respectively; Isolation Ni represents Ni 2ϩ -NTA isolation.
Furthermore, the binding of 125 I-GDF-5 and 125 I-BMP-2 to BMPR-IB was competed by unlabeled GDF-5 or BMP-2 (Fig.  2C), indicating that the binding site to BMPR-IB is shared by these ligands. MC3T3-E1 cells are known to respond to BMP-4 and OP-1/BMP-7. In this cell type, ActR-I and BMPR-IA were shown to bind 125 I-OP-1/BMP-7 and 125 I-BMP-4, respectively (27). However, 125 I-GDF-5 did not efficiently bind to MC3T3-E1 cells, and we could not detect the cross-linked complexes with 125 I-GDF-5 in this cell line using any of the serine/threonine kinase receptor antisera (data not shown).
We then investigated the binding of GDF-5 to cell lines of nonskeletal origins. In U-1240 MG glioblastoma cells and Mv1Lu mink lung epithelial cells, GDF-5 bound only to BMPR-IB among the six type I receptors (Fig. 3). These cells have previously been shown to endogenously express ActR-I and BMPR-IA (27), but the cross-linked complexes with 125 I-GDF-5 were not immunoprecipitated by the antiserum against ActR-I or BMPR-IA. Moreover, we found that 125 I-BMP-2 bound to BMPR-IA as well as BMPR-IB in the Mv1Lu cells (data not shown). Taken together, the binding profile of GDF-5 in nontransfected cells is different from that of BMP-2, BMP-4, or OP-1/BMP-7; i.e. GDF-5 preferentially binds to BMPR-IB but not to ActR-I and BMPR-IA.
We next studied the binding of GDF-5 in COS-1 cells transfected with the receptor cDNAs. Similar to BMP-4 and OP-1/ BMP-7, BMPR-IB bound GDF-5 without the transfection of type II receptors. However, other type I receptors did not bind GDF-5 in the absence of type II receptors (Fig. 4). Binding of GDF-5 to the type II receptors was similar to that of OP-1/ BMP-7; i.e. GDF-5 bound to BMPR-II as well as ActR-II and ActR-IIB1 (Fig. 4). When the type I receptors were co-transfected with the type II receptors, the binding profiles were complicated; GDF-5 formed complexes only with BMPR-IB in the presence of BMPR-II (Fig. 5A), but it formed complexes not only with BMPR-IB but also with BMPR-IA in the presence of ActR-II (Fig. 5B). In the presence of ActR-IIB1, binding of GDF-5 to ActR-I, BMPR-IA, and BMPR-IB could be seen, but BMPR-IB was not likely to form a tight complex with ActR-IIB1, because co-immunoprecipitation of neither ActR-IIB1 by the BMPR-IB antiserum nor BMPR-IB by the ActR-IIB1 antiserum could be observed (Fig. 5C).
Analysis of the signaling activity of GDF-5 using the p3TP-Lux construct in the R mutant Mv1Lu cells was in good agreement with the binding data observed in the transfected COS-1 cells. GDF-5 transduced the signal through BMPR-IB together with BMPR-II or ActR-II (Fig. 6). The complex of BMPR-IA and ActR-II mediated a less efficient signal than those containing BMPR-IB. GDF-5 did not transduce the p3TP-Lux signal through ActR-IIB1 and the type I receptors (data not shown). In contrast, activin was shown to activate the p3TP-Lux signal through ActR-IIB and ActR-I (25). These results in the nontransfected and transfected cells indicate that BMPR-IB and BMPR-II can act as functional type I and type II receptors for GDF-5, respectively. ActR-II may also possibly serve as a type II receptor for GDF-5 in certain cell types.
Cartilage and bone formation could be induced by GDF-5 in rodent thigh in vivo (14). Moreover, osteoblasts in primary culture were shown to respond to GDF-5, increasing ALP activity (14), and they formed bone-like nodules, where positive immunohistochemical staining could be seen by the antibodies against BMPR-IA and BMPR-IB. 4 In adult tissues, BMPR-IB is found mainly in brain, whereas BMPR-IA and BMPR-II are expressed in various tissues (24,26,28,36). BMPR-IB is expressed in the process of bone formation during embryogenesis and after bone fracture, although the expression profile appears to be limited compared with that of BMPR-IA (55,56). These data suggest that the bone and cartilage formation by GDF-5 observed in vivo is induced through BMPR-IB.
Binding of GDF-5 was observed in U-1240 MG glioblastoma cells and Mv1Lu mink lung epithelial cells. In addition to bone and cartilage, BMPR-IB and BMPR-II are expressed in various soft tissues during embryogenesis (36,55). Moreover, both receptor transcripts are expressed in adult brain. Chang et al. (3) showed that although GDF-5 (CDMP-1) is predominantly expressed in skeletal muscles, GDF-6 (CDMP-2) is found in various soft tissues. Thus, GDF-5 or other highly related proteins may have broad physiological functions in different tissues. Compatible with the limited expression of GDF-5, the mutations in the GDF-5 gene result in the abnormalities only in skeletal tissues in both mouse and human (1,19). The other GDF-5-like proteins, including GDF-6 and GDF-7, may play important roles in nonskeletal tissues.
Among the TGF-␤ superfamily, BMPs are classified by their ability to form bone and cartilage in vivo. However, BMPs include heterogenous members, and they can be subdivided into subgroups based on their amino acid sequence similarities. Moreover, the biological activities appear to be different among the members in the BMP family. Present data showed that binding profiles of GDF-5 are more limited than those of BMP-2, BMP-4, and OP-1/BMP-7, which suggests different biological functions of GDF-5 both in vitro and in vivo. Future studies including the comparison of the receptor binding properties of the other members in the BMP family, will be needed to understand the in vivo functions of various members of the BMP family.
FIG. 6. Signal transduction by GDF-5 through type I and type II receptors. Transcriptional activation was measured using R mutant Mv1Lu cells transfected with p3TP-Lux promoterreporter construct. cDNAs for ActR-I, BMPR-IA, BMPR-IB, ActR-II, and BMPR-II were co-transfected with p3TP-Lux into the cells, which was followed by the stimulation with GDF-5 (300 ng/ml) for 20 h. The experiments were repeated three times with similar results each time.