(Received for publication, October 2, 1996, and in revised form, May 29, 1997)
From the Department of Biochemistry, School of
Dentistry, Showa University, Tokyo 142, Japan, the § Faculty
of Pharmaceutical Sciences, Hokkaido University, Sapporo 060, Japan,
the ¶ Yamanouchi Pharmaceutical Co., Ltd., Tokyo 103, Japan,
and the
Genetics Institute Inc.,
Cambridge, Massachusetts 02140
Members of the transforming growth
factor (TGF)- superfamily bind the transmembrane serine/threonine
kinase complex consisting of type I and type II receptors. Their
intracellular signals are propagated via respective type I receptors.
Bone morphogenetic protein (BMP)-2, a member of the TGF-
superfamily, induces ectopic bone formation when implanted into
muscular tissues. Two type I receptors (BMPR-IA and BMPR-IB) have been
identified for BMP-2. We have reported that BMP-2 inhibits the terminal
differentiation of C2C12 myoblasts and converts their differentiation
pathway into that of osteoblast lineage cells (Katagiri, T., Yamaguchi, A., Komaki, M., Abe, E., Takahashi, N., Ikeda, T., Rosen, V., Wozney,
J. M., Fujisawa-Sehara, A. and Suda, T. (1994) J. Cell Biol. 127, 1755-1766). In the present study, we examined the
involvement of functional BMP-2 type I receptors in signal transduction
in C2C12 cells, which expressed mRNA for BMPR-IA, but not for
BMPR-IB in Northern blotting. TGF-
type I receptor (T
R-I)
mRNA was also expressed in C2C12 cells. Subclonal cell lines of
C2C12 that stably expressed a kinase domain-truncated BMPR-IA
(
BMPR-IA) differentiated into myosin heavy chain-expressing myotubes
but not into alkaline phosphatase (ALP)-positive cells, even in the
presence of BMP-2. In contrast, the differentiation of the
BMPR-IA-transfected C2C12 cells into myotubes was suppressed by
TGF-
1, as in the parental C2C12 cells. BMP-2 did not efficiently
suppress the mRNA expression of muscle-specific genes such as
muscle creatine kinase, MyoD, and myogenin, nor did it induce the
expression of ALP mRNA in the
BMPR-IA-transfected C2C12 cells.
In contrast, TGF-
1 inhibited mRNA expression of the
muscle-specific genes in those cells. When wild-type BMPR-IA was
transiently transfected into the
BMPR-IA-transfected C2C12 cells, a
number of ALP-positive cells appeared in the presence of BMP-2.
Transfection of wild-type BMPR-IB or T
R-I failed to increase the
number of ALP-positive cells. These results suggest that the
BMP-2-induced signals, which inhibit myogenic differentiation and
induce osteoblast differentiation, are transduced via BMPR-IA in C2C12
myoblasts.
Demineralized bone matrix can induce ectopic bone formation when
implanted into muscular tissues (1). Bone morphogenetic proteins
(BMPs)1 purified and cloned
from demineralized bone matrix are the factors responsible for this
ectopic bone formation (2-7). The deduced amino acid sequence of BMPs
indicates that they are members of the transforming growth factor-
(TGF-
) superfamily (3, 6-8). Genetic studies show that BMPs and
their related molecules are essential for normal skeletal development
in vertebrates (9-11). When the BMP-5 or growth/differentiation factor
(GDF)-5 gene was mutated in mice, they developed abnormal skeletons
(12, 13). A genetic mutation of the human homologue of mouse GDF-5 was
also identified in a patient with chondrodysplasia (14). Gene targeting showed that BMP-7-deficient mice had a defect in skeletal pattern formation besides those in the eye and kidney (15, 16).
Several lines of evidence indicate that the role of BMPs is not restricted to skeletal development but is expanded to diverse developmental events in animals (9-11). DPP, a homologue of human BMP-2 and BMP-4, acts as a morphogen in dorsal-ventral patterning of the Drosophila embryo. BMP-4 acts as a ventralizing factor in the Xenopus embryo. In BMP-4-deficient mice, the mesoderm did not differentiate during embryogenesis (17). It was also suggested that BMPs act as signaling molecules of epithelium-mesenchyme interaction in the development of various organs in vertebrates.
The intracellular signals of members of the TGF- superfamily are
transduced via transmembrane serine/threonine kinase receptors (9, 11,
18, 19). These are classified into two groups known as type I and type
II receptors that are distinguishable by their amino acid sequences and
functional features. Members of the TGF-
superfamily are associated
with specific sets of the type I and type II receptors. Formation of a
type I-type II receptor complex is required for ligand-induced
signalings. Studies on the receptors for TGF-
have revealed the
sequential mechanism of receptor activation (20). The TGF-
type II
receptor (T
R-II) first binds to the ligand and then associates and
phosphorylates TGF-
type I receptor (T
R-I). T
R-I is
phosphorylated by T
R-II at serine and threonine residues in the GS
domain, which is located upstream of the kinase domain. This
phosphorylated T
R-I propagates the ligand-induced signals by
phosphorylating downstream substrates. The structure of the type I
receptor is conserved in all type I receptors of the TGF-
superfamily, suggesting that the BMP signals are transduced by the BMP
type I receptor like those of TGF-
. Two type I receptors (BMPR-IA
and BMPR-IB) and one type II receptor (BMPR-II) have been identified
for BMP-2 and BMP-4 (21-27). BMP-7 can interact with activin type I
and type II receptors in addition to BMP-2/BMP-4 receptors (26,
28).
The ectopic bone-inducing activity of BMPs suggests that they are
involved at an early stage of osteoblast differentiation during
skeletal development. We reported that BMP-2 not only stimulates the
maturation of committed osteoblast progenitor cells but also induces
the differentiation of pluripotent fibroblastic cells into osteoblastic
cells (29-32). To examine the molecular mechanism of ectopic bone
formation induced by BMPs in muscular tissues, we established a model
system using C2C12 myoblasts that reflects an early stage of osteoblast
differentiation during bone formation in muscular tissues (33). In this
model, BMP-2 not only inhibits the differentiation of C2C12 myoblasts
into multinucleated myotubes but also induces the expression of typical
osteoblast phenotypes such as alkaline phosphatase (ALP) activity,
parathyroid hormone response, and osteocalcin production. TGF- also
inhibited terminal differentiation of C2C12 cells but did not induce
any of the osteoblast phenotypes (33). These results suggested that
some BMP-2-specific intracellular signalings are involved in the
induction of osteoblast differentiation of myoblasts. However, the
BMP-2 signaling pathway that regulates the myogenic and osteogenic
differentiation of myoblasts is not clear.
A kinase domain-truncated BMPR-IA abolished the ventralizing activity
of BMP-4 in a dominant negative manner in the Xenopus embryo
(23, 34, 35). We therefore examined whether a dominant negative BMP
type I receptor could inhibit the BMP-2-induced signalings in
myoblasts. We report here that overexpression of the kinase domain-truncated BMPR-IA in C2C12 myoblasts blocks the BMP-2-induced inhibition of myogenic differentiation and induction of osteogenic differentiation. BMP-2 signaling was restored by transiently
transfecting wild-type BMPR-IA into kinase domain-truncated mouse
BMPR-IA (BMPR-IA)-transfected C2C12 cells but not by transfecting
BMPR-IB or T
R-I. These results suggest that the BMP-2-induced
signals, namely, the inhibition of myogenic differentiation and the
induction of osteoblast differentiation, are transduced via BMPR-IA in
C2C12 myoblasts.
Recombinant human BMP-2 was produced and purified from the conditioned media of Chinese hamster ovary cells as described (36).
PlasmidsThe BMPR-IA cDNA was generated by the
standard PCR protocol using a set consisting of vector-derived upstream
and mutated downstream primers. The latter had a stop codon at position
706 (encodes 188 amino acids) in the original mouse TFR-11 (23) followed by an EcoRI restriction site
(5
-AAGAATTCAACGACCCCTGCTTGAGATACT-3
). The kinase domain-truncated
mouse BMPR-IB (
BMPR-IB) (24) cDNA was also generated by a
similar method, and it encoded 147 amino acids. Mutations of those
products were verified by DNA sequencing. These truncated receptor
cDNAs were subcloned into the EcoRI site of the
mammalian expression vector, pMIKHygB (provided Dr. K. Maruyama), which
expresses an insert under the control of the SR
promoter and has a
hygromycin-resistant gene. Wild-type BMPR-IA, BMPR-IB, or T
R-I (37)
receptor was subcloned into pEF-BOS and expressed under the control of
the human EF1
promoter (38). Constitutively active BMPR-IB
(aBMPR-IB) was generated by a substitution of glutamine at 203 for
aspartic acid by PCR. For the immunoprecipitation experiment, the
wild-type BMPR-IB was tagged by introducing epitope sequences for human
c-myc (EQKLISEEDL) into the carboxyl terminus.
C2C12 myoblasts (39) were purchased from the American Type Culture Collection (Rockville, MD). Subclonal cell lines of C2C12 (see below) were maintained in Dulbecco's modified Eagle's medium (DMEM; Sigma) containing 15% fetal bovine serum (FBS; Life Technologies, Inc., Grand Island, NY) and antibiotics (100 units/ml penicillin-G and 100 µg/ml streptomycin) at 37 °C in a humidified atmosphere of 5% CO2 in air. To examine the effects of BMP-2 on muscle and osteoblast differentiation of the subclonal cell lines, C2C12 cells were inoculated at a density of 2 × 104 cells/cm2. After an overnight incubation, the medium was replaced with DMEM containing 5% FBS and 300 ng/ml BMP-2, and then the cells were cultured for an additional 3 days.
TransfectionTo establish subclonal C2C12 cell lines that
constitutively express BMPR-IA or
BMPR-IB, C2C12 myoblasts were
inoculated onto 100-mm dishes at a density of 5 × 105/dish 1 day before transfection by modified calcium
phosphate precipitation according to an instruction manual (Stratagene, CA). In brief, C2C12 cells were incubated overnight with the calcium phosphate-DNA precipitates containing 10 µg of pMIKHygB,
pMIKHygB/
BMPR-IA, or pMIKHygB/
BMPR-IB. They were split at a 1:20
ratio and selected for 10 days in the presence of 700 µg/ml
hygromycin B (Wako Pure Chemical Industries, Osaka, Japan). Colonies
were isolated with penicillin cups and passaged into stable cell lines.
The subclonal cell lines, Vec-16 and Vec-19,
IA-12,
IA-14 and
IA-16 and
IB-2, and
IB-12, were established from cultures of
C2C12 cells transfected with pMIKHygB, pMIKHygB/
BMPR-IA, and
pMIKHygB/
BMPR-IB, respectively.
For rescue experiments, subclonal cells that stably expressed
BMPR-IA were transiently transfected with the wild-type type I
receptor (BMPR-IA, BMPR-IB, or T
R-I) using DOTAP
(N-[1-(2,3-dioleoyloxy)propyl]-N,N,N-trimethylammonium methyl sulfate) (Boehringer Mannheim, Germany). Cells were inoculated onto 48-well plates at a density of 2.5 × 104/well 1 day before transfection. The cells were then incubated for 10 h
with OPTI-MEM (Life Technologies, Inc.) containing a mixture of the
expression vector (pEF-BOS, pEF-BOS/BMPR-IA, pEF-BOS/BMPR-IB, pEF-BOS/T
R-I) and DOTAP according to an instruction manual. The cells were transferred to DMEM containing 15% FBS and 300 ng/ml BMP-2
and were cultured for an additional 3 days. Constitutively active
BMPR-IB (pEF-BOS/aBMPR-IB) was also transfected into the subclonal
cells that stably expressed
BMPR-IA by the same method described
above, and the cells were cultured without BMP-2. Cells were fixed and
histochemically stained for ALP as described below.
Total cellular and poly (A)+
RNAs were extracted from cells using TRIZOL (Life Technologies, Inc.)
and a QuickPrep Micro mRNA purification kit (Pharmacia P-L
Biochemicals Inc., Milwaukee, WI), respectively. Twenty µg of total
RNA or 2.5 µg of poly (A)+ RNA was resolved by
electrophoresis in a 1.2% agarose-formaldehyde gel and transferred
onto a Hybond-N membrane (Amersham International, Amersham, UK). The
membrane was hybridized with [32P]-labeled cDNA
probes for mouse BMPR-IA (23), rat BMPR-IB (24), mouse TR-I (37),
mouse ALP (40), mouse osteocalcin (41), mouse MCK, mouse MyoD (42), and
mouse myogenin (43). The hybridized probes were removed from the
membrane by boiling in 0.2% sodium dodecyl sulfate (SDS) and then
sequentially rehybridized with the respective probes.
To histochemically analyze the expression levels of myosin heavy chain (MHC) and ALP, MHC and ALP were double stained. Cells were fixed in 3.7% formaldehyde and then stained immunohistochemically for MHC using anti-MHC antibody (MF-20; Developmental Studies Hybridoma Bank, Iowa City, IA) (44) as described (33). Before color development, cultures were histochemically stained for ALP using naphthol AS-MX phosphate (Sigma) and fast blue BB salt (Sigma) (33). Cells were rinsed with phosphate-buffered saline, and then the reaction products of biotin-streptavidin were visualized with an AEC substrate kit (Histofine, Nichirei Co., Tokyo, Japan).
ALP ActivityALP activity in the cell layer was measured as described (33). Cell layers were sonicated in 50 mM Tris-HCl, 0.1% Triton X-100, pH 7.5. ALP activity in the lysate was measured at 37 °C in a buffer containing 0.1 M 2-amino-2-methyl-1-propanol (Sigma) and 2 mM MgCl2, pH 10.5, for 30 min using p-nitrophenyl phosphate as the substrate. The enzyme activity was expressed as micromoles of p-nitrophenol (p-NP) produced per min per mg of protein. The protein content was determined using a BCA protein assay kit (Pierce).
Receptor Affinity Labeling Assay125I-labeled BMP-2 was prepared using chloramine-T as described previously (45). In short, cell layers were incubated for 2 h at 37 °C with 5 ng/ml 125I-BMP-2. Receptors were cross-linked to bound ligands with disuccinimidyl substrate (Pierce) and solubilized by sample buffer (25 mM Tris-HCl, pH 6.5, 5% glycerol, 1% SDS, 1% 2-mercaptoethanol, 0.05% bromphenol blue) for SDS-polyacrylamide gel electrophoresis (PAGE). Aliquots were subjected to SDS-PAGE. The autoradiogram was analyzed using a BAS 2000 BioImaging Analyzer (Fuji Photo Film, Tokyo, Japan).
For the immunoprecipitation of receptor proteins, affinity-labeled cells were lysed for 1 h in TNE buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM EDTA) containing 0.5% Triton X-100 and 10 µg/ml aprotinin. For the isolation of receptor complexes, cell extracts were clarified by centrifugation and precipitated with c-myc antibody (9E10; BAbCO, Richmond, CA) followed by binding to protein G-Sepharose (Zymed Laboratories, San Francisco, CA). Immunoprecipitates were washed five times with TNE buffer containing 0.1% Triton X-100 and subjected to SDS-PAGE and autoradiography.
First, we
examined the mRNA levels of the two BMP type I receptors in C2C12
myoblasts by Northern blotting (Fig. 1).
C2C12 cells expressed two major forms of BMPR-IA mRNAs of ~6 and
3.6 kb. C2C12 cells also expressed TR-I mRNA of ~6 kb.
However, a BMPR-IB transcript was not detected in C2C12 cells by
Northern blotting.
Transfection of a Kinase Domain-truncated BMPR-IA into C2C12 Cells Specifically Blocks BMP-2 Signals
To examine the role of BMPR-IA
on the BMP-2 signaling pathways in C2C12 cells, we constructed
BMPR-IA and
BMPR-IB expression vectors and stably transfected
them into C2C12 cells. The subclonal cell lines,
IA-12,
IA-14 and
IA-16,
IB-2 and
IB-12, and Vec-16 and Vec-19, were established
from the C2C12 cultures transfected with
BMPR-IA (
IA lines),
BMPR-IB (
IB lines), and its empty vectors (Vec lines). In the
parental C2C12 cells, 300 ng/ml BMP-2 almost completely inhibited the
differentiation of C2C12 cells into myotubes and converted their
differentiation pathway into ALP-positive osteoblast-like cells (33).
In Vec-16, Vec-19,
IB-2, and
IB-12 cells, myotube formation was
almost completely inhibited, and ALP activity was induced by BMP-2 at
concentrations above 300 ng/ml, like in the parental C2C12 cells (Fig.
2, A and B). The
cells transfected with truncated BMPR-IB (
IB-2 and
IB-12) also
differentiated into ALP-positive cells but not into mature myotubes in
the presence of BMP-2. In contrast, a number of myotubes were formed in
cells transfected with truncated BMPR-IA (
IA-12,
IA-14, and
IA-16), even in the presence of 1000 ng/ml of BMP-2 (Fig.
2A). In the C2C12 cells transfected with
BMPR-IA, little ALP activity was induced by 1000 ng/ml BMP-2 (Fig. 2B).
Fig. 3 shows the histochemical changes
induced by BMP-2 and TGF-1 in Vec-16,
IA-12, and
IB-12 cells.
Elongated MHC-positive cells appeared in all the cultures in the
absence of BMP-2 and TGF-
1 (Fig. 3, a, d, and
g). In Vec-16 and
IB-12 cells, BMP-2 inhibited the
formation of MHC-positive myotubes and induced ALP-positive osteoblast-like cells (Fig. 3, b and h). In
contrast,
IA-12 cells differentiated into MHC-positive, but not into
ALP-positive cells, even in the presence of 300 ng/ml BMP-2 (Fig.
3e). When these empty vector-,
BMPR-IA-, and
BMPR-IB-transfected cells were cultured with TGF-
1,
differentiation into MHC-positive cells was greatly suppressed in all
the cultures examined (Fig. 3, c, f, and
i).
We examined the expression of receptor proteins in the subclones
transfected with BMPR-IA (
IA-12),
BMPR-IB (
IB-12), or the
empty vector (Vec-16) by the affinity labeling assay using 125I-BMP-2. Several 125I-BMP-2-cross-linked
proteins with lower and higher molecular masses were detected in
IA-12 cells, which disappeared by adding cold BMP-2 (Fig.
4, lanes 1 and 2).
However, no appreciable signals of the proteins labeled with
125I-BMP-2 were observed in
IB-12 (Fig. 4, lane
3) and Vec-16 cells (lane 5).
The Effects of BMP-2 and TGF-
To further determine
the role of BMPR-IA on BMP-2 signalings in C2C12 cells, we examined
mRNA expression of BMP-2 and TGF- receptors and muscle- and
osteoblast-specific phenotypes by Northern blotting in Vec-16,
IA-12, and
IB-12 cells. When the mRNA levels of BMPR-IA were
analyzed using an extracellular domain probe, endogenous BMPR-IA
mRNA was detected in all the cell lines at an equivalent level
(Fig. 5). Truncated BMPR-IA mRNA was
also detected in
IA-12 cells (Fig. 5). A kinase domain probe of
BMPR-IA was hybridized with the endogenous mRNA but not with the
truncated BMPR-IA mRNA (data not shown). Truncated BMPR-IB mRNA
was detected only in
IB-12 cells, but endogenous BMPR-IB mRNA
was not detected in any of these cell lines. All three cell lines
expressed T
R-I mRNA, irrespective of the presence and absence of
BMP-2 or TGF-
1. In Vec-16 cells, BMP-2 suppressed the expression of
muscle-specific mRNAs such as MCK and myogenin and induced
osteocalcin mRNA, a specific marker for osteoblasts. These
muscle-specific mRNAs were expressed in
IA-12 cells even in the
presence of BMP-2. Although myogenin mRNA was weakly detected in
BMP-2-treated
IB-12 cells, BMP-2 completely suppressed the
expression of MCK mRNA and strongly induced osteocalcin mRNA.
However, TGF-
1 suppressed the expression of the muscle-specific
mRNAs in all cell lines tested.
IA-12 cells did not express
osteocalcin mRNA, irrespective of the presence or absence of
BMP-2.
The Impaired ALP-inducing Activity of BMP-2 in
We performed a rescue experiment to examine
whether the blocked signals of BMP-2 in BMPR-IA-transfected cells
are rescued by the overexpression of wild-type type I receptors. The
wild-type BMPR-IA, BMPR-IB, and T
R-I were transiently transfected
into
IA-12 and
IA-14 cells, and they were incubated with or
without BMP-2. The BMP-2 signals were evaluated by the induction of
ALP. A number of ALP-positive cells appeared when the wild-type BMPR-IA was overexpressed in
IA-12 and
IA-14 cells (Fig.
6, b and f). However, transfection with either an empty vector or wild-type BMPR-IB
or T
R-I failed to increase the number of ALP-positive cells in
IA-12 and
IA-14 cells (Fig. 6, a, c,
d, e, g, and h). ALP-positive cells did not appear in the absence of BMP-2 (data not
shown).
To confirm that C2C12 cells are capable of expressing exogenously
introduced wild-type BMPR-IB proteins on the cell surface, we examined
the binding ability of exogenously introduced wild-type BMPR-IB
proteins in IA-12 cells by the affinity cross-linking and
immunoprecipitation using 125I-BMP-2. A weak but
significant band corresponding to the wild-type BMPR-IB was observed in
the immunoprecipitates of
IA-12 cells transfected with the wild-type
BMPR-IB (Fig. 7a). To further
examine the presence of the signaling pathway from BMPR-IB in C2C12
cells, we transiently transfected the constitutively active BMPR-IB
into
IA-12 cells. A number of ALP-positive cells appeared in the
culture even in the absence of BMP-2 (Fig. 7b).
In this study, we examined the involvement of the functional type I receptor for BMP-2 in C2C12 myoblasts in the molecular mechanism of osteoblast differentiation induced by BMPs. We reported that BMP-2 inhibits the differentiation of C2C12 cells into multinucleated myotubes and converted their differentiation pathway into osteoblast lineage cells (33). This model appears to reflect an early stage of osteoblast differentiation from non-osteogenic cells during BMP-induced bone formation in muscular tissues.
The intracellular signals of the TGF- superfamily appear to be
propagated via their own type I receptors. Two type I receptors, BMPR-IA and BMPR-IB, have been identified for BMP-2/BMP-4. C2C12 cells
expressed BMPR-IA mRNA but not BMPR-IB mRNA as detectable levels by Northern blotting. BMPR-IB mRNA was barely detected by
reverse transcription-PCR with excessive amplification. Moreover, reducing the mRNA expression of TGF-
type I receptor causes
resistance to the growth suppression induced by TGF-
in mink lung
epithelial cells (46). This suggests that when the type I receptor is
expressed at a low level, ligand-induced signals cannot be transduced.
The present study suggests that the BMP-2 signals are mainly transduced via BMPR-IA but not via BMPR-IB in C2C12 cells because BMPR-IB was
expressed at an extremely low level in those cells. To elucidate the
involvement of BMPR-IA in the BMP-2 signaling in C2C12 cells, a kinase
domain-truncated BMPR-IA was constructed and stably transfected into
C2C12 cells. This truncated BMPR-IA acts as a dominant negative receptor for BMP-4 signaling in the Xenopus embryo (23, 34, 35). When the kinase domain-truncated BMPR-IA was overexpressed in
C2C12 myoblasts, they differentiated into MHC-positive myotubes but not
into osteoblast-like ALP-positive cells, even in the presence of BMP-2.
In contrast, the differentiation into myotubes of the truncated
BMPR-IA-transfected cells was inhibited by TGF-
1. Moreover, the
impaired ALP-inducing activity of BMP-2 was rescued when the wild-type
BMPR-IA was overexpressed in the cells expressing truncated BMPR-IA.
These results suggest that BMPR-IA transduces two distinct signals
specific for BMP-2 in C2C12 myoblasts, namely, the inhibition of
myogenic differentiation and the induction of osteoblast
differentiation.
It is believed that the kinase domain-truncated type I receptor for the
TGF- superfamily inhibits the ligand-induced signals through
competition during the formation of the complex of endogenous type I
and type II receptors. C2C12 cells expressed BMPR-II
mRNA.2 This suggests that
the BMP-2 signals are transduced by the complex of BMPR-IA and BMPR-II
in these cells. It was reported that both BMPR-IA and BMPR-IB formed
complexes with BMPR-II or DAF-4, a type II receptor for a homologue of
BMP-2/BMP-4 in C. elegans, in a ligand-dependent
manner when overexpressed together on the surface of COS cells (21, 22,
25-27). In contrast to the truncated BMPR-IA, the kinase
domain-truncated BMPR-IB did not inhibit the BMP-2 signaling in C2C12
cells even if its mRNA was abundantly expressed. Moreover, the
truncated BMPR-IB was not cross-linked with 125I-BMP-2.
These results indicate that the truncated BMPR-IB does not bind BMP-2
with a high affinity in C2C12 cells. Interestingly, Nohno et
al. (27) reported that a truncated BMPR-II (BRK-3) did not
efficiently cross-link to 125I-BMP-4 when expressed alone
in COS cells. However, coexpression of the truncated BMPR-II with the
type I receptor enhanced the binding to 125I-BMP-4.
Although we could not detect endogenous type I nor type II BMP
receptors in
IB-12 and Vec-16 cells by cross-linking using 125I-BMP-2, these endogenous receptors were detected in the
truncated BMPR-IA-overexpressed cells (Fig. 4). Taking together, these
results suggest that the endogenous receptors require sufficient amount of type I and type II receptors to bind ligands with a high
affinity.
In the present study, wild-type BMPR-IB failed to rescue the impaired
ALP-inducing activity of BMP-2 in C2C12 cells transfected with
truncated BMPR-IA. The reason for this is not clear, but there are at
least two possible explanations. First, the downstream signaling
pathways of BMPR-IA and BMPR-IB may be different, and C2C12 cells lack
the pathway for BMPR-IB. To test this notion, we examined the effects
of a constitutively active mutant BMPR-IB on the induction of ALP
activity. Wieser et al. (47) and Attisano et al.
(48) reported that similar mutant receptors of TR-I and activin
type-IB constitutively activate their kinases and transduce the
ligand-induced signals in the absence of the ligands and appropriate
type II receptors. Transfection of constitutively active BMPR-IB into
truncated BMPR-IA-expressing C2C12 cells induced ALP activity even in
the absence of BMP-2. Moreover, the stable transfection of
constitutively active BMPR-IB into C2C12 cells inhibited their
differentiation into in the absence of BMP-2 (49). These results
clearly indicate that C2C12 cells have a functional downstream
signaling pathway for BMPR-IB.
Secondly, BMPR-IB may not be able to associate with BMPR-II (or other
type II receptors that are associated with BMPR-IA), or BMPR-IB may not
be activated by BMPR-II in C2C12 cells. Although C2C12 cells expressed
BMPR-II mRNA, it is not clear whether BMPR-IA and BMPR-IB share the
same BMPR-II that was expressed at the physiological level in C2C12
cells. If BMPR-IB associated with the same type II receptor for BMPR-IA
in C2C12 cells, a kinase domain-truncated BMPR-IB should have inhibited
the BMP-2-induced signals in these cells. The affinity binding assay
followed by immunoprecipitation indicates that the wild-type BMPR-IB
was expressed on the cell surface and bound BMP-2 in
BMPR-IA-expressing C2C12 cells (Fig. 7a). Moreover, we
recently found that the wild-type BMPR-IB-overexpressed C2C12 cells
responded to another member of the BMP
subfamily.3 These results
suggest that the wild-type BMPR-IB can transduce BMP signaling when its
appropriate ligand is present. Interestingly, Liu et al.
(26) reported that BMPR-IB·BMPR-II complex could bind BMP-2 and BMP-7
although BMP-2 could not activate the transcription of reporter genes
in COS cells. These results suggest that BMPR-IB is associated with a
different type II receptor from that for BMPR-IA. Further studies are
needed to prove this hypothesis.
At present, a mediator of the osteoblast differentiation-inducing signals of BMP-2 that acts at a downstream site of BMP type I receptor has not been elucidated. Genetic studies in Drosophila have identified Mothers against dpp (Mad), a gene that acts downstream of the signals of DPP, a Drosophila homologue of mammalian BMP-2/BMP-4 (50, 51). Mad-related proteins mimic the BMP-4 and activin effects in Xenopus (52, 53). Recently, a human homologue of Mad was identified as a possible signaling molecule for BMP-2 and BMP-4 in mammals (53, 54). The Mad-related protein was expressed in the cytoplasm and accumulated into the nucleus by the BMP-2/BMP-4 signals that were transduced by BMPR-IA or BMPR-IB (53, 54). These results suggest that Mad and/or its related proteins also mediate the osteoblast differentiation-inducing signals of BMP-2 via the activated type I receptor in some cells, including C2C12 myoblasts. Further studies are necessary to confirm this notion.
In conclusion, both BMP-2-induced signals, namely, the inhibition of myogenic differentiation and the induction of osteoblast differentiation, are transduced via BMPR-IA in C2C12 myoblasts. The downstream signaling pathway of BMPR-IA is involved in the ectopic bone formation induced by BMP-2 in muscular tissues. This in vitro model of C2C12 myoblasts is a useful system with which to investigate the molecular mechanism of not only osteoblast differentiation, but also of the BMP-2 signaling pathways.
We thank Dr. K. Maruyama (Tokyo Medical and Dental University) for providing pMIKHygB. We are also grateful to Drs. A. Yamaguchi (Showa University) and M. Doi (Yamanouchi Pharmaceutical Co.) for valuable discussion. The antibody MF-20 was obtained from the Developmental Studies Hybridoma Bank maintained at Johns Hopkins University School of Medicine, Baltimore, MD and at the University of Iowa, Iowa City, IA, under the contract N01-HD-6-2915 from the NICHD, National Institutes of Health.