FLRG, a new ADAM12-associated protein, modulates osteoclast differentiation

BACKGROUND INFORMATION: FLRG (follistatin-related gene) is a secreted glycoprotein that is highly homologous with follistatin. These proteins are involved in the regulation of various biological effects mediated by their binding to TGF-beta (transforming growth factor-beta) superfamily members, activin A and bone morphogenetic proteins. To characterize further the function of FLRG, we used a yeast two-hybrid screen to look for other possible functional partners. RESULTS: We report a direct interaction between the cysteine-rich domain of FLRG and ADAM12 (a disintegrin and metalloprotease 12). ADAMs are metalloprotease-disintegrin proteins that have been implicated in cell adhesion, protein ectodomain shedding, matrix protein degradation and cell fusion. Several studies have reported that ADAM12 protein, as well as activin A, are important regulators of osteoclast differentiation. We observed that the expressions of ADAM12 and activin A are modulated during osteoclast formation, whereas the FLRG expression seemed to remain quite constant. We showed that the FLRG protein inhibits osteoclast differentiation from murine primary spleen cells and macrophage RAW264.7 cells cultured in the presence of RANK-L (receptor activator of nuclear factor kappaB ligand) and M-CSF (macrophage colony-stimulating factor). Addition of FLRG protein to precursors significantly reduces the number of osteoclasts, as well as the average number of nuclei in each osteoclast. CONCLUSIONS: Our study indicates that the FLRG protein may contribute to bone formation by inhibiting osteoclast differentiation.     

Testing the antagonistic effect of follistatin on BMP family members in ovine granulosa cells

Follistatin was first demonstrated as an activin-binding protein, neutralizing its actions. However, there is emerging evidence that follistatin inhibits the action of other members of the transforming growth factor beta(TGFbeta) / bone morphogenetic protein (BMP) superfamily. Recently, numerous BMP factors have been shown to play important roles in regulating folliculogenesis and ovulation rate in mammals, and such a potential antagonistic role of follistatin is of particular interest in the context of ovarian function. Using a biological test based on progesterone production by ovine primary granulosa cells in culture, we show that follistatin was a strong antagonist of activin A, but not BMP-2 or BMP-4 actions. In contrast, noggin, a known specific BMP antagonist, had no effect on activin A but strongly neutralized BMP-2 and BMP-4 actions. BMP-6 action was only slightly reduced by both follistatin and noggin. Our data led to the conclusion that follistatin would not represent a determinant physiological modulator of the biological effect of BMP factors on granulosa cells.     

A novel BMP expressed in developing mouse limb, spinal cord, and tail bud is a potent mesoderm inducer in Xenopus embryos

The bone morphogenetic proteins (BMPs) play critical roles in patterning the early embryo and in the development of many organs and tissues. We have identified a new member of this multifunctional gene family, BMP-11, which is most closely related to GDF-8/myostatin. During mouse embryogenesis, BMP-11 is first detected at 9.5 dpc in the tail bud with expression becoming stronger as development proceeds. At 10.0 dpc, BMP-11 is expressed in the distal and posterior region of the limb bud and later localizes to the mesenchyme between the skeletal elements. BMP-11 is also expressed in the developing nervous system, in the dorsal root ganglia, and dorsal lateral region of the spinal cord. To assess the biological activity of BMP-11, we tested the protein in the Xenopus ectodermal explant (animal cap) assay. BMP-11 induced axial mesodermal tissue (muscle and notochord) in a dose-dependent fashion. At higher concentrations, BMP-11 also induced neural tissue. Interestingly, the activin antagonist, follistatin, but not noggin, an antagonist of BMPs 2 and 4, inhibited BMP-11 activity on animal caps. Our data suggest that in Xenopus embryos, BMP-11 acts more like activin, inducing dorsal mesoderm and neural tissue, and less like other family members such as BMPs 2, 4, and 7, which are ventralizing and anti-neuralizing signals. Taken together, these data suggest that during vertebrate embryogenesis, BMP-11 plays a unique role in patterning both mesodermal and neural tissues.     

BMP-2 induces the expression of activin betaA and follistatin in vitro

Activins are members of the transforming growth factor beta (TGF-beta) superfamily and have been shown to be multifunctional regulators of development and cell differentiation. Increasing evidence suggests activin betaA is involved in skeletal development. Using differential display PCR we have identified activin betaA as a gene associated with recombinant human bone morphogenetic protein-2 (rhBMP-2) induced differentiation of a mouse limb bud cell line, MLB13MYC clone 17, from a prechondroblastic to an osteoblastic phenotype. The expression of activin betaA peaks at 24 h of rhBMP-2 treatment, before detection of osteocalcin mRNA expression. Cycloheximide treatment inhibits induction of activin betaA, indicating a requirement for new protein synthesis. The induction of the mRNA encoding follistatin, an activin binding protein, was also examined. Follistatin mRNA increases within 18 h of rhBMP-2 treatment, as activin betaA mRNA increases but before it peaks. Treatment of MLB13MYC clone 17 cells with purified activin betaA concomitant with rhBMP-2 does not affect markers of chondrocyte or osteoblast differentiation, nor does treatment with purified activin betaA alone. This suggests that activin betaA exerts its effect via a paracrine mechanism. In situ hybridization analysis demonstrates that activin betaA expression is localized to cells in the developing interphalangeal joints of embryonic mouse limbs. This is consistent with in vivo induction by BMP-2 which is also expressed in the developing joints. Activin betaA, therefore, is downstream from BMP-2 in the cascade of events that result in skeletal development.     

Follistatin regulates bone morphogenetic protein-7 (BMP-7) activity to stimulate embryonic muscle growth

Bone morphogenetic proteins (BMPs) can either promote growth of embryonic muscle by expanding the Pax-3-expressing muscle precursor population or restrict its development by inducing apoptosis. Follistatin, a proposed BMP antagonist, is expressed in embryonic muscle. Deficiency in Follistatin results in muscle defects and postnatal asphyxia. Here, we report that during chick limb development Follistatin enhances BMP-7 action to induce muscle growth but prevents the ability of BMP-7 to induce apoptosis and muscle loss. Follistatin, unlike another BMP-binding protein, Noggin, promotes Pax-3 expression and transiently delays muscle differentiation and thus exerts proliferative signalling during muscle development. We provide data which show that Follistatin binds BMP-7 and BMP-2 at low affinities and that the binding is reversible. These data suggest that Follistatin acts to present BMPs to myogenic cells at a concentration that permits stimulation of embryonic muscle growth.     

Regulation of myostatin in vivo by growth and differentiation factor-associated serum protein-1: a novel protein with protease inhibitor and follistatin domains

Myostatin, a member of the TGFbeta superfamily, is a potent and specific negative regulator of skeletal muscle mass. In serum, myostatin circulates as part of a latent complex containing myostatin propeptide and/or follistatin-related gene (FLRG). Here, we report the identification of an additional protein associated with endogenous myostatin in normal mouse and human serum, discovered by affinity purification and mass spectrometry. This protein, which we have named growth and differentiation factor-associated serum protein-1 (GASP-1), contains multiple domains associated with protease-inhibitory proteins, including a whey acidic protein domain, a Kazal domain, two Kunitz domains, and a netrin domain. GASP-1 also contains a domain homologous to the 10-cysteine repeat found in follistatin, a protein that binds and inhibits activin, another member of the TGFbeta superfamily. We have cloned mouse GASP-1 and shown that it inhibits the biological activity of mature myostatin, but not activin, in a luciferase reporter gene assay. Surprisingly, recombinant GASP-1 binds directly not only to mature myostatin, but also to the myostatin propeptide. Thus, GASP-1 represents a novel class of inhibitory TGFbeta binding proteins.     

A novel role for fibronectin type I domain in the regulation of human hematopoietic cell adhesiveness through binding to follistatin domains of FLRG and follistatin

FLRG and follistatin belong to the family of follistatin proteins involved in the regulation of various biological effects, such as hematopoiesis, mediated by their binding to activin and BMP, both members of the TGFbeta family. To further characterize the function of FLRG, we searched for other possible functional partners using a yeast two-hybrid screen. We identified human fibronectin as a new partner for both FLRG and follistatin. We also demonstrated that their physical interaction is mediated by type I motifs of fibronectin and follistatin domains. We then analyzed the biological consequences of these protein interactions on the regulation of hematopoiesis. For the first time, we associated a biological effect with the regulation of human hematopoietic cell adhesiveness of both the type I motifs of fibronectin and the follistatin domains of FLRG and follistatin. Indeed, we observed a significant and specific dose-dependent increase of cell adhesion to fibronectin in the presence of FLRG or follistatin, using either a human hematopoietic cell line or primary cells. In particular, we observed a significantly increased adhesion of immature hematopoietic precursors (CFC, LTC-IC). Altogether these results highlight a new mechanism by which FLRG and follistatin regulate human hematopoiesis.     

Skin cell induction of calcitonin gene-related peptide in embryonic sensory neurons in vitro involves activin

Target skin cells induce the neuropeptide calcitonin gene-related peptide (CGRP) in na?ve embryonic dorsal root ganglion (DRG) neurons in vitro, but the molecular basis of that induction is not known. Recombinant activin or bone morphogenetic proteins (BMPs) dramatically increase the number of sensory neurons with CGRP and substance P in vitro (X. Ai et al., 1999, Mol. Cell. Neurosci. 14, 506-518). These experiments were designed to test if activin or BMPs accounted for the CGRP-inductive activity by skin cells. To identify factors from skin that induce CGRP, we developed a bioassay in which embryonic DRG neurons isolated before peripheral target contact in vivo are challenged in vitro with specific factors. Conditioned medium from an embryonic rat skin cell line induced neuronal CGRP expression, and induction was blocked by follistatin, implicating transforming growth factor family members. Immunoblot analysis revealed that the skin cell line medium contained several activin and bone morphogenetic protein moieties. Antibody specific to activin neutralized most of the CGRP-inductive activity in skin conditioned medium. These data indicate that the CGRP-inductive action of skin cells involves activin and establish activin as a candidate regulator of this sensory neuropeptide phenotype during development.     

Site-specific interaction of bone morphogenetic protein 2 with procollagen II

Bone morphogenetic proteins (BMPs) play a critical role in embryo development, organogenesis, and regeneration of damaged tissues. Biological activity of BMPs depends on their local concentration, which is regulated by intracellular enzymatic processing of pro-BMPs, and then the binding of secreted BMPs to antagonizing extracellular proteins. It has been suggested that BMPs interact with structural proteins of the extracellular matrix, but this process is poorly understood. To study interactions of BMPs with fibrillar collagens in detail we expressed recombinant procollagen II variants in which specific domains that correspond to the D-periods were deleted. Subsequently, the procollagen II variants were used in biosensor and immuno-precipitation binding assays to map the regions of procollagen II with a high affinity for the BMP-2. Our data suggest that interaction of BMP-2 with procollagen II is site-specific, and that the high-affinity binding site is located in the D4-period of the collagen triple helix. We hypothesize that the binding of BMP-2 to collagen II reflects a general mechanism of interaction between the fibrillar collagens and morphogens that belong to the transforming growth factor (TGF)-beta superfamily.     

Type I receptor binding of bone morphogenetic protein 6 is dependent on N-glycosylation of the ligand

Bone morphogenetic proteins (BMPs), together with transforming growth factor (TGF)-beta and activins/inhibins, constitute the TGF-beta superfamily of ligands. This superfamily is formed by more than 30 structurally related secreted proteins. The crystal structure of human BMP-6 was determined to a resolution of 2.1 A; the overall structure is similar to that of other TGF-beta superfamily ligands, e.g. BMP-7. The asymmetric unit contains the full dimeric BMP-6, indicating possible asymmetry between the two monomeric subunits. Indeed, the conformation of several loops differs between both monomers. In particular, the prehelix loop, which plays a crucial role in the type I receptor interactions of BMP-2, adopts two rather different conformations in BMP-6, indicating possible dynamic flexibility of the prehelix loop in its unbound conformation. Flexibility of this loop segment has been discussed as an important feature required for promiscuous binding of different type I receptors to BMPs. Further studies investigating the interaction of BMP-6 with different ectodomains of type I receptors revealed that N-glycosylation at Asn73 of BMP-6 in the wrist epitope is crucial for recognition by the activin receptor type I. In the absence of the carbohydrate moiety, activin receptor type I-mediated signaling of BMP-6 is totally diminished. Thus, flexibility within the binding epitope of BMP-6 and an unusual recognition motif, i.e. an N-glycosylation motif, possibly play an important role in type I receptor specificity of BMP-6.     

Involvement of activin/BMP system in development of human pituitary gonadotropinomas and nonfunctioning adenomas

Roles of activin/bone morphogenetic protein (BMP) system in the pathogenesis of human pituitary adenoma remain unknown although these factors stimulate follicle-stimulating hormone (FSH) secretion in the normal pituitary. Here we demonstrated that type-I and -II subunit mRNAs of activin/BMP receptors are expressed in Pit-1-negative FSH-producing (FSH-oma) and nonfunctioning pituitary adenomas (NF-oma). Basal levels of serum FSH standardized by luteinizing hormone (LH) were markedly high in FSH-omas in contrast to NF-omas. However, gonadotropin-releasing hormone (GnRH)-induced increment of FSH standardized by that of LH was not changed in FSH-omas, suggesting that imbalanced FSH secretion by FSH-oma is not attributable to GnRH regardless of the expression of GnRH receptor. Although activin betaA subunit was detected in neither adenoma, the betaB subunit was expressed highly in FSH-omas and, to lesser extent, in NF-omas. As for BMPs, BMP-6 and -7 were detected in NF-omas while BMP-4 and -15 were not detected in either type of adenoma. In the presence of pituitary activin/BMP system, the levels of co-expressing follistatin mRNA in the tumors were reduced in FSH-oma compared with NF-oma, suggesting that endogenous follistatin is involved in FSH overproduction through inhibition of activin/BMP system independently of GnRH.     

Characterization and cloning of a receptor for BMP-2 and BMP-4 from NIH 3T3 cells.

The bone morphogenetic proteins (BMPs) are a group of transforming growth factor beta (TGF-beta)-related factors whose only receptor identified to date is the product of the daf-4 gene from Caenorhabditis elegans. Mouse embryonic NIH 3T3 fibroblasts display high-affinity 125I-BMP-4 binding sites. Binding assays are not possible with the isoform 125I-BMP-2 unless the positively charged N-terminal sequence is removed to create a modified BMP-2, 125I-DR-BMP-2. Cross-competition experiments reveal that BMP-2 and BMP-4 interact with the same binding sites. Affinity cross-linking assays show that both BMPs interact with cell surface proteins corresponding in size to the type I (57- to 62-kDa) and type II (75- to 82-kDa) receptor components for TGF-beta and activin. Using a PCR approach, we have cloned a cDNA from NIH 3T3 cells which encodes a novel member of the transmembrane serine/threonine kinase family most closely resembling the cloned type I receptors for TGF-beta and activin. Transient expression of this receptor in COS-7 cells leads to an increase in specific 125I-BMP-4 binding and the appearance of a major affinity-labeled product of approximately 64 kDa that can be labeled by either tracer. This receptor has been named BRK-1 in recognition of its ability to bind BMP-2 and BMP-4 and its receptor kinase structure. Although BRK-1 does not require cotransfection of a type II receptor in order to bind ligand in COS cells, complex formation between BRK-1 and the BMP type II receptor DAF-4 can be demonstrated when the two receptors are coexpressed, affinity labeled, and immunoprecipitated with antibodies to either receptor subunit. We conclude that BRK-1 is a putative BMP type I receptor capable of interacting with a known type II receptor for BMPs.     

Possible involvement of protein kinases and Smad2 signaling pathways on osteoclast differentiation enhanced by activin A.

Bone tissues reportedly contain considerable amounts of activin A and follistatin, an activin A-binding protein. In the present study, we found that follistatin strongly inhibited osteoclast formation in cocultures of mouse bone marrow cells and primary osteoblasts induced by 1alpha,25 dihydroxyvitamin D(3), prostaglandin E(2), and interleukin-1alpha. Antibody aganist activin A also inhibited the osteoclast formation. Furthermore, activin A synergistically stimulated osteoclast differentiation mediated by receptor activator NF-kappaB ligand (RANKL). RT-PCR analysis revealed that osteoblasts produced not only activin A but also follistatin. Western blot analysis of a panel of phosphorylated proteins revealed that activin A stimulated the phosphorylation of p44/42 mitogen activated protein (MAP) kinase (ERK1/2) and p38 MAP kinase in macrophage colony-stimulating factor-dependent bone marrow macrophages (M-BMMPhis). In addition, phosphorylation of Smad2 was observed in M-BMMPhis stimulated with activin A. These findings indicate that the phosphorylation of p44/42 MAP kinase, p38 MAP kinase, and Smad2 is involved in activin A-enhanced osteoclast differentiation induced by RANKL. Taken together, these results suggest that both activin A and follistatin produced by osteoblasts may play an important role in osteoclast differentiation through MAP kinases and Smad2 signaling pathways.