JBJS | Article

The Journal of Bone & Joint Surgery, Volume 83, Issue 1 suppl 1

Scientific Article | March 01, 2001

The Crystal Structure of the BMP-2:BMPR-IA Complex and the Generation of BMP-2 Antagonists

Joachim Nickel, PhD; Matthias K. Dreyer, PhD; Thomas Kirsch, PhD; Walter Sebald, PhD

Investigation performed at Physiologische Chemie II, Biozentrum der Universität Würzburg, Würzburg, Germany
Joachim Nickel, PhD
Matthias K. Dreyer, PhD
Walter Sebald, PhD
Physiologische Chemie II, Biozentrum der Universität Würzburg, Am Hubland, 97074 Würzburg, Germany. E-mail address for W. Sebald: sebald@biozentrum.uni-wuerzburg.de
Thomas Kirsch, PhD
Medigene, Lochhamer Str. 11, 82152 Martinsried, Germany
In support of their research or preparation of this manuscript, one or more of the authors received grants or outside funding from Deutsche Forschungsgemeinschaft (DFG) Grant SFB 487 TP B1 and B2. None of the authors received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the authors are affiliated or associated.

The Journal of Bone & Joint Surgery. 2001; 83:S7-S14

5 Recommendations (Recommend) | 3 Comments | Saved by 3 Users Save Case

Article

Figures

References

text A A A

Abstract

Background: Bone morphogenetic proteins (BMPs) and growth and differentiation factors (GDFs) belong to the large transforming growth factor-ß (TGF-ß) superfamily of multifunctional cytokines. Signaling of the BMPs requires the binding of the BMP to the BMP cell surface receptors BMPR-IA, BMPR-IB, and BMPR-II. Similar to other cytokines, members of the TGF-ß superfamily exhibit stringent specificity in their ligand-receptor interactions, which may be a reason for the qualitative and quantitative differences in cellular responses. To understand how BMPs and GDFs activate their receptors, it is important to determine structure and binding mechanisms of ligand-receptor complexes. We have used BMP-2 as a key representative of the BMPs to identify the epitopes for type I and type II receptor binding by mutational interaction analyses and have solved the crystal structure of a BMP-2:BMPR-IA receptor ectodomain complex.

Methods: To identify amino acid side chains involved in receptor binding, a collection of in vitro mutagenized human BMP-2 variants was prepared and subjected to interaction analyses with use of the receptor ectodomains of BMPR-IA, BMPR-II, and ActR-II immobilized on a biosensor system. The biological activity of the BMP-2 variants was measured by BMP-2 dependent expression of alkaline phosphatase (ALP) in C2C12 cells. For crystallization, a complex of BMP-2 and the ectodomain of BMPR-IA was formed in solution, purified, and crystallized as described ¹² .

Results: The ligand-receptor interaction analysis of the BMP-2 variants identified distinct epitopes for type I and type II receptor binding. Because the structure of TGF-ß-like proteins has been compared with that of an open hand, the binding epitope for the type I receptor was-on the basis of its location-termed "wrist" epitope. The crystal structure of the BMP-2:BMPR-IA ectodomain complex revealed a key feature of the ligand-receptor interaction: a large hydrophobic residue (Phe85) within a hydrophobic patch of BMPR-IA fit into a hydrophobic pocket composed of residues of both BMP-2 monomers. A second epitope identified by alanine mutagenesis scanning was termed the "knuckle" epitope on the basis of its location on the outer side of the "finger" segments of BMP-2. Mutations in either the wrist epitope or the knuckle epitope produced variants with altered biological activities. Variants with antagonistic properties were exclusively generated by mutations in the knuckle epitope of BMP-2.

Conclusions and Clinical Relevance: The identification and characterization of the two receptor binding epitopes in BMP-2 provide new insight into the primary steps of BMP-receptor activation. Because of the structural similarities between members of the TGF-ß superfamily, it can be assumed that the data presented in this work are transferable to other TGF-ß receptor systems. Because of the association with various diseases, the generation of antagonists of other TGF-ß superfamily members might generate potent tools for basic research and therapeutic approaches.

Figures in this Article

Introduction

Introduction | Materials and Methods | Results | Discussion | References

Bone morphogenetic protein-2 (BMP-2), a member of the transforming growth factor-ß (TGF-ß) superfamily of multifunctional cytokines, induces bone formation and regeneration in adult vertebrates ^22,23 and plays an important role in the early embryonal development of animals ⁸ . Signaling of TGF-ß superfamily members requires the binding of the BMP molecule to cell surface receptors consisting of two types of transmembrane serine/threonine kinase receptor chains classified as type I and type II ^6,18 . The type II receptor transphosphorylates and thus activates the type I receptor chain, which then phosphorylates Smad proteins ⁷ . On phosphorylation, these signal transducers (the Smad proteins) translocate into the nucleus and mediate transcription of responsive genes. The mechanisms of receptor binding for BMP-2, TGF-ß, and the activins are different ¹⁸ .

The ordered sequential binding mechanism of TGF-ßs involves (a) the binding of the ligand to the type II receptor and (b) within the membrane, the recruitment of the type I receptor into the complex ^27,30 . Accordingly, the type II receptor represents the high-affinity receptor for TGF-ßs.

In contrast, BMP-2 binds with high affinity to the type I receptors BMPR-IA, BMPR-IB, and probably also to ActR-I ^13,17,28 . The type II receptors BMPR-II and ActR-II bind solute BMP-2 with lower affinity. Another difference between the TGF-ß and BMP-2 receptor systems was observed in the oligomerization pattern of the two receptor types. TGF-ß type I and type II receptors have been shown to be completely homodimeric in the absence of ligand ^15,16,29 , whereas BMP-2 receptors show a more flexible oligomerization pattern. A certain level of preexisting heterodimeric and homodimeric receptor complexes could be detected in live cells in the absence of BMP-2. Ligand binding leads to elevated levels of hetero-oligomerization and homo-oligomerization ³ . The more flexible oligomerization pattern might be important for the multiplicity in biological activities demonstrated for the BMP-receptor system. For a better understanding of these diversities in biological function mediated by distinct BMP and GDF members, it is important to quantify binding affinities of ligand-receptor interactions and to characterize the epitopes for type I and type II receptor binding.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1-A:: Figs. 1-A and 1-B The structure of the bone morphogenetic protein (BMP)-2:BMPR-IA ectodomain complex. Fig. 1-A The two BMP-2 monomers forming the homodimer are shown in blue and magenta, and the two ectodomain molecules of BMPR-IA are in gold. The view is from the side with the C-termini of the receptor chains pointing toward the cell membrane. The N-termini of the ligand and the receptor molecules are indicated.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 1-B:: Figs. 1-A and 1-B The structure of the bone morphogenetic protein (BMP)-2:BMPR-IA ectodomain complex. Fig. 1-B Close-up view of the BMP-2:BMPR-IA binding interface. Residue Phe85 of the receptor (golden) and residues forming the hydrophobic pocket on the ligand are indicated. (Reprinted, with permission, from Kirsch T, Sebald W, Dreyer MK. Crystal structure of the BMP-2-BRIA ectodomain complex. Nat Struct Bio 2000; 7:492-6.)

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2-A:: Figs. 2-A and 2-B Interaction of bone morphogenetic protein (BMP)-2 variants with type I (BMPR-IA) and type II (BMPR-II) receptor ectodomains. Fig. 2-A Amino acid exchanges in the constructed BMP-2 variants are indicated above the wild type BMP-2 sequence. Variants with reduced binding affinity to BMPR-IA are indicated in blue, and those with decreased binding affinities for BMPR-II are indicated in red in the respective substituted positions. Positions without color indicate no measurable alteration in binding affinities for both receptor types or that the variant could not be isolated in amounts required for functional analysis. The location of secondary structure elements as ß-sheets (ß1-ß9) and a-helix (a3) was adapted from Scheufler et al. (1998) ²⁵ .

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 2-B:: Figs. 2-A and 2-B Interaction of bone morphogenetic protein (BMP)-2 variants with type I (BMPR-IA) and type II (BMPR-II) receptor ectodomains. Fig. 2-B Plot of the relative binding affinities to BMPR-IA (as K d variant / K d BMP-2) versus those to BMPR-II (as EQ 45 variant / EQ 45 BMP-2). (Reprinted, with permission, from Kirsch T, Nickel J, Sebald W. BMP-2 antagonists emerge from alterations in the low-affinity binding epitope for receptor BMPR-II. EMBO J 2000; 3314-24.)

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 3:: Location of residues determining type I or type II receptor binding. On the space-filling model of the dimeric bone morphogenetic protein (BMP)-2 molecule (monomers shown in different gray shades), the residues determining type II receptor binding (knuckle epitope) are indicated in red. Residues determining the wrist epitope are shown on one monomer in blue and on the other monomer in purple. The location of wrist and knuckle epitopes is shown in a small ribbon model of BMP-2 (top view). (Reprinted, with permission, from Kirsch T, Nickel J, Sebald W. BMP-2 antagonists emerge from alterations in the low-affinity binding epitope for receptor BMPR-II. EMBO J 2000; 3314-24.)

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 4-A:: Figs. 4-A and 4-B Biological activity and antagonistic properties of bone morphogenetic protein (BMP) variants. Fig. 4-A The dose-dependent induction of alkaline phosphatase activity in serum-starved C2C12 cells is shown for wild type BMP-2 (black circles), P50A (a type I receptor mutant, blue triangles), and A34D as a representative of a type II receptor mutant (red diamonds). The background absorption at 405 nm (0.08 0.02) was not subtracted to indicate the signal-to-noise ratio.

View Large | Download Slide(.ppt)
Add to My JBJS

Fig. 4-B:: Figs. 4-A and 4-B Biological activity and antagonistic properties of bone morphogenetic protein (BMP) variants. Fig. 4-B In a similar assay, the dose-dependent inhibition of wild type BMP-2 (10 n M BMP-2) induced alkaline phosphatase activity is shown for the type II receptor mutants L90A (red triangles) and A34D (pink circles). (Reprinted, with permission, from Kirsch T, Nickel J, Sebald W. BMP-2 antagonists emerge from alterations in the low-affinity binding epitope for receptor BMPR-II. EMBO J 2000; 3314-24.)

Materials and Methods

Introduction | Materials and Methods | Results | Discussion | References

Preparation of BMP-2 Variants

BMP-2 cDNA was prepared, corresponding to residues 283-396 of the mature protein plus a N-terminal oligonucleotide. The resulting BMP-2 mutant proteins were expressed in Escherichia coli , isolated from inclusion bodies, renatured, and purified as described ¹⁰ .

Preparation of Recombinant Receptor Ectodomains

The extracellular domain of BMPR-IA was expressed in E. coli and purified as described ¹¹ . The C-terminal His6-tagged extracellular domains of ActR-II and BMPR-II were expressed in SF9-insect cells and purified from the culture medium by standard techniques including Ni-NTA-Agarose (Qiagen, Hilden, Germany) and BMP-2-affinity chromatography ¹⁰ . Purified receptors were N-biotinylated with sulfo-NHS-LC-biotin (Pierce, Rockford, IL, U.S.A.) as described ²⁶ .

Biosensor Interaction Analysis

The binding of the BMP-2 variants to immobilized receptor ectodomains was examined with use of the BIA2000 system (Biacore, Uppsala, Sweden) ¹⁰ . Equilibrium binding of BMP-2 variants to ActR-II and BMPR-II at 45 n M (EQ ₄₅ ) was measured twice in duplicate with a maximal SD of 20%. The rate constants K _d of the interaction between BMP-2 variants and BMPR-IA were calculated from the association rate constant k _on and the dissociation rate constant k _off of mean values of at least 12 measurements with three different concentrations of the ligand. Because of an unproved stoichiometry of ligand-receptor interaction, only apparent and not absolute values are presented.

Crystallization

BMP-2 and the ectodomain of BMPR-IA were expressed in E. coli and purified as described ¹¹ . Crystals were obtained in hanging drops as described ¹² .

Results

Introduction | Materials and Methods | Results | Discussion | References

Detailed information concerning ligand-receptor interaction (the binding of BMP-2 to the BMPR-IA receptor) could be achieved by solving the crystal structure of the BMP-2:BMPR-IA ectodomain complex ( Fig. 1-A ). The receptor molecules bind to the finger-helix-grooves of the BMP-2 dimer in a way that each receptor chain is in contact with both BMP-2 monomers. The C-termini of the receptor chains are clearly separated by 65 Å. The overall structure of BMPR-IA as the first solved structure of a type I receptor chain can be compared with that of an open left hand where the thumb is represented by the helix, the three middle fingers by the central ß-sheet, and the little finger, slightly bent to the palm of the hand, by the ß1-ß2-sheet including loop 1 ¹² . As expected, the general fold of the BMPR-IA ectodomain is similar to that of the murine ActR-II and to a number of snake venom neurotoxins and has been named the three-finger-toxin fold. Despite the structural similarity of receptors and toxins, their sequence identities are low and no functional similarities have been found to date. Major structural differences between BMPR-IA and mActR-II are loop 1 and helix a1. The latter structure, which is missing in ActR-II, contains residue Phe85, which sticks out of the helix and fits "knob-into-hole"- like into a hydrophobic pocket on the ligand, where it is surrounded by hydrophobic residues of both BMP-2 monomers ( Fig. 1-B ). Since all of the pocket-forming residues are invariant or at least highly conserved within the TGF-ß superfamily and a large hydrophobic residue corresponding to Phe85 of BMPR-IA is found in all type I receptor chains except the activin receptor-like kinase I (ALK-I), we propose that the "knob-into-hole" motif is the key feature in ligand-type I receptor recognition. Consequently, we assume that type I receptor binding is always mediated by this feature irrespective of binding affinity.

In parallel to the crystallization, a mutational analysis scan of the BMP-2 molecule was performed to obtain additional information about the contribution of the substituted side chains on binding energy and to identify the epitope for type II receptor binding. In a first round of mutagenesis, 20 residues with surface-exposed side chains were chosen covering net-like the whole BMP-2 molecule. After the first results, mutagenesis was expanded to juxtaposed residues of variants with considerable changes in affinities for either type I or type II receptor binding. Finally, charged residues were introduced, in the expectation of a more disruptive potential. From the originally designed BMP-2 mutants, 42 variants substituted at 40 different positions could be isolated and purified in a yield sufficient for analyses of receptor binding and biological activity ( Fig. 2-A ). With use of the immobilized ectodomain of BMPR-IA, differences in the rate constants for ligand-receptor association (k _on ) and dissociation (k _off ) of the BMP-2 variants were analyzed by means of a biosensor system. The binding affinity of wild type BMP-2 evaluated from the concentration dependence of BMP-2 equilibrium binding yields an apparent K _d of ~1 n M¹⁰ . This affinity is in a range of the high-affinity binding reported for other ligand-receptor interactions, e.g., interleukin-4 (IL-4) and the IL-4 receptor a chain ^2,26 . For BMP-2, the high affinity results mainly from the slow dissociation rate of the ligand (apparent k _off ~4 × 10 ^-4 S ^-1 ), suggesting a half-life of the complex of ~0.5 hours. The association rate (apparent k _on ~6 × 10 ⁵ M ^-1 S ^-1 ) is comparable with that of other receptor systems ¹⁰ .

A subset of variants (marked blue, see Fig. 2-A ) exhibited a specifically decreased binding affinity due to either increased k _off rates (e.g., variants W31A, W31C, A52R, and S69R) with values up to 30-fold higher or decreased k _on rates with values up to 10-fold lower than that of wild type BMP-2 (e.g., V26A, H54E, F49A, and P50A). The altered rate values were observed for BMPR-IA but not for the BMPR-II-interaction and therefore cannot be caused by an instability or impurity of the variant. Binding of BMP-2 and BMP-2 variants to the immobilized ectodomain of BMPR-II could also be recorded despite the low affinity, which was evaluated from the concentration dependence of equilibrium binding with an apparent K _d of ~100 n M . Because of the quite high kinetic constants for this interaction (k _on > 10 ⁶M^-1 S ^-1 , k _off > 10 ^-2 S ^-1 ), a reliable evaluation of k _on and k _off values was prevented. Therefore, differences in binding affinity of wild type BMP-2 and BMP-2 variants to BMPR-II were evaluated from changes in the equilibrium binding at a concentration of 45 n M of the ligand (EQ ₄₅ ). Five mutants (marked red, Fig. 2-A ) revealed an up to 15 times reduced EQ ₄₅ (e.g., A34D, H39D, S88A, L90A, and L100A). These alterations were specific for the interaction with BMPR-II, whereas binding to BMPR-IA was not affected. As a summary, the relative binding affinities of the tested mutants to BMPR-IA or BMPR-II are shown in Fig. 2-B . Variants with a reduced binding affinity to BMPR-IA are marked in blue, and those exhibiting a reduced binding affinity to BMPR-II are marked in red. The amino acid positions of type I receptor and type II receptor mutants are distributed along the whole sequence of BMP-2 and represent two non-overlapping subsets ( Fig. 2-A ). But, as shown in a space-filling model, these subsets form two separate epitopes on the surface of the homodimeric BMP-2 molecule ( Fig. 3 ). Determinants for BMPR-IA interaction co-localize in an epitope comprising residues from both BMP-2 monomers, which fits nicely to the complementary data obtained from the crystal structure determination ¹² . Because the monomeric structure of TGF-ß-like proteins has been compared with that of an open hand ^5,19 , with the central a-helix (a3) being the wrist and the two aligned two-stranded ß-sheets representing the four fingers, the identified epitope for type I receptor binding was designated the "wrist" epitope. The second epitope determining type II receptor binding is located on the back of the hand near the outer finger segments and is therefore termed the "knuckle" epitope.

To test the biological activity of the BMP-2 variants, C2C12 cells were used that, at serum starvation and addition of BMP-2, undergo differentiation toward an osteoblastic lineage. During differentiation, ALP expression is induced by BMP-2 in a dose-dependent manner ^1,20 . In this assay, small alterations in biological activity can be measured with a high reproducibility. Wild type BMP-2 induces a high ALP expression with an ED50 of 20 10 n M ( Fig. 4-A ). Several BMP-2 variants with reduced affinity for type I receptor binding, as well as variants with reduced type II receptor binding affinity, revealed a clearly reduced biological activity at a concentration of 250 n M¹⁰ . Some variants, e.g., P50A as a representative for a type I receptor mutant and A34D as an example for a type II receptor mutant, did not induce ALP activity at all. Most of the tested variants exhibited normal BMP-2-like or only slightly reduced responses. A few others showed reduced activities from 2 to 30% of that of wild type BMP-2 ¹⁰ .

In a similar experimental setup, C2C12 cells were stimulated with 10 n M of wild type BMP-2 and increasing concentrations of the variants to test their antagonistic properties by inhibiting wild type BMP-2 mediated ALP induction. For most of the variants tested, ALP activities were found to be increased to different extents. A reduction of ALP activity was detected only by adding the variants S88A, L100A, L90A, or A34D ¹⁰ . The latter two revealed the most severe antagonistic effects. At a concentration of 250 n M , these mutants reduced the ALP activity to ~3% (L90A) and to <0.5 % (A34D) of that value induced by wild type BMP-2 in the absence of the mutants ( Fig. 4-B ). The half maximal inhibition (IC50) could be achieved at concentrations of 20-40 n M , which is the range of the ED50 (10-20 n M ) of BMP-2; this finding led to the assumption that the inhibitory effect is mediated by the competition for a common receptor binding site. All of the variants with antagonistic properties belong exclusively to the subset of mutants with a decreased affinity for BMPR-II binding (marked red, Fig. 2-B ). Therefore, antagonistic properties are generated by mutations in the low-affinity binding site, the knuckle epitope of the BMP-2 molecule.

Discussion

Introduction | Materials and Methods | Results | Discussion | References

The identification and characterization of two separate binding epitopes in BMP-2 and the detection of variants with antagonistic characteristics provide important insights into the primary steps of BMP receptor activation. Receptor binding epitopes have not been described to date for any other of the closely related TGF-ß superfamily members that signal by way of type I and type II receptor serine/threonine-kinases. The wrist epitope identified in the homodimeric BMP-2 molecule encompasses a large area that correlates with the high affinity for BMPR-IA binding. In accordance with the lower affinity to BMPR-II, the knuckle epitope seems to be smaller. Binding residues in the knuckle epitope of BMP-2 are provided by only one monomer and are located in sheets b3, b4, b7, and b8 ( Fig. 2-A ). The wrist epitope is highly discontinuous and is composed of different elements of both monomers. The distances between the juxtaposed knuckle and wrist epitopes are only 10-15 Å, whereas those between the two type I (~40 Å) and type II (~55 Å) receptors are much larger ¹⁰ . Since the small ectodomain of the receptors is linked to the transmembrane region only by a small spacer consisting of less than 12 residues, the short distances between the knuckle and wrist epitopes might be important for the interaction of the type I and type II receptor serine-kinases in the cytosol. The identification and detailed molecular characterization of the two binding epitopes in BMP-2 might also facilitate a better understanding in binding affinity and specificity of other members of the TGF-ß superfamily. Other well characterized members, GDF-5 and BMP-7, bind to a set of receptor chains that partially overlap with those that transduce signals of BMP-2 ^14,21,24,28 . Like BMP-2, both factors exhibit a high affinity for type I receptor binding, whereas type II receptor chains are bound only with low affinity. The binding affinity to the whole complex consisting of both type I and type II receptor chains is higher than that to either receptor chain alone. These similarities in ligand-receptor interaction suggest that the binding epitopes of these factors are located at the same sites. Additionally, most of the residues located in the wrist epitope of BMP-2 are invariant or replaced by isofunctional side chains in BMP-7 and GDF-5 (10 of 15 and 12 of 15, respectively), suggesting that binding specificity is determined only by a small subset of residues whereas binding affinity might be provided by the same set of residues in the wrist epitope of all BMP and GDF proteins. For the knuckle epitope, a similar specification into affinity and specificity-determining residues can be observed.

If these factors activate their respective receptors by the same mechanism as BMP-2 (i.e., by binding by a high-affinity wrist epitope and a low-affinity knuckle epitope), then variants with antagonistic properties might be generated easily by amino acid substitution in the knuckle epitope. For other members of the TGF-ß superfamily that bind to the type II receptor chains with high affinity (e.g., TGF-ßs or activins), the location of the low-affinity binding epitope is still unproved. Recently, mutations of two amino acid residues (W52A and D55A) in TGF-ß1 have been described to cause as much as a 5-fold reduction of biological activity in a mink lung cell assay ⁹ . These mutations are located in the N-terminal loop of helix a3, an area that in the BMP-2 contains five binding determinants of the wrist epitope. This strongly suggests that the binding epitope in TGF-ß1 is at a site similar to that in the wrist epitope in BMP-2. The antagonistic properties of the TGF-ß1 variants W52A and D55A indicate that in case of a high-affinity binding to type II receptor chains as demonstrated for the TGF-ß and activin receptor system, antagonists might be generated by mutations in an area homologous to the wrist epitope of BMP-2.

The definition of binding epitopes in BMP-2 and the finding of BMP-2 antagonists revealed valuable insights in the primary steps of BMP/GDF-receptor activation. It will be important to test whether these findings are applicable to other members of the TGF-ß superfamily. Since members of this superfamily have been associated with various diseases, the development of variants with antagonistic properties might in general produce powerful tools for basic research and therapeutic approaches.

References

Introduction | Materials and Methods | Results | Discussion | References

ChalauxE, Lopez-Rovira T, Rosa JL, Bartrons R,Fentura F. JunB is involved in the inhibition of myogenic differentiation by bone morphogenetic protein-2. J Biol Chem,273: 537-S543. 1998;273537 1998 [PubMed]

CunninghamBC,Wells JA. Comparison of a structural and a functional epitope. J Mol Biol,234: 554-S563. 1993;234554 1993 [PubMed]

GilboaL, Nohe A, Geissendörfer T, Sebald W, Henis YI,Knaus P. Bone morphogenetic protein receptor complexes on the surface of live cells: a new oligomerization mode for serine/threonine kinase receptors. Mol Biol Cell,11: 1023-S1035. 2000;111023 2000 [PubMed]

GreenwaldJ, Fischer WH, Vale WW,Choe S. Three-finger toxin fold for the extracellular ligand-binding domain of the type II activin receptor serine kinase. Nat Struct Biol,6: 18-S22. 1999;618 1999 [PubMed]

GriffithDL, Keck PC, Sampath TK, Rueger DC,Carlson WD. Three-dimensional structure of recombinant human osteogenetic protein 1: structural paradigm for the transforming growth factor beta superfamily. Proc Natl Acad Sci U S A,93: 878-S883. 1996;93878 1996 [PubMed]

HeldinCH. Dimerization of cell surface receptors in signal transduction. Cell,80: 213-S223. 1995;80213 1995 [PubMed]

HeldinCH, Miyazono K,ten Dijke P. TGF-beta signaling from cell membrane to nucleus through SMAD proteins. Nature,1390: 465-S471. 1997;1390465 1997

HoganBL. Bone morphogenetic proteins in development. Curr Opin Genet Dev,6: 432-S438. 1996;6432 1996 [PubMed]

HuangSS, Zhou M, Johnson FE, Shieh HS,Huang JS. An active site of transforming growth factor-ß(1) for growth inhibition and stimulation. J Biol Chem,274: 27754-S27758. 1999;27427754 1999 [PubMed]

KirschT, Nickel J,Sebald W. BMP-2 antagonists emerge from alterations in the low-affinity binding epitope for receptor BMPR-II. EMBO J,19: 3314-S3324. 2000;193314 2000 [PubMed]

KirschT, Nickel J,Sebald W. Isolation of recombinant BMIP receptor IA ectodomain and its 2:1 complex with BMP-2. FEBS Lett,468: 215-S219. 2000;468215 2000 [PubMed]

KirschT, Sebald W,Dreyer MK. Crystal structure of the BMP-2-BRIA ectodomain complex. Nat Struct Bio,7: 492-S496. 2000;7492 2000

KoenigBB, Cook JS, Wolsing DH, Ting J, Tiesman JP, Correa PE, Olson CA, Pecqet AL, Ventura F,Grant RA. Characterization and cloning of a receptor for BMP-2 and BMP-4 from NIH 3T3 cells. Mol Cell Biol,14: 5961-S5974. 1994;145961 1994 [PubMed]

LiuF, Ventura F, Doody J,Massague J. Human type II receptor for bone morphogenetic proteins (BMPs): extension of the two-kinase receptor model to the BMPs. Mol Cell Biol,15: 3479-S3486. 1995;153479 1995 [PubMed]

LuoK,Lodish HF. Signaling by chimeric erythropoietin-TGFb receptors: homodimerization of the cytoplasmic domain of the type I TGF-ß receptor and heterodimerization with the type II receptor are both required for intracellular signal transduction. EMBO J,15: 4485-S4496. 1996;154485 1996 [PubMed]

LuoK,Lodish HF. Positive and negative regulation of type II TGF-ß receptor signal transduction by autophosphorylation on multiple serine residues. EMBO J,16: 1970-S1981. 1997;161970 1997 [PubMed]

Macias-Silva M, Hoodless PA, Tang SJ, Buchwald M,Wrana JL. Specific activation of Smad1 signaling pathways by the BMP7 type I receptor, ALK2. J Biol Chem,273: 25628-S25636. 1998;27325628 1998 [PubMed]

MassagueJ. TGF-ß signal transduction. Annu Rev Biochem,67: 753-S791. 1998;67753 1998 [PubMed]

MittlPR, Priestle JP, Cox DA, McMaster G, Cerletti N,Grütter MG. The crystal structure of TGF-ß 3 and comparison to TGF-ß 2: implications for receptor binding. Protein Sci,5: 1261-S1271. 1996;51261 1996 [PubMed]

NamikiM, Akiyama S, Katagiri T, Suzuki A, Ueno N, Yamaji N, Rosen V, Wozney JM,Suda T. A kinase domain-truncated type I receptor blocks bone morphogenetic protein-2-induced signal transduction in C2C12 myoblasts. J Biol Chem,272: 22046-S22052. 1997;27222046 1997 [PubMed]

NishitohH, Ichijo H, Kimura M, Matsumoto T, Makishima F, Yamaguchi A, Yamashita H, Enomoto S,Miyazono K. Identification of type I and type II serine/threonine kinase receptors for growth/differentiation factor-5. J Biol Chem,271: 21345-S21352. 1996;27121345 1996 [PubMed]

ReddiAH. Bone morphogenetic proteins: an unconventional approach to isolation of first mammalian morphogens. Cytokine Growth Factor Rev,8: 11-S20. 1997;811 1997 [PubMed]

ReddiAH. Role of morphogenetic proteins in skeletal tissue engineering and regeneration. Nat Biotechnol,16: 247-S252. 1998;16247 1998 [PubMed]

RosenzweigBL, Imamura T, Okadome T, Cox GN, Yamashita H, ten Dijke P, Heldin CH,Miyazono K. Cloning and characterization of a human type II receptor for bone morphogenetic proteins. Proc Natl Acad Sci U S A,92: 7632-S7636. 1995;927632 1995 [PubMed]

ScheuflerC, Sebald W,Hülsmeyer M. Crystal structure of human bone morphogenetic protein-2 at 2.7 Å resolution. J Mol Biol,287: 103-S115. 1998;287103 1998

ShenBJ, Hage T,Sebald W. Global and local determinants for the kinetics of interleukin-4/interleukin-4 receptor alpha chain interaction: a biosensor study employing recombinant interleukin-4-binding protein. Eur J Biochem,240: 252-S261. 1996;240252 1996 [PubMed]

ten DijkeP, Yamashita H, Ichijo H, Franzen P, Laiho M, Miyazono K,Heldin CH. Characterization of type I receptors for transforming growth factor-ß and activin. Science,264: 101-S104. 1994;264101 1994 [PubMed]

ten DijkeP, Yamashita H, Sampath TK, Reddi AH, Estevez M, Riddle DL, Ichijo H, Heldin CH,Miyazono K. Identification of type I receptors for osteogenic protein-1 and bone morphogenetic protein-4. J Biol Chem,269: 16985-S16988. 1994;26916985 1994 [PubMed]

Weis-Garcia F,Massague J. Complementation between kinase-defective and activation-defective TGF-ß receptors reveals a novel form of receptor cooperativity essential for signaling. EMBO J,15: 276-S289. 1996;15276 1996 [PubMed]

WranaJL, Attisano L, Wieser R, Ventura F,Massague J. Mechanism of activation of the TGF-ß receptor. Nature,370: 341-S347. 1994;370341 1994 [PubMed]