Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 5 (9), e12846

Binding Between Crossveinless-2 and Chordin Von Willebrand Factor Type C Domains Promotes BMP Signaling by Blocking Chordin Activity

Affiliations

Binding Between Crossveinless-2 and Chordin Von Willebrand Factor Type C Domains Promotes BMP Signaling by Blocking Chordin Activity

Jin-Li Zhang et al. PLoS One.

Abstract

Background: Crossveinless-2 (CV2) is an extracellular BMP modulator protein of the Chordin family, which can either enhance or inhibit BMP activity. CV2 binds to BMP2 via subdomain 1 of the first of its five N-terminal von Willebrand factor type C domains (VWC1). Previous studies showed that this BMP binding is required for the anti-, but not for the pro-BMP effect of CV2. More recently, it was shown that CV2 can also bind to the BMP inhibitor Chordin. However, it remained unclear which domains mediate this binding, and whether it accounts for an anti- or pro-BMP effect.

Principal findings: Here we report that a composite interface of CV2 consisting of subdomain 2 of VWC1 and of VWC2-4, which are dispensable for BMP binding, binds to the VWC2 domain of Chordin. Functional data obtained in zebrafish embryos indicate that this binding of Chordin is required for CV2's pro-BMP effect, which actually is an anti-Chordin effect and, at least to a large extent, independent of Tolloid-mediated Chordin degradation. We further demonstrate that CV2 mutant versions that per se are incapable of BMP binding can attenuate the Chordin/BMP interaction.

Conclusions: We have physically dissected the anti- and pro-BMP effects of CV2. Its anti-BMP effect is obtained by binding to BMP via subdomain1 of the VWC1 domain, a binding that occurs in competition with Chordin. In contrast, its pro-BMP effect is achieved by direct binding to Chordin via subdomain 2 of VWC1 and VWC2-4. This binding seems to induce conformational changes within the Chordin protein that weaken Chordin's affinity to BMP. We propose that in ternary Chordin-CV2-BMP complexes, both BMP and Chordin are directly associated with CV2, whereas Chordin is pushed away from BMP, ensuring that BMPs can be more easily delivered to their receptors.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Figure 1
Figure 1. BIAcore analysis of CV2/Chordin interaction.
Sensograms showing the binding of 100,200,300,400,800 and 1200 nM CV2 (A) or CV2-N (B) to immobilized Chordin; 50, 100, 150, 200, 400 and 600 nM Chordin to immobilized CV2 (C); 100, 200 and 300 nM CV2-N (D, a,b,c) or CV2 (D,d,e,f) to immobilized VWC2 of Chordin; 10, 20, 30, 40, 80, 120 nM VWC2 of Chordin to immobilized CV2-N (E); 200 nM CV2-VWC1-4 (F) or CV2-VWC1-4M (G) to immobilized BMP2 (Fa, G,a) or VWC2 of Chordin (Fb, G,b); 300 nM CV2 ΔClip-SD1(H) or CV2 ΔSD2 (I) to BMP2 (Ha, Ia) or VWC2 of Chordin (Hb, Ib). (J) and (K) show schemes of the CV2 mutant constructs. Abbreviation: RU, resonance unit.
Figure 2
Figure 2. Binding of CV2 to Chordin promotes BMP signaling in vivo.
(A) Graphical illustration of proportions of phenotypes generated upon injection of different CV2 constructs. For classification from C5 (strong dorsalization) to V4 (strong ventralization), see Mullins et al and Kishimoto et al (38). Numbers of analyzed embryos are indicated above the columns. (B–H) Representative zebrafish embryos after mRNA injections as in (A). Lateral views of live embryos at 32 hpf.
Figure 3
Figure 3. The pro-BMP activity of CV2-T5P/I21R requires the presence of Chordin.
(A) Graphical illustration of proportions of dorsalized and ventralized phenotypes generated upon injection of cv2-T5P/I21R or bmp2b mRNA (in blue letters), cv2, chd or nog1 MOs (in red letters), or mixtures of the mRNA/MOs in different combinations. Numbers of analyzed embryos are indicated above the columns; columns are numbered. (B–J) Representative zebrafish embryos after mRNA and/or MO injection as in (A). Right panels show overall morphology of live embryos in lateral views at 32 hpf, left panels show embryos at 80% epiboly (mid gastrula stage), after whole mount otx2/eve1 double in situ hybridizations (domains indicated in B), lateral views, dorsal to the right. The expression of the dorsal marker otx2 is reduced in ventralized and expanded in dorsalized embryos, whereas the ventral marker eve1 shows contrary shifts.
Figure 4
Figure 4. CV2 can antagonize the dorsalizing effect of Chordin, but not of Noggin1, while the anti-Chordin effect of CV2 is both dependent and independent of Tolloid/Bmp1a.
(A) Graphical illustration of proportions of dorsalized and ventralized phenotypes generated upon injection of cv2-T5P/I21R mRNA into wild-type embryos (lane 9), into embryos after chordin or noggin1 overexpression (lanes 1–4), or into bmp2b (lanes 5,6) or tolloid/bmp1a morphants (lanes 7,8), in comparison to the response of the tolloid/bmp1a double morphants to co-injection of chordin MO (lanes 10,11) or sizzled MO (lanes 12,13). Numbers of analyzed embryos are indicated above the columns; columns are numbered. (B–O) Representative zebrafish embryos after mRNA and/or MO injections as in (A). Lateral views on live embryos at 32 hpf. (P) Western blot detecting C-terminally-Myc-tagged Chordin (upper panel) or GFP (lower panel) in zebrafish embryos after mRNA injections. Lanes were loaded with precipitates of extracts from 50 mid-gastrula stage embryos (85% epiboly stage; 9 hpf) that had been injected at the 1-cell stage with 300 pg chordin and – as injection control - gfp mRNA per embryo, or with the same amounts of chordin and gfp mRNA, plus 600 pg mRNA encoding CV2, Tolloid or TsgW67G, respectively. Sibling embryos from the various injections had been allowed to develop further and displayed dorsalization of similar strengths at 32 hpf. In the upper blot, in addition to Chordin, Myc-tagged full-length CV2 and its C-terminal fragment CV2-C could be detected.
Figure 5
Figure 5. Inhibition of Chordin/BMP2 binding by CV2 ΔClip mutant.
(A) BIAcore analysis. 500, 1000 and 3000 nM CV2 ΔClip were first perfused over Chordin immobilized on the BIAcore Chip for 120 seconds, in the second phase the same concentration of CV2 ΔClip (e,f,g, as a negative control) or CV2 ΔClip plus 100 nM BMP2 (b,c,d) were perfused. For a positive control, binding of 100 nM BMP2 alone to immobilized Chordin in a separate experiment is overlaid (a). (B) Co-immunoprecipitation. Mixtures of BMP2, His-tagged Chordin and different concentrations of His-tagged CV2 ΔClip were co-immunoprecipitated by anti-His-tag antibody and Protein-A sepharose (lanes 2–4), followed by Western blotting to detect BMP2 (upper panel) or His-tagged proteins (lower panel) with anti-BMP2 and anti-His-tag antibodies, respectively. Lanes 1, 5 and 6 are loading controls. The CV2 ΔClip protein was auto-catalytically processed into N-and C-terminal fragments, which were associated by disulfide bonds and separated in the reducing SDS gel. Only the His-tagged C-terminal fragment was detected in the Western blot. Abbreviation: RU, resonance unit.

Similar articles

See all similar articles

Cited by 14 articles

See all "Cited by" articles

References

    1. Umulis D, O'Connor MB, Blair SS. The extracellular regulation of bone morphogenetic protein signaling. Development. 2009;136:3715–3728. - PMC - PubMed
    1. Hammerschmidt M, Mullins M, editors. Dorsoventral patterning in the zebrafish: Bone morphogenetic proteins and beyond. Berlin Heidelberg: Springer-Verlag; 2002. pp. 72–95. - PubMed
    1. De Robertis EM, Kuroda H. Dorsal-ventral patterning and neural induction in Xenopus embryos. Annu Rev Cell Dev Biol. 2004;20:285–308. - PMC - PubMed
    1. De Robertis EM. Spemann's organizer and self-regulation in amphibian embryos. Nat Rev Mol Cell Biol. 2006;7:296–302. - PMC - PubMed
    1. Sasai Y, Lu B, Steinbeisser H, De Robertis EM. Regulation of neural induction by the Chd and Bmp-4 antagonistic patterning signals in Xenopus. Nature. 1995;377:757. - PubMed

Publication types

MeSH terms

LinkOut - more resources

Feedback