spacer gif spacer gif spacer gif spacer gif spacer gif
 QUICK SEARCH:   [advanced]


spacer gif
     Home     Help     Feedback     Subscriptions     Archive     Search     Table of Contents    

First published online 26 January 2006
doi: 10.1242/dev.02250


Development 133, 801-811 (2006)
Published by The Company of Biologists 2006


This Article
Right arrow Summary Freely available
Right arrow Full Text
Right arrow Full Text (PDF)
Right arrow Supplementary Material
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Rentzsch, F.
Right arrow Articles by Hammerschmidt, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Rentzsch, F.
Right arrow Articles by Hammerschmidt, M.

Crossveinless 2 is an essential positive feedback regulator of Bmp signaling during zebrafish gastrulation

Fabian Rentzsch1,*,{dagger}, Jinli Zhang2,*, Carina Kramer1, Walter Sebald2 and Matthias Hammerschmidt1,{ddagger}

1 Max-Planck-Institute of Immunobiology, Stuebeweg 51,79108 Freiburg, Germany.
2 Department of Physiological Chemistry II, Biocenter, University of Wuerzburg, Am Hubland, 97074 Wuerzburg, Germany.


Figure 1
View larger version (65K):

[in a new window]
 
Fig. 1. cvl2 is co-expressed with bmps and positively regulated by Bmp signaling. (A) Domain composition of Cvl2 compared with the CR domain proteins Chordin (Sasai et al., 1994Go), Neuralin (Coffinier et al., 2001Go) and Kielin (Matsui et al., 2000Go). Blue boxes indicate signal peptide; red boxes indicate CR domains; yellow boxes indicate vWFd domain; gray box indicates the Thrombospondin domain; numbers one the right indicate total number of amino acid residues. (B) Temporal expression profile of cvl2 determined by RT-PCR. cvl2 is only zygotically expressed. (C-K) Spatial expression pattern determined by in situ hybridization; developmental stage is indicated in lower left-hand corner; genotype of embryos is indicated on the right-hand side. (C-E) Lateral views, ventral leftwards. (F,H) Frontal views, focal plane of H is indicated in G. (G) Animal pole view on head, ventral leftwards. Filled arrows in E-H indicate anterior lateral mesoderm; open arrows in D,F indicate prechordal plate; arrowhead in E indicates ventral tailbud mesoderm. (I-K) Vegetal views, ventral leftwards, with dorsal borders of the expression domain indicated by arrows. (I,J) cvl2 expression is expanded dorsally in chordin mutants. (I,K) Ventral expression is absent in bmp2b mutants. (L-N) Late cvl2 expression at 34 hpf in otic vesicle (arrowhead in L) and pharyngeal pouches (arrows in L) is elevated in hsp70::bmp2b transgenic (M), but downregulated in hsp70::noggin3 transgenic (N) embryos after a 30 minutes heatshock (28°C to 39°C) at 26 hpf.

 

Figure 2
View larger version (97K):

[in a new window]
 
Fig. 2. cvl2 is required for ventral specification and Bmp signaling. (A) Western blot; translation of myc-tagged cvl2 mRNA is suppressed by cvl2 ATG-MO. (B) Western blots, revealing strongly reduced pSmad1/5 levels in embryos of 80% epiboly stage after injection of cvl2 MO1 (upper panel) or co-injection of cvl2 MO1+2 (lower panel). (C,D) Whole-mount immunostaining, lateral view at 80% epiboly; cvl2 MO1 leads to a reduction of ventral {alpha}-pSmad1/5 immunoreactivity. (E,F) p-Smad1/5 intensity along dorsoventral axis of cvl2 morphant (gray in F) and control sibling (black in F) at 80% epiboly stage. Signal intensities were measured from vegetal views, along a yellow line indicated in E for one of the investigated controls (n=7). (G-X) Whole-mount in situ hybridization with probes indicated in lower left corners. (G-N) Eighty percent epiboly stage. (G,H,K-N) Lateral views; (I,J) vegetal views, dorsal rightwards. Ratios of affected embryos were: eve1, 33/39; cvl2, 66/71; gata2, 31/32; otx2, 45/53. (J) Only the ventral expression of cvl2 is reduced in the morphant, while expression in the prechordal plate is normal. (O,P) Ten-somite stage, view of tailbud, anterior leftwards; expression domain of the blood precursor marker gata1 is reduced in cvl2 morphant (29/37 embryos). (Q-X) Five-somite stage, anterior leftwards; the krox20 expression domains are ventrally expanded in cvl2 morphant (R, 53/68 embryos), but rescued to near wild-type condition in embryo co-injected with cvl2 MO and Drosophila cvl2 mRNA (T, 46/47 embryos). (U-X) Low doses of cvl2 MO1 enhance the mild dorsalization of tolloid/minifin mutants (40/43 in cross of two mfn homozygotes, 22/97 in cross of two mfn heterozygotes), while having no or a much weaker effect in wild-type embryos injected in parallel (0/33). Control embryos are injected with a mismatch cvl2 morpholino. Abbreviations: Coom, Coomassie Blue staining of gel loaded with same amounts of protein; PonRed, Ponceau Red staining of blots after immunoreaction.

 

Figure 3
View larger version (91K):

[in a new window]
 
Fig. 3. Loss of cvl2 only weakly affects chordin/dino mutants. (A-D) Lateral views of 32 hpf embryos; different classes of dino mutants displaying an increasing loss of ventral tail fin tissue (arrows in C,D). Percentages indicate the fraction of embryos displaying the shown tail fin phenotype among uninjected mutants (n=132) and mutants from the same crosses injected with cvl2 MO (n=215). Injection of cvl2 MO2 increases the percentage of dino mutants with class III phenotype, reflecting a stronger dorsalization within the ventral tissue fin tissue. (E,F) krox20 staining, five-somite stage, animal view, anterior leftwards. cvl2 MO1 clearly expands the neuroectoderm in ogontm305 mutants (F, 41/42 embryos), but only has a weak effect in chordin morphants (E, 55/55 embryos), indicating that ventralized traits of chordin can neither be rescued nor enhanced upon loss of Cvl2 function.

 

Figure 4
View larger version (96K):

[in a new window]
 
Fig. 4. Proteolytic cleavage converts Cvl2 from an anti- to a pro-Bmp factor. (A) Western blot with anti-Myc antibody after SDS-PAGE with embryonic protein extracts of cvl2-myc mRNA-injected embryos. Under reducing conditions [+mercaptoethanol (ME)], full-length Cvl2 and a C-terminal cleavage fragment are detectable. Under non-reducing conditions (without ME), the cleavage products remain associated, and only the high molecular weight band is detectable. (B) Illustration of Cvl2 proteins encoded by the different mRNAs. The protein encoded by cvl2-WT is either cleaved, with the two fragments remaining associated via disulfide bounds of thus far unidentified residues (top), or uncleaved, resembling Cvl2-CM (lower panel). (C) Western Blot showing that Cvl2-CM is not cleaved in embryos, and that Cvl2-N is readily synthesized. (D-G) Lateral views of 32 hpf embryos; (H-O) animal views, dorsal rightwards, after in situ hybridization at 80% epiboly for the ventral marker eve1 (H-K) or the dorsal marker chordin (chd; L-O) (Schulte-Merker et al., 1997Go). Injection of cvl2-wt mRNA leads to both weakly ventralized (V1 in D and right embryos in H,L) and weakly (C1 in D) or moderately dorsalized embryos (C2 in D and left embryos in H,L). Injection of cvl2-CM mRNA causes strong dorsalization (F,J,N), while cvl2-N mRNA causes ventralization (G,K,O). Arrows in I,K,M,O indicate borders of the eve1 or chd expression domains. Ratios of obtained phenotypes were: (D-G) see Table 1; (H) dorsalization, 17/44; wild type, 22/44; mild ventralization, 5/44; (J) 52/55; (K) 21/43; (L) dorsalization, 21/56; wild type, 29/56; mild ventralization 6/56; (N) 45/48; (O) 21/36.

 

Figure 5
View larger version (114K):

[in a new window]
 
Fig. 5. The ventralizing effect of Cvl2-N depends on Bmp2b, but is partly independent of Chordin. (A-F) krox20 staining. (A) Wild type, (B) wild type injected with cvl2-N mRNA, (C) bmp2b morphant, (D) bmp2b morphant injected with cvl2-N mRNA, (E) chd mRNA injected embryo and (F) chd plus cvl-N mRNA injected embryo. cvl2-N mRNA can rescue dorsalization caused by chd mRNA but not by bmp2b MO. (G,H) Lateral views of 32 hpf embryos. The ventralization of (G) chd MO injected embryos is enhanced by co-injection of (H) cvl2-N mRNA.

 

Figure 6
View larger version (41K):

[in a new window]
 
Fig. 6. Biochemical characterization of Cvl2 proteins. (A) SDS-PAGE gel showing purified Cvl2-N (lanes 2, 5), Cvl2-CM (lanes 3, 6) and Cvl2-WT (lanes 4, 7) proteins expressed in SF9 cells under reducing (lanes 2-4) and non-reducing (lanes 5-7) conditions; bands in lane 4 correspond to the uncleaved protein (NC), and the N- and C-terminal cleavage products. (B) Biacore sensograms showing the binding of 50 (a), 100 (b), 300 (c) nM Cvl2-N to immobilized Bmp2. (C) Sensograms showing binding of 8 (a), 25 (b) and 75 (c) nM Chordin to immobilized Bmp2. (D) Binding of Chordin to Cvl2-N saturated immobilized Bmp2. At time zero, perfusion with 500 nM Cvl2-N was initiated. The saturation binding of Cvl2-N was set as zero. After 120 seconds, perfusion was continued with 500 nM Cvl2-N plus Chordin. In different cycles, 25 (b) and 75 (d) nM Chordin were applied. In between, sensograms were recorded in the absence of Chordin (a,c). Perfusion with buffer started at 240 seconds. The subtle shift upon perfusion with highly concentrated Chordin (b,d) might result from a slow, equilibrium-driven replacement of Cvl2-N by Chordin. (E-H) Western blots with anti-Myc antibody. (E) Differential distribution of Cvl2 proteins in HEK 293 cell cultures. Cells were transfected with empty vector (lanes 1,2), or plasmids encoding Cvl2-WT (lanes 3-5) or Cvl2-N (lanes 6,7). Cleaved/associated Cvl2-WT is characterized by a high molecular weight band in the absence of ME (lane 3), and a low molecular weight band in the presence of ME (lane 4). It is found only in the supernatant, whereas uncleaved Cvl2-wt (high molecular weight band in presence of ME) is found only in the ECM fraction (lanes 4,5). Cvl2-N is present only in the supernatant (lanes 6,7). (F) Elution profiles of Cvl2 proteins from heparin-coated sepharose beads. Cvl2-CM, Cvl2-CM{Delta}393-396 and Cvl2-N display decreasing affinity to heparin. (G) Removal of a putative heparin binding site (amino acids 393-396) from Cvl2-CM leads to accumulation of the protein in the supernatant of HEK 293 cell cultures. (H) In injected zebrafish embryos, Cvl2-CM and Cvl2-CM{Delta}393-396 are present in similar amounts, although the dorsalizing effect of Cvl2-CM{Delta}393-396 is much weaker (compare with Table 1).

 





© The Company of Biologists Ltd 2006