XB-ART-14799Development April 1, 1998; 125 (8): 1447-56.
Xvent-1 and Xvent-2 are members of a novel homeobox subfamily that have been implicated in dorsoventral patterning in Xenopus mesoderm and are thought to function in BMP signalling. Here we investigate the requirement for Xvent function by employing two dominant-negative strategies. Loss of Xvent function dorsalizes ventral mesoderm, induces secondary embryonic axes and directly neuralizes ectoderm. We further find that (1) Xvents act as transcriptional repressors, (2) Xvents function in an additive fashion and (3) a surprising number of genes are able to rescue dominant-negative Xvent phenotypes including Bmp-4, Smad-1 and wild-type Xvents and Xhox3, but not Xwnt-8. The results show that Xvent-1 and Xvent-2 are essential for ventral mesoderm formation and for preventing neural differentiation. A model is suggested to explain how Bmp-4 positional information is converted into distinct cellular responses.
PubMed ID: 9502725
Article link: Development
Genes referenced: bmp4 evx1 hist1h4d smad1 ventx1.2 ventx2.1 ventx2.2 wnt8a
Article Images: [+] show captions
|Fig. 1. Schematical drawing of constructs used in this study. Constructs for Xvent-1 and Xvent-2 were made in similar fashion and are referred to as Xvent-1/2. (Xvent-1/2) Wild-type genes, homeodomain is shown as a black box. (VPXvent-1/2) The Nterminus is replaced by the VP16 transcriptional activator domain, indicated by a light-dashed box. (VPXvent-1(fs)) VP16 activation domain followed by frame-shifted Xvent-1 was used as a control for VPXvent-1/2 injections. (EveXvent-1/2) The N-terminus is replaced by the transcriptional repressor domain of even-skipped indicated by a dark-dashed box. (EveXvent-1(fs)) even-skipped repression domain followed by frame-shifted Xvent-1 was used as a control for EveXvent-1/2 injections. (Xvent-1/2P(40)) Leu-Pro mutation in the position 40 of the homeodomain (indicated by white vertical line).|
|Fig. 2. Loss of Xvent gene function induces secondary embryonic axes. Embryos at the 4- to 8-cell stage were microinjected (A-D) radially or (E-L) ventrally with the RNAs indicated. The amount of RNA in ng/blastomere is given in parenthesis. Microcephaly caused by microinjection of (A) EveXvent-1 (0.1) (47%, n=62) or (B) EveXvent-2 (0.1) (46%, n=67). Embryos microinjected with control construct mRNA EveXvent-1(fs) (0.25) were normal (n=44). Microcephalic phenotypes caused by wild-type mRNA injections of (C) Xvent-1 (0.15) and (D) Xvent-2 (1) are shown for comparison. Secondary axis formation resulting from injections of (E) VPXvent-1 (0.1) and (F) VPXvent-2 (0.5). Rescue of secondary axis formation by mRNA (G) coinjection of VPXvent-1 (0.1) and Xvent-1 (0.15) as well as (H) coinjection of VPXvent-2 (0.5) and Xvent-2 (1). (I) Embryos microinjected with Xvent-1P(40) (2.5) mRNA are normal. (J) Embryo microinjected with Xvent-2P(40) (2.5) showing secondary embryonic axis. (K) Normal embryo resulting from mRNA injection of control construct VPXvent-1(fs) (0.5). (L) Rescue of secondary axis formation by mRNA coinjection of Xvent-2P(40) (2.5) and Xvent-2 (1).|
|Fig. 3. VPXvents do not interfere with mesoderm induction by activin in animal caps. Embryos were microinjected animally with lacZ (0.5 ng/blastomere), VPXvent-1(0.1 ng/blastomere) or VPXvent-2 (0.5 ng/blastomere) with or without activin RNA as indicated. Animal caps were cut at stage 8-9 and cultivated until sibling embryos reached stage 20. In this experiment, ventral control injections with both VPXvent-1 and VPXvent-2 induced more than 50% secondary axes. (A) Coinjection of VPXvent constructs does not inhibit elongation of animal caps, induced by activin (0.1 ng/blastomere; top; 100% elongation activin alone (n=17), 100% elongation VPXvent-1 + activin (n=15), 89% elongation VPXvent-2 + activin (n=18)). VPXvent constructs alone do not induce elongation of animal caps (bottom; lacZ, n=13 ; VPXvent-1, n=16; VPXvent-2, n=16). (B) VPXvent constructs do not inhibit activininduced mesodermal marker expression in animal caps. Total RNA was isolated from animal caps at stage 20 and analysed by RT-PCR for expression of marker genes indicated. Low and high activin correspond to 0.07 and 0.1 ng/blastomere.|
|Fig. 4. Characterisation of secondary embryonic axes. Embryos were microinjected ventrally at the 4- to 8-cell stage with VPXvent-1 (0.1 ng/blastomere), VPXvent-2 (0.5 ng/blastomere) or tBR RNA (0.25 ng/blastomere) as indicated. (A-C) Immune whole-mount staining of embryos with muscle-specific antibody 12-101. Note that all secondary embryonic axes contain muscle. (D-F) Histological analysis of VPXvent-1/2 and tBR-induced secondary embryonic axes. (D) VPXvent-1-induced secondary axes containing notochord. (E) VPXvent-2-induced secondary embryonic axes typically contain no notochord. (F) tBR-induced secondary embryonic axes typically contain no notochord. mu, muscle; no, notochord.|
|Fig. 5. Loss of Xvent-1/2 function dorsalizes ventral mesoderm. Top, schematic drawing of VMZ assay. Embryos were either non-injected (Embryo, DMZ, Co), or injected radially at the 4- to 8-cell stage with mRNAs encoding VPXvent-1 (0.1 ng/blastomere), VPXvent-2 (0.5 ng/blastomere) and tBR (0.25 ng/blastomere). Dorsal (DMZ) or ventral (VMZ) marginal zones were explanted at early gastrula stage and incubated until sibling embryos reached stage 25. Total RNA was isolated and analysed by RT-PCR assays for expression of marker genes indicated. Histone H4 was used for normalisation.|
|Fig. 6. Analysis of Xvent loss-of-function in animal caps. Top, schematical drawing of animal cap assay. Embryos were either noninjected (Embryo, Co), or injected animally at the 4- to 8-cell stage with the RNAs indicated. Animal caps were explanted at early gastrula stage and incubated until the indicated stage. (A) Loss of Xvent-1/2 function neuralises ectoderm. Doses in ng/blastomere were 0.1 VPXvent-1 RNA, 0.5 VPXvent-2 RNA and 0.25 tBR RNA. Total RNA was isolated at stage 31 and analysed by RT-PCR for expression of marker genes indicated. (B) Induction of organizerspecific genes in animal caps. Doses in ng/blastomere were 0.1 VPXvent-1 RNA and 0.5 VPXvent-2 RNA. Total RNA was isolated at early gastrula stage and analysed by RT-PCR for expression of marker genes indicated. Histone H4 was used for normalisation.|
|Fig. 7. Rescue of VPXvent-1 and VPXvent-2 action by different ventralizing factors. 4- to 8-cell-stage embryos were microinjected ventrally with VPXvent-1 (0.1 ng/blastomere, left column) or VPXvent-2 mRNA (0.5 ng/blastomere, right column), either (A,B) alone or (C-L) in combination with the mRNAs or DNA indicated (doses in parenthesis in ng/blastomere): (C) Xvent-2 (1), (D) Xvent-1 (0.15), (E,F) Xhox3 (0.2), (G,H) Bmp-4 (0.3), (I,J) CSKAXwnt-8 (0.05), (K,L) dnXwnt-8 (0.3). Note failure of rescue by CSKAXwnt- 8 (I,J) and secondary embryonic axes containing anterior structures in (K,L).|
|Fig. 8. Additive effects of Xvent genes in gainand loss-of-function. (A) Xvent genes ventralize mesoderm in an additive fashion. Embryos were either non-injected (Co) or microinjected radially at the 4- to 8-cell stage with Xvent-1 (range 0.125-0.25 ng/blastomere) or Xvent-2 mRNA (range 0.37-075 ng/blastomere), or injected with a mixture (same range of both). The average dorsoanterior index (Kao and Elinson, 1988) of embryos was 4.8 (uninjected control, n=22), 4.4 (Xvent-1, n=34), 3.0 (Xvent- 2, n=47) and 2.6 (Xvent-1 + Xvent-2, n=33). (B) Xvent genes function in an additive fashion in downregulation of dorsal and upregulation of ventral marker genes in DMZ. Embryos were either non-injected (Co), or microinjected radially at the 4- and 8-cell stage with the doses indicated in ng/blastomere with Xvent-1 or Xvent-2 mRNAs, or injected with a mixture. Control (Co) VMZ and DMZs were explanted at early gastrula stage and fixed immediately for total RNA preparation. Expression of marker genes indicated was assayed by RT-PCR. (C) VPXvent-1/2 mRNA injections induce dorsal marker genes additively. Embryos were either non-injected (Co), or microinjected radially at the 4- to 8-cell stage with the doses indicated in ng/blastomere with VPXvent-1 or VPXvent-2 mRNAs, or injected with a mixture. Control (Co) DMZ and VMZs were explanted at early gastrula stage and fixed immediately for total RNA preparation. Expression of marker genes indicated was assayed by RT-PCR. Histone H4 was used for normalisation.|
|Fig. 9. Model of Xvent gene action in response to different Bmp-4 concentrations. (Left) High Bmp-4 activity in lateroventral marginal zone drive the expression of both Xvent genes. Sum of Xvent gene activity is high and specifies cells as the most ventral type of mesoderm: blood and mesenchyme, repressing muscle and notochord. (Right) Low Bmp-4 activity in dorsolateral marginal zone is not sufficient for Xvent-1 but for Xvent-2 expression. The sum of Xvent gene activity is low and specifies cells as muscle, but represses notochord.|