XB-ART-38128BMC Dev Biol 2008 Jul 28;8:74. doi: 10.1186/1471-213X-8-74.
Show Gene links Show Anatomy links
GATA4 and GATA5 are essential for heart and liver development in Xenopus embryos.
BACKGROUND: GATA factors 4/5/6 have been implicated in the development of the heart and endodermal derivatives in vertebrates. Work in zebrafish has indicated that GATA5 is required for normal development earlier than GATA4/6. However, the GATA5 knockout mouse has no apparent embryonic phenotype, thereby questioning the importance of the gene for vertebrate development. RESULTS: In this study we show that in Xenopus embryos GATA5 is essential for early development of heart and liver precursors. In addition, we have found that in Xenopus embryos GATA4 is important for development of heart and liver primordia following their specification, and that in this role it might interact with GATA6. CONCLUSION: Our results suggest that GATA5 acts earlier than GATA4 to regulate development of heart and liver precursors, and indicate that one early direct target of GATA5 is homeobox gene Hex.
PubMed ID: 18662378
PMC ID: PMC2526999
Article link: BMC Dev Biol
Species referenced: Xenopus
Genes referenced: gata4 gata5 gata6 hhex myl2 nkx2-5 nr1h5 odc1 tbx5 tnni3
Morpholinos: gata4 MO1 gata4 MO4 gata5 MO1 gata5 MO2 gata5 MO4
Article Images: [+] show captions
|Figure 1. Injection of G5 MO, but not of C1 and C2 control MOs, cause severe size reduction or loss of heart and liver in Xenopus embryos. A: Translation of injected GATA5-GR.HA mRNA, detected by Western blotting with anti-HA antibody, is blocked by G5 but not by other MOs indicated. G5UTR MO ("G5U") does not affect translation of GATA5-GR.HA, as this constructs lacks the 5'UTR sequence. 1 ng of mRNA was injected into 1- or 2-cell embryos and 105 minutes later the embryos were injected with 10 ng of G5, or 50 ng of G5UTR or G4 MOs. E-Uninjected embryos. B: Injection of 5 ng of G5 MO results in a loss (B1,2) or severe reduction (B3-5) of cardiac and liver precursors, as revealed by whole-mount in situ hybridisation for MLC2 (purple) and Hex (turquoise). Posterior injection of G5 MO (5ng) has no obvious effects (B6). Control MOs 1 or 2 (C1 or C2, 50 ng/embryo) have no effect on normal development of heart and liver precursors (B7-9). B1-4, 8,9: ventral view; B5-7: lateral view (anterior to the left). In B3-5 arrows point to the remnants of the heart. C: G5 MO (1 ng/embryo) causes heart defects in X. tropicalis (Xt) embryos (arrow points to the heart remnant), and C1 MO (10 ng) has no effect on normal heart development (arrowhead in C2). Complete bleaching of X. tropicalis embryos has removed the morphological landmarks, and cement gland (cg) and eyes (e) have been indicated to add visualisation.|
|Figure 2. G5SP MO creates GATA5 protein lacking the C-terminal Zn finger and causes heart and liver defects. A: G5SP MO causes dose-dependent splicing out of exon 4 in both Xenopus laevis and Xenopus tropicalis. The injected dose is indicated (in ng). 3, cDNA that contains exon 4 and regions of exons 3 and 5 determined by target sites of the primers; 3, cDNA without exon 4. Below are shown the sequences of the wt 3 and 3 cDNAs showing in-frame splicing in both species. B: G5SP MO causes a dose-dependent reduction in the level of the wt full-length mRNA (including exon 4; 3) and concomitant increase in the level of the mRNA that lacks exon 4 (3), as revealed by RT-PCR with primers based in exons 2 and 4. Injection of 9 ng of G5SP MO causes partial loss (~50%) of wt GATA5 mRNA. The dose in ng used per embryo is given for each MO. -PCR, control with no cDNA input. M-DNA marker. ODC- Orhithine Decarboxylase loading control. Embryos were collected for RNA analysis at st. 15. C: Injection of 9 ng of G5SP MO into the same group of embryos analysed in (B) causes severe reduction of heart and liver st. 37 (C1,2). Injection of 50 ng of C1 MO has no effect on heart and liver development (C3). D: The dGATA5 protein can neither activate transcription nor can it significantly affect the ability of GATA5 or GATA4 to activate a firefly luciferase reporter driven by 2 GATA sites in animal cap explants. Dual luciferase assays were performed 3 hours after excision of explants, and firefly luciferase activity was normalised to renilla luciferase activity resulting from TK-RL DNA. A representative experiment (out of 3) is shown; whilst the levels of induction varied between experiments, the trend (activation by GATA4 or GATA5 and lack of substantial effect by dGATA5) remained consistent. E: Schematic representation of the effect of G4/5SP MOs (exon-specific part shown as red line) on the domain structure of their targets. TAD-Trans Activation Domain; NLS-Nuclear Localisation Signal; N, C-Zn fingers. Below-Western blot showing efficient translation of the dGATA5 protein in embryos, detected with anti-HA antibody. E-uninjected embryos. d, wt-embryos injected with the d- or wtGATA5-GR.HA mRNA, respectively.|
|Figure 3. Summary of effects of GATA4 and GATA5 MOs in Xenopus embryos. A: Examples of phenotypic classes caused by GATA5 and GATA4 MOs. Ventral views of embryos injected with G5UTR MO or G4 MO are shown (50 ng/embryo). Heart and liver precursors have been revealed by MLC2 (BCIP) and FOR1 (BM purple) probes, respectively. FOR1 was developed first. Ventral views are shown, with anterior at the top. B, C: Summary of frequencies with which the heart and liver phenotypes were observed for GATA5 MOs (B) and for GATA4 MOs (C). The doses of splicing MOs are indicated and for other MOs are as in Figs. 1 and 4.|
|Figure 4. GATA4 MOs cause defects in heart and liver development. A: Translation of injected GATA4-GR.HA mRNA, detected by Western blotting with anti-HA antibody, is blocked by G4 but not by other MOs indicated. 1 ng of mRNA was injected into 1- or 2-cell embryos and 105 minutes later the embryos were injected with 50 ng of indicated MOs. E-Uninjected embryos. B: Injection of 50 ng/embryo of G4 MO leads to a reduction in heart and liver precursors and to cardiac morphogenesis defects such as cardia bifida, highlighted by arrows in B3 (B1-5). The same dose (50 ng/embryo) of the C2 MO has no effect on heart and liver development (B6,7). B1-3,5,7: ventral view. B4,6: lateral view. Heart and liver precursors were revealed by MLC2 and Hex probes (BM purple and BCIP, respectively). C: 10 ng/embryo of C2 MO has no effect on heart in X. tropicalis embryos, but the same dose of G4 MO causes heart defects. D: G4SP MO causes dose-dependent splicing out of exon 4 in both Xenopus laevis and Xenopus tropicalis. The injected dose is indicated (in ng). 3, cDNA that contains exon 4 and regions of exons 3 and 5 determined by target sites of the primers; 3, cDNA without exon 4. Below are shown the sequences of the wt 3 and 3 cDNAs showing in-frame splicing in both species. E (top): Injection of 80 or 100 ng/embryo of G4SP MO, but not of 80 ng of C1 MO, causes splicing out of exon 4 with ~90% efficiency, as detected by RT-PCR analyses of mRNA from st. 15 embryos. +/- RT-indicates presence or absence of Reverse Transcriptase in samples that were analysed by PCR.E (bottom): 80 ng of C1 MO has no effect on heart and liver development (E1), whereas the same amount of the G4SP MO causes cardia bifida and liver defects (E2-4). In E2 remnants of cardiac tissue detected by weak expression of cardiac marker in severely affected embryo with cardia bifida are shown by arrows. Ventral views are shown. Heart was labelled by cTnI (purple) and liver with Hex (light blue/turquoise) probes.|
|Figure 5. GATA5 is required for stable specification of heart and liver precursors and is sufficient to directly induce Hex. A: GATA5 MOs (20/32 embryos for G5 (5 ng) and 14/25 embryos for G5UTR (50 ng) showed reduced expression), but not GATA4 MOs (19/19 embryos for G4 (50 ng) and 22/22 embryos for G4SP (100 ng) expressed Nkx2.5) or GATA6 MO (10 ng; 14/14 embryos with normal expression), cause early deficiency of heart precursors at st. 22 (arrowheads in A1-3). Heart field was revealed by Nkx2.5 expression. Ventral views are shown, with anterior at the top. B: Liver precursors in anterior endoderm (see Additional File 3) expressing Hex are affected by 10 ng of G5 MO (31/39), but not by 50 ng of C1 MO (20/23 normal). Embryos injected uniformly are shown before (B1,3) and after (B) staining for lineage tracer (pink). Ventral views are shown, with anterior to the left. C: G4SP MO does not effect Hex expression. Embryos injected with indicated MOs were analysed at st. 22 for expression of Hex, splicing out of exon 4 of GATA4 and for ODC by RT-PCR. D: GATA5-GR induces expression of Hex mRNA in animal caps explants in the presence or absence of cycloheximide (Chx), which was added at st. 9. All samples were treated with dexamethasone at the same time to activate GATA5-GR protein. E-st. 11 control embryos.|
|Additional file 2. G5 MO reduces Tbx5 expression. Injection of 5 ng of G5 MO leads to a reduction in Tbx5 expression at st. 23/24 (arrowhead), whereas 50 ng of C1 MO has no effect. Ventral views are shown.|
|Additional file 3. St. 22 Hex-expressing anterior ventral endoderm is fated to give rise to liver precursors. A: st. 22 embryo, hybridised with Hex probe, is aligned along a graticule. B: strategy used for fate mapping the st. 224 Hex expressing ventral endoderm domain. In brief, embryos were aligned along the graticule using the pharyngeal arches, cement gland and a ventral protrusion as landmarks. The embryo was bisected caudal to the pharyngeal arches and the ventral protrusion and a small spot of DiI (Sigma) injected onto the exposed ventral endoderm. Bisected embryos were then grown until they reached st. 35. In situ hybridisation was performed to ascertain whether the position of the DiI corresponded to the position of Hexexpression. C, D: visible and fluorescent light views of st.35 half-embryos. E: merged images C, D. F: in s|
|Additional file 4. High doses of GATA5 MOs reduce gastrula-stage Hex expression. G5 and G5SP MO cause downregulation of Hex expression in gastrulae (st. 11) only at high doses (20 ng for G5 (30/37 embryos) and 44 ng for G5SP (27/33 embryos), whereas 50 ng of C1 MO has no effect (15/17 embryos with normal expression). G5 MOs cause gastrulation defects which cause a delay in blastpore closure and the shape of Hex domain of expression. Blastopore is highlighted by dashed line.|
|Figure 6. Injection of GATA4 and GATA5 MOs in endoderm interferes with heart morphogenesis. Top: experimental scheme showing targeted co-injections of MOs and rhodamine-dextran (red fluorescence). The drawing of ventral view of Nieuwkoop and Faber stage 41 embryo  is from http://www.xenbase.org webcite. Cardiac actin/GFP embryos were injected at 32-cell stage with indicated MOs and the fate of injected cells was recorded at st. 41 by rhodamine fluorescence (B, D, F, H, J). A, C, E, G, I: GFP fluorescence was used to observe heart development (cardiac actin/GFP is expressed in all striated muscles, and head muscles are visible in all images), and GFP images were merged with rhodamine images to highlight injected areas. Heart is indicated by arrows, and in specimens where heart was partially injected (C, G), uninjected area is highlighted by an arrowhead. All properly targeted embryos were similarly affected by G4SP and G5SP MOs (n = 8, 10, respectively), but not by C1 MO (n = 7). Ventral views are shown, with anterior at the top.|
|Figure 7. Interaction between GATA4 and GATA6 in heart and liver development. Suboptimal doses of G4 (A, 25 ng) or G6 (B, 5 ng) MOs have only minor effects on heart and liver development at st. 37. 50 ng of G4 MO (C, D) or 10 ng of G6 MO cause heart and liver defects. Similar defects are seen in embryos coinjected with 25 ng of G4 MO and 5 ng of G6 MOs (F). G, H: 50 ng/embryo of C1 MO has no effect on heart and liver. I-frequencies of the phenotypes observed. Heart and liver precursor have been analysed as in Fig. 1 with MLC2 and Hex markers.|
References [+] :
Aries, Essential role of GATA-4 in cell survival and drug-induced cardiotoxicity. 2004, Pubmed