Spagnoli FM and Brivanlou AH (2008), The Gata5 target, TGIF2, defines the pancreatic...

XB-ART-36968

Development 2008 Feb 01;1353:451-61. doi: 10.1242/dev.008458.

Show Gene links Show Anatomy links

The Gata5 target, TGIF2, defines the pancreatic region by modulating BMP signals within the endoderm.

Spagnoli FM , Brivanlou AH .

???displayArticle.abstract???
Mechanisms underlying regional specification of distinct organ precursors within the endoderm, including the liver and pancreas, are still poorly understood. This is particularly true for stages between endoderm formation and the initiation of organogenesis. In this report, we have investigated these intermediate steps downstream of the early endodermal factor Gata5, which progressively lead to the induction of pancreatic fate. We have identified TGIF2 as a novel Gata5 target and demonstrate its function in the establishment of the pancreatic region within dorsal endoderm in Xenopus. TGIF2 acts primarily by restricting BMP signaling in the endoderm to allow pancreatic formation. Consistently, we found that blocking BMP signaling by independent means also perturbs the establishment of pancreatic identity in the endoderm. Previous findings demonstrated a crucial role for BMP signaling in determining dorsal/ventral fates in ectoderm and mesoderm. Our results now extend this trend to the endoderm and identify TGIF2 as the molecular link between dorsoventral patterning of the endoderm and pancreatic specification.

???displayArticle.pubmedLink??? 18094028
???displayArticle.link??? Development

Species referenced: Xenopus
Genes referenced: a2m amy2a babam2 bmp4 c1h19orf25 chrd fabp2 foxa2 foxf1 fzd7 gal.2 gata4 gata5 gata6 hdgf hhex ins msx1 msx2 myc myod1 ncam1 nkx2-5 odc1 pdx1 rcc2 smad1 smad2 sox2 tbx2 tbx5 tbxt tgif2 tuba4b vegt ventx2.2 washc2a wnt8a wnt8b zfpm1
???displayArticle.morpholinos??? chrd.1 MO1 tgif2 MO1 tgif2 MO2

???attribute.lit??? ???displayArticles.show???

	Fig. 2. Whole-mount in situ hybridization of selected array clones. (A) Xenopus embryos at early gastrula (st. 10), tailbud (st. 30) stage and dissected whole gut (st. 43) were hybridized with antisense probe for Fz7/clone 10B11. Arrowheads indicate expression in the endoderm; the dotted line depicts the dorsal blastopore lip. Staining in endodermal derivatives, such as the hepatic and pancreatic regions, is indicated by a bracket in the middle panel. (B) Embryos at early gastrula (st. 10), tadpole (st. 35) stage and dissected whole gut (st. 43) were hybridized with HDGF/clone 7C5. Arrowheads point to expression in the endoderm; and the bracket in the middle panel indicates endodermal derivatives, such as the hepatic and pancreatic regions. (C) Embryos at gastrula (st. 10.5), tadpole (st. 37) stage and dissected whole gut (st. 42) were hybridized with clone 7G9 [EST similar to hypothetical protein (BG410109)] indicating expression in the ectoderm and hepatic rudiment (bracket in the middle panel). (D) Embryos at gastrula (st. 10), tadpole (st. 36) stage and dissected whole gut (st. 42) were hybridized with clone 8C11 [EST similar to hypothetical protein KIAA0592 (BG410148)] showing expression in the dorsal ectoderm, brain, pronephros, liver and duodenum. (E) Embryos at gastrula (st. 10), early tailbud (st. 27) stage and dissected whole gut (st. 42) were hybridized with clone 7B7 [EST similar to hypothetical protein (BC094159)] indicating expression in the ectoderm, brain, branchial arches and pancreas. (F-J) In situ hybridizations on serial sections of gastrula and neurula stage embryos using Xenopus TGIF2 (xTGIF2) and Gata5 probes show expression of both genes in a subset of endodermal cells. xTGIF2 is also strongly expressed in the ectoderm and mesoderm. (F,F′) In situ hybridization on section of stage 10 embryo using antisense probe for TGIF2/clone 8B1 (* indicates the dorsal side). Red dashed box outlines the TGIF2-positive endodermal region, which is magnified in F′. Arrowheads point to endodermal cells. (G) Gata5 in situ hybridization on section of stage 10 embryo. Arrowheads point to endodermal cells. (H,H′) In situ hybridization on section of stage 10.5 embryo shows enriched staining in dorsal endodermal cells (* indicates the dorsal side). Red dashed box outlines the TGIF2-positive endodermal region, which is magnified in H'. Arrowheads point to endodermal cells. (I,I′) In situ hybridization on neurula stage whole (I) and sectioned (I′) embryos shows TGIF2 staining in anterior endodermal cells (red dashed box in I′). (J) Gata5 in situ hybridization on section of stage 17 embryo. (K) Dissected whole gut (st. 43) stained with TGIF2 probe shows expression in the pancreatic rudiment. a, anterior; d, duodenum; lv, liver; pa, pancreas; p, posterior.
	Fig. 3. xTGIF2 behaves as a modifier of the endoderm. (A) Xenopus eight-cell stage embryos were injected into both ventral (VV) or dorsal vegetal (VD) blastomeres with xTGIF2 (1 ng) mRNA. Vegetal explants were dissected at early gastrula stage (stage 10). Uninjected VV or VD pole halves were used as control for their regional differences in the expression of endodermal markers. All explants were collected at tailbud stage and assayed for expression of the indicated markers by RT-PCR analysis. (B) Animal caps injected with VegT (60 pg) mRNA and/or xTGIF2 mRNA, as indicated. The explants were cultured until stage 30 (tailbud) and analyzed for expression of the indicated markers. (C) Eight-cell stage embryos were injected into ventral vegetal (VV) blastomeres with xTGIF2 (1 ng) mRNA. Uninjected VV or VD pole halves were used as control. All explants assayed at tailbud stage for expression of the indicated mesodermal markers by RT-PCR analysis. (D-E′) Lineage-tracing analysis of endodermal cells injected with xTGIF2 mRNA. (D) Ventral vegetal explants isolated from uninjected or xTGIF2+lacZ-co-injected embryos were stained for β-gal and assayed for Pdx1 expression by in situ hybridization at tailbud stage. The β-gal staining (blue) and ectopic Pdx1 (red) expression co-localize in endodermal explants (see arrowheads). (E) Whole-mount in situ hybridization analysis of Pdx1 expression in xTGIF2+lacZ-co-injected embryos. White dashed outline demarcates the region of the embryo magnified in E′. Arrowheads indicate purple cells stained for β-gal (blue) and positive for Pdx1 (red). (E′) Transverse section through the stained endodermal region shown in E′.
	Fig. 4. xTGIF2 is required in vivo for establishing the pancreatic region within the endoderm. (A) Both dorsal vegetal (VD) blastomeres of eight-cell stage embryos were injected with a combination (5 ng each) of two antisense morpholino oligonucleotides (TGIF2-Mo) targeting both Xenopus laevis TGIF2 pseudoalleles (see Fig. S3 in the supplementary material). TGIF2-Mo-injected VD, uninjected VD and ventral vegetal (VV) explants were dissected at stage 10 and assayed for expression of the indicated markers by RT-PCR analysis. (B) Whole-mount in situ hybridization using Pdx1 probe. Embryos injected with TGIF2-Mo showed reduction of Pdx1 expression domain (80%; n=40). Pdx1 staining was rescued in embryos injected with TGIF2-Mo and mouse Tgif2 (mTgif2; 500 pg) mRNA (70%; n=25). Embryos left untreated show normal Pdx1 expression domain in the pancreatic/duodenum endoderm, as indicated by the white bracket. (C) Whole-mount in situ hybridization with Hex. Embryos injected with TGIF2-Mo or left untreated were stained at stage 34. Hex expression domain in the hepatic endoderm is indicated by the yellow bracket. (D) Analysis of the expression of the pancreatic differentiation markers, insulin and amylase, by whole-mount in situ hybridization. Arrowheads indicate normal domains of expression of insulin and amylase in the pancreatic bud of tadpole gut tubes.
	Fig. 1. Schematic of experimental strategy. (A) Xenopus embryos were injected into both animal blastomeres at the two-cell stage with Gata5 (500 pg) mRNA. Ectodermal explants were isolated at stage 9 from both Gata5-injected and uninjected embryos and cultured until stage 28 (tailbud). Transcriptional differences were analyzed on 5000-clone gastrula-stage cDNA microarray (Munoz-Sanjuan et al., 2002). An aliquot of each RNA sample was assayed for expression of the indicated Gata5 targets by RT-PCR. ODC was used as loading control. (B) Pie chart of the classification of the genes upregulated by Gata5 based on the GO molecular function categories. The majority of the upregulated clones in the Gata5 array fall into four main categories: (1) catalytic activity (26%); (2) binding activity (26%), including the large group of nucleic-acid-binding as well as protein-binding factors; (3) hypothetical/unknown function (13%), including full-length sequences conserved in the mouse and human databases the function of which is unknown; (4) no database match (13%), including a number of clones with no hits in database searches, which might be either genes unique to the frog or partial cDNAs. Edd, endodermin; -RT, minus reverse transcriptase; ODC, ornithine decarboxylase.
	Fig. 5. Inhibitory effects of xTGIF2 on TGFβ and BMP signalings in Xenopus. (A) Partial secondary axis (indicated by *) was observed in tadpole stage embryos injected with xTGIF2 (1 ng) mRNA into one ventral vegetal cell. (B) xTGIF2 (1 ng) mRNA and activin (100 pg) mRNA were injected separetely or together into the animal pole of two-cell stage embryos. Animal caps were analyzed at gastrula stage (stage 11) for the expression of indicated markers by RT-PCR. (C) Animal caps injected with xTGIF2 (1 ng) mRNA were analyzed at gastrula stage (stage 11) for the expression of indicated markers by RT-PCR. (D) Luciferase assay with BMP inducible VENT2-luciferase (VENT2-Luc.) reporter construct. Two-cell stage embryos were injected with VENT2-Luc. alone or in combinations with BMP4 (200 pg) and/or xTGIF2 (1 ng) mRNAs, as indicated. Embryos were harvested at the onset of gastrulation and assayed for luciferase activity. (E) Immunoprecipitation (IP) of flag-xTGIF2 and endogenous Smad1 or Smad2. Four-cell stage embryos were injected into the the vegetal pole (ventrally for Smad1 IP; dorsally for Smad2 IP) with flag-xTGIF2 (1 ng), lysates were prepared at gastrula stage (stage 11) and immunoprecipitated with anti-flag antibody and analyzed by immunoblot (IB) with anti-Smad1 and anti-Smad2 antibodies. The expression of flag-xTGIF2, Smad1 and Smad2 was checked by immunoblotting on the crude extracts used for the IP reaction. As loading control, the membranes were stripped and reprobed with anti-α-tubulin. Chd, chordin; EK, epidermal keratin; xbra, brachyury.
	tgif2 (TGFB-induced factor homeobox 2 ) gene expression in bisected Xenopus laevis embryo, mid-sagittal section, assayed via in situ hybridization, NF stage 10, dorsal right, animal hemisphere up.
	tgif2 (TGFB-induced factor homeobox 2) gene expression in isolated gut tube of Xenopus laevis embryo, assayed via in situ hybridization, NF stage 43.
	hdgf (hepatoma-derived growth factor) gene expression in bisected Xenopus laevis embryo, mid-sagittal section, assayed via in situ hybridization, NF stage 10.5, dorsal right, animal hemisphere up.

	hdgf (hepatoma-derived growth factor) gene expression in isolated gut tube of a Xenopus laevis embryo, assayed via in situ hybridization, NF stage 43.
	c1h19orf25 (chromosome 1 C19orf25 homolog) gene expression in Xenopus laevis embryo,, assayed via in situ hybridization, NF stage 10, blastoporal right, animal hemisphere up.
	c1h19orf25 (chromosome 1 C19orf25 homolog) gene expression in Xenopus laevis embryo, assayed via in situ hybridization, NF stage 37 & 38, lateral view, anterior left, dorsal up.
	c1h19orf25 (chromosome 1 C19orf25 homolog) gene expression in isolated gut tube of a Xenopus laevis embryo, assayed via in situ hybridization, NF stage 43.
	fam21a (family with sequence similarity 21 member A )g ene expression in Xenopus laevis embryo, assayed via in situ hybridization, NF stage10, lateral view, blastopore down.

	fam12a (family with sequence similarity 21 member A ) gene expression in isolated gut tube of a Xenopus laevis embryo, assayed via in situ hybridization, NF stage 42.
	rcc2 (regulator of chromosome condensation 2) gene expression in Xenopus laevis embryo, assayed via in situ hybridization, NF stage10, lateral view, blastopore down.
	rcc2 (regulator of chromosome condensation 2) gene expression in Xenopus laevis embryo, assayed via in situ hybridization, NF stage 27, lateral view, anterior left, dorsal up.
	rcc2 (regulator of chromosome condensation 2) gene expression in isolated gut tube of a Xenopus laevis embryo, assayed via in situ hybridization, NF stage 42.
	ins (insulin) gene expression in Xenopus laevis embryos, assayed via in situ hybridization, NF stage 41 lateral view, anterior left, dorsal up; and NF stage 42 ventral view, anterior up.
	Figure 6. Conserved role for the mouse Tgif2. (A) BRE-luciferase activity in mammalian cells. Mouse C2C12 cells were transfected with Renilla luciferase and BRE-Luc. reporter constructs alone or in combination with mTgif2, as indicated. Transfected cells were stimulated 48 hours after transfection with BMP4 recombinant protein, as indicated. Renilla luciferase was used for normalization. (B) Eight-cell stage embryos were injected into both ventral (VV) or dorsal vegetal (VD) blastomeres with xTGIF2 or mTgif2 mRNAs. Explants were collected at tailbud stage and assayed for expression of the indicated markers by RT-PCR analysis. (C) BTC6 cells were transfected with increasing amount of shRNA targeted against mTgif2 or with control shRNA, and analyzed 48 hours later by real-time RT-PCR for the expression of endogenous mTgif2 level and indicated markers. The plot shows the regulation (expression ratio) of target genes in shTGIF2-transfected cells versus shControl-transfected cells. All the values were normalized to the reference gene SDHA (succinate dehydrogenase), and calculated using the software REST (Pfaffl et al., 2002). Data were determined in triplicate.
	Figure 7. BMP inhibits pancreatic fate within the endoderm. (A) Real-time RT-PCR analysis of dorsal vegetal (VD) explants injected with TGIF2-Mo, TGIF2-Mo in combination with mTgif2 (500 pg) mRNA for the rescue, and a combination of two antisense morpholino oligonucleotides targeting both Xenopus laevis chordin (Chd-Mo) pseudoalleles (Oelgeschlager et al., 2003). The plot shows the regulation (expression ratio) of target genes in VD-injected cells versus VD-uninjected cells. The analysis was done as described in Fig. 6. (B) Eight-cell stage embryos were injected into both ventral (VV) blastomeres with xTGIF2 (1 ng) mRNA and DN-Alk3 (1 ng). All explants were collected at tailbud stage and assayed for expression of the indicated markers by RT-PCR analysis. (C) Eight-cell stage embryos were injected into both dorsal vegetal (VD) blastomeres with increasing amount of CA-Alk3 (0.5 and 1 ng). All explants were collected at tailbud stage and assayed for expression of the indicated markers by RT-PCR analysis. (D) Schematic diagrams of gastrula, neurula and tadpole stages embryos. At gastrula stage, signals from the organizer and dorsal endoderm counteract the ventralizing factor BMP, establishing the region where pancreatic fate is specified. At neurula stage, ongoing intracellular inhibition of BMP signals by TGIF2 defines the region where prospective pancreatic buds are formed (dorsal bud is boxed in black; ventral bud boxed in red). At tailbud stage, Pdx1 marks both pancreatic buds. V, ventral; D, dorsal; A, anterior; P, posterior.
	Fig. S1. Confirmation of selected array clones by RT-PCR analysis. Twenty-five percent of the array clones that resulted upregulated by Gata5 was validated by RT-PCR analysis, reasoning that this was a representative sample size. RT-PCR was performed on cDNA synthesized from ectodermal explants expressing Gata5. ODC was used as loading control and Pdx1 as positive control for Gata5 induction. The primers used for RT-PCR analysis are available at the http://xenopus.rockefeller.edu/.
	Fig. S2. Establishment of the temporal hierarchy in the activation of Gata5 targets. (A-C) A hormone-inducible version of Gata5 (Gata5-GR) was injected into the animal pole of two-cell stage embryos and the animal caps were dissected at blastula stage. Half of the Gata5-injected and uninjected explants were exposed to dexamethasone for 2 hours at three developmental stages (stage 11/gastrula; stage 18/neurula; stage 30/tailbud) before being processed for RT-PCR analysis. ODC was used as loading control, Sox17α and Pdx1 as positive controls for Gata5 induction at gastrula and tailbud stages, respectively.
	Fig. S3. Morpholinos targeting two distinct xTGIF2 pseudoalleles. (A) Alignment of Xenopus tropicalis TGIF2 and the two Xenopus laevis TGIF2 sequences targeted by the antisense morpholino oligonucleotides used in this study (yellow). The ATG is indicated in bold. Morpholino antisense oligonucleotides were directed against the 5′UTR of the two Xenopus laevis TGIF2 pseudoalleles: MO-1 (5′-CAGGGGGAGATTCAAGGAAGATGAA-3′) and MO-2 (5′-CCTAGTGATAGACCACGAGATGGAC-3′). Standard control morpholino corresponds to 5′-CCTCTTACCTCAGTTACAATTTATA-3′. In green is the Xenopus tropicalis TGIF2 morpholino sequence that was used in a recent systematic MO-based functional screen carried out in Xenopus tropicalis. In the X. tropicalis morphants, general phenotypic defects, including shortening of the A-P axis and defects in anterior and dorsal tissues, were described (Rana et al., 2006). Similar defects were also visible in TGIF2-Mo-injected embryos (see Fig. 4B). (B) Xenopus TGIF2 translation from the mRNA encoding xTGIF2 allele 1 is specifically inhibited by TGIF2-MO1, but not by control MO. 10 µM morpholino was added to the in vitro translation reactions. Rana, A. A., Collart, C., Gilchrist, M. J. and Smith, J. C. (2006). Defining synphenotype groups in Xenopus tropicalis by use of antisense morpholino oligonucleotides. PLoS Genet. 2, e193.
	ig. S4. Immunoprecipitation of flag-xTGIF2 with both Smad1 and Smad2. Flag-tagged TGIF2 or Myc-tagged FAST1 were injected into vegetal blastomeres (dorsally for Smad2 IP; ventrally for Smad1 IP) of four-cell stage embryos. Embryonic lysates were immunoprecipitated with anti-Smad1 or anti-Smad2 and analyzed by immunoblot with anti-flag and anti-myc antibodies. Myc-Fast1 was used as control for the specificity of the immunoprecipitation reactions, for its ability to interact exclusively with Smad2 (Chen et al., 1997). Flag-xTGIF2 immunoprecipitated complexes are indicated with *, and the myc-Fast1 complex with . A portion of the lysates was blotted with anti-flag and anti-myc directly (without immunoprecipitation) to assess the expression of flag-xTGIF2 and myc-Fast1, respectively. In addition, equal expression of Smad1 and Smad2 was tested by immunoblotting on the crude extracts using anti-Smad1 and anti-Smad1 antibodies. TrueBlot IP beads and HRP-conjugated secondary antibodies were used to minimize interference by the heavy and light chains of the immunoprecipitating antibody in the IP/IB experiments. Chen, X., Weisberg, E., Fridmacher, V., Watanabe, M., Naco, G. and Whitman, M. (1997). Smad4 and FAST-1 in the assembly of activin-responsive factor. Nature389, 85-89.