|
Figure 2. RA directly activates ndrg1a expression in Xenopus archenteron roof endoderm cells.(A–O) RA signaling is necessary and sufficient to activate ndrg1a expression in the archenteron roof endoderm cells as well as in the pancreas, oesophagus, stomach, and duodenum primordia. Embryos were treated with 0.25 µM BMS453 or 5 µM RA at stage 10 for one hour, collected at stages indicated in the control panels, and subjected to whole mount in situ hybridization analyses of ndrg1a expression. (A–I) Whole mount in situ hybridization on bisected embryos, anterior toward the left. (J–O) Lateral view, head toward the left. Red arrows in panels K and N and red triangles in panels L and O point to the loss or expansion of ndrg1a expression in pancreas, oesophagus, stomach, and duodenum upon BMS453 or RA treatment, respectively. (P–S) RA directly activates ndrg1a expression in Xenopus archenteron roof endoderm cells. Embryos were treated with cycloheximide (CHX, 10 µg/ml), 5µM RA, or both at stage 15 for one hour, fixed at stage 20, and bisected for whole mount in situ hybridization analyses of ndrg1a expression. For the combined treatment, CHX was added 15 min before the application of RA. Anterior is toward the left. The numbers of embryos showing the illustrated phenotypes are given in the corresponding images.
|
|
Figure 3. The efficiency of ndgr1a MO1 and MO2 was verified in vivo.(A–D) Fertilized Xenopus eggs were co-injected with 3.5 picomoles of coMO or ndrg1a MO1 in combination with an mRNA containing MO1 binding site followed by GFP coding sequence (MO1BS-GFP) and evaluated for live GFP translation at stage 18 with an Olympus SZX16 fluorescence microscope. (A, B) GFP. (C, D) Bright field view. (E–H) ndrg1a MO2 specifically inhibited the translation of GFP following its binding site (MO2BS-GFP), as revealed by the in vivo assay as above. (E, F) GFP. (G, H) Bright field view.
|
|
Figure 4. ndrg1a knockdown in Xenopus embryos specifically disturbed pancreas, oesophagus, stomach, and duodenum formation, which phenocopied BMS435 treatment.(A) 3.5 picomoles of coMO, ndrg1a MO1 or ndrg1a MO2 was vegetally injected into all four blastomeres at four cell stage of development and collected at stages 36 and 42 for whole mount staining with marker genes indicated on the left side or in the images 1–6. (A1–12 and A16–42) Lateral view, head toward the left. (A13–15) Dorsal view. The dorsal structures, such as the neural tube, notochord, and somites were removed after whole mount in situ hybridization. (B) Morphology of ndrg1a MO1 injected and BMS453 treated embryos collected when control siblings developed to stage 42. The numbers of embryos showing the illustrated phenotypes are given in the corresponding images.
|
|
Figure 5. TALEN mediated disruption of ndrg1 in Xenopus tropicalis confirmed MO mediated ndrg1a knockdown phenotype in Xenopus laevis.(A) The TALEN targeting site in Xenopus tropicalis ndrg1 locus (GenBank accession no. NM_001008145.1) was designed in the third exon highlighted in gray with the flanking introns in plain text. The TALEN recognition sequences are highlighted in yellow. The underlined sequences are PCR primers (P1 and P2) used for the evaluation of the gene targeting efficiency. (B) Somatic mutations induced by ndrg1 TALENs in Xenopus tropicalis embryos. Deletions (Δ) are indicated by red dashes and insertions (+) by lowercase red letters against a gray background. The numbers in parentheses show the number of deleted or inserted base pairs. The frequency of the mutation in the sequenced samples is shown in the square brackets. Up to 90% (28/31) of ndrg1 loci sequenced were disrupted. (C) Whole mount in situ hybridization analysis of pdx1 and insulin expression in stage 40 Xenopus tropicalis embryos with or without injection of ndrg1 TALEN mRNAs. All images are lateral view with head toward the left. The numbers of embryos showing the illustrated phenotypes are given in the corresponding images. (D) Xenopus tropicalis froglets once subjected to the injection of ndrg1 TALEN mRNAs showed pancreas, stomach, and duodenum hypoplasia (upper right image), while their liver and gall bladder developed normally (lower right image). Among 25 froglets sacrificed and dissected, 6 showed pancreas aplasia as illustrated in D (upper right image) and the rest showed severe pancreas hypoplasia, while all of them showed stomach and duodenum hypoplasia. The pancreas in control froglet is outlined by white dashed lines. Abbreviations: du, duodenum; gb, gall bladder; he, heart; li, liver; st, stomach.
|
|
Figure 6. Nuclear β-catenin localization in archenteron roof endoderm cells appears to be suppressed by RA activated Ndrg1a in Xenopus laevis.Embryos were treated with 5 µM RA, 0.25 µM BMS453 for one hour at stage 10, or vegetally injected with 3.5 picomoles of ndrg1a MO1 at four cell stage and collected at stage 20 for immunofluorescence. (A) Left panel is a schematic drawing illustrating a midsagittal section of stage 20 embryos (after Hausen and Riebesell [71]). The dashed blue lines outline the archenteron roof endoderm where ndrg1a is expressed. Right panels are representative immunofluorescence images showing β-catenin signals (red channel) and DAPI staining (blue channel) in the outlined archenteron roof endoderm cells. (B) Quantification data obtained from three independent experiments. Nine embryos in total (three for every experiment) from each group were sectioned to evaluate the mean percentage of β-catenin positive cells in the outlined archenteron roof endoderm illustrated in the left panel of A. For each embryo, the outlined archenteron roof endoderm cells in the 30 consecutive parasagittal sections central to the median plane were scanned for nuclear β-catenin signals. *, p<0.05. **, p<0.01 (Student’s t-test, two-tailed distribution).
|
|
Figure 7. Ndrg1a represses Wnt/β-catenin signaling, thus specially allowing pancreas, oesophagus, stomach, and duodenum specification.(A, B) Xenopus laevis embryos were vegetally injected with the reagents indicated on the top, treated with 10 µM dexamethasone (Dex) at stage 15, collected at stages 36 (A) and 42 (B), and subjected to whole mount staining with probes indicated on the left side. Doses of the reagents injected are as follows: ndrg1a MO1, 3.5 picomoles; GR-ΔNTcf3 mRNA, 0.25 ng; GR-LEFΔN-βCTA mRNA, 0.5 ng. Red arrows in images A19, 20, 24, 25, B14, 15, 24, 25, 29, 30, 44, and 45 point to either ectopic or loss of expression of marker genes indicated upon inhibition of Wnt/β-catenin signaling. (A11–15) Dorsal view. The dorsal structures, such as the neural tube, notochord, and somites were removed after whole mount in situ hybridization. All the rest images in A and B are lateral view with head toward the left. The numbers of embryos showing the illustrated phenotypes are given in the corresponding images.
|
|
Figure 8. Overexpression of Atf3 phenocopied ndrg1a knockdown.(A) Xenopus laevis embryos were vegetally injected with 0.25 ng of β-catenin mRNA, 0.5 ng of GR-LEFΔN-βCTA mRNA or 3.5 picomoles of ndrg1a MO1 at 4-cell stage, treated with 10 µM Dex at stage 11, and subjected to RT-PCR analysis of atf3 expression at stage 30. Ornithine decarboxylase (odc) was used as the RNA loading control. (B, C) Xenopus laevis embryos were vegetally injected with 0.3 ng of atf3 mRNA at 4-cell stage and collected at stages 36 (B) and 42 (C) for whole mount staining with probes indicated on the left side. (B5, 6) Dorsal view. The dorsal structures, such as the neural tube, notochord, and somites were removed after whole mount in situ hybridization. All the rest images in B and C are lateral view with head toward the left. The numbers of embryos showing the illustrated phenotypes are given in the corresponding images. Abbreviation: CE, control embryos.
|
|
Figure 9. atf3 MO was able to rescue ndrg1a knockdown phenotype.Pdx1 partially rescued ndrg1a knockdown or Atf3 overexpression phenotypes. (A, B) Xenopus laevis embryos were vegetally injected with the reagents indicated on the top at 4-cell stage and collected at stages 36 (A) and 42 (B) for whole mount staining with probes indicated on the left side. The doses of the reagents injected are as follows: atf3 MO, 2 picomoles; pdx1 mRNA, 75 pg; ndrg1a MO1, 3.5 picomoles; atf3 mRNA, 300 pg. (A9–16) Dorsal view. The dorsal structures, such as the neural tube, notochord, and somites were removed after whole mount in situ hybridization. All the rest images in A and B are lateral view with head toward the left. The numbers of embryos showing the illustrated phenotypes are given in the corresponding images. (C) The data obtained suggest an epistasis that RA activated Ndrg1a represses Wnt/β-catenin signaling and consequently releases the inhibitory effect of Wnt/β-catenin, which may be partially mediated by Atf3, on pancreas, oesophagus, stomach, and duodenum formation.
|
|
Figure 1.
Xenopus ndrg1a is expressed early in the archenteron roof endoderm and late in pancreas, oesophagus, stomach, and duodenum primordia, as revealed by whole mount in situ hybridization.(A) Lateral view of a blastula. (B) Whole mount in situ hybridization on bisected stage 18 neurula, anterior toward the left. (C–E) Lateral view, head toward the left. Note that the foregut ndrg1a signal in E covers pharynx, oesophagus, pancreas, stomach, and duodenum. (F, G) Transversal sections of a stage 28 embryo at the levels illustrated by the dashed lines in E. (H, I) Lateral view, head toward the left. (J) Ventral view, head toward the left. Abbreviations: ar, archenteron roof endoderm; de, dorsal endoderm; dp, dorsal pancreatic bud; ey, eye; gb, gall bladder; osd, oesophagus, stomach and duodenum anlagen; p, proctodaeum; pa, pancreas; pd, pronephric duct; ph, pharynx; pr, pronephric anlage; pt, pronephric proximal tubules; vp, ventral pancreatic bud.
|
