XB-ART-12236Curr Biol September 9, 1999; 9 (17): 946-54.
BACKGROUND: Signals from anterior endodermal cells that express the homeobox gene Hex initiate development of the most rostral tissues of the mouse embryo. The dorsal/anterior endoderm of the Xenopus gastrula, which expresses Hex and the putative head-inducing gene cerberus, is proposed to be equivalent to the mouse anterior endoderm. Here, we report the origin and signalling properties of this population of cells in the early Xenopus embryo. RESULTS: Xenopus anterior endoderm was found to derive in part from cells at the centre of the blastocoel floor that express XHex, the Xenopus cognate of Hex. Like their counterparts in the mouse embryo, these Hex-expressing blastomeres moved to the dorsal side of the Xenopus embryo as gastrulation commenced, and populated deep endodermal adjacent to Spemann''s organiser. Experiments involving the induction of secondary axes confirmed that XHex expression was associated with anterior development. Ventral misexpression of XHex induced ectopic cerberus expression and conferred anterior signalling properties to the endoderm. Unlike the effect of misexpressing cerberus, these signals could not neuralise overlying ectoderm. CONCLUSIONS: XHex expression reveals the unexpected origin of an anterior signalling centre in Xenopus, which arises in part from the centre of the blastula and localises to the deep endoderm adjacent to Spemann''s organiser. Signals originating from these endodermal cells impart an anterior identity to the overlying ectoderm, but are insufficient for neural induction. The anterior movement of Hex-expressing cells in both Xenopus and mouse embryos suggests that this process is a conserved feature of vertebrate development.
PubMed ID: 10508583
Article link: Curr Biol
Genes referenced: cer1 frzb2 gsc hhex mlc1 ncam1 odc1 otx2 tbxt wnt8a
Article Images: [+] show captions
|Figure 1. Expression of XHex in blastulae and early gastrulae. (a,b) Animal pole view of XHex expression in late blastulae (stage 9) showing transcripts in a central region of the embryo and, in some embryos, in a superficial population of cells (arrowheads in panel a; see also panel f). (c) Lateral view of a stage 9 embryo highlighting the central location of the XHex-expressing cells. In (a–c), the specimens were cleared to allow visualisation of cells within the embryo. (d) Cryosection of an embryo that had been subjected to in situ hybridisation. Note the localisation of transcripts in cells of the blastocoel floor. (e–g) In situ hybridisation of Xenopus embryos fixed at 60–90 min intervals. (e) XHex expression began throughout the animal hemisphere of midblastulae (stage 8; 7.5 h post fertilisation). (f) Later, XHex transcript levels declined in the ectoderm (stage 9; 8.5 h post fertilisation), and (g) by the early gastrula stage (stage 10; 10 h post fertilisation), ectodermal expression was undetectable. In (e–g), the embryos were bisected manually after the in situ hybridisation procedure. (h) The ectodermal expression of XHex at stage 8 was confirmed by in situ hybridisation to sectioned embryos. In this example, the vegetal expression of XHex was absent, suggesting that ectodermal expression preceded expression in the blastocoel floor. (i,j) In situ hybridisation to sectioned embryos at (i) stage 9 and (j) stage 9.5 showing XHex expression in (i) the blastocoel floor and more vegetally located endoderm and later in (j) presumptive dorsal endoderm. (k,l) Transcripts of (k) XHex and (l) cerberus were detected in two domains in early gastrulae (stage 10). Cells of the blastocoel floor and a second domain of superficial cells at the blastopore lip expressed both genes. In (k,l), the specimens were cleared to allow visualisation of cells within the embryo.|
|Figure 2. Temporal expression of XHex and its expression pattern in early gastrulae compared with that of cerberus. (a) RNase protection analysis revealed that XHex transcript levels peaked during early gastrula stages. A probe against the ornithine decarboxylase (ODC) gene was used as a loading control. A sample of tRNA was used as a negative control. (b,c) Results of in situ hybridisation showing XHex expression in (b) animal pole and (c) dorsal views. Transcripts were located exclusively on the dorsal side of the embryo in a wedge-shaped domain extending from the blastopore lip (arrow in panel c) to the floor of the blastocoel. (d,e) Sagittal cryosections of embryos like those in (b,c) revealed that XHex transcripts were confined to the deep non-involuting endodermal population of cells. Note the absence of XHex expression in the involuting tissue immediately adjacent to the dorsal blastopore lip (arrow in panel d). (f) Animal pole view of stage 10.5 gastrula shows that cerberus expression is detected in a wider domain of the deep endoderm than is XHex (b). Embryos in (b,c,f), specimens cleared to allow visualisation of cells within the embryo.|
|Figure 3. Movement of XHex-expressing cells from a central domain to the dorsal side of the embryo. Cells in a central region of the blastocoel floor marked with Dil at stage 9 moved to the dorsal side of the embryo by stage 10.5, and came to underlie anterior ectoderm adjacent to the ventrally displaced blastocoel by stage 11.5. (a) An embryo fixed shortly after labelling a central population of cells at the floor of the blastocoel. (b) An embryo (stage 10.5) analysed after formation of a crescent-shaped dorsal blastopore lip (arrow). Embryos at this stage were recovering from manipulation and were somewhat distorted. Nevertheless, labelled cells were always positioned on the dorsal side of the embryo above the blastopore lip. (c–e) Embryos fixed at later gastrula stages (stage 11–11.5), after recovery from surgery. Labelled cells were positioned in anterior endodermal tissue, immediately adjacent to the ventrally displaced blastocoel (BLC). (f) Drawing of the specimen in (e) to show the position of labelled cells with respect to the germ layers of the embryo; 55 embryos were analysed in this experiment. (g,h) Two examples of embryos that were ventralised by treatment with ultraviolet (UV) radiation. Central blastomeres were labelled and the embryos analysed at early gastrula stages. Cells did not move from their original central location in UV-treated embryos.|
|Figure 4. Association of XHex expression with development of anterior head structures and persistent expression in UV-ventralised embryos. (a–c) Induction of ectopic XHex expression by ventrovegetal injection of β-catenin RNA. (a) Injected embryos formed complete secondary axes, which included the most anterior structures. (b) Ectopic XHex expression was always detected in such embryos. (c) Higher-magnification view of one such embryo. Ectopic XHex transcripts as well as endogenous expression of the gene were detected in deeply located endodermal cells. Note the absence of detectable expression in cells located superficially to the XHex-expression domain. (d–f) Induction of partial secondary axes by lateroventral expression of a truncated BMP receptor. (d) The induced partial secondary axes lacked anterior structures, and (e) no ectopic XHex expression was detected. (f) Lineage tracing of cells expressing the truncated BMP receptor (red) confirmed that no XHex expression was detected in the injected cells. In this specimen, XHex transcripts are evident immediately adjacent to the dorsal blastopore lip and as faintly staining cells deeper within the embryo. Embryos in (b,c,e) were cleared, but the embryo in (f) was not. (g) RT–PCR showing persistence of XHex and cerberus transcript levels when dorsoaxial development was compromised following UV irradiation. The average dorsoanterior index of treated embryos was 0, and the effective treatment is confirmed by the abolition of goosecoid expression. The last lane shows a control PCR in which the reverse transcription step was omitted (−RT). The RT–PCR to detect expression of elongation factor 1α (EF1α) served as a loading control.|
|Figure 5. The presumptive anterior endoderm confers anterior character to overlying ectoderm. (a) Schematic illustration of the experiment, showing the deep endoderm cells that were dissected and recombined with gastrula-stage ectoderm. At stage 9, a central region of vegetal tissue was isolated and recombined with ectoderm from early gastrulae. All recombinants were allowed to develop to late tadpole stages (stage 35) before analysis. (b) Dorsoanterior endoderm imparts an anterior character to recombined ectoderm. This was marked by cement gland formation (arrows). (c) Cement gland formation did not occur in combinations of ventral endoderm and ectoderm. (d) Central vegetal tissue from late blastulae (the XHex-expressing cells evident in Figure 1a,b) induced cement glands (arrows) when recombined with gastrula ectoderm. (e)XHex-expressing ventral endoderm behaved as dorsoanterior endoderm and induced cement gland formation (arrows). Ventral endodermal tissue was derived from Xenopus embryos injected at the one-cell stage with 0.5 ng XHex RNA. (f,g) Molecular analysis of endoderm–ectoderm recombinants by RT–PCR. Recombinants containing (f) stage 10 dorsoanterior endoderm, stage 9 central region endoderm and (g) XHex-expressing ventral endoderm (derived from embryos injected with 0.5 ng XHex RNA) expressed CG13, a cement-gland-specific marker. Note that gastrula endoderm recombinants do not express NCAM, Otx2 or MLC, demonstrating that they are not neural and that the endodermal tissue was not contaminated with dorsal mesoderm The absence of NCAM expression in lane 2 of (f) was confirmed in longer exposures of the gel (see Supplementary material). Isolated endoderm alone did not express CG13 (data not shown). In one experiment, the dorsal endoderm recombinants did express MLC, and in this case NCAM and Otx2 were also detected, presumably as a response to neural-inducing signals derived from contaminating dorsal mesoderm. Data are representative of results obtained when no dorsal mesoderm markers were detected. Stage 9 endoderm induced expression of MLC, but no neural markers were observed.|
|Figure 6. Molecular markers induced by ectopic ventral expression of XHex. (a)XHex RNA (0.5–3.0 ng total) was injected into both ventral blastomeres at the four-cell stage. Injected embryos were bisected (left panel) or marginal-zone regions isolated at early gastrula stages (right panel) and molecular markers assayed by RT–PCR. Within the ventral half-embryo XHex expression (lane 3, XHex ventral) resulted in the induction of Otx2, goosecoid and cerberus (compare with lane 2, ventral), genes normally expressed exclusively in dorsoanterior regions of early gastrulae. Xbra and Xwnt8 levels were mostly unaffected. When assayed in isolated XHex-injected ventral marginal zones (XHex VMZ, lane 6; free of underlying endoderm), both Otx2 and goosecoid were detected, but Xbra transcript levels were suppressed (compare with VMZ, lane 5). Xwnt8 levels were again unaffected. Expression of cerberus was not induced in the ventral mesoderm in response to XHex (lane 6, XHex VMZ). Dorsal and ventral halves (lanes 1 and 2, respectively), and dorsal marginal zone (DMZ) and VMZ (lanes 5 and 6, respectively) served as controls for the accuracy of dissection. Whole embryo and −RT samples provided positive and negative controls for RT–PCR. (b–e) Localised ventral expression of XHex (red cells in panels b and c; 0.5 ng XHex RNA injected) resulted in the ectopic activation of cerberus. (d) Lateral view of a representative embryo (cleared) shows two seemingly identical domains of cerberus following injection of XHex. (e) A section through a similar embryo revealed that ectopic XHex expression induced cerberus only in the deep endodermal cell layer. Some XHex-expressing cells (red) are evident on the left side of the sectioned embryo. (f) Ectopic ventral expression of cerberus did not induce XHex. (g) Consistent with the RT–PCR analysis of VMZ isolates, localised XHex expression (red) resulted in a loss of detectable Xbra expression when assayed by whole-mount in situ hybridisation at early gastrula stages.|
|Longer exposure of Figure 5f to show absence of NCAM in Lane 2.|