Xenopus nodal-related signaling is essential for mesendodermal patterning during early embryogenesis.
Previously, we showed that Xenopus nodal-related factors (Xnrs) can act as mesoderm inducers, and that activin induces Xnr transcription, suggesting that Xnrs relay or maintain induction processes initiated by activin-like molecules. We used a dominant negative cleavage mutant Xnr2 (cmXnr2) to carry out loss-of-function experiments to explore the requirement for Xnr signaling in early amphibian embryogenesis, and the relationship between activin and Xnrs. cmXnr2 blocked mesoderm induction caused by Xnr, but not activin, RNA. In contrast, cmXnr2 did suppress mesoderm and endoderm induction by activin protein, while Xnr transcript induction was unaffected by cmXnr2, consistent with an interference with the function of Xnr peptides that were induced by activin protein treatment. The severe hyperdorsalization and gastrulation defects caused by Xnr2 in whole embryos were rescued by cmXnr2, establishing a specific antagonistic relationship between the normal and cleavage mutant proteins. Expression of cmXnr2 resulted in delayed dorsal lip formation and a range of anterior truncations that were associated with delayed and suppressed expression of markers for dorsoanterior endoderm, in which the recently recognized head organizer activity resides. Reciprocally, Xnr2 induced dorsoanterior endodermal markers, such as cerberus, Xhex-1 and Frzb, in animal cap ectoderm. The migratory behavior of head mesendoderm explanted from cmXnr2 RNA-injected embryos was drastically reduced. These results indicate that Xnrs play crucial roles in initiating gastrulation, probably by acting downstream of an activin-like signaling pathway that leads to dorsal mesendodermal specification, including setting up the head organizer.
PubMed ID: 10375512
Article link: Development.
Grant support: GM56238 NIGMS NIH HHS , HD28632 NICHD NIH HHS
Genes referenced: a2m act3 actl6a bmp4 cer1 chrd dkk1 egr2 en2 fn1 frzb hhex inhba lhx1 mixer myod1 nodal nodal1 nodal2 otx2 pdx1 sia1 t tgfb1 wnt8a
Article Images: [+] show captions
|Fig. 1. cmXnr2 specifically blocks mesoderm induction by Xnrs. (A) Schematic representation of putative dominant negative block of Xnr function by cmXnr2 (open box, pro region; shaded box, mature ligand region). The cleavage site alteration is indicated. TGFb-type molecules form dimers (disulfide bond indicated). Protease cleavage releases active dimeric ligand; dimerization of cmXnr2 with Xnrs abrogates this process. (B) cmXnr2 blocks elongation of animal caps induced by Xnr2 and Xnr4 (stage 25). Parentheses: RNA dose (ng). b-globin was a control RNA. (C) cmXnr2 blocks mesoderm induction by Xnr1, Xnr2 and Xnr4, but not activin or BVg1 RNA. RT-PCR analysis of animal cap RNA for the pan-mesodermal marker, Xbra (stage 10.5), and dorsal mesodermal marker, muscle actin (stage 25). Parentheses: RNA dose (ng). Fibroblast growth factor receptor (FGFR) RNA was assessed as a loading control. RT(+) and RT(-) indicate whole embryo RNA transcribed with or without reverse transcriptase (RT).|
|Fig. 2. Rescue of Xnr2-induced gastrulation defects by coinjected cmXnr2 RNA. (A) Embryos injected into two dorsal blastomeres with Xnr2 and control b-globin RNAs develop gross gastrulation defects, absence of recognizable axes and protrusions (arrowheads). (B) Examples of axial rescue by coinjected cmXnr2 RNA.|
|Fig. 3. cmXnr2 blocks mesendoderm induction by activin protein. (A) cmXnr2 blocks animal cap elongation induced by activin protein. Animal caps explanted from embryos injected with or without cmXnr2 RNA were dissected and treated with 3 ng/ml of human activin A protein, and analyzed at stage 22. (B) cmXnr2 blocks mesendoderm induction by activin protein. RT-PCR analysis of stage 10.5 and stage 25 animal cap RNA for mesodermal (Xbra, chordin, Xlim-1, Xnr1 and muscle actin) and endodermal (cerberus, Mixer, edd/endodermin, and XlHbox8) markers. Note that induction of Xnr1 by activin protein is not blocked by cmXnr2.|
|Fig. 4. cmXnr2 induces anterior truncations in Xenopus embryos. (A,B) cmXnr2 delays dorsal lip formation and (C) impairs neurulation. Control (left) or cmXnr2 (right) RNA was injected into two dorsal blastomeres at the 4-cell stage (2 ´ 1 ng). White arrowheads: dorsal lip. Asterisk: normal blastopore. Red arrowheads: elevating neural folds in control embryos. (D) Series of anterior truncations induced by cmXnr2 RNA (500 pg). I, normal uninjected embryo at stage 35; II-VI, embryos injected with cmXnr2 RNA into two dorsal blastomeres. The distribution of resultant embryos between the classes is indicated (%). Characteristic features and estimated anteroposterior deficiencies index (APDI) of the classes are: I, normal embryo (APDI 5); II, reduced forehead and slightly cyclopic (APDI 4); III, cyclopia and reduced cement gland (APDI 3); IV, no eye pigment or no cement gland (APDI 2); V, truncation of posterior and anterior axial structures (APDI 1), and VI, significant A-P axis shortening (APDI 0). (E) In situ hybridization of cmXnr2 RNA-injected embryos with region-specific neural markers (XOtx2, forebrain/midbrain; XEn2, midbrain/hindbrain boundary; Krox20, hindbrain rhombomeres 3 and 5). Yellow arrowheads indicate expression domains. Fig. 5. cmXnr2 does not grossly ventralize mesoderm, but affects gastrulation movements. (A) Dorsal and ventral marginal zones (DMZ, VMZ) were analyzed at stage 25 by RT-PCR for expression of the dorsal mesodermal marker, muscle actin, and the ventral mesodermal marker, globin. Unlike BMP4 RNA, neither cmXnr2 RNA nor pCSKA-cmXnr2 induced globin or reduced actin expression in DMZs. Parentheses: RNA or DNA dose (ng). FGFR was used as a loading control. (B) ‘Exogastrulae’ induced by cmXnr2 RNA injected into two dorsal blastomeres (500 pg/blastomere) at the 4-cell stage. (C) At stage 10.25, involuting head mesendoderm (gray region) was dissected from uninjected, b-globin and cmXnr2 RNA-injected embryos, and cultured on fibronectin-coated dishes. (D) Morphology of explants 1 and 3 hours after explantation. No significant cell migration is seen in the explant from cmXnr2 RNA-injected embryos after 3 hours. A|
|Fig. 5. cmXnr2 does not grossly ventralize mesoderm, but affects gastrulation movements. (A) Dorsal and ventral marginal zones (DMZ, VMZ) were analyzed at stage 25 by RT-PCR for expression of the dorsal mesodermal marker, muscle actin, and the ventral mesodermal marker, globin. Unlike BMP4 RNA, neither cmXnr2 RNA nor pCSKA-cmXnr2 induced globin or reduced actin expression in DMZs. Parentheses: RNA or DNA dose (ng). FGFR was used as a loading control. (B) ‘Exogastrulae’ induced by cmXnr2 RNA injected into two dorsal blastomeres (500 pg/blastomere) at the 4-cell stage. (C) At stage 10.25, involuting head mesendoderm (gray region) was dissected from uninjected, b-globin and cmXnr2 RNA-injected embryos, and cultured on fibronectin-coated dishes. (D) Morphology of explants 1 and 3 hours after explantation. No significant cell migration is seen in the explant from cmXnr2 RNA-injected embryos after 3 hours.|
|Fig. 6. cmXnr2 suppresses anterior mesendodermal specification. (A) Uninjected (U) and embryos injected equatorially at the 4-cell stage (4 ´ 1 ng) with cmXnr2 RNA (cm) were collected at the indicated stages, and analyzed by RT-PCR for Xbra (general mesoderm), Xwnt8, XmyoD (ventrolateral mesoderm), Xlim-1, chordin (dorsal mesoderm), cerberus, Xhex, Frzb and dickkopf-1/dkk-1 (anterior mesendoderm). FGFR: loading control. (B) The signals obtained above were quantitated using ‘NIH Image’ software, and normalized to the intensity of FGFR. The normalized expression profiles of the markers fall into two categories, type I and II (represented by cerberus and Xlim-1, respectively), as indicated to the right of panel A. Note that the peak of expression in cmXnr2- injected embryos is delayed in both type I and II, but that overall induction (area under the line) is more suppressed in type I than type II.|
|Fig. 7. cmXnr2 suppresses expression of anterior endodermal markers in vegetal explants. (A) Experimental design. RNAs were injected vegetally into 4-cell embryos, vegetal explants dissected at stage 9 and tested at stage 10.5 or 25 for endodermal marker expression. (B) RT-PCR analysis for early (cerberus, Xhex1, Frzb, Mixer, Xsox17ß and Siamois), and late (endodermin/edd and XlHbox8) endodermal markers. RNA doses injected per blastomere are indicated. cmXnr2-injected VEs were staged by reference to undissected cmXnr2-injected embryos (see text). Suppression of anterior endodermal markers by cmXnr2 was abolished by coinjected Xnr2 RNA. Note that Siamois expression was not affected by cmXnr2.|
|Fig. 8. Induction of anterior endodermal markers by mouse nodal and Xnrs. Animal caps injected with nodal, Xnr1 or Xnr2 RNA were analyzed at stage 10.5 by RT-PCR for the expression of anterior mesendodermal markers (cerberus, Xhex1 and Frzb) and the general endodermal marker Xsox17ß. Siamois, an early zygotic marker of dorsovegetal endoderm, was not induced.|