Effect of activin and lithium on isolated Xenopus animal blastomeres and response alteration at the midblastula transition

Abstract

Dorsoventral mesoderm patterning in the amphibian embryo involves a series of interactions mediated by several peptide growth factors. Animal blastomeres isolated at the 8-cell stage are useful for studying mesoderm patterning, since they contain the prospective (uninduced) mesoderm region and allow examination of the default state of animal cells. When activin is applied to these dorsal and ventral animal half explants, a competence prepattern for responding to activin is observed. In order to investigate the characteristics of prepatterning, we treated animal blastomeres with the embryo dorsalizing agent LiCl. Treatment with lithium alone did not induce normal trunk mesoderm in either blastomere. Lithium did, however, alter the competence of animal blastomeres to activin. Dorsal mesoderm was formed in the ventral blastomeres, as well as in the dorsal blastomeres. This result reveals that the early dorsoventral polarity in the animal hemisphere is not fixed. Using goosecoid(gsc) and Xwnt-8 genes as dorsal and ventral mesoderm markers, it was verified that lithium modifies the competence to activin. Unexpectedly, lithium treatment on its own resulted in gsc expression in the animal half explants. This suggests that embryo goosecoid expression may be induced by the effect of dorsal determination activity, but not by mesoderm induction. However, lithium induced also the expression of brachyury (Xbra) gene at very low levels. This would indicate the formation of dorsal-anterior mesoderm, which was not identified by the tissue observations. Expression of Xwnt-8, a ventral mesoderm marker usually induced in blastula animal caps by activin, was hardly induced in the blastomere explants. We isolated whole animal half explants at the 8-cell stage and exposed to activin at different stages. It was found that the same concentration of activin induces gsc before the midblastula stage, and induces Xwnt-8 at later stages. This suggests that the response of animal blastomeres alters depending on the stage of activin signaling.