March 1, 1995;
Regulation of Spemann organizer formation by the intracellular kinase Xgsk-3.
Dorsal axis formation in the Xenopus embryo
can be induced by the ectopic expression of several Wnt family members. In Drosophila, the protein encoded by the Wnt family gene, wingless, signals through a pathway that antagonizes the effects of the serine/threonine kinase zeste-white 3/shaggy. We describe the isolation and characterization of a Xenopus homolog of zeste-white 3/shaggy, Xgsk-3
. A kinase-dead mutant of Xgsk-3
, Xgsk-3K-->R, has a dominant negative effect and mimics the ability of Wnt to induce a secondary axis by induction of an ectopic Spemann organizer
. Xgsk-3K-->R, like Wnt, induces dorsal axis formation when expressed in the deep
vegetal cells, which do not contribute to the axis. These results indicate that the dorsal fate is actively repressed by Xgsk-3
, which must be inactivated for dorsal axis formation to occur. Furthermore, our work suggests that the effects of Xgsk-3K-->R are mediated by an additional intercellular signal.
[+] show captions
Fig. 1. Amino acid sequence of Xgsk-3. The predicted amino acid sequences of Xgsk-3, GSK-
3b, and zw3/shaggy are compared, with non-identical residues in GSK-3b and zw3/shaggy
indicated. The putative kinase region encompasses residues 54 to 325. The lysine residue
which was altered in the Xgsk-3K®R mutant is indicated by an asterisk at position 85.
Fig. 2. Temporal and spatial expression of Xgsk-3. (A,B) RNA levels were determined with the RNase protection assay using a mixture of
Xgsk-3 and EF-1a probes. EF-1a is a ubiquitously expressed gene in the Xenopus embryo; EF-1a levels increase from the mid-blastula
transition at stage 8 (Krieg et al., 1989). (A) Analysis of 20 mg of total RNA from unfertilized eggs and embryos at the following stages: stage 7
(blastula); stage 9 (late blastula); stage 11 (mid-gastrula); stage 15 (mid-neurula); stage 19 (late neurula). (B) Analysis of RNA isolated from
dissected embryos (10 embryos per sample). 32-cell-stage embryos were dissected into dorsal and ventral halves (lanes 1 and 2). Stage-9
embryos were dissected into dorsal and vental halves (lanes 3 and 4), or the animal hemisphere was dissected away from the rest of the embryo
(lanes 5 and 6). Note that the total amount of RNA in the samples from 32-cell-stage and stage-9 embryos cannot be compared because the EF-
1a probes are not at the same specific activity. (C) Animal caps were explanted at the late blastula stage from uninjected embryos or from
embryos injected in the animal pole of each cell at the 2-cell stage with 1 ng of Xgsk-3K®R RNA. Animal caps were cultured alone or with
XbFGF for approximately 20 hours. Total RNA was extracted from the caps and analyzed by RNase protection using the muscle actin probe.
The muscle actin probe also protects a portion of the cytoplasmic actin (cyto. actin) gene which is ubiquitously expressed at this stage and thus
serves as an internal control.
Fig. 3. Xgsk-3K®R causes dorsal axis duplication. Both cells of two-cell embryos were injected laterally with 2 ng per blastomere of DXgsk-3
RNA (A) or 0.5 ng per blastomere of Xgsk-3K®R RNA (B,C) and allowed to develop for 3 days. (B) A dorsal view and (C) a dorsoanterior
view of embryos with different degrees of axis duplication.
Fig. 4. Xgsk-3K®R rescues dorsal axis formation in UV-irradiated embryos. Fertilized eggs were UV-irradiated for 60 seconds within 40
minutes after fertilization. At the 4-cell stage, one cell was injected with the indicated RNA and the embryos were allowed to develop for three
days. (A) Uninjected embryo; (B) Embryo injected with 2 ng DXgsk-3 RNA; (C) Embryo injected with 1 ng Xgsk-3K®R RNA; (D) The
dorsoanterior index (DAI) of the uninjected (n=54) (upper panel) and Xgsk-3K®R RNA injected (n=22) (lower panel) embryos was scored.
The percentage of embryos with each score is shown.
Fig. 5. Regulation of dorsal genes by Xgsk-3K®R and Xgsk-3. 4-cell embryos were injected with RNA into either the two dorsal or two ventral blastomeres, allowed to develop to stage 10 and stained by in situ hybridization for goosecoid (A-C), or to stage 12 and stained for Xnot (D-F). (A,D) Uninjected embryos; (B,E) Embryos injected ventrally with 1 ng per blastomere Xgsk-3K®R RNA. In B, the arrows indicate the two dorsal lips. (C,F) Embryos injected dorsally with 2 ng per blastomere Xgsk-3 RNA. Note the faint, dispersed Xnot staining between the arrows in F.
Fig. 6. Xgsk-3K®R rescues dorsal axis formation from deep vegetal cells. Fertilized eggs were UV irradiated for 60 seconds within 40 minutes
after fertilization. At the 32-cell stage, the embryos were injected in two adjacent cells of the same tier with 200 pg of b-galactosidase (b-gal)
RNA per cell in tier C (n=24) (A) or 200 pg b-gal and 1 ng of Xgsk-3K®R RNA per cell in tier C (n=39) (B) or tier D (n=38) (C). After 3 days,
the DAI of the embryos was scored. The percentage of embryos with each score is shown in D. The scores of embryos injected with b-gal RNA
in tier C (n=24) and tier D (n=25) are combined.
Fig. 7. Model for the function of Xgsk-3 in dorsal mesoderm
patterning. (A) A low level of Vg1, acting in combination with FGF
(not shown in the figure) induces ventral mesoderm (R. Cornell, T.
Musci, and D. K., unpublished data; and see Kimelman et al., 1992).
Dorsal mesoderm may arise by one of two pathways: (B) Wnt-like
signals are suggested to inactivate Xgsk-3, leading to the activation
of dorsal-specific genes, possibly through signals from an
unidentified morphogen (M). Low level Vg1 signaling is still
required for mesoderm induction since neither addition of Xwnt-8
(Smith and Harland, 1991; Sokol et al., 1991) or inhibition of Xgsk-3
is able to induce mesoderm directly. (C) High levels of Vg1 can also
induce dorsal mesoderm (Thomsen and Melton, 1993) either by
overriding the effects of Xgsk-3, or by inhibiting Xgsk-3. The
intracellular signaling pathway used by noggin is still unclear,
although it may be similar to the Wnt pathway since noggin, like
Xwnt-8, is unable to directly induce mesoderm (Smith and Harland,