Fig. 1. The Brat Protein. (A) The predicted amino acid sequence of Brat encodes a
protein of 456 amino acids. The DNA-binding domain, T domain, is located between
the arrows and is highlighted in bold. (B) A comparison of the T domain of Brat with
six other T domain containing proteins. Identical amino acids are indicated by dots,
and spaces (dash) were introduced to maximize homology. The degree of identity
between these domains and the T domain of Brat is: Xenopus Brachyury (Xbra,
47%), mouse Brachyury (T, 45%), zebrafish T (Zf-T, 46%), Drosophila T-related
gene (Trg, 45%), mouse Tbx2 (54%), and Drosophila optomotor blind (Omb, 51%).
The accession number for Brat is U89707.
Fig. 2. Expression pattern of the Brat gene in Xenopus development.
(A) A Developmental northern blot on total embryonic RNA
revealed a single Brat transcript of approximately 3 kb that is
expressed maternally and zygotically into mid-neurulation (Stage
15). Stages analyzed were egg, blastula (stages 7,9), gastrula (stage
11), neurula (stages 15,18) tailbud tadpole (stage 26) and swimming
tadpole (stage 38). Five embryo equivalents were loaded per lane and
RNA recovery and integrity was equal across the blot (not shown).
(B) Localization of Brat mRNA in oocytes by whole-mount in situ
hybridization. Brat mRNA is expressed throughout oogenesis and is
localized to the vegetal pole by stage II. Pigmented oocytes (P) are
shown in the upper panel, animal pole up, to orient the stain relative
to the animal and vegetal pole. Albino oocytes (A) are shown in the
lower panel to clearly display the domain occupied by Brat
transcripts. During the in situ procedure, the cortex was partially
removed from several stage VI oocytes to demonstrate that the bulk
of Brat transcripts are sequestered in the cortex. Arrows indicate the
edge where the cortex was torn off, revealing the underlying,
unstained cytoplasm. (C) Vegetal localization of Brat mRNA in early
cleavage stage embryos. A northern blot of RNA from animal and
vegetal halves of stage 6 embryos. Seven explants or three embryos
were analyzed and the blot was reprobed with Histone H4 to contrast
the localization of Brat mRNA with that of a cytoplasmic mRNA.
Fig. 3. Comparison of the spatial distribution of Brat and Xbra
transcripts during gastrulation by whole-mount in situ hybridization.
Brat expression is shown in the upper panel, Xbra expression is
shown in the lower panel. (A,D) Zygotic expression of Brat and Xbra
is first detected just prior to the appearance of the dorsal blastopore
lip (arrow), stage 10-. (B,E) By stage 10+ both Brat and Xbra are
expressed throughout the marginal zone. (C,F) During late
gastrulation, stage 12, Brat is expressed in the ventrolateral
mesoderm surrounding the yolk plug, but not in the axial mesoderm
of the future notochord (arrow). Xbra is also expressed in the
ventrolateral mesoderm, but, in contrast to Brat, it is expressed in the
Fig. 4. Analysis of mesoderm
induction by Brat and Xbra in
animal caps. (A) Comparison of
Xbra and Brat activities reveals
that both induce the ventrolateral
markers Xwnt-8 (Christian et al.,
1991) and XMyoD (Frank and
Harland, 1991); only Brat
however, induces the dorsal
mesoderm markers XFKH-1
(Dirkson and Jamrich, 1992) and
chordin (Sasai et al., 1994). (B) A
more detailed examination of
mesoderm induction by Brat. At
low doses of injected mRNA (14-
42 pg), Brat induced ventralposterior
mesoderm, as scored by
the expression of Xwnt-8, Xhox3
(Ruiz i Altaba and Melton, 1989),
Xlhbox6 (Wright et al., 1990), Xbra, and XMyoD. At intermediate doses (42-140 pg) XFKH-1 is induced, followed by organizer-specific genes
goosecoid (gsc) (Cho et al., 1991) and chordin at higher doses (350-700 pg). (C) In tadpole stage 28 animal caps Brat induces globin, a marker
for blood (the most ventral mesoderm derivative), and muscle actin, which marks dorsal mesoderm. Animal caps were injected with synthetic
mRNA at the 2-cell stage and mesoderm induction was analyzed by RT-PCR at either stage 11 (A,B) or stage 28 (C).
Fig. 5. Brat induces mesodermal tissues. (A) Control-injected animal
caps form solid balls of atypical epidermis. (B) At low doses, 50 pg,
Brat-injected animal caps form vesicles, characteristic of ventral
mesoderm. Caps in (A) and (B) were scored at stage 28.
(C-F) Whole-mount staining for muscle with the 12/101 antibody.
(C) Sibling stage 24 whole embryo. Muscle tissue is stained in the
segmented somites. (D) CS2-injected animal caps, 250 pg mRNA.
No muscle tissue formed, n=11. (E) Brat-injected animal caps, 500
pg mRNA. Muscle formed in 33% of the injected caps, n=27.
(F) BVg1-injected animal caps, 5 pg mRNA. Muscle staining is seen
in all caps, n=22.
Fig. 6. RT-PCR analysis of endoderm induction by Brat in animal
caps at stage 28. Brat induces the endodermal markers, intestinal
fatty acid binding protein (IFABP) and Xlhbox8. IFABP is a general
marker of endoderm while Xlhbox8 is an anterior endoderm marker
of pancreas and liver.
Fig. 7. Brat is induced by peptide growth factors in the FGF and
TGF-b families. (A) Brat displays dose-dependent induction by FGF
or activin protein. Animal caps were excised at stage 8-9 and
exposed to the indicated concentrations of FGF or activin protein.
Caps were harvested at stage 11, and Brat induction was analyzed on
a northern blot. RNA recovery was equal across the blot (not shown).
(B) Brat is induced by BVg1 and BMP-4. Caps were injected with 30
pg of BVg1 mRNA or 1.8 ng BMP-4 mRNA, excised at stage 9,
harvested at stage 11, and scored for Brat expression by RT-PCR.
EF1-alpha expression is a control for RNA recovery and cDNA
synthesis. Maternal Brat transcripts in the animal pole (see Fig 2c)
account for the background signal in vector-injected animal caps.
Fig. 8. A repressor form of Brat, Brat-EnR, inhibits expression of
dorsal and ventral mesodermal markers. (A) Schematic diagram of
wild-type Brat (upper) and a chimera of the N terminus of Brat and
the Drosophila engrailed repressor domain, Brat-EnR (lower). The Tbox
is red, while the engrailed repressor domain is black.
(B) Diagram illustrating the experimental design. 2 ng Brat-EnR and
0.5 ng LacZ mRNA was injected into the marginal zone of both
blastomeres along the first cleavage plane at the 2-cell stage. Albino
embryos were used in which dorsal-ventral differences are not
apparent and therefore the site of injection is randomized with respect
to the dorsal or ventral side. (C,F) Wild-type whole-mount expression
patterns of Xbra and gsc. (D,G) Vegetal view of marker expression in
embryos injected with Brat-EnR. (E,H) Animal view of the same
embryos in D,G to show b-gal staining in the region coinjected with
Brat-EnR and Lac Z mRNA. Note that the expression of gsc and Xbra
is absent on the side expressing Brat-EnR. Xlim expression was also
absent in these embryos (data not shown). Since the site of injection
along the dorsal-ventral axis was randomized, gsc expression was
present opposite the b-gal stain in 50% of the embryos (n=30), as
expected. Xbra expression, on the contrary, was always eliminated in
the injected cells, whether injected dorsally or ventrally.
Fig. 9. Brat-EnR disrupts mesodermal patterning. (A) Vegetal
view stage 11 wild-type embryo, positioned with the dorsal
blastopore lip at the bottom. (B) Dorsal injection of Brat-EnR
inhibits dorsal blastopore lip formation. (C) Ventrally injected
embryo shows normal dorsal lip formation, but ventral lip
formation is blocked. (D) Stage 35 wild-type embryo.
(E) Embryos injected in the dorsal marginal zone with Brat-
EnR mRNA do not develop anterior head structures (note the
lack of eyes) and the neural plate does not close (arrow).
(F) Ventral injection of Brat-EnR disrupts posterior mesoderm
development, but anterior development is not affected. A total
of 2 ng of Brat-EnR was injected into two dorsal or two ventral
blastomeres at the 4-cell stage.
Fig. 10. Rescue of Brat-EnR phenotypes by Brat but not by Xbra.
(A) Wild-type stage 34 embryo. (B) Embryos injected with 250 pg
Brat-EnR into the lateral marginal zone. At this dose, the embryos
have kinked backs and forked tails, and sometimes incomplete
closure of the neural plate. (C) A 2:1 ratio of wild-type Brat: Brat-
EnR rescues the embryonic defects caused by Brat-EnR. (D) A 2:1
ratio of Xbra: Brat-EnR does not rescue the Brat-EnR defects. In fact,
no dose of Xbra tested (0.25:1 to 5:1) rescued the Brat-EnR