|
Fig. 1. (A) Chimeric Vg1 constructs. Schematic representation of
activin bB, BMP2, Vg1, and chimeric BMP2-Vg1 and activin-Vg1.
Members of the TGF-b superfamily, these genes contain an signal
sequence, pro-region, tetrabasic cleavage site and mature region.
Activin bB and BMP2, but not Vg1, form disulfide-linked dimers
that are subsequently cleaved, releasing the mature C-terminal
peptide as a secreted bioactive dimer. The chimeric constructs, fused
four amino acids downstream of the cleavage site, are designed to
facilitate processing and secretion of mature Vg1. (B) Western blot
analysis of oocyte supernatants. Oocytes were injected with the
indicated mRNA and conditioned supernatants (10 ml) analyzed by
western blotting of reducing SDS-PAGE using activin-specific (left)
or Vg1-specific (right) antisera. Activin bB mRNA directs secretion
of mature protein. While Vg1 mRNA directs no secretion and
BMP2-Vg1 directs low levels of secretion, activin-Vg1 mRNA
results in secretion of abundant mature Vg1 protein.
|
|
Fig. 2. (A) Mature Vg1 induces morphogenetic movements in ectodermal explants. Blastula-stage animal pole explants were treated with 10%
supernatant (1/10 dilution) until the gastrula stage and cultured to the neurula stage (stage 15). Mature Vg1 (Vg) results in a dramatic
elongation indicative of mesoderm induction (Symes and Smith, 1987). Activin bB (bB) results in similar elongation, while supernatant of
uninjected oocytes (CT) have no effect. Scale bar, 300 mm. (B) Dose-dependent induction of mesodermal markers by mature Vg1. Blastulastage
animal pole explants were treated with increasing doses of mature Vg1 supernatant (1%, 3%, 10%, 30%), or control (30%) or activin bB
supernatant (10%), cultured to the neurula stage and analyzed by RT-PCR. At low doses (1%) the general marker Xbra is induced; at
intermediate doses (3%) the ventrolateral marker Xwnt8 and the dorsolateral marker cardiac actin (M. Actin) are induced; and at high doses (10-
30%) the dorsoanterior markers goosecoid (Gsc) and noggin (Nog), and the neural marker NCAM are induced. Activin bB treatment results in
a similar response and control supernatants have no effect. EF1a is a loading and reverse transcription control and the embryo and embryo-RT
are additional positive and negative controls. (C) Induction of embryoids by mature Vg1. Blastula animal pole explants treated with a high dose
of mature Vg1 and cultured to the late tadpole stage (stage 40), differentiated into embryoids displaying a rudimentary axial organization, with
anterior-posterior pattern and head structures. The embryoid shown (top right) has a clear head-to-tail pattern and pigmented eye and cement
gland. A histological section (bottom right) reveals differentiated notochord (no), somitic muscle (sm), neural tube (nt), eye (e) and cement
gland (cg). Treatment with supernatant of uninjected oocytes has no effect and explants form atypical epidermis (left). Scale bar, 150 mm.
|
|
Fig. 3. Mature Vg1 is not bound by the activin type II receptor.
Uninjected oocytes, or oocytes expressing a myc-tagged activin type
II receptor (XARmyc) were bound with radiolabeled activin bB or
mature Vg1, complexes chemically crosslinked, immunoprecipitated
with a myc-specific antiserum and visualized by 15% reducing SDSPAGE
and fluorography. Co-precipitating activin is released from
complexes by reducing the crosslinker, and is therefore resolved as
an ~14´103 Mr band. Uninjected oocytes did not bind radiolabeled
activin bB (lane 1) and XARmyc-expressing oocytes displayed no
binding in the absence of radiolabeled ligand (lane 2). Receptorexpressing
oocytes strongly bound activin (lane 3) and this binding
was competed by a 10-fold excess of unlabeled activin (lane 4). In
contrast, a 10-fold excess of unlabeled mature Vg1 was unable to
compete activin binding (lane 5). Furthermore, a 10-fold higher dose
of radiolabeled Vg1 than that required for activin binding did not
result in detectable binding (lane 6). The higher molecular weight
bands (lanes 3, 5) may be due to non-reduced ligand-receptor
complexes and co-precipitating unprocessed activin.
|
|
Fig. 4. Follistatin does not inhibit endogenous dorsal mesoderm induction. Both blastomeres
of 2-cell-stage embryos were injected with b-galactosidase mRNA (A) or follistatin mRNA
(B) (1 ng total mRNA) and cultured to the tadpole stage. Follistatin resulted in a loss of
posterior structures with maintenance or enhancement of dorsoanterior structures.
Histological analysis reveals the presence of dorsal mesodermal tissues in both b-
galactosidase (C)- and follistatin (D)-injected embryos (no, notochord; sm, somitic muscle;
nt, neural tube; cg, cement gland). (E) Follistatin does not inhibit expression of mesodermal
markers at the gastrula stage. 2-cell-stage embryos were injected with a total of 2 ng of b-
galactosidase (bGal) or follistatin (XFS) mRNA, harvested at the mid-gastrula stage (stage
11) and analyzed by RT-PCR. The expression of a general mesodermal marker, brachyury
(Xbra), the dorsal markers, goosecoid (Gsc) and noggin (Nog), and a ventrolateral marker,
Xwnt8, were unaffected by follistatin. EF1a is a loading and reverse transcription control.
Scale bars, 1 mm (A,B) and 0.5 mm (C,D).
|
|
Fig. 5. Follistatin does not inhibit mature Vg1 activity. Blastulastage
animal pole explants were treated with control (A,B), activin
bB (C,D), or mature Vg1 (E,F) supernatants following mixture and
preincubation with control (A,C,E) or follistatin (B,D,F)
supernatants. While follistatin fully blocks activin-induced
morphogenetic movements, induction by mature Vg1 was
unaffected. (G) Follistatin fails to inhibit cardiac actin (M. Actin)
induction by mature Vg1. Blastula animal pole explants were treated
with mature Vg1 or activin only, or with these supernatants
preincubated with follistatin. Upon reaching the neurula stage,
samples were analyzed by RT-PCR. While follistatin fully inhibits
cardiac actin induction by activin, no effect on mature Vg1 activity is
detected. EF1a is a loading and reverse transcription control. Scale
bar, 300 mm.
|
|
Fig. 6. Mature Vg1 does not dorsalize gastrula ventral marginal zone
explants. Dorsal or ventral marginal zone explants were prepared
from blastulae (stage 8) or early gastrulae (stage 10.25) and
incubated in supernatants (30%) containing activin, mature Vg1 or
noggin, or in a supernatant of uninjected oocytes (control). At the
neurula stage, samples were analyzed by RT-PCR. While all three
factor induced cardiac actin (M. Actin) in blastula ventral marginal
zone explants, only noggin did so in gastrula-stage explants. EF1a is
a loading and reverse transcription control.
|