Click here to close
Hello! We notice that you are using Internet Explorer, which is not supported by Xenbase and may cause the site to display incorrectly.
We suggest using a current version of Chrome,
FireFox, or Safari.
Abstract
We have used the Xenopus embryo as a test system for analyzing the activity of SpAN, a sea urchin metalloprotease in the astacin family containing BMP1 and tolloid. Embryos expressing SpAN initiated gastrulation on a time scale indistinguishable from controls, but invagination of the vegetal pole was subsequently delayed by several hours. At tailbud stages the most severely affected embryos were completely ventralized, lacking all dorsal structures. Molecular analysis of injected embryos, using probes for both dorsal (xgsc and xnot) and ventral (xhox3 and xwnt8) mesoderm, indicates that SpAN ventralizes dorsal mesoderm during gastrula stages. These results mirror those previously obtained with BMP4, suggesting that SpAN may enhance the activity of this ventralizing factor. Consistent with this suggestion, we have shown that SpAN blocks the dorsalizing activity of noggin and chordin, two inhibitory binding proteins for BMP4, but not that of a dominant-negative receptor for BMP4. In contrast, a dominant-negative SpAN, in which the metalloprotease domain has been deleted, dorsalizes ventral mesoderm, a phenotype that can be rescued by coexpressing either SpAN or XBMP1. This suggests that SpAN is mimicking a Xenopus metalloprotease responsible for regulating the activity of Xenopus BMPs during gastrulation. Moreover, our results raise the possibility that SpAN may function to facilitate BMP signaling in early sea urchin embryos.
FIG. 1. SpAN disrupts gastrulation and dorsoanterior development.
Injection of 1.6 ng of span mRNA at the one-cell stage does
not disrupt the appearance of the dorsal blastopore lip (arrows) at
early gastrula stages (B) when compared to embryos injected with
1.6 ng of Dspan mRNA (A) or uninjected controls (not shown). At
later stages during gastrulation, span mRNA-injected embryos (D)
exhibit a delay in blastopore closure (arrowheads) when compared
to controls (C). Note the greater size of the vegetal yolk plug in
span mRNA-injected embryos. By tailbud stages span mRNAinjected
embryos exhibit no dorsoanterior structures (F).
FIG. 2. SpAN ventralizes dorsal mesoderm. (A) Total RNA isolated from uninjected embryos and embryos injected with 1.6 ng of either
Dspan or span mRNA was analyzed by RNase protection using probes for either muscle-specific a-actin (m-actin), which also detects
cytoskeletal actin (c-actin), or a combination of blood-specific aT4-globin and the uniformly expressed ornithine decarboxylase (ODC). The
results show a significant decrease in expression of a-actin mRNA in span mRNA-injected embryos, confirming the loss of dorsal
structures, and a small increase in the expression of aT4-globin. (B) Animal caps were isolated from injected (1.6 ng of Dspan or span mRNA)
or uninjected mid-blastulae (stage 8), which were then incubated in the presence (1) or absence (2) of human activin A until control
embryos had reached tailbud stage (stage 27). Total RNA was isolated and analyzed by RNase protection using probes for a-actin (m-actin)
and aT4-globin. The results show that SpAN ventralizes the mesoderm in activin-treated caps, lowering the expression of muscle a-actin
and increasing expression of aT4-globin.
FIG. 3. SpAN ventralizes dorsal mesoderm during gastrulation. Dorsal marginal zones (DMZs) were isolated from both control and span
mRNA-injected (1.6 ng/embryo) early gastrulae, incubated until control embryos reached the indicated stages, and analyzed for the
expression of both dorsal- and ventral-specific genes by RNase protection (A). Levels of ornithine decarboxylase (ODC) mRNA were
determined as a loading control. The dorsal-specific genes goosecoid (gsc) and xnot are initially induced in span mRNA-injected DMZs, but
expression is subsequently down-regulated. In contrast, expression of the ventral-specific genes xhox3 and xwnt8 is increased in span
mRNA-injected embryos as gastrulation progresses. Gastrulae were also analyzed for xnot (B, C) and xwnt8 (D, E) expression by whole
mount in situ hybridization. In control stage 12 embryos, xnot was clearly localized to the dorsal blastopore lip (B) but was absent or very
diffuse in span mRNA-injected embryos (C), consistent with the loss of dorsal structures. In contrast, xwnt8 was expressed in ventral and
lateral mesoderm, but not dorsal mesoderm, of control stage 12 embryos (D), but expression expanded into the dorsal mesoderm of span
mRNA-injected embryos (E), demonstrating that ventralization of the dorsal marginal zone has occurred by this stage.
FIG. 4. SpAN blocks the dorsalizing activity of chordin and
noggin. Embryos were injected with the indicated mRNAs (see text
for amounts) and VMZs were cut from early gastrulae. These were
cultured until sibling control embryos had reached stage 27 and
then analyzed for expression of muscle-specific a-actin (m-actin)
and blood-specific aT4-globin. Levels of cytoskeletal actin (c-actin)
were determined as a loading control for muscle actin, and ODC
was determined as a loading control for aT4-globin. chordin (lane
3), noggin (lane 5), and a dominant-negative BMP receptor
(DxBMPR; lane 7) dorsalize ventralmesoderm as shown by the high
levels of a-actin expression and lower levels of aT4-globin expression
when compared to control VMZs (lane 1). SpAN was not able
to ventralize dorsal mesoderm induced by DxBMPR(lane 8) indicating
that SpAN acts upstream of the BMP receptor and requires a
functional BMP signaling pathway to act. SpAN, however, is able
to overcome the dorsalizing activity of both chordin (lane 4) and
noggin (lane 6).
FIG. 5. DNSpAN dorsalizes ventralmesoderm. (A) Schematic
diagram illustrating the structure of SpAN (top) and DNSpAN
(bottom). The C-terminal EGF- (E), CUB- (C), and threonine-rich (T)
domains of SpAN were fused to the signal sequence (S) of XBMP1,
thereby deleting the proregion (P) and metalloprotease (MP) domains.
(B) Injection of 1.5 ng of dnspan mRNA blocks the ventralizing
activity of 150 pg of coinjected span mRNA in activin-treated
animal caps. Embryos were injected at the two-cell stage and
incubated until stage 27 when they were analyzed by RNase
protections using a probe for a-actin. Levels of cytoskeletal actin
(c-actin) were determined as a loading control. (C, D) VMZs
isolated at stage 10 and incubated until stage 40, whereas controls
differentiate ventral-type mesoderm (C), DNSpAN injected VMZs
(D) differentiate muscle (mu) and neural tissue (nt). (E) RNase
protection analysis using probes for muscle-specific a-actin (mactin)
and blood-specific aT4-globin demonstrates that DNSpAN
expressing VMZs are dorsalized. (F) RNase protection analysis on
isolated VMZs using a probe for a-actin (m-actin), demonstrating
that coexpressed SpAN or XBMP1 block the dorsalizing activity of
DNSpAN.
