J Cell Biol
June 1, 1996;
Activities of the Wnt-1 class of secreted signaling factors are antagonized by the Wnt-5A class and by a dominant negative cadherin in early Xenopus development.
When overexpressed in Xenopus embryos, Xwnt-1
, -3A, -8 and -8b define a functional class of Wnts (the Wnt-1
class) that promotes duplication of the embryonic axis, whereas Xwnt-5A
, -4, and -11 define a distinct class (the Wnt-5A
class) that alters morphogenetic movements (Du, S., S. Purcell, J. Christian, L. McGrew, and R. Moon. 1995. Mol. Cell. Biol. 15:2625-2634). Since come embryonic cells may be exposed to signals from both functional classes of Wnt during vertebrate development, this raises the question of how the signaling pathways of these classes of Wnts might interact. To address this issue, we coexpressed various Xwnts and components of the Wnt-1
class signaling pathway in developing Xenopus embryos. Members of the Xwnt-5A
class antagonized the ability of ectopic Wnt-1
class to induce goosecoid
expression and a secondary axis. Interestingly, the Wnt-5A
class did not block goosecoid
expression or axis induction in response to overexpression of cytoplasmic components of the Wnt-1
signaling pathway, beta-catenin or a kinase-dead gsk-3
, or to the unrelated secreted factor, BVg1. The ability of the Wnt-5A
class to block responses to the Wnt-1
class may involve decreases in cell adhesion, since ectopic expression of Xwnt-5A
leads to decreased Ca2+-dependent cell adhesion and the activity of Xwnt-5A
to block Wnt-1
class signals is mimicked by a dominant negative N-cadherin
. These data underscore the importance of cell adhesion in modulating the responses of embryonic cells to signaling molecules and suggest that the Wnt-5A
functional class of signaling factors can interact with the Wnt-1
class in an antagonistic manner.
[+] show captions
References [+] :
Figure 1. Xwnt-5A antagonizes Wnt-1 class activity. (A) Control embryos possess single axes at stage 25. (B) Embryos injected with prolactin
followed by Xwnt-8 RNA have duplicated axes. (C) Embryos injected with Xwnt-5A followed by Xwnt-8 RNA have single axes.
(D) Uninjected stage 10 control embryos possess a single site of gsc expression. (E) Embryos injected with prolactin followed by Xwnt-8
RNA display multiple sites of gsc expression. (F) Embryos injected with Xwnt-5A followed by Xwnt-8 RNA have a single site of gsc expression.
(Arrows) Axes (A, B, and C) and sites of gsc expression (D, E, and F).
Figure 2. Detection of gsc
transcripts by RT-PCR in
stage 10 embryos ventralized
by UV irradiation. (A)
Xwnt-5A inhibits the rescue
of gsc expression by Xwnt-8.
(B) AN-cadherin blocks the
rescue of gsc expression by
Xwnt-8. ( C) Neither Xwnt-ll
nor AN-cadherin block the
rescue of gsc expression by
BVgl. (A, B, and C; lane 1)
Control embryos. (A, B, and
C; lane 2) Uninjected embryos
UV irradiated to eliminate
endogenous gsc expression.
(Lane 3) UV-irradiated embryos injected with prolactin followed by Xwnt-8 RNA (A and B) or prolactin followed by BVgl RNA
(C). (Lane 4) UV-irradiated embryos injected with Xwnt-5A (A), AN-cadherin (B), or Xwnt-ll (C) RNA followed by Xwnt-8 (A and B)
or BVgl (C) RNA. (Lane 5) UV-irradiated embryos injected with wild type N-cadherin (B) or AN-cadherin RNA (C) followed by
Xwnt-8 (B) or BVgl RNA (C). All EFla lanes serve as controls for RT-PCR (see Materials and Methods).
Figure 3. Xwnt-4 antagonizes the activity of Xwnt-8 after MBT.
The ability of CSKA-Xwnt-8 to interfere with notochord formation
(Christian and Moon, 1993) was scored in stage 25 embryos
by anti-Tor 70 whole mount immunocytochemistry to detect the
notochord and by F-spondin in situ hybridization to detect the
floorplate. (A) Uninjected embryos possess a normal notochord
and floorplate. (B) Transcription of Xwnt-8 from the cytoskeletal
actin promoter vector after MBT leads to loss of notochord and
floorplate staining. (C) Xwnt-4 interferes with the Xwnt-8 activity,
restoring the formation of the notochord and the floorplate. (Arrowheads)
F-spondin floorplate staining. (Straight arrows) Tor 70 notochord
staining. (Curved arrows) anterior limit of the notochord.
Figure 4. Overexpression of Xwnt-5A inhibits open face dorsal lip explant elongation and decreases Ca2+-dependent cell reaggregation.
(A) Uninjected control dorsal lip explants elongate normally. (B) Dorsal lip explants overexpressing Xwnt-5A RNA do not elongate.
(C) Xwnt-8 does not inhibit dorsal lip explant elongation. (D) Xwnt-5A blocks dorsal lip explant elongation when coexpressed
with Xwnt-8 at a 1:1 ratio of injected RNAs. (E) Dissociated control dorsal lip explants reaggregate in a CaZ÷-dependent manner. In the
absence of Ca 2÷, reaggregation is inhibited (inset). (F) Xwnt-5A inhibits Ca2÷-dependent cell reaggregation, which can be rescued by
co-expressing Xwnt-5A with N-cadherin (inset). (G) Xwnt-8 has no appreciable effect on caa÷-dependent cell reaggregation. (H) Xwnt-
5A blocks Ca2÷-dependent cell reaggregation when coexpressed with Xwnt-8 at a 1:1 ratio of injected RNAs. dN-cadherin blocks Ca 2÷dependent
cell reaggregation in a manner similar to Xwnt-5A (inset).
Figure 5. The inhibition of elongation of the gastrula organizer explants by Xwnt-5A does not require Xwnt-5A to be expressed in all
cells of the explant. Embryos were injected into the marginal zone of one dorsal cell at the four-cell stage with either fl-galactosidase
RNA mixed with rhodamine dextran (A), or with Xwnt-5A RNA mixed with rhodamine dextran (B). Open face explants of the entire
dorsal marginal zone were prepared at stage 10 and visualized by fluorescence microscopy after control embryos had developed to stage
13-15. Control explants (A) elongate extensively (6 of 8 explants), while explants expressing Xwnt-5A (B) display no convergence and
extension movements (10 of 15 explants), or reduced elongation (3 of 15 explants). In both panels, pseudocolor imaging reveals that the
injected dextran and RNAs were restricted to the yellow-red cells.
Figure 6. AN-cadherin but not N-cadherin inhibits the induction of ectopic gsc expression by Xwnt-8 as assayed by in situ hybridization
for gsc in stage 10 gastrula embryos. (A) Uninjected control embryos possess a single site of gsc expression. (B) Embryos injected with
prolactin followed by Xwnt-8 RNA possess two sites of gsc expression. (C) Embryos injected with AN-cadherin followed by Xwnt-8
RNA possess a single site of gsc expression. (D) Embryos injected with N-cadherin followed by Xwnt-8 RNA possess two sites of gsc expression.
(Arrows) Representative sites of gsc expression.
Figure 7. Members of the Wnt-5A class and AN-cadherin do not block the induction of ectopic gsc expression by BVgl, as assayed by in
situ hybridization for gsc in stage 10 gastrula embryos. (A) Uninjected control embryos possess a single site of gsc expression. (B) Embryos
injected withprolactin followed by BVgl RNA possess two sites ofgsc expression. (C) Embryos injected with Xwnt-ll or AN-cadherin
(D) RNA followed by BVgl RNA also possess multiple sites of gsc expression. (Arrows) Sites of gsc expression.
Tortes et al. Wnt Antagonism 1133
Downloaded from jcb.rupress.org on May 8, 2015
Published June 1, 1996
Loss of epithelial differentiation and gain of invasiveness correlates with tyrosine phosphorylation of the E-cadherin/beta-catenin complex in cells transformed with a temperature-sensitive v-SRC gene.