57
Signaling Pathways in AP Patterning
signaling in AP patterning of the whole embryo, which
includes mesendodermal tissues, came from studies of organizer-enriched molecules, including the secreted factors
Cerberus and the Frizzled-like gene Frzb-1 (Bouwmeester et
al., 1996; Leyns et al., 1997 ). Their expression domains were
complementary to that of zygotic wnt8a, and both genes,
cer1 and frzb1, were shown to promote head formation
in Xenopus embryos. While Frzb-1 was demonstrated to
inhibit Wnt/β-Catenin signaling by direct binding to the
ligands (Leyns et al., 1997; Lin et al., 1997 ), the action of
Cerberus initially was unclear. The breakthrough came
when it was discovered that induction of a complete secondary axis on the ventral side required not only inhibition of
BMP signaling but also simultaneous inhibition of the Wnt
pathway. BMP inhibitors induced only a secondary trunk,
whereas inhibition of both BMP and Wnt signals produced
a secondary axis with both the head and the trunk (Glinka
et al., 1997). Functional cDNA expression library screening
for endogenous molecules to synergize with BMP inhibitors
to induce a complete secondary axis identif ed Dickkopf
(Dkk1), a secreted molecule that was required for head formation during Xenopus development (Glinka et al., 1998).
Dkk1 turned out to be a Wnt inhibitor via binding to the
Wnt co-receptor Lrp5/Lrp6 to prevent Wnt signaling (Mao
et al., 2001; Semenov et al., 2001; Yamamoto et al., 2008).
Further investigations demonstrated that Cerberus also
inhibited both Wnt and BMP signals by direct binding to
ligands of both families to promote head formation (Glinka
et al., 1997; Piccolo et al., 1999; Silva et al., 2003). This
series of studies helped revise our understanding on how AP
patterning is achieved. Instead of the two-signal model of an
activator and a transformer in neural AP patterning, signal
antagonists are required to actively repress posteriorizing
infuence of the Wnt signal to ensure the normal development of anterior structures. In these cases, the Wnt antagonists may regulate anterior mesendoderm as well as neural
tissues, hence complicating the interpretation of whether the
antagonists act directly to infuence AP neural patterning.
Once it was known that inhibition of Wnt signaling is
crucial for head development, the assays for genes to either
directly regulate head formation or to cooperate with BMP
inhibitors to induce a complete secondary axis were used
to identify other molecules that may regulate Wnt signal
transduction. The expression and the endogenous functions
of these molecules in head-trunk formation in Xenopus
embryos were then analyzed. These studies revealed a surprisingly large number of regulators of Wnt signaling that
have relevant expression domains for patterning the AP
axis of the embryo and act to modulate Wnt signaling to
affect head or trunk formation. Moreover, novel regulatory
mechanisms were uncovered that show how the array of the
new genes modulate Wnt/β-Catenin signaling at different
subcellular levels. For example, the type I transmembrane
receptors Kremen1 and 2 were shown to bind to Dkk1 and
Lrp5/6 to form a ternary complex that enhances endocytosis and removal of Lrp5/6 from the plasma membrane to
inhibit Wnt signaling (Davidson et al., 2002; Mao et al.,
2002). Knockdown of Kremen1/2 leads to head truncation.
Interestingly, in the absence of Dkk1, Kremen2 associates
with Lrp6 to enhance Wnt/β-Catenin signaling (Hassler et
al., 2007). Kremen therefore seems to modulate Wnt signaling levels in a context-dependent manner to regulate early
Xenopus development. Regulation of the Wnt receptors on
the plasma membrane can also be achieved by other molecules. The ER-localized protein Shisa is expressed in the
organizer, can prevent maturation and surface expression
of the Wnt receptor Frizzled, and functions to promote
head formation in Xenopus embryos (Yamamoto et al.,
2005). The organizer-expressed protein tyrosine phosphatase
receptor-type kappa (PTPRK) suppresses Wnt signaling by
regulating surface levels of both Lrp6 and Frizzled via the
transmembrane E3 ubiquitin ligase ZNRF3 to control head
formation (Chang et al., 2020). The ER transmembrane protein TMEM79 interacts with and inhibits the deubiquitinase,
USP8, to facilitate degradation of Frizzled receptor (Chen
et al., 2020). Bighead, a secreted organizer-specif c protein,
binds to Lrp6 to remove it from the cell surface. Gain- and
loss-of-function assays reveal that Bighead is required for
head formation (Ding et al., 2018). Besides the receptors, the
availability of the Wnt ligands is modulated by a variety of
factors. Tiki1 is a transmembrane protein that cleaves Wnt to
promote oxidation and oligomerization of Wnt ligands and
prevents them from binding to the Wnt receptors ( Zhang et
al., 2012). The secreted factor Notum is a Wnt deacylase and
promotes deacylated Wnt ligands to form oxidized oligomers
to prevent Wnt signaling ( Zhang et al., 2015). Both Tiki1 and
Notum are required for head formation, though they may
act in different embryonic regions, with Tiki1 in the organizer and Notum broadly in the ectoderm and weakly in the
mesoderm. Cytoplasmic and nuclear signaling molecules of
the Wnt pathway are also regulated by different factors that
infuence head-trunk development. The phosphatase Pgam5
interacts with and dephosphorylates Dishevelled to prevent
Wnt signaling (Rauschenberger et al., 2017). The GPCR proteins Flop1 and 2 promote β-Catenin degradation (Miyagi
et al., 2015). March2, a membrane-associated E3 ubiquitin
ligase, promotes degradation of Dishevelled via the adaptor
protein Dapper to regulate head formation (Lee et al., 2018).
Idax, a Dvl binding protein, prevents interaction between
Dvl and Axin to inhibit the pathway (Michiue et al., 2004).
NF2/Merlin, a FERM-domain containing protein, also
inhibits Wnt signaling upstream of β-Catenin ( Zhu et al.,
2015). Custos binds to and controls cytoplasmic to nuclear
shuttling of β-Catenin to inhibit Wnt signaling ( Komiya
et al., 2014). The studies of head-trunk formation in Xenopus
therefore prove a fruitful ground for exciting discoveries of
novel Wnt regulators and detailed mechanisms of Wnt regulation in development.
5.3.4. MULTIPLE PATHWAYS AND SIGNAL INTEGRATION
Although RA, FGF, and Wnt are the main signals being
investigated in depth for their roles in AP patterning of the
neural plate and the body axis, other signals also regulate
Signaling Pathways in AP Patterning
signaling in AP patterning of the whole embryo, which
includes mesendodermal tissues, came from studies of organizer-enriched molecules, including the secreted factors
Cerberus and the Frizzled-like gene Frzb-1 (Bouwmeester et
al., 1996; Leyns et al., 1997 ). Their expression domains were
complementary to that of zygotic wnt8a, and both genes,
cer1 and frzb1, were shown to promote head formation
in Xenopus embryos. While Frzb-1 was demonstrated to
inhibit Wnt/β-Catenin signaling by direct binding to the
ligands (Leyns et al., 1997; Lin et al., 1997 ), the action of
Cerberus initially was unclear. The breakthrough came
when it was discovered that induction of a complete secondary axis on the ventral side required not only inhibition of
BMP signaling but also simultaneous inhibition of the Wnt
pathway. BMP inhibitors induced only a secondary trunk,
whereas inhibition of both BMP and Wnt signals produced
a secondary axis with both the head and the trunk (Glinka
et al., 1997). Functional cDNA expression library screening
for endogenous molecules to synergize with BMP inhibitors
to induce a complete secondary axis identif ed Dickkopf
(Dkk1), a secreted molecule that was required for head formation during Xenopus development (Glinka et al., 1998).
Dkk1 turned out to be a Wnt inhibitor via binding to the
Wnt co-receptor Lrp5/Lrp6 to prevent Wnt signaling (Mao
et al., 2001; Semenov et al., 2001; Yamamoto et al., 2008).
Further investigations demonstrated that Cerberus also
inhibited both Wnt and BMP signals by direct binding to
ligands of both families to promote head formation (Glinka
et al., 1997; Piccolo et al., 1999; Silva et al., 2003). This
series of studies helped revise our understanding on how AP
patterning is achieved. Instead of the two-signal model of an
activator and a transformer in neural AP patterning, signal
antagonists are required to actively repress posteriorizing
infuence of the Wnt signal to ensure the normal development of anterior structures. In these cases, the Wnt antagonists may regulate anterior mesendoderm as well as neural
tissues, hence complicating the interpretation of whether the
antagonists act directly to infuence AP neural patterning.
Once it was known that inhibition of Wnt signaling is
crucial for head development, the assays for genes to either
directly regulate head formation or to cooperate with BMP
inhibitors to induce a complete secondary axis were used
to identify other molecules that may regulate Wnt signal
transduction. The expression and the endogenous functions
of these molecules in head-trunk formation in Xenopus
embryos were then analyzed. These studies revealed a surprisingly large number of regulators of Wnt signaling that
have relevant expression domains for patterning the AP
axis of the embryo and act to modulate Wnt signaling to
affect head or trunk formation. Moreover, novel regulatory
mechanisms were uncovered that show how the array of the
new genes modulate Wnt/β-Catenin signaling at different
subcellular levels. For example, the type I transmembrane
receptors Kremen1 and 2 were shown to bind to Dkk1 and
Lrp5/6 to form a ternary complex that enhances endocytosis and removal of Lrp5/6 from the plasma membrane to
inhibit Wnt signaling (Davidson et al., 2002; Mao et al.,
2002). Knockdown of Kremen1/2 leads to head truncation.
Interestingly, in the absence of Dkk1, Kremen2 associates
with Lrp6 to enhance Wnt/β-Catenin signaling (Hassler et
al., 2007). Kremen therefore seems to modulate Wnt signaling levels in a context-dependent manner to regulate early
Xenopus development. Regulation of the Wnt receptors on
the plasma membrane can also be achieved by other molecules. The ER-localized protein Shisa is expressed in the
organizer, can prevent maturation and surface expression
of the Wnt receptor Frizzled, and functions to promote
head formation in Xenopus embryos (Yamamoto et al.,
2005). The organizer-expressed protein tyrosine phosphatase
receptor-type kappa (PTPRK) suppresses Wnt signaling by
regulating surface levels of both Lrp6 and Frizzled via the
transmembrane E3 ubiquitin ligase ZNRF3 to control head
formation (Chang et al., 2020). The ER transmembrane protein TMEM79 interacts with and inhibits the deubiquitinase,
USP8, to facilitate degradation of Frizzled receptor (Chen
et al., 2020). Bighead, a secreted organizer-specif c protein,
binds to Lrp6 to remove it from the cell surface. Gain- and
loss-of-function assays reveal that Bighead is required for
head formation (Ding et al., 2018). Besides the receptors, the
availability of the Wnt ligands is modulated by a variety of
factors. Tiki1 is a transmembrane protein that cleaves Wnt to
promote oxidation and oligomerization of Wnt ligands and
prevents them from binding to the Wnt receptors ( Zhang et
al., 2012). The secreted factor Notum is a Wnt deacylase and
promotes deacylated Wnt ligands to form oxidized oligomers
to prevent Wnt signaling ( Zhang et al., 2015). Both Tiki1 and
Notum are required for head formation, though they may
act in different embryonic regions, with Tiki1 in the organizer and Notum broadly in the ectoderm and weakly in the
mesoderm. Cytoplasmic and nuclear signaling molecules of
the Wnt pathway are also regulated by different factors that
infuence head-trunk development. The phosphatase Pgam5
interacts with and dephosphorylates Dishevelled to prevent
Wnt signaling (Rauschenberger et al., 2017). The GPCR proteins Flop1 and 2 promote β-Catenin degradation (Miyagi
et al., 2015). March2, a membrane-associated E3 ubiquitin
ligase, promotes degradation of Dishevelled via the adaptor
protein Dapper to regulate head formation (Lee et al., 2018).
Idax, a Dvl binding protein, prevents interaction between
Dvl and Axin to inhibit the pathway (Michiue et al., 2004).
NF2/Merlin, a FERM-domain containing protein, also
inhibits Wnt signaling upstream of β-Catenin ( Zhu et al.,
2015). Custos binds to and controls cytoplasmic to nuclear
shuttling of β-Catenin to inhibit Wnt signaling ( Komiya
et al., 2014). The studies of head-trunk formation in Xenopus
therefore prove a fruitful ground for exciting discoveries of
novel Wnt regulators and detailed mechanisms of Wnt regulation in development.
5.3.4. MULTIPLE PATHWAYS AND SIGNAL INTEGRATION
Although RA, FGF, and Wnt are the main signals being
investigated in depth for their roles in AP patterning of the
neural plate and the body axis, other signals also regulate