58
Xenopus
AP axis formation. Both the Nodal and the BMP branches
of the TGF-β pathway have been shown to modulate AP patterning (Piccolo et al., 1999; Polevoy et al., 2019). Ectopic
stimulation of Nodal or BMP signaling leads to head truncation of Xenopus embryos, whereas Cerberus inhibits
both Nodal and BMP in addition to Wnt. Crescent, another
organizer-specifc secreted factor, belongs to the soluble
Frizzled-related protein (sFRP) family and also regulates
both Wnt and BMP signals (Pera and De Robertis, 2000;
Ploper et al., 2011; Shibata et al., 2005; Shibata et al., 2000).
The Nodal-like ligand Derriere is also involved in posterior
development. Derriere is expressed in the marginal zone
with high dorsal levels during gastrulation and is detected
subsequently in the posterior region of neurula embryos.
Ectopic expression of Derriere leads to the reduction of head
structures (Sun et al., 1999). In addition, insulin-like growth
factor (IGF) signaling regulates head formation via inhibition of the Wnt pathway (Richard-Parpaillon et al., 2002).
Therefore, many pathways participate in controlling AP patterning during Xenopus embryogenesis.
As expected, the signals do not work independently and
often cooperate with each other. Activation of posterior
markers in neuralized animal caps or embryos exhibits a
degree of co-dependence between the Wnt, RA, and FGF
signaling pathways so that in the absence of one, the induction of hindbrain or spinal cord markers by another signal is
often compromised (Domingos et al., 2001; McGrew et al.,
1997; Roche et al., 2009; Shiotsugu et al., 2004). Redundancy
of multiple signals in vivo may make detection of signal
cooperation more complex in whole embryos. Quite often
a regulatory protein for AP patterning can simultaneously
modulate different pathways: Cerberus antagonizes Nodal,
BMP, and Wnt signals (Piccolo et al., 1999), and Crescent
regulates Wnt as well as BMP pathways (Ploper et al., 2011).
The transmembrane protein Shisa controls maturation and
cell surface localization of both FGFR and Frizzled receptors (Yamamoto et al., 2005). Cytoplasmic kinases and
phosphatases can also modulate phosphorylation of an array
of proteins involved in distinct signaling pathways. On the
other hand, closely related protein members of the same family can have different regulatory activities for distinct pathways. Examples include Noggin 4, which blocks Wnt instead
of BMP signaling (Eroshkin et al., 2016); Noggin 2, which
inhibits Nodal and Wnt in addition to BMP (Bayramov et
al., 2011); and Dkk3, which regulates TGF-β and FGF pathways instead of Wnt pathways (Pinho and Niehrs, 2007).
In addition to direct interaction of components of different pathways, crosstalk between signals can happen at the
transcription level. Expression of Fgfr1 and Fgfr4 depends
on RA signaling (Shiotsugu et al., 2004), whereas Wnt3a
activates transcription of FGF3 and FGF8 via the transcription factor Meis3 (Gutkovich et al., 2010), and FGF and Wnt
ligands can be co-regulated by the same transcription factors involved in AP patterning (e.g. Tbx6 activates expression of fgf8 and wnt8a, whereas JunB stimulates expression
of both fgf3 and wnt8a (Lou et al., 2006; Yoshida et al.,
2016). Distinct signals can also converge directly in the
gene regulatory networks that control expression of caudalizing factors. DNA binding motifs for Ets and TCF, the
transcription factors downstream of FGF and Wnt signals,
respectively, are both present and often positioned in close
proximity in the regulatory region of cdx4 and cdx1, encoding caudal-like homeodomain proteins involved in posterior
development (Haremaki et al., 2003; Isaacs et al., 1998;
Kjolby and Harland, 2017; Kjolby et al., 2019). FGF, Wnt,
and RA have all be shown to regulate hox gene expression
(Dekker et al., 1992; Durston, 2019; In der Rieden et al.,
2010; Janssens et al., 2010) and may potentially cooperate
at the regulatory regions of the hox genes via binding by
respective effector transcription factors of these signals,
although the exact mechanisms of regulatory input have not
been investigated in detail for all the signals.
Control of AP patterning by multiple pathways begs the
questions whether all signals function similarly and the
redundancy simply exists to ensure robustness of the patterning system, or different signals have non-overlapping activities to inf uence specifc aspects of AP development. Most
of the molecular studies on AP neural formation focused
on a limited set of markers, hence restricting a distinction
between the different effects by various signals. Conclusions
drawn from explant assays are frequently considered to
equate to those using whole embryos, but the expression
levels of the markers are often analyzed in explants and
the positions of the marker domains assessed in manipulated embryos. Results from explant and embryo studies are
therefore not always congruent, especially in the spinal cord
region. For example, although posterior hox genes can be
blocked effciently by functional reduction of a caudalizing
factor in explants, the expression domains of the same hox
genes can remain the same in manipulated embryos (Curran
and Grainger, 2000). The embryonic regions most sensitive
to altered signaling levels are in the fore-, mid- and hindbrain, with their respective markers shifting anteriorly or
posteriorly relative to the control side when the signal levels are modifed. The comparative insensitivity of the spinal
cord to signal manipulation may refect functional redundancy of multiple signals, but it can also indicate that the
signals have limited AP patterning capacity in the trunk
region of the embryos. Simultaneous modulation of several
pathways may help to resolve the issue.
5.4. SYNTHESIS OF THE CLASSICAL
AND MODERN STUDIES AND
SOME UNRESOLVED ISSUES
Classical experimental embryologists used microsurgery,
explant, or transplant studies to address several crucial
questions concerning AP embryonic patterning in amphibians. What regions of early embryonic tissues give rise to
anterior or posterior structures at later stages? When do the
tissues acquire AP characteristics? How can naïve tissues
be specif ed along a particular developmental path; can this
developmental trajectory be altered; and, if so, how? How
Xenopus
AP axis formation. Both the Nodal and the BMP branches
of the TGF-β pathway have been shown to modulate AP patterning (Piccolo et al., 1999; Polevoy et al., 2019). Ectopic
stimulation of Nodal or BMP signaling leads to head truncation of Xenopus embryos, whereas Cerberus inhibits
both Nodal and BMP in addition to Wnt. Crescent, another
organizer-specifc secreted factor, belongs to the soluble
Frizzled-related protein (sFRP) family and also regulates
both Wnt and BMP signals (Pera and De Robertis, 2000;
Ploper et al., 2011; Shibata et al., 2005; Shibata et al., 2000).
The Nodal-like ligand Derriere is also involved in posterior
development. Derriere is expressed in the marginal zone
with high dorsal levels during gastrulation and is detected
subsequently in the posterior region of neurula embryos.
Ectopic expression of Derriere leads to the reduction of head
structures (Sun et al., 1999). In addition, insulin-like growth
factor (IGF) signaling regulates head formation via inhibition of the Wnt pathway (Richard-Parpaillon et al., 2002).
Therefore, many pathways participate in controlling AP patterning during Xenopus embryogenesis.
As expected, the signals do not work independently and
often cooperate with each other. Activation of posterior
markers in neuralized animal caps or embryos exhibits a
degree of co-dependence between the Wnt, RA, and FGF
signaling pathways so that in the absence of one, the induction of hindbrain or spinal cord markers by another signal is
often compromised (Domingos et al., 2001; McGrew et al.,
1997; Roche et al., 2009; Shiotsugu et al., 2004). Redundancy
of multiple signals in vivo may make detection of signal
cooperation more complex in whole embryos. Quite often
a regulatory protein for AP patterning can simultaneously
modulate different pathways: Cerberus antagonizes Nodal,
BMP, and Wnt signals (Piccolo et al., 1999), and Crescent
regulates Wnt as well as BMP pathways (Ploper et al., 2011).
The transmembrane protein Shisa controls maturation and
cell surface localization of both FGFR and Frizzled receptors (Yamamoto et al., 2005). Cytoplasmic kinases and
phosphatases can also modulate phosphorylation of an array
of proteins involved in distinct signaling pathways. On the
other hand, closely related protein members of the same family can have different regulatory activities for distinct pathways. Examples include Noggin 4, which blocks Wnt instead
of BMP signaling (Eroshkin et al., 2016); Noggin 2, which
inhibits Nodal and Wnt in addition to BMP (Bayramov et
al., 2011); and Dkk3, which regulates TGF-β and FGF pathways instead of Wnt pathways (Pinho and Niehrs, 2007).
In addition to direct interaction of components of different pathways, crosstalk between signals can happen at the
transcription level. Expression of Fgfr1 and Fgfr4 depends
on RA signaling (Shiotsugu et al., 2004), whereas Wnt3a
activates transcription of FGF3 and FGF8 via the transcription factor Meis3 (Gutkovich et al., 2010), and FGF and Wnt
ligands can be co-regulated by the same transcription factors involved in AP patterning (e.g. Tbx6 activates expression of fgf8 and wnt8a, whereas JunB stimulates expression
of both fgf3 and wnt8a (Lou et al., 2006; Yoshida et al.,
2016). Distinct signals can also converge directly in the
gene regulatory networks that control expression of caudalizing factors. DNA binding motifs for Ets and TCF, the
transcription factors downstream of FGF and Wnt signals,
respectively, are both present and often positioned in close
proximity in the regulatory region of cdx4 and cdx1, encoding caudal-like homeodomain proteins involved in posterior
development (Haremaki et al., 2003; Isaacs et al., 1998;
Kjolby and Harland, 2017; Kjolby et al., 2019). FGF, Wnt,
and RA have all be shown to regulate hox gene expression
(Dekker et al., 1992; Durston, 2019; In der Rieden et al.,
2010; Janssens et al., 2010) and may potentially cooperate
at the regulatory regions of the hox genes via binding by
respective effector transcription factors of these signals,
although the exact mechanisms of regulatory input have not
been investigated in detail for all the signals.
Control of AP patterning by multiple pathways begs the
questions whether all signals function similarly and the
redundancy simply exists to ensure robustness of the patterning system, or different signals have non-overlapping activities to inf uence specifc aspects of AP development. Most
of the molecular studies on AP neural formation focused
on a limited set of markers, hence restricting a distinction
between the different effects by various signals. Conclusions
drawn from explant assays are frequently considered to
equate to those using whole embryos, but the expression
levels of the markers are often analyzed in explants and
the positions of the marker domains assessed in manipulated embryos. Results from explant and embryo studies are
therefore not always congruent, especially in the spinal cord
region. For example, although posterior hox genes can be
blocked effciently by functional reduction of a caudalizing
factor in explants, the expression domains of the same hox
genes can remain the same in manipulated embryos (Curran
and Grainger, 2000). The embryonic regions most sensitive
to altered signaling levels are in the fore-, mid- and hindbrain, with their respective markers shifting anteriorly or
posteriorly relative to the control side when the signal levels are modifed. The comparative insensitivity of the spinal
cord to signal manipulation may refect functional redundancy of multiple signals, but it can also indicate that the
signals have limited AP patterning capacity in the trunk
region of the embryos. Simultaneous modulation of several
pathways may help to resolve the issue.
5.4. SYNTHESIS OF THE CLASSICAL
AND MODERN STUDIES AND
SOME UNRESOLVED ISSUES
Classical experimental embryologists used microsurgery,
explant, or transplant studies to address several crucial
questions concerning AP embryonic patterning in amphibians. What regions of early embryonic tissues give rise to
anterior or posterior structures at later stages? When do the
tissues acquire AP characteristics? How can naïve tissues
be specif ed along a particular developmental path; can this
developmental trajectory be altered; and, if so, how? How