128
Xenopus
roles have been described for Kremen2, which mediates
its effects through canonical Wnt signaling (Hassler et al.
2007). Taken together, the combined activity of these proteins and signaling pathways establishes an embryonic territory, the neural plate border, which constitutes the earliest
recognizable domain of the neural crest in vertebrates.
The neural plate border region is characterized by the
expression of a set of transcription factors implicated in
intricate regulatory interactions that include positive and
negative feedback loops, feed-forward loops, and mutual
repression. Although there is substantial inter-species variation in spatiotemporal expression patterns and the exact
types of regulatory interactions, all vertebrates nonetheless
express a common suite of neural plate border genes. For
Xenopus, these include members of the Pax (pax3, pax7)
(Milet et al. 2013), Msx (msx1, msx2) (Khadka et al. 2003;
Monsoro-Burq et al. 2005), and Zic (zic1, zic2, zic3, zic5)
(Kuo et al. 1998; Nakata et al. 1998) families, as well as
dlx3 (Feledy et al. 1999; Luo et al. 2001), gbx2.2 (Li et al.
2009 ), irx1 (Glavic et al. 2004), snai1 (Aybar et al. 2003),
tfap2a (de Crozé et al. 2011), and hes4/hairy2 ( Vega-López
et al. 2015). Recently, there has been evidence that miRNAs
(mir-301a, mir-338) and miRNA-associated proteins are
also expressed at the neural plate border and may be important for maintaining pluripotency (Gessert et al. 2010; Ward
et al. 2018). Together, these factors progressively ref ne the
spatial boundaries of the neural crest domain to distinguish
it from adjacent tissues, activate downstream target genes
responsible for maintaining the neural crest in a stem cell
state, and eventually endow these cells with the ability to
migrate (Bae et al. 2014).
Work in Xenopus has suggested that the core regulatory
circuitry of the neural plate border includes pax3, msx1, zic1,
and tfap2a (Hong and Saint-Jeannet 2007; Plouhinec et al.
2014; Pla and Monsoro-Burq 2018). There is evidence that
these transcription factors are some of the earliest targets of
signaling through Wnt, BMP, FGF, and retinoic acid pathways
and pattern the neural plate border by mediating the effects
of these signaling pathways on downstream targets (Tríbulo
et al. 2003). Some neural plate border factors are capable of
driving expression of downstream neural crest genes in neuralized (i.e. BMP-attenuated) ectodermal explants. For example, expression of either tfap2a or msx1 in neuralized explants
is suffcient to activate expression of neural crest genes pax3,
msx1, sox9, sox10, snai2 (Monsoro-Burq et al. 2005; Sato et
al. 2005; Hong and Saint-Jeannet 2007; de Crozé et al. 2011).
Gain- and loss-of-function experiments have revealed some of
the epistatic relationships among neural plate border factors,
as well as the signaling pathways through which they mediate
regulatory control. For example, msx1 is required for activation of pax3, whereas tfap2a can coordinate with canonical
Wnt signaling to activate pax3 (de Crozé et al. 2011). pax3
then works with msx1 to promote FGF signaling by activating
early neural crest genes such as snai2. Unlike the Tfap2a and
Msx1 factors, neither Pax3 nor Zic1 alone are capable of promoting formation of neural crest cells in neuralized explants.
However their combined activity is suffcient to drive robust
expression of sox9, sox10, snai2, and meis3, followed by
tubb2b (n-tubulin) and th (tyrosine hydroxylase) during differentiation into neural and melanocyte derivatives, respectively
(Maeda et al. 2001). Combined Pax3/Zic1 activity can also
promote Cadherin switching, EMT, and production of melanocytes in ectodermal explants (Milet and Monsoro-Burq
2012) and can reprogram other tissues such as ventral ectoderm toward a neural crest progenitor state (Sato et al. 2005;
Milet et al. 2013). Importantly, however, these effects on the
neural crest are highly dose dependent. For example, although
increased levels of Pax3 can promote ectopic neural crest,
high pax3 expression converts neural crest-fated ectoderm to
an alternative fate, hatching gland (Hong and Saint-Jeannet
2007). Another important function of neural crest GRN components at these stages is to refne spatial boundaries between
neural crest and other ectodermal populations. This can occur
through transcriptional repression of sox2 and sox3 by Snai1
at the lateral edges of the neural plate (LaBonne and BronnerFraser 1998, 2000; Aybar et al. 2003; Langer et al. 2008) and
by Nbx/Nkx1–2-mediated inhibition of sox2 and otx2 ( Kurata
and Ueno 2003). Conversely, Prdm12 is expressed in the preplacodal ectoderm and can inhibit expression of neural crest
genes such as foxd3, snai2, and sox8 (Matsukawa et al. 2015).
Beyond the core neural plate border subcircuit consisting
of pax3-msx1-zic1-tfap2a, studies in Xenopus have uncovered additional factors essential for regulating the establishment of defnitive neural crest. These include Pbx1 and Meis1
(Maeda et al. 2002), Pcdh7 (Rashid et al. 2018), Klf4/Neptune
(Kurauchi et al. 2010), and Znf703, which act downstream of
Pax3 (Hong and Saint-Jeannet 2017) and activate the neural
crest genes snai2 and sox10. Similarly, Hmga2 (Macri et al.
2009, 2016) and Fbxw7/Cdc4 (Almeida et al. 2010) act
downstream of Pax3 and Msx1 and upstream of myc, snai1,
and snai2. More recently, unbiased genome-wide analyses
have described the global transcriptomic landscape at the
neural plate border. These studies have suggested that, once
activated, Pax3 creates an autoregulatory loop and, together
with Zic1, activates defnitive neural crest regulatory factors
including ednra, gbx2.2, sox8, sox9, sox10, twist1, tfap2b,
snai2, and foxd3 (Bae et al. 2014; Plouhinec et al. 2014).
Many of these interactions were found to promote Wnt and
retinoic acid signaling, confrming observations from earlier work (Bae et al. 2014; Plouhinec et al. 2014). Moreover,
some of these interactions are likely direct, as in the case
with Pax3 and Zic1 directly regulating expression of snai1/
snai2 (Bae et al. 2014; Plouhinec et al. 2014).
The proteins Snai1 and Snai2 are widely recognized for
their roles in regulating early neural crest stem cell development. They have been historically categorized as transcriptional repressors (Barrallo-Gimeno and Nieto 2005, 2009),
and this activity is key to sharpening the medial-lateral
boundaries of the early neural crest by repressing expression of pro-neural soxb1 (sox2, sox3) factors (LaBonne and
Bronner-Fraser 2000; Acloque et al. 2011). Another mechanism by which Snai1 and Snai2 factors control the spatial
boundaries and balance of neural crest proliferation is by
regulating apoptosis. In Xenopus Snai2 has been proposed
Xenopus
roles have been described for Kremen2, which mediates
its effects through canonical Wnt signaling (Hassler et al.
2007). Taken together, the combined activity of these proteins and signaling pathways establishes an embryonic territory, the neural plate border, which constitutes the earliest
recognizable domain of the neural crest in vertebrates.
The neural plate border region is characterized by the
expression of a set of transcription factors implicated in
intricate regulatory interactions that include positive and
negative feedback loops, feed-forward loops, and mutual
repression. Although there is substantial inter-species variation in spatiotemporal expression patterns and the exact
types of regulatory interactions, all vertebrates nonetheless
express a common suite of neural plate border genes. For
Xenopus, these include members of the Pax (pax3, pax7)
(Milet et al. 2013), Msx (msx1, msx2) (Khadka et al. 2003;
Monsoro-Burq et al. 2005), and Zic (zic1, zic2, zic3, zic5)
(Kuo et al. 1998; Nakata et al. 1998) families, as well as
dlx3 (Feledy et al. 1999; Luo et al. 2001), gbx2.2 (Li et al.
2009 ), irx1 (Glavic et al. 2004), snai1 (Aybar et al. 2003),
tfap2a (de Crozé et al. 2011), and hes4/hairy2 ( Vega-López
et al. 2015). Recently, there has been evidence that miRNAs
(mir-301a, mir-338) and miRNA-associated proteins are
also expressed at the neural plate border and may be important for maintaining pluripotency (Gessert et al. 2010; Ward
et al. 2018). Together, these factors progressively ref ne the
spatial boundaries of the neural crest domain to distinguish
it from adjacent tissues, activate downstream target genes
responsible for maintaining the neural crest in a stem cell
state, and eventually endow these cells with the ability to
migrate (Bae et al. 2014).
Work in Xenopus has suggested that the core regulatory
circuitry of the neural plate border includes pax3, msx1, zic1,
and tfap2a (Hong and Saint-Jeannet 2007; Plouhinec et al.
2014; Pla and Monsoro-Burq 2018). There is evidence that
these transcription factors are some of the earliest targets of
signaling through Wnt, BMP, FGF, and retinoic acid pathways
and pattern the neural plate border by mediating the effects
of these signaling pathways on downstream targets (Tríbulo
et al. 2003). Some neural plate border factors are capable of
driving expression of downstream neural crest genes in neuralized (i.e. BMP-attenuated) ectodermal explants. For example, expression of either tfap2a or msx1 in neuralized explants
is suffcient to activate expression of neural crest genes pax3,
msx1, sox9, sox10, snai2 (Monsoro-Burq et al. 2005; Sato et
al. 2005; Hong and Saint-Jeannet 2007; de Crozé et al. 2011).
Gain- and loss-of-function experiments have revealed some of
the epistatic relationships among neural plate border factors,
as well as the signaling pathways through which they mediate
regulatory control. For example, msx1 is required for activation of pax3, whereas tfap2a can coordinate with canonical
Wnt signaling to activate pax3 (de Crozé et al. 2011). pax3
then works with msx1 to promote FGF signaling by activating
early neural crest genes such as snai2. Unlike the Tfap2a and
Msx1 factors, neither Pax3 nor Zic1 alone are capable of promoting formation of neural crest cells in neuralized explants.
However their combined activity is suffcient to drive robust
expression of sox9, sox10, snai2, and meis3, followed by
tubb2b (n-tubulin) and th (tyrosine hydroxylase) during differentiation into neural and melanocyte derivatives, respectively
(Maeda et al. 2001). Combined Pax3/Zic1 activity can also
promote Cadherin switching, EMT, and production of melanocytes in ectodermal explants (Milet and Monsoro-Burq
2012) and can reprogram other tissues such as ventral ectoderm toward a neural crest progenitor state (Sato et al. 2005;
Milet et al. 2013). Importantly, however, these effects on the
neural crest are highly dose dependent. For example, although
increased levels of Pax3 can promote ectopic neural crest,
high pax3 expression converts neural crest-fated ectoderm to
an alternative fate, hatching gland (Hong and Saint-Jeannet
2007). Another important function of neural crest GRN components at these stages is to refne spatial boundaries between
neural crest and other ectodermal populations. This can occur
through transcriptional repression of sox2 and sox3 by Snai1
at the lateral edges of the neural plate (LaBonne and BronnerFraser 1998, 2000; Aybar et al. 2003; Langer et al. 2008) and
by Nbx/Nkx1–2-mediated inhibition of sox2 and otx2 ( Kurata
and Ueno 2003). Conversely, Prdm12 is expressed in the preplacodal ectoderm and can inhibit expression of neural crest
genes such as foxd3, snai2, and sox8 (Matsukawa et al. 2015).
Beyond the core neural plate border subcircuit consisting
of pax3-msx1-zic1-tfap2a, studies in Xenopus have uncovered additional factors essential for regulating the establishment of defnitive neural crest. These include Pbx1 and Meis1
(Maeda et al. 2002), Pcdh7 (Rashid et al. 2018), Klf4/Neptune
(Kurauchi et al. 2010), and Znf703, which act downstream of
Pax3 (Hong and Saint-Jeannet 2017) and activate the neural
crest genes snai2 and sox10. Similarly, Hmga2 (Macri et al.
2009, 2016) and Fbxw7/Cdc4 (Almeida et al. 2010) act
downstream of Pax3 and Msx1 and upstream of myc, snai1,
and snai2. More recently, unbiased genome-wide analyses
have described the global transcriptomic landscape at the
neural plate border. These studies have suggested that, once
activated, Pax3 creates an autoregulatory loop and, together
with Zic1, activates defnitive neural crest regulatory factors
including ednra, gbx2.2, sox8, sox9, sox10, twist1, tfap2b,
snai2, and foxd3 (Bae et al. 2014; Plouhinec et al. 2014).
Many of these interactions were found to promote Wnt and
retinoic acid signaling, confrming observations from earlier work (Bae et al. 2014; Plouhinec et al. 2014). Moreover,
some of these interactions are likely direct, as in the case
with Pax3 and Zic1 directly regulating expression of snai1/
snai2 (Bae et al. 2014; Plouhinec et al. 2014).
The proteins Snai1 and Snai2 are widely recognized for
their roles in regulating early neural crest stem cell development. They have been historically categorized as transcriptional repressors (Barrallo-Gimeno and Nieto 2005, 2009),
and this activity is key to sharpening the medial-lateral
boundaries of the early neural crest by repressing expression of pro-neural soxb1 (sox2, sox3) factors (LaBonne and
Bronner-Fraser 2000; Acloque et al. 2011). Another mechanism by which Snai1 and Snai2 factors control the spatial
boundaries and balance of neural crest proliferation is by
regulating apoptosis. In Xenopus Snai2 has been proposed