95
Notch Signaling in Early Embryogenesis
Constitutively active Notch1 constructs and time-controlled
Notch1/RBPJ activation or blockade resulted in a variety of
changes in markers of ectodermal-, neural-, mesodermal-,
and endodermal-derived cell types at neurula and tailbud stages (Tables 7.5, 7.9), indicating that the response to
Notch1/RBPJ signaling changes over time. We perturbed
the Notch pathway in several ways to address a possible role
in controlling the boundaries between germ layers, analyzing the consequences during gastrulation when they segregate, and also in early neurulation (Revinski et al., 2010)
(Table 7.5). Activation and blockade delayed gastrulation,
indicating that Notch1 activity is tightly balanced to keep
morphogenetic movements at a normal pace. Germ layers
were specifed, but they did not develop properly because
their MZ boundaries were shifted. Consequently, cells at the
boundaries allocated incorrectly and changed their specif -
cation to the incorrect germ layer. In NICD1 mRNA-injected
embryos, the presumptive neural ectoderm and supra-blastoporal endoderm were expanded at the expense of mesoderm,
whereas notch1 knock-down produced the opposite changes
( Revinski et al., 2010). Both dll1 STU ( Revinski et al., 2010)
and dll1 knock-down (Kinoshita et al., 2011) inhibited the
pan-mesodermal marker tbxt in its normal circumblastoporal domain. In embryos injected with dll1 STU, this was
accompanied by the animal displacement of the tbxt domain
as expanded supra-blastoporal endoderm (sox17 -positive)
took its place. The neural ectoderm (sox2 expressing) also
was reduced in response to this animal-ward expansion of
the presumptive mesoderm (Revinski et al., 2010).
We proposed that notch1 is involved in the segregation
between neural ectoderm, mesoderm, and endoderm by controlling their boundaries in the MZ (Revinski et al., 2010)
(Figure 7.3B,C ) in the following ways. (1) Refning the limit
of involution between the IMZ and the NIMZ favored neural
ectoderm at the expense of mesoderm (type A decisions).
(b) In the IMZ, by refning mesoderm segregation from the
supra-blastoporal endoderm, favoring endoderm over mesoderm (type B decisions). Strikingly, dll1 STU shifted the
limit of involution animal-wards, favoring endomesodermal
development over neural ectoderm, but notch1 knock-down
expanded the mesoderm at the expense of both endoderm
and neural ectoderm; perturbing Notch/RBPJ signaling during gastrulation had a similar effect (Contakos et al., 2005).
This indicates that Dll1 is involved in type A but not in type
B decisions. According to this model, pre-involuted IMZ
cells present Dll1 to the neighboring cells on the other side
of the limit of involution, thus preventing them from adopting the same fate (endomesoderm) by triggering the Notch
pathway, which instead promotes neural ectoderm specif -
cation (Revinski et al., 2010). Interestingly, in animal caps
assays, Notch1ΔE alone weakly induced neural ectoderm
but strongly enhanced ectodermal competence for neural
induction (Coffman et al., 1993). Therefore, Dll1 signaling
from pre-involuted IMZ might enhance the competence of
their neighbors above the limit of involution to respond to
neural inducers and become neural instead of mesoderm,
sharpening the boundary between both populations. Once
the mesodermal cells involute, they no longer express Dll1,
ending this activity. While it appears that Dll1 controls the
limit of involution (type A decisions), it remains unknown
which notch1-dependent mechanisms underlie mesodermal
versus endodermal (type B) decisions. More work is needed
to discern the possible role of the diverse Dll/Jag ligands and
non-canonical Notch pathways in the segregation of germ
layers.
7.3.2.2. Which Notch Targets Are Involved
in Germ Layer Segregation?
The IMZ expresses several hes genes during gastrulation
(Figure 7.2). Most of their patterns are similar to that of
dll1, but the dorsal hes5.1 and hes5.3 boundaries are more
distant from the organizer, and hes7.2 is more abundant in
the organizer. Only hes4 is broadly expressed in the NIMZ,
but hes5.10 (at early gastrula) and hes2 (at mid-gastrula) are
expressed in scattered cells (Figure 7.2). Except for hes5.8
and hes5.9, whose regulation by Notch signaling has not
been studied, all the hes5 genes expressed in the MZ, as well
as hes2, hes4, and hes7.2, are positively regulated by Notch
in several contexts, although a few of them were tested for
Notch responsiveness in the MZ. In contrast, hes6.1 and
hes7.3/esr5 are down-regulated in the neural plate and the
IMZ, respectively (Tables 7.2, 7.3). Interestingly, hes5.10
is later expressed throughout the non-neural ectoderm and
responds positively to NICD1, although it is not clear whether
RBPJ is involved (Deblandre et al., 1999). The early expression of hes6.1, hes7.2, and hes7.3/esr5 in the IMZ might be
related to their role during somitogenesis (see Section 3.6).
Only a few of the hes genes expressed in the MZ have been
experimentally tested for their role in the MZ (hes4, hes5.1,
hes5.6, hes6.1, hes7.3/esr5) (Table 7.6 ). One clear candidate
for positioning the limit of involution is hes4, which f rst
is broadly expressed in the presumptive ectoderm of the
blastula, then progressively confned to the boundary with
the mesoderm during gastrulation, accumulating transcripts
in the whole NIMZ with highest levels dorsally in a pattern complementary to tbxt (pan-mesoderm) (López et al.,
2005; Aguirre et al., 2013). hes4 might be one of the Notch
targets involved in type A decisions because: (1) the hes4
NIMZ domain was expanded by NICD1 (López et al., 2005)
and down-regulated by blockade of Notch1/RBPJ signaling (unpublished results); (2) hes4 overexpression blocked
gastrulation movements, impeding MZ cell involution; (3)
hes4 overexpression repressed tbxt throughout the entire
IMZ (López et al., 2005; Cui, 2005; Aguirre et al., 2013);
and (4) hes4 knock-down expanded the tbxt domain toward
the animal pole, indicating that it is required for the correct
placement of the ectoderm-mesoderm boundary (Aguirre et
al., 2013 ) ( Figure 7.3B , C ) ( Tables 7.5 , 7.6 ).
Overexpression and dominant-negative experiments indicate that hes7.3/esr5 promotes and hes5.1 inhibits mesoderm specifcation and they repress each other (Kinoshita
et al., 2011) (Tables 7.4, 7.6 ). In animal cap explants, NICD1
induced hes5.1 but not hes7.3/esr5 (Kinoshita et al., 2011)
(Tables 7.2, 7.3), Nodal2 induced hes7.3/esr5 but not hes5.1,
Notch Signaling in Early Embryogenesis
Constitutively active Notch1 constructs and time-controlled
Notch1/RBPJ activation or blockade resulted in a variety of
changes in markers of ectodermal-, neural-, mesodermal-,
and endodermal-derived cell types at neurula and tailbud stages (Tables 7.5, 7.9), indicating that the response to
Notch1/RBPJ signaling changes over time. We perturbed
the Notch pathway in several ways to address a possible role
in controlling the boundaries between germ layers, analyzing the consequences during gastrulation when they segregate, and also in early neurulation (Revinski et al., 2010)
(Table 7.5). Activation and blockade delayed gastrulation,
indicating that Notch1 activity is tightly balanced to keep
morphogenetic movements at a normal pace. Germ layers
were specifed, but they did not develop properly because
their MZ boundaries were shifted. Consequently, cells at the
boundaries allocated incorrectly and changed their specif -
cation to the incorrect germ layer. In NICD1 mRNA-injected
embryos, the presumptive neural ectoderm and supra-blastoporal endoderm were expanded at the expense of mesoderm,
whereas notch1 knock-down produced the opposite changes
( Revinski et al., 2010). Both dll1 STU ( Revinski et al., 2010)
and dll1 knock-down (Kinoshita et al., 2011) inhibited the
pan-mesodermal marker tbxt in its normal circumblastoporal domain. In embryos injected with dll1 STU, this was
accompanied by the animal displacement of the tbxt domain
as expanded supra-blastoporal endoderm (sox17 -positive)
took its place. The neural ectoderm (sox2 expressing) also
was reduced in response to this animal-ward expansion of
the presumptive mesoderm (Revinski et al., 2010).
We proposed that notch1 is involved in the segregation
between neural ectoderm, mesoderm, and endoderm by controlling their boundaries in the MZ (Revinski et al., 2010)
(Figure 7.3B,C ) in the following ways. (1) Refning the limit
of involution between the IMZ and the NIMZ favored neural
ectoderm at the expense of mesoderm (type A decisions).
(b) In the IMZ, by refning mesoderm segregation from the
supra-blastoporal endoderm, favoring endoderm over mesoderm (type B decisions). Strikingly, dll1 STU shifted the
limit of involution animal-wards, favoring endomesodermal
development over neural ectoderm, but notch1 knock-down
expanded the mesoderm at the expense of both endoderm
and neural ectoderm; perturbing Notch/RBPJ signaling during gastrulation had a similar effect (Contakos et al., 2005).
This indicates that Dll1 is involved in type A but not in type
B decisions. According to this model, pre-involuted IMZ
cells present Dll1 to the neighboring cells on the other side
of the limit of involution, thus preventing them from adopting the same fate (endomesoderm) by triggering the Notch
pathway, which instead promotes neural ectoderm specif -
cation (Revinski et al., 2010). Interestingly, in animal caps
assays, Notch1ΔE alone weakly induced neural ectoderm
but strongly enhanced ectodermal competence for neural
induction (Coffman et al., 1993). Therefore, Dll1 signaling
from pre-involuted IMZ might enhance the competence of
their neighbors above the limit of involution to respond to
neural inducers and become neural instead of mesoderm,
sharpening the boundary between both populations. Once
the mesodermal cells involute, they no longer express Dll1,
ending this activity. While it appears that Dll1 controls the
limit of involution (type A decisions), it remains unknown
which notch1-dependent mechanisms underlie mesodermal
versus endodermal (type B) decisions. More work is needed
to discern the possible role of the diverse Dll/Jag ligands and
non-canonical Notch pathways in the segregation of germ
layers.
7.3.2.2. Which Notch Targets Are Involved
in Germ Layer Segregation?
The IMZ expresses several hes genes during gastrulation
(Figure 7.2). Most of their patterns are similar to that of
dll1, but the dorsal hes5.1 and hes5.3 boundaries are more
distant from the organizer, and hes7.2 is more abundant in
the organizer. Only hes4 is broadly expressed in the NIMZ,
but hes5.10 (at early gastrula) and hes2 (at mid-gastrula) are
expressed in scattered cells (Figure 7.2). Except for hes5.8
and hes5.9, whose regulation by Notch signaling has not
been studied, all the hes5 genes expressed in the MZ, as well
as hes2, hes4, and hes7.2, are positively regulated by Notch
in several contexts, although a few of them were tested for
Notch responsiveness in the MZ. In contrast, hes6.1 and
hes7.3/esr5 are down-regulated in the neural plate and the
IMZ, respectively (Tables 7.2, 7.3). Interestingly, hes5.10
is later expressed throughout the non-neural ectoderm and
responds positively to NICD1, although it is not clear whether
RBPJ is involved (Deblandre et al., 1999). The early expression of hes6.1, hes7.2, and hes7.3/esr5 in the IMZ might be
related to their role during somitogenesis (see Section 3.6).
Only a few of the hes genes expressed in the MZ have been
experimentally tested for their role in the MZ (hes4, hes5.1,
hes5.6, hes6.1, hes7.3/esr5) (Table 7.6 ). One clear candidate
for positioning the limit of involution is hes4, which f rst
is broadly expressed in the presumptive ectoderm of the
blastula, then progressively confned to the boundary with
the mesoderm during gastrulation, accumulating transcripts
in the whole NIMZ with highest levels dorsally in a pattern complementary to tbxt (pan-mesoderm) (López et al.,
2005; Aguirre et al., 2013). hes4 might be one of the Notch
targets involved in type A decisions because: (1) the hes4
NIMZ domain was expanded by NICD1 (López et al., 2005)
and down-regulated by blockade of Notch1/RBPJ signaling (unpublished results); (2) hes4 overexpression blocked
gastrulation movements, impeding MZ cell involution; (3)
hes4 overexpression repressed tbxt throughout the entire
IMZ (López et al., 2005; Cui, 2005; Aguirre et al., 2013);
and (4) hes4 knock-down expanded the tbxt domain toward
the animal pole, indicating that it is required for the correct
placement of the ectoderm-mesoderm boundary (Aguirre et
al., 2013 ) ( Figure 7.3B , C ) ( Tables 7.5 , 7.6 ).
Overexpression and dominant-negative experiments indicate that hes7.3/esr5 promotes and hes5.1 inhibits mesoderm specifcation and they repress each other (Kinoshita
et al., 2011) (Tables 7.4, 7.6 ). In animal cap explants, NICD1
induced hes5.1 but not hes7.3/esr5 (Kinoshita et al., 2011)
(Tables 7.2, 7.3), Nodal2 induced hes7.3/esr5 but not hes5.1,