133
Evolution of the Vertebrate Neural Crest
LaBonne 2018 ). These fndings indicate that neural crest cells
and pluripotent blastula cells are similar at the gene regulatory level and provide novel insights into the broad multi-germ
layer developmental potential that characterizes the neural
crest. Indeed, work in Xenopus has demonstrated that neural
crest cells can not only contribute to ectoderm and mesoderm
but also endoderm ( Buitrago-Delgado et al. 2015 ). The latter provides new context to the contributions of neural crest
cells to endoderm-derived organs such as the adrenal gland,
thyroid, and thymus ( Pearse and Polak 1971 ; Le Douarin and
Teillet 1974 ; Polak et al. 1974 ; Le Douarin and Jotereau 1975 ;
Bockman and Kirby 1984 ; Le Douarin and Kalcheim 1999 ).
The similarities between neural crest and blastula stem
cells extend beyond shared transcription factor circuitry. FGF/
MAPK signaling is required for establishment of neural crest
development and maintenance of pluripotency and proper
lineage restriction of blastula stem cells. Both processes are
accompanied by a decrease in MAPK signaling and increase
in PI3K/Akt signaling (Geary and LaBonne 2018). This can be
replicated in vitro as Pax3/Zic1-mediated reprogramming of
animal cap cells to a neural crest state results in robust MAPK
signaling with low levels of PI3K (Geary and LaBonne 2018).
Similar results are obtained with transcription factors.
Both Snai1 and Sox5 regulate pluripotency factors and lineage
restriction in the blastula and the formation of neural crest
cells both in vitro and in vivo. Sox5 has been shown to partner
with BMP R-Smads to regulate expression of target genes in
the blastula (ventx2.2, id3), neural crest (msx1), and epidermis
(krt12.4/epk) (Nordin and LaBonne 2014; Buitrago-Delgado
et al. 2015). Similarly, inhibition of Snai1 function revealed
that it regulates expression of most of the core pluripotency
network (sox2/3, oct25/60, ventx2.2, tfap2a, id3) in the blastula in vivo and is essential to direct proper lineage restriction
of pluripotent blastula cells toward an endomesodermal progenitor state (Buitrago-Delgado et al. 2015). Shared features
between neural crest and blastula stem cells are also found
at the epigenetic level. For example, Hdac1 activity is crucial
for expression of the sox-oct-myc-vent pluripotency axis and
proper lineage restriction of all three germ layers, as well as
for establishment of the neural crest (Rao and LaBonne 2018).
In addition, low levels of H3K9 and H3K27 acetylation are
characteristic of both pluripotent blastula stem cells and neural crest cells (Rao and LaBonne 2018). Importantly, increased
hdac1 activity was found to enhance reprogramming to a
neural crest state, which has implications for regenerative
medicine (Rao and LaBonne 2018).
While there are many similarities between neural crest
and blastula stem cells, there are also key differences. One
of these involves a change in the deployment of Sox transcription factors. SoxB1 (Sox2, Sox3), but not SoxE factors
(Sox8, Sox9, Sox10), are active in pluripotent blastula cells
(Buitrago-Delgado et al. 2018). Later, as expression of soxB1
genes becomes restricted to the neural plate, soxE genes are
turned on in neural crest cells (Buitrago-Delgado et al. 2018).
This suggests a functional “hand-off” from SoxB1 to SoxE
transcription factor activity during the emergence of neural
crest progenitors from pluripotent blastula cells. Consistent
with this idea, there are distinct functional requirements
for these gene families in each population. Ectopic expression of sox9 or sox10 in the blastula disrupts pluripotency,
whereas forced expression of soxB1 genes results in loss of
neural crest. By contrast, increased expression of soxE genes
promote excess neural crest. Finally, there is evidence that
this Sox factor “hand-off” is essential for the emergence of
neural crest cells because SoxB1 factors cannot replace the
activity of SoxE factors to rescue neural crest development
(Buitrago-Delgado, Schock et al. 2018).
The extensive shared features between neural crest
and blastula stem cells suggest a new model for neural
crest origins in which neural crest cells do not “re-gain”
developmental potential in response to inductive cues but
rather retain aspects of the pluripotent regulatory state of the
blastula stem cells from which they are derived. This new
model helps explain why neural crest cells exhibit greater
developmental potential than the ectoderm in which they
originate. In this model, the neural crest should be thought
of not as a population of ectoderm-derived cells but rather
as a population derived from pluripotent blastula cells that
retain their position at or near the top of Waddington’s landscape ( Figure 8.2B). Retention of this blastula-stage potential into later stages of development allows neural crest cells
to add novelty to the basic chordate body plan by contributing cell types and features associated with all three germ
layers. Interestingly, Raven recognized this capacity from
his amphibian transplantation experiments over 80 years
ago (Figure 8.2C ), when he presciently observed that:
The Neural Crest Material in the early stages was omnipotent [e.g. pluripotent] as to the “faculty of differentiation”
but had no “tendencies of differentiation” of its own. At
the transition from the stage of a yolk plug of intermediate
size to one with a small yolk plug there is a change in the
potency . . . the faculty of differentiation is restricted . . .
as regards to tissues that characterize other germ layers
(endomesoderm, notochord, intestine)
( Raven 1935 ; Horstadius 1950 ).
8.5. FUTURE DIRECTIONS AND OUTLOOK
This chapter provides an overview of the many advances in
our understanding of vertebrate neural crest cells derived
from studies in amphibians, in particular Xenopus. This
historical overview reveals just how much our knowledge of neural crest development owes to the over 150
years of research using this key model organism. That
said, much remains to be learned about how the neural
crest develops and evolves, and many outstanding questions remain. Many of these questions center on control
of neural crest fate and potential, some of the very same
issues faced by early experimentalists such as Raven and
Hörstadius. How, for example, do neural crest cells ultimately decide what fate to adopt and when to do so? Are
these choices binary and sequential, or are multiple paths
open to all or some cells? What are the gene regulatory
Evolution of the Vertebrate Neural Crest
LaBonne 2018 ). These fndings indicate that neural crest cells
and pluripotent blastula cells are similar at the gene regulatory level and provide novel insights into the broad multi-germ
layer developmental potential that characterizes the neural
crest. Indeed, work in Xenopus has demonstrated that neural
crest cells can not only contribute to ectoderm and mesoderm
but also endoderm ( Buitrago-Delgado et al. 2015 ). The latter provides new context to the contributions of neural crest
cells to endoderm-derived organs such as the adrenal gland,
thyroid, and thymus ( Pearse and Polak 1971 ; Le Douarin and
Teillet 1974 ; Polak et al. 1974 ; Le Douarin and Jotereau 1975 ;
Bockman and Kirby 1984 ; Le Douarin and Kalcheim 1999 ).
The similarities between neural crest and blastula stem
cells extend beyond shared transcription factor circuitry. FGF/
MAPK signaling is required for establishment of neural crest
development and maintenance of pluripotency and proper
lineage restriction of blastula stem cells. Both processes are
accompanied by a decrease in MAPK signaling and increase
in PI3K/Akt signaling (Geary and LaBonne 2018). This can be
replicated in vitro as Pax3/Zic1-mediated reprogramming of
animal cap cells to a neural crest state results in robust MAPK
signaling with low levels of PI3K (Geary and LaBonne 2018).
Similar results are obtained with transcription factors.
Both Snai1 and Sox5 regulate pluripotency factors and lineage
restriction in the blastula and the formation of neural crest
cells both in vitro and in vivo. Sox5 has been shown to partner
with BMP R-Smads to regulate expression of target genes in
the blastula (ventx2.2, id3), neural crest (msx1), and epidermis
(krt12.4/epk) (Nordin and LaBonne 2014; Buitrago-Delgado
et al. 2015). Similarly, inhibition of Snai1 function revealed
that it regulates expression of most of the core pluripotency
network (sox2/3, oct25/60, ventx2.2, tfap2a, id3) in the blastula in vivo and is essential to direct proper lineage restriction
of pluripotent blastula cells toward an endomesodermal progenitor state (Buitrago-Delgado et al. 2015). Shared features
between neural crest and blastula stem cells are also found
at the epigenetic level. For example, Hdac1 activity is crucial
for expression of the sox-oct-myc-vent pluripotency axis and
proper lineage restriction of all three germ layers, as well as
for establishment of the neural crest (Rao and LaBonne 2018).
In addition, low levels of H3K9 and H3K27 acetylation are
characteristic of both pluripotent blastula stem cells and neural crest cells (Rao and LaBonne 2018). Importantly, increased
hdac1 activity was found to enhance reprogramming to a
neural crest state, which has implications for regenerative
medicine (Rao and LaBonne 2018).
While there are many similarities between neural crest
and blastula stem cells, there are also key differences. One
of these involves a change in the deployment of Sox transcription factors. SoxB1 (Sox2, Sox3), but not SoxE factors
(Sox8, Sox9, Sox10), are active in pluripotent blastula cells
(Buitrago-Delgado et al. 2018). Later, as expression of soxB1
genes becomes restricted to the neural plate, soxE genes are
turned on in neural crest cells (Buitrago-Delgado et al. 2018).
This suggests a functional “hand-off” from SoxB1 to SoxE
transcription factor activity during the emergence of neural
crest progenitors from pluripotent blastula cells. Consistent
with this idea, there are distinct functional requirements
for these gene families in each population. Ectopic expression of sox9 or sox10 in the blastula disrupts pluripotency,
whereas forced expression of soxB1 genes results in loss of
neural crest. By contrast, increased expression of soxE genes
promote excess neural crest. Finally, there is evidence that
this Sox factor “hand-off” is essential for the emergence of
neural crest cells because SoxB1 factors cannot replace the
activity of SoxE factors to rescue neural crest development
(Buitrago-Delgado, Schock et al. 2018).
The extensive shared features between neural crest
and blastula stem cells suggest a new model for neural
crest origins in which neural crest cells do not “re-gain”
developmental potential in response to inductive cues but
rather retain aspects of the pluripotent regulatory state of the
blastula stem cells from which they are derived. This new
model helps explain why neural crest cells exhibit greater
developmental potential than the ectoderm in which they
originate. In this model, the neural crest should be thought
of not as a population of ectoderm-derived cells but rather
as a population derived from pluripotent blastula cells that
retain their position at or near the top of Waddington’s landscape ( Figure 8.2B). Retention of this blastula-stage potential into later stages of development allows neural crest cells
to add novelty to the basic chordate body plan by contributing cell types and features associated with all three germ
layers. Interestingly, Raven recognized this capacity from
his amphibian transplantation experiments over 80 years
ago (Figure 8.2C ), when he presciently observed that:
The Neural Crest Material in the early stages was omnipotent [e.g. pluripotent] as to the “faculty of differentiation”
but had no “tendencies of differentiation” of its own. At
the transition from the stage of a yolk plug of intermediate
size to one with a small yolk plug there is a change in the
potency . . . the faculty of differentiation is restricted . . .
as regards to tissues that characterize other germ layers
(endomesoderm, notochord, intestine)
( Raven 1935 ; Horstadius 1950 ).
8.5. FUTURE DIRECTIONS AND OUTLOOK
This chapter provides an overview of the many advances in
our understanding of vertebrate neural crest cells derived
from studies in amphibians, in particular Xenopus. This
historical overview reveals just how much our knowledge of neural crest development owes to the over 150
years of research using this key model organism. That
said, much remains to be learned about how the neural
crest develops and evolves, and many outstanding questions remain. Many of these questions center on control
of neural crest fate and potential, some of the very same
issues faced by early experimentalists such as Raven and
Hörstadius. How, for example, do neural crest cells ultimately decide what fate to adopt and when to do so? Are
these choices binary and sequential, or are multiple paths
open to all or some cells? What are the gene regulatory