34
Xenopus
formation (Heim et al., 2014). Additional evidence from f sh
suggests that this self-assembling process is triggered by the
formation of the “chromosome bouquet,” a polarized cluster
of telomeres (Elkouby et al., 2016). Genetically, this event
is upstream of mitochondrial cloud formation (Elkouby
et al., 2016; Escobar-Aguirre et al., 2017). Experiments in
Xenopus show that Velo1 transitions from a mitochondrial
cloud/amyloid state to a non-amyloid hydrogel-like state
during formation of the germ plasm domain of the cloud,
possibly based on phosphorylation of Velo1 (Boke et al.,
2016; Nijjar and Woodland, 2013b). This more mobile state
may be a prerequisite for germ plasm morphogenesis during
development, in contrast to the more stable “architectural”
form present in the oocyte.
3.6. CONCLUDING REMARKS
The general picture of early Xenopus development is well
appreciated but is now being investigated with increasing
sophistication at the cellular, molecular, and genetic levels.
Thus far, a relatively small number of maternal mRNAs,
both non-localized and localized to different extents, have
been found to exert strong effects on development. But there
clearly is greater underlying complexity than is immediately
evident. These mRNAs have predominant roles in primordial germ cell specifcation, axis formation, and germ layer
induction and patterning (Figure 3.2).
A broad view of early Xenopus development suggests
that non-localized (or animally enriched) maternal RNAs
would initiate general ectodermal identity throughout the
embryo (in addition to housekeeping functions), within
which is embedded the competence to respond to various
inducers to alter germ layer identity. Maternal transcription factors Pou5f3.3 and Sox3 may function in this regard
as an adjunct to their potential role in contributing to the
MZT, but the distinction (or lack thereof) between the two
activities is unclear. Also, maternal Foxh1 may act to mark
some genes for later activation/repression by Nodal/Smad2
or other signals, but other roles in ectoderm differentiation
are unknown.
mRNAs localized widely throughout the vegetal hemisphere, such as vegt and otx1, are thought to establish
endodermal identity in vegetal cells and indirectly activate
the expression of mesendoderm-inducing Nodal proteins.
Maternally expressed signaling ligands such as Gdf1 may
“prime” this mesendoderm induction or, more likely, control the range, activity, or specifcity of Nodal/Nodal-related
proteins. Other localized mRNAs of the more “intermediate” pattern may be dedicated in some way to dorsal-ventral
patterning, with trim36 and dnd1 functioning to stimulate
microtubule polymerization in vegetal subcortical cytoplasm for cortical rotation, controlling the distribution and
activity of maternal β-Catenin, possibly through Wnt11b.
Dorsal β-Catenin would then interact with ubiquitous
maternal TCFs to de-repress/activate Nodal, initiating its
spatiotemporal regulation, and to regulate genes important
for organizer function. Last, mRNAs restricted to the germ
plasm generate proteins that generally inhibit somatic fate
and specify cellular properties, such as global transcriptional activity and cell migration, of a small subset of cells
destined to become the PGCs.
The use of the Xenopus embryo has been extraordinarily
successful in characterizing genes involved in early vertebrate development, and part of this success of Xenopus has
been the ease of “tinkering” with development—trying out
various hypotheses and rejecting erroneous ideas quickly.
It is unlikely that many of the discoveries described in this
chapter would have been made as readily without the ability
to test the functions of genes at the biochemical and cellular levels almost at a whim. There are several areas in
which studying the maternal control of early development
in Xenopus holds good potential for further uncovering new
knowledge regarding embryonic development.
First, although there are many efforts to map gene regulatory network models for different aspects of cell differentiation, in Xenopus, this analysis can be easily extended
to include the contributions of maternally provided genes.
In Xenopus, it should be possible to construct a complete
description of development, beginning with contributions of
maternally localized mRNAs. Also, although there has been
tremendous progress in modeling embryonic development
using stem cells, many of these approaches use growth factor stimulation of cells to initiate self-organizing processes.
These approaches were pioneered by the late Xenopus investigator Yoshiki Sasai, focusing on brain and retinal development (Eiraku et al., 2011, 2008). Methods have recently
been developed to approximate post-implantation-like states
in mouse embryos and mouse and human pluripotent cell
aggregates ( Zhu and Zernicka-Goetz, 2020). These experiments are possible because of the robust ability of differentiating cells to undergo self-organization, a well-noted
but not well-understood phenomenon. Developmental biologists familiar with the work of Spemann and the Holtfreters
should not be surprised that self-organization seems to be
the norm. Because studies in early Xenopus embryos do
not require artif cial stimulation or culture conditions, Wnt/
β-Catenin activation and other signaling mechanisms can
thus be investigated under endogenous activation conditions.
Second, the realization that many types of ribonucleoprotein granules are formed through liquid-like phase transitions
to a hydrogel state has allowed the formulation of a “solidstate” model of information transfer via spatially distinct iterations of these states along the nucleocytoplasmic transport
route (Kato and McKnight, 2016). These studies are still in the
early stages, and we still do not know the full extent to which
mRNA localization depends on hydrogel states (Neil et al.,
2021). This intriguing possibility has recently been suggested
by experiments showing that intermediate f lament head
domains also form hydrogels that can coaggregate with lowcomplexity domain-containing RNA binding proteins (Zhou
et al., 2021). Such an observation could also help explain the
structural roles of vegt and other localized mRNAs in addition to germ plasm morphogenesis. These questions could all
be proftably addressed using Xenopus oocytes.
Xenopus
formation (Heim et al., 2014). Additional evidence from f sh
suggests that this self-assembling process is triggered by the
formation of the “chromosome bouquet,” a polarized cluster
of telomeres (Elkouby et al., 2016). Genetically, this event
is upstream of mitochondrial cloud formation (Elkouby
et al., 2016; Escobar-Aguirre et al., 2017). Experiments in
Xenopus show that Velo1 transitions from a mitochondrial
cloud/amyloid state to a non-amyloid hydrogel-like state
during formation of the germ plasm domain of the cloud,
possibly based on phosphorylation of Velo1 (Boke et al.,
2016; Nijjar and Woodland, 2013b). This more mobile state
may be a prerequisite for germ plasm morphogenesis during
development, in contrast to the more stable “architectural”
form present in the oocyte.
3.6. CONCLUDING REMARKS
The general picture of early Xenopus development is well
appreciated but is now being investigated with increasing
sophistication at the cellular, molecular, and genetic levels.
Thus far, a relatively small number of maternal mRNAs,
both non-localized and localized to different extents, have
been found to exert strong effects on development. But there
clearly is greater underlying complexity than is immediately
evident. These mRNAs have predominant roles in primordial germ cell specifcation, axis formation, and germ layer
induction and patterning (Figure 3.2).
A broad view of early Xenopus development suggests
that non-localized (or animally enriched) maternal RNAs
would initiate general ectodermal identity throughout the
embryo (in addition to housekeeping functions), within
which is embedded the competence to respond to various
inducers to alter germ layer identity. Maternal transcription factors Pou5f3.3 and Sox3 may function in this regard
as an adjunct to their potential role in contributing to the
MZT, but the distinction (or lack thereof) between the two
activities is unclear. Also, maternal Foxh1 may act to mark
some genes for later activation/repression by Nodal/Smad2
or other signals, but other roles in ectoderm differentiation
are unknown.
mRNAs localized widely throughout the vegetal hemisphere, such as vegt and otx1, are thought to establish
endodermal identity in vegetal cells and indirectly activate
the expression of mesendoderm-inducing Nodal proteins.
Maternally expressed signaling ligands such as Gdf1 may
“prime” this mesendoderm induction or, more likely, control the range, activity, or specifcity of Nodal/Nodal-related
proteins. Other localized mRNAs of the more “intermediate” pattern may be dedicated in some way to dorsal-ventral
patterning, with trim36 and dnd1 functioning to stimulate
microtubule polymerization in vegetal subcortical cytoplasm for cortical rotation, controlling the distribution and
activity of maternal β-Catenin, possibly through Wnt11b.
Dorsal β-Catenin would then interact with ubiquitous
maternal TCFs to de-repress/activate Nodal, initiating its
spatiotemporal regulation, and to regulate genes important
for organizer function. Last, mRNAs restricted to the germ
plasm generate proteins that generally inhibit somatic fate
and specify cellular properties, such as global transcriptional activity and cell migration, of a small subset of cells
destined to become the PGCs.
The use of the Xenopus embryo has been extraordinarily
successful in characterizing genes involved in early vertebrate development, and part of this success of Xenopus has
been the ease of “tinkering” with development—trying out
various hypotheses and rejecting erroneous ideas quickly.
It is unlikely that many of the discoveries described in this
chapter would have been made as readily without the ability
to test the functions of genes at the biochemical and cellular levels almost at a whim. There are several areas in
which studying the maternal control of early development
in Xenopus holds good potential for further uncovering new
knowledge regarding embryonic development.
First, although there are many efforts to map gene regulatory network models for different aspects of cell differentiation, in Xenopus, this analysis can be easily extended
to include the contributions of maternally provided genes.
In Xenopus, it should be possible to construct a complete
description of development, beginning with contributions of
maternally localized mRNAs. Also, although there has been
tremendous progress in modeling embryonic development
using stem cells, many of these approaches use growth factor stimulation of cells to initiate self-organizing processes.
These approaches were pioneered by the late Xenopus investigator Yoshiki Sasai, focusing on brain and retinal development (Eiraku et al., 2011, 2008). Methods have recently
been developed to approximate post-implantation-like states
in mouse embryos and mouse and human pluripotent cell
aggregates ( Zhu and Zernicka-Goetz, 2020). These experiments are possible because of the robust ability of differentiating cells to undergo self-organization, a well-noted
but not well-understood phenomenon. Developmental biologists familiar with the work of Spemann and the Holtfreters
should not be surprised that self-organization seems to be
the norm. Because studies in early Xenopus embryos do
not require artif cial stimulation or culture conditions, Wnt/
β-Catenin activation and other signaling mechanisms can
thus be investigated under endogenous activation conditions.
Second, the realization that many types of ribonucleoprotein granules are formed through liquid-like phase transitions
to a hydrogel state has allowed the formulation of a “solidstate” model of information transfer via spatially distinct iterations of these states along the nucleocytoplasmic transport
route (Kato and McKnight, 2016). These studies are still in the
early stages, and we still do not know the full extent to which
mRNA localization depends on hydrogel states (Neil et al.,
2021). This intriguing possibility has recently been suggested
by experiments showing that intermediate f lament head
domains also form hydrogels that can coaggregate with lowcomplexity domain-containing RNA binding proteins (Zhou
et al., 2021). Such an observation could also help explain the
structural roles of vegt and other localized mRNAs in addition to germ plasm morphogenesis. These questions could all
be proftably addressed using Xenopus oocytes.