294
Xenopus
identities and stem cell proliferation genes (Edwards-Faret
et al. 2021). Even more exciting, the dependency of proper
spinal cord regeneration on the presence of NSPCs was demonstrated by ablation of NSPCs in R-stages using the nitroreductase/metronidazole system in the transgenic line (Xla.
Tg(Dre.gfap::mCherry-Nitroreductase) Larra , which led to the
loss of functional recovery, determined by the decrease in
swimming capacity (Edwards-Faret et al. 2021).
In contrast, NR-stages have reduced levels of Sox2
and a delayed and poor proliferative response of Sox2/3 +
cells, concomitant with no increase in neurogenic markers
(Muñoz et al. 2015). Interestingly, transplantation of R-stage
Sox2/3+ cells into the NR-stage injury site provided further evidence that Sox2/3 + cells are NSPCs (Méndez-Olivos
et al. 2017). Transplanted cells were able to self-organize,
proliferate, and differentiate into neurons in the host NRenvironment, with axons growing in the grafted tissue and
into the host spinal cord. Surprisingly, regeneration of axons
coming from the host was also observed, suggesting that
transplanted cells are able to provide a permissive environment for axon outgrowth (Méndez-Olivos et al. 2017). It
remains to be elucidated which intrinsic and extrinsic factors are provided by the donor cells that are responsible for
this regeneration in NR-stages.
20.3.4. AXON REGENERATION
For successful spinal cord regeneration and functional
recovery, nerve regeneration and reconstitution of lost connections is necessary. As early as the beginning of the 1960s,
X. laevis axons traversing the injury gap were seen, but their
origin was not possible to identify (Sims 1962). In the mid1980s, it was determined that some of the growing axons
were from serotonergic neurons (Beattie et al. 1990), whose
projections are formed early during X. laevis development
(Van Mier et al. 1986), indicating the axons were regenerating. Later, it was demonstrated that metamorphosis is preceded by a new wave of neurogenesis and the generation of
new neuronal connections to allow for locomotor changes
(Kollros 1981; Sillar et al. 2008), showing the possibility
that these neurons could compensate for the interrupted spinal cord tracts following SCI.
It was not until 2006 that Gibbs and Szaro were able to
demonstrate that axonal regeneration occurs and that the
axons originate in the hindbrain, mainly in the reticular
and raphe nuclei (Gibbs and Szaro 2006). Additionally, they
showed that axon regeneration capacity depends on metamorphosis progression; inhibiting metamorphosis allows
animals to regenerate axons and reestablish swimming,
while induction of metamorphosis prematurely in R-stage
animals by thyroid hormone 3,3′,5′-Triiodo-L-thyronine
(T3) treatment results in impaired axon regeneration and
promotes a differential transcriptional response in the
hindbrain compared to untreated animals. Therefore,
intrinsic transcriptional changes in the neuron soma
accompany axon R- and NR-responses in the spinal cord
(Gibbs et al. 2011). The loss of regenerative capacity after
metamorphosis correlates with the progressive changes in
the cells lining the central canal of the spinal cord (EdwardsFaret et al. 2018), concomitant with the increase in thyroid
hormone (TH) levels (Brown and Cai 2007) and the downregulation of sox2 (Gaete et al. 2012; Muñoz et al. 2015)
and lin28, which is an heterochronic gene that controls
developmental timing (Faunes et al. 2017). Interestingly,
overexpression of lin28 regulates X. laevis metamorphosis by disturbing TH function, resulting in improved digit
regeneration (Faunes et al. 2017). These observations
add to the interest of studying genetic interactions during
metamorphosis.
20.3.5. IDENTIFICATION OF BIOLOGICAL PROCESSES
INVOLVED IN SPINAL CORD REGENERATION
AND COMPARISON TO OTHER MODELS
The different regenerative abilities of X. laevis prior to
and during metamorphosis, and its genome availability
(Session et al. 2016), have made it possible to perform highthroughput experiments using RNA-seq at 1, 2, and 6 dpt,
and Proteomics at 1 dpt, to compare the response to SCI
in R- versus NR-stages (Lee-Liu et al. 2014, 2018). Even
though both stages display a massive transcriptomic change
response, they differ in terms of timing and levels of gene
expression. R-stages show a rapid response, with most of
the transcript changes observed at 1 dpt, whereas NR-stages
present a later response, with most of the changes observed
at 6 dpt (Table 20.2) (Lee-Liu et al. 2014).
Gene ontology enrichment analysis shows that most of
the genes that change their expression levels in response to
injury, which are mostly differentially regulated between
R- and NR-stages, participate in biological processes such
as metabolism, response to stress, cell cycle, development,
immune response and infammation, neurogenesis, and axonal regeneration (Table 20.2). These biological processes
are related to the SCI secondary injury cellular phases
mentioned previously, cell death and inf ammation, cell
proliferation and tissue replacement, and tissue remodeling (Burda and Sofroniew 2014). Therefore, the different
transcriptional responses of R- and NR-stages could lead
to dissimilar cellular responses and be responsible for the
regenerative capacity at the R-stages and its loss at the
NR-stage. Even though the contribution of transcript and
protein-level changes has been only partially tested in
X. laevis during SCI, their comparison with other animal
models or paradigms is important for guiding future directions in the understanding of R- and NR-stage cell response
after SCI. Here we interpret these fndings in the context
of what is known about spinal cord regeneration in other
experimental paradigms.
1. Cell death: In X. laevis tail amputation experiments, an early and controlled process of cellular
apoptosis takes place. Apoptosis is required during
the frst 24 h, as its inhibition during the f rst day
abolishes tail regeneration, whereas later inhibition
Xenopus
identities and stem cell proliferation genes (Edwards-Faret
et al. 2021). Even more exciting, the dependency of proper
spinal cord regeneration on the presence of NSPCs was demonstrated by ablation of NSPCs in R-stages using the nitroreductase/metronidazole system in the transgenic line (Xla.
Tg(Dre.gfap::mCherry-Nitroreductase) Larra , which led to the
loss of functional recovery, determined by the decrease in
swimming capacity (Edwards-Faret et al. 2021).
In contrast, NR-stages have reduced levels of Sox2
and a delayed and poor proliferative response of Sox2/3 +
cells, concomitant with no increase in neurogenic markers
(Muñoz et al. 2015). Interestingly, transplantation of R-stage
Sox2/3+ cells into the NR-stage injury site provided further evidence that Sox2/3 + cells are NSPCs (Méndez-Olivos
et al. 2017). Transplanted cells were able to self-organize,
proliferate, and differentiate into neurons in the host NRenvironment, with axons growing in the grafted tissue and
into the host spinal cord. Surprisingly, regeneration of axons
coming from the host was also observed, suggesting that
transplanted cells are able to provide a permissive environment for axon outgrowth (Méndez-Olivos et al. 2017). It
remains to be elucidated which intrinsic and extrinsic factors are provided by the donor cells that are responsible for
this regeneration in NR-stages.
20.3.4. AXON REGENERATION
For successful spinal cord regeneration and functional
recovery, nerve regeneration and reconstitution of lost connections is necessary. As early as the beginning of the 1960s,
X. laevis axons traversing the injury gap were seen, but their
origin was not possible to identify (Sims 1962). In the mid1980s, it was determined that some of the growing axons
were from serotonergic neurons (Beattie et al. 1990), whose
projections are formed early during X. laevis development
(Van Mier et al. 1986), indicating the axons were regenerating. Later, it was demonstrated that metamorphosis is preceded by a new wave of neurogenesis and the generation of
new neuronal connections to allow for locomotor changes
(Kollros 1981; Sillar et al. 2008), showing the possibility
that these neurons could compensate for the interrupted spinal cord tracts following SCI.
It was not until 2006 that Gibbs and Szaro were able to
demonstrate that axonal regeneration occurs and that the
axons originate in the hindbrain, mainly in the reticular
and raphe nuclei (Gibbs and Szaro 2006). Additionally, they
showed that axon regeneration capacity depends on metamorphosis progression; inhibiting metamorphosis allows
animals to regenerate axons and reestablish swimming,
while induction of metamorphosis prematurely in R-stage
animals by thyroid hormone 3,3′,5′-Triiodo-L-thyronine
(T3) treatment results in impaired axon regeneration and
promotes a differential transcriptional response in the
hindbrain compared to untreated animals. Therefore,
intrinsic transcriptional changes in the neuron soma
accompany axon R- and NR-responses in the spinal cord
(Gibbs et al. 2011). The loss of regenerative capacity after
metamorphosis correlates with the progressive changes in
the cells lining the central canal of the spinal cord (EdwardsFaret et al. 2018), concomitant with the increase in thyroid
hormone (TH) levels (Brown and Cai 2007) and the downregulation of sox2 (Gaete et al. 2012; Muñoz et al. 2015)
and lin28, which is an heterochronic gene that controls
developmental timing (Faunes et al. 2017). Interestingly,
overexpression of lin28 regulates X. laevis metamorphosis by disturbing TH function, resulting in improved digit
regeneration (Faunes et al. 2017). These observations
add to the interest of studying genetic interactions during
metamorphosis.
20.3.5. IDENTIFICATION OF BIOLOGICAL PROCESSES
INVOLVED IN SPINAL CORD REGENERATION
AND COMPARISON TO OTHER MODELS
The different regenerative abilities of X. laevis prior to
and during metamorphosis, and its genome availability
(Session et al. 2016), have made it possible to perform highthroughput experiments using RNA-seq at 1, 2, and 6 dpt,
and Proteomics at 1 dpt, to compare the response to SCI
in R- versus NR-stages (Lee-Liu et al. 2014, 2018). Even
though both stages display a massive transcriptomic change
response, they differ in terms of timing and levels of gene
expression. R-stages show a rapid response, with most of
the transcript changes observed at 1 dpt, whereas NR-stages
present a later response, with most of the changes observed
at 6 dpt (Table 20.2) (Lee-Liu et al. 2014).
Gene ontology enrichment analysis shows that most of
the genes that change their expression levels in response to
injury, which are mostly differentially regulated between
R- and NR-stages, participate in biological processes such
as metabolism, response to stress, cell cycle, development,
immune response and infammation, neurogenesis, and axonal regeneration (Table 20.2). These biological processes
are related to the SCI secondary injury cellular phases
mentioned previously, cell death and inf ammation, cell
proliferation and tissue replacement, and tissue remodeling (Burda and Sofroniew 2014). Therefore, the different
transcriptional responses of R- and NR-stages could lead
to dissimilar cellular responses and be responsible for the
regenerative capacity at the R-stages and its loss at the
NR-stage. Even though the contribution of transcript and
protein-level changes has been only partially tested in
X. laevis during SCI, their comparison with other animal
models or paradigms is important for guiding future directions in the understanding of R- and NR-stage cell response
after SCI. Here we interpret these fndings in the context
of what is known about spinal cord regeneration in other
experimental paradigms.
1. Cell death: In X. laevis tail amputation experiments, an early and controlled process of cellular
apoptosis takes place. Apoptosis is required during
the frst 24 h, as its inhibition during the f rst day
abolishes tail regeneration, whereas later inhibition