303
Tumor Formation and Regulation in Xenopus
1997; Yang et al., 1998). Relatively recently, the technique for
inducing tumors in embryos was used to study novel mechanisms in tumor formation. By investigating ITLSs, it was
found that transmembrane voltage potential is a crucial cellular parameter for tumor detection and control, highlighting
opportunities for novel therapeutic intervention (Chernet and
Levin, 2013). The same research group also identif ed the
importance of long-range gap junctional signaling in regulating tumorigenesis (Chernet, Fields and Levin, 2015).
21.2.4. TUMOR–IMMUNE SYSTEM INTERACTIONS
Xenopus appears to be well positioned for investigating the
role of the immune system against cancer. The majority of
the mechanisms and cell types governing both the innate and
the adaptive immune system are highly conserved between
Xenopus and humans (Robert and Ohta, 2009), and experiments in Xenopus laevis have exposed crucial roles for CD8+,
NK, and NKT-like cells in anti-tumor defenses (Goyos
and Robert, 2009; Banach and Robert, 2017; Banach et al.,
2019). In addition, an elegant strategy for modeling tumors in
Xenopus is used based on the subcutaneous transplantation
of a lymphoid tumor cell/collagen mixture matrix in isogenic
tadpoles, allowing the study of the tumor microenvironment,
neovascularization, and immune cell interactions (HaynesGimore et al., 2015). Interestingly, in light of the recent promising developments in the use of immune checkpoint inhibition
for treating a range of aggressive cancers in humans (Taube
et al., 2014), the Xenopus genome contains the orthologues for
both the human receptors PDCD1 (PD1) and CTLA4 as well
as their respective ligands CD274 (PDL1) and CD86.
21.2.5. GENETICALLY ENGINEERED XENOPUS MODELS
With the invention of nuclease-based genome editing systems
like ZFNs, TALENs, and CRISPR/Cas9, cancer research
experienced an unprecedented push forward. Next to mice and
zebrafsh, numerous genetically engineered organisms have
emerged for cancer modeling, ranging from yeast (reviewed by
Guaragnella et al., 2014), Drosophila (reviewed by Mirzoyan
et al., 2019), and swine (reviewed by Watson et al., 2016) to
Xenopus tropicalis (Van Nieuwenhuysen et al., 2015; Naert
et al., 2016; Naert, Dimitrakopoulou, et al., 2020). Especially
favored by its diploid genome, genetically engineered Xenopus
tropicalis might become extremely useful in deciphering factors that drive cancer or expose genetic cancer vulnerabilities.
21.3. CHALLENGES TO OVERCOME
As with other animal models, GEXMs can experience some
criticism from the scientifc and medical communities.
Therefore, in this section, we will highlight some concerns
that might arise and how to answer potential reviewers’ critique regarding the use of GEXM in cancer research.
An obvious criticism that may arise is that of the greater
evolutionary distance that exists between amphibia and mammals. Without any doubt, mice models currently better ref ect
the patient situation. However, Xenopus —like zebraf sh—can
perfectly serve as complementary model, since generating
disease models (even in F0) is extremely straightforward, and
fndings can be eventually extrapolated to mammalian organisms such as mice. Additionally, it is worth mentioning that in
contrast to its teleost counterpart, Xenopus does have limbs,
lungs, and urinary bladder, allowing the modeling of diseases
and or cancer in these organ systems. An additional major
shortcoming of current GEXMs is that the genetic insults are
not introduced in a tissue-, organ-, or cell-type-specif c manner. While inducible Cre-Lox technology would be applicable
in Xenopus, there is limited reason or beneft to introduce this
technology for cancer modeling in Xenopus, given the extensive genetic tools currently available in the mouse. However,
the ease with which targeted injections can be performed
in the external developing Xenopus embryos, especially with
the existence of a detailed fate map (Dale and Slack, 1987;
Moody, 1987), makes tissue-restricted genome engineering
possible (Van Nieuwenhuysen et al., 2015). This in turn also
lowers the danger of abrogating embryonic processes leading
to lethality. Of note, we want to point out that even for eff cient
guide RNAs injected in one- or two-cell-stage embryos, the
signifcant occurrence of small functional in-frame mutations,
as well as the possibility to carefully dose the concentration
of the genome-editing reagents, might be suffcient to retain
normal functionality in order to avoid perturbing embryonic
development (Naert, Tulkens, et al., 2020).
Another major shortcoming is the fact that cancer formation in GEXMs generated via multiplexing injections results
from the simultaneous introduction of mutations in genes,
while in normal oncogenesis, this is a sequential process,
with intermediate selection of certain cell sub-populations.
A fnal drawback that one might notice is the limitation of
only being able to generate loss-of-function mutations, as
knock-in engineering remains challenging (Aslan et al.,
2017). For that, we refer to recently published technologies
that may be used in the future to induce expression of oncogenes (described in section 6.3).
Of note, an issue that might arise when using CRISPRor TALEN-mediated genome editing is the phenomenon of
off-target effects. However, when targeting cancer-related
genes, the genetic insult will be under positive selection.
Therefore, if a developing tumor does not show the intended
biallelic inactivation of the TSG or monoallelic inactivation combined with loss of heterozygosity (LOH), tumor
induction may have been the result of an off-target effect.
Similarly, when introducing activating mutations in a protooncogene, this should be apparent in at least one of the two
targeted alleles of the sampled tumor.
21.4. GENETICALLY ENGINEERED XENOPUS
MODELS FOR CANCER RESEARCH
21.4.1. GENERATION OF CLINICALLY RELEVANT GEXMS
From genome-wide studies, it is becoming clear that specifc genetic alterations found in human cancers are often
Tumor Formation and Regulation in Xenopus
1997; Yang et al., 1998). Relatively recently, the technique for
inducing tumors in embryos was used to study novel mechanisms in tumor formation. By investigating ITLSs, it was
found that transmembrane voltage potential is a crucial cellular parameter for tumor detection and control, highlighting
opportunities for novel therapeutic intervention (Chernet and
Levin, 2013). The same research group also identif ed the
importance of long-range gap junctional signaling in regulating tumorigenesis (Chernet, Fields and Levin, 2015).
21.2.4. TUMOR–IMMUNE SYSTEM INTERACTIONS
Xenopus appears to be well positioned for investigating the
role of the immune system against cancer. The majority of
the mechanisms and cell types governing both the innate and
the adaptive immune system are highly conserved between
Xenopus and humans (Robert and Ohta, 2009), and experiments in Xenopus laevis have exposed crucial roles for CD8+,
NK, and NKT-like cells in anti-tumor defenses (Goyos
and Robert, 2009; Banach and Robert, 2017; Banach et al.,
2019). In addition, an elegant strategy for modeling tumors in
Xenopus is used based on the subcutaneous transplantation
of a lymphoid tumor cell/collagen mixture matrix in isogenic
tadpoles, allowing the study of the tumor microenvironment,
neovascularization, and immune cell interactions (HaynesGimore et al., 2015). Interestingly, in light of the recent promising developments in the use of immune checkpoint inhibition
for treating a range of aggressive cancers in humans (Taube
et al., 2014), the Xenopus genome contains the orthologues for
both the human receptors PDCD1 (PD1) and CTLA4 as well
as their respective ligands CD274 (PDL1) and CD86.
21.2.5. GENETICALLY ENGINEERED XENOPUS MODELS
With the invention of nuclease-based genome editing systems
like ZFNs, TALENs, and CRISPR/Cas9, cancer research
experienced an unprecedented push forward. Next to mice and
zebrafsh, numerous genetically engineered organisms have
emerged for cancer modeling, ranging from yeast (reviewed by
Guaragnella et al., 2014), Drosophila (reviewed by Mirzoyan
et al., 2019), and swine (reviewed by Watson et al., 2016) to
Xenopus tropicalis (Van Nieuwenhuysen et al., 2015; Naert
et al., 2016; Naert, Dimitrakopoulou, et al., 2020). Especially
favored by its diploid genome, genetically engineered Xenopus
tropicalis might become extremely useful in deciphering factors that drive cancer or expose genetic cancer vulnerabilities.
21.3. CHALLENGES TO OVERCOME
As with other animal models, GEXMs can experience some
criticism from the scientifc and medical communities.
Therefore, in this section, we will highlight some concerns
that might arise and how to answer potential reviewers’ critique regarding the use of GEXM in cancer research.
An obvious criticism that may arise is that of the greater
evolutionary distance that exists between amphibia and mammals. Without any doubt, mice models currently better ref ect
the patient situation. However, Xenopus —like zebraf sh—can
perfectly serve as complementary model, since generating
disease models (even in F0) is extremely straightforward, and
fndings can be eventually extrapolated to mammalian organisms such as mice. Additionally, it is worth mentioning that in
contrast to its teleost counterpart, Xenopus does have limbs,
lungs, and urinary bladder, allowing the modeling of diseases
and or cancer in these organ systems. An additional major
shortcoming of current GEXMs is that the genetic insults are
not introduced in a tissue-, organ-, or cell-type-specif c manner. While inducible Cre-Lox technology would be applicable
in Xenopus, there is limited reason or beneft to introduce this
technology for cancer modeling in Xenopus, given the extensive genetic tools currently available in the mouse. However,
the ease with which targeted injections can be performed
in the external developing Xenopus embryos, especially with
the existence of a detailed fate map (Dale and Slack, 1987;
Moody, 1987), makes tissue-restricted genome engineering
possible (Van Nieuwenhuysen et al., 2015). This in turn also
lowers the danger of abrogating embryonic processes leading
to lethality. Of note, we want to point out that even for eff cient
guide RNAs injected in one- or two-cell-stage embryos, the
signifcant occurrence of small functional in-frame mutations,
as well as the possibility to carefully dose the concentration
of the genome-editing reagents, might be suffcient to retain
normal functionality in order to avoid perturbing embryonic
development (Naert, Tulkens, et al., 2020).
Another major shortcoming is the fact that cancer formation in GEXMs generated via multiplexing injections results
from the simultaneous introduction of mutations in genes,
while in normal oncogenesis, this is a sequential process,
with intermediate selection of certain cell sub-populations.
A fnal drawback that one might notice is the limitation of
only being able to generate loss-of-function mutations, as
knock-in engineering remains challenging (Aslan et al.,
2017). For that, we refer to recently published technologies
that may be used in the future to induce expression of oncogenes (described in section 6.3).
Of note, an issue that might arise when using CRISPRor TALEN-mediated genome editing is the phenomenon of
off-target effects. However, when targeting cancer-related
genes, the genetic insult will be under positive selection.
Therefore, if a developing tumor does not show the intended
biallelic inactivation of the TSG or monoallelic inactivation combined with loss of heterozygosity (LOH), tumor
induction may have been the result of an off-target effect.
Similarly, when introducing activating mutations in a protooncogene, this should be apparent in at least one of the two
targeted alleles of the sampled tumor.
21.4. GENETICALLY ENGINEERED XENOPUS
MODELS FOR CANCER RESEARCH
21.4.1. GENERATION OF CLINICALLY RELEVANT GEXMS
From genome-wide studies, it is becoming clear that specifc genetic alterations found in human cancers are often