235
Discovering the Function of CHD Genes
of earlier developmental pathways, such as LR patterning, in
addition to cardiogenesis.
16.2.1. PREVIOUS STUDIES OF VERTEBRATE
CARDIOGENESIS IN XENOPUS
Disruptions at various stages in early embryonic development can cause cardiac structural anomalies. Much work
has gone into characterizing cardiac tissue specif cation
and morphogenesis across many organisms. Several studies on cardiac morphogenesis performed in Xenopus models
have connected CHD patient variants directly with cardiac
development.
Vertebrate cardiac development follows a conserved
developmental pathway. Briefy, the mesoderm is specif ed
at gastrulation (Sater and Jacobson 1989). As the embryo
elongates, mesodermal cardiac precursors migrate to the
midline and specify into the cardiac progenitors of the f rst
and second heart felds. The frst heart feld eventually forms
the two atria and the trabeculated ventricle. The second heart
f eld becomes the outf ow tract (Buckingham, Meilhac, and
Zaffran 2005; Gessert and Kühl 2009).
Several known cardiogenesis genes originally discovered through genetic screens in model organisms and CHD
patient genetic analyses have been studied in Xenopus.
nkx2–5 and its co-factor gata4 have been associated with
atrio-septal defects (ASDs) (Cleaver, Patterson, and Krieg
1996; Durocher, Schwartz, and Nemer 1997; Schott et al.
1998; Benson et al. 1999; Bartlett et al. 2007). msh-2 (nkx2–
5) was originally identifed in a screen for novel homeobox
genes in Drosophila melanogaster (Bodmer, Jan, and Jan
1990). The gene was renamed tinman (tin) in Drosophila
due to tin mutants lacking cardiac primordia (Bodmer
1993 ). Xenopus nkx-2.5, like tinman, is expressed in heart
and gut tissues (Tonissen et al. 1994). The nkx2–5 co-factor
gata4 was originally identifed in a cDNA screen for novel
GATA family members in Xenopus expressed in the developing heart (Kelley et al. 1993). X. laevis studies showed that
both nkx2–5 and gata4 overexpression resulted in embryos
with enlarged hearts. Authors proposed that nkx2–5 and
gata4 cause this myocardial hyperplasia through pathways
involved in myocardial progenitor proliferation or recruitment (Cleaver, Patterson, and Krieg 1996 ).
In contrast, the T-box family gene tbx5 was f rst associated with cardiac development via patient studies. Tbx5 was
initially identifed in mice and was found to be expressed in
early cardiac and forelimb tissues, but researchers focused
on its role in limb formation (Gibson-Brown et al. 1996;
Chapman et al. 1996 ). Patients with Holt-Oram syndrome
(HOS) have defects in limb formation and cardiac septation.
A study looking for the HOS gene mapped TBX5 within a
deleted region of chromosome 12q2, the locus responsible for
HOS (Basson et al. 1997; Mcdermott et al. 2005). Studies
in Xenopus showed that both overexpression and knockdown
of tbx5 was associated with heart tube formation abnormalities and heart-looping anomalies, loss of cardiac mass, and
edema (Brown et al. 2005; Horb and Thomsen 1999).
Another T-box gene, tbx20, was identifed in a screen for
novel family members involved in FGF patterning of the
mesoderm in developing zebrafsh (Kevin J.P. Griffn et al.
1998) and was found to be expressed in the same tissues as
nkx2–5, gata4, and tbx5 ( Griffn et al. 2000; Brown et al.
2005). Families with TBX20 mutations present with cardiomyopathy, ASD, and mitral valve malformations (Kirk et al.
2007 ). hrT (tbx20) knockdown in zebrafsh causes abnormal
cardiac chamber formation (Szeto, Griffn, and Kimelman
2002 ). tbx20 knockdown in Xenopus is associated with heart
looping abnormalities, edema, and cardiac chamber differentiation, though specifcation of cardiac tissue is normal
(Brown et al. 2005).
Similar to tbx5, ets1 was not directly associated with
cardiac structuring until patients with a CHD-related syndrome underwent genetic analysis. ETS1 was initially identifed as a potential oncogene (Leprince et al. 1983). It was
subsequently associated with cranial neural crest migration,
vasculogenesis, immune cell differentiation, and endothelial
differentiation (Meyer et al. 1997; Tahtakran and Selleck
2003; Sumanas and Lin 2006; Barton et al. 1998; Wang
et al. 2005). Later studies mapped ETS1 within the region
of chromosome 11 that is affected in patients with Jacobsen
syndrome, a syndrome that includes CHD (Penny et al.
1995; Grossfeld et al. 2004; Ye et al. 2009). When investigated under in the context of Jacobsen syndrome, researchers found that ets1 is necessary for cardiac mesoderm and
neural crest specifcation (Ye et al. 2009; Nie and Bronner
2015). Studies in Xenopus found ets1 to be expressed in
several tissues, including neural crest and the developing
heart (Meyer et al. 1997 ). Depleting ets1 in the mesoderm,
cardiac mesoderm, or cardiac neural crest leads to loss of
endocardial specif cation, poor cardiac morphogenesis (single-chamber heart, no trabeculation, and no aortic septation)
due to delayed heart tube formation, and small malformed
outfow tracts, respectively (Nie and Bronner 2015). These
CHD genes and the roles they play in cardiac morphogenesis
account for only a small portion of genes and mechanisms
identifed in CHD patients. More recently identif ed candidate genes may have unknown functions or unpredictable
roles in development prior to or in conjunction with cardiac
morphogenesis.
16.2.2. IDENTIFYING ADDITIONAL CANDIDATE CHD GENES
Analysis of large CHD/HTX patient cohorts using genotyping microarrays and whole exome sequencing (WES) has
identifed many new candidate genes. We can eff ciently
identify copy-number variants (CNVs) with genotyping
microarrays (Alkan, Coe, and Eichler 2011). CNVs are a
type of genetic perturbation that involves duplications or
deletions that can span multiple genes. In such cases, it can
be diffcult to pinpoint disease causality to one or more
genes.
An early study of CNVs from HTX patients discovered
36 relatively small CNVs affecting 61 genes. Morpholino
knockdown of a subset of these genes in Xenopus tropicalis
Discovering the Function of CHD Genes
of earlier developmental pathways, such as LR patterning, in
addition to cardiogenesis.
16.2.1. PREVIOUS STUDIES OF VERTEBRATE
CARDIOGENESIS IN XENOPUS
Disruptions at various stages in early embryonic development can cause cardiac structural anomalies. Much work
has gone into characterizing cardiac tissue specif cation
and morphogenesis across many organisms. Several studies on cardiac morphogenesis performed in Xenopus models
have connected CHD patient variants directly with cardiac
development.
Vertebrate cardiac development follows a conserved
developmental pathway. Briefy, the mesoderm is specif ed
at gastrulation (Sater and Jacobson 1989). As the embryo
elongates, mesodermal cardiac precursors migrate to the
midline and specify into the cardiac progenitors of the f rst
and second heart felds. The frst heart feld eventually forms
the two atria and the trabeculated ventricle. The second heart
f eld becomes the outf ow tract (Buckingham, Meilhac, and
Zaffran 2005; Gessert and Kühl 2009).
Several known cardiogenesis genes originally discovered through genetic screens in model organisms and CHD
patient genetic analyses have been studied in Xenopus.
nkx2–5 and its co-factor gata4 have been associated with
atrio-septal defects (ASDs) (Cleaver, Patterson, and Krieg
1996; Durocher, Schwartz, and Nemer 1997; Schott et al.
1998; Benson et al. 1999; Bartlett et al. 2007). msh-2 (nkx2–
5) was originally identifed in a screen for novel homeobox
genes in Drosophila melanogaster (Bodmer, Jan, and Jan
1990). The gene was renamed tinman (tin) in Drosophila
due to tin mutants lacking cardiac primordia (Bodmer
1993 ). Xenopus nkx-2.5, like tinman, is expressed in heart
and gut tissues (Tonissen et al. 1994). The nkx2–5 co-factor
gata4 was originally identifed in a cDNA screen for novel
GATA family members in Xenopus expressed in the developing heart (Kelley et al. 1993). X. laevis studies showed that
both nkx2–5 and gata4 overexpression resulted in embryos
with enlarged hearts. Authors proposed that nkx2–5 and
gata4 cause this myocardial hyperplasia through pathways
involved in myocardial progenitor proliferation or recruitment (Cleaver, Patterson, and Krieg 1996 ).
In contrast, the T-box family gene tbx5 was f rst associated with cardiac development via patient studies. Tbx5 was
initially identifed in mice and was found to be expressed in
early cardiac and forelimb tissues, but researchers focused
on its role in limb formation (Gibson-Brown et al. 1996;
Chapman et al. 1996 ). Patients with Holt-Oram syndrome
(HOS) have defects in limb formation and cardiac septation.
A study looking for the HOS gene mapped TBX5 within a
deleted region of chromosome 12q2, the locus responsible for
HOS (Basson et al. 1997; Mcdermott et al. 2005). Studies
in Xenopus showed that both overexpression and knockdown
of tbx5 was associated with heart tube formation abnormalities and heart-looping anomalies, loss of cardiac mass, and
edema (Brown et al. 2005; Horb and Thomsen 1999).
Another T-box gene, tbx20, was identifed in a screen for
novel family members involved in FGF patterning of the
mesoderm in developing zebrafsh (Kevin J.P. Griffn et al.
1998) and was found to be expressed in the same tissues as
nkx2–5, gata4, and tbx5 ( Griffn et al. 2000; Brown et al.
2005). Families with TBX20 mutations present with cardiomyopathy, ASD, and mitral valve malformations (Kirk et al.
2007 ). hrT (tbx20) knockdown in zebrafsh causes abnormal
cardiac chamber formation (Szeto, Griffn, and Kimelman
2002 ). tbx20 knockdown in Xenopus is associated with heart
looping abnormalities, edema, and cardiac chamber differentiation, though specifcation of cardiac tissue is normal
(Brown et al. 2005).
Similar to tbx5, ets1 was not directly associated with
cardiac structuring until patients with a CHD-related syndrome underwent genetic analysis. ETS1 was initially identifed as a potential oncogene (Leprince et al. 1983). It was
subsequently associated with cranial neural crest migration,
vasculogenesis, immune cell differentiation, and endothelial
differentiation (Meyer et al. 1997; Tahtakran and Selleck
2003; Sumanas and Lin 2006; Barton et al. 1998; Wang
et al. 2005). Later studies mapped ETS1 within the region
of chromosome 11 that is affected in patients with Jacobsen
syndrome, a syndrome that includes CHD (Penny et al.
1995; Grossfeld et al. 2004; Ye et al. 2009). When investigated under in the context of Jacobsen syndrome, researchers found that ets1 is necessary for cardiac mesoderm and
neural crest specifcation (Ye et al. 2009; Nie and Bronner
2015). Studies in Xenopus found ets1 to be expressed in
several tissues, including neural crest and the developing
heart (Meyer et al. 1997 ). Depleting ets1 in the mesoderm,
cardiac mesoderm, or cardiac neural crest leads to loss of
endocardial specif cation, poor cardiac morphogenesis (single-chamber heart, no trabeculation, and no aortic septation)
due to delayed heart tube formation, and small malformed
outfow tracts, respectively (Nie and Bronner 2015). These
CHD genes and the roles they play in cardiac morphogenesis
account for only a small portion of genes and mechanisms
identifed in CHD patients. More recently identif ed candidate genes may have unknown functions or unpredictable
roles in development prior to or in conjunction with cardiac
morphogenesis.
16.2.2. IDENTIFYING ADDITIONAL CANDIDATE CHD GENES
Analysis of large CHD/HTX patient cohorts using genotyping microarrays and whole exome sequencing (WES) has
identifed many new candidate genes. We can eff ciently
identify copy-number variants (CNVs) with genotyping
microarrays (Alkan, Coe, and Eichler 2011). CNVs are a
type of genetic perturbation that involves duplications or
deletions that can span multiple genes. In such cases, it can
be diffcult to pinpoint disease causality to one or more
genes.
An early study of CNVs from HTX patients discovered
36 relatively small CNVs affecting 61 genes. Morpholino
knockdown of a subset of these genes in Xenopus tropicalis