Received:
February 24, 2022.
Revised:
July 22, 2022.
Accepted:
July 26, 2022
© The Author(s) 2022. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com
Human Molecular Genetics
, 2022, Vol. 31, 24, 4217–4227
https://doi.org/10.1093/hmg/ddac174
Advance access publication date 28 July 2022
Original Article
ETS1 loss in mice impairs cardiac outflow tract
septation via a cell migration defect autonomous to the
neural crest
Lizhu Lin
1
,
†
,
Antonella Pinto
2
,
†
,
Lu Wang
1
,
†
,
Kazumi Fukatsu
1
,
Yan Yin
1
,
Simon D. Bamforth
3
,
Marianne E. Bronner
4
,
Sylvia M. Evans
5
,
Shuyi Nie
6
,
Robert H. Anderson
3
,
†
,
Alexey V. Terskikh
2
,
†
and
Paul D. Grossfeld
1
,
7
,
†
*
,
1
Department of Pediatrics, UCSD School of Medicine, La Jolla, CA 92093, USA
2
Department of Biology, Sanford-Burnham-Prebys Institute of Medical Discovery, La Jolla, CA 92037, USA
3
Cardiovascular Research Centre, Institute of Genetic Medicine, Newcastle University, Newcastle upon Tyne NE1 3BZ, UK
4
Department of Biology, California Institute of Technology, Pasadena, CA 91125, USA
5
Department of Pharmacology, Skaggs School of Pharmacy and Pharmaceutical Sciences, UCSD, La Jolla, CA 92093, USA
6
Department of Biology, Georgia Institute of Technology, Atlanta, GA 30332, USA
7
Division of Cardiology, Rady Children’s Hospital, San Diego, CA 92123, USA
*
To whom correspondence should be addressed at: 3020 Children’s Way, MC 5004, San Diego, CA 92123, USA. Email: pgrossfeld@health.ucsd.edu
†
These authors contributed equally.
Abstract
Ets1
deletion in some mouse strains causes septal defects and has been implicated in human congenital heart defects in Jacobsen
syndrome, in which one copy of the
Ets1
gene is missing
.
Here, we demonstrate that loss of
Ets1
in mice results in a decrease in neural
crest (NC) cells migrating into the proximal outflow tract cushions during early heart development, with subsequent malalignment of
the cushions relative to the muscular ventricular septum,resembling double outlet right ventricle (DORV) defects in humans.Consistent
with this,we find that cultured cardiac NC cells from
Ets1
mutant mice or derived from iPS cells from Jacobsen patients exhibit decreased
migration speed and impaired cell-to-cell interactions. Together, our studies demonstrate a critical role for ETS1 for cell migration in
cardiac NC cells that are required for proper formation of the proximal outflow tracts. These data provide further insights into the
molecular and cellular basis for development of the outflow tracts, and how perturbation of NC cells can lead to DORV.
Introduction
Proper heart development requires precise specification and
migration of cardiac neural crest (NC) cells (
1
–
5
)
.
Cardiac NC
cells are essential for the remodeling of the extra-pericardial
arterial channels and septation of the outflow tract (OFT). This
remodeling not only separates the pulmonary and systemic
circulations, but also underlies the development of the arterial
valves (
6
,
7
). During early embryonic development, cardiac crest
cells arise from the lateral folds of the caudal hindbrain, located
between the otic placodes and the third somite. Beginning on
E8.5 in the mouse, NC cells undergo an epithelial–mesenchymal
transition, to emigrate from the neural tube and migrate through
the posterior pharyngeal arches to the developing heart. In animal
models, impairment of cardiac crest function causes congenital
heart defects (CHDs) (
5
,
7
).
ETS1, a member of the ETS family of transcription factors (
8
),
has been implicated as a causal factor in CHDs. As case in point,
multiple lines of evidence suggest that ETS1 is the cause of CHDs
in Jacobsen syndrome: (1) all patients with Jacobsen syndrome
and CHDs have a deletion that spans the
Ets1
gene (
9
); (2) the
cardiac ‘critical region’,as defined by interstitial deletions in distal
11q is
∼
1 megabase and contains only seven annotated genes
including
Ets1
(
10
); (3) studies in
Ciona intestinalis
and
Drosophila
suggest an important role for ETS1 in heart development, specifi-
cally in determining cardiac cell fate and migration (
11
,
12
) and (4)
at least one patient has been identified with double outlet right
ventricle (DORV) and hypoplasia of the left ventricle with a
de novo
loss-of-function frameshift mutation in the
Ets1
gene (
13
).
Previous studies have reported that deletion of
Ets1
causes sep-
tal defects in mice in a C57BL/6 background (
10
,
14
,
15
). The mech-
anism(s) underlying how deletion of
Ets1
causes CHDs remains
unknown. In this study, we trace the migration pattern of murine
cardiac NC cells
in vivo
during early stages of heart development
not previously studied.
Ets1
is highly expressed in pre-migratory
and migrating cardiac crest cells and its deletion causes migration
defects, leading to impaired OFT remodeling. Not only does loss
of ETS1 results in a decrease in the number of migrating cardiac
NC cells populating the proximal OFT cushions, but also causes a
failure to build the muscular shelf in the roof of the developing
right ventricle required for normal septation of the proximal
OFTs, resulting in a DORV phenotype not previously described.
We demonstrate this is due to impaired muscularization of the
cushions, leaving a fibrous outlet septum and continuity between
the aortic and pulmonary valvar leaflets, with a ventricular septal
defect in the subaortic position. We also observe a bisinuate aortic
valve in three out of four
Ets1
knockout (KO) mutant embryos,
4218
|
Human Molecular Genetics
, 2022, Vol. 31, No. 24
indicative of a defect in the development of the intercalated valvar
swellings.
In vitro
studies using cultured cardiac crest cells demon-
strate that loss of ETS1 causes decreased migration velocity and
impaired cell-to-cell interactions.
In vivo
studies demonstrate
increased N-cadherin expression in ETS1-deficient neural crest
cells, a likely direct downstream target of ETS1, and suggestive
that N-cadherin may be a genetic modifier of the ETS1 gene affect-
ing the development of CHDs. We also show for the first time that
conditional deletion of
Ets1
in the NC recapitulates the global
Ets1
mutant cardiac phenotype, consistent with a NC-autonomous
mechanism underlying the cardiac phenotype. Finally, we demon-
strate that NC cells derived from iPS cells lacking one copy of the
Ets1
gene from patients with Jacobsen syndrome and congenital
heart disease display NC cell migration defects, consistent with
our murine studies, also providing a likely molecular basis for the
incomplete penetrance of the cardiac phenotype in patients with
Jacobsen syndrome. Together, the results demonstrate a critical
role for ETS1 in cardiac NC cells for normal heart development,
extending the results of previous studies and highlighting the fact
that a subset of cardiac NC cells is required for formation of the
proximal outflow cushions and subsequent OFT septation.
Results
Global deletion of
Ets1
in a pure C57BL/6J
background causes DORV
Ets1
is highly expressed in cardiac NC cells during early
murine development (
10
). Moreover, deletion of
Ets1
in C57BL/6J
mice causes septal defects (
10
,
14
). To better define the three-
dimensional anatomy of the heart defects, we performed cardiac
MRIs on E15.5 mouse embryos lacking ETS1 compared with
controlhearts.Asshownin
Figure 1
, there is exclusive origin
of the aorta from the right ventricle, giving the ventriculo-
arterial connection of DORV (
Fig. 1A and B
). Of the four E15.5
Ets1
KO hearts analysis by MRIs, three of them exhibited DORV
phenotypes and one exhibited an overriding aorta phenotype.
In addition, serial section analysis through the OFTs in E14.5
embryonic hearts demonstrated all four
Ets1
KO E14.5 hearts
have DORV. Three out of four
Ets1
KO embryos (
Fig. 1D and F
)
analyzed have a bisinuate and bi-leaflet aortic valves, whereas
a trileaflet aortic valve was observed in all six control hearts
(
Fig. 1C and E
). Combined with the observed reduction in NC cells
at the aortic valve, our results indicate that ETS1 is required in
cardiac crest cells for proper septation of the OFT. In addition,
heterozygous ETS1 KO mice demonstrate a patch of white fur
on their abdomen, consistent with a NC cell defect as described
previously (
14
).
Loss of ETS1 causes NC migration defects in
Ets1
−
/
−
mutant mice
We first examined the role of ETS1 in cardiac NC migration.
To track the cardiac crest, we used
Pax3Cre;tdTomato
reporter
mouse embryos (
16
). Pax3Cre is an established reporter that is
expressed in the dorsal neural tube and migrating NCCs reflecting
the endogenous expression of the NC transcription factor Pax3
(
17
). The fidelity of the reporter was confirmed by immunohis-
tochemistry analysis against ETS1. We observed co-expression of
ETS1 in
Pax3Cre–Tdtomato
expressing cardiac NC cells migrating
from the neural tube toward the developing heart (
Supplementary
Material, Fig. S1
). Next, we compared cardiac crest cells from
control and
Ets1
KO embryos at pre-migratory, migratory and post-
migratory stages (E8.5–E11.5) using the
Pax3Cre;tdTomato
mouse
embryos. Three embryos from each genotype and developmental
stage were analyzed. The results demonstrate that prior to neural
tube closure at E8.5, a similar number of
tdTomato
-expressing pre-
migratory cardiac NC cells are present in the neural folds of both
Ets1
KO and control embryos (
Fig. 2A and B
). At E9.5, when the
neural tube closes and cardiac NC cells have started migration,
tdTomato
-expressing cells are observed in the developing OFT of
control embryos (
Fig. 2C
). Despite the migration of cardiac crest
cells from the dorsal neural tube in
Ets1
KO embryos, there
were fewer
tdTomato
cNCCs in the developing cardiac OFT in
Ets1
KO mutant (
Fig. 2D
) compared to control embryos, suggesting a
delay of NC migration into the heart. At E10.5, in the control
(
Fig. 2E
), the
tdTomato
cNCCs are present in the developing OFT
cushions (
Fig. 2E
,arrows),whereasinthe
Ets1
KO mutant, there
is a nearly complete absence of t
dTomato
cells in the proximal
component of the OFT cushions (
Fig. 2F
). The lack of
tdTomato
cNCCs in the proximal outflow cushion persists at E11.5, although
less severe compared to earlier stages, consistent with delayed
migration. In the control, the
Pax3Cre–tdTomato
expressing cells
are present throughout the proximal–distal axis of outflow cush-
ions (
Fig. 2G
), whereas in the
Ets1
KO mutant, cardiac crest cells
are found predominantly in the distal outflow cushions, with
fewer
Pax3Cre–tdTomato
-expressing cells in the proximal cushions
(
Fig. 2H
). Quantitation of
tdTomato
cells in the OFT for the sec-
tions shown in
Figure 2C–H
from E9.5 through E11.5 is shown in
Figure 2I
.
These results were further confirmed by immunohistochem-
istry analysis against another NC transcription factor, SOX10.
SOX10 is critical for specification of NC cells and is a direct
downstream target of ETS1 in chick (
18
). In control embryos, most
of the migratory
Pax3Cre–tdTomato
expressing cells co-expressed
SOX10 protein (
Fig. 3A
and A
). They are observed streaming in
a linear fashion from the neural tube toward the heart. In
Ets
1
KO mutant embryos, there was a loss of the linear migratory
pattern (
Fig. 3B and B
’). Most of the
tdTomato
cells lacked SOX10
expression, suggesting that ETS1 plays a conserved regulatory role
in the transcription of SOX10 in mice, and that loss of SOX10
expression contributes to the migration defects of
Ets1
KO NCCs.
Loss of ETS1 causes decreased migration velocity
and abnormal cell adhesion in cultured cardiac
NC cells
To fully characterize their migration properties, we performed
time-lapse imaging of cultured cardiac crest cells as described
previously (
19
) using a
Pax3Cre;tdTomato
reporter strain from
E8.5
Ets1
KO and control embryos. To analyze cell migration,
we explanted the neural tube from the somite 1–3 region of
E8.5 mouse embryos, which was then cultured for 24 h. We
then followed the migration of individual cardiac crest cells by
imaging at 10 min intervals for 3 h (see Materials and Methods
for details). The speed of migration in the
Ets
1KOmutantswas
significantly decreased compared to controls (0.19
μ
m/min vs
0.8
μ
m/min vs
P
<
0.0001,
Fig. 4
). Since the loss of ETS1 results in
a disruption of the linear stream pattern of migration, we also
assessed cell-to-cell interactions during their migration (
Table 1
).
In control cells, 8.5% of cells make contact and a large percentage
of them (68.8%) separate within 2 h. A slightly higher percentage
of cardiac crest cells from
Ets
1 KO mutant embryo made cell–
cell contacts (12.7%), and more strikingly, 81% of those contacts
failed to separate within 2 h. Taken together, these results suggest
that cardiac crest from
Ets
1 KO mutants have increased cell–cell
adhesion compared to controls. N-cadherin plays a crucial role in
NC migration (
20
,
21
) and previous studies have demonstrated that
N-cadherin over-expression in the chick blocked their migration
Human Molecular Genetics
, 2022, Vol. 31, No. 24
|
4219
Figure 1.
Ets1
KO mutant hearts exhibit DORV and a bisinuate (two leaflet) aortic valve. (
A
,
B
) Cardiac MRI studies demonstrate the
Ets1
KO mutant
heart DORV phenotype. Control (A), Ets1 mutant (B). LCC, left common carotid; LSA, left subclavian artery; PA, pulmonary artery; RCC, right common
carotid; RSA, right subclavian artery and Tr, trachea. Scale bars: 500
μ
m. (
C
–
F
) Serial section analysis through the OFTs in E14.5 hearts revealed a two
leaflet (bisinuate) aortic valve in three out of four KO embryos analyzed (asterisks). Control hearts (C, E) and
Ets1
mutant (D, F). NC cells are stained with
LacZ in C and D (in blue), and tdTomato in E and F (in magenta). Scale bars: 200
μ
m.
Table 1.
Quantitative assessment of cell-to-cell interaction in cultured cardiac NC cells
Control
ETS-1
KO
Total number of cells
189
166
Number of cells that make contact
16
21
Number of cells that make contact and then separate (over 2 h)
11 (68.8%)
4 (19.0%)
Number of cells that make contact but do not separate (over 2 h)
5 (31.2%)
17 (81.0%)
(
20
,
21
). We hypothesized that the cell migration defect observed
in vivo
may result from impaired expression of N-cadherin. To test
this, we assessed N-cadherin expression by immunofluorescence
on sections from E8.5 and E9.5 embryos, corresponding to the
early migration stage. The results show that N-cadherin protein
levels were markedly increased in the migratory NCCs in
Ets
1KO
mutants compared to controls (
Fig. 5A–D
), indicating a potential
molecular basis for the migration defect caused by loss of ETS1.
Furthermore, bioinformatics analysis of the promoter region of
the mouse N-cadherin gene identified three likely ETS binding
sites, suggesting that N-cadherin is a direct downstream target of
ETS1 (
Fig. 5E
).
Loss of ETS1 impairs myocardialization in the
developing OFT
Myocardialization is the process by which cardiac myocytes from
the OFT myocardium invade the OFT cushions in order to com-
plete the process of septation, and for the subsequent formation
of the subpulmonary muscular infundibulum. To investigate the
role of ETS1 in myocardialization, we performed immunofluo-
rescent analysis using an antibody against alpha-actinin, as a
marker for cardiac myocytes. In E13.5 control embryos, cardiac
myocytes in the myocardium of the OFT are seen invading the OFT
cushions (
Fig. 6A
and A
). They are in close proximity to
Pax3Cre–
tdTomato
expressing cNCCs. In
Ets
1KOmutants,therewasagap
between cardiac myocytes and
Pax3Cre–tdTomato
expressing car-
diac crest cells, indicative of impaired myocardialization at E13.5
(
Fig. 6B and B
’). By E15.5 the muscularization of OFT cushions
is complete in the control embryos, with the newly produced
myocardium forming the free-standing subpulmonary infundibu-
lar sleeve. By this stage, the
Pax3Cre–tdTomato
are almost com-
pletely surrounded by cardiomyocytes (
Fig. 6C and C
), persisting
as the fibrous tendon of the infundibulum. In the
Ets
1KO mutant,
in contrast, there were far fewer cardiomyocytes in contact with
Pax3Cre–tdTomato
cell, with the cardiac NC cells persisting as a
fibrous outlet septum in the setting of DORV (
Fig. 6D and D
).
By E18.5, while there was an infundibular sleeve consisting of
cardiac myocytes in the control (
Fig. 7A
and A
), the sleeve was
severely reduced in
Ets
1 KO mutant. Instead of the tendon of the
infundibulum, the fibrous outlet septum now produced continu-
ity between leaflets of the aortic and pulmonary valves, which
were in a side-by-side orientation (
Fig. 7B and B
). Interestingly,
this structure appears to be the same as the cartilaginous struc-
ture described previously by Gao
et al.
(
14
).
NC-specific conditional deletion of
Ets1
recapitulates the global KO cardiac phenotype
Previous studies have implicated a NC cell-autonomous mech-
anism by which loss of ETS1 causes septal defects. To exam-
ine the role of ETS1 specifically in the NC, we generated an
Ets1
NC-specific mutant mouse line in which
Ets1
is deleted via
Cre-LoxP-mediated recombination (
Supplementary Material, Fig.
S2
). The original
Ets1
floxed mice we obtained were in a mixed
genetic background. We bred the mice with C57BL/6J mice for
4220
|
Human Molecular Genetics
, 2022, Vol. 31, No. 24
Figure 2.
Cardiac NC cells in sections from embryos with
Pax3Cre;tdTomato
reporter. At E8.5, the
tdTomato
labeled NCCs are present in the neural fold in
both controls and
Ets1
KO mutants (
A
,
B
, arrows). At E9.5, the
tdTomato-labeled
cNCCs are present in the developing cardiac OFT in controls (
C
,arrows).
There are fewer
tdTomato
cNCCs in the developing cardiac OFT in the
Ets1
KO mutant (
D
). At E10.5, the
tdTomato
cNCCs have populated the developing
OFT cushions in controls (
E
, arrows), but there are fewer
tdTomato-
labeled cNCCs in the outflow cushions in the
Ets1
KO mutants (
F
, arrows). At E11.5,
there are abundant
tdTomato
cNCCs in the proximal outflow cushions in controls (
G
, arrows), but in the
Ets1
KO mutants, the
tdTomato
cNCCs are observed
only in the distal and intermediate outflow cushions (
H
, arrow), and not in the proximal cushion. Scale bars: 100
μ
m(
I
), quantitation of
tdTomato
cNCCs
in the OFT shown in the sections. The number of
tdTomato
cNCCs in the OFT in controls and
Ets1
KO mutants was counted, demonstrating decreased
tdTomato
cNCCs in the developing cardiac OFT in the
Ets1
KO mutant from E9.5 through E11.5, compared to controls.
∗∗
P
<
0.01;
∗∗∗
P
<
0.001.
12 more generations and then used a
Pax3Cre
driver to delete
Ets1
specifically in NC cells (
17
). Analysis of
Pax3Cre;Ets1
con-
ditional knockout (cKO) embryos revealed DORV (
Fig. 8B and D
)
with incomplete penetrance. Of the eight
Pax3Cre;Ets1
conditional
deletion embryos analyzed, four of them exhibited DORV (
Sup-
plementary Material, Table S1
). To determine if the 50% pene-
trance could be due to inefficient expression of the Cre leading to
incomplete deletion of the gene, we assessed ETS1 protein expres-
sion by immunohistochemistry analysis on sections from E10.5
Pax3Cre;Ets1
conditional deletion mutants and
Pax3Cre
controls.
ETS1 is co-expressed with SOX10 in control embryos (
Supplemen-
tary Material, Fig. S3A
and A
). In contrast, there was a nearly
complete absence of ETS1 protein signal in SOX10 expressing
cardiac crest cells in
Pax3Cre;Ets1
conditional deletion mutants
(
Supplementary Material, Fig. S3B and B
), demonstrating efficient
deletion of the floxed allele by E10.5. To test further for efficiency
of deletion we used a
Sox2Cre
driver to generate a global deletion
of
Ets1
using the
Ets1
floxed allele.
Sox2Cre
-mediated deletion of
the
Ets1
-floxed allele resulted in DORV with complete penetrance,
reproducing the global KO phenotype (data not shown). These
results indicate an early temporal requirement for ETS1 for nor-
mal cardiac crest function. There is likely variability in the exact
timing of the deletion of the
Ets1
-floxed allele, such that it is
deleted prior to the time it is required in only a subset of embryos.
Generation of NC cells from human-induced
pluripotent stem cells (iPSCs) from patients with
Jacobsen syndrome and congenital heart disease
Our studies in mice demonstrate a clear link between DORV
and the deficit of Ets1 in the cardiac NC. To explore the
effect of loss of ETS1 in NC migration in human patients with
congenital heart disease, we generated NC cells from human
iPSCs from three healthy subjects and two patients with Jacobsen
syndrome and hypoplastic left heart syndrome (HLHS): patient
Human Molecular Genetics
, 2022, Vol. 31, No. 24
|
4221
Figure 3.
Loss of ETS1 causes NCCs migration defects. (
A
–
B
) Immunos-
taining for SOX10 protein (green), a marker for migrating NCCs, in
sections from E9.5 control and
Ets1
KO embryos with
Pax3Cre;tdTomato
reporter. In control (
A
,
A
arrows), the Pax3-Cre-expressing migrating
NCCs cells are co-expressed with SOX10 protein and the
tdTomato
positive-expressing SOX10 cells are observed in a linear configuration
from the neural tube toward the developing heart. However, in
Ets
1KO
mutants the number of the
tdTomato
positive-expressing SOX10 cells is
markedly reduced (
B
,
B
arrows), and the linear migration pattern is
disrupted. Furthermore, most of the
tdTomato
positive-expressing cells
lacking SOX10 expression are located more peripherally. Scale bars:
100
μ
m.
Figure 4.
Time-lapse image analysis of explanted cultured cNCCs from
controls and
Ets1
KO mutant embryos. Quantitation of migration velocity
of cNCCs. Total, total distance and NET, net distance traveled. The data
represent the mean
±
SD of seven independent experiments of cultured
cNCCs from control and
Ets1
KO mutant embryos (
n
=
5). Statistical
analysis was performed by two-tailed unpaired
t
-test at confident level
of 95%.
∗∗∗
P
<
0.001.
8389 has a 14.0 Mb terminal deletion and multiple craniofacial
defects; patient 192 has an 8.0 Mb terminal deletion and mild
hypertelorism (
Supplementary Material, Table S2
). To characterize
iPSC-derived NCCs, we employed a neurosphere-based migration
method, as published previously (
22
–
25
). When the iPSC-derived
neurospheres were plated on Matrigel, a permissive substrate
for NC cells, a migratory cell population emerged mimicking
early events of NC delamination from the dorsal neural tube.
These migratory cells expressed known NC markers, including
SOX10, SOX9, Nestin and Pax3 and downregulated the expression
of pluripotency markers such as OCT4 and SOX2 in both
control and patient cells (
Supplementary Material, Fig. S4
). Upon
neuralization, expression of the pluripotency marker OCT4 was
suppressed, while SOX2 levels were reduced by 70% in both
control and patient cells (
Supplementary Material, Fig. S4
).
Loss of ETS1 negatively affects the migration
profile of hiPSC-derived NC cells
Based on our data from ETS1 KO mice indicating a NC cell
migration defect, we assessed the migratory performance of the
hiPSC-derived NCCs
in vitro
. We studied the ability of all the iPSC-
NCCs lines to invade a matrix and move towards a central cell-free
zone. A 3-D matrix approach was chosen to recapitulate the best
in vivo
environment. Cell invasion was quantified by obtaining
snapshot images with an automated microscope every 24 h. By
measuring the number of cells in the cell-free detection zone
over time, we observed an impaired migratory phenotype of NC
derived from iPSCs from the two patients compared to the three
controls (
Fig. 9
). At 48 h, NC cells from the two patients had a
slower migration rate compared to wild-type controls. To further
test the possibility that the lack of Ets1 in migratory NC cells was
responsible for the migration defects, we silenced the ETS1 gene in
control and patient iPSC-NCCs using siRNA for 72 h. ETS1-silenced
cells showed a decrease in ETS1 mRNA (
Supplementary Material,
Fig. S5
). This negatively affected the migration rate of cells into
the extracellular matrix, resulting in a reduction of cell counts in
the cell-free zone (
Fig. 10
and
Supplementary Material, Fig. S6
),
consistent with our findings in mice that loss of ETS1 is sufficient
to impair NC cell migration.
Discussion
Global deletion of ETS1 in C57/B6 mice caused ventricular septal
defects, which is the most common CHD that occurs in Jacobsen
syndrome, thus identifying the ETS1 transcription factor as a
candidate gene for causing CHDs in this setting (
9
,
10
,
14
). Previous
studies have led to the hypothesis that the heart defects in ETS1
KO mice were due to impaired migration of NC cells involving
the endocardial cushions (
14
,
15
). Our studies endorse and extend
these observations, providing additional insights regarding the
anatomical background of the lesions produced.
In this study, we performed a 3-D analysis of ETS1 KO embry-
onic hearts, which revealed DORV with a subaortic intraventricu-
lar communication and side-by-side arterial trunks. To date, we
have never observed common arterial trunk in ETS1 KO mice,
suggesting that loss of ETS1 affects OFT septation but NOT the
separation of the distal OFT into the aorta and pulmonary trunk.
It is unknown if, in mammals, there are distinct subsets of cardiac
NC cells that are required for these two developmental processes,
or if a common subset of NC cells can execute both the functions.
In support of the idea that ETS1 is important for cardiac
crest migration, here we demonstrate that ETS1 KO mice have
decreased numbers of cardiac NC cells in the developing prox-
imal OFT cushions, consistent with delayed migration. This is
detectable as early as E9.5 and is consistent with previous studies
4222
|
Human Molecular Genetics
, 2022, Vol. 31, No. 24
Figure 5.
Immunofluorescence analysis of sections from controls and
Ets1
KO mutants with
Pax3Cre;R26R-tdTomato
reporter. N-cadherin staining
demonstrates upregulation of N-cadherin expression in migratory NCCs in
Ets1
KO mutant at E8.5 (
A
,
B
, arrows) and E9.5 (
C
,
D
,arrows).
Pax3Cre–
tdTomato
(red), anti-N-cadherin (green), nucleus (DAPI, blue). Scale bars: 100
μ
m. (
E
) Potential binding sites of ETS1 in the promoter regions of N-cadherin
predicted by JASPAR (a, b and c, three potential ETS1 binding sites).
that have demonstrated a critical role for the proximal OFT cush-
ions in the development of the ventricular OFTs (
26
). Specifically,
our findings implicate a critical role for ETS1 in building the shelf
in the roof of the right ventricle that subsequently allows for the
aorta to achieve its connection with the left ventricle. Of note, we
also performed cell proliferation and apoptosis studies on E11.5
ETS1 global KO mice, and there was no difference compared to
wild-type controls (data not shown). In addition, it is possible that
ectopic migration of cardiac NC cells could also contribute to the
decreased cardiac NC cells observed in the proximal cushions.
Although not systematically studied, we did observe weak expres-
sion of SOX10 in the optic node in ETS1 global KO mice (data
not shown), suggestive of ectopic migration in a small subset of
cardiac NC cells.
Interestingly, surgical repair of this form of DORV is performed
utilizing the procedure (
27
) in which a tunnel is built such that the
interventricular communication is used to place the aortic root in
continuity with the left ventricular cavity. In addition, three out
of four ETS1 KO mice analyzed have a bisinuate aortic valve with
decreased NC cells in the valvar leaflets, consistent with previous
studies demonstrating a critical role for cNCCs in arterial valvar
development (
6
). In addition to decreased numbers of NCCs in
the proximal cushions prior to the time of OFT septation, ETS1
KO mice demonstrate abnormal myocardialization, resulting in
the presence of a fibrous outlet septum, with continuity between
the leaflets of the aortic and pulmonary valves. Impaired myocar-
dialization suggests that normally cardiac crest cells populating
the proximal OFTs generate a yet to be identified factor(s) that is
required for cardiac myocyte migration to the cushions.
Our studies implicate a unique subset of cardiac crest cells
that are required for septation of the OFTs, independent of those
required for separation of the distal OFT into the pulmonary trunk
and aorta. It is unclear whether this failure may be specifically
due to decreased migration of a multifunctional population of
cardiac crest cells to the proximal cushions, which can execute
separation of the OFTs. Alternatively, there may be a unique
subset of predetermined cardiac crest that populates the proximal
cushions and are required for myocardialization. Consistent with
this model, recent studies in chick by Gandhi
et al.
indicate that
there are distinct functional populations of pre-migratory cardiac
crest cells that exist in the left and right sides of the neural
folds (
28
), such ablation of left versus right-sided neural folds
gives rise predominantly to DORV versus common arterial trunk,
respectively.
Human Molecular Genetics
, 2022, Vol. 31, No. 24
|
4223
Figure 6.
Immunofluorescence analysis of sections from controls and
Ets1
KO mutants with the
Pax3Cre;tdTomato
reporter. Alpha-actinin staining
demonstrated cardiomyocytes adjacent to and protruding into
tdTomato
positive cNCCs in the OFT cushion in control E13.5 embryos (
A
). In the mutant,
the cardiac myocytes are separated from the
tdTomato
positive cNCCs (
B
).
A
and
B
show higher magnification images boxed in A and B. By E15.5, the
tdTomato
positive cNCCs in OFT cushions are fully surrounded by cardiomyocytes, demonstrated completion of muscularization of OFT cushions (
C
),
but in the
Ets1
KO mutant, there were markedly fewer cardiomyocytes in contact with
tdTomato
positive cNCCs in the OFT cushion, indicating impaired
muscularization. a-Actinin (green), nucleus (DAPI, blue). Scale bars: 100
μ
m.
Figure 7.
Sirius Red analysis of sections from controls and
Ets1
KO
mutants. In E18.5
Ets1
KO mutants, the supraventricular crest (SC) in
control hearts (
A
,
A
) was replaced with fibrotic tissue (
B
,
B
,arrows),with
continuity between the aortic and pulmonary valve leaflets. Scale bars:
500
μ
m.
Our studies demonstrate that conditional deletion of ETS1 in
the NC recapitulates the DORV phenotype, with 50% penetrance.
These results are consistent with a cell-autonomous role for ETS1
in the NC. The incomplete penetrance is likely due to variation
in the precise time that ETS1 is deleted by the PAX3/Cre driver,
relative to the exact time in cardiac development that ETS1 is
required for normal cardiac crest migration. Normally, cardiac
crest cells emigrate from the neural tube beginning at about E8.5.
Our
in vivo
studies demonstrate abnormal NC migration already
by E9.5, consistent with an early requirement of ETS1 for proper
migration.
Our
in vitro
studies demonstrate decreased migration velocity
and evidence for increased cell adhesion in ETS1 KO cultured NC
cells. We subsequently determined that loss of ETS1 results in an
Figure 8.
Histological analysis of cardiac defects in E14.5
Pax3Cre; Ets1
cKO hearts. Pax3Cre; R26R-LacZ-stained showing contribution of cNCCs
to OFT cushions, aortic valve leaflets and pulmonary trunk in an E14.5
control heart (
A
,
C
,
E
) and a
Pax3Cre;Ets1
cKO heart (
B
,
D
,
F
)demonstrating
the DORV phenotype: the aorta is aligned with the right ventricle with an
interventricular communication (IVC) (F, arrow). There are fewer cNCCs
in the aorta valve leaflets in the mutant (D) compared to control embryos
(C). Nuclei were counterstained with nuclear fast red. Scale bars: 100
μ
m.
increase in the expression of N-cadherin, a critical cell adhesion
molecule required for normal NC migration. Previous studies have
demonstrated that N-cadherin expression is downregulated prior
to the onset of NC migration (
21
), and that increased N-cadherin
expression impairs migration (
21
). Interestingly, polymorphisms
in the N-cadherin gene regulatory region have been correlated
with varying levels of N-cadherin expression (
29
). This raises
the intriguing possibility that N-cadherin may be a previously
unrecognized genetic modifier for the development of ventricular
OFT defects.
Our studies using human NC cells derived from iPS cells from
Jacobsen syndrome patients missing a copy of ETS1 and CHDs
demonstrate a NC migration defect, consistent with our stud-
ies in mice. Our results also provide novel insights into why
4224
|
Human Molecular Genetics
, 2022, Vol. 31, No. 24
Figure 9.
Effect of Ets1 reduction on NC cell migration
in vitro
. The plot shows the cell count of cells migrated in the detection zone every 24 h, for each
cell lines (three NCC wild-type and two NCC Jacobsen), cultured in standard conditions. The graph shows a significant reduction of cell migration in
Jacobsen lines. All the graphs were generated with GraphPad Prism. The statistical analysis was performed with ANOVA Tukey’s multiparametric test;
∗∗
P
<
0.01,
∗∗∗
P
<
0.001 average wild-type vs NC 8389;
$$$
P
<
0.001 average wild-type vs NC 192.
Figure 10.
Migration assay for each individual line, before and after treatment with siRNA. The plot shows the cell count of cells migrated in the detection
zone every 24 h, for each cell line growing in standard conditions (red lines) or silenced for ETS1 for 72 h (blue lines). The graph shows a significant
reduction of cell migration when ETS1 is silenced. All the graphs were generated with GraphPad Prism. The statistical analysis was performed with
ANOVA Sidak’s multiparametric test;
∗
P
<
0.05,
∗∗
P
<
0.01,
∗∗∗
P
<
0.001 siRNA vs controls for each cell line.
only homozygous KO mice exhibit DORV, while human patients
with Jacobsen syndrome have CHDs with only one copy of ETS1
deleted. Specifically, we observed different ETS1 mRNA and pro-
tein levels in NC cells derived from the iPS cells from the two
patients with Jacobsen syndrome and HLHS. The patient with the
largest deletion, 8389, had a markedly lower level of ETS1 mRNA
and protein (analogous to a mouse null mutant), compared to the
controls, and the most severe migration defect. Consistent with
this, the siRNA inhibition of Ets1 in the control cells reduced the
number of migrating cells to that of patient 8389 NOT treated with
siRNA. Patient 192, with a smaller deletion, had similar mRNA
levels as controls, and an intermediate migration defect compared
to patient 8389. This would suggest that in this patient, the cell
migration defect may be due to decreased function of the Ets1
protein derived from the ‘normal’ allele, compared to the controls.
In our iPSC model we observed a large effect on migration with
only 10–15% of cells expressing clearly detectable levels of Ets1
protein. This suggests the intriguing possibility that Ets1
+
cells
lead the migration of Ets1
−
cells in a 3-D environment.
Taken together, our studies demonstrate a critical cell
autonomous role for ETS1 in regulating early cardiac crest
migration and OFT septation, advancing our understanding of the
molecular and cellular basis for the pathogenesis of one anatomic
variant of DORV in humans.
Limitations of the study
Our studies on global and conditional deletion mice were per-
formed on homozygous KO mice, whereas in human patients with
Human Molecular Genetics
, 2022, Vol. 31, No. 24
|
4225
JS and CHDs there is one normal ETS1 allele. As described above,
patient 8389 had severely diminished ETS1 mRNA and protein
levels, analogous to the mouse homozygous KO ‘null’ mutant.
In contrast, patient 192 had mRNA and protein levels similar to
controls, suggesting that the ETS1 protein encoded by the intact
chromosome in this patient has diminished function. Both mech-
anisms may contribute to the cardiac phenotype, whereas other
patients without a cardiac phenotype may have a more normal
functioning allele from the intact chromosome with normal levels
of mRNA and protein. Additionally, the differences between mice
and humans could also be due to a lower threshold for sensitivity
to the ETS gene dosage in humans than in mice.
Our
in vivo
studies on NC migration were performed on global
ETS1 KO mice. Although conditional deletion of ETS1 recapitu-
lated the DORV phenotype observed in the global KO mice, it is
still possible that ETS1 regulates additional factors through a non-
cell-autonomous mechanism that can contribute to the DORV
phenotype. Our studies on human NCCs derived from iPS cells
were performed on two patients with HLHS. Previous studies have
implicated a NCC component contributing to the pathogenesis
of HLHS (
30
). Therefore (in order to contain costs) we chose to
generate iPS cells from two Jacobsen syndrome patients with
HLHS to be used for the present study as well as for future
studies to investigate the role of ETS1 in other cardiac cellular
lineages that are also implicated in causing CHDs, such as the
endocardium (
31
). In addition, although we did not specifically
assess the aortic valve morphology in the NC conditional deletion,
conditional deletion of ETS1 in the endothelium did not affect
aortic valve morphology, consistent with the loss of ETS1 in the
NC as the cause of the bisinuate aortic valve that we observed
in the ETS1 global KO. We have demonstrated that endothelial
deletion of ETS1 leads to a coronary defect and ventricular non-
compaction (
32
). The relative contribution of ETS1 deficiency in
the NC versus the endocardium in the pathogenesis of HLHS and
other CHDs will continue to be an area of great interest for future
studies (
31
).
Materials and Methods
Mice
Global ETS1 null mice have been described previously (
33
). Mice
from the Jackson Laboratory are Pax3Cre (
17
) (#005549), Rosa26
tdTomato (
16
) (#007908) and C57BL6/J (#000664). Ets1 floxed mice
were obtained from Dr Michael Ostrowski and have been reported
previously (
23
).
Mice were maintained on a C57BL6/J background, and genotyp-
ing was performed as described. Animal care and experimental
procedures were performed according to the NIH Guide for the
Care and Use of Laboratory Animals as well as institutional
guidelines and approved by the Institutional Animal Care and Use
Committee at UC San Diego (UCSD IRB Protocol #S01049).
Immunohistochemistry and histological analyses
For immunohistochemistry, mouse embryos were dissected out in
PBS and fixed in 4% paraformaldehyde overnight, then saturated
with 20% sucrose and frozen in OCT component and cut into
8-
μ
m-thick sections on a cryostat. Sections were blocked and
stained with antibodies. We used antibodies to the following
proteins: anti-ETS1 (Cell Signal, D808A, #14069), anti-N-cadherin
(Abcam, #ab76057), anti-Sox10 (Santa Cruz Biotechnology, SC-
17342) and anti-Sarcomeric alpha-actinin (Abcam, ab68167). Don-
key anti-rabbit Alexa Fluor 488 secondary antibody was from Jack-
son ImmunoResearch Laboratories Inc. For histological analyses,
mouse embryos were fixed in 4% paraformaldehyde overnight,
dehydrated and embedded in paraffin, cut into 8-mm-thick sec-
tions using a microtome and stained with hematoxylin–eosin
according to standard protocol.
Time-lapse imaging of migrating cNCCs
in explant culture
Mouse embryos were treated by dispase to dissociate the tissue
gently. After treatment, neural tubes from somite 1–3 regions
(cardiac NC) were dissected out and cut into 100
×
300
μ
m pieces.
Each piece of cardiac NC tissue was placed on fibronectin-coated
glass bottom slides and incubated in 5% CO
2
and 21% O
2
at 37
◦
C
for 16 h, and then the time-lapse imaging capture was initiated.
The time-lapse imaging was performed at 10 min intervals for 3 h
at 37
◦
C.
X-gal staining of mouse embryos
Mouse embryos were fixed in 4% paraformaldehyde for 30 min on
ice, permeabilized in PBS containing 0.02% nadeoxycholate and
0.01% NP-40 for 4 h at room temperature and were then subjected
to 5-bromo-4-chloro-3-indolyl-d-galactoside (X-gal) staining.
Analysis of collagen
Sections were stained with Sirius Red/Fast Green Collagen stain-
ing kit (Chondrex Inc., #9046) according to manufacturer’s proto-
cols.
Manufacture and characterization of human
iPSCs
Sendai reprogramming of primary human fibroblasts was con-
ducted using CytoTune-iPS Sendai Reprogramming Kits according
to manufacturer’s instructions (Invitrogen, A13780-02, A16517,
A16518) and manufactured by ReGen Theranostics. All skin biop-
sies were undertaken by institutional regulations. Briefly, skin
biopsy samples were dissected into pieces and incubated in the
six-well plate under a glass coverslip. Fibroblasts were main-
tained in DMEM/high glucose/glutamax (Invitrogen, 10 566-016).
After 7–14 days fibroblasts were removed by TrypLE (Invitrogen,
12 604 013), washed and expanded into a 100 mm dish. After
infection with Sendai 2.0 (DMEM, essential amino acids and 10%
FBS) fibroblasts were cultured in E7 media (STEMCELL Technolo-
gies, 05910). Colonies were selected and expanded via mechani-
cal dissociation and maintained in mTeSR1 on GelTrex (Thermo
Fisher, purity
≥
98%) coated plates. All clones were subjected to
etoposide sensitivity assay testing, analysis of TRA-1-60 surface
expression by flow cytometry, karyotyping, DNA fingerprinting
and mycoplasma screening.
Generation of NC cells from iPSCs
To generate NC cells from human iPSCs,we used the neurospheres
protocol, as previously described (
23
,
24
) using AggreWell system
(StemCell Technologies). Briefly, stem cells were detached as sin-
gle cells using Accutase, counted to calculate the necessary num-
ber of cells dependent on the size of spheres and number of wells,
and transferred to the AggreWell plate. After 24 h incubation,
neuralization was initiated by carefully moving the spheres to low
attachment well plate in serum-free chemically defined media
[1:1 ratio of DMEM/F12 glutamax (Gibco): neurobasal medium,
0.05X B27 supplement without vitamin A (Gibco), 10% BIT 9500
(StemCell Technologies), 5
μ
g/ml insulin (Sigma), 20 ng/ml bFGF
(Chemicon), 20 ng/ml EGF (Sigma), 1 mm glutamine (Gibco)]. Half
of the medium was changed every day for 5 days. Then, the
neurospheres were plated on Matrigel-coated culture dishes and
4226
|
Human Molecular Genetics
, 2022, Vol. 31, No. 24
allowed to migrate for 5 days in the same medium as described
above, supplemented with 20 ng/ml EGF, 20 ng/ml bFGF, 5
μ
g/ml
insulin and 5 mm nicotinamide. The medium was changed every
other day. We carried out stem cell neuralization under 3% O
2
conditions.
Invasion assay
To assess the invasion ability of NC cells through an extracellu-
lar matrix we used the Oris™ Cell Migration Assay (CMA1.101,
Platypus Technologies), which is based on a 96-well plate with
‘stopper’ barriers that create a central cell-free detection zone
for cell migration. NC cells were plated in a 96-well plate coated
with Matrigel in the presence of the ‘stopper’ tool. We seeded
each cell line at the same cell density to minimize variations. For
the controls, we plated cells in standard conditions (2-D Matrigel
without the stopper tools) and quantified them at each time point
to identify potential differences in cell proliferation rate among
the lines. Cells were incubated overnight to allow attachment to
the surface. The ‘stopper’ tool was carefully removed to create
the detection zone at the center of each well and a 3-D matrix
was created adding 30
μ
L of Matrigel, homogeneously distributed
on top of the cell monolayer and solidified at 37
◦
C for 30 min.
Cell medium was added on top of the Matrigel coating. Each
condition had four replicates per experiment, and each experi-
ment was reproduced two times separately. Cell tracking was done
by taking pictures in the bright field every 24 h, using an auto-
mated microscope (Celigo High Throughput Micro-Well Image
Cytometer, Nexcelom) and the number of migrated cells was
quantified by counting the cells in the detection zone using Celigo
software. GraphPad Prism was used for plotting and statistical
analysis, performed using the appropriate two-way ANOVA test
as indicated in each figure. At the end of the invasion assay cells
were fixed and stained with DAPI, Sox10 and ETS1, as described
below.
ETS1 silencing and metalloproteinases inhibition
ETS1 silencing gene in iPSC-derived NCC was achieved by using
Accell ETS1 siRNA SMARTpool (E-003887, Dharmacon) following
the manufacturer’s instructions. ETS1 siRNA was used at 1
μ
M
and delivered in the NC medium. Incubation of 72 h achieved
ETS1 silencing at 37
◦
Cunder3%O
2
conditions. The result of the
silencing was assessed and confirmed by qPCR and immunocyto-
chemistry. Silenced cells were then plated for the invasion assay
as described above.
Immunocytochemistry
Cells were imaged at the end of the differentiation protocol from
iPSC and at the end of the invasion assay as follows: Media was
aspirated, and cells rinsed once with PBS for 2 min, followed by
fixation with 4% PFA for 10 min at room temperature. Cells were
then washed twice with PBS and blocked and permeabilized with
PBSAT (0.5% Triton X-100 and 2% BSA diluted in PBS) for 1 h.
Incubation with the following primary antibodies was carried out
overnight: goat anti-SOX2 (1:500, R&D), rabbit anti-SOX9 (1:100,
Millipore), rabbit anti-SOX10 (1:500, R&D), goat anti-Nestin (1:500,
Santa Cruz Biotechnology), mouse anti-Integrin
α
4 (1:500, Milli-
pore), rabbit anti-Pax3 (1:500, R&D) and rabbit anti-ETS1 (1:250,
Cell Signaling). Cells were washed three times with PBSAT and
then incubated with the corresponding secondary antibody in
blocking solution for 1 h at room temperature. After washing three
times with PBSAT, nuclei were stained with DAPI (diluted 1:000 in
PBSAT) for 10 min. Cells were then washed two times with PBS
and kept in PBS at 4
◦
C until ready to be analyzed. Images were
acquired using IC200 microscope (Vala Sciences) and analyzed
with Acapella (PerkinElmer).
Supplementary Material
Supplementary Material
is available at
HMG
online.
Acknowledgements
The authors would like to thank Dr Michael Ostrowski for pro-
viding the
Ets1
floxed mice. The authors also would like to thank
the UCSD School of Medicine Microscopy Core (Grant: NINDS P30
NS047101) for their invaluable technical support.
Funding
American Heart Association (#16GRNT26700008); The Hertz Fam-
ily Foundation; The Rady Children’s Hospital Foundation; The 11q
Research and Resource Group; The Chloe Duyck Memorial Fund;
The Warren J. and Betty C. Zable Foundation and the cast and crew
of ‘How I Met Your Mother’.
Disclosures
The authors have no disclosures.
References
1.
Creazzo, T.L., Godt, R.E., Leatherbury, L., Conway, S.J. and Kirby,
M.L. (1998) Role of cardiac neural crest cells in cardiovascular
development.
Annu. Rev. Physiol.
,
60
, 267–286.
2.
Conway, S.J., Henderson, D.J. and Copp, A.J. (1997) Pax 3 is
required for cardiac neural crest migration in the mouse:
evidence from the splotch (Sp2H) mutant.
Development
,
124
,
505–514.
3.
Kirby, M.L. (1989) Plasticity and predetermination of mesen-
cephalic and trunk neural crest transplanted into the region of
the cardiac neural crest.
Dev. Biol.
,
134
, 402–412.
4.
Hutson, M.R. and Kirby, M.L. (2007) Model systems for the study
of heart development and disease. Cardiac neural crest and
conotruncal malformations.
Semin. Cell Dev. Biol.
,
18
, 101–110.
5.
Waldo, K., Miyagawa-Tomita, S., Kumiski, D. and Kirby, M.L.
(1998) Cardiac neural crest cells provide new insight into septa-
tion of the cardiac outflow tract: aortic sac to ventricular septal
closure.
Dev. Biol.
,
196
, 129–144.
6.
Anderson, R.H., Henderson, D.J. and Chaudhry, B. (2017) Devel-
opment and maintenance of the arterial valves.
Eur. Heart J.
,
38
,
687–689.
7.
Bruneau, B.G. and Srivastava, D. (2014) Congenital heart disease:
entering a new era of human genetics.
Circ. Res.
,
114
, 598–599.
8.
Garrett-Sinha, L.A. (2013) Review of Ets 1 structure, function, and
roles in immunity.
Cell. Mol. Life Sci.
,
70
, 3375–3390.
9.
Favier, R., Akshoomoff, N., Mattson, S. and Grossfeld, P. (2015)
Jacobsen syndrome: advances in our knowledge of phenotype
and genotype.
Am.J.Med.Genet.C Semin.Med.Genet.
,
169
, 239–250.
10.
Ye, M., Coldren, C., Liang, X., Mattina, T., Goldmuntz, E., Benson,
D.W., Ivy, D., Perryman, M.B., Garrett-Sinha, L.A. and Grossfeld,
P. (2010) Deletion of ETS-1, a gene in the Jacobsen syndrome
critical region, causes ventricular septal defects and abnormal
ventricular morphology in mice.
Hum. Mol. Genet.
,
19
, 648–656.
11.
Davidson, B. (2007) Ciona intestinalis as a model for cardiac
development.
Semin. Cell Dev. Biol.
,
18
, 16–26.
Human Molecular Genetics
, 2022, Vol. 31, No. 24
|
4227
12.
Alvarez, A.D., Shi, W., Wilson, B.A. and Skeath, J.B. (2003) pannier
and pointedP2 act sequentially to regulate Drosophila heart
development.
Development
,
130
, 3015–3026.
13.
Glessner, J.T., Bick, A.G., Ito, K., Homsy, J., Rodriguez-Murillo, L.,
Fromer, M., Mazaika, E., Vardarajan, B., Italia, M., Leipzig, J.
et al.
(2014) Increased frequency of de novo copy number variants
in congenital heart disease by integrative analysis of single
nucleotide polymorphism array and exome sequence data.
Circ.
Res.
,
115
, 884–896.
14.
Gao, Z., Kim, G.H., Mackinnon, A.C., Flagg, A.E., Bassett, B., Earley,
J.U. and Svensson, E.C. (2010) Ets1 is required for proper migra-
tion and differentiation of the cardiac neural crest.
Development
,
137
, 1543–1551.
15.
Ye, M., Yin, Y. and Fukatsu, K. (2016) Evidence that deletion of
ETS-1, a gene in the Jacobsen Syndrome (11q-) cardiac critical
region, causes congenital heart defects through impaired car-
diac neural crest cell function. In Nakanishi, T., Markwald, R.R.,
Baldwin, H.S., Keller, B.B., Srivastava, D. and Yamagishi, H. (eds),
Etiology and Morphogenesis of Congenital Heart Disease: From Gene
Function and Cellular Interaction to Morphology [Internet]
. Springer,
Tokyo, Chapter 52.
16.
Madisen, L., Zwingman, T.A., Sunkin, S.M., Oh, S.W., Zariwala,
H.A., Gu, H., Ng, L.L., Palmiter, R.D., Hawrylycz, M.J., Jones, A.R.
et al.
(2010) A robust and high-throughput Cre reporting and
characterization system for the whole mouse brain.
Nat. Neu-
rosci.
,
13
, 133–140.
17.
Engleka, K.A., Gitler, A.D., Zhang, M., Zhou, D.D., High, F.A. and
Epstein, J.A. Insertion of Cre into the Pax3 locus creates a new
allele of Splotch and identifies unexpected Pax3 derivatives.
Dev.
Biol.
,
280
, 396–406.
18.
Betancur, P., Bronner-Fraser, M. and Sauka-Spengler, T. (2010)
Genomic code for Sox10 activation reveals a key regulatory
enhancer for cranial neural crest.
Proc. Natl. Acad. Sci. U. S. A.
,
107
, 3570–3575.
19.
Epstein, J.A., Li, J., Lang, D., Chen, F., Brown, C.B., Jin, F., Lu,
M.M., Thomas, M., Liu, E., Wessels, A.
et al.
(2000) Migration of
cardiac neural crest cells in Splotch embryos.
Development
,
127
,
1869–1878.
20.
Punovuori, K., Malaguti, M. and Lowell, S. (2021) Cadherins in
early neural development.
Cell. Mol. Life Sci.
,
78
, 4435–4450.
21.
Taneyhill, L.A. and Schiffmacher, A.T. (2017) Should I stay or
should I go? Cadherin function and regulation in the neural
crest.
Genesis
,
55
, 10.
22.
Menendez, L., Kulik, M.J., Page, A.T., Park, S.S., Lauderdale, J.D.,
Cunningham, M.L. and Dalton, S. (2013) Directed differentiation
of human pluripotent cells to neural crest stem cells.
Nat. Protoc.
,
8
, 203–212.
23.
Cimadamore, F., Curchoe, C.L., Alderson, N., Scott, F., Salvesen, G.
and Terskikh, A.V. (2009) Nicotinamide rescues human embry-
onic stem cell-derived neuroectoderm from parthanatic cell
death.
Stem Cells
,
27
, 1772–1781.
24.
Curchoe, C.L., Maurer, J., McKeown,S.J., Cattarossi, G.,
Cimadamore, F., Nilbratt, M., Snyder, E.Y., Bronner-Fraser,
M. and Terskikh, A.V. (2010) Early acquisition of neural crest
competence during hESCs neuralization.
PLoS One
,
5
, e13890.
25.
Lee, G., Chambers, S.M., Tomishima, M.J. and Studer, L. (2010)
Derivation of neural crest cells from human pluripotent stem
cells.
Nat. Protoc.
,
5
, 688–701.
26.
Anderson, R.H., Mori, S., Spicer, D.E., Brown, N.A. and Mohun, T.J.
(2016) Development and morphology of the ventricular outflow
tracts.
World J. Pediatr. Congenit. Heart Surg.
,
7
, 561–577.
27.
McGoon, D.C., Rastelli, G.C. and Wallace, R.B. (1970) Discontinuity
between right ventricle and pulmonary artery: surgical treat-
ment.
Ann. Surg.
,
172
, 680–689.
28.
Gandhi, S., Ezin, M. and Bronner, M.E. (2020) Reprogramming
axial level identity to rescue neural crest-related congenital
heart defects.
Dev. Cell
,
53
, 300–315.
29.
Ruedel, A., Stark, K., Kaufmann, S., Bauer, R., Reinders, J., Roven-
sky, J., Blazicková, S., Oefner, P.J. and Bosserhoff, A.K. (2014) N-
cadherin promoter polymorphisms and risk of osteoarthritis.
FASEB J.
,
28
, 683–691.
30.
Ahola, J., Koivusalo, A., Sairanen, H., Jokinen, E., Rintala, R.J. and
Pakarinen, M.P. (2009) Increased incidence of Hirschsprung’s dis-
ease in patients with hypoplastic left heart syndrome—a com-
mon neural crest-derived etiology?
J. Pediatr. Surg.
,
44
, 1396–1400.
31.
Grossfeld, P., Nie, S., Lin, L., Wang, L. and Anderson, R.H. (2019)
Hypoplastic left heart syndrome: a new paradigm for an old
disease?
J. Cardiovasc. Dev. Dis.
,
6
, 10.
32.
Wang, L., Lin, L., Qi, H., Chen, J. and Grossfeld, P. (2022) Endothe-
lial loss of ETS1 impairs coronary vascular development and
leads to ventricular non-compaction.
Circ. Res.
,doi:
10.1161/CIR-
CRESAHA.121.319955
33.
Barton, K., Muthusamy, N., Fischer, C., Ting, C.N., Walunas, T.L.,
Lanier, L.L. and Leiden, J.M. (1998) The Ets-1 transcription factor
is required for the development of natural killer cells in mice.
Immunity
,
9
, 555–563.