European Journal of Medical Genetics 53 (2010) 6–13

Contents lists available at ScienceDirect

European Journal of Medical Genetics

journal homepage: http://www.elsevier.com/locate/ejmg

Genetic factors in esophageal atresia, tracheo-esophageal fistula and the

VACTERL association: Roles for FOXF1 and the 16q24.1 FOX transcription
factor gene cluster, and review of the literature
Charles Shaw-Smith a, b, *
a
Wellcome Trust Sanger Institute, Hinxton, Cambridge CB10 1SA, UK
b
Institute of Child Health, 30, Guilford Street, London WC1N 1EH, UK

a r t i c l e i n f o a b s t r a c t

Article history: Esophageal atresia with/without tracheo-esophageal fistula is a relatively common malformation,
Received 24 July 2009 occurring in around 1 in 3500 births. In around half of cases, additional malformations are present,
Accepted 4 October 2009 forming either a syndrome of known genetic aetiology, or a recognised association, of which the VAC-
Available online 12 October 2009
TERL association (Vertebral anomalies, Anal atresia, Cardiac malformations, Tracheo-Esophageal fistula,
Renal and Limb malformations) is the most recognised. Recently, microdeletions of the FOX gene cluster
Keywords:
at 16q24.1, comprising four genes, FOXF1, MTHFSD, FOXC2 and FOXL1, were reported to cause a phenotype
Esophageal atresia
resembling VACTERL association, with vertebral anomalies, gastro-intestinal atresias (esophageal,
Tracheo-esophageal fistula
VACTERL association duodenal and anal), congenital heart malformations, and urinary tract malformations, as well as a rare
16q24.1 microdeletion lethal developmental anomaly of the lung, alveolar capillary dysplasia. This article reviews these new
FOXF1 data alongside other genetic causes of syndromic esophageal atresia, and also highlights information
Sonic hedgehog from relevant mouse models, particularly those for genes in the Sonic Hedgehog pathway.
Ó 2009 Elsevier Masson SAS. All rights reserved.

1. Introduction Genes involved in critical developmental processes, such as tran-

Esophageal atresia with or without tracheo-esophageal fistula is
a relatively common malformation, occurring in around 1 in 3500
births [50]. In around half of cases, additional malformations are
present, forming either a syndrome of known genetic aetiology, of
which CHARGE syndrome [OMIM 214800], Feingold syndrome
[OMIM 164280] and AEG syndrome [OMIM 206900] are examples;
or a recognised association, of which the VACTERL association
[OMIM 192350] is the best recognised. Esophageal atresia and the
VACTERL association have been the subject of several recent
reviews [5,16,50]; however, new data concerning the FOX tran-
scription factor gene cluster at 16q24.1 [52] have provided addi-
tional insights, making another look at genetic aetiologies of
esophageal atresia and the VACTERL association worthwhile.
A significant factor which has expedited this research is the
application of high-resolution array-based methods to patients
with congenital malformations [25,49]. Chromosomal deletions in
humans result in haploinsufficiency and may therefore disrupt the
normal function of any dosage-sensitive genes which they harbour.
* Tel.: þ44 1223 834244; fax: þ44 1223 494919.
E-mail address: css@sanger.ac.uk
scription factors and signalling molecules, often prove to be dosage
sensitive. Examples in humans include SOX2 [OMIM 184429] in eye
and foregut development, and N-MYC [OMIM 164840] and, most
recently, members of the FOX transcription factor gene cluster at
16q24.1 in foregut and lung development [52].
The application of array-based comparative genomic hybrid-
ization (array-CGH) to the detection of haploinsufficiency in
humans has therefore proved to be very useful for the identifica-
tion, amongst other things, of developmentally critical, dosage-
sensitive, genes. This technique is now fully integrated into clinical
cytogenetics laboratories [2], and in the future may well supersede
analysis of G-banded metaphase cells by light microscopy. It has the
ability to detect copy number loss (and gain) with extremely high
resolution, far higher than was previously achievable in a routine
clinical cytogenetics laboratory only a few years ago. This has had
two related, beneficial effects: the first is that more pathogenic
deletions are being identified; the second is that the deletions are
much smaller, and so contain orders of magnitude fewer candidate
genes.
Some of the results from the studies using human subjects have
been confirmed in other model systems. For example, the efforts of
mouse developmental biologists have been hugely valuable in
characterizing gene function in developmental processes by

1769-7212/$ – see front matter Ó 2009 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmg.2009.10.001
 C. Shaw-Smith / European Journal of Medical Genetics 53 (2010) 6–13
targeted mutagenesis of candidate genes in mice. In the field of
foregut and lung development, as in many other fields, work on
signalling molecules and transcription factors in a wide range of
pathways is gradually building up a picture of key events in
morphogenesis of the foregut, lung and other organ systems
[6,33,44]. This work offers a panel of candidate genes to human
geneticists as they analyze the loci which emerge from array data.
In this review, I focus initially on phenotypes associated with
microdeletions at 16q24.1, and the role of the individual genes in
the FOX transcription factor cluster. I then go on to place these new
data in the context of other single gene disorders associated with
syndromic esophageal atresia. The contribution of mouse models to
our understanding is discussed, followed by a section focussing on
the role of Sonic hedgehog and genes in that pathway.
2. FOXF1 microdeletions at 16q24.1
Recently it was shown that microdeletions encompassing the
FOX cluster at 16q24.1 result in alveolar capillary dysplasia
together with a broad spectrum of additional malformations [52].
Alveolar capillary dysplasia with misalignment of pulmonary veins
(ACD/MPV) is a rare lethal disorder associated chiefly with failure
of development of the intrinsic pulmonary vasculature [[52] and
references therein]. There is minimal response to respiratory
supportive measures and death usually occurs within the first
month of life. In around 80% of cases, additional malformations are
present, and the elucidation of the genetic pathology at 16q24.1 in
a significant proportion of these cases has enabled some inter-
esting genotype–phenotype correlations to be made. Patients with
deletions spanning the entire FOX cluster (FOXF1, MTHFSD, FOXC2
and FOXL1) at 16q24.1 (these are hereinafter referred to as ‘whole
cluster deletions’) have, in addition to ACD/MPV, esophageal
atresia, tracheo-esophageal fistula, and other gastro-intestinal tract
atresias (duodenal and anal atresia). Congenital heart defects,
urinary tract malformations, vertebral or axial malformations and
single umbilical artery are also present, making the resemblance to
the VACTERL association a compelling one. This raises the possi-
bility that some cases of 16q24.1 microdeletion may be mis-
diagnosed as VACTERL association, and the accompanying
diagnosis of ACD/MPV overlooked. The diagnosis of ACD/MPV can
be made only by histology, either at post mortem or by examina-
tion of lung biopsy tissue; and even then may be missed, as it is
rare and requires some expertise on the part of the pathologist to
make it.
Of the four genes in the 16q24.1 FOX cluster, two, FOXF1 [OMIM
601089] and FOXC2 [OMIM 602402] have now been reasonably well
studied both in humans and the mouse; the remaining two,
MTHFSD and FOXL1 [OMIM 603252] have been less well studied.
Mutations in FOXC2 cause lymphoedema–distichiasis syndrome
[OMIM 153400]. A comparison of the phenotypes due to FOXF1 and
FOXC2 mutations in humans, and to Foxf1 and Foxc2 inactivation in
mouse is shown in Table 1, together with the phenotypes resulting
from deletions of the whole FOX cluster in humans. For some organ
systems, clear interpretation and correlation for the two genes in
human and mouse is possible. For example, in the vertebral/axial
system, FOXC2 mutations, whole cluster deletions, and Foxc2
knockout mice all result in vertebral/axial malformations
[3,11,19,52]; neither humans with FOXF1 mutations nor Foxf1
knockout mice appear to manifest these. Similarly, FOXF1 and Foxf1,
but not FOXC2 or Foxc2, are clearly associated with abnormalities of
intrinsic pulmonary vasculature and lung lobation [23,28,34] while
inactivation of FOXC2 and Foxc2, but not FOXF1 or Foxf1, is associ-
ated with cleft palate [3,19]. Mutations in both FOXF1 and FOXC2, as
well as whole cluster deletions, result in urinary tract
7
malformations [3,52] but there is no reference to this organ system
in the literature on Foxf1 and Foxc2 mouse mutants.
In the cardiovascular system and gastro-intestinal tract, these
correlations are not as simple. FOXC2 is clearly implicated in
cardiovascular malformations, albeit with low penetrance:
tetralogy of Fallot in mutation cases [3] and interrupted aortic arch
in whole cluster deletion cases [52], correlating well with mouse
mutants, which have aortic arch malformations [19]. However, one
of the human FOXF1 mutation cases had a partial AV canal defect
[52]; and no such malformations have been reported in Foxf1
mutant mice.
In the GI tract, FOXF1 mutations appear consistently to cause
intestinal malrotation, as well as a collection of other GI tract
anomalies: duodenal stenosis, congenital short bowel, and annular
pancreas, but no gastro-intestinal atresia [52]. This contrasts with
heterozygous Foxf1 null mice, which have esophageal atresia and
tracheo-esophageal fistula on selected genetic background (CD1),
but in which neither intestinal malrotation, nor indeed any other GI
tract anomaly, have been reported [35]. In contrast again, humans
with whole cluster deletions have not only EA/TEF but also
duodenal and anal atresia [52]. No GI tract malformations have
been reported in humans or mice with FOXC2/Foxc2 mutations. The
possible implications of these observations, and their relationship
to the Sonic Hedgehog pathway, are discussed further below.
Not only whole cluster deletions, but also deletions upstream of
the 16q24.1 FOX cluster, are associated with an abnormal pheno-
type. Such deletions reproduce the alveolar capillary dysplasia
phenotype (cases D9 and D10, Fig. 1) and are also associated, in one
case, with anal atresia (case D9, Fig. 1), suggesting that the ‘critical
interval’ for GI atresias may not actually include the FOX cluster
itself (Fig. 1). Future studies will focus on refining this interval by
the ascertainment of additional deletion cases.
Mutations in MTHFSD and FOXL1 have not been described in
humans. Targeted disruption of Foxl1 in mice is associated with
hyperplasia of the gastric mucosa, abnormal crypt architecture and
epithelial cell positioning in the small intestine and retarded
growth [32,54]; its role in humans will be unclear until mutations
or single gene deletions of FOXL1 are described in humans.
Comparison of the phenotypes of patients with FOXF1, FOXC2
and whole cluster deletions leads to a hypothesis that the pheno-
type in patients with whole cluster deletions results from the
contiguous deletion of FOXF1 and FOXC2 with perhaps additional
effects from the other genes or elements in the cluster. This model
appears to hold for vertebral, cardiac and renal malformations, but
not in the GI tract, where malformations seen in FOXF1 mutation
cases are different from those seen in whole cluster deletion
patients. This apparent puzzle is discussed further in the section on
the Sonic Hedgehog pathway, below.
Aside from ACD/MPV, there are some phenotypic differences
between 16q24.1 microdeletion syndrome and VACTERL associa-
tion. Hypoplastic left heart syndrome has not previously been
reported as a cardiac manifestation of VACTERL association. Limb
malformations, particularly of the thumb and radius, are a part of
the VACTERL association but do not appear to be present in patients
with 16q24.1 microdeletions, while other malformations not part of
VACTERL do occur, such as cleft lip and palate in one patient [52].
3. Syndromic esophageal atresia/tracheo-esophageal fistula
as a single gene disorder
Four disorders in which EA/TEF feature in a significant number
of cases, and for which the genetic defect has been elucidated, have
been described: these are Feingold syndrome [OMIM 164280],
CHARGE syndrome [OMIM 214800], AEG syndrome [OMIM
206900] and VACTERL-H syndrome [OMIM 276950] (see Table 2
8 C. Shaw-Smith / European Journal of Medical Genetics 53 (2010) 6–13

Table 1
Phenotypes associated with haploinsufficiency of FOXF1 and FOXC2 in humans and of Foxf1, Foxc2 and Shh in mouse.

Gene/ Vertebrae Anal Cardiac Tracheo- Renal Limb Respiratory Cranio-facial Other References
locus Esophageal
FOXF1 – – AV canal – Hydronephrosis, – ACD/MPV, – Intestinal malrotation, [52]
defect, PDA hydroureter, abnormal lung annular pancreas, duodenal
dilated bladder lobation stenosis, congenital short
bowel,
Foxf1 – – – EA/TEF – – ACD-like but no – Gall bladder small with [23,28,35]
MPV. Lung abnormal smooth muscle
lobation
anomalies
FOXC2 Scoliosis, rib – Tetralogy of – Duplex kidney, – – Cleft palate Lymphoedema–distichiasis, [3,11]
fusion Fallot, VSD, pyelonephritis, varicose veins, ptosis
PDA urinary tract
infections
Foxc2 Split neural – Interrupted – – – – Cleft palate, – [19]
arches, absent aortic arch, hypoplastic/
spinous aortic fused middle
processes, rib coarctation, ear bones
fusions aortic atresia,
VSD
Deletions Butterfly Anal HLHS, EA/TEF Hydronephrosis, – ACD/MPV, Cleft lip, cleft Single umbilical artery [52]
of vertebra, rib atresia Tetralogy of renal pneumothorax palate,
whole fusions Fallot, pelvicaliectasis pulmonary brachycephaly
FOX Interrupted lymphangiectasia
cluster aortic arch,
PDA
Shh Agenesis of Anal Abnormalities EA/TEF Renal agenesis Limb Lung lobation Cyclopia Intestinal malrotation, [9,29,46]
axial skeleton atresia of cardiac truncation annular pancreas, duodenal
looping defects stenosis, abnormal gut
innervation, intestinal
transformation of stomach

AV=atrioventricular, VSD=ventricular septal defect, PDA=patent ductus arteriosus, HLHS=hypoplastic left heart syndrome, ACD/MPV=alveolar capillary dysplasia/misalign-
ment of pulmonary veins.

and references therein). For the purposes of comparison, the
16q24.1 locus is added to this list in Table 2, accepting the fact that
the genetic mechanism in this disorder has not been fully eluci-
dated, and that it may be a contiguous gene deletion syndrome.
There are some striking similarities between the five disorders,
as well as some differences. Vertebral and urinary tract malfor-
mations occur in all five. All except AEG syndrome feature cardio-
vascular malformations, with aortic arch abnormalities featuring
prominently, and likewise, anal atresia or stenosis occur in these
four. GI tract malformations are very similar in Feingold syndrome
and 16q24.1 microdeletions, with EA/TEF, duodenal atresia, anal
atresia, and annular pancreas occurring in both. There is least
concordance for limb malformations, which are of different types in
the three conditions in which they occur, VACTERL-H, Feingold and
Charge syndromes. There is clearly a danger in overstating the
similarities between these diverse conditions, especially as the
column labelled ‘other’ contains many anomalies not otherwise
classified in the table, but nonetheless, similarities both across and
within organ systems, such as those between Feingold syndrome
and 16q24.1 microdeletion syndrome, are striking and hint at
similarities in pathogenesis.
Esophageal atresia/tracheo-esophageal fistula occur occasion-
ally in other single gene disorders. These include Opitz B/GGG
syndrome [OMIM 300000], due to mutations in MID1 [OMIM
300552] at Xp22.3 [42], in which laryngeal malformations (cleft,
diastema) are more usual; McKusick–Kaufman syndrome [OMIM
236700], which is allelic with Bardet–Biedl syndrome [BBS-6,
OMIM 209900] and due to mutations in MKKS [OMIM 604896]
[51,53], a gene related to members of the chaperonin family; and
Oculo-Auriculo-Vertebral Spectrum [OAVS, OMIM 164210], for
which epigenetic dysregulation of BPAX1 has been proposed as an
aetiological mechanism [14]. Putative single gene or sporadic
disorders featuring esophageal atresia for which there is as yet no
confirmed genetic aetiology include Fryns syndrome [OMIM
229850] [1], and congenital microgastria–limb reduction complex
[OMIM 156810] [31], discussed further below. Finally, there are
several reports of a syndrome of multiple gastro-intestinal atresias,
including esophageal atresia/tracheo-esophageal fistula, with gall
bladder agenesis and neonatal diabetes mellitus, associated in
some but not all cases with pancreatic anomalies [8]. Parental
consanguinity and sibling recurrence point to autosomal recessive
inheritance, but there are as yet no mapping data.
4. Mouse models featuring esophageal
atresia/tracheo-esophageal fistula
A list of mouse models featuring esophageal atresia or tracheo-
esophageal fistula is provided in Table 3. Prominent on this list are
members of the Sonic Hedgehog pathway, including Sonic
Hedgehog itself. Shh/ mice have an extensive set of malforma-
tions which overlaps very significantly not only with the VACTERL
association, but also with the set of malformations seen in patients
with microdeletions at 16q24.1 (excluding ACD/MPV) [9,52].
Mutations in SHH in humans cause holoprosencephaly [HPE3,
OMIM 142945] [40] and microphthalmia [OMIM 611638] [48].
Mutations have not been reported in patients with malformations
of the gastro-intestinal tract. The GLI family of transcription factors
are downstream effectors of SHH and in humans, mutations in
members of this gene family result in syndromes well known to
clinical geneticists. Greig cephalopolysyndactyly [GCPS, OMIM
175700] and Pallister–Hall [PHS, OMIM 146510] syndromes are
caused by mutations in GLI3 [20], while mutations in GLI2 are
associated with holoprosencephaly [47]. None of these syndromes
manifests foregut malformations, although anal atresia is
 C. Shaw-Smith / European Journal of Medical Genetics 53 (2010) 6–13 9

FLJ30679
FOXC2
FOXL1
FOXF1
MTHFSD
LOC732275

MAP1LC3B
CRISPLD2

ZCCHC14
KIAA0513
KIAA1609

C16ofr74
COX4NB
ATP2C2

FBX031
COX4i1
KIA182
COTL1

USP10

GINS2

JPH3
IRF8
83 84 85 86
D1 Mb from 16pter

D2
D3
D4
D5
D6
D7
D8
D9
D10

FLJ30679
FOXC2
MTHFSD
FOXF1

FOXL1
LOC732275

MAP1LC3B
CRISPLD2

ZCCHC14
KIAA1609

KIAA0513

C16ofr74
COX4NB
ATP2C2

FBX031
COX4i1
KIA182
COTL1

USP10

GINS2

JPH3
IRF8

83 84 85 86 M b from 16pter
D1
D2
D3
D4
D5
D8
D9
D10

Fig. 1. Microdeletions at 16q24.1 in patients with gastro-intestinal atresias Three patients in the recent report of microdeletions at 16q24.1 presented with gastro-intestinal atresias:
D1, D3 and D9. One of these, D9, is a deletion upstream of the FOX transcription factor cluster, suggesting a ‘critical interval’ for gastro-intestinal atresias at this locus, that does not
span the FOX cluster itself. More deletion cases are needed in order to confirm or refute these preliminary ideas.

a component of Pallister–Hall syndrome; however, mice with
inactivation of both Gli2 and Gli3 do have foregut malformations
(summarized in Table 3), and again it seems worthwhile to search
for mutations in GLI family members in patients with syndromic
esophageal atresia.
Mice with Noggin/ mutations have a wide range of malfor-
mations including esophageal atresia and tracheo-esophageal
fistula [27,43]. Tantalizingly, a clear link between the NOGGIN locus
at 17q22 and esophageal atresia has been established in humans
[36], with most recently, the critical interval narrowed to a 5.9 Mb
region encompassing NOGGIN [41]. The phenotype in these dele-
tion cases has been tentatively delineated to include deafness and
skeletal malformations (symphalangism, joint contractures) as well
as EA/TEF, and so mutations in NOGGIN could usefully be sought in
deletion-negative patients with this phenotype.
Finally, a mouse model was published recently in which muta-
tion of the proprotein convertase enzyme Pcsk5 was shown to be
associated with malformations in the VACTERL spectrum, including
defects of tracheo-oesophageal septation [60]. Convincing muta-
tions in this gene in humans with VACTERL association have yet to
be identified.
5. Dissection of Shh and Foxf1 function in model organisms
Key emerging players in the pathogenesis of esophageal atresia,
tracheo-esophageal fistula and the VACTERL association are
members of the Sonic Hedgehog pathway, and of the FOX tran-
scription factor gene cluster at 16q24.1. Studies in model organisms
including Xenopus and mouse have demonstrated key roles for
Sonic Hedgehog itself, and for Foxf1, in mediating mesoderm–
endoderm cross talk during very early development of the gut tube.
A critical early step is the subdivision of lateral plate mesoderm into
its somatic and visceral components, and evidence from Xenopus
[56] and mouse [34] assigns a key role to Foxf1 in this process.
Subsequently, the dorsal mesentery, from which the early gut tube
is suspended from the dorsal abdominal wall, and the splanchnic
mesoderm lining the gut tube, develop from the visceral compo-
nent of the lateral plate mesoderm. Signalling between the visceral
mesoderm, which develops into smooth muscle and other
specialized tissues in the gut wall, and the gut endoderm, which
provides the epithelial cell lining of the gastro-intestinal tract, is
critical for correct development of the gut [37]. There is compelling
evidence that Foxf1 is a downstream effector of the Sonic hedgehog
 10
Table 2
Syndromic esophageal atresia due to single gene disorders in humans, with microdeletion at 16q24.1 for comparison.

Syndrome/gene Vertebrae Anal atresia/other Cardiac Tracheo- Renal Limb Other References
GI malformations Esophageal
Feingold syndrome Fused cervical Anal atresia, also Tricuspid EA, TEF Hydronephrosis, Short middle Microcephaly, learning [7]
N-MYC vertebrae, absent duodenal atresia stenosis/atresia, cystic dysplasia, phalanges of 2nd difficulties, deafness, short
sacral vertebrae, or stenosis, interrupted aortic dilatation of renal and 5th digits, stature, asplenia or polysplenia

C. Shaw-Smith / European Journal of Medical Genetics 53 (2010) 6–13

absent rib pairs annular pancreas arch, VSD, PDA pelvis, small kidneys hypoplastic thumbs,
toe syndactyly (2nd
and 3rd, or 4th and 5th
toes)

VACTERL- Lumbar spina bifida Anal atresia Aortic coarctation, EA, TEF Unilateral renal Bilateral radial agenesis, Fanconi anaemia, hydrocephalus, [18,30,38]
hydrocephalus occulta, cervical ASD, VSD, agenesis, renal absent thumbs Arnold–Chiari malformation; cleft
FANCB vertebral defects dysplasia palate, incomplete
lung lobation

Charge syndrome Vertebral body Anal stenosis HLHS, Tetralogy of EA, TEF Renal agenesis, Monodactyly, tibial Coloboma of eye, structural anomalies [21,57]
CHD7 anomalies, Fallot, DORV, AVSD, horseshoe kidney, aplasia, bifid femora of external ear, deafness, agenesis of
kyphoscoliosis right-sided descending vesico-ureteric reflux, semi-circular canals, choanal atresia,
aorta, Shone’s complex, renal cysts cleft lip, cleft palate, cryptorchidism,
ASD, VSD, PDA micropenis, hydrocephalus, corpus
callosum agenesis, seizures, learning
difficulties

AEG syndrome Hemi-vertebrae, – – EA, TEF Hypoplastic kidneys, – Anophthalmia, microphthalmia, [59]
SOX2 butterfly vertebrae, duplex kidneys hypospadias, cryptorchidism
fused ribs, absent ribs,
extra ribs

16q24.1 Butterfly vertebra, rib Anal atresia, HLHS, Tetralogy of Fallot, EA, TEF Hydronephrosis, – ACD/MPV, cleft lip, cleft palate, [52]
microdeletion fusions posterior placement Interrupted aortic arch, renal pelvicaliectasis brachycephaly, single umbilical artery
encompassing of anus, duodenal PDA
FOXF1, MTHFSD, atresia, annular
FOXC2 and FOXL1 pancreas

AVSD=atrioventricular septal defect, ASD=atrial septal defect, VSD=ventricular septal defect, PDA=patent ductus arteriosus, HLHS=hypoplastic left heart syndrome, ACD/MPV=alveolar capillary dysplasia/misalignment of
pulmonary veins, DORV=double outlet right ventricle.
 C. Shaw-Smith / European Journal of Medical Genetics 53 (2010) 6–13 11

Table 3
Knockout mouse models featuring esophageal atresia and/or tracheo-esophageal fistula, with phenotype associated with inactivation of counterpart human gene for
comparison.

Mouse Lung/foregut phenotype Phenotype (other) Reference Human Locus Mutations Lung/ Phenotype (other) Reference
gene gene foregut
phenotype
Shh Esophageal atresia, tracheo- Abnormalities of CNS; cyclopia; [9,29,46] SHH 7q36 YES NR Holoprosencephaly [HPE3, [40,48]
esophageal fistula, lung distal limb truncation; OMIM 142945],
anomalies abnormalities of axial skeleton, microphthalmia [OMIM
renal agenesis, abnormalities of 611638]
heart looping, intestinal
malrotation, annular pancreas,
duodenal stenosis, intestinal
transformation of the stomach,
abnormal gut innervation,
imperforate anus

Foxf1 Esophageal atresia, tracheo- Axial skeletal defects [22,35] FOXF1 16q24.1 YES Alveolar Intestinal malrotation, atrio- [52]
esophageal fistula, lung lobe capillary ventricular canal defect, renal
fusion defects, pulmonary dysplasia malformations
vascular defects

Gli2 Gli2/ mice are normal. Gli2 NR [39] GLI2 2q14 YES NR Holoprosencephaly [47]
/ Gli3þ/ mice have
esophageal atresia tracheo-
esophageal fistula. Gli2/ Gli3
/ mice have absent
oesophagus, trachea and lungs

Gli3 See Gli2 entry Gli3/ mice are allelic with Xt. [39] GLI3 7p13 YES NR Greig cephalopolysyndactyly [20]
They have synpolydactyly and syndrome [GCPS, OMIM
brain malformations. 175700]; Pallister–Hall
syndrome [PHS, OMIM 146510]
(imperforate anus, polydactyly,
hypopituitarism, hypothalamic
hamartoblastoma)

Noggin Esophageal atresia, tracheo- Failure of closure of neural tube, [27,43] NOGGIN 17q22 YES NR Proximal symphalangism with [4,17,26]
esophageal fistula exencephaly, wide, club-shaped multiple synostoses; stapes
limbs, shortened, abnormal ankylosis with broad thumbs
body axis, lethality at birth and toes; brachydactyly type B2

Sox2 Esophageal atresia, tracheo- Neurodegeneration, impaired [13,44] SOX2 3q26.3 YES Esophageal Anophthalmia, genitourinary [59]
esophageal fistula neurogenesis atresia malformations

NR ¼ not recorded.

pathway in this process, both in the lung and in the gut. In the lung,
ectopic epithelial expression of Shh activates Foxf1 expression [35];
in Shh/ mice, Foxf1 mRNA is absent from its usual sites of
expression in mesenchyme of trachea and oesophagus, and lungs,
although midgut and hindgut do retain some expression [35].
Similarly, in the intestine, but not the stomach, of both Gli2 and Gli3
null mice, Foxf1 levels are reduced [32]. The Foxf1 promoter, as well
as that of Foxl1, contains binding sites for Gli factors which mediate
transcriptional activation [32]. It does appear that expression of
Foxf1 is not exclusively controlled by the Sonic Hedgehog pathway,
as evidenced by its retained expression in the intestine of Shh/
mice [35], but clearly there is some regulation of Foxf1 by this
pathway, and the close link between the two is evident in the
phenotypic similarities of mouse mutants and human patients with
FOXF1 mutations and deletions [52]. More detailed analysis of GI
tract malformations in Shh/ mice has revealed more extensive
abnormalities: intestinal malrotation, annular pancreas, duodenal
stenosis, abnormal gut innervation, and intestinal transformation
of stomach [46]. Strikingly, the first three malformations on this list
are also seen in patients with FOXF1 mutations [52]. Sonic Hedgehog
therefore appears to provide a link between the set of malforma-
tions seen in FOXF1 mutation cases and the set seen in whole
cluster deletion cases: the malformations that occur are all seen in
Shh/ mice. The relationship between Shh and Foxf1 is further
explored in Fig. 2.
In mice, the phenotypic consequences of Shh, Gli transcription
factor and Foxf1 inactivation are shown in Table 3; the resemblances
to VACTERL association have already been commented upon. How
do we relate these observations from model organisms to relevant
human phenotypes? For SHH, human mutations so far described
have resulted in holoprosencephaly and microphthalmia, but these
have so far been mis-sense mutations or in-frame deletions. Where
are patients with inactivating mutations and deletions of this gene?
One possibility is that these mutations are lethal at an early
embryonic stage; a second possibility is that investigators have not
looked in the right phenotypic group. One reported phenotype is the
Microgastria–Limb Reduction complex [OMIM 156810], which
features terminal transverse limb defects similar to those seen in
Shh/ mice, intestinal malrotation, esophageal and anal atresia,
megacolon, abnormal lung lobation, heart defects, renal and CNS
anomalies. This shows some similarities to the phenotype of Shh/
mutant mice, but, against this, three reported cases of this condition
have been in discordant twin pairs, suggesting that the aetiology
may be related to the twinning process [31].
Concerning FOXF1 and the 16q24.1 FOX gene cluster, the inter-
esting question remains: why does FOXF1 inactivation result in
intestinal malrotation, but deletion of the entire FOX cluster result
in gastro-intestinal atresias, whereas Foxf1 inactivation in mice
results in esophageal atresia? Future studies will focus on nar-
rowing the ‘critical interval’ responsible for occurrence of GI atre-
sias (esophageal, duodenal and anal) in patients with
microdeletions at 16q24.1. It will be interesting to determine
whether FOXL1, which also has a role in intestinal development,
contributes to this phenotype in humans.
12 C. Shaw-Smith / European Journal of Medical Genetics 53 (2010) 6–13
Fig. 2. Epithelial–mesenchymal interactions mediated by the Sonic Hedgehog
pathway in the gut. Sonic hedgehog effects signalling between the epithelial and
mesenchymal cell in the gastro-intestinal tract, and acts indirectly upon Foxf1 via its
receptors, Smo and Ptch, and Gli transcription factors. The effects of Foxf1 are mediated
within the mesenchyme, but there is currently much uncertainty surrounding the
identity of the downstream effectors of Foxf1. Shh, Sonic Hedgehog; Ptch, Patched;
Smo, Smoothened.
6. Chromosomal imbalances
A comprehensive review of chromosomal anomalies in esoph-
ageal atresia/tracheo-esophageal fistula was published recently
[12], and in this section, attention is drawn to just a few of these.
Trisomies for chromosomes 18 and 21 have been associated with
EA/TEF; interestingly more cases have been associated with trisomy
18 despite the fact that it is much rarer than trisomy 21 [50]. The
locus perhaps closest to revealing a new causative gene is 17q22,
harbouring NOG, discussed above [36]. A link between chromo-
somal deletions at chromosome 13q32, and VACTERL-type mal-
formations has previously been postulated [58], although in fact
deletions within the region 13q22-13qter have been associated
with a very broad spectrum of malformations [24,45] and the
suggested link with VACTERL association probably reflects ascer-
tainment bias, at least to some extent. The zinc finger transcription
factor ZIC2 [OMIM 603073] appears to be responsible for the major
CNS malformations occurring in patients with deletions at this
locus [45]. Just one case with esophageal atresia has been reported
[58]; nonetheless, there does appear to be a definite link with anal
atresia and peno-scrotal transposition mapping to 13q33, and the
critical region for this malformation appears to exclude ZIC2 [15].
Esophageal atresia/tracheo-esophageal fistula are rare associations
in patients with deletions at chromosome 22q11.2 [10]. Finally,
microdeletions spanning the FOX transcription factor gene cluster
at chromosome 16q24.1 should now be added to this list [52], but
further genotype–phenotype studies are required in order to clarify
precisely which genes or sequence elements are responsible.
7. Concluding remarks
The recent delineation of the role of the 16q24.1 locus in the
aetiology of severe developmental malformations illustrates very
clearly the role that high-resolution microarrays can play in
improving our understanding of developmental disorders. All of the
deletions reported in affected patients are below the level of cyto-
genetic resolution, but very easily detectable using microarrays.
Recent work on chromosomal anomalies in EA/TEF suggests that
there are other haploinsufficient susceptibility loci for these and
related malformations, notably at 13q and 17q, and quite naturally
as a result of clinical studies, our knowledge will grow. There is also
a clear utility in research studies focussed on a particular malfor-
mations: our own study, the Genetics of Oesophageal Atresia http://
www.ich.ucl.ac.uk/ich/academicunits/MMU/CustomMenu_01 now
has high-resolution array data on close to 100 patients with syn-
dromic esophageal atresia, with some potentially interesting loci.
These data will be published in due course.
A complementary approach is the sequencing of genes in the
same cohort of patients, looking for mutations both in known
(N-MYC, SOX2 and others) and in candidate (SHH, NOG and others)
genes. Sequencing can be carried out on a large scale on whole-
genome amplified DNA, as has recently been described for the
whole X chromosome [55]. Ultimately, the goal of these studies is to
complete the ‘cytogenetic map’ for esophageal atresia and tracheo-
esophageal fistula. This work is now well under way, and will be
useful not only for its own sake but also as a complement to studies
of non-genetic factors in EA/TEF, which have to date been restricted
mostly to large-scale epidemiological studies.
Acknowledgements
C.S.-S. is the recipient of a Wellcome Trust Intermediate Clinical
Fellowship, and gratefully acknowledges additional funding from
the Addenbrooke’s Charitable Trust and Tracheo-Oesophageal
Fistula Support (TOFS), the UK patient support group for patients
with esophageal atresia and tracheo-esophageal fistula. Thanks to
Vicki Martin for technical assistance and to Jianwen Que for
detailed suggestions on the manuscript.
References
[1] S. Ayme, C. Julian, D. Gambarelli, B. Mariotti, A. Luciani, N. Sudan, et al., Fryns
syndrome: report on 8 new cases. Clin. Genet. 35 (1989) 191–201.
[2] A.L. Beaudet, J.W. Belmont, Array-based DNA diagnostics: let the revolution
begin. Annu. Rev. Med. 59 (2008) 113–129.
[3] G. Brice, S. Mansour, R. Bell, J.R. Collin, A.H. Child, A.F. Brady, et al., Analysis of
the phenotypic abnormalities in lymphoedema–distichiasis syndrome in 74
patients with FOXC2 mutations or linkage to 16q24. J. Med. Genet. 39 (2002)
478–483.
[4] D.J. Brown, T.B. Kim, E.M. Petty, C.A. Downs, D.M. Martin, P.J. Strouse, et al.,
Autosomal dominant stapes ankylosis with broad thumbs and toes, hyperopia,
and skeletal anomalies is caused by heterozygous nonsense and frameshift
mutations in NOG, the gene encoding noggin. Am. J. Hum. Genet. 71 (2002)
618–624.
[5] H.G. Brunner, H. van Bokhoven, Genetic players in esophageal atresia and
tracheoesophageal fistula. Curr. Opin. Genet. Dev. 15 (2005) 341–347.
[6] W.V. Cardoso, J. Lu, Regulation of early lung morphogenesis: questions, facts
and controversies. Development 133 (2006) 1611–1624.
[7] J. Celli, H. van Bokhoven, H.G. Brunner, Feingold syndrome: clinical review and
genetic mapping. Am. J. Med. Genet. A 122A (2003) 294–300.
[8] L. Chappell, S. Gorman, F. Campbell, S. Ellard, G. Rice, A. Dobbie, et al., A further
example of a distinctive autosomal recessive syndrome comprising neonatal
diabetes mellitus, intestinal atresias and gall bladder agenesis. Am. J. Med.
Genet. A 146A (2008) 1713–1717.
[9] C. Chiang, Y. Litingtung, E. Lee, K.E. Young, J.L. Corden, H. Westphal, et al.,
Cyclopia and defective axial patterning in mice lacking Sonic hedgehog gene
function. Nature 383 (1996) 407–413.
 C. Shaw-Smith / European Journal of Medical Genetics 53 (2010) 6–13
[10] M.C. Digilio, B. Marino, P. Bagolan, A. Giannotti, B. Dallapiccola, Microdeletion
22q11 and oesophageal atresia. J. Med. Genet. 36 (1999) 137–139.
[11] J. Fang, S.L. Dagenais, R.P. Erickson, M.F. Arlt, M.W. Glynn, J.L. Gorski, et al.,
Mutations in FOXC2 (MFH-1), a forkhead family transcription factor, are
responsible for the hereditary lymphedema–distichiasis syndrome. Am.
J. Hum. Genet. 67 (2000) 1382–1388.
[12] J.F. Felix, D. Tibboel, A. de Klein, Chromosomal anomalies in the aetiology of
oesophageal atresia and tracheo-oesophageal fistula. Eur. J. Med. Genet. 50
(2007) 163–175.
[13] A.L. Ferri, M. Cavallaro, D. Braida, A. Di Cristofano, A. Canta, A. Vezzani, et al.,
Sox2 deficiency causes neurodegeneration and impaired neurogenesis in the
adult mouse brain. Development 131 (2004) 3805–3819.
[14] S. Fischer, H.J. Ludecke, D. Wieczorek, S. Bohringer, G. Gillessen-Kaesbach,
B. Horsthemke, Histone acetylation dependent allelic expression imbalance of
BAPX1 in patients with the oculo-auriculo-vertebral spectrum. Hum. Mol.
Genet. 15 (2006) 581–587.
[15] N.M. Garcia, J. Allgood, L.J. Santos, D. Lonergan, J.R. Batanian, M. Henkemeyer,
et al., Deletion mapping of critical region for hypospadias, penoscrotal trans-
position and imperforate anus on human chromosome 13. J. Pediatr. Urol. 2
(2006) 233–242.
[16] D. Genevieve, L. de Pontual, J. Amiel, S. Sarnacki, S. Lyonnet, An overview of
isolated and syndromic oesophageal atresia. Clin. Genet. 71 (2007) 392–399.
[17] Y. Gong, D. Krakow, J. Marcelino, D. Wilkin, D. Chitayat, R. Babul-Hirji, et al.,
Heterozygous mutations in the gene encoding noggin affect human joint
morphogenesis. Nat. Genet. 21 (1999) 302–304.
[18] S.T. Holden, J.J. Cox, I. Kesterton, N.S. Thomas, C. Carr, C.G. Woods, Fanconi
anaemia complementation group B presenting as X linked VACTERL with
hydrocephalus syndrome. J. Med. Genet. 43 (2006) 750–754.
[19] K. Iida, H. Koseki, H. Kakinuma, N. Kato, Y. Mizutani-Koseki, H. Ohuchi, et al.,
Essential roles of the winged helix transcription factor MFH-1 in aortic arch
patterning and skeletogenesis. Development 124 (1997) 4627–4638.
[20] J.J. Johnston, I. Olivos-Glander, C. Killoran, E. Elson, J.T. Turner, K.F. Peters, et al.,
Molecular and clinical analyses of Greig cephalopolysyndactyly and Pallister–
Hall syndromes: robust phenotype prediction from the type and position of
GLI3 mutations. Am. J. Hum. Genet. 76 (2005) 609–622.
[21] M.C. Jongmans, R.J. Admiraal, K.P. van der Donk, L.E. Vissers, A.F. Baas,
L. Kapusta, et al., CHARGE syndrome: the phenotypic spectrum of mutations in
the CHD7 gene. J. Med. Genet. 43 (2006) 306–314.
[22] V.V. Kalinichenko, L. Lim, D.B. Stolz, B. Shin, F.M. Rausa, J. Clark, et al., Defects in
pulmonary vasculature and perinatal lung hemorrhage in mice heterozygous
null for the Forkhead Box f1 transcription factor. Dev. Biol. 235 (2001) 489–506.
[23] V.V. Kalinichenko, Y. Zhou, D. Bhattacharyya, W. Kim, B. Shin, K. Bambal, et al.,
Haploinsufficiency of the mouse Forkhead Box f1 gene causes defects in gall
bladder development. J. Biol. Chem. 277 (2002) 12369–12374.
[24] M. Kirchhoff, A.M. Bisgaard, R. Stoeva, B. Dimitrov, G. Gillessen-Kaesbach,
J.P. Fryns, et al., Phenotype and 244k array-CGH characterization of chromo-
some 13q deletions: an update of the phenotypic map of 13q21.1-qter. Am.
J. Med. Genet. A 149A (2009) 894–905.
[25] D.H. Ledbetter, Cytogenetic technology – genotype and phenotype. N. Engl.
J. Med. 359 (2008) 1728–1730.
[26] K. Lehmann, P. Seemann, F. Silan, T.O. Goecke, S. Irgang, K.W. Kjaer, et al., A new
subtype of brachydactyly type B caused by point mutations in the bone morpho-
genetic protein antagonist NOGGIN. Am. J. Hum. Genet. 81 (2007) 388–396.
[27] Y. Li, Y. Litingtung, P. Ten Dijke, C. Chiang, Aberrant Bmp signaling and noto-
chord delamination in the pathogenesis of esophageal atresia. Dev. Dyn. 236
(2007) 746–754.
[28] L. Lim, V.V. Kalinichenko, J.A. Whitsett, R.H. Costa, Fusion of lung lobes and
vessels in mouse embryos heterozygous for the forkhead box f1 targeted
allele. Am. J. Physiol. Lung Cell. Mol. Physiol. 282 (2002) L1012–L1022.
[29] Y. Litingtung, L. Lei, H. Westphal, C. Chiang, Sonic hedgehog is essential to
foregut development. Nat. Genet. 20 (1998) 58–61.
[30] F.E. Lomas, J.E. Dahlstrom, J.H. Ford, VACTERL with hydrocephalus: family with
X-linked VACTERL-H. Am. J. Med. Genet. 76 (1998) 74–78.
[31] I.W. Lurie, C.A. Magee, C.C. Sun, C. Ferencz, ‘Microgastria–limb reduction’
complex with congenital heart disease and twinning. Clin. Dysmorphol. 4
(1995) 150–155.
[32] B.B. Madison, L.B. McKenna, D. Dolson, D.J. Epstein, K.H. Kaestner, FoxF1 and
FoxL1 link hedgehog signaling and the control of epithelial proliferation in the
developing stomach and intestine. J. Biol. Chem. 284 (2009) 5936–5944.
[33] Y. Maeda, V. Dave, J.A. Whitsett, Transcriptional control of lung morphogen-
esis. Physiol. Rev. 87 (2007) 219–244.
[34] M. Mahlapuu, M. Ormestad, S. Enerback, P. Carlsson, The forkhead transcrip-
tion factor Foxf1 is required for differentiation of extra-embryonic and lateral
plate mesoderm. Development 128 (2001) 155–166.
[35] M. Mahlapuu, S. Enerback, P. Carlsson, Haploinsufficiency of the forkhead gene
Foxf1, a target for sonic hedgehog signaling, causes lung and foregut malfor-
mations. Development 128 (2001) 2397–2406.
[36] A.J. Marsh, D. Wellesley, D. Burge, M. Ashton, C. Browne, N.R. Dennis, et al.,
Interstitial deletion of chromosome 17 (del(17)(q22q23.3)) confirms a link
with oesophageal atresia. J. Med. Genet. 37 (2000) 701–704.
13
[37] V.A. McLin, S.J. Henning, M. Jamrich, The role of the visceral mesoderm in the
development of the gastrointestinal tract. Gastroenterology 136 (2009) 2074–
2091.
[38] A.R. Meetei, M. Levitus, Y. Xue, A.L. Medhurst, M. Zwaan, C. Ling, et al.,
X-linked inheritance of Fanconi anemia complementation group B. Nat. Genet.
36 (2004) 1219–1224.
[39] J. Motoyama, J. Liu, R. Mo, Q. Ding, M. Post, C.C. Hui, Essential function of Gli2
and Gli3 in the formation of lung, trachea and oesophagus. Nat. Genet. 20
(1998) 54–57.
[40] L. Nanni, J.E. Ming, M. Bocian, K. Steinhaus, D.W. Bianchi, C. Die-Smulders, et
al., The mutational spectrum of the sonic hedgehog gene in hol-
oprosencephaly: SHH mutations cause a significant proportion of autosomal
dominant holoprosencephaly. Hum. Mol. Genet. 8 (1999) 2479–2488.
[41] H. Puusepp, O. Zilina, R. Teek, K. Mannik, S. Parkel, K. Kruustuk, et al., 5.9 Mb
microdeletion in chromosome band 17q22–q23.2 associated with tracheo-
esophageal fistula and conductive hearing loss. Eur. J. Med. Genet. 52 (2009)
71–74.
[42] N.A. Quaderi, S. Schweiger, K. Gaudenz, B. Franco, E.I. Rugarli, W. Berger, et al.,
Opitz G/BBB syndrome, a defect of midline development, is due to mutations
in a new RING finger gene on Xp22. Nat. Genet. 17 (1997) 285–291.
[43] J. Que, M. Choi, J.W. Ziel, J. Klingensmith, B.L. Hogan, Morphogenesis of the
trachea and esophagus: current players and new roles for noggin and Bmps.
Differentiation 74 (2006) 422–437.
[44] J. Que, T. Okubo, J.R. Goldenring, K.T. Nam, R. Kurotani, E.E. Morrisey, et al.,
Multiple dose-dependent roles for Sox2 in the patterning and differentiation
of anterior foregut endoderm. Development 134 (2007) 2521–2531.
[45] C. Quelin, C. Bendavid, C. Dubourg, C. de la Rochebrochard, J. Lucas, C. Henry,
et al., Twelve new patients with 13q deletion syndrome: genotype–phenotype
analyses in progress. Eur. J. Med. Genet. 52 (2009) 41–46.
[46] M. Ramalho-Santos, D.A. Melton, A.P. McMahon, Hedgehog signals regulate
multiple aspects of gastrointestinal development. Development 127 (2000)
2763–2772.
[47] E. Roessler, Y.Z. Du, J.L. Mullor, E. Casas, W.P. Allen, G. Gillessen-Kaesbach, et
al., Loss-of-function mutations in the human GLI2 gene are associated with
pituitary anomalies and holoprosencephaly-like features. Proc. Natl. Acad. Sci.
U.S.A. 100 (2003) 13424–13429.
[48] L.A. Schimmenti, J. de la Cruz, R.A. Lewis, J.D. Karkera, G.S. Manligas,
E. Roessler, et al., Novel mutation in sonic hedgehog in non-syndromic
colobomatous microphthalmia. Am. J. Med. Genet. A 116A (2003)
215–221.
[49] C. Shaw-Smith, R. Redon, L. Rickman, M. Rio, L. Willatt, H. Fiegler, et al.,
Microarray based comparative genomic hybridisation (array-CGH) detects
submicroscopic chromosomal deletions and duplications in patients with
learning disability/mental retardation and dysmorphic features. J. Med. Genet.
41 (2004) 241–248.
[50] C. Shaw-Smith, Oesophageal atresia, tracheo-oesophageal fistula, and the
VACTERL association: review of genetics and epidemiology. J. Med. Genet. 43
(2006) 545–554.
[51] A.M. Slavotinek, E.M. Stone, K. Mykytyn, J.R. Heckenlively, J.S. Green, E. Heon,
et al., Mutations in MKKS cause Bardet–Biedl syndrome. Nat. Genet. 26 (2000)
15–16.
[52] P. Stankiewicz, P. Sen, S.S. Bhatt, M. Storer, Z. Xia, B.A. Bejjani, et al., Genomic
and Genic deletions of the FOX gene cluster on 16q24.1 and inactivating
mutations of FOXF1 cause alveolar capillary dysplasia and other malforma-
tions. Am. J. Hum. Genet. (2009).
[53] D.L. Stone, A. Slavotinek, G.G. Bouffard, S. Banerjee-Basu, A.D. Baxevanis,
M. Barr, et al., Mutation of a gene encoding a putative chaperonin causes
McKusick–Kaufman syndrome. Nat. Genet. 25 (2000) 79–82.
[54] M. Takano-Maruyama, K. Hase, H. Fukamachi, Y. Kato, H. Koseki, H. Ohno,
Foxl1-deficient mice exhibit aberrant epithelial cell positioning resulting from
dysregulated EphB/EphrinB expression in the small intestine. Am. J. Physiol.
Gastrointest. Liver Physiol. 291 (2006) G163–G170.
[55] P.S. Tarpey, R. Smith, E. Pleasance, A. Whibley, S. Edkins, C. Hardy, et al., A
systematic, large-scale resequencing screen of X-chromosome coding exons in
mental retardation. Nat. Genet. 41 (2009) 535–543.
[56] H.T. Tseng, R. Shah, M. Jamrich, Function and regulation of FoxF1 during
Xenopus gut development. Development 131 (2004) 3637–3647.
[57] I. Van de Laar, D. Dooijes, L. Hoefsloot, M. Simon, J. Hoogeboom, K. Devriendt,
Limb anomalies in patients with CHARGE syndrome: an expansion of the
phenotype. Am. J. Med. Genet. A 143A (2007) 2712–2715.
[58] L.E. Walsh, G.H. Vance, D.D. Weaver, Distal 13q Deletion Syndrome and the
VACTERL association: case report, literature review, and possible implications.
Am. J. Med. Genet. 98 (2001) 137–144.
[59] K.A. Williamson, A.M. Hever, J. Rainger, R.C. Rogers, A. Magee, Z. Fiedler, et al.,
Mutations in SOX2 cause anophthalmia–esophageal–genital (AEG) syndrome.
Hum. Mol. Genet. 15 (2006) 1413–1422.
[60] D. Szumska, G. Pieles, R. Essalmani, M. Bilski, D. Mesnard, K. Kaur, et al.
VACTERL/caudal regression/Currarino syndrome-like malformations in mice
with mutation in the proprotein convertase Pcsk5. Genes Dev. 22 (2008)
1465–1477.