Journal of Pediatric Surgery 50 (2015) 12511259

Contents lists available at ScienceDirect

Journal of Pediatric Surgery

journal homepage: www.elsevier.com/locate/jpedsurg

Lung maturity in esophageal atresia: Experimental and

clinical study,
Ana Catarina Fragoso a,b,c,, Leopoldo Martinez a,b, Jos Estevo-Costa c, Juan A. Tovar a,b
a
Department of Pediatric Surgery, Hospital Universitario La Paz, Madrid, Spain
b
Department of Congenital Malformations, INGEMM and IdiPaz Research Laboratory, Madrid, Spain
c
Faculty of Medicine, University of Porto, Porto, Portugal

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

Key words: Introduction: Esophageal atresia and tracheoesophageal stula (EA-TEF) survivors suffer respiratory morbidity of
esophageal atresia unclear pathogenesis. Defective lung morphogenesis has been described in the rat model. This study examined
tracheoesophageal stula fetal lung growth and maturity in rats and patients with EA-TEF.
lung maturity Methods: Pregnant rats received either adriamycin or vehicle. Control and adriamycin-exposed lungs, with and
lung hypoplasia
without EA-TEF, were weighed and processed for RT-PCR, DNA quantication, immunouorescence and immunoblot
newborn
adriamycin rat model
analysis of TTF1, VEGF, Sp-B, and -sma. Twenty human lungs were also processed for immunouorescence and
Alcian-blue staining.
Results: Lungs from fetuses with EA-TEF (E21) showed decreased total DNA; FGF7 and TTF1 mRNA expressions were
upregulated at E15 and E18, respectively. Protein expression and immunouorescent distribution of maturity
markers were similar. Lungs from stillborns with EA-TEF showed decreased epithelial expression of Sp-B and VEGF
whereas those from newborns tended to have less Sp-B and more VEGF and mucous glands.
Discussion: The lungs of rats with EA-TEF were hypoplastic but achieved near-normal maturity. Stillborns with
EA-TEF exhibited an apparently disturbed differentiation of the airway epithelium. Newborns with EA-TEF
demonstrated subtle differences in the expression of differentiation markers, and increased number of mucous
glands that could inuence postnatal respiratory adaptation and explain some respiratory symptoms of
EA-TEF survivors.
2015 Elsevier Inc. All rights reserved.

Esophageal atresia with tracheoesophageal stula (EA-TEF) is a
congenital malformation with an incidence of 1 per 3000 live births. Now-
adays, with survival approaching 95%, the focus of interest is shifting to
functional and quality of life issues throughout childhood and adulthood.
Approximately fty percent of neonates with EA-TEF have one or
more additional skeletal, anal, cardiovascular or renal malformations.
Respiratory malformations occur in 6% of patients [1], in 13.2% of autop-
sies of newborns with EA-TEF [2], and in up to 47% of VACTERL (verte-
bral, anal, cardiac, tracheoesophageal, renal, limbs) association
patients [3,4]. The relatively low incidence and the scarce clinical rele-
vance of some of these malformations invite to ascribe the respiratory
symptoms in these children to tracheomalacia, aspiration related to
impaired esophageal motility, esophageal stricture, recurrence of
tracheoesophageal stula or gastroesophageal reux. However, upon
adequate testing, up to 75% of EA-TEF survivors have abnormal pulmo-
nary function apparently not related to either associated conditions or
This study was supported in part by IdiPaz grants and by the Spanish Health Institute
Carlos III (grant N. RD08/0072: Maternal, Child Health and Development Network).
The authors declare that they do not have any competing or nancial interests.
Corresponding author at: Faculty of Medicine, University of Porto, Alameda Hernni
Monteiro, 4200319, Porto, Portugal.
E-mail address: catarina.fragoso@gmail.com (A.C. Fragoso).
to postoperative sequelae or prematurity [5]. In fact, EA-TEF survivors
suffer persistent unspecic respiratory symptoms (chronic cough,
asthma-like wheezing, and bronchial hyperresponsiveness) that some-
times do not improve or even become more frequent with age [68].
Additionally mild respiratory function disorders (obstructive/restrictive)
have been demonstrated in children, adolescents and adults born with
EA-TEF, but their origin remains unclear [5,6,9].
The scarcity of fetal or neonatal human tissue from affected individuals
prompted the development of animal models. In 1996, Diez-Pardo et al
described a toxicologic model of EA-TEF/VACTERL association. After
adriamycin administration to pregnant rats during the appropriate gesta-
tional days, a signicant proportion of fetuses accurately reproduced the
malformations of the human VACTERL association [10]. This invaluable
research tool was later extended to the mouse [11].
Because EA-TEF is the consequence of an abnormal division of the
foregut into esophagus and trachea, we hypothesize that the distur-
bance in the early embryonic development of the foregut that results
in EA-TEF may also interfere with the emergence of lung buds, and con-
sequently lung morphogenesis. This hypothesis was further supported
recently by demonstration of lung hypoplasia and abnormal control of
airway branching in the lungs of fetal rats with EA-TEF [12,13].
The present study aimed at gaining new insights into lung develop-
ment in EA-TEF. Lung growth and biochemical maturity were assessed

http://dx.doi.org/10.1016/j.jpedsurg.2015.06.015
0022-3468/ 2015 Elsevier Inc. All rights reserved.
1252 A.C. Fragoso et al. / Journal of Pediatric Surgery 50 (2015) 12511259
in rat fetuses with esophageal atresia and in lung specimens from
human fetuses and newborns with this malformation.
1. Materials and methods
Approval of the institution research ethical committee was obtained
for the study (HULP PI-1501). Parental consent had been required for
the retention of tissue in each case.
1.1. EA-TEF rat model
Time-dated pregnant Sprague-Dawley rats (OFA Charles River Labo-
ratories. Cerdanyola, Spain) were treated once a day from gestational
day 79 (morning of sperm in vaginal smear was considered day
0) by intraperitoneal injection, either of 1.75 mg/Kg adriamycin
(Farmiblastina Pharmacia, Madrid, Spain) or vehicle.
1.1.1. Fetal harvesting and dissection
Cesarean section was performed on E15, E18 and E21 before eutha-
nizing the dams with intracardiac injection of potassium chloride. At the
elected time endpoints, fetuses were recovered, weighed and dissected
under a microscope to document the presence of EA-TEF. Lungs were
harvested, weighed, photographed and xed or snap frozen and stored
at 80 until further use. Three groups of offsprings were compared:
control (C, n = 66), adriamycin-treated with EA-TEF (adria EA, n =
76) and adriamycin-treated without EA-TEF (adria noEA, n = 67).
1.1.2. Lung wet/dry weight ratio
Lungs from E21 fetuses were weighed immediately after dissection
to determine the wet weight and after being placed in an incubator
for 72 hours, to assess the dry weight.
1.1.3. Total DNA extraction and quantication
DNA was extracted from snap-frozen tissue with DNeasy blood and
tissue extraction kit (cat# 69504, Qiagen, Las Rozas, Spain), and the pu-
ried (RNAseA 100 mg/ml, cat#19101, Qiagen, Las Rozas, Spain) total
DNA content was determined with a spectrophotometer (Nano Drop;
Fisher Scientic, Madrid, Spain) at 260/280 m.
1.1.4. mRNA extraction and cDNA synthesis
Total mRNA was isolated from snap-frozen lungs using High Pure
RNA Tissue Kit (Roche Applied Science, Mannheim, Germany). Concen-
tration and purity of RNAs were determined spectrophotometrically;
250 ng of total RNAs was retrotranscripted to complementary DNAs
(cDNA) by reverse transcription reactions using a High Capacity RNA-
to-cDNA Kit (Applied Biosystems, Carlsbad, CA). All cDNAs were stored
at 80, until further use.
1.1.5. Real time reverse transcriptase polymerase chain reaction (real time
RT-PCR)
TTF1 and FGF7 lung expressions were quantied in a LightCycler 480 SYBR
Green I Master (Roche Applied Science, Mannheim, Germany) using the
following primer for TTF1: forward 5 CACCTTACCAGGACACCATG and reverse
3 GCCCATGCCGCTCATATTCA and forward 5 AAGTGAAAGGGACCCAGGAG
and reverse 3 GCCACAATTCCAACTGCCAC for FGF7. All RT-PCR reactions
were run in duplicate. The RT-PCR conditions were: 95 for 5 minutes,
followed by 50 cycles at 95 for 10 seconds, 60 for 10 seconds and
72 for 10 seconds. Results were normalized to the expression of
the 18S. The relative mRNA levels were determined by calculating
the threshold cycle for TTF1 and FGF7 gene using the threshold
cycle method.
1.1.6. Western blot
TTF1, Sp-B, VEGF and -smooth muscle actin protein levels were
measured in homogenized lungs (E21) in cell disruption buffer (PARIS
Kit, Ambion, USA). The protein content of 50 g of tissue was measured
using a protein assay kit (Pierce; BCA Protein Assay Kit, Rockford, IL,
USA). Immunoblotting was performed with 12% SDS-polyacrylamide
gel with the anti-TTF1 1:300 (H-190:sc13040; Santa Cruz Biotechnolo-
gies, Santa Cruz, CA, USA), Sp-B 1:300 (H-300: sc-13978; Santa Cruz
Biotechnologies, Santa Cruz, CA, USA), VEGF 1:400 (07-1420, Merck
Millipore, Germany) and -smooth muscle actin antibody 1:200
(1A4:sc-32251; Santa Cruz Biotechnologies, Santa Cruz, CA, USA).
Values were normalized to anti-Cu/Zn superoxide-dismutase (1:1000,
Stressgen, Belgium).
1.1.7. Immunouorescence staining
Lungs from E21 were xed overnight in 4% paraformaldehyde. After
inclusion in parafn, 5 m sections were stained. Immunouorescence
staining was performed using standard techniques with anti-TTF1 rabbit
polyclonal antibody (H-190:sc13040. Santa Cruz Biotechnologies, Santa
Cruz, CA, USA) after a dilution 1:100; anti-SpB rabbit polyclonal anti-
body (H-300: sc-13978; Santa Cruz Biotechnologies, Santa Cruz, CA,
USA) 1:50; anti-VEGF rabbit polyclonal antibody (07-1420, Merck
Millipore, Germany) 1:400 and anti--smooth muscle actin mouse
monoclonal antibody (1A4:sc-32251; Santa Cruz Biotechnologies,
Santa Cruz, CA, USA) 1:200. Briey, antigen recovery was performed
with sodium citrate (10 mM, pH 6) in microwave for 10 minutes follow-
ed by a 2 hours incubation with 10% horse serum, 1% albumin TBS. Sec-
tions were incubated overnight at 4 C with primary antibodies then
washed and incubated with universal secondary antibody (Vectstain
Universal Quick Kit; PK-8800; Vector Laboratories, Inc. Burlingame, CA,
USA) for 1 hour, and at last with Streptavidin Alexa uor 488 conjugate
(S-32354, Molecular Probes, Invitrogen, Carlsbad CA, USA) for
45 minutes. The sections were mounted and the nuclei counterstained
using Vectashield Mounting Medium for uorescence with DAPI (H
1200; Vector Laboratories, Inc. Burlingame, CA, USA). Negative control
slides were stained by the same procedure, with the primary antibody
omitted. Images were obtained from a minimum of 6 different slides
from each rat group with a Leica LMD6000 uorescence microscope
(Leica Microsystems, Germany).
1.2. Stillborn and neonatal human lung samples
Blocks of lung tissue taken at postmortem examination, between
1985 and 2012, were obtained from the pathology department of the
hospital. The inclusion criteria for the EA-TEF group were: fetuses of
33 or more weeks of gestational age, with the malformation but without
primary lung disease and/or bilateral renal anomalies, oligohydramnios
or chromosomal anomalies. Same criteria were used for selecting
the lungs for the control group except for the presence of the
tracheoesophageal malformation. Twenty lung samples, 8 from still-
birth and 12 from deceased newborns were examined. The gestational
age ranged from 33 weeks to 42 weeks (with 18 of N 33 weeks gestation
and both groups with comparable gestational ages).
1.2.1. Immunouorescence staining
The same protocol and antibody were used to assess Sp-B immuno-
reactivity in the human 5 m lung sections. For VEGF and -smooth
muscle actin a double immunouorescence staining was carried out
with the following primary antibodies: anti-VEGF rabbit polyclonal an-
tibody (07-1420, Merck Millipore, Germany) 1:400 and anti--smooth
muscle actin mouse monoclonal antibody (1A4:sc-32251; Santa Cruz
Biotechnologies, Santa Cruz, CA, USA) 1:100; the secondary antibodies
used were Alexa Fluor 488 goat anti-rabbit (A-11034; Molecular
Probes, Life Technologies, USA) and Alexa Fluor 594 rabbit anti-mouse
(A-11062; Molecular Probes, Life Technologies, USA). The sections
were mounted and the nuclei counterstained using Vectashield Mounting
Medium for uorescence with DAPI (H 1200; Vector Laboratories, Inc.
Burlingame, CA, USA). Negative control slides were stained by the same
procedure, with the primary antibody omitted. Images were obtained
 A.C. Fragoso et al. / Journal of Pediatric Surgery 50 (2015) 12511259 1253

Fig. 1. a. Lung wet/dry weight ratio and puried total lung DNA. The lung wet/dry weight ratio was signicantly higher in the adriamycin-exposed groups with the highest ratio
showed by the adria EA-TEF lungs even in comparison with the adria noEA group [control (n = 21; 6.62 0.4) vs adria EA-TEF (n = 16; 8.85 1.3) * p b 0.0001; control vs adria
noEA (n = 8; 7.57 0.8) * p = 0.005; adria EA-TEF vs adria noEA * p = 0.01]. The puried total DNA content in the adriamycin-exposed groups was signicantly lower than in
controls. A markedly lower total DNA was quantied in the adria EA-TEF lungs [control (n = 9; 5.31 0.8) vs adria EA-TEF (n = 9; 3.14 0.3) * p b 0.0001; control vs adria noEA
(n = 8; 4.06 0.8) * p = 0.01; adria EA-TEF vs adria noEA * p = 0.04]. b. Real time RT-PCR analysis of expression of TTF1 and FGF7 mRNA. The adriamycin-exposed groups (EA-
TEF and adria noEA lungs) exhibited a similar prole of TTF1 mRNA expression during the studied time points, with a signicant overexpression at the canalicular stage (E18) in
contrast to the control group [control (n = 10; 0.95 0.2) vs adria EA-TEF (n = 10; 1.26 0.3) * p = 0.02; control vs adria noEA (n = 10; 1.31 0.4) * p = 0.04]. Regarding FGF7
mRNA, the adria EA-TEF lungs showed a signicantly higher expression at the pseudoglandular stage of development compared to control and adria noEA lungs [adria EA-TEF
(n = 8; 1.81 0.6) vs control (n = 10; 1.07 0.5) * p b 0.01; adria EA-TEF vs adria noEA (n = 9; 1.06 0.2) * p = 0.03]. c. Protein expression of TTF1, Sp-B, VEGF and -sma
protein on E21 (Western blot). At the end of the gestation, there were no signicant differences of expression in the TTF1, Sp-B, VEGF and -smooth muscle actin protein although the
esophageal atresia with tracheoesophageal lungs showed a relatively lower Sp-B and VEGF protein expression [TTF1: control n = 15 (0.49 0.2), EA-TEF n = 17 (0.46 0.1), adria noEA
n = 18 (0.53 0.2). Sp-B: control n = 22 (0.49 0.2), EA-TEF n = 18 (0.40 0.2); adria noEA n = 21 (0.42 0.2). VEGF: control n = 14 (0.45 0.3), EA-TEF n = 14 (0.37 0.2), adria
noEA n = 20 (0.44 0.2). -sma: control n = 11 (1.09 0.5), EA-TEF n = 12 (1.12 0.4), adria noEA n = 17 (1.14 0.3)].
1254 A.C. Fragoso et al. / Journal of Pediatric Surgery 50 (2015) 12511259

Fig. 2. Immunouorescent localization of TTF1, Sp-B, VEGF and -sma antibodies in control (a), adria EA-TEF (b) and adria noEA (c) rat lungs at E21. 2.1. TTF1 protein: Nuclear TTF1 immunore-
activity was noticed in the alveolar and bronchiolar epithelial cells in a quite uniform pattern in all groups (20). 2.2. Sp-B protein: No differences among groups were observed in the cytoplasmic
Sp-B staining pattern (20). 2.3. VEGF protein: Cytoplasmic VEGF immunoreactivity was observed mainly in the bronchial epithelium of both groups. Some scattered immunoreactivity was
detected in mesenchymal and alveolar epithelial cells in control lungs in contrast to the adriamycin-exposed groups (20). 2.4. -sma protein: All groups showed clear immunoreactivity for
-smooth muscle actin antibody in the airway wall and pulmonary arteries/arterioles. There were no relevant differences between control, adria EA-TEF or adria noEA lungs (20).

from a minimum of 6 different slides from each lung specimen with a TEF ones. Conversely, total DNA was signicantly decreased in
Leica LMD6000 uorescence microscope (Leica Microsystems, Germany). adriamycin-exposed lungs and particularly in the EA-TEF group.

1.2.2. Alcian-blue staining protocol 2.1.2. Real time RT-PCR analysis of expression of TTF1 and FGF7 mRNA
Sections of 5 m from lung specimens from control and EA-TEF new- These results are shown in Fig. 1b. At E15, the adria EA-TEF lungs
borns were stained following the standard protocol. Images were expressed a signicantly higher FGF7 mRNA and at E18, both EA-TEF
obtained from 6 or more slides from each lung. and adria noEA lungs signicantly overexpressed TTF1 mRNA.

2.1.3. Protein expression (Western blot analysis)

1.3. Statistical methods
Data are presented as mean SD (standard deviation). Comparison
between groups was performed with U MannWhitney test or unpaired
t test where appropriate. The statistical signicance level was set at 5%.
2. Results
2.1. Rat experiments
2.1.1. Lung wet/dry weight ratio and Puried Total DNA quantication
As shown in Fig. 1a, the lung wet/dry weight ratio was signicantly
higher in the adriamycin-exposed lungs and particularly in adria EA-
These results are also shown in Fig. 1c.
Control, adria EA-TEF and adria noEA lungs showed similar levels of
TTF1 and -sma protein expression. Despite the absence of signicant
differences in Sp-B protein expression between groups, the adriamycin-
exposed lungs tended to express less Sp-B content. The expression of
VEGF protein was slightly lower in the lungs from fetuses with EA-TEF
than those found in the control and adria noEA groups although without
statistical signicance.
2.1.4. Immunolocalization
Nuclear TTF1 immunoreactivity was uniformly distributed in the
alveolar and bronchiolar epithelial cells in all groups (Fig. 2.1 a, b, c).
No relevant differences among groups were observed in the
 A.C. Fragoso et al. / Journal of Pediatric Surgery 50 (2015) 12511259 1255

Table
Clinical data from fetuses and newborns, with (EA group) and without (control group) EA-TEF included in the study.

EA Fetal age (wk) Malformation Origin Body weight (g)

1 35 EA-TEF In-utero death 1440

2 38 EA-TEF/VACTERL In-utero death 1200
3 39 EA-TEF In-utero death 2000
4 39 EA-TEF In-utero death 1340
5 33 EA-TEF Postnatal death (1st day) 1450
6 36 EA-TEF/VACTERL Postnatal death (1st day) 1980
7 42 EA-TEF/cardiopathy Postnatal death (1st day) 3050
8 35 EA-TEF Postnatal death (3th day) 1200
9 36 EA-TEF/VACTERL Postnatal death (5th day) 2200
10 35 EA-TEF/CHARGE Postnatal death (9th day) 2000

Control Cause of death

1 37 Abruptio placentario In-utero death 3420

2 38 Maternal hypertension In-utero death 2520
3 38 Maternal diabetes/cardiopathy In-utero death 2300
4 40 Umbilical cord circular (5) In-utero death 3360
5 35 Fetal hypoxia Postnatal death (1st day) 3100
6 39 Fetal hypoxia Postnatal death (2th day) 4020
7 33 Fetal hypoxia Postnatal death (5th day) 1660
8 38 Erythroblastosis Postnatal death (5th day) 2220
9 39 Fetal hypoxia Postnatal death (5th day) 3250
10 39 Massive intestinal volvulus Postnatal death (5th day) 2680

cytoplasmic Sp-B staining pattern although immunoreactivity was
more disperse in adriamycin-exposed lungs (Fig. 2.2 a, b, c).
Cytoplasmic VEGF immunoreactivity was observed mainly in the
bronchial epithelium although some scattered uptake was present in
mesenchymal and alveolar epithelial cells of control lungs. In the
adriamycin-exposed lungs there was only epithelial staining (Fig. 2.3
a, b, c). All groups showed similar immunoreactivity for -sma antibody
in the airway wall and pulmonary arteries/arterioles (Fig. 2.4 a, b, c).
2.2.3.3. -sma. A similar staining pattern was observed in the airway
walls and the tips of alveolar septa. Immunoreactivity detected in
arterioles was similar in both groups (Fig. 4.2 a, b).
2.2.4. Alcian-blue staining of lung specimens from control and EA-TEF newborns
As depicted in Fig. 5, lungs from newborns with EA-TEF exhibited
denitely more mucous glands and goblet cells in the airway epithelium
than controls.

2.2. Stillborn and neonatal human lung specimens 3. Discussion

Clinical data are presented in the Table.
2.2.1. Immunolocalization
These results are depicted in Fig. 3 (a, control; b, EA-TEF).
2.2.2. Lungs from stillborns
2.2.2.1. Sp-B. Scarce immunoreactivity was detected in the lungs of still-
borns with EA-TEF in contrast to the control lungs that exhibited cytoplas-
mic Sp-B staining in the peripheral alveolar and epithelial cells (Fig. 3.1 a, b).
2.2.2.2. VEGF. A strong VEGF immunoreactivity was detected in the pul-
monary arterioles of the EA-TEF lungs in contrast with a minor signal
noticed in the bronchial epithelial cells. The lungs of stillborns without
the malformation showed an intense VEGF expression in the bronchial
epithelial cells and an almost inexistent staining in the pulmonary
vasculature (Fig. 3.2 a, b).
2.2.2.3. -sma. The expressed -sma in the airway walls was similar in
both groups (Fig. 3.3 a, b).
2.2.3. Lungs from newborns
These results are shown in Fig. 4 (a, control; b, EA-TEF).
2.2.3.1. Sp-B. Less cytoplasmatic immunoreactivity was detected in the
lungs of newborns with EA-TEF specically in the alveolar and bronchial
epithelial cells (Fig. 4.1 a, b).
2.2.3.2. VEGF. Epithelial VEGF immunoreactivity was moderately
increased in EA-TEF specimens (Fig. 4.2 a, b).
Accumulating evidence suggests that lung damage during early
development, may result in persistent structural changes and impaired
function during postnatal life. Although the respiratory symptoms
suffered by children born with EA-TEF may be explained by GER,
tracheomalacia and surgical complications, or even by intrauterine
growth retardation or prematurity, several long term studies on respirato-
ry function in children, adolescents or adults with the malformation did
not nd a clear correlation between the abovementioned conditions and
some of the respiratory symptoms [5,14]. We postulate the hypothesis
that these might have a fetal origin since disturbed regulation during cri-
tical periods of early life may lead to developmental adaptations potentially
responsible for permanent structural, physiological and epigenetic changes
with lifetime consequences [15]. Additionally, airway abnormalities are de-
termined very early in life, even in utero, and these will increase the risk of
subsequent bronchoconstriction and asthma in early childhood, predispo-
sing to decient lung function. Finally, and not less crucial for normal post-
natal lung function is a normally-functioning host defense mechanism
since maladaptive responses of the newborn lung may trigger subsequent
lung injury resulting in further disruption to the growing organ [15].
The complex phenomenon of lung development incorporates two
related but differently controlled processes, i.e., lung growth and lung
maturation [16]. The experiments shown herein revealed that
adriamycin-exposed lungs were markedly hypoplastic as demonstrated
by increased wet/dry weight ratios and decreased total DNA content
that reveal decreased number of cells with increased water contents.
It is relevant to point out that these changes were signicantly more
pronounced in the lungs of rats with EA-TEF, suggesting that the EA-
TEF malformation itself contributes to the defective lung growth.
Transcriptional control of differentiation genes and epithelial-
mesenchymal interactions mediated by growth factors are critical for
lung development modulating airway branching and concurrent capillary
1256 A.C. Fragoso et al. / Journal of Pediatric Surgery 50 (2015) 12511259

Fig. 3. Immunouorescent localization of Sp-B, VEGF and -sma antibodies in lungs from control (a) and EA-TEF (b) stillborns. 3.1. Sp-B: Poor Sp-B immunoreactivity was detected in the
lungs of stillborns with EA-TEF in contrast with the control lungs that exhibited cytoplasmic Sp-B staining in the peripheral alveolar and epithelial cells (20). 3.2. VEGF: A strong VEGF
immunoreactivity was detected in the pulmonary arterioles of the EA-TEF lungs in contrast to a scarce signal noticed in the bronchial epithelial cells. The lungs of stillborns without the
malformation showed intense VEGF expression in the bronchial epithelium and almost absent staining in the pulmonary vasculature (20). 3.3. -sma: The -smooth muscle actin
expression in the airway walls was similar in both groups (20).

network development. The emergence of the tracheobronchial anlage
from the oor of the primitive foregut requires coordinated transcrip-
tional activation and repression of several key developmental genes
such as HNF-3b, Shh, Ptch, Gli2 and Gli3 as well as TTF1 (Nkx2.1) [17].
TTF1 functions as a lung master gene playing distinct roles at various
stages of development and is required for early patterning of the anteri-
or foregut, growth and differentiation of the primordial lung during em-
bryonic period, maturation of the respiratory epithelium before birth
and synthesis of pulmonary surfactant [17,18]. In addition, mutations in
TTF1 are associated with tracheoesophageal stula, lung dysmorphogenesis,
and respiratory failure at birth [19].
In the present experiments, adriamycin-exposed lungs overexpressed
TTF1 mRNA in the canalicular stage of development. Because this stage is
characterized by the initiation of differentiation of the epithelial cells
lining the ducts, capillary growth and the rst appearance of type II
cells with lamellar bodies containing lung surfactant [20], a likely
 A.C. Fragoso et al. / Journal of Pediatric Surgery 50 (2015) 12511259 1257

Fig. 4. Immunouorescent localization of Sp-B, VEGF and -sma antibodies in control (a) and EA-TEF (b) newborn lungs. 4.1. Sp-B: Less cytoplasmic immunoreactivity was detected in the
lungs of newborns with EA-TEF specically in the alveolar and bronchial epithelial cells (20). 4.2. Immunouorescence colocalization of VEGF and -sma: In newborns with EA-TEF, the
epithelial VEGF immunoreactivity was increased and a uniform staining pattern of -sma in the airway walls and the tips of alveolar septa was observed. Immunoreactivity detected in
arterioles was similar in both groups (20).

interpretation would be that the hypoplastic lungs modulate TTF1 tran-
scriptional signal in order to accelerate differentiation and achieve normal
epithelial maturation and function at term through epithelium-specic
gene expression [21].
Along with transcription factors, broblast growth factors (FGF) also
intervene in both early and late lung development in a way that distur-
bance in FGF signaling during embryonic organogenesis results in obvi-
ous abnormalities of epithelial branching and cyto-differentiation [17].
Two of the most relevant FGFs involved in lung morphogenesis are
FGF10 and FGF7 (keratinocyte growth factor) that share the same
mesenchymal FGFR2IIIb receptor. Indeed, abnormal FGF10 signaling
has been reported in the lungs from rat fetuses with EA-TEF [13]. FGF7
stimulates lung uid secretion that is critical for lung branching and
growth and latter on, promotes maturation of fetal alveolar type II
cells and consequently surfactant lipid and protein expression [17,18].
The signicant overexpression of FGF7 mRNA observed in the lungs
from fetuses with EA-TEF at the pseudoglandular stage, at the time
when branching of the bronchiolar epithelium is active (whereas the
1258 A.C. Fragoso et al. / Journal of Pediatric Surgery 50 (2015) 12511259
Fig. 5. Alcian-blue staining of lung specimens from control (a) and EA-TEF (b) newborns. Lungs from newborns with EA-TEF exhibited more submucosal mucous glands and goblet cells in
the airway epithelium (10).
ductular cells remain relatively undifferentiated) [19], suggests an
enhanced stimulus to the branching process and uid secretion to com-
pensate for the marked hypoplasia in the lung and for the probable
unbalanced uid dynamics/intrapulmonary pressures owing to the
tracheoesophageal stula.
As far as lung maturity near term is concerned, TTF1 was used in this
investigation as a marker of differentiation of epithelial cells of the de-
veloping airways, Sp-B as a pneumocytes type II and, to a lesser extent,
a Clara cells marker and VEGF (vascular endothelial growth factor) as
vasculogenesis and angiogenesis marker also important for the growth
of alveolar cells as for surfactant production [22]. Finally, -smooth
muscle actin was used as a myobroblast and smooth muscle cells
differentiation essential to provide tone to maintain the lung liquid at
a positive pressure to stimulate prenatal growth and stabilize lung
cellular architecture [23,24].
As judged by the expression of TTF1, Sp-B, VEGF and -smooth mus-
cle actin proteins, the lungs of control and adriamycin-exposed fetuses
behave normally demonstrating biochemical markers consistent with
overall mature organs. Nevertheless, the lungs of fetuses with EA-TEF
showed minimally reduced levels of expression of Sp-B and VEGF to-
gether with somewhat different patterns of localization (disperse Sp-B
and scarce mesenchymal VEGF) that despite the lack of statistical signi-
cance, might suggest some minor disturbance. In summary, despite
being markedly hypoplastic, the lungs of rats with EA-TEF were able
to overexpress TTF1 and FGF7 in different stages of lung development
until achieving full lung maturity.
The human necropsy results in the present study require more cau-
tious interpretation. Transposition of animal evidence to the equivalent
human condition is always questionable. There are obvious species dif-
ferences, the stillborns and newborns studied had different ages and
sizes and the tissue handling, xation, sectioning and staining of nec-
ropsy material lack the controlled conditions met in the experimental
laboratory. For obvious reasons, growth of the lung could not be
assessed. However, some data on differentiation and maturation could
be analyzed: the normal expression of alpha smooth muscle actin in
the lungs from stillborns with EA-TEF contrasts with a noticeable de-
crease of Sp-B and epithelial VEGF proteins accompanied by increased
endothelial VEGF immunoreactivity suggest a disturbed epithelial dif-
ferentiation status and consequently some degree of relative lung im-
maturity. Increased VEGF expression in the small pulmonary arteries
may reveal an attempt to stimulate the pulmonary vascular bed, a
nding previously described in newborns with congenital diaphrag-
matic hernia in which the lungs are severely hypoplastic [25]. The
lungs of newborns with EA-TEF also showed a trend toward decreased
Sp-B content and an increased airway epithelial VEGF protein expres-
sion. These abnormal expressions may impair lung function per se and
inuence postnatal adaptation but they may also cause greater suscep-
tibility to injury and thereby disturb lung growth. In fact, Sp-B deciency
may induce alveolar collapse and respiratory distress at birth. In turn,
VEGF levels are increased in acute respiratory distress, chronic pulmo-
nary obstructive disease and asthma [26] and its overexpression is
associated with mucosal edema, hypervascularity and airway hyperre-
activity [27]. Finally, the denitely higher density of goblet cells and mu-
cous glands observed in the lungs of newborns with EA-TEF suggests
increased mucus production that, together with squamous metaplasia
of the tracheal epithelium and reduced ciliary bed previously described
 A.C. Fragoso et al. / Journal of Pediatric Surgery 50 (2015) 12511259
by Emery JL et al. could account for the frequent respiratory infections,
chronic cough and asthma-like symptoms observed in EA-TEF survivors
as mucus production is a primary defense mechanism for maintaining
healthy lungs through clearance of particles and pathogens from the
airways by cilia beating within the periciliary layer [2831].
In summary, putting together the ndings in rats and human indi-
viduals with EA-TEF, there were no changes in the amount and distribu-
tion of alpha smooth muscle actin, there were various degrees of
decrease of surfactant B protein and nally, epithelial VEGF was also de-
creased in both rats and stillborns with the malformation, but it was in-
creased in affected newborns. Newborns with EA-TEF also showed
increased numbers of goblet cells and mucous glands in their airways.
Some of these differences could be accounted for by the different
stage of lung development at term in both species, since the alveolar
phase begins postnatally in rodents. Furthermore, the discrepancies
observed between stillborn and newborn lungs may be explained by a
more severe and therefore fatal outcome in the former group.
In conclusion, experimental results had some resemblance to human
ndings. The subtle differences in the expression of the Sp-B and VEGF
epithelial markers observed in fetuses and newborns with the malfor-
mation could inuence the adaptative response and host defense of
the lung facilitating respiratory infections, chronic cough, and asthma-
like symptoms or wheezing. Further studies are needed in order to
better understand the pathophysiology involved in the respiratory
morbidity observed in survivors to neonatal repair of EA-TEF.
Acknowledgments
We are grateful to Dr. Ignacio Rodriguez from the Department
of Pathology (La Paz Hospital, Madrid) for his help in the human
specimens selection.
References
[1] Spitz L. Oesophageal atresia. Orphanet J Rare Dis 2007;11:224.
[2] Montedonico S, Dez-Pardo JA, Lassaletta L, et al. Respiratory malformations associated
with esophageal atresia. Cir Pediatr 1999;12:614.
[3] de Jong EM, Felix JF, Deurloo JA, et al. Non-VACTERL-Type anomalies are frequent in
patients with esophageal atresia/tracheo-esophageal stula and full or partial
VACTERL association. Birth Defects Res A Clin Mol Teratol 2008;82:927.
[4] Usui N, Kamata S, Ishikawa S, et al. Anomalies of the tracheobronchial tree in
patients with esophageal atresia. J Pediatr Surg 1996;31:25862.
[5] Legrand C, Michaud L, Salleron J, et al. Long-term outcome of children with
oesophageal atresia type III. Arch Dis Child 2012;97:80811.
[6] Gadzinsky V, Jnsson L, Ekerljung L, et al. Long-term respiratory symptoms following
oesophageal atresia. Acta Pediatr 2011;100:12225.
1259
[7] Sistonen SJ, Pakarinen M, Rintala R. Long-term results of esophageal atresia: Helsinki
experience and review of literature. Pediatr Surg Int 2011;27:11419.
[8] Lilja HE, Wester T. Outcome in neonates with esophageal atresia treated over the last
20 years. Pediatr Surg Int 2008;24:5316.
[9] Sistonen S, Malmberg P, Malmstrm K, et al. Repaired oesophageal atresia:
respiratory morbidity and pulmonary function in adults. Eur Respir J 2010;36:
110612.
[10] Diez-Pardo JA, Baoquan Q, Navarro C, et al. A new rodent experimental model of
esophageal atresia and tracheoesophageal stula: preliminary report. J Pediatr
Surg 1996;31:498502.
[11] Ioannides AS, Chaudhry B, Henderson DJ, et al. Dorsoventral patterning in oesopha-
geal atresia with tracheo-esophageal stula: evidence from a new mouse model. J
Pediatr Surg 2002;37:18591.
[12] Xiaomei L, Aras-Lopez R, Martinez L, et al. Lung hypoplasia in rats with esophageal
atresia and tracheo-esophageal stula. Pediatr Res 2012;71:23540.
[13] Fragoso AC, Aras-Lopez R, Martinez L, et al. Abnormal control of lung branching in
experimental esophageal atresia. Pediatr Surg Int 2013;29:1717.
[14] Calisti A, Oriolo L, Nanni L, et al. Mortality and long term morbidity in esophageal
atresia: the reduced impact of low birth weight and maturity on surgical outcome.
J Perinat Med 2004;32:1715.
[15] Narang I, Bush A. Early origins of chronic obstructive pulmonary disease. Semin Fetal
Neonatal Med 2012;17:1128.
[16] DiFiore LW, Wilson JM. Lung development. Semin Pediatr Surg 1994;3:2213.
[17] Kumar VH, Ryan RM. Growth factors in the fetal and neonatal lung. Front Biosci
2004;9:46480.
[18] Whitsett JA, Matsuzaki Y. Transcriptional regulation of perinatal lung maturation.
Pediatr Clin North Am 2006;53:8737.
[19] Whitsett JA. Disorders of lung morphogenesis. Pediatr Res Rev 2006;7(Suppl 1):
S248.
[20] Mendelson CR. Role of transcription factors in fetal lung development and surfactant
preotein gene expression. Annu Rev Physiol 2000;62:875915.
[21] Stahlman MT, Gray ME, Whitsett JA. Expression of thyroid transcription factor-1
(TTF1) in fetal and neonatal human lung. J Histochem Cytochem 1996;44:6738.
[22] Ad hoc statement committee, American Thoracic Society. Mechanisms and limits of
induced postnatal growth. Am J Respir Crit Care Med 2004;170:31943.
[23] Sparrow MP, Lamb JP. Ontogeny of airway smooth muscle: structure, innervation,
myogenesis and function in the fetal lung. Respir Physiol Neurobiol 2003;137:
36172.
[24] Yamada M, Kurihara H, Kinoshita K, et al. Temporal expression of alpha-smooth
muscle actin and debrin in septal interstitial cells during alveolar maturation. J
Histochem Cytochem 2005;53:73544.
[25] Shehata SM, Mooi WJ, Okazaki T, et al. Enhanced expression of vascular endothelial
growth factor in lungs of newborns with congenital diaphragmatic hernia and
pulmonary hypertension. Thorax 1999;54:42731.
[26] Papaioannou AI, Kostikas K, Kollia P, et al. Clinical implications for vascular endothe-
lial growth factor in the lung: friend or foe? Respir Res 2006;17:128.
[27] Voelkel NF, Vandivier RW, Tuder RM. Vascular endothelial growth factor in the lung.
Am J Physiol Lung Cell Mol Physiol 2006;290:20921.
[28] Evans CM, Koo JS. Airway mucus: the good, the bad, the sticky. Pharmacol Ther
2009;121:33248.
[29] Nadel JA. Mucous hypersecretion and relationship to cough. Pulm Pharmacol Ther
2013;26:5103.
[30] Emery JL, Haddadin AJ. Squamous epithelium in the respiratory tract of children
with tracheo-esophageal stula. Arch Dis Child 1971;46:23642.
[31] Ordoez CL, Khashayar R, Wong HH, et al. Mild and moderate asthma is associated
with airway goblet cell hyperplasia and abnormalities in mucin gene expression.
Am J Respir Crit Care Med 2001;163:51723.