Neonatal exposure to hyperoxia leads to persistent disturbances in pulmonary histone signatures associated with NOS3 and STAT3 in a mouse model
Chao et al. Clinical Epigenetics
Neonatal exposure to hyperoxia leads to persistent disturbances in pulmonary histone signatures associated with NOS3 and STAT3 in a mouse model
Cho-Ming Chao 3 4
Rhea van den Bruck 0
Samantha Lork 0
Janica Merkle 0
Laura Krampen 0
Patrick P Weil 0
Malik Aydin 0
Saverio Bellusci 4
Andreas C. Jenke 0 1 2 5
Jan Postberg 0 1
0 Department of Pediatrics, HELIOS Medical Center Wuppertal, Center for Clinical & Translational Research (CCTR), Center for Biomedical Education & Research (ZBAF), Witten/Herdecke University , Wuppertal , Germany
1 Equal contributors
2 EKO Children's Hospital , Oberhausen , Witten/Herdecke University , Alfred-Herrhausen Str. 40, Witten , Germany
3 University Children's Hospital Gießen, Division of General Pediatrics and Neonatology, Justus-Liebig-University , Gießen , Germany
4 Excellence Cluster Cardio-Pulmonary System (ECCPS), Member of the German Center for Lung Research (DZL), Department of Internal Medicine II, Universities of Giessen and Marburg Lung Center (UGMLC) , Aulweg 130, 35392 Giessen , Germany
5 EKO Children's Hospital , Oberhausen , Witten/Herdecke University , Alfred-Herrhausen Str. 40, Witten , Germany
Background: Early pulmonary oxygen exposure is one of the most important factors implicated in the development of bronchopulmonary dysplasia (BPD). Methods: Here, we analyzed short- and long-term effects of neonatal hyperoxia on NOS3 and STAT3 expression and corresponding epigenetic signatures using a hyperoxia-based mouse model of BPD. Results: Early hyperoxia exposure led to a significant increase in NOS3 (median fold change × 2.37, IQR 1.54-3.68) and STAT3 (median fold change × 2.83, IQR 2.21-3.88) mRNA levels in pulmonary endothelial cells with corresponding changes in histone modification patterns such as H2aZac and H3K9ac hyperacetylation at the respective gene loci. No complete restoration in histone signatures at these loci was observed, and responsivity to later hyperoxia was altered in mouse lungs. In vitro, histone signatures in human aortic endothelial cells (HAEC) remained altered for several weeks after an initial long-term exposure to trichostatin A. This was associated with a substantial increase in baseline eNOS (median 27.2, IQR 22.3-35.6) and STAT3α (median 5.8, IQR 4.8-7.3) mRNA levels with a subsequent significant reduction in eNOS expression upon exposure to hypoxia. Conclusions: Early hyperoxia induced permanent changes in histones signatures at the NOS3 and STAT3 gene locus might partly explain the altered vascular response patterns in children with BPD.
Bronchopulmonary dysplasia (BPD) is a chronic lung
disease of prematurely born infants and remains a
leading cause of morbidity and mortality. Currently, there is
no curative therapy available. Based on the
severitybased definition of BPD (inclusion of infants with mild
BPD), 68% of premature infants born with a gestational
age (GA) ≤ 28 weeks develop BPD [
]. The risk of
developing BPD correlates inversely with the gestational
age (GA) and birth weight (BW) [
]. Since premature
infants are born with a lung which is in the canalicular or
saccular stages of development, the lung structure is
therefore not adequate to provide sufficient ventilation
and gas exchange. Thus, mechanical ventilation and high
oxygen concentration are often necessary at birth.
Barotrauma induced by mechanical ventilation as well as
oxygen toxicity and inflammation are major contributing
factors responsible for the pulmonary damages in the
immature lung. In addition, some studies have suggested a
strong genetic component in BPD [
]. In fact, clinical
studies suggest that decisions during the first minutes of
] or even events before the delivery might be crucial
for the later development of BPD. In line with these
clinical data are observations that hyperoxia in rat pups leads
to increased endothelilal nitric oxide synthase (eNOS)
levels, nitric oxide (NO) activity, hyperemia [
possibly eNOS uncoupling eventually leading to BPD.
Importantly, BPD does not only affect neonatal
mortality but also leads to long-term morbidity, e.g.,
increased susceptibility to upper respiratory infections
during the first year of life [
] and pulmonary
]. Again, neonatal oxygen seems to be an
important pathological factor since it increases for example
the sensitivity to influenza A virus infection by
suppressing epithelial expression of Ear1 [
]. In fact, extremely
and moderately preterm infants face a 3.6 times
increased risk of being hospitalized due to respiratory
infection in the first year of life [
Whereas it seems clear from twin studies that genetic
susceptibility plays an important role for disease
], the exact mechanism remains unclear but
epigenetic mechanisms seem to be part of it since
several recent publications have reported abnormalities of
histone acetylase activity and the chromatin remodeling
pathway in BPD patients [
]. Thus, persistent
changes in epigenetic signatures might be also at least
partly responsible for the later development of
pulmonary hypertension and the increased susceptibility to
respiratory tract infections. Since we have recently shown
that epigenetic signatures at NOS3 encoding eNOS in
human umbilical artery endothelial cells are substantially
shaped by prenatal events such as placental insufficiency
and that there exists an interdependency between NOS3
and the gene activity of signal transducer and activator
of transcription 3 alpha (STAT3α) as well as
Stat3transcription factor-binding in the NOS 3 promoter.
Moreover, in human, a 27-nt non-coding RNA becomes
co-expressed with NOS3 and post-transcriptionally
regulates STAT3 expression . We thus aimed to utilize
the same gene loci to analyze the effect of early neonatal
oxygen exposure on the pulmonary vascular endothelial
epigenetic landscape and the associated consequences
for oxygen exposure later in life.
Therefore, we analyzed NOS3 gene expression in
response to early neonatal hyperoxia at first and then
studied the respective changes in epigenetic signatures as
consequences of this first early oxygen exposure on later
events of hyperoxia and hypoxia later in life.
Wildtype mice (males and females, CD1 background)
were crossed to generate time-pregnant females. All
mice received food and water ad libitum.
Hyperoxia injury (BPD mouse model)
Newborn pups were exposed to either normoxia (NOX)
or hyperoxia (HOX) within 24 h after birth (P0) (Fig. 1).
In experimental group 1, dams and pups were kept in
NOX from P0 to P16. In experimental group 2, dams
and pups were kept in NOX from P0 to P15. From P15
to P16, the dams and the pups were exposed to HOX
(85% O2). In experimental group 3, newborn pups were
subjected to HOX (85% O2) injury from P0 to P8 in a
chamber (Proox Model 110, Biospherix). To minimize
oxygen toxicity and bias, nursing dams were rotated
every 24 h between NOX and HOX. Afterwards, nursing
dams and pups were exposed to NOX (21% O2) from P8
to P16. From P15 to P16, the dams and the pups were
re-exposed to HOX (85% O2). In experimental groups 1
and 3, lungs were harvested at P8. In all groups, lungs
were harvested at P16. All dams and pups received food
and water ad libitum.
Lung perfusion, isolation, tissue processing, and histology
Mice were euthanized by intraperitoneal injection of a
solution made of ketamin, dormitor, heparin, and saline.
After sternotomy, the lung was perfused transcardially
(through the right ventricle) by using PBS (1×), then
isolated and incubated for 30 min at 4 °C in COLD
medium, stored at − 20 °C overnight, and finally stored
at − 80 °C till further tissue processing.
For histology, the lung was flushed from the right
ventricle to remove blood cells then perfused through the
trachea with a pressure of 20 cm H2O with 5 ml 4%
PFA. The trachea was tied off with a string, and the lung
was removed and placed in 4% PFA for max. 24 h at 4 °
C. Lungs were then progressively dehydrated (30, 50, 70,
and 99% ethanol, 3 h each), incubated in xylole, then in
paraffine overnight and finally embedded with a Leica
embedding machine (EG 1150C). Paraffin blocks were
kept cold, and 5 μm sections were cutted. Hematoxylin
and eosin staining was performed according to protocols
Positive selection for vascular endothelial cells
Positive selection for vascular endothelial cells was
performed by magnetic separation with human CD146
Microbeads (Miltenyi®, Bergisch-Gladbach, Germany)
following manufacturers’ instructions with minor
modifications as described previously [
Total RNA was isolated using Trizol (Sigma-Aldrich)
and isopropanol precipitation and further purification
on columns. Next, RNA integrity was assayed using the
Agilent Bioanalyzer 2000. Only samples with
nonfragmented RNA were included.
Gene expression analyses
Gene expression analyses were performed using
quantitative real-time PCR (qPCR) analyses on a Rotor-Gene
6000 (Qiagen). For PCR reactions, QuantiTect SYBR
Green qPCR Master Mix (Qiagen) containing Hot Start
Taq DNA polymerase and SYBR Green was used.
Primers were used as described earlier [
expression of gene of interest was normalized using three
housekeeping genes (BACT, GAPDH, PECAM1). PCR
conditions were as follows: 95 °C for 15 min, 40× [95 °C
for 15 s, 60 °C for 30 s]. Melting of PCR product was
done using a gradient from 55 to 95 °C rising in 0.5 °C
Primary antibodies used in this study were (1) Rabbit
anti-H2A.Zac (Diagenode pAb-173-050), (2) Rabbit
antiH3K9ac (Active Motif pAb#39137), and (3) Rabbit
antiH3K4me3 (Diagenode pAbCSP-030-050).
Chromatin purification and ChIP assays
Chromation purification and ChIP assays were
performed as described previously [
PCR analyses were performed using a Rotorgene 6000
(Qiagen). The relative amounts of specifically
immunoprecipitated DNA were estimated as “percent of input”
and quantified using individual standard curves for each
amplicon. Primer pairs were used as described earlier
DNA methylation signatures of promoter segments of
NANOG, NFE2L2, and STAT3 were analyzed using a
Qiagen Pyromark Q24 sequencer as previously described
(Plos One Jenke et al.). Briefly, after standard sodium
bisulfite conversion using the EZ DNA Methylation-Gold
Kit (Zymo Research, USA) pyrosequencing methylation
analysis was conducted using the PyroMark Q24
(Qiagen, Germany) according to the manufacturer’s
protocol. Therefore, we designed and made use of the
following oligonucleotides: 1. for NFE2L2—primer F1
(5′- gga gtt aga ggg gat agt ggt t-3′), 5′-biotinylated
primer (5′-acc cca cca aat caa aac ttc ct-3′), and
sequencing primer S1 (5′-agg gta aag gag gat g-3′); 2. for
STAT3—primer F1 (5′-ggt gta ggg tgg ggt tat t-3′),
5′biotinylated primer (5′-acc cta tat atc tcc tcc tat cct-3′),
and sequencing primer S1 (5′-ggg tgg ggt tat ttt t-3′); 3.
for NANOG (DNA methylation positive
control)—primer F1 (5′-gta gga tag gaa tgg ggg ttg-3′),
5′-biotinylated primer (5′-acc tta aat tta ccc caa att cta c-3′), and
sequencing primer S1 (5′-aat ggg ggt tgg gga-3′). No
reliable NOS3 DNA methylation assays meeting our
quality standards could be designed. The level of
methylation was analyzed using PyroMark Q24 2.0.6 Software
(Qiagen). Non-CpG cytosine residues and a standard
fully methylated DNA (Zymo Research, USA) were used
as controls to verify bisulfite conversion.
Small RNA-seq and analyses pipeline
Total RNA was purified as described above. For
multiplexing, we made use of different multiplex sequencing
barcodes for sequencing in a single lane as described
]. Briefly, total RNA was separated by polyacrylamide
gel electrophoresis. Gel fragments corresponding to 15
to 35 nt RNA molecules were cut, and RNA was eluted.
These small RNAs were directly used for the
construction of sequencing libraries in four steps: step 1: ligation
of DNA oligos to the 3′-end of the RNA; step 2: ligation
of RNA or, respectively, chimeric RNA/DNA oligos to
the 5′-end of RNAs; step 3: cDNA library synthesis by
reverse transcriptase; and step 4: amplification of the
cDNA library. Subsequently, after final quality checks by
microcapillary electrophoresis and qPCR, the libraries
were sequenced on an Illumina Hiseq 2000 platform
(single end, 50 bp). This work has benefited from the
facilities and expertise of the high-throughput sequencing
core facility of IMAGIF Gif-sur-Yvette (Centre de
Recherche de Gif—www.imagif.cnrs.fr). The initial data
analysis pipeline was as follows: CASAVA-1.8.2 was used
for demultiplexing, Fastqc 0.10.1 for read quality
assessment and Cutadapt-1.3 for adaptor trimming, resulting
in an average sequence number for each developmental
time point sample of 7.58 Mbp. File conversions,
filtering, and sorting as well as mapping (Bowtie2) were done
using “Galaxy” [
], a platform for data intensive
biomedical research (https://usegalaxy.org/), and
Immunofluorescence staining for eNOS
Paraffin sections were deparaffinized. Antigen retrieval
was performed for 15 min in citrate buffer using a rice
cooker, then slides were cooled down on ice for 20 min.
After washing slides three times in TBST (TBS buffer +
0.1% Tween 20) for 5 min, slides were blocked with 3%
bovine serum albumin (BSA) and 0.4% Triton X-100 [in
Tris-buffered saline (TBS)] at room temperature (RT)
for 1 h and then incubated with primary antibody
against eNOS (ThermoFisher Scientific, PA1-037; 1:100)
at 4 °C overnight. Afterwards, slides were again washed
three times in TBST for 5 min, incubated with secondary
antibody (Goat Anti-Rabbit Alexa 555, ThermoFisher
Scientific, A-21429; 1:500) in room temperature for 1 h,
and then washed three times in TBST before being
mounted with Prolong Gold Anti-fade Reagent with
DAPI (4,6-diamidino-2-phenylindole; ThermoFisher
Scientific, P36931). Fluorescent images were acquired
using Leica DM5500 B fluorescence microscope
connected to Leica DFC360 FX camera.
Data were compared using Mann-Whitney U test
according to normality assumptions on univariate analysis
followed by Bonferroni correction for multivariate
analysis. Categorical variables were compared using the
Fisher exact test. Statistical analyses were performed
with GraphPad Prism 5.0.
Human arterial endothelial cells (HAEC) were bought
from PromoCell (Heidelberg, Germany) and cultivated
upon manufacturer’s recommendations. For cDNA
synthesis, we used 500 ng RNA per sample using the
QuantiTect Reverse Transcription kit (Qiagen). DNA was
isolated by phenol:chloroform:isoamylic alcohol
extraction followed by precipitation with isopropanol. For in
vitro experiments, HAEC (PromoCell) were cultivated
upon manufacturer’s recommendations. For HDAC
inhibition, HAEC were treated with 1 μM trichostatin A
(TSA) for either 72 h or twice for 6 h with an 60-h
interval in between. Hypoxia experiments were performed
using a hypoxia incubator chamber (STEMCELL
Technologies, Grenoble, France) exposing cells to a ppO2 of
0.12 bar corresponding to an oxygen fraction of 12% for
24 h. Control experiments were performed under
normoxic conditions (ppO2 = 0.21 bar).
Neonatal hyperoxia from P0 to P8 leads to increased
NOS3 expression—in both whole lung tissue and CD146
For our study, we used a well-established mouse model
for bronchopulmonary dysplasia. Newborn pups were
exposed to 85% oxygen from P0 to P8 then stayed in
normoxia from P8 to P16. Lungs were analyzed at P8
and P16. In neonatal mouse lungs, exposure to
hyperoxia (85% oxygen) from P0 to P8 led to a significant
increase in NOS3 (median fold change × 2.37, IQR
1.54–3.68, p = 0.003), GPX1 (median fold change × 1.73,
IQR 1.28–2.08, p = 0.001), and STAT3 α and β (median
fold change × 2.83, IQR 2.21–3.88, p = 0.001) mRNA
levels at P8. Other genes, such as HIF1A for example,
were not differentially expressed on the mRNA level
(Fig. 2a). At P16, 8 days after hyperoxia, mRNA levels
of GPX1 and STAT3 dropped to levels comparable to
the normoxia group at P8. NOS3 mRNA levels also
dropped but tended to be slightly above the level of
the normoxia group at P8—even though this
difference did not reach statistical significance except for
primer pair matching to exon10/11 (Fig. 2a). In
addition, we aimed to determine whether expression
profiles upon oxygen exposure differ between whole
lung tissue and pulmonary endothelial cells.
Therefore, we isolated vascular endothelial cells from whole
lung tissue using magnetic separation with human
CD146 microbeads. Results obtained from
CD146positive cell population are shown in Fig. 2b. For all
analyzed genes of interest, we did not see any
differences in expression profiles. After hyperoxia, histology
showed simplification of alveoli with increased
airspace and dilated alveoli (Fig. 3c, d) compared to
the normoxia group (Fig. 3a, b) indicating
hypoalveologenesis due to disturbed secondary septa formation.
This could be associated with the vascular
malformation seen in children with BPD [
Responsivity to later hyperoxia at P15 is altered in mouse lungs that have been exposed to hyperoxia in the neonatal period
Exposure to hyperoxia later in life at P15 led to
substantial differences in mRNA levels compared to mice
previously exposed to hyperoxia during the neonatal period
and those without. More specifically, there was less
increase in NOS3 mRNA levels (median fold change ×
1.90, IQR 0.92–2.72) in mice that were previously
exposed to hyperoxia compared to mice exposed to
normoxia (median fold change × 3.93, IQR 2.50–8.51).
Interestingly, the increase in STAT3 mRNA was similar
in both groups—even though it tended to be higher in
the hyperoxia (median 3.73, IQR 1.38–7.29) compared
to the normoxia group (median 2.30, IQR
0.74-4.53)—when compared to the median fold change
(× 2.83, IQR 2.21–3.88) after the first exposure. With
respect to the other genes investigated, such as NFE2L2
(Fig. 4c), GPX1, and HIF1A (data not shown), we found
no significant differences between mice with and
without previous hyperoxia exposure (Fig. 4). Since in
human, a small 27 nt-RNA is apparently involved in the
negative regulation of NOS3 mRNA via STAT3 mRNA
targeting, we performed holistic analyses of microRNA
(miR) profiles in purified lung epithelial cell obtained
from experiments under normoxic (group 1), hyperoxic
(groups 2/5), and after repetitive hyperoxic treatment
(group 6). Notably, an ortholog of human 27 nt-RNA
could not be identified in the mouse genome. Therefore,
we aimed to identify differentially expressed murine
candidate miRs, which potentially could target STAT3 mRNA
in a way reminiscent of human 27 nt-RNA and human
STAT3 mRNA. In total, we identified 301 different miRs
from which 101 were non-marginally expressed and thus
were considered for differential analyses. Figure 4d shows
the relative average enrichment of these 101 miRs and
below the group-specific quantitative profiles as a heat
map. Interestingly, almost general drop of miR abundance
was observed after hyperoxic treatment (groups 2/5) when
compared with normoxia (group 1), whereas the
abundance of many miRs increases markedly, when a repetitive
oxygen treatment was applied (group 6). Notably, we
identified several miRs differentially overexpressed under these
conditions, where strong evidence exists for STAT3
mRNA targeting in mice (Fig. 4e), i.e., mmu-let-7b-5p,
mmu-miR-181a-5p, mmu-miR-93-5p, mmu-miR-17-5p,
Modifications in histone signatures at the NOS3 and STAT3 gene locus in response to hyperoxia exposure are not restored
We then further analyzed in pulmonary vascular
endothelial cells whether changes in histone and/or CpG
signatures were associated with the observed expression
patterns. Whereas we did not see any significant
differences in CpG methylation (Table 1), histone
modifications patterns at the NOS3 and STAT3 gene locus
changed substantially. Upon initial exposure to
hyperoxia, we noted an increase in H2aZac at both loci and
an additional decrease in H3K4me3 at the NOS3 locus.
No significant alterations were observed for H3k9ac.
Importantly, levels for H2aZac did not return to a baseline
level but remained elevated and levels for H3K9ac
increased significantly after 8 days of normoxia at both
loci (Fig. 4a, b). Interestingly, upon second exposure to
hyperoxia, we observed an even further increase in
H2aZac, H3K4me, and H3K9ac in mice with previous
excessive oxygen exposure when compared to mice
raised in a normoxic environment, which showed an
acetylation/methylation pattern very similar to the initial
pattern observed upon the primary excessive oxygen
exposure (Fig. 4a, b). At the NFE2L2 locus, no changes
were noted (Fig. 4c).
Long-lasting disturbances in histone modification at the
NOS3 gene locus lead to disruption of the regulation of eNOS expression upon physiological stimuli later on
To further analyze whether persistent histone
modifications are primarily responsible for the altered eNOS
response pattern of vascular endothelial cells upon
physiological stimuli later on, we used human aortic
endothelial cells (HAEC) which were exposed to
trichostatin A (TSA). To mimic persistent environmental
stimulation, we incubated HAEC with TSA for 72 h and
compared these to unexposed HAEC and HAEC that
were exposed twice for 6 h with an interval of 60 h in
between to simulate a more physiological situation.
Whereas persistent stimulation led to a change in
histone acetylation as for example at H3K9, this effect
was much weaker in HAEC subjected to repetitive
stimulation (Fig. 5). Importantly, we also noticed
significant differences between both groups in eNOS and
STAT3α expression. Whereas after repetitive stimulation
eNOS (median 0.52, 0.4–0.67) and STAT3α (median
0.23, 0.2-0.29) dropped to levels below baseline
measurement levels after 4 weeks (Fig. 6a, c), we noticed a
substantial increase in baseline eNOS (median 27.2, 22.3–
35.6) and STAT3α (median 5.8, 4.8–7.3) expression in
HAEC initially subjected to long-term TSA treatment
(Fig. 6b, d). Interestingly, the drop to baseline expression
of eNOS and STAT3α in the group with repetitive
stimulation took significantly more time compared with
the first stimulus.
In the last part of this study, we analyzed the response
of the TSA-exposed HAEC to a physiological stimulus.
Since in contrast to pulmonary vascular endothelial cells,
eNOS expression in peripheral arterial endothelial cells
is stimulated by hypoxia, HAEC—either exposed to two
6-h treatments with TSA or one 72-h TSA
treatment—were exposed to hypoxia after cultivation for 3 weeks
under normal growth conditions. As can be seen in
Fig. 7, HAEC with a previous long-term exposure to
TSA did show a pathological response to hypoxia with a
significant reduction in eNOS expression whereas cells
The mean percentage of methylation at the respective CpGs located in the different promoter regions is shown
with a previous short-term repetitive exposure to TSA
showed a response pattern very similar to control
HAEC. This pathological response was accompanied by
altered histone modifications at the NOS3 gene locus
(Fig. 7d, e). Importantly, as previously mentioned, HAEC
with previous long-term exposure to TSA showed
persistence of H3K9ac at the NOS3 gene locus whereas in
cells with short-term repetitive TSA treatment histone
marks at the NOS3 gene locus were restored.
Interestingly, we also noticed high levels of the repressive
histone mark H3K27me3 which persisted during hypoxia.
Bronchopulmonary dysplasia (BPD) remains one of the
main problems of prematurity even in the
postsurfactant area. Importantly, BPD is also the main
reason for pulmonary hypertension in the pediatric
population. Early oxygen exposure has been identified as one of
the most important risk factors for the development of
BPD. Even though the exact mechanism remains
unclear, it seems to involve high oxygen exposure possibly
by inducing excessive pulmonary vasodilatation and
production of oxygen radicals followed by tissue damage.
However, to date, no pathophysiological link has been
identified between early oxygen exposure and later
pathological response patterns of the pulmonary
vasculature in pulmonary hypertension.
Here, we provide the first evidence that excessive
oxygen exposure early in life leads to permanent changes in
the epigenetic signatures at the NOS3 locus in pulmonary
endothelial artery cells in a mouse model. These changes
seem to be associated with a non-physiological expression
pattern of eNOS upon oxygen exposure later in life.
Moreover, non-physiological alterations of epigenetic signatures
at the NOS3 gene locus induce similar effects in human
aortic endothelial cells (HAEC) in vitro.
Specifically, in the first part of the study, we
demonstrated that NOS3 mRNA levels increased in pulmonary
artery cells upon neonatal oxygen exposure in newborn
mouse. This was paralleled by hyperemia and later
disruption of the alveolar structure as shown by histology.
This is in line with previous reports demonstrating
increased protein concentrations [
] and increased NOS3
mRNA levels upon early excessive oxygen exposure in
various models [
] and human tissue samples [
Interestingly, we did not observe any changes in RNA
expression of NFE2L2, a transcription factor that
regulates the expression of antioxidant proteins that protect
against oxidative damage triggered by injury and
inflammation. This fits to the previous observation that SOD3,
an Nrf2-independent antioxidant, was not found to be
upregulated as one might expect in the O2-exposed
neonatal mice compared with room air [
]. In fact, this
possibly reflects the inefficiency of counteractive
antioxidant mechanisms in the neonatal lung.
The observed increase of NOS3 mRNA levels was
paralleled by an increase in H3K9 and H2Az acetylation
very similar to our previous observations in HUAEC
]. Importantly, we did not observe a restitution of
acetylation patterns at the NOS3 gene locus after 5 days
of normoxia. Moreover, upon a second exposure to
hyperoxia later in life, we noticed a dramatic increase
in H3K9 and H2Az acetylation and also H3K4
methylation in mice that have been exposed to neonatal
hyperoxia. However, NOS3 mRNA levels did not increase but
were even found to be substantial lower in these mice
when compared to healthy controls. Obviously, the
substantial increase in activating histone marks at the
NOS3 gene locus did not lead to an increase in
transcription in these mice. This is in line with a previous
report demonstrating a suppressive effect on NOS3
mRNA synthesis after 12 h incubation with TSA [
As we have previously shown the most reasonable
explanation for this phenomenon is an increased activity
of counteracting processes mainly due to a negative
feedback loop, which in humans presumably involves a
co-processed 27 nt-ncRNA encoded in NOS3 intron 5
], which targets and suppresses STAT3 mRNA
. Our analyses allow us to speculate that a negative
regulatory function in mice, where a STAT3-targeting
27 nt-RNA was not identified, could be fulfilled by one
or several of the differentially expressed miRs, which
we identified by microRNA expression profiling.
Moreover, we also observed increased levels of
H3K27me3—a repressive histone mark—which might
function as an additional counteracting process
responsible for the absent increase in NOS3 mRNA
levels. Considering a recent report demonstrating
eNOS uncoupling and decreased NO production with
an enhanced superoxide production in adult rats after
neonatal hyperoxia [
], the observed downregulation
of NOS3 mRNA production in our study might be a
protective mechanism downregulating eNOS
expression upon oxygen exposure to prevent further
superoxide production and tissue damage. Overall, the
observed changes in epigenetic signatures are in line
with recent reports such as the study by Zhu et al.
who found that H3K27 trimethylation is present in
a BPD mouse model and involved in RUNX3
downregulation, a gene associated with pulmonary
In the second part of the study, we provide further
evidence for the relevance of the observed histone
modifications for the regulation of NOS3 expression using an
in vitro cell culture model. For correct interpretation of
the data, it is important to mention that the
physiological response pattern of human peripheral arterial
endothelial cells is opposite to the response pattern of
pulmonary endothelial artery cells, i.e., vasodilatation
and NOS3 upregulation upon hypoxia and not
hyperoxia. To induce a response pattern comparable to the
mouse model upon hyperoxia, we therefore subjected
HAEC to hypoxia after exposure to TSA. Similar to the
mouse model, we observed long persistence of altered
acetylation patterns after long-term exposure to TSA
whereas histone signatures were restored after repetitive
short-term exposure. In HAEC with long-term exposure
to TSA, NOS3 mRNA levels remained high
corresponding to an increase H3K9ac at the NOS3 gene locus.
However, upon exposure to hypoxia, we observed a
substantial decrease in NOS3 mRNA levels in these cells
even though levels or H3K9ac further increased.
Currently, similar to the constant increase in H3K27me3 at
the NOS3 gene locus in these cells, we have no good
explanation for this phenomenon. Possibly, this just
reflects a complete disturbance and malfunction of the
epigenetic code, but this needs to be further elucidated.
One major limitation of our study is the lack of
functional studies such as measurement of NO. Since many
previous studies have already demonstrated that
upregulation of NOS3 corresponds very well with increased
NO production in the hyperoxia mouse model [
believe that mRNA levels are sufficient physiological to
characterize the cellular response patterns. In addition,
even if that is not the case, it does not change the
disturbances observed on the level of mRNA expression and
in epigenetic signatures. The second major limitation of
this study is the lack of further functional analysis on
the role of miRNA on the development of BPD.
However, since this seems to be rather complex [
considering the focused rather than generalistic approach of
this study, this would have been beyond the primary
scope of this study.
This study provides further evidence for the
longterm relevance of early life events on changes in
epigenetic signatures using a newborn mouse model of
hyperoxia-induced BPD. It demonstrates the
occurrence and relevance of persisting epigenetic marks for
regulation of vascular endothelial response patterns
upon physiological stimuli later in life. Even though
currently very theoretical, this suggests that treatment
of BPD and its sequelae by modification of epigenetic
mechanisms might at least be an option.
BPD: Bronchopulmonary dysplasia; BW: Birth weight; eNOS: Endothelial nitric
oxide synthase; GA: Gestational age; HAEC: Human arterial endothelial cell;
HDAC: Histone deacetylase; HOX: Hyperoxia; NO: Nitric oxide;
NOX: Normoxia; PCR: Polymerase chain reaction; TSA: Trichostatin A
We would like to sincerely thank Claudia Förster and Prof. Klaus-Peter Zimmer
for their constant support.
Rhea van den Bruck received an internal research grant from the University
Availability of data and materials
AJ, CC, SM, JM, LK, PW, MA, and RB have performed data acquisition and
analyzed the data. AJ and CC wrote the manuscript. AJ, CC, and JP took part
in the main study design and supervision of the study. SB took part in the
supervision of the study. SB and JP critically reviewed the manuscript. All
authors read and approved the final manuscript.
All animal experiments were performed in accordance with the National
Institutes of Health Guidelines for the Use of Laboratory Animals. Animal
experiments were approved by the Federal Authorities for Animal Research
of the Regierungspräsidium Giessen, Hessen, Germany, protocol 105/2011.
This study was conducted with the approval of the Witten/Herdecke
University Ethics board.
All authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published
maps and institutional affiliations.
1. Northway WH Jr, Rosan RC , Porter DY . Pulmonary disease following respirator therapy of hyaline-membrane disease . Bronchopulmonary dysplasia. N Engl J Med . 1967 ; 276 ( 7 ): 357 - 68 .
2. Stoll BJ , Hansen NI , Bell EF , Shankaran S , Laptook AR , Walsh MC , Hale EC , Newman NS , Schibler K , Carlo WA , et al. Neonatal outcomes of extremely preterm infants from the NICHD Neonatal Research Network . Pediatrics. 2010 ; 126 ( 3 ): 443 - 56 .
3. Baraldi E , Filippone M. Chronic lung disease after premature birth . N Engl J Med . 2007 ; 357 ( 19 ): 1946 - 55 .
4. Bhandari A , Bhandari V . Pitfalls, problems, and progress in bronchopulmonary dysplasia . Pediatrics . 2009 ; 123 ( 6 ): 1562 - 73 .
5. Shaw GM , O'Brodovich HM . Progress in understanding the genetics of bronchopulmonary dysplasia . Semin Perinatol . 2013 ; 37 ( 2 ): 85 - 93 .
6. Vento M , Moro M , Escrig R , Arruza L , Villar G , Izquierdo I , Roberts LJ 2nd, Arduini A , Escobar JJ , Sastre J , et al. Preterm resuscitation with low oxygen causes less oxidative stress, inflammation, and chronic lung disease . Pediatrics . 2009 ; 124 ( 3 ): e439 - 49 .
7. Potter CF , Kuo NT , Farver CF , McMahon JT , Chang CH , Agani FH , Haxhiu MA , Martin RJ . Effects of hyperoxia on nitric oxide synthase expression, nitric oxide activity, and lung injury in rat pups . Pediatr Res . 1999 ; 45 ( 1 ): 8 - 13 .
8. Kondrikov D , Elms S , Fulton D , Su Y. eNOS-beta-actin interaction contributes to increased peroxynitrite formation during hyperoxia in pulmonary artery endothelial cells and mouse lungs . J Biol Chem . 2010 ; 285 ( 46 ): 35479 - 87 .
9. Simpson SJ , Hall GL , Wilson AC. Lung function following very preterm birth in the era of 'new' bronchopulmonary dysplasia . Respirology . 2015 ; 20 ( 4 ): 535 - 40 .
10. Mourani PM , Sontag MK , Younoszai A , Miller JI , Kinsella JP , Baker CD , Poindexter BB , Ingram DA , Abman SH . Early pulmonary vascular disease in preterm infants at risk for bronchopulmonary dysplasia . Am J Respir Crit Care Med . 2015 ; 191 ( 1 ): 87 - 95 .
11. O 'Reilly MA , Yee M , Buczynski BW , Vitiello PF , Keng PC , Welle SL , Finkelstein JN , Dean DA , Lawrence BP . Neonatal oxygen increases sensitivity to influenza A virus infection in adult mice by suppressing epithelial expression of Ear1 . Am J Pathol . 2012 ; 181 ( 2 ): 441 - 51 .
12. Domm W , Misra RS , O'Reilly MA . Affect of early life oxygen exposure on proper lung development and response to respiratory viral infections . Front Med . 2015 ; 2 : 55 .
13. Bhandari V , Bizzarro MJ , Shetty A , Zhong X , Page GP , Zhang H , Ment LR , Gruen JR , Neonatal Genetics Study G. Familial and genetic susceptibility to major neonatal morbidities in preterm twins . Pediatrics . 2006 ; 117 ( 6 ): 1901 - 6 .
14. Londhe VA , Sundar IK , Lopez B , Maisonet TM , Yu Y , Aghai ZH , Rahman I. Hyperoxia impairs alveolar formation and induces senescence through decreased histone deacetylase activity and up-regulation of p21 in neonatal mouse lung . Pediatr Res . 2011 ; 69 ( 5 Pt 1 ): 371 - 7 .
15. Ni W , Lin N , He H , Zhu J , Zhang Y . Lipopolysaccharide induces upregulation of TGF-alpha through HDAC2 in a rat model of bronchopulmonary dysplasia . PLoS One . 2014 ; 9 ( 3 ): e91083 .
16. Postberg J , Kanders M , Forcob S , Willems R , Orth V , Hensel KO , Weil PP , Wirth S , Jenke AC . CpG signalling, H2A.Z/H3 acetylation and microRNAmediated deferred self-attenuation orchestrate foetal NOS3 expression . Clin Epigenetics . 2015 ; 7 ( 1 ): 9 .
17. Jenke AC , Postberg J , Raine T , Nayak KM , Molitor M , Wirth S , Kaser A , Parkes M , Heuschkel RB , Orth V , et al. DNA methylation analysis in the intestinal epithelium-effect of cell separation on gene expression and methylation profile . PLoS One . 2013 ; 8 ( 2 ): e55636 .
18. Postberg J , Kanders M , Forcob S , Willems R , Orth V , Hensel KO , Weil PP , Wirth S , Jenke AC . CpG signalling, H2A.Z/H3 acetylation and microRNAmediated deferred self-attenuation orchestrate foetal NOS3 expression . Clin Epigenetics . 2015 ; 7 : 9 .
19. Weil PP , Jaszczyszyn Y , Baroin-Tourancheau A , Postberg J , Amar L . Holistic and affordable analyses of MicroRNA expression profiles using tagged cDNA libraries and a multiplex sequencing strategy . Methods Mol Biol . 2017 ; 1654 : 179 - 96 .
20. Blankenberg D , Von Kuster G , Coraor N , Ananda G , Lazarus R , Mangan M , Nekrutenko A , Taylor J. Galaxy : a web-based genome analysis tool for experimentalists . Curr Protoc Mol Biol . 2010;Chapter 19(Unit 19 10): 11 - 21 .
21. Giardine B , Riemer C , Hardison RC , Burhans R , Elnitski L , Shah P , Zhang Y , Blankenberg D , Albert I , Taylor J , et al. Galaxy: a platform for interactive large-scale genome analysis . Genome Res . 2005 ; 15 ( 10 ): 1451 - 5 .
22. Goecks J , Nekrutenko A , Taylor J , Galaxy T. Galaxy : a comprehensive approach for supporting accessible, reproducible, and transparent computational research in the life sciences . Genome Biol . 2010 ; 11 ( 8 ): R86 .
23. Vitsios DM , Enright AJ . Chimira: analysis of small RNA sequencing data and microRNA modifications . Bioinformatics . 2015 ; 31 ( 20 ): 3365 - 7 .
24. Thebaud B. Angiogenesis in lung development, injury and repair: implications for chronic lung disease of prematurity . Neonatology . 2007 ; 91 ( 4 ): 291 - 7 .
25. Thebaud B , Abman SH . Bronchopulmonary dysplasia: where have all the vessels gone? Roles of angiogenic growth factors in chronic lung disease . Am J Respir Crit Care Med . 2007 ; 175 ( 10 ): 978 - 85 .
26. Feng X , Guo Z , Nourbakhsh M , Hauser H , Ganster R , Shao L , Geller DA . Identification of a negative response element in the human inducible nitric-oxide synthase (hiNOS) promoter: the role of NF-kappa B-repressing factor (NRF) in basal repression of the hiNOS gene . Proc Natl Acad Sci U S A . 2002 ; 99 ( 22 ): 14212 - 7 .
27. Davis CW , Gonzales LW , Ballard RA , Ballard PL , Guo C , Gow AJ . Expression of nitric oxide synthases and endogenous NO metabolism in bronchopulmonary dysplasia . Pediatr Pulmonol . 2008 ; 43 ( 7 ): 703 - 9 .
28. Poonyagariyagorn HK , Metzger S , Dikeman D , Mercado AL , Malinina A , Calvi C , McGrath-Morrow S , Neptune ER . Superoxide dismutase 3 dysregulation in a murine model of neonatal lung injury . Am J Respir Cell Mol Biol . 2014 ; 51 ( 3 ): 380 - 90 .
29. Rossig L , Li H , Fisslthaler B , Urbich C , Fleming I , Forstermann U , Zeiher AM , Dimmeler S. Inhibitors of histone deacetylation downregulate the expression of endothelial nitric oxide synthase and compromise endothelial cell function in vasorelaxation and angiogenesis . Circ Res . 2002 ; 91 ( 9 ): 837 - 44 .
30. Zhang MX , Ou H , Shen YH , Wang J , Wang J , Coselli J , Wang XL . Regulation of endothelial nitric oxide synthase by small RNA . Proc Natl Acad Sci U S A . 2005 ; 102 ( 47 ): 16967 - 72 .
31. Yan L , Hao H , Elton TS , Liu Z , Ou H . Intronic microRNA suppresses endothelial nitric oxide synthase expression and endothelial cell proliferation via inhibition of STAT3 signaling . Mol Cell Biochem . 2011 ; 357 ( 1-2 ): 9 - 19 .
32. Yzydorczyk C , Comte B , Huyard F , Cloutier A , Germain N , Bertagnolli M , Nuyt AM . Developmental programming of eNOS uncoupling and enhanced vascular oxidative stress in adult rats after transient neonatal oxygen exposure . J Cardiovasc Pharmacol . 2013 ; 61 ( 1 ): 8 - 16 .
33. Zhu Y , Fu J , Yang H , Pan Y , Yao L , Xue X . Hyperoxia-induced methylation decreases RUNX3 in a newborn rat model of bronchopulmonary dysplasia . Respir Res . 2015 ; 16 : 75 .
34. Nardiello C , Morty RE . MicroRNA in late lung development and bronchopulmonary dysplasia: the need to demonstrate causality . Mol Cell pediatr . 2016 ; 3 ( 1 ): 19 .