This study shows that bile acids (BA) may contribute to an increase in alveolar epithelial permeability in a dose-dependent manner. The key findings are: 1) BA upregulates mitogen-activated protein kinase (MAPK), cytosolic phospholipase A2 (cPLA2), cyclooxygenase-2 (COX-2), and prostaglandin E2 (PGE2) generation. 2) This leads to a reduction in the tight junction proteins occludin, zonula occludens-1 (ZO-1), and E-cadherin, which are responsible for maintaining epithelial integrity. 3) The increase in alveolar permeability and

1. ORIGINAL ARTICLE
Bile acids increase alveolar epithelial permeability via
mitogen-activated protein kinase, cytosolic phospholipase A2,
cyclooxygenase-2, prostaglandin E2 and junctional proteins
KANG-CHENG SU,1,3
* YU-CHUNG WU,2,4
* CHUN-SHENG CHEN,3
MING-HUI HUNG,3
YI-HAN HSIAO,3
CHING-MIN TSENG,3
SHI-CHUAN CHANG,1,3
YU-CHIN LEE2,3
AND DIAHN-WARNG PERNG2,3
1
Institute of Emergency and Critical Care Medicine, School of Medicine, and 2
School of Medicine, National Yang-Ming
University, 3
Department of Chest Medicine and 4
Division of Thoracic Surgery, Department of Surgery, Taipei Veterans
General Hospital, Taipei, Taiwan
ABSTRACT
Background and objective: Bile acid (BA) aspiration
is associated with various lung diseases. It was hypoth-
esized that BA may induce changes in alveolar epithe-
lium permeability and contribute to the pathogenesis
of lung injury.
Methods: Human alveolar epithelial cells were grown
in monolayer and stimulated with a major component
of BA, chenodeoxycholic acid (CDCA). Transepithelial
electrical resistance (TER) and paracellular fluxes were
measured to assess permeability alteration. Prostag-
landin E2 (PGE2) production was measured, and its
effect on TER and junctional proteins (JP) was also
examined. Reverse transcription polymerase chain
reaction andWestern blots were used to investigate the
expression of messenger RNA and JP.
Results: CDCA induced significant p38 and c-Jun
N-terminal kinase (JNK) phosphorylation, cytosolic
phospholipase A2 (cPLA2) and cyclooxygenase-2
(COX-2) messenger RNA expression, PGE2 production,
TER reduction and decay of JP (including occludin,
zonula occludens-1 (ZO-1) and E-cadherin, in which
ZO-1 had maximal change). CDCA also increased para-
cellular fluxes, which was abolished by dexametha-
sone. Both CDCA and PGE2 contributed to TER
reduction in an identical trend and a dose–response
manner.PGE2 also reduced ZO-1 expression,which was
similar to that observed by CDCA stimulation. Pre-
treatment with inhibitors of p38 (SB203580), JNK
(SP600125), cPLA2 (mepacrine) and COX-2 (NS398) as
well as dexamethasone reversed the CDCA-induced
PGE2 production, TER reduction and decay of ZO-1.
Conclusions: The increase in alveolar permeability
was associated with decay of JP. BA may induce perme-
ability alteration through the upregulation of mitogen-
activated protein kinase, cPLA2, COX-2, PGE2 and JP,
which may contribute to the pathogenesis of
BA-associated lung injury.
Key words: alveolar permeability, junctional protein, paracel-
lular flux, prostaglandin E2, transepithelial electrical resistance.
Abbreviations: BA, bile acid; CDCA, chenodeoxycholic acid;
COX-2, cyclooxygenase-2; cPLA2, cytosolic phospholipase A2;
HAEC, human alveolar epithelial cells; JNK, c-Jun N-terminal
kinase; JP, junctional proteins; MAPK, mitogen-activated protein
kinase; PGE2, prostaglandin E2; TER, transepithelial electrical
resistance; TJ, tight junctions; ZO-1, zonula occludens-1.
INTRODUCTION
Duodenogastroesophageal reflux aspiration is a
major risk factor that can cause or complicate various
lung diseases.1–5
We recently found that bile acids (BA)
were much higher in asthmatics with gastroesopha-
geal reflux symptoms6
and in patients with suspected
ventilator-associated pneumonia.7
We also illustrated
that BA promoted interleukin-8 production and con-
tributed to transforming growth factor-b1 production
and subsequent fibroblast proliferation in human
Correspondence: Diahn-Warng Perng, School of Medicine,
National Yang-Ming University, Department of Chest Medicine,
Taipei Veterans General Hospital, 201, Section 2, Shih-Pai Road,
Taipei 11217, Taiwan. Email: dwperng@vghtpe.gov.tw
*Dr Yu-Chung Wu and Dr Kang-Cheng Su made equal contri-
butions to this study.
Received 27 January 2012; invited to revise 20 March, 10 June
and 23 September 2012; revised 19 April, 8 August and 13
October 2012; accepted 3 January 2013 (Associate Editor: Shu
Hashimoto).
SUMMARY AT A GLANCE
BA-related lung inflammation is seldom dis-
cussed. This study shows that BA may contribute
to an increase in alveolar epithelial permeability in
a dose-dependent manner. This process involves
upregulation of MAPK, cPLA2, COX-2 and PGE2
generation, and decay of occludin, ZO-1 and
E-cadherin.
bs_bs_banner
© 2013 The Authors
Respirology © 2013 Asian Pacific Society of Respirology
Respirology (2013) 18, 848–856
doi: 10.1111/resp.12086

2. alveolar epithelial cells (HAEC) and human bronchial
epithelial cells, respectively.6–8
In an animal model, BA
aspiration produced a pneumonitis which lead to
non-cardiac pulmonary oedema of much greater
severity than that induced by normal gastric con-
tents.9
The mechanism and the extent of damage in
lung alveoli remains unclear.
Increased alveolar permeability is the hallmark of
acute lung injury.10–12
A possible mediator of this
damage is prostaglandin E2 (PGE2), which increases
in bronchoalveolar lavage fluid after acid aspiration-
induced lung injury in rats.13
PGE2, a lipid me-
tabolite synthesized from arachidonic acid by
cyclooxygenase-2 (COX-2), is released from epithelial
cells in response to stimuli and can modulate the
immune and inflammatory response during acute
and chronic lung disease.14–19
The permeability of
epithelial layers depends on tight junctions (TJ),
which are located at the apicolateral border of the
epithelial cells and maintain the epithelial integrity
in order to protect the underlying tissue against
external stimuli.20–22
Adherence junctions, located
below the TJ, are required in the assembly of TJ.23,24
Both are constituted mainly by junctional proteins
(JP), including transmembrane proteins, such as
occludin, claudin in TJ, E-cadherin in adherence
junctions and other peripheral membrane proteins,
such as zonula occludens-1 and -2 (ZO-1 and
ZO-2).20–23,25,26
All of these structures are essential to
regulate epithelial permeability. PGE2 can alter per-
meability of kidney epithelial cells27
and impair
barrier function by disassembling JP in colorectal
cells.28
Human alveolar epithelium is especially likely to be
exposed to BA and the effects of BA on alveolar per-
meability remain unclear. In this study, we tested the
hypothesis that BA may alter alveolar permeability
through the production of PGE2 and examined its
signal pathway as well as the expression of JP.
METHODS
Enzymatic dissociation of alveolar cells from
lung parenchyma
Preparation of the HAEC from lung parenchyma,
which was obtained from surgical lobectomy for
lung cancer, has been described previously.7
Briefly,
after enzymatic dissociation, filtration and centrifu-
gation, the media containing non-adherent type II
epithelial cells was resuspended and re-plated in a
flask for 1 h at 37°C to allow adherence of the con-
taminating fibroblast cells. Type II cell-enriched
medium was collected, centrifuged and resuspended
with specific alveolar cell culture medium (see the
online supporting information). Cells (100 mL) were
seeded in 24-well culture inserts (coated with type IV
collagen, 50 mg/cm2)
at a density of 1 ¥ 105
cells/mL
and grown in culture medium. The cells were incu-
bated for 7 to 9 days until a confluent alveolar mon-
olayer was formed and used for experiments. This
study was approved by our hospital’s institutional
review board.
Measurement of epithelial cell permeability
The leakiness of intercellular TJ contributes to altera-
tions in epithelial permeability, which is usually rep-
resented by transepithelial electrical resistance (TER)
and paracellular tracer fluxes. TER across the conflu-
ent monolayer was measured using EVOM micro
volt-ohm-meter (World Precision Instruments,
Owslebury, UK) with a chopstick-style electrode in
repeated measurements before and hours after
stimulation with chenodeoxycholic acid (CDCA, from
Sigma, St. Louis, MO, USA), the major component of
BA.7
The measurements were performed with a con-
stant temperature of 25°C. When the alveolar monol-
ayer reached a TER of 1500–2000 W·well, baseline
values of TER were obtained by subtracting the con-
tribution of the filter and the bathing solution.Various
concentrations of CDCA (0–200 mmol/L) or PGE2
(0–100 ng/mL) were added to the apical part of the
culture inserts, and TER was measured at different
time points. To reverse TER altered by CDCA
(150 mmol/L), COX-2 inhibitor NS398 (N-[-2-
cyclohexyloxy]-4-nitrophenyl methanesulfonamide,
Sigma, 0.1–10 mmol/L) and dexamethasone (10–
100 mmol/L) were pretreated for 1 h, followed by
CDCA stimulation. TER was measured at the indi-
cated time points (1, 2, 3 and 6 h after stimulation).
The apical-to-basolateral paracellular fluxes, repre-
sented by tracer markers (Cascade blue-Dextran,
10 kDa, 2 mg/mL, Molecular Probes, Eugene, OR,
USA),29
were added to the apical side of culture inserts
and were agitated gently, after stimulation with
150 mmol/L CDCA on the apical side. After incubation
for 30 min, the medium in the basal culture well was
replaced with fresh medium and collected. After 2 h,
the basal medium was collected again. The amount of
the diffused marker was determined with a Perkin-
Elmer 2000 (Wellesley, MA, USA) fluorescence
spectrophotometer (excitation wavelength = 365 nm,
emission wavelength = 440 nm).
Cell viability by
3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl
tetrazolium bromide assay
After measurement of TER treated with CDCA
(0–200 mmol/L) for 6 h and 24 h, 100 mL of cell
suspensionwasaddedintoeachwellina96-wellplate,
followed by adding 20 mL 3-(4,5-dimethylthiazol-2-
yl)-2,5-diphenyl tetrazolium bromide, 5 mg/mL in
phosphate-buffered saline) solution apically and
incubation for 4 h at 37°C in a 5% CO2 atmosphere.
Solubilizing solution (dimethyl sulfoxide) was added
(100 mL/well) with a gentle shake for 10 min, and the
absorbance was analysed by spectrophotometry in a
microtitre plate reader at a wavelength of 540 nm.
PGE2 generation
CDCA was added to culture inserts at varying concen-
trations (0–200 mmol/L). PGE2 was collected from the
supernatants after incubation for 1, 3 and 6 h. To sup-
press the effect of PGE2 release induced by CDCA,
cells were pretreated for 1 h with cytosolic phospholi-
pase A2 (cPLA2) inhibitor (mepacrine, 1–20 mmol/L),
Bile acids increase lung permeability 849
© 2013 The Authors
Respirology © 2013 Asian Pacific Society of Respirology
Respirology (2013) 18, 848–856

3. COX-2 inhibitor (NS398, 0.1–10 mmol/L), c-Jun N-
terminal kinase (JNK) inhibitor (SP600125, 25 mmol/
L)30
and p38 inhibitor (SB203580, 10 mmol/L)6
(all pur-
chased from Sigma), followed by adding CDCA
150 mmol/L to the apical compartment for 6 h. Subse-
quently, the supernatants were collected and stored at
-80°C. PGE2 were determined by the PGE2 enzyme-
linked immunosorbent assay, according to the manu-
facturer’s instructions (R&D Systems, Minneapolis,
MN, USA).
Western blot analysis of mitogen-activated
protein kinases
HAEC were exposed to CDCA in the presence or
absence of inhibitors of mitogen-activated protein
kinase (MAPK) activity. Cells were pre-incubated for
1 h with dexamethasone (100 mmol/L), p38 inhibitor
SB203580 (10 mmol/L) and JNK inhibitor SP600125
(25 mmol/L) (Cell Signaling Technology, Beverly, MA,
USA), and subsequently treated with 150 mmol/L
CDCA for 30 min. At the end of treatment, cells were
lysed on ice in a lysis buffer (see the online supporting
information). The procedure to detect MAPK activity
has been described in detail previously.31
cPLA2 and COX-2 mRNA expression
After treatment with the absence or presence of CDCA
(100–200 mmol/L) in 24-well culture plates for the
time indicated (1, 3 and 6 h), the supernatants were
removed. Total cellular RNA was isolated from cell
monolayer using High Pure RNA Isolation Kit (Roche
Molecular Biochemicals, Mannheim, Germany) for
reverse transcription polymerase chain reaction as
described previously.14
Briefly, the RNA (1 mg) was
reverse transcribed into complementary DNA using
Superscript II RNase H-
reverse transcriptase. An
aliquot of complementary DNA was then subjected
to 28 cycles of polymerase chain reaction for
glyceraldehyde-3-phosphate dehydrogenase using a
standard procedure: denaturing at 94°C for 2 min,
hybridizing at 55°C for 30 s and elongating at 72°C for
1.5 min. Similarly, another aliquot of complementary
DNA was then subjected to 35 cycles of polymerase
chain reaction (COX-1, COX-2 and cPLA2, respec-
tively) using a standard procedure denaturing at 94°C
for 1 min, annealing at 55°C for 30 s and elongating at
72°C for 1.2 min. The primer sequence for cPLA2,
COX-1, COX-2 and glyceraldehyde-3-phosphate
dehydrogenase is shown in Table S1 (see supplemen-
tary information available online). The amplified
products were electrophoresed in a 2% agarose gel
containing ethidium bromide (0.5 g/mL) and viewed
under an UV illuminator. The image was photo-
graphed, stored and analysed by a photodocumenta-
tion system using Photo-Capt software (ETS Vilber-
Lourmat Inc., Marne LuVallee Cedex, France). Each
band was quantified by calculating the ratio of target
complementary DNA signal to the glyceraldehyde-3-
phosphate dehydrogenase control. The mRNA
expression was analysed by densitometry and pre-
sented as a fold of control.
Western blot analysis of JP
HAEC were treated with CDCA (150 mmol/L) or PGE2
(100 ng/mL) and incubated for 2.5 h and 3.5 h. To
suppress the protein alterations induced by CDCA,
HAEC were pretreated with p38 inhibitor (SB203580,
10 mmol/L), JNK inhibitor (SP600125, 25 mmol/L),
COX-2 inhibitor (NS398, 1–10 mmol/L) and dexam-
ethasone (100 mmol/L) for 1 h, followed by adding
150 mmol/L CDCA for 3.5 h of incubation. At the end
of treatment, cells were lysed on ice in a lysis buffer.
The protein concentration was determined using a
bicinchoninic acid protein assay (Pierce Chemicals;
Rockford, IL, USA) with bovine serum albumin as the
standard. Equal amounts of total cell lysates were
solubilized in a sample buffer by boiling for 10 min,
fractionated on a 7.5% sodium dodecylsulphate-
polyacrylamide gel, and transferred onto a nitrocellu-
lose membrane. The membrane was washed with
0.1% Tween 20 supplemented with Tris-buffered
saline solution and incubated in a blocking buffer.
Specific antibody for ZO-1 (1:1000 dilution), occludin
(1:400) (both from Zymed Laboratories, South San
Francisco, CA, USA) and E-cadherin (Abcam, Cam-
bridge, UK) were then applied at room temperature
for 3 h, with gentle shaking. Monoclonal antibody
against anti-b-actin (1:5000, Sigma) was used for
standardization. After washing three times with Tris-
buffered saline, blots were incubated with a 1:2000
dilution of a horseradish peroxidase-conjugated sec-
ondary antibody (Cell Signaling Technology) for 1 h.
The protein bands were viewed using enhanced
chemiluminescence (Amersham Pharmacia Biotech,
San Francisco, CA, USA). The densitometric analysis
of immunoblots was performed by using the Quantity
One Software (Bio-Rad Laboratories, Hercules, CA,
USA).
Immunofluorescent staining of ZO-1
The immunofluorescent staining has been described
previously.18
Briefly, HAEC were grown on a culture
plate until confluence, followed by treatment with
150 mmol/L CDCA or 100 ng/mL PGE2 for 3.5 h. Spe-
cific antibody for ZO-1 (1:100 dilution) was then
applied at room temperature for 3 h, followed by
incubation with fluorescein-isothiocyanate-labelled
antirabbit immunoglobulin G (Novocastra Laborato-
ries, Newcastle upon Tyne, UK) for 1 h at room
temperature. Colour was developed using 3,30-
diaminobenzidine tetrahydrochloride peroxidase
substrate for 10 min. After washing, fluorescent
images were obtained using a fluorescent microscope
(Nikon, TE 300, Tokyo, Japan) for fluorescein-
isothiocyanate detection.
Statistical analysis
Results are expressed as means Ϯ standard error of
the mean. Differences between groups were assessed
using either Mann–Whitney U-test (densitometry) or
one-way analysis of variance for multiple compari-
sons (TER, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl
tetrazolium bromide, PGE2, dextran). P < 0.05 was
considered significant for all tests.
K-C Su et al.850
© 2013 The Authors
Respirology © 2013 Asian Pacific Society of Respirology
Respirology (2013) 18, 848–856

4. RESULTS
Effects of CDCA on TER alteration and
cell viability
Figure 1a shows that CDCA transiently decreased TER
of HAEC in a concentration- and time-dependent
manner. CDCA 150 mmol/L induced the maximal
transitory response after 1 h of incubation. CDCA
concentrations below 100 mmol/L did not provoke
any change in TER, whereas CDCA 200 mmol/L
induced an irreversible decrease that was associated
with a 39.5% reduction in cell viability (Fig. 1b), indi-
cating the lower viability results from a cytotoxic
effect of CDCA.
Effects of CDCA on PGE2 generation
PGE2 has been associated with reduced TER in
Madin–Darby canine kidney cells incubated with the
epidermal growth factor.27
We therefore determined
whether this autacoid is produced during CDCA-
induced TER reduction. Figure 2 shows that PGE2
increases in a time- and CDCA concentration-
dependent manner, suggesting that it plays a role in
TER reduction.
Effects of CDCA and PGE2 on TER alteration
Figure 3a shows that the CDCA-induced transient
decrease in TER gradually recovered to approximate
baseline levels at 24 h. Figure 3b shows that PGE2
affected TER in a concentration- and time-dependent
manner, with a dose of 100 ng/mL eliciting the
maximal response. The similar effect of CDCA and
PGE2 on TER suggests that TER reduction might be
mediated via PGE2.
Effects of CDCA on MAPK phosphorylation
MAPK family members modulate inflammation in
various tissues. Figure 4 indicates that a 30-min
treatment with CDCA (150 mmol/L) markedly
induces the phosphorylation of p38 (Fig. 4a) and
JNK (Fig. 4b). Dexamethasone (100 mmol/L), p38
inhibitor SB203580 (10 mmol/L) and JNK inhibitor
SP600125 (25 mmol/L) abolished CDCA-induced
phosphorylation.
Effects of CDCA on cPLA2 and COX-2
mRNA expression
To investigate signalling pathways involved in CDCA-
induced inflammation, the expression of cPLA2 and
COX-2 mRNA was evaluated (Fig. 5). CDCA induced
the expression of both genes in a dose-dependent
manner, and the increase was evident at each time
point (Fig. 5a). Treatment with CDCA 150 mmol/L
generated the maximal response, in accordance with
its effects on TER reduction and PGE2 production.
Densitometry readings showed that this dose of
CDCA significantly increased in both cPLA2 and
Figure 1 Effects of chenodeoxycholic acid (CDCA) on transepi-
thelial electrical resistance (TER) alteration and cell viability.
CDCA (0–200 mmol/L) was added to alveolar monolayer, followed
by incubation and measurements of TER across epithelial mon-
olayer at the indicated time points. *P < 0.05, vs control, CDCA
100 and CDCA 125, respectively. ***P < 0.001, vs control, CDCA
100, CDCA 125 and CDCA 150, respectively (a). The cell viability
was evaluated by MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl
tetrazolium bromide) assay at 6 h of incubation. **P < 0.01, vs
control, CDCA 100, CDCA 125 and CDCA 150, respectively (b).
Values are means Ϯ standard error for three separate experi-
ments performed in triplicate. , control; , CDCA 100; ,
CDCA 125; , CDCA 150; , CDCA 200 (mmol/L).
Figure 2 Effects of chenodeoxycholic acid (CDCA) on prostag-
landin E2 (PGE2) generation. PGE2 level was determined in the
supernatants of epithelial cells treated with CDCA (0–200 mmol/L)
for 1, 3 and 6 h. *P < 0.05, **P < 0.01, ***P < 0.001. Values are
means Ϯ standard error for three separate experiments per-
formed in triplicate. , control; , CDCA 100; , CDCA 150; ,
CDCA 200 (mmol/L).
Bile acids increase lung permeability 851
© 2013 The Authors
Respirology © 2013 Asian Pacific Society of Respirology
Respirology (2013) 18, 848–856

5. COX-2 mRNA expression (Fig. 5b). By contrast, CDCA
did not elicit any change in COX-1 mRNA expression
(data not shown).
Effects of inhibitors to reverse alteration
in permeability
Figure 6 shows that dexamethasone and the COX-2
inhibitor NS398 blocked the changes in permeability
in a dose-dependent manner. Treatment with
10 mmol/L NS398 (Fig. 6a) and 100 mmol/L dexam-
ethasone (Fig. 6b) led to the most significant suppres-
sion of CDCA-dependent effects at each time point.
Treatment with 150 mmol/L CDCA significantly
increased the paracellular fluxes, as indicated by a
dextran passage, at both 30 min and 2 h, whereas
100 mmol/L dexamethasone (but not the lower dose
of 10 mmol/L) significantly attenuated the CDCA-
induced increase in paracellular fluxes (Fig. 6c).
Effect of inhibitors on CDCA-induced
PGE2 generation
We next examined the CDCA-induced inflammatory
pathway by blocking mediators related to PGE2 pro-
duction.Figure 7showsthatCDCA-inducedPGE2 pro-
duction was significantly suppressed by inhibitors
of cPLA2 (mepacrine, 1 mmol/L), COX-2 (NS398,
10 mmol/L), JNK (SP600125, 25 mmol/L) and p38
(SB203580, 10 mmol/L) at 6 h.The ability of mepacrine
to suppress PGE2 production was similar at a concen-
tration range of 1–20 mmol/L (data not shown).
Effects of CDCA and PGE2 on JP
Treatment with CDCA significantly downregulated
occludin, ZO-1 and E-cadherin at 3.5 h (Fig. 8a).
Among these TJ, ZO-1 levels were the most reduced,
as indicated by densitometry readings (see Figure S1
in the supplementary information available online).
Similarly, PGE2 100 ng/mL also caused a significant
decrease in ZO-1 at 3.5 h (Fig. 8b). Moreover, dexam-
ethasone (100 mmol/L), NS398 (10 mmol/L), SP600125
(25 mmol/L) and SB203580 (10 mmol/L) significantly
suppressed the CDCA-induced ZO-1 decay, in accord-
ance with its inhibitory effect on CDCA-induced PGE2
production (Fig. 8c,d, Fig. S1).
Immunostaining of ZO-1 in alveolar cells
Figure 9 shows the results of immunocytochemical
studies of the ZO-1 protein in cells on the culture plate
at 3.5 h post-treatment. The ZO-1 protein presented
as a continuous band around the unstimulated cells
(Fig. 9a), whereas 150 mmol/L CDCA (Fig. 9b) and
100 ng/mL PGE2 (Fig. 9c) contributed to the disrup-
tion of ZO-1 band. The results were compatible with
those of the Western blotting analysis for ZO-1.
Figure 3 Effects of chenodeoxycholic acid (CDCA) and prostag-
landin E2 (PGE2) on TER alteration. CDCA (0–150 mmol/L) ,
control; , CDCA 100; , CDCA 125; , CDCA 150 (mmol/L).
(a) and PGE2 (0–100 ng/mL) , control; , PGE2 0.1; , PGE2
1; , PGE2 10; , PGE2 100 ng/mL. (b) were added to alveolar
monolayer, followed by incubation and measurements of
transepithelial electrical resistance at the indicated time
points. * P < 0.05, and *** P < 0.001, vs control. Values are
mean Ϯ standard error for four to six separate experiments per-
formed in triplicate.
Figure 4 Effects of chenodeoxycholic acid (CDCA) on mitogen-
activated protein kinase phosphorylation. Cells were treated with
buffer alone (control) or CDCA (150 mmol/L) in the absence or
presence of dexamethasone (100 mmol/L), p38 inhibitor
(SB203580, 10 mmol/L) and c-Jun N-terminal kinase (JNK) inhibi-
tor (SP600125, 25 mmol/L) for 30 min, then harvested for Western
blotting. Representative immunoblots of three separate experi-
ments for phosphorylated p38 (a) and JNK (b) are shown.
K-C Su et al.852
© 2013 The Authors
Respirology © 2013 Asian Pacific Society of Respirology
Respirology (2013) 18, 848–856

6. Signal pathway of CDCA-induced alteration
in permeability
Figure 10 summarizes our findings and illustrates the
signalling pathway associated with CDCA-induced
inflammation in HAEC. Briefly, CDCA stimulated the
cells to induce the phosphorylation of p38 and JNK,
mRNA expression of cPLA2 and COX-2, production of
PGE2, and decay of JP. Pretreatment with inhibitors of
p38 (SB203580), JNK (SP600125), cPLA2 (mepacrine)
and COX-2 (NS398) as well as dexamethasone sup-
pressed CDCA-induced PGE2 production, restored
TER reduction and attenuated ZO-1 decay.
DISCUSSION
We recently discovered that CDCA concentration in
tracheal aspirates from patients with suspected
ventilator-associated pneumonia was significantly
higher than normal and that CDCA might contribute
to interleukin-8 production in HAEC.7
In the present
study, we demonstrated the direct impact of CDCA on
alveolar permeability. CDCA increased the perme-
ability of HAEC, and it was associated with the activa-
tion of p38 and JNK kinase-associated signalling
cPLA2 upregulation, COX-2 mRNA expressions and
PGE2 production. The increase in epithelial perme-
ability was associated with decay of occludin, ZO-1
and E-cadherin.
Compared with conjugated BA (glycochenodeoxy-
cholic acids), unconjugated BA (CDCA) are approxi-
mately 5-fold more potent stimulators of PGE2
synthesis by oesophageal squamous cells.32
In the
intestinal epithelial cells, CDCA is a particularly
potent BA that can reduce TER and increase paracel-
lular permeability, a process that involves occludin
Figure 5 Effects of chenodeoxycholic acid (CDCA) on cytosolic
phospholipase A2 (cPLA2) and cyclooxygenase-2 (COX-2) mes-
senger RNA (mRNA) expression. The mRNA expression of cPLA2,
COX-2 and glyceraldehyde 3-phosphate dehydrogenase
(GAPDH) was performed by reverse transcription polymerase
chain reaction (PCR) after incubation with CDCA (0–200 mmol/L)
for 1, 3 and 6 h (a). Densitometry readings were derived from the
PCR gels at 1 h and expressed as fold of control (b). *P < 0.05,
***P < 0.01. Values are means Ϯ standard error for three sepa-
rate experiments. , cPLA2; , COX-2.
Figure 6 Effects of inhibitors to reverse alteration in permeabil-
ity. With 1 h pretreatment in the absence or presence of NS398
(0.1–10 mmol/L). , control; , CDCA 150; , CDCA-NS398
0.1; , CDCA-NS398 1; , CDCA-NS398 10 (mmol/L). (a) and
dexamethasone (10–100 mmol/L). , control; , CDCA 150;
, CDCA+Dex10; , CDCA+Dex100 (mmol/L). (b), transepithe-
lial electrical resistance was measured in cells incubated with
chenodeoxycholic acid (CDCA) (150 mmol/L) at the indicated time
points. *P < 0.05, **P < 0.01, †
P < 0.05, vs CDCA 150 mmol/L. With
1 h pretreatment in the absence or presence of dexamethasone
(10–100 mmol/L), the CDCA-induced paracellular fluxes (c), repre-
sented by a dextran passage, were measured at 30 min and 2 h.
*P < 0.05, **P < 0.01. Values are means Ϯ standard error for
three separate experiments performed in triplicate. , control; ,
CDCA 150; , CDCA+Dex10; , CDCA+Dex100 (mmol/L).
Bile acids increase lung permeability 853
© 2013 The Authors
Respirology © 2013 Asian Pacific Society of Respirology
Respirology (2013) 18, 848–856

7. dephosphorylation and redistribution.29
In this study,
CDCA stimulated PGE2 production, and the pattern of
CDCA- or PGE2-induced TER alteration and ZO-1
decay were found to be identical. We conclude that a
potential mechanism is through PGE2 production and
decay of JP. This was further confirmed by the fact that
dexamethasone and NS398 suppressed TER reduc-
tion and PGE2 production, and prevented the ZO-1
decay. In canine kidney cells, PGE2 alone reduces TER
and inactivates epidermal growth factor-induced Raf-
1/extracellular signal-regulated kinase 1/2 signal
transduction and the expression of claudin-4.27
In our
previous study, we also found that exposing fungal
antigens to human bronchial epithelial cells induced
PGE2 production and perturbed occludin expres-
sion.18
These findings suggest that PGE2 is associated
with expression of JP, and that different mechanisms
related to permeability-associated JP exist in discrete
types of epithelium.
cPLA2 can liberate arachidonic acid, which is
metabolized to generate prostaglandins and leukot-
rienes.33
The MAPK family is known to play a key role in
the phosphorylation of cPLA2.
34,35
In alveolar epithe-
lium, low magnitude stretch can induce cPLA2 phos-
phorylation independently through the extracellular
signal-regulated kinase kinase/extracellular signal-
regulated kinase and PI3K-Akt pathways.36
Our previ-
ous study has shown that neutrophil elastase
stimulated human bronchial epithelial cells to
produce PGE2 via p44/42, not p38.14
However, BA
stimulated human bronchial epithelial cells to induce
transforming growth factor-b production via p38,6,8
and HAEC to generate interleukin-8 via p38 and JNK,
rather than via p44/42.7
In this study, inhibition of
cPLA2, p38 and JNK reversed PGE2 production, and
could, therefore, potentially prevent the increase in
alveolar permeability.The MAPK family plays a pivotal
role in modulating lung inflammation, but the precise
signalling pathway that is activated may be deter-
mined in a stimulus- and epithelial tissue-specific
manner.
Gastroesophageal reflux and aspiration are the
major routes by which the lung epithelium is exposed
to BA, and the true prevalence of this may be under-
estimated.2
However, the causative concentration of
BA and the extent of lung injury is difficult to identify
in humans because of the large variation in the
amounts and contents of aspiration. The aspiration
contents may include gastric acids, non-acid contents
(such as BA and pepsin) and various microorgan-
isms.2
All of these have the potential to induce lung
injury with a single component alone or with a com-
Figure 7 Effect of inhibitors on chenodeoxycholic acid (CDCA)-
induced prostaglandin E2 (PGE2) production. After 1 h pretreat-
ment with mepacrine (1 mmo/L), NS398 (10 mmol/L), SP600125
(25 mmol/L) and SB203580 (10 mmol/L), the CDCA-induced PGE2
production was measured in the supernatants at 6 h. * P < 0.05,
vs CDCA alone. Values are means Ϯ standard error for three
separate experiments performed in triplicate.
Figure 8 Effect of chenodeoxycholic acid (CDCA) on junctional
proteins. After treatment with CDCA (150 mmol/L) (a) or prostag-
landin E2 (PGE2) (100 ng/mL) (b), cells were incubated for the time
indicated, then harvested for Western blotting of junctional pro-
teins. With 1 h pretreatment in the absence or presence of NS398
(1–10 mmol/L) or dexamethasone (100 mmol/L) (c) or SB203580
(10 mmol/L) or SP600125 (25 mmol/L) (d), cells were treated with
CDCA (150 mmol/L) for 3.5 h, then harvested for Western blotting
of zonula occludens-1 (ZO-1). Blots presented are representative
of three separate experiments.
K-C Su et al.854
© 2013 The Authors
Respirology © 2013 Asian Pacific Society of Respirology
Respirology (2013) 18, 848–856

8. bination of such substances. Our study documents
the critical role that BA play in HAEC injury, and we
suggest that these findings deserve further investiga-
tion by clinicians.
In conclusion, alteration in the BA-induced perme-
ability in HAEC occurs through the upregulation of
MAPK, cPLA2, COX-2, PGE2 and JP, which may con-
tribute to the pathogenesis of BA-associated lung
injury. Treatment with clinically available inhibitors,
such as dexamethasone, can reverse the process,
and thus has great promise as a future therapeutic
strategy.
Figure 9 Immunofluorescence of the
zonula occludens-1 (ZO-1) protein. Cells
were incubated in the absence (a) or
presence of chenodeoxycholic acid
(150 mmol/L) (b) or PGE2 (100 ng/mL) (c)
for 3.5 h. Arrowheads indicate the dis-
rupted ZO-1 band. No staining was
observed in negative control (d). ZO-1
protein is photographed by green fluores-
cence, and cell nuclei are counterstained
with propidium iodide. (magnification
400¥).
Figure 10 The signal pathway of cheno-
deoxycholic acid (CDCA)-induced altera-
tion in permeability. AA, arachidonic
acid; COX-2, cyclooxygenase-2; cPLA2.
cytosolic phospholipase A2; JNK, c-Jun
N-terminal kinase; PGE2, prostaglandin E2.
Bile acids increase lung permeability 855
© 2013 The Authors
Respirology © 2013 Asian Pacific Society of Respirology
Respirology (2013) 18, 848–856

9. Acknowledgements
This work was assisted in part by the Division of Experimental
Surgery of the Department of Surgery and Department of Pathol-
ogy and Laboratory Medicine, Taipei Veterans General Hospital.
This study was supported by research grants from Taipei
Veterans General Hospital (V96C1-065). We thank Miss Mo-Tzu
Wu, Miss Tun-Yun Hsueh and Dr Kuo-Ting Chang for their assist-
ance in lab work.
REFERENCES
1 Hills BA, Chen Y, Masters IB et al. Raised bile acid concentrations
in SIDS lungs at necropsy. Arch. Dis. Child. 1997; 77: 120–3.
2 Sweet MP, Patti MG, Hoopes C et al. Gastro-oesophageal reflux
and aspiration in patients with advanced lung disease. Thorax
2009; 64: 167–73.
3 Nagase T, Uozumi N, Ishii S et al. Acute lung injury by sepsis and
acid aspiration: a key role for cytosolic phospholipase A2. Nat.
Immunol. 2000; 1: 42–6.
4 Marik PE. Aspiration pneumonitis and aspiration pneumonia. N.
Engl. J. Med. 2001; 344: 665–71.
5 Zecca E, Costa S, Lauriola V et al. Bile acid pneumonia: a ‘new’
form of neonatal respiratory distress syndrome? Pediatrics 2004;
114: 269–72.
6 Perng DW, Chang KT, Su KC et al. Exposure of airway epithelium
to bile acids associated with gastroesophageal reflux symptoms:
a relation to transforming growth factor-beta1 production and
fibroblast proliferation. Chest 2007; 132: 1548–56.
7 Wu YC, Hsu PK, Su KC et al. Bile acid aspiration in suspected
ventilator-associated pneumonia. Chest 2009; 136: 118–24.
8 Perng DW, Wu YC, Tsai CC et al. Bile acids induce CCN2 produc-
tion through p38 MAP kinase activation in human bronchial epi-
thelial cells: a factor contributing to airway fibrosis. Respirology
2008; 13: 983–9.
9 Porembka DT, Kier A, Sehlhorst S et al. The pathophysiologic
changes following bile aspiration in a porcine lung model. Chest
1993; 104: 919–24.
10 Matthay MA, Folkesson HG, Clerici C. Lung epithelial fluid trans-
port and the resolution of pulmonary edema. Physiol. Rev. 2002;
82: 569–600.
11 Matthay MA, Clerici C, Saumon G. Invited review: active fluid
clearance from the distal air spaces of the lung. J. Appl. Physiol.
2002; 93: 1533–41.
12 Ware LB, Matthay MA. The acute respiratory distress syndrome.
N. Engl. J. Med. 2000; 342: 1334–49.
13 Terao Y, Nakamura T, Morooka H et al. Effect of
cyclooxygenase-2 inhibitor pretreatment on gas exchange after
hydrochloric acid aspiration in rats. J. Anesth. 2005; 19: 257–9.
14 Perng DW, Wu YC, Tsai MC et al. Neutrophil elastase stimulates
human airway epithelial cells to produce PGE2 through activa-
tion of p44/42 MAPK and upregulation of cyclooxygenase-2. Am.
J. Physiol. Lung Cell. Mol. Physiol. 2003; 285: L925–30.
15 Tilley SL, Coffman TM, Koller BH. Mixed messages: modulation
of inflammation and immune responses by prostaglandins and
thromboxanes. J. Clin. Invest. 2001; 108: 15–23.
16 Vancheri C, Mastruzzo C, Sortino MA et al. The lung as a privi-
leged site for the beneficial actions of PGE2. Trends Immunol.
2004; 25: 40–6.
17 Montuschi P, Kharitonov SA, Ciabattoni G et al. Exhaled leukot-
rienes and prostaglandins in COPD. Thorax 2003; 58: 585–8.
18 Tai HY, Tam MF, Chou H et al. Pen ch 13 allergen induces secre-
tion of mediators and degradation of occludin protein of human
lung epithelial cells. Allergy 2006; 61: 382–8.
19 N’Guessan PD, Hippenstiel S, Etouem MO et al. Streptococcus
pneumoniae induced p38 MAPK- and NF-kappaB-dependent
COX-2 expression in human lung epithelium. Am. J. Physiol.
Lung Cell. Mol. Physiol. 2006; 290: L1131–8.
20 Tsukita S, Furuse M, Itoh M. Multifunctional strands in tight
junctions. Nat. Rev. Mol. Cell Biol. 2001; 2: 285–93.
21 Van Itallie CM, Anderson JM. Claudins and epithelial paracellu-
lar transport. Annu. Rev. Physiol. 2006; 68: 403–29.
22 Cereijido M, Contreras RG, Shoshani L et al. Tight junction and
polarity interaction in the transporting epithelial phenotype.
Biochim. Biophys. Acta 2008; 1778: 770–93.
23 Kypta R, Bernfield M, Burridge K et al. Cell junctions, cell adhe-
sion, and the extracellular matrix. In: Alberts B, Johnson A, Lewis
J et al. (eds) Molecular Biology of the Cell, 4th edn. Garland
Science, New York, 2002; 1065–127.
24 Gumbiner B, Simons K. A functional assay for proteins involved
in establishing an epithelial occluding barrier: identification of a
uvomorulin-like polypeptide. J. Cell Biol. 1986; 102: 457–68.
25 Harhaj NS, Antonetti DA. Regulation of tight junctions and loss
of barrier function in pathophysiology. Int. J. Biochem. Cell Biol.
2004; 36: 1206–37.
26 Fink MP, Delude RL. Epithelial barrier dysfunction: a unifying
theme to explain the pathogenesis of multiple organ dysfunction
at the cellular level. Crit. Care Clin. 2005; 21: 177–96.
27 Flores-Benitez D, Rincon-Heredia R, Razgado LF et al. Control of
tight junctional sealing: roles of epidermal growth factor and
prostaglandin E2. Am. J. Physiol. Cell Physiol. 2009; 297: C611–20.
28 Tanaka MN, Diaz BL, de Souza W et al. Prostaglandin E2-EP1 and
EP2 receptor signaling promotes apical junctional complex dis-
assembly of Caco-2 human colorectal cancer cells. BMC Cell
Biol. 2008; 9: 63.
29 Raimondi F, Santoro P, Barone MV et al. Bile acids modulate
tight junction structure and barrier function of Caco-2 monolay-
ers via EGFR activation. Am. J. Physiol. Gastrointest. Liver Physiol.
2008; 294: G906–13.
30 Han B, Mura M, Andrade CF et al. TNFalpha-induced long pen-
traxin PTX3 expression in human lung epithelial cells via JNK. J.
Immunol. 2005; 175: 8303–11.
31 Perng DW, Wu YC, Chang KT et al. Leukotriene C4 induces TGF-
beta1 production in airway epithelium via p38 kinase pathway.
Am. J. Respir. Cell Mol. Biol. 2006; 34: 101–7.
32 Zhang F, Altorki NK, Wu YC et al. Duodenal reflux induces
cyclooxygenase-2 in the esophageal mucosa of rats: evidence for
involvement of bile acids. Gastroenterology 2001; 121: 1391–9.
33 Borsch-Haubold AG. Regulation of cytosolic phospholipase A2
by phosphorylation. Biochem. Soc. Trans. 1998; 26: 350–4.
34 Piomelli D. Arachidonic acid in cell signaling. Curr. Opin. Cell
Biol. 1993; 5: 274–80.
35 Lin LL, Wartmann M, Lin AY et al. cPLA2 is phosphorylated and
activated by MAP kinase. Cell 1993; 72: 269–78.
36 Letsiou E, Kitsiouli E, Nakos G et al. Mild stretch activates cPLA2
in alveolar type II epithelial cells independently through the
MEK/ERK and PI3K pathways. Biochim. Biophys. Acta 2011;
1811: 370–6.
Supporting information
Additional Supporting Information may be found in the online
version of this article at the publisher’s web-site:
Figure S1 Effect of CDCA on junctional proteins illustrated by
densitometry. After treatment with CDCA (150 mmol/L) (a) or PGE2
(100 ng/mL) (b), cells were incubated for the time indicated, then
harvested for Western blotting of junctional proteins. With 1 h
pretreatment in the absence or presence of NS398 (1–10 mmol/L)
or dexamethasone (100 mmol/L) (c) or SB203580 (10 mmol/L) or
SP600125 (25 mmol/L) (d), cells were treated with CDCA
(150 mmol/L) for 3.5 h, then harvested for Western blotting of
ZO-1. Densitometry readings were derived from the Western
blots in Figure 8 and expressed as fold of control. Data
(means Ϯ SE) are presented as fold of the control for three sepa-
rate experiments. *P < 0.05, ***P < 0.001.
Table S1 The PCR primers.
K-C Su et al.856
© 2013 The Authors
Respirology © 2013 Asian Pacific Society of Respirology
Respirology (2013) 18, 848–856