prostaglandins mediate the fetal pulmonary response to ... · prostaglandins mediate the fetal...

doi: 10.1152/ajplung.00297.2011302:L664-L678, 2012. First published 27 January 2012;Am J Physiol Lung Cell Mol Physiol

Alana J. Westover, Stuart B. Hooper, Megan J. Wallace and Timothy J. M. Mossintrauterine inflammationProstaglandins mediate the fetal pulmonary response to

You might find this additional info useful...

54 articles, 23 of which you can access for free at: This article citeshttp://ajplung.physiology.org/content/302/7/L664.full#ref-list-1

including high resolution figures, can be found at: Updated information and serviceshttp://ajplung.physiology.org/content/302/7/L664.full

can be found at: Molecular PhysiologyAmerican Journal of Physiology - Lung Cellular and about Additional material and information

http://www.the-aps.org/publications/ajplung

This information is current as of October 9, 2012.

ESSN: 1522-1504. Visit our website at http://www.the-aps.org/. Society, 9650 Rockville Pike, Bethesda MD 20814-3991. Copyright © 2012 the American Physiological Society. components of the respiratory system. It is published 24 times a year (twice monthly) by the American Physiologicalthe broad scope of molecular, cellular, and integrative aspects of normal and abnormal function of cells and

publishes original research coveringAmerican Journal of Physiology - Lung Cellular and Molecular Physiology

at Monash U

niversity on October 9, 2012

http://ajplung.physiology.org/D

ownloaded from

http://ajplung.physiology.org/

Prostaglandins mediate the fetal pulmonary responseto intrauterine inflammation

Alana J. Westover,1 Stuart B. Hooper,1,2 Megan J. Wallace,1,2 and Timothy J. M. Moss1,2

1Ritchie Centre, Monash Institute of Medical Research and 2Department of Obstetrics and Gynaecology, Monash University,Clayton, Victoria, Australia

Submitted 7 September 2011; accepted in final form 21 January 2012

Westover AJ, Hooper SB, Wallace MJ, Moss TJM. Prostaglan-dins mediate the fetal pulmonary response to intrauterine inflamma-tion. Am J Physiol Lung Cell Mol Physiol 302: L664–L678, 2012.First published January 27, 2012; doi:10.1152/ajplung.00297.2011.—Intra-amniotic (IA) lipopolysaccharide (LPS) induces intrauterine andfetal lung inflammation and increases lung surfactant and compliancein preterm sheep; however, the mechanisms are unknown. Prostaglan-dins (PGs) are inflammatory mediators, and PGE2 has establishedroles in fetal lung surfactant production. The aim of our first study wasto determine PGE2 concentrations in response to IA LPS and pulmo-nary gene expression for PG synthetic [prostaglandin H synthase-2(PGHS-2) and PGE synthase (PGES)] and PG-metabolizing [prosta-glandin dehydrogenase (PGDH)] enzymes and PGE2 receptors. Oursecond study aimed to block LPS-induced increases in PGE2 with aPGHS-2 inhibitor (nimesulide) and determine lung inflammation andsurfactant protein mRNA expression. Pregnant ewes received an IAsaline or LPS injection at 118 days of gestation. In study 1, fetalplasma and amniotic fluid were sampled before and at 2, 4, 6, 12, and24 h after injection and then daily, and fetuses were delivered 2 or 7days later. Amniotic fluid PGE2 concentrations increased (P � 0.05)12 h and 3–6 days after LPS. Fetal lung PGHS-2 mRNA and PGESmRNA increased 2 (P � 0.0084) and 7 (P � 0.014) days after LPS,respectively. In study 2, maternal intravenous nimesulide or vehicleinfusion began immediately before LPS or saline injection and con-tinued until delivery 2 days later. Nimesulide inhibited LPS-inducedincreases in PGE2 and decreased fetal lung IL-1� and IL-8 mRNA(P � 0.002) without altering lung inflammatory cell infiltration.Nimesulide decreased surfactant protein (SP)-A (P � 0.05), -B (P �0.05), and -D (P � 0.0015) but increased SP-C mRNA (P � 0.023).Thus PGHS-2 mediates, at least in part, fetal pulmonary responses toinflammation.

preterm lung; surfactant; chorioamnionitis; respiratory distress syn-drome; nimesulide

INTRAUTERINE INFLAMMATION is a major contributing factor tomany preterm deliveries (16). Preterm birth is typically asso-ciated with neonatal lung immaturity and increased risk ofrespiratory distress syndrome (RDS). However, epidemiologi-cal studies (2, 4) show that intrauterine inflammation decreasesthe risk of RDS. In sheep, intrauterine inflammation inducesstructural changes in the fetal lung and a profound increase inpulmonary surfactant, resulting in a marked improvement inneonatal lung function in preterm lambs (21, 38, 40). Theseresults demonstrate the likely biological mechanism for pro-tection against RDS in preterm infants after exposure to in-flammation in utero.

The underlying mechanisms by which inflammation in-duces preterm lung “maturation” are not known. It has beensuggested that inflammation may induce lung maturation byeliciting a fetal stress response, resulting in increased cor-tisol levels, known to have a critical role in lung maturation(30). However, it appears that inflammation-induced in-creases in surfactant in the preterm lungs occur via amechanism independent of glucocorticoid signaling. This isbecause the characteristics of the fetal lung response toinflammation are different from those of glucocorticoid-induced lung maturation in preterm lambs (20) and the smallincrease in fetal plasma cortisol induced by intra-amnioticLPS injection is unlikely to be sufficient to induce lungmaturation (42).

Prostaglandins are critical mediators of inflammation, andsome [particularly prostaglandin E2 (PGE2)] are known toincrease surfactant production by the preterm fetal lung (1, 36).Prostaglandins are produced from arachidonic acid by isoformsof the enzyme prostaglandin H synthase (PGHS). PGHS-1 isconstitutively expressed in most cells of the body and serves ahomeostatic role, whereas PGHS-2 is an inducible form of theenzyme responsible for production of prostaglandins in re-sponse to physiologic stresses such as inflammation (41, 53). Ingestational tissues and the fetus, a major PGHS-2 metabolite isPGE2, which is produced by PGE synthase (PGES). Thisenzyme is functionally coupled to PGHS-2 and is induced byinflammation. PGE2 signals via membrane-bound G-protein-coupled prostaglandin receptors that have been classified intofour subtypes, EP1–4 (46).

PGE receptors and the principal prostaglandin metaboliz-ing enzyme prostaglandin dehydrogenase (PGDH) are allpresent in the fetal lung and may mediate normal fetal lungdevelopment (10, 36, 46). Furthermore, in human fetal lungexplants dilation of terminal air sacs and differentiation ofthe pulmonary epithelium are accelerated by addition ofprostaglandins and inhibited by indomethacin, a nonspecificPGHS inhibitor (19). In vivo, administration of indometha-cin to pregnant sheep inhibits fetal pulmonary surfactantproduction, indicating a critical role for PGs in lung devel-opment (8, 28).

We aimed to determine if intrauterine inflammation in-duced by intra-amniotic LPS injection in pregnant sheep1) increases PGE2 concentrations in amniotic fluid, fetallung liquid, and the fetal circulation; and 2) alters mRNAlevels of prostaglandin synthetic and metabolic enzymes andprostaglandin receptors in the fetal lungs. In a separateexperiment, we aimed to determine if a preferential PGHS-2inhibitor, nimesulide, could 1) inhibit the inflammation-induced increases in PGE2 concentrations, 2) alter the in-

Address for reprint requests and other correspondence: T. J. M. Moss, theRitchie Centre, Monash Institute of Medical Research, PO Box 5418, Clayton,VIC, Australia 3168 (e-mail: [email protected]).

Am J Physiol Lung Cell Mol Physiol 302: L664–L678, 2012.First published January 27, 2012; doi:10.1152/ajplung.00297.2011.

1040-0605/12 Copyright © 2012 the American Physiological Society http://www.ajplung.orgL664

at Monash U



ownloaded from


flammatory response, and 3) alter the fetal lung response tointra-amniotic LPS injection. We hypothesized that expo-sure of preterm fetal sheep to intra-amniotic LPS wouldincrease PGE2 levels and alter gene expression of PGHS-2,PGES, PGDH, and the EP receptor subtypes in the pretermfetal lungs, indicating pulmonary PGs may mediate the fetallung response to intrauterine inflammation. Furthermore, wehypothesized that nimesulide infusion would inhibit LPS-induced PGE2 production, modulate the fetal pulmonaryinflammatory response, and reduce the effect of inflamma-tion on fetal lung development.

METHODS

Animal Experimentation

The relevant Monash University Animal Ethics Committee ap-proved all experimental procedures. All pregnant ewes underwentaseptic surgery at 112 days of gestation (term is �147 days) to havecatheters inserted into the fetal trachea, a jugular vein and carotidartery, the amniotic cavity, and a maternal jugular vein.

Study 1. On day 118 of gestation, ewes were randomly assignedto receive saline (4 ml: n � 12) or 20 mg LPS (Escherichia coli055:B5; Sigma) solubilized in 4 ml saline (n � 11) into theamniotic cavity via the indwelling catheter. Fetal lung liquid (3ml), amniotic fluid (3 ml), and arterial blood (3.5 ml) samples werecollected immediately before and at 2, 4, 6, 12, 24, and 48 h aftersaline (n � 6) or LPS (n � 5) administration in one (2-day) cohortof sheep, and immediately before and at 2, 4, 6, 12, and 24 h andthen daily until 7 days after intra-amniotic injection in another

(7-day) cohort (saline, n � 6; LPS n � 6). Three milliliters of eacharterial blood sample were centrifuged at 3,000 rpm for 10 min toisolate plasma; the remaining 0.5 ml were used for measurement ofblood gases, electrolytes, and metabolites (Radiometer ABL 500blood-gas analyzer; Radiometer, Copenhagen, Denmark). Fetalplasma, lung liquid, and amniotic fluid samples were mixed withindomethacin (20 �l) and immediately diluted 1:1 with 0.12 Mmethyloxyamine hydrochloride (Sigma) in sodium acetate buffer(1 M, pH 5.6) containing 10% ethanol for subsequent PGE2

measurement. Samples were left overnight at room temperatureand then stored at �20°C until analysis.

At 120 days (2-day cohort) or 125 days (7-day cohort), the ewe andfetus were humanely killed (pentobarbitone, 5 g; injected into thematernal jugular catheter). The fetus was delivered and weighed, andthe chest was opened. An endotracheal tube was tied into the fetaltrachea, and lung compliance was assessed by inflation with air to 40cmH2O pressure and sequentially lowering the pressure to 20, 10, 5,and 0 cmH2O while the corresponding lung volumes were recorded.The lungs were removed from the chest; the left lung was choppedand pieces were frozen in liquid nitrogen; and the right lung was fixedby bronchial instillation of 4% paraformaldehyde at 20 cmH2Opressure. A sample of chorioamnion was collected and fixed in 4%paraformaldehyde.

Study 2. On day 118 of gestation, ewes were randomly assigned toone of two cohorts. The first received a continuous maternal intrave-nous infusion of nimesulide (20 mg·kg�1·day�1) and a single injec-tion of either saline (4 ml; n � 6) or LPS (20 mg solubilized in 4 mlsaline; n � 6) into the amniotic sac. In the other cohort, ewes receiveda continuous maternal intravenous infusion of the vehicle solution(90% polyethylene glycol 400, 10% ethanol; 1 ml/h) and a single

Table 1. Oligonucleotide primer sequences

Gene Name and Genbank Accession Number Sequences (5=-3=)QuantitativeRT-PCR cDNA

Concentration, ng/�lAnnealing

Temperature, °C

18S (X01117) F: GTC TGT GAT GCC CTT AGA TGT C N/A* 60R: AAG CTT ATG ACC CGC ACT TAC

EP1 (AF035415) F: AGT CGT TCG CCA ACT CGT 1,000 59R: CCC AAG TGC TCC TGG TTA G

EP2 (AJ876410) F: GCT GGA TCA CTG GAA GTA TGC 1,000 60R: GTC TGC TTG TGT CCG ATG

EP3 (AF035417) F: CCG TTG CTG ATA ATG ATG TTG 1,000 59R: CTG TGT ATG TCT TGC AGT GCT C

EP4 (AJ617580) F: CCA TCG TGG TGC TCT GTA AG 1,000 59R: CCG CAT ACC AGT GTG TAG AAA G

IL-1� (NM_001009465) F: CGA TGA GCT TCT GTG TGA TG 1,000 59R: CTG TGA GAG GAG GTG GAG AG

IL-6 (NM_001009392) F: CGC AAA GGT TAT CAT CAT CC 1,500 60R: CCC AGG AAC TAC CAC AAT CA

IL-8 (NM_001009401) F: CCT CAG TAA AGA TGC CAA TGA 1,000 59R: TGA CAA CCC TAC ACC AGA CC

PGDH (NM_001034419) F: GCA CAG CAG CCT GTT TAT TG 1,000 60R: CAT GAG ATT AGC AGC CAT CG

PGES (NM_174443) F: AGT GAG GCT GCG GAA GAA 500 60R: CCA CAT CTG GGT CGT TCC

PGHS-2 (NM_001009432) F: GGT CGC TGG TCG TCG GAA 1,000 60R: CTC TCT GCT CTG GTC AAG TGA A

SP-A (AF211856) F: CAT CAA GTC CTG CAG TCA CA 50 60R: GCC CAT TGG TAG AGA AGA CC

SP-B (AF035417) F: GTC CTC TGC TGG ACA AGA TG 100 59R: GGA GAG GTC CTG TGT CTG AG

SP-C (AF076634) F: GTG AAC ATC AAA CGC CTT C 200 58R: TGT GAA GAC CCA TGA GCA

SP-D (AJ133002) F: ATG ACC GAT ACC AGG AAG GA 200 60R: GCC CAG TTG GAA TAG ACC AG

EP, prostaglandin E receptor; PGDH, prostaglandin dehydrogenase; PGES, prostaglandin E synthase; PGHS, prostaglandin H synthase; SP, sur-factant protein; F, forward primer; R, reverse primer. *PCR for 18S rRNA was performed using the same primer and cDNA concentrations as the geneof interest.

L665PROSTAGLANDINS IN FETAL LUNG INFLAMMATION IN SHEEP

AJP-Lung Cell Mol Physiol • doi:10.1152/ajplung.00297.2011 • www.ajplung.org

at Monash U



ownloaded from


injection of either saline (n � 6) or LPS (n � 5), as above, into theamniotic sac. Nimesulide [N-(4- nitro-2-phenoxyphenyl)-methanesul-fonilamide] is a nonsteroidal anti-inflammatory drug that preferen-tially inhibits PGHS-2 relative to PGHS-1 activity to the order of 733in sheep (22) and 70 in human enzymes (5). Indeed, nimesulide isparticularly selective in sheep and has a lower IC50 for ovine (0.03 at100 �M arachidonic acid; Ref. 22) than human PGHS-2 (1.27 at 100�M arachidonic acid; Ref. 5). A continuous infusion of nimesulide(20 mg·kg�1·day�1) to laboring ewes is known to greatly reduce fetalPGE2 levels without adverse fetal effects (17) that occur with inhibi-tion of constitutively generated prostaglandins using nonspecificPGHS inhibitors (43). Nimesulide or vehicle infusions commenced 1h before intra-amniotic LPS or saline injection.

Amniotic fluid (3 ml) and arterial blood (3.5 ml) samples werecollected immediately before and at 2, 4, 6, 12, 24, and 48 h aftersaline or LPS administration. Samples were processed as in study 1 formeasurement of blood gases, electrolytes, and PGE2 concentrations.

At 120 days, the ewe and fetus were humanely killed for the sameassessments described in study 1.

PGE2 Radioimmunoassay

PGE2 concentrations were measured in amniotic fluid andplasma samples by radioimmunoassay (7, 13). Assay mixtures

consisted of samples (0.2 ml) or standard PGE2 concentrationsdissolved in 0.2 ml of charcoal-stripped plasma. All standards andsamples were assayed in duplicate after incubation at 4°C over-night with 3H-PGE2 tracer (�5,000 cpm; 0.1 ml/tube) and antise-rum raised in sheep against the methyl oxime of PGE2. Thecross-reactivities of the antiserum at 50% displacement were 4.9%

Table 2. Study 1: Fetal blood gases

Fetal Arterial Blood Measurement/Cohort/Treatment

pH2 day

Saline 7.36 � 0.004LPS 7.37 � 0.004

7 daySaline 7.36 � 0.002LPS 7.37 � 0.002

PaCO2, mmHg2 day

Saline 45.66 � 0.72LPS 47.48 � 0.33

7 daySaline 51.92 � 0.49LPS 50.51 � 0.30

PO2, mmHg2 day

Saline 22.10 � 0.42LPS 20.14 � 0.67

7 daySaline 21.47 � 0.25LPS 23.24 � 0.54

tHB, g/dl2 day

Saline 10.33 � 0.10LPS 10.44 � 0.16

7 daySaline 11.06 � 0.06LPS 10.52 � 0.22

Glucose, mmol/l2 day

Saline 0.93 � 0.04LPS 0.86 � 0.03

7 daySaline 0.69 � 0.03LPS 0.74 � 0.03

Lactate, mmol/l2 day

Saline 1.30 � 0.03LPS 1.74 � 0.11

7 daySaline 1.41 � 0.03LPS 1.43 � 0.10

Values are means � SE across all time points. tHB, hemoglobin.

Fig. 1. Study 1: effect of intra-amniotic LPS on expression of mRNA forproinflammatory cytokines in the fetal lung. Levels of interleukin (IL)-1� (A), IL-6(B), and IL-8 (C) mRNA, as determined by quantitative RT-PCR, were highest infetal lungs from the 2-day LPS group. Data are expressed as means � SE. *P �0.05 vs. saline controls in the same cohort.

L666 PROSTAGLANDINS IN FETAL LUNG INFLAMMATION IN SHEEP


at Monash U



ownloaded from


with PGE1, 0.47% with 15-keto-PGE2, 0.15% with 13,14-dihydro-15-keto-PGE2, 0.11% with PGA2, and �0.006% with PGD2,PGF1�, PGF2�, 6-keto-PGF1�, and thromboxane B2. Intra-amnioticLPS resulted in an increase in the viscosity of lung liquid, render-ing volume measurement inaccurate at early time points and thuspreventing measurement of PGE2 concentrations in fetal lungliquid at these times. By 7 days after intra-amniotic LPS, lungliquid viscosity had normalized, allowing a single measurement atthis time. In study 1, the sensitivities of the amniotic fluid, lungliquid, and plasma assays were 5.4, 3.2, and 4.4 ng/ml, respec-tively. In study 2, sensitivities of the amniotic fluid and fetalplasma assays were 1.3 and 1.5 ng/ml, respectively.

Lung Morphology

Study 1. Secondary septal crests project out from the primarysaccule walls into the lung lumen to form alveoli and can beidentified by elastin deposition (12, 14). The relative volumedensity of secondary septal crests is a relative measure of alveolarnumber per unit volume of lung. Blocks of lung tissue from theright upper, middle, and lower lung lobes were embedded inparaffin, cut into 5-�m sections, and mounted onto glass slides.Two sections from each of the right lung lobes from each fetuswere stained with Hart’s resorcin-fuchsin stain for elastin, incu-bated for 4 h, and counterstained with tartrazine (0.25%) insaturated picric acid to identify secondary septal crests. Fivenonoverlapping fields from each section were captured at 40magnification, and a linear point-counting grid (19 26) wassuperimposed to determine relative volume density of secondaryseptal crests for each lobe; this was expressed relative to the tissuevolume. Similarly, a point-counting technique was used to deter-mine the ratio of tissue to air space volumes for each lung lobe.Lung lobes were analyzed separately to identify potential regionaldifferences; none were identified so data were combined to providea whole lung value. Lung morphometric analysis was not per-formed for study 2.

Immunohistochemistry

Immunostaining of CD45 (a protein tyrosine phosphatase ex-pressed on the surface of leukocytes; Ref. 52) was performed toassess inflammation. Lung and chorioamnion tissue sections weremounted as above, paraffin was removed using xylene, and sec-tions were rehydrated in graded alcohol. Antigen retrieval wasperformed by incubation with 0.01 M sodium citrate (pH 6) in amicrowave for 20 min at 900 W, and endogenous peroxidase wasblocked by incubation with 3% H2O2 in distilled water for 20 min.The slides were incubated in 10% normal horse serum for 30 minto block nonspecific binding and then incubated at 4°C overnightwith mouse anti-ovine CD45 primary antibody (MCA2220PE;AbD Serotec) diluted 1:50 in antibody diluent (S0809; Dako).Unbound antibody was removed by rinsing with PBS/0.1% Tween20, and the slides were incubated with the secondary antibody,biotinylated goat anti-mouse IgG (Dako) diluted 1:500 in antibodydiluent, for 1 h, and then in ABC complex (Vectastain ABC kit;Vector Laboratories) for 30 min. Immunostaining was visualizedby incubation with 3,3=-diaminobenzidine (D4293; Sigma) for 15min, and sections were counterstained with hematoxylin. Imagesfrom five nonoverlapping fields from each section (2 sections fromeach of the right lung lobes and 2 sections of chorioamnion perfetus) were viewed under a light microscope and captured at 40magnification using a digital camera (Nikon Eclipse 80i; CoherentScientific). Quantification of CD45 immunohistochemistry wasperformed using the Image-Pro Plus (Media Cybernetics) softwareprogram. For each field of view, the number of CD45-positive cellswas counted. Mean values for each tissue specimen were calcu-lated. Lung lobes were analyzed separately, but data were com-bined to provide a single measure for each subject.

Real-time PCR

Total RNA was extracted from frozen portions of the left lung lobewith the RNeasy Maxi kit (75162; Qiagen) according to the manu-facturer’s instructions. Reverse transcription was performed on 1 �gof RNA with the Superscript III First Strand synthesis system forreal-time PCR kit as specified by the manufacturer (18080–051;Invitrogen).

Quantitative RT-PCR was used to measure gene expression asdescribed by Sozo et al. (49) under optimized primer-specific condi-tions (Table 1).

Study 1. The messenger RNA levels for each fetus were normal-ized to the 18S rRNA values for that fetus and are expressedrelative to the mean mRNA levels for that gene in the saline controlfetuses.

Study 2. When all samples would not fit onto one PCR plate,amplification of the gene of interest and 18S was performed on acalibrator cDNA sample in quintuplicate in each assay to permitcomparisons between all groups. The messenger RNA levels foreach fetus were normalized to the 18S rRNA values for that fetusand are expressed relative to the mean mRNA levels in vehicle saline fetuses.

Statistical Analysis

Results are presented as means � SE. Analysis of blood gasesand electrolytes, radioimmunoassay data, and lung pressure-volume curves was by two-way repeated measures ANOVA withpost hoc comparisons performed using Fisher’s least significancedifference (LSD) test. Area under the curve was calculated usingthe trapezoid rule (GraphPad Software). Body and organ weightswere compared between control and LPS groups by Student’st-test.

Study 1. Analysis of lung morphology was by two-way ANOVAwith post hoc comparisons made using Fisher’s LSD test. Com-parisons of mRNA levels and day 7 lung liquid PGE2 concentra-tions between control and LPS groups were by Student’s t-test.Statistical comparisons between 2-day and 7-day cohorts were notperformed.

Study 2. Comparisons of mRNA levels between treatment groupsand infusion cohorts were by two-way ANOVA with post hoccomparisons performed using the Fisher LSD test.

RESULTS

Study 1

Fetal blood gases and organ weights. Intra-amniotic LPSdid not affect fetal pH, PaCO2

, PaO2, hemoglobin, or fetal blood

Table 3. Study 1: CD45 expression in the chorioamnionand fetal lung

Tissue/ Cohort/TreatmentNo. CD45 Cells/Field

of ViewP Value (Compared with

Corresponding Saline Group)

Chorioamnion2 day

Saline 2.30 � 0.55LPS 13.87 � 1.48 0.0002

7 daySaline 0.53 � 0.46LPS 10.17 � 2.19 0.016

Lung2 day

Saline 2.71 � 0.26LPS 13.95 � 0.89 �0.0001

7 daySaline 0.76 � 0.22LPS 5.92 � 0.49 �0.0001



at Monash U



ownloaded from


glucose concentrations (Table 2). Fetal blood lactate concen-trations were significantly elevated 12 h after LPS administra-tion (LPS: 2.24 � 0.27 mmol/l; saline: 1.35 � 0.076 mmol/l;P � 0.007) but returned to control levels by 24 h. There wereno statistically significant differences in fetal body weights or

in lung, heart, liver, spleen, thymus, or kidney-to-body weightratios between groups in either cohort. Brain-to-body weightratios were significantly lower (P � 0.013) after intra-amnioticLPS in the 7-day cohort but were not different between groupsin the 2-day cohort.

Fig. 2. Study 1: effect of intra-amniotic LPS on preterm lung compliance and surfactant protein mRNA levels. A: deflation limbs of pressure-volume curves showgreater lung gas volumes at 7 days but not 2 days after LPS administration. B: there was increased expression of surfactant protein (SP)-A mRNA in the fetallungs at 2 and 7 days after LPS. C: SP-B mRNA tended to be elevated 2 days after LPS but was not significantly different between groups in either cohort.D: expression of SP-C was elevated at 2 and 7 days after LPS. E: SP-D mRNA was elevated 2 days after LPS. Data are expressed as means � SE. Data foreach cohort are expressed relative to the mean value for the saline group. *P � 0.05 vs. saline controls in the same cohort.



at Monash U



ownloaded from


Inflammation in the lung and chorioamnion. Fetal lunginterleukin (IL)-1� mRNA levels were significantly higher 2days (Fig. 1A; P � 0.0092) and 7 days (Fig. 1A; P � 0.045)after LPS, compared with saline groups. IL-6 (Fig. 1B) andIL-8 (Fig. 1C) mRNA levels in fetal lung tissue tended to behigher in LPS groups, but values were not significantly differ-ent from saline groups.

The number of CD45-positive cells per field of view in thechorioamnion and the fetal lungs was higher in the 2-day and7-day LPS groups compared with corresponding saline groups(Table 3; P � 0.02).

Fig. 3. Study 1: effect of intra-amniotic LPS on lung morphometry. Secondaryseptal crest densities were not different between saline and LPS groups in the2-day cohort. However, there were significantly more septal crests in the lungs7 days after LPS than in the respective saline group, and a significantly greaterseptal crest density in lungs from the 7-day cohort than the 2-day cohort (A).Tissue-to-air space ratio was not different between groups in the 2-day cohortand increased with increasing gestational age in both 7-day groups (B).However, LPS significantly reduced the normal gestational-age related in-crease. Data are expressed as means � SE. *P � 0.05 vs. saline.

Fig. 4. Study 1: effect of intra-amniotic LPS on expression of prostaglandinsynthesizing and metabolizing enzymes in the fetal lungs. Relative abundanceof prostaglandin H synthase type-2 (PGHS-2; A) mRNA was higher thancontrol 2 days after LPS and prostaglandin E synthase (PGES; B) mRNA waselevated 7 days after LPS administration. C: expression of prostaglandindehydrogenase (PGDH) mRNA in the fetal lung was not different betweengroups. Data for each cohort (means � SE) are expressed relative to the meanvalue for the saline group. *P � 0.05 vs. saline controls in the same cohort.



at Monash U



ownloaded from


Lung compliance and surfactant protein mRNA. Deflationlimbs of pressure-volume curves showed greater lung gasvolumes in the LPS group than saline group in the 7-day cohort(P � 0.002), but no difference between LPS and saline groups2 days after intra-amniotic injection (Fig. 2A). Surfactantprotein (SP)-A (Fig. 2B; 2-day cohort: P � 0.028, 7-daycohort: P � 0.033) and SP-C (Fig. 2D; 2-day cohort: P �0.034, 7-day cohort: P � 0.021) mRNA levels were higher in2-day and 7-day LPS groups than controls. SP-B (Fig. 2C)mRNA levels tended to be higher 2 days after LPS (P � 0.058)but were not significantly different between groups in eithercohort. SP-D (Fig. 2E; P � 0.0004) mRNA levels were higher2 days, but not 7 days, after LPS than in the correspondingsaline group.

Lung morphometry. The relative volume density of second-ary septal crests in the fetal lungs was significantly higher inthe 7-day LPS group than in the 7-day saline group (Fig. 3A;P � 0.027). Values were not different between groups in the2-day cohort, but there were significantly more septal crests infetal lungs from the 7-day cohort than the 2-day cohort (Fig.3A; P � 0.0001). The tissue-to-air space volume ratio was notdifferent between saline and LPS groups in the 2-day cohort(Fig. 3B). The tissue-to-air space ratio was significantly lower

after LPS in the 7-day cohort (Fig. 3B; P � 0.0001) and wasgreater in the 7-day cohort than the 2-day cohort (Fig. 3B; P �0.0001).

Prostaglandin-metabolizing enzymes. The relative abun-dance of PGHS-2 mRNA in the lungs was significantly higherin the LPS group than in the saline group for the 2-day cohort(Fig. 4A; P � 0.0084). PGES gene expression in the lung wassignificantly higher in the 7-day LPS group than the 7-daysaline group (Fig. 4B; P � 0.014). No significant differences inPGDH expression were observed between LPS and salinegroups in either cohort (Fig. 4C).

PGE receptor subtypes. Expression of mRNA for the PGEreceptors EP1–3 (Figs. 5, A–C) in the fetal lung was notsignificantly different between LPS and saline groups in eitherthe 2-day or 7-day cohort. However, EP4 receptor mRNAexpression in the fetal lung was significantly higher in the2-day LPS group than the corresponding saline group (Fig. 5D;P � 0.011); EP4 mRNA levels tended to be higher 7 days afterLPS, although expression was not significantly different be-tween groups (P � 0.11).

PGE2 concentrations. Mean amniotic fluid PGE2 concentra-tions increased slightly but not significantly within the first 24h after LPS but were significantly higher in the 7-day LPS

Fig. 5. Study 1: effect of intra-amniotic LPS on expression of PGE receptor (EP) subtypes in the fetal lungs. PGE receptor-1 (A), 2 (B), and 3 (C) mRNA levelsin the lung were not significantly different between groups in either cohort. D: PGE receptor-4 mRNA was higher than control 2 days after LPS. Data areexpressed as means � SE. *P � 0.05 vs. saline controls in the same cohort.



at Monash U



ownloaded from


group than in the saline group at 2–6 days after intra-amnioticinjection (Fig. 6A; P � 0.05). The cumulative PGE2 response(area under the curve) was significantly higher (P � 0.039) inthe 7-day LPS group (3,726 � 854 nM/168 h) than in the 7-daysaline group (1,170 � 476 nM/168 h).

Fetal plasma PGE2 concentrations tended to be higher in7-day LPS-exposed fetuses, compared with the 7-day salinegroup, from 12 h to 2 days after intra-amniotic injection butvalues were not significantly different between groups at any ofthe time points (Fig. 6B). Area under the curve was notsignificantly different between 7-day saline (160 � 59 nM/168h) and LPS groups (361 � 151 nM/168 h).

The PGE2 concentration in fetal lung fluid 7 days afterintra-amniotic LPS (3.03 � 0.43 nM) was higher than in the7-day saline group (1.58 � 0.85 nM), but values were notsignificantly different (P � 0.081).

Study 2

Fetal blood gases and organ weights. Intra-amniotic LPS ormaternal nimesulide infusion did not affect fetal pH, PaCO2

,

PaO2, or fetal blood hemoglobin or glucose concentrations

(Table 4). In the vehicle cohort, fetal blood lactate concentra-tions were significantly elevated at 6 h (LPS: 2.40 � 0.49mmol/l; saline: 1.08 � 0.11 mmol/l; P � 0.021) and 12 h(LPS: 1.93 � 0.24; saline: 1.07 � 0.033 mmol/l; P � 0.023)after LPS administration. There were no statistically significanttreatment or infusion effects on fetal body weights or on lung,heart, liver, spleen, or kidney-to-body weight ratios; however,nimesulide caused a significant increase (P � 0.045) in thy-mus-to-body weight ratios.

Inflammation in the lung and chorioamnion. Fetal lungIL-1� (P � 0.002; Fig. 7A) and IL-8 (P � 0. 0017; Fig. 7C)mRNA levels were significantly lower in the nimesulide co-hort, for both saline and LPS groups, compared with thevehicle cohort. IL-6 (Fig. 7B) mRNA levels in fetal lung tissuewere not significantly different between saline and LPS groupsin either vehicle or nimesulide cohorts, or between cohorts.

The number of CD45-positive cells per field of view wassignificantly higher (P � 0.0008) in the chorioamnion afterLPS in the vehicle cohort but not after maternal nimesulide

Table 4. Study 2: Fetal blood gases

Fetal Arterial Measurement/Cohort/Treatment

pHVehicle

Saline 7.37 � 0.003LPS 7.34 � 0.01

NimesulideSaline 7.31 � 0.007LPS 7.38 � 0.005

PaO2, mmHgVehicle

Saline 41.43 � 1.34LPS 42.23 � 0.84


PaO2, mmHgVehicle

Saline 22.04 � 0.37LPS 20.43 � 0.80


tHB, g/dlVehicle

Saline 10.34 � 0.18LPS 11.92 � 0.30


Glucose, mmol/lVehicle

Saline 0.77 � 0.03LPS 0.90 � 0.04


Lactate, mmol/lVehicle

Saline 1.06 � 0.03LPS 1.88 � 0.90


Values are means � SE across all time points.

Fig. 6. Study 1: effect of intra-amniotic LPS on amniotic fluid and fetal plasmaPGE2 concentrations. Intra-amniotic (IA) LPS significantly increased PGE2

concentrations from 2–6 days from administration in amniotic fluid (A) but notin fetal plasma (B). Data are expressed as means � SE. *P � 0.05 vs. saline.



at Monash U



ownloaded from


infusion (Table 5). In the fetal lungs, LPS significantly in-creased (P � 0.0001) the number of CD45-positive cells perfield of view in both vehicle and nimesulide cohorts (Table 5).

Surfactant protein mRNA. Intra-amniotic LPS significantlyincreased SP-A (Fig. 8A; P � 0.013) and SP-B (Fig. 8B; P �0.031) mRNA levels in the vehicle and nimesulide cohorts, butthere was a strong trend for levels to be lower after nimesulideinfusion than vehicle (P � 0.05 for SP-A and SP-B). SP-C(Fig. 8C) and SP-D (Fig. 8D) mRNA levels were not differentbetween saline and LPS groups in either cohort. However,nimesulide significantly increased SP-C (P � 0.023) geneexpression but caused a profound reduction in SP-D mRNAexpression (P � 0.0015) compared with vehicle.

Prostaglandin-metabolizing enzymes. PGHS-2 mRNA lev-els were similar in LPS and saline groups in the nimesulidecohort but were 2.4-fold higher in the LPS group comparedwith saline in the vehicle cohort, although this difference wasnot statistically significant (Fig. 9A). PGES gene expression inthe lung was significantly higher (P � 0.049) after LPS in boththe vehicle and nimesulide cohorts but was not significantlydifferent between cohorts (Fig. 9B). PGDH mRNA levels werenot different between LPS and saline groups in either cohort,but nimesulide significantly reduced (P � 0.008) PGDHmRNA levels compared with vehicle (Fig. 9C).

PGE receptor subtypes. Expression of mRNA for the PGEreceptors EP1–4 in the fetal lung was not significantly differentbetween LPS and saline groups in either the vehicle or nime-sulide cohort (Fig. 10, A-D). However, nimesulide significantlyreduced mRNA levels of EP1 (P � 0.002), EP2 (P � 0.014),and EP3 (P � 0.005), but not EP4, compared with vehicle.

PGE2 concentrations. In the vehicle cohort, mean amnioticfluid PGE2 concentrations were significantly higher at 24 (P �0.023) and 48 (P � 0.013) h after intra-amniotic LPS, com-pared with the saline group (Fig. 11A). Amniotic fluid PGE2

concentrations at 24 h (P � 0.018) and 48 h (P � 0.0002) afterLPS were significantly lower in the nimesulide cohort than inthe vehicle cohort. The area under the curve was not signifi-cantly different between groups (vehicle saline: 122 � 20nM/48 h; vehicle LPS: 386 � 116 nM/48 h; nimesulide saline: 166 � 41 nM/48 h; nimesulide LPS: 201 � 85nM/48 h).

Fig. 7. Study 2: effect of nimesulide infusion on inflammation-inducedexpression of mRNA for proinflammatory cytokines in the fetal lungs.Levels of IL-1� (A), IL-6 (B), and IL-8 (C) mRNA in the fetal lungs werenot different between saline and LPS groups. However, there was asignificant decrease in IL-1� and IL-8 gene expression in the nimesulidecohort compared with the vehicle cohort. Data are expressed as means �SE. **P � 0.01.

Table 5. Study 2: CD45 expression in the chorioamnion andthe fetal lung

Tissue/Infusion/TreatmentNo. CD45 Cells/Field

of ViewP Value

(Infusion)P Value

(Treatment)

ChorioamnionVehicle

Saline 1.15 � 0.33LPS 9.94 � 2.17

NimesulideSaline 3.57 � 1.04LPS 4.95 � 1.28 0.33 0.0008

LungVehicle

Saline 1.54 � 0.30LPS 6.4 � 2.29

NimesulideSaline 3.21 � 1.05LPS 6.98 � 0.58 0.35 �0.0001

Values are means � SE.



at Monash U



ownloaded from


Fetal plasma PGE2 concentrations were not significantlydifferent between saline and LPS groups in either cohort;however, nimesulide significantly reduced fetal plasma PGE2

concentrations from vehicle levels at 4 (P � 0.044), 6 (P �0.017), and 12 (P � 0.048) h after LPS (Fig. 11B). Intra-amniotic LPS did not significantly alter the area under thecurve in either cohort, but nimesulide significantly reduced(P � 0.024) the area under the curve compared with levels inthe vehicle groups (vehicle saline: 51 � 17 nM/48 h; vehicle LPS: 163 � 66 nM/48 h; nimesulide saline: 5 � 5 nM/48 h;nimesulide LPS: 1 � 1 nM/48 h).

DISCUSSION

Increases in amniotic fluid PGE2 concentrations and ele-vated PGHS-2, PGES, and EP4 mRNA levels in the fetal lungsaccompany intra-uterine inflammation and consequent altera-tions in fetal lung development. Maternal nimesulide infusionreduced fetal lung inflammation and altered mRNA levels ofprostaglandin synthetic and metabolic enzymes, EP receptormRNA, and SP mRNA levels in the fetal lungs. These obser-vations demonstrate that PGHS-2-induced increases in PGE2

play a role in mediating inflammation-induced alterations infetal lung development.

The effect of intra-amniotic LPS on fetal lung inflammationobserved in study 1 is consistent with previous studies (26, 29).

Similarly, the changes we observed in pressure-volume curvesand surfactant protein mRNA in study 1 reflect the functional“maturation” of the lungs demonstrated in these previousstudies (26, 29), which is consistent with a reduced risk ofrespiratory distress syndrome in preterm infants born afterexposure to inflammation in utero (2, 31).

Disparate studies have shown that prostaglandins have arole in regulating pulmonary surfactant synthesis and secre-tion. Human fetal lung epithelial cells spontaneously syn-thesize SP-A in vitro, and its mRNA levels can be decreasedby incubation with the PGHS inhibitor indomethacin andincreased by PGE2 (1). Surfactant lipid secretion in newbornrabbit lung slices is also inhibited by indomethacin andstimulated by PGE2 (34). In vivo, administration of indo-methacin to pregnant rabbits (8) or sheep (28) appears toinhibit fetal pulmonary surfactant production, as indicatedby decreased saturated phosphatidylcholine concentrations.We aimed to inhibit LPS-induced PGE2 production to elu-cidate the role of prostaglandins in the fetal lung’s responseto inflammation. The increase in PGE2 concentrations in theamniotic fluid 48 h after LPS was inhibited by nimesulide,and circulating PGE2 concentrations were reduced to lowlevels 4 –12 h after LPS injection and remained lower thanvehicle responses for the duration of the maternal nimesu-lide infusion.

Fig. 8. Study 2: effect of nimesulide infusion on inflammation-induced expression of surfactant proteins in the fetal lungs. LPS increased expression of SP-A (A)and SP-B (B) mRNA in the fetal lungs, but mRNA levels were lower in the nimesulide groups (P � 0.05). Nimesulide infusion significantly increased SP-C(C) mRNA levels in fetal lungs and significantly reduced SP-D (D) gene expression from responses in the vehicle cohort. Data are expressed as means � SE.Data are expressed relative to the mean value for the vehicle saline group. *P � 0.05; ** P � 0.001.



at Monash U



ownloaded from


The 2-day experimental timeline of study 2 allowed us toinvestigate events when intra-amniotic LPS-induced increasesin PGE2 concentrations first occur in the amniotic fluid andwhen the inflammatory response to intra-amniotic LPS isgreatest (29). In study 2, intra-amniotic LPS did not induce anincrease in IL-1�, IL-6, or IL-8 mRNA levels in the fetal lungs,in contrast to the increase in IL-1� and IL-8 mRNA levels 2days after LPS in study 1. It is possible that an increase in basallevels of IL-1� and IL-8 mRNA levels in the vehicle cohortmasked an LPS-induced increase in cytokine expression be-cause PEG-400 has been observed to stimulate cytokine ex-pression (47). Regardless, nimesulide infusion abolished IL-1�gene expression and significantly reduced IL-8 mRNA levelscompared with vehicle-infused animals. This is consistent withstudies (32) in nimesulide-treated human macrophages, whichdemonstrated attenuated cytokine expression in response toinfluenza virus. Our observed decrease in IL-1� and IL-8mRNA levels in nimesulide-treated saline and LPS groupsindicates a reduction in the basal inflammatory state of the fetallungs when PGHS-2 is inhibited. We observed equivalentpulmonary inflammatory cell recruitment after LPS exposurein vehicle and nimesulide groups, despite different levels ofmRNA for the chemoattractant cytokine IL-8. This is consis-tent with previous studies (24) in sheep demonstrating that fetalpulomonary inflammatory cell recruitment is not IL-8 depen-dent. Additionally, lung inflammatory cell recruitment is ele-vated in adult PGHS-2 knockout mice (relative to wild type) inresponse to allergen exposure (9, 15), demonstrating a role forPGHS-2 in pulmonary inflammatory cell recruitment. Thereare likely differences between pharmacological PGHS-2 inhi-bition in fetal sheep and gene deletion in adult mice, but ourdata suggest that inflammatory cell recruitment in the lungs offetal sheep is not mediated by PGHS-2, suggesting differencesbetween fetal and adult animals in this respect.

Intra-amniotic LPS injection increased PGHS-2 mRNA3-fold in fetal lung tissue after 2 days, which returned near tocontrol levels by 7 days after treatment. This time course isconsistent with the acute inflammatory response of the fetalsheep lungs to intra-amniotic injection of LPS (23) and theestablished role of PGHS-2 in inflammation (11, 56). Ourobservation confirms previous data from fetal sheep showing asimilar time course of elevation in pulmonary PGHS-2 mRNAlevels within the first 3 days after intra-amniotic LPS (39). Inthat study, the levels of PGHS-2 protein were significantlygreater than control 5 h after intra-amniotic LPS and tended tobe higher 24 h after treatment before returning to baselinelevels 3 days after endotoxin exposure (39). We did notmeasure PGHS-2 protein levels because of this previouslydemonstrated discordance between mRNA and protein levels.PGHS-2 undergoes rapid protein degradation as a consequenceof suicide inactivation (48, 55); thus, protein levels may notaccurately reflect enzyme activity and local tissue prostaglan-din production. The increased PGE2 synthetic capacity of fetallung tissue is indicated by elevated PGHS-2 mRNA levels andis supported by our demonstrated increase in PGES mRNAlevels 7 days after intra-amniotic injection of LPS. Fetal lungprostaglandin metabolism is unlikely to be altered by inflam-mation since PGDH mRNA levels were not changed by intra-amniotic LPS. We did not attempt to measure lung tissue PGE2

levels because such measurements cannot be simply performedin tissue homogenates. Tissue disruption in-and-of-itself would

Fig. 9. Study 2: effect of nimesulide infusion on expression of prostaglandinsynthesizing and metabolizing enzymes in the fetal lungs. Relative abundanceof PGHS-2 (A) mRNA was not different between saline and LPS groups orafter vehicle or nimesulide maternal infusion. Expression of PGES (B) mRNAwas elevated after LPS in both vehicle and nimesulide cohorts. PGDH mRNAlevels in the fetal lung were significantly lower after nimesulide maternalinfusion than vehicle maternal infusion. Data are expressed as means � SE.Data are expressed relative to the mean value for the vehicle saline group.*P � 0.05; ** P � 0.001.



at Monash U



ownloaded from


liberate significant amounts of prostaglandins, and their labilityin vivo means that concentrations measured in such an assaywould unlikely reflect actual fetal lung levels. Rather thanperforming tissue analyses to investigate the role of prostaglan-dins in the fetal lung response to inflammation, we aimed toinhibit prostaglandin production in vivo and assess effects onfetal lung inflammation and development. Although nimesu-lide selectively inhibits PGHS-2 relative to PGHS-1, and thisselectivity is greater in sheep than humans (5, 22), there areeffects of the drug not mediated by PGHS inhibition (6). Thusthe effects in our experiment cannot be solely attributed toPGHS-2 inhibition; non-PGHS-mediated effects may thereforeaccount for at least some of the differences between vehicleand nimesulide cohorts.

Nimesulide tended to reduce LPS-induced increases inPGHS-2 and PGES gene expression, but this effect was notstatistically significant. Gene expression of PGDH was signif-icantly lower after maternal nimesulide infusion, consistentwith observations in neonatal rat lungs in response to indo-methacin (50), indicating that pulmonary prostaglandin catab-olism is not affected by LPS, but is dependent on PGHSactivity.

Of the four prostaglandin receptor subtypes, only mRNA forEP4 increased in response to LPS, consistent with previousstudies of EP4 expression in response to LPS (in murine

fibroblast and macrophage-like cells lines; Ref. 3) and PGE2

(in rat bone marrow in vivo; Ref. 54). This is interesting, giventhat PGE2 stimulates SP-A mRNA expression in human fetallung tissue in vitro in association with cAMP formation (1), aconsequence of EP2 and EP4 activation. Nimesulide infusionreduced mRNA levels of EP1–3 but not EP4. The effect on EP1

expression is interesting given its reported role in mediatingPGE2-induced surfactant production by rat alveolar epithelialcells in vitro (36).

Nimesulide infusion resulted in a strong trend for a reductionin SP-A and SP-B expression and a significant reduction inSP-D mRNA levels, and LPS-induced increases in expressionof these SPs were also inhibited, suggesting a critical role forinducible prostaglandins in fetal lung development. Our obser-vations are consistent with a high rate of neonatal deathdescribed for PGHS-2 knockout mice, which was attributed topatency of the ductus arteriosus (33); however, examination ofthe pulmonary surfactant system in these animals has not beenreported. The effect of nimesulide on SP-C gene expressionwas the opposite to that observed for the other surfactantproteins, suggesting independent regulation by prostaglandins.The longer term outcomes of PGHS-2 inhibition on surfactantproduction and lung compliance remain unknown but deservefurther investigation.

Fig. 10. Study 2: effect of nimesulide infusion on expression of PGE receptor subtypes in the fetal lungs. PGE receptor-1 (A), 2 (B), and 3 (C) mRNA levelsin the lung were not altered by intra-amniotic LPS but were significantly reduced by maternal nimesulide infusion. D: PGE receptor-4 mRNA levels were notsignificantly altered by LPS administration or by the nimesulide infusion. Data are expressed as means � SE. Data are expressed relative to the mean value forthe vehicle saline group. *P � 0.05; ** P � 0.001.



at Monash U



ownloaded from


Nimesulide’s preference for PGHS-2 reduces disruption tohomeostatic functions and thereby reduces, but does not abol-ish, adverse effects. Indeed, there is a possibility for nimesulideto be acting via a mechanism independent to prostaglandinssince PGHS-2 gene expression in saline groups is the samebetween cohorts, but mRNA levels of IL-1�, IL-8, SP-C,SP-D, and EP1–3 are significantly altered after saline in thenimesulide cohort. While most research has focused on theprostaglandin pathway, nimesulide can also induce glucocor-ticoid receptor phosphorylation (51), suggesting that some ofthe anti-inflammatory effects of nimesulide may be due toglucocorticoid signaling.

We did not measure mRNA levels of prostaglandin syn-thetic and metabolizing enzymes in the chorioamnion, as wewere unable to recover sufficient RNA from these samples.

However, it has been shown that expression of both PGHS-2(18, 27) and PGES (45) increases in the fetal membranes inresponse to inflammation. The fetal membranes produceprostaglandins and the human amnion specifically producessignificantly more PGE2 compared with the chorion, de-cidua, and placenta (35). This indicates that the fetal mem-branes and, in particular, the amnion, are the likely source ofthe LPS-induced increase we observed in PGE2 concentra-tions in the amniotic fluid. In addition to potential localproduction of prostaglandins by the fetal lungs (indicated byour studies), it is possible that PGE2 generated by theamnion reaches the fetal respiratory system to influencelung development.

The relative volume density of secondary septal crestswas greater in preterm lambs 7 days after exposure to LPS,which may potentially represent a rebound in alveolariza-tion. Initially, inflammation decreases the number of distalair spaces in preterm lambs, but there is recovery of alveolarnumber by later in gestation (25), which persists postnatally(37, 38). It is also possible that the increase in relative septalcrest density after LPS may be due to the overall reductionin tissue-to-air space ratio, rather than an increase in abso-lute number of secondary septal crests. This would beconsistent with trends observed previously by us (37, 44),whereby alveolar wall thickness was reduced by LPS. Lungstructural effects of intra-amniotic LPS take longer than 2days to manifest, so we did not assess lung structure afternimesulide infusion.

In summary, we found that PGE2 concentrations increasein the amniotic fluid between 2 and 6 days after intra-amniotic LPS administration in sheep. Increases in fetallung PGHS-2, PGES, and EP4 gene expression were ob-served in association with the changes in lung developmentthat accompanied the inflammatory response. Maternalnimesulide infusion inhibited increases in PGE2 concentra-tions in response to LPS, modulated the inflammatory re-sponse in the fetal lungs and altered surfactant proteinmRNA levels. These data suggest that prostaglandins me-diate inflammation-induced changes in lung development.However, the precise degree to which prostaglandins con-tribute to inflammation-induced fetal lung maturation re-mains unknown, as does the underlying mechanism, andwarrants further investigation.

ACKNOWLEDGMENTS

We are grateful to Jan Loose, Valerie Zahra, Karyn Rodgers, and AlisonMoxham for assistance.

GRANTS

This work was funded by the National Health and Medical ResearchCouncil of Australia and the Victorian Government’s Operational Infrastruc-ture Support Program.

DISCLOSURES

No conflicts of interest, financial or otherwise are declared by theauthor(s).

AUTHOR CONTRIBUTIONS

Author contributions: A.J.W. and T.J.M. performed experiments; A.J.W.and T.J.M. analyzed data; A.J.W., S.B.H., M.J.W., and T.J.M. interpretedresults of experiments; A.J.W. prepared figures; A.J.W. drafted manuscript;A.J.W., S.B.H., M.J.W., and T.J.M. edited and revised manuscript; A.J.W.,

Fig. 11. Study 2: effect of nimesulide infusion on PGE2 concentrations. A: inthe amniotic fluid, PGE2 concentrations were increased at 24 and 48 h afterLPS in the vehicle cohort but not in the nimesulide group that received LPS.B: PGE2 concentrations were elevated at 4, 6, and 12 h in fetal plasma in thevehicle LPS group compared with the nimesulide LPS group. Data areexpressed as means � SE. *P � 0.05 vs. vehicle saline group; P � 0.05between LPS groups; P � 0.001 between LPS groups.



at Monash U



ownloaded from


S.B.H., M.J.W., and T.J.M. approved final version of manuscript; S.B.H.,M.J.W., and T.J.M. conception and design of research.

REFERENCES

1. Acarregui M, Snyder J, Mitchell M, Mendelson C. Prostaglandinsregulate surfactant protein A (SP-A) gene expression in human fetal lungin vitro. Endocrinology 127: 1105–1113, 1990.

2. Andrews W, Goldenberg R, Faye-Petersen O, Cliver S, Goepfert A,Hauth J. The Alabama Preterm Birth study: polymorphonuclear andmononuclear cell placental infiltrations, other markers of inflammation,and outcomes in 23- to 32-wk preterm newborn infants. Am J ObstetGynecol 195: 803–808, 2006.

3. Arakawa T, Laneuville O, Miller C, Lakkides K, Wingerd B, DeWittD, Smith W. Prostanoid receptors of murine NIH 3T3 and RAW 264.7cells. J Biol Chem 271: 29569–29575, 1996.

4. Bachurski C, Ross G, Ikegami M, Kramer B, Jobe A. Intra-amnioticendotoxin increases pulmonary surfactant components and induces SP-Bprocessing in fetal sheep. Am J Physiol Lung Cell Mol Physiol 280:L279–L285, 2001.

5. Barnett J, Chow J, Ives D, Chiou M, Mackenzie R, Osen E, Nguyen B,Tsing S, Bach C, Freire J, Chan H, Sigal E, Ramesha C. Purification,characterization and selective inhibition of human prostaglandin G/Hsynthase 1 and 2 expressed in the baculovirus system. Biochim BiophysActa 1209: 130–139, 1994.

6. Bennett A, Villa G. Nimesulide: an NSAID that preferentially inhibitsCOX-2, and has various unique pharmacological activities. Expert OpinPharmacol 1: 277–286, 2000.

7. Burgess K, Ralph M, Jenkin G, Thorburn G. Effect of oxytocin andestradiol on uterine prostaglandin release in nonpregnant and early-pregnant ewes. Biol Reprod 42: 822–833, 1990.

8. Bustos R, Ballejo G, Giussi G, Rosas R, Isa J. Inhibition of fetal lungmaturation by indomethacin in pregnant rabbits. J Perinat Med 6: 240–245, 1978.

9. Carey M, Germolec D, Bradbury J, Gooch R, Moorman M, Flake G,Langenbach R, Zeldin D. Accentuated T helper type 2 airway responseafter allergen challenge in cyclooxygenase-1�/� but not cyclooxygenase-2�/� mice. Am J Respir Crit Care Med 167: 1509–1515, 2003.

10. Conner C, Kelly R, Hume R. Regulation of prostaglandin availability inhuman fetal lung by differential localisation of prostaglandin H synthase-1and prostaglandin dehydrogenase. Histochem Cell Biol 116: 313–319,2001.

11. Eisengart C, Mestre J, Naama H, Mackrell P, Rivadeneira D, MurphyE, Stapleton P, Daly J. Prostaglandins regulate melanoma-induced cyto-kine production in macrophages. Cell Immunol 204: 143–149, 2000.

12. Filby C, Hooper S, Wallace M. Partial pulmonary embolization disruptsalveolarization in fetal sheep. Respir Res 11: 42, 2010.

13. Fowden A, Harding R, Ralph M. The nutritional regulation of plasmaprostaglandin E concentrations in the fetus and pregnant ewe late ingestation. J Physiol 394: 1–12, 1987.

14. Fukuda Y, Ferrans V, Crystal R. The development of alveolar septa infetal sheep lung. An ultrastructural and immunohistochemical study. Am JAnat 167: 405–439, 1983.

15. Gavett S, Madison S, Chulada P, Scarborough P, Qu W, Boyle J,Tiano H, Lee C, Langenbach R, Roggli V, Zeldin D. Allergic lungresponses are increased in prostaglandin H synthase-deficient mice. J ClinInvest 104: 721–732, 1999.

16. Goldenberg R, Hauth J, Andrews W. Intrauterine infection and pretermdelivery. N Engl J Med 342: 1500–1507, 2000.

17. Grigsby P, Poore K, Hirst J, Jenkin G. Inhibition of premature labor insheep by a combined treatment of nimesulide, a prostaglandin synthasetype 2 inhibitor, and atosiban, an oxytocin receptor antagonist. Am JObstet Gynecol 183: 649–657, 2000.

18. Hansen W, Sato T, Mitchell M. Tumor necrosis factor-alpha stimulatesincreased expression of prostaglandin endoperoxide H synthase type 2mRNA in amnion-derived WISH cells. J Mol Endocrinol 20: 221–231,1998.

19. Hume R, Kelly R, Cossar D, Giles M, Hallas A, Gourlay M, Bell J.Self-differentiation of human fetal lung in organ culture: the role ofprostaglandins PGE2 and PGF2�. Exp Cell Res 194: 111–117, 1991.

20. Jobe A, Ikegami M. Fetal responses to glucocorticoids. In: Endocrinol-ogy of the Lung. Totowa, NJ: Humana, 2000, p. 45–57.

21. Jobe A, Newnham J, Willet K, Moss T, Ervin M, Padbury J, Sly P,Ikegami M. Endotoxin induced lung maturation in preterm lambs is notmediated by cortisol. Am J Respir Crit Care Med 162: 1656–1661, 2000.

22. Johnson J, Wimsatt J, Buckel S, Dyer R, Maddipati K. Purification andcharacterization of prostaglandin H synthase-2 from sheep placental cot-yledons. Arch Biochem Biophys 324: 26–34, 1995.

23. Kallapur S, Jobe A, Ball M, Nitsos I, Moss T, Hillman N, NewnhamJ, Kramer B. Pulmonary and systemic endotoxin tolerance in pretermfetal sheep exposed to chorioamnionitis. J Immunol 179: 8491–8499,2007.

24. Kallapur S, Moss T, Auten RJ, Nitsos I, Pillow J, Kramer B, MaedaD, Newnham J, Ikegami M, Jobe A. IL-8 signaling does not mediateintra-amniotic LPS-induced inflammation and maturation in preterm fetallamb lung. Am J Physiol Lung Cell Mol Physiol 297: L512–L519, 2009.

25. Kallapur S, Nitsos I, Moss T, Kramer B, Newnham J, Ikegami M,Jobe A. Chronic endotoxin exposure does not cause sustained structuralabnormalities in the fetal sheep lungs. Am J Physiol Lung Cell Mol Physiol288: L966–L974, 2005.

26. Kallapur S, Willet K, Jobe A, Ikegami M, Bachurski C. Intra-amnioticendotoxin: chorioamnionitis precedes lung maturation in preterm lambs.Am J Physiol Lung Cell Mol Physiol 280: L527–L536, 2001.

27. Kim G, Romero R, Kuivaniemi H, Tromp G, Haddad R, Kim Y, KimM, Nien J, Hong J, Espinoza J, Santolaya J, Yoon B, Mazor M, KimC. Expression of bone morphogenetic protein 2 in normal spontaneouslabor at term, preterm labor and preterm PROM. Am J Obstet Gynecol193: 1137–1143, 2005.

28. Kitterman J, Liggins G, Clements J, Campos G, Lee C, Ballard P.Inhibitors of prostaglandin synthesis, tracheal fluid, and surfactant in fetallambs. J Appl Physiol 51: 1562–1567, 1981.

29. Kramer BW, Moss TJ, Willet KE, Newnham JP, Sly PD, KallapurSG, Ikegami M, Jobe AH. Dose and time response after intra-amnioticendotoxin in preterm lambs. Am J Respir Crit Care Med 164: 982–988,2001.

30. Laatikainen T, Raisanen I, Salminen K. Corticotrophin-releasing hor-mone in amniotic fluid during gestation and labor and in relation to fetallung maturation. Am J Obstet Gynecol 142: 664–669, 1988.

31. Lahra M, Beeby P, Jeffery H. Maternal versus fetal inflammation andrespiratory distress syndrome: a 10-year hospital cohort study. Arch DisChild Fetal Neonatal Ed 94: F13–F16, 2009.

32. Lee S, Cheung C, Nicholls J, Hui K, Leung C, Uiprasertkul M, TipoeG, Lau Y, Poon L, Ip N, Guan Y, Peiris J. Hyperinduction of cycloox-ygenase-2-mediated proinflammatory cascade: a mechanism for the patho-genesis of avian influenza H5N1 infection. J Infect Dis 198: 525–535,2008.

33. Loftin C, Trivedi D, Tiano H, Clark J, Lee C, Epstein J, Morham S,Breyer M, Nguyen M, Hawkins B, Goulet J, Smithies O, Koller B,Langenbach R. Failure of ductus arteriosus closure and remodeling inneonatal mice deficient in cyclooxygenase-1 and cyclooxygenase-2. ProcNatl Acad Sci USA 98: 1059–1064, 2001.

34. Marino P, Rooney S. Surfactant secretion in a newborn rabbit lung slicemodel. Biochim Biophys Acta 620: 509–519, 1980.

35. Mitchel M, Bibby J, Hicks B, Turnbull A. Specific production ofprostaglandin E by human amnion in vitro. Prostaglandins 15: 377–382,1978.

36. Morsy M, Isohama Y, Miyata T. Prostaglandin E2 increases surfactantsecretion via the EP1 receptor in rat alveolar type II cells. Eur J Phar-macol 426: 21–24, 2001.

37. Moss T, Davey M, Harding R, Newnham J. Effects of intra-amnioticendotoxin on lung structure and function two months after term birth insheep. J Soc Gynecol Investig 9: 220–225, 2002.

38. Moss T, Newnham J, Willett K, Kramer B, Jobe A, Ikegami M. Earlygestational intra-amniotic endotoxin: lung function, surfactant, and mor-phometry. Am J Respir Crit Care Med 2002; 165: 805-11.

39. Moss T, Nitsos I, Kramer B. Fetal Pulmonary COX-2 Increases in anOvine Model of Chorioamnionitis. Houston, TX: Society for GynecologicInvestigation, 2004, p. 112A.

40. Moss T, Nitsos I, Kramer B, Ikegami M, Newnham J, Jobe A.Intra-amniotic endotoxin induces lung maturation by direct effects on thedeveloping respiratory tract in preterm sheep. Am J Obstet Gynecol 187:1059–1065, 2002.

41. Murakami M, Nakatani Y, Tanioka T, Kudo I. Prostaglandin E syn-thase. Prostaglandins Other Lipid Mediat 68–69: 383–399, 2002.

42. Nitsos I, Moss T, Cock M, Harding R, Newnham J. Fetal responses tointra-amniotic endotoxin in sheep. J Soc Gynecol Investig 9: 80–85, 2002.

43. Norton M, Merrill J, Cooper B, Kuller J, Clyman R. Neonatal com-plications after the administration of indomethacin for preterm labor. NEngl J Med 329: 1602–1607, 1993.



at Monash U



ownloaded from


44. Pillow J, Jobe A, Collins R, Hantos Z, Ikegami M, Moss T, NewnhamJ, Willet K, Sly P. Variability in preterm lamb lung mechanics afterintra-amniotic endotoxin is associated with changes in surfactant pool sizeand morphometry. Am J Physiol Lung Cell Mol Physiol 287: L992–L998,2004.

45. Premyslova M, Li W, Alfaidy N, Bocking A, Campbell K, Gibb W,Challis J. Differential expression and regulation of microsomal prosta-glandin E2 synthase in human fetal membranes and placenta with infectionand in cultured trophoblast cells. J Clin Endocrinol Metab 88: 6040–6047,2003.

46. Schmitz T, Cox L, Li C, Levine B, Ford S, McDonald T, NathanielszP. Prostaglandin E2 receptor expression in fetal baboon lung at 0.7gestation after betamethasone exposure. Pediatr Res 61: 421–426, 2007.

47. Singh V, Grace M, Parekh V, Whitnall M, Landauer M. Effects ofgenistein administration on cytokine induction in whole-body gammairradiated mice. Int Immunopharmacol 9: 1401–1410, 2009.

48. Smith W, Lands W. Oxygenation of polyunsaturated fatty acids duringprostaglandin biosynthesis by sheep vesicular glands. Biochemistry 11:3276–85, 1972.

49. Sozo F, Hooper S, Wallace M. Thrombospondin-1 expression andlocalization in the developing ovine lung. J Physiol 584: 625–635,2007.

50. Takizawa T, Ikeda Y, Togashi H, Yamamoto M, Arishima K, AkahoriF, Masaoka T. Inhibitory effect of indomethacin on neonatal lungcatabolism of prostaglandin E2: possible mechanism of the re-opening ofthe ductus arteriosis after indomethacin therapy. J Toxicol Sci 21: 243–248, 1996.

51. Tavares I, Bishai P, Bennett A. Activity of nimesulide on constitutiveand inducible cyclooxygenases. Arzneimittelforschung 45: 1093–1095,1995.

52. Thomas M. The leukocyte common antigen family. Annu Rev Immunol 7:339–369, 1989.

53. Vane J, Bakhle Y, Botting R. Cyclooxygenases 1 and 2. Annu RevPharmacol Toxicol 38: 97–120, 1998.

54. Weinreb M, Grosskopf A, Shir N. The anabolic effect of PGE2 in ratbone marrow cultures is mediated via the EP4 receptor subtype. Am JPhysiol Endocrinol Metab 276: E376–E383, 1999.

55. Wu G, Wei C, Kulmacz R, Osawa Y, Tsai A. A mechanistic study ofself-inactivation of the peroxidase activity in prostaglandin H synthase-1.J Biol Chem 274: 9231–9237, 1999.

56. Zeldin D, Wohlford-Lenane C, Chulada P, Alyce Bradbury J, Scarbor-ough P, Roggli V, Langenbach R, Schwartz D. Airway inflammation andresponsiveness in prostaglandin H synthase-deficient mice exposed to bacte-rial lipopolysaccharide. Am J Respir Cell Mol Biol 25: 457–465, 2001.



at Monash U



ownloaded from


prostaglandins mediate the fetal pulmonary response to ... · prostaglandins mediate the fetal...

Documents