http://rsx.sagepub.com/Reproductive Sciences

http://rsx.sagepub.com/content/early/2013/11/11/1933719113508820The online version of this article can be found at:

 DOI: 10.1177/1933719113508820

published online 12 November 2013Reproductive SciencesKannan, John P. Newnham, Boris W. Kramer, Alan H. Jobe, Jeffrey A. Keelan and Matthew W. Kemp

Masatoshi Saito, Matthew S. Payne, Yuichiro Miura, Demelza J. Ireland, Sarah Stock, Suhas G. Kallapur, Paranthaman S.Model

Polymyxin B Agonist Capture Therapy for Intrauterine Inflammation: Proof-of-Principle in a Fetal Ovine  

Published by:

http://www.sagepublications.com

On behalf of: 

  Society for Gynecologic Investigation

can be found at:Reproductive SciencesAdditional services and information for    

  http://rsx.sagepub.com/cgi/alertsEmail Alerts:

 

http://rsx.sagepub.com/subscriptionsSubscriptions:  

http://www.sagepub.com/journalsReprints.navReprints:  

http://www.sagepub.com/journalsPermissions.navPermissions:  

What is This? 

- Nov 12, 2013OnlineFirst Version of Record >>

at OhioLink on November 13, 2013rsx.sagepub.comDownloaded from at OhioLink on November 13, 2013rsx.sagepub.comDownloaded from

Original Article

Polymyxin B Agonist Capture Therapyfor Intrauterine Inflammation: Proof-of-Principle in a Fetal Ovine Model

Masatoshi Saito, PhD1,2, Matthew S. Payne, PhD1,Yuichiro Miura, PhD1, Demelza J. Ireland, PhD1, Sarah Stock, PhD1,Suhas G. Kallapur, MD1,3, Paranthaman S. Kannan, PhD3,John P. Newnham, MD1, Boris W. Kramer, PhD1,4, Alan H. Jobe, PhD1,3,Jeffrey A. Keelan, PhD1, and Matthew W. Kemp, PhD1

AbstractIntrauterine infection is a leading cause of preterm birth (PTB), most notably in deliveries occurring before 32 weeks gestation.Preterm infants exposed to intrauterine inflammation are more likely to have a host of neurological, respiratory, gastrointestinal,and visual pathologies. Preventing preterm delivery and protecting the fetus from injury is thus likely to require treatment of bothintrauterine infection and inflammation. Polymyxin B (PMXB) is a cationic peptide antibiotic that binds Escherichia coli lipopoly-saccharides (LPS) and prevents inflammatory activation. We hypothesized that intraamniotic administration of PMXB wouldselectively inhibit LPS-driven inflammation, serving as a proof-of-principle for targeted agonist capture therapy as a treatment forPTB and fetal injury. In vitro studies with primary fetal ovine keratinocytes demonstrated a significant and sustained reduction intumor necrosis factor a and interleukin 8 messenger RNA expression after treatment with PMXB and LPS, relative to cellstreated with LPS alone. In vivo studies with fetal sheep demonstrated a significant reduction in proinflammatory cytokines in theamniotic fluid and fetal lung (but not fetal skin or chorioamnion) in LPS þ PMXB-treated animals, relative to those treated withLPS alone. These data are consistent with a partial resolution of LPS-driven intrauterine inflammation. They suggest the potentialfor agonist capture as a conceptual means of resolving the proparturition inflammation caused by infection of the amniotic cavity.

Keywordspreterm birth, inflammation, sheep, infection, fetus, agonist

Introduction

Preterm birth (PTB) remains a primary cause of neonatal death

and disease.1 The incidence of PTB varies markedly depending

on a host of interacting factors including geographical location,

ethnicity, and socioeconomic status. Of all births worldwide,

14.9 million or 11.1% were estimated to be born preterm in

2010; 2.6% of births in urban China, 7.8% in Hong Kong,

7% in the United Kingdom, 12% in the United States, and

15.5% in Indonesia are classified as preterm.1-4 Globally, PTB

accounts for some 28% of the estimated 4 million neonatal

deaths that occur each year5; in Australia, PTB is responsible

for an estimated 70% of total neonatal deaths.6 Despite marked

improvements in global rates of <5 mortality, to date little

progress has been made in reducing the rates of neonatal

mortality.7

The need for therapies to prevent both preterm delivery and fetal

injury is underscored by the profound societal cost of PTB. Preterm

infants are at an increased risk (in proportion, generally speaking,

to the degree of prematurity) of acute and chronic diseases of the

respiratory (pneumonia, bronchopulmonary dysplasia and chronic

lung disease), ocular (retinopathy of prematurity), neurological

(intra-ventricular hemorrhage, cerebral palsy, executive function/

learning deficit), and gastrointestinal (necrotizing enterocolitis)

systems.7-15 Interestingly, several studies now exist to suggest that

those born late preterm (36-38 weeks gestation) are also at an ele-

vated risk of sudden infant death syndrome and exhibit lower

1 School of Women’s and Infants’ Health, The University of Western Australia,

Perth, Western Australia2 Division of Perinatal Medicine, Tohoku University Hospital, Sendai, Japan3 Division of Pulmonary Biology, Cincinnati Children’s Hospital Medical

Centre, University of Cincinnati School of Medicine, Cincinnati, OH, USA4 Department of Paediatrics, School of Oncology and Developmental Biology,

Maastricht University Medical Centre, Maastricht, the Netherlands

Corresponding Author:

Matthew W. Kemp, School of Women’s and Infants’ Health, The University of

Western Australia, Perth, Western Australia 6009, Australia.

Email: matthew.kemp@uwa.edu.au

Reproductive Sciences00(0) 1-9ª The Author(s) 2013Reprints and permission:sagepub.com/journalsPermissions.navDOI: 10.1177/1933719113508820rs.sagepub.com

at OhioLink on November 13, 2013rsx.sagepub.comDownloaded from

intelligence quotient scores than contemporaries delivered

between 39 and 41 weeks, potentially due to exposure to a sub-

optimal uterine environment.16,17 High on-going rates of PTB

also convey a significant financial cost. The annual societal

economic burden of prematurity in the United States is esti-

mated at US$26.2 billion. A single very early (delivery before

32 weeks) preterm infant born in England or Wales that sur-

vives to adulthood is estimated to cost the public £61,781 more

than a contemporary, term-born survivor.4,18

Uterine infection and inflammation is involved in at least

40% of all cases of prematurity and is the causal pathway that

is best understood from a clinical, molecular, and patho-

physiological perspective (for review see7,8,11,19). Microbial

pathogen-associated molecular patterns (including peptidogly-

can, lipopolysaccharides [LPS], and lipoteichoic acids) are

recognized by innate immune system receptors that trigger a

series of signaling cascades leading to the local production of

proparturition mediators including, proinflammatory cytokines/

chemokines (interleukins [IL] 1b, IL-6, IL-8, tumor necrosis fac-

tor [TNF] a), prostaglandins, and matrix metalloproteinases by

gestational tissues.7,8,11 Within the fetus, systemic inflammation,

or the fetal inflammatory response syndrome, is characterized by

cord blood IL-6 in excess of 11 pg/mL and is indicative of uter-

ine infection and adverse neonatal outcomes.20

An appreciation for the role of inflammation in both preterm

labor and fetal injury has led to increased efforts to develop

PTB prophylaxis that resolve both intrauterine infection and

resultant intrauterine inflammation. The majority of anti-

inflammatory approaches tested to date involve the disruption

of inflammatory signaling (eg, inhibitors of nuclear factor

kB(NF-kB) signaling, and IL-1 antagonists) or administration

of anti-inflammatory agents such as IL-10.21,22 Although a

number of these approaches show some promise, their clinical

application may be hampered by concerns relating to their

nonspecific modulation of fetal inflammatory signaling. In

addition to playing a key role in normal pregnancy, appropri-

ately regulated inflammation (controlled by ubiquitous media-

tors such as NF-kB) also regulate normal fetal immunological

development and organogenesis.23-25

The microorganisms involved in PTB, and the ligand

receptor interactions responsible for the induction of pathologi-

cal inflammation are increasingly well understood.8,19,26,27 We

hypothesized that inhibition of inflammatory signaling by

binding of key PAMP ligands by an inert agent may serve as

a selective, effective, and nontoxic means of resolving intrau-

terine inflammation.

Polymyxin B (PMXB) is a cyclic, cationic peptide antibiotic

that prevents LPS-driven inflammatory activation by binding

the lipid A core of LPS.28,29 A number of earlier studies have

demonstrated the utility of PMXB in resolving LPS-driven

inflammation. Relative to LPS stimulation alone, Stokes and

colleagues demonstrated that PMXB blocked TNF-a produc-

tion by rat alveolar macrophages following costimulation with

LPS in both in vitro experiments and in an in vivo rat model.29

More recently, PMXB, or synthetic derivatives, was used as

LPS binding agents to prevent inflammatory activation in a

number of settings including hemoperfusion and as a treatment

for endotoxemia in horses.30-33 As a proof of principle study,

we hypothesized that administration of PMXB would inhibit

LPS-driven intrauterine inflammation in a well-characterized

ovine model of human pregnancy.

Materials and Methods

Animals

All procedures with animals were performed at The University

of Western Australia (Perth, Western Australia) after approval

by The University of Western Australia Animal Ethics Com-

mittee. Twenty four date-mated Australian merino ewes with

singleton pregnancies were randomized to receive (1) a single

ultrasound-guided intraamniotic injection of 10 mg PMXB

(81271; Sigma Aldrich, St Louis, Missouri; PMXB group;

n¼ 8) or (2) a single ultrasound-guided intraamniotic injection

of 10 mg LPSs from Escherichia coli O55: B5 (LPS; L2880,

Sigma Aldrich; LPS group; n ¼ 8); or (3) a single ultrasound-

guided intraamniotic injection of 10 mg PMXB immediately

followed by 10 mg LPS (PMXBþ LPS group; n¼ 8). Success-

ful placement of intraamniotic injections were confirmed with

electrolyte (Cl�) analysis of amniotic fluid sample using a Sie-

mens Rapidlab 1265 Analyzer (Siemens, Munich, Germany).

Injections were performed at 122 day + 1 day gestational age

(GA). Fetuses were surgically delivered after 2 days and both

ewe and fetus were euthanized with an intravenous injection

of pentobarbitone (100 mg/kg). Fetal tissues for protein and

messenger RNA (mRNA) expression analyses were snap

frozen in liquid nitrogen. No fetal losses occurred during the

experiment. Fetal lung tissues for histological analysis were

inflation fixed in 10% neutral-buffered formalin for 24 hours

before paraffin embedding.

Fetal Epidermal Keratinocyte Isolation and Culture

Primary ovine epidermal keratinocytes were collected from the

skin of 124 d GA fetuses (no in utero exposure to either LPS

or PMXB) following surgical delivery and euthanasia, as previ-

ously described.26,34 Briefly, full-thickness fetal skin was cut

into 3-mm2 pieces and washed 3 times in 70% ethanol and 3

times in sterile phosphate-buffered saline (PBS) containing 5

mg/mL streptomycin and 3.3mg/mL penicillin. Tissue pieces

were immersed in a 3% solution of iodine for 5 minutes then

incubated for 48 hours in sterile PBS containing 1U/mL

Dispase II (17105-014; Life Technologies, Carlsbad, Califor-

nia). Epidermal sheets were removed from the dermis using ster-

ile forceps and placed into 5 mL 0.1% trypsin in EDTA.

Epidermal sheets were disrupted by incubating cells for 20 min-

utes in a shaking water bath at 37�C. Trypsinization was halted

with the addition of saline containing 5% fetal calf serum. Cells

were pelleted by centrifugation (1000 rpm, 10 minutes, 24�C)

and washed 3 times in sterile PBS. Isolated keratinocytes were

cultured in 25 cm2 flasks and cultured at 37�C in a 5% CO2

atmosphere in Keratinocyte-SFM media (17005-042; Life

2 Reproductive Sciences 00(0)

at OhioLink on November 13, 2013rsx.sagepub.comDownloaded from

Technologies) containing 1 � Penicillin-Streptomycin (15140-

122; Life Technologies). All cells used in this study were

passage 3.

In Vitro PMXB Dose–Response Study

To confirm dose–response studies performed previously,28,29

keratinocytes were subcultured into 12-well culture plates at

a density of 2 � 105 cells/well, 16 hours prior to commencing

the study. The >95% viability of cells added to each well was

confirmed with trypan blue staining. Keratinocytes were

exposed to 2 mL keratinocyte-SFM media containing (1) 1

mg LPS; (2) 1 mg LPS and 1 mg PMXB (1:1 w/w); (3) 1 mg LPS

and 10 mg PMXB (1:10, w/w); or (4) 1 mg LPS and 100 mg

PMXB (1:100 w/w). Each exposure was performed once in

triplicate for 0, 1, 2, 3, 6, and 12 hours.

Assessment of In Vivo LPS Concentration

LPS concentrations in amniotic fluid were determined using a

Pierce LAL Chromogenic Endotoxin Quantitation Kit (88282;

Thermo Scientific, Rockford, Illinois) incorporating a 3-point

standard curve.

Hematology

Complete blood counts and differential analyses were

performed by VetPath Laboratory Services (Perth, Western

Australia) using an automated Coulter counter adapted for

ovine specimens.

Fetal Lung Histology

Tissues from 5, randomly selected animals from each group

were analyzed for histological changes. Sections of 5 mm thick-

ness from formalin-fixed lung (right upper lobe) tissues

embedded in paraffin blocks were stained with hematoxylin

and eosin. Immunofluorescent staining of fetal lung for CD3

(A0452, Dako, Glostrup, Denmark, working concentration

1:100) was performed as previously published.35 CD3 cells

in fetal lung were quantified by counting positively stained

cells in 5 randomly selected, nonoverlapping fields at 20�objective magnification.

Isolation and Purification of RNA

Total RNA was extracted from keratinocyte monolayers or

fetal tissue homogenized in liquid nitrogen using TRIzol

(15596-018; Life Technologies). Extracted RNA was treated

with Turbo DNase (AM2238; Life Technologies) to remove

any residual DNA, following manufacturer’s instructions.

RNA yields from fetal tissues were normalized to 100 ng/mL

using nuclease-free water (AM9937; Life Technologies).

Relative Quantification of mRNA Expression

Polymerase chain reaction (PCR) primers specific for ovine IL-8

and TNF-a were used to perform quantitative PCR (qPCR) reac-

tions using Power SYBR Green PCR Master Mix (4367659;

Life Technologies) on complementary DNA generated from

600 ng DNase-treated RNA isolated from cultured epidermal

keratinocytes. Power SYBR Green qPCR cycling conditions

were: 1 � 5 minute initial denaturation (95�C), 35 � 30-

second denaturation (95�C), 30-second annealing (60�C), 30-

second extension (60�, 1 � final 0.5�C stepped temperature

ramp 60�C-95�C). Reactions were performed on a Qiagen

(Venlo, the Netherlands) Rotorgene in a final volume of 20

mL as described previously.26,34 The Cq values were normalized

to ovine-specific glyceraldehyde 3-phosphate dehydrogenase to

calculate difference in Cq (dCq) and fold-change values. Reac-

tion efficiencies and specificities were within limits proposed in

the Minimum Information for Publication of Quantitative Real-

Time PCR Experiments (MIQE) guidelines.36

The PCR primers and hydrolysis probes specific for ovine

IL-1b, IL-6, IL-8, TNF-a, MCP-2, Hepcidin (a regulator of iron

homeostasis that is increased in response to inflammation),

serum amalyoid A3 (SAA-3) protein (a marker of inflamma-

tion), and C-reactive protein (CRP, Custom Targets, Life Tech-

nologies) sequences were used to perform quantitative PCRs

on DNase-treated RNA isolated from fetal tissue. Reactions

were performed using an EXPRESS One-Step SuperScript

qRT-PCR Kit (1178101K; Life Technologies) with 400 ng

total RNA template in a total volume of 20 mL in accordance

with the manufacturer’s instructions. Reaction cycling condi-

tions consisted of 15 minutes reverse transcription at 50�C and

an initial denaturation at 95�C for 20 seconds, followed by 40

cycles of 95�C for 3 seconds and 60�C for 30 seconds (acqui-

sition phase). All reactions were performed in fast 96-well

plates on a ViiA7 real-time PCR thermocycler (Life Technolo-

gies). The Cq values were normalized to 18s rRNA and

expressed as fold changes relative to pooled control values

+ standard deviation (SD). Reaction efficiencies were within

limits proposed in the MIQE guidelines.36 Six randomly

selected animals from each group were analyzed for changes

in cytokine/chemokine mRNA expression. The dCq values

were used for statistical analyses as described subsequently.

Enzyme-Linked Immunosorbent Assays

Six randomly selected animals from each group were analyzed

for changes in cytokine/chemokine protein expression. Quanti-

fication of IL-1b and IL-8 protein concentrations in amniotic

fluid and alveolar wash samples was performed as previously

described.37 All samples were diluted 1:5 in assay buffer and

thoroughly mixed by vortexing prior to analysis. Quantification

of IL-6 protein concentration in ovine amniotic fluid was

performed using an identical protocol with the following

modifications: Plate wells were coated overnight at 4�C with

5 mg/mL capture antibody (MCA1659; ABD Serotech,

Kidlington, United Kingdom). Recombinant sheep IL-6 protein

Saito et al 3

at OhioLink on November 13, 2013rsx.sagepub.comDownloaded from

standards (Protein Express, Cincinnati, Ohio) and amniotic

fluid samples were diluted 1:2 in assay buffer and incubated

overnight at 4�C. The detection antibody (AHP424; ABD

Serotech) was diluted 1:750 in assay buffer. All samples and

standards were assayed in duplicate.

Statistical Analyses

All values are expressed as mean + 1 SD. All statistical anal-

yses were performed using IBM SPSS Statistics for Windows,

Version 20.0 (IBM Corp, Armonk, New York). Data were

assessed for normality with Shapiro-Wilk Tests. Normally dis-

tributed (parametric) data were screened for outliers with

Dixon Q-Parameter and differences tested for significance with

1-way analysis of variance (ANOVA) employing an a value of

.05. Multiple post hoc comparisons were performed using

Tukey test. Apparent differences in nonparametric data were

tested for significance with Kruskal-Wallis 1-way ANOVA

employing an a value of .05. Multiple post hoc comparisons

were performed using Mann-Whitney tests with an a value cor-

rected for n multiple comparisons.

Results

In Vitro PMXB Dose Response for LPS

Relative to treatment with culture media containing LPS only,

expression of IL-8 and TNF-a mRNA was significantly reduced

(P < .05) in ovine keratinocytes exposed to culture media supple-

mented with LPS and PMXB (Figure 1). A 1:1 or 1:10 ratio of

LPS:PMXB resulted in the most consistent suppression of IL-8

and TNF-a mRNA expression. As no significant difference

(P > .05) in IL-8 or TNF-a expression was identified between

a 1:1 and 1:10 LPS–PMXB dose, the lower 1:1 LPS–PMXB

ratio was selected for subsequent in vivo studies.

Fetal Weight, Hematology, and Amniotic Fluid EndotoxinConcentrations

No differences (P > .05) were observed in fetal birth weight (kg),

arterial cord blood pH, pO2 (mm Hg), pCO2 (mm Hg), or

hemoglobin (g/L) between PMXB-, PMXB þ LPS-, LPS-

exposed animals, respectively (Table 1). No differences (P >

.05) were observed in fetal arterial cord total white blood cells

(1011/L), neutrophil (109/L), lymphocyte (109/L), monocyte

(109/L), or eosinophil counts (109/L) between PMXB-, PMXB

þ LPS-, and LPS-exposed animals, respectively (Table 1). No

significant difference (P ¼ .345) in amniotic fluid LPS concen-

tration was observed between PMXB þ LPS- and LPS-exposed

animals. Amniotic fluid LPS concentrations from PMXB þLPS- and LPS-exposed animals were significantly increased

(P < .000) relative to amniotic fluid from PMXB-exposed ani-

mals (Figure 2).

Fetal Lung Histology

Analysis of hematoxylin and eosin-stained lung (right upper

lobe) sections from LPS-exposed animals demonstrated

numerous focal areas of consolidation and extensive airway

inflammatory cells (Figure 3A; panel C, arrows), consistent

with alveolar inflammation. Similar consolidation/airway

involvement was absent from sections taken from PMXB- and

PMXB þ LPS-exposed animals (Figure 3A; panels A and B,

respectively). The number of CD3-positive cells was signifi-

cantly increased in lung tissue of LPS-exposed animals (15.0

Figure 1. Expression of interleukin (IL) 8 (panel A) and tumor necrosis factor (TNF) a (panel B) mRNA (fold change) in primary fetal ovinekeratinocytes exposed to LPS þ polymyxin B (PMXB), relative to LPS-exposure alone. *, P < .05. Note logarithmic scale: values <1 representreduced mRNA expression relative to LPS stimulation alone; values >1 represent increased mRNA expression relative to LPS stimulation alone.mRNA indicates messenger RNA; LPS, lipopolysaccharide

Table 1. Fetal Arterial Cord Blood Measurements at Delivery.

PMXB PMXB þ LPS LPS

Fetal birth weight, kg 2.8 + 0.4 3.0 + 0.2 2.6 + 0.4Arterial cord blood, pH 7.2 + 0.05 7.2 + 0.05 7.2 + 0.1pO2, mm Hg 83.0 + 9 77 + 6 83.0 + 11pCO2, mm Hg 4.60 + 1.3 5.2 + 2.4 5.3 + 3.7Hemoglobin, g/L 130.0 + 15.5 122.0 + 9.4 122.0 + 15.0Total white blood cells,

1011/L3.7 + 1.9 3.5 + 1.0 3.1 + 1.7

Neutrophils, 109/L 0.4 + 0.2 0.4 + 0.3 0.3 + 0.3Lymphocytes, 109/L 2.6 + 1.6 2.3 + 0.7 2.1 + 1.0Monocytes, 109/L 0.2 + 0.1 0.2 + 0.2 0.2 + 0.1Eosinophils, 109/L 0.06 + 0.05 0.04 + 0.04 0.02 + 0.04

Abbreviations: LPS, lipopolysaccharide; PMXB, polymyxin B.

4 Reproductive Sciences 00(0)

at OhioLink on November 13, 2013rsx.sagepub.comDownloaded from

+ 14.0). Relative to LPS-exposed fetuses, significantly fewer

CD3-positive cells were present in the lung of PMXB- (3 +1.3; P ¼ .004) and PMXB þ LPS- (3.7 + 2.6; P ¼ .009)

exposed fetuses. No difference (P¼ 1.0) was detected between

CD3-positive cells in PMXB- and PMXB þ LPS-exposed ani-

mals. Representative images are shown in Figure 3B.

Cytokine and Acute-Phase Protein mRNA Expression

Expression of IL-1b, IL-6, IL-8, and MCP-2 mRNA was signif-

icantly increased (P < .010) in chorioamnion tissue from LPS-

and LPS þ PMXB-exposed animals, relative to chorioamnion

tissue from PMXB-control animals. No significant difference

in TNF-a mRNA expression (P¼ .270) was identified between

treatment groups. No significant difference in expression of IL-

1b, TNF-a, IL-6, IL-8, or MCP-2 mRNA was identified

between LPS- and LPS þ PMXB-exposed animals (data not

shown). Expression of IL-1b, TNF-a, IL-6, IL-8, and MCP-2

mRNA was significantly increased (P < .015) in lung tissue

(right lower lobe) from LPS-exposed animals, relative to both

PMXB- and PMXB þ LPS-exposed animals. No significant

difference (P > .05) in IL-1b or IL-6 mRNA expression was

observed in PMXB þ LPS-exposed lung, relative to lung from

fetuses exposed to PMXB (Figure 4; panel A). No significant

difference (P > .05) in IL-1b, TNF-a, or IL-6 mRNA expres-

sion was identified in PMXB þ LPS- or LPS-exposed skin,

relative to skin from fetuses exposed to PMXB. The IL-8 and

MCP-2 mRNA expression was equivalently increased (P <

.05) in fetal skin exposed to both PMXB þ LPS and LPS,

relative to skin from fetuses exposed to PMXB (Figure 4; panel

B). No significant differences in assessed cytokine/chemokine

mRNA expression were identified in the fetal spleen PMXB-,

PMXB þ LPS-, or LPS-exposed fetuses (Figure 4; panel C).

Hepcidin and SAA-3 mRNA expression were significantly

(P < .02) increased in liver from fetuses exposed to LPS,

relative to fetuses exposed to PMXB. No differences in liver

hepcidin or SAA-3 mRNA were detected between PMXB- and

PMXB þ LPS-exposed animals (Figure 5). An apparent

increase in both liver CRP mRNA in LPS-exposed fetuses,

relative to PMXB-exposed fetuses, and apparent reduction in

liver CRP mRNA in PMXB þ LPS-exposed fetuses (relative

to LPS-exposed fetuses) approached statistical significance

(P ¼ .02 and .08 [Pcrit ¼ .02]).

Amniotic Fluid and Alveolar Wash CytokineConcentrations

The concentration of IL-6 in amniotic fluid was significantly

increased in LPS-exposed animals, relative to animals exposed

to either PMXB (P ¼ .007) or PMXB þ LPS (P ¼ .018). The

concentration of amniotic fluid IL-8 was significantly

increased in LPS-exposed animals, relative to animals exposed

to either PMXB (P¼ .000) or PMXBþ LPS (P¼ .014), and in

PMXB þ LPS-exposed animals, relative to PMXB-exposed

animals (P ¼ .011; Figure 6). The concentration of IL-8 in

alveolar wash samples from LPS-exposed fetuses (140 + 10

ng/mL) was significantly increased relative to PMXB (52 +54 pg/mL; P ¼ .001) and PMXB þ LPS (0.19 + 0.13 ng/

mL; P¼ .002). In contrast, there was no statistically significant

difference in the concentration of IL-8 in the alveolar wash

between animals exposed to PMXB and PMXB þ LPS (P ¼.999).

Discussion

Intrauterine inflammation (frequently caused by microbial

colonization of the amniotic environment7) is hypothesized to

be a primary cause of PTB, most notably in deliveries prior

to 32-week gestation.7,19,20 Much remains to be understood

with regard to the organ-specific short- and long-term out-

comes of exposure to aberrant antenatal inflammation, as high-

lighted by research into the impact of antenatal inflammation

on fetal lung development.15 However, pathological intrauter-

ine inflammation is hypothesized to alter the developmental

trajectory of several fetal organs including the lung,15 brain,38

and gastrointestinal tract.39 Antenatal exposure to pathological

inflammation also programs alterations in glucose and lipid

metabolism that persist into neonatal life.40 In pregnancy,

inflammation is involved both in normal fetal and in gestational

development as well as the pathological processes underlying

PTB and changes in fetal organogenesis.23-25,41 Developing

anti-inflammatory treatments that resolve pathogenic intrauter-

ine inflammation without affecting fetal or gestational develop-

ment is thus likely to constitute an important step in preventing

PTB and fetal injury.

The cationic peptide antibiotic PMXB binds the lipid A core

of LPS, inhibiting proinflammatory activation of the innate arm

of the immune system.28-30,32 Rather than attempting to use

potent pharmacological agents to block intrauterine inflamma-

tion, we hypothesized that adapting the concept of inhibiting

inflammation by binding and functionally inactivating micro-

bial agonist (which we term agonist capture therapy) may be

a strategy for PTB prevention. As a proof-of-principle experi-

ment, we asked whether intra-amniotic administration of PMXB

Figure 2. Amniotic fluid endotoxin concentration at delivery. *, P < .05versus polymyxin B (PMXB) group. n ¼ 5/group.

Saito et al 5

at OhioLink on November 13, 2013rsx.sagepub.comDownloaded from

would inhibit LPS-driven intrauterine inflammation. The admin-

istration of a 1:1 weight ratio of PMXB to LPS partially blocked

intra-amniotic inflammation. The strengths of this study relate to

the well-established sheep model of pregnancy and the signifi-

cant body of data available describing the fetal ovine inflamma-

tory response to intrauterine LPS. The limitations of this study

include the use of a single time point at which inflammation was

assessed and the simultaneous administration of PMXB and LPS

to the intervention group. Administration of PMXB both before

and after LPS administration would likely provide additional

data of interest on both the conceptual utility of agonist capture

therapy and clarify the tissue-specific response identified in the

present study.

Elevated concentrations of IL-6 and IL-8 in the amniotic

fluid are associated with adverse neonatal neurological out-

comes. In a landmark study involving 94 neonates, the presence

of brain white matter lesions were found to be significantly

associated with elevated median concentrations of IL-1b, IL-

6, and TNF-a in amniotic fluid samples obtained by amniocent-

esis.42 A number of subsequent studies in animal models have

demonstrated an association between elevated cytokine con-

centrations in the amniotic fluid and fetal neurological

injury.38,43 In the present study, we demonstrated that, relative

to animals in the LPS group, animals exposed to PMXBþ LPS

had significantly reduced concentrations of both IL-6 and IL8.

Although IL-8 remained elevated relative to animals exposed

to PMXB only, there was no significant difference in the

concentration of IL-6 in the amniotic fluid between animals

in the PMXB and animals in the PMXB þ LPS groups.

Infection44 and/or the introduction of microbial agonists into

the amniotic fluid are both associated with the increased expres-

sion of proinflammatory mediators in fetal tissues including the

membranes,45,46 lung,47-49 skin,26,50 and liver.40 We previously

hypothesized that inflammation in fetal tissues may be trans-

duced to injurious systemic inflammation and have demon-

strated that the lung is a key organ in the transmission of

proinflammatory intra-amniotic stimuli to cause systemic fetal

inflammation.35

In the present study, fetuses exposed to PMXB þ LPS had

large and significant reductions in the expression of IL-1b,

TNF-a, IL-6, IL-8, and MCP-2 mRNA in the lung (but not skin

or chorioamnion), relative to fetuses exposed to LPS only. In

concordance with the reduced pulmonary inflammation, histo-

logical analysis of fetal lung tissues demonstrated a reduction

in airway consolidation and CD3þ infiltration in animals

exposed to PMXB þ LPS, relative to those exposed to LPS

alone. The reasons for the decreased respiratory inflammation

in the presence of robust inflammatory responses in the skin

and chorioamnion are unclear. Previous studies in this model

suggest that LPS-driven increases in cytokine mRNA expres-

sion in the fetal lung lag behind that identified in the chorioam-

nion by at least 5 hours,45 presumably due to a delay in LPS

accessing the fetal airways. We speculate that the observed dif-

ferential tissue response in the present study may be derived

from the immediate exposure of the skin and amnion to

unbound LPS following intra-amniotic injection; in contrast,

a delay in LPS reaching the fetal lung likely allows PMXB

more time to bind and functionally inactivate the LPS, resulting

in significantly less inflammatory activation in respiratory

tissues.

No significant changes in fetal arterial cord blood plasma

IL-6 and IL-8 concentrations or the expression of IL-1b,

TNF-a, IL-6, IL-8, or MCP-2 mRNA in the fetal spleen were

identified in any exposure group, irrespective of exposure. In

Figure 3. A, Representative images of hematoxylin and eosin-stained fetal lung from PMXB- (panel A), PMXB þ LPS- (panel B), and LPS- (panelC) exposed fetuses. Scale ¼ 100 mm. n ¼ 5/group. B, Representative images of CD3 (immunofluorescent) probed fetal lung from PMXB- (panelA), PMXB þ LPS- (panel B), and LPS- (panel C) exposed fetuses. Panel D, CD3 counts (cells/high-powered view). *, P < .05 versus LPS group.Scale ¼ 100 mm. n ¼ 5/group. LPS indicates lipopolysaccharide; PMXB, polymyxin B.

6 Reproductive Sciences 00(0)

at OhioLink on November 13, 2013rsx.sagepub.comDownloaded from

contrast, LPS exposure yielded variable, but statistically signif-

icant increases in liver CRP, hepcidin, and SAA-3 mRNA

expression in the fetal liver relative to animals in both PMXB

and PMXB þ LPS groups. The apparent reductions identified

in hepatic acute-phase protein expression in the PMXB þ LPS

group, relative to the LPS group, were not statistically signifi-

cant (P > .05).

Conclusions

The administration of PMXB decreased LPS-driven inflamma-

tion in the fetal lung but not the fetal skin, chorioamnion, or

liver. Taken together, these data suggest that the concept of

in utero agonist capture therapy offers some promise as a

means of resolving intrauterine inflammation in a targeted

fashion. Two points of particular interest may be drawn from

these data. First, a substantial decrease in pulmonary inflamma-

tion alone was associated with a large reduction in amniotic

fluid IL-6 and IL-8 concentrations. It is tempting to speculate

that the inflammatory state of the fetal lung is thus a major

determinant of amniotic fluid inflammation, although further

work is required to confirm this hypothesis. Second, although

the use of PMXB substantially decreased pulmonary and

amniotic fluid inflammation, this alone was insufficient to

prevent the induction of systemic fetal inflammation, as

demonstrated by increased acute-phase protein expression in

the liver of PMXB þ LPS-treated animals. These data suggest

that successful dampening of fetal inflammation likely requires

inhibiting the production of proinflammatory mediators in

most and perhaps all the fetal tissues that may come into

contact with proinflammatory agonist. This observation is per-

tinent to both our agonist capture strategy and alternative

approaches employing pharmacological methods to downregu-

lated inflammatory signaling pathways. Moreover, given data

Figure 4. Cytokine/chemokine messenger RNA (mRNA) expressionin fetal lung (panel A), skin (panel B), and spleen (panel C). *, P < .05versus PMXB group. �, P < .05 versus PMXB þ lipopolysaccharide(LPS) group. n ¼ 6/group. PMXB indicates polymyxin B.

Figure 6. Concentration of amniotic fluid IL-6 and IL-8 protein. *, P <.05 versus PMXB group. �, P < .05 versus PMXB þ lipopolysaccharide(LPS) group. n¼ 6/group. IL indicates interleukin; PMXB, polymyxin B.

Figure 5. Acute-phase protein messenger RNA (mRNA) expressionin the fetal liver. *, P < .05 versus polymyxin B (PMXB) group. n ¼ 6/group.

Saito et al 7

at OhioLink on November 13, 2013rsx.sagepub.comDownloaded from

demonstrating that bacterial agonist remains in utero for an

extended period,51 the approach used to control inflammation

will likely need to be effective for several weeks at a time

(or able to be repeatedly delivered) and nontoxic to the fetus.

Acknowledgments

The authors wish to acknowledge the generous support received from

Siemens Australia for their donation of Rapidlab 1265 consumables

used in this study. The authors wish to thank Sara and Andrew Ritchie

(Icon Agriculture, Darkan, Western Australia) for the expert provision

of date-mated sheep used in this study.

Authors’ Note

Findings were presented at the 61st Annual Meeting, Society for

Gynaecologic Investigation, Miami, FL, March 16-19, 2011 and the

8th World Congress on Developmental Origins of Health and Disease,

Singapore, November 17-20, 2013.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to

the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for

the research, authorship, and/or publication of this article: This work

was funded by a grant from the Ramaciotti Foundations, Australia, to

MWK and JAK.

References

1. March of Dimes P, Save the Children, WHO. Born Too Soon: The

Gobal Action Report on Preterm Birth. Geneva: World Health

Organization; 2012.

2. Blencowe H, Cousens S, Oestergaard MZ, et al. National,

regional, and worldwide estimates of preterm birth rates in the

year 2010 with time trends since 1990 for selected countries: a

systematic analysis and implications. Lancet. 2012;379(9832):

2162-2172.

3. Newnham JP, Sahota DS, Zhang CY, et al. Preterm birth rates in

Chinese women in China, Hong Kong and Australia—the price of

Westernisation. Aust N Z J Obstet Gynaecol. 2011;51(5):426-431.

4. Petrou S, Eddama O, Mangham L. A structured review of the

recent literature on the economic consequences of preterm birth.

Arch Dis Child Fetal Neonatal Ed. 2011;96(3):F225-F232.

5. Lawn JE, Cousens S, Zupan J, Lancet Neonatal Survival Steering

Team. 4 Million neonatal deaths: when? Where? Why? Lancet.

2005;365(9462):891-900.

6. Tracy SK, Tracy MB, Dean J, Laws P, Sullivan E. Spontaneous

preterm birth of liveborn infants in women at low risk in Australia

over 10 years: a population-based study. BJOG. 2007;114(6):

731-735.

7. Goldenberg RL, Culhane JF, Iams JD, Romero R. Epidemiology

and causes of preterm birth. Lancet. 2008;371(9606):75-84.

8. Bastek JA, Gomez LM, Elovitz MA. The role of inflammation

and infection in preterm birth. Clin Perinatol. 2011;38(3):

385-406.

9. Fawke J. Neurological outcomes following preterm birth. Semin

Fetal Neonatal Med. 2007;12(5):374-382.

10. McAdams RM, Juul SE. Cerebral palsy: prevalence, predictabil-

ity, and parental counseling. Neo Rev. 2011;12(10):e564-e574.

11. Romero R, Espinoza J, Kusanovic JP, et al. The preterm parturi-

tion syndrome. BJOG. 2006;113(suppl 3):17-42.

12. Shatrov JG, Birch SCM, Lam LT, Quinlivan JA, McIntyre S,

Mendz GL. Chorioamnionitis and cerebral palsy: a meta-analysis.

Obstet Gynecol. 2010;116(2 pt 1):387-392.

13. Vermeulen GM, Bruinse HW, de Vries LS. Perinatal risk factors

for adverse neurodevelopmental outcome after spontaneous

preterm birth. Eur J Obstet Gynecol Reprod Biol. 2001;99(2):

207-212.

14. Wood NS, Costeloe K, Gibson AT, et al. The EPICure study:

associations and entecedents of neurological and developmental

disability at the 30 months of age following extremely preterm

birth. Arch Dis Childh Fetal Neonatal Ed. 2005;90(2):F134-F140.

15. Jobe AH. Effects of chorioamnionitis on the fetal lung. Clin

Perinatol. 2012;39(3):441-457.

16. Smith GCS, Pell JP, Dobbie R. Risk of sudden infant death syn-

drome and week of gestation of term birth. Pediatrics. 2003;

111(6 pt 1):1367-1371.

17. Yang S, Platt RW, Kramer MS. Variation in child cognitive

ability by week of gestation among healthy term births. Am J

Epidemiol. 2010;171(4):399-406.

18. Mangham LJ, Petrou S, Doyle LW, Draper ES, Marlow N. The

cost of preterm birth throughout childhood in England and Wales.

Pediatrics. 2009;123(2):e312-e327.

19. Agrawal V, Hirsch E. Intrauterine infection and preterm labor.

Semin Fetal Neonatal Med. 2012;17(1):12-19.

20. Gotsch F, Romero R, Kusanovic JP, et al. The fetal inflammatory

response syndrome. Clin Obstet Gynecol. 2007;50(3):652-683.

21. Keelan JA. Pharmacological inhibition of inflammatory pathways

for the prevention of preterm birth. J Reprod Immunol. 2011;

88(2):176-184.

22. Rinaldi SF, Hutchinson JL, Rossi AG, Norman JE. Anti-

inflammatory mediators as physiological and pharmacological

regulators of parturition. Expert Rev Clin Immunol. 2011; 7(5):

675-696.

23. Hayden MS, West AP, Ghosh S. NF-kB and the immune

response. Oncogene. 2006;25(51):6758-6780.

24. Jiang YJ, Lu B, Crumrine D, Man MQ, Elias PM, Feingold KR.

IL-1a accelerates stratum corneum formation and improves per-

meability barrier homeostasis during murine fetal development.

J Dermatol Sci. 2009;54(2):88-98.

25. Jiang YJ, Lu B, Crumrine D, Elias PM, Feingold KR. IL-6 Stimu-

lates but is not essential for stratum corneum formation and

permeability barrier development during gestation. Exp Derma-

tol. 2010;19(8):e31-e36.

26. Kemp MW, Saito M, Nitsos I, Jobe AH, Kallapur SG, Newnham

JP. Exposure to in utero lipopolysaccharide induces inflammation

in the fetal ovine skin. Reprod Sci. 2010;18(1):88-98.

27. Jones HE, Harris KA, Azizia M, et al. Differing prevalence and

diversity of bacterial species in fetal membranes from very

preterm and term labor. PLoS One. 2009;4(12):e8205.

28. Coyne CP, Fenwick BW. Inhibition of lipopolysaccharide-induced

macrophage tumor necrosis factor-a synthesis by polymyxin B

sulfate. Am J Veterinary Res. 1993;54(2):305-314.

8 Reproductive Sciences 00(0)

at OhioLink on November 13, 2013rsx.sagepub.comDownloaded from

29. Stokes DC, Shenep JL, Fishman M, Hildner WK, Bysani GK,

Rufus K. Polymyxin B prevents lipopolysaccharide-induced

release of tumor necrosis factor-a from alveolar macrophages.

J Infect Dis. 1989;160(1):52-57.

30. Coyne CP, Moritz JT, Fenwick BW. Inhibition of lipopolysacchar-

ide-induced TNF-a production by semisynthetic polymyxin-B

conjugated dextran. Biotechnol Ther. 1994;5(3-4):137-162.

31. Jaber BL, Barrett TW, Cendoroglo Neto M, Sundaram S, King

AJ, Pereira BJ. Removal of cytokine inducing substances by

polymyxin-B immobilized polystyrene-derivative fibers during

in vitro hemoperfusion of 10% human plasma containing Staphy-

lococcus aureus challenge. ASAIO J. 1998;44(1):48-53.

32. Barton MH. Use of Polymyxin B for treatment of endotoxemia in

horses. Comp Cont Educ Pract 2000;22(11):1056.

33. Iwagaki A, Porro M, Pollack M. Influence of synthetic antiendo-

toxin peptides on lipopolysaccharide (LPS) recognition and LPS-

induced proinflammatory cytokine responses by cells expressing

membrane-bound CD14. Infect Immun. 2000;68(3):1655-1663.

34. Kemp MW, Saito M, Kallapur SG, et al. Inflammation of the fetal

ovine skin following in utero exposure to ureaplasma parvum.

Reprod Sci. 2011;18(11):1128-1137.

35. Kemp MW, Senthamarai Kannan P, Saito M, et al. Selective

exposure of the fetal lung and skin/amnion (but not gastro-

intestinal tract) to LPS elicits acute systemic inflammation in fetal

sheep. PLoS One. 2013;8(5):e63355.

36. Bustin SA, Benes V, Garson JA, et al. The MIQE guidelines: min-

imum information for publication of quantitative real-time PCR

experiments. Clin Chem. 2009;55(4):611-622.

37. Zhang L, Saito M, Jobe A, et al. Intra-amniotic administration of

E. coli lipopolysaccharides causes sustained inflammation of the

fetal skin in sheep. Reprod Sci. 2012;19(11):1181-1189.

38. Burd I, Balakrishnan B, Kannan S. Models of fetal brain injury,

intrauterine inflammation, and preterm birth. Am J Reprod Immu-

nol. 2012;67(4):287-294.

39. Wolfs TGAM, Buurman WA, Zoer B, et al. Endotoxin induced

chorioamnionitis prevents intestinal development during gesta-

tion in fetal sheep. PLoS One. 2009;4(6):e5837.

40. Vlassaks E, Gavilanes AWD, et al. Antenatal exposure to

chorioamnionitis affects lipid metabolism in 7-week-old sheep.

J Dev Orig Health Dis. 2012;3(2):103.

41. Liggins GC. Cervical ripening as an inflammatory reaction. In:

Ellwood DA, Anderson ABM, eds. The Cervix in Pregnancy and

Labour, Clinical and Biochemical Investigations. Edinburgh:

Churchill Livingstone; 1981:1.

42. Bo Hyun Y, Jong Kwan J, Romero R, et al. Amniotic fluid inflam-

matory cytokines (interleukin-6, interleukin-1b, and tumor necro-

sis factor-a), neonatal brain white matter lesions, and cerebral

palsy. Am J Obstet Gynecol. 1997;177(1):19-26.

43. Burd I, Bentz AI, Chai J, et al. Inflammation-induced preterm

birth alters neuronal morphology in the mouse fetal brain. J Neu-

rosci Res. 2010;88(9):1872-1881.

44. Gravett MG, Witkin SS, Haluska GJ, Edwards JL, Cook MJ,

Novy MJ. An experimental model for intraamniotic infection and

preterm labor in rhesus monkeys. Am J Obstet Gynecol. 1994;

171(6):1660-1667.

45. Kramer BW, Moss TJ, Willet KE, et al. Dose and time response

after intraamniotic endotoxin in preterm lambs. Am J Respir Crit

Care Med. 2001;164(6):982-988.

46. Grigsby PL, Novy MJ, Adams Waldorf KM, Sadowsky DW,

Gravett MG. Choriodecidual inflammation: a harbinger of the

preterm labor syndrome. Reprod Sci. 2010;17(1):85-94.

47. Kramer BW, Kramer S, Ikegami M, Jobe AH. Injury, inflamma-

tion, and remodeling in fetal sheep lung after intra-amniotic endo-

toxin. Am J Physiol Lung Cell Mol Physiol. 2002;283(2):

L452-L459.

48. Kallapur SG, Willet KE, Jobe AH, Ikegami M, Bachurski CJ.

Intra-amniotic endotoxin: Chorioamnionitis precedes lung

maturation in preterm lambs. Am J Physiol Lung Cell Mol Phy-

siol. 2001;280(3):L527-L536.

49. Davies JK, Shikes RH, Sze CI, et al. Histologic inflammation in

the maternal and fetal compartments in a rabbit model of acute

intra-amniotic infection. Am J Obstet Gynecol. 2000;183(5):

1088-1093.

50. Kim YM, Romero R, Chaiworapongsa T, Espinoza J, Mor G, Kim

CJ. Dermatitis as a component of the fetal inflammatory response

syndrome is associated with activation of Toll-like receptors in

epidermal keratinocytes. Histopathology. 2006;49(5):506-514.

51. Newnham JP, Kallapur SG, Kramer BW, et al. Betamethasone

effects on chorioamnionitis induced by intra-amniotic endotoxin

in sheep. Am J Obstet Gynecol. 2003;189(5):1458-1466.

Saito et al 9

at OhioLink on November 13, 2013rsx.sagepub.comDownloaded from