Smoking during pregnancy influences the maternal immune ...

sfct

Research www.AJOG.org

BASIC SCIENCE: OBSTETRICS

Smoking during pregnancy influences the maternalimmune response in mice and humansJelmer R. Prins, MD; Machteld N. Hylkema, PhD; Jan Jaap H. M. Erwich, PhD; Sippie Huitema;Gerjan J. Dekkema; Frank E. Dijkstra; Marijke M. Faas, PhD; Barbro N. Melgert, PhD

OBJECTIVE: During pregnancy the maternal immune system has to adaptits response to accommodate the fetus. The objective of this study was toanalyze the effects of smoking on the maternal immune system.

STUDY DESIGN: First-trimester decidual tissue and peripheral blood ofsmoking and nonsmoking women were analyzed by real time reverse tran-scription–polymerase chain reaction (RT-PCR) and flow cytometry. Amouse model was used to further analyze the effects of smoking. Murinetissue was analyzed by flow cytometry, real-time RT-PCR, andimmunohistochemistry.

RESULTS: Smoking caused lower percentages of viable pups in mice
and lower birthweights in humans. Smoking mothers, both mice and
Obstet Gynecol 2012;207:76.e1-14.

tostpu

aac

0002-9378/$36.00 • © 2012 Mosby, Inc. All rights reserved. • http://dx.doi.org/10.10

human, had more natural killer cells and inflammatory macrophages lo-cally, whereas systemically they had lower percentages of regulatory Tcells than nonsmoking controls.

CONCLUSION: Maternal smoke exposure during pregnancy influenceslocal and systemic immune responses in both women and mice. Suchchanges may be involved in adverse pregnancy outcomes in smokingindividuals.

Key words: macrophages, maternal immune system, natural killer
cells, smoking
Cite this article as: Prins JR, Hylkema MN, Erwich JJHM, et al. Smoking during pregnancy influences the maternal immune response in mice and humans. Am J

ds

Despite the well-known fact thatsmoking adversely affects preg-

nancy outcome,1 about 30% of pregnantwomen still smoke in The Netherlands.2,3

Fetal reactions to cigarette smoke were al-ready described in 1935.4 After a mothertarted smoking a cigarette, an increase inetal heart rate was found, and it was con-luded that a toxic agent, probably nico-ine, crossed the placenta to the fetus.4

Currently, it is clear that smoking duringpregnancy causes lower birth weights,an increased incidence of spontaneousabortions and preterm birth, increased

From the Department of Obstetrics and GynecoPathology and Medical Biology (Dr Hylkema, Msof Medical Biology, Department of Pathology anUniversity Medical Center Groningen and UniverPharmacokinetics, Toxicology, and Targeting (DNetherlands.

Received Feb. 2, 2012; revised April 3, 2012; ac

This study was supported by the Dutch Kidney Fde Cock Foundation, and Guide Graduate Scho

The authors report no conflict of interest.

Presented, in part, at the 57th Annual ScientificInvestigation, March 24-27, 2010, Orlando, FL,Reproductive Immunology, August 15-19, 2010

Reprints not available from the authors.

placental pathology, and increased stillbirthrates.5-9 Interestingly, it has also been shownhat smoking women have a lower incidencef pregnancies complicated by preeclamp-ia.10 The mechanisms responsible for theoxic and protective effects of smoking onregnancy and preeclampsia are not yetnderstood.Tobaccosmokecontainsmanysubstances

nd may affect the fetus directly, but it maylso affect placental development and de-idualization.1,11,12 Alternatively, smoking

may achieve its effects by modulation of ma-ternalimmunologicalandendocrineparam-

y (Drs Prins and Erwich), Department ofitema, Mr Dekkema, and Mr Dijkstra), Divisionedical Biology (Drs Faas and Melgert),of Groningen, and Department ofelgert), University of Groningen, The

ted April 9, 2012.

ndation (grant 2266, to B.N.M.), Jan Kornelis

ting of the Society for Gynecologicat the 11th International Congress of

airns, Queensland, Australia.

16/j.ajog.2012.04.017

JULY 2012 Ameri

eters. During pregnancy the maternal im-munesystemplaysakeyrole inthesuccessofpregnancy by adapting to accommodate thesemiallogeneic fetus.13 Adaptations of thematernalimmunesystemtakeplacelocallyatthe implantation site as well as in the periph-eral circulation. Disturbances in this adapta-tiontowardmaternaltolerancehavebeenas-sociated with adverse pregnancy outcomesand other reproductive pathologies.13,14 In-

eed, in nonpregnant women and animals,moking alters immune responses.15-19

Therefore, adverse pregnancy outcomes insmoking women could be the result of im-munological changes caused by smoking.

Theexactinfluenceofsmokingonthema-ternal immune system during pregnancy isnot known. Only 1 study investigated the ef-fects of smoking on circulating maternal leu-kocytes during pregnancy, showing thatsmoking pregnant women had higher per-centages of CD3�cells, lower percentages ofCD56� cells, and a lower expression ofCD54 (activation marker) on monocytes inperipheral blood in early pregnancy com-pared with nonsmoking pregnant women.20

It was concluded that smoking acts both asan inflammatory and antiinflammatorystimulus on the maternal immune system

logHu

d Msityr M

cep

ouol.

Meeand, C

during pregnancy.20

can Journal of Obstetrics & Gynecology 76.e1

http://dx.doi.org/10.1016/j.ajog.2012.04.017

cmrboa

bswivmm

rfocsdnmwdetvtwpbpudd

bdTfkustrdtt

uT

(rppwtuwrp

Research Basic Science: Obstetrics www.AJOG.org

Important players in the adaptation ofthe maternal immune response are T cells(especially regulatory T cells), natural killer(NK) cells, and monocytes and macro-phages.13,21-24 RegulatoryT(Treg)cells area subpopulation of T cells that regulateimmune responses. They are important inthe process of tolerance to self and foreignantigens, andtheythereforeplayaroleinau-toimmune diseases, inflammatory diseases,transplantation, and pregnancy.13,25,26 NKells are key players in the regulation of re-odeling of spiral arteries and also have a

ole in regulation of invasion of tropho-lasts.27 Functional and numerical changesf NK and Treg cells are both associated withdverse pregnancy outcomes.13,23,28

Monocytes and macrophages are alsoinvolved in remodeling of spiral arteriesand placental development.29,30 It has

ecome clear that there are specific sub-ets of monocytes and macrophages,hich are thought to have specific roles

n inflammation, tissue remodeling, andascularization, respectively, M1 (inflam-atory) and M2 (remodeling/regulatory)acrophages.24,31,32 Macrophages in the

placenta during healthy pregnancy appearto have a remodeling/regulatory role andare thus mainly of the M2 subtype.24

The aim of this study was to analyzethe effects of smoking on the maternalimmune system during pregnancy lo-cally in the uterus and systemically. Inhumans, smoking effects on the localand systemic maternal immune systemwere investigated in first-trimester de-cidual tissue and peripheral blood ofsmoking women. In this study, we usedunique decidual material, collected dur-ing routine chorionic villous samplingbetween 10 and 12 weeks of pregnancy,which allowed us to study smoking ef-fects in deciduas of ongoing pregnancieswith known pregnancy outcomes. We fur-thermore used a mouse model to comparethe results with human pregnancy andmore specifically study changes in periph-eral and local immune responses.

MATERIALS AND METHODSStudy designFirst-trimester decidual tissue with a knownpregnancy outcome of 42 women, 21
smoking women, and 21 control women
76.e2 American Journal of Obstetrics & Gynecolog

was selected from a tissue collection andanalyzed by real-time reverse transcrip-tion–polymerase chain reaction (RT-PCR).In addition, data on peripheral blood T cellsubsets of 6 smoking and 12 nonsmokingpregnant women were selected from a con-trol cohort of a previously published study.33

Tomorespecificallystudychangesinperiph-eral blood and local immune responses, weused a mouse model. Therefore, 6 pregnantmice were exposed to smoke and 6 controlpregnant mice were exposed to fresh air.

To analyze the effects of smoking on thelocal maternal immune system, decidualtissue was analyzed by immunohisto-chemistry and real-time RT-PCR, anduterus-draining lymph nodes were ana-lyzed by flow cytometry. For the analysis ofthe systemic immune response, systemiclymph nodes and spleens were analyzed byflow cytometry.

Human studyCollection of human decidual tissue. Duringoutine chorionic villus sampling, per-ormed vaginally between 10 and 12 weeksf gestation for maternal age, a previouslyhromosomal abnormal child, or serumcreening-related risk for Down syn-rome, surplus material, which was noteeded for karyotyping, was obtained. Im-ediately after sampling, decidual tissueas mechanically separated from villi un-er a microscope by a qualified, experi-nced laboratory technician to minimizerophoblast contamination. The chorionicilli appeared as free-floating white struc-ures with fluffy, filiforme branches,hereas decidual tissue had a more amor-hous appearance and lacked distinctranches. Earlier studies showed that tro-hoblast cells were seldom found in decid-al samples.34,35 Tissue was stored imme-iately at –20°C until further processing asescribed before.34

Patients were informed that surplusmaterial could be used for research. Fol-low-up of these pregnancies was avail-able by a questionnaire returned by thepatient postpartum. From this material,we selected decidual tissue from smokingwomen, and for each case we used a non-smoking control woman matched for ma-ternal age, parity, and gestational age attime of sampling. Information about
smoking was obtained by medical history 5
y JULY 2012

taking. For cases and controls, only preg-nancies with a known pregnancy outcomelonger than 30 weeks’ gestation were se-lected. In total, decidual tissue of 21 smok-ing and 21 nonsmoking women was se-lected for analysis (Table 1).

To determine whether smoking had adose-dependent effect, we divided thetotal smoking group into a heavy (�10cigarettes per day) and moderate smok-ing group (�10 cigarettes per day). Thelow percentage of male offspring iswithin normal variation and may havebeen influenced by the relatively highmaternal age because there is evidencethat a higher maternal age is associatedwith a shift from a male to a female ma-jority in newborns.36,37

PCR on human decidual tissue. Total ri-onucleic acid (RNA) was isolated fromecidual tissue using a RNA isolationrizol kit (Invitrogen, Carlsbad, CA). To

urther purify RNA, a Turbo DNA freeit and column chromatography weresed according to the manufacturer’s in-tructions (respectively, Applied Biosys-ems, Foster City, CA, and Bio-Rad Labo-atories, Hercules, CA). Complementaryeoxyribonucleic acid (cDNA) was reverseranscribed using a Superscript-II reverseranscriptase kit (Invitrogen).

As a housekeeping gene (HPRT), wesed the forward primer 5=-GGCAGTA-AATCCAAAGATGGTCAA-3=-GTCT-

GGCTTATATCCAACACTTCGT-3= (In-vitrogen), and probe: 6-FAM 5=-CAA-GCTTGCTGGTGAAAAGGACCCC-3=TAMRA(Eurogentic,Herstal,Belgium).Forthe other genes, interleukin (IL) 6, IL10,Tbet21 (Th1 response), Gata 3 (Th2 re-sponse), Ror�t (Th17 response), Foxp3Tregcell),CD56(NKcell),CD68(panmac-ophage marker), NOS2 (iNOS, M1 macro-hage), and CD206 (MRC-1, M2 macro-hage), On-Demand gene expression assaysere used (Applied Biosystems). PCR reac-

ions were performed in triplicate in a vol-me of 10 �L consisting of 0.1 �L of MilliQater, 5 �L PCR mix (Eurogentic), 1 �L of

everse and forward primer each, 0.4 �Lrobe, and 2.5 �L cDNA for the housekeep-

ing gene.For the other genes, the total volume

of 10 �L contained 2 �L of MilliQ water,
�L PCR mix (Eurogentic), 0.5 �L assay

Amtwaoa

Himdons

wgtwgtsLwgeas

–r

CwsEmsouMfi

Obs

www.AJOG.org Basic Science: Obstetrics Research

mix, and 2.5 �L cDNA. Runs were per-formed by a 7900HT Fast real-time PCRsystem (Applied Biosystems), and RNAdata were normalized to HPRT messen-ger RNA (mRNA) expression using 2-�Ct.

fter the run, it appeared that HPRTRNA expression of 4 samples (3 con-

rols, 1 case) was undetectable; these casesere excluded from this study. Undetect-

ble cycle threshold (Ct) values of the genef interest (more than 40) were interpreteds the maximum Ct value (40).

uman flow cytometry. To analyze thenfluence of smoking on the systemic

aternal immune system in humans,ata on T-cell subsets were selected fromf a previously published control preg-ant cohort and reanalyzed for effects ofmoking.33 In short, peripheral blood

samples were obtained from pregnantwomen well before the onset of labor,and T-cell subsets were analyzed usingfluorescently labeled antibodies againstCD4, CD25, and Foxp3 as described be-fore.33 In total, 6 smoking pregnant

omen were selected and matched byestational age, parity, and maternal ageo 2 nonsmoking control pregnantomen per case, leading to a controlroup of 12. Results obtained by flow cy-ometry were reanalyzed using Winlistoftware and cell subsets were calculated.ymphocytes were selected based on for-ard and side scatter plots. Additionalates were used based on fluorescencemission wavelengths of the antibodies,nd negative isotype controls were used to

TABLE 1Characteristics of pregnant women

Characteristic

Age, y...................................................................................................................

Primiparity, %...................................................................................................................

Birthweight, g...................................................................................................................

Sex of child (% male)...................................................................................................................

Duration of pregnancy, d...................................................................................................................

Smoking average, (cigarettes per day)...................................................................................................................

Data are mean � SD.a P �.001 compared with controls when analyzing birthweigh

Prins. Smoking and the maternal immune response. Am J

et the gates (Supplemental Figure 1).

Animal studyAnimals. Female and male C57Bl6 mice,aged 8-10 weeks, were obtained fromHarlan (Horst, The Netherlands). Allmice were held under specific pathogen-free conditions on a 12:12 light-dark cy-cle and were administered food and wa-ter ad libitum. All animal protocols wereapproved by the local committee onanimal experimentation (University ofGroningen, Groningen, The Nether-lands) and were performed under strictgovernmental and international guide-lines on animal experimentation.

For experimental purposes, female micewere treated with 1.5 IU pregnant mare’sserum gonadotropin and 1.25 IU humanchorionic gonadotrophin to induce simul-taneous cycling. To induce pregnancy, 2females were housed with 1 male. Matingwas confirmed by vaginal plug detection.All female mice had been mated, 4 of the 6smoking females (pregnancy rate 67%),and 5 of the 6 nonsmoking females (preg-

TABLE 2Pregnancy outcome in smoke-expo

Variable Nonsm

Pregnancy rate, % 83...................................................................................................................

Number of implantations 14.4 �...................................................................................................................

Number of resorptions 0.60...................................................................................................................

Viable pups, % 96...................................................................................................................

Maternal weight, g 21.5 �...................................................................................................................

Data are mean � SD.a P � .01 in a Student t test; b P � .05.

r the decidual mRNA expression dat

onsmoking� 18 )

Smoking(n � 20 )

37.6 � 2.1 36.8 � 4.3.........................................................................................................................

20.0 25.0.........................................................................................................................

388.6 � 591.6 3339.4 � 600.2.........................................................................................................................

28.6 35.0.........................................................................................................................

273.1 � 18.1 270.9 � 31.1.........................................................................................................................

0 9.2 � 6.3b

.........................................................................................................................

centiles for gestational age; b P �.001 compared with controls.

tet Gynecol 2012.

Prins. Smoking and the maternal immune response. Am J Obs

JULY 2012 Ameri

nancy rate 83%) ended up being pregnant(Table 2). The morning after mating wascalled day 0.5 of gestation. Mated femaleswere separated from males and were killedat day 11.5 of gestation. At day 11.5paraaortic lymph nodes (uterus-draininglymph nodes), inguinal (systemic) lymphnodes, and spleens were removed. In addi-tion, implantation sites were removed andfixed in 4% paraformaldehyde, zinc fixa-tive 38 or immediately snap frozen at

80°C for immunohistochemical andeal-time RT-PCR analysis.

igarette smoke exposure. Cigarette smokeas generated using a TE-10 smoke expo-

ure system of Teague Enterprises Smokexposure System (Woodland, CA). Fe-ale mice were exposed to fresh air or

moke in sessions of 7 hours continu-usly per day, from 7 days before matingntil the day the animals were killed.ice were exposed to 5 cigarettes the

rst day, 10 cigarettes the second day, 25

d and control mice

ing (n � 6) Smoking (n � 6)

67..................................................................................................................

1 14.0 � 3.4..................................................................................................................

.6 3.4 � 1.5a

..................................................................................................................

74b

..................................................................................................................

4 19.6 � 0.8..................................................................................................................

10 cigaretteser day (n � 11)

>10 cigarettesper day (n � 9)

37.5 � 2.8 35.8 � 5.7..................................................................................................................

27.3 22.2..................................................................................................................

612.0 � 574.8 3165.0 � 548.5a

..................................................................................................................

36.4 33.3..................................................................................................................

268.6 � 41.8 273.6 � 14.3..................................................................................................................

4.8 � 2.6b 15.0 � 4.3b

..................................................................................................................

se

ok

.........

2..........

� 0.........

.........

1..........

fo a

N(n

<p

......... .........

......... .........

3 3......... .........

......... .........

......... .........

......... .........

t per

tet Gynecol 2012.


o

fbphv


cigarettes the third day, 50 cigarettes thefourth, and 60 cigarettes on the fifth dayand thereafter. Total particulate mattercounts were at least 100. Kentucky 2R4Fresearch-reference filtered cigarettes(The Tobacco Research Institute, Uni-versity of Kentucky, Lexington, KY)were used.

Mouse decidual real-time RT-PCR. Frozenimplantation sites were sectioned trans-versely at 4 �m and stained with hematox-ylin. Frozen tissue of 1 control mouse wasnot available. Decidual and placental tissuewas identified, selected, and cut from theslides using laser dissection microscopy(Leica, Rijswijk, The Netherlands). Decid-ual and placental tissue was collected sep-arately. Total RNA was isolated from de-cidual and placental tissue using an mRNAeasy kit (QIAGEN, Valencia, CA) accord-ing to the manufacturer’s instructions.cDNA was reverse transcribed using aSuperscript-II reverse transcriptase kit(Invitrogen).

For the housekeeping gene (B2M) andfor the other genes (IL6, IL10, Tbet21[Th1 response], Gata 3 [Th2 response],Ror�t [Th17 response], Foxp3 [Tregcells], Nk1.1 [NK cells], CD68 [pan mac-rophage marker], NOS2 [M1- macro-phages; iNOS], and CD206 [M2-macro-phages; MRC-1]), On-Demand geneexpression assays were used (AppliedBiosystems). PCRs were performed intriplicate in a volume of 10 �L consisting

f 2 �L of MilliQ water, 5 �L PCR mix(Eurogentic), 0.5 �L assay mix, and 2.5�L cDNA. Runs were performed by a7900HT Fast real-time PCR system (Ap-plied Biosystems), and mRNA data werenormalized to B2M mRNA expressionusing 2-�Ct. Undetectable Ct values(greater than 40) were interpreted as themaximum Ct value (40).

Mouse flow cytometry. Immediately afterthe animals were killed, paraaortic lymph

A, Higher CD56 mRNA expression in smoking pregurther analyzing dose effects of smoking on mRNAars), and heavy smokers (�10 cigarettes per dayregnant women compared with nonsmoking pregnigher ratios of M1 macrophage markers in heavyariance and Dunnet post hoc testing on log transfo

Prins. Smoking and the maternal immune response. Am J Obstet


nodes (uterus-draining lymph nodes), in-guinal (systemic) lymph nodes, andspleens were removed and thoroughly dis-persed into single cell suspensions in fluo-rescence-activated cell sorting (FACS) buf-fer (consisting of PBS with 2% fetal calfserum). Red blood cells in cell suspensionswere lysed using NH4Cl solution.

For flow cytometric analysis, 1 millioncells were stained according to standardmethods. In short, cells were incubatedfor 5 minutes with 10% normal mouseserum to block a-specific binding. Cellsfrom the lymph nodes were then incu-bated with a cocktail of fluorescently la-beled antibodies for 30 minutes, includ-ing the following: Pacific Blue-labeledanti-CD3 (17A2; BioLegend, San Diego,CA), PerCP-labeled anti-CD4 (RM4-5;BD Pharmingen, San Diego, CA), Alexa700-labeled anti-CD8 (53-6.7, Bioleg-end), allophycocyanin (APC)-labeledanti-CD25 (3C7; BD Pharmingen), andPE-Cy7-labeled anti-NK1.1 (PK 136;BioLegend). After surface staining, in-tracellular staining was performed ac-cording to the Foxp3-staining kit in-structions (eBioscience, San Diego, CA)using fluorescein isothiocyanate (FITC)-labeled anti-Foxp3 (FJK-16s; eBioscience).

Cells from the spleen were not only in-cubated with the T-cell antibody cocktaildescribed in the previous text but werealso incubated with a monocyte anti-body cocktail: PerCP-labeled anti-CD4(RM4-5; BD Pharmingen), APC-labeledanti-CD11b (M1/70; BD Pharmingen),phycoerythrin (PE)-labeled anti-CD43(1B11; Biolegend), PE-Cy7-labeled anti-Gr-1 (RB6-8C5;Biolegend),PacificBlue-la-beled anti-F4/80 (BM8; Biolegend), andFITC-labeled anti-MHC-II (2G9; BDPharmingen).

Cells were analyzed on a LSRII (Beck-ton Dickinson, Lincoln Park, NJ) flowcytometer. Results were analyzed withFacs Diva software (Beckton Dickinson).

t women (gray bars, n � 20) compared with nonsmressions in nonsmokers (white bars, n � 18), mod

9, black bars), a higher expression of Gata3, CDwomen. C, There was a dose effect of smoking on mokers. Data are expressed as mean � SEM, statised data. Asterisk indicates P � .05. Double asteris

Gynecol 2012.

y JULY 2012

Lymphocytes were selected based on for-ward and side scatter plots and subse-quently staining for CD3, CD4, CD25,and Foxp3. To identify cells positive forthe labeled antibodies, additional gateswere set based on fluorescence emissionwavelengths of the used antibodies andnegative controls (Supplemental Figure1). For monocytes all viable cells weregated for CD11b and F4/80. CD11b-hiand F4/80-intermediate cells were con-sidered monocytes, and these were gatedfor CD43 and GR1 to distinguish theGR1-hi and GR1-lo subset (Supplemen-tal Figure 2).

(Immuno)histochemical analysis mice ute-rine tissue. Implantation sites were fixedin paraformaldehyde and zinc fixative38

for 24 hours, transferred into 70% etha-nol, and embedded in paraffin accordingto standard methods and sectionedtransversely at 7 �m. Slides were dew-axed in xylene and rehydrated. For gen-eral histological examination, sectionswere stained with hematoxylin-eosin.First, myometrial, decidual, and placentalsizes, and ratios of decidual spiral arteryvessel to lumen area were measured.39

T cells and Foxp3� cells were stained us-ing primary antibodies against CD3(ab16669; Abcam, Cambridge, MA), andFoxp3 (FJK16s; eBioscience). Antigenretrieval was performed by incubatingsections (15 minutes at 400 W) in citratebuffer.

Sections were incubated with primaryantibodies diluted 1:400 (CD3) or 1:50(Foxp3) overnight at 4°C. Thereaftersections were incubated for 30 minuteswith the appropriate secondary and ter-tiary peroxidated antibodies (CD3), orslides were incubated with a biotinylatedsecondary antibody and were incubatedthereafter with a streptavidin-peroxidasecomplex (Foxp3). Positive staining waslocalized using diaminobenzidine tetra-

g pregnant women (white bars, n � 18). B, Whente smokers (�10 cigarettes per day, n � 11, grayCD68, IL6, and iNOS was found in heavy smokingA expressions of macrophage subset markers, withl analysis was performed using 1-way analysis ofndicate P � .01.

nan okinexp era

, n � 56,ant RNsm ticarm ks i


FIGURE 1mRNA expression in first-trimester decidual tissue

JULY 2012 American Journal of Obstetrics & Gynecology 76.e5

Obs


chloride (DAB). NK cells were stainedusing biotinylated lectin for dolichos bi-florus (Sigma, St. Louis, MO) diluted1:800 and were incubated overnight at4°C. Sections were then incubated with astreptavidin-peroxidase complex for 30minutes at room temperature, and posi-tive staining was localized using 3-amino-9-ethyl-carbazole (AEC).

Subsets of macrophages were stainedusing primary antibodies against F4/80(pan macrophage marker) (MCA497;Serotec, Raleigh, NC), IRF5 (M1 macro-phage lineage marker) (10547-1-AP;Proteintech, Manchester, UK), andYM-1 (M2 macrophage lineage marker)(Mouse Chitinase 3-like; R&D SystemsEurope Ltd, Abingdon, UK). Antigenretrieval was performed by incubatingsections (15 minutes at 400W) in ci-trate buffer. Sections were incubatedwith diluted antibody solution, 1:50(F4/80), 1:100 (IRF5), or 1:400 (YM-1)for 1 hour at room temperature (IRF5,YM-1) or were incubated overnight at4°C (F4/80). Thereafter sections wereincubated for 30 minutes with the ap-propriate secondary and tertiary per-oxidated antibodies (IRF5, YM-1), orsections were incubated with a biotin-ylated secondary antibody and were in-cubated with ABC elite kit (VectorLaboratories, Burlingame, CA) (F4/80). Positive staining was localized us-ing DAB (IRF5) or AEC (F4/80, YM-1). All sections were counterstainedusing hematoxylin and all incubationsteps were followed by 3 washes in PBSfor 5 minutes.

Except for the Foxp3 staining, whichwas performed in triplicate, everystaining was performed in duplicatewith at least 7 sections in between toavoid duplicate counting of cells. Sec-tions were scanned using a NanoZo-omer (Hamamatsu, Shizuoka, Japan).Morphometric analysis was performedby setting a staining threshold manu-ally, based on negative control sec-tions, and measuring the total stainedarea using ImageJ (National Institutesof Health, Bethesda, MD).

Statistical analysisFor statistical analysis, SPSS (SPSS, Inc.,
Chicago, IL) was used. All real-time RT-

PCR data were log transformed beforestatistical analysis. Statistical compari-sons between human smoking andnonsmoking groups were performedusing 1-way analysis of variance andDunnett post hoc testing. Statisticalcomparisons between smoke-exposedand control mice groups were per-formed using Student t tests. Pearsoncorrelation coefficients were calculatedto analyze the correlation betweennumber of cigarettes and mRNA ex-pression of several genes. P � .05 wasconsidered to be significant, and P �.10 was considered a statistical trend.

RESULTSSmoking influences the humanmaternal immune systemlocally and systemicallySmoking affects immune parameters in hum-an decidual tissue. To investigate the ef-fect of human maternal smoking on lo-cal immune parameters during preg-nancy, first-trimester decidual tissuewas analyzed for mRNA expression ofdifferent immune markers. Whencomparing the total smoking groupwith the control group, a significantlyhigher CD56 mRNA expression in

TABLE 3Characteristics of pregnant women

Characteristic Nonsm

Age, y 32.8...................................................................................................................

Primigravidity, % 25.0...................................................................................................................

Birthweight, g 3798.6...................................................................................................................

Sex of child (% male) 53.0...................................................................................................................

Duration of pregnancy, d 261.7...................................................................................................................

Data are mean � SD.


TABLE 4Percentages of T-cell subsets in hu

Variable No

CD4�CD25� 36...................................................................................................................

CD4�CD25�Foxp3– (Teff cells) 38...................................................................................................................

CD4�Foxp3� (Treg cells) 11...................................................................................................................

Data are mean � SD. Subsets are expressed as a percentagea P � .05.

Prins. Smoking and the maternal immune response. Am J Obs

y JULY 2012

smoking women was found as com-pared with nonsmoking women. Nosignificant differences in expression ofthe other genes were found (Figure 1,A). However, when further dividingthe smoking women into a moderate(�10 cigarettes per day) and a heavysmoking (�10 cigarettes per day) sub-group, a significant increase was seenin IL6, Gata 3 (Th2 response), CD56(NK cells), and iNOS (M1 macro-phages) and a trend toward higherCD68 (pan macrophages) mRNA ex-pression in heavy smoking womencompared with nonsmoking women(Figure 1, B). Because iNOS (M1 mac-rophages) was increased in smokingwomen, we analyzed iNOS/CD68 andCD206/CD68 ratios to identify M1 andM2 macrophages as well as the iNOS/CD206 ratio as a measure for the ratioof M1/M2 macrophages (Figure 1, C).In moderate smokers a significanthigher CD206/CD68 ratio was found,whereas in heavy smokers significantlyincreased iNOS/CD206 and iNOS/CD68 macrophage ratios were found(Figure 1, C). Interestingly, when cal-culating correlation coefficients (CC),significant positive correlations be-

r the peripheral blood data

ng (n � 12) Smoking (n � 6)

4.6 30.0 � 4.2..................................................................................................................

50.0..................................................................................................................

465.5 3273.0 � 637.6..................................................................................................................

60.0..................................................................................................................

22.5 265.3 � 23.3..................................................................................................................

tet Gynecol 2012.

an peripheral blood

moking (n � 12) Smoking (n � 6)

9.5 40.7 � 9.9..................................................................................................................

9.8 30.8 � 9.4..................................................................................................................

5.6 6.1 � 4.7a

..................................................................................................................

D4� cells.

fo

oki

�.........

.........

�.........

.........

�.........

m

ns

.1 �.........

.4 �.........

.9 �.........

of C

tet Gynecol 2012.

dcws

b1tirl

sif

dl


tween the number of cigarettes andmRNA expressions were found for IL6(CC 0.64 P � .001), CD56 (CC 0.58,P � .001), and CD68 (CC 0.55, P � .001)(data not shown).

Lower percentages of Treg cells in peripheralblood of smoking pregnant women. To ana-lyze the influence of smoking on the sys-temic human maternal immune re-sponse, data of a previously publishedstudy were reanalyzed (Table 3).33

Smoking pregnant women had signifi-cantly lower percentages of Treg cells(CD4�Foxp3�) than nonsmokingpregnant women (P � .05) (Table 4). No

ifferences in percentages of effector Tells (Teff) (CD4�CD25�Foxp3–) cellsere found between smoking and non-

FIGURE 2Expression of mRNA in decidual tis

A, Higher expressions of IL10 and Ror�t were oifferences were found in macrophage subset ra

og-transformed data. Asterisk indicates P � .0Prins. Smoking and the maternal immune response. Am J Ob

moking pregnant women (Table 4). b

Smoking adversely affects pregnancyin mice and humansEffects of smoking on pregnancy outcomein mice and humans. It is well estab-lished that smoking during pregnancyadversely affects human pregnancy out-come.1,7 In this study, no differences inbirthweights were found between the to-tal human smoking group and controlgroup (Table 1). However, percentilebirthweights in the heavy smoking groupwere significantly (P � .001) lower than

irthweights in the control group (Table, not all data shown). Our results showhat smoke exposure to pregnant micenduced a significantly higher number ofesorptions (Table 2) and significantlyower percentages of viable pups. Num-

e of smoke-exposed and control pre

rved in smoke-exposed mice (gray bars) compa. Data are expressed as mean � SEM, and statis

Gynecol 2012.

ers of implantations were similar in

JULY 2012 Ameri

moke- and air-exposed mice, suggest-ng that smoking does not adversely af-ect implantation.

Smoking influences the maternalimmune system locally andsystemically in miceSmoking affects immune parameters in mur-ine decidual tissue. In contrast to thehuman decidual PCR data, mRNA ex-pressions of IL10 and Ror�t (Th17 re-sponse) were significantly increased insmoke-exposed mice (Figure 2, A). Therewere no significant differences in macro-phage subset ratios between smoke-ex-posed and control mice (Figure 2, B). ThemRNA expressions of all tested genes inplacental tissue were similar in smoke-ex-posed and control mice (results not

nt mice

with control mice (white bars). B, No statisticall analysis was performed using Student t test on

su gna

bse redtios tica5.stet

shown). We further analyzed T-cell sub-


.

c


sets locally in uterus-draining lymphnodes using flow cytometry. Our datashow that locally, in the uterine draininglymph nodes, smoking had no significanteffect on percentages of any T-cell subsetsin mice (results not shown).

To further analyze whether smokingcaused local changes in placentas and de-ciduas of mice, we histologically and im-munohistochemically analyzed murineuterine tissue. Histological analysisshowed comparable myometrial, decidual,and placental area sizes and similar ratiosof decidual spiral artery vessel to lumenarea in smoke-exposed and control mice

FIGURE 3Quantification of immune parametestaining of decidual tissue of smok

A, Higher numbers of CD3� cells, natural killerontrol mice. B, Lower Foxp3/CD3 and higher IRF

expressed as mean � SEM, and statistical analPrins. Smoking and the maternal immune response. Am J Ob

(results not shown). Immunohistochemi-


cal staining of deciduas showed presence ofT cells (CD3), Treg cells (Foxp3), NK cells(DBA lectin), macrophages (F4/80), andmacrophage subsets (IRF5 [M1] andYM-1 [M2]). All negative controls did notshow staining. Decidual tissue of smoke-exposed mice showed significantly morecells positive for CD3 (T cells), DBA (NKcells), and IRF5 (M1 macrophages) and asa result of that a higher IRF5/F4/80 ratio(Figure 3).

Smoking influenced the systemic percent-ages of different T cell subsets in pregnantmice. To analyze the systemic effects of

after immunohistochemicalexposed and control mice

ls (DBA lectin), and M1 macrophages (IRF5) in s4/80 ratios in decidual tissue of smoke-exposedwas performed using Student t test on log-tran

Gynecol 2012.

smoking on the immune response of preg-

y JULY 2012

nant mice, we analyzed changes in T-cellsubsets in inguinal lymph nodes andspleens of pregnant mice. Percentages ofTeff (CD3�CD4�CD25�Foxp3–) weresignificantly lower in systemic lymphnodes of smoke exposed mice as comparedwith control mice. Furthermore, these Teffcells showed a trend to be lower in thespleens of smoke-exposed mice (P � .05)(Table 5). Our data also showed that smoke-exposed mice show a trend toward lowerpercentages of Treg cells in the spleens (P �06) (Table 5). Percentages of CD8� cellsshowedatrendtowardhigherpercentages insmoke-exposedmiceinbothinguinal lymph

ke-exposed mice (gray bars) as compared withe (n � 4) compared with control mice. Data are

med data. Asterisk indicates P � .05.

rse-

cel mo5/F mic

ysis sforstet

nodes and spleen (Table 5).

stcls(nsTpm

tsmhtiiefmsb

sssIflcwpic

bdamihmpaiasd

snscsm

cw

Obs


Smoking only influences numbers of Gr1-lomonocytes in pregnant mice. Our datahow that smoke-exposed mice had arend toward lower numbers of mono-ytes (Figure 4, A), mainly caused byower numbers of the GR1-lo monocyteubset, the nonclassical monocyte subsetFigure 4, B). There was no difference inumbers of GR1-hi monocytes, the clas-ical monocyte subset (Figure 4, C).here were no other differences in ex-ression of activation markers on theonocyte subsets (data not shown).

COMMENTIt has been known for a long time thatsmoking during pregnancy is associatedwith adverse pregnancy outcomes.1,9 Inhis study, we analyzed the effects ofmoking on the local and systemic hu-

an maternal immune system duringuman pregnancy. To further analyzehe effects of smoking on the maternalmmune system, we used a mouse smok-ng model. We indeed found that smokexposure during pregnancy adversely af-ects pregnancy outcome in human and

ice as shown by higher numbers of re-orptions in pregnant mice and lowerirthweights in heavy smoking women.In this study, we further showed that

moking affects the maternal immuneystem locally and systemically in theame way in pregnant humans and mice.n both humans and mice, decidual in-ammatory macrophages (M1) and NKells were higher in smokers as comparedith nonsmoking controls. Systemically,ercentages of Treg cells were lower

n smoking mice and women than inontrols.

In this study, we found higher num-ers of DBA lectin-positive NK cells inecidual tissue of smoke-exposed micend higher mRNA expression of NK cellarker CD56 in decidual tissue of smok-

ng women. The higher numbers andigher mRNA expression of NK cellarkers might be involved in the adverse

regnancy outcomes in smoking micend women. NK cells are 1 of the keymmune players during pregnancy andre regarded as especially important inupporting vascular remodeling and
ecidualization.23,40
Higher numbers of NK cells are asso-ciated with adverse pregnancy out-comes.28,41 The mechanism by whichmoking affects uterine NK (uNK) cellumbers remains unknown from thistudy. Theoretically, numbers of NKells could increase in the deciduas ofmoking women and smoke-exposed

ice as a response to oxidative stress,

FIGURE 4Absolute numbers of monocyte subof smoke-exposed and control mic

, A trend (P � .076) toward lower absolute nuice compared with control mice. B, Significan

ifference in numbers of GR1-hi monocytes in splice. Data are expressed as mean � SEM, and s

n log-transformed data. Asterisk indicates P �rins. Smoking and the maternal immune response. Am J Ob

TABLE 5Percentages of peripheral T-cell su

Variable

Nonsmok

Inguinal

CD4� 51.8 � 1....................................................................................................................

CD4�CD25� 6.0 � 1....................................................................................................................

CD4�CD25�Foxp3– (Teff cells) 1.3 � 0....................................................................................................................

CD4�Foxp3� (Treg cells) 5.2 � 1....................................................................................................................

CD8� 47.4 � 1....................................................................................................................

Data are mean � SD. All T-cell subsets are CD3� and exprea P � .1; b P � .05.


JULY 2012 Ameri

aused by impaired vascularization,hich is caused by smoking.42

The higher numbers of NK cells in de-cidual tissue of smoking mice andwomen could also be a compensatory re-sponse for impairment of these NK cellsby smoking as in a nonpregnant popula-tion, it has been reported that smokingimpairs NK cell cytotoxic activity.43 Al-

ts in spleens

ers of monocytes in spleens of smoke-exposedower numbers of GR1-lo monocytes and C, nos of smoke-exposed mice compared with controltical analysis was performed using Student t test.ynecol 2012.

ets in mice

(n � 5) Smoking (n � 4)

Spleen Inguinal Spleen

57.1 � 2.6 49.2 � 2.4a 54.0 � 3.2a

..................................................................................................................

7.4 � 1.2 5.0 � 1.5 6.2 � 0.9..................................................................................................................

1.5 � 0.5 0.93 � 0.09b 0.92 � 0.2a

..................................................................................................................

7.5 � 0.5 4.4 � 1.6 6.7 � 0.6a

..................................................................................................................

35.8 � 3.2 50.6 � 1.9a 40.2 � 2.1b

..................................................................................................................

as a percentage of CD3� cells.

tet Gynecol 2012.

see

A mbm tly ld eenm tatiso .05P stet G

bs

ing

3.........

2.........

2.........

2.........

5.........

ssed


aflb

bhmavmlvn

aspsspsccma

sleIrm

mfmipn(dh


though this compensatory increase maybe necessary to stimulate vascularizationof the decidua and placenta, it mighthave adverse effects on inflammatory pa-rameters and thereby cause adversepregnancy outcome. Whether smokingindeed affects uNK cell function is un-known. Effects of smoking on uNK cellsmay be different from effects of smokingon NK cells in nonpregnant women be-cause uNK cells produce cytokines andregulate their actions differently fromsystemic NK cells.43

Interestingly, contrary to the higherlocal numbers of NK cells, systemicallywe found lower percentages of NK cellsin spleens of smoke-exposed mice. Thesesystemically lower percentages of NKcells in smoke-exposed mice are in linewith previous human research, showinglower percentages of CD56� NK cells inperipheral blood of smoking pregnantwomen.20 However, as explained above,systemic and uterine NK cells have dif-ferent functions and capacities. Futureresearch focusing on the influence ofsmoking on the function of decidual andsystemic NK cells during is needed toelucidate these intriguing findings.

Other findings that might be evidencefor an altered decidual and placental vas-cularization in smoke-exposed mice andsmoking women would be the changedratios of inflammatory (M1)/regulatory(M2) macrophage subsets in the deciduaof smoke-exposed mice and smokingwomen. It has been shown that in decid-ual tissue there are 2 different subsets ofmacrophages.24,31,44 These subsets arectivated in different ways and have dif-erent functions.44 Higher expressionevels for M2 macrophage markers haveeen found in first-trimester decidua.45

This regulatory, tissue-repairing macro-phage subset may be higher during preg-nancy to promote maternal toleranceand coregulate vascular remodeling. Thegreater presence of M1 subset markers inthe decidua of smoke-exposed mice andsmoking women indicates a skewing to-wards an inflammatory environment inthe decidua, and furthermore, it mightindicate impaired vascularization causedby smoke exposure.

Interestingly, systemically, we found
lower numbers and percentages of
76.e10 American Journal of Obstetrics & Gynecolo

GR1-lo monocytes in smoke-exposedmice. A previous study showed angio-genic properties of these GR1-lo mono-cytes and also showed that GR1-lomonocytes can transform into (M2)macrophages regulating tissue remodel-ing and angiogenesis.46 The lower num-

ers of the M2 precursor monocytes andigher mRNA expression of M1 markersight both reflect the same change from

regulatory remodeling macrophage en-ironment toward an inflammatoryacrophage environment. This could

ead to altered decidual remodeling andascular changes leading to adverse preg-ancy outcome.We found significantly lower percent-

ges of Treg cells in peripheral blood ofmoking women. In line with this lowerercentage in human peripheral blood,moke exposure in mice did result in atatistical trend (P � .085) toward lowerercentages of Treg cells in spleens ofmoke-exposed mice. Furthermore, lo-ally, we found a lower Foxp3�/CD3�ell ratio in deciduas of smoke-exposedice. Because it is known that Treg cells

re essential for pregnancy success,13 thelower systemic percentages of Treg cellsand lower Foxp3�Treg/CD3� T-cellratios in smoking pregnant women andin smoke-exposed mice are possibly in-volved in the increase in adverse preg-nancy outcomes in smoking women andmice. The higher Ror�t expression, sug-gesting increased Th17 responses, in thedecidua of smoke-exposed mice is in linewith previous research showing induc-tion of a Th17-like inflammatory envi-ronment in smoke-exposed nonpreg-nant mice.19 It is known that increases inthe Th17 cell subset and imbalances be-tween the different T-cell subsets canlead to adverse pregnancy outcomes.47

The higher Ror�t mRNA expression inmice could therefore also reflect a mech-anism responsible for the adversely af-fected pregnancy outcomes in smoke-exposed mice.

The lower numbers of Treg cells andhigher expression of Ror�t could be ex-plained by the higher IL6 mRNA expres-sion because it is known that IL6 has animportant role in the differentiation ofnaïve T cells into Treg or Th17 cells.48,49

Our data of increased IL6 in the human a

gy JULY 2012

deciduas is in line with the previously de-scribed effect of smoking on IL6 levels innonpregnant mice.19 During pregnancy,IL6 is produced systemically and also lo-cally by uterine, decidual cells and mac-rophages, and it has been shown that IL6has constant levels during pregnancywith a slight increase at the end ofpregnancy.50-52

Interestingly, it is further shown thatlevels of IL6 are higher in placental anddecidual tissue from abortion-proneCBA/J � DBA/2J mice compared withnormal pregnant mice.51 In humanpregnancy, higher amniotic levels of IL6are associated with preterm birth53

Therefore, the higher mRNA expressionof IL6 we found could be responsible foradverse pregnancy outcomes. This couldbe the result of higher numbers of M1(inflammatory) macrophages becauseM1 macrophages are known to produceIL6. However, whether higher expres-sion of IL6 is a cause or consequence ofthe more inflammatory macrophages re-mains to be answered.

The gestational time points used inour experiments were different for miceand humans. In humans, we analyzedfirst-trimester decidual tissue, and pe-ripheral blood was taken from third-tri-mester pregnant women, whereas mu-rine decidual tissue, lymph nodes, andspleens were obtained midregnancy. Al-though, we did find similar changes innumbers of NK cells, macrophage sub-sets, and Treg cells, we could not confirmthe higher mRNA expression levels ofIL10 and Ror�t that were found inmoke exposed mice. Furthermore, theower percentages of Teff cells in smoke-xposed mice and higher expression ofL6 mRNA in smoking women did noteach significance in the comparable hu-an/mouse group.The different gestational time pointsight be responsible for differences we

ound in the effects of smoking betweenice and human because it is known that

mmunological parameters vary duringregnancy.13 In addition, the smallumber of mice used in this experiment6 per group) could explain some of theifferences. Furthermore, the fact thatumans are continuously challenged by
ntigens, contrary to the mice used in


our experiments, and the fact that weused syngeneic mating couples in ourmice experiments, which is contrary tothe human situation, might also be re-sponsible for differences in immunolog-ical findings. Our mice and human data,however, clearly show the adverse effectsof smoking.

In summary, smoke exposure duringpregnancy influences the local and sys-temic maternal immune response in bothhuman and mice and does this in a dose-dependent manner. The higher numbersand/or expression of IL6, decidual NKcells, M1 macrophages, and lower Treg/CD3 cell ratio in addition to the systemi-cally lower percentages of regulatory T cellsin smoke-exposed mice and smokingwomen may be responsible for the increasein adverse pregnancy outcomes. If thesedifferences in peripheral and local im-mune responses after smoking could alsoexplain the lower incidence of preeclamp-sia in smoking women remains an openquestion. However, the immunologicalchanges seen in smoke-exposed pregnantwomen and mice will help us define themechanisms causing complicated preg-nancies and ultimately finding noveltherapies. f

ACKNOWLEDGMENTSWe would like to acknowledge A. Lech (PhD stu-dent) and M. Geerlings (technician) for helpingwith the animal experiments and J. Seinen andJ.D. Heeringa (biology students) for their help withthe PCR experiments. The acknowledged per-sons did not receive any compensation.

REFERENCES1. Dechanet C, Anahory T, Mathieu Daude JC,et al. Effects of cigarette smoking on reproduc-tion. Hum Reprod Update 2011;17:76-95.2. Goedhart G, van der Wal MF, Cuijpers P,Bonsel GJ. Psychosocial problems and contin-ued smoking during pregnancy. Addict Behav2009;34:403-6.3. Jaddoe VW, Verburg BO, de Ridder MA, etal. Maternal smoking and fetal growth charac-teristics in different periods of pregnancy: thegeneration R study. Am J Epidemiol 2007;165:1207-15.4. Herriot A, Billewicz WZ, Hytten FE. Cigarettesmoking in pregnancy. Lancet 1962;1:771-3.5. Ness RB, Grisso JA, Hirschinger N, et al. Co-caine and tobacco use and the risk of sponta-neous abortion. N Engl J Med 1999;340:333-9.6. Pollack H, Lantz PM, Frohna JG. Maternal
smoking and adverse birth outcomes among
singletons and twins. Am J Public Health2000;90:395-400.7. Andres RL, Day MC. Perinatal complicationsassociated with maternal tobacco use. SeminNeonatol 2000;5:231-41.8. DiFranza JR, Lew RA. Effect of maternal cig-arette smoking on pregnancy complicationsand sudden infant death syndrome. J FamPract 1995;40:385-94.9. Voigt M, Zels K, Guthmann F, Hesse V, Gor-lich Y, Straube S. Somatic classification of ne-onates based on birth weight, length, and headcircumference: quantification of the effects ofmaternal BMI and smoking. J Perinat Med2011;39:291-7.10. England L, Zhang J. Smoking and risk ofpreeclampsia: a systematic review. Front Biosci2007;12:2471-83.11. Card JP, Mitchell JA. The effects of nicotineadministration on deciduoma induction in therat. Biol Reprod 1978;19:326-31.12. Khorram O, Han G, Magee T. Cigarettesmoke inhibits endometrial epithelial cell prolif-eration through a nitric oxide-mediated path-way. Fertil Steril 2010;93:257-63.13. Guerin LR, Prins JR, Robertson SA. Regu-latory T-cells and immune tolerance in preg-nancy: a new target for infertility treatment?Hum Reprod Update 2009;15:517-35.14. Munoz-Suano A, Hamilton AB, Betz AG.Gimme shelter: the immune system duringpregnancy. Immunol Rev 2011;241:20-38.15. Brandsma CA, Hylkema MN, Geerlings M,et al. Increased levels of (class switched) mem-ory B cells in peripheral blood of current smok-ers. Respir Res 2009;10:108.16. Corre F, Lellouch J, Schwartz D. Smokingand leucocyte-counts. Results of an epidemio-logical survey. Lancet 1971;2:632-4.17. van Eeden SF, Hogg JC. The response ofhuman bone marrow to chronic cigarette smok-ing. Eur Respir J 2000;15:915-21.18. Melgert BN, Postma DS, Geerlings M, et al.Short-term smoke exposure attenuates ovalbu-min-induced airway inflammation in allergicmice. Am J Respir Cell Mol Biol 2004;30:880-5.19. Melgert BN, Timens W, Kerstjens HA, et al.Effects of 4 months of smoking in mice withovalbumin-induced airway inflammation. ClinExp Allergy 2007;37:1798-808.20. Luppi P, Lain KY, Jeyabalan A, DeLoia JA.The effects of cigarette smoking on circulatingmaternal leukocytes during pregnancy. Clin Im-munol 2007;122:214-9.21. Aluvihare VR, Kallikourdis M, Betz AG. Reg-ulatory T cells mediate maternal tolerance to thefetus. Nat Immunol 2004;5:266-71.22. Hanna J, Goldman-Wohl D, Hamani Y, et al.Decidual NK cells regulate key developmentalprocesses at the human fetal-maternal inter-face. Nat Med 2006;12:1065-74.23. Moffett-King A. Natural killer cells and preg-nancy. Nat Rev Immunol 2002;2:656-63.24. Nagamatsu T, Schust DJ. The immuno-modulatory roles of macrophages at the mater-
nal-fetal interface. Reprod Sci 2010;17:209-18.
JULY 2012 Americ

25. Sakaguchi S, Miyara M, Costantino CM,Hafler DA. FOXP3� regulatory T cells in the hu-man immune system. Nat Rev Immunol 2010;10:490-500.26. Sakaguchi S, Sakaguchi N, Asano M, ItohM, Toda M. Immunologic self-tolerance main-tained by activated T cells expressing IL-2 re-ceptor alpha-chains (CD25). Breakdown of asingle mechanism of self-tolerance causes var-ious autoimmune diseases. J Immunol 1995;155:1151-64.27. Plaisier M. Decidualisation and angiogene-sis. Best Pract Res Clin Obstet Gynaecol2011;25:259-71.28. Lash GE, Bulmer JN. Do uterine naturalkiller (uNK) cells contribute to female reproduc-tive disorders? J Reprod Immunol 2011;88:156-64.29. Harris LK. Review: trophoblast-vascular cellinteractions in early pregnancy: how to remodela vessel. Placenta 2010;31(Suppl):S93-8.30. Shakhawat A, Shaikly V, Elzatma E, Mavra-kos E, Jabeen A, Fernandez N. Interaction be-tween HLA-G and monocyte/macrophages inhuman pregnancy. J Reprod Immunol 2010;85:40-6.31. Houser BL, Tilburgs T, Hill J, Nicotra ML,Strominger JL. Two unique human decidualmacrophage populations. J Immunol 2011;186:2633-42.32. Martinez FO, Gordon S, Locati M, Man-tovani A. Transcriptional profiling of the humanmonocyte-to-macrophage differentiation andpolarization: new molecules and patterns ofgene expression. J Immunol 2006;177:7303-11.33. Prins JR, Boelens HM, Heimweg J, et al.Preeclampsia is associated with lower percent-ages of regulatory T cells in maternal blood. Hy-pertens Pregnancy 2009;28:300-11.34. Huisman MA, Timmer A, Zeinstra M, et al.Matrix-metalloproteinase activity in first trimes-ter placental bed biopsies in further compli-cated and uncomplicated pregnancies. Pla-centa 2004;25:253-8.35. Huisman MA, Timmer B, Stegehuis J, SwartB, Aarnoudse JG, Erwich JH. Vascularization infirst-trimester chorionic villi in complicated anduncomplicated pregnancies. Am J Obstet Gy-necol 2010;202:88.e1-7.36. James WH. The human sex ratio. Part 1: areview of the literature. Hum Biol 1987;59:721-52.37. Juntunen KS, Kvist AP, Kauppila AJ. A shiftfrom a male to a female majority in newbornswith the increasing age of grand multiparouswomen. Hum Reprod 1997;12:2321-3.38. Beckstead JH. A simple technique for pres-ervation of fixation-sensitive antigens in paraf-fin-embedded tissues: addendum. J His-tochem Cytochem 1995;43:345.39. Xie X, He H, Colonna M, Seya T, Takai T,Croy BA. Pathways participating in activation ofmouse uterine natural killer cells during preg-nancy. Biol Reprod 2005;73:510-8.40. Blois SM, Klapp BF, Barrientos G. Decidu-
alization and angiogenesis in early pregnancy:
an Journal of Obstetrics & Gynecology 76.e11


unravelling the functions of DC and NK cells. JReprod Immunol 2011;88:86-92.41. Zenclussen AC, Fest S, Sehmsdorf US, Ha-gen E, Klapp BF, Arck PC. Upregulation of de-cidual P-selectin expression is associated withan increased number of Th1 cell populations inpatients suffering from spontaneous abortions.Cell Immunol 2001;213:94-103.42. Jauniaux E, Burton GJ. Morphological andbiological effects of maternal exposure to to-bacco smoke on the feto-placental unit. EarlyHum Dev 2007;83:699-706.43. Mian MF, Lauzon NM, Stampfli MR, Moss-man KL, Ashkar AA. Impairment of human NK cellcytotoxic activity and cytokine release by cigarettesmoke. J Leukoc Biol 2008;83:774-84.44. Nagamatsu T, Schust DJ. The contribution
of macrophages to normal and pathological
76.e12 American Journal of Obstetrics & Gynecolo

pregnancies. Am J Reprod Immunol 2010;63:460-71.45. Gustafsson C, Mjosberg J, Matussek A, etal. Gene expression profiling of human decidualmacrophages: evidence for immunosuppres-sive phenotype. PLoS One 2008;3:e2078.46. Nahrendorf M, Swirski FK, Aikawa E, et al.The healing myocardium sequentially mobilizestwo monocyte subsets with divergent and com-plementary functions. J Exp Med 2007;204:3037-47.47. Saito S, Nakashima A, Shima T, Ito M. Th1/Th2/Th17 and regulatory T-cell paradigm in preg-nancy. Am J Reprod Immunol 2010;63:601-10.48. Bettelli E, Carrier Y, Gao W, et al. Reciprocaldevelopmental pathways for the generation ofpathogenic effector TH17 and regulatory T
cells. Nature 2006;441:235-8.
gy JULY 2012

49. Korn T, Bettelli E, Oukka M, Kuchroo VK.IL-17 and Th17 Cells. Annu Rev Immunol2009;27:485-517.50. Robertson SA, Seamark RF, Guilbert LJ,Wegmann TG. The role of cytokines in gesta-tion. Crit Rev Immunol 1994;14:239-92.51. Zenclussen AC, Blois S, Stumpo R, et al.Murine abortion is associated with enhancedinterleukin-6 levels at the feto-maternal inter-face. Cytokine 2003;24:150-60.52. Diehl S, Rincon M. The two faces of IL-6 onTh1/Th2 differentiation. Mol Immunol 2002;39:531-6.53. El-Bastawissi AY, Williams MA, Riley DE,Hitti J, Krieger JN. Amniotic fluid interleukin-6and preterm delivery: a review. Obstet Gynecol
2000;95:1056-64.


APPENDIX

SUPPLEMENTAL FIGURE 1Gating strategy lymphocytes

A, Lymphocytes were selected based on forward and side scatter plots. Further T-cell subsets were defined using additional gates. For example,CD3�CD4�Foxp3� Treg cells were gated based on positive staining for antibodies against B, CD3, C, CD4, and D, Foxp3. E, Isotype controls were usedto more precisely define positions of gates.Prins. Smoking and the maternal immune response. Am J Obstet Gynecol 2012.

JULY 2012 American Journal of Obstetrics & Gynecology 76.e13


SUPPLEMENTAL FIGURE 2Gating strategy monocytes

A, Monocytes were selected based on CD11b and F4/80. CD11b-hi and F4/80-intermediate cells were considered monocytes. B, The CD11b-hi andF4/80-intermediate monocytes were gated for CD43 and GR1 to distinguish the GR1-hi and GR1-lo subset.Prins. Smoking and the maternal immune response. Am J Obstet Gynecol 2012.

76.e14 American Journal of Obstetrics & Gynecology JULY 2012

Smoking during pregnancy influences the maternal immune ...

Documents

Transcript of Smoking during pregnancy influences the maternal immune ...