Levels of PAH–DNA adducts in placental tissue and the risk of fetal neural tube defects in a Chinese population

Page 1

Ld

YRa

b

c

d

a

ARRAA

KPDNPF

1

swTcawcncshait1N

UT

0h

Reproductive Toxicology 37 (2013) 70–75

Contents lists available at SciVerse ScienceDirect

Reproductive Toxicology

journa l homepage: www.e lsev ier .com/ locate / reprotox

evels of PAH–DNA adducts in placental tissue and the risk of fetal neural tubeefects in a Chinese population

ue Yuana,b, Lei Jina,b, Linlin Wanga,b, Zhiwen Lia,b, Le Zhanga,b, Huiping Zhuc,ichard H. Finnell c, Guodong Zhoud, Aiguo Rena,b,∗

Department of Epidemiology and Health Statistics, School of Public Health, Peking University, Beijing 100191, ChinaInstitute of Reproductive and Child Health/Ministry of Health Key Laboratory of Reproductive Health, Peking University, Beijing 100191, ChinaDell Pediatric Research Institute, Department of Nutritional Sciences, The University of Texas at Austin, Austin, TX 78723, USAInstitute of Biosciences and Technology, Texas A&M University Health Science Center, Houston, TX 77030, USA

r t i c l e i n f o

rticle history:eceived 4 November 2012eceived in revised form 9 January 2013ccepted 26 January 2013vailable online 15 February 2013

a b s t r a c t

We examined the relationship between PAH–DNA adduct levels in the placental tissue, measured bya highly sensitive 32P-postlabeling assay, and the risk of fetal neural tube defects (NTDs). We furtherexplored the interaction between PAH–DNA adducts and placental PAHs with respect to NTD risk. Placen-tal tissues from 80 NTD-affected pregnancies and 50 uncomplicated normal pregnancies were includedin this case-control study. Levels of PAH–DNA adducts were lower in the NTD group (8.12 per 108

8

eywords:olycyclic aromatic hydrocarbons (PAHs)NA adductseural tube defects (NTDs)lacentaetus

nucleotides) compared to controls (9.92 per 10 nucleotides). PAH–DNA adduct concentrations belowthe median was associated with a 3-fold increased NTD risk. Women with a low PAH–DNA adduct level inconcert with a high placental PAH level resulted in a 10-fold elevated risk of having an NTD-complicatedpregnancy. A low level of placental PAH–DNA adducts was associated with an increased risk of NTDs; thisrisk increased dramatically when a low adduct level was coupled with a high placental PAH concentration.

. Introduction

Neural tube defects (NTDs) are among the most common anderious of all congenital abnormalities. A neural tube is defectivehen it fails to close between 18 and 28 days post-conception [1].

his serious error in the development of the central nervous systeman result in death, or permanent damage to the brain, spinal cord,nd spinal nerves. Common types of NTDs include anencephaly,hich occurs when the cranial region of the neural tube fails to

lose, and spina bifida, which occurs when the caudal portion of theeural tube remains unfused and open [2]. Most infants with anen-ephaly are stillborn or die shortly after birth, while infants withpina bifida may survive yet suffer with life-long disabilities [3]. Itas been estimated that more than 300,000 infants are born withn NTDs each year worldwide. NTDs have higher birth prevalencen middle- and low-income countries [3]. In certain areas of China,

he birth prevalence of NTDs was reported to be as high as 13.9 per000 births [4]. The exact mechanism underlying the etiology ofTDs remains unknown.

∗ Corresponding author at: Institute of Reproductive and Child Health, Pekingniversity Health Science Center, Beijing 100191, China.el.: +86 10 82801140; fax: +86 10 82801141.

E-mail addresses: [email protected], [email protected] (A. Ren).

890-6238/$ – see front matter © 2013 Elsevier Inc. All rights reserved.ttp://dx.doi.org/10.1016/j.reprotox.2013.01.008

© 2013 Elsevier Inc. All rights reserved.

While genetic factors are believed to play a significant rolein the complex etiology of NTDs, it is also known that selectedenvironmental exposures are important contributors to NTD risk.Polycyclic aromatic hydrocarbons (PAHs) are ubiquitous envi-ronmental contaminants. Several derivatives of PAHs have beenshown to be teratogenic in animal models, producing NTDs anda variety of other malformations in the exposed embryos [5].There is also growing evidence from human environmental andoccupational studies on risks of NTDs in association with pre-natal exposure to PAH [6–9]. PAHs can exert genotoxic effectsthrough metabolic activation and subsequent binding to DNA,thus forming bulky PAH–DNA adducts, a widely used indicator ofDNA damage that has been associated with increased cancer risk[10–12]. PAH–DNA adducts reflect individual variation in exposure,absorption, metabolic activation, and DNA repair. They providean informative biomarker of the biologically effective PAH dosefrom all sources, including tobacco smoking, car exhaust, incom-plete coal or biomass combustion from cooking and heating, andoccupational exposures [13]. PAH–DNA adduct levels in umbilicalcord blood or placental tissue have been associated with a varietyof adverse birth outcomes, including aberrant child behavior [14],

neurodevelopment deficits [15], and impaired fetal growth [16,17].

Our previous preliminary study has explored the associationbetween levels of PAH–DNA adducts in maternal venous blood andthe risk of NTDs. We observed statistically significantly elevated

Page 2

ve Tox

Pmi(tetbahw

pmwtatcpc

2

2

tcwwaiciaaftc

2

pweuoTsiommgwbbTatDww2

Y. Yuan et al. / Reproducti

AH blood levels, along with lower PAH–DNA adduct level inothers of NTD cases, when compared with mothers of healthy

nfants [8]. However, the sample size of this study was smallcases/controls = 35/18). In addition, peripheral blood has a shorturnover and may only be appropriate for short-term exposurestimates. In contrast, the placenta is an exchange organ betweenhe mother and the fetus, and may therefore serve as a betteriospecimen for exposure assessment, as organic pollutants mayccumulate in placental tissue over the length of the pregnancy. Weave found that increased PAH concentrations in placental tissueas significantly associated with an elevated risk of NTDs [7].

We hypothesized that elevated levels of PAH–DNA adducts inlacental tissue are associated with increased risks of NTDs, anday interact with PAH concentrations in placental tissue. Hereine describe our efforts at assessing levels of PAH–DNA adducts in

he placentas of the NTD-affected fetuses or newborns (cases) rel-tive to those of the healthy infants (controls). The objectives ofhe present study were to examine the association between pla-ental PAH–DNA adduct level and the risk of NTDs, and to assessossible interaction between PAH–DNA adducts and placental PAHoncentrations on the risk of NTDs.

. Materials and methods

.1. Subjects and sampling

The population-based samples of cases and controls used inhe study have been previously described elsewhere [7]. Briefly,ases were women who had pregnancies complicated by an NTD,hile controls were women who gave birth to healthy term infantsith no congenital malformations. 80 placentas of NTDs cases (36

nencephalic; 44 spina bifida) and 50 placentas of controls werencluded in this study. All subjects were recruited from four ruralounties of Shanxi Province (Pingding, Xiyang, Taigu, and Zezhou)n Northern China from 2005 to 2007. Placentas were collectedt delivery or termination of NTD-affected pregnancies, immedi-tely stored at −20 ◦C, and transferred on dry ice to our laboratoryor processing and analyses. The study protocol was approved byhe institutional review board of Peking University, and informedonsent was obtained from all mothers.

.2. Laboratory analyses

DNA was extracted from placental tissue using a standardrocedure. Approximately 1 g of placental tissue on the fetal sideas taken from an edge of the placenta. In a few cases of very

arly terminations, the sides of the placenta were somewhatnrecognizable; therefore, we sampled the placental tissue fromne side to the opposite side to ensure the fetal side was included.he apparatus used to prepare placental tissue was repeatedlyterilized with an alcohol solution and dried with a paper towelnbetween every placental sample in order to avoid contaminationf the samples. Sampled tissue was thawed, minced, and washedanually with ice-cold PBS (pH 7.0) or saline to remove blood. Theinced placental tissue was washed again followed by centrifu-

ation. The tissue pellet was homogenized in an electric blenderith a solution of PBS and centrifuged again. DNA was extracted

y treating this placental tissue pellet free of maternal and fetallood with a mixture of 2 ml digestion buffer [25 mM EDTA, 10 mMris-Cl (pH 7.4), 10 mM NaCl, 1% SDS, 200 �g/ml proteinase K,nd 20 �g/ml RNase A], overnight at 37 ◦C, followed by depro-einization with phenol/chloroform/isoamyl alcohol (25:24:1).

NA was precipitated with ethanol/NaCl, spooled out, washedith 70% ethanol, and air-dried. DNA concentration and purityere measured spectrophotometrically by absorbance at 260 and

80 nm. The A260/A280 ratio for all samples was between 1.8 and 2.0.

icology 37 (2013) 70–75 71

PAH–DNA adducts were determined by the nuclease P1-enhanced 32P-postlabeling method that was originally developedby Reddy and Randerath [18], and was standardized by Phillipset al. [19,20], with minor modifications. Briefly, aliquots of 5 �gDNA samples were dried over Speedvac and digested by a mixtureof micrococcal endonuclease (MN) and spleen phosphodiesterase(SPD) at 37 ◦C overnight. Nuclease P1 was used for adduct enrich-ment, followed by incubating the mixture with [�-32P]ATP and T4polynucleotide kinase. The labeled DNA adducts were separated bymultidirectional anion-exchange thin layer chromatography (TLC)on 10 cm × 10 cm PEI-cellulose plates. Solvent systems used forTLC were the following: D1: 2.3 M sodium phosphate, pH 5.7; D2:4.05 M lithium formate + 7.65 M urea, pH 3.5; D3: 0.72 M sodiumphosphate, 0.42 M tris, 7.65 M urea, pH 8.2. The radioactivitywas determined by phosphoimager (Typhoon Trio, GE Healthcare)after exposing to a storage phosphor screen (35 cm × 43 cm) atroom temperature for 2 h. DNA adduct levels were quantified andexpressed as adducts per 108 nucleotides.

In order to assess any potential dose–response relationship ofadduct formation and to serve as negative and positive controls,we also measured calf thymus DNA samples, which were treatedwith B[a]P (0 �M, 1 �M, and 5 �M) in vitro. To three differenttubes were added 100 �g DNA sample with 2.0 ml DNA bindingbuffer [150 mM Tris-Cl (pH 7.4), 150 mM KCl, 5 mM MgCl2], fol-lowed by NADPH (100 mM), B[a]P solution (0 �M, 1 �M, and 5 �M)and 2 mg human liver microsomes. After incubation at 37 ◦C for2 h, 50 �l of Tris-Cl (1 M, pH 8.0), 10 �l of EDTA (100 mM), and 2 mgproteinase were added, and the mixture were incubated at 37 ◦Cfor an additional 45 min. Then DNA samples were extracted withthe phenol/chloroform/isoamyl alcohol procedure. Adduct levelsof the B[a]P treated samples were 0.6, 8.2, 36.4 adducts per 108

nucleotides for 0 �M, 1 �M and 5 �M, respectively.PAHs in the placental tissue were analyzed by gas chromato-

graph and mass spectrometry (Agilent 7890A-5975C). The medianconcentration of PAHs in placental tissue was 597 ng/g lipid and392 ng/g lipid in case and control group, respectively, which wasdescribed elsewhere in detail [7].

2.3. Statistical analysis

Maternal demographic information between the NTD cases andthe control group were compared using a chi-square test. Becausethe adduct levels were not normally distributed, the medianwith interquartile range was used to describe the distribution.Differences in adduct levels between case and control groups wereassessed using the Mann-Whitney U test. Tertile of total PAH–DNAadducts in controls was used as cutoff value in the dose–responseanalysis. Differences in proportions of adduct levels presentedbetween different groups were assessed with the Pearson’s �2

test, or Fisher’s exact test if cell expectation was less than 5. ACochran-Armitage test was used for trend analysis.

Risk of NTDs associated with PAH–DNA adduct levels wasestimated from the odds ratio with 95% confidence interval (CI),using an unconditional logistic model. Only two factors – hyper-thermia, which was found to be associated with NTD risk in thepresent population [7], and maternal age, which was one of themost important demographic variables – were adjusted in themultivariate model. Other variables, such as maternal education,occupation, parity, periconceptional folate supplementation, andmaternal passive smoking, were not included in the model becausethey were not found to be statistically significant in our univariateanalyses. We examined the interaction between PAH–DNA adducts

and PAH concentrations in the placental tissue by introducinga multiplicative term of the two variables. Since the interactionterm was significant, we included the term in the multivariateanalyses. Each of the two variables was used for stratification when

Page 3

72 Y. Yuan et al. / Reproductive Toxicology 37 (2013) 70–75

Table 1Select characteristics of women who had pregnancies affected by NTDs (cases) andwomen who delivered healthy infants (controls).

Characteristic Casesa Controlsa P valueb

(n = 78) (n = 50)

Maternal age (y)<25 31 (39) 15 (32)25–29 23 (29) 11 (23)≥30 25 (32) 21 (45) 0.34

Maternal educationPrimary or lower 15 (19) 5 (10)Junior high 54 (68) 41 (82)High school or above 10 (13) 4 (8) 0.23

OccupationNon-farmer 13 (17) 4 (8)Farmer 64 (83) 45 (92) 0.16

Parity1 45 (60) 23 (46)≥2 30 (40) 27 (54) 0.12

Periconceptional folate supplementationNo 71 (91) 41 (87)Yes 7 (9) 6 (13) 0.55

Maternal passive smokingNo 29 (37) 26 (52)Yes 50 (63) 24 (48) 0.087

Hyperthermia during pregnancyNo 50 (67) 46 (92)Yes 25 (33) 4 (8) 0.001

a Data are presented in number (percentage). Total number may not be equal tot

t

araPSP

3

3

a

Table 2Concentration of PAH–DNA adducts (per 108 nucleotides) in the placental tissue ofcases with NTDs and healthy controls.

N Median Interquartile range P valuea

Total NTDs 80 8.12 6.43–9.93 0.033Anencephaly 36 8.37 5.63–10.0 0.059Spina bifida 44 8.10 6.70–9.67 0.083

Controls 50 9.92 6.81–12.3

Fa

he total of cases or controls due to missing or unknown data.b Cases and controls are compared by Pearson’s chi-square test, or Fisher’s exact

est if any cell expectation was less than 5.

ssessing the effect of the other factor on NTD risks, followed byecoding them to a dummy variable, which stands for low PAH/highdducts, low PAH/low adducts, high PAH/high adducts and highAH/low adducts, with cutoff values being the medians of controls.tatistical analyses were conducted using SPSS 18.0. A two-tailedvalue of <0.05 was considered to indicate statistical significance.

. Results

.1. Characteristics

The study cohort consisted of 80 NTD cases (36 anencephalicnd 44 spina bifida) and 50 controls. Distribution of maternal

ig. 1. Profiles of DNA adducts from (A) human placenta and (B) calf thymus DNA treated wfter exposing to a storage phosphor screen at r.t. for 2 h.

a In comparison with the median of controls, Mann-Whiney U test.

characteristics was summarized in Table 1. There were no sig-nificant differences between the two groups with regard tomaternal age, educational level, occupation, parity, folic acid sup-plementation, or passive smoking exposure. A significant higherproportion of case mothers (33%) reported hyperthermia (feveror influenza) during early pregnancy, compared to 8% of controlmothers.

3.2. Levels of PAH–DNA adducts and NTDs

Fig. 1 represents a typical profile of DNA adducts from humanplacenta (Pane A). DNA samples from cases or controls exhibitedqualitatively similar profiles. A positive control sample obtainedfrom calf thymus DNA treated with B[a]P displayed similar patternof adducts (1–3), but not adduct 4 (Fig. 1, panel B). Adducts 1–3were summed up as total PAH–DNA adducts for further analyses.

The median level of adducts in case placentas were 8.12adducts/108 nucleotides, significantly lower than the level of 9.92adducts/108 nucleotides in control placentas (P < 0.05) (Table 2).Subtypes of NTDs, namely anencephaly and spina bifida, alsoshowed lower adduct levels than controls (8.37 adducts/108

nucleotides for anencephaly; 8.10 adducts/108 nucleotides forspina bifida), although the differences did not reach statistical sig-nificance.

The association between lower levels of PAH–DNA adducts withan increased risk of NTDs showed a dose–response relationship(Table 3). When the highest tertile was used as the referent, a1.48-fold (95% CI, 0.59–3.75) and 2.67-fold (95% CI, 1.07–6.64)increases in the risk of NTDs were observed for women whoseplacental concentrations of PAH–DNA adducts were in the second

and lowest tertile, respectively. Trend analysis showed a statisti-cally significant linear trend for the associated risks (Ptrend = 0.029).The dose–response relationship was present for both anencephaly(Ptrend = 0.081) and spina bifida (Ptrend = 0.054) subtypes, although

ith 1 �M B[a]P. Images were from a phosphoimager (Typhoon Trio, GE Healthcare)

Page 4

Y. Yuan et al. / Reproductive Toxicology 37 (2013) 70–75 73

Table 3Tertiles of PAH–DNA adducts in placental tissue and the risk of neural tube defects (NTDs). .

Total NTDs Anencephaly Spina bifida Controls

N Odds ratio N Odds ratio N Odds ratio N

1st tertile 40 2.67 (1.07–6.64) 18 2.57 (0.84–7.84) 22 2.75 (0.95–7.98) 162nd tertile 25 1.48 (0.59–3.75) 11 1.40 (0.44–4.47) 14 1.56 (0.52–4.67) 18

1.00.08

bl

3l

wPO1thPPaotatataht(9t

TA

3rd tertile 15 1.0 (ref.) 7P for trend 0.029

oth trends did not reach statistical significance, apparently due toimited subgroup sample sizes.

.3. Interaction between PAH–DNA adducts and placental PAHevels

Reduced PAH–DNA adduct levels were significantly associatedith increased NTD risks for total NTDs (OR = 3.2, 95% CI 1.4–7.2,< 0.01), as well as the individual subtypes (for anencephaly,R = 3.2, 95% CI 1.2–8.4, P < 0.05; for spina bifida, OR = 3.8, 95% CI.4–9.8, P < 0.01), when the effect of PAH concentration in placen-al tissue was adjusted (Table 4). Controlling for maternal age andyperthermia did not substantially alter the relationship betweenAH–DNA adducts and risk of NTDs (OR = 3.0, 95% CI 1.3–7.0,< 0.05). This relationship was not statistically significant for thenencephaly subtype (OR = 2.5, 95% CI 0.87–7.0, P = 0.09) becausef the attenuated analysis power. When stratified by PAH concen-ration in placental tissue, association between PAH–DNA adductsnd risk of NTDs was only seen in the higher placental PAH concen-ration subgroup. When stratified by PAH–DNA adducts levels, thessociation between placental PAH and NTDs was only observed inhe lower adduct subgroup. When compared to those who had highdduct but low placental PAH levels, only those with low adduct butigh placental PAH levels exhibited significantly elevated risk for

otal NTDs (OR = 10.0, 95% CI 2.8–35.6, P < 0.001), for anencephalyOR = 10.5, 95% CI 1.9–57.2, P < 0.01), and for spina bifida (OR = 9.7,5% CI 2.2–42.3, P < 0.01). Adjustment for maternal age and hyper-hermia did not change the associations substantially.

able 4ssociations of placental PAH–DNA adducts, PAH, and possible joint effect with the risk o

Crude odds ratio (95% CI)

Total NTDs Anencephaly Spi

Both PAH–DNA adducts and PAH were included in the regression modelPAHb 4.6 (2.0–10.5)*** 4.8 (1.7–13.7)** 4.9PAH–DNA adductsc 3.2 (1.4–7.2)** 3.2 (1.2–8.4)* 3.8

Stratified by PAHLow PAH

PAH–DNA adductsc 1.3 (0.35–5.1) 1.7 (0.27–10.3) 1.1High PAH

PAH–DNA adductsc 5.0 (1.9–13.4)** 3.9 (1.3–12.3)* 6.0Stratified by PAH–DNA adductsLow adducts

PAHb 7.5 (2.6–21.4)*** 6.3 (1.8–22.2)** 8.7High adducts

PAHb 2.0 (0.55–7.3) 2.7 (0.47–15.3) 1.6Combined PAH–DNA adducts and PAH

High adducts and low PAH 1.0 (reference)Low adducts and low PAH 1.3 (0.35–5.1) 1.7 (0.27–10.3) 1.1High adducts and high PAH 2.0 (0.55–7.3) 2.7 (0.47–15.3) 1.6Low adducts and high PAH 10.0 (2.8–35.6)*** 10.5 (1.9–57.2)** 9.7

* P < 0.05, significance of regression coefficient, logistic regression.** P < 0.01, significance of regression coefficient, logistic regression.

*** P < 0.001, significance of regression coefficient, logistic regression.a Adjusted for hyperthermia during pregnancy and maternal age (continuous).b In comparison with PAH concentration below median of control placentas.c In comparison with PAH–DNA adducts concentration above median of control placen

(ref.) 8 1.0 (ref.) 161 0.054

4. Discussion

This study examined the association between placental levels ofPAH–DNA adducts and NTD risk in a Chinese population in whichexposure to coal-burning emissions is prevalent [21]. Contrary toour hypothesis, we found that PAH–DNA adduct level in the placen-tal tissue was inversely associated with the risk of NTDs, i.e., highlevel of PAH–DNA adducts is associated with a reduced risk. Theassociation showed a dose–response pattern. On the other hand,women with a low level of PAH–DNA adducts along with a highPAH level in placental tissue had a 10-fold elevated risk of havinga pregnancy affected by NTDs when compared with those with ahigh level of adducts but a low level of placental PAHs.

The finding that a low PAH–DNA adduct level is associatedwith an elevated NTD risk is consistent with our previous study,in which we found that the blood PAH–DNA adduct level waslower in NTDs case mothers than in control mothers [8]. In anotherstudy from this same region of China, we found that the ratio ofplacental DDT metabolites to parent compound, as indicated by(p,p′-DDE + p,p′-DDD)/p,p′-DDT, was lower in the case group thanin the control group [7]. These findings are indicative of a slowerelimination of xenobiotics from the maternal body in the casegroup, which may indicate a more persistent fetal exposure toPAHs or metabolites, resulting in increased risk of birth defects. The

preponderance of available toxicity data showed potential healthrisk secondary to retained xenobiotics and persistent exposure indeveloping systems. Thus, there is an ever-increasing discussionunderway about possible interventions to facilitate elimination of

f neural tube defects (NTDs).

Adjusted odds ratio (95% CI)a

na bifida Total NTDs Anencephaly Spina bifida

(1.8–13.3)** 4.2 (1.7–10.3)** 4.2 (1.4–13.1)* 4.2 (1.4–12.4)**

(1.4–9.8)** 3.0 (1.3–7.0)* 2.5 (0.87–7.0) 4.3 (1.5–12.7)**

(0.22–5.7) 1.4 (0.28–7.3) 3.2 (0.30–35.0) 1.1 (0.14–7.8)

(2.0–19.6)** 4.5 (1.6–12.6)** 3.1 (0.89–10.9) 7.3 (2.0–26.3)**

(2.5–30.1)** 6.5 (2.0–21.2)** 4.3 (0.94–19.5) 7.8 (1.9–31.0)**

(0.32–7.5) 1.6 (0.37–6.5) 1.6 (0.23–11.5) 1.2 (0.21–6.7)

1.0 (reference)(0.22–5.7) 1.2 (0.27–5.5) 1.7 (0.22–12.8) 1.1 (0.17–7.4)(0.32–7.5) 1.8 (0.44–7.7) 3.0 (0.44–20.4) 1.2 (0.22–7.0)(2.2–42.3)** 8.2 (2.1–32.7)** 8.5 (1.4–52.3)* 8.9 (1.7–46.0)*

tas.

Page 5

7 ive Tox

po

btFtggeifirtgcfipeAiciilclbtt

ibiwharleesAv[tsooiasm

Nobcehnttlo

4 Y. Yuan et al. / Reproduct

ersistent toxicants from the humans to diminish the body burdenf these compounds [22].

Placental levels of PAH–DNA adducts and PAHs are determinedy both maternal exposure and the metabolic rate. A two-step bio-ransformation pathway is involved in the metabolism of PAHs.irst, the parent compound is generally oxidized by Phase I enzymeshrough hydroxylation catalyzed by cytochrome P450 monooxy-enase enzymes. Subsequently, the intermediates conjugate tolucuronide, glutathione, or other moieties, catalyzed by Phase IInzymes (e.g., glutathione S-transferase). These reactions greatlyncrease the water solubility of the transformed compound, andacilitate their excretion [23]. As our cases and controls have sim-lar socioeconomic status and living conditions (all of them fromural areas of Shanxi Province), it is reasonable to assume thathey shared similar routes of PAH exposures. Therefore, individualenetic background and metabolic rate of PAHs may play a criti-al role in the pollutant metabolism and excretion. Based on ournding of a lower level of PAH–DNA adducts in the placenta, weostulate a slower metabolic rate may present in NTD case moth-rs, resulting in a prolonged exposure of the fetus to xenobiotics.low PAH–DNA adduct level in placental tissue may also indicate

ncreased PAH passage from mother to fetus. If true, elevated pla-ental DNA adducts may reflect the protective effect of the placentan lowering the risk of NTDs during pregnancy. In a study conductedn Czech Republic, mothers exposed to tobacco smoke had higherevels of DNA adducts in placenta but lower levels of DNA adducts inord blood, while mothers unexposed to tobacco smoke had lowerevels of DNA adducts in placenta but higher DNA adducts in cordlood [24]. Whether higher placental DNA adducts mean a protec-ive effect to the fetus needs to be more fully examined in studieshat will require the use of animal models.

The interaction between PAH–DNA adducts and placental PAHss relatively strong; women bearing a lower PAH–DNA adduct levelut a higher PAH level in the placental tissue were at a remarkably

ncreased risk for NTDs. The association between PAH exposureith elevated risk of NTDs and other congenital malformationsas been reported in several epidemiological studies [6–9,25,26],s well as in animal experiments [5,27]. Few prior studies haveeported any association between congenital malformations withevel of PAH–DNA adducts, which represent DNA damage fromxposure to PAHs. Fetal toxicity from PAHs may include anti-strogenic effects [28], P450 enzymes induction [29], oxidativetress [30], or direct effects of co-existing carbon monoxide [31].nimal studies have demonstrated that NTDs induced by in uteroalproic acid exposure was mediated by oxidative DNA damage32]. Another study measuring cord blood DNA methylation foundhat PAH exposure alters global methylation [33]. The results of ourtudy suggest that PAHs may not exert their effect via the formationf PAH–DNA adducts with regard to embryonic neural tube devel-pment. The combined effect of PAH–DNA adducts and PAH leveln the placental tissue suggests a possible non-genotoxic mech-nism. Excess retained xenobiotics and slower metabolism has aynergistic-like effect, which may compromise normal develop-ent, particularly of the neural tube.However, the higher PAH–DNA adducts with reduced risk of

TD does not indicate a protective role with respect to fetal devel-pment. Evidence has shown that prenatal PAH exposure measuredy PAH–DNA adduct formation in umbilical cord blood was asso-iated with impairment in fetal and child development [34],specially, neurobehavioral development [15]. PAH–DNA adductsave also been reported to associate with somatic mutations inewborns [35]. Although the individuals bearing higher concentra-

ions of PAH–DNA adducts and higher levels of PAHs in the placentalissue did not show an increased the risk of NTDs, better surveil-ance efforts are required in order to identify other deleteriousutcomes.

icology 37 (2013) 70–75

We did not find any statistically significant correlation betweenplacental levels of PAH–DNA adducts and placental concentrationsof PAHs. To our knowledge, no prior study has simultaneouslyreported PAH–DNA adducts and PAHs levels in placental tissue.Although it is undeniable that exposure to PAHs will inducePAH–DNA adducts, the relationship between these two biomark-ers in the same tissue was not clear, as complex biological andbiochemical factors are involved in the process of their formation.Even adducts themselves were not significantly correlated in differ-ent tissues. For example, PAH–DNA adduct levels in lung tissue donot correlate with adduct levels in white blood cells [36,37]. In ourstudy, PAH–DNA adducts were quantified in genomic DNA fromplacental tissue, while PAHs were measured in lipids separatedfrom placental tissue. Despite the same tissue of origin, hetero-geneity exists within its different constituents. Other explanationsfor the lack of correlation between PAH–DNA adducts and PAH lev-els in the placental tissue includes the fact that inhibitory effectof non-carcinogenic PAH mixtures may alter metabolism of car-cinogenic PAHs and reduce the formation of DNA adducts [38,39].The study area has some of the largest known concentrations ofPAHs emissions from industrial sources, coupled with the fact thatcoal is commonly used indoors for cooking and heating, combiningfor extensive potential for population exposure [40]. We did notfind a relationship between ambient PAH exposure measured byan index of indoor air pollution from coal combustion (IAPCC) [9]and placental PAH–DNA adducts. A previous study reported thattotal placental DNA adducts in mothers exposed to tobacco smokewere higher compared with unexposed mothers [24], along withinconsistent findings that placental PAH–DNA adducts were notcorrelated with urinary nicotine, cotinine or hours of passive smokeexposure [41].

This study investigated placental PAH–DNA adducts and theassociated risk of NTDs. These data extended our understandingof the etiology of some instances of NTDs, yet the hypothesisthat lower PAH–DNA adducts in the placenta represent a slowermetabolic rate, remains to be tested. A drawback of this study isabout the time window – the development of NTDs occurs early ingestation, yet the placentas were sampled later in pregnancy. Thedevelopment of the placenta from trophoblast cells has its continu-ity, and biomarkers present in the placenta may be more relevant tofetal development, as compared to maternal blood or cord blood. Inthe present study, the placental tissues were sampled at differentgestational stages among NTD cases and controls; this differencemay be a source of error that could not be eliminated with routinecase-control design, as the placentas of healthy controls can only besampled at term. Therefore, further studies are needed to confirmthe association between PAH–DNA adducts and the risk of NTDs.

5. Conclusion

In summary, a low level of PAH–DNA adducts in placental tis-sue was found to be associated with an increased risk of NTDs, withan especially high risk of NTDs being observed among those with alower level of PAH–DNA adducts coupled with a higher level of pla-cental PAH. Further studies are needed to replicate these findingsand to reveal mechanisms underlying these observations.

Conflict of interest statement

The authors declare they have no competing financial or non-financial interests.

Role of the funding sources

The funding agencies have no role in study design, implemen-tation, data analysis, and interpretation.

Page 6

ve Tox

A

NRDaw

R

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[Environmental Science and Technology 2006;40:702–8.

Y. Yuan et al. / Reproducti

cknowledgments

This study was supported in part by a grant from The Nationalatural Science Foundation of China (Grant No. 31071315) to Dr.en. Additional support was provided by NIH grant (ES021006) tors. Finnell and Zhu. The authors would like to thank the maternalnd child health care workers in the four counties for their helpith subject recruitment and data collection.

eferences

[1] Botto LD, Moore CA, Khoury MJ, Erickson JD. Neural-tube defects. The NewEngland Journal of Medicine 1999;341:1509–19.

[2] Lazareff JA. Neural Tube Defects. Singapore: World Scientific Pub Co. Inc.; 2009.[3] Christianson AL, Howson CP, Modell B. Global Report on Birth Defects: The

hidden toll of dying and disabled children. March of Dimes Birth Defects Foun-dation; 2006.

[4] Li Z, Ren A, Zhang L, Ye R, Li S, Zheng J, et al. Extremely high prevalence ofneural tube defects in a 4-county area in Shanxi Province, China. Birth DefectsResearch Part A, Clinical and Molecular Teratology 2006;76:237–40.

[5] Barbieri O, Ognio E, Rossi O, Astigiano S, Rossi L. Embryotoxicity ofbenzo(a)pyrene and some of its synthetic derivatives in Swiss mice. CancerResearch 1986;46:94–8.

[6] Langlois PH, Hoyt AT, Lupo PJ, Lawson CC, Waters MA, Desrosiers TA, et al.Maternal occupational exposure to polycyclic aromatic hydrocarbons and riskof neural tube defect-affected pregnancies. Birth Defects Research Part A, Clin-ical and Molecular Teratology 2012;94:693–700.

[7] Ren A, Qiu X, Jin L, Ma J, Li Z, Zhang L, et al. Association of selected persis-tent organic pollutants in the placenta with the risk of neural tube defects.Proceedings of the National Academy of Sciences of the United States of America2011;108:12770–5.

[8] Naufal Z, Li ZW, Zhu L, Zhou GD, McDonald T, He LY, et al. Biomarkersof exposure to combustion by-products in a human population in Shanxi,China. Journal of Exposure Science and Environmental Epidemiology 2010;20:310–9.

[9] Li Z, Zhang L, Ye R, Pei L, Liu J, Zheng X, et al. Indoor air pollution from coalcombustion and the risk of neural tube defects in a rural population in ShanxiProvince, China. American Journal of Epidemiology 2011;174:451–8.

10] Sagiv SK, Gaudet MM, Eng SM, Abrahamson PE, Shantakumar S, TeitelbaumSL, et al. Polycyclic aromatic hydrocarbon–DNA adducts and survival amongwomen with breast cancer. Environmental Research 2009;109:287–91.

11] Castano-Vinyals G, Talaska G, Rothman N, Alguacil J, Garcia-Closas M, Dose-meci M, et al. Bulky DNA adduct formation and risk of bladder cancer. CancerEpidemiology, Biomarkers and Prevention 2007;16:2155–9.

12] Rybicki BA, Rundle A, Savera AT, Sankey SS, Tang D. Polycyclic aro-matic hydrocarbon–DNA adducts in prostate cancer. Cancer Research2004;64:8854–9.

13] Vineis P, Perera F. DNA adducts as markers of exposure to carcinogens and riskof cancer. International Journal of Cancer 2000;88:325–8.

14] Perera FP, Wang S, Vishnevetsky J, Zhang B, Cole KJ, Tang D, et al. Polycyclicaromatic hydrocarbons–aromatic DNA adducts in cord blood and behav-ior scores in New York city children. Environmental Health Perspectives2011;119:1176–81.

15] Perera F, Li TY, Zhou ZJ, Yuan T, Chen YH, Qu L, et al. Benefits of reducing prenatalexposure to coal-burning pollutants to children’s neurodevelopment in China.Environmental Health Perspectives 2008;116:1396–400.

16] Perera FP, Tang D, Rauh V, Lester K, Tsai WY, Tu YH, et al. Relationshipsamong polycyclic aromatic hydrocarbon–DNA adducts, proximity to the WorldTrade Center, and effects on fetal growth. Environmental Health Perspectives2005;113:1062–7.

17] Sram RJ, Binkova B, Dejmek J, Chvatalova I, Solansky I, Topinka J. Associa-tion of DNA adducts and genotypes with birth weight. Mutation Research2006;608:121–8.

18] Reddy MV, Randerath K. Nuclease P1-mediated enhancement of sensitivityof 32P-postlabeling test for structurally diverse DNA adducts. Carcinogenesis1986;7:1543–51.

19] Phillips DH, Castegnaro M. Standardization and validation of DNA adductpostlabelling methods: report of interlaboratory trials and production of rec-ommended protocols. Mutagenesis 1999;14:301–15.

20] Phillips DH, Arlt VM. The 32P-postlabeling assay for DNA adducts. Nature Pro-tocols 2007;2:2772–81.

[

icology 37 (2013) 70–75 75

21] Liao Y, Wang J, Wu J, Driskell L, Wang W, Zhang T, et al. Spatial analysis of neuraltube defects in a rural coal mining area. International Journal of EnvironmentalHealth Research 2010;20:439–50.

22] Genuis SJ. Elimination of persistent toxicants from the human body. Humanand Experimental Toxicology 2011;30:3–18.

23] Shimada T. Xenobiotic-metabolizing enzymes involved in activation and detox-ification of carcinogenic polycyclic aromatic hydrocarbons. Drug Metabolismand Pharmacokinetics 2006;21:257–76.

24] Topinka J, Milcova A, Libalova H, Novakova Z, Rossner Jr P, Balascak I, et al.Biomarkers of exposure to tobacco smoke and environmental pollutants inmothers and their transplacental transfer to the foetus. Part I: Bulky DNAadducts. Mutation Research 2009;669:13–9.

25] Lupo PJ, Symanski E, Langlois PH, Lawson CC, Malik S, Gilboa SM, et al.Maternal occupational exposure to polycyclic aromatic hydrocarbons andcongenital heart defects among offspring in the national birth defects preven-tion study. Birth Defects Research Part A, Clinical and Molecular Teratology2012;94:875–81.

26] Lupo PJ, Langlois PH, Reefhuis J, Lawson CC, Symanski E, Desrosiers TA, et al.Maternal occupational exposure to polycyclic aromatic hydrocarbons: effectson gastroschisis among offspring in the National Birth Defects PreventionStudy. Environmental Health Perspectives 2012;120:910–5.

27] Incardona JP, Collier TK, Scholz NL. Defects in cardiac function precedemorphological abnormalities in fish embryos exposed to polycyclic aro-matic hydrocarbons. Toxicology and Applied Pharmacology 2004;196:191–205.

28] Santodonato J. Review of the estrogenic and antiestrogenic activity of poly-cyclic aromatic hydrocarbons: Relationship to carcinogenicity. Chemosphere1997;34:835–48.

29] Nebert DW, Dalton TP. The role of cytochrome P450 enzymes in endogenoussignalling pathways and environmental carcinogenesis. Nature Reviews Cancer2006;6:947–60.

30] Rossner Jr P, Milcova A, Libalova H, Novakova Z, Topinka J, Balascak I, et al.Biomarkers of exposure to tobacco smoke and environmental pollutants inmothers and their transplacental transfer to the foetus. Part II. Oxidative dam-age. Mutation Research 2009;669:20–6.

31] Viau C, Hakizimana G, Bouchard M. Indoor exposure to polycyclic aro-matic hydrocarbons and carbon monoxide in traditional houses in Burundi.International Archives of Occupational and Environmental Health 2000;73:331–8.

32] Tung EW, Winn LM. Valproic acid increases formation of reactive oxygenspecies and induces apoptosis in postimplantation embryos: a role for oxidativestress in valproic acid-induced neural tube defects. Molecular Pharmacology2011;80:979–87.

33] Herbstman JB, Tang D, Zhu D, Qu L, Sjodin A, Li Z, et al. Prenatal expo-sure to polycyclic aromatic hydrocarbons. Benzo[a]pyrene–DNA adducts andgenomic DNA methylation in cord blood. Environmental Health Perspectives2012;120:733–8.

34] Tang D, Li TY, Liu JJ, Chen YH, Qu L, Perera F. PAH–DNA adducts in cord bloodand fetal and child development in a Chinese cohort. Environmental HealthPerspectives 2006;114:1297–300.

35] Perera F, Hemminki K, Jedrychowski W, Whyatt R, Campbell U, Hsu Y, et al. Inutero DNA damage from environmental pollution is associated with somaticgene mutation in newborns. Cancer Epidemiology, Biomarkers and Prevention2002;11:1134–7.

36] Weston A, Bowman ED, Shields PG, Trivers GE, Poirier MC, Santella RM, et al.Detection of polycyclic aromatic hydrocarbon–DNA adducts in human lung.Environmental Health Perspectives 1993;99:257–9.

37] Vanschooten FJ, Hillebrand MJX, Vanleeuwen EE, Vanzandwijk N, Jansen HM,Denengelse L, et al. Polycyclic aromatic hydrocarbon DNA adducts in whiteblood-cells from lung-cancer patients – no correlation with adduct levels inlung. Carcinogenesis 1992;13:987–93.

38] Agency for Toxic Substances and Disease Registry., Toxicological profile forpolycyclic aromatic hydrocarbons (PAHs). 1995 [cited 2012 November 26].Available from: http://www.atsdr.cdc.gov/toxprofiles/tp69.pdf

39] Rice JE, Hosted Jr TJ, Lavoie EJ. Fluoranthene and pyrene enhancebenzo[a]pyrene–DNA adduct formation in vivo in mouse skin. Cancer Letters1984;24:327–33.

40] Xu S, Liu W, Tao S. Emission of polycyclic aromatic hydrocarbons in China.

41] Sanyal MK, Mercan D, Belanger K, Santella RM. DNA adducts in human pla-centa exposed to ambient environment and passive cigarette smoke duringpregnancy. Birth Defects Research Part A, Clinical and Molecular Teratology2007;79:289–94.