Adverse reproductive outcomes from exposure to environmental mutagens

Ž .Mutation Research 428 1999 203–215www.elsevier.comrlocatermolmut

Community address: www.elsevier.comrlocatermutres

Adverse reproductive outcomes from exposure to environmentalmutagens

ˇ a,) a b c aRadim J. Sram , Blanka Binkova , Pavel Rossner , Jirı Rubes , Jan Topinka ,´ ´ ¨ ˇ́ ˇJan Dejmek a

a Laboratory of Genetic Ecotoxicology, Regional Institute of Hygiene of Central Bohemia, cro Institute of Experimental Medicine,Academy of Sciences of the Czech Republic, Vıdenska 1083, 142 20 Prague 4, Czech Republic´ ´

b National Institute of Public Health, Prague, Czech Republicc Veterinary Research Institute, Brno, Czech Republic

Received 29 November 1998; accepted 20 February 1999

Abstract

The effect of environmental pollution on reproductive outcomes has been studied in the research project ‘TepliceŽ .Program’ analyzing the impact of air pollution on human health. Genotoxicity of urban air particles -10 mm PM10 in in

vitro system was determined by the analysis of DNA adducts. The highest DNA binding activity was observed in aromaticŽ .fraction, identifying DNA adducts of carcinogenic polycyclic aromatic hydrocarbons PAHs presumably diolepoxide-de-

w x Ž w x . w x w x Ž .rived from: 9-hydroxybenzo a pyrene 9-OH-B a P , benzo a pyrene-r-7,-dihydrodiol-t-9,10-epoxide q anti-BPDE ,w x Ž w x . Ž . w x Ž w x . w x Ž w x .benzo b fluoranthene B b F , chrysene CHRY , benz a antracene B a A , indeno 1,2,3-cd pyrene I cd P . Reproductive

studies were conducted in both females and males. A study of the effects of PM10 exposure on pregnancy outcomes foundŽ . 3the relationship between the intrauterine growth retardation IUGR and PM10 levels over 40 mgrm in the first gestational

Ž 3 3 .month Odds Ratio for 40–50 mgrm s1.6, )50 mgrm s1.9 . Selected biomarkers were analyzed in venous blood, cordŽ . Žblood chromosomal aberrations, comet assay and placenta DNA adducts, genetic polymorphisms of GSTM1 and NAT2

.genotypes of women enrolled in a nested case-control study. DNA adduct levels were higher in polluted vs. controldistricts, in smoking vs. nonsmoking mothers, and in GSTM1 null genotype, which was more pronounced in polluteddistrict. No effect of air pollution was observed by cytogenetic analysis of chromosomal aberrations or by comet assay. The

w x w x w x w xAbbreÕiations: AB.C., aberrant cells; Anti-BPDE, benzo a pyrene-r-7,t-8-dihydrodiol-t-9,10-epoxide " ; B a A, benz a anthracene;w x w x w x w x w x w x w x w xB ab F, B aj F, B k F, benzo b,j,k fluoranthenes; B ghi P, benzo ghi perylene; B a P, benzo a pyrene; CHEST, chick embryotoxicity

w x w xscreening test; DB ah A, dibenz ah anthracene; ED50, dose inducing in 50% of exposed embryos malformation andror death; FISH,fluorescence in situ hybridization; GC-MS, gas chromatography-mass spectrometry; GSTM1, glutathione-S-transferase M1; HiVol air

w x w xsampler, high volume air sampler; HPLC, high pressure liquid chromatography; 9-OH-B a P, 9-hydroxybenzo a pyrene; IUGR, intrauterinew x w xgrowth retardation; I cd P, indeno 1,2,3-cd pyrene; LBW, low birth weight; NAT2, N-acetyltransferase 2; NO , nitrogen oxides; O.R.,x

Odds Ratio; PAHs, polycyclic aromatic hydrocarbons; PCR, polymerase chain reaction; PM2.5, air particles -2.5 mm; PM10, air particles-10 mm; RFLP, restriction fragment length polymorphism; SO , sulfur dioxide2

) Corresponding author. Tel.: q42-02-475-2596; Fax: q42-02-475-2785; E-mail: [email protected]

0027-5107r99r$ - see front matter q 1999 Elsevier Science B.V. All rights reserved.Ž .PII: S1383-5742 99 00048-4

ˇ ( )R.J. Sram et al.rMutation Research 428 1999 203–215´204

reproductive development of young men was followed by measures of semen quality, adjusted for ambient SO exposure.2

The analysis identified significant associations with air pollution for -13% morphologically normal sperm, -29% spermwith normal head shape, -24% motile sperm. Analysis of aneuploidy in human sperm by FISH showed, aneuploidy YY8was associated with season of heaviest air pollution. These findings are suggestive for an influence of air pollution on YY8disomy. All these results indicate that air pollution may increase DNA damage in human population, which may be even

Ž . Ž .higher for susceptible groups. Biomarkers of exposure DNA adducts and susceptibility GSTM1 and NAT2 may indicatethe risk of presumable low environmental exposure. Pregnancy outcome and semen studies imply that relatively low air

Ž 3.pollution higher than 40 mg PM10rm can significantly increase the adverse reproductive outcomes affecting bothgenders. q 1999 Elsevier Science B.V. All rights reserved.

Keywords: Air pollution; Pregnancy outcome; Sperm morphology; PM10; Polycyclic aromatic hydrocarbon; DNA adduct; GSTM1 andNAT2 genotypes

1. Introduction

The Northern Bohemia brown coal basin wasperceived as one of the worst environmentally pol-luted regions in Europe. Conifers in Krusne Horyˇ ´Ž .Ore Mountains forming the northern border of thisregion have been essentially destroyed. This processstarted more than 20 years ago. As first conse-quences of environmental pollution on the healthwere remarkable increase of allergies, immunodefi-ciencies and respiratory diseases in children. Unfa-vorable effect of environment on pregnant womenwas understood as the reason for the increase of birth

defects and higher number of children with low birthweight. Exploratory analysis of data collected priorto 1999 suggested a higher incidence of cancer andreproductive and behavioral effects in this regionw x1,2 .

The Northern Bohemia brown coal basin com-prises several mining districts. The coal in this re-gion is of low quality with high sulfur content, and isfrom surface-mines with open pits. It is primarilyused to produce steam and power for this heavyindustrialized region. Coal-fire power plants pro-duced 50% of the electricity used in the CzechRepublic. After major political changes in the Czech

Fig. 1. Czech Republic, Districts of Teplice and Prachatice.

ˇ ( )R.J. Sram et al.rMutation Research 428 1999 203–215´ 205

Republic in November 1989, a new research pro-gram, the Teplice Program, was developed to evalu-ate the short-term and long-term health impact of airpollution on the population. Teplice, one of themining districts in Northern Bohemia, was desig-nated as a model district for investigation of thehealth effects of air pollution. The district of Prachat-ice, an agricultural area without heavy industry, was

Ž . w xselected as a control district Fig. 1 3 .The Teplice Program was initiated by the Czech

Ministry of Environment in 1990 to provide scientif-ically valid information needed to assess environ-mental health problems in the Northern Bohemiabasin area. In collaboration with the U.S. Environ-

Ž .mental Protection Agency US EPA , the programstarted in 1991. Simultaneously, this program wasincorporated into PHARE II, as ECrHEAr18-CZproject ‘Impact of Environmental Pollution on the

Ž .Health of Population Teplice Program ’. This pro-gram has succeeded in bringing together many dif-ferent research organization and government labora-tories in the Czech Republic, United States and ECcountries to accomplish the multidisciplinary pro-gram.

The hypothesis in the Teplice Program has beenthat the air pollution in the Teplice district adverselyaffects the health of population. The principal objec-tive of this program was to assess human exposure totoxic air pollutants, to relate ambient concentrationsof pollutants to health risks. First studies indicatedhigher level of carcinogenic PAHs, responsible for

w xan increased mutagenicity of ambient air 4 as wellw xas DNA adducts in women working outdoor 5,6 .

The Teplice Program was originally composed from25 different projects. We review here the results ofadverse reproductive outcomes from exposure to en-vironmental mutagens.

2. Air pollution concentrations and sourcecontributions

Air pollution monitoring was carried out continu-ously during this study at both regions, in Tepliceand Prachatice cities. The daily concentrations of thefollowing pollutants were collected: sulfur dioxideŽ . Ž .SO , nitrogen oxides NO , particles -10 mm2 xŽ . Ž .PM10 , particles -2.5 mm PM2.5 , polycyclic

Ž .aromatic hydrocarbons PAHs , toxic metals, sulfatesw xand crustal elements 7 . These data were used to

characterize the ambient exposures of population toair pollutants for the health studies.

The ambient monitoring and source characteriza-tion data were used in receptor modeling calcula-tions, the results of which indicate that residentialspace heating and power plant emissions accountedfor most of respirable particle mass concentrations.The source of pollution from residential heating wastwo to three times more significant than from powerplants. Mobile sources account for the total mass ofPM2.5 more in summer than in winter. Stevens et al.w x8 concluded: ‘‘ . . . The extremely high concentra-

Žtions of sulfur dioxide and particulate matter espe-.cially respirable particles measured in the Czech

Republic in these studies may pose a health risk topeople living in the region.’’

During the last eight years, the concentrations ofSO were substantially decreased, and they are now2

approximately 20% of those seen in the winter 1989.Surprisingly, the concentrations of PM did not10

Fig. 2. Seasonal concentrations of PM10 in the districts of Tepliceand Prachatice, period 1993–1998.


Žsubstantially change in the period 1994–1998 Fig..2 .

3. Genotoxicity and embryotoxicity of urban airparticulate matter

In order to assess a possible health risk associatedwith a complex mixture of hundreds of organiccompounds adsorbed to air particles we used thebiomarker-directed fractionated procedure to evalu-ate biological activities of different chemical com-

w xpound classes 9 . Urban air particles PM10 werecollected using an Anderson Hi-Vol air sampler in

Žboth Teplice and Prachatice during the winter Oc-. Ž .tober–March and summer April–September sea-

sons in the years 1993–1994. Particulate organicmatters were extracted by dichloromethane and crudeextracts were fractionated according to the procedure

w xof Lewtas et al. 10 by acid-base partitioning intoorganic bases, acids and neutral fractions. The neu-tral organic fractions were subsequently fractionatedinto five subfractions according to increasing polar-ity by silica gel column chromatography.

The principal aim of this study was to comparethe DNA binding activities of the different classes ofcompounds using an in vitro acellular assay coupledwith 32 P-postlabeling and an embryotoxicity assay

Ž .using Chick Embryotoxicity Screening Test CHESTw x 3211 . DNA adducts were determined by P-post-labeling analysis using nuclease P1 and butanol ex-traction procedures. Qualitative measurement of32 P-postlabeled DNA adducts was determined by

Ž .high pressure liquid chromatography HPLC toidentify some of the major DNA adducts originatedfrom the in vitro assay. Using CHEST, embryotoxic-ity was defined as the sum of dead and malformed

Ž . Žembryos and expressed calculated as ED 50 a doseinducing in 50% of exposed embryos malformation

. w xandror death 11 .Both the assays evaluating DNA binding activities

and embryotoxicity showed that the different organiccompound classes behaved qualitatively similarly.For both assays, the highest toxicity was observedfor the neutral fractions from which the aromaticsubfractions were the most toxic for both the locali-ties and seasons. These subfractions contained mainlyPAHs and their methyl-derivatives as was confirmed

by GC-MS analysis. These results are in agreementwith the other studies showing that PAHs accountfor most of the mutagenic activity of neutral fraction

w xof urban air 12,13 . A good correlation betweenDNA adduct levels formed in the presence of the S9metabolic activation system and the ED 50 wasobserved for all of the different complex mixtures of

Ž .organic compounds tested rs0.773, p-0.001 .The major adduct spots were analyzed by HPLC

analysis of 32 P-labeled DNA adducts. DNA adductmaps and HPLC profiles were similar for the sam-ples from both the districts and seasons. The majorDNA adducts resulting from the crude extracts wereidentical to those derived from the aromatic frac-tions. We tentatively identified presumably diolepox-

w x Žide-derived adducts from: 9-OH-benzo a pyrene 9-w x . w xOH-B a P , benzo a pyrene-r-7,t-8-dihydrodiol-t-

w x Ž . w x9,10-epoxide q anti-BPDE , benzo b fluorantheneŽ w x . w x Ž w x . w xB b F , benzo j fluoranthene B j F , benzo k fluo-

Ž w x . Ž . w xranthene B k F , chrysene CHRY , benzo a anthra-Ž w x . w x Ž w x .cene B a A , and indeno 1,2,3-cd pyrene I cd P .

The results confirmed the similarities of the majorubiquitous emission sources of organic compoundsin both districts, which are presumably residentialhome heating and motor vehicles in winter seasonsand motor vehicles in the summer period.

4. Pregnancy outcome

The main aim of this study was to evaluate theimpact of air pollution and lifestyle variables on all

Žhospitalized pregnancies in two districts Teplice and.Prachatice using biomarkers as a measure of expo-

sure. The hypothesis was that the pregnancy out-comes would be generally worse in the district with

Ž .higher pollution level Teplice . Exposure was esti-mated by air pollution monitoring and modeling,questionnaire information and selected biomarkers.Based on the power analysis, the main reproductive

Žeffects were chosen, namely, low birth weight be-. Ž .low 2500 g , premature births below 37 weeks , and

Žintrauterine growth retardation small for gestational. w xage 14 .Two basic approaches were combined:Ž .a In the prospective cohort study, all hospital-

ized pregnancies terminated in Teplice and Prachat-Ž .ice districts during a 2-year period 1994–1996


were involved. Personal and lifestyle data were ob-tained via self-administered questionnaires, togetherwith information about reproductive history, workand environmental exposures, health status, diseases,and medications.

Ž .b Selected biomarkers were analyzed in venousblood, cord blood, and placentas of the women en-rolled in a nested case-control study. Women withpregnancies less than 37 weeks or with babiesweighing less than 2500 g were classified as cases.An equal number of women giving birth to babieswith normal parameters were selected systematically

Žfrom the whole cohort as controls each fifth noncase.birth was selected . Biomarkers such as DNA

adducts, chromosomal aberrations, comet assay, celland humoral immunity markers, and vitamins A, E,and C were analyzed in blood and placenta obtainedfrom cases and controls. Metabolic genotypesŽ .GSTM1, and NAT2 were determined.

Preliminary analysis of the data collected duringthe first 18 months of the study resulted in twogeneral observations:

Ž .1 Air pollution level in the polluted regionŽ .namely SO , PM10, PAHs was highest during2

winter and lower during summer, as measured bycontinuous air monitoring.

Ž . Ž .2 The prevalence of low birth weight LBWŽ .and intrauterine growth retardation IUGR was

greater for infants conceived during winter monthsŽ .than for those conceived in summer Fig. 3 . These

preliminary observations were consistent with recenthypotheses for the etiology of IUGR, that initialchanges leading to fetal growth retardation may betriggered in very early pregnancy, around the time of

w ximplantation 15,16 .The study group included all singleton full-term

births of European origin in Teplice district. Informa-tion on reproductive history, health and lifestyle wasobtained from maternal questionnaire. The meanconcentration of PM10 for each month of gestationwas calculated using continuous monitoring data.Each mother’s PM10 level was estimated using aver-ages for each of nine consecutive 30-day periods

Ž .after the estimated date of conception EDC ; theseperiods correspond roughly to the nine months ofgestation. The PM10 data were divided into three

Ž 3.categories for analysis: low Ls-40 mgrm ,Ž 3. Žmedium Ms40 to -50 mgrm and high Hs)

3.50 mgrm .Each month of pregnancy was analyzed sepa-

Žrately, allowing some factors to vary over time e.g.,particulate matter, maternal smoking and alcohol

.consumption, and season . Thirty-day averages ofPM10, calculated for nine time periods for eachpregnancy, varied between 29 and 86 mgrm3. Thevalues were highest in winter due to inversions and

Fig. 3. Prevalence of adverse outcomes in pregnancies conceived during summer and winter periods in the district of Teplice. LBWs lowbirth weight, IUGRs intrauterine growth retardation.


the use of coal for heating, and lowest in summermonths. The correlation between season and particu-late matter levels was significant, ranging between

Ž .0.49–0.60 p-0.01 for monthly average PM10.The study included 1943 women who gave birth

to infants between the 37th and 43rd gestationalweeks in the period 1994–1996. A total of 9.8%infants in the study were below the 10th percentile ofbirth weight for gestational age. Many potentiallyconfounding factors were included into transitionallogistic models. The final model contains maternalheight and pre-pregnancy weight, parity, marital sta-tus, education and maternal smoking, season and the

Ž .year of the study. Adjusted Odds Ratios OR andtheir 95% confidence intervals were estimated using

w xProc logistic 17 .Evaluating crude ORs, we observed the effect of

air pollution on IUGR at medium and high PM10Ž .levels in the first month of pregnancy Fig. 4 . After

adjustment, both the medium and high exposurewere statistically significant for PM10: the medium

Ž .level OR was 1.62 CI: 1.07–2.46 and the OR for

Ž .the high level was 2.64 CI: 1.48–4.71 . ThoughPM10 levels were lower in Prachatice than in Teplice,a similar significant association between IUGR andPM10 level in the first month of pregnancy wasobserved. These data suggest that exposure to partic-

Ž .ulate matter or an associated air pollutants early inpregnancy may adversely affect fetal growth. Re-gardless of which particular toxicant associated withparticulate matter could affect fetal growth, the bio-logical mechanisms remain to be explained. Theeffective components would need to be inhaled andabsorbed into the maternal bloodstream. Highly bio-

Ž .logically active compounds e.g., PAHs might inter-fere with some processes affecting development ornutrition of the fetus.

Effects during the first month of gestation wereobserved for monthly average concentrations ofPM10 above 40 mgrm3. These levels are also ob-served in other industrialized parts of the Czech

w x w xRepublic 18 and other countries 19 . Therefore,potential adverse effects of pollutants may be rele-vant for a significant number of births anywhere.

Fig. 4. Adjusted OR of IUGR by PM10 levels during pregnancy.


Fig. 5. Prevalence of IUGR for PM10 levels in the first month of gestation by maternal smoking and season.

Data on indoorroutdoor exposures from theTeplice region suggest that PM10 levels in non-smokers’ homes are approximately 50–80% of out-door PM10 levels. Concentrations of PM10 in smok-ers’ homes are substantially higher and therefore lessaffected by outdoor levels. Therefore, it is importantto control for active and passive smoking in this typeof studies. Active smoking of the mother and herexposure to the environmental tobacco smoke sub-stantially increases the IUGR among their childrenŽ .Fig. 5 .

5. Biomarkers and pregnancy

The impact of air pollution on pregnancy outcomewas analyzed by several biomarkers in venous bloodof mothers, cord blood and placenta of the womenenrolled in a nested case-control study.

5.1. Chromosomal aberrations

In the period 1992–1994, the level of chromoso-mal aberrations was analyzed in pregnant mothers in

the course of pregnancy, mothers in the course ofdelivery and in newborns cord blood. The increase ofchromosomal aberrations in the Teplice district com-pared to Prachatice district in pregnant mothers was

Ž .observed Table 1 . No differences between theŽ .groups of mothers sampled during delivery as well

as groups of newborn children were found. Thefrequency of aberrant cells in newborns from bothdistricts corresponds to spontaneous level of chromo-somal aberrations in peripheral blood lymphocytes ofnewborns in the Czech Republic. The level of chro-mosomal aberrations in mothers in both the regionswere significantly lower than the frequency of aber-rant cells for the same age group, which was 1.59%

w xAB.C. 20 . It seems that exposure to air pollutantsmay have more effect on chromosomes of mothersduring pregnancy than at the end of pregnancy andin the newborn.

5.2. Comet assay

ŽComet assay the single cell gel electrophoresis.assay was selected as a biomarker of exposure to


Table 1Cytogenetic analysis of peripheral lymphocytes in pregnant women and cord blood, period 1992–1994

a bDistrict Pregnant Mothers Cord bloodc dN % AB.C. BrC N % AB.C. BrC N % AB.C. BrC

U UTeplice 131 1.54 0.017 54 1.22 0.016 86 0.89 0.010Prachatice 48 1.04 0.012 20 1.05 0.012 29 1.00 0.011

Up-0.05.

a Pregnant—samples during pregnancy.b Mothers—samples in the course of delivery.cAB.C.—aberrant cells.d BrC—breaks per cell.Cultivation 48 h, analyzed 100 metaphasesrsubject.

evaluate the impact of air pollution and lifestylevariables on hospitalized pregnancies in both thedistricts. We supposed that the DNA damage de-tected as single strand breaks would be generallyhigher in the district with higher pollution. A total of322 pregnancies in Teplice and 220 in Prachatice

w xwere enrolled 21 . Venous and cord blood wereanalyzed using the alkaline Comet assay procedurewith lysis for 60 min, unwinding for 40 min andelectrophoresis for 24 min. The modified procedurewith unwinding prolonged to 60 min and elec-trophoresis to 40 min was used later. Peripheralwhite blood cells were analyzed using an imageanalyzer system. No differences for Comet parame-ters in the blood of mothers and children betweenpolluted and control districts were found. The pro-longation of alkali unwinding and electrophoresis didnot increase sensitivity of the assay. No effects ofpre-maturity, low birth weight, ethnicity, smoking orGSTM1 polymorphism were observed for any of theComet parameters. Multiple regression analyses were

Ž .performed for the European population ns285 . Astatistical model was fitted to determine the relation-ship between the Comet parameters of mothers andtheir children, the only ones, which were statisticallyand biologically significant.

The Comet assay was not a particularly sensitivetechnique in determining the effects of environmen-tal pollution at the DNA level if peripheral white

w xblood cells were used 21 .

5.3. DNA adducts

The placenta bulky DNA adducts have been stud-ied in relation to metabolic genotypes for glu-

Ž .tathione-S-transferase M1 GSTM1 and N-acetylŽ .transferase 2 NAT2 , cotinine and vitamins A, C

Žand E in 158 mothers 113 nonsmokers and 45.smokers . DNA adduct levels were determined using

the butanol extraction enrichment procedure of 32 P-postlabeling. GSTM1 and NAT2 genotypes were

Ž .studied using polymerase chain reaction PCR andRFLP techniques.

All sets of data were analyzed by multiple regres-sion analysis. Using this approach more completerelationship could be observed as compared to de-

w xscriptive statistics in the original paper 22 . TotalDNA adduct levels were significantly higher in thepolluted region and in smoking mothers. HigherDNA adduct levels were detected in the group ofmothers carrying the GSTM1-null genotype. Thisfinding seems more relevant to subjects living in

w xpolluted industrial regions 23 .The biomarker data has been employed for the

Ž .following reasons: a to test the biomarkers for theŽ .prediction of environmental exposure, b to evaluate

the effect of metabolic andror nutritional status onŽ .DNA adduct levels, and c to test the association

between the biomarkers of exposure, susceptibilityand pregnancy outcome. For sound valid results,many social, demographic and lifestyle character-

Žistics were controlled e.g., maternal age, parity,pre-pregnancy weight, education, marital status, ac-

.tive and passive smoking . Environmental exposuresof the mothers during the last three months of gesta-tion were estimated using data of continuously moni-toring of PM10, PAHs, and SO .2

Multiple regression models using stepwise proce-dure for evaluation of interaction of all possible


variables was used. DNA adducts as outcomes re-sponse variables were utilized. As predictors wereapplied air pollutants, smoking status, genotypes,antioxidant vitamin levels, and IUGR. Using thestepwise approach the final models contained onlycovariates which contribute significantly to the modelpredictivity. Analyzing the effect of carcinogenicPAHs concentrations and vitamin C levels in non-smoking women, an inverse relationship betweenvitamin C levels and DNA adduct levels was foundŽ .bsy0.649, p-0.05 . The higher DNA adductlevels was observed in nonsmoking mothers deliver-

Ž .ing children with IUGR bsy0.741, ps0.01 .The DNA adduct levels was significantly affected byGSTM1 resulting in higher DNA adduct levels inplacentas of the GSTM1-null nonsmoking mothersŽ .bsy0.513, p-0.05 . We also at first observed

Žthe effect of passive smoking on DNA adducts bs.0.031, p-0.01 . DNA adducts were also increased

by the exposure to carcinogenic PAHs in the lastŽ .month of pregnancy bs0.043, p-0.05 .

Ž .Summing up these results Table 2 , multipleregression analysis indicates that DNA adduct levelsin placentas of all women are affected by smoking,NAT2 genotype, vitamin C level, IUGR and educa-tion. In the group of smoking women the DNAadduct levels are affected by NAT2 genotype, car-cinogenic PAHs in the last month of pregnancy,IUGR, the length of gestation and education. Theeffect of the education on DNA adduct levels mayindicate the role of better knowledge of good dietand healthy lifestyle. The results of multiple regres-

sion analysis showed that DNA adducts togetherwith analysis of genotypes are sensitive biomarkersof exposure. These biomarkers may be used to findvarious factors which may affect the development of

Ž .pregnancy. It is probably relevant to expect believethat the DNA adduct levels correspond to the expo-sure during the last month or, in the best case, duringthe last three months of pregnancy.

A surprising association was observed betweenDNA adduct levels and IUGR which are anticipatedto be induced by PM10 already in the first month of

w xpregnancy 14 .DNA adduct data in placenta are complementary

with in vitro DNA binding activity and embryotoxic-w xity studies 11 which proved genotoxic and embry-

otoxic potential of the organic extracts from the airin the Teplice and Prachatice districts. The comple-mentarity of both results may be used to put forwarda hypothesis about the risk factors after modellinggenotoxicity of complex mixtures in human cells invitro.

We can propose:- to use the analysis of DNA adducts and identifi-cation of specific DNA adducts,- to determine mutations using the analysis ofHPRT and p53 with the identification of specificmutation spectra,- to use cytogenetic endpoints as chromosomalaberrations, SCE or micronuclei and determine thesensitivity of cytogenetic endpoints,

Ž- to determine genotypes to be studied metabolic.polymorphisms .

Table 2Multiple regression analysis of factors affecting DNA adducts in placenta

Group All Nonsmokers Smokers

N 104 77 27R 0.5084 0.2842 0.9530

B P-level B P-level B P-level

IUGR y0.971 0.0004 y1.115 0.0079 y0.889 0.0015Length of gestation 0.198 0.0001Education y0.153 0.0374 y1.103 0.0001Smoking 0.704 0.0006Vitamin C y0.753 0.0047 y0.323 0.0019NAT2 y0.453 0.0026 y0.240 0.0202 y0.560 0.0021Carcinogenic PAHs y0.371 0.0012Ž .in ninth month of pregnancy


The results of in vitro studies might help us toformulate hypotheses about risk factors in selectingthe population to study as an exposed and controlgroups and help us in the selection of validated andsensitive biomarkers.

The availability of extensive exposure measure-ments is critical to molecular epidemiology studies.However, it is very difficult to establish monitoringof all relevant pollutants on a routine basis as it isexpensive and very laborious. Still for the future, itshould be the basic requirement for molecular epi-demiology studies. It seems that to realize propermolecular epidemiology studies with a broad spec-trum of the biomarkers of exposure, effects andsusceptibility, only the international collaboration ofmany laboratories could be successful.

6. Semen quality

This study examined whether exposure to highlevels of air pollution over the entire process of

Ž .spermatogenesis about 90 days is associated withw xabnormal semen parameters 24 . Certain compo-

nents of air pollution, e.g., PAHs, have altered malereproductive function in test species providing addi-

w xtional rationale for the undertaking 25,26 . In theoverall study, 325 young men from both communi-ties, Teplice and Prachatice were examined. Surveysin the fall and late winter included interviews, physi-

cal examinations and collection of semen samples.Semen quality measures included concentration, vol-ume, motility and morphology, computer-aided mo-tion analysis and sperm chromatin structure assay.Men were classified into exposure groups using rela-tive levels of ambient sulfur dioxide.

Semen was collected on site, measures includedsperm concentrations and volume, percentage ofmotile sperm, percentage of sperm with normal mor-phology and percentage with normal head morphol-ogy. For the morphology evaluation 200–300 spermsper sample were assessed from air dried preparations

w xand classified into WHO categories 27 . To estimatethe relevant exposure scenarios SO data were2

grouped by district, sampling time, early fall or latewinter, year and exposure window, as shown inTable 3. An examination of these exposure windowsshowed that mean SO values were similar regard-2

less of the exposure window used. There were nosignificant relationship observed between measures

Žof sperm production semen volumes, sperm concen-.trations or total sperm count and air pollution in the

logistic regression model.The most notable findings of this report are the

significant relationships observed between air pollu-Ž .tion and sperm morphology Table 3 with the lower

air pollution levels in the late winter in PrachaticeŽ . Ž .with ORs of 0.2 0.1–0.7 and 0.5 0.2–2.0 and

increasing with medium levels to an OR of 4.1Ž . Ž1.2–13.9 and high levels to an OR of 10.1 2.8–

.36.0 in Teplice. The results were consistent for the

Table 3Semen outcomes by level of exposure late winter surveys versus all fall surveysa

Prachatice Teplice

Very low Low Medium High

Late winter 94 Late winter 93 Late winter 94 Late winter 93Ž Ž Ž Žns48, mean ns28, mean ns63, mean ns47, mean

. . . .SO s17.5 SO s41.5 SO s79.4 SO s164.02 2 2 2

Ž . Ž . Ž . Ž .Total sperm count -40 millionrsample 0.6 0.2–1.7 1.5 0.5–4.5 1.1 0.5–2.5 0.7 0.3–1.9Ž . Ž . Ž . Ž .Percent motile -24% 2.0 0.7–5.5 2.0 0.6–6.8 9.8 3.6–27.2 3.5 1.1–11.5Ž . Ž . Ž . Ž .Sperm morphology -13% normal 0.2 0.1–0.7 0.5 0.2–2.0 4.1 1.2–13.9 10.1 2.8–36.0Ž . Ž . Ž . Ž .Sperm head shape -29% normal 0.5 0.2–1.6 0.2 0.0–1.2 6.1 2.0–18.4 1.4 0.4–5.0

a For all fall: Teplice: Fall, 1992, mean SO s23.8, ns36; Fall, 1993, mean SO s29, ns44; Prachatice: Fall, 1992, mean SO s8.9,2 2 2

ns17; Fall, 1993, mean SO s6, ns42.2Ž .Odds Ratios 95% confidence intervals .


linear regression: sperm morphology, very low bsŽ . Ž .3.31 0.12, 6.52 , low bs5.90 1.88, 9.93 , medium

Ž .bs5.80 y8.84, y2.75 , ps0.0002 and high bsŽ .y7.25 y10.62, y3.89 , ps0.0001. These analy-

ses suggest a strong relationship between sperm mor-phology and air pollution with a possible seasonaleffect in the opposite direction.

An increase in abnormal sperm head morphologyw Ž .xwas also noted with medium ORs6.1 2.0–18.4

w Ž .xbut not high exposure ORs1.4 0.4–5.0 , respec-Ž .tively Table 3 . The linear regression was consistent

with these findings: medium exposure bs9.58Ž .y13.29, y5.86 , ps0.0001; high bsy3.84Ž .y7.95, y0.26 , ps0.07. Thus alteration of spermhead shape may be a significant component of thesperm morphology effect, but does not account forall of it.

The main effects associated with air pollutionappear to be postmeiotic effects on sperm motilityand morphology. Effects on sperm morphology sug-gest effects during spermiogenesis when the normalspermatids are transformed into differentiated spermcells. Severe alterations in motility and morphologycan be associated with infertility. Accumulating evi-dence suggests that exposure to environmental pollu-tion may contribute to a decline in sperm counts

w xworldwide 28,29 . In general, these data suggest thatexposure to air pollution for one spermatogenic cyclemay increase the risk of altered semen quality. Thisappears to be reversible since the young men evalu-ated six months after high pollution episodes haveimproved semen quality.

In the cohort of 25 subjects from the Teplicedistrict aneuploidy human sperm by FISH were ana-

w xlyzed 30 . Semen samples were provided by 10 menwho reported smoking 20 cigarettesrday, and 15nonsmokers who reported no more than minimalexposure to passive smoking. Sperm FISH aneu-ploidy was determined with the use of three probes

Ždirectly labeled for chromosome X, Y and 8 Table.4 . There were no statistically significant differences

between smokers and nonsmokers with regard tooccupational exposure, health status, passive smok-ing, age at first seminal emission and other activities.Smokers produced significantly elevated aggregatefrequencies of sperm, aneuploidy for chromosomes

Ž .X, Y and 8 p-0.01 . The frequencies of spermwith Y disomy were significantly elevated among

Table 4w xAneuploidy in human sperm by FISH, effect of smoking 30

Sperm aneuploidy Grouptype Nonsmokers Smokers

Ž . Ž .ns15 ns10

DisomyX 5.8"0.6 5.9"0.8

UY 2.2"0.3 4.5"0.6X–Y 14.0"1.4 20.4"4.48 4.7"0.6 6.5"1.0

DiploidyX–X–8–8 2.3"0.6 4.7"1.4Y–Y–8–8 2.6"0.6 2.5"1.1X–Y–8–8 19.0"5.3 23.9"5.7

Up-0.001.

4 Ž .Note: Values are frequency per 10 sperm mean"SE ; all nucleiare confirmed as sperm heads by the presence of flagella.

Žsmokers in comparison with nonsmokers p -.0.001 . There have been no previous report relating

an exogenous exposure to increased frequencies of Ydisomy in sperm. The finding of increased frequen-cies of Y disomy in sperm may have implications forthe risk of having an aneuploid child. The proportionof affected children may be higher in men whosmoke more.

Relationship between air pollution in the districtof Teplice, other covariates of interest and spermaneuploidy was described using Poisson and linear

w xregression modeling 31 . YY8 aneuploidy was sig-nificantly associated with the season of heaviest air

Žpollution ORs5.25, 95% CI 2.5–11.0; linear mod-eling of normally transformed YY8, coefficient 1.44,

.p-0.0001 . No other cytogenetic endpoints weresignificantly associated with seasonal air pollution.These findings are suggestive of an influence ofseasonal air pollution on YY8 disomy. Rubes putˇforward an idea for further study if disomy of chro-mosome Y in sperm could be used as a marker ofsperm injury by chronic exposure to environmentalpollutants.

The sperm disomy findings for the Teplice non-smokers compared closely with those of a group of

Ž .Californian non-smokers Table 5 . The slides forboth cohorts were prepared and scored in the same

w xlaboratory 32,33 using the same probes and scoringcriteria, but different scorers. Overall, there were nodifferences in the levels of sperm diploidies or Y


Table 5Comparison of sperm aneuploidy levels between Teplice and

w xCalifornia nonsmokers 32,33

Sperm aneuploidy Grouptype Teplice California

Ž . Ž .ns15 ns14

DisomyUX 5.8"0.6 3.1"0.5

Y 2.2"0.3 3.1"0.7UX–Y 14.0"1.4 9.5"0.7U8 4.7"0.6 6.6"0.7

DiploidyX–X–8–8 2.3"0.6 2.2"0.4Y–Y–8–8 2.6"0.6 1.7"0.4X–Y–8–8 19.0"5.3 10.6"1.4

Up-0.001.

4 Ž .Note: Values are frequency per 10 sperm mean"SE ; all nucleiare confirmed as sperm heads by the presence of flagella.

disomy. The Czech non-smokers, however, were ele-vated in the levels of X, X–Y and 8 disomiesŽ .ps0.02 . These differences may be related to thegeographic and ethnic differences, pollution differ-ences, the slight age difference, or cultural differ-ences between the two cohorts.

7. Conclusion

Data generated from the Teplice Program indicatethat air pollution in Central Europe may represent avery specific problem. We should expect health im-pact on adverse reproductive outcomes, both in fe-males and males. Probably the most prominent envi-ronmental pollutants are organic compounds, espe-cially PAHs. Another risk factor is lifestyle. We canspeculate that in fact smoking and mutagens fromdiet could be a more significant risk factor than airpollution.

All these ideas indicate that this region seems tobe suitable for systematic molecular epidemiologystudies. It seems that pregnant women, newbornchildren and young men are groups most susceptibleto the air pollution expressed as PM10 or PAHs.Precise determination of their health risk for diseaseprevention is an important task for molecular epi-demiology.

Acknowledgements

The authors would like to thank Prof. R.J. Alber-tini, University of Vermont, USA, and Prof. W.W.Au, University of Texas Medical Branch at Galve-ston, USA, for their critical comments in the prepara-tion of this manuscript. The authors are grateful toDr. Sally Perreaut-Darney and Dr. Sherry G. Selevanfor their support in those studies. The authors ac-knowledge the technical support of the staff of theDistrict Institutes of Hygiene in Teplice and Prachat-ice for providing of the air monitoring under thesuperiority by Dr. I. Benes and Dr. J. Nozicka. Thisˇ ˇ ˇstudy was supported by the Czech Ministry of Envi-

Ž .ronment Teplice Program II—VaV 340r1r1997 ,ŽU.S. EPArU.S. AID and CEC PHARE II, ECr

.HEAr18-CZ .

References

ˇw x1 R.J. Sram, I. Roznıckova, V. Albrecht, A. Berankova, E.´ ˇ ´ˇ ´ ´ ´Machovska, Monitoring congenital anomalies in populations´

Ž .exposed to environmental mutagens, in: A. Kappas Ed. ,Mechanisms of Environmental Mutagenesis-Carcinogenesis,Plenum Press, New York, 1990, pp. 255–266.

ˇw x2 R.J. Sram, New ethical problems related to environmental´ˇpollution, in: R.J. Sram, V. Bulyshenkov, L. Prilipko, Y.´

Ž .Christen Eds. , Ethical Issues of Molecular Genetics inPsychiatry, Springer, Berlin, 1991, pp. 94–105.

ˇw x3 R.J. Sram, I. Benes, B. Binkova, J. Dejmek, D. Horstman, F.´ ˇ ´Kotesovec, D. Otto, S.D. Perreault, J. Rubes, S.G. Selevan, I.ˇˇ ˇSkalık, R.K. Stevens, J. Lewtas, Teplice Program—the im-´pact of air pollution on human health, Environ. Health

Ž .Perspect. 104 1996 699–714.w x4 R. Watts, J. Lewtas, R. Stevens, T. Hartlage, J. Pinto, R.

William, K. Hattaway, I. Mıskova, I. Benes, F. Kotesovec,´ˇ ´ ˇ ˇˇˇR. Sram, Czech–U.S.EPA health study: assessment of per-´

sonal and ambient air exposure to PAH and organic muta-gens in the Teplice district of Northern Bohemia, Int. J.

Ž .Environ. Anal. Chem. 56 1994 271–287.ˇw x5 B. Binkova, J. Lewtas, I. Mıskova, J. Lenıcek, R. Sram,´ ´ˇ ´ ´ˇ ´

DNA adducts and personal air monitoring of carcinogenicpolycyclic aromatic hydrocarbons in an environmentally ex-

Ž .posed population, Carcinogenesis 16 1995 1037–1046.ˇw x6 B. Binkova, J. Lewtas, I. Mıskova, P. Rossner, M. Cerna, G.´ ´ˇ ´ ¨ ´

ˇMrackova, K. Peterkova, J. Mumford, S. Meyer, R. Sram,ˇ ´ ´ ´Biomarker studies in Northern Bohemia, Environ. Health

Ž . Ž .Perspect. 104 3 1996 591–597.w x7 J.P. Pinto, R.K. Stevens, R.D. Willis, R. Kellogg, Y. Ma-

ˇmane, J. Novak, I. Santroch, I. Benes, J. Lenıcek, V. Bures,´ ˇ ´ˇ ˇCzech air quality monitoring and receptor modeling study,

Ž .Environ. Sci. Technol. 32 1998 843–854.


w x8 R.K. Stevens, J.P. Pinto, Y. Mamane, J. Novak, I. Benes,´ ˇMonitoring and modeling methods for developing air pollu-tion control strategies: a case study in northwest Czech

Ž .Republic, in: I. Alligrini, F. DeSantis Eds. , NATO ASISeries, Partnership Sub-Series, 2: Environment, Springer,Heidelberg 1996, pp. 151–166.

w x9 B. Binkova, J. Lenıcek, I. Benes, P. Vidova, O. Gajdos, M.´ ´ˇ ˇ ´ ˇˇFried, R.J. Sram, Genotoxicity of coke oven and urban air´

particulate matter in in vitro acellular assay coupled with32 P-postlabeling and HPLC analysis of DNA adducts.

w x10 J. Lewtas, J. Chuang, M. Nishioka, B. Peterson, Bioassay-di-rected fractionation of the organic extract of SRM 1649urban air particulate matter, Int. J. Environ. Anal. Chem. 39Ž .1990 245–256.

ˇw x11 B. Binkova, D. Vesely, D. Vesela, R. Jelınek, R.J. Sram,´ ´ ´ ´ ´Genotoxicity and embryotoxicity of urban particulate mattercollected during winter and summer period in two differentdistricts of the Czech Republic, Mutat. Res., in press.

w x12 D.M. DeMarini, M.L. Shelton, D.A. Bell, Mutation spectrain Salmonella of complex mixtures: comparison of urban air

Ž . Ž .to benzo a pyrene, Environ. Mol. Mutagen. 24 1994 262–275.

w x13 D.M. DeMarini, Mutation spectra of complex mixtures, Mu-Ž .tat. Res. 411 1998 11–18.

ˇw x14 J. Dejmek, S.G. Selevan, I. Benes, I. Solansky, R.J. Sram,ˇ ´ ´Fetal growth and maternal exposure to particulate matterduring pregnancy, Environ. Health Perspect., in press.

w x15 T.Y. Khong, F. De Wolf, W.B. Robertson, I. Brosens, Inade-quate maternal vascular response to placentation in pregnan-cies complicated by pre-eclampsia and by small-for-gesta-

Ž .tional age infants, Br. J. Obstet. Gynecol. 93 1986 1049–1059.

w x16 J.J. Duvekot, E.C. Cheriex, F.A.A. Pieters, Severely im-paired growth is preceded by maternal haemodynamic mal-adaptaton in very early pregnancy, Acta Obstet. Gynecol.

Ž .Scand. 74 1995 693–697.w x17 R.B. Schlesinger, Chapter Six: Deposition and clearance of

Ž .inhaled particles, in: McClellan, R.F. Henderson Eds. , Con-cepts in Inhalation Toxicology, Hemisphere Publ., New York,1989.

w x18 System Monitoring of Health Status of Population in relationto Environment: Report 1997, Natl. Inst. of Public Health,

Ž .Prague, 1998 in Czech .w x19 T.J. Woodruff, The relationship between selected causes of

post-neonatal infant mortality and particulate air pollution inŽ .the United States, Environ. Health Perspect. 105 1997

608–612.ˇw x20 P. Rossner, R.J. Sram, H. Bavorova, D. Ocadlıkova, M.¨ ´ ´ ˇ ´ ´

ˇ ˇCerna, E. Svandova, Spontaneous level of chromosomal´ ´aberrations in peripheral blood lymphocytes of control indi-viduals of the Czech Republic population, Toxicol. Lett. 96,

Ž .97 1998 137–142.

ˇw x21 R.J. Sram, K. Podrazilova, J. Dejmek, G. Mrackova, T.´ ´ ˇ ´Pilcık, Single cell gel electrophoresis assay: sensitivity ofˇ́peripheral white blood cells in human population studies,

Ž .Mutagenesis 13 1998 99–103.w x22 J. Topinka, B. Binkova, G. Mrackova, Z. Stavkova, V.´ ˇ ´ ´ ´

Peterka, I. Benes, J. Dejmek, J. Lenıcek, Influence of GSTM1ˇ ´ˇand NAT2 henotypes on placental DNA adducts in an envi-ronmentally exposed population, Environ. Mol. Mutagen. 30Ž .1997 184–195.

ˇw x23 R.J. Sram, Effect of glutathione S-transferase M1 polymor-´phisms on biomarkers of exposure and effects, Environ.

Ž . Ž .Health Perspect. 106 1 1998 231–239.w x24 S.G. Selevan, L. Borkovec, V.L. Slott, Z. Zudova, J. Rubes,´ ˇ

D.P. Evenson, S.D. Perreault, Air pollution and semen qual-ity in young men residing in two Czech communities, Epi-demiology, in press.

w x25 K.M. MacKenzie, D.M. Angevine, Infertility in mice ex-Ž . Ž .posed in utero to benzo a pyrene, Biol. Reprod. 24 1981

183–191.w x26 E. Ford, C. Huggins, Selective destruction in testis induced

Ž . Ž .by 7,12- dimethylbenz a anthracene, J. Exp. Med. 118 196327–40.

w x27 World Health Organisation, WHO Laboratory Manual for theExamination of Human Semen and Semen-Cervical MucusInteractions, 3rd edn., Univ. Cambridge Press Syndicate,Cambridge, UK, 1992.

w x28 J. Toppari, J.C. Larsen, P. Christianen, A. Giwercman, P.Grandjean, L.J. Guillette Jr., M. Jegou, T.K. Jensen, P.Jouannet, N. Keiding, H. Leffers, J.A. McLachlan, O. Meyer,J. Muller, E.R. Meyts, T. Scheike, R. Sharper, J. Sumpter,N.E. Skakkebaek, Male reproductive health and environmen-

Ž . Ž .tal xenoestrogens, Environ. Health Perspect. 104 4 1996741–776.

w x29 J. Auger, J.M. Kunstman, F. Czyglik, P. Jouannet, Decline insemen quality among fertile men in Paris during the past 20

Ž .years, N. Engl. J. Med. 332 1995 281–285.w x30 J. Rubes, X. Lowe, D. Moore, S.D. Perreault, V. Slott, D.ˇ

Evenson, S.G. Selevan, A.J. Wyrobek, Smoking cigarettes isassociated with increased sperm disomy in teenage men,

Ž .Fertil. Steril. 70 1998 715–723.w x31 W.A. Robbins, J. Rubes, S.G. Selevan, S.D. Perreault, Sea-ˇ

sonal air pollution and sperm aneuploidy in healthy 18-year-olds, An International Conference ‘Hazardous Substancesand Male Reproductive Health’, New York, 1998.

w x32 J. Rubes, X. Lowe, M. Cassel, D. Moore, S.D. Perreault, V.ˇSlott, D. Evenson, Z. Zudova, L. Borkovec, Aneuploidy´detection in sperm nuclei using fluorescence in situ hy-

Ž .bridization, Arch. Zootecnia 45 1996 323–327.w x33 S.G. Selevan, A.J. Wyrobek, Sperm aneuploidy in smokers

and nonsmokers living in the Teplice district of the CzechŽ .Republic, Cytogenet. Cell. Genet. 74 1996 240.

Documents

Adverse reproductive outcomes from exposure to environmental mutagens