Ecotoxicology and Environmental Safety 94 (2013) 14–20

Contents lists available at SciVerse ScienceDirect

Ecotoxicology and Environmental Safety

0147-65http://d

n CorrE-m

acmielli1 Pr

Eldorad

journal homepage: www.elsevier.com/locate/ecoenv

Comparative genotoxicity of airborne particulate matter (PM2.5)using Salmonella, plants and mammalian cells

Kelly Cristina Tagliari de Brito a,1, Clarice Torres de Lemos a, Jocelita Aparecida Vaz Rocha a,Ana Cristina Mielli b, Cristina Matzenbacher a, Vera Maria Ferrão Vargas a,n

a Programa de Pesquisas Ambientais, Fundação Estadual de Proteção Ambiental Henrique Luís Roessler (FEPAM), Avenida Salvador França,1707, 90690-000 Porto Alegre, RS, Brazilb Toxicology Laboratory, Medical Investigation Laboratory, Clinics Hospital School of Medicine, University of São Paulo, Avenida Dr. Arnaldo, 455, 01246-903São Paulo, SP, Brazil

a r t i c l e i n f o

Article history:Received 28 January 2013Received in revised form16 April 2013Accepted 18 April 2013Available online 31 May 2013

Keywords:Airborne particulate matter (PM2.5)Salmonella/microsome assayMicronucleus testComet assayTradescantia pallida

13/$ - see front matter & 2013 Published by Ex.doi.org/10.1016/j.ecoenv.2013.04.014

esponding author. Fax: +55 51 33346765.ail addresses: kelly_tagliari@hotmail.com (K.C@yahoo.com.br (A.C. Mielli), ecorisco@fepam.esent address: FEPAGRO-IPVDF, Estrada do do Sul, RS, Brazil.

a b s t r a c t

This study compared genotoxicity in bacteria, plants and cell cultures in areas at risk of exposure toairborne pollution. Genotoxicity of moderately polar organic extracts of PM2.5 from areas with urbanairborne pollution (Site 1) and urban-industrial pollution (Site 2) was evaluated using microsuspensionassays in Salmonella/microsome, micronucleus test with Tradescantia pallida (Trad-MN) with acuteexposure, and in V79 (V79-MN) cells, Comet assay in V79 and human lymphocyte, besides Trad-MNin situ at Site 1. In the Salmonella/microsome assay all samples presented frameshift mutagenic activity(−/+S9), most intense at Site 2 (rev/m3). The presence of nitro-PAHs and hydroxylamines in PM2.5 wasshown by positive mutagenic responses with YG1021 and YG1024. In tests with Trad-MN, no significantgenotoxic responses were found (MN %). In V79-MN a genotoxic response was not found. The Cometassay damages were found in the DNA at Site 1 in both cell systems. Non-detection of genotoxicity withTrad-MN at sites or in environmental samples from polluted areas detected using other biomarkerssuggests the need for careful evaluation when biomonitoring genotoxic compounds using plants. Themicrosuspension assay in Salmonella/microsome was sensitive to detect and identify different classes ofairborne mutagenic compounds present in fine particulate matter in Porto Alegre city, showing thatmonitoring air quality with PM2.5 using this methodology is relevant.

& 2013 Published by Elsevier Inc.

1. Introduction

The great world paradigm is sustainable development. Growingurbanization in many countries is increasing levels of airbornepollution from different sources, often compounds or their mix-tures, known genotoxic agents that can have adverse effectson human health (Coronas et al., 2009; Lemos et al., 2012;WHO, 2009).

Suspended particles are important indicators of environmentalpollution and their potential to cause health problems is directlylinked to their size, chemical composition and origin whichdetermine their transport and removal in air, and capacity topenetrate and be deposited in different areas of the humanrespiratory system. When PM2.5 is inhaled, finer particles, foundin cigarette smoke or in gases produced by industries and

lsevier Inc.

.T.d. Brito),rs.gov.br (V.M.F. Vargas).o Conde, 6000, 92990-000

automobiles can affect the bronchioles and interfere in pulmonaryrespiratory capacity (De Martinis et al., 1999; Vargas et al., 2011).

In fine airborne particulate matter there is often an associationof polycyclic aromatic hydrocarbons (PAHs) from combustionprocesses, industrial activities and, especially, incomplete burningof automotive fuel (engine exhaust, diesel or gasoline), some ofthem classified by the International Agency for Research on Cancer(IARC) as mutagenic and/or carcinogenic (IARC, 2012). The muta-genicity of airborne particles is generally associated with PAHs but,despite their contribution to the effect, recent reviews showedthat these mutagens may not be the predominant form of atmo-spheric pollutants, and nitroaromatic compounds, aromaticamines and aromatic ketones often found in moderately or highpolar atmospheric organic fractions are relevant (Claxton et al.,2004). Nitro-polycyclic aromatics (nitro-PAHs) are persistentenvironmental mutagens and can be found in airborne suspendedparticles from direct sources such as diesel and gasoline exhaust,or may be products of atmospheric reactions in the presence ofNO2 and NO3 radicals (De Martinis et al., 1999; Lewtas, 2007).

The international and Brazilian regulating agencies establishlimits or recommendations for air quality standards based on the

Fig. 1. Location of studied sites in the Porto Alegre city area, Rio Grande do Sul,Brazil.

K.C.T.d. Brito et al. / Ecotoxicology and Environmental Safety 94 (2013) 14–20 15

concentration of airborne particles with different diameters: inBrazil for total suspended particles (TSP) as 240–150 mg/m3 24-hmean and 80–60 mg/m3 annual mean; and for the inhalableparticles smaller than 10 mm (PM10), mean values of 150 mg/m3

for 24 h and 50 mg/m3 annually (Brazil, 1990). The World HealthOrganization Air Quality Guidelines (WHO, 2006) recommendsmore restrictive values for PM10 (50 mg/m3, 24 h mean and 20 mg/m3, annual mean) and PM2.5 (25 mg/m3, 24 h mean and 10 mg/m3

annual mean). However, studies evaluating air samples that werewithin the limits established in Brazil showed high mutagenicresponses, evidencing the need to review these parameters(Coronas et al., 2009; Vargas, 2003; Vargas et al., 2011).

Studies with exposure biomarkers together with the chemicalcharacterization of the airborne particulate matter and its sourcesof emission help understand the effects of biologically activepollutants and to choose the most appropriate atmosphericstandards for the safety of human health (Vargas et al., 2011).The Salmonella/microsome test, measuring genetic alterations atmolecular level was chosen to study the mutagenic potential ofenvironmental samples, due to its widespread use and continuousimprovement. Its use is recommended by international organiza-tions to diagnose genotoxins in water sources, sediments, atmo-spheric compartment and for public water supply (FederalRegister, 1989; OECD, 1997). The Ames test is most widely used,since it is sensitive to detect and identify mutagens commonlypresent in air, such as PAHs, nitrocompounds and aromaticamines. Using multiple strains and microsuspension method,mutagens can be detected in small samples like atmospheric ones(Claxton et al., 2004; Vargas, 2003).

Many cytogenetic assays, such as single-cell gel electrophoresis(SCGE) assay (Comet assay) or the evaluation of micronuclei havebeen used successfully to monitor populations exposed to muta-gens, and it is an effective warning system for genetic disease orcancers (Rojas et al., 1999). The Comet assay detects DNA double-and single-strand breaks, lesions in alkali-labile sites and incom-plete repair of excision sites in different eukaryotic cell types, bothin vivo and in vitro, supplying information about recent levels ofexposure to genotoxic substances where part of the damage canstill be repaired (Rojas et al., 1999; Singh et al., 1988).

The micronucleus (MN) test is often used for environmentalbiomonitoring to detect clastogenic agents that can break thechromosomes and aneugenic agents that induce aneuploidy orabnormal chromosomal segregation. Cell cultures such as strainV79 from Chinese hamster lung cells, and human lymphocytes, arewidely used to evaluate the potential to induce damaged geneticmaterial caused by chemicals or environmental samples (Lemosand Erdtmann, 2000; Lemos et al., 2011; Nunes et al., 2011).These cell types have shown sensitivity to detect toxicity andgenotoxicity caused by complex mixtures, and are among themost recommended in vitro biological systems in internationallyrecognized cytogenetic test protocols (OECD, 2010). In studies ofareas without a history of data, it is useful to perform thegenotoxic evaluation in biological systems that have a differentsensitivity.

Trad-MN involves hybrid clones of Tradescantia sp. (BNL e KU)and/or Tradescantia pallida allowing the quantification of MN inmother cells of pollen grain in flower sprigs exposed to aclastogenic and/or aneugenic agent present in various samples(Carvalho-Oliveira et al., 2005; Mielli et al., 2009). The MN in thisbioassay are formed in the initial phases of the meiotic cycle(prophase I) of pollen mother cells, considered the most sensitiveto the action of genotoxic agents and visualized in the tetradphase. They are frequently used in evaluations of atmosphericgenotoxicity (Carvalho-Oliveira et al., 2005; Klumpp et al., 2006).In Brazil, the use of Tradescantia pallida to replace Clone BNL 4430(Ma et al., 1978) in a micronuclei test, in situ or in the laboratory,

has been satisfactory and convenient because it is more adapted tothe weather conditions and resistant to insects found in SouthAmerica (Pereira et al., 2013; Suyama et al., 2002).

Evaluations of genotoxicity in different organisms and atdifferent cellular levels are important tools for the prevention ofpollution and to further the conservation and sustainable use ofthe environment. These studies allow the early detection ofsubstances with a capacity to react to genetic material, whichmay cause cumulative damage leading to alterations in differentorganisms. The association of assays evaluating genes in high riskareas for exposure to atmospheric mutagenic substances is inno-vative and this study aims to combine three genotoxicity assays,Salmonella/microsome test, MN and Comet assay.

2. Material and methods

2.1. Site

The study was performed on urban/industrial particulate matter of the city ofPorto Alegre (30101′59″S; 51113′48″W), the capital of Rio Grande do Sul state, insouth Brazil. The region is characterized by a humid subtropical climate, with anannual mean temperature of 20 1C, (76% mean relative air humidity ) and rainfallindex of 1324 mm a year (INMET, 2011; Vargas et al., 2011).

2.2. PM2.5 sampling and organic compounds extraction

PM2.5 samples collected at two sites in the city of Porto Alegre were evaluated(Fig. 1). Site 1 (30103′12.93″S, 51110′28.79″W), Jardim Botânico District, is locatednearby a weather station, about 1 km from an avenue with moderate traffic (about4000 vehicles/h), away from major industrial areas and near the city's BotanicalGarden. Site 2 (29158′46.94″S, 51110′6.79″W), Anchieta District, is located in amixed residential/industrial neighborhood, with small to medium-sized industries(textiles, food, metallurgy, latex, paper and plastic), affected by about 5000 vehi-cles/h and 2 km from a federal highway.

K.C.T.d. Brito et al. / Ecotoxicology and Environmental Safety 94 (2013) 14–2016

The samplings were performed by pools in August, September and October of2010, and January, February and March of 2011. PM2.5 samples were collected inTeflons

filters (TX40HI20WW, 254 mm�203 mm) using a high-volume collector(AVG MP10, 1200/CCV) for 2471 h, on the dates defined in Table A.1, resulting inmonthly pools comprising three to four samplings. The filters were stabilized andweighed (ABNT, 1995) before and after sampling to calculate the PM2.5 valuesexpressed in units of mg/m3 of sample air (Table A.1).

A quarter of the monthly pool of each filter was used to extract the organiccompounds. Each pool was submitted to extraction by sonication with dichlor-omethane (DCM, CASRN. 75-09-2) removing the moderately polar organic com-pounds considered the most representative fraction of mutagenicity (De Martiniset al., 1999; Vargas et al., 1998), extracting a wide range of compounds fromairborne particulate matter (Marvin and Hewitt, 2007). According to Claxton et al.(1992), this methodology allows extracting percent values of extractable masswithin the appropriate levels. During the second and fourth weeks of November/2010 and first week of December/2010 individual organic extracts from the filterswere prepared. The percentage of extractable organic matter (EOM percent) andthe mass were compared to a quarter of the volume of air sampled (EOM in mg/m3)(Table A.1). Prior to bioassay performance, the organic extract was dried withgaseous nitrogen and resuspended in dimethyl sulfoxide (DMSO, CASRN. 67-68-5).

2.3. Salmonella/microsome assay

Samples from the pools of organic extracts August–October/2010; and January–March/2011 or from individual filters (second and fourth weeks of November/2010and first week of December/2010), were evaluated for mutagenicity through theSalmonella/microsome assay (Maron and Ames, 1983), using the microsuspensionmethod (Kado et al., 1983) at 1.25, 2.5, 5, 10, 20 and 40 mg/plate concentrationsassessed in duplicate. Assays were performed using frameshift strains, TA98 in theabsence (−S9) and the presence (+S9) of microsomal standard fraction (4%)prepared from Sprague Dawley rat liver (Aroclor activator 1254, Moltox S.A.,USA), the YG1021 (nitroreductase-overproduction) and YG1024 (O-acetyltransfer-ase-overproduction) derivative strains of TA98 in the absence S9 (–S9), as specificto detect nitrocompounds (Watanabe et al., 1989). The negative control used wasDMSO (5 mL/plate) and the positive one (4-nitroquinoline oxide – 4NQO, 0.05 mg/plate, CASRN. 56-57-5; 2-aminofluorene – 2AF, 1 mg/plate, CASRN. 153-78-6; and 2-nitrofluorene – 2NF, 0.15 mg/plate, CASRN. 607-57-8, Sigma Chemical Company, St.Louis, MO) according to strain. Cytotoxic activity was also assessed by a survivalcurve of the test organism compared with different concentrations of the samples(Vargas et al., 1998).

The sample was considered mutagenic when a significant ANOVA (p≤0.05) wasobserved, accompanied by a positive dose-response curve (p≤0.05) assessed bySALANAL software (Salmonella Assay Analysis, version 1.0, Integrated LaboratorySystems of Research Triangle Institute, RTP, NC, USA) choosing the linear orBernstein model analyzes the mutagenicity of the linear portion of the dose-response curve (Bernstein et al., 1982). The results were expressed as number ofrevertants per mass of PM2.5 organic extracts (rev/mg) and volume of air sampled(rev/m3) calculated by multiplying the rev/mg by EOM in mg/m3. The cytotoxicitywas positive when the percentage of surviving cells in the sample was less than 60percent of the colonies compared with the negative control.

2.4. Micronucleus test in germinative cells of Tradescantia pallida

2.4.1. Chronic exposure – in situRooted stems of Tradescantia pallida, approximately 15–20 cm long, were

cultivated in commercial soil in an area free of urban atmospheric pollution(Floricultura BlumenGarten Plantas e Flores Ltda.). The cultivation substrate usedfor these plants was a mixture of: two parts special vegetal soil (Humosolo – VidaProdutos Biológicos band), one part substrate (Florestal brand), one part vermiculite(Terra Mater brand), 1/4 part worm humus (Vitaplan brand). Flowerpots were addedwith 4 kg of that mixture (Mielli et al., 2009). These plants remained exposed at Site1, close to the air samplers, during the same period as the air filter collection in 2010(August/September), until the first young inflorescences (without open flowers)appeared, which were collected in the fourth week of October; second, third andfourth weeks of November; first and fifth weeks of December/2010. The plants weregiven plenty of water at least twice a week. At the time of cultivation, younginflorescences of Tradescantia pallida were also collected and these supplied thebaseline index of reference micronuclei. The collected inflorescences were fixed in1:3 acetic acid–ethanol solutions for 24 h and stored in 70 percent ethanol formicronuclei analysis.

2.4.2. Acute exposure – short durationFloral stems with young inflorescences (without open flowers) of T. pallida,

from the garden of the Administrative Center at the University of São Paulo-USP,Medical School, São Paulo, remained in the laboratory for 24 h in glass containerswith tap water at ambient temperature, obeying for the adaptation period (Maet al., 1994; Mielli et al., 2009). After this time, the stems were divided into fiveexperimental groups with an average of 15 floral stems per group and exposed for

6 h to aqueous test solutions (200 mL): tap water (negative control); tap water+DMSO 3% (negative control); formaldehyde 1% (positive control); mixture of poolsof organic extracts (1 mg/L) from Site 1 and also from Site 2, obtained whilestudying chronic exposure (August-October/2010). This concentration was definedfrom data in the literature (Carvalho-Oliveira et al., 2005; Alves et al., 2011). Therecovery period after acute exposure to test solutions also occurred in a glasscontainer with tap water for 24 h at ambient temperature. The young inflores-cences were removed from the stems, fixed in an acetic acid/ethanol solution (1:3)and conserved in an ethanol solution 70 percent for later preparation and analysisof the microscopy slides.

Preparation and analysis of slides in both exposures were performed in thesame manner (Ma et al., 1994; Mielli et al., 2009). Initially floral buds in the youngtetrad stage of the inflorescences stored in ethanol 70 percent solution weredissected. The anthers obtained were macerated with a glass stick on a slide formicroscopy after dripping acetic carmine on them, and 8–10 slides were preparedper group. The slides were analyzed with 100� and 400� magnifications underoptical microscopy. For each group, at least 1500 tetrads were evaluated, and eachslide was considered a sample of the group, the results being expressed in terms ofMN/100 tetrads (MN %).

The data obtained for the different groups were evaluated using analysis ofvariance (ANOVA), followed by post-hoc Tukey (parametric data). The statisticalprogram used was SPSS vs. 13.0 and the differences were considered significant atp≤0.05 relative to negative control.

2.5. Micronucleus test in V79 cellular strain

Aliquots of cellular suspension containing 5�104 cells of V79 strain with MEMmedium supplemented by fetal bovine serum, penicillin G, streptomycin, L-glutamine, NaHCO3 were distributed in 25 cm2 culture flasks. Exposure to theorganic extracts of the pools for August, September and October of 2010, at Site1 and Site 2, was a mass of 40 mg of extract, corresponding to the largest dose usedin the Salmonella/microsome assay, at a final concentration of 8 mg/mL. The flaskswere incubated at 37 1C and 5% CO2 for 24 h (two cycles of cellular division).Negative controls were sterile distilled water and DMSO 3%, and bleomycin (0.1 mg/mL) as a positive control. After exposure, the cultures were washed in PBS andtrypsinized. The cells were collected and centrifuged for 5 min, resuspended insodium citrate and formaldehyde. The material was fixed in methanol:glacial aceticacid and staining was performed with Giemsa.

Fenech's criteria (Fenech, 1993) were used to analyze the micronuclei (MN),and 2000 cells were analyzed per sample and controls. Statistical analysis wasperformed using the Z test that compares two POISSON curves. The result wasconsidered positive when a significant difference was observed in relation tonegative control, with at least po0.05 (Zar, 1996).

2.5.1. CytotoxicityThe Plating Efficiency test (PE) was used to evaluate cytotoxicity. After

treatment, two hundred cells, were washed and incubated in MEM at 37 1C with5% of CO2 for seven days. The colonies were fixed with methanol and acetic acid(3:1) and stained with Crystal Violet 1%. Survival of the colonies was expressed inpercentage of the negative control (sterile distilled water). The result wasconsidered cytotoxic when cell survival was o70 percent (Cardozo et al., 2006).

2.6. Comet assay in V79 cells and Human Blood

In the alkaline Comet assays (Collins, 2004), the cellular suspension containing5�104 cells strain and 150 ml of whole human peripheral blood were exposed for3 h to the organic extracts of air from the pools of August, September and October2010, at Site 1 and Site 2, at a mass of organic extract of 40 mg, as already defined initem 2.5 at the final concentration of 66 mg/mL. Analysis was performed in 100 cellsto calculate the intensity of the tail using IV Comet Assay software. The correlationbetween the two sites sampled and the negative control was determined byanalysis of variance (ANOVA), Dunnett Test, in a GraphPad Prism program,version 5.0.

3. Results and discussion

The volume of air collected (Table A.1) in the filters was similarin most samples, while the concentration of PM2.5 presentedgreater variability, from 2 to 79 mg/m3. At both sites evaluated (Site1 and Site 2) it can be seen that higher mean concentration ofPM2.5 in mg/m3 and extractable organic matter (EOM) in mg/m3,were detected in August/2010. The filters collected on August 23,2010, presented high concentrations of PM2.5, in agreement withthe period of occurrence of a dense dust cloud over the state dueto slash-and-burn in the North, Northeast and Midwest of Brazil,

K.C.T.d. Brito et al. / Ecotoxicology and Environmental Safety 94 (2013) 14–20 17

and Bolivia (INPE, 2012). During this period the quality of air in themetropolitan Region of Porto Alegre was found to have deterio-rated. As to the mean obtained from the PM2.5 concentration atSite 1 and Site 2, 35 and 49 mg/m3 respectively, the August/2010pools clearly presented higher levels than recommended by WHO(2006), where the daily mean is 25 mg/m3. However, at Site 2 in2010, a place influenced by mixed urban-industrial atmosphericpollution, with vehicular traffic considered moderate for the Cityand near to industries, 80% of the constituents of the poolssurpassed the recommended limit, reaching concentrationsgreater than three times the daily level advocated, during theslash-and-burn period. During this period, also, the mean tem-perature was colder (17 1C) than during the air sampling period of2011 (25 1C), and this may contribute to a greater effect ofpollution as described in the literature (Ducatti and Vargas,2003; Pereira et al., 2010).

In the Salmonella microsuspension assay, all samples presentedmutagenic responses with or without metabolization (Table 1).The differences in the mutagenicity responses observed in thetreatments and strains used in this study show the complexity ofchemical interactions of organic extract of the atmospheric parti-culate material. The highest values in rev/mg of sample were foundat Site 1, in the September/2010 pool (−S9) and in March/2011(+S9). At Site 2, the highest mutagenicity indexes occurred in theMarch/2011 pool, but for mutagens acting directly (–S9) orindirectly (+S9). The responses expressed in rev/m3 at both sitesevaluated, the highest mutagenicity values (−/+S9) were found inAugust/2010, with higher values at Site 2. In previous studies,several authors evaluated atmospheric mutagenicity beginningwith PM10, using the Salmonella microsuspension assay (TA98 –

S9), during winter/2007 (15 and 11 rev/m3), spring/2007 (9 and13 rev/m3) and summer/2008 (7 and 9 rev/m3), at Site 1 and Site2 respectively (Käffer et al., 2012). Compared to the present resultsin PM2.5 in 2010 and 2011, the mutagenicity values increased inrev/m3 both in winter (1.8 and 4.8-fold), and in spring (1.8 and 1.5-fold), at both Sites respectively. In summer the indexes weresimilar. These results raised suspicion of synergy at Site 2 fromdifferent anthropic sources of pollutants (urban-industrial), and,again, reflect the influence of atmospheric pollutants from slash-and-burn in compromising air quality. All the cytotoxic samples

Table 1Mutagenic and cytotoxic activity of the organic extracts of air samples (pools filters PM2.Brazil.

Sitesa Samples (pools) TA98b

Revertants/mg7SD

−S9 +S9

Site 1 August 2010 5.070.37 4.170.53September 2010 17.670.90 7.870.56October 2010 5.170.54 3.370.44January 2011 1.470.15 1.970.23February 2011 4.370.45 4.070.47March 2011 5.470.70 8.570.31

Site 2 August 2010 6.170.51 3.670.64September 2010 8.470.46 3.170.18October 2010 5.870.58 4.470.55January 2011 1.370.18 1.670.19February 2011 7.070.52 7.870.39March 2011 18.971.64 30.371.01

a Site 1: Jardim Botânico district; Site 2: Anchieta district.b −S9, +S9¼absence, presence of S9 mix; Negative control (5 mL/plate DMSO): 28.87

controls (0.05 mg/plate 4NQO): 775.67141.23 revertants/plate (TA98 –S9), (1 mg/plate 2deviation (SD).

c First dosage cytotoxic (mg/plate); -¼non-cytotoxic samples in the presence of leasd Bernstein model in SALANAL software (Salmonella Assay Analysis, version 1.0, Inte

presented a Bernstein regression model in SALANAL software(Table 1).

Several authors recognize the contribution of nitrocompounds todirect frameshift mutagenicity for urban atmospheric samples,associating this capacity with the presence of PAHs derivatives, suchas mono or dinitro PAHs (Claxton et al., 2004; Ducatti and Vargas,2003; Vargas et al., 1998). These classes of mutagenic compounds canbe identified through Salmonella assay by using specific strains suchas YG1021 (pYG216) and YG1024 (pYG219), which overexpress highlyactive enzymes with a high sensitivity to nitrocompounds such asnitroarenes or dinitrorenes, hydroxylamines and aromatic amines,respectively (Watanabe et al., 1989). All extracts evaluated showedhigher mutagenicity when the strains with overproduction ofnitroreductase enzymes and O-acetyltransferases compared to theirparental line (TA98) were used, indicating the participation of nitro-PAHs in the direct mutagenic responses, as well as the presence ofhydroxyalamine compounds detected by strain YG1024 (Fig. 2). Ingeneral, a difference is seen in the nitrocompound composition at thetwo sites evaluated: at Site 1, a mixture of mono and dinitroarenes, inwhich the response in YG1024 predominated in September 2010.YG1024 presents greater sensitivity to dinitroarenes and hydroxyla-mines; at Site 2, the predominance of the response in strain YG1021,which presents a higher sensitivity for mononitroarenes. Previousstudies detected a similar response at the same site in the urbanregion of Porto Alegre (Site 1), during the hot season, when there ismore marked mutagenic activity due to the presence of mono anddinitroarene compounds in different sizes of atmospheric particlessuch as PTS and PM10 (Vargas et al., 2011). Another study also foundthat the main source of atmospheric pollution in these Porto Alegrestudy areas was the presence of nitro-PAHs (Käffer et al., 2012).

Other evidence of nitro-PAHs at the sites was that in severalsamples mutagenicity, evaluated by the Salmonella microsuspen-sion method, diminished or was not modified in the presence of S9mix, indicating the predominance of mutagens that act directly onthe organic extracts of PM2.5 (Table 1 and Fig. 3). However, wemust also take into account the relevant participation of thepresence of PAHs, evidenced in the mutagenicity with metaboliza-tion, in the contribution of airborne pollution at both sitesevaluated in this study and in previous results (Ducatti andVargas, 2003; Käffer et al., 2012). Moreover, it is important to

5), evaluated by Salmonella/microsome assay in two sites from Porto Alegre City, RS,

Cytotoxicityc

Revertants/m37SD TA98

−S9 +S9 −S9 +S9

26.671.97 21.872.82 – –

6.570.33 2.970.21 20d –d

24.072.55 15.472.08 40d –

1.770.18 2.370.28 – –d

5.470.56 4.970.58 – –d

4.170.53 6.470.23 40d –

53.474.47 31.575.61 – –

19.371.05 7.270.41 – –

19.971.97 15.171.87 – d –

3.470.48 4.370.51 – –

8.570.63 9.570.48 – –

10.670.92 17.070.57 40d 10d

7.74 revertants/plate (TA98 –S9), 38.0712.56 revertants/plate (TA98 +S9); PositiveAF): 326.37135.84 revertants/plate (TA98 +S9). The values are mean7standard

t one dosage. Cell survival is less than 60 percent of negative control.grated Laboratory Systems of Research Triangle Institute, RTP, NC, USA).

0

10

20

30

40

50

60

0

10

20

30

40

50

60

August 2010 October 2010

September 2010

February 2011

March 2011 January 2011

µg/m

3

Rev

erta

nts/m

3

TA98 -S9TA98 +S9PM2.5

0

10

20

30

40

50

60

0

10

20

30

40

50

60

August 2010 October 2010

September 2010

March 2011 January 2011

February 2011

µg/m

3

Rev

erta

nts/m

3

TA98 -S9TA98 +S9PM2.5

SITE 1 SITE 2

Fig. 3. Mutagenic activity of atmospheric organic extracts (PM2.5) evaluated by the microsuspension assay in Salmonella/microsome, of the pools of filters in the JardimBotânico (Site 1) and Anchieta (Site 2) districts in the city of Porto Alegre, RS, Brazil, for strains TA98 (revertants/m3), organized by mean concentration of PM2.5 (mg/m3). –S9+S9, absence and presence of metabolization fraction (S9 mix).

020406080

100120140160180200

August 2010 September 2010

October 2010 January 2011 February 2011 March 2011

Rev

erta

nts/

µg

TA98 YG1021 YG1024

020406080

100120140160180200

August 2010 September 2010

October 2010 January 2011 February 2011 March 2011

Rev

erta

nts/

µg

TA98 YG1021 YG1024

SITE 1 SITE 2

Fig. 2. Mutagenic activity of the atmospheric organic extracts (PM2.5) evaluated by the microsuspension assay in Salmonella/microsome, of the pools of filters in the JardimBotânico (Site 1) and Anchieta (Site 2) districts in the city of Porto Alegre, RS, Brazil for strains TA98, YG1021 and YG1024 in the absence of S9 fractions (–S9). Organic extractsevaluated at concentrations of 1.25, 2.5, 5, 10, 20 and 40 mg/plate in duplicate. Negative control (5 mL/plate DMSO): 28.877.74 revertants/plate (TA98 –S9), 79.7746.24revertants/plate (YG1021), 56.2760.35 revertants/plate (YG1024); Positive controls: 4NQO (0.05 mg/plate): 775.67141.23 revertants/plate (TA98 –S9), 2NF (0.15 mg/plate):6075.271786.11 revertants/plate (YG1021), 4181.371538.02 revertants/plate (YG1024). The values are mean7standard deviation (SD).

Table 2Number of tetrads analyzed and frequency of micronuclei (MN %) obtained for T.pallida in chronic exposure (in situ) during the period from August to December2010, in the Jardim Botânico district (Site 1).

Samples Number oftetrads analyzed

Micronuclei/100 tetrads7S.D. (MN %) a

Weekly analysisNegative controlb 900 2.670.514th October 900 3.070.582nd November 2100 2.371.603rd November 1500 3.471.094th November 2700 3.770.821st December 3000 2.170.925th December 3000 4.573.08

Monthly analysisNegative Controlb 900 2.670.51October 900 3.070.58November 6300 3.271.29December 6000 3.372.53

a Tukey test used for weekly (p¼0.076) and monthly (p¼0.929) analysis.b Negative Control: Tradescantia pallida cultivated in flowerpots containing

commercial soil in a non-polluted area.

K.C.T.d. Brito et al. / Ecotoxicology and Environmental Safety 94 (2013) 14–2018

highlight the diminished air quality in Porto Alegre, especially atSite 1, considered an area of reference with the absence ofmutagenicity in previous atmospheric studies (Vargas et al.,1998) which, over time, has shown an increase in these valuesrelated to the greater number of vehicles and airborne pollutionpeaks at the site (Käffer et al., 2012; Vargas, 2003; Vargas et al.,2011).

The responses obtained, both in the urban area (Site 1) and inthe urban-industrial one (Site 2), indicate a clear, directly propor-tional tendency between the PM2.5 concentration and the intensityof mutagenic response, especially during the coldest period eval-uated (Fig. 3). This relationship was not found in previous studies atthese sites and in other urban areas, under petrochemical and oilrefinery influence, evaluating TSP and PM10 (Coronas et al., 2009;Ducatti and Vargas, 2003; Pereira et al., 2010). This shows theimportance of studies using PM2.5, since the capacity to penetratedeeply into the respiratory tract, increasing its mutagenic action, isdirectly related to the diminished diameter of airborne particles(Claxton et al., 2004; Pagano et al., 1996; Vargas et al., 2011).

The Trad-MN assay is considered the most sensitive mutageni-city test system using plants to detect environmental genotoxicpollutants. This assay is often used to detect the genotoxicity ofdifferent samples, especially atmospheric ones (Carvalho-Oliveiraet al., 2005; Klumpp et al., 2006; Mielli et al., 2009; Mišík et al.,2011; Pereira et al., 2013; Suyama et al., 2002). Table 2 shows themicronuclei frequency (MN %) found in young inflorescences of T.pallida exposed chronically at Site 1, during the period in whichthe mutagenicity of airborne organic extracts was evaluated usingbacteria. Our results were not statistically significant in theanalysis of weekly MN % (p¼0.076). The results of the monthlyMN % analyses (Table 2) were also not statistically significant

(p¼0.929). The mutagenicity of organic extracts of some indivi-dual filters collected on the same day as the young inflorescencesof T. pallida were removed from Site 1 (second and third weeks ofNovember/2010 and first week of December/2010) was also eval-uated using the microsuspension assay in Salmonella. All thesesamples showed a positive mutagenic response in rev/mg, of5.970.27 and 5.070.36 in the second week of November; in thethird week of November, it was 3.170.61 and 3.070.38; and in the

C-

C- DM

SO

Site 1A

ug

Site 1

Sep

Site 1

Oct

Site 2

Aug

Site 2

Sep

Site 2

Oct0

10

20

30

TAIL

INTE

NSI

TY

***

*

C-

C- DM

SO

Site 1

Aug

Site 1

Sep

Site 1O

ct

Site 2

Aug

Site 2

Sep

Site 2

Oct0

5

10

15

20

25

TAIL

INTE

NSI

TY

***

Fig. 4. Genotoxic responses obtained by the Comet assay, based on the intensity ofthe tail observed in V79 cells (a) and in human peripheral blood (b) exposed toorganic extracts of air in the Jardim Botânico (Site 1) and Anchieta (Site 2) districts,frommonthly pools of August, September and October 2010. C-, C- DMSO: NegativeControls. *, ***: Statistically significant responses where *Po0.05 and ***Po0.001.

K.C.T.d. Brito et al. / Ecotoxicology and Environmental Safety 94 (2013) 14–20 19

first week of December 4.570.37 and 3.870.31, in the absence andpresence of metabolization, respectively. It has been suggested bysome authors that Trad-MN should be used in situ only to detect hotspots with mutagenic/carcinogenic airborne pollution in urbanareas with intense traffic on small surfaces, such as tunnels(Klumpp et al., 2006; Mišík et al., 2011). According to some authors,results were always negative in situ monitoring of urban areas withassay Trad-MN (Monarca et al., 1999). Since the results found inplants exposed in situ are highly variable, and the statistical resultsin our study are non-significant (high SD), with low discriminatorypower, airborne monitoring techniques are not appropriate. It isnecessary to standardize the culture and exposure programs toavoid the adverse influence of environmental conditions (Klumppet al., 2006; Mielli et al., 2009; Pereira et al., 2013).

The evaluation of genotoxicity (MN %) via the Trad-MN assay inorganic extracts of PM2.5 at the same sites in acute exposure assays(short duration) resulted in no significant differences in theresponses for the concentration of 1 mg/L with MN % at Site 1(6.475.60) and Site 2 (9.174.66) relative to negative controls (tapwater 5.671.95 and DMSO 3% 6.171.54; positive control8.271.81). Some studies found genotoxicity in the Trad-MN assay,both in situ and in PM10 extracts, only at sites very close to intensevehicular traffic (Klumpp et al., 2006; Monarca et al., 1999). Positiveresponses were compared for mutagenicity of airborne extractsPM10 in the Trad-MN and Salmonella/microsome assays, andsignificance was only detected for MN in the samples that pre-sented high mutagenicity values, and close to places with intensetraffic, which did not occur in our study (Monarca et al., 1999).

The results observed using V79 cells are shown in Fig. 4(a), andthose for human peripheral blood in Fig. 4(b), referring to damageevaluated by the Comet assay. A positive genotoxic responseprovoked by samples from Site 1 was found, in September instrain V79 and in October in both cellular systems. In themicronucleus test using V79, no positive genotoxic response wasfound in any of the samples analyzed. In the plating efficiency testonly the sample from Site 1, in August, was cytotoxic.

The responses using the Comet test, alkaline version, refer todamage caused by breaks in the DNA chain, alkali-labile sites andresults of incomplete repair. The two cellular systems used wereconcordant in 83% of the responses. The V79 cells were moresensitive to detect damage with 33% of positive responses, asagainst only 17% in human peripheral blood. The absence ofresponses in the micronucleus assay, in V79, at the concentrationevaluated, could be because the damage found by the Comet assaydoes not correspond to clastogenic events suggesting genotoxicitythrough other types of damage or incomplete repair. It should alsobe observed that in the MN-V79 assay, the final concentration ofextract evaluated was smaller. Evaluating these results of cellulargenotoxicity in mammals, compared to those obtained in theSalmonella/microsome assay, we find a certain concordance ofdata observed at Site 1, in September/October. However, theseassays were not sensitive to evaluate the compounds present atSite 2. The specific sensitivity of the Salmonella/microsome assayto detect the intensity of special events that have occurred, such asslash-and-burn, should also be underscored.

4. Conclusions

The quality of air in Porto Alegre, based on studies in thepresence of genotoxins, is becoming worse as the urban areagrows. This is reflected in the increased flow of vehicles amongother factors as a consequence of pollution. Methodologies toinvestigate atmospheric genotoxicity with molecular assays, plantsand cell cultures are an important tool to obtain early geneticdamage responses for possible mitigating actions before the

population health is compromised. However, they must be care-fully interpreted based on multiple factors.

The fact that no genotoxicity was detected in our work by Trad-MN assay, at sites or in environmental samples from polluted areasdetected using other biomarkers, suggests the need for a carefulevaluation of the biomonitoring of genotoxic compounds withplants. In assays with T. pallida, greater control of environmentalconditions is indicated, with strict standardization to obtain theplants used in assays as well as further information and adaptationof the statistical analyses so that no adverse effects will occurmaking it impossible to perform or interpret the assays.

Genotoxic assays with mammalian cell lines were partiallysensitive to detect genetic damage concerning places, air pollutionevents and type of environmental sample evaluated.

The Salmonella microsuspension assay proved sensitive todetect and identify different classes of airborne mutagenic com-pounds present in PM2.5 of Porto Alegre, even when the concen-trations were within the recommended air quality limit. Sincemutagenic events occur primarily in carcinogenesis provoked bygenotoxic compounds, it is relevant to measure atmosphericquality in fine particles using the Salmonella/microsome assay.

Acknowledgements

The authors are grateful to FEPAM's air sampling and qualityteam, the Biologist Eliane Tigre PhD (USP) for her help in analysesof Trad-MN, Luciana F. Santana (Scientific Initiation), GeographerEliana Sarmento and Conselho Nacional de Desenvolvimento

K.C.T.d. Brito et al. / Ecotoxicology and Environmental Safety 94 (2013) 14–2020

Científico e Tecnológico (CNPq), for the Post-Doctoral scholarship(K.C.T. de Brito). This work was supported by CNPq and FEPAM.

Appendix A. Supporting information

Supplementary data associated with this article can be found inthe online version at http://dx.doi.org/10.1016/j.ecoenv.2013.04.014.

References

ABNT, Associação Brasileira de Normas Técnicas, 1995. Material particulado emsuspensão na atmosfera, Determinação da concentração de partículas inaláveispelo método do amostrador de grande volume acoplado a um separadorinercial de partículas. NBR 13412, Particulate matter as PM10 in the atmo-sphere, Determination of the concentration of inhalable particles by themethod of high volume sampling with an inlet for inertial particle separation,Method of test. June 1995, 8 p.

Alves, N.O., Loureiro, A.L.M., Santos, F.C., Nascimento, K.H., Dallacort, R., Vasconcel-los, P.C., Hacom, S.S., Artaxo, P., Medeiros, S.R.B., 2011. Genotoxicity andcomposition of particulate matter from biomass burning in the easternBrazilian Amazon region. Ecotoxicol. Environ. Saf. 74, 1427–1433.

Bernstein, L., Kaldor, J., McCann, J., Pike, M.C., 1982. An empirical approach to thestatistical analysis of mutagenesis data from the Salmonella test. Mutat. Res. 97,267–281.

Brazil, Conselho Nacional do Meio Ambiente, Resolução CONAMA n◦ 003, 1990.Padrões Nacionais de Qualidade do Ar, Diário Oficial da República Federativa doBrasil, Brasília, pp. 15937–15939.

Cardozo, T.R., Rosa, D.P., Feiden, I.R., Rocha, J.A.V., Oliveira, N.C.A., Pereira, T.S.P.,Pastoriza, T.F., Marques, D.M., Lemos, C.T., Terra, N.R., Vargas, V.M.F., 2006.Genotoxicity and toxicity assessment in urban hydrographic basins. Mutat. Res.603 (1), 83–96.

Carvalho-Oliveira, R., Pozo, R.M.K., Lobo, D.J.A., Lichtenfels, A.J.F.C., Martins-Junior, H.A.,Bustilho, J.O.W.V., Saiki, M., Sato, I.M., Saldiva, P.H.N., 2005. Diesel emissionssignificantly influence composition and mutagenicity of ambient particles: a casestudy in São Paulo, Brazil. Environ. Res. 98, 1–7.

Claxton, L.D., Creason, J., Leroux, B., Agurell, E., Bagley, S., Bryant, D.W., Courtois, Y.A.,Douglas, G., Clare, C.B., Goto, S., Quillardet, P., Jagannath, D.R., Kataoka, K., Mohn, G.,Nielsen, P.A., Ong, T., Pederson, T.C., Shimizu, H., Nylund, L., Tokiwa, H., Vink, G.J.,Wang, Y., Warshawsky, D., 1992. Results of the IPCS collaborative study oncomplex mixtures. Mutat. Res. 276, 23–32.

Claxton, L.D., Matthews, P., Warren, S., 2004. The genotoxicity of ambient outdoorair, a review: Salmonella mutagenicity. Mutat. Res. 567, 347–399.

Collins, A.R., 2004. The comet assay for DNA damage and repair: principles,applications, and limitations. Mol. Biotechnol. 26, 249–261.

Coronas, M.V., Pereira, T.S., Rocha, J.A.V., Lemos, A.T., Fachel, J.M.G., Salvadori, D.M.F.,Vargas, V.M.F., 2009. Genetic biomonitoring of a urban population exposed tomutagenic airborne pollutants. Environ. Int. 35, 1023–1029.

De Martinis, B.S., Kado, N.Y., Carvalho, L.R.F., Okamoto, R.A., Gundel, L.A., 1999.Genotoxicity of fractionated organic material in airborne particles from SãoPaulo, Brazil. Mutat. Res. 446, 83–94.

Ducatti, A., Vargas, V.M.F., 2003. Mutagenic activity of airborne particulate matteras an indicative measure of atmospheric pollution. Mutat. Res. 540, 67–77.

Federal Register, 1989. Guidelines establishing test procedures for the analysis ofpollutants under the clean water act, Federal Register, USA, Part III: EPA,December 4, vol. 54(231), pp. 50216–50223.

Fenech, M., 1993. The cytokinesis-block micronucleus technique: a detaileddescription of the method and its application to genotoxicity studies in humanpopulation. Mutat. Res 285, 35–44.

IARC, International Agency for Research on Cancer, Monographs, Supplement, 2012.Available from: ⟨http://monographs.iarc.fr/ENG/Classification/index.php⟩.

INMET, Instituto Nacional de Meteorologia, 2011. Available from: ⟨http://www.inmet.gov.br⟩ (accessed 12.02.11).

INPE, Instituto Nacional de Pesquisas Espaciais, 2012. Available from: ⟨http://www.inpe.br/queimadas⟩ (accessed 10.05.12).

Kado, N.Y., Langley, D., Eisenstadt, E., 1983. A simple modification of the Salmonellaliquid-incubation assay. Increased sensitivity for detecting mutagens in humanurine. Mutat. Res. 121, 25–32.

Käffer, M.I., Lemos, A.T., Apel, M.A., Rocha, J.A.V., Martins, S.M.A., Vargas, V.M.F.,2012. Use of bioindicators to evaluate air quality and genotoxic compounds inan urban environment in southern Brazil. Environ. Pollut. 163, 24–31.

Klumpp, A., Ansel, W., Klumpp, G., Calatayud, V., Garrec, J.P., He, S., Peñuelas, J.,Ribas, P., Ro-Poulsen, H., Rasmussen, S., Sanz, M.J., Vergne, P., 2006. Tradescantia

micronucleus test indicates genotoxic potential of traffic emissions in Europeancities. Environ. Pollut. 139, 515–522.

Lemos, C.T., Erdtmann, B., 2000. Cytogenetic evaluation of aquatic genotoxicity inhuman cultured lymphocytes. Mutat. Res. 467, 1–9.

Lemos, A.O., Oliveira, N.C.D., Lemos, C.T., 2011. In vitro micronuclei tests to evaluatethe genotoxicity of surface water under the influence of tanneries. Toxicol. inVitro. 25, 761–766.

Lemos, A.T., Coronas, M.V., Rocha, J.A., Vargas, V.M.F., 2012. Mutagenicity ofparticulate matter fractions in areas under the impact of urban and industrialactivities. Chemosphere 89 (9), 1126–1134.

Lewtas, J., 2007. Air pollution combustion emissions: characterization of causativeagents and mechanisms associated with cancer, reproductive, and cardiovas-cular effects. Mutat. Res. 636, 95–133.

Ma, T.H., Cabrera, G.L., Chen, R., Gill, B.S., Sandhu, S.S., Vandenberg, A.L., Salamone,M.F., 1994. Tradescantia micronucleus bioassay. Mutat. Res. 310, 221–230.

Ma, T.H., Sparrow, A.H., Schairer, L.A., Nauman, A.F., 1978. Effect of 1,2-dibroethane(DBE) on meiotic chromosomes of Tradescantia. Mutat. Res. 58, 251–258.

Maron, D.M., Ames, B.N., 1983. Revised methods for the Salmonella mutagenicitytest. Mutat. Res. 113, 173–215.

Marvin, C.H., Hewitt, LM., 2007. Analytical methods in bioassay-directed investiga-tions of mutagenicity of air particulate material. Mutat. Res. 636, 4–35.

Mielli, A.C., Matta, M.E.M., Nersesyan, A., Saldiva, P.H.N., Umbuzeiro, G.A., 2009.Evaluation of the genotoxicity of treated urban sludge in the Tradescantiamicronucleus assay. Mutat. Res. 672 (1), 51–54.

Mišík, M., Ma, T-H., Nersesyan, A., Monarca, S., Kim, J.K., Knasmueller, S., 2011.Micronucleus assays with Tradescantia pollen tetrads: an update. Mutagenesis26 (1), 215–221.

Monarca, S., Feretti, D., Zanardini, A., Falistocco, E., Nardi, G., 1999. Monitoring ofmutagens in urban air samples. Mutat. Res. 426 (2), 189–192.

Nunes, E.A., Lemos, C.T., Gavronski, L., Moreira, T.N., Oliveira, N.C.D., Silva, J., 2011.Genotoxic assessment on river water using different biological systems.Chemosphere 84 (1), 47–53.

OECD, Organisation for Economic Co-operation and Development, 1997. Guidelinesfor the Testing of Chemicals/Section 4: Health Effects, Bacterial ReverseMutation Test: No. 471 Adopted 21st July.

OECD, Organisation for Economic Co-operation and Development, 2010. Guidelinesfor the Testing of Chemicals In Vitro Mammalian Cell Micronucleus Test: No.487 Adopted 22st July.

Pagano, P., Zaiacomo, T., Scaecella, E., Bruni, S.M., 1996. Calamosca, Mutagenicactivity of total and particle-sized fractions of urban particulate matter.Environ. Sci. Technol. 30 (12), 3512–3516.

Pereira, B.B., De Campos Júnior, E.O., Morelli, S., 2013. In situ biomonitoring of thegenotoxic effects of vehicular pollution in Uberlândia, Brazil, using a Trades-cantia micronucleus assay. Ecotoxicol. Environ. Saf. 87, 17–22.

Pereira, T.S., Gotorb, G.N., Beltramia, L.S., Nollab, C.G., Rocha, J.A.V., Brotob, F.P.,Comellasb, L.R., Vargas, V.M.F., 2010. Salmonella mutagenicity assessment ofairborne particulate matter collected from urban areas of Rio Grande do SulState, Brazil, differing in anthropogenic influences and polycyclic aromatichydrocarbon levels. Mutat. Res. 702, 78–85.

Rojas, E., Lopez, M.C., Valverde, M., 1999. Single cell gel electrophoresis assay:methodology and applications. J. Chromatogr. B 722, 225–254.

Singh, N.P., McCoy, M.T., Tice, R.R., Schneider, E.L., 1988. A simple technique forquantification of low levels of DNA damage in individual cells. Exp. Cell Res.175, 184–191.

Suyama, F., Guimarães, E.T., Lobo, D.J.A., Rodrigues, G.S., Domingos, M., Alves, E.S.,Carvalho, H.A., Saldiva, P.H.N., 2002. Pollen mother cells of Tradescantia Clone#4430 and Tradescantia pallida var. purpurea are equally sensitive to theclastogenic effects of X-rays. Braz. J. Med. Biol. Res. 35, 127–129.

Vargas, V.M.F., Horn, R.C., Guidobono, R.R., Mittelstaedt, A.B., Azevedo, I.G., 1998.Mutagenic activity of airborne particulate matter from the urban area of PortoAlegre. Genet. Mol. Biol. 21, 247–253.

Vargas, V.M.F., 2003. Mutagenic activity as a parameter to assess ambient air qualityfor protection of the environment and human health. Mutat. Res. 544, 313–319.

Vargas, V.M.F., Brito, K.C.T., Coronas, M.V., 2011. Genetic biomarkers applied toenvironmental air quality: ecological and human health aspects. In: Mazzeo, N.A. (Ed.), Air Quality Monitoring, Assessment and Management. Intechopen access publisher, intechweb.org, http://www.intechopen.com/books/air-quality-monitoring-assessment-and-management.

Watanabe, M., IshidateJr., M., Nohmi, M.T., 1989. A sensitive method for thedetection of mutagenic nitroarenes: construction of nitroreductase-overproducing derivates of Salmonella typhimurium strains TA98 and TA100.Mutat. Res. 216, 211–220.

WHO, World Health Organization. Air quality guidelines for particulate matter,ozone, nitrogen dioxide and sulfur dioxide. Global update 2005, Geneva, 2006.

WHO, Wolrd Health Organization, World Cancer Report, 2008. InternationalAgency for Research on Cancer. Lyon, 2009.

Zar, J.H., 1996. Biostatistical Analysis, third ed. Prentice-Hall, Englewood Cliffs, NJ.