Characterization of urban aerosol: seasonal variation of mutagenicity and genotoxicity of PM2.5, PM1 and semi-volatile organic compounds

Mutation Research 809 (2016) 16–23

Contents lists available at ScienceDirect

Mutation Research/Genetic Toxicology and Environmental Mutagenesis journal homepage: www.elsevier.com/locate/gentox Community address: www.elsevier.com/locate/mutres

Characterization of urban aerosol: seasonal variation of mutagenicity and genotoxicity of PM2.5 , PM1 and semi-volatile organic compounds Clara Bocchi ∗ , Cristina Bazzini, Federica Fontana, Giancarlo Pinto, Anna Martino, Francesca Cassoni Agenzia Regionale Prevenzione Ambiente Energia, Emilia-Romagna, Sezione di Parma, Italy

a r t i c l e

i n f o

Article history: Received 17 May 2016 Received in revised form 19 July 2016 Accepted 26 July 2016 Available online 30 July 2016 Keywords: Mutagenicity Genotoxicity Urban PM2.5 PM1 gaseous phase Salmonella test Comet assay Micronucleus assay Chemical composition

a b s t r a c t Urban particulate matter (PM) is an environmental public health concern as it has been classified by the IARC as carcinogenic to humans (group 1) and it’s well known that pollutants are more associated with the finest fractions of PM. In this study we characterize urban aerosol in Bologna, county town of Emilia-Romagna in the north of Italy, collecting PM2.5 , PM1 and semi-volatile organic compounds using polyurethane foam. Samples were collected in three different seasons (winter, summer and autumn) and were extracted with acetone. On these three fractions we assessed mutagenicity using Salmonella reverse mutation test and genotoxicity by alkaline comet assay and micronucleus assay in human lung cancer cell line, A549. Organic extracts were also characterized for alkanes, polycyclic aromatic hydrocarbons (PAHs), nitrated and oxygenated PAHs. We also evaluated associations between the physicochemical parameters of samples and their genotoxicity. The particulate samples, collected in autumn and winter, indicated the presence of both base pair substitution and frameshift mutagens using TA98 and TA100 strains of Salmonella typhimurium and the mutagenicity was more associated with the finest fraction. Enhanced mutagenic response was observed in the absence of enzyme activation. Only a third of comet and a half of micronucleus assays gave positive results that, unlike Salmonella’s ones, are not season-related. These results were compared with environmental chemicals concentrations and we found that Salmonella’s data correlated with PAHs detected on PM filters and with mass concentrations, whereas the DNA damage correlate only with PAHs extracted from polyurethane foams. The use of different assays was sensitive to detect and identify different classes of airborne mutagenic/genotoxic compounds present in aerosol, showing that monitoring air quality using this methodology is relevant. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Chronic exposure to airborne pollutants such as heavy metals, polycyclic aromatic hydrocarbons and particulate matter may cause adverse health effects. Harmful effects of PM are specifically associated with finest particles that can adsorb high concentrations of toxic air pollutants and are easily inhaled into the lungs. Many results indicated that ultrafine particles play an important role as carriers of mutagens into the lung [1]. However, combined effects of fine particles and air pollutants on human health remain unclear [2].

∗ Corresponding author at: Arpae, Via Spalato, 4, 43125 Parma, Italy. E-mail address: [email protected] (C. Bocchi). http://dx.doi.org/10.1016/j.mrgentox.2016.07.007 1383-5718/© 2016 Elsevier B.V. All rights reserved.

Among airborne genotoxic substances, polycyclic aromatic hydrocarbons (PAHs) and their derivatives, oxygenated and nitrated PAHs, have a great importance. PAHs derivatives can be both directly emitted in atmosphere or be byproducts of combustion and photo-oxidation of parent PAHs (oxy-PAHs) or be produced by interaction between PAHs and NOx (nitro-PAHs) [3]. Fuel combustion (e.g. motor vehicle exhaust and residential heating), waste incineration and industrial plants can be the sources of those pollutants that can be both in the gas-phase and be adsorbed on airborne particulate matter, in according to chemical and weather conditions [4]. Many studies demonstrate that organic complex mixtures of urban PM have mutagenic and genotoxic activities in different short-term tests on bacteria, mammalian cells cultured in vitro and laboratory animals [5–7]and the mutagenicity of airborne particulate can result from at least 500 identified

C. Bocchi et al. / Mutation Research 809 (2016) 16–23

compounds from various chemical classes [8], but the volatile and semi-volatile compounds, not adsorbed on to the particles, can be genotoxic and mutagenic, as well. So it’s clear that no single technique can explain the complexity of genotoxic effects of air pollutants and a lot of data show that the mutagenic/genotoxic properties of a complex mixture cannot be accurately determined on the sole basis of its composition in chemical compounds, therefore the use of complementary approaches is highly recommended [9]. Our previous report [10] confirmed the mutagenicity of PM2.5 organic extracts detected with Salmonella reverse mutation test, sampled in eight cities of Emilia-Romagna region with a seasonal trend that featured higher levels during the autumn-winter period and lower levels in spring-summer. In this research we investigate the DNA-damaging ability of chemically characterized organic-soluble extracts prepared from two particle fractions (PM2.5 and PM1 ) and from polyurethane foams (PUF) to characterize the respirable fraction of urban aerosol. Organic extracts (OE) of PM and PUF were tested for mutagenic activities using bacterial reverse mutation test with Salmonella typhimurium, TA98 and TA100 strains, to detect point mutations (base substitution and frameshift mutations). To analyse their genotoxicity human lung alveolar epithelial cells (A549), as a model system for the potential inhalation health effects, were exposed to OE and the alkaline comet assay and the cytochalasin block micronucleus test were conducted to identify DNA strand breaks and chromosomal changes, such as acentric chromosome and chromatid fragments or chromosome lagging in anaphase. Extracts were chemically characterized for the presence of PAHs, nitro-PAHs, oxy-PAHs and alkanes, associations between those chemical parameters and genotoxicity were also evaluated. 2. Materials and methods 2.1. Sampling site Samples were collected in an urban background site in Bologna, a big town in the Po Valley near Apennines mountains (northern Italy), three times a year (winter, summer and autumn) over a three-year period. PM and gaseous phase were sampled for at least 20 to a maximum of 31 days. Samples were collected, 24 h daily, using AMS Analitica high volume samplers (Air Flow PM2.5-HVS), PM2.5 was collected at a flow rate of 500 L/minute, glass filters for PM1 and PUF for gaseous phase were in the same sampler, with customized sampling head, operating at a flow of 200 L/min. Particles were collected on Nupore Filtration Systems glass-fiber filter paper: 150 mm diameter for PM2.5 and 102 mm for PM1. Semi-volatile compounds were sampled for the first three campaigns with SKC polyurethane foams and for the last five using Restek corporation foams. The amount of air particulate matter was determined weighing filters before and after air sampling. 2.2. Samples extraction Samples were extracted with acetone using a Soxhlet extractor. PUFs and filters, of each season, were pooled together to obtain three extracts per campaign. Extracts were evaporated using a rotary evaporator and redissolved in dimethyl sulphoxide (DMSO) 50 m3 /mL for Salmonella assay, 1000 m3 /mL for human cell line assays. 2.3. Organic micropollutants analysis The analysis of organic micropollutants were performed in the ARPA laboratory of Ravenna, as a part of the Supersito project activ-

17

ities [11]. PAHs, nitro-, oxy-PAHs and alkanes were detected in the same organic extracts used in genotoxicity assays. OEs were purified using solid-phase extraction cartridges packed with silica gel adsorbent, deactivated to 3% with water [12]. Aliquots of sample were loaded onto the cartridges, which were then eluted in three fractions with different polarity: • linear alkanes were eluted with 25 ml of hexane into the first fraction • PAHs and nitro-PAHs were eluted together with 30 ml of dichloromethane/hexane (1:1) • the last fraction was eluted with 20 ml of dichloromethane and contains oxy-PAHs The three different fractions obtained, collected in glass test tubes, were evaporated in a stream of nitrogen and then transferred into vials for further analysis. 2.3.1. PAHs analysis PAHs were analysed on Thermo Scientific DFS Magnetic Sector GC-HRMS system gas chromatograph equipped with a capillary column and a mass spectrometer high resolution. Isotope labelled (deuterated) PAH standards were used for quantification. The following PAHs were determined: naphtalene, acenaphthylene, acenaphtene, 2,6 dimethylnaphtalene, dibenzofuran, fluorene, 1-methylfluorene, phenanthrene, 9-methylphenanthrene, 1-methylphenanthrene, anthracene, fluoranthene, pyrene, retene, benzo(g,h,i)fluoranthene, cyclopenta(cd)pyrene, triphenylene, benzo(a)anthracene, chrysene, 1-methylchrysene, benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(e)pyrene, benzo(a)pyrene, perylene, indeno(1,2,3-cd)pyrene, dibenz(a,h + a,c)anthracene, dibenzo(a,i)phenanthrene, benzo(ghi)perylene; dibenzo(a,l)pyrene, dibenzo(a,e)fluoranthene, dibenzo(a,e)pyrene, coronene, dibenzo(a,i)pyrene, dibenzo(a,h)pyrene. PAHs underlined were determined from winter 2013 campaign on. Comparisons between PAHs concentrations and PM extracts genotoxicity were made considering PAHs from fluoranthene to dibenzo(a,h)pyrene, total PAHs were used for comparisons with PUF extracts genotoxicity. 2.3.2. Nitro- and oxy-PAHs analysis The analytical determination of nitro- and oxy- PAHs was carried out by high resolution gas chromatography interfaced to a mass spectrometer triple quadrupole at low resolution (HRGC/GC/MS/MS). They were analysed on a Thermo Scientific HRGC/MS/MS gas chromatograph equipped with a capillary column and a mass spectrometer. The quantitative analysis was performed with an external calibration, using a mixture of reference standards. Nitro-PAHs and oxy-PAHs detected: 2-nitronaphtalene, xanthone, 1H-phenalene-1-one, 1,8-naphtalic anhydride, 9nitroanthracene, 9-nitrophenanthrene, 2 + 3-nitrofluoranthene, benz(a)anthracene-7,12-dione, 1-nitropyrene, 7nitrobenz(a)anthracene, 6-nitrochrysene, 3-nitrobenzanthrone, 1,6-dinitropyrene, 1,8-dinitropyrene, 6-nitrobenzo(a)pyrene. Organic extracts genotoxicity was compared with the total of nitro- and oxy-PAHs measured. 2.3.3. Alkanes analysis The n-alkanes were analysed on an Agilent 6890 N gas chromatograph equipped with a capillary column and a mass spectrometer (MSD, Agilent 5975). The straight chained n-alkanes detected ranged from n-decane (C10) to n-tetracontane (C40).

18

C. Bocchi et al. / Mutation Research 809 (2016) 16–23

2.4. Salmonella mutagenicity assay Mutagenicity was studied using Salmonella typhimurium, strains TA98 and TA100, both with and without an Aroclor-induced ratliver homogenate activation − S9 mix − [13]; three doses of each sample were tested in triplicate: 2, 4, 8 cubic meter of air equivalent per plate. DMSO was used as a negative control, whereas standard mutagens were used as positive controls: sodium azide 1.5 ␮g/plate (TA100 without S9), 2-nitrofluorene 2 ␮g/plate (TA98 without S9) and 2-aminoanthracene 1 ␮g/plate (both strains with S9 mix) according to OECD Guidelines [14]. We applied the twofold rule to judge the positivity (i.e. mutagenicity) of samples [15]. Results are calculated by linear regression analysis of the linear part of the dose-response curves, with a coefficient of determination (R2 ) greater than or equal to 0.6, and are expressed as induced revertants per cubic meter of air (rev/m3 ) and per microgram of airborne particulate mass sampled (rev/␮g).

CB-MN test was carried out according to the OECD guideline [18] and Fenech et al. [19]. The micronuclei were scored in binucleated cells (BN) at 400-fold magnification by one observer, BN were selected according to criteria described by Fenech [19], for each dose the number of micronucleated binucleated cells in 2000 BN (1000 cells per replica) were examined. To rule out cytotoxic effects the cytochalasin block proliferation index (CBPI), that indicates the number of cell cycles per cell during the period of exposure to cytochalasin B, was evaluated according to OECD guideline [18] by assessing 500 cells. Results are calculated by linear regression analysis of dose-response curve (R2 ≥ 0.6) and are expressed as MN/m3 of air sampled. The frequency of binucleated cells with one or more micronuclei in the treated cultures were compared with their respective controls by using the chi-squared test (p < 0.05), to assess a positive result both chi-squared test and coefficients of determination values (R2 ≥ 0.6) were used. 2.6. Statistical analysis

2.5. Cell culture A549 human lung adenocarcinoma type II alveolar epithelial cells were grown in RPMI-1640 medium supplemented with 10% (v/v) fetal bovine serum, 1% penicillin/streptomycin and 1% lglutamine. Cells were incubated in a humidified incubator at +37 ◦ C and 5% CO2 . For the experiments cells were trypsinized at 80% confluency and seeded in a 24-wells culture plates. 2.5.1. Alkaline comet assay (single cell gel electrophoresis) Single cell gel electrophoresis, or the comet assay, was used to detect DNA damage, strand breaks and incomplete excision repair sites, induced either directly or at alkaline-sensitive sites by OE at the level of individual cells. Cells were seeded 2*105 cells/ml per well in 24-wells culture plate and exposed for 24 h to three concentrations of PM1 , PM2.5 and PUF organic extracts: 2.5, 5 and 10 m3 equivalent/ml [16]. As positive control, cells were treated with ethyl methanesulfonate (EMS, 0.1 M). Cell viability was assessed, after incubation, by trypan blue assay, trypan penetrates and dyes only dead cells. Duplicate slides were prepared for each dose and, in each slide, 100 randomly selected nuclei (200 cells per dose) were acquired and analysed with specific image analyser software (Comet Assay IV; Perspective Instruments, Suffolk, UK). Results are expressed as percentage of DNA in the tail (Tail Intensity − %TI) per m3 of air sampled, calculated by linear regression analysis of dose-response curve, with a R2 ≥ 0.6. Significant differences between treatments and controls were assessed with non-parametric statistic using notched box plots. Notched box plots it’s a useful and quick graphic way to display and to see differences among data. The notches display the confidence intervals around the medians, if the notches do not overlap, the difference between two medians is statistically significant at a 0.05 level. The notched box plot formula is: median ± 1.57 x IQR/sqrt(n), where IQR is the interquartile range (25–75 percentile) and n is the number of observations [17]. To assess a positive result both notched box plots and coefficients of determination values (R2 ≥ 0.6) was used. 2.5.2. Cytokinesis block micronucleus test Clastogenicity and aneugenicity of extracts were quantified on human alveolar lung cells using micronucleus assay. Cells were seeded 6*104 cell per well (3*104 cell/ml) in 24-wells culture plate and exposed for 24 h to three concentrations of PM1 , PM2.5 , and PUF organic extracts (4, 8 and 16 m3 equivalent/ml), duplicate wells were prepared for each dose. Mitomycin C 0.6 ␮M was used as a positive control.

Linear regression was used to analyse correlation between chemical parameters and genotoxicity results. Concentrations of PM mass, PAHs, their derivatives and alkanes from the different campaigns were compared with one way ANOVA with Tukey’s post hoc multiple comparisons. Differences between PM1 and PM2.5 concentrations were revealed with Student’s t-test for paired samples. 3. Results 3.1. Airborne particulate matter concentrations Table 1 summarizes the average of PM1 and PM2.5 mass concentration in the different seasons and shows the results of comparisons between particles concentrations achieved with Student’s t-test for paired samples. In both PM fractions the mass concentration is higher in autumn than in summer, PM1 winter values are equal to those of the others two seasons and PM2.5 winter mass concentrations are lower than autumn, but higher than summer ones (ANOVA, with Tukey’s post hoc, p < 0.05). The highest PM concentrations were detected in autumn 2011, followed by winter 2013 and autumn 2012. 3.1.1. PAHs, and their derivatives, concentrations No difference was found among total PAHs concentrations detected in particulate and in gaseous extracts. The same results were found with nitro- + oxy-PAHs concentrations. The highest amount of PAHs was found in PM collected in autumn 2011, instead the highest nitro- + oxy-PAHs concentrations were found in winter 2013 (Fig. 3), the lower concentrations were detected, especially regarding PAHs, in summer campaigns (ANOVA, with Tukey’s post hoc, p < 0.05). 3.1.2. Alkanes concentrations Airborne alkanes can be both natural and anthropogenic as they are emitted in asphalt roofing tar, vehicle exhaust, tyre abrasion, in wood, coal and cigarette combustion or can come from leaves, pollens, bacteria and insects. To discriminate the influence from natural and man-made sources it is possible to use the carbon preference index (CPI) that is the concentration ratio of alkanes with odd and even carbon number: CPI = (odd carbon number alkanes)/(even carbon number alkanes). In general, CPI values > 2.0 indicate predominant biogenic sources, such as leaves waxes of plants, instead emissions from utilization of fossil fuel exhibit CPI values close to 1.0. Urban environments, with large contribution from anthropogenic emissions, generally have CPI ranging from 1.0 to 2.0 [20].

C. Bocchi et al. / Mutation Research 809 (2016) 16–23

19

Table 1 Mean concentration (± standard deviation) of PM1 and PM2.5 in every monitoring campaign, 24 h sampling and p-values of paired sample t-test. Campaign

PM1 (␮g/m3 )

Autumn 2011 (14Nov − 07Dec2011) Summer 2012 (13June − 11July2012) Autumn 2012 (23Oct − 12Nov2012) Winter 2013 (28Jan − 08Feb and 13–27Feb2013) Summer 2013 (06–28May2013) Autumn 2013 (27Sept − 18Oct2013) Winter 2014 (27Jan − 27Feb2014) Summer 2014 (12May − 12June2014)

31.90 14.94 19.97 23.47 7.35 12.19 10.04 8.61

± ± ± ± ± ± ± ±

6.09 2.89 11.47 8.54 2.10 4.94 5.81 3.34

p Student’s t-test

PM2.5 (␮g/m3 )

0.001 0.000 0.282 0.000 0.011 0.012 0.000 0.000

48.62 18.78 25.83 28.67 9.28 18.54 18.34 11.82

± ± ± ± ± ± ± ±

13.76 4.89 11.15 9.25 3.07 10.14 9.94 4.51

Table 2 Values of carbon preference index (CPI) in gaseous and particulate phases in different seasons. Sample

Aut. ‘11

Sum. ‘12

Aut. ‘12

Win. ‘13

Sum. ‘13

Aut. ‘13

Win. ‘14

Sum. ‘14

PM1 PM2.5 PUF

2.0 1.9 0.8

3.3 5.3 0.6

1.5 1.4 0.3

1.4 1.2 1.8

2.1 3.9 2.1

1.5 2.2 0.9

1.6 2.0 0.7

0.6 7.5 1.4

Table 3 Results of Salmonella assays performed before the beginning of this study. Revertants per cubic meter of air sampled (slope of linear regression), induced in Salmonella typhimurium, strains TA98 and TA100 with (+S9) and without metabolic activation, by organic extracts of PM10 and PUF sampled in Piacenza in July 2010. July 2010

TA98

TA98 + S9

TA100

TA100 + S9

PUF PM10

4 0

3 0

0 0

0 0

As shown in Table 2 the CPI values are lower in autumn and winter (average 1.4 for both seasons) and higher in summer when the average CPI value, considering both PM and gaseous aerosol, is 3.0. No difference was found among alkanes concentrations detected in particulate and in gaseous extracts, even in different seasons alkanes concentrations almost stay the same, in fact only autumn 2012 has a higher alkanes concentration than summer 2012 (ANOVA, with Tukey’s post hoc, p < 0.05). 3.2. Salmonella mutagenicity assay No toxicity and/or mutagenicity was found in extracts obtained from filters and polyurethane foams before sampling (blank extracts) using Salmonella assay, but data obtained from PUF sampled in winter and summer 2013 were invalidated, due to a high concentration of organic pollutants in blank extracts that were also toxic towards A549 cells (see paragraph 3.3.2.). No PUF extracts, at any dose, showed a 2-fold or greater increase in mutant frequency as compared to the respective controls; only summer 2012 extract showed a dose-depended response in TA98 strain with metabolic activation (+S9), but with a treated/control ratio of 1.6. These results are in contrast to what was expected for PUF sampled in summer when, owing to the weather conditions (sunlight, UV radiation, temperature), PAHs in the gas phase are significantly higher than those in particle phase [21]. Moreover in a previous reverse mutation test on PM10 and on gaseous phase, collected in summer 2010 in Piacenza, a town in the same region of Bologna, we found that PUF, but not PM, extract induced mutagenicity in TA98 strain of Salmonella (Table 3). Almost every extract of PM collected in autumn and winter exhibited a linear dose–response relationship whether S9 was present or not, in both strains (TA98 and TA100), and PM extract did not show mutagenic activity in summer campaigns. Exceptions are: autumn 2013 and winter 2014, when PM1 OE did not induce revertants in TA100 strain with S9 and, summer 2013 when both

strains revealed direct mutagenicity in both fractions and PM2.5 extract induced revertants also in TA98 + S9. This result could be due to the fact that in summer 2013 filters were collected in May (see Table 1). This seasonal trend with higher mutagenic activities, of PM organic extracts, in autumn-winter and lower in warmer periods of the year is typical for Salmonella reversion assay [10,22]. As shown in Figs. 1 and 2, in both PM fractions, the S9 mix decreased the mutagenicity expressed as revertants per cubic meter and also per microgram air-equivalent (Student’s t-test, paired samples; p < 0.05), this means that the most of substances present in ambient PM can act directly on DNA. Considering both PM sizes there is an association between increased mass per unit volume of air (␮g/m3 ) and the mutagenicity (mean of rev/m3 ) of organic extracts, as coefficients of determination (R2 ) between these two parameters are for PM1 = 0,88 and for PM2.5 = 0,93. Anyway, a higher mutagenic potency does not always correspond to a higher concentration of airborne particles (Fig. 1), this means that the number of rev/m3 , sometimes, could be due more to a specific activity of particles, i.e. a different number of rev/␮g, rather than their atmospheric mass, this is clear especially in PM1 fraction. Comparing samples with mutagenic activity, that is particulate collected in winter and in autumn campaigns, PM1 has a greater number of revertants per microgram than PM2.5 (Student’s t-test, paired samples; p < 0.05), but if we consider mutagenicity on a m3 basis there is no difference between the two fractions. This result underlines and confirms that organic mutagenic substances are more associated with the finest fractions of airborne particles [23,24].

3.2.1. Chemicals and mutagenicity Results of mutagenicity assays in the next graphs are expressed as mean of the slopes of linear regression curves. The possible association between mutagenicity and the chemical components was evaluated by coefficients of linear regression analysis (R2 ) and a significant association was found in PM fractions with total PAHs detected, as R2 ranges from 0.73 to 0.93 (Figs. 1 and 3 and Table 4), but the correlation is similar with and without S9, revealing that there are chemical classes of pollutants that act directly on DNA. Nitro- plus oxy-PAHs, in both particulate fractions, only weakly correlate (R2 are about 0.6) with TA100 strain without S9 (direct activity). Alkanes seem to correlate with mutagenicity induced in PM1 (R2 ranges from 0.50 to 0.73) and only with TA100-S9 in PM2.5 .

C. Bocchi et al. / Mutation Research 809 (2016) 16–23

70

120

60

100

50

80

40

60

30

40

20

rev/m3

140

20

10

Win.'13

TA100

Aut.'13

PUF

PM1

Win.'14

PM µg/m3

PM2,5

PUF

PM1

PM2,5

PUF

PM1

Sum.'13

TA100+S9

PM2,5

0

PUF

PM1

PUF

PM1

Aut.'12

TA98+S9

PM2,5

PUF

PM1

Sum.'12

TA98

PM2,5

PUF

PM1

PM2,5

PUF

PM1

PM2,5

Aut.'11

inv.

PM2,5

inv.

0

PM (µg/m3) alkanes (ng/m3)

20

Sum.'14

total alkanes

Fig. 1. Mutagenicity expressed as number of revertants per m3 of air sampled (slope of linear regression ± confidence interval) induced by OEs in Salmonella typhimurium, strains TA98 and TA100 with (+S9) and without metabolic activation and comparison between revertants and PM concentration (␮g/m3 ) and alkanes concentration (from C10 to C40, ng/m3 ). Inv.: invalidated data.

4

rev/µg

3 2 1

Aut.'11

Win.'13

TA98+S9

Sum.'13

TA100

Aut.'13

Win.'14

PM2,5

PM1

PM2,5

PM1

PM2,5

PM1

PM2,5

PM1

PM1

Aut.'12

TA98

PM2,5

PM1

Sum.'12

PM2,5

PM2,5

PM1

PM2,5

PM1

0

Sum.'14

TA100+S9

140

10

120

8

revertants/m3

100 80

6

60

4

40 2

20

Sum.'12

Aut.'12

Win.'13

.

Sum.'13

mean TA98+S9&TA100+S9

Aut.'13

mutagenic PAHs

Win.'14

PUF

PM1

PM2,5

PUF

PM1

PM2,5

PUF

PM2,5

0 PM1

PUF

PM1

inv

PM2,5

PM1

PM2,5

PUF

PM2,5

PM1

PUF

PM1

PM2,5

PUF

PM1

PM2,5

Aut.'11

mean TA98&TA100

.

PUF

inv

0

nitro-+oxy-PAHs and PAHs (ng/m3)

Fig. 2. Mutagenicity expressed as number of revertants per microgram of particulate (slope of linear regression ± confidence interval) induced in Salmonella typhimurium, strains TA98 and TA100 with (+S9) and without metabolic activation.

Sum.'14

Total nitro-+oxy-PAHs

Fig. 3. Comparison between mutagenic activity observed in Salmonella typhimurium, strains TA98 and TA100 with (+S9) and without metabolic activation (mean of revertants/m3 ± SD), and mutagenic PAHs levels (sum from fluoranthene to dibenzo(a,h)pyrene for PM and from naphtalene to dibenzo(a,h)pyrene for semi-volatile compounds − PUF) and nitro-PAHs plus oxy-PAHs (ng/m3 ). Inv.: invalidated data. PAHs concentration in gaseous phase collected in autumn 2011 it’s not reported as it was fifteen times higher than the mean PAHs concentrations.

No correlation was found between PUF extracts genotoxicity and chemicals concentrations (Figs. 1 and 3). 3.3. Genotoxicity assays on human cell lines 3.3.1. Alkaline comet assay Human alveolar A549 cells were exposed to organic extracts and nearly a third of them induced DNA strand breaks measured by increased percentage of DNA in tail, in comet assay (Notched

box plot, p < 0.05, and R2 >0.6) but, different from Salmonella assays, genotoxicity results seem to be not season-related (Fig. 4). Furthermore data show no difference in strand breaks levels in A549 cells between PM2.5 and PM1 collected in the same period (Student’s t, paired samples) nor among samples, of each fraction, collected in the three different seasons (ANOVA). The relationships between DNA damage and concentrations of PM chemical species are illustrated in Figs. 4 and 5, where it’s clear that for PM1 and PM2.5 there is no relationship between increase

C. Bocchi et al. / Mutation Research 809 (2016) 16–23

21

Table 4 Correlations (coefficients of determination − R2 ) between chemical parameters and mutagenicity induced by OEs in Salmonella assays.

PAHs

PM1

PM2.5 R2 0.93 0.87 0.73 0.78 0.87 0.88 0.92

TA98 TA98 + S9 TA100 TA100 + S9 mean of indirect mutagenicity (+S9) mean of direct mutagenicity mean of revertants

TA98 TA98 + S9 TA100 TA100 + S9 mean of indirect mutagenicity (+S9) mean of direct mutagenicity mean of revertants PM2.5

nitro-PAHs + oxyPAHs

PM1 R2 0.36 0.16 0.56 0.23 0.18 0.46 0.39

TA98 TA98 + S9 TA100 TA100 + S9 mean of indirect mutagenicity (+S9) mean of direct mutagenicity mean of revertants

TA98 TA98 + S9 TA100 TA100 + S9 mean of indirect mutagenicity (+S9) mean of direct mutagenicity mean of revertants

R2 0.30 0.19 0.64 0.08 0.15 0.45 0.34

PM2.5

PM1

Alkanes

R2 0.93 0.88 0.76 0.76 0.86 0.92 0.94

R2 0.73 0.66 0.69 0.50 0.63 0.74 0.75

TA98 TA98 + S9 TA100 TA100 + S9 mean of indirect mutagenicity (+S9) mean of direct mutagenicity mean of revertants

TA98 TA98 + S9 TA100 TA100 + S9 mean of indirect mutagenicity (+S9) mean of direct mutagenicity mean of revertants

R2 0.42 0.30 0.76 0.09 0.22 0.58 0.46

70 60

4

50 3

40

2

30 20

1

10

Sum.'12

Aut.'12

TI%/m3

Win.'13

Sum.'13

PM µg/m3

Aut.'13

Win.'14

PM1 PM2,5 PUF

PM1 PM2,5 PUF

0 PM1 PM2,5 PUF

PUF

PM1 PM2,5

PM1 PM2,5 PUF

PUF

PM1 PM2,5

PUF

PM1 PM2,5

Aut.'11

inv.

PM1 PM2,5

inv.

0

PUF

DNA damage (TI%/m3)

5

PM (μ μg/m3) and alkanes (ng/m3)

Fig. 4. DNA damage induced in A549 cells by OEs of PM and PUF expressed in % tail DNA per cubic meter (slope of linear regression ± confidence intervals) in the different seasons and comparison of damage with mutagenic PAHs levels (sum from fluoranthene to dibenzo(a,h)pyrene for PM and from naphtalene to dibenzo(a,h)pyrene for semivolatile compounds − PUF) and nitro-PAHs plus oxy-PAHs (ng/m3 ). Inv.: invalidated data. PAHs concentration in gaseous phase collected in autumn 2011 it’s not reported as it was fifteen times higher than the mean PAHs concentrations.

Sum.'14

total alkanes

Fig. 5. Comparison between genotoxic activity revealed by comet assay (TI%/m3 of air sampled − slope of linear regression ± conf. int.) and total alkanes concentrations (ng/m3 ) and between damage and PM1 and PM2.5 concentrations (␮g/m3 ) in the different seasons. Inv.: invalidated data.

22

C. Bocchi et al. / Mutation Research 809 (2016) 16–23

contaminated with organic compounds, therefore tests on those PUF were invalidated. 4. Discussion 4.1. Mutagenicity

Fig. 6. Comparison between PAHs and their derivatives, nitro- + oxy-PAHs, concentrations (ng/m3 ) and micronuclei inducted per cubic meter of air sampled (slope of linear regression ± confidence intervals).

Fig. 7. Comparison between PM1 , PM2.5 mass (␮g/m3 ) and alkanes (ng/m3 ) concentrations and increase of MN number per cubic meter of air sampled (slope of linear regression ± confidence intervals).

of percent of DNA in the tail (slope of linear regression model) and mass concentration, PAHs and their derivatives or alkanes per cubic meter of air. But, if we consider organic PUF extracts, DNA damage correlate with total PAHs (R2 = 0.7) and alkanes (R2 = 0.9) concentrations (Figs. 4 and 5). 3.3.2. Micronucleus assay The cytokinesis block micronucleus assay was used to investigate the capability of PM sampled in winter and gaseous phase collected in summer campaigns to induce chromosome breakage or loss in A549 cells. Of these samples both fractions of PM collected in the winter of 2014 increased the number of micronucleated BN cells in a dose-dependent manner. Semi-volatile compounds collected in summer 2012 and 2014 induced DNA damage in comet assay but they resulted toxic in this assay, this could be due to a longer incubation of cells in culture and/or to cytochalasin B addition that could increase cytotoxicity (Fig. 6). No association was found between the induction of micronuclei and the organic micropollutants evaluated and with the particulate matter mass (Fig. 6 and 7). Unfortunately we had to invalidate every test (Salmonella, comet and micronucleus test) on PUF sampled in winter and in summer 2013, as blank extracts of PUF purchased in the winter 2013 showed a high concentration of organic pollutants and were genotoxic and highly toxic on A549 cells (data not shown). When we tried to understand the reason for these results we found out that foams, used for those two campaigns, could have been stored improperly before purchasing them and for that could have been

As we already know, the organic extracts from ambient air particles have direct-acting and indirect acting mutagenicity in bacterial mutagenicity assays that show higher potency in winter and autumn than in summer campaigns [25]. This difference among seasons could be due to meteorological factors as temperature inversion, sunlight, etc. and, in the cold period, to increased emissions from car traffic and house heating [26]. For both particle sizes and strains of Salmonella the OE directacting mutagenic potency is higher than the indirect one, the occurrence of direct mutagens adsorbed onto air particulate has been extensively demonstrated in urban samples collected throughout the world, and nitrocompounds, as nitro-PAHs, and oxygenated PAHs, too, could be implicated [8]. For both PM fractions there is an association between average PM concentrations (␮g/m3 ) and the mean of revertants per cubic meter, this can be seen especially for PM2.5 . However the level of measured particle mass could not explain the level of mutagenic activity elicited, as the mutagenic potency of particulate matter depends more upon the type and percentage of mutagenic substances than upon the mass itself. PM1 shows a higher number of revertants per microgram than PM2.5 underlying that mutagenic substances are more associated with finest particles, confirming that they play a meaningful role in causing adverse health effects in the exposed population. On the contrary only one foam sampled in summer 2012 showed weak mutagenic activity in only one test (TA98 + S9) this could be explained by the fact that in particles and in vapour phase there are different substances. This difference, between the mutagenic activity of PM and semi-volatile compounds, shows that mutagenic substances, in atmosphere, are predominantly present in the solid phase. When attempting to correlate PM mutagenicity results with chemical levels, no correlation was found with nitro + oxy-PAHs and not even with some test with alkanes concentrations. A good correlation was found with PAHs known to be mutagenic in the Salmonella assay, however, the correlation is good even with those tests that don’t reveal PAHs activity. Therefore, chemical analysis alone is insufficient for predicting the level of mutagens in the atmosphere, and it’s clear that the level of each measured pollutant could not explain the level of mutagenic activity elicited. 4.2. Genotoxicity Results of comet assays don’t show any seasonal trend and, for that, don’t always correspond to those of Salmonella, other studies have shown increased levels of strand breaks in cultured cells without assessment of temporal, spatial or particle size differences [27]. Organic extracts of PM generated significant DNA breakage and micronucleus formation in a dose-dependent manner in winter 2014 samples in both PM fractions, showing that in organic extracts we find substances that can break DNA or induce chromosome loss, too. The only correlations found between genotoxicity and physical − chemical parameters are between tail intensity values (comet assay) and alkanes and PAHs in gas phase. It’s important to underline that mutagenicity and genotoxicity tests often give different results: the highest mutagenic activity was measured in particulate collected in autumn 2011 where we found

C. Bocchi et al. / Mutation Research 809 (2016) 16–23

no DNA damage and, furthermore, the only samples that generate micronuclei are winter 2014 PM1 and PM2.5 extracts that are the least mutagenic of all (ANOVA, with Tukey’s post hoc, p < 0.05). Even when we look at correlations we see that PAHs and alkanes concentrations in gaseous phase correlate with comet data, but they don’t explain mutagenicity in Salmonella, on the contrary, PAHs levels in PM correlate with Salmonella results but not with DNA damage assessed by comet assay. These data show that using only one test could lead to a loss of information about genotoxic and mutagenic activity of airborne pollutants and that semi-volatile compounds have clastogen effects contrary to particulate matter that has a great mutagenic activity. These work leads to three important conclusions: mutagenicity and genotocixity change upon sample type (PM2.5 , PM1 filter or PUF extract) and upon seasons; using complementary tests gives a wide range of results that, as many chemical substances can be revealed, may even disagree amongst themselves and, last but not least, chemical analysis have limits when compared to bioassays. 5. Conflict of interest The authors declare no conflict of interest.

[8]

[9]

[10]

[11]

[12]

[13]

[14] [15]

Acknowledgements [16]

This study is a part of Supersito project funded by the Regione Emilia-Romagna and the Agenzia Regionale Prevenzione, Ambiente ed Energia dell’Emilia-Romagna (Arpae), Italy. We would like to thank these organisations and, in particular: Vanes Poluzzi, Silvia Ferrari, Claudio Maccone, Isabella Ricciardelli and Arianna Trentini from the regional department Aree urbane of Arpae in Bologna, for sampling; Ivan Scaroni, Patrizia Casali and Alberto Santolini for chemical analysis (Arpae’s Ravenna department). We would also like to thank Gian Paolo Ilari and Bryan Levandowski for proofreading.

[17]

[18] [19]

[20]

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.mrgentox.2016. 07.007.

[21]

[22]

References [1] C.F. Lu, X.Y. Yuan, L.Z. Li, W. Zhou, J. Zhao, Y.M. Wang, S.Q. Peng, Combined exposure to nano-silica and lead induced potentiation of oxidative stress and DNA damage in human lung epithelial cells, Ecotoxicol. Environ. Saf. 122 (2015) 537–544, http://dx.doi.org/10.1016/j.ecoenv.2015.09.030. [2] C. Lepers, M. Dergham, L. Armand, S. Billet, A. Verdin, V. Andre, D. Pottier, D. Courcot, P. Shirali, F. Sichel, Mutagenicity and clastogenicity of native airborne particulate matter samples collected under industrial, urban or rural influence, Toxicol. In Vitro 28 (2014) 866–874, http://dx.doi.org/10.1016/j.tiv. 2014.03.011. [3] A. Albinet, E. Leoz-Garziandia, H. Budzinski, E. Villenave, Polycyclic aromatic hydrocarbons (PAHs), nitrated PAHs and oxygenated PAHs in ambient air of the Marseilles area (South of France): concentrations and sources, Sci, Total Environ. 384 (2007) 280–292, http://dx.doi.org/10.1016/j.scitotenv.2007.04. 028. [4] D. Traversi, E. Festa, C. Pignata, G. Gilli, Aero-dispersed mutagenicity attributed to particulate and semi volatile phase in an urban environment, Chemosphere 124 (2015) 163–169, http://dx.doi.org/10.1016/j.chemosphere. 2014.12.033. [5] S.M. ElAssouli, M.H. AlQahtani, W. Milaat, Genotoxicity of air borne particulates assessed by comet and the Salmonella mutagenicity test in Jeddah, Saudi Arabia, Int. J. Environ. Res. Public Health 4 (3) (2007) 216–223, http://dx.doi.org/10.3390/ijerph2007030004. [6] L.D. Claxton, G.M. Woodall Jr., A review of the mutagenicity and rodent carcinogenicity of ambient air, Mutat. Res. 636 (200) (2016) 36–94, http://dx. doi.org/10.1016/j.mrrev.2007.01.001. [7] P. Poli, A. Buschini, N. Campanini, M.V. Vettori, F. Cassoni, S. Cattani, C. Rossi, Urban air pollution: use of different mutagenicity assays to evaluate

[23]

[24]

[25]

[26]

[27]

23

environmental genetic hazard, Mutat. Res. 298 (1992) 113–123, http://dx.doi. org/10.1016/0165-1218(92)90036-Y. C. Ramos de Rainho, S. Machado Corrêa, J. Luiz Mazzei, C. Alessandra Fortes Aiub, I. Felzenszwalb, Genotoxicity of polycyclic aromatic hydrocarbons and nitro-derived in respirable airborne particulate matter collected from urban areas of Rio de Janeiro (Brazil), Biomed. Res. Int. 2013 (2013), http://dx.doi. org/10.1155/2013/765352. A. Tarantini, A. Maitre, E. Lefebvre, M. Marques, C. Marie, J.L. Ravanat, T. Douki, Relative contribution of DNA strand breaks and DNA adducts to the genotoxicity of benzo[a]pyrene as a pure compound and in complex mixtures, Mutat. Res. 671 (2009) 67–75, http://dx.doi.org/10.1016/j. mrfmmm.2009.08.014. F. Cassoni, C. Bocchi, A. Martino, G. Pinto, F. Fontana, A. Buschini, The Salmonella mutagenicity of urban airborne particulate matter (PM2.5) from eight sites of the Emilia-Romagna Regional Monitoring Network (Italy), Sci. Total Environ. 324 (2004) 79–90, http://dx.doi.org/10.1016/j.scitotenv.2003. 10.030. Arpae Emilia-Romagna Temi ambientali, Supersito http://www.arpa.emr.it/ index.asp?idlivello=1459, web site in Italian language with an English summary (accessed 28.04.16). EPA United States Environmental Protection Agency, Hazardous Waste Test Methods, 1996, https://www.epa.gov/sites/production/files/2015-12/ documents/3630c.pdf (accessed 28.04.16). D.M. Maron, B.N. Ames, Revised methods for the Salmonella mutagenicity test, Mutat. Res. 113 (1983) 173–215, http://dx.doi.org/10.1016/01651161(83)90010-9. OECD guidelines for the testing of chemicals, section 4, test No. 471, Bact. Reverse Mutat. Test (1997), http://dx.doi.org/10.1787/9789264071247-en. K.C. Chu, K.M. Patel, A.H. Lin, R.E. Tarone, M.S. Linhart, V.C. Dunkel, Evaluating statistical analysis and reproducibility of microbial mutagenicity assays, Mutat. Res. 85 (1981) 119–132, http://dx.doi.org/10.1016/01651161(81)90027-3. M.E. Gutiérrez-Castillo, D.A. Roubicek, M.E. Cebriàn-GarcÍa, A. De ˇ P. Ostrosky-Wegman, Effect of chemical Vizcaya-RuÍz, M. Sordo-Cedeno, composition on the induction of DNA damage by urban airborne particulate matter, Environ. Mol. Mutagen. 47 (2006) 199–211, http://dx.doi.org/10. 1002/em.20186. J.M. Chambers, W.S. Cleveland, B. Kleiner, P.A. Tuk, Comparing data distributions, in: J.M. Chambers (Ed.), Graphical Methods for Data Analysis, Wadsworth Int. Group, Belmont California, 1983, pp. 61–63. OECD Guidelines for the Testing of Chemicals, Section 4, Test No. 487 : In Vitro Mammalian Cell Micronucleus Test 2014, doi:10.1787/9789264224438-en. M. Fenech, W.P. Chang, M. Kirsch-Volders, N. Holland, S. Bonassi, E. Zeiger, HUMN project: detailed description of the scoring criteria for the cytokinesis-block micronucleus assay using isolated human lymphocyte cultures, Mutat. Res. 534 (2003) 65–75, http://dx.doi.org/10.1016/S13835718(02)00249-8. ˇ P. Kotianová, H. Puxbaum, H. Bauer, A. Caseiro, I.L. Marr, G. Cík, Temporal patterns of n-alkanes at traffic exposed and suburban sites in Vienna, Atmos. Environ. 42 (2008) 2993–3005, http://dx.doi.org/10.1016/j.atmosenv.2007.12. 048. Z.G. Yi, X.R. Huang, J.Q. Bi, P.P. Guo, L.L. Zheng, Characteristics of PAHs in the atmosphere in winter and summer in the urban and suburban of Fuzhou, Huan. Jing. Ke. Xue. 34 (2013) 1252–1257 (Article in Chinese). P. Poli, A. Buschini, F.M. Restivo, A. Ficarelli, F. Cassoni, I. Ferrero, C. Rossi, Comet assay application in environmental monitoring: DNA damage in human leukocytes and plant cells in comparison with bacterial and yeast tests, Mutagenesis 14 (1999) 547–556, http://dx.doi.org/10.1093/mutage/14. 6.547. A. Valavanidis, K. Fiotakis, T. Vlachogianni, Airborne particulate matter and human health: toxicological assessment and importance of size and composition of particles for oxidative damage and carcinogenic mechanisms, J. Environ. Sci. Health C Environ. Carcinog. Ecotoxicol. Rev. 26 (2008) 339–362, http://dx.doi.org/10.1080/10590500802494538. A. Buschini, F. CassonI, E. Anceschi, L. Pasini, P. Poli, C. Rossi, Urban airborne particulate: genotoxicity evaluation of different size fractions by mutagenesis tests on microorganisms and comet assay, Chemosphere 44 (2001) 1723–1736, http://dx.doi.org/10.1016/S0045-6535(00)00550-6. Arpae Emilia-Romagna, Temi Ambientali, Tossicologia Ambientale, Mutagenesi Ambientale, Rete regionale di monitoraggio della mutagenicità del particolato atmosferico in ambiente urbano, 2016, http://www.arpa.emr. it/dettaglio generale.asp?id=2872&idlivello=1611, in Italian language (accessed 28.04.16). C. Rossi, P. Poli, A. Buschini, F. Cassoni, S. Cattani, E. de’ Munari, Comparative investigations among meteorological conditions, air chemical-physical pollutants and airborne particulate mutagenicity: a long-term study (1990–1994) from a northern Italian town, Chemosphere 30 (1995) 1829–1845, http://dx.doi.org/10.1016/0045-6535(95)00065-G. P. Møller, P.H. Danielsen, D.G. Karottki, K. Jantzen, M. Roursgaard, H. Klingberg, D.M. Jensen, D.V. Christophersen, J.G. Hemmingsen, Y. Cao, S. Loft, Oxidative stress and inflammation generated DNA damage by exposure to air pollution particles, Mutat. Res. Rev. Mutat. Res. 762 (2014) 133–166, http:// dx.doi.org/10.1016/j.mrrev.2014.09.001.