Identification of PM10 characteristics involved in cellular responses in human bronchial epithelial cells (Beas-2B)

Environmental Research 149 (2016) 48–56

Contents lists available at ScienceDirect

Environmental Research journal homepage: www.elsevier.com/locate/envres

Identification of PM10 characteristics involved in cellular responses in human bronchial epithelial cells (Beas-2B) Rosette Van Den Heuvel a,n, Elly Den Hond a, Eva Govarts a, Ann Colles a, Gudrun Koppen a, Jeroen Staelens b, Maja Mampaey c, Nicole Janssen d, Greet Schoeters a,e a

Flemish Institute for Technological Research (VITO), Environmental Risk and Health Unit, Boeretang 200, 2400 Mol, Belgium Flanders Environment Agency (VMM), Unit Air, Kronenburgstraat 45, 2000 Antwerp, Belgium c LNE (Environment, Nature and Energy Department), Flemish Government, Koning Albert II-laan 20, 1000 Brussels, Belgium d National Institute for Public Health and the Environment (RIVM), P.O. Box, 2720 BA, Bilthoven, The Netherlands e University of Antwerp, Department of Biomedical Sciences, 2000 Antwerp, Belgium b

art ic l e i nf o

a b s t r a c t

Article history: Received 1 December 2015 Received in revised form 18 April 2016 Accepted 20 April 2016 Available online 10 May 2016

Notwithstanding evidence is present that physicochemical characteristics of ambient particles attribute to adverse health effects, there is still some lack of understanding in this complex relationship. At this moment it is not clear which properties (such as particle size, chemical composition) or sources of the particles are most relevant for health effects. This study investigates the in vitro toxicity of PM10 in relation to PM chemical composition, black carbon (BC), endotoxin content and oxidative potential (OP). In 2013–2014 PM10 was sampled (24 h sampling, 108 sampling days) in ambient air at three sites in Flanders (Belgium) with different pollution characteristics: an urban traffic site (Borgerhout), an industrial area (Zelzate) and a rural background location (Houtem). To characterize the toxic potential of PM10, airway epithelial cells (Beas-2B cells) have been exposed to particles in vitro. Different endpoints were studied including cell damage and death (cell viability) using the Neutral red Uptake assay, the production of pro-inflammatory molecules by interleukin 8 (IL-8) induction and DNA-damaging activity using the FPG-modified Comet assay. The endotoxin levels in the collected samples were analysed and the capacity of PM10 particles to produce reactive oxygen species (OP) was evaluated by electron paramagnetic resonance (EPR) spectroscopy. Chemical characteristics of PM10 (BC, As, Cd, Cr, Cu, Mn, Ni, Pb, Zn) and meteorological conditions were recorded on the sampling days. PM10 particles exhibited dose-dependent cytotoxicity in Beas-2B cells and were found to significantly induce the release of IL-8 in samples from the three locations. Oxidatively damaged DNA was observed in exposed Beas-2B cells. Endotoxin levels above the detection limit were detected in half of the samples. OP was measurable in all samples. Associations between PM10 characteristics and biological effects of PM10 were assessed by single and multiple regression analyses. The reduction in cell viability was significantly correlated with BC, Cd and Pb. The induction of IL-8 in Beas-2B cells was significantly associated with Cu, Ni and Zn and endotoxin. Endotoxin levels explained 33% of the variance in IL-8 induction. A significant interaction between ambient temperature and endotoxin on the pro-inflammatory activity was seen. No association was found between OP and the cellular responses. This study supports the hypothesis that, on an equal mass basis, PM10 induced biological effects differ due to differences in PM10 characteristics. Metals (Cd, Cu, Ni and Zn), BC, and endotoxin were among the main determinants for the observed biological responses. & 2016 Elsevier Inc. All rights reserved.

Keywords: Particulate matter Cytotoxicity Pro-inflammatory response Oxidative potential Endotoxin Comet assay

1. Introduction

n

Correspondence to: VITO, Boeretang 200, 2400 Mol, Belgium. E-mail addresses: [email protected] (R. Van Den Heuvel), [email protected] (E. Den Hond), [email protected] (E. Govarts), [email protected] (A. Colles), [email protected] (G. Koppen), [email protected] (J. Staelens), [email protected] (M. Mampaey), [email protected] (N. Janssen), [email protected] (G. Schoeters). http://dx.doi.org/10.1016/j.envres.2016.04.029 0013-9351/& 2016 Elsevier Inc. All rights reserved.

Exposure to particulate matter (PM) air pollution has been significantly associated with increased mortality and morbidity in numerous epidemiological studies (Brunekreef and Holgate, 2002; Pope et al., 2002; Anderson et al., 2012). A consistent association has been found between ambient levels of the PM mass concentration and various health outcomes such as cardiovascular and

R. Van Den Heuvel et al. / Environmental Research 149 (2016) 48–56

respiratory diseases. These associations are particularly demonstrated for PM mass concentrations, such as PM10 and PM2.5 concentrations which are currently used for air quality legislation in many countries. More recently, it has been demonstrated by both epidemiological and toxicological research that PM mass is not necessarily the major factor causing adverse health effects (Peng et al., 2009; Schlesinger et al., 2006; Bell et al., 2014). Several studies have associated health effects with specific PM characteristics (Gordon, 2007; Gray et al., 2015). However, at this moment it is not clear which properties (such as particle size, chemical composition) or sources of the particles are most relevant for health effects. In the last decades, more research focused on clarifying the mechanisms of particle toxicology that link specific chemical and physical characteristics of PM with biological responses in vitro (Steerenberg et al., 2006; Sandström et al., 2005; Valavanidis et al., 2008; Steenhof et al., 2011). In vitro toxicity studies of PM aim to provide more information on the toxic properties of the complex mixture of air pollutants. Therefore, in vitro studies using cells that are treated with various types of ambient particles are a useful technique investigating ambient air toxicity and may yield a tool for a more health-relevant approach for air quality follow-up. A number of studies correlated PM chemical composition with different biological endpoints such as inflammation, perturbation of cell cycle, oxidative stress or DNA damage. However, although several studies provided some evidence, there is still no consistent agreement regarding the determinants of the biological responses. Cytotoxicity and the induction of pro-inflammatory cytokines in lung cells (e.g. IL-8, IL-6, TNF-α) have been associated with the presence of metals in particulate matter (Akhtar et al., 2010; Healy et al., 2012; Steenhof et al., 2011; Osornio-Vargas et al., 2011). The increase in the production of pro-inflammatory molecules in human airway epithelial cells is related to the activation of membrane TLRs (Toll-like Receptors) and is mediated through an NFкβ- dependent signalling pathway (Silbajoris et al., 2011; Bengalli et al., 2013). Genotoxic properties of PM have been associated with the presence of PAHs (polyaromatic hydrocarbons) (Brits at al., 2004). More recently, oxidative stress caused by the generation of Reactive Oxygen Species (ROS) has been proposed as one of the main mechanism underlying PM induced toxicity (Møller et al., 2010). In addition, the presence of biological components such as bacterial endotoxin in ambient air has been recognized as a modulator of the inflammatory response in vitro (Ferguson et al., 2013; Allen et al., 2011; Michael et al., 2013). In 2013–2014 PM10 sampling campaigns were initiated in three different areas in Flanders with different pollution pressure: an urban area (Borgerhout), a background site (Houtem) and an industrial area (Zelzate) located in the Ghent Canal zone. A panel of biological endpoints was tested on the collected PM samples including cell viability, inflammation and oxidatively damaged DNA in bronchial epithelial cells (Beas-2B). Beside chemical constituents of PM, the endotoxin level and oxidative potential (OP) of samples were analysed. This study aimed to provide more insights in the underlying PM characteristics responsible for the biological responses.

2. Methods 2.1. PM sampling The sampling campaigns were conducted at three locations in Flanders: an industrial area (Zelzate), an urban area (Borgerhout) and a background location (Houtem). The site in Zelzate is representative for an industrial background due to its location near

49

the industrial canal zone Ghent-Terneuzen. The urban site in Borgerhout is located 700 m from the ring road of the city of Antwerp and is representative for urban traffic background. The site is used as an urban background station for routine PM10 monitoring in the framework of the air quality directive (2008/50/ EC), but is located near (30 m) to a busy 2
2-lane road (33,500 vehicles day  1 in Feb 2010; Mishra et al., 2012). The measurement site in Houtem is generally considered as a rural background site. The station is located in the middle of the polders 500 m from the French-Belgian border, 2 km from the centre of Houtem (700 inhabitants), slightly more than 8 km from the Belgian coast. PM samples were collected using a high-volume sampler, Digitel DHA-80 (Digitel, Hegnau, Switzerland) with a PM10 pre-separator. Particles were collected on Teflon T38 filters (Schleicher & Schuell, Dassel, Germany). The filters were changed automatically after 24 h. Sampling in Borgerhout (urban) and Houtem (background) was conducted during the period April 2013-May 2014. The sampling schedule was designed to sample with an interval of 6 days in Borgerhout and 12 days in Houtem, yielding 61 days in Borgerhout and 34 days in Houtem over a 1-year period. The sampling occurred simultaneously on the two locations (Borgerhout and Houtem). In the industrial area (Zelzate), 13 filters were collected between April 2013 and January 2014. In total 108 filters were collected for biological effect studies. All filters were stored at  20 °C until further handling. 2.2. Chemical characterisation and meteorological conditions Information on the chemical characterisation of ambient air at the three sampling sites throughout the campaign was provided by the Flanders Environment Agency (VMM). Daily PM10; PAHs in TSP (Total Suspended Matter) (fluoranthene, pyrene, benzo(a)anthracene, chrysene, benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(a)pyrene, benzo(g,h,i)-perylene and indeno(1,2,3-cd)pyrene); heavy metals (ng/m3) in PM10 (Pb, Zn, Cu, Ni, As, Mn, Cd, Cr) and black carbon (BC) (mg/m3) in TSP are available for the sampling days. Meteorological conditions such as air temperature, wind speed and precipitation amount were monitored and recorded during the sampling campaigns by VMM and KMI (Royal Meteorological Institute from Belgium). 2.3. PM preparation To perform toxicological characterisation using a battery of biological tests and oxidative potential (OP) analyses on the same sample, each daily Teflon filter was sectioned into parts. A small part of the Teflon filter was used for OP analyses. Biological tests were performed using the particle phase of the filters. A portion of the filters was used for particle extraction as described elsewhere (Salonen et al., 2004; Steerenberg et al., 2006; Jalava et al., 2005). Briefly, the collected PM10 were extracted from the Teflon substrate into pure methanol by sonication for 30 min. The suspension was evaporated using a rotavapor at 40 °C, 337 mbar and 100 rpm. The suspension was sonicated and transferred to a pre-weighed 1.5 mL conical tube. To enable in vitro testing on equal PM mass concentration, the tube was further evaporated under N2 to dryness. The conical tube was subsequently weighted again to determine the weight of the particles. The particles were kept in the freezer at  20 °C. All daily PM samples (n¼ 108) were used for in vitro exposure. The different assays were performed once on each sample. In addition, the extraction procedure was performed using blank filters to evaluate filter toxicity in Beas-2B cells.

50

R. Van Den Heuvel et al. / Environmental Research 149 (2016) 48–56

2.4. Endotoxin The presence of microbial endotoxin activity in PM10 samples was assessed using a quantitative LAL (Limulus Amebocyte Lysate) test kit according to the manufacturer's instructions (QCL-1000 kit, Lonza). Endotoxin analyses were performed in duplo on suspended PM10 samples (100 mg PM10/mL) that were used for the in vitro toxicity testing. The mean endotoxin concentration of two replicate wells per PM sample was calculated. Extracts from blank filters were used as negative control. No endotoxin levels above the detection limit were found in these samples. 2.5. Oxidative potential (OP) The method used for the detection of ROS (radical oxidant species) generation capacity of PM was the electron paramagnetic resonance (EPR) spectroscopy. This method is based on the trapping of PM induced hydroxyl radicals (OH) mainly generated via Fenton-type reaction in the presence of H2O2, using the spin trap 5,5-dimethyl-1-pyrroline-N-oxide (DMPO). The analysis was done by RIVM using a direct method without PM extractions from the filters (Yang et al., 2014; Hellack et al., 2014). OP was analysed on five blank filters as background control. The mean value of the latter was subtracted from the levels measured on sampled filters. 2.6. Biological effect studies 2.6.1. Cell culture The Beas-2B cell line was obtained from American Type Culture Collection (ATCC number: CRL-9609, LGC Promochem, Teddington, UK) and was originally derived from normal bronchial epithelial cells obtained from autopsy of noncancerous individuals. These cells were cultured as described before (Verstraelen et al., 2009; Remy et al., 2014). BEAS-2B cells were grown in T-75 culture flasks precoated with a mixture of 10 lg/mL human plasma fibronectin (Invitrogen, Paisley, UK), 30 lg/mL PureColTM (Advance Biomatrix, USA), and 10 lg/mL bovine serum albumin (Sigma–Aldrich Chemie Gmbh, Steinheim, Germany) dissolved in growth medium. The mixture was added at a ratio of 0.2 mL/cm2 surface area. The growth medium consisted of bronchial epithelial cell basal medium (500 mL, Lonza, Verviers, Belgium) supplemented with: 0.5 mL retinoic acid, 0.5 mL epinephrine, 0.5 mL triiodothyronine, 0.5 mL human recombinant epidermal growth factor, 0.5 mL insulin, 0.5 mL gentamicin sulfate amphotericin-B, 0.5 mL transferrin, 0.5 mL hydrocortisone, and 2 mL bovine pituitary extract (BulletKit, Lonza). The cells were kept in a humidified atmosphere at 37 °C and 5% CO2. Before reaching 80% confluence, cells were subcultured using 0.25% trypsin/0.53 mM versene (EDTA) solution (LGC Promochem) containing 0.5% polyvinylpyrrolidone (Sigma– Aldrich Chemie Gmbh). Medium was refreshed every 2–3 days and cells were subcultured every 4–5 days (1500–3000 cells/cm2). 2.6.2. Experimental setup Beas-2B cells were seeded at 2
104 cells per well in 96-well plates (200 mL/well; cell culture area 0.32 cm2/well) and allowed to adhere for 24 h. PM samples were suspended in PBS (-Ca/Mg) by overnight shaking to a final concentration of 10 mg/mL. Samples were further diluted in Beas-2B cell culture medium to 1 mg/mL and stirred overnight. All PM samples were stirred before further dilution and addition to the cells. The cells were exposed for 24 h to a concentration range of particles: 100, 50, 25, 12.5 mg PM10/mL medium (corresponding to 62.5, 31.3, 15.6, 7.8 μg/cm2). Each concentration was tested in six replicate wells. Untreated cells were used as a negative control. On each plate, cells were exposed to an increasing concentration (0.2, 0.4, 0.8, 1.6, 3.12, 6.25, 12.5, 25, 50 mg/mL) of CdCl2 as a positive

control. In addition, Lipopolysaccharide (LPS) from Escherichia coli exposure was used a positive control for cytokine induction in Beas-2B cells (twofold serial dilution between 2.5 and 0.02 ng/mL). After exposure, the culture medium from the replicate wells was collected, centrifuged (2500 rpm, 10 min) to remove the suspended particles and stored at  80 °C until for further cytokine analysis. The viability of the cells was determined using the Neutral-Red Uptake test. 2.6.3. Neutral red uptake (NRU) assay Inhibition of cell growth was measured by the neutral red cytotoxicity assay. The procedure was based on INVITTOX protocol no. 64 on Neutral Red Uptake (NRU) Cytotoxicity assay. The uptake of NR into the lysosomes/endosomes and vacuoles of living cells is used as a quantitative indication of cell number and viability. After 24 h exposure, NRU is determined for each treatment concentration and compared to that determined in control cultures. Cell viability (mean of 6 replica wells) was expressed as percentage viable cells compared to the unexposed control cells (100%). The extracts from blank filters did not induce cell viability reduction. CdCl2 induced a dose-dependent decrease in Beas-2B cell viability. IC50 values ranged between 1.5 and 4.5 mg CdCl2/mL (IC50 : the concentration producing 50% reduction of NRU). 2.6.4. IL-8 analyses Interleukin-8 (IL-8) release by exposed Beas-2B cells was determined as an indicator of pro-inflammatory potency. IL-8 is a major pro-inflammatory mediator involved in cell attraction. IL-8 was measured in the supernatants using enzyme-linked immunosorbent assay (ELISA) kits (Life Technologies, IL-8 Human Antibody Pair, Novexs). The cell culture medium of the 6 replicate wells per PM10 concentration was pooled and the IL-8 levels in the pooled samples were analysed in duplicate. The mean IL-8 concentration (pg/mL) of two replicates was calculated. Results are expressed as the ratio between IL-8 production after PM10 exposure and IL-8 levels in non-exposed cells. The average basal IL-8 concentration (795% confidence limit (CL)) produced by nonexposed cells was 54.4 74.7 pg/mL. LPS induced a significant dose-dependent increase in IL-8. LPS at 1.25 ng/mL induced IL-8 levels in Beas-2B cells of 237.2 756.2 pg/mL (average 7 95% CL). 2.6.5. FPG-modified Comet assay Beas-2B cells were seeded in 24-well plates (105 cells/mL/well; cell culture area 2 cm2/well) 24 h before exposure to PM10. Cells were exposed to particles (100 μg PM10/mL, corresponding to 50 μg PM10/cm2) during 24 h. Exposed cells were trypsinized to produce a single-cell suspension. Beas-2B cells were mixed with 0.8% Low Melting Point Agarose (LMPA) (Life Technologies, Foster City, CA, USA). For analysis of regular strand breaks, 100 mL of this mixture was placed onto a Gelbonds Film (Lonza, Basel, Switzerland) with the size of a conventional glass microscope slide. In order to assess DNA lesions mediated by ROS, the Comet assay was performed by treatment of slides with an endonuclease FPG (formamidopyrimidine glycosylase), which further enhances DNA migration in case of oxidative damage. For determination of oxidatively damaged DNA, a second gel was made on a separate Gelbonds Film. Gels were immersed in a cold lysis solution (2.25 M NaCl, 90 mM Na2EDTA, 9 mM Tris, 1% Triton X-100, 10% DMSO) and kept overnight at 4 °C until further analysis. One gel was incubated in enzyme solution (1/20.000 dilution of FPG stock solution (19.14 mg/mL; kindly provided by Dr. Gunnar Brunborg, National Institute of Public Health, Oslo, Norway) in enzyme buffer) while the second was incubated in enzyme buffer . (100 mM NaCl, 40 mM HEPES, 0.5 mM Na2EDTA, 0.2 mg/mL BSA) The slide was incubated for 30 min at 37 °C. Before electrophoresis, slides were placed for 40 min in the electrophoresis tank in a cold (4 °C)

R. Van Den Heuvel et al. / Environmental Research 149 (2016) 48–56

alkaline solution (300 mM NaOH, 1 mM Na2EDTA) to unwind the DNA. Electrophoresis was performed during 20 min at 4 °C and 1 V/cm (0.7 V/cm across the elevated platform of the electrophoresis tank). After electrophoresis, gels were stained with SYBRs Gold Nucleic Acid Gel Stain (Life Technologies) and image analysis was performed using a Zeiss Axioplan 2 microscope (Zeiss, Germany) equipped with a semi-automatic Metafer 3.5.3 system (MetaSystems GmbH, Altlussheim, Germany). At least 100 comets per slide were automatically scored and the relative amount of DNA damage (percentage DNA in the tail compared to the total amount in the whole comet) was assessed by the system. Results were displayed as median percentage DNA in tail. The difference in median percentage DNA in tail between FPG-treated and non-treated slides served a measure of ROS induced DNA damage (oxidatively damaged DNA). Peripheral blood cells treated with Ro 12-9786 (9 mM) and exposed to light (2.5 min, 13.5 cm distance) were used as a positive control. The mean (7 95% CL) oxidatively damaged DNA expressed as median percentage DNA in tail in the positive controls was 71.473.4%. The mean (7 95% CL) percentage DNA in tail in non-exposed Beas-2B cells was 34.1 73.0% and 3.4 71.0% respectively with and without FPG treatment. 2.7. Statistics For statistical analysis, we used Statistica (version 10) and SAS software (version 9.3) (SAS Institute Inc., Cary, NC, USA). Indicated sampling days (n) o108 (3 areas together), o61 (urban area), o34 (rural area) or o13 (industrial area) reflect either missing data in PM characteristic or missing data in biological endpoint. Concentration-response relationships between PM exposure and the different toxicological responses were assessed by oneway analysis of variance (ANOVA) followed by Dunnett’s post hoc test. The statistical significance level was set at p ¼0.05. Biological effect data obtained at the highest exposure concentration of 100 mg PM10/mL were used in the regression analyses. Correlations between different cellular responses and PM characteristics were primarily assessed using Spearman rank correlation. In addition, multiple linear regression analysis was conducted. Variables with a non-Gaussian distribution were logarithmically transformed. Explanatory variables were identified by a stepwise regression procedure with the p-values for variables to enter in the model set at 0.20, and to stay in the model at 0.05. PAKs concentrations were not included in the model, since PAKs are not soluble in the cell culture medium. Values below the detection limit (DL) were set at DL/2 in the statistical analyses. Interactions between explanatory variables and temperature and wind speed on the biological responses were also considered.

51

3. Results 3.1. Toxicity of PM10 Biological responses for each site during the sampling campaign are summarised in Table 1. Human bronchial epithelial cells were exposed to a concentration range of PM10 particles. The viability of the cells was significantly concentration-dependent in 92 out of 108 sampled days (ANOVA, p o0.01). The overall average reduction (795% CL) in cell viability at the highest exposure concentration was 2472.9%. The cytotoxic effect of PM10 was comparable between the three sites. The pro-inflammatory capacity of particles as measured by IL-8 induction (ratio PM10-exposed vs negative control) in Beas-2B cells increased significantly with increasing PM10 exposure concentration (ANOVA, p o0.05). At the highest PM10 exposure concentration (100 mg PM10/mL), the average il-8 level ( 795% CL) measured in the cell culture medium was 106.9 719.0 pg/mL. The highest pro-inflammatory activity was seen in the urban site with a 2.6 fold increase in IL-8 production. Oxidatively damaged DNA in PM10 exposed Beas-2B cells was measured by the FPG-modified Comet assay. Besides regular strand breaks, PM10 samples induced oxidatively damaged DNA in Beas-2B cells. The median percentage of DNA in tail was significantly increased after FPG treatment indicating that ROS generated by PM10 induced an increase in DNA damage (Paired t-test, p o0.01). 3.2. PM10 characterisation Chemical composition of PM for each site on the sampling days are shown in Table 2. The metals were highly inter-correlated with correlation coefficients ranging between 0.20 and 0.79. Spearman rank correlations are shown in Table 3. BC was significantly correlated with metal concentrations and OP. Since the PAH congeners were strongly correlating (r ranging between 0.792 and 0.978), they were categorized as non-carcinogenic (sum of pyrene and fluoranthene) and carcinogenic (sum of benzo(a)anthracene, chrysene, benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(a) pyrene, dibenzo(a,h)anthracene, benzo(g,h,i)-perylene and indeno (1,2,3-cd)pyrene) PAHs. Endotoxin concentrations above the detection limit were detected in more than half of the samples and ranged between oDL and 8 EU/100 mg PM10. Exposure of Beas-2B cells to a concentration range of LPS, showed that concentrations 40.8 EU/mL significantly induced IL-8 production (data not shown). Such high endotoxin levels were found in 10 out 105 daily samples. All samples showed oxidative potential (OP) above the detection limit. The highest OP was seen in the urban area while the

Table 1 In vitro biological responses (cell viability, IL-8 induction and oxidatively damaged DNA) in Beas-2B cells after exposure to 100 mg PM10/ mL. Results are expressed as mean 7 95% CL (confidence limit) per sampling location. The number of sampling days with in vitro data is given between brackets. Sampling days (n) o61 in the urban area,o 34 in the rural area and o 13 in the industrial area indicate missing results in the biological outcome. Cell viability is expressed as the percentage viable cells compared to the negative control. IL-8 induction is expressed as the ratio of IL-8 production in exposed cells compared to unexposed Beas-2B cells. Oxidatively damaged DNA is expressed as median percentage DNA in tail.

Cell viability* IL-8 induction** Oxidatively damaged DNA***

Urban area Borgerhout

Rural area Houtem

Industrial area Zelzate

Sum of 3 areas

767 2.6 (61) 2.62 7 0.6 (58) 29.5 7 4.2 (41)

787 6.5 (34) 1.80 7 0.5 (34) 38.2 7 8.0 (20)

68 713.6 (13) 1.06 70.4 (13) 39.2 78.8 (12)

75.60 72.9 (108) 2.16 70.4 (105) 33.5 73.5 (73)

Mean 7 95% confidence limit. The number of sampling days with in vitro data is given between brackets. *

Percentage viable cells compared to negative control. Ratio IL-8 production in exposed cells compared to unexposed Beas-2B cells. *** Median percentage DNA in tail. **

52

R. Van Den Heuvel et al. / Environmental Research 149 (2016) 48–56

Table 2 Values of PM mass concentration, chemical composition, endotoxin concentration and oxidative potential (OP) at urban, rural and industrial sites in Flanders at the selected sampling days (The data represent the median (P50), 25th (P25) and 75th percentile(P75) ( n ¼number of PM10 sampling days intended for toxicity studies with physicochemical data. Sampling days (n) o61 in the urban area,o 34 in the rural area or o 13 in the industrial area indicate missing data.). Urban area Borgerhout

Rural area Houtem

Industrial area Zelzate

n

P25

P50

P75

n

P25

P50

P75

n

P25

P50

P75

PM10

mg/m³

61

17.00

21.05

30.35

30

12.25

17.00

26.00

13

24.00

29.00

34.00

BC

mg/m³

60

1.43

1.87

3.24

33

0.28

0.42

1.31

13

0.98

1.25

1.79

As Cd Cr Cu Mn Ni Pb Sb Zn

ng/m³ ng/m³ ng/m³ ng/m³ ng/m³ ng/m³ ng/m³ ng/m³ ng/m³

59 59 59 59 59 59 59 59 29

0.40 0.18 5.45 14.81 6.49 2.14 8.74 1.19 50.06

0.76 0.25 9.05 22.56 9.88 3.07 13.26 2.09 68.09

1.25 0.52 11.12 32.17 16.14 4.61 19.74 3.94 89.35

33 33 33 33 33 33 33 33 33

0.30 0.10 0.70 2.40 2.20 1.00 2.60 0.30 0.50

0.40 0.10 1.10 3.60 3.80 2.50 4.70 1.40 3.50

0.80 0.20 2.50 6.70 12.90 3.80 8.30 16.50 14.40

13 13 13 13 13 13 13

0.35 0.35 2.10 6.60 5.80 1.50 6.50

0.35 0.35 3.20 10.60 17.50 2.40 13.50

0.35 0.60 4.10 12.50 32.10 7.00 52.40

13

19.80

29.30

55.50

Fluoranthene Pyrene Benzo(a)anthracene Chrysene Benzo(b)fluoranthene Benzo(k)fluoranthene Benzo(a)pyrene Benzo(ghi)perylene Indeno(1,2,3-cd)pyrene

ng/m³ ng/m³ ng/m³ ng/m³ ng/m³ ng/m³ ng/m³ ng/m³ ng/m³

35 35 35 35 35 35 35 35 35

0.243 0.126 0.052 0.098 0.049 0.022 0.03 0.051 0.024

0.338 0.225 0.104 0.15 0.081 0.038 0.05 0.074 0.046

0.616 0.388 0.193 0.257 0.178 0.079 0.119 0.15 0.103

14 14 14 14 14 14 14 14 14

0.052 0.009 0.005 0.021 0.019 0.003 0.000 0.007 0.005

0.1115 0.0595 0.0485 0.124 0.081 0.0325 0.0265 0.0515 0.039

0.277 0.094 0.149 0.214 0.256 0.102 0.069 0.112 0.073

6 6 6 6 6 6 6 6 6

0.221 0.153 0.181 0.349 0.214 0.081 0.092 0.126 0.107

0.869 0.646 0.52 0.529 0.454 0.239 0.39 0.306 0.23

0.545 0.3885 0.405 0.407 0.366 0.194 0.3425 0.2195 0.17

Sum PAHs* Sum cPAHs ** Sum non-cPAHs***

ng/m³ ng/m³ ng/m³

35 35 35

0.79 0.40 0.39

1.23 0.58 0.64

1.73 1.01 0.98

14 14 14

0.19 0.09 0.09

0.70 0.44 0.17

1.14 0.93 0.41

6 6 6

2.10 1.44 0.47

2.88 2.06 1.01

4.13 2.67 1.46

OP Endotoxin°

AU/m3 EU/mg

22 61

2394 0.96

2327 2.28

5056 5.03

28 34

540 o DL

856 1.39

1470 2.95

13 13

1034 o DL

2366 o DL

4036 1.94

BC ¼ black carbon; DL¼ detection limit; EU ¼Endotoxin Unit; AU ¼arbitrary unit;; °values below DL were set at DL/2 in the statistical analyses. * Sum of fluoranthene, pyrene, benzo(a)anthracene, chrysene, benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(a)pyrene, benzo(g,h,i)-perylene and indeno(1,2,3-cd) pyrene . ** Sum of carcinogenic PAHs (fluoranthene, pyrene). *** Sum of non-carcinogenic PAHs (benzo(a)anthracene, chrysene, benzo(b)fluoranthene, benzo(k)fluoranthene, benzo(a)pyrene, benzo(g,h,i)-perylene and indeno(1,2,3cd)pyrene ).

rural area PM10 showed the lowest OP. OP was significantly correlated with all other PM10 characteristics (metals, BC and endotoxin) except Ni (Table 3). 3.3. Correlations with PM characteristics In order to identify which PM10 characteristics may provide a possible explanation for the observed adverse effects, the biological results were compared to the PM10 characteristics and to the meteorological conditions. Associations between the biological effects and the physicochemical parameters were initially tested using nonparametric Spearman rank correlation. The correlation coefficients between the PM characteristics and the toxic responses are presented in Table 4. In the statistical analyses, biological responses in Beas-2B cells were correlated with water soluble compounds. Scatter plots of selected chemical constituents with biological endpoints are shown in Fig. 1 . The viability of the airway cells was significantly negative correlated with BC and the metals Cd and Pb. Positive significant correlations for IL8 induction were observed with Cu, Ni, Zn and endotoxin . A significant positive correlation was observed between the oxidatively damaged DNA and Cd while a negative correlation was seen with Ni. However, looking at the scatterplot

showed that the correlation with Cd is dependent on one day with a high ambient Cd concentration. When excluding this outlier, the significance disappeared. The observed in vitro effects were linked with the PM composition using multiple regression analyses (Table 5). A significant positive association was found between the pro-inflammatory activity of the samples and the endotoxin concentration which explained 33% of the total variance in IL-8 induction. No significant association was found between cell viability and PM10 characteristics in the multiple regression analyses. The positive association between oxidatively damaged DNA and Cd concentrations was confirmed in the multiple model but disappeared after exclusion of the high Cd value. The interaction of the explanatory variables with the ambient temperature and the wind speed was also tested. A significant interaction between temperature and endotoxin on IL-8 induction was seen. IL-8 induction increases with temperature for a given endotoxin concentration.

4. Discussion PM10 samples collected in an urban area, an industrial area and a background location were toxicologically characterized using

R. Van Den Heuvel et al. / Environmental Research 149 (2016) 48–56

53

Table 3 Spearman rank correlation coefficients (r) between PM10 characteristics in all the samples from the three locations. Significant correlations (p r 0.05) are indicated in bold.

AS

CD

CR

CU

MN

NI

PB

ZN

BC

ENDOTOXIN

OP

r p n r p n r p n r p n r p n r p n r p n r p n r p n r p n r p n

AS

CD

CR

CU

MN

NI

PB

ZN

BC

ENDOTOXIN

OP

100.00

0.547 o .0001 102 100.00

0.430 o .0001 103 0.242 0.0149 100 100.00

0.571 o.0001 103 0.445 o.0001 100 0.752 o.0001 103 100.00

0.327 0.0007 105 0.420 o .0001 102 0.365 0.0001 103 0.659 o .0001 103 100.00

0.289 0.0027 105 0.201 0.0427 102 0.382 o.0001 103 0.326 0.0008 103 0.137 0.1609 105 100.00

0.539 o .0001 105 0.608 o .0001 102 0.496 o .0001 103 0.669 o .0001 103 0.598 o .0001 105 0.102 0.2981 105 100.00

0.360 0.0017 73 0.268 0.0248 70 0.787 o.0001 73 0.776 o.0001 73 0.497 o.0001 73 0.254 0.0298 73 0.563 o.0001 73 100.00

0.467 o .0001 104 0.351 0.0003 101 0.727 o .0001 102 0.799 o .0001 102 0.409 o .0001 104 0.230 0.0184 104 0.534 o .0001 104 0.568 o .0001 73 100.00

0.024 0.8064 105 -0.144 0.1483 102 0.109 0.2704 103 0.145 0.1424 103 0.145 0.1380 105 -0.174 0.0742 105 0.103 0.2917 105 0.213 0.0702 73 0.065 0.5082 104 100.00

0.337 0.0090 59 0.373 0.0039 58 0.707 o.0001 57 0.762 o.0001 57 0.445 0.0004 59 0.172 0.1908 59 0.618 o.0001 59 0.594 o.0001 46 0.803 o.0001 58 0.258 0.0481 59 100.00

105 0.547 o.0001 102 0.430 o.0001 103 0.571 o.0001 103 0.327 0.0007 105 0.289 0.0027 105 0.539 o.0001 105 0.360 0.0017 73 0.467 o.0001 104 0.024 0.8064 105 0.337 0.0090 59

102 0.242 0.0149 100 0.445 o .0001 100 0.420 o .0001 102 0.201 0.0427 102 0.608 o .0001 102 0.268 0.0248 70 0.351 0.0003 101 -0.144 0.1483 102 0.373 0.0039 58

103 0.752 o .0001 103 0.365 0.0001 103 0.382 o .0001 103 0.496 o .0001 103 0.787 o .0001 73 0.727 o .0001 102 0.109 0.2704 103 0.707 o .0001 57

103 0.659 o .0001 103 0.326 0.0008 103 0.669 o .0001 103 0.776 o .0001 73 0.799 o .0001 102 0.145 0.1424 103 0.762 o .0001 57

105 0.137 0.1609 105 0.598 o .0001 105 0.497 o .0001 73 0.409 o .0001 104 0.145 0.1380 105 0.445 0.0004 59

105 0.102 0.2981 105 0.254 0.0298 73 0.230 0.0184 104 -0.174 0.0742 105 0.172 0.1908 59

105 0.563 o .0001 73 0.534 o .0001 104 0.103 0.2917 105 0.618 o .0001 59

73 0.568 o .0001 73 0.213 0.0702 73 0.594 o .0001 46

104 0.065 0.5082 104 0.803 o .0001 58

108 0.258 0.0481 59

59

Sampling days (n) o108 indicate that there are missing data in PM characteristic or missing data in biological endpoint. BC: black carbon; OP: oxidative potential.

Table 4 Spearman rank correlation coefficients (r) between PM10 characteristics and biological endpoints (cell viability, IL-8 induction, oxidatively damaged DNA) in all the samples from the three locations. Cell viability

As Cd Cr Cu Mn Ni Pb Zn Endotoxin BC OP

IL8 induction

Oxidatively damaged DNA

N

R

p-value

N

R

p-value

N

R

p-value

105 102 103 103 105 105 105 73 105 104 59

 0.07  0.28  0.06  0.08  0.09 0.02  0.21 0.16  0.01  0.28  0.20

0.49 o 0.01 0.57 0.42 0.38 0.80 0.03 0.17 0.92 o 0.01 0.14

102 99 100 100 102 102 102 73 105 101 58

0.51  0.16 0.18 0.26 0.19 0.87 0.06 0.41 0.76 0.06 0.11

0.07 0.12 0.07 0.01 0.06 0.02 0.58 o 0.01 o 0.01 0.54 0.40

72 71 70 70 72 72 72 47 72 71 41

 0.08 0.26  0.21  0.13  0.10  0.28 0.00  0.16 0.06  0.05 0.11

0.51 0.03 0.08 0.30 0.38 0.02 0.99 0.27 0.63 0.68 0.50

Sampling days (n) o108 indicate that there are missing data in PM characteristic or missing data in biological endpoint. BC: black carbon; OP: oxidative potential.

different cellular responses (cell viability, pro-inflammatory activity, oxidatively damaged DNA) in airway bronchial epithelial cells. Beas-2B cells have been frequently used to study effects of particulate matter (Gualtieri et al., 2010; Oh et al., 2011; Dergham et al., 2015; Dieme et al., 2012; Cachon et al., 2014; Lepers et al., 2014). Endotoxin concentrations and OP were analyses in the PM10 samples. By exposing cells to the same PM mass, only the relation between the biological effects and the chemical composition of PM10 is reflected and not the relation to the ambient PM10 concentration. We found associations between specific constituents of PM10 and their toxic effects in Beas-2B cells.

4.1. Metals Water soluble metals have been associated with PM induced cytotoxicity and pro-inflammatory effects in cell lines (Akhtar et al., 2014; Perrone et al., 2013). A specific role of metal components present on PM in the induction of cellular responses has been suggested via the induction of oxidative stress (Donaldson and MacNee, 2001). Metals induce the formation of reactive oxygen species (ROS), via a Fenton like reaction. In this study the cytotoxic effect of PM10 was related to Cd and Pb. A correlation between cell viability reduction and metals (Al,

54

R. Van Den Heuvel et al. / Environmental Research 149 (2016) 48–56

A.

D.

120

12 Urban site 10 Il-8 induction

Cell viability

100 80 60 40

Urban site Rural site

20

Rural site Industrial site

8 6 4 2

Industrial site 0

0

0

5

10

15

20

0

10

20

BC (µg/100 µg PM10 )

E.

B. 120

40

50

12

100

10 Il-8 induction

Cell viability

30

Ni (ng/100µg PM10)

80 60 40

Urban site

20

Rural site

8 Urban site

6

Rural site

4

Industrial site

2

Industrial site 0 0

50

100

150

200

250

300

0 0.000

350

2.000

4.000

F.

12 10

Il-8 induction

8.000

10.000

8

80

Urban site

70

Rural site

60

Industrial site

6 4 2

% DNA in tai

C.

6.000

Endotoxin EU/100µg PM10 )

Pb (ng/100 µg PM10 )

outlier

50 40

Urban site

30

Rural site

20

Industrial site

10 0 0

200

400

600

800

1000

1200

1400

0 0

Zn (ng/100µg PM10 )

5

10

15

Cd (ng/100 µg PM10)

Fig. 1. Scatterplots of significant relationships between PM10 characteristic and cellular response in Beas-2B cells after exposure to 100 mg PM10/mL. A: Cell viability versus BC (r¼  0.28, p o 0.1); B: Cell viability versus Pb concentration (r¼  0.21, po 0.1); C: IL-8 induction versus Zn concentration (r¼ 0.41, po 0.01); D: IL-8 induction versus Ni concentration (r ¼0.87, p ¼0.02); E: IL-8 induction versus endotoxin concentration (r ¼0.76, p o0.01); F: Oxidatively damaged DNA versus Cd concentration (r ¼0.26, p ¼0.03). Cell viability is expressed as the percentage viable cells compared to the negative control. IL-8 induction is expressed as the ratio of IL-8 production in exposed cells compared to unexposed Beas-2B cells. Oxidatively damaged DNA is expressed as percentage DNA in tail.

Ba, Ca, Cr, Cu, Fe, Mg, Mn, Na, Si, Ti, Zn) has been reported by Akhtar et al. (2014). In accordance with literature we found a positive correlation between IL-8 induction and metal concentrations (Cu, Ni and Zn). In a Japanese study correlations were found between IL-8 production in Beas-2B with Cr and Mn (Huang et al., 2003). In primary human bronchial epithelial cells (NHBE), Il-6 and IL-8 production was significantly correlated with the group of elements Cr, Al, Si, Ti, Fe and Cu (Becker et al., 2005). The role of metal components on PM10 in the induction of IL-6, IL-8 and TNF-α (tumor necrosis factor α) was also demonstrated in human bronchial epithelial cells by inhibitor studies (Carter et al., 1997; Nakayama Wong et al., 2011). Single and multiple regression models indicated that oxidatively damaged DNA increased with increasing concentrations of Cd in PM10, which is in agreement with the literature (Perrone et al., 2010). However, significance disappeared when excluding the Cd outlier. Metal-induced DNA damage can be ascribed to the oxidative activity of metals (Møller et al., 2008).

4.2. Endotoxin Endotoxins are lipopolysaccharides (LPS) present in the outer wall of Gram-negative bacteria. Endotoxin has pro-inflammatory and adjuvant-like properties and airborne endotoxin has been associated with respiratory diseases (Rylander, 2002). Endotoxin concentrations per mg collected coarse dust have been reported and showed average endotoxin concentration of 2.14 EU/mg (Huang et al., 2003) and 10.3 EU/mg (Long et al., 2001) which is similar to the level of 2. EU/mg in PM10 in our study. Several studies reported significant associations between endotoxin content in PM and the production of cytokines/chemokines in vitro (Steenhof et al., 2011; Michael et al., 2013; Gualtieri et al., 2010; den Hartigh et al., 2010). Huang et al. (2003) showed that polymyxin-B pretreatment significantly reduced PM induced TNF-α production in Beas-2B cells, indicating the major role of endotoxin. In this study, we also observed a significant positive correlation between the endotoxin concentration and pro-inflammatory capacity of the samples. The endotoxin concentration in 10 out of

R. Van Den Heuvel et al. / Environmental Research 149 (2016) 48–56

Table 5 Significant (po 0.05) associations between biological response and PM10 characteristics after multiple regression analyses. Toxic effect

N

Estimate

95% CL

R2

55

increased il-6 production in Beas-2B cells (Veranth et al., 2006). We could not confirm this relationship in the current study.

p

5. Conclusion IL-8 induction Endotoxin

Oxidatively damaged DNA Cd

104

1.10*

1.07–1.13

0.33

o0.0001

71

3.97**

0.57–7.36

0.07

0.023

Sampling days (n) o108 indicate that there are missing data in PM characteristic or missing data in biological endpoint. Interpretation estimate: the mean toxic effect is multiplied with the estimate for an increase of the PM characteristic with the interquartile range (IQR). The IQR of the explanatory variables is calculated as the distance from the 25th percentile (P25) to the 75th percentile (P75). The effect estimates obtained from the multiple regression models are then recalculated to express the change in the effect for an increase of the explanatory variable with the IQR instead of a unit increase. * Effect calculated for an increase of endotoxin with an IQR of 0.31 EU/100 mg PM10 ** Effect calculated for an increase of Cd with an IQR of 1.3 ng/100 mg PM10

105 samples was sufficiently high to induce IL-8 production in Beas-2B cells. On these days, the presence of endotoxin in the air may partly explain the pro-inflammatory potential of the samples.

This study supports the hypothesis that, on an equal mass basis, PM10 induced biological effects differ due to differences in PM10 characteristics.. Metals (Cd, Cu, Ni and Zn), BC, and endotoxin were among the main determinants for the observed biological responses Despite the large sample size of the present study, multiple linear regression analyses did not allow to draw a straightforward conclusion on associations between in vitro cell viability and oxidatively damaged DNA and specific PM components. One of the limitations of this study is that chemical characterisation of PM10 and the in vitro responses were performed on different filters collected on the same day. Ideally, chemical characterisation of the PM samples used for in vitro exposure would be more suitable and could improve the outcome of correlation analyses. In addition, different PM sizes were not considered in the current study. It is likely that differences in composition of PM sizefractions play a role in the toxic responses. The in vitro responses represent an estimation of hazard and potency of the particles and serve as an indication of potential health impacts. Currently PM is defined in health policy for ambient PM by particle size and PM mass concentration, but studies, focusing on the correlation between PM characteristics and their toxic effects on cells may lead to new strategies to reduce the most critical health-relevant elements of ambient air.

4.3. Oxidative potential ROS can be generated from soluble transition metals (Cu, Cr, Fe, Zn, etc.) or organic compounds, i.e. polycyclic aromatic hydrocarbons (PAHs), that are present on the surface of particles. A number of methods are available to measure radical generating properties of PM including acellular and intracellular assays (Møller et al., 2014). We measured ROS generation by electron spin resonance (ESR) in acellular conditions. The formation of ROS induced by specific constituents of PM has been suggested to be an important initiating factor in PM-induced cytotoxicity, inflammation and oxidatively damaged DNA in lung epithelial cell lines (Cachon et al., 2014; Wessels et al., 2010; Oh et al., 2011; Dergham et al., 2015). The current study did not demonstrate a significant association between OP and either cell viability reduction, IL-8 induction or oxidatively damaged DNA. Despite the fact OP was correlated with all other PM10 parameters of which some (e.g. BC, Cd, Cu, Pb and Zn) were associations with toxic responses, no correlations were found. This might be explained by the fact that OP was measured directly on the filters containing the particles while biological effects were assessed in the epithelial cells. 4.4. Black carbon Black carbon (BC) is a component of PM and is formed by the incomplete combustion of fossil fuels, biofuels, and biomass. It is an indicator of soot. BC and EC (elemental carbon) represent the same fraction of PM. BC is determined by an optical method and measured online while filter-based EC levels are determined by a thermo-optical method. Recently, BC has been linked to a series of adverse human health effects (Bell et al., 2014) and has been proposed as a valuable additional air quality indicator (Janssen et al., 2011). Despite the rising interest in BC, the number of studies correlating BC levels with biological responses is limited. In the current study, BC was significantly associated with cytotoxicity. Similar results were reported by Perrone et al. (2013) who found cytotoxicity as measured by MTT response to be related with EC. It has previously been reported that EC is associated with

Funding sources The study was supported by Joaquin (Joint Air Quality Initiative) funded by the INTERREG IVB North-West Europe Programme (Grant number INTERREG IV-B 247H)and by the Flemish Government under contract numbers VMM/LUC/2012/ Genotox and LNE/OL201300015/14001/M&G.

Acknowledgement Special thanks go to S. Van Uytsel for filter extractions and to D. Ooms for in vitro cell culture work.

References Akhtar, U., Rastogi, N., McWhinney, R., Urch, B., Chow, C.-W., Evans, G., Scott, J., 2014. The combined effects of physicochemical properties ofsize-fractionated ambient particulate matter on in vitro toxicity in human A549 lung epithelial cells. Toxicol. Rep. 1, 145–156. Akhtar, U.S., McWhinney, R.D., Rastogi, N., Abbatt, J.P., Evans, G.J., Scott, J.A., 2010. Cytotoxic and proinflammatory effects of ambient and source-related particulate matter (PM) in relation to the production of reactive oxygen species (ROS) and cytokine adsorption by particles. Inhal. Toxicol. 22 (Suppl. 2), S37–S47. Allen, J., Bartlett, K., Graham, M., Jackson, P., 2011. Ambient concentrations of airborne endotoxin in two cities in the interior of British Columbia, Canada. J. Environ. Monit. 13, 631–640. Anderson, J.O., Thundiyil, J.G., Stolbach, A., 2012. Clearing the air: a review of the effects of particulate matter air pollution on human health. J. Med. Toxicol. 8, 166–175. Becker, S., Dailey, L.A., Soukup, J.M., Grambow, S.C., Devlin, R.B., Huang, Y.C., 2005. Seasonal variations in air pollution particle-induced inflammatory mediator release and oxidative stress. Environ. Health Perspect. 113, 1032–1038. Bell, M.L., Ebisu, K., Leaderer, B.P., Gent, J.F., Lee, H.J., Koutrakis, P., Wang, Y., Dominici, F., Peng, R.D., 2014. Associations of PM2.5 constituents and sources with hospital admissions: analysis of four counties in Connecticut and Massachusetts (USA) for persons 4/ ¼65 years of age. Environ. Health Perspect. 122, 138–144. Bengalli, R., Molteni, E., Longhin, E., Refsnes, M., Camatini, M., Gualtieri, M., 2013. Release of IL-1 beta triggered by Milan summer PM10: molecular pathways

56

R. Van Den Heuvel et al. / Environmental Research 149 (2016) 48–56

involved in the cytokine release. Biomed. Res. Int. 2013, 158093. Brits, E., Schoeters, G., Verschaeve, L., 2004. Genotoxicity of PM10 and extracted organics collected in an industrial, urban and rural area in Flanders, Belgium. Env. Res. 96, 109–118. Brunekreef, B., Holgate, S.T., 2002. Air pollution and health. Lancet 360, 1233. Cachon BF, Firmin S, Verdin A, Ayi-Fanou L, Billet S, Cazier F, Martin P J, Aissi F, Courcot D, Sanni A, Shirali P. 2014. Proinflammatory effects and oxidative stress within human bronchial epithelial cells exposed to atmospheric particulate matter (PM(2.5) and PM( 42.5)) collected from Cotonou, Benin. Environ. Pollut. 185 340–351. Carter, J.D., Ghio, A.J., Samet, J.M., Devlin, R.B., 1997. Cytokine production by human airway epithelial cells after exposure to an air pollution particle is metal-dependent. Toxicol. Appl. Pharm. 146, 180–188. Dergham, M., Lepers, C., Verdin, A., Cazier, F., Billet, S., Courcot, D., Shirali, P., Garcon, G., 2015. Temporal-spatial variations of the physicochemical characteristics of air pollution Particulate Matter (PM2.5–0.3) and toxicological effects in human bronchial epithelial cells (BEAS-2B). Environ. Res. 137, 256–267. Dieme, D., Cabral-Ndior, M., Garcon, G., Verdin, A., Billet, S., Cazier, F., Courcot, D., Diouf, A., Shirali, P., 2012. Relationship between physicochemical characterization and toxicity of fine particulate matter (PM2.5) collected in Dakar city (Senegal). Environ. Res. 113, 1–13. Donaldson, K., MacNee, W., 2001. Potential mechanisms of adverse pulmonary and cardiovascular effects of particulate air pollution (PM10). Int. J. Hyg. Environ. Health 203, 411. Ferguson, M.D., Migliaccio, C., Ward, T., 2013. Comparison of how ambient PMc and PM2.5 influence the inflammatory potential. Inhal. Toxicol. 25, 766–773. Gordon, T., 2007. Linking health effects to PM components, size, and sources. Inhal. Toxicol. 19 (Suppl. 1), S3–S6. Gray, D.L., Wallace, L.A., Brinkman, M.C., Buehler, S.S., La, L.C., 2015. Respiratory and cardiovascular effects of metals in ambient particulate matter: a critical review. Rev. Environ. Contam. Toxicol. 234, 135–203. Gualtieri, M., Ovrevik, J., Holme, J.A., Perrone, M.G., Bolzacchini, E., Schwarze, P.E., Camatini, M., 2010. Differences in cytotoxicity versus pro-inflammatory potency of different PM fractions in human epithelial lung cells. Toxicol. Vitr. 24, 29–39. den Hartigh, L.J., Lame, M.W., Ham, W., Kleeman, M.J., Tablin, F., Wilson, D.W., 2010. Endotoxin and polycyclic aromatic hydrocarbons in ambient fine particulate matter from Fresno, California initiate human monocyte inflammatory responses mediated by reactive oxygen species. Toxicol. Vitr. 24, 1993–2002. Healy, D.A., Hellebust, S., Silvari, V., Lopez, J.M., Whittaker, A.G., Wenger, J.C., Heffron, J.J.A., Sodeau, J.R., 2012. Using a pattern recognition approach to link inorganic chemical fingerprints of ambient Pm2.5–0.1 with in vitro biological effects. Air Qual. Atmos. Health 5, 125–147. Hellack, B., Yang, A., Cassee, F., Janssen, N.A., Schins, R.P., Kuhlbusch, T.A.J., 2014. Intrinsic hydroxyl radical generation measurements directly from sampled filters as a metric for the oxidative potential of ambient particulate matter. J. Aerosol Sci. 72, 47–55. Huang, S.L., Hsu, M.K., Chan, C.C., 2003. Effects of submicrometer particle compositions on cytokine production and lipid peroxidation of human bronchial epithelial cells. Environ. Health Perspect. 111, 478–482. Jalava, P.I., Salonen, R.O., Hälinen, A.I., Sillanpää, M., Sandell, E., Hirvonen, M.R., 2005. Effects of sample preparation on chemistry, cytotoxicity, and inflammatory responses induced by air particulate matter. Inhal. Toxicol. 17 (2), 107–117. Janssen, N.A., Hoek, G., Simic-Lawson, M., Fischer, P., van Bree, L., ten Brink, H., Keuken, M., Atkinson, R.W., Anderson, H.R., Brunekreef, B., Cassee, F.R., 2011. Black carbon as an additional indicator of the adverse health effects of airborne particles compared with PM10 and PM2.5. Environ. Health Perspect. 119 (12), 1691–1699. Lepers, C., Dergham, M., Armand, L., Billet, S., Verdin, A., Andre, V., Pottier, D., Courcot, D., Shirali, P., Sichel, F., 2014. Mutagenicity and clastogenicity of native airborne particulate matter samples collected under industrial, urban or rural influence. Toxicol. in Vitro 28, 866–874. Long, C.M., Suh, H.H., Kobzik, L., Catalano, P.J., Ning, Y.Y., Koutrakis, P., 2001. A pilot investigation of the relative toxicity of indoor and outdoor fine particles: in vitro effects of endotoxin and other particulate properties. Environ. Health Perspect. 109, 1019–1026. Møller, P., Folkmann, J.K., Forchhammer, L., Brauner, E.V., Danielsen, P.H., Risom, L., Loft, S., 2008. Air pollution, oxidative damage to DNA, and carcinogenesis. Cancer Lett. 266, 84–97. Møller, P., Jacobsen, N.R., Folkmann, J.K., Danielsen, P.H., Mikkelsen, L., Hemmingsen, J.G., Vesterdal, L.K., Forchhammer, L., Wallin, H., Loft, S., 2010. Role of oxidative damage in toxicity of particulates. Free Radic. Res. 44, 1–46. Møller, P., Danielsen, P.H., Karottki, D.G., Jantzen, K., Roursgaard, M., Klingberg, H., Jensen, D.M., Christophersen, D.V., Hemmingsen, J.G., Cao, Y., Loft, S., 2014. Oxidative stress and inflammation generated DNA damage by exposure to air pollution particles. Mutat. Res. Rev. Mutat. Res. 762, 133–166. Michael, S., Montag, M., Dott, W., 2013. Pro-inflammatory effects and oxidative stress in lung macrophages and epithelial cells induced by ambient particulate matter. Environ. Pollut. 183, 19–29. Mishra, V.K., Kumar, P., Van Poppel, M., Bleux, N., Frijns, E., Reggente, M., Berghmans, P., Panis, LI, Samson, R., 2012. Wintertime spatio-temporal variation of

ultrafine particles in a Belgian city. Sci. Total Environ. 431, 307–313. Nakayama Wong, L.S., Aung, H.H., Lame, M.W., Wegesser, T.C., Wilson, D.W., 2011. Fine particulate matter from urban ambient and wildfire sources from California's San Joaquin Valley initiate differential inflammatory, oxidative stress, and xenobiotic responses in human bronchial epithelial cells. Toxicol. in Vitro 25, 1895–1905. Oh, S.M., Kim, H.R., Park, Y.J., Lee, S.Y., Chung, K.H., 2011. Organic extracts of urban air pollution particulate matter (PM2.5)-induced genotoxicity and oxidative stress in human lung bronchial epithelial cells (BEAS-2B cells). Mutat. Res. 723, 142–151. Osornio-Vargas, A.R., Serrano, J., Rojas-Bracho, L., Miranda, J., Garcia-Cuellar, C., Reyna, M.A., Flores, G., Zuk, M., Quintero, M., Vazquez, I., Sanchez-Perez, Y., Lopez, T., Rosas, I., 2011. In vitro biological effects of airborne PM2.5 and PM10 from a semi-desert city on the Mexico-US border. Chemosphere 83, 618–626. Peng, R.D., Bell, M.L., Geyh, A.S., McDermott, A., Zeger, S.L., Samet, J.M., Dominici, F., 2009. Emergency admissions for cardiovascular and respiratory diseases and the chemical composition of fine particle air pollution. Environ. Health Perspect. 117, 957–963. Perrone, M.G., Gualtieri, M., Ferrero, L., Lo, P.C., Udisti, R., Bolzacchini, E., Camatini, M., 2010. Seasonal variations in chemical composition and in vitro biological effects of fine PM from Milan. Chemosphere 78, 1368–1377. Perrone, M.G., Gualtieri, M., Consonni, V., Ferrero, L., Sangiorgi, G., Longhin, E., Ballabio, D., Bolzacchini, E., Camatini, M., 2013. Particle size, chemical composition, seasons of the year and urban, rural or remote site origins as determinants of biological effects of particulate matter on pulmonary cells. Environ. Pollut. 176, 215–227. Pope, C.A., Burnett, R., Thun, M., 2002. Lung cancer, cardiopulmonary mortality and long-term exposure to fine particulate air pollution. J. Am. Med. Assoc. 287, 1132–1141. Remy, S., Verstraelen, S., Van Den Heuvel, R., Nelissen, I., Lambrechts, N., Hooyberghs, J., Schoeters, G., 2014. Gene expressions changes in bronchial epithelial cells: markers for respiratory sensitizers and exploration of the NRF2 pathway. Vitr. Toxicol. in Vitro 28, 209–217. Rylander, R., 2002. Endotoxin in the environment—exposure and effects. J. Endotoxin Res. 8, 241–252. Salonen, R.O., Halinen, A.I., Pennanen, A.S., Hirvonen, M.R., Sillanpaa, M., Hillamo, R., Shi, T., Borm, P., Sandell, E., Koskentalo, T., Aarnio, P., 2004. Chemical and in vitro toxicologic characterization of wintertime and springtime urban-air particles with an aerodynamic diameter below 10 microm in Helsinki. Scand. J. Work Environ. Health 30 (Suppl. 2), S80–S90. Sandström, T., Cassee, F.R., Salonen, R., Dybing, E., 2005. Recent outcomes in European multicentre projects on ambient particulate air pollution. Toxicol. Appl. Pharmacol. 207, 261. Schlesinger, R.B., Kunzli, N., Hidy, G.M., Gotschi, T., Jerrett, M., 2006. The health relevance of ambient particulate matter characteristics: coherence of toxicological and epidemiological inferences. Inhal. Toxicol. 18, 95. Silbajoris, R., Osornio-Vargas, A.R., Simmons, S.O., Reed, W., Bromberg, P.A., Dailey, L.A., Samet, J.M., 2011. Ambient particulate matter induces interleukin-8 expression through an alternative NF-kappaB (nuclear factor-kappa B) mechanism in human airway epithelial cells. Environ. Health Perspect. 119, 1379–1383. Steenhof, M., Gosens, I., Strak, M., Godri, K.J., Hoek, G., Cassee, F.R., Mudway, I.S., Kelly, F.J., Harrison, R.M., Lebret, E., Brunekreef, B., Janssen, N.A., Pieters, R.H., 2011. In vitro toxicity of particulate matter (PM) collected at different sites in the Netherlands is associated with PM composition, size fraction and oxidative potential–the RAPTES project. Part Fibre Toxicol. 8, 26. Steerenberg, P.A., van Amelsvoort, L., Lovik, M., Hetland, R., Alberg, T., Halatek, T., Bloemen, H., Rydzynski, K., Swaen, G., Schwarze, P., Dybing, E., Cassee, F.R., 2006. Relation between sources of particulate air pollution and biological effect parameters in samples from four European cities: an exploratory study. Inhal. Toxicol. 18, 333. Valavanidis, A., Fiotakis, K., Vlachogianni, T., 2008. 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, 339–362. Veranth, J.M., Moss, T.A., Chow, J.C., Labban, R., Nichols, W.K., Walton, J.C., Watson, J. G., Yost, G.S., 2006. Correlation of in vitro cytokine responses with the chemical composition of soil-derived particulate matter. Environ. Health Perspect. 114, 341–349. Verstraelen, S., Nelissen, I., Hooyberghs, J., Witters, H., Schoeters, G., Van Cauwenberge, P., Van Den Heuvel, R., 2009. Gene profiles of a human bronchial epithelial cell line after in vitro exposure to respiratory (non-)sensitizing chemicals: identification of discriminating genetic markers and pathway analysis. Toxicology 255, 151–159. Wessels, A., Birmili, W., Albrecht, C., Hellack, B., Jermann, E., Wick, G., Harrison, R. M., Schins, R.P., 2010. Oxidant generation and toxicity of size-fractionated ambient particles in human lung epithelial cells. Environ. Sci. Technol. 44, 3539–3545. Yang, A., Jedynska, A., Hellack, B., Kooter, I., Hoek, G., Brunekreef, B., Kuhlbusch, T.A., Cassee, F., Janssen, N.A., 2014. Measurement of the oxidative potential of PM2.5 and its constituents: the effect of extraction solvent and filter type. Atmos. Environ. 83, 35–42.