Accepted Manuscript

Measurement of the oxidative potential of PM2.5 and its constituents: The effect ofextraction solvent and filter type

Aileen Yang, Aleksandra Jedynska, Bryan Hellack, Ingeborg Kooter, Gerard Hoek,Bert Brunekreef, Thomas AJ. Kuhlbusch, Flemming R. Cassee, Nicole AH. Janssen

PII: S1352-2310(13)00802-9

DOI: 10.1016/j.atmosenv.2013.10.049

Reference: AEA 12541

To appear in: Atmospheric Environment

Received Date: 11 June 2013

Revised Date: 25 September 2013

Accepted Date: 23 October 2013

Please cite this article as: Yang, A., Jedynska, A., Hellack, B., Kooter, I., Hoek, G., Brunekreef, B.,Kuhlbusch, T.A., Cassee, F.R, Janssen, N.A., Measurement of the oxidative potential of PM2.5 andits constituents: The effect of extraction solvent and filter type, Atmospheric Environment (2013), doi:10.1016/j.atmosenv.2013.10.049.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service toour customers we are providing this early version of the manuscript. The manuscript will undergocopyediting, typesetting, and review of the resulting proof before it is published in its final form. Pleasenote that during the production process errors may be discovered which could affect the content, and alllegal disclaimers that apply to the journal pertain.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Measurement of the oxidative potential of PM2.5

and its constituents: The effect of extraction

solvent and filter type

Aileen Yang a,b*, Aleksandra Jedynska c, Bryan Hellack d, Ingeborg Kooter c, Gerard Hoek b,

Bert Brunekreef b,e, Thomas AJ Kuhlbusch d,f, Flemming R Cassee a,b, Nicole AH Janssen b

a National Institute for Public Health and the Environment (RIVM), P.O. Box 1, 3720BA

Bilthoven, the Netherlands;

b Institute for Risk Assessment Sciences, Division Environmental Epidemiology, Utrecht

University, P.O. Box 80.178, 3508TD Utrecht, the Netherlands;

c The Netherlands Organisation for Applied Scientific Research (TNO), Princetonlaan 6,

3584CB Utrecht, the Netherlands;

d Air Quality & Sustainable Nanotechnology, Institute of Energy and Environmental

Technology (IUTA) e.V., Blierheimer Str. 60, 47229 Duisburg, Germany;

e Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht,

Utrecht, the Netherlands

f CENIDE, University Duisburg-Essen, Duisburg, Germany

Abstract

The capacity of Particulate Matter (PM) to oxidise target molecules, defined as its oxidative

potential (OP), has been proposed as a biologically more relevant metric than PM mass.

Different assays exist for measuring OP and their methodologies vary in the choice of

extraction solvent and filter type. Little is known about the impact of extraction and filter type

on reported OP. Four a-cellular assays; electron spin resonance (ESR), dithiothreitol (DTT),

ascorbate acid depletion (AA) and reductive acridinium triggering (CRAT) assay were chosen

to evaluate whether these differences affect the OP measurement, the correlation between OP

from different assays and the association with PM chemical composition. We analysed 15

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

urban 48-72 hour PM2.5 samples collected on quartz and Teflon filters. The choice of

extraction solvent had only a significant effect on OPDTT, while all OP measures for quartz

filters were heavily attenuated. OP values derived from quartz were, however, highly

correlated with those derived from Teflon. OPDTT correlated highly with OPCRAT, and OPESR

correlated highly with OPAA. These correlations were affected by the choice of filter type.

Correlations between OP and PM chemical composition were not affected by filter type and

extraction solvent. These findings indicate that the measurement of relative OP reactivity is

not greatly influenced by filter type and extraction solvent for the investigated assays. This

robustness is also promising for exploratory use in monitoring and subsequent

epidemiological studies.

1 Introduction 1

It has been recognized that increased exposure to ambient particulate matter (PM) is 2

associated with a wide range of adverse health effects (Brunekreef and Holgate, 2002; Pope 3

III and Dockery, 2006). Currently, PM is regulated based on mass concentration, whereas 4

evidences indicate that the chemical composition, surface area and other characteristics of PM 5

are more closely linked to the induction of toxic responses (Nel, 2005). Oxidative stress, 6

initiated by the presence and formation of reactive oxygen species (ROS), or free radicals, has 7

been considered an important mechanism to the particle-induced health effects (Delfino et al., 8

2005). Although the human body is capable of dealing with ROS, diseases can overwhelm or 9

impair this host defence mechanism (Delfino et al., 2011). In that case, ROS can trigger a 10

cascade of events eventually leading to for example airway and pulmonary inflammation. 11

This in turn can cause a range of adverse effects like cell and tissue damage (Nel, 2005). 12

Oxidative potential (OP) is defined as a measure of the capacity of PM to oxidise target 13

molecules, i.e. by generating ROS in environments without living cells. It has been proposed 14

as a metric that is better related to biological responses to PM exposures and thus could be 15

more informative than mass alone (Borm et al., 2007). Several methods for testing OP have 16

been developed, but no consensus has been reached yet as to which assay is most appropriate 17

(Ayres et al., 2008). Furthermore, few inter-assay comparisons have been published. 18

Various assays exist to assess the oxidative capacity of PM, each with a different sensitivity to 19

the ROS generating compounds. Electron spin resonance (ESR) with 5,5-dimethylpyrroline-20

N-oxide (DMPO) as a spin trap, measures the ability of PM to induce hydroxyl radicals (•OH) 21

(Shi et al., 2003a, 2003b) in the presence of H2O2. The consumption of dithiothreitol (DTT) is 22

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

based on the ability of redox active compounds to transfer electrons from DTT to oxygen 23

(Cho et al., 2005; Kumagai et al., 2002). Other common assays involve measuring the ability 24

of PM to deplete antioxidants such as vitamin C, glutathione and uric acid (Mudway et al., 25

2004). Due to high detection sensitivity, fluorescent-based probes have also been used to 26

quantify PM-related ROS. These are based on the principle that a fluorescent product is 27

generated when the non-fluorescent probe molecule reacts with ROS. The most common used 28

probe is 2,7-dichlorofluorescein (DCFH) (Landreman et al., 2008). Another system with high 29

sensitivity consists of chemiluminogenic compounds, where certain acridinium esters, e.g. 4-30

methoxyphenyl-10-methylacridinium-9-carboxylate, have a high selectivity for superoxide 31

(Yamaguchi et al., 2010). Using this principle, a ROS assay using acridinium esters (CRAT) 32

as a redox probe was developed (Zomer et al., 2011) . 33

PM is usually collected on filters. After sampling, various methods are used to extract the PM 34

from the filters into suspension and then used for physical, biological, chemical and 35

toxicological analyses. The choice of extraction solvent varies between laboratories ranging 36

from deionised water to organic solvents (e.g. dichloromethane, methanol). This has an effect 37

on the efficiency of which PM species will be extracted and, as such, their toxicological 38

properties (Eiguren-Fernandez et al., 2010; Verma et al., 2012). Polytetrafluoroethylene 39

(PTFE), also known as Teflon, glass and quartz are frequently used PM sampling filter types. 40

Quartz is often used when PM is sampled for composition analysis. In the literature, most a-41

cellular assays for assessment of oxidative potential assessment have been applied to Teflon 42

filters and information on the usage of quartz filters is limited. Presently, quartz filters are 43

used as reference material within the PM10 standard (EN 12341) in the national air 44

monitoring networks in EU, while quartz, glass fibre, PTFE and PTFE-coated glass fibre are 45

allowed within the PM2.5 standard (EN 14907) 46

(“http://ec.europa.eu/environment/air/quality/legislation/assessment.htm,” 2012). 47

In the framework of the OPERA project (Oxidative Potential Exposure and Risk Assessment); 48

we aim to evaluate the value of OP as a health relevant PM metric for air quality assessment 49

and regulation. Given the different filter types and extraction methods, our primary goal for 50

this study was to assess if the aforementioned differences in methodologies might affect the 51

measurement of OP. Additionally, we aimed to see how this influenced the correlation 52

between OP from different assays and the association with PM composition. We evaluated 53

this for four acellular OP assays. ESR, depletion of ascorbate assay and the reductive 54

acridinium triggering (CRAT) assay. 55

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

2 Materials and Method 56

2.1 Study design 57

Ambient PM2.5 samples were collected at two locations in the Netherlands using Harvard 58

Impactors operating at 10 L/min flow (Air Diagnostics and Engineering Inc., Naples, Maine, 59

USA). The sampling sites were located at an urban background site in Rotterdam and along a 60

busy highway in Amsterdam. The sample volume was calculated using elapsed time 61

indicators and flow readings. The sampling periods were five (day 1 – 5) 48-72-hour 62

measurements in Rotterdam and ten (day 6 – 15) 48-72-hour measurements in Amsterdam 63

from August to November 2011. Five collocated pump units, each containing four Harvard 64

Impactors of which two used 37-mm Teflon (2 µm pore size, PVC support ring, Pall Corp., 65

NY, USA) and two 37-mm quartz filters (prebaked, QMA, Whatman - GE Healthcare 66

Biosciences Corp) were operated simultaneously. Thus, ten Teflon filters and ten quartz filters 67

were collected for each measurement period. 68

To determine particle mass concentrations, filters were weighed before and after sampling, in 69

accordance to EN14907, in a climatised room at an average temperature of 20 ± 1 °C and 50 ± 70

5 % relative humidity using a microbalance (Model MT5, Mettler-Toledo Ltd., Greifensee, 71

Switzerland) with 1 µg precision. Until processing, the filters were stored in petri dishes at 72

4°C in the dark. The OP analyses were completed by September 2012. 73

2.2 Extraction procedures 74

In our experiments, the Teflon filters were extracted with methanol (HPLC grade, Biosolve 75

BV, Valkenswaard, Netherlands) and traceSELECT ultrapure water (Sigma, Zwijndrecht, 76

Netherlands). The quartz filters were only extracted with methanol, and additionally analysed 77

directly on filter with ESR. See Supplement information (SI) Table S 1 for an overview of 78

how filter duplicates were allocated to different extraction procedure and OP analysis. 79

Methanol extraction 80

The filters were immersed with methanol in a petri dish and extracted in an ultrasonic bath 81

(Branson 5510 Ultrasonic cleaner, 40 kHz). The extract was then transferred to a rounded 82

glass flask and reduced in volume using the evaporator set (RV 10 Basic Rotary Evaporators, 83

IKA Works, VWR, USA) at 30 °C until about 1 mL was left. The filtrates were then 84

transferred to Eppendorf vials and dried overnight at 30 °C under a constant flow of nitrogen. 85

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

The quartz filter extracts were filtered through a 0.45µm PTFE syringe filter (VWR, Breda, 86

Netherlands), to remove the high amount of observed quartz fibres. 87

Water extraction 88

Water based extractions were only performed on Teflon filters using the method by Shi et al. 89

(Shi et al., 2003b). Briefly, the filters were immersed in 2 mL deionised water followed by 90

five minutes shaking, five minutes sonication (Bandelin Sonorex RK-52, 60/120 KW, 35 91

KHz) in a sonication water bath and finally five minutes vortexing (rpm 2800). 92

We attempted water extraction on the quartz filters, but it was proven difficult as the filters 93

readily absorbed water, thus requiring large quantities of water (> 5 mL). Due to the low mass 94

loading (79 – 1000 µg of PM) on several filters, adding such a substantial amount of water 95

might result in too heavily diluted suspensions. This is without taking into consideration loss 96

due to extraction. In addition, sonication destroyed the quartz fibres making it difficult to 97

separate quartz fibres from particles when filtering the suspensions. 98

2.3 Elemental composition analysis 99

Teflon filter duplicates for each sampling day, and two lab blanks were analysed with energy 100

dispersive X-ray fluorescence spectrometry (ED-XRF) at Cooper Environmental Services 101

(Portland, OR, USA) to get the elemental composition. Not enough quartz filters were 102

obtained due to pump failure, thus additional composition analysis was not possible. All 103

elements of interest (see SI Table S 3) were above detection limit (LOD) in all samples except 104

for aluminium (Al) and nickel (Ni) (1-3 samples < LOD). The limit of detection was 105

calculated as 3 times the standard deviation of the lab blanks. The coefficients of variation 106

(CV) values, as a measure for the precision of duplicate measurements, were less than 25%. 107

CV values were calculated as the sum of the squared absolute differences of the duplicates, 108

divided by two times the number of duplicates. The square root of this value was then divided 109

by the mean and multiplied by 100 to get the percentage (Eeftens et al., 2012). 110

Carbon analysis (EC/OC) was performed on quartz filers using a Thermal Optical 111

Transmittance (TOT) analyser (Sunset Laboratory Inc., Tigard, OR, USA). From each filter, a 112

punched sample of 1.5 cm2 was analysed with the temperature protocol EUSAAR_2; a 113

standard protocol developed for European aerosols (Cavalli et al., 2010). Optical 114

transmittance is used to correct for charring. For quality assurance, we analysed every 4th 115

filter twice. 116

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

2.4 Oxidative potential assays 117

2.4.1 DTT 118

The DTT assay measures the presence of reactive oxygen species via formation of the DTT-119

disulfide due to transfer electrons from DTT to ROS by recycling chemicals such as quinones 120

(Cho et al., 2005). Typically compounds which react in this assay are organic species (e.g. 121

quinones), but previous studies have shown that transition metals can also oxidize DTT 122

(Charrier and Anastasio, 2012; Lin and Yu, 2011) 123

In brief, aliquots of PM2.5 water-reconstituted methanol and water extracts were incubated 124

with DTT (Sigma, Zwijndrecht). The reaction was stopped at designated time points (0, 10, 125

20, 30 minutes), adding 5,5′-Dithiobis(2-nitrobenzoic acid) (DTNB) (Sigma, Zwijndrecht). 126

The absorbance at 412 nm is recorded, and the rates are calculated using linear regression of 127

the data as seen from a plot of absorbance against time. The results are expressed as nmol 128

DTT/min per sampled volume or per µg of PM. Domestic oil burning furnace (DOFA, 129

obtained from US EPA, RTP, NC) with a fixed concentration was used as a positive control 130

and ultrapure water as a negative control 131

2.4.2 Ascorbate (AA) depletion assay 132

The AA assay is a simplified version of the synthetic respiratory tract lining fluid (RTFL) 133

assay (Zielinski et al., 1999), where only ascorbate acid is used. This assay can be used to 134

quantify the transition metal-based redox activities, but has also shown to be sensitive to 135

quinones (Roginsky et al., 1999). 136

AA analysis was performed according to the protocol by Mudway et al. (2011), but 137

TraceSELECT ultrapure water was used instead of Chelex-resin treated water. Briefly, PM 138

extracts are incubated in the spectrophotometer (spectraMAX 190: Molecular Devices, 139

Sunnyvale, USA) for 10 min at 37 °C. After adding ascorbate acid, the absorption at 265 nm 140

is measured every 2 min for 2 hours. The 96-wells plate is auto shaken for 3 seconds before 141

each measurement. The maximum depletion rate of ascorbic acid is determined by linear 142

regression of the linear section data, plotted as absorbance against time. The results are 143

expressed as nmol/s of max AA depletion per sampled volume or per µg of PM. Domestic oil 144

burning furnace (DOFA obtained from US EPA, RTP, NC) with a fixed concentration was 145

used as a positive control and ultrapure water as a negative control 146

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

2.4.3 ESR 147

Two different ESR measurements approaches were applied, one for the different filter extracts 148

and an improved method for the quartz fibre filters without any extraction procedure (ESR-149

direct). Both approaches are based on the trapping of PM induced (hydroxyl radicals) OH• 150

mainly generated via Fenton-type reaction in presence of H2O2 and 5,5-dimethyl-1-pyrroline-151

N-oxide (DMPO) as spin trap. Transition metals are especially sensitive to H2O2 oxidation 152

and generation of OH-radical. 153

The preparation and analyses of the extracted filter samples were done according to the 154

method by Shi et al. (2003b) without the described filtering step of the sample after 155

incubation and prior the ESR analyses. Briefly, PM suspensions are mixed with the chemical 156

ingredients (H2O2 and DMPO), followed by incubation for 15 minutes at 37 °C in a heated 157

shaking water bath prior to ESR analysis. The ESR quantification was conducted with the 158

Analysis Software (2.0 or higher, Magnettech GmbH, Berlin) on first derivation of ESR 159

signals of DMPO–OH quartet as the average of total amplitudes and expressed in arbitrary 160

units (A.U.), expressed per sampled volume or per µg of PM. Tempol (4-hydroxy-2,2,6,6-161

tetramethylpiperidinyloxy) was used as an internal standard, and CuSO4 as a positive control. 162

2.4.4 CRAT assay 163

The chemiluminescence reaction of acridinium ester under slightly basic conditions forms the 164

basis of the CRAT assay. ROS production is measured from the interaction of reductants and 165

oxidants (Zomer et al., 2011). The CRAT assay uses DTT as the reducing agent leading to 166

formation of hydrogen peroxide, which in turn reacts with acridinium ester after addition of a 167

buffer. The light emitted during this reaction is measured for about 1 second with 168

luminescence meter. This assay is sensitive to oxidants such as ferric or cupric ions or organic 169

species (e.g. quinones). 170

Briefly, 50 µL of sample with known concentration is incubated with 50 µL of 10 mM DTT 171

for 10 minutes at 438 rpm in the plate reader. After incubation, 30 µL of acridinium ester (0.5 172

µg/mL in 0.1M HNO3,) is added. The luminescence is measured in the kinetic mode after 173

adding 50 µL 400 mM carbonate buffer pH 9.4 during 1 second. ROS production assay is 174

performed using Mithras LB 940 Luminescence meter (Berthold) and 96 white micro plates. 175

During the measurement the 1,2-Naphtoquinone (NQ) is used as calibration line. The results 176

are quantified as equivalent of 1,2-Naphtoquinone per sample PM mass (pmol NQ/µg). 9,10-177

Phenanthrenequinone is used for quality control. 178

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

A more detailed methodology description of the assays is provided in the Supplement 179

Information (SI). 180

2.5 Statistical analysis 181

Correlations for each OP method for different extraction methods and filter types were 182

calculated using Spearman rank correlation coefficient (rS). p < 0.05 was considered 183

statistically significant. We did not differentiate between sampling site and sampling day due 184

to too few data points. The paired sample t-test was used to examine whether filter type and 185

extraction solvent differed significantly for each OP method. 186

3 Results and discussion 187

To investigate measurement precision, CVs for all duplicates used for OP analysis were 188

calculated (See Table 1). Quartz filter extracts gave generally the highest CV values (15 – 189

38%), and the water extracted PM from Teflon filters the lowest (8 – 22%). The DTT and 190

CRAT assay gave respectively the lowest and highest CV values. Corresponding CV values 191

of the PM mass concentrations are shown in SI Table S 2. 192

3.1 Quartz filter versus Teflon filter 193

To our knowledge, no study has been published on the effect of filter type on oxidative 194

potential of PM2.5. Irrespective of assay type, OP/m3 values from quartz filters were 195

significantly lower than those from Teflon filters (see Table 2). The AA and CRAT assay 196

showed the largest difference in OP values, as the average OPCRAT/m3 and OPAA/m3 for quartz 197

filters was 63% and 66% lower respectively. The DTT assay yielded the lowest decrease 198

(21%) in OP values, while OPESR/m3 was 47% lower for quartz filters. 199

The attenuation of the OP values for the metal-driven assays (AA and ESR) might suggest 200

lower extraction efficiency of the OP reactive components from the quartz filters compared to 201

Teflon filters. OP values from the DTT assay were less affected, indicating that reactive 202

components for this assay (e.g. organic species) were readily extracted from the quartz filters. 203

The filtering of the quartz filters extracts, which was inevitable due to the high amount of 204

quartz fibres, also led to removal of insoluble PM species. This might have contributed further 205

to the attenuation of the OP values. Yet, as seen in Table 3 and Figure 1, we found high 206

correlations between OP values derived from Teflon and those from quartz filters. This 207

suggests that the choice of sample collection media may have a profound effect on the 208

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

absolute value for each OP assay, but not on the relative measure of OP reactivity for these 209

four assays. 210

3.2 Methanol extracts versus water extracts 211

We only observed a significant effect of the choice of extraction solvent for the DTT assay (p 212

<0.01, paired t-test), with a lower OPDTT/m3 for the water extracts. Water extracts were not 213

analysed with the CRAT assay. As mentioned in section 2.2, we did not perform water 214

extraction on quartz filters, thus the comparison has only been done for Teflon extracts. 215

Several studies assessed the choice of extraction solvent on OP assessment. Eiguren-216

Fernandez et al. (2010) compared the DTT activity of ambient PM with two extraction 217

methods for Teflon filters; dichloromethane extraction including filtering and water 218

extraction. They found the water extracts to be over an order of magnitude more reactive than 219

the dichloromethane extracts (Eiguren-Fernandez et al., 2010). Verma et al. (2012) measured 220

the DTT consumption of water and methanol PM2.5 extracts from quartz filters Both extracts 221

were filtered, but the methanol extract still produced significantly higher DTT reactivity 222

(expressed per µg of PM mass) than the water extract (Verma et al., 2012). Moreover, 223

Rattanavara et al. (2011) also tested out the difference in extraction solvent for the DTT assay, 224

and found for 1,4-Naphtoquinone, a greater DTT consumption when extracted with methanol 225

compared to water. Methanol and dichloromethane have similar extraction properties, but the 226

former has higher polarity and is therefore also able to extract the hydrophilic compounds in 227

addition to the hydrophobic organic components. Since we also did not filter the methanol 228

extracts, the methanol-insoluble components are also retained. Daher et al. (2011) examined 229

the DTT activity for different sampling methods and reported the highest OPDTT/mg for the 230

Biosampler, which is considered most efficient in collecting both the insoluble and soluble 231

PM species. Consistent with our results, this suggests that the use of methanol to retrieve DTT 232

reactive components might be more efficient than using water. 233

The choice of extraction solvent had no significant effect on the ESR and AA assay, showing 234

that the OP reactive components (i.e. water-soluble transition metals) are equally effectively 235

extracted with water as with methanol. 236

We also used ESR to assess OP for quartz filters without the extraction step (ESR-direct), 237

which shortened analysis time. ESR-direct had the lowest OP values among the ESR assays. 238

This might be caused by the fact that not all components are available for reacting with H2O2 239

(and subsequently DMPO). Despite the attenuation of the ESR-direct signal, the difference in 240

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

OP between methanol quartz extract and ESR-direct was not significant. More importantly, 241

ESR-direct was very highly correlated with OPESR/m3and OPAA/m3 measured from PM 242

suspensions. This indicates that the ESR-direct may be a promising method to use in future 243

OP studies. 244

3.3 Correlations between OP measurement methods 245

Furthermore, the effect of filter type and extraction solvent between the different OP methods 246

by was explored by calculating Spearman rank correlation coefficient shown in Table 3. 247

Generally, the correlations between different OP values were affected by filter type (lower for 248

quartz) and less so by extraction fluid. As discussed in section 3.1, the filtering of PM 249

suspension of the quartz filters might be one of the reasons for the attenuated OP reactivity. 250

Two of the chosen OP assays, ESR and AA, are mainly sensitive towards the transition metals 251

which trigger the formation of OH radicals (Godri et al., 2009; Shi et al., 2003b), and the 252

strong correlations between OPESR and OPAA confirms this. Similar strong correlations were 253

found for the DTT and CRAT assay, which rely on organic compound-mediated, and the 254

latter also to a certain extent metals (Zomer et al., 2011). To what extent metals influence the 255

DTT reactivity is a complex issue yet to be solved (Charrier and Anastasio, 2012; Lin and Yu, 256

2011). 257

Few inter-assay comparisons have been published so far. A comparison study was done by 258

Künzli et al. (2006) where OP was assessed for water extracted PM2.5 Teflon samples from 20 259

European cities using the ESR with DMPO as spin trap and RTFL assay reporting the AA 260

dependent OP value. A moderate correlation was found (Pearson’s correlation, r = 0.65) 261

between OPESR and OPAA, consistent with our results, although we applied the simplified AA-262

only assay. We observed moderate (rS = 0.63 – 0.69) to very high correlations (rS >0.90) 263

between OPAA/m3and OPESR/m3 for both filter types and extraction solvent, with the highest 264

correlation for Teflon filter extracts. 265

Mudway et al. (2011) compared the DTT, AA depletion from a complex RTFL assay, and the 266

AA-only assay for methanol extracts from TEOM filters. They observed no significant 267

correlations between DTT and the two AA assays as with our results, but found a quantitative 268

association (Pearson’s correlation, r = 0.74) between the simplified AA assay and the AA 269

depletion from RTFL assay (Mudway et al., 2011). 270

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

We found no significant correlations between OPDTT/m3 and OPAA/m3, or between OPDTT/m3 271

and OPESR/m3. This suggests that a combination of OPDTT and OPESR, or OPDTT and OPAA 272

assay, might provide complementary results regarding their oxidative properties. 273

3.4 Correlations between OP and PM chemical constituents 274

The results of the correlations between PM chemical composition and OP methods are shown 275

in Table 4. Generally, we found no major impact of filter type and choice of extraction fluid 276

on these correlations. 277

OPESR/m3 and OPAA/m3 were strongly correlated with Cu, Fe, and EC, and moderately with 278

Zn. These correlations were also found when OP is expressed per µg of PM for these two 279

assays (see SI Table S 5). OPESR/m3 and OPAA/m3 were also not correlated with PM mass 280

concentration. High correlations between OPESR, OPAA and the transition metal 281

concentrations are consistent with previous observations (Boogaard et al., 2011; Godri et al., 282

2009; Künzli et al., 2006; Nawrot et al., 2009; Shi et al., 2003b). 283

OPDTT/m3 correlated moderately (rS = 0.61 – 0.68) with Cu, Fe, Mn and Zn, and highly (rS = 284

0.86 -0.96) with K, Br and OC. When expressed per µg of PM, no significant positive 285

correlations were found for the latter, but the correlations with Cu and Fe remained. 286

OPCRAT/m3 correlated moderately to highly (rS = 0.66 – 0.88) to the same elements as 287

OPDTT/m3, with the exception of Mn. Both OPCRAT/m3 and OPDTT/m3 correlated very highly 288

with PM mass concentration. This strong correlation with PM mass concentrations may also 289

have contributed to the high correlations between DTT, CRAT and Br, K, and OC. This is 290

obvious when the correlations are expressed per µg of PM (see SI Table S 5) and the 291

mentioned correlations disappear. As seen in SI Table S 7, DTT normalised by PM mass 292

resulted in low OP variation between the samples, which might have led to the high 293

correlations to PM mass concentrations when expressed per m3. Another reason of the high 294

correlations might be the low sample variation for some of the PM components (see SI Table 295

S 4). 296

The DTT assay has been reported to be predominately reactive towards PM quinone content 297

and insensitive towards metals (Cho et al., 2005). However, recent studies have shown that 298

DTT might be associated with the water-soluble metals, but the overall interactions between 299

DTT and metal ions have proven to be complex (Charrier and Anastasio, 2012; Lin and Yu, 300

2011; Ntziachristos et al., 2007). 301

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

The PM chemical constituents in our study are derived by ED-XRF, which only provides the 302

total elemental concentrations in the samples. As mentioned earlier, due to pump failure, we 303

did not have enough quartz filters to do elemental analysis for filter comparison. Less detailed 304

information is provided with only ED-XRF, as the redox activities of PM associated elements 305

depend on the chemical speciation and oxidative state of the metals (Shi et al., 2003a). 306

However, the total elemental concentrations do offer an insight, albeit limited, on the 307

interactions between the different OP assays and PM constituents, and also to which extent 308

choice of extraction solvent and filter type affected these correlations. Furthermore, it was not 309

within the scope of our paper to look at the complex interactions between the different OP 310

assays and the various fractions of soluble/insoluble PM components. 311

4 Conclusions 312

We evaluated the effect of the choice of extraction solvent and filter type on four OP assays. 313

Extraction solvent only had a significant effect on OPDTT, but not on OPESR and OPAA. We 314

observed high correlations between OPDTT and OPCRAT, and between OPESR and OPAA. These 315

correlations were affected by filter type and to a lesser extent choice of extraction fluid. OPESR 316

and OPAA were highly correlated with Cu, Fe, Zn and EC, but not with PM mass. OPDTT and 317

OPCRAT were highly correlated with PM mass, OC, Br, K, and S. These correlations were not 318

affected by the choice of filter type and extraction fluid. 319

Despite the difference in extraction procedure, which likely led to the heavy attenuation of the 320

OP values for the quartz filters, we still found strong correlations with OP/m3 values obtained 321

from Teflon filter for each assay, and between the expected assay types. This indicates that 322

the measurement of the relative OP reactivity is not greatly influenced by filter type for the 323

four assays we applied in our study. These findings are promising for exploratory use in 324

monitoring and subsequent epidemiological studies. 325

The ESR direct method, where ROS formation is measured directly on the filter, showed 326

promising results, with high correlations to the ESR results from suspensions. By omitting the 327

extraction step, analysis time is also shortened, which is an advantage for routine monitoring 328

work. 329

330

Acknowledgements 331

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

We would like to thank John Boere, Paul Fokkens, Daan Leseman from RIVM, Kees 332

Meliefste (IRAS) and Maciej Strak for helping with the sampling campaigns and lab work. 333

We would also like to thank Saskia van der Zee (GGD Amsterdam) for assistance in locating 334

the sampling site in Amsterdam. Lastly, thanks to Astrid John and Mareile Renker (IUTA) 335

for performing the EC/OC analysis. The OPERA project is funded by the RIVM Strategic 336

Research Program (SOR) 337

338

AUTHOR INFORMATION 339

Corresponding Author 340

*Phone: +31 30 2743316; fax: +31 30 2744451; e-mail: aileen.yang@rivm.nl 341

342

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

14

REFERENCES

Ayres, J.G., Borm, P., Vincent, C., Donaldson, Ken, Ghio, Andy, Harrison, Roy M, Hider,

Robert, Kelly, Frank, Kooter, Ingeborg M., 2008. Evaluating the Toxicity of Airborne

Particulate Matter and Nanoparticles by Measuring Oxidative Stress Potential—A

Workshop Report and Consensus Statement, Inhalation Toxicology, Informa

Healthcare. Inhalation Toxicology 20, 75–99.

Boogaard, H., Janssen, N.A.H., Fischer, P.H., Kos, G.P.A., Weijers, E.P., Cassee, F.R., van

der Zee, S.C., de Hartog, J.J., Brunekreef, B., Hoek, G., 2011. Contrasts in Oxidative

Potential and Other Particulate Matter Characteristics Collected Near Major Streets

and Background Locations. Environmental Health Perspectives 120, 185–191.

Borm, P.J.A., Kelly, F., Künzli, N., Schins, R.P.F., Donaldson, K., 2007. Oxidant generation

by particulate matter: from biologically effective dose to a promising, novel metric.

Occupational and Environmental Medicine 64, 73–74.

Brunekreef, B., Holgate, S.T., 2002. Air pollution and health. Lancet 360, 1233–1242.

Cavalli, F., Viana, M., Yttri, K.E., Genberg, J., Putaud, J.-P., 2010. Toward a standardised

thermal-optical protocol for measuring atmospheric organic and elemental carbon: the

EUSAAR protocol. Atmospheric Measurement Techniques 3, 79–89.

Charrier, J.G., Anastasio, C., 2012. On dithiothreitol (DTT) as a measure of oxidative

potential for ambient particles: evidence for the importance of soluble \newline

transition metals. Atmospheric Chemistry and Physics 12, 9321–9333.

Cheung, K.L., Ntziachristos, L., Tzamkiozis, T., Schauer, J.J., Samaras, Z., Moore, K.F.,

Sioutas, C., 2010. Emissions of Particulate Trace Elements, Metals and Organic

Species from Gasoline, Diesel, and Biodiesel Passenger Vehicles and Their Relation

to Oxidative Potential. Aerosol Science and Technology 44, 500.

Cho, A.K., Sioutas, C., Miguel, A.H., Kumagai, Y., Schmitz, D.A., Singh, M., Eiguren-

Fernandez, A., Froines, J.R., 2005. Redox activity of airborne particulate matter at

different sites in the Los Angeles Basin. Environmental Research 99, 40–47.

Daher, N., Ning, Z., Cho, A.K., Shafer, M., Schauer, J.J., Sioutas, C., 2011. Comparison of

the Chemical and Oxidative Characteristics of Particulate Matter (PM) Collected by

Different Methods: Filters, Impactors, and BioSamplers. Aerosol Science and

Technology 45, 1294–1304.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

15

Delfino, R., Staimer, N., Vaziri, N., 2011. Air pollution and circulating biomarkers of

oxidative stress. Air Quality, Atmosphere & Health 4, 37–52.

Delfino, R.J., Sioutas, C., Malik, S., 2005. Potential Role of Ultrafine Particles in

Associations between Airborne Particle Mass and Cardiovascular Health. Environ

Health Perspect 113, 934–946.

DiStefano, E., Eiguren-Fernandez, A., Delfino, R.J., Sioutas, C., Froines, J.R., Cho, A.K.,

2009. Determination of metal-based hydroxyl radical generating capacity of ambient

and diesel exhaust particles. Inhalation Toxicology 21, 731–738.

Eeftens, M., Tsai, M.-Y., Ampe, C., Anwander, B., Beelen, R., Bellander, T., Cesaroni, G.,

Cirach, M., Cyrys, J., de Hoogh, K., De Nazelle, A., de Vocht, F., Declercq, C.,

Dėdelė, A., Eriksen, K., Galassi, C., Gražulevičienė, R., Grivas, G., Heinrich, J.,

Hoffmann, B., Iakovides, M., Ineichen, A., Katsouyanni, K., Korek, M., Krämer, U.,

Kuhlbusch, T., Lanki, T., Madsen, C., Meliefste, K., Mölter, A., Mosler, G.,

Nieuwenhuijsen, M., Oldenwening, M., Pennanen, A., Probst-Hensch, N., Quass, U.,

Raaschou-Nielsen, O., Ranzi, A., Stephanou, E., Sugiri, D., Udvardy, O., Vaskövi, É.,

Weinmayr, G., Brunekreef, B., Hoek, G., 2012. Spatial variation of PM2.5, PM10,

PM2.5 absorbance and PM coarse concentrations between and within 20 European

study areas and the relationship with NO2 – Results of the ESCAPE project.

Atmospheric Environment 62, 303–317.

Eiguren-Fernandez, A., Shinyashiki, M., Schmitz, D.A., DiStefano, E., Hinds, W., Kumagai,

Y., Cho, A.K., Froines, J.R., 2010. Redox and electrophilic properties of vapor- and

particle-phase components of ambient aerosols. Environmental Research 110, 207–

212.

Godri, K.J., Duggan, S.T., Fuller, G.W., Baker, T., Green, D., Kelly, F.J., Mudway, I.S.,

2009. Particulate Matter Oxidative Potential from Waste Transfer Station Activity.

Environ Health Perspect 118, 493–498.

http://ec.europa.eu/environment/air/quality/legislation/assessment.htm [WWW Document],

2012. URL http://ec.europa.eu/environment/air/quality/legislation/assessment.htm

(accessed 1.6.13).

Kumagai, Y., Koide, S., Taguchi, K., Endo, A., Nakai, Y., Yoshikawa, T., Shimojo, N., 2002.

Oxidation of Proximal Protein Sulfhydryls by Phenanthraquinone, a Component of

Diesel Exhaust Particles. Chemical Research in Toxicology 15, 483–489.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

16

Künzli, N., Mudway, I.S., Götschi, T., Shi, T., Kelly, F.J., Cook, S., Burney, P., Forsberg, B.,

Gauderman, J.W., Hazenkamp, M.E., Heinrich, J., Jarvis, D., Norbäck, D., Payo-Losa,

F., Poli, A., Sunyer, J., Borm, P.J.A., 2006. Comparison of Oxidative Properties, Light

Absorbance, and Total and Elemental Mass Concentration of Ambient PM2.5

Collected at 20 European Sites. Environ Health Perspect 114, 684–690.

Landreman, A.P., Shafer, M.M., Hemming, J.C., Hannigan, M.P., Schauer, J.J., 2008. A

Macrophage-Based Method for the Assessment of the Reactive Oxygen Species

(ROS) Activity of Atmospheric Particulate Matter (PM) and Application to Routine

(Daily-24 h) Aerosol Monitoring Studies. Aerosol Science and Technology 42, 946.

Li, N., Sioutas, C., Cho, A., Schmitz, D., Misra, C., Sempf, J., Wang, M., Oberley, T.,

Froines, J., Nel, A., 2002. Ultrafine Particulate Pollutants Induce Oxidative Stress and

Mitochondrial Damage. Environ Health Perspect 111, 455–460.

Lin, P., Yu, J.Z., 2011. Generation of Reactive Oxygen Species Mediated by Humic-like

Substances in Atmospheric Aerosols. Environ. Sci. Technol. 45, 10362–10368.

Mudway, I., Fuller, G.W., Green, D., Dunster, C., Kelly, F.J., 2011. Report: Quantifying the

London Specific Component of PM10 Oxidative Activity - Defra, UK. Defra -

Department of Environment Food and Rural Affairs.

Mudway, I.S., Stenfors, N., Duggan, S.T., Roxborough, H., Zielinski, H., Marklund, S.L.,

Blomberg, A., Frew, A.J., Sandström, T., Kelly, F.J., 2004. An in vitro and in vivo

investigation of the effects of diesel exhaust on human airway lining fluid

antioxidants. Archives of Biochemistry and Biophysics 423, 200–212.

Nawrot, T.S., Kuenzli, N., Sunyer, J., Shi, T., Moreno, T., Viana, M., Heinrich, J., Forsberg,

B., Kelly, F.J., Sughis, M., Nemery, B., Borm, P., 2009. Oxidative properties of

ambient PM2.5 and elemental composition: Heterogeneous associations in 19

European cities. Atmospheric Environment 43, 4595–4602.

Nel, A., 2005. Air Pollution-Related Illness: Effects of Particles. Science 308, 804–806.

Ntziachristos, L., Froines, J.R., Cho, A.K., Sioutas, C., 2007. Relationship between redox

activity and chemical speciation of size-fractionated particulate matter. Part Fibre

Toxicol 4, 5–5.

Pope III, C.A., Dockery, D.W., 2006. Health effects of fine particulate air pollution: Lines

that connect. Journal of the Air and Waste Management Association 56, 709–742.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

17

Rattanavaraha, W., Rosen, E., Zhang, H., Li, Q., Pantong, K., Kamens, R.M., 2011. The

reactive oxidant potential of different types of aged atmospheric particles: An outdoor

chamber study. Atmospheric Environment 45, 3848–3855.

Roginsky, V.A., Barsukova, T.K., Stegmann, H.B., 1999. Kinetics of redox interaction

between substituted quinones and ascorbate under aerobic conditions. Chemico-

Shi, T., Knaapen, A., Begerow, J., Birmili, W., Borm, P., Schins, R., 2003a. Temporal

variation of hydroxyl radical generation and 8-hydroxy-2’-deoxyguanosine formation

by coarse and fine particulate matter. Occup Environ Med 60, 315–321.

Shi, T., Schins, R., Knaapen, A., Kuhlbusch, T., Pitz, M., Heinrich, J., Borm, P., 2003b.

Hydroxyl radical generation by electron paramagnetic resonance as a new method to

monitor ambient particulate matter composition. J. Environ. Monitor. 5, 550.

Verma, V., Pakbin, P., Cheung, K.L., Cho, A.K., Schauer, J.J., Shafer, M.M., Kleinman,

M.T., Sioutas, C., 2011. Physicochemical and oxidative characteristics of semi-

volatile components of quasi-ultrafine particles in an urban atmosphere. Atmospheric

Environment 45, 1025–1033.

Verma, V., Rico-Martinez, R., Kotra, N., King, L., Liu, J., Snell, T.W., Weber, R.J., 2012.

Contribution of Water-Soluble and Insoluble Components and Their

Hydrophobic/Hydrophilic Subfractions to the Reactive Oxygen Species-Generating

Potential of Fine Ambient Aerosols. Environ. Sci. Technol. 46, 11384–11392.

Yamaguchi, S., Kishikawa, N., Ohyama, K., Ohba, Y., Kohno, M., Masuda, T., Takadate, A.,

Nakashima, K., Kuroda, N., 2010. Evaluation of chemiluminescence reagents for

selective detection of reactive oxygen species. Analytica Chimica Acta 665, 74–78.

Zielinski, H., Mudway, I.S., Bérubé, K.A., Murphy, S., Richards, R., Kelly, F.J., 1999.

Modeling the interactions of particulates with epithelial lining fluid antioxidants.

American Journal of Physiology - Lung Cellular and Molecular Physiology 277, L719

–L726.

Zomer, B., Collé, L., Jedyńska, A., Pasterkamp, G., Kooter, I., Bloemen, H., 2011.

Chemiluminescent reductive acridinium triggering (CRAT)—mechanism and

applications. Analytical and Bioanalytical Chemistry 401, 2945–2954.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Figure 1 Association between extraction methods and filter types for the different OP assays

normalised by sampled volume. Correlation analysis from this figure is given in Table 3.

Removing the outliner for CRAT assay results in an rS=0.90.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

1

TABLES

Table 1 Coefficient of variation (CV) values for filter duplicates used in the different OP methods, expressed as OP/m3. (Met = Methanol, Q = quartz, T = Teflon).

Ascorbate

(nmol AA/s/m3)

DTT

(nmol DTT/min/m3)

ESR

(A.U./m3)

CRAT

(pmol NQ/m3)

Met

Q

Met

T

Water

T

Met

Q

Met

T

Water

T

Met

Q

Met

T

Water

T

Direct

Q

Met

Q

Met

T

Duplicates 14a 15 15 14a 15 15 14a 15 15 14c 14a 15

CV (%) 22 21 17 15 11 9 19 29b 22 29 36 22

a Filter omitted from calculations due to technical problems during sampling. b High CV caused by one poor duplicate with an individual CV of 82%; without this duplicate, the CV decreases to 13%. c One ESR signal was below detection limit.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

2

Table 2: Effect of extraction fluid and filter type on OP values for four assays. T= Teflon, Q= quartz, Met = Methanol, n=15.

Ascorbate

(nmol AA/s/m3)

DTT

(nmol DTT/min/m3)

ESR

(A.U./m3)

CRAT

(pmol NQ/m3)

Met Q

Met T

Water T

Met Q

Met T

Water T

Met Q

Met T

Direct

Q

Water T

Met Q

Met T

Mean 5.5* 16.1 18.4 2.3* 2.9 2.2# 2058* 3883 1777 3885 636* 1732

Std.dev 3.8 10.9 14.8 1.3 1.7 1.3 1204 2068 1026 2489 632 1419

Min 0.8 3.23 2.2 0.4 0.6 0.5 499 762 407 648 38 98

Max 12.3 35.0 43.5 5.3 7.2 5.2 4866 7961 3733 8176 2192 4488 *Filter type: Tested against methanol-Teflon, paired t-test, significant at p<0.01 #Extraction solvent: tested against methanol-Teflon, paired t-test, significant at p<0.01.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

3

Table 3 Spearman’s correlation coefficient between OP methods, the lower shaded area is expressed per m3 and the upper part is expressed per µg. (n=15)

Ascorbate DTT ESR CRAT

Methanol Quartz

Methanol Teflon

Water Teflon

Methanol Quartz

Methanol Teflon

Water Teflon

Methanol Quartz

Methanol Teflon

Water Teflon

ESR Direct

Methanol Quartz

Methanol Teflon

Methanol Quartz

0.94** 0.96** 0.04 0.53* 0.60* 0.92** 0.86** 0.88** 0.90** 0.10 0.36

Methanol Teflon

0.94** 0.96** 0.03 0.61* 0.63* 0.93** 0.94** 0.93** 0.88** 0.22 0.51* Ascorbate

Water Teflon

0.93** 0.95** 0.03 0.54* 0.62* 0.89** 0.87** 0.91** 0.91** 0.14 0.45

Methanol Quartz

0.30 0.21 0.27 0.36 0.48 0.20 -0.07 0.00 0.09 0.23 0.08

Methanol Teflon

0.41 0.36 0.36 0.88** 0.72** 0.61* 0.50 0.49 0.40 0.61* 0.50 DTT

Water Teflon

0.31 0.27 0.30 0.91** 0.97** 0.62* 0.58* 0.58* 0.55* 0.49 0.56*

Methanol Quartz

0.69** 0.63* 0.66** 0.79** 0.81** 0.80** 0.87** 0.93** 0.83** 0.31 0.54*

Methanol Teflon

0.88** 0.91** 0.94** 0.38 0.49 0.44 0.76** 0.94** 0.87** 0.19 0.46

Water Teflon

0.93** 0.95** 0.96** 0.34 0.46 0.41 0.75** 0.97** 0.81** 0.29 0.60* ESR

ESR Direct

0.94** 0.91** 0.94** 0.29 0.34 0.28 0.69** 0.90** 0.91** -0.10 0.17

Methanol Quartz

0.39 0.36 0.37 0.91** 0.93** 0.93** 0.85** 0.52* 0.48 0.35 0.70**

CRAT Methanol

Teflon 0.53* 0.56* 0.53* 0.63* 0.82** 0.81** 0.83** 0.64** 0.67** 0.51* 0.78**

**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

4

Table 4 Spearman’s correlation coefficient between PM characteristics and OP methods (n=15), expressed per m3. (Met = Methanol, Q = Quartz, T= Teflon).

Ascorbate DTT ESR CRAT

(nmol AA/s/m3) (nmol DTT/min/m3) (A.U./m3) (pmol NQ/m3)

Met Q

Met T

Water T

Met Q

Met T

Water T

Met Q

Met T

Water T

Direct

Q Met Q

Met T

Mass Teflon 0.27 0.22 0.23 0.85** 0.95** 0.97** 0.72** 0.35 0.34 0.16 0.90** 0.78**

Mass quartz 0.29 0.25 0.26 0.85** 0.97** 0.98** 0.75** 0.37 0.38 0.20 0.91** 0.80**

EC 0.88** 0.86** 0.93** 0.38 0.35 0.32 0.70** 0.93** 0.90** 0.87** 0.41 0.46

OC 0.19 0.13 0.24 0.88** 0.95** 0.96** 0.80** 0.35 0.40 0.18 0.88** 0.76**

Al 0.43 0.25 0.39 0.80** 0.74** 0.72** 0.64** 0.35 0.35 0.39 0.69** 0.41

Br 0.29 0.27 0.33 0.77** 0.86** 0.88** 0.63* 0.41 0.39 0.19 0.79** 0.73**

Cr 0.30 0.25 0.29 0.57* 0.41 0.46 0.51 0.29 0.31 0.34 0.33 0.29

Cu 0.83** 0.80** 0.88** 0.62* 0.66** 0.63* 0.89** 0.91** 0.91** 0.83** 0.70** 0.73**

Fe 0.80** 0.79** 0.86** 0.63* 0.68** 0.64* 0.89** 0.89** 0.89** 0.81** 0.69** 0.74**

K 0.26 0.25 0.26 0.82** 0.87** 0.90** 0.71** 0.40 0.38 0.23 0.82** 0.79**

Mn 0.65** 0.51 0.59* 0.67** 0.65** 0.61* 0.69** 0.54* 0.54* 0.65** 0.55* 0.43

Ni -0.34 -0.50 -0.38 0.15 0.13 0.20 -0.16 -0.41 -0.44 -0.40 0.01 -0.11

Pb 0.39 0.48 0.47 0.33 0.56* 0.54* 0.36 0.54* 0.51 0.25 0.48 0.52*

S -0.15 -0.19 -0.21 0.72** 0.70** 0.78** 0.39 -0.13 -0.10 -0.23 0.70** 0.39

Si 0.34 0.21 0.31 0.56* 0.41 0.40 0.56* 0.36 0.37 0.41 0.50 0.35

Ti -0.08 -0.18 -0.08 0.34 0.09 0.20 0.09 -0.13 -0.20 -0.12 0.05 -0.09

V -0.21 -0.33 -0.25 -0.05 -0.10 -0.05 -0.29 -0.31 -0.37 -0.33 -0.15 -0.28

Zn 0.63* 0.63* 0.75** 0.63* 0.68** 0.67** 0.76** 0.79** 0.83** 0.66** 0.69** 0.66** **. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Highlights:

• We examined effects of filter type and extraction solvent on oxidative potential (OP). • Extraction solvent had a significant effect on OPDTT, but not on OPESR or OPAA. • OP values measured from quartz filter extracts were heavily attenuated for all assays. • However, OP values from quartz filters were highly correlated with those from Teflon. • OP measured with ESR direct method showed promising results.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Measurement of the oxidative potential of PM2.5

and its constituents: The effect of extraction

solvent and filter type

Aileen Yang a,b*, Aleksandra Jedynska c, Bryan Hellack d, Ingeborg Kooter c, Gerard Hoek b,

Bert Brunekreef b,e, Thomas AJ Kuhlbusch d,f, Flemming R Cassee a,b, Nicole AH Janssen b

a National Institute for Public Health and the Environment (RIVM), Bilthoven, the

Netherlands;

b Institute for Risk Assessment Sciences, Division Environmental Epidemiology, Utrecht

University, Utrecht, the Netherlands;

c The Netherlands Organisation for Applied Scientific Research (TNO), Utrecht, the

Netherlands;

d Air Quality & Sustainable Nanotechnology, Institute of Energy and Environmental

Technology (IUTA) e.V., Duisburg, Germany;

e Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht,

Utrecht, the Netherlands

f CENIDE, University Duisburg-Essen, Duisburg, Germany

Supporting Information

Number of Pages 8

Number of Tables 7

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Table S 1 Overview of the filter allocation for the OP assays.

Teflon filters Quartz filters

Methanol extraction

AA+ DTT 2 (duplicate) 2 (duplicate)

CRAT + ESR 2 (duplicate) 2 (duplicate)

Water extraction

AA+DTT 2 (duplicate)

ESR 2 (duplicate)

Non-extracted

ESR- direct 2 (duplicate)

XRF 2 (duplicate)

EC/OC 2 (duplicate)

Note: + indicates analysis from same filter

Table S 2 Coefficient of variation (CV) for mass concentration of the filters used in the different OP

methods. AA and DTT used the same set of filter duplicates, while ESR and CRAT were analysed using

another set of filter duplicates. (Met = Methanol, Q = Quartz, T = Teflon).

AA/DTT

(µg /m3)

ESR/CRAT

(µg/m3)

Met

Q

Met

T

Water

T

Direct

Q

Met

T

Met

Q

Duplicates 14* 15 15 15 15 14*

CV (%) 19 4 3 8 6 9

*Filter omitted from calculations due to technical problems during sampling.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Protocol for OP assays

DTT

PM2.5 extracts from Teflon filters were diluted with traceSELECT ultrapure water (Sigma,

Zwijndrecht, Netherlands) to a standard of 500µg/mL. The quartz filter sample extracts are

each diluted with 1 mL ultrapure water. The incubation volume is 500µL. All PM samples are

analysed in duplicates. Domestic oil burning furnace (DOFA obtained from US EPA, RTP,

NC) was used as a positive control and ultrapure water as a negative control.

For every sample, 100µL of 0.5M potassium phosphate buffer (pH 7.4), 10µL PM-stock (for

Teflon and quartz) are added to a 1.5mL Eppendorf vial. The solution is then incubated for 10

minutes at 37°C before adding 50µL 1mM DTT (DL-Dithiothreithol, Sigma, Zwijndrecht) to

start the reaction. At designated time points (0, 10, 20 and 30 minutes), 50µL of the reaction

mixture is withdrawn from the tubes and added into a precooled vial containing 50µL icecold

10% TCA (Trichloroacetic acid ACS reagent, Sigma, Zwijndrecht, Netherlands). 200µL of

2mL Tri-HCl containing 50µl DTNB-stock (5,5′-Dithiobis(2-nitrobenzoic acid), Sigma,

Zwijndrecht) is then added into the cold Eppendorf containing the stopped reaction mixture.

The solution is mixed using a vortex before 250µL is transferred to the dedicated well on the

96 wells plate. The absorbance at 412nm is recorded on the spectrophotometer

(SpectraMax190). The rate of DTT consumption, here considered as the OP, is calculated

using linear regression of the data in the linear range as seen from a plot of absorbance against

time. The results are expressed as nmol DTT/min per sampled volume (m3) or per PM mass

(µg).

AA

PM derived from Teflon filters are diluted to a concentration of 12.5 µg/mL in ultrapure

water. After dilution, aliquots of 160 µL of PM sample are added to a 96-well flat-bottomed

UV plates (UV-star: VWR, Breda, Netherlands) in quadruplicate. The quartz filter sample

extracts are each diluted with 1 mL ultrapure water, then 20µL of the samples are added to the

UV-plate. The final incubation volume for both filter types is 200 µL.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

160 µL of ultrapure water is used as the negative control, and 160 µL of Domestic oil burning

furnace (DOFA) suspension, in the same concentration as the samples, as the positive control.

The samples are incubated in the spectrophotometer (spectraMAX 190: Molecular Devices,

Sunnyvale, USA) for 10 min at 37 °C, the optimal temperature for the depletion of ascorbic

acid. After incubation, 20 µl of 2mM ascorbic acid is added, and the absorption at 265 nm

measured every 2 min for 2 hours. The 96-wells plate is auto shaken for 3 seconds before

each measurement.

ESR

DMPO (5.5-dimethyl-pyrrolineN-oxide) and Tempol (4-hydroxy-2,2,6,6-

tetramethylpiperidinyloxy) were purchased from Alexis (Switzerland). All other reagents, like

hydrogen peroxide (H2O2), Cu(II)SO4, Ni(II)SO4, Fe(II)SO4, Fe(III)SO4 and the BCR 723

Road Dust (all solid) were purchased from Sigma-Aldrich (Germany). DMPO and H2O2 was

prepared and used in a chloride and calcium free phosphate Buffer (PBS, Sigma-Aldrich).

Filter extracts (Teflon and quartz) or quartz filters (ESR direct) are directly placed in a vessel,

completely covered by the chemical ingredients (DMPO, H2O2 and dH2O) and agitated by a

vortexer for 30 seconds before being placed in a heating shaking water bath for incubation at

37 °C for 15 minutes. After incubation, the suspension is vortexed again for 15 seconds, then

transferred into a 50 µl glass capillary without any filtration. The DMPO-OH quartet signal is

measured with a MS-300 ESR spectrometer (Magnettech, Berlin, Germany). The following

ESR settings at room temperature were used for all measurements: magnetic field: 3.365 G,

sweep width: 100 G, scan time: 30 s, number of scans: 3, modulation amplitude: 2 G, receiver

gain: 1.000.

Quantification was carried out on first derivation of EPR signal of DMPO–OH quartet as

average of total amplitudes of the four peaks in arbitrary units (A.U.). As blank, a mixture of

deionised water (dH2O), H2O2 and DMPO (plus a blank filter) were used and a limit of

detection (LOD) was calculated for each experiment as mean of n blank value plus threefold

standard deviation (SD) of the blank.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Table S 3 PM components above the LOD and their corresponding CV values .

Component LOD

(ng)

< LOD

(n)

CV

(%)

Al 0.00 1 8

Br 0.19 0 7

Cr 1.02 0 12

Cu 1.22 0 5

Fe 3.57 0 6

K 5.09 0 6

Mn 1.17 0 4

Ni 0.60 1 18

Pb 0.00 0 5

S 0.00 0 6

Si 61.00 0 19

Ti 2.24 0 21

V 0.35 0 19

Zn 8.34 0 5

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Table S 4 Summary of the XRF values for PM elements. n= 15 for all except EC/OC where n = 13.

per m3 per µg of PM

Mean

Std.

Deviation Min Max Mean

Std.

Deviation Min Max

Al 19.89 11.91 0.00 43.54 1.20 0.65 0.00 2.65

Br 3.87 1.91 1.59 8.91 0.25 0.04 0.18 0.33

Cr 2.13 1.00 0.52 4.15 0.15 0.12 0.06 0.54

Cu 11.55 6.91 1.40 22.07 0.73 0.47 0.18 1.74

Fe 266.77 149.72 37.79 499.87 17.15 10.11 5.43 39.84

K 110.15 78.22 30.32 346.95 6.51 1.68 3.96 9.05

Mn 6.48 2.40 2.56 11.12 0.45 0.17 0.22 0.76

Ni 1.26 0.94 0.26 3.28 0.10 0.10 0.02 0.36

Pb 7.84 4.46 2.13 18.87 0.53 0.37 0.26 1.77

S 797.69 474.56 311.84 2333.43 52.36 18.22 30.27 86.42

Si 49.98 24.59 8.51 95.70 3.81 3.45 1.28 14.87

Ti 2.99 1.99 1.16 8.09 0.24 0.25 0.06 1.05

V 1.85 1.27 0.32 4.68 0.15 0.15 0.02 0.55

Zn 29.10 12.56 5.86 50.50 1.92 0.79 0.76 3.66

EC 1.54 0.57 0.61 2.31 0.11 0.07 0.03 0.25

OC 3.52 1.45 1.66 7.06 0.23 0.06 0.16 0.33

Mass

Teflon 16.46 8.90 5.21 38.36

Mass

quartz 16.06 9.07 5.13 37.45

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Table S 5 Spearman’s correlation coefficient between PM characteristics and OP methods (n=15),

expressed per µg of PM. (Met = Methanol, Q = Quartz, T= Teflon).

Ascorbate

(nmol AA/s/ µg)

DTT

(nmol DTT/min/µg)

ESR

(A.U./µg)

CRAT – ROS

(pmol NQ/µg)

Met

Q

Met

T

Water

T

Met

Q

Met

T

Water

T

Met

Q

Met

T

Water

T

Direct

Q

Met

Q

Met

T

EC 0.83** 0.81** 0.85** 0.06 0.07 0.34 0.70** 0.88** 0.77** 0.96** -0.35 -0.06

OC 0.42 0.42 0.47 -0.06 -0.23 0.11 0.32 0.57* 0.40 0.73** -0.57* -0.45

Al 0.47 0.33 0.37 0.26 0.35 0.59* 0.45 0.30 0.25 0.49 0.05 0.05

Br 0.05 -0.03 0.03 -0.08 -0.52* -0.33 -0.02 0.13 0.05 0.27 -0.52* -0.59*

Cr 0.57* 0.52* 0.57* 0.55* 0.25 0.38 0.59* 0.38 0.40 0.65** -0.18 0.07

Cu 0.81** 0.91** 0.87** 0.10 0.60* 0.68** 0.88** 0.86** 0.93** 0.71** 0.51* 0.70**

Fe 0.83** 0.93** 0.90** 0.15 0.62* 0.69** 0.90** 0.86** 0.93** 0.76** 0.46 0.67**

K -0.11 0.06 -0.08 0.65** 0.18 0.29 0.10 0.11 0.03 0.04 0.22 0.06

Mn 0.66** 0.64* 0.73** 0.27 0.30 0.46 0.60* 0.57* 0.52* 0.85** -0.31 0.03

Ni -0.25 -0.45 -0.31 -0.28 -0.68** -0.46 -0.38 -0.42 -0.41 -0.17 -0.70** -0.56*

Pb 0.20 0.12 0.16 -0.11 -0.25 -0.23 0.08 0.05 0.15 0.12 -0.50 -0.27

S -0.41 -0.44 -0.44 -0.06 -0.25 -0.41 -0.47 -0.33 -0.49 -0.17 -0.47 -0.72**

Si 0.45 0.46 0.35 0.44 0.22 0.40 0.50 0.53* 0.34 0.61* -0.11 -0.09

Ti 0.01 -0.20 -0.06 0.13 -0.46 -0.35 -0.06 -0.20 -0.19 0.19 -0.70** -0.66**

V 0.09 -0.18 -0.04 -0.29 -0.60* -0.37 -0.17 -0.16 -0.18 0.12 -0.69** -0.67**

Zn 0.81** 0.80** 0.84** -0.05 0.35 0.50 0.73** 0.81** 0.79** 0.87** -0.14 0.18

**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Table S 6 Spearman correlation coefficients between the chemical composition and other PM characteristics, the lower area is expressed per m3 and the upper

shaded area is expressed per µµµµg of PM (n=15).

Mass Teflon

Mass Quartz EC OC Al Br Cr Cu Fe K Mn Ni Pb S Si Ti V Zn

Mass Quartz 0.99**

EC 0.29 0.29 0.80** 0.29 0.63* 0.62* 0.64* 0.68* -0.01 0.74** 0.01 0.26 -0.09 0.60* 0.53 0.44 0.90**

OC 0.95** 0.95** 0.46 0.38 0.80** 0.40 0.23 0.31 0.18 0.68* 0.29 0.25 0.15 0.59* 0.74** 0.47 0.75**

Al 0.71** 0.71** 0.36 0.69** -0.18 0.37 0.26 0.30 0.04 0.53* 0.06 -0.25 -0.17 0.45 0.23 0.07 0.45

Br 0.93** 0.90** 0.45 0.92** 0.68** 0.18 -0.13 -0.10 0.16 0.20 0.39 0.39 0.35 0.36 0.59* 0.62* 0.19

Cr 0.36 0.33 0.70** 0.68* 0.40 0.35 0.33 0.43 0.26 0.76** 0.11 0.20 -0.08 0.63* 0.40 0.18 0.33

Cu 0.56* 0.57* 0.95** 0.66* 0.59* 0.63* 0.44 0.99** 0.14 0.42 -0.62* 0.00 -0.65** 0.29 -0.39 -0.38 0.69**

Fe 0.56* 0.58* 0.93** 0.69** 0.59* 0.64* 0.50 0.99** 0.17 0.53* -0.57* -0.01 -0.60* 0.33 -0.31 -0.37 0.71**

K 0.90** 0.89** 0.25 0.90** 0.60* 0.85** 0.35 0.56* 0.59* 0.05 -0.41 -0.05 -0.08 0.61* -0.16 -0.42 -0.08

Mn 0.55* 0.56* 0.52 0.55* 0.87** 0.51 0.55* 0.68** 0.70** 0.53* 0.10 0.08 0.02 0.49 0.42 0.11 0.73**

Ni 0.28 0.23 -0.14 0.31 0.31 0.31 0.16 -0.23 -0.25 0.06 0.07 0.28 0.43 -0.15 0.72** 0.78** -0.20

Pb 0.61* 0.60* 0.39 0.53 0.33 0.68** 0.01 0.48 0.45 0.62* 0.29 -0.03 -0.07 -0.02 0.31 0.46 0.24

S 0.80** 0.79** -0.05 0.73** 0.55* 0.63* 0.29 0.10 0.10 0.60* 0.31 0.48 0.29 0.05 0.48 0.34 -0.26

Si 0.33 0.33 0.33 0.42 0.46 0.37 0.25 0.57* 0.56* 0.51* 0.45 -0.05 0.06 0.00 0.18 0.08 0.33

Ti 0.24 0.16 0.44 0.53 0.38 0.29 0.67** 0.03 0.06 0.13 0.32 0.54* -0.08 0.41 -0.09 0.69** 0.14

V 0.07 -0.01 0.06 0.07 0.15 0.19 0.07 -0.20 -0.25 -0.16 -0.08 0.88** 0.01 0.23 -0.07 0.50 0.03

Zn 0.60* 0.61* 0.69** 0.74** 0.63* 0.68** 0.30 0.88** 0.86** 0.68** 0.64** -0.20 0.65** 0.16 0.53* -0.05 -0.26

**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Table S 7 Effect of extraction fluid and filter type on OP values for four assays. T= Teflon, Q= quartz, Met = Methanol, n=15.

Ascorbate (nmol AA/s/µµµµg of PM)

DTT (nmol DTT/min/ µµµµg of PM)

ESR (A.U./ µ µ µ µg of PM)

CRAT (pmol NQ/µµµµg of PM)

Met Q

Met T

Water T

Met Q

Met T

Water T

Met Q

Met T

Water T

Direct Q

Met Q

Met T

Mean 0.41 1.14 1.33 0.15 0.17 0.13 133 265 133 267 32 95

Std. dev 0.32 0.95 1.28 0.07 0.04 0.02 55 163 88 201 21 59

Min 0.07 0.22 0.12 0.07 0.11 0.10 66 85 29 76 5 12

Max 1.16 3.24 4.18 0.35 0.23 0.18 249 595 351 760 67 180