A New Method for the Assessment of the Oxidative Potential of Both Water-Soluble and Insoluble PM

: Water-soluble and insoluble fractions of airborne particulate matter (PM) exhibit different toxicological potentials and peculiar mechanisms of action in biological systems. However, most of the research on the oxidative potential (OP) of PM is focused exclusively on its water-soluble fraction, since experimental criticisms were encountered for detaching the whole PM (soluble and insoluble species) from ﬁeld ﬁlters. However, to estimate the actual potential effects of PM on human health, it is essential to assess the OP of both its water-soluble and insoluble fractions. In this study, to estimate the total OP (TOP), an efﬁcient method for the detachment of intact PM 10 from ﬁeld ﬁlters by using an electrical toothbrush was applied to 20 PM 10 ﬁlters in order to obtain PM 10 water suspensions to be used for the DCFH, AA and DTT oxidative potential assays (OP DCFH , OP AA and OP DTT ). The contribution of the insoluble PM 10 to the TOP was evaluated by comparing the TOP values to those obtained by applying the three OP assays to the water-soluble fraction of 20 equivalent PM 10 ﬁlters. The OP of the insoluble fraction (IOP) was calculated as the difference between the TOP and the WSOP. Moreover, each PM 10 sample was analyzed for the water-soluble and insoluble fractions of 10 elements (Al, Cr, Cs, Cu, Fe, Li, Ni, Rb, Sb, Sn) identiﬁed as primary elemental tracers of the main emission sources in the study area. A principal component analysis (PCA) was performed on the data obtained to identify the predominant sources for the determination of TOP, WSOP, and IOP. Results showed that water-soluble PM 10 released by trafﬁc, steel plant, and biomass burning is mainly responsible for the generation of the TOP as well as of the WSOP. This evidence gave strength to the reliability of the results from OP assays performed only on the water-soluble fraction of PM. Lastly, the IOP DCFH and IOP DTT were found to be principally determined by insoluble PM 10 from mineral dust.


Introduction
Air pollution is widely recognized as a key topic in public health protection actions [1,2]. Exposure to particulate matter (PM) is one of the major global health concerns [3,4], since it may adversely affect human health, leading to the development of several chronic and acute pathologies, such as cardiovascular and respiratory diseases, lung cancer, bronchitis, diabetes, and neurodevelopmental disorders [5][6][7][8]. A growing number of epidemiological studies reported associations of PM pollution with health effects also at low levels, often below current air quality standards [9,10]. Therefore, the evaluation of exposure to PM and associated health risks is crucial for planning targeted mitigation strategies and policies to protect human health.
The main cytotoxicity mechanism involved in developing damaging health effects and promoting chronic diseases is the ability of PM to induce oxidative stress, due to the and/or insoluble fraction to the OP. However, with the aim of investigating the possible toxicological effects related to PM exposure under real-world conditions, it would be useful to apply OP assays directly to a PM water suspension for the assessment of the overall OP of PM (water-soluble and insoluble species).
To this aim, a non-invasive, simple, and efficient method, recently validated by Massimi et al. [46] for the recovery of the elements in PM 10, was used for the detachment and suspension in water of intact PM 10 from field filters by using an electrical toothbrush [47,48]. The retrieved PM 10 water suspension was used for the assessment of the overall OP DCFH , OP AA , and OP DTT of intact PM 10 samples (water-soluble and insoluble fraction) and for the evaluation of the real contribution of insoluble particles to PM redox properties.  10 [49,50].

Analytical Procedure
PM 10 filters collected from one of the two sample lines were processed by following a previously optimized and detailed chemical fractionation procedure, involving the water-extraction of PM 10 filters and the acid digestion of the residue, followed by elemental analysis of both the water-soluble and insoluble fractions [49,50,52,53]. After the removal of the supporting polymethyl pentene ring from each sampled filter, PM 10 filters were immersed in 10 mL of deionized water (produced by Arioso UP 900 Integrate Water Purification System, Seoul, Korea) and then extracted by rotating agitation (Rotator, Glas-Col, Hangzhou Yooning Instrunents, Hangzhou, Zhejiang, China) for 30 min at 60 rpm. The obtained solution was then filtered through a nitrocellulose filter (NC filter; pore size 0.45 µm, Merck Millipore Ltd., Billerica, MA, USA) to obtain the PM 10 water-soluble fraction. Subsequently, the PM 10 filters containing the residue and the nitrocellulose filter used for the filtration were acid-digested in a microwave oven (Ethos Touch Control with Q20 rotor, Milestone, Sorisole, Bergamo, Italy) using 2 mL of HNO 3 (67%, Promochem, Wesel, Germany) and 1 mL of H 2 O 2 (30% Suprapur, Merck Millipore Ltd., Billerica, MA, USA). The digested solutions were diluted to 50 mL with deionized water and filtered by nitrocellulose syringe filters (diameter 25 mm, pore size 0.45 µm, GVS Filter Technology, Morecambe, England, UK) to obtain the PM 10 insoluble fraction. Water-soluble and insoluble fractions of PM 10 samples were analyzed for the determination of 10 elements (Al, Cr, Cs, Cu, Fe, Li, Ni, Rb, Sb, Sn) in the two fractions using quadrupole inductively coupled plasma mass spectrometry (ICP-MS, Bruker 820-MS, Billerica, MA, USA). Further information about the instrumental conditions and performance of the method is reported in Astolfi et al. [54]. The analyzed elements were selected as they were already identified as primary elemental tracers of the predominant emission sources in the Terni basin [49,50]. The minimum detection limit (MDL), concentrations, the mean and standard deviation of PM 10 mass, and of the elements analyzed in the water-soluble and insoluble fraction of PM 10 collected at PR are reported in Table 1. PM 10 mass concentration data were obtained from ARPA Umbria reports available online (www.arpa.umbria.it, accessed deviation of PM10 mass, and of the elements analyzed in the water-soluble and insoluble fraction of PM10 collected at PR are reported in Table 1. PM10 mass concentration data were obtained from ARPA Umbria reports available online (www.arpa.umbria.it, accessed on 25 January 2022). Average solubility percentages (%) of the elements in the 20 PM10 samples are shown in Figure 1. Equivalent PM10 filters from the second sample line were instead subjected to PM10 toothbrush detachment and suspension in water, as thoroughly described in Massimi et al. [46]. This method has been already applied only by Süring et al. [47,48], who used this procedure to quantify allergen-loaded particles in PM10 by flow cytometry. Briefly, each filter was put in a polystyrene Petri dish of 50 mm diameter and overlaid with 10 mL of deionized water. Subsequently, the filter was held with PTFE tweezers and brushed for 2 min using an electrical toothbrush with a sensitive brush head (Braun, Germany, Oral-B Vitality Sensitive). The obtained total PM10 water suspension was then used in this study for the application of the three OP assays. The performance of this method for the recovery of the elements in PM10 was evaluated by Massimi et al. [46] after applying the chemical fractionation procedure described above to the obtained PM10 water suspensions and to equivalent PM10 samples not subjected to the toothbrush detachment, revealing an efficiency of approximatively 70% for the recovery of the elements. Moreover, this method is described in Süring et al. [47,48].

Oxidative Potential Measurements
The OP DCFH , OP AA , and OP DTT assays were used to assess the OP of the 20 pairs of PM10 equivalent samples.To obtain the water-soluble OP (WSOP) and total OP (TOP, water-soluble and insoluble PM10), the three OP assays were simultaneously applied to the water-soluble fraction of PM10 filters obtained by following the chemical fractionation procedure described in Section 2.2, and to the PM10 aqueous suspension retrieved by applying the toothbrush detachment to the equivalent PM10 filters. Furthermore, the OP of the insoluble fraction (IOP) was calculated as the difference between the TOP and the WSOP. The average contribution percentages (%) of the two solubility fractions to the TOP DCFH , TOP AA , and TOP DTT are shown in Figure 1. The OP analytical measurements performed in this study followed the validated and frequently used procedures [26,42,50] largely described below. Equivalent PM 10 filters from the second sample line were instead subjected to PM 10 toothbrush detachment and suspension in water, as thoroughly described in Massimi et al. [46]. This method has been already applied only by Süring et al. [47,48], who used this procedure to quantify allergen-loaded particles in PM 10 by flow cytometry. Briefly, each filter was put in a polystyrene Petri dish of 50 mm diameter and overlaid with 10 mL of deionized water. Subsequently, the filter was held with PTFE tweezers and brushed for 2 min using an electrical toothbrush with a sensitive brush head (Braun, Germany, Oral-B Vitality Sensitive). The obtained total PM 10 water suspension was then used in this study for the application of the three OP assays. The performance of this method for the recovery of the elements in PM 10 was evaluated by Massimi et al. [46] after applying the chemical fractionation procedure described above to the obtained PM 10 water suspensions and to equivalent PM 10 samples not subjected to the toothbrush detachment, revealing an efficiency of approximatively 70% for the recovery of the elements. Moreover, this method is described in Süring et al. [47,48].

Oxidative Potential Measurements
The OP DCFH , OP AA , and OP DTT assays were used to assess the OP of the 20 pairs of PM 10 equivalent samples.To obtain the water-soluble OP (WSOP) and total OP (TOP, water-soluble and insoluble PM 10 ), the three OP assays were simultaneously applied to the water-soluble fraction of PM 10 filters obtained by following the chemical fractionation procedure described in Section 2.2, and to the PM 10 aqueous suspension retrieved by applying the toothbrush detachment to the equivalent PM 10 filters. Furthermore, the OP of the insoluble fraction (IOP) was calculated as the difference between the TOP and the WSOP. The average contribution percentages (%) of the two solubility fractions to the TOP DCFH , TOP AA , and TOP DTT are shown in Figure 1. The OP analytical measurements performed in this study followed the validated and frequently used procedures [26,42,50] largely described below.

OP AA
For both the PM 10 equivalent samples, the OP AA followed the method reported by Fang et al. [21]. Phosphate buffer measuring 300 µL (0.5 mM) and 100 µL of AA reagent (2 mM; Sigma-Aldrich, St. Louis, MO, USA) was added to 2.5 mL of sample solution. Then, the absorbance of the reaction mixture was recorded at 265 nm wavelength at different reaction times (0, 10, and 20 min) using UV-Vis absorption spectrometry (Varian Cary 50 Bio UV-Vis; Varian Inc., Palo Alto, CA, USA). Blanks were always measured in parallel. OP AA was calculated as the AA depletion rate per sampled volume (nmol AA min −1 ·m −3 ) according to Equation (1).
where ơAbs is the slope of the absorbance of operative blanks vs. time (min −1 ), Abs 0 is the initial absorbance calculated from the intercept of the linear regression of absorbance vs. time, N 0 is the number of AA moles added into the reaction mixture (200 nmol), ơAA s and ơAA b are the rates of AA consumption for the sample and for the blank, respectively (nmol·min −1 ), V e and V a are the extraction volume and sample volume added to the reaction mixture, respectively, and V s is the PM sampled volume (m 3 ).

OP DTT
To perform OP DTT , the solution of PM 10 equivalent samples was split into three aliquots of 0.7 mL each (2.1 mL in total), then incubated at 37 • C with 0.1 mL of DTT (1 mM; Sigma-Aldrich, USA) and 0.2 mL of potassium phosphate buffer (1 M). Then, 1 mL of trichloroacetic acid (10% TCA; Sigma-Aldrich, USA) was added to the mixture at different reaction times (0, 10, and 20 min) to stop the DTT reaction. An aliquot of the reaction mixture (1 mL) was taken and mixed with 2 mL of tris-buffer (0.08 M, containing EDTA 4 mM) and with 50 µL of 5,5-dithiobis-2-nitrobenzoic acid (DTNB; Sigma-Aldrich, USA) to form 2-nitro-5-mercaptobenzoic acid (TNB) for a colorimetric reaction with the residual DTT. The obtained solution was then measured at 412 nm using the UV-Vis spectrometer. Furthermore, blanks were measured in parallel to samples. OP DTT was expressed as DTT consumption rate per sampled PM volume (nmol DTT min −1 ·m −3 ), according to Equation (2).
where ơAbs is the slope of the absorbance of operative blanks vs. time (min −1 ), Abs 0 is the initial absorbance calculated from the intercept of the linear regression of absorbance vs. time, N 0 is the number of DTT moles added in the reaction mixture (100 nmol), ơDTT s and ơDTT b are the rates of DTT consumption for the sample and for the blank, respectively (nmol·min −1 ), V e and V a are the extraction volume and sample volume added to the reaction mixture, respectively, and V s is the PM sampled volume (m 3 ).

OP DCFH
DCFH solution was prepared by dissolving 4.873 mg of the 2 ,7 -dichlorofluorescein diacetate (DCFH-DA; Sigma-Aldrich, USA) in 5 mL of ethanol (EtOH, 96%) in the dark. Then, 20 mL NaOH 0.01 M were added to favor the de-acetalization reaction. The obtained solution was kept in the dark at room temperature for at least 30 min before use. DCFH reagent measuring 125 µL (5 µM) and 5 mL of HRP (0.5 units mL −1 ) dissolved in a sodium phosphate buffer (pH 7.4; 25 mM) were added to 1.5 mL of the solution of PM 10 equivalent samples. The reaction mixture was placed in the thermostatically controlled water bath at 37 • C for 5 min. The DCFH became fluorescent dichlorofluorescein (DCF) upon reaction with ROS. Hence, the concentration of DCF was measured using fluorescent spectroscopy (Jasco FP-920; excitation at 427 nm, emission at 530 nm). Standard H 2 O 2 solutions (5 × 10 −8 , 1 × 10 −7 , 2 × 10 −7 , 5 × 10 −7 , and 1 × 10 −6 M) were used to obtain a calibration curve to convert the fluorescence intensity into H 2 O 2 equivalents, which were used as indicators of the reactive species reactivity, thus obtaining OP DCFH values (nmol H 2 O 2 m −3 ).

Data Analysis
The paired sample t-test was used to observe the significance of the differences between WSOP and TOP results obtained by each OP method applied to the 20 couples of equivalent PM 10 samples in order to evaluate the significance of the contribution of insoluble particles to the generation of the TOP. A p-value less than 0.05 was considered statistically significant.
A principal component analysis (PCA) was carried out on the matrix of the data (580 data points) composed of 20 PM 10 samples and 29 variables: OP DCFH , OP AA and OP DTT for the water-soluble OP (WSOP AA , WSOP DTT and WSOP DCFH ), insoluble OP (IOP AA , IOP DTT and IOP DCFH ) and total OP (TOP AA , TOP DTT and TOP DCFH ), and 10 elements in the water-soluble (Al_ws, Cr_ws, Cs_ws, Cu_ws, Fe_ws, Li_ws, Ni_ws, Rb_ws, Sb_ws, Sn_ws) and insoluble fraction (Al_i, Cr_i, Cs_i, Cu_i, Fe_i, Li_i, Ni_i, Rb_i, Sb_i, Sn_i) of the PM 10 samples. The matrix of the data was transformed by column mean centering and row and column autoscaling to correct variations in the different scaling of the variables before performing the PCA [55,56]. The principal component analysis was performed using the statistical software CAT (Chemometric Agile Tool) based on the R-project for statistical computing, Ver. 3.0, 32-bit.

WSOP vs. TOP
The contribution of the PM 10 insoluble fraction to the TOP was assessed comparing the results obtained by the OP DCFH , OP AA , and OP DTT assays performed on the water-soluble fraction of PM 10 (WSOP) to the results achieved by applying the three assays to the aqueous suspension of intact PM 10 (water-soluble and insoluble PM 10 ; TOP). WSOP, TOP, and IOP (insoluble OP obtained from the difference between TOP and WSOP) obtained results are shown in Figure 2 and reported in Table 2 along with the percentage of the contribution of IOP to TOP. TOP values were found to be significantly higher with respect to those of the WSOP for all the three OP assays. In detail, for the OP DCFH , OP AA , and OP DTT assays, the p-value between the TOP and the WSOP is 0.00016, 0.00096, and 0.0022, respectively, confirming a significant contribution of the PM insoluble fraction to the redox properties of the 20 PM 10 samples.
OP AA (Figure 2b) showed higher values for both the WSOP and the TOP on 7, 10, 11, 21, and 23 May, when the highest concentrations of water-soluble Cu were recorded (Table 1), to which the OP AA is well-known to be selectively responsive [21,57,58]. TOP values were found to be significantly higher with respect to those of the WSOP for all the three OP assays. In detail, for the OP DCFH , OP AA , and OP DTT assays, the p-value between the TOP and the WSOP is 0.00016, 0.00096, and 0.0022, respectively, confirming a significant contribution of the PM insoluble fraction to the redox properties of the 20 PM10 samples.
OP AA (Figure 2b) showed higher values for both the WSOP and the TOP on 7, 10, 11, 21, and 23 May, when the highest concentrations of water-soluble Cu were recorded (Table 1), to which the OP AA is well-known to be selectively responsive [21,57,58].  On 23 May, TOP AA was significantly higher than WSOP AA (3.9 and 0.59, respectively). This was probably due to the higher contribution of crustal dust to PM 10 . In fact, on that day, much higher concentrations of all the elements released in the insoluble fraction of PM 10 by resuspension of coarse mineral dust [59][60][61], such as Al, Cs, Fe, Li e Rb (Table 1), were recorded. It is well-known that OP AA predominantly responds to coarse particles [26,62], and it is assumed that a relevant contribution to the TOP was due to a higher contribution of soil dust on this day. OP DCFH and OP DTT showed higher TOP with respect to WSOP on 21 and 17 May, respectively, when the highest concentrations of crustal dust elements were recorded, which therefore confirms a significant contribution of soil dust to the TOP. Both the assays exhibited the highest WSOP and TOP on 11 and 21 May, when higher concentrations of Fe, Cr, and Ni from the steel plant [49,63], and Cu, Sb, and Sn from traffic [64][65][66] were released in water-soluble PM 10 . Although the insoluble fraction of these elements (especially Fe, to which the DTT was highly sensitive) [67] presented significantly higher concentrations, the OP assays appeared to be predominantly responsive to their water-soluble fraction, which therefore plays a crucial role in the generation of the TOP.
The discussed findings partially agree with the findings of Frezzini et al. [42], which revealed that the insoluble fraction of PM provided a significant contribution to the TOP AA and left open questions regarding the TOP DCFH and TOP DTT , forwhich the insoluble fraction did not seem to contribute significantly to the TOP, probably due to issues related to the experimental procedure. In detail, the OP AA values were significantly influenced by the PM insoluble fraction, probably due to the higher sensitivity of this assay to the coarser insoluble particles [26,50]. These particles may have moved from the PTFE filters immersed in the aqueous solution during the extraction procedure, and reacted with the solutions used for the OP AA assay. Conversely, the insoluble fraction did not determine a significantly higher OP DCFH and OP DTT . This might be explained by the higher sensitivity of these two assays through the fine PM particles [67,68], which are almost predominantly soluble in water [61] and, accordingly, contributed more to WSOP rather than to TOP. Another possible explanation is that particles were probably fixed too deeply in the immersed PTFE filter to fully react with the reagents of the OP DCFH and OP DTT assays to generate significantly higher TOP with respect to WSOP.
The application of the TOP assays directly to the aqueous suspension of intact particles detached from PM 10 filters overcame the operative limits of performing OP assays directly on solutions containing the immersed PM filter. In fact, the described method allowed the authors to detach and carry in suspension a high fraction of the particles sampled on the filter, approximatively 70% in reference to the recovery of the elements [46].
These results revealed the non-negligible contribution of the insoluble particles to the TOP DCFH and TOP DTT , in agreement with previous findings that demonstrated the relevance of the PM insoluble fraction to the total OP of PM [27,28,45].

Principal Component Analysis
To identify tracers of emission sources that can determine the above-discussed OP results, explorative PCA was performed on the WSOP, TOP, and IOP values for OP AA , OP DTT , and OP DCFH , and the element concentrations in the water-soluble and insoluble fraction of the 20 PM 10 samples. Five significant components accounting for 78.4% were obtained from the PCA. The variance explained by each component is of 29.7%, 17.6%, 13.3%, 9.1% and 8.5%, respectively.
The biplot of PC1/PC2 is reported in Supplementary Materials ( Figure S1), along with the scores and loadings obtained from the PCA (Table S1). The first component (PC1), which represents 29.7% of the total variance, separates out the samples (scores) in which the highest element concentrations (loadings) were recorded, and largely explains the data variability due to daily variations in atmospheric conditions. However, PC1 only distinguishes samples in which the highest concentrations of all the considered variables were found, as distinct from those characterized by the lowest concentrations. On the contrary, PC2, and PC3, which represent, respectively, 17.6% and 13.3% of the total variance, separate the variables depending on their concentration variability among the sampling days. PC2/PC3 are graphically summarized in the biplot of Figure 3.
PC2 and PC3 account for 31.1% of the total variance. In the biplot, the variables are grouped in 4 main clusters, each one containing tracers of a specific emission source. The cluster on the lower-left part of the biplot includes the insoluble elements typically released by the steel plant and by non-exhaust traffic, such as brake abrasion (Cu_i, Cr_i, Fe_i, Ni_i, Sb_i, and Sn_i) [64][65][66]69,70]. The cluster on the upper-left part of the biplot contains the same elemental tracers of the group described above in their water-soluble fraction (Cu_ws, Cr_ws, Fe_ws, Ni_ws, Sb_ws, and Sn_ws). The group on the upper-right part of the biplot contains the water-soluble elements mainly released by biomass burning emission sources (Cs_ws, Li_ws, and Rb_ws) [66,[71][72][73]. Lastly, the cluster on the lower-right part of the biplot includes insoluble elements considered as tracers of mineral dust (Al_i, Cs_i, Li_i, and Rb_i) [50,74,75].
Results show that all the three OP assays for the WSOP and TOP are included in the group of traffic and steel plant water-soluble tracers and in the same direction of biomass burning elemental tracers along PC2. On the other hand, WSOP and TOP assays do not seem to respond to the insoluble fraction of the elements released by traffic and the steel plant. This constitutes a first indication that the total OP is dominated by the water-soluble species present in the solution (Figure 1). Although it is known that WSOP AA and WSOP DTT respond predominantly to transition metals, such as Cu and Fe [62,67,76], the TOP, when using both the assays, was found to be sensitive mainly to the watersoluble fraction of these elements, even if Cu and Fe, as well as the other non-exhaust traffic and steel plant elemental tracers, were almost exclusively in their insoluble fraction ( Figure 1). Moreover, the predominant dependence of OP DCFH and OP DTT on traffic, steel plant, and biomass burning water-soluble particles and the higher sensitivity of OP AA to non-exhaust traffic is confirmed [50,[77][78][79]. The response of the OP DCFH and OP DTT assays to biomass burning tracers confirms previous findings [21,50,80,81] and highlights the central role of the water-soluble species in determining PM redox properties, even when TOP is considered.

Conclusions
It is well known that the different PM components are responsible for several effects that can trigger harmful oxidative reactions and inflammation on living organisms, This represents a key finding, since it gives strength to the reliability of the OP results, even if performed only on the water-soluble fraction of PM 10 . Overall, the application of the OP assays to the intact particles detached from PM 10 filters (water-soluble and insoluble fraction) reveals that water-soluble PM 10 released by traffic, steel plant, and biomass burning was primarily responsible for the generation of the TOP.
Finally, it is worth noting that IOP DTT and IOP DCFH are clustered with the insoluble Al, Cs, Li, and Rb, tracers of mineral dust, thus showing that considering only the insoluble fraction of PM 10 , OP DTT , and OP DCFH are mainly responsive to soil particles [26,50,82,83]. The results obtained in this study are compared and tabulated with the findings of existing works in the literature in Supplementary Materials (Table S2).

Conclusions
It is well known that the different PM components are responsible for several effects that can trigger harmful oxidative reactions and inflammation on living organisms, contributing to genotoxicity and cytotoxic mechanisms responsible for cell damage. However, most of the studies conducted so far have exclusively evaluated the toxicity of the water-soluble fraction of PM, which is considered more bioavailable and is more easily extractable. Since it has been shown that insoluble species can contribute significantly to the toxicological potential of PM, the evaluation of the OP of both PM fractions (water-soluble and insoluble) is essential for the assessment of the total health risk induced by PM.
The application of an innovative method for the detachment and suspension in water of intact PM 10 (water-soluble and insoluble species) from the sampled filters allowed us to evaluate the total OP of PM 10 collected at a highly polluted industrial site, and to identify specifically the contribution to the OP of every single fraction (water-soluble and insoluble). By performing explorative PCA, it was possible to evaluate the response of the OP DCFH , OP AA , and OP DTT assays to the various elemental fractions, and to identify the different emission sources responsible for the WSOP, TOP, and IOP.
The results confirmed the different sensitivity of the three OP assays to the various PM 10 components. In particular, the dependence of OP DCFH and OP DTT on traffic, steel plant, and biomass burning water-soluble particles, and the sensitivity of OP AA to nonexhaust traffic, was reinforced.
Moreover, the application of the OP assays to the intact particles detached from PM 10 filters (water-soluble and insoluble fraction) reveals that water-soluble PM 10 released by traffic, steel plant, and biomass burning is primarily responsible for the generation of the TOP. This represents a key finding since it gives strength to the reliability of the OP results, even if performed only on the water-soluble fraction of PM 10 . In addition, the insoluble fraction of PM 10 was found to be able to contribute to the IOP DCFH and IOP DTT , which seem to respond to mineral particles.
Further studies will have to be carried out to verify these results and to confirm the representativeness of the WSOP and TOP assays by comparison with biological endpoints.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/atmos13020349/s1, Figure S1. Biplot of PC1/PC2 from the PCA performed on the WSOP, TOP and IOP values for OPAA, OPDTT and OPDCFH, and element concentrations in the water-soluble (_ws) and insoluble (_i) fraction of the 20 PM10 samples; Table S1. Scores and loadings of the five significant components obtained by performing the PCA on the matrix of the data (580 data points) composed of 20 PM10 samples and 29 variables; Table S2. Results of this study compared with other findings available in the literature.