Investigating the Oxidative Potential and In Vitro Toxicity of Ambient Water-Soluble PM10 in an Eastern Mediterranean Site

Abstract

In this study, the acellular dithiothreitol (DTT) assay, the in vitro cellular DCFH-DA assay on human lung epithelial cells, and gene expression measurements were used to assess the toxicity of water-soluble (WS) PM₁₀ relating to reactive oxygen species (ROS) in summer at an Eastern Mediterranean urban site. Large influences from anthropogenic sources on health risks were observed with acellular and cellular assays. Anthropogenic biomass burning (BB) and natural dust events increased human pulmonary exposure to the oxidative potential (OP_dose,T) of WS-PM₁₀ by 209 and 47%, respectively, compared to regular periods. OP_v ^DTT and ROS_v results were positively correlated in anthropogenic-dominant samples, while showed no significant correlation in the remaining samples. As a result, the BB and dust event had higher and lower levels of cellular ROS_v compared with the nonevent period, respectively. Source apportionment results suggest that specific organic contents (e.g., PAHs) had relatively low contents in samples less influenced by anthropogenic sources, possibly explaining the divergence in acellular and cellular results. Heavy metals were dominant contributors of OP_v ^DTT throughout the campaign, and a Chelex method is recommended over a EDTA method for quantification of their summed OP_v ^DTT.

Keywords: EDTA; biomass burning; cytotoxicity; dust event; oxidative potential; particulate matter; reactive oxygen species.

1

Allocation of WS-PM₁₀ induced (a) OP_v ^DTT and (b) cellular ROS_v to different pollution sources. The solid and dashed arrows denote the mineral dust event and the BB event, respectively. The unit of ROS_v mean fold changes of fluorescence when PM₁₀ in unit volume of air is extracted in unit volume of water. The cellular ROS_v was measured independently two times with three replicas of each treatment. Raw fluorescence data were normalized to the corresponding blank fluorescence signal from the same experiment. A dashed line is plotted as a reference at y = 1. The standard deviation is shown as shadow. The solid and dashed arrows denote the mineral dust event and BB event, respectively.

2

Decreased and remaining OP_v ^DTT with the method of (a) Chelex ion exchange, (b) EDTA chelation. The solid and dashed arrows denote the mineral dust event and the BB event, respectively.

3

(a) Time series of OP_metals ^M and OP_metals ^R; (b) the relationship between (OP_metals ^M–OP_metals ^R) and the proportion of OP_metals ^R in OP_v ^DTT. The OP_metals ^M was quantified by the Chelex ion exchange method. The solid and dashed arrows denote the mineral dust event and the BB event, respectively.

4

OP^DTT dosage in the human respiratory tract in different periods per unit (a) time, and (b) mass of PM₁₀.

5

Correlation between OP_v ^DTT and cellular ROS_v in anthropogenic-dominant samples and the remaining samples. Anthropogenic-dominant: anthropogenic sources (BB, traffic, and industrial dust) accounted for >50% of both OP_v ^DTT and ROS_v in 14 samples, and they are listed in Table S1.

6

Gene_v following exposure to WS-PM₁₀ for (a) Cyp1a1, (b) Cat, (c) Gclc, (d) Gpx, (e) HO1, and (f) SOD2 relative to the control group. The unit of gene_v means fold changes of gene expression when PM₁₀ in unit volume of air is extracted in unit volume of water. Experiments were performed independently two times with three replicas of each treatment. Error bars indicate standard error. Bars with different letters have significantly (p < 0.05) different mean values, which were assessed with one-way ANOVA followed by post hoc tests with Tukey HSD (under variance homogeneity) or with Tamhane’s T2 (under variance heterogeneity).