In vitro toxicity assessment of atmospheric particulate matter on human lung and hepatic cells with agar membrane-based sampling and exposure strategy

ABSTRACT

In vitro toxicology research into the health effects of atmospheric particulate matter (PM) depends on the extraction of PM from filters. Previously, we proposed an efficient PM sampling and extraction method based on low-cost agar membrane and performed an exposure study with RAW264.7 macrophages. Here, we extended the application of this strategy, employing lung and hepatic cells, to validate its reliability and reproducibility. One traditional strategy using polytetrafluoroethylene (PTFE) filter was adopted for comparison. Cytotoxicity and proteomics results showed that the PM extracted by two methods induced comparable toxicity to A549 and BEAS-2B cells, while the PM extracted from the agar membranes induced higher toxicity to HepG2 and HL7702 cells than that from the PTFE filters. The differences in the investigations might be associated with the cell's sensitivity to the extracted suspended particles. This work indicated the importance of assessing PM cytotoxicity from the perspective of extraction methods and cell lines.

KEYWORDS:

• Atmospheric particulate matter
• agar membrane
• lung cells
• hepatic cells
• cytotoxicity

1. Introduction

Atmospheric particulate matter (PM) is a prominent factor threatening human health, especially fine particulate matter (PM_2.5) which is considered respirable and can penetrate deeply into the respiratory system and possibly enter the bloodstream [1]. Epidemiological studies have largely demonstrated that PM is closely related to the development of many major diseases, such as stroke, cardiovascular disease, chronic obstructive pulmonary disease, lung cancer, etc. [2–7]. In vitro toxicology research is normally adopted to explore the potential toxicity mechanism of PM, which is simple, cost-effective and convenient.

Until now, for in vitro toxicity investigations, the commonly used strategy is detaching PM from the filters and then adding PM extracts into cell culture medium [8]. This method has the following advantages: (a) is less demanding on experimental conditions, (b) can employ on any cells, (c) can address spatially, temporally, and seasonally variable PM samples, thus helping us to broaden our understanding of the potential cytotoxicity of PM and to find out the toxic drivers. However, the traditional method also has its limitations. Firstly, using different filters can affect the mass or components of PM collected [9]. In our previous work, we also reminded that polytetrafluoroethylene (PTFE) filter would absorb free polycyclic aromatic hydrocarbons (PAHs) in the gas phase of the atmosphere due to its strong hydrophobicity [10]. Secondly, different extraction solutions can differ compositionally from PM; generally, water (including buffer solution) is mainly to extract water-soluble components of PM, including inorganic salts, formic acid, acetic acid and other soluble organic matters, while organic solvents (dichloromethane, n-hexane, acetone, etc.) are mainly to extract fat-soluble components of PM, such as PAHs, fatty acid, long-chain alkane, etc. [8,11–13]. Thirdly, fibrous filters which capture PM by the combination of physical barriers and adhesion [14] cannot extract completely as some particles embedded in fibers. Therefore, the traditional strategy requires some vigorous extraction methods, such as sonication, probe sonication, and Soxhlet extraction [15–18]; however, these operation procedures will influence the morphology, component, and other physical or chemical properties of PM. At the same time, fibers of the fiber filters can be detached during the extraction process [19,20]. And then, these various collection filters, extraction solutions and extraction methods can influence toxicity results, and may misrepresent the toxicity responses to actual PM samples [20–22]. As there is no standard protocol for PM collection and extraction procedures, it is difficult for interlaboratory comparison and integration of PM toxicology researches in a global perspective.

In our previous study, we prepared a freeze-dried agar membrane with a flat surface and porous interior, which was biocompatible and not easy to breed bacteria; then we proposed a new strategy for PM sampling and extraction based on the agar membrane for the first time, which had a good comparability in PM sampling efficiency and low disturbance in extraction [10]. Since PM collected on the agar membranes could be directly suspended in culture medium by vortexing, so there was no need for concentration and reconstitution when applied in in vitro cell exposure. We compared this strategy with a traditional one, in which PM collected on the PTFE filters was extracted with water by ultrasonication followed by lyophilized and resuspended in culture medium, in terms of the responses of RAW264.7 macrophage cell line. The experimental results suggested that the agar membrane-based strategy was a potential solution for a standard PM sampling and exposure method with comparability.

As a follow-up of preliminary study, the key objective here was to further verify the reliability and reproducibility of the agar membrane-based PM sampling and exposure strategy. Human lung and hepatic cells were chosen, as lung epithelial cells were a first line of defense against PM exposure through the respiratory tract [23–25], while liver was a target organ for PM toxicity, which played an initial role in metabolism and detoxification of environmental pollutants [26]. On the other hand, there are limited studies comparing the responses of human normal and cancerous cells when exposed to PM, especially, hepatoma cells and normal hepatocytes. Therefore, herein, we determined the cytotoxicity effects of PM extracts from two methods on human lung adenocarcinoma cell-line A549, normal lung epithelial cell-line BEAS-2B, hepatic carcinoma cell-line HepG2 and human normal liver cell-line HL7702. Meanwhile, the differentially expressed proteins (DEPs) and enriched KEGG pathways in the cells after PM exposure were investigated by proteomic analyses, using iTRAQ quantitative proteomic technique, which provides plentiful information to elucidate the toxicological mechanism of PM.

2. Materials and methods

2.1 PM collection and extraction methods

PM was collected with the home-made agar membranes (prepared from 10% agarose solution by freeze-drying) [10], and commercial PTFE filters (Xingya Co. Ltd., China) simultaneously using four Andersen eight-stage non-viable cascade impactors (TE-20-800, Tisch, U.S.A.). The sampling was carried out every 10 days for 24 h each day, between November 2018 and December 2019, at the Xianlin campus of Nanjing University, Nanjing, China. The PM was segregated into nine size fractions by the Anderson impactors: >9.0 μm, 5.8–9.0 μm, 4.7–5.8 μm, 3.3–4.7 μm, 2.1–3.3 μm, 1.1–2.1 μm, 0.7–1.1 μm, 0.4–0.7 μm, and <0.4 μm. The morphology characterization and chemical composition analysis of the PM collected could be found in our previous work [10]. And the fine PM collected on three layers of agar and PTFE films (0.4–0.7 μm, 0.7–1.1 μm, and 1.1–2.1 μm PM deposited) were extracted and combined. For the agar membranes, the PM samples were suspended directly in culture medium after vortexing, which is denoted as ‘Agar-PM’. For the PTFE filters, the PM samples were resuspended in culture medium after extraction with double distilled water by sonicating and then freeze-drying in vacuum, which is denoted as ‘PTFE-PM’. The final concentration of the PM samples was 600 μg mL⁻¹. In our previous work [10], the metal element concentrations, light scattering intensities and light absorption intensities of the PM extracts were measured. Herein, the number and density of particles in the PM extracts were further characterized. Specifically, 10 μL drops of Agar-PM and PTFE-PM were respectively applied to the surface of aluminum sheets, dried at room temperature, and then photoed by scanning electron microscopy (SEM; S-4800, Hitachi, Japan) at 5 kV. Unsampled agar membranes and PTFE filters were extracted referred to the respective extraction procedures, and they were considered as ‘blank’ extracts with a PM concentration of 0 μg mL⁻¹.

2.2 Cell culture

A549 and BEAS-2B cells were purchased from Cobioer Biotech (Nanjing, China) and were cultured in Ham’s F-12K medium (Gibco, U.S.A.) supplemented with 10% fetal bovine serum (FBS) (Gibco, U.S.A.) and 100 U/100 mL penicillin/streptomycin (Gibco, U.S.A.). HepG2 and HL7702 cells were kindly provided by State Key Laboratory of Pharmaceutical Biotechnology (Nanjing, China) and were cultured in DMEM high glucose medium (Gibco, U.S.A.) supplemented with 10% FBS and 100 U/100 mL penicillin/streptomycin. All the cells were cultured at 37°C under 5% CO₂ humidified atmosphere.

2.3 Cell viability

The cell viability was assessed by cell counting kit-8 (CCK-8; Beyotime, China) assay. The cells were seeded into 96-well plates at a density of 5 × 10³ cells per well in 100 μL of the complete culture medium and cultured for 24 h. After incubation, the cells were treated with PM extracts at the concentrations of 0, 50, 75, 100, 200, 300 μg mL⁻¹ for 24 h. After the exposure, the suspensions were removed, and the cells were rinsed twice with 100 μL PBS. Then, 100 μL of the 10% CCK-8 was added to each well and incubated at 37°C for 1 h. The absorbance was measured at 450 nm via a microplate reader (Cytation 3, Biotek, U.S.A.). The viability of the treated cells was calculated as a percentage by comparing the absorbance of untreated control cells.

2.4 Intracellular ROS measurement

The intracellular ROS generation was measured using 2,7-dichlorodihydrofluorescein diacetate (DCFH-DA) probe (Beyotime, China). Briefly, the cells were seeded into 6-well plates at a density of 1 × 10⁵ cells per well in 2 mL of the complete culture medium and cultured for 24 h. After incubation, the cells were treated with PM extracts at the concentrations of 0, 50, 75, 100, 200, 300 μg mL⁻¹. After 24 h exposure, the cells were collected and rinsed twice with PBS. Then the cells incubated with 200 μL DCFH-DA (10 μM in serum-free medium) in the dark for 20 min. Next, the cells were washed three times with 200 μL cold serum-free DMEM medium and resuspended in 400 μL PBS for detection. The fluorescent intensity was detected by a flow cytometry (Cytomics FC 500, Beckman Coulter, U.S.A.), and 1 × 10⁴ cells were collected for each sample. The ROS level of the treated cells was calculated as a fold change by comparing the DCFH-DA fluorescent intensity of untreated control cells.

2.5 iTRAQ analysis

The cells were seeded with a density of 4 × 10⁵ cells per dishes and cultured for 24 h. After incubation, the cells were treated with PM extracts at the concentrations of 0, 200 μg mL⁻¹ for 24 h. To analyze the protein expression, the cells were lysed in RIPA lysis buffer (Beyotime, China), and the protein concentrations were determined by BCA protein assay (Beyotime, China). For each sample, 200 μg protein was firstly precipitated with acetone at −20°C for 1 h, then centrifuged to discard the supernatant. The precipitated protein was dissolved in 100 mM triethylamonium bicarbonate (TEAB) buffer. The protein solution was reduced with 4 μL dithiothreitol (1 M in TEAB buffer) at 56°C for 1 h and blocked with 10 μL iodoacetamide (1 M in TEAB buffer) at room temperature for 30 min in the dark. Then the protein sample was ultrafiltered and digested with trypsin at 37°C overnight. The obtained tryptic peptides were labeled with the iTRAQ reagent-8 plex multiplex kit (AB Sciex, U.S.A.) according to the instructions. At last, labeled samples were measured by an LC-MS/MS (Nano LC-Triple TOF^TM 5600+, AB Sciex, U.S.A.).

2.6 Apoptosis determination

HepG2 and HL7702 cells were seeded at a density of 1 × 10⁵ cells per well in 6-well plates and cultured for 24 h. After incubation, the cells were treated with a series concentration of PM extracts for 24 h. All the cells were then collected and resuspended in PBS. The apoptosis rate was determined by an Annexin V-FITC/Propidium Iodide (PI) apoptosis detection kit (Beyotime, China). Firstly, 195 μL Annexin V-FITC binding solution was added to each sample to completely suspend the cells. Then the cells were incubated with 5 μL Annexin V-FITC and followed by stained with 10 μL PI. Finally, the fluorescent intensity was detected by the Cytomics FC 500 flow cytometry (Beckman Coulter, U.S.A.), and 1 × 10⁴ cells were collected for each sample.

2.7 Western blotting analysis

For verifying the proteomics results, the protein expressions of some key proteins in NF-κB signaling pathway (including p50, p65, phospho-p65, IκBα) were tested by western blotting. Firstly, the proteins were boiled with loading buffer at 95°C for 10 min. Next, the proteins were separated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), and subsequently transferred to polyvinylidene difluoride (PVDF) membranes. The PVDF membranes were blotted in 5% BSA at room temperature for 2 h and then incubated with the primary antibodies at 4°C overnight. The membranes were then incubated with horseradish peroxidase-labeled secondary antibodies at room temperature for 2 h. Thereafter the membranes were determined by ECL kit (Beyotime, China) and visualized using a luminescence imaging system (Tanon, China). The band density was measured using Image J software (NIH, U.S.A.).

2.8 Statistical and bioinformatics analysis

Each experiment was repeated at least three times. One-way analysis of variance (ANOVA) was used to determine whether there were significant differences across the treatment groups, and statistically significant differences were signed with *: p < 0.05, **: p < 0.01, and ***: p < 0.001. Bioinformatics analysis was referred to our previous work [10,22]. The ProteinPilot^TM software version 4.2 (AB Sciex, U.S.A.) was applied to process raw data against the homo sapiens from UniProt database (https://www.uniprot.org/). Only peptides with a global false discovery rate (FDR) <0.01 and a 95% confidence interval (CI) were accepted for proteins annotation. The treatment/control ratios were applied to determine the fold-changes for proteins. The DEPs were identified based on the principle that fold-change >1.50 or <0.67 and p value <0.05. DAVID bioinformatics resources version 6.8 (https://david.ncifcrf.gov/) was applied for Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway and Gene Ontology (GO) annotation analysis. Hypergeometric tests were used to determine the enrichment of GO terms and KEGG pathways, and a p value <0.05 was considered statistically significant.

3. Results and discussions

3.1 Cytotoxicity assessment of lung cells after exposed to Agar-PM and PTFE-PM

The cell viability of A549 and BEAS-2B cells did not decrease significantly after exposed to PM for 24 h (Figure 1a). When the PM concentration was increased to 500 μg mL⁻¹, the cell viability subsequently decreased, showing a dose-dependent manner (Figure S1). Especially, the viability of A549 and BEAS-2B cells affected by Agar-PM was similar to that by PTFE-PM, illustrating that the two PM extraction methods were comparable in terms of PM-induced cytotoxicity to these two lung cells. Figure S2 showed the SEM images of Agar-PM and PTFE-PM extracts, and it was found that the number of particles in Agar-PM was higher than that in PTFE-PM, and the particle density of Agar-PM was more uniform than that of PTFE-PM. Meanwhile, probably due to the aggregation of particles on the fibers of the PTFE filters, there were some large particle aggregates in PTFE-PM, which was consistent with our previous observation of particles on the agar and the PTFE films [10]. It concluded that more particles were separated from the agar membranes than from the PTFE filters. Thus, the result indicated that the difference in the number of particles between the two extracts had no significant effect on A549 and BEAS-2B cells. However, the viability of BEAS-2B cells influenced by Agar-PM and PTFE-PM was lower than that of A549 cells at higher levels of exposure (Figure S1). It suggested that normal BEAS-2B cell was more sensitive to the PM exposure than cancerous A549 cell, which was consistent with existing literature [27,28]. The impacts of PM on cellular oxidative stress were assessed by measuring the generation of intracellular ROS. As shown in Figure 1b, after 24 h exposure, the intracellular ROS levels of A549 cells also did not change significantly. Meanwhile, the intracellular ROS levels induced by Agar-PM were comparable to those by PTFE-PM, which was tallied with the results of the viability experiment. As for BEAS-2B cells, the changes of intracellular ROS levels were similar to those of A549 cells, but not the same, as it showed a slight increase in ROS levels at higher concentrations (Figure S1).

Figure 1. Cell viability (% of untreated control) (a) and intracellular ROS level (fold change compared to untreated control) (b) of A549 and BEAS-2B cells after exposed to PTFE-PM and Agar-PM for 24 h.

3.2 Cytotoxicity assessment of hepatic cells after exposed to Agar-PM and PTFE-PM

The cell viability of HepG2 cells decreased in a dose-dependent manner after exposed to PTFE-PM and Agar-PM for 24 h (Figure 2a). However, in HL7702 cells, the viability did not change significantly after exposed to Agar-PM, whereas the viability of cells slightly increased after exposed to PTFE-PM (Figure 2a). When extended the exposure time to 48 h, the cell viability of HL7702 cells was significantly decreased (Figure S3), showing a time-dependent manner. It seemed that HL7702 cells exhibited apparent tolerance to PM exposure in comparison with HepG2 cells, which was consistent with a previous study where dust extract induced higher cytotoxicity to HepG2 cells than HL7702 cells [29]. In addition, the intracellular ROS levels of HepG2 and HL7702 cells were also measured. The data showed that the intracellular ROS levels of both cells increased significantly compared with the untreated cells at 24 h exposure (Figure 2b). At the same PM concentration, the intracellular ROS level induced by Agar-PM was significantly higher than that induced by PTFE-PM. The cytotoxicity investigation results suggested that the PM extracted from the agar membranes was more toxic to these two hepatic cells than the PM extracted from the PTFE filters. Moreover, it was also observed that PM induced higher ROS level in HepG2 than in HL7702 cells. Previously, Zhang et al. reported that PM_2.5 effectively promoted the ROS production of HepG2 cells and inhibited proliferation of HL7702 cells; however, there was no comparison between HepG2 and HL7702 cells [30]. Xiang et al. compared the ROS levels and oxidative stress biomarker levels of HepG2 and HL7702 cells exposed to organic dust extract for 24 h and suggested that dust extract contributed to higher oxidative stress to HepG2 than HL7702 cells [29]. In general, in this work, the ROS results combined with the cell viability results revealed that there were differences in cellular response between cancerous and normal hepatic cell lines when exposed to PM.

Figure 2. Cell viability (% of untreated control) (a) and intracellular ROS level (fold change compared to untreated control) (b) of HepG2 and HL7702 cells after exposed to PTFE-PM and Agar-PM for 24 h.

3.3 DEPs of lung/hepatic cells after exposed to Agar-PM and PTFE-PM

To elucidate the differences in cytotoxicity between normal and cancerous lung/hepatic cells, the detailed toxicological proteomics was then discussed. The iTRAQ-based proteomic analyses of A549 and BEAS-2B cells treated with fine PM extracts identified a total of 5303 and 4749 proteins, respectively. Of these proteins, 134 and 100 DEPs were selected in A549 cells after exposed to PTFE-PM and Agar-PM, respectively. Correspondingly in BEAS-2B cells, 266 and 249 DEPs were selected. The Venn diagram analysis illustrated the numbers of total, up-regulated and down-regulated DEPs in A549 and BEAS-2B cells after exposed to PM (Figure 3). The results showed that PM induced more DEPs in BEAS-2B cells compared with A549 cells.

Figure 3. Venn diagram analysis of total, up-regulated and down-regulated DEPs in A549 (I) and BEAS-2B (II) cells after exposed to PM for 24 h; and top 15 GO enrichment terms in A549 (III) and BEAS-2B (IV) cells (A: PTFE-PM, B: Agar-PM). The numeric label above the bar showed the fold enrichment. Note: Separate Venn diagrams and GO enrichment diagrams could be found in the Supporting Information.

The results of GO annotation of the DEPs in A549 and BEAS-2B cells after exposed to PM extracts are itemized in the Supporting Information (Tables S1-S12). Meanwhile, Figure 3 presents the respective top 15 GO enriched terms for molecular functions (MF), biological processes (BP) and cellular components (CC). Although the specific terms were somewhat different, or the number of enriched proteins was different in some common terms, the results revealed that both PTFE-PM and Agar-PM could induce toxicity to lung epithelial cells mainly in the cell adhesion, biosynthetic and metabolic processes. This explained why, in A549 and BEAS-2B cells, no significant increase in ROS levels was observed, and cell viabilities were not significantly affected, especially when stimulated at low concentrations.

The iTRAQ-based proteomic analyses of HepG2 and HL7702 cells treated with fine PM extracts identified a total of 4630 and 3892 proteins, respectively. Of these proteins, 136 and 82 DEPs were selected in HepG2 cells after exposed to PTFE-PM and Agar-PM, respectively. Correspondingly in HL7702 cells, 240 and 187 DEPs were selected. The Venn diagram analysis illustrated the numbers of total, up-regulated and down-regulated DEPs in HepG2 and HL7702 cells after exposed to PM (Figure 4). Similarly, it was observed that PM induced more DEPs in HL7702 cells compared with HepG2 cells. The proteomics results suggested that the normal cells were more sensitive to PM than the cancerous cells.

Figure 4. Venn diagram analysis of total, up-regulated and down-regulated DEPs in HepG2 (I) and HL7702 (II) cells after exposed to PM for 24 h; and top 15 GO enrichment terms in HepG2 (III) and HL7702 (IV) cells (A: PTFE-PM, B: Agar-PM). The numeric label above the bar showed the fold enrichment. Note: Separate Venn diagrams and GO enrichment diagrams could be found in the Supporting Information.

The results of GO annotation of the DEPs in HepG2 and HL7702 cells after exposed to PM extract are itemized in the Supporting Information (Tables S13-S24). At the same time, the respective top 15 GO enriched terms for MF, BP, and CC in HepG2 and HL7702 cells are also shown in Figure 4. In the main, the data revealed that both PTFE-PM and Agar-PM affected a great quantity of proteins related to cell adhesion, metabolic processes, and biological regulation (including protein, DNA and RNA binding).

3.4 Enriched KEGG pathways of lung cells after exposed to Agar-PM and PTFE-PM

The enriched KEGG pathways based on DEPs in A549 and BEAS-2B cells after exposed to Agar-PM and PTFE-PM for 24 h are shown in Figure 5. Only a few KEGG pathways were enriched in A549 cells whether exposed to PTFE-PM or Agar-PM. Among them, ribosome and spliceosome pathways have been reported by Kumar et al. to be impacted by steel industry PM [31]. However, other pathways in A549 cells affected by PM in our present study were dissimilar from those affected by PM from heavy fuel oil and diesel fuel shipping emission [32] or in urban air [23], which was due to the intrinsic differences in the chemical composition, size distribution and mechanical properties of PM from different sources. Base excision repair (BER) pathway was a common enriched pathway in A549 cells treated with Agar-PM and PTFE-PM, and the KEGG maps are shown in Figure S4. BER pathway was one of the major pathways for the repair of DNA damage from oxidation, alkylation and deamination [33,34]. Our result indicated that the BER signaling pathway might be involved in DNA damage induced by PM in A549 cells.

Figure 5. KEGG pathways analysis of DEPs in A549 and BEAS-2B cells after exposed to PM for 24 h (A: PTFE-PM, B: Agar-PM).

As for BEAS-2B cells, there were more KEGG pathways enriched by DEPs compared with A549 cells. And the majority of the enriched pathways were involved in metabolic processes. Meanwhile, we noticed that ‘protein processing in endoplasmic reticulum (ER)’ pathway was enriched both in BEAS-2B and A549 cells after PM exposure. Their representative KEGG maps are shown in Figure S5 and Figure S6, respectively. ER was involved in synthesis, modification and folding of protein. As shown in Figure S5, the most enriched protein along this pathway in BEAS-2B cells was the ER-associated protein degradation (ERAD), which played an important role in cellular homeostasis by monitoring the degradation of misfolded molecules in the ER [35]. Inactive ERAD resulted in the accumulation of misfolded proteins in the lumen and membrane of the ER [36], which was called ER stress. Previous studies suggested that PM could cause ER stress in human lung cells, endothelial cells and skin cells [37–39]. Sun’s group also found that protein processing in the ER was one of the most critical pathways in PM-induced toxicity in human umbilical vein endothelial cells (HUVECs) [40].

3.5 Enriched KEGG pathways of hepatic cells after exposed to Agar-PM and PTFE-PM

The enriched KEGG pathways based on DEPs in HepG2 and HL7702 cells after exposed to Agar-PM and PTFE-PM for 24 h are shown in Figure 6. For HepG2 cells, Agar-PM induced more pathway terms than PTFE-PM. The enriched KEGG pathways induced by PTFE-PM were mainly involved in carbohydrate metabolism. Ye et al. previously discovered that liposoluble extracts of PM_2.5 had a significant effect on metabolic pathways (including lipid, amino acid, nucleotide and carbohydrate metabolism) in HepG2 cells [41]. As for HL7702 cells, the data showed that enriched pathways affected by both PTFE-PM and Agar-PM also mainly took part in amino acid and carbohydrate metabolism.

Figure 6. KEGG pathways analysis of DEPs in HepG2 and HL7702 cells after exposed to PM for 24 h (A: PTFE-PM, B: Agar-PM).

However, the enriched KEGG pathways in HepG2 cells induced by Agar-PM were also involved in apoptosis, immune and inflammation response. Therefore, the cell apoptosis rates of HepG2 and HL7702 cells were further determined. After exposed to Agar-PM for 24 h, the apoptosis rate of HepG2 cells increased significantly, particularly under high concentration stimulation; whereas the apoptosis rate did not significantly increase after exposed to PTFE-PM (Figure 7a). And the data showed that there was no apparent apoptosis in HL7702 cells regardless of exposure to Agar-PM or PTFE-PM (Figure 7a), which was consistent with the cell viability and proteomics results.

Figure 7. Apoptosis rate (%) of HepG2 and HL7702 cells after exposed to Agar-PM and PTFE-PM for 24 h (a). Expressions of p65, phospho-p65, p50 and IκBα in HepG2 cells after exposed to Agar-PM and PTFE-PM for 24 h (b).

It was worth noting that NF-κB pathway was involved in both of the most enriched pathways in HepG2 cells after exposed to Agar-PM (Figures S7-S8). Furthermore, to validate the proteomics results, the expression levels of key proteins were determined by western blotting. As shown in Figure 7b, after exposed to Agar-PM, the expression levels of p50 and phospho-p65 were significantly up-regulated, and the expression level of IκBα was significantly down-regulated in HepG2 cells. However, the expression levels of p50, phospho-p65 and IκBα did not change after exposed to PTFE-PM. The NF-κB transcription factor family includes NFKB1 (p50/p105), NFKB2 (p52/p100), RelA (p65), c-Rel and RelB [42,43]. In canonical NF-κB signaling pathway, when IκB kinases (IKKs) are activated, IKKs can phosphorylate IκBs, inducing IκBα degradation in the cytoplasm, and releasing p50 and p65 [43]. Thus permitting NF-κB to translocate to the nucleus where it activates the expression of a variety of genes [44]. The results preliminarily indicated that the PM extracted from the agar membranes could activate NF-κB pathway in HepG2 cells. Previous study also showed that PM_2.5 suspension exposure could induce activation of NF-κB pathway in HepG2 cells [45]. However, further studies are still needed to find out whether tiny insoluble particles play a role in it. Overall, the PM had a more significant effect on the cellular mechanism of HepG2 cells compared with HL7702 cells, thus inducing lower viability and higher apoptosis rate at the cellular level, indicating that HepG2 cells was more prone to stress response to PM. The proteomic findings could shed new light on the mechanisms of PM toxicity to human hepatic cells.

4. Conclusions

In this study, we extended the applications of the PM sampling and exposure strategy based on the agar membrane to investigate the cytotoxicity of PM extracts on two human lung cell lines and two human hepatic cell lines. Meanwhile, the traditional strategy, in which fine PM collected on the PTFE filters was extracted with water by ultrasonication for comparison. The cytotoxicity results showed that the two PM extraction methods were comparable in terms of PM-induced cytotoxicity in A549 and BEAS-2B cells. However, the PM extracted from the agar membranes induced higher toxicity to HepG2 and HL7702 cells than the PM extracted from the PTFE filters. This difference might be associated with the cells’ sensitivity to the suspended particles, as more particles were extracted from the agar membranes than from the PTFE filters. Moreover, the cytotoxicity results indicated that normal and cancerous cell lines behaved differently when exposed to PM. And the quantitative proteomics data showed that PM induced more DEPs in the two normal cell lines compared to the two cancerous cell lines, indicating that normal cell lines were more sensitive to PM than cancerous cell lines. Furthermore, the results of proteomics screened out that only the PM extracted from the agar membranes affected NF-κB signaling pathway in HepG2 cells. It is suggested that the agar membrane-based PM sampling and exposure strategy could provide a complementary approach to investigate the cytotoxicity mechanism of PM.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/26395940.2023.2233699

Data availability statement

All data reported here can be made available on request.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China (91643105, 91543129, 21874065, 22176085), and the Natural Science Foundation of Jiangsu Province (BK20181261).