Skip to main content
Download PDF
Download ePub

The chemical composition of secondary organic aerosols regulates transcriptomic and metabolomic signaling in an epithelial-endothelial in vitro coculture

Particle and Fibre Toxicology volume 21, Article number: 38 (2024) Cite this article

Abstract

Background

The formation of secondary organic aerosols (SOA) by atmospheric oxidation reactions substantially contributes to the burden of fine particulate matter (PM2.5), which has been associated with adverse health effects (e.g., cardiovascular diseases). However, the molecular and cellular effects of atmospheric aging on aerosol toxicity have not been fully elucidated, especially in model systems that enable cell-to-cell signaling.

Methods

In this study, we aimed to elucidate the complexity of atmospheric aerosol toxicology by exposing a coculture model system consisting of an alveolar (A549) and an endothelial (EA.hy926) cell line seeded in a 3D orientation at the air‒liquid interface for 4 h to model aerosols. Simulation of atmospheric aging was performed on volatile biogenic (β-pinene) or anthropogenic (naphthalene) precursors of SOA condensing on soot particles. The similar physical properties for both SOA, but distinct differences in chemical composition (e.g., aromatic compounds, oxidation state, unsaturated carbonyls) enabled to determine specifically induced toxic effects of SOA.

Results

In A549 cells, exposure to naphthalene-derived SOA induced stress-related airway remodeling and an early type I immune response to a greater extent. Transcriptomic analysis of EA.hy926 cells not directly exposed to aerosol and integration with metabolome data indicated generalized systemic effects resulting from the activation of early response genes and the involvement of cardiovascular disease (CVD) -related pathways, such as the intracellular signal transduction pathway (PI3K/AKT) and pathways associated with endothelial dysfunction (iNOS; PDGF). Greater induction following anthropogenic SOA exposure might be causative for the observed secondary genotoxicity.

Conclusion

Our findings revealed that the specific effects of SOA on directly exposed epithelial cells are highly dependent on the chemical identity, whereas non directly exposed endothelial cells exhibit more generalized systemic effects with the activation of early stress response genes and the involvement of CVD-related pathways. However, a greater correlation was made between the exposure to the anthropogenic SOA compared to the biogenic SOA. In summary, our study highlights the importance of chemical aerosol composition and the use of cell systems with cell-to-cell interplay on toxicological outcomes.

Graphical Abstract

Background

Air pollution by aerosols is a complex mixture of compounds in the gaseous and particulate phases. It is well established that outdoor air pollution has adverse health outcomes, resulting in high levels of mortality and morbidity worldwide [1,2,3]. Multiple studies have shown a strong correlation between exposure to ambient fine particulate matter (PM2.5) and the occurrence of respiratory and cardiovascular diseases (CVD). Due to its size, PM2.5 can penetrate and remain deep in the lungs and induce direct pulmonary effects, as well as effects on other organs or systemic effects, such as those in the cardiovascular system [4, 5]. A cohort study of healthy individuals revealed endothelial cell dysfunction as a key player in the induction of CVD after PM2.5 inhalation [6]. Supporting evidence was also found in the in vivo experiments by Davel, Lemos [7] and Tamagawa, Bai [8] who demonstrated vascular endothelial dysfunction after PM exposure.

A major fraction of ambient PM is, however, composed of secondary organic aerosols (SOA), generated from low volatility products of atmospheric oxidation reactions of gaseous anthropogenic or biogenic organic compounds [9]. SOA formation can occur either through nucleation and growth of new molecular clusters [10] or condensation of oxidized vapors on preexisting particles [e.g., soot particles (SP)] [11, 12]. The composition of SOA includes a vast range of organic compounds, which are functionalized with moieties of particularly carboxylic acids, carbonyls and alcohols. However, the individual SOA composition varies upon SOA precursor and atmospheric aging conditions [13, 14]. Previous studies investigated several biological endpoints in in vitro exposures or the oxidative potential of SOA in acellular assays from combustion aerosols, such as from biomass burning or road traffic, or SOA precursors of biogenic or anthropogenic origin [15]. Recently, exposure to SOA has been shown to contribute substantially to mortality associated with air pollution [16]. Pye, Ward-Caviness [17] demonstrated a correlation between annual average SOA concentrations and county-level cardiorespiratory death rates. The importance of understanding the relationship between aerosol composition and adverse biological effects for determining the toxicity mechanisms of different SOA has been demonstrated in studies with human airway cell lines [18,19,20,21] and lung lining fluid [22]; however, this topic elusive in model systems that enable cell-to-cell signaling.

In this follow-up work, we aimed to elucidate the complexity of atmospheric aerosol toxicology by applying controlled model aerosols for realistic in vitro exposure at the air-liquid interface (ALI) by considering cell-to-cell signaling possibly driving systemic adverse effects. To achieve this goal, we have exposed a coculture model system consisting of a lung epithelial cell line (A549) and an endothelial cell line (EA.hy926) in a 3D orientation. A549 cell line is a widely used model cell line for nanoparticle toxicological studies, particularly concerning lung metabolism and carcinogenesis. When cultured at the ALI, it has been shown that A549 cells can express important characteristics of alveolar epithelial type II cells that are essential for the defense of inhaled toxicants (e.g., tight epithelial phenotype, surfactant production, detoxification by metabolic enzymes) [23]. To mimic the blood-air barrier as well as to investigate the interaction between the cells, we seeded EA.hy926 cells on the basolateral side of the insert membrane. This coculture model system was exposed to different dilutions (undiluted, 1:3; 1:10 or 1:30) of either an anthropogenic SOA from an aromatic precursor (naphthalene), mixed with SP (SOANAP-SP), or a biogenic SOA from an aliphatic precursor (β-pinene) mixed with SP (SOAβPIN-SP). Comparisons were made with fresh SP. Recently, we were able to detect secondary genotoxicity and increased angiogenic potential in endothelial cells only after exposure to SOANAP-SP [21], thus we hypothesized that non directly exposed endothelial cells may be differentially affected by the type of SOA. Therefore, we compared functional, transcriptional, and metabolic changes in the coculture model system after the exposure to SP, SOAβPIN-SP or SOANAP-SP at the ALI system. An aerosol dilution of 1:3 was the focus of the study as we previously observed low toxic results and detected DNA breaks in the non-directly exposed EA.hy926 cells [21]. A549 and EA.hy926 cells were separated by an insert membrane, which enabled us to trace their specific transcriptomic changes. Moreover, secreted mediators, including cytokines and metabolites found in the sampled media, were used to validate the observed transcriptomic outcomes and to draw possible conclusions on the type-specific molecular effects of SOA.

Results and discussion

Experimental set-up

We have previously demonstrated that the exposure of our coculture model system (A549/EA.hy926 cells) to SOANAP-SP revealed greater toxic effects than SOAβPIN-SP and pure SP. Calculated cellular deposition of the aerosol dilutions (undiluted, 1:3, 1:10 and 1:30) ranged from 0.9 ng/cm2 (1:30 dilution) to 28 ng/cm2 (undiluted) for SOANAP-SP and from 0.6 ng/cm2 (1:30 dilution) to 17 ng/cm2 (undiluted) for SOAβPIN-SP. According to calculations made by Paur, Cassee [24] for in vitro nanotoxicology studies, a cellular deposition of 0.75 ng/cm2 is suggested for a realistic ambient exposure up to 130 ng/cm2 for worst-case occupational exposure which corresponds to a daily dose Therefore, our estimated cellular deposition is covering different exposure conditions ranging from ambient to mild occupational settings [21, 24]. In addition, the exposure time was set at 4 h to obtain first indications of acute stress responses of the cells. At the functional level, we observed that SOANAP-SP augmented the secretion of lactate dehydrogenase (LDH), malondialdehyde (MDA) and interleukin 8 (IL8) as signs for cytotoxicity, oxidative stress, and inflammation. Moreover, we detected secondary genotoxicity and a greater angiogenic potential in EA.hy926 cells exposed to SOANAP-SP, suggesting a possible activation of the non-directly exposed endothelial cells in the coculture system. Table S1 summarizes previously published results [21]. In this work we elucidate the transcriptional and metabolic changes in our coculture model system after SOA exposure and focused on aerosol dilution 1:3 that corresponds to low toxic exposures [21].

Characterization of the generated SOA revealed similar physical but distinct chemical properties

To investigate the role of SOA to influence cellular mechanisms, we took advantage of previously described in-depth characterization of the two models of SOA [21]. Briefly, the aging of both SOA precursors (biogenic: β-pinene and anthropogenic: naphthalene) together with soot particles was conducted under the same OH exposure, which is equivalent to approximately 3 days of atmospheric photochemical aging. Using pure CASTs soot as primary particles, a substantial fraction of the photooxidation products of β-pinene and naphthalene condensed on the SP, resulting in SOAβPIN-SP and SOANAP-SP. Consequently, the particle size distributions and the particle diameters (means ∼ 115 nm) of the two SOA were remarkably similar. However, there were differences in the chemical properties of the aerosol particles and the gas phase.

Particularly in the semi-volatile fraction of the PM, which was investigated by two-dimensional gas chromatography-mass spectrometry, the chemical base structure of the SOA precursors was maintained. While SOAβPIN-SP contained aliphatic and cyclic SOA species, functionalized aromatic compounds were detected in SOANAP-SP. As two-dimensional gas chromatography-mass spectrometry is limited to semi-volatile compounds in aerosol particles, we investigated the abundance of this compound class via direct-infusion ultrahigh resolution mass spectrometry with the concept of the maximum carbonyl ratio (MCR) [25]. According to the electrospray ionization in positive mode (+ ESI) results for the two SOA-SP, a similar range of #O is covered, but the #O distribution in SOAβPIN-SP is more shifted to a larger #O than that in SOANAP-SP. In contrast, the DBE in SOANAP-SP has a multimodal distribution and correlates with the detection of dimers and trimers to larger oligomers [20], whereas the DBE in SOAβPIN-SP has a single mode at 5 and does not significantly exceed a DBE of 10. These molecular properties allow us to conclude that for SOANAP-SP > 50% of the peak intensity corresponds to oxygenated (O-containing) compounds belonging to unsaturated carbonyls, while for SOAβPIN-SP, the percentage is only ≈ 10% (Fig. 1). As a complementary ionization technique, ESI in negative mode (-ESI) is capable of detecting compounds with acidic functional groups, such as carboxylic acids, as well as 1,3-dicarbonyls and other derivatives with acidic hydrogen. A similar observation can be made in the -ESI results, but with a less distinct pattern for oligomeric structures (Figure S1). Moreover, the percentage of compounds with MCR ≥ 1 was lower than that with + ESI for both aerosols, revealing that a substantial number of carbonyl moieties with unsaturated carbon backbones are carboxyl groups or conjugated carbonyls.

Fig. 1
figure 1

Maximum carbonyl ratio (MCR) and aromaticity of SOANAP-SP and SOAβPIN-SP. Number of oxygen atoms (#O) vs. double bond equivalents (DBE) in SOANAP-SP (top) and SOAβPIN-SP (bottom) with pie charts depicting the percentage of peaks in individual bins of the maximum carbonyl ratio (MCR) and aromaticity index (AI) for ESI in positive mode

Full size image

To better understand the gas phase composition, we investigated the differences in the molecular composition by proton-transfer-reaction (PTR) high-resolution time-of-flight mass spectrometry (PTR-TOFMS; Table 1 and S2). Small molecules with carbon numbers ranging from one (e.g., formaldehyde) to ten (e.g., pinonaldehyde) had higher oxygen contents for volatile organic compounds (VOC) in SOANAP-SP (CHO1: 137 ± 12 ppb; CHOn>1: 173 ± 13 ppb) than in SOAβPIN-SP (CHO1: 336 ± 26 ppb; CHOn>1: 130 ± 13 ppb). However, oxygenated VOCs were overall more abundant in SOAβPIN-SP than in the other samples, which may be explained by the oxidation mechanism. In the OH-initiated photooxidation of β-pinene, acetone is one of the major first-generation oxidation products [26] and the most abundant individual VOC of SOAβPIN-SP in our study, followed by other small C1 and C2 species, i.e., formaldehyde (CH2O), methanol (CH4O), formic acid (CH2O2) and acetic acid (CH4O2).

Table 1 Chemical properties of the gas-phase of atmospheric photooxidation (aging) by OH radicals in a PAM (potential aerosol mass) reactor of SP (CAST soot; 1 mg/m3) together with either naphthalene (4 mg/m3) or β-pinene (4 mg/m3), forming SOANAP-SP and SOAβPIN-SP, respectively, measured by proton-transfer-reaction (PTR) high-resolution time-of-flight mass spectrometry PTR-TOFMS. The results are shown as the mean ± SD of n = 4 independent experiments
Full size table

In total, these five most abundant VOC accounted for 55% of the total VOC detected by PTR-TOFMS at SOAβPIN-SP. Generally, CH and N-containing VOC species originated from primary emissions of CAST or from unreacted SOA precursors, which was particularly the case for SOANAP-SP. Here, naphthalene exceeded the concentration of all the other detected VOC and accounted for 53% of the total CH species. The other abundant individual VOC for SOANAP-SP included an unknown species with the sum of the formula C5H2O2, methanol (CH4O), butene (C4H8) and formic acid (CH2O2). Despite the distinct differences in VOC concentration compared to that of SOAβPIN-SP, for SOANAP-SP, the five most abundant VOC accounted for 57% of the total detected VOCs. Taken together, the key findings over the entire molecular range were the prevailing aromaticity of the naphthalene precursor and its greater tendency to form unsaturated, highly oxygenated species compared to those of the β-pinene precursor.

ALI exposure to SOANAP-SP induced the strongest changes in both epithelial (A549) and endothelial (EA.hy926) human lung cells

As a result of the generated physically similar but chemically distinct SOA, we are now able to assess how different chemistries impact cell effects. To identify the molecular mechanisms induced by an anthropogenic (SOANAP-SP) and a biogenic (SOAβPIN-SP) model of SOA, as well as by SP alone, in our coculture system, the directly exposed A549 cells and the non-directly exposed EA.hy926 cells were separately harvested after 4 h to perform RNA-Seq (Fig. 2A). The results showed distinct gene expression patterns in A549 and EA.hy926 cells exposed to all aerosols (both SOA and SP) and to the clean air (CA) control (Fig. 2B and S2A). Next, we conducted differential gene expression analysis between cells exposed to SOAβPIN-SP, SOANAP-SP, or SP and those exposed to CA control. In A549 cells, we found 881 DEGs (298 genes upregulated and 583 genes downregulated) after SP exposure (Figure S2B and S3A), 1601 DEGs (532 genes upregulated and 1069 genes downregulated) after SOAβPIN-SP exposure (Figure S2B and S3B) and 2080 DEGs (805 upregulated and 1275 downregulated) after SOANAP-SP exposure (Figure S2B and S3C), indicating the differential impact of exposure both in terms of the number of regulated genes and the overall patterns of up- and downregulation, which varied from approximately 33–34% of upregulated DEGs following SP and SOAβPIN-SP exposure to 39% after SOANAP-SP treatment. The most up- and downregulated genes in A549 cells after the exposure to SP, SOAβPIN-SP or SOANAP-SP compared to those in the CA control are summarized in Tables S2, S3 and S4, respectively. Consistently, comparison of DEGs between SP, SOAβPIN-SP and SOANAP-SP revealed the most uniquely expressed genes after exposure to SOANAP-SP (50%) followed by SOAβPIN-SP (32%), whereas the exposure to SP showed the greatest overlap of DEGs (60%) between all three aerosol types in A549 cells (Fig. 2C). Gene Ontology (GO) analysis of the significantly upregulated genes in A549 cells revealed several enriched processes, such as “transmembrane transporter activity”, “receptor binding”, “immune responses”, and “metabolic processes”, especially for SOANAP-SP (Fig. 2D). Notably, channel and transporter activity are known to play important roles in pulmonary homeostasis, and their dysregulation could contribute to pulmonary diseases [27]. Within those GO groups, we detected several genes whose expression was significantly upregulated after exposure to both SOA in our dataset (MFSD2A, TRPC6 or KCNK1) that were recently correlated specifically with lung injury, airway inflammation and lung cancer [28,29,30]. Moreover, the importance of immunological and metabolic pathways activated in response to SOA exposure has already been noted at the protein level in monocultures of lung epithelial cells and murine macrophages [18, 31]. However, to our knowledge, a more in-depth characterization of the biological effects of direct and indirect exposure of cocultured cells is still lacking. By analyzing endothelial cells from the EA.hy926 cell line separately, we focused on cell-to-cell interactions with lung epithelial A549 cells. In EA.hy926 cells, we found 1167 DEGs (571 upregulated genes and 596 downregulated genes) after SP exposure (Figure S2B and S3D), 1239 DEGs (555 upregulated genes and 684 downregulated genes) after SOAβPIN-SP exposure (Figure S2B and S3E) and 1587 DEGs (674 upregulated genes and 913 downregulated genes) after SOANAP-SP exposure (Figure S2B and S3E). This finding is consistent with the greater toxicogenomic effects that we observed in A549 cells after exposure to SOANAP-SP. Interestingly, exposure to SOANAP-SP induced greater imbalances between the up- and downregulated genes (42.5% and 57.5% respectively) that were not detected after exposure to SP (49% up- and 51% downregulation) or SOAβPIN-SP (45% up- and 55% downregulation). The most up- and downregulated genes in the EA.hy926 cells after exposure to SP, SOAβPIN-SP or SOANAP-SP compared to the CA control are summarized in Tables S5,S6 and S7, respectively. An overlap of approximately 41% of similar DEGs after all three aerosol exposures was observed in EA.hy926 cells. The most uniquely expressed genes were detected after exposure to SOANAP-SP (25%), followed by SOAβPIN-SP (8%) and SP (7%) (Fig. 2E). For the significantly upregulated genes in EA.hy926 cells, GO analysis revealed enrichment of GO terms, such as “receptor activity”, “transcription activity”, “angiogenesis”, and “inflammation” for all aerosol types; however, there was a greater –log10(p-value) for SOANAP-SP (Fig. 2F). These findings suggested an early involvement of endothelial cell biology in systemic effects and are consistent with the findings of a study in healthy individuals highlighting endothelial cell injury and systemic inflammation caused by PM2.5 exposure [6].

Fig. 2
figure 2

Transcriptional analysis of aerosol-induced effects in A549 and EA.hy926 cells. (A) Schematic representation of the experimental setup. A coculture model system consisting of A549 lung epithelial cells on one side of the insert membrane and EA.hy926 endothelial cells on the other side were exposed for 4 h at the air-liquid interface (ALI) to two SOA, namely, SOAβPIN-SP and SOANAP-SP and SP. Transcriptional changes in A549 and EA.hy926 cells were also analyzed. (B) Principal component analysis (PCA) of all genes with batch correction and normalization to colors based on exposure to clean air (CA), SOANAP-SP, and SOAβPIN-SP; n = 4 independent experiments and SP of n = 3 experiments of A549 and EA.hy926 cells. (C) Venn diagram of the unique and overlapping genes associated with the up- and downregulated genes (adjusted p-value ≤ 0.05, and log2FC ≥ 0.5 and log2FC ≤ -0.5) in A549 cells after exposure to SP, SOAβPIN-SP or SOANAP-SP compared to the CA control. (D) Gene Ontology (GO) analysis of differentially expressed genes (DEGs) related to biological processes and molecular functions between A549 cells exposed to different aerosols (SP, SOAβPIN-SP, and SOANAP-SP) and the CA Ctrl. (E and F) Same as in (C and D) but based on the analysis of EA.hy926 cells

Full size image

Aerosol exposure to the tested aerosols induced stress- and inflammation-related gene expression in both epithelial (A549) and endothelial (EA.hy926) cells

To expand our understanding of the cellular effects to different SOA types, we next characterized the aerosol-specific impact on biological processes. We first performed a gene set enrichment analysis (GSEA) on all the genes, which detected small but consistent changes in a predefined set of genes. We sorted the resulting terms by a normalized enrichment score (NES) of either ≥ 2 or ≤ -2, categorized the remaining terms into their respective GO ‘parent’ terms (Excel Sheets S1 and S2) and clustered the comprised significant DEGs in a heatmap. In A549 cells, exposure to SOANAP-SP induced the upregulation of genes associated with the enriched terms for “response to stimuli”, “inflammatory response”, and “cell motility” (Fig. 3A and Excel Sheet S1). Moreover, we found more highly expressed cellular stress response genes, such as cAMP responsive element binding protein 1 (CREBL1) and fos proto-oncogene (FOS) (Fig. 3B), which are known to be involved in early endoplasmic reticulum (ER) stress and oxidative stress. This effect was less detectable after exposure to SP or SOAβPIN-SP. Several studies have linked cellular ER stress [32] and reactive oxygen species (ROS) generation [33] to redox-active quinones. Given that SOANAP-SP contains more quinones, such as 1,2- and 1,4-naphthoquinone, we hypothesized that these quinones would impact on gene expression. Moreover, the significantly enhanced expression of the gene NADPH oxidase 1 (NOX1) (Fig. 3B) possibly indicates the occurrence of ROS [34], which is consistent with the observed elevated levels of the oxidative stress marker malondialdehyde (MDA) after exposure to SOANAP-SP compared to those after exposure to SP and SOAβPIN-SP [21]. The increased expression of the gene prostaglandin-endoperoxidase synthase 2 (PTGS2) is likely due to its protective function against pulmonary oxidative stress [35] and acute lung injury [36]. This finding is in line with a study showing enhanced cellular expression of PTGS2 after exposure to diesel exhaust (PM2.5) [37]. In addition to demonstrating an increase in stress-related gene expression, we also observed the induction of inflammatory response genes [e.g., CXC motif chemokine ligand 1, 3 and 8 (CXCL1, CXCL3 and CXCL8) and CXC motif chemokine receptor 4 (CXCR4)] (Fig. 3B) after exposure to SOANAP-SP, which are among the other powerful neutrophil chemoattractants and are involved in the pathogenesis of airway inflammation, asthma, cancer and angiogenesis [38,39,40]. Similarly, EA.hy926 cells demonstrated strong enrichment of genes within the categories, e.g., response to stimuli, inflammatory responses, blood vessel development and cell population proliferation (Fig. 3C and Excel Sheet S2). Consistent with the findings in A549 cells and consistently with the observation of secondary genotoxicity in EA.hy926 cells by comet assay in a previous study [21], we found significantly upregulated oxidative stress and ER stress-related genes, such as activating transcription factor-4 (ATF4) (Fig. 3D). ATF4 serves as the master transcriptional regulator of the cellular response to ER stress or amino acid starvation, controls genes involved in metabolism and protection from oxidative stress [41] and apoptosis or senescence [42] and can be linked to angiogenesis [43] and metastasis [44]. Moreover, the activation of ErbB receptor feedback inhibitor 1 (ERRFI1), the FosB proto-oncogene (FOSB), and nuclear receptor subfamily 4 group a member 1 (NR4A1), has been associated with immediate early genes, especially after exposure to SOANAP-SP (Fig. 3D). Early genes exhibit rapid transcription in response to acute stress, such as hypoxia or hyperglycemia [45, 46], or proliferation-inducing signals possibly accompanied by genomic instability and activated DNA damage responses [47]. In addition to the activation of downstream genes related to early stress responses, genes involved in the inflammatory response are also often targeted. Here, we detected significant enrichment of adrenomedullin (ADM), B-cell translocation gene 1 (BTG1), CXCL8 and interleukin 6 (IL6) after exposure to SOANAP-SP and, to a lesser extent, to SOAβPIN-SP (Fig. 3D). ADM is activated in the presence of proinflammatory cytokines (e.g., IFNγ, IL1β, and IL6), that contributes to endothelial and epithelial barrier functions [48], and impacts the blood and lymphatic vasculature [49]. In endothelial cells, overexpression of BTG1 promotes cell migration and tube formation [50], upregulation of CXCL8 and IL6 can be associated with cellular dysfunctions [51], and elevated serum levels serve as markers for adverse prognosis [52, 53]. Several studies have shown a correlation between endothelial inflammatory activation and endothelial dysfunction and the development of atherosclerosis or hypertension [54, 55].

Fig. 3
figure 3

Induction of stress- and inflammatory-related genes in A549 and EA.hy926 cells. (A) Heatmap comparing gene expression values among different aerosol exposures (SP, SOAβPIN-SP and SOANAP-SP) compared to those of the clean air (CA) control originating from GO overrepresentative analysis of A549 cells. (B) Boxplots of the genes depicted from the heatmap in Fig. 3A showing the normalized counts (TMM) in A549 cells after exposure to CA, SP, SOAβPIN-SP or SOANAP-SP. (C – D) Same as in (A and B) but based on the analysis of EA.hy926 cells. (E and F) Dot plots showing the level of expression (fold change) of genes activated (filled dots) or suppressed (empty dots) by exposure to SP, SOAβPIN-SP or SOANAP-SP compared to CA exposure in the “mucin-related genes” and “innate immunity” categories for A549 cells I and in the “atherosclerosis” and “endothelial cell biology” categoriesfor EA.hy926 cells (F). * adjusted p-value < 0.05, ** adjusted p-value < 0.01 and *** adjusted p-value < 0.001

Full size image

Intriguingly, by examining the most up- or downregulated genes (Table S3), we detected high expression of mucin-related genes, such as fc gamma binding protein (FCGBP), mucin 1 (MUC1) and mucin 5AC (MUC5AC) (Fig. 3E), which have been previously associated with stress-related airway remodeling processes, in the epithelial cells after SOANAP-SP exposure [56, 57]. Moreover, we found differential expression of more mucin-related genes, such as mucin 13 (MUC13), mucin 5B (MUC5B), solute carrier family 26 (SLC26A9), SLIP, SAM pointed domain containing ETS transcription factor (SPDEF) and vesicle-associated membrane protein 9 (VAMP9). In addition, SP exposure also led to the induction of innate immune-specific genes in A549 cells; however, this effect was lower than that of SOAβPIN-SP and SOANAP-SP exposure (Fig. 3E). Collectively, these findings possibly point toward induced stress-related airway remodeling in AT2-secretory like A549 cells to more MUC5A+ goblet-secretory-like cells and an early type I immune response after exposure to SOANAP-SP. In 2005, Kunzli, Jerrett [58] published the first epidemiological study showing a correlation between atherosclerosis and ambient air pollution (PM2.5). In our model system, endothelial cells also responded by upregulating several atherosclerosis marker genes after all aerosol exposures (Fig. 3F). Marked upregulation of the genes baculoviral IAP repeat containing 3 (BIRC3), epidermal growth factor receptor (EGFR3), interleukin 1 alpha (IL1A) and serpin family b member 2 (SERPINB2) was detected after exposure to SOANAP-SP (Fig. 3F). When evaluating more known endothelial cell biology genes, we further found altered genes involved in angiogenesis (e.g., fms-related receptor tyrosine kinase 1; FLT1), vasodilation and constriction (e.g., endothelin 1; EDN1), coagulation (e.g., plasminogen activator; PLAT, tissue factor pathway inhibitor; TFPI) and platelet activation (e.g., interleukin 11; IL11, interleukin 6; IL6) (Fig. 3F). These findings revealed a strong increase in proatherosclerotic, procoagulation and platelet-activated gene expression. Taken together, our results suggested early activation of stress-related and pathogenic-specific genes in the direct exposed A549 epithelial cells, but also in the beneath laying EA.hy926 endothelial cells, highlighting the importance of using multicellular coculture systems for exposure studies.

Circulating cytokine validated the observed toxicogenomic responses

Transcriptional changes were validated by analyzing the conditioned cell culture media for cytokines. These cytokines were chosen based on the RNAseq results and subsequently correlated with PM-induced cell responses [20, 59]. The observed cytokine secretion and transcription profiles were concordant with cytokines IL-1β, IL6, IL8 and IL12, which were found among the top upregulated genes in A549 and EA.hy926 cells after exposure to SOANAP-SP. These cytokines can induce the activation of several observed upregulated genes (CXCR4 and ADM), as described above (Fig. 4, Tables S5 and S8). Moreover, significant differences in cytokine secretion were detected for IL12(p70) and IL23 after SOAβPIN-SP exposure and for IFNγ, IL12(p70), IL23 and CXCL11 after SOANAP-SP exposure. No significant change was observed for TNFα (Fig. 4). It is suggested that the secretion of IL1β, IL6 and IL8 by epithelial cells is coupled to the organic compounds in the particle phase, as previously shown for diesel engine exhaust PM [60] and ambient PM2.5 [61]. This possibly explains the minor secretion of cytokines by epithelial cells exposed to uncoated pure SP. Additionally, the activation of the aryl hydrocarbon receptor (AHR) by polycyclic aromatic hydrocarbons (PAHs), such as naphthalene or its ring-retaining photooxidation products, was shown to be responsible for inflammatory responses via the upregulation of IL6 through nuclear factor kappa-B NFκB signaling [62] and the induction of lung cancer in mice [63]. Here, we observed significantly increased expression of the AHR and NFκB2 genes after exposure to SOANAP-SP in A549 cells (AhR: log2FC: 0.93, p < 0.001; NFkB2: log2FC: 0.51, p = 0.012); however, at our tested time points we were not able to detect any upregulation of the cytochrome P450 genes, which are known to play an important role in the metabolism of early-generation aging products of naphthalene, e.g., 1-naphthol and 1,2-dihydroxynaphthalene [64]. Moreover, we detected enrichment of cytokines from the IL12 family (IL12 and IL23) and CXCL11 in the cell culture media, especially after exposure to SOANAP-SP, which plays a critical role in the development of T-cell-mediated immunity [65], and elevated levels of these cytokines appeared in patients with asthma, acute respiratory diseases and autoimmunity [66,67,68].

Fig. 4
figure 4

Circulating cytokine validated the observed toxicogenomic responses. Cytokine validation of IFN-γ, IL1-β, IL6, IL8, IL12(p70), IL23, I-TAC/CXCL11 and TNFα in the cell culture media of the coculture system after exposure to CA, SP, SOAβPIN-SP and SOANAP-SP. The data are presented as the fold change (compared to the CA control) ± SEM from three independent experiments, and the significance of the differences are outlined

Full size image

Integrative analysis of circulating metabolites and gene expression revealed important role of non-directly exposed endothelial cells

In recent years, invaluable information has been obtained from metabolomics to mechanistically understand the impact of PM2.5 exposure. Thus, we next investigated the metabolic signature defined by exposure to SP, SOAβPIN-SP or SOANAP-SP in the collected sample media. We identified a total of 312 metabolites, with 51 differentially expressed metabolites after SP exposure (Fig. 5A and Excel Sheet S3), 32 after SOAβPIN-SP exposure (Fig. 5B and Excel Sheet S3) and 32 after SOANAP-SP exposure (Fig. 5C and Excel Sheet S3), with respect to the CA control. Here, we observed distinct metabolite profiles, with only five metabolites commonly found to a greater or lesser extent in the exposure media after exposure to both SOA types and SP and with metabolites allocated to cellular and potential aerosol origins (Fig. 5D). In particular, the accumulation of chemical compounds of non-endogenous origin, such as i.e., 1,2,3-cyclohexanetriol after the exposure to SOAβPIN-SP or 3,4,5-trihydroxybenzyl methyl ether and 3-hydroxy-2-naphthoic acid after the exposure to SOANAP-SP can possibly be correlated with oxidation compounds of the respective aerosols. Interestingly, in the toxicogenomic database [69], 3-hydroxy-2-naphthoic acid has been associated with an interaction of G-protein-coupled receptor 35 (GPR35), whose increased activity has been linked to various pathologies in the inflammatory and cardiovascular systems [70] or to a poor prognosis in non-small lung cancer [71].

Fig. 5
figure 5

Integrative analysis of circulating metabolites and gene expression revealed important role of non-directly exposed endothelial cells. (A-C) Volcano plots showing the relationships between differentially abundant metabolite expression (p-value ≤ 0.1 and log2FC ≥ 0.5 and log2FC ≤ -0.5) in the cell culture media after exposure to SP (A), SOAβPIN-SP (B) or SOANAP-SP (C) and the CA control. (D) Venn diagram of the unique and overlapping metabolites identified as up- or downregulated (p-value ≤ 0.1, log2FC ≥ 0.5 and log2FC ≤ -0.5) in the cell culture media after exposure to SP, SOAβPIN-SP or SOANAP-SP compared to the CA control. GC/MS analysis was performed on n = 3 independent experiments

Full size image

Like a study by Breitner, Schneider [72] investigating the abundance of metabolites in blood plasma after short-term exposure to PM2.5 in a cardiac catheterization cohort, we detected a decrease in arginine and glycine. In contrast, ornithine was significantly enriched after exposure to SOANAP-SP. This finding was consistent with the results for glycine degradation (SOANAP-SP, p < 0.001; SOAβPIN-SP, p = 0.048 and SP, p = 0.013), the urea cycle (SOANAP-SP, p = 0.006 and SOAβPIN-SP, p = 0.01) and arginine degradation (SOANAP-SP, p < 0.006; and SP, p = 0.015) under the 5 most altered canonical pathways identified by IPA analysis (Table 2). An increased consumption of glycine by cells was observed to be a sign of glutathione (GSH) demand induced through oxidative stress [73] or angiogenesis through vascular endothelial growth factor (VEGF) signaling [74]. Moreover, a reduced serum level of glycine in patients with COPD has been correlated with airflow obstruction via the development of emphysema [75].

Table 2 Canonical pathway enrichment analysis of the metabolome data
Full size table

In addition, the reduced availability of arginine and increased abundance of ornithine in the cell culture media suggested enhanced activity of the enzyme arginase within the urea cycle. This has already been observed in conditions of acute and chronic stress [76] and in a study by Liang, Ladva [77] in which asthmatic persons were exposed to traffic-related pollutants. Interestingly, reduced secretion of 4-aminobutyric acid (GABA) was detected after all aerosol exposures, and increased uptake of GABA in airway epithelial progenitor cells induced differentiation into goblet cells [78], supporting our RNASeq analysis. As anticipated, we detected alterations in gene and metabolite expression; however, the degree to which transcriptomic and metabolic changes occur has remained elusive. Therefore, to integrate the metabolome data with the gene expression data, we performed an IPA comparison analysis with a pathway-based approach [79]. We found that eight common pathways were significantly altered in the gene expression of A549 and/or EA.hy926 cells and in the metabolome of the sampled media after all aerosol exposures (Table 3). The iNOS signaling pathway was altered in EA.hy926 cells after all exposures at the transcript level (SOANAP-SP: p = 0.013; SOAβPIN-SP: p = 0.007 and SP: p = 0.019) and at the metabolome level after exposure to SOANAP-SP (p = 0.037) as well as SOAβPIN-SP (p = 0.048). The PDGF signaling pathway was significantly regulated in both cell lines following exposure to SOANAP-SP (A549: p = 0.029; EA.hy926: p = 0.004) and in only EA.hy926 cells after exposure to SOAβPIN-SP (p = 0.039) or SP (p = 0.027). However, at the metabolome level, the PDGF pathway was regulated only after exposure to SOAβPIN-SP (p = 0.048). These pathways have previously been shown to be involved in vascular endothelial cell migration [80] and the development of fibrotic diseases [81]. Notably, after exposure to SOANAP-SP, we found an alteration in phosphatidylinositol 3-kinase (PI3K)/protein kinase B (AKT) signaling in the gene expression of EA.hy926 cells in response to SOANAP-SP (p = 0.016) and in the metabolome data after exposure to both SOANAP-SP (p = 0.042) and SOAβPIN-SP (p = 0.048). This indicated the involvement of DNA damage and oxidative stress [82] as well as the typical activation of endothelial cells for effective angiogenesis [83], as demonstrated at the functional level by the observed secondary genotoxicity and induced angiogenesis [21]. Taken together, these findings indicate that exposure to SP, SOAβPIN-SP and SOANAP-SP is associated with significant metabolomic changes that are distinct for single metabolites, leading to altered phenotypes and highlighting the pivotal role of not directly exposing EA.hy926 cells.

Table 3 Canonical pathway enrichment analysis of the transcriptome and metabolome data via IPA software
Full size table

Limitations of the study

Despite the importance of these findings, there are limitations to our study that should be considered. First, by choosing A549 cells as our epithelial lung cell line we are aware of several possible drawbacks, including a low expression of tight junctions and adherence proteins, inability to form a fully polarized epithelium, as well as, being of carcinogenic origin [84]. Therefore, additional studies are needed to observe the effects of SOA exposure on other epithelial cell lines or even primary human epithelial cells. However, due to the complexity of our experimental setup, we found with A549 cells a robust and metabolic competent cell line that is suitable for large and repetitive studies. Second, the focus on only one aerosol dilution (1:3) and one time point (4 h exposure) complicates the risk assessment of the aerosols. Future studies comparing short- and long-term exposures and different aerosol doses should help characterize meaningful adverse effects.

Conclusion

In this study, we have revealed the importance of the chemical identities of SOA in inducing different toxicogenomic and metabolic effects in an epithelial-endothelial coculture system (Fig. 6). By conducting exposure studies with physically similar but chemically distinct model SOA, we identified aerosol-specific effects depending on chemical composition, especially in directly exposed epithelial cells (A549). The exposure to SOA derived from an anthropogenic aromatic precursor (naphthalene) resulted in the upregulation of genes in A549 cells related to oxidative stress (e.g., NOX1 and PTGS2) and to an early type I immune response (e.g., CXCL8 and CXCR4) in A549. In addition, transcriptomic insights revealed a stress-related airway remodeling of AT2-secretory like cells to more MUC5A+ goblet-secretory-like cells. Milder effects were observed in A549 cells exposed to a biogenic, aliphatic SOA precursor (β-pinene) and fresh SP. Remarkably, we detected early activation of stress- (e.g., FOSB and NR4A1) and CVD-related genes (e.g., ADM and BTG1) in the underlying endothelial cells (EA.hy926). This effect was significant after exposure to SOANAP-SP and minor after exposure to SOAβPIN-SP or SP and in line with the previously detected DNA breaks and increased angiogenic potential. Cytokine and metabolic analyses confirmed the induction of a systemic proinflammatory state, and the integration analysis highlighted the key role of cells that are not directly exposed in the overall molecular toxicological response. We observed alterations especially in signaling pathways that are known to play key roles in the onset of CVD (e.g., iNOS, PDGF and PI3/AKT signaling). Our findings thus emphasize the importance of enabling cellular crosstalk in exposure assessment studies combined with systems biology approaches to better determine potential human health risks beyond the lung due to exposure to SOA.

Fig. 6
figure 6

Proposed pathomechanisms triggered by aerosol exposure. Aerosols confer most of their effects via damage to the lung (inflammation, oxidative stress, genotoxicity, and airway remodeling toward MUC5AC+ goblet-secretory like cells). This effect seemed to be dependent on the chemical composition of the specific aerosols (organic content, PAHs, functional groups, oxidation state, aromaticity) and was greatest after exposure to SOANAP-SP, followed by SOAβPIN-SP and SP. The primary target organ damaged by aerosols converges at the cardiovascular level by inducing dysfunctions in endothelial cells (secondary genotoxicity), inflammation, metabolic reprogramming, and angiogenesis. These effects seemed to be more independent of the aerosol type and systemic treatment, resulting in the activation of coagulation, atherosclerosis and/or hypertension via iNOS, PDGF and/or PI3/AKT signaling, respectively

Full size image

Materials and methods

Aerosol generation and PM (SOA) characterization

A detailed description of the aerosol generation and characterization has been given previously [20, 21, 85]. Briefly, secondary organic aerosols (SOA) of either naphthalene (Sigma-Aldrich, 147141-25G, 99%) or β-pinene (Sigma-Aldrich, 402753-10G, ≥ 99%) were produced by mixing their pure vapor with soot particles (SP) from a Combustion Aerosol Standard generator (CAST, model 5201 C; Jing Ltd., Zollikofen, CH). The mixtures were subsequently processed in an oxidation flow reactor (potential aerosol mass reactor, PAM, Aerodyne Research Inc., Billerica, MA, US) [86, 87] for the simulation of atmospheric photooxidation (aging) dominated by OH radicals. This process resulted in two SOA types, SOAβPIN-SP and SOANAP-SP. Pure SP were fed into the PAM and used as condensation nuclei in the SOA experiments, without precursor addition or aging, as a reference. Methanolic extracts of SOA-SP were examined by ultrahigh-resolution Fourier-transform ion cyclotron resonance mass spectrometry (SolariX, 7 T FT-ICR MS, Bruker Daltoniks, Bremen, GE) with electrospray ionization in positive and negative mode with detailed parameters discussed previously elsewhere [20] to determine the elemental composition. The elemental composition was determined by the exact mass identified at a signal-to-noise ratio (S/N) of 9. The following settings were applied for elemental composition assignment: CcHhOoNax; for positive mode ESI: 2 ≤ c ≤ 100, 2 ≤ h ≤ 100, o ≤ 20, and x ≤ 1; and for negative mode ESI: 2 ≤ c ≤ 50, 2 ≤ h ≤ 100, o ≤ 16, and x ≤ 0 with a maximum error of 2 ppm. To describe the degree of unsaturation, the concept of double bond equivalents (DBE) is used, considering that the octet rule (except for hydrogen) is obeyed and that unsaturation is caused by covalent carbon – carbon bonds.

$$DBE = 1 + C - 0.5H + 0.5N$$
(1)

The MCR is calculated from the relation of DBE to the number of oxygen atoms (#O) via the sum formula.

$$MCR = {{DBE} \over {\# O}}$$
(2)

Gas phase characterization

VOC in the gas phase were analysed and quantified via proton-transfer-reaction (PTR) high-resolution time-of-flight mass spectrometry (PTR-TOFMS; IONICON Analytik GmbH, Innsbruck, AT) via measurements of the filtered gas phase at the outlet of the PAM at a 1:10 dilution using previously described settings and procedure in [88].

Cell culture and ALI exposure

Detailed descriptions of the cell culture and the ALI exposure systems used are given in Offer, Hartner [21]. Briefly, A549 human alveolar epithelial cells (ATCC®, CCL-185™) and EA.hy926 human endothelial cells were routinely cultured in high-glucose Gibco Dulbecco’s Modified Eagle Medium: Nutrient Mixture F-12 (DMEM/F-12) (ThermoFisher Scientific, 31331-028) supplemented with 5% (v/v) fetal bovine serum (FBS) (ThermoFisher Scientific, 10500-064), 100 U/mL penicillin, and 100 µg/mL streptomycin (P/S; Sigma-Aldrich, P4333) in a humidified incubator at 37 °C and 5% CO2. For the coculture exposure experiments, A549 cells were seeded on transferrable 24 mm Transwell® inserts with a polyester membrane (0.4 μm pore-size, Type #3450, Corning, NY, US) 96 h before the exposure experiments at a density of 1.8 × 105 cells/mL per insert (3.8 × 104 cells/cm2 growth area) with 1.5 mL of cell culture medium provided at the basal compartment and 1 mL at the apical side of the Transwell® plate. Forty-eight hours after the initial cell seeding, the culture medium on the apical side was removed to establish ALI conditions, and fresh cell culture medium (1.5 mL) was added to the basal compartment. After an additional 24 h later, the inserts were inverted, 1 × 105 EA.hy926 cells per insert (0.21 × 104 cells/cm2 growth area) were seeded in 750 µl of medium, and after 1 h, the insert was returned, after which fresh medium (1.5 mL) was added to the basolateral compartment of the Transwell® plate. On day 5, the day of the exposure experiments, the inserts were placed in the exposure modules of the ALI exposure system (Vitrocell® Automated Exposure Station Standard Version, Vitrocell Systems, Waldkirch, GE) with 1.8 mL of serum-free DMEM/F12 medium supplemented with 1% P/S and 15 mM N-2-hydroxyethylpiperazine-N-2-ethane sulfonic acid (HEPES) buffer solution (ThermoFisher Scientific, 15630-056) in the basolateral compartment. The cells were then exposed for 4 h to conditioned (85% r.h. 37 °C) and 1:3 diluted aerosol (PM in the gas phase), named SP, SOAβPIN-SP and SOANAP-SP with a 100 mL/min flowrate over every position in the ALI exposure system, as described in Mülhopt, Dilger [89]. Due to trumpet-shaped flow-guiding elements in every position of the ALI exposure system, the aerosols are depositing onto cells by diffusion and sedimentation [90]. In addition to the aerosol exposures, each system had a separate clean air (CA; purified compressed laboratory air) exposure sector serving as a control. Aerosol exposure to SP was conducted in n = 3 independent experiments, to SOAβPIN-SP in n = 4 independent experiments and to SOANAP-SP in n = 4 independent experiments. To avoid possible sources of inherent variability in the deposited aerosols due to the ALI exposure system, the same insert positions were always used per module. After exposure, the effects of the aerosols and CA on cells were examined via several assays (RNASeq analysis), and the exposure medium was collected (from the sample media) on ice and frozen at -80 °C for later analyses (cytokine and metabolome analysis).

RNA extraction, library construction and sequencing

Transwell® inserts were placed in a 6-well plate (ThermoScientific, 140675) containing RNAprotect Cell Reagent (QIAGEN, Hilden, 76526) immediately after exposure. Then, the A549 cells on the apical side and the EA.hy926 cells on the basolateral side were carefully scraped with a cell scraper (Sigma, SIAl0010) and separately stored at -20 °C until RNA extraction. Total RNA was extracted from the cells using an RNA Plus Mini Kit (QIAGEN, Hilden, 74136) according to the manufacturer’s instructions. The quality of the extracted RNA was assessed with a Nanodrop and TapeStation (Agilent Technologies, CA, US) and only replicates with high RNA integrity (RIN > 9.8) were processed for RNA sequencing (RNASeq) at the Crown Genomics Institute of the Nancy and Stephen Grand Israel National Center for Personalized Medicine, Weizmann Institute of Science. Total RNA (500 ng for each sample) was processed using the inhouse poly A-based RNAseq protocol (INCPM mRNA Seq). SR reads were sequenced on 1 lane(s) of an Illumina NovaSeq Sp (100 cycles) with an output of ~ 17 million reads per sample. Libraries were evaluated by Qubit (Thermo FisherScientific Inc., MA, US) and TapeStation (Agilent Technologies, CA, US) instruments. A detailed description of the RNA extraction, library construction and sequencing procedure was previously described in [20]. Shortly, P Poly-A/T stretches and Illumina adapters were trimmed from the reads using cutadapt [91], resulting reads shorter than 30 bp were discarded. Reads were mapped to the H. sapiens reference genome GRCh38 using STAR [92], supplied with gene annotations downloaded from Ensembl (and with EndToEnd option and outFilterMismatchNoverLmax was set to 0.04). The percentage of the reads aligned uniquely to the genome was 97%. Reads with the same UMI were removed using the PICARD MarkDuplicate tool using the BARCODE_TAG parameter. Expression levels for each gene were quantified using htseq-count (version 0.11.2) [93], using the gtf above. Only uniquely mapped reads were used to determine the number of reads falling into each gene (intersection-strict mode). The sequencing data have been deposited in the NCBI´s Gene Expression Omnibus [94] and are accessible through the GEO Series accession number GSE226350 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE226350).

Differential gene expression analysis

All the statistical analyses were performed in the R programming language (version 4.1.2). For the A549 and EA.hy926 samples, genes with a count per million (CPM) reads ≥ 30 in at least 3 samples (the number of the fewest replicates of a given condition) were retained. Gene expression levels were normalized by the trimmed mean of M-values (TMM) method using the calcNormFactors function of the EdgeR package in R [95]. Moreover, samples were corrected based on the processed sequencing batch, and differential expression analyses were performed using the limma-voom package in R [96]. Significantly differentially expressed genes (DEGs) of all aerosol types compared to those of the clean air (CA) control were selected with a controlled FALSE positive rate (BH method) of 5% (FDR ≤ 0.05). Upregulated genes were selected at a minimum log2-fold change of 0.5, and downregulated genes were selected at a minimum log2-fold change of -0.5. Principal component analyses (PCA) were performed using batch effect-corrected normalized values of the genes expressed by A549 and EA.hy926 cells. Heatmaps were drawn on the normalized expression matrix using the heatmap.2 function from the gplots package in R with Euclidean distance and complete linkage for hierarchical clustering [97]. To visualize the statistical significance versus the log2-fold change, volcano plots were generated by using the EnhancedVolcano plot function in R [98]. Venn diagrams were constructed using the vennDiagram function based on the overlapping DEGs (FDR ≤ 0.05, log2-fold change ≤ -0.5/ ≥ 0.5).

Gene Ontology (GO) overrepresentation analysis and GO gene set enrichment analysis

Gene Ontology (GO) overrepresentation analysis of the biological process (BP) and molecular function (MF) categories was performed using the enrichGO function from the clusterProfile package in R [99] based on DEGs (FDR ≤ 0.05, log2-fold change ≥ 0.5) of A549 and EA.hy926 cells, respectively. Important GO terms were selected based on an FDR ≤ 0.05, and the top ten results were plotted by bar plots showing the –log10(p-value) of the analysis. Gene set enrichment analysis (GSEA) was performed by the gseaGO function from the clusterProfile package in R, which computes the normalized enrichment score (NES) for each cell line and condition. For our analysis, only pathways with a NES above 2 or below − 2 were selected and further analysed by categorization into their respective GO ‘parent’ terms [100, 101], which was followed by clustering of the significant DEGs (FDR ≤ 0.05, log2-fold change ≥ 0.5) in a heatmap generated with the heatmap function of the complexHeatmap package in R [102].

Plotting specific gene expression across aerosol exposure

Using heatmaps, we identified genes within the functional GO gene sets (“response to stimuli” and “inflammatory response”) and created boxplots on the basis of the normalized counts (TMM). Moreover, categories were generated with the names “mucin-related gene” and “innate immunity”, in which we plotted the fold change in DEGs in A549 cells (FDR ≤ 0.05, log2-fold change ≥ 0.5 or ≤ -0.5) that have been previously published within those contexts [56]. For EA.hy926 cells, we focused on genes (FDR ≤ 0.05, log2-fold change ≥ 0.5 or ≤ -0.5) that have been used as markers for “atherosclerosis” or “endothelial cell biology”, including “angiogenesis, vasodilation and -constriction”, coagulation, and platelet activation, by Qiagen RT2 Profiler PCR Arrays. The genes were plotted that were detected in our dataset were determined to be differentially expressed (FDR ≤ 0.05, log2-fold change ≥ 0.5 or ≤ -0.5).

Cytokine detection

Several cytokines, such as CXCL11/I-TAC, IFNγ, IL-12p70, IL-1β, IL-23, IL-6, IL-8 and TNF-α, were measured in the frozen collected sample media by using the MILLIPLEX MAP Human High Sensitivity T-Cell Panel-Immunology Multiplex Assay. Cytokines were measured with a Milliplex Magpix instrument (Luminex, Merck KGaA, Darmstadt, GE). Analysis was performed according to the manufacturer’s instructions. Twenty-five microliters of the cell medium was used for the analysis. The calibration curves for each cytokine were calculated within the assay, and the detection ranges for each cytokine were different: CXCL11/I-TAC: 1.46–6000 pg/mL, IFNγ: 0.61–2500 pg/mL, IL-12p70: 0.49–2000 pg/mL, IL-1β: 0.49–2000 pg/mL, IL-23: 7.23–32,500 pg/mL, IL-6: 0.18–750 pg/mL, IL-8: 0.31–1250 pg/mL and TNF-α 0.43–1750 pg/mL. Cytokine concentrations were determined by fluorescence intensity, and fluorescence data were analyzed with Millipore Milliplex Analyst version 3.4 according to the manufacturer’s recommendations. Statistical analyses were performed with the calculated concentrations of each sample.

Metabolomic sample preparation and untargeted GC/MS of sample media

Metabolomes were detected from the frozen collected sample media from three independent experiments per aerosol type. The derivatization method and GC/MS setup used were modified from the methods described previously Pink, Verma [103]. Then, 50 µL of the thawed collected sample media was added to 450 µL of a cold methanol/water mixture (88.9%, 11.1%) on ice. After vortexing, the sample was centrifuged for 5 min (18000 rcf, 4 °C). One hundred microliters were then transferred to a 1.5 mL GC vial and dried in a Speedvac. To obtain water-free samples, 100 µL of dichloromethane was added, and the samples were subsequently dried in a Speedvac. The extract was solubilized in 50 µL of a methoxyamine hydrochloride solution (20 mg/mL in pyridine) by sonication and incubated at 50 °C for 60 min before being placed on an orbital shaker at room temperature overnight. Afterwards, 25 µL of MSTFA (N-methyl-N-(trimethylsilyl) trifluoroacetamide) supplemented with 1% trimethylchlorosilane (TMCS) was added, and the mixture was further incubated for 60 min at 40 °C. For analysis, the samples were transferred to 100 µL inserts and frozen at − 80 °C until analysis. The derivatized metabolites were separated and semi-quantified by GC/MS analysis using an HP 8890/5977B GC/MS (Agilent Technologies, Waldbronn, Germany) instrument equipped with a 30 m × 320 μm (i.d.) Optima 5 column coated with a 5% phenyl/ 95% methylpolysolixane cross-linked stationary phase (0.25 μm film thickness; MACHEREY-NAGEL, Düren, Germany). Helium was used as the carrier gas at a flow rate of 1.5 mL/min. Two microliters of each sample were injected (injector temperature: 250 °C) in splitless mode with a solvent cutoff time of 6 min. The oven temperature was maintained at 80 °C for 10 min and then linearly increased at a rate of 5 °C/min up to 330 °C. The MS instrument was operated in electron impact ionization mode at 70 eV with the quadrupole temperature set at 150 °C and the source temperature set at 230 °C. Full scans were acquired by repetitive scanning over the mass range from 60 to 550 Da at a scan rate of 500 ms/scan.

GC/MS spectrum and data analysis of sample media

A detailed description of the conversion of the GC/MS spectra, the database and the peak identification was previously described in Pink, Verma [104]. Briefly, the GC/MS spectra were evaluated by running the R package eRah [105], and peak identification was performed with the GOLM metabolome database (version 2011-11-21) [106] supplemented with data from NIST14 (NIST 14 Mass Spec Library and Search Programs – User Manual (sisweb.com)). Metabolome profiling was conducted in three independent experiments (n = 3). The resulting metabolites were annotated with the metabolite ID conversion tool of MetaboAnalyst 5.0 software and subjected to PCA and volcano plotting [107]. Moreover, the list of altered metabolites (log2FC ≤ -0.5/ ≥ 0.5; p ≤ 0.05) was uploaded alone or together with the RNASeq data (log2FC ≤ -0.5/ ≥ 0.5; p ≤ 0.05) in the Ingenuity Pathway Analysis (IPA, Qiagen) software for integrated analysis of canonical pathways.

Data availability

The sequencing data have been deposited in the NCBI´s Gene Expression Omnibus and are accessible through the GEO Series accession number GSE226350. Metabolome data is provided within the manuscript or supplementary information files.

References

  1. Manisalidis I, Stavropoulou E, Stavropoulos A, Bezirtzoglou E. Environmental and Health impacts of Air Pollution: a review. Front Public Health. 2020;8:14.

    Article  PubMed  PubMed Central  Google Scholar 

  2. Lelieveld J, Pozzer A, Poschl U, Fnais M, Haines A, Munzel T. Loss of life expectancy from air pollution compared to other risk factors: a worldwide perspective. Cardiovasc Res. 2020;116(11):1910–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  3. Pozzer A, Anenberg SC, Dey S, Haines A, Lelieveld J, Chowdhury S. Mortality attributable to Ambient Air Pollution: a review of global estimates. GeoHealth. 2023;7(1):e2022GH000711.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  4. Al-Kindi SG, Brook RD, Biswal S, Rajagopalan S. Environmental determinants of cardiovascular disease: lessons learned from air pollution. Nat Reviews Cardiol. 2020;17(10):656–72.

    Article  Google Scholar 

  5. Brook RD, Rajagopalan S, Pope CA 3rd, Brook JR, Bhatnagar A, Diez-Roux AV, et al. Particulate matter air pollution and cardiovascular disease: an update to the scientific statement from the American Heart Association. Circulation. 2010;121(21):2331–78.

  6. Pope CA 3rd, Bhatnagar A, McCracken JP, Abplanalp W, Conklin DJ, O’Toole T. Exposure to fine Particulate Air Pollution is Associated with Endothelial Injury and systemic inflammation. Circ Res. 2016;119(11):1204–14.

  7. Davel AP, Lemos M, Pastro LM, Pedro SC, de André PA, Hebeda C, et al. Endothelial dysfunction in the pulmonary artery induced by concentrated fine particulate matter exposure is associated with local but not systemic inflammation. Toxicology. 2012;295(1–3):39–46.

    Article  CAS  PubMed  Google Scholar 

  8. Tamagawa E, Bai N, Morimoto K, Gray C, Mui T, Yatera K, et al. Particulate matter exposure induces persistent lung inflammation and endothelial dysfunction. Am J Physiol Lung Cell Mol Physiol. 2008;295(1):L79–85.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  9. Fine PM, Sioutas C, Solomon PA. Secondary particulate matter in the United States: insights from the Particulate Matter supersites Program and related studies. J Air Waste Manag Assoc. 2008;58(2):234–53.

    Article  CAS  PubMed  Google Scholar 

  10. Zhang R, Khalizov A, Wang L, Hu M, Xu W. Nucleation and growth of nanoparticles in the atmosphere. Chem Rev. 2012;112(3):1957–2011.

    Article  CAS  PubMed  Google Scholar 

  11. Donahue NM, Robinson AL, Stanier CO, Pandis SN. Coupled partitioning, dilution, and chemical aging of semivolatile organics. Environ Sci Technol. 2006;40(8):2635–43.

    Article  CAS  PubMed  Google Scholar 

  12. Schnitzler EG, Dutt A, Charbonneau AM, Olfert JS, Jager W. Soot Aggregate Restructuring due to Coatings of secondary Organic Aerosol Derived from aromatic precursors. Environ Sci Technol. 2014;48(24):14309–16.

    Article  CAS  PubMed  Google Scholar 

  13. Hallquist M, Wenger JC, Baltensperger U, Rudich Y, Simpson D, Claeys M, et al. The formation, properties and impact of secondary organic aerosol: current and emerging issues. Atmos Chem Phys. 2009;9(14):5155–236.

    Article  CAS  Google Scholar 

  14. Rudich Y, Donahue NM, Mentel TF. Aging of organic aerosol: bridging the gap between laboratory and field studies. Annu Rev Phys Chem. 2007;58:321–52.

    Article  CAS  PubMed  Google Scholar 

  15. Lei R, Wei Z, Chen M, Meng H, Wu Y, Ge X. Aging effects on the toxicity alteration of different types of Organic aerosols: a review. Curr Pollution Rep. 2023;9(3):590–601.

    Article  CAS  Google Scholar 

  16. Nault BA, Jo DS, McDonald BC, Campuzano-Jost P, Day DA, Hu WW, et al. Secondary organic aerosols from anthropogenic volatile organic compounds contribute substantially to air pollution mortality. Atmos Chem Phys. 2021;21(14):11201–24.

    Article  CAS  Google Scholar 

  17. Pye HOT, Ward-Caviness CK, Murphy BN, Appel KW, Seltzer KM. Secondary organic aerosol association with cardiorespiratory disease mortality in the United States. Nat Commun. 2021;12(1):7215.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Han J, Wang S, Yeung K, Yang D, Gu W, Ma Z, et al. Proteome-wide effects of naphthalene-derived secondary organic aerosol in BEAS-2B cells are caused by short-lived unsaturated carbonyls. Proc Natl Acad Sci U S A. 2020;117(41):25386–95.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Park M, Joo HS, Lee K, Jang M, Kim SD, Kim I, et al. Differential toxicities of fine particulate matters from various sources. Sci Rep. 2018;8(1):17007.

    Article  PubMed  PubMed Central  Google Scholar 

  20. Pardo M, Offer S, Hartner E, Di Bucchianico S, Bisig C, Bauer S, et al. Exposure to naphthalene and beta-pinene-derived secondary organic aerosol induced divergent changes in transcript levels of BEAS-2B cells. Environ Int. 2022;166:107366.

    Article  CAS  PubMed  Google Scholar 

  21. Offer S, Hartner E, Bucchianico SD, Bisig C, Bauer S, Pantzke J, et al. Effect of Atmospheric Aging on Soot particle toxicity in lung cell models at the air-liquid interface: Differential Toxicological impacts of Biogenic and Anthropogenic secondary Organic aerosols (SOAs). Environ Health Perspect. 2022;130(2):027003.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  22. Shahpoury P, Zhang ZW, Arangio A, Celo V, Dabek-Zlotorzynska E, Harner T, et al. The influence of chemical composition, aerosol acidity, and metal dissolution on the oxidative potential of fine particulate matter and redox potential of the lung lining fluid. Environ Int. 2021;148:106343.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  23. Rothen-Rutishauser B, Blank F, Muhlfeld C, Gehr P. In vitro models of the human epithelial airway barrier to study the toxic potential of particulate matter. Expert Opin Drug Metab Toxicol. 2008;4(8):1075–89.

    Article  CAS  PubMed  Google Scholar 

  24. Paur HR, Cassee FR, Teeguarden J, Fissan H, Diabate S, Aufderheide M, et al. In-vitro cell exposure studies for the assessment of nanoparticle toxicity in the lung-A dialog between aerosol science and biology. J Aerosol Sci. 2011;42(10):668–92.

    Article  CAS  Google Scholar 

  25. Zhang Y, Wang K, Tong H, Huang R-J, Hoffmann T. The maximum carbonyl ratio (MCR) as a new index for the structural classification of secondary organic aerosol components. Rapid Commun Mass Spectrom. 2021;35(14):e9113.

    Article  CAS  PubMed  Google Scholar 

  26. Vereecken L, Peeters J. A theoretical study of the OH-initiated gas-phase oxidation mechanism of beta-pinene (C10H16): first generation products. Phys Chem Chem Phys. 2012;14(11):3802–15.

    Article  CAS  PubMed  Google Scholar 

  27. Bartoszewski R, Matalon S, Collawn JF. Ion channels of the lung and their role in disease pathogenesis. Am J Physiol-Lung C. 2017;313(5):L859–72.

    Article  Google Scholar 

  28. Wong BH, Mei D, Chua GL, Galam DL, Wenk MR, Torta F, et al. The lipid transporter Mfsd2a maintains pulmonary surfactant homeostasis. J Biol Chem. 2022;298(3):101709.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  29. Chen QZ, Zhou YB, Zhou LF, Fu ZD, Yang CT, Zhao L et al. TRPC6-dependent Ca2 + signaling mediates airway inflammation in response to oxidative stress via ERK pathway. Cell Death Dis. 2020;11(3).

  30. Ko EA, Kim YW, Lee D, Choi J, Kim S, Seo Y, et al. Expression of potassium channel genes predicts clinical outcome in lung cancer. Korean J Physiol Pharmacol. 2019;23(6):529–37.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Tuet WY, Chen YL, Fok S, Champion JA, Ng NL. Inflammatory responses to secondary organic aerosols (SOA) generated from biogenic and anthropogenic precursors. Atmos Chem Phys. 2017;17(18):11423–40.

    Article  CAS  Google Scholar 

  32. Wang X, Thomas B, Sachdeva R, Arterburn L, Frye L, Hatcher PG, et al. Mechanism of arylating quinone toxicity involving Michael adduct formation and induction of endoplasmic reticulum stress. Proc Natl Acad Sci U S A. 2006;103(10):3604–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  33. Gopinath P, Mahammed A, Ohayon S, Gross Z, Brik A. Understanding and predicting the potency of ROS-based enzyme inhibitors, exemplified by naphthoquinones and ubiquitin specific protease-2. Chem Sci. 2016;7(12):7079–86.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Carnesecchi S, Dunand-Sauthier I, Zanetti F, Singovski G, Deffert C, Donati Y, et al. NOX1 is responsible for cell death through STAT3 activation in hyperoxia and is associated with the pathogenesis of Acute Respiratory Distress Syndrome. Int J Clin Exp Patho. 2014;7(2):537–51.

    CAS  Google Scholar 

  35. Seta F, Rahmani M, Turner PV, Funk CD. Pulmonary oxidative stress is increased in Cyclooxygenase-2 Knockdown mice with mild Pulmonary Hypertension Induced by Monocrotaline. PLoS ONE. 2011;6(8).

  36. Fukunaga K, Kohli P, Bonnans C, Fredenburgh LE, Levy BD. Cyclooxygenase 2 plays a pivotal role in the resolution of acute lung injury. J Immunol. 2005;174(8):5033–9.

    Article  CAS  PubMed  Google Scholar 

  37. Faber SC, McNabb NA, Ariel P, Aungst ER, McCullough SD. Exposure effects beyond the epithelial barrier: Transepithelial induction of oxidative stress by Diesel Exhaust particulates in lung fibroblasts in an Organotypic Human Airway Model. Toxicol Sci. 2020;177(1):140–55.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  38. Mukaida N. Pathophysiological roles of interleukin-8/CXCL8 in pulmonary diseases. Am J Physiol Lung Cell Mol Physiol. 2003;284(4):L566–77.

    Article  CAS  PubMed  Google Scholar 

  39. Shieh JM, Tsai YJ, Tsou CJ, Wu WB. CXCL1 regulation in human pulmonary epithelial cells by tumor necrosis factor. Cell Physiol Biochem. 2014;34(4):1373–84.

    Article  CAS  PubMed  Google Scholar 

  40. Al-Alwan LA, Chang Y, Mogas A, Halayko AJ, Baglole CJ, Martin JG, et al. Differential roles of CXCL2 and CXCL3 and their receptors in regulating normal and asthmatic airway smooth muscle cell Migration. J Immunol. 2013;191(5):2731–41.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Harding HP, Zhang Y, Zeng H, Novoa I, Lu PD, Calfon M, et al. An integrated stress response regulates amino acid metabolism and resistance to oxidative stress. Mol Cell. 2003;11(3):619–33.

    Article  CAS  PubMed  Google Scholar 

  42. Hiramatsu N, Messah C, Han J, LaVail MM, Kaufman RJ, Lin JH. Translational and posttranslational regulation of XIAP by eIF2alpha and ATF4 promotes ER stress-induced cell death during the unfolded protein response. Mol Biol Cell. 2014;25(9):1411–20.

    Article  PubMed  PubMed Central  Google Scholar 

  43. Wang YG, Alam GN, Ning Y, Visioli F, Dong ZH, Nor JE, et al. The unfolded protein response induces the angiogenic switch in human tumor cells through the PERK/ATF4 pathway. Cancer Res. 2012;72(20):5396–406.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  44. Dey S, Sayers CM, Verginadis II, Lehman SL, Cheng Y, Cerniglia GJ, et al. ATF4-dependent induction of heme oxygenase 1 prevents anoikis and promotes metastasis. J Clin Invest. 2015;125(7):2592–608.

    Article  PubMed  PubMed Central  Google Scholar 

  45. Zhao M, Wang S, Zuo A, Zhang J, Wen W, Jiang W, et al. HIF-1alpha/JMJD1A signaling regulates inflammation and oxidative stress following hyperglycemia and hypoxia-induced vascular cell injury. Cell Mol Biol Lett. 2021;26(1):40.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Xu DZ, Patten RD, Force T, Kyriakis JM. Gene 33/RALT is induced by hypoxia in cardiomyocytes, where it promotes cell death by suppressing phosphatidylinositol 3-kinase and extracellular signal-regulated kinase survival signaling. Mol Cell Biol. 2006;26(13):5043–54.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  47. Guo H, Golczer G, Wittner BS, Langenbucher A, Zachariah M, Dubash TD, et al. NR4A1 regulates expression of immediate early genes, suppressing replication stress in cancer. Mol Cell. 2021;81(19):4041–58. e15.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  48. Garcia-Ponce A, Chanez Paredes S, Castro Ochoa KF, Schnoor M. Regulation of endothelial and epithelial barrier functions by peptide hormones of the adrenomedullin family. Tissue Barriers. 2016;4(4):e1228439.

    Article  PubMed  PubMed Central  Google Scholar 

  49. Karpinich NO, Hoopes SL, Kechele DO, Lenhart PM, Caron KM. Adrenomedullin function in vascular endothelial cells: insights from genetic mouse models. Curr Hypertens Rev. 2011;7(4):228–39.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  50. Iwai K, Hirata K, Ishida T, Takeuchi S, Hirase T, Rikitake Y, et al. An anti-proliferative gene BTG1 regulates angiogenesis in vitro. Biochem Biophys Res Commun. 2004;316(3):628–35.

    Article  CAS  PubMed  Google Scholar 

  51. Pinheiro-de-Sousa I, Fonseca-Alaniz MH, Teixeira SK, Rodrigues MV, Krieger JE. Uncovering emergent phenotypes in endothelial cells by clustering of surrogates of cardiovascular risk factors. Sci Rep. 2022;12(1):1372.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  52. Ridker PM, Rifai N, Stampfer MJ, Hennekens CH. Plasma concentration of interleukin-6 and the risk of future myocardial infarction among apparently healthy men. Circulation. 2000;101(15):1767–72.

    Article  CAS  PubMed  Google Scholar 

  53. Xiong X, Liao X, Qiu S, Xu H, Zhang S, Wang S, et al. CXCL8 in Tumor Biology and its implications for clinical translation. Front Mol Biosci. 2022;9:723846.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  54. Ying ZK, Xu XH, Bai YT, Zhong JX, Chen MJ, Liang YJ, et al. Long-term exposure to concentrated ambient PM2.5 increases mouse blood pressure through abnormal activation of the sympathetic nervous system: a role for hypothalamic inflammation. Environ Health Perspect. 2014;122(1):79–86.

    Article  PubMed  Google Scholar 

  55. Adar SD, Sheppard L, Vedal S, Polak JF, Sampson PD, Diez Roux AV, et al. Fine particulate air pollution and the progression of carotid intima-medial thickness: a prospective cohort study from the multi-ethnic study of atherosclerosis and air pollution. PLoS Med. 2013;10(4):e1001430.

    Article  PubMed  PubMed Central  Google Scholar 

  56. Jackson ND, Everman JL, Chioccioli M, Feriani L, Goldfarbmuren KC, Sajuthi SP, et al. Single-cell and Population Transcriptomics reveal pan-epithelial remodeling in type 2-High asthma. Cell Rep. 2020;32(1):107872.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  57. Basil MC, Cardenas-Diaz FL, Kathiriya JJ, Morley MP, Carl J, Brumwell AN, et al. Human distal airways contain a multipotent secretory cell that can regenerate alveoli. Nature. 2022;604(7904):120–.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  58. Kunzli N, Jerrett M, Mack WJ, Beckerman B, LaBree L, Gilliland F, et al. Ambient air pollution and atherosclerosis in Los Angeles. Environ Health Perspect. 2005;113(2):201–6.

    Article  CAS  PubMed  Google Scholar 

  59. Wang J, Huang J, Wang L, Chen C, Yang D, Jin M, et al. Urban particulate matter triggers lung inflammation via the ROS-MAPK-NF-kappaB signaling pathway. J Thorac Dis. 2017;9(11):4398–412.

    Article  PubMed  PubMed Central  Google Scholar 

  60. Totlandsdal AI, Herseth JI, Bolling AK, Kubatova A, Braun A, Cochran RE, et al. Differential effects of the particle core and organic extract of diesel exhaust particles. Toxicol Lett. 2012;208(3):262–8.

    Article  CAS  PubMed  Google Scholar 

  61. Fuentes-Mattei E, Rivera E, Gioda A, Sanchez-Rivera D, Roman-Velazquez FR, Jimenez-Velez BD. Use of human bronchial epithelial cells (BEAS-2B) to study immunological markers resulting from exposure to PM(2.5) organic extract from Puerto Rico. Toxicol Appl Pharmacol. 2010;243(3):381–9.

    Article  CAS  PubMed  Google Scholar 

  62. Chen PH, Chang H, Chang JT, Lin P. Aryl hydrocarbon receptor in association with RelA modulates IL-6 expression in non-smoking lung cancer. Oncogene. 2012;31(20):2555–65.

    Article  CAS  PubMed  Google Scholar 

  63. Matsumoto Y, Ide F, Kishi R, Akutagawa T, Sakai S, Nakamura M, et al. Aryl hydrocarbon receptor plays a significant role in mediating airborne particulate-induced carcinogenesis in mice. Environ Sci Technol. 2007;41(10):3775–80.

    Article  CAS  PubMed  Google Scholar 

  64. Bao L, Liu W, Li Y, Wang X, Xu F, Yang Z, et al. Carcinogenic metabolic activation process of Naphthalene by the cytochrome P450 enzyme 1B1: a computational study. Chem Res Toxicol. 2019;32(4):603–12.

    Article  CAS  PubMed  Google Scholar 

  65. Vignali DA, Kuchroo VK. IL-12 family cytokines: immunological playmakers. Nat Immunol. 2012;13(8):722–8.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  66. Wong JJM, Tan HL, Zhou J, Lee JH, Leong JY, Yeo JG, et al. Large scale cytokine profiling uncovers elevated IL12-p70 and IL-17A in severe pediatric acute respiratory distress syndrome. Sci Rep. 2021;11(1):14158.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  67. Lee HS, Park DE, Lee JW, Chang Y, Kim HY, Song WJ, et al. IL-23 secreted by bronchial epithelial cells contributes to allergic sensitization in asthma model: role of IL-23 secreted by bronchial epithelial cells. Am J Physiol Lung Cell Mol Physiol. 2017;312(1):L13–21.

    Article  PubMed  Google Scholar 

  68. Kameda M, Otsuka M, Chiba H, Kuronuma K, Hasegawa T, Takahashi H, et al. CXCL9, CXCL10, and CXCL11; biomarkers of pulmonary inflammation associated with autoimmunity in patients with collagen vascular diseases-associated interstitial lung disease and interstitial pneumonia with autoimmune features. PLoS ONE. 2020;15(11):e0241719.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  69. Davis AP, Wiegers TC, Johnson RJ, Sciaky D, Wiegers J, Mattingly CJ. Comparative toxicogenomics database (CTD): update 2023. Nucleic Acids Res. 2023;51(D1):D1257–62.

    Article  CAS  PubMed  Google Scholar 

  70. Li H, Nguyen H, Venkata SPM, Koh JY, Kowluru A, Li L, et al. Novel role of GPR35 (G-Protein–coupled receptor 35) in the regulation of endothelial cell function and blood pressure. Hypertension. 2021;78(3):816–30.

    Article  CAS  PubMed  Google Scholar 

  71. Wang W, Han T, Tong W, Zhao J, Qiu X. Overexpression of GPR35 confers drug resistance in NSCLC cells by β-arrestin/Akt signaling. Onco Targets Ther. 2018;11:6249–57.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  72. Breitner S, Schneider A, Devlin RB, Ward-Caviness CK, Diaz-Sanchez D, Neas LM, et al. Associations among plasma metabolite levels and short-term exposure to PM2.5 and ozone in a cardiac catheterization cohort. Environ Int. 2016;97:76–84.

    Article  CAS  PubMed  Google Scholar 

  73. Vivancos PD, Wolff T, Markovic J, Pallardo FV, Foyer CH. A nuclear glutathione cycle within the cell cycle. Biochem J. 2010;431:169–78.

    Article  Google Scholar 

  74. Guo D, Murdoch CE, Xu H, Shi H, Duan DD, Ahmed A, et al. Vascular endothelial growth factor signaling requires glycine to promote angiogenesis. Sci Rep. 2017;7(1):14749.

    Article  PubMed  PubMed Central  Google Scholar 

  75. Godbole S, Bowler RP. Metabolome features of COPD: a scoping review. Metabolites. 2022;12(7).

  76. Luiking YC, Ten Have GA, Wolfe RR, Deutz NE. Arginine de novo and nitric oxide production in disease states. Am J Physiol Endocrinol Metab. 2012;303(10):E1177–89.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  77. Liang DH, Ladva CN, Golan R, Yu TW, Walker DI, Sarnat SE, et al. Perturbations of the arginine metabolome following exposures to traffic-related air pollution in a panel of commuters with and without asthma. Environ Int. 2019;127:503–13.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  78. Wang A, Zhang QY, Wang YM, Li X, Li K, Li Y et al. Inhibition of Gabrp reduces the differentiation of airway epithelial progenitor cells into goblet cells. Exp Ther Med. 2021;22(1).

  79. Cavill R, Jennen D, Kleinjans J, Briedé JJ. Transcriptomic and metabolomic data integration. Brief Bioinform. 2015;17(5):891–901.

    Article  PubMed  Google Scholar 

  80. Zhu T, Yao Q, Wang W, Yao H, Chao J. iNOS induces vascular endothelial cell Migration and apoptosis Via Autophagy in Ischemia/Reperfusion Injury. Cell Physiol Biochem. 2016;38(4):1575–88.

    Article  CAS  PubMed  Google Scholar 

  81. Layton TB, Williams L, Yang N, Zhang M, Lee C, Feldmann M et al. A vasculature niche orchestrates stromal cell phenotype through PDGF signaling: importance in human fibrotic disease. P Natl Acad Sci USA. 2022;119(13).

  82. Karimian A, Mir SM, Parsian H, Refieyan S, Mirza-Aghazadeh-Attari M, Yousefi B, et al. Crosstalk between Phosphoinositide 3-kinase/Akt signaling pathway with DNA damage response and oxidative stress in cancer. J Cell Biochem. 2019;120(6):10248–72.

    Article  CAS  PubMed  Google Scholar 

  83. Pober JS, Sessa WC. Evolving functions of endothelial cells in inflammation. Nat Rev Immunol. 2007;7(10):803–15.

    Article  CAS  PubMed  Google Scholar 

  84. Papazian D, Wurtzen PA, Hansen SW. Polarized Airway Epithelial models for Immunological Co-culture studies. Int Arch Allergy Immunol. 2016;170(1):1–21.

    Article  CAS  PubMed  Google Scholar 

  85. Hartner E, Paul A, Käfer U, Czech H, Hohaus T, Gröger T, et al. On the complementarity and informative value of different Electron ionization Mass Spectrometric techniques for the Chemical Analysis of Secondary Organic Aerosols. ACS Earth Space Chem. 2022;6(5):1358–74.

    Article  CAS  Google Scholar 

  86. Kang E, Root MJ, Toohey DW, Brune WH. Introducing the concept of potential Aerosol Mass (PAM). Atmos Chem Phys. 2007;7(22):5727–44.

    Article  CAS  Google Scholar 

  87. Bruns EA, El Haddad I, Keller A, Klein F, Kumar NK, Pieber SM, et al. Inter-comparison of laboratory smog chamber and flow reactor systems on organic aerosol yield and composition. Atmos Meas Tech. 2015;8(6):2315–32.

    Article  CAS  Google Scholar 

  88. Pardo M, Czech H, Offer S, Sklorz M, Di Bucchianico S, Hartner E et al. Atmospheric aging increases the cytotoxicity of bare soot particles in BEAS-2B lung cells. Aerosol Sci Technol. 2023;57(5):367–383.

  89. Mülhopt S, Dilger M, Diabaté S, Schlager C, Krebs T, Zimmermann R, et al. Toxicity testing of combustion aerosols at the air–liquid interface with a self-contained and easy-to-use exposure system. J Aerosol Sci. 2016;96:38–55.

    Article  Google Scholar 

  90. Karg EW, Ferron GA, Bauer S, Di Bucchianico S, Zimmermann R. Is the particle deposition in a cell exposure facility comparable to the lungs? A computer model approach. Aerosol Sci Technol. 2020;54(6):668–84.

    Article  CAS  Google Scholar 

  91. Martin M. Cutadapt removes adapter sequences from high-throughput sequencing reads. 2011. 2011;17(1):3.

  92. Dobin A, Davis CA, Schlesinger F, Drenkow J, Zaleski C, Jha S, et al. STAR: ultrafast universal RNA-seq aligner. Bioinformatics. 2013;29(1):15–21.

    Article  CAS  PubMed  Google Scholar 

  93. Anders S, Pyl PT, Huber W. HTSeq—a Python framework to work with high-throughput sequencing data. Bioinformatics. 2014;31(2):166–9.

    Article  PubMed  PubMed Central  Google Scholar 

  94. Edgar R, Domrachev M, Lash AE. Gene expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic Acids Res. 2002;30(1):207–10.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  95. Robinson MD, Oshlack A. A scaling normalization method for differential expression analysis of RNA-seq data. Genome Biol. 2010;11(3):R25.

    Article  PubMed  PubMed Central  Google Scholar 

  96. Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, et al. Limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 2015;43(7):e47.

    Article  PubMed  PubMed Central  Google Scholar 

  97. Gregory R, Warnes BB, Bonebakker L, Gentleman R, Liaw WHA, Lumley T, Maechler M. Arni Magnusson, Steffen Moeller, Marc Schwartz and Bill Venables. gplots: Various R Programming Tools for Plotting Data. R package version 2.17.0. 2015.

  98. Blighe K, Rana S, Lewis M, EnhancedVolcano. Publication-ready volcano plots with enhanced colouring and labeling. 2022 [.

  99. Wu T, Hu E, Xu S, Chen M, Guo P, Dai Z, et al. clusterProfiler 4.0: a universal enrichment tool for interpreting omics data. Innov (Camb). 2021;2(3):100141.

    CAS  Google Scholar 

  100. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, et al. Gene ontology: tool for the unification of biology. The Gene Ontology Consortium. Nat Genet. 2000;25(1):25–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  101. Gene Ontology C. The Gene Ontology resource: enriching a GOld mine. Nucleic Acids Res. 2021;49(D1):D325–34.

    Article  Google Scholar 

  102. Gu Z, Eils R, Schlesner M. Complex heatmaps reveal patterns and correlations in multidimensional genomic data. Bioinformatics. 2016;32(18):2847–9.

    Article  CAS  PubMed  Google Scholar 

  103. Pink M, Verma N, Zerries A, Schmitz-Spanke S. Dose-dependent response to 3-Nitrobenzanthrone exposure in Human Urothelial Cancer cells. Chem Res Toxicol. 2017;30(10):1855–64.

    Article  CAS  PubMed  Google Scholar 

  104. Pink M, Verma N, Kersch C, Schmitz-Spanke S. Identification and characterization of small organic compounds within the corona formed around engineered nanoparticles. Environ Sci-Nano. 2018;5(6):1420–7.

    Article  CAS  Google Scholar 

  105. Domingo-Almenara X, Brezmes J, Vinaixa M, Samino S, Ramirez N, Ramon-Krauel M, et al. eRah: a computational Tool Integrating Spectral Deconvolution and Alignment with quantification and identification of metabolites in GC/MS-Based metabolomics. Anal Chem. 2016;88(19):9821–9.

    Article  CAS  PubMed  Google Scholar 

  106. Kopka J, Schauer N, Krueger S, Birkemeyer C, Usadel B, Bergmuller E, et al. GMD@CSB.DB: the Golm Metabolome Database. Bioinformatics. 2005;21(8):1635–8.

    Article  CAS  PubMed  Google Scholar 

  107. Pang Z, Zhou G, Ewald J, Chang L, Hacariz O, et al. Using MetaboAnalyst 5.0 for LC-HRMS spectra processing, multi-omics integration and covariate adjustment of global metabolomics data. Nat Protoc. 2022;17(8):1735–1761.

Download references

Acknowledgements

We thank W.H. Brune (Pennsylvania State University) for providing the potential aerosol mass (PAM) reactor, which was funded by the University of Rostock.

Funding

We thank the Helmholtz International Laboratory aeroHEALTH (InterLabs-0005; https://www.aerohealth.eu), the Helmholtz Association of German Research Centers, the Helmholtz Virtual Institute of Complex Systems in Environmental Health (InhaleHICE) and the Projektverbund “ Messung, Charakterisierung, Bewertung ultrafeiner Partikel” (UFP) for granting this project. C.B. acknowledges the support of the Swiss National Science Foundation (grant P2FRP3_178112). YR acknowledges support from the EASVOLEE project (Horizon Europe Framework Programme, under grant agreement No 101095457) and by a research grant from the Estate of Benjamin Carasso and the Pudgin Memorial Trust.

Open Access funding enabled and organized by Projekt DEAL.

Author information

Authors and Affiliations

  1. Joint Mass Spectrometry Center (JMSC) at Comprehensive Molecular Analytics (CMA), Helmholtz Zentrum München, Ingolstädter Landstr. 1, D-85764, Neuherberg, Germany

    Svenja Offer, Sebastiano Di Bucchianico, Hendryk Czech, Jana Pantzke, Christoph Bisig, Stefanie Bauer, Elias J. Zimmermann, Sebastian Oeder, Elena Hartner, Thomas Gröger, Jürgen Schnelle-Kreis, Martin Sklorz & Ralf Zimmermann

  2. Joint Mass Spectrometry Center (JMSC) at Analytical Chemistry, Institute of Chemistry, University of Rostock, Albert-Einstein-Str. 27, D-18059, Rostock, Germany

    Svenja Offer, Sebastiano Di Bucchianico, Hendryk Czech, Jana Pantzke, Eric Schneider, Elias J. Zimmermann, Elena Hartner, Thomas Gröger, Christopher P. Rüger & Ralf Zimmermann

  3. Department Life, Light & Matter (LLM), University of Rostock, D-18051, Rostock, Germany

    Sebastiano Di Bucchianico, Eric Schneider, Christopher P. Rüger & Ralf Zimmermann

  4. Department of Earth and Planetary Sciences, Faculty of Chemistry, Weizmann Institute of Science, 234 Herzl Street, POB 26, Rehovot, ISR-7610001, Israel

    Michal Pardo & Yinon Rudich

  5. Institute and Outpatient Clinic of Occupational, Social and Environmental Medicine, Friedrich-Alexander University of Erlangen-Nuremberg, Henkestr. 9-11, D-91054, Erlangen, Germany

    Rasha Alsaleh, Christian Kersch & Simone Schmitz-Spanke

  6. Institute of Energy and Climate Research, Forschungszentrum Jülich GmbH, Troposphere (IEK-8), Wilhelm- Johen-Str, D-52428, Jülich, Germany

    Till Ziehm, Thorsten Hohaus & Astrid Kiendler-Scharr

Authors
  1. Svenja Offer
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Sebastiano Di Bucchianico
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Hendryk Czech
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Michal Pardo
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Jana Pantzke
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Christoph Bisig
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Eric Schneider
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Stefanie Bauer
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Elias J. Zimmermann
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Sebastian Oeder
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Elena Hartner
    View author publications

    You can also search for this author in PubMed Google Scholar

  12. Thomas Gröger
    View author publications

    You can also search for this author in PubMed Google Scholar

  13. Rasha Alsaleh
    View author publications

    You can also search for this author in PubMed Google Scholar

  14. Christian Kersch
    View author publications

    You can also search for this author in PubMed Google Scholar

  15. Till Ziehm
    View author publications

    You can also search for this author in PubMed Google Scholar

  16. Thorsten Hohaus
    View author publications

    You can also search for this author in PubMed Google Scholar

  17. Christopher P. Rüger
    View author publications

    You can also search for this author in PubMed Google Scholar

  18. Simone Schmitz-Spanke
    View author publications

    You can also search for this author in PubMed Google Scholar

  19. Jürgen Schnelle-Kreis
    View author publications

    You can also search for this author in PubMed Google Scholar

  20. Martin Sklorz
    View author publications

    You can also search for this author in PubMed Google Scholar

  21. Astrid Kiendler-Scharr
    View author publications

    You can also search for this author in PubMed Google Scholar

  22. Yinon Rudich
    View author publications

    You can also search for this author in PubMed Google Scholar

  23. Ralf Zimmermann
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

S.D.B., H.C., T.H., A.K.-S., Y.R. and R.Z. conceived and supervised the study. R.A., C.B. S.B., S.D.B., H.C., T.G., E.H., T.H., C.K., A.K.-S., S.Oe., S.O., J.P., M.P., Y.R., C.R., E.S., J.S.-K., M.S., S.S.-S, T.Z., E.J.Z. and R.Z. designed and performed the experiments. R.A., S.D.B., H.C., C.K., S.O., M.P., Y.R and T.Z analyzed and evaluated the data. A.K.-S., Y.R, and R.Z. provided funding; S.O. wrote the original draft with the help of S.D.B., H.C. and R.Z. Revision and approval were conducted by all authors.

Corresponding author

Correspondence to Sebastiano Di Bucchianico.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

^Deceased: Astrid Kiendler-Scharr.

Electronic supplementary material

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Offer, S., Di Bucchianico, S., Czech, H. et al. The chemical composition of secondary organic aerosols regulates transcriptomic and metabolomic signaling in an epithelial-endothelial in vitro coculture. Part Fibre Toxicol 21, 38 (2024). https://doi.org/10.1186/s12989-024-00600-x

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12989-024-00600-x

Keywords

Particle and Fibre Toxicology

ISSN: 1743-8977

Contact us