https://ift.tt/XgJneYq
Unraveling toxicity of nanoparticles from different subway materials in lung epithelial cells and macrophages
1. Introduction
Ambient particulate matter (PM) is classified as a human group 1 carcinogen (Straif et al., 2013) and is estimated to be a leading cause for over 4 million deaths attributed to environmental factors per year worldwide (Cohen et al., 2017). While the World Health Organization (WHO) publishes air quality guidelines regarding coarse and fine particles regularly, statements concerning ultrafine particles were included for the first time in 2021, emphasizing the need to start monitoring these small particles (World Health Organization, 2021). The ultrafine particles, also called nanoparticles, are defined as being smaller than 100 nm and they often dominate the total particle number concentration (PNC) in the air (Stölzel et al., 2007; Midander et al., 2012). Although some studies have shown an association between ultrafine particles and an increase in mortality, available studies show mixed results and difficulties in exposure estimation are likely since ultrafine particles have a larger spatial and temporal variability compared to fine particles (Stölzel et al., 2007; Vallabani et al., 2023).
Compared to coarse and fine particles, the small size of nanoparticles plays a crucial role in toxicity. Nanoparticles display unique properties such as increased surface-to-volume ratios and surface reactivity, also with respect to airborne exposure scenarios. Moreover, these particles may also deposit into deep alveolar lung tissues and escape tracheobronchial clearance mechanisms (Geiser and Kreyling, 2010). Alveolar macrophages play a significant role in the local immune defense of the lungs, but nanoparticles have also been shown to largely be able to escape the macrophage surveillance, leading to higher retention in the lung (Oberdörster, 2001). Furthermore, by crossing the air-blood barrier, nanoparticles are a potential hazard to other organs such as the liver and spleen, but also e.g., kidney, testis, and brain (Lee et al., 2023; Hadrup et al., 2024). They have also been shown to overcome the blood-brain barrier or take the olfactory nerve as a route to enter the brain (Maher, 2019). In this context, several studies have reported that particle exposure plays a role in the pathogenesis of neurodegenerative processes such as dementia (Maher, 2019; Calderón-Garcidueñas et al., 2022). Moreover, metal solid nanoparticles have been found in the brains of children and young adults in Mexico City, which was shown to be associated with neuropathology and potential lethality (Calderón-Garcidueñas et al., 2022).
Nanoparticles can enter cells via different entry and uptake routes. Once inside the cells, nanoparticles often end up in acidic lysosomes, and for some metallic nanoparticles the acidic environment will lead to high dissolution/metal ion release. Hence, the particle structure can “smuggle” toxic ions into cells and contribute to an increased intracellular ion load (a process sometimes referred to as a “Trojan-horse mechanism”). (Limbach et al., 2007; Horie et al., 2009). Toxicity is often driven by the generation of reactive oxygen species (ROS), which e.g., can be formed via Fenton-like reactions (Paunovic et al., 2020). Moreover, nanoparticles as well as larger PM have been shown to cause DNA damage and inflammation, processes that play a role in carcinogenesis (Nishanth et al., 2011; Oberdörster, 1995). Hence, it is of high relevance to identify both major sources of these particles in our environment and their impact on human health. In this context, many studies have reported significantly increased levels of airborne PM in the subway compared to above-ground locations in various cities, including Barcelona, Athens, Mexico City, and Stockholm, among many others (Martins et al., 2017; Velasco et al., 2019; Tu and Olofsson, 2021; Bendl et al., 2023). Worldwide, many commuters use the subway daily, leading to a steady expansion of the railway network in many cities. Consequently, employees working in the subway are growing in number as well. Subway PM originates to a large extent from non-exhaust emissions from wear by the rail network (Bendl et al., 2023; Cha et al., 2016). Proper ventilation of subway stops is also challenging due to often deep-lying subway stations underground (Pan et al., 2013). In fact, outdoor air quality levels regularly exceed levels recommended by the WHO, which raises concerns about possible effects on human health. Moreover, it should be noted that the air in the subway stations is not covered by the legislation or guidelines for ambient air.
A rather specific property of subway particles, compared to other urban particles, is that they contain high amounts of metallic elements, primarily iron but also manganese, copper, and chromium (Bendl et al., 2023; Chillrud et al., 2004; Font et al., 2019). The exact elemental composition does not only differ between subway systems, but also varies between materials such as rails or wheels. Noteworthy, a multi-analytical approach by Midander et al. reported that over 98 % of the mean PNC in the Stockholm subway is below 250 nm in size (Midander et al., 2012). Moreno et al. showed that most particles from the subway are nanosized, mainly consisting of iron (Moreno et al., 2015), which is supported by a study by Bendl et al., reporting the presence of iron oxide nanoparticles around 30 nm (Bendl et al., 2023). Moreover, a recent study investigating the non-exhaust particle emissions from two different subway lines of the subway system in Toronto reported that a “Chemical Mass Balance (CMB) model apportioned 92 % and 55 % of PM2.5 to iron-rich compounds” such as wheels, rails and contact rails (Van Ryswyk et al., 2024). This specific metal-rich composition of subway particles is assumed to be a leading cause of their toxicity (Kim et al., 2010). However, we still lack profound knowledge on the toxicity of subway particles and their respective health impact.
Loxham et al. reviewed data on the toxicity of subway particles in 2019, describing 27 relevant publications on this topic including both short- and long-term exposure studies (Loxham and Nieuwenhuijsen, 2019). Several of the studies available indicated health impacts such as significantly increased fibrinogen plasma levels after 2 h of exposure (Klepczyńska Nyström et al., 2010) or a significant correlation between oxidative stress markers and working years (Grass et al., 2010). Thus, researchers suggest an influence on inflammation and other particle-driven effects by chronic exposures (Grass et al., 2010; Bigert et al., 2008). In vitro studies, in particular, have shown several cellular pathways and processes impacted by particle exposure, highlighting the potential health hazards associated with such exposures. Early studies by Karlsson et al. compared particles from a subway station to particles collected from street as well as from tire-road wear and wood combustion in vitro. They found the highest ROS production and DNA damage upon exposures to subway particles (Karlsson et al., 2006, 2008). Also, Seaton et al. reported that subway particles led to increased DNA damage, oxidative stress, cytotoxicity as well as inflammation compared to particles from welding fumes or urban sources (Seaton et al., 2005). The “RAPTES” study (Risk of Airborne Particles: a Toxicological–Epidemiological hybrid Study) conducted in the Netherlands found associations between eosinophilic lung inflammation, impaired lung function and subway-related particles containing iron and copper (Strak et al., 2012). Researchers furthermore propose that the unique properties of subway PM compared to particles from other airborne sources, such as the high metal content, may lead to distinct toxicological effects (Steenhof et al., 2011, 2014; Cooper and Loxham, 2019).
Overall, the study by Loxham et al., in general, concluded that “there is little direct evidence that underground railway particulate matter exposure is more harmful than ambient particulate matter exposure”. In contrast, a recent estimation of human health risks associated with PM10 and PM2.5-bound trace metals reports that the findings “support the view that indoor air pollutant levels in subway systems can have a serious impact on the cancerous and non-cancerous risks to human health” (Roy et al., 2022).
Taken together, the available literature showed mixed conclusions, and the toxicological studies available were mainly performed using larger, micron-sized particles. Further, studies on subway nanoparticles are rare (Vallabani et al., 2023; Bendl et al., 2023) and there is a lack of understanding of the toxicity of such particles. Thus, there is a clear need for additional studies on subway toxicity in order to provide a better understanding on potential health impacts. To the best of our knowledge, studies investigating nanoparticles generated from different subway materials in a controlled manner are currently not available at all. Such materials include wheels, brakes, and the conductor rail that provides power to the train (hereafter referred to as the "third rail").
This study aimed to explore the toxicity of nanoparticles from different subway materials, including third rail, rail, and wheel, and to compare this to iron nanoparticles. Nanoparticles were generated using spark discharge. Particles were then assessed for oxidative potential (OP), cytotoxicity, genotoxicity, and inflammatory potential in monocyte-derived macrophages (dTHP-1) and lung epithelial cells (A549). The aim of this study was to contribute to a better understanding of possible effects on human health in subway environments.
2. Materials and methods
2.1. Cell culture and reagents
Human monocytic THP-1 cells were originally obtained from the European Collection of Authenticated Cell Cultures (ECACC, Sigma-Aldrich, CB-88081201, leukemic monocyte) and cultured in RPMI-1640 medium (Gibco) supplemented with 10% fetal bovine serum (FBS), 1% penicillin/streptomycin and 2 mM L-glutamine. Cells were cultured in T75 tissue culture flasks, standing at 37 °C with 5% CO2 in a humidified incubator and passaged once a week. Cell density was kept between 5 × 105-1.5 × 106 cells/mL. Differentiation into monocyte-derived macrophages (dTHP-1) was performed by incubation of THP-1 cells with 50 ng/mL phorbol 12-myristate 13-acetate (PMA, Sigma) in RPMI-1640 medium at 37 °C with 5% CO2 for 48 h.
A549 cells were purchased from the American Type Culture Collection (ATCC, product number CCL-185) and were cultured in DMEM medium (Gibco) supplemented with 10% FBS, 1% penicillin/streptomycin, and 1 mM sodium pyruvate. Cells were passaged three times per week in T75 tissue culture flasks and kept in a humidified incubator at 37 °C and 5% CO2.
2.2. Subway material nanoparticles and spark discharge
All materials used in this study, including third rail, rail, and wheel were obtained from the respective components of the Stockholm subway system. The materials were used as electrodes in a spark discharge particles generator (DNP Digital 3000, Palas GmbH, Karlsruhe, Germany) to generate nanoparticles under laboratory conditions and a constant N2 flow (99.996 vol%, 3 lpm). Particles were collected on 37 mm polycarbonate filters (WHAT10417309) for 2 h using a total dust sampler at 1.95 lpm to obtain 2–3 mg per filter. Iron electrodes (purity >99.99 %) and copper electrodes (purity >99.99%) from Goodfellow (Coraopolis, USA) were used to generate iron and copper nanoparticles as a reference, respectively. Filters were stored in closed vessels in a desiccator flushed with argon.
2.3. Preparation of particles for cell exposure
Filters were placed into a glass bottle (100 mL) and 1 mL MilliQ water was added on top of the filter, whereby vigorous mixing on a vortex was applied multiple times. The filter was additionally rinsed with 500 μL MilliQ until no further particles could be removed by visual inspection. The particle suspension was pooled with another 500 μL MilliQ water to rinse the glass tube and then transferred to a glass test tube. The particle suspension was sonicated in a water bath sonicator (ultrasonic cleaner USC 200T, 45 kHz, VWR) for 20 min at 30 °C and vortexed subsequently. The concentration of particles in suspension was calculated by the difference of filter dry weight before and after removing particles. An empty filter served as a blank control for particle exposure, following the same procedure. Particle suspensions were stored as aliquots at −80 °C. They were thawed and sonicated for 20 min at 30 °C in a water bath sonicator and vortexed again vigorously prior to cell exposure.
2.4. Transmission and scanning electron microscopy
Particle size was determined by Transmission Electron Microscopy (TEM) and Scanning Electron Microscopy (SEM) imaging. Samples were prepared with a particle concentration of 50 μg/mL in MilliQ water by pipetting 3 μL onto a glow-discharged carbon-coated and formvar-stabilized 400 mesh copper grid (Ted Pella). After 30 s, the sample was blotted off and washed with MilliQ water. TEM was performed with a Hitachi HT7700 transmission electron microscope, operating at 100 kV and equipped with a 2kx2k Veta CCD camera (Olympus Soft Imaging system). For SEM imaging, samples were prepared as follows: first, particles deposited on collection filters were stripped by sonication for 5 min at 35 kHz in 4 mL of ethanol in a 5 mL glass vial. Afterwards, particles were left deposited on the bottom of the vial overnight. Following this, about 3 mL of ethanol was removed by hand using a Pasteur pipette. Finally, the remaining volume of ethanol was evaporated in mild vacuum conditions (20 mbar) for 12 h. Once the ethanol was completely removed, about 1 mg of free powder for each sample was taken from the bottom of the vial, using a metallic but not magnetic spatula. The powders were then deposited on an aluminum stub covered with carbon tape for electron microscopy (Agar Scientific). SEM images were acquired using a Zeiss EVO MA10 electron microscope equipped with LaB6 filament. All the images were acquired with secondary electrons to highlight the morphological details of the investigated particulates. All the measurements were performed using the following power parameters: V = 20 kV and I = 300 pA. The primary particle size of the agglomerates was analyzed by using TEM images and Image J. In each image, the diameters of 25–50 nanoparticles were manually marked and the length was measured by Image J.
2.5. μ-ED-XRF
Micro Energy Dispersive X-Ray Fluorescence analysis (μ-ED-XRF) was used to investigate the chemical composition of the produced particulate. The specific target of the analysis was the characterization of the inorganic fraction, with a particular focus on d-block elements. The analysis was performed by means of the Bruker M4 Tornado spectrometer equipped with a Rh-source. All the measurements were performed directly on the samples deposited on a zero background Teflon sample holder without any further manipulation. Measurements were carried out using the following power parameters: V = 50 kV and I = 200 μA (detector deadtime <15%). Each sample was analyzed on an area of 10 × 10 mm, with a spatial resolution of 25 μm and three measurement repetitions. The used instrument does not provide information on light elements (Z < 11), therefore results are provided as normalized over the 11 < Z < 92 range. The detection limits for the samples at the experimental conditions are in order of 0.01 wt% (100 ppm) for each of the measured elements.
2.6. Acellular oxidative potential and ROS formation
To determine the oxidative potential of particles, the dithiothreitol (DTT) assay, as well as the acellular dichlorodihydrofluorescein diacetate (DCFH-DA) assay, were used. The DTT assay was performed similarly as previously described (Cho et al., 2005). In short, reactions with DTT were performed in 1.5 mL centrifuge tubes containing 0.8 mL blank filter or particle filter water suspension, 0.1 mL 1 M potassium phosphate buffer (pH 7.4), and 0.1 mL 1 mM DTT (total volume 1.0 mL). The concentration of nanoparticles in the reaction vials was 28 μg/mL for the three subway materials and 0.12 μg/mL for the copper nanoparticles used as control. The vials were incubated in a shaking water bath at 37 °C. At different time points (0, 12, 24, and 36 min), a 0.2 mL aliquot was removed from the respective incubated vial and quenched with 0.2 mL 10% trichloroacetic acid in a 24 well plate. After the final aliquot was taken, 20 μL of 10 mM 5,5′-dithiobis-(2-nitrobenzoic acid) (DTNB) was added to each sample. DTNB reacts with the remaining DTT to form 2-nitro-5-thiobenzoic acid (TNB). After ∼10 min, a 0.75 mL mixture of 0.4 M Tris-Base and 20 mM ethylenediamine tetraacetic acid (EDTA, pH 8.9) was added to stabilize the solution. The absorption of TNB at 412 nm was then measured with a microplate reader (CLARIOstar Plus, BMG Labtech) within 2 h, which is directly proportional to the remaining DTT concentration. The OP is given by the DTT consumption rate. Measured OP values were normalized to the mass of nanoparticles in the reaction vial to calculate the mass-normalized OP. All samples were analyzed in triplicates and each batch was accompanied by a blank in duplicates. Possible absorbance from the nanoparticles themselves was measured and corrected for. As another test, the acellular DCFH-DA assay was performed as previously described in detail (Kessler et al., 2021) (although only the version without HRP was used). The nanoparticles were thawed and diluted to 100 μg/mL, sonicated for 20 min at 37 °C, and mixed with the DCFH according to the protocol. NiO nanoparticles (50 μg/mL) were used as a positive control. The fluorescent signal was measured at 485/530 nm excitation/emission after 30 min incubation and fold change compared to control (blank filter) was calculated. Controls containing only nanoparticles were included to investigate potential interference with the signal from the particles themselves. However, no influence on the readings was observed.
2.7. Cell viability
An Alamar Blue assay was performed to measure cell viability. In brief, 5.5 × 104 THP-1 cells per well were seeded in a 96 well plate and differentiated. A549 cells (1 × 104 cells/well) were seeded in a 96 well plate for 24 h at 37 °C. After thawing particles, a stock of 200 μg/mL was prepared in the respective cell culture medium and sonicated again for 20 min at 30 °C in a water bath sonicator. Thereafter, cells were exposed to 10 μg/mL-200 μg/mL particle suspension for 24 h. The control was set up using a blank filter and 50 μg/mL cobalt nanoparticles (Sigma) were used as a positive control. After the exposure, medium was removed and 100 μL of 10 % Alamar Blue reagent (Invitrogen, Carlsbad, CA) in cell medium was added to each well and incubated for 2 h at 37 °C. The fluorescent signal was then measured at 540/590 nm excitation/emission with a plate reader (Tecan Infinite F2000, Magellan 7.2 Software, Grodig, Austria). Cell-free wells exposed to particles were included to rule out interference of particles with the assay, and fluorescent values were normalized to the blank control which was set to 100 % viability.
2.8. Genotoxicity (alkaline comet assay)
To assess genotoxicity, the alkaline comet assay was in principle performed as previously described (Gliga et al., 2014). Here, 3.3 × 105 THP-1 cells per well were seeded in 24 well plates and differentiated for 48 h to dTHP-1 cells. A549 cells were seeded in 24 well plates with 6 × 104 cells per well for 24 h. After thawing and preparing the final particle concentrations (10, 50, and 100 μg/mL) in the respective cell medium, particles were sonicated for 20 min at 30 °C in a water bath sonicator. Both cell lines were exposed to the particles for 3 h and 24 h. A blank filter was used as control and 50 μg/mL NiO nanoparticles served as a positive control. Briefly described, microscope slides were coated with 0.3 % low-gelling agarose, and cells were embedded in 0.75 % low-gelling agarose and spread onto the coated slides. Therein, cells were lysed in Triton X-100 lysis buffer overnight at 4 °C and subsequently underwent alkaline treatment for 40 min in electrophoresis buffer. Electrophoresis took place for 30 min at 29 V, whereafter neutralization was performed followed by overnight drying and fixation in methanol for 5 min. Afterwards, cells were stained with SYBR Green (1:10.000 in TAE Buffer) for 20 min and the slides were dried completely before scoring. The slides were scored blinded using a fluorescent microscope (Leica DMLB, Wetzlar, Germany) using the Comet assay IV software (Perceptive Instruments, Suffolk, UK).
2.9. Gene expression (Rt-qPCR)
Real-time quantitative PCR (Rt-qPCR) was applied to investigate markers of DNA damage, oxidative stress, and inflammation at the gene level. In brief, 1.65 × 106 THP-1 cells per well were seeded in 6 well plates and differentiated for 48 h, and 3 × 105 A549 cells per well were seeded in 6 well plates for 24 h. Both cell lines were exposed to 100 μg/mL subway nanoparticles for 24 h, whereby blank filters were used as a control. RNA from exposed cells was extracted on ice, following instructions of the manufacturer’s protocol (RNeasy® Mini Kit 250, Qiagen, Germany) and stored at −80 °C upon further transcription into cDNA. The final RNA concentration was measured by using a photo spectrometer (NanoDrop). RNA (1 μg) was then reverse transcribed into cDNA, following instructions of the manufacturer’s protocol (High Capacity cDNA Reverse Transcription Kit, Thermo Fisher Scientific, USA) and stored at −20 °C until usage. Gene expression was measured with respect to GAPDH as a reference gene and of GADD45a, HMOX1, IL-1β, IL-6, IL-8, and TNF-α (supplement S1). The master mix included 10 μL SYBR green per reaction and 0.3 μM of the respective forward and reverse primers as well as 0.01 μM ROX, with a total reaction volume of 18 μL. 2 μL cDNA (or water as control) was added to each well. The RT-qPCR program was performed at an annealing temperature of 57 °C (supplement S2). Results were analyzed with the QuantStudio 5 Software (Thermo Fisher Scientific, USA). Relative gene expression was calculated according to the 2−ΔΔCt method. Results are displayed as log2 fold change in a heatmap.
2.10. Inflammation (cytokine release with four-plex immunoassay)
Inflammation was investigated in terms of cytokine release via a four-plex electrochemiluminescence immunoassay kit (V-PLEX Human Proinflammatory Panel II 4-Plex) from Meso Scale Discovery (Rockville, MD) to detect the proinflammatory cytokines IL-1β, IL-8, IL-6, and TNF-a. Our first analysis showed that for dTHP1, the levels of IL-8 were too high for proper quantification. Additional experiments were done to test various dilutions, with the conclusion that × 500 enabled all samples included in the test to fall within the standard curve for IL-8. The samples were therefore further diluted (in total 500 times), and IL-8 was again measured (V-PLEX IL-8). The assay was performed as described in the manufacturer’s protocol and was previously described by McCarrick et al. (2021). Regarding cell seeding, 5.5 × 104 THP-1 cells per well were seeded in a 96 well plate and differentiated to dTHP-1 cells for 48 h at 37 °C. A549 cells (1 × 104 cells/well) were seeded in a 96 well plate for 24 h. After the exposure to 100 μg/mL particles and 1 μg/mL lipopolysaccharides (LPS) as positive control for 24 h, the cell medium was gently collected, triplicates were pooled and then stored at – 80 °C until it was used for the cytokine release studies. Results are displayed as log2 fold change in a heatmap.
2.11. Statistics
All experiments were conducted with a minimum of three independent experiments. For cell viability and DNA damage (with at least three doses), a two-way ANOVA was applied to test the significance of concentration and/or particle type. A Holm-Šídák’s multiple comparisons post hoc test was performed to explore differences between particle types. A Dunnett’s post hoc test for multiple comparisons was employed to investigate which doses resulted in significantly different values compared to the control group. For Rt-qPCR and cytokine release (with one dose) a one-way ANOVA followed by Dunnett’s multiple comparison test was used to test for significance between exposure (fold change) and control. GraphPad Prism Software 9 was used for all statistical tests and the graphical representation of respective results. Mean values are depicted with the standard error (SE) of the mean. A p-value of <0.05 was considered significant.
4. Discussion
The overall goal of this study was to increase the understanding of the toxicity of nanoparticles formed in subway systems. Thus, nanoparticles were generated by spark-discharge from electrodes of the same material as third rail, rail and wheel in the subway system in Stockholm. A range of toxicity endpoints was explored using both monocyte-derived macrophages (dTHP-1) and lung epithelial cells (A549). Our main finding is that the nanoparticles caused DNA strand breaks in both dTHP-1 cells and A549 cells (statistically significant for three of the four nanoparticles), but they were not potent in inducing ROS. Also, clear changes in gene expression of the pro-inflammatory marker IL-8 were noted. Third rail, which contains more silicon, was slightly more cytotoxic and caused more potent changes in IL-8 gene expression but slightly less DNA damage. Overall, large and consistent differences in the toxicity between nanoparticles from different materials were not observed.
The way we produced the nanoparticles should be considered novel for these materials. No other studies have been found on subway-related nanoparticles generated by spark-discharge. In general, we did not find any clear acellular ROS generation by the nanoparticles, and furthermore, the cytotoxic effects were rather limited. This contrasts with results on so-called “quasi-ultrafine” particles collected from the underground in the Netherlands (Strak et al., 2012), for which clear cytotoxic effects were observed in exposed macrophages (Steenhof et al., 2011). Also, a clear ROS generating ability was observed by using the DTT assay. Likely, the “quasi-ultrafine” particles used in the study were different in terms of composition (copper etc.). Moreover, all nanoparticles used in our study were highly agglomerated, and this could affect the toxicity. The method for dispersion of the stock suspension was a 20 min sonication in a water bath sonicator. Possibly, using a probe sonicator could have achieved a better dispersion. The ROS generation in our study was tested by acellular assays, but also by studying the gene expression of the oxidative stress marker HMOX1. Here, a significant increase was noted for third rail. Despite their limited ROS-generating ability, the nanoparticles showed a dose-depended increase in DNA breaks. This result thus confirms the previous data on micron-sized particles from the subway environment (Karlsson et al., 2005, 2006). However, there were no significant changes observable in gene expression of the DNA damage marker GADD45a. GADD45a is involved in a complex signaling network associated with cell cycle control and cell death (Liebermann and Hoffman, 2008). As part of the nucleotide excision repair pathway, this DNA damage marker is diversely regulated (Smith et al., 2000). The reason for the lack of response, despite effects observed in the comet assay, might be because the comet assay is able to pick up single strand breaks that may not be severe enough for inducing effects on cell cycle and apoptosis and thus, GADD45a signaling.
In this study, we used two cell lines for all endpoints, which allows for a clear comparison between these two. Particularly, changes in gene expression were very different in dTHP-1 and A549 cells. The changes were more obvious in dTHP-1 cells and the oxidative stress marker HMOX1 was e.g., slightly increased in dTHP-1 cells whereas a slight repression was observed in A549 cells. However, this result should not be seen as an effect in lung cells in general, as this effect in the A549 cancer cell line can be attributed to a mutation in the kelch-like ECH-associated protein 1 (KEAP1) gene. KEAP1 usually represses the transcription factor nuclear factor erythroid 2-related factor 2 (NRF2), which activates pathways concerning oxidative stress in terms of antioxidant mechanisms or inflammation. In the A549 cell line, KEAP1 underwent a loss of function mutation. Consequently, NRF2 levels in A549 cells are increased, whereby antioxidant proteins and detoxification mechanisms are more pronounced. This makes the A549 cell line more robust to related stimuli (Singh et al., 2006). As HMOX-1 is inducible by NRF2, HMOX-1 baseline levels in A549 cells are assumed to be elevated. Thus, A549 cells are not just potentially leading to an underestimation in terms of inflammation but are also more resistant to induced oxidative stress (Kweon et al., 2006). Still, A549 cells are valuable to use in nanoparticle studies since this is a well-used model for alveolar cells, which enables the comparison to other studies. A549 cells also have been used to study effects in air-liquid interface (ALI) approaches, which also could be interesting for future comparisons. The more pronounced effects observed in monocyte-derived macrophages were expected, as they are a part of the immune system and have a role in the clearance of particles and pathogens from the lungs. It should also be noted that macrophage-differentiated THP-1 cells are efficient as phagocytes, and thus, it is likely that the uptake of nanoparticles is higher for these cells (Rudolph et al., 2021). Increased cytokine release of (d)THP-1 compared to A549 cells has also been observed in other studies, e.g. in a study on cigarette smoke exposure (Holownia et al., 2016), as well as in a study on cerium oxide (CeO2) nanoparticles (Cappellini et al., 2020). As A549 cells have been shown to display impaired oxidative stress mechanisms, this could indeed be another indicator for differences in inflammation with respect to higher levels of cytokine release between A549 and dTHP-1 cells. The changes in gene expression of inflammatory markers, particularly IL-8, were not clearly translated to an increase in cytokine release. This could be due to several factors, including the higher sensitivity of transcriptional effects compared to cytokine release and the fact that various post-transcriptional mechanisms critically determine the outcome of immune responses (Liu et al., 2023). Additionally, our data showed significantly greater variation in cytokine release measurements for dTHP-1 cells, which contributed to uncertainty and less possibility of finding statistically significant effects.
With respect to the concentrations tested, a study modeling human alveolar lung retention of particles suggested that for a full working lifetime exposure to nanoparticles (20 nm) at 0.1 mg/m3, the highest lung surface concentration was 4.9 μg/cm2 (Gangwal et al., 2011). This suits the concentrations applied in this study well, where main effects were studied at 10–100 μg/mL corresponds approximately to 3–31 μg/cm2. In a study exploring human lung doses following high (5 mg/m3) exposure to welding fume nanoparticles in an occupational setting, McCarrick et al. showed that the tracheobronchial lung dose was estimated to be 0.89 μg/cm2 after 6 h exposure (Mccarrick et al., 2022). Since the clearance is relatively efficient in the tracheobronchial region, this amount decreases rapidly when exposure stops, whereas the lung dose builds up in the alveolar region. Thus, after one working year, the tracheobronchial and alveolar retention was estimated to be 1.15 and 2.85 μg/cm2, respectively (Mccarrick et al., 2022). The concentrations of nanoparticles in subway systems are clearly much lower compared to the concentration used for simulating welding occupational settings. The PM levels in the Stockholm subway system were recently monitored for over two years. The study shows that the average PM10 and PM1 levels were 169 μg/m³ and 31 μg/m³, respectively (Tu and Olofsson, 2024). We are currently modeling the lung retention of particles in subway systems and plan to compare these results to the doses used in this study. Although the concentrations used in the study are high compared to the expected lung dose received by subway commuters, they fall within the range of doses often used to test particle toxicity, thus allowing for comparisons between studies.
Although we noted some differences between the nanoparticles tested such as that third rail caused a slightly stronger cytotoxic response and a more pronounced difference in IL-8 gene expression, there was no large and consistent difference in toxicity between the nanoparticles tested observable. This is likely explained by the fact that there were only small differences in chemical composition, except for the higher amount of silicon in third rail. A study on subway materials from Barcelona analyzed, among others, rails and wheels. The authors could show that these materials “significantly contributes to increased PM levels” in the subway including iron, chromium, manganese, and nickel (Font et al., 2019). However, it should be mentioned that the authors assume chromium being mainly represented as Cr(III), which is significantly less toxic than Cr(IV). Also in other European cities, manganese, chromium, and copper were found to be elevated in subway material next to the predominant metal iron (Martins et al., 2016). Other studies could show that the inflammatory effect of particles in macrophages is dependent on the exact material, revealing the importance to study the exact cellular mechanisms (Oberdörster, 1995). The method of generating nanoparticles by spark discharge using actual subway materials such as rails is novel. On the other hand, in practice, nanoparticles are also generated by friction. For both electric discharge and friction between two contacting bodies, there is a temperature driven nanoparticle formation. Although the current methodology uniquely allows us to study the toxicity of the contacting friction pairs individually, it does not include the combination of two materials generated by friction. We plan to conduct a further study that includes both the brake material and the collector shoe in contact with the third rail. We are also conducting a study on nanoparticles collected from the Stockholm subway, and it will be interesting to compare the results to those of the present study. The fact that subway materials differ a lot with respect to material from different cities suggests that comparative studies with nanoparticles collected from different cities would be of interest.
By comparing the toxicity of various subway particles, we can assess the influence of nanoparticles specifically and further, also distinguish if specific materials contribute most to the toxicity, like rails or wheels. Noteworthy, another study has shown that most particles measured in a subway train cabin were nanoparticles, majorly from wheels and rails (Lee et al., 2018). This is especially interesting with respect to health policies and the maintenance of subway systems, whereby toxicity studies can impact stakeholders and material decisions.
CRediT authorship contribution statement
Jana Kuhn: Writing – review & editing, Writing – original draft, Visualization, Investigation, Formal analysis. N.V. Srikanth Vallabani: Writing – review & editing, Supervision, Methodology, Investigation. Andrea Montano Montes: Writing – review & editing, Supervision. Ana T. Juárez-Facio: Writing – review & editing, Visualization, Investigation. Micol Introna: Writing – review & editing, Investigation. Sarah S. Steimer: Writing – review & editing, Supervision. Anil Patel: Investigation. Divya Bharathi Manem: Investigation. Bozhena Tsyupa: Investigation. Alessandro Mancini: Writing – review & editing, Supervision. Ulf Olofsson: Writing – review & editing, Supervision, Funding acquisition, Conceptualization. Karine Elihn: Writing – review & editing, Supervision, Funding acquisition, Conceptualization. Hanna L. Karlsson: Writing – review & editing, Visualization, Supervision, Funding acquisition, Conceptualization.
February 12, 2025 at 02:46PM
https://ift.tt/0uRqoi5