Volume 10 Number 9 Environmental September 2023 Pages 2213–2612 Science Nano rsc.li/es-nano ISSN 2051-8153 PAPER Nerea Roher et al. Polystyrene nanoplastics target lysosomes interfering with lipid metabolism through the PPAR system and affecting macrophage functionalization Environmental Science Nano This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online PAPER View Journal | View Issue Polystyrene nanoplastics target lysosomes Cite this: Environ. Sci.: Nano, 2023, interfering with lipid metabolism through the Open Access Article. Published on 14 June 2023. Downloaded on 10/18/2023 1:46:49 PM. 10, 2245 PPAR system and affecting macrophage functionalization† Irene Brandts,‡ab Roger Solà,‡a Marlid Garcia-Ordoñez,a Alex Gella,bd Albert Quintana,bd Beatriz Martin,c Anna Esteve-Codina, ce Mariana Telesab and Nerea Roher *abf Nanoplastics (NPs) are currently a main concern for environmental, animal and human health due to their potential to accumulate in different environmental compartments and provoke effects in living organisms. Nevertheless, neither these effects nor the interaction of NPs with the cellular machinery are well characterized, and only scattered information is available. In the present work, we focused on the interaction between NPs and fish cells, both intestinal cells and macrophages, in order to understand which cell organelles are targeted by polystyrene (PS)-NPs and how this could impact cell function. PS-NPs can pass through phospholipid membranes, entering cells via endocytosis, phagocytosis or passive transport. Once internalized, we found that PS-NPs co-localize with lysosomes but not with mitochondria. Moreover, using two types of fluorescent probe (H2DCFDA and DHE) we demonstrated that NPs did not trigger the production of reactive oxygen species (ROS), which was corroborated by the fact that neither the oxidative consumption ratio (OCR) nor the extracellular acidification rate (ECAR) in mitochondrial respiration were altered. RNASeq data revealed clear interference by PS-NPs with lipid metabolism, peroxisomes and PPAR signaling. The M1/M2 balance critically determines tissue homeostasis when exposed to exogenous agents Received 29th November 2022, such as microorganisms or pollutants. Thus, the expression of different genes (il1β, tnfα, il6, il10, il12, cox2, Accepted 13th June 2023 mmp9, ppar a, b and g) was further assessed to characterize the macrophage phenotype M1 or M2, induced by PS-NPs. Overall, in this study we demonstrate that PS-NPs co-localize within lysosomes, both DOI: 10.1039/d2en01077a in macrophages and in intestinal cells of rainbow trout, but do not trigger ROS production nor alter rsc.li/es-nano mitochondrial respiration. In macrophages, PS-NPs modulate polarization towards the M2-like phenotype. Environmental significance Small-sized plastic particles have been reported around the planet in all kinds of environments. Nevertheless, studies of the effects of nanoplastics (NPs) on vertebrate species are still scarce, probably because nanoplastics are more difficult to detect and to quantify. Few studies evaluate the impact of nanoplastics in relevant cell types, such as intestinal cells that first encounter the nanoplastic in aquatic organisms or in macrophages relevant to mounting a proper immune response also at the mucosal surfaces. Here we dissect the interaction of PS-NPs with cell machinery, and we show how this accumulation, mainly in the lysosomal compartment, interferes with the lipidic metabolism of macrophages and may modify their functional phenotype to a specific M2-like functional phenotype. We propose that modifications on the tissue microenvironment by nanoplastic accumulation may modulate the functional response of macrophages, thus modifying the global immune response. a Institut de Biotecnologia i Biomedicina, Universitat Autònoma de Barcelona, 08193 Barcelona, Spain. E-mail: nerea.roher@uab.cat 1. Introduction b Department de Biologia Cel·lular, Fisiologia i Immunologia, Universitat Autònoma de Barcelona, 08193 Barcelona, Spain Nanoplastics (NPs) are the particle size-fraction under 1000 c CNAG-CRG, Centre for Genomic Regulation, Barcelona Institute of Science and nm of plastic litter,1 and could potentially be the most Technology, 08028 Barcelona, Spain hazardous. They have been found in different environments d Institut de Neurociències, Universitat Autònoma de Barcelona, Spain such as oceans, rivers, soils, mountain glaciers and polar e Universitat Pompeu Fabra (UPF), Barcelona, Spain f ice.2–4 Their nanoscale physicochemical properties allow CIBER de Bioingeniería, Biomateriales y Nanomedicina (CIBER-BBN), Spain † Electronic supplementary information (ESI) available. See DOI: https://doi.org/ them to cross biological barriers, including the intestinal 10.1039/d2en01077a wall,5 reaching and accumulating in different tissues such as ‡ Both authors equally contributed. liver, muscle or brain.6–8 Aquatic organisms are highly This journal is © The Royal Society of Chemistry 2023 Environ. Sci.: Nano, 2023, 10, 2245–2258 | 2245 View Article Online Paper Environmental Science: Nano exposed to NPs, as the external environment directly bathes described both in mammals and in fish,24 and associated internal surfaces. In particular, fish share an intimate with specific pro-inflammatory (M1) and anti-inflammatory relationship between gills, intestine and their water (M2) states and metabolic profiles.25 M1 macrophages surroundings, which represent an extremely wide surface basically rely on glycolytic metabolism, while M2 fuel the This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. interacting with the external milieu. For example, the trout immune response through lipid metabolism. It has been gill surface is estimated to be 0.1–0.4 m2 kg−1 body weight, previously described that nanomaterials can reprogram and the salmon digestive tube is estimated to be 0.675 m2 macrophage phenotypes.26 Moreover, Merkley and coauthors kg−1 body weight.9,10 This pinpoints the digestive tube as the found that PS-NPs can modify the macrophage polarization largest organ surface exposed to external media in fish. In status towards a glycolytic metabolic profile.27 addition, there is a short distance between blood and water Among different manufactured plastic polymers, PS is one (e.g., 6 μm on average in the gill lamellae of rainbow trout)9 of the most commonly fabricated, and is predominantly used Open Access Article. Published on 14 June 2023. Downloaded on 10/18/2023 1:46:49 PM. and between the lumen of the intestine and the lamina in packaging, contributing to a significant fraction of plastic propria of the microvillus. This large area and short distance pollution.28 Moreover, PS has been one of the polymers envisage the easy uptake and wide impact of NPs directly on identified in the nano-fraction of the few real environmental exposed surfaces, as well as subsequent impact on internal samples evaluated to date.2,3 Several studies have reported tissues. In zebrafish larvae, NPs mainly accumulate in the gut effects of PS-NPs in fish and have described alterations in the and pancreas after acute exposure by immersion,11 oxidative status, endocrine disruption, dysregulation of energy highlighting the gastrointestinal system as one of the main metabolism, interference with the immune system, gut portals of entry of NPs. dysbiosis and behavioral alterations.29,30 However, knowledge The intestinal epithelium consists of a single layer of of where NPs accumulate within the cell and how this epithelial cells; thus intestinal cell cultures are excellent accumulation impacts the cell function is scattered and models to understand the final fate of NPs inside the cell sometimes controversial. Different studies have reported the and to understand how the cell machinery copes with this impact of PS-NPs on oxidative status,31,32 mitochondrial emergent pollutant.12,13 At a functional level, and in addition respiration,27 glucose and lipid metabolism33,34 or apoptosis.32 to digesting and absorbing nutrients, the intestine is critical Oxidative stress and ROS production have been suggested for water and electrolyte balance, endocrine regulation of as one of the main mechanisms of NPs effects and are digestion and metabolism, and immunity.14,15 Moreover, the synthesized in large amounts in M1 macrophages to eliminate digestive tube is a complex tissue intimately associated with pathogens.35,36 ROS are byproducts of biological reactions that different immune cell types such as intra-epithelial occur during the processes of respiration in organelles such lymphocytes (IEL) and macrophages.16 Accordingly, there is a as mitochondria and peroxisomes, two organelles that interact large population of resident macrophages within the lamina in a coordinated way in fatty acid and ROS metabolism.37 propria of the intestine, which could also suffer the impact Fatty acid β-oxidation is a multistep process by which fatty of NPs, once these NPs cross the one-cell epithelial barrier. In acyl-CoA esters are shortened. However, mitochondria and addition to disturbing the local immune response, this could peroxisomes contain a distinct set of enzymes for each lead to NPs being transported from the intestine to immune- β-oxidation reaction step. The peroxisomal β-oxidation relevant organs (spleen and head kidney), altering the products can only be fully oxidized to CO2 and H2O after they systemic immune response to pathogens. Previous studies on have been shuttled to mitochondria. Oxidative status the effects of NPs on the mammalian immune system imbalance has attracted the attention of researchers because reported that NPs were detected in macrophages, indicating excessive ROS production leads to oxidative damage and an attempt by the host's immune system to clear and remove activates a variety of stress-sensitive signaling pathways in the NPs.17 Macrophages are dispersed in tissues of both cells. Taking all the above into account, the hypothesis of the lymphoid and non-lymphoid origin and can adapt their present work is that intestinal cells and macrophages attempt functions in response to signals from their to clear PS-NPs using the cellular machinery for lipid uptake microenvironment.18–20 Relevantly, the immune functions of and lipid metabolism, and that this could alter normal cell macrophages can be regulated by lipids, developing into function, potentially affecting macrophage polarization. In this different functional phenotypes depending on the input study, we present exhaustive data on the subcellular fate of stimuli. Macrophages take up lipoproteins (e.g. LDL or PS-NPs in two different and functionally distant cell types that oxLDL) and cholesterol from the extracellular environment are at the forefront of interaction with PS-NPs. We describe via phagocytosis, macropinocytosis and scavenger receptor- the main subcellular organelles targeted by PS-NPs and how mediated pathways,21 processing them into lysosomes and the initial steps of an acute exposure to PS-NPs impact cell mitochondria.22 When the excess of lipids and cholesterol is machinery and specific functions such as lipid metabolism. not eliminated from macrophages, it leads to the formation The resilience of biological systems to environmental of “foam cells”, the accumulation of which is a characteristic perturbation is remarkable, but given that the presence of NPs feature observed in some pathologies: for example, during is extensive in air, water and soil and that NPs can be the development of atherosclerotic plaque in humans.23 M1 incorporated in trophic webs, the long-term impact of this still and M2 are the best characterized macrophage phenotypes, enigmatic contaminant could be extremely important. 2246 | Environ. Sci.: Nano, 2023, 10, 2245–2258 This journal is © The Royal Society of Chemistry 2023 View Article Online Environmental Science: Nano Paper 2. Materials and methods aimed to remove PS-NPs attached to the cell surface. 2.1. Nanoplastic properties and characterization in exposure Subsequently, two volumes of complete medium were added, media and the cells were pelleted by centrifugation at 300×g for 5 min at RT. The pellets were resuspended in PBS or HBSS for This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. Two types of NPs were purchased from Bangs Laboratories flow cytometry (FACSCalibur BD), and 10 000 events were (Fisher, IN, USA): fluorescently labelled PS-NPs (50 nm size – counted. The data was analyzed using FlowJo 10.4 (Leland ref. FSDG001 Dragon Green) and non-labelled PS-NPs (44 nm Stanford University) and plotted with Prism 8.01 (GraphPad size – ref. PS02002). The hydrodynamic size and zeta Prism). potential were assessed by dynamic light scattering (DLS) To obtain spatial information and confirm that the (Zetasizer Pro, Malvern) in ultrapure water and cell culture fluorescent PS-NPs were inside the cells, we performed media (DMEM) at 0 and 24 h. A summary of the PS-NPs confocal microscopy (Zeiss LSM 700). RT-HKM or RTgutGC Open Access Article. Published on 14 June 2023. Downloaded on 10/18/2023 1:46:49 PM. characterization is presented in ESI† Table S1. Full were seeded on Ibidi plates (μ-dish 35 mm) and incubated at characterization information in different conditions and 17 or 20 °C and 5% or 0% CO2, respectively. The next day, corresponding figures were previously published for both the cells at approximately 50–60% confluence were placed in fluorescent PS-NPs11 and the non-labelled PS-NPs.7 minimal media. PS-NPs at 25 μg ml−1 were added 2 h later and the cells were incubated for 16 h at 17 or 20 °C. Prior to 2.2. Cell culture observation in the microscope, the cells were incubated with Rainbow trout head kidney macrophages (RT-HKM) were media containing Lysotracker Red (75 nM) or MitoTracker isolated from Oncorhynchus mykiss (109 ± 18 g body weight) Deep Red (25 nM), for 1 h and 45 min, respectively, as following previously described procedures.38 Primary indicated in the manufacturer's instructions. After staining, adherent cultures were established in DMEM + GlutaMAX, the media was replaced by fresh, prewarmed media. The cells with 10% heat-inactivated FBS and 100 μg ml−1 Primocin were then stained with Hoechst (nuclei) and CellMask Deep (Invitrogen) at 17 °C and 5% CO2. Experiments for NP Red or WGA (membrane) (Life Technologies). The images uptake, protein and gene expression were performed on day were analyzed using Imaris software v8.2.1 (Bitplane) and 5 when the macrophages were fully differentiated. ImageJ (National Institute of Health, USA). RTgutGC cells were obtained from Dr Carolina Tafalla's lab39 and cultured at 20 °C, without CO2 in L-15 medium 2.4. Oxygen consumption measurements with GlutaMAX and supplemented with 10% of FBS and 1% Oxygen consumption rate (OCR) was measured using a penicillin–streptomycin. Sub-confluent cultures (at an Seahorse XFp Extracellular Flux Analyzer (Agilent approximate density of 80%) were detached by adding TrypLE Technologies, Inc), at 21 °C. RT-HKM or RTgutGC cells were Express enzyme (Gibco, cat.12605010) and left to incubate for plated at 1.5 × 105 cells/well or 4 × 104 cells/well, respectively, 5 minutes. Detached cells were split up into new flasks at a and grown for 6 h. Cells were placed in minimal media for 2 1 : 2 seeding ratio. Experiments for NP uptake and gene h and PS-NPs were then added, at 25 μg mL−1, and left to expression were performed at 80% confluency. incubate for 16 h. The cells were assayed in Agilent Seahorse ZFL cells were cultured as previously11 in DMEM + XF DMEM Medium (Agilent Technologies, Inc) supplemented GlutaMAX, with 10% heat-inactivated FBS and 1% penicillin– with 25 mM glucose, 1 mM pyruvate and 4 mM glutamine. streptomycin. First, basal measurements of oxygen consumption rate were obtained. Next, oligomycin (2 mM) was injected, followed by 2.3. Cytometry and confocal microscopy sequential injections of FCCP (5 mM) and antimycin A/ rotenone (1.25 mM and 1 mM, respectively) to disrupt To evaluate the uptake of PS-NPs in cells, RT-HKM, RTgutGC mitochondrial respiration. Altogether, this injection series or ZFL cells were incubated in their respective minimal allowed for the determination of basal respiration, spare media for 2 h, then a range of concentrations of PS-NPs (0, capacity, ATP-linked respiration and maximal respiration. 0.05, 1, 5, 10, 25, 50 and 100 μg mL−1; see figure legend for specific concentrations) was added to the media and left for 16 h. 2.5. Detection of reactive oxygen species To assess whether PS-NP uptake affected the number of To measure ROS, cells were incubated with 10 μM 2′,7′- lysosomes or mitochondria, cells were treated as stated above dichlorofluorescein diacetate (DCFDAH2; ThermoFisher) or and then washed with PBS or HBSS. Cells were then 10 μM dihydroethidium (DHE, Sigma) in the dark at 17 or 21 incubated with media containing Lysotracker/Red (75 nM) or °C for 30 min in the presence of 25 or 50 μg mL−1 PS-NPs (1 Mitotracker Deep Red (25 nM), for 1 h and 45 minutes, h pre-treatment) and/or doxorubicin (10 μM) (45 min) as a respectively, as indicated in the manufacturer's instructions. positive control. Cells then were trypsinised as explained in After treatment and staining, RT-HKM or RTgutGC cells were Section 2.2 for cytometry analysis. A probe control was always washed in PBS or HBSS, respectively, and incubated under included in the experimental design, and ROS production their appropriate conditions (Section 2.1) with 1 mg ml−1 (GeoMean Fluorescence Intensity and fold change) was trypsin (Gibco) for 15 min. This strong trypsinization step calculated using this control. ROS production was measured This journal is © The Royal Society of Chemistry 2023 Environ. Sci.: Nano, 2023, 10, 2245–2258 | 2247 View Article Online Paper Environmental Science: Nano by cytometry using a CytoFLEX LX (Beckman Coulter). ligation product was enriched with 15 PCR cycles. The size Viability was monitored by staining the cells in parallel with and quality of the libraries were assessed in a High BD Via-Probe Red Nucleic Acid Stain (VPR) or Propidium Sensitivity DNA Bioanalyzer assay (Agilent). The libraries Iodide (PI). were sequenced on NovaSeq 6000 (Illumina) in paired-end This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. mode with a read length of 2 × 101 bp, following the 2.6. RNA extraction and Q-PCR manufacturer's protocol for dual indexing. Image analysis, base calling and quality scoring of the run were processed RT-HKM or RTgutGC cells were plated to 70% confluence, using the manufacturer's software Real Time Analysis cultured in minimal media for 2–3 h and then exposed to (NovaSeq 6000 RTA 3.4.4). Illumina readings were mapped 25 μg ml−1 PS-NPs (12 h) and subsequently to Poly I:C or against the O. mykiss reference genome (Omyk_1.0) using LPS (10 μg ml−1, 12 h). Total RNA was extracted using aligner STAR version 2.7.8a42 with ENCODE parameters. Open Access Article. Published on 14 June 2023. Downloaded on 10/18/2023 1:46:49 PM. TriReagent (Sigma) following the manufacturer's Annotated genes were quantified using RSEM version instructions and RNA yield and quality were determined on 1.3.043 with default parameters, using release 104 of O. a Nanodrop ND-1000 (Thermo Fisher Scientific) and mykiss EN-SEMBL annotation. Differential expression integrity was assessed on an Agilent 2100 Bioanalyzer using analysis was performed with the limma v3.42.3 R package, an RNA 6000 Nano Lab-Chip kit (Agilent Technologies). For using TMM normalization. The voom function44 was used RNASeq analysis, see Section 2.7. to estimate the mean–variance relationship and to compute For Q-PCR, RT-HKM and RTgutGC, cells were exposed to observation-level weights. The voom-transformed counts 25 or 50 μg ml−1 PS-NPs (16 h). cDNA was synthesized from 1 were used to fit the linear model and contrasts were μg of high-quality total RNA using an iScript cDNA synthesis extracted. Genes with an adjusted p-value < 0.05 and kit (Bio-Rad). Quantitative real-time PCR (Q-PCR) was absolute fold change |FC| > 1.5 were considered to be performed at 60 °C annealing temperature in a CFX384 touch differentially expressed (DEG). Those genes were used to real-time PCR detection system (Bio-Rad) using an iTaq perform a functional enrichment analysis with gprofiler2 Universal SYBR Green Supermix kit (Bio-Rad). Each PCR v0.1.8.45 To use human databases as a reference, the gorth mixture consisted of 5 μl of SYBR green supermix, 0.4 μM of function from gprofiler2 was used to translate gene specific primers, 2 μl of diluted cDNA and 2.6 μl of water identifiers between organisms (orthologous gene mappings (Sigma-Aldrich) in a final volume of 10 μl. The reference gene are based on the information retrieved from the Ensembl was ef1-α. The dilution factor for all the genes tested was 1 : database). Additionally, gene set enrichment analysis (GSEA) 10 (ppara, pparb, pparg, tnfa, il1b, il6, il12, cox2, il10, mmp9, was performed with a list of genes pre-ranked with a cathepsinD, ef1-a, gas7 or b-actin (ESI† Table S1)). Reference t-statistic and human Reactome database, using the R gene stability was calculated using NormFinder (Andersen package fgsea v1.12.0. et al., 200440) and the best reference gene (gas7) was selected. Amplification efficiencies for the primers were also calculated 2.8. Statistics (ESI† Table S1). All the samples (N = 3 or 6 per treatment) were run in triplicate, and data were analyzed for individual Results are expressed as mean ± standard deviation and replicates using the Livak method.41 statistical data analysis was undertaken using GraphPad Prism 7.0. The data was first checked for normality using the 2.7. RNA Seq Shapiro–Wilk test. Afterwards, different treatments were compared using one-way ANOVA or a t-test depending on the Total RNA was extracted from RT-HKM cells using experimental design. TriReagent (Sigma-Aldrich), as indicated in Section 2.6, and RNA yield and quality were determined on a Nanodrop ND- 3. Results and discussion 1000 (Thermo Fisher Scientific). The samples were quantified with a Qubit® RNA BR Assay kit (Thermo Fisher 3.1. Uptake and accumulation of PS-NPs in macrophages and Scientific) and the RNA integrity was estimated with an RNA intestinal cells 6000 Nano Bioanalyzer 2100 Assay (Agilent). For all samples In this study, we used an in vitro approach to study the RIN was between 9 and 10. The RNA-Seq libraries were interaction of PS-NPs with the cell machinery using two prepared with KAPA Stranded mRNA-Seq Illumina Platforms relevant cell types: primary macrophages isolated from head Kit (Roche) following the manufacturer's recommendations. kidney, RT-HKM, and intestinal cells, RTgutGC. The Briefly, 500 ng of total RNA was used for poly-A fraction internalization of PS-NPs was very efficient and dose enrichment with oligo-dT magnetic beads, following mRNA dependent in both cell types (Fig. 1a–d). However, some fragmentation. Strand specificity was achieved during the differences were observed regarding the total fluorescent second strand synthesis performed in the presence of dUTP signal intensity (MFI). RTgutGC cells (Fig. 1c) were able to instead of dTTP. The blunt-ended double-stranded cDNA internalize more fluorescent PS-NPs than RT-HKM (Fig. 1a) at was 3′ adenylated and Illumina platform compatible a given dose (e.g. at 25 μg ml−1 HKM they reach an MFI value adaptors with unique dual indexes and unique molecular of 1000 while RTgutGC reach 4000). The uptake of PS-NPs in identifiers (Integrated DNA Technologies) were ligated. The zebrafish liver cells (ZFL) has been described in detail 2248 | Environ. Sci.: Nano, 2023, 10, 2245–2258 This journal is © The Royal Society of Chemistry 2023 View Article Online Environmental Science: Nano Paper This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. Open Access Article. Published on 14 June 2023. Downloaded on 10/18/2023 1:46:49 PM. Fig. 1 PS-NP uptake by HKM, RTgutGC and ZFL cells. Uptake of fluorescent PS-NPs by HKM (a and b), RTgutGC (c and d) and ZFL (e and f). The mean fluorescence intensity (MFI) of the dose–response (a, c and e) and the percentage of positive cells (b, d and f) are shown. Cells were incubated for 16 h with fluorescent PS-NPs at different concentrations (0.05, 1, 10, 25, 50 and 100 μg ml−1). Data show the mean ± standard deviation (SD). Each dot represents an independent experiment (n = 3 to 6). This journal is © The Royal Society of Chemistry 2023 Environ. Sci.: Nano, 2023, 10, 2245–2258 | 2249 View Article Online Paper Environmental Science: Nano previously11 and, when compared to RT-HKM and RTgutGC that specifically label lysosomes and mitochondria (see cells, ZFL showed an intermediate capacity of internalization Section 2.2). As shown in Fig. 2, PS-NPs accumulated within (at 25 μg ml−1 we observed an MFI of 3000; Fig. 1e). The the cytosol in both RT-HKM and RTgutGC cells, positive fluorescent cell percentage reaches around 100% at 5 predominantly co-localizing with lysosomes (Fig. 2, right- This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. μg ml−1 in all cell types (Fig. 1b–f). Differences in uptake hand panels, white arrows). Confocal images corroborated could be attributed to functional differences among cell the differential ability of macrophages and intestinal cells to types, as RTgutGC is a cell line derived from epithelial take up PS-NPs in a qualitative way, with RTgutGC cells intestinal cells, which have the main function of showing higher PS-NP accumulation (Fig. 2). We used incorporating substances present in the intestinal lumen. No MitoTracker to investigate co-localization of PS-NPs in the associated toxicity was found at the assayed doses and time mitochondria, as this organelle has been claimed to be one in any of the cell types (ESI† Fig. S1). PS-NP toxicity has been of the main targets of PS-NPs, where they would interfere Open Access Article. Published on 14 June 2023. Downloaded on 10/18/2023 1:46:49 PM. reported at higher doses in different cell lines: up to 75 μg with the mitochondrial function by inducing the production ml−1 in ZFL,11 up to 200 μg ml−1 in Caco-246 and up to 500 μg of ROS.48 However, we did not observe any trace of co- ml−1 in THP-1.47 localization (Fig. 3, composites 1 and 2). To further characterize the interaction of PS-NPs with both lysosomes and mitochondria, we used cytometry to assess whether 3.2. PS-NPs localize within the lysosomes in intestinal cells incubation with PS-NPs would modify the abundance of and macrophages and modulate lysosomal abundance either of the two organelles. We detected a clear and We investigated the intracellular fate of PS-NPs in RTgutGC significant increase in the fluorescent intensity due to and RT-HKM by using flow cytometry to obtain quantitative LysoTracker and no modification of the fluorescent signal information and confocal microscopy to gain morphological/ due to MitoTracker (Fig. 4a), indicating an increase in dimensional information, and combining fluorescent dyes lysosomes after cellular exposure to PS-NPs, but no changes Fig. 2 Confocal imaging of HKM (upper panel) and RTgutGC (lower panel) showing the internalization and co-localization of fluorescent PS-NPs, after 16 h of incubation at 25 μg ml−1 PS-NPs. Magenta corresponds to cell mask staining membranes; blue to Hoechst (nucleus); red to LysoTracker (lysosomes) and green to dragon green (PS-NPs). White arrows point to co-localization of PS-NPs and lysosomes. 2250 | Environ. Sci.: Nano, 2023, 10, 2245–2258 This journal is © The Royal Society of Chemistry 2023 View Article Online Environmental Science: Nano Paper This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. Open Access Article. Published on 14 June 2023. Downloaded on 10/18/2023 1:46:49 PM. Fig. 3 Confocal imaging of fluorescent PS-NP uptake by HKM and organelle staining. Representative confocal microscopy imaging of HKM showing the internalization and co-localization of fluorescent PS-NPs, after 16 h of incubation at 25 μg ml−1 PS-NPs. Magenta fluorescence corresponds to cell mask staining membranes; blue to Hoechst (nucleus); red to MitoTracker (mitochondria) and green to dragon green (PS-NPs). in mitochondria. Cytometry data was consistent with status, even if not through direct interaction with microscopy, where strong staining of mitochondria and more mitochondria, we measured ROS production using a discrete and less abundant staining of lysosomes were combination of fluorescent chemical probes and cytometry. observed in HKM (Fig. 2 and 3). The gating strategy and This approach allowed us to evaluate ROS production and representative flow cytometry plots of HKM treated with cell viability in parallel, to discard uncontrolled side effects. fluorescent PS-NPs and stained with Mito- and LysoTracker We assessed ROS levels using a DHE probe for HKM and a are shown in Fig. 4b. Moreover, the confocal images shown DCFDH2 probe for RTgutGC (see Section 2.4). As a positive in Fig. 3 also point to an increase in the number of control for ROS production, we used doxorubicin as a lysosomes (LysoTracker) when the cells were treated with PS- consistent ROS-inducing stimulus.51 It is worth mentioning NPs, compared with LysoTracker staining in control cells that we also carried out these assays with TBH as a positive (Fig. 2). control for ROS production, but this compound did not consistently induce ROS in both intestinal cells and macrophages (data not shown). We set up a specific ROS 3.3. Mitochondrial respiration rates and ROS production in assay for each cell line (DHE for HKM and DCFDH2 for intestinal cells and head kidney macrophages are not altered RTgutGC) using cytometry, and viability was monitored in by PS-NPs parallel, staining the cells with PI or VPR. We did not detect To explore the direct interaction of PS-NPs with ROS production due to PS-NP treatment in HKM or in mitochondria, we assessed mitochondrial respiration in RT- RTgutGC cells (Fig. 6a and b, respectively). These results HKM and RTgutGC. The mitochondrial oxygen consumption demonstrated that PS-NPs did not trigger the production of rate (OCR) was assessed by using a Seahorse XFp extracellular ROS, under present conditions, which was corroborated by flux analyzer (Fig. 5). OCR results revealed that respiration the fact that there were no alterations in the oxidative was similar for RT-HKM (Fig. 5a and b) and RTgutGC consumption ratio (OCR) or extracellular acidification rate (Fig. 5c and d) cells treated or not with 25 μg mL−1 of PS-NPs. (ECAR), either in HKM or in intestinal RTGut cells. A more detailed analysis of mitochondrial respiration revealed that basal respiration, spare capacity, ATP-linked respiration and maximal respiration were unaffected by the 3.4. PS-NPs modulate the lipid metabolism in macrophages presence of PS-NPs in both RT-HKM (Fig. 5b) and RTgutGC through the PPAR system and promote an M2-like cells (Fig. 5d). polarization state This result contrasts with previous studies that reported Given that RTgutGC cells and HKM showed the same type modifications in mitochondrial respiration in human liver of interaction with the cell machinery after PS-NP treatment, and lung cells after exposure to PS-NPs, at similar doses we decided to further investigate the functional response of and times to those used in this study.49,50 This could be macrophages to PS-NP exposure by using RNASeq. We explained by the fact that not every cell type may confront planned an experimental design that would allow us to the same external factor in an identical way and may need characterize both the macrophage response to PS-NPs and different doses or times to reach the damage threshold. To the response to PS-NPs combined with immune stimuli, by further evaluate whether PS-NPs would alter cell oxidative using bacterial (LPS) and viral (Poly(I:C)) mimics. This journal is © The Royal Society of Chemistry 2023 Environ. Sci.: Nano, 2023, 10, 2245–2258 | 2251 View Article Online Paper Environmental Science: Nano This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. Open Access Article. Published on 14 June 2023. Downloaded on 10/18/2023 1:46:49 PM. Fig. 4 Uptake of fluorescent PS-NPs by HKM and organelle staining. HKM were stained with Lyso- and MitoTracker and exposed to 25 μg ml−1 PS-NPs for 16 h. (a) Differences in GeoMean fluorescence intensity between treatments (PS-NPs, PS-NPs + Lyso/MitoTracker). Data show the mean ± standard deviation and significant differences were assessed using a t-test with Welch's correction and indicated as: *p < 0.05, **p < 0.01. Lyso/ MitoTracker and Lyso/MitoTracker + PS-NPs channels (FIT-C, APC-A, PE-A). (b) Representative dot plots of PS-NPs and Lyso/MitoTracker staining of HKM by cytometry. Macrophages showed high transcriptional plasticity when functional categories were observed (Fig. 8a). Moreover, they were treated with both bacterial and viral-like stimuli, when macrophages were treated with PS-NPs alone, we presenting large changes in gene expression: 5354 altered observed significant up- or down-regulation in only 93 genes in LPS-stimulated macrophages and 7365 genes in genes, most of them coding for enzymes or transporters Poly(I:C)-treated macrophages (adjusted p = 0.05). However, involved in lipid transport and catabolism (Fig. 7 and 8b). the pre-exposure of macrophages to PS-NPs did not have When we analyzed the functional enrichment in this gene any significant impact on the global gene expression list using either Reactome or KEGG databases, we observed response to subsequent LPS or Poly(I:C) stimulation, as functional categories related to lipid catabolism, peroxisome similar percentages of up- and down-regulated genes and and PPAR signaling as the most significatively enriched 2252 | Environ. Sci.: Nano, 2023, 10, 2245–2258 This journal is © The Royal Society of Chemistry 2023 View Article Online Environmental Science: Nano Paper This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. Open Access Article. Published on 14 June 2023. Downloaded on 10/18/2023 1:46:49 PM. Fig. 5 Mitochondrial respiration rates of HKM and RTgutGC cells. Representative pharmacological profile of oxygen consumption rate (OCR) from RT-HKM (a) and RTgutGC (c) cells. Respiratory capacities of RT-HKM (b) and RTgutGC (d) cells (calculated from Fig. 5a and c, respectively). Data are presented as the mean ± SEM with n = 3 in each group. (Fig. 8b). As PS is a neutral chemically inert polymer of no ROS production and Arg2 and PPARβ expression (see the lipophilic nature, it is possible that PS-NPs are detected by summary in ESI† Table S3). Our RNASeq data indicates a cells as lipid entities, therefore activating the cell machinery clear induction of FA oxidation associated also with changes responsible for lipid catabolism. As previously stated, in the expression of the PPAR signaling module (Fig. 8). As macrophages can be functionally divided into at least two can be seen in Fig. 9, a Q-PCR profiling of trout macrophages different phenotypes, – a pro-inflammatory phenotype (M1) after PS-NP exposure shows that pro-inflammatory cytokines and an anti-inflammatory phenotype (M2), but this main (e.g. il1b, il12 or tnfa) were not up-regulated while the main characterization has begun to be updated with new subsets anti-inflammatory cytokine, il-10, was up-regulated. The of functional phenotypes influenced by the expression of pparβ was also up-regulated, while pparγ was microenvironment. For example, macrophages infiltrated in down-regulated and no pparα expression was detected. tumors have a specific functional profile related to this Odegaard et al. showed that disruption of PPARγ in myeloid microenvironment.18 Moreover, it has been described that cells impaired M2 macrophage activation,53 supported by the nanomaterials can also modify the macrophage polarization data of Wu and coauthors, where a PPAR-γ antagonist favors status.52 M1 macrophages are mainly characterized by a the M1 phenotype.54 This modulation of PPAR-γ could high expression of pro-inflammatory cytokines, glycolytic influence macrophage polarization, as well as affecting lipid metabolism, ROS production and PPARα, while M2 metabolism. The expression of cathepsinD and mmp9 was also macrophages are characterized by low expression of pro- up-regulated in macrophages after PS-NP treatment. The inflammatory cytokines, high expression of anti- MMP9 protein is involved in the breakdown of the inflammatory cytokines, lipidic metabolism (FA oxidation), extracellular matrix in normal physiological processes, as well This journal is © The Royal Society of Chemistry 2023 Environ. Sci.: Nano, 2023, 10, 2245–2258 | 2253 View Article Online Paper Environmental Science: Nano This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. Open Access Article. Published on 14 June 2023. Downloaded on 10/18/2023 1:46:49 PM. Fig. 6 ROS production in RT-HKM and RTgutGC. DHE (RT-HKM) and H2DCFDA (RTgutGC). Fold change of mean fluorescence intensity (MFI) with respect to control plus probe in HKM (a) and RTgutGC (b). Histograms represent the GeoMean fluorescence intensity in each channel with only PS-NPs exposure (25 μg ml−1 or 50 μg ml−1) and PS-NPs plus 10 μM doxorubicin (DOX). Significant differences with the control were evaluated through one-way ANOVA and represented as: *p < 0.05. as in pathological processes and mmp9 expression is Concluding remarks associated with M2 polarization.55 Cathepsin D is a ubiquitous lysosomal protease that degrades proteins in PS-NPs directly interact with lysosomes, as shown by their co- lysosomal compartments and inactivates different localization as well as the increase in lysosomes after chemokines, such as macrophage inflammatory protein-1 exposure to PS-NPs. Lysosomes have a key lipid sorting alpha (Chemokine (C–C motif) ligand (CCL) 3), CCL4 and function derived from their ability to process and sort CCL21.56 The expression of cathepsinD has been associated exogenous and endogenous lipids, and the presence of PS- with the M1 phenotype (ESI† Table S3). On the other hand, NPs in this organelle could impact lipid metabolism. the RNASeq data showed up-regulation of ccl25b (Fig. 7), Moreover, as shown by RNASeq data, PS-NPs significantly which has been related to the polarization of macrophages alter the expression of transcripts related to lipid from M1 to M2 phenotype in a tumor microenvironment metabolism, as well as to PPAR pathways. PPARs play a (TME).18 Moreover, the CCR9–CCL25 chemokine axis also relevant role in intracellular lipid metabolism by up- strongly correlates with the up-regulation activity of MMP regulating the expression of enzymes involved in the proteins,57 which matches the up-regulation of mmp9 gene conversion of fatty acids into acyl-coA esters. Even though expression found in the present study (ESI† Table S3). The our data show no trace of the impact of PS-NPs in anti-inflammatory profile, together with induced FA oxidation mitochondrial function, they could be affecting mitochondria could suggest that trout macrophages exposed to PS-NPs in an indirect way, as suggested by changes in the expression polarized towards a M2-like phenotype, but we observe some of peroxisomal receptors. We might hypothesize that the important differences in this trend, such as down-regulation lipophilic nature of PS-NPs could suggest that they are of pparγ and the increase in ccl25 and cathepsinD expression. somehow partially confounded by lipid substances inside the Due to these significant differences, we propose that PS-NPs cell, either by their own nature or by the biomolecules that induce a specific polarization state associated with PS-NP might be bound to the NPs, and therefore activate the exposure as a distinct microenvironment. cellular machinery and pathway related to lipid metabolism. 2254 | Environ. Sci.: Nano, 2023, 10, 2245–2258 This journal is © The Royal Society of Chemistry 2023 View Article Online Environmental Science: Nano Paper This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. Open Access Article. Published on 14 June 2023. Downloaded on 10/18/2023 1:46:49 PM. Fig. 7 Transcriptome analysis showing the effects of PS-NPs (25 μg ml−1) on RT-HKM. The heatmap represents the transcript abundance of all DEGs. Our data show a series of gene expression changes in fully match the previously characterized M1 and M2 macrophages after exposure to PS-NPs. However, they do not macrophage phenotypes, although key characteristics from This journal is © The Royal Society of Chemistry 2023 Environ. Sci.: Nano, 2023, 10, 2245–2258 | 2255 View Article Online Paper Environmental Science: Nano This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. Open Access Article. Published on 14 June 2023. Downloaded on 10/18/2023 1:46:49 PM. Fig. 8 Transcriptome analysis of the effects of PS-NPs and PS-NPs combined with immune stimuli on HKM. (a) Bar chart showing the total amount of DEGs (up- or down-regulated) under different treatments with LPS and poly(I:C). (b) Main affected pathways and genes after PS-NP exposure. The Y-axis represents the genes and pathways, and the X-axis the fold change. both profiles are displayed. As the macrophagic phenotype is dependent on the lipidic substrates, we believe that a specific determined by their microenvironment, which is highly macrophage phenotype is attained after PS-NP exposure. 2256 | Environ. Sci.: Nano, 2023, 10, 2245–2258 This journal is © The Royal Society of Chemistry 2023 View Article Online Environmental Science: Nano Paper This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. Open Access Article. Published on 14 June 2023. Downloaded on 10/18/2023 1:46:49 PM. Fig. 9 Analysis of gene expression in HKM after PS-NPs treatment. Cells were incubated with PS-NPs (25 μg ml−1 and 50 μg ml−1) for 16 h. gas7 was used as a reference gene for PPARs, il6, il10, il12 and mmp9. β-Actin was used as a reference gene for cox2, tnfa, il1b and cath d. Data show fold change mean ± standard deviation (SD) and significant differences with the control are represented as: *p < 0.05. Author contributions AQ and PT17/0009/0019 (ISCIII/MINECO and FEDER) to AEC. MT is supported by a Ramon y Cajal contract (ref. RYC2019- IB, RS, MG, AG and NR performed the experiments; NR 026841-I) and IB is supported by a PhD grant from Generalitat designed the experiments and IB, MG, RS, AG, MT and NR de Catalunya (2018FI_B_00711). analysed the data; RS, NR, BM and AC analysed the RNASeq data; AQ provided reagents and machinery for the Seahorse experiments; NR wrote the manuscript draft; NR conceived References ideas and NR and MT oversaw the research. All authors were 1 N. B. Hartmann, T. Hüffer, R. C. Thompson, M. Hassellöv, A. involved in discussions and contributed to the final writing Verschoor, A. E. Daugaard, S. Rist, T. Karlsson, N. Brennholt, of the manuscript. M. Cole, M. P. Herrling, M. C. Hess, N. P. Ivleva, A. L. Lusher and M. Wagner, Environ. Sci. Technol., 2019, 53, 1039–1047. Conflicts of interest 2 A. T. Halle, L. Jeanneau, M. Martignac, E. Jardé, B. Pedrono, The authors declare that they have no known competing L. Brach and J. Gigault, Environ. Sci. Technol., 2017, 51, financial interests or personal relationships that could have 13689–13697. appeared to influence the work reported in this paper. 3 M. Llorca, A. Vega-Herrera, G. Schirinzi, K. Savva, E. Abad and M. Farré, J. Hazard. Mater., 2021, 404, 124022. Acknowledgements 4 D. Materić, R. Holzinger and H. Niemann, Sci. Total Environ., 2022, 846, 157371. This research was supported through Plan Nacional de 5 P. Stapleton, AIMS Environ. Sci., 2019, 6, 367–378. Investigación: RTI2018-096957-B-C21 MINECO/FEDER and 6 X. Fan, X. Wei, H. Hu, B. Zhang, D. Yang, H. Du, R. Zhu, X. 2017-SGR-695 (AGAUR) to NR, and PID2020-114977RB-I00 to Sun, Y. Oh and N. Gu, Chemosphere, 2022, 288, 132607. This journal is © The Royal Society of Chemistry 2023 Environ. Sci.: Nano, 2023, 10, 2245–2258 | 2257 View Article Online Paper Environmental Science: Nano 7 I. Brandts, M. Cánovas, A. Tvarijonaviciute, M. Llorca, A. 34 W. Lai, D. Xu, J. Li, Z. Wang, Y. Ding, X. Wang, X. Li, N. Xu, Vega, M. Farré, J. Pastor, N. Roher and M. Teles, Environ. Res., K. Mai and Q. Ai, J. Hazard. Mater., 2021, 419, 126454. 2022, 212, 113433. 35 A. Covarrubias, V. Byles and T. Horng, Cell Res., 2013, 23, 8 M. Teng, X. Zhao, C. Wang, C. Wang, J. C. White, W. Zhao, L. 984–985. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. Zhou, M. Duan and F. Wu, ACS Nano, 2022, 16, 8190–8204. 36 Q. Hu, H. Wang, C. He, Y. Jin and Z. Fu, Environ. Pollut., 9 E. O. Koppang, A. Kvellestad and U. Fischer, Fish mucosal 2021, 269, 116075. immunity: gill, in Mucosal Health in Aquaculture, 2015, pp. 37 M. Fransen, C. Lismont and P. Walton, Int. J. Mol. Sci., 93–133, DOI: 10.1016/B978-0-12-417186-2.00005-4. 2017, 18, 1126. 10 E. D. Stevens, G. N. Wagner and A. Sutterlin, J. Fish Biol., 38 N. Roher, A. Callol, J. V. Planas, F. W. Goetz and S. A. 1999, 55, 517–526. MacKenzie, Innate Immun., 2011, 17, 16–28. 11 I. Brandts, M. Garcia-Ordoñez, L. Tort, M. Teles and N. Roher, 39 A. Kawano, C. Haiduk, K. Schirmer, R. Hanner, L. E. J. Lee, B. Open Access Article. Published on 14 June 2023. Downloaded on 10/18/2023 1:46:49 PM. Environ. Sci.: Nano, 2020, 7, 2410–2422. Dixon and N. C. Bols, Aquacult. Nutr., 2011, 17, e241–e252. 12 M. Minghetti, C. Drieschner, N. Bramaz, H. Schug and K. 40 C. L. Andersen, J. L. Jensen and T. F. Ørntoft, Cancer Res., Schirmer, Cell Biol. Toxicol., 2017, 33, 539–555. 2004, 64, 5245–5250. 13 M. Revel, C. Roman and A. Châtel, Ecotoxicology, 2021, 30, 41 K. J. Livak and T. D. Schmittgen, Methods, 2001, 25, 402–408. 421–430. 42 A. Dobin and T. R. Gingeras, Comment on “TopHat2: accurate 14 W. H. Karasov and A. E. Douglas, Compr. Physiol., 2013, 3, alignment of transcriptomes in the presence of insertions, 741–783. deletions and gene fusions” by Kim et al., DOI: 10.1101/000851. 15 R. K. Buddington, A. Krogdahl and A. M. Bakke-Mckellep, 43 B. Li and C. N. Dewey, BMC Bioinf., 2011, 12, 323. Acta Physiol. Scand., Suppl., 1997, 638, 67–80. 44 C. W. Law, Y. Chen, W. Shi and G. K. Smyth, Genome Biol., 16 D. Parra, F. E. Reyes-Lopez and L. Tort, Front. Immunol., 2014, 15, R29. 2015, 6, 354. 45 L. Kolberg, U. Raudvere, I. Kuzmin, J. Vilo and H. Peterson, 17 K. Kenesei, K. Murali, Á. Czéh, J. Piella, V. Puntes and E. F1000Research, 2020, 9(ELIXIR), 709. Madarász, J. Nanobiotechnol., 2016, 14, 55. 46 J. Domenech, M. de Britto, A. Velázquez, S. Pastor, A. 18 C. Zhang, M. Yang and A. C. Ericsson, Front. Immunol., Hernández, R. Marcos and C. Cortés, Biomolecules, 2021, 11, 2021, 12, 620510. 1442. 19 D. M. Mosser and J. P. Edwards, Nat. Rev. Immunol., 2008, 8, 47 I. Florance, N. Chandrasekaran, P. M. Gopinath and A. 958–969. Mukherjee, Ecotoxicol. Environ. Saf., 2022, 238, 113612. 20 P. J. Murray and T. A. Wynn, Nat. Rev. Immunol., 2011, 11, 48 L. Rubio, I. Barguilla, J. Domenech, R. Marcos and A. 723–737. Hernández, J. Hazard. Mater., 2020, 398, 122900. 21 I. Tabas and K. E. Bornfeldt, Circ. Res., 2016, 118, 653–667. 49 S. Lin, H. Zhang, C. Wang, X.-L. Su, Y. Song, P. Wu, Z. Yang, 22 J. Yan and T. Horng, Trends Cell Biol., 2020, 30, 979–989. M.-H. Wong, Z. Cai and C. Zheng, Environ. Sci. Technol., 23 A. Remmerie and C. L. Scott, Cell. Immunol., 2018, 330, 27–42. 2022, 56, 12483–12493. 24 M. Forlenza, I. R. Fink, G. Raes and G. F. Wiegertjes, Dev. 50 S. E. Lee, Y. Yi, S. Moon, H. Yoon and Y. S. Park, Metabolites, Comp. Immunol., 2011, 35, 1246–1255. 2022, 12, 897. 25 A. S. Wentzel, J. J. E. Janssen, V. C. J. de Boer, W. G. van Veen, M. 51 T. A. Jacinto, G. S. Meireles, A. T. Dias, R. Aires, M. L. Porto, Forlenza and G. F. Wiegertjes, Front. Immunol., 2020, 11, 152. A. L. Gava, E. C. Vasquez, T. M. C. Pereira, B. P. Campagnaro 26 X. Miao, X. Leng and Q. Zhang, Int. J. Mol. Sci., 2017, 18, 336. and S. S. Meyrelles, Biol. Res., 2018, 51, 33. 27 S. D. Merkley, H. C. Moss, S. M. Goodfellow, C. L. Ling, J. L. 52 D. Reichel, M. Tripathi and J. M. Perez, Nanotheranostics, Meyer-Hagen, J. Weaver, M. J. Campen and E. F. Castillo, Cell 2019, 3, 66–88. Biol. Toxicol., 2022, 38, 31–41. 53 J. I. Odegaard, R. R. Ricardo-Gonzalez, M. H. Goforth, C. R. 28 W. P. de Haan, A. Sanchez-Vidal, M. Canals and N. S. S. Party, Morel, V. Subramanian, L. Mukundan, A. R. Eagle, D. Vats, F. Mar. Pollut. Bull., 2019, 140, 523–535. Brombacher, A. W. Ferrante and A. Chawla, Nature, 2007, 447, 29 C. Barría, I. Brandts, L. Tort, M. Oliveira and M. Teles, Mar. 1116–1120. Pollut. Bull., 2020, 151, 110791. 54 H. Wu, X. Ni, Q. Xu, Q. Wang, X. Li and J. Hua, 30 H. Jacob, M. Besson, P. W. Swarzenski, D. Lecchini and M. J. Gastroenterol. Hepatol., 2020, 35, 1998–2008. Metian, Environ. Sci. Technol., 2020, 54, 4733–4745. 55 E. Zajac, B. Schweighofer, T. A. Kupriyanova, A. Juncker- 31 Z. Liu, Y. Huang, Y. Jiao, Q. Chen, D. Wu, P. Yu, Y. Li, M. Cai Jensen, P. Minder, J. P. Quigley and E. I. Deryugina, Blood, and Y. Zhao, Aquat. Toxicol., 2020, 220, 105420. 2013, 122, 4054–4067. 32 H. Wu, J. Guo, Y. Yao and S. Xu, Fish Shellfish Immunol., 56 M. Wolf, I. Clark-Lewis, C. Buri, H. Langen, M. Lis and L. 2022, 125, 1–8. Mazzucchelli, Am. J. Pathol., 2003, 162, 1183–1190. 33 N. R. Brun, P. van Hage, E. R. Hunting, A.-P. G. Haramis, S. C. 57 H. Mir and S. Singh, CCL25 signalling in the tumor Vink, M. G. Vijver, M. J. M. Schaaf and C. Tudorache, environment, in Tumor environment. The role of chemokines – Commun. Biol., 2019, 2, 382. Part B, 2021. 2258 | Environ. Sci.: Nano, 2023, 10, 2245–2258 This journal is © The Royal Society of Chemistry 2023