Journal of Hazardous Materials 435 (2022) 129053 Contents lists available at ScienceDirect Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat Research Paper Hazard assessment of abraded thermoplastic composites reinforced with reduced graphene oxide Savvina Chortarea a, Ogul Can Kuru a, Woranan Netkueakul a, Marco Pelin b, Sandeep Keshavan c, Zhengmei Song d, Baojin Ma d, Julio Gómes e, Elvira Villaro Abalos f, Luis Augusto Visani de Luna g, h, i, Thomas Loret g, h, i, Alexander Fordham g, h, i, Matthew Drummond g, h, Nikolaos Kontis j, George Anagnostopoulos j, George Paterakis j, Pietro Cataldi h, l, Aurelia Tubaro b, Costas Galiotis j, k, Ian Kinloch h, l, Bengt Fadeel c, Cyrill Bussy g, h, i, Kostas Kostarelos g, h, m, Tina Buerki-Thurnherr a, Maurizio Prato n, o, p, Alberto Bianco d, Peter Wick a, * a Swiss Federal Laboratories for Materials Science and Technology (Empa), Laboratory for Particles-Biology Interactions, 9014 St. Gallen, Switzerland b Department of Life Sciences, University of Trieste, 34127 Trieste, Italy c Nanosafety & Nanomedicine Laboratory, Institute of Environmental Medicine, Karolinska Institutet, 177 77 Stockholm, Sweden d CNRS, Immunology, Immunopathology and Therapeutic Chemistry, UPR3572, University of Strasbourg, ISIS, 67000 Strasbourg, France e Avanzare Innovacion Tecnologica S.L. 26370 Navarrete, Spain f Instituto de Tecnologías Químicas de La Rioja (InterQuímica), 26370 Navarrete, Spain g Nanomedicine Lab, Faculty of Biology, Medicine & Health, University of Manchester, Manchester M13 9PT, United Kingdom h National Graphene Institute, University of Manchester, Manchester M13 9PL, United Kingdom i Lydia Becker Institute of Immunology and Inflammation, Faculty of Biology, Medicine and Health, University of Manchester, Manchester Academic Health Science Centre, Manchester M13 9PL, United Kingdom j Institute of Chemical Engineering Sciences, Foundation of Research and Technology-Hellas (FORTH/ICE-HT), 26504 Patras, Greece k Department of Chemical Engineering, University of Patras, 26504 Patras, Greece l Department of Materials, University of Manchester, Manchester M13 9PL, United Kingdom m Catalan Institute of Nanoscience and Nanotechnology (ICN2), and Barcelona Institute of Science and Technology (BIST), Barcelona 08193, Spain n Department of Chemical and Pharmaceutical Sciences, University of Trieste, 34127 Trieste, Italy o Center for Cooperative Research in Biomaterials (CIC biomaGUNE), Basque Research and Technology Alliance (BRTA), 20014 Donostia San Sebastián, Spain p Basque Foundation for Science (IKERBASQUE), 48013 Bilbao, Spain H I G H L I G H T S G R A P H I C A L A B S T R A C T • Graphene related materials are exten­ sively used to reinforce polymers. • Abrasion of composite materials is an important end-of-life scenario to be assessed prior to their commercial use. • This study provides a qualitative and quantitative analysis of the released materials after abrasion. • The hazard of abraded thermoplastics is assessed focusing on the most likely exposure routes. • Life cycle considerations and impact of reduced graphene oxide in the context of reinforced polymers are addressed. Abbreviations: GRM, graphene related materials; PA6, Polyamide 6; rGO, reduced graphene oxide. * Correspondence to: Swiss Laboratories for Materials Science and Technology, Lerchenfeldstrasse 5, CH-9014 St. Gallen, Switzerland. E-mail address: peter.wick@empa.ch (P. Wick). https://doi.org/10.1016/j.jhazmat.2022.129053 Received 13 February 2022; Received in revised form 22 April 2022; Accepted 29 April 2022 Available online 5 May 2022 0304-3894/© 2022 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by- nc-nd/4.0/). S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 A R T I C L E I N F O A B S T R A C T Editor: Dr. S Nan Graphene-related materials (GRMs) are subject to intensive investigations and considerable progress has been made in recent years in terms of safety assessment. However, limited information is available concerning the Keywords: hazard potential of GRM-containing products such as graphene-reinforced composites. In the present study, we Graphene-related materials conducted a comprehensive investigation of the potential biological effects of particles released through an Thermoplastic polymer composites abrasion process from reduced graphene oxide (rGO)-reinforced composites of polyamide 6 (PA6), a widely used Hazard assessment engineered thermoplastic polymer, in comparison to as-produced rGO. First, a panel of well-established in vitro models, representative of the immune system and possible target organs such as the lungs, the gut, and the skin, was applied. Limited responses to PA6-rGO exposure were found in the different in vitro models. Only as- produced rGO induced substantial adverse effects, in particular in macrophages. Since inhalation of airborne materials is a key occupational concern, we then sought to test whether the in vitro responses noted for these materials would translate into adverse effects in vivo. To this end, the response at 1, 7 and 28 days after a single pulmonary exposure was evaluated in mice. In agreement with the in vitro data, PA6-rGO induced a modest and transient pulmonary inflammation, resolved by day 28. In contrast, rGO induced a longer-lasting, albeit mod­ erate inflammation that did not lead to tissue remodeling within 28 days. Taken together, the present study suggests a negligible impact on human health under acute exposure conditions of GRM fillers such as rGO when released from composites at doses expected at the workplace. Graphene and its derivatives (graphene-related materials or GRMs) weak mechanical performance in some areas, which can all be consid­ have been among the fastest growing areas of nanoscience and tech­ erably improved by the use of GRMs as nanofillers (Xu and Gao, 2010; nology over the past decade. This atomically thin, two-dimensional form Pant et al., 2013; O’Neill et al., 2014). of carbon has generated considerable excitement since its initial dis­ Although GRMs may provide solutions to current challenges in many covery. Due to its outstanding physicochemical properties, namely, fields, the inevitable occupational or consumer exposure to these ma­ excellent stretchability (20% of its initial length), high intrinsic me­ terials has raised scientific and public concern for their potential to chanical stiffness and strength (1 TPa and 130 GPa, respectively), affect human health. As a result, various studies have been initiated to extraordinary electrical conductivity (mobility of charge carriers assess the safety of GRMs in different biological systems, with the re­ 200,000 cm2/ V/s), large surface area (2630 m2/g), and superior ther­ ported investigations often showing contradictory or inconclusive re­ mal conductivity (~5000 W/mK), it has raised high expectations on sults. Several cellular and animal studies with different experimental future applications (George et al., 2018; Novoselov et al., 2012; Geim setups have shown that different types of GRMs can induce mitochon­ and Novoselov, 2007). Many of these properties are superior to those of drial dysfunction, cell apoptosis, genotoxicity, inflammatory or oxida­ other materials, but most importantly, the combination of all these tive stress responses (Mittal et al., 2016; Chen et al., 2012; Li et al., 2018; characteristics is exceptional. It is of no surprise that GRMs are currently Liu et al., 2013), while others reported the absence of adverse effects being explored for a multitude of applications, such as in electronics, (Kucki et al., 2016, 2018; Drasler et al., 2018) as reviewed by Fadeel energy, photonics, composite, filtration, sensors, or biomedicine (Ferrari et al. (2018) Overall, this diversity in research outcomes indicate that et al., 2015). As a result of the increased demand, production of GRMs GRMs are an incredibly diverse family of materials, whose safety has increased from 14 tons in 2009 to nearly 1200 tons per year within thresholds is dependent on the complex interplay of several physico­ just a decade. By 2025, the market value is projected to reach between chemical properties such as lateral dimension, thickness, oxidative state, ~US$180 million to US$2.1 billion per year (Reiss et al., 2019; Kong functional groups, dispersion state, purity, administrative doses, expo­ et al., 2019), with about 30% of the annual growth rate for graphene sure route and duration (Wick et al., 2014; Ou et al., 2016). Despite the reinforced composites (Reiss et al., 2019). large number of studies addressing the potential impact of GRMs on In this respect, GRMs have been proposed as alternative fillers (i.e., human health, knowledge gaps on their structure-activity relationships materials added in the polymer matrix to improve specific properties) in remain. At present, no classification system exists to determine the exact polymer composites to replace conventional filler agents, such as carbon level at which each material property may contribute to GRM toxicity. black and carbon fibers, for reinforcement purpose (Stankovich et al., Thus, GRM hazard assessment prediction based on physicochemical 2006). More specifically, graphene oxide (GO) and reduced graphene characteristics is precluded. oxide (rGO) are ideal for polymer composites due to their high dis­ It is further acknowledged that during the life cycle (production, use, persibility and the presence of functional groups that have chemical disposal) of GRM-reinforced composites, possible degradation of the affinity to the surrounding polymers and increase filler/matrix adhesion solid polymer matrix, i.e., through mechanical abrasion, hydrolysis, (Ferrari et al., 2015; Stankovich et al., 2006). Several studies have water exposure, elevated temperatures or increased UV light intensity, shown that the addition of small GRM amounts (0.1–5%) can signifi­ can potentially lead to GRM release and subsequent unintentional cantly enhance the strength, electrical conductivity or thermal transport exposure of workers and/or consumers (Froggett et al., 2014; Duncan of the composites compared to the neat polymer matrix (Zhao et al., and Pillai, 2015; Zepp et al., 2020; Lankone et al., 2020; Waquier et al., 2010; Bortz et al., 2012; Song et al., 2011; Wang et al., 2015a, 2015b; 2020). In contrast to pristine GRMs, which have been extensively Shi et al., 2018). Among the different polymer matrices, polyamide 6 studied, the nanosafety assessment of materials released from (PA6, also known as Nylon-6), a widely used engineered thermoplastic GRM-containing products have not been yet explored and only limited polymer, has received great attention, as an ideal matrix for the devel­ information is available about intrinsic hazards of dissociated materials opment of graphene-reinforced polymer composites (Gong et al., 2015; from graphene-based composites. Moreover, GRMs used as nanofillers in Fu et al., 2015). Its attractive properties, such as good processability, composites may be transformed during the manufacturing process heat and chemical resistance together with its beneficial mechanical and/or during abrasion process in terms of their physicochemical characteristics, i.e., stiffness and toughness, make PA6 appealing for characteristics such as lateral dimension, thickness, and defects. As a composite applications (e.g., automotive and sports equipment) (Fu result, the transformed GRMs may cause different biological responses et al., 2015; Mouhmid et al., 2006). Despite its exciting characteristics, when compared to their pristine form. PA6 has poor dimensional stability, low electrical conductivity, and The aim of the present study was therefore to investigate the 2 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 potential biological effects of particles released from a real-life GRM- Eckmann and colleagues (Carosio et al., 2018). The GO water dispersion enabled product (i.e., rGO-reinforced PA6 composites that will be soon was ultracentrifuge at 20.000 rpms in a SIGMA3–30 K ultracentrifuge, found in the vehicle industry) through an abrasion process, and compare and the wet solid was collected and filtered off and air-dried. these effects to those induced by the rGO initially incorporated in the PA6 thermoplastic polymer matrix. Since GRMs and nanomaterials in 1.2.2. rGO preparation and functionalization general, can enter the human body via different routes (i.e., inhalation, GO powder was thermally reduced in argon atmosphere at 200ºC for ingestion, or dermal absorption) (Fadeel et al., 2018; Oberdörster et al., 20 min. Aminopropyl silane functionalized rGO was prepared as fol­ 2005; Pelin et al., 2018) we performed a multi-system and lows; Ten g of rGO was suspended in a mixture of ethanol/water (30/70) multi-endpoint in vitro human toxicity analysis (Fig. 1). Using and stirred for 1 h. Concentrated hydrochloric acid was added to adjust well-established cell lines and cell-type specific endpoints, we first pH between 3 and 3.5. Five mL of (3-aminopropyl) triethoxysilane assessed the impact of abraded rGO-reinforced PA6 (PA6-rGO) com­ (APTES) were added and the mixture was stirred overnight at 60ºC. The posite materials on cells representative of key target organs related to suspension was vacuum filtered and washed with ethanol to avoid free the three main exposure routes (i.e., lungs, gut, and skin), and the im­ silane molecules. The powder was dried at 80ºC for 24 h. mune system, our primary defense mechanism against foreign intrusion. As inhalation of airborne materials is a major concern at the workplace 1.2.3. Preparation of PA6-rGO composites and characterization where such GRM-enabled products are likely to be processed, we further For the preparation of the concentrated masterbatches, an Alphatech assess the possible effects of PA6-rGO composite materials on the lungs AD30mm LD48 counter rotating twin screw extruder was utilised. The after single pulmonary exposure in mice. Based on this tiered approach functionalised rGO (2.5 wt%) was dispersed in PA6 by melt com­ combining in vitro and in vivo data, the objective of the present study was pounding, resulting in concentrated master batches. The screw speed to provide a comprehensive, multi-laboratory hazard assessment of was kept at 125 rpm and the feeder speed at 30 rpm. commercial GRM-enabled products at doses expected at the workplace. The process temperature in the different barrels in extrusion pro­ cesses are presented in Table 1. 1. Materials and methods The extruded material was then compression moulded to form the film for the abrasion. 1.1. Chemicals and reagents Raman Spectroscopy. Raman spectra of rGO and composite materials (before abrasion) were acquired on a confocal RENISHAW in via Raman All chemicals and reagents used were obtained from Merck-Sigma- microscope at room temperature. The system uses a CCD detector and a Aldrich (Switzerland), unless otherwise stated. holographic notch filter, using excitation wavelength of 532 nm. Scans were acquired from 1000 to 3500 cm-1, performing maps of 25 spectra and making an average spectrum which is the one that is presented in 1.2. Production and characterization of rGO and PA6/PA6-rGO this paper. Analysis and deconvolution of spectra were obtained with composites Wire 4.2 software. X-ray Photoelectron Spectroscopy. XPS analysis was performed in a 1.2.1. GO preparation KRATOS Axis Ultra DLD. GO dispersion was prepared using the modified Hummers method, Fourier-transform infrared spectroscopy. Infrared spectra were using natural graphite sieved at 60 and 100 mesh as described by Fig. 1. Schematic overview of the experimental strategy. The main steps in the study include: 1) Material generation where rGO is produced by thermal reduction from GO dispersions synthe­ sized by the modified Hummers method. Then, rGO is functionalized with aminopropyl silane and used to generate an industrially relevant rGO-reinforced polyamide 6 composite (PA6- rGO; 2.5 wt% rGO). Abrasion particles are produced from neat PA6 and PA6-rGO by Taber abrasion to mimic a mechanical abrasion. 2) Extensive material characterization is per­ formed using a wide spectrum of analytical methods to support data interpretation and comparability across studies. 3) In vitro hazard assessment of the different materials is con­ ducted in robust and widely used cell models representing potential key target organs including the lung, intestine, skin and immune system. 4) In vivo hazard is assessed in mice with a focus on short- and long-term toxicity in the lungs, representing the most relevant exposure route. Finally, all the data is carefully analysed in order to achieve a better under­ standing of the health hazard of abraded PA6- rGO versus rGO alone. (parts of the figure were created with Bio­ render.com). 3 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 Table 1 Temperatures of the different extruder zones, during melt compounding. Extruder zones Z1 Z2 Z3 Z4 Z5 Z6 Z7 Z8 Z9 Z10 Temperature (ºC) 190 190 210 210 210 220 220 220 220 230 recorded in the 4000–450 cm− 1 range using a PerkinElmer FT-IR For this positive control, the fluorescence spectra were recorded Spectrum Two spectrometer. immediately (<1 min) and 10 min after adding H2O2. Thermogravimetric analysis. TGA was performed in a METTLER TOLEDO TGA/SDTA 851e in the range of 25–900 ºC 10º/min at air 1.3. Abrasion process and particle collection atmosphere. Brunauer-Emmett-Teller. The specific surface area (SSA) of GRMs was The abrasion process was performed as described by Netkueakul determined by BET using an autosorb-6 Quantachrome instruments. The et al. Netkueakul et al. (2020) Briefly, PA6 and PA6-rGO composites samples were degassed in an autosorb degasser (Quantachrome in­ were abraded using a Taber abraser (Model 5135, Taber, North Tona­ struments) at 250 ºC for 8 h. wanda, NY) equipped with a S-42 sandpaper strip wrapped around a Transmission electron microscopy. TEM were recorded using a JEOL CS-0 wheel with an applied weight of 1 kg to simulate the sanding microscope (JEM-2010). GRM samples were dispersed in isopropyl process on the surface of the composites. The released particles were alcohol then sonicated with a Hielscher UP200S sonicator for 15 min collected on 0.2 µm nucleopore filter (Whatman, UK) via a vacuum inlet and drop casted onto copper grids for TEM. (flow rate 10 lpm) with a small suction area of 40 mm2, connected to conductive silicon tubing (TSI, United States). 1.2.4. Characterization of rGO and abraded particles Particle Size Distribution. The particle size distribution of the release Scanning Electron Microscopy. The morphology of rGO powder and particles was measured online using an APS (Model 3321, TSI) and a abraded particles was characterised using an SEM Zeiss Evo50 micro­ SMPS consisting of a differential mobility analyser (DMA) (Model 3080, scope (acceleration voltage of 10 kV). The powders were deposited on TSI, Shoreview, MN) and a condensation particle counter (CPC) (Model conductive carbon tape and a thin layer of gold (≈ 10 nm) was sputter- 3775, TSI). The flow rate was generated by the vacuum line and moni­ coated on top of the samples. tored in the range of 9–11 L/min using a mass flow controller (Model Atomic Force Microscopy. AFM images were collected by a contact GFM37, Aalborg, NY). In the present experimental setup, the APS ana­ mode (Bruker, Dimension-Icon) to analyze the thickness of the tested lysed the aerodynamic diameter of particles in the range of 0.5–20 µm, materials. Images were obtained using ScanAsyst-Air probes (silicon tips while the SMPS analysed particles with electrical mobility diameter on silicon nitride cantilever, Bruker) with 0.4 N/m nominal spring ranging from approximately 10 to 570 nm. At least three particle size constant of the cantilever. For the AFM visualization, rGO, PA6 and PA6- distributions measurements were collected from each sample. The rGO were dispersed via sonication (10 min) to ethanol. The rGO flakes collection of abraded particles for nanosafety analysis and for further were then coated to Si/SiO2 (300 nm) substrate using spin coating material characterization was performed without the APS and SMPS to method. reduce possible particle loss. Optical Microscopy. For the optical measurements a Nikon Eclipse LV150N - Digital imaging combined with an advanced optical system 1.4. Material dispersion was employed. Raman Spectroscopy. Raman spectra of the samples were obtained For all experiments, freshly prepared stock dispersions of the tested using a Renishaw inVia Raman spectrometer equipped with a 514 nm materials (rGO, abraded PA6 and PA6-rGO particles) of 0.5 mg/mL in laser. Four different spectra were acquired in different locations of the 0.1% bovine serum albumin (BSA) were prepared by sonication for samples giving consistent results. 45 min prior to usage. Stock dispersions were diluted to the final EPR spectroscopy. EPR spectra were recorded with an ESP 300E working experimental concentrations (2.5–40 μg/mL) in complete cell spectrometer (Bruker) operating at X-band and equipped with a high culture medium and used immediately for exposure experiments. sensitivity resonator (4119HS-W1). Spectra were recorded with 2 mW microwave power, a modulation frequency of 100 kHz, a modulation 1.5. Endotoxin content amplitude of 0.1 mT, a conversion time of 81.92 ms and time constant of 20.48 ms. 1.5.1. Endotoxin Chromogenic assay Fluorescence spectroscopy. The presence of ROS in PA6, rGO and PA6- The endotoxin concentration in PA6, PA6-rGO and rGO suspensions rGO dispersed in 0.1% BSA (at a concentration of 40 µg/mL) was was measured using the Pierce™ LAL Chromogenic Endotoxin Quanti­ assessed following the fluorescence changes of dihydrorhodamine 123 tation kit (sensitivity 0.1 EU/mL; Thermo Fisher Scientific, MA, USA), (DHR123, Sigma-Aldrich) signals using a Jasco FP8003 fluorimeter following the manufacturer’s instructions, at concentrations 0–80 μg/ equipped with a swig xenon 450 W lamp. The non-fluorescent DHR123 mL. is oxidized by ROS to fluorescent rhodamine 123 leading to a significant increase of photoluminescence. In a typical assay, PA6, rGO and PA6- 1.5.2. TET Assay rGO were prepared in 0.1% BSA (1.0 mg/mL) and stocked in the 4 ℃ PA6, PA6-rGO, and rGO samples were assessed for endotoxin content for one week. Then, the stock solutions (80 µL) were diluted with a 0.1% with the TNF-α expression test (TET), which enables unequivocal BSA solution (1920 µL) to 40 µg/mL, following the addition of DHR123 detection of endotoxin with a sensitivity comparable to that of the (0.4 µL of a stock solution at 1 mM in DMSO) (final concentration of conventional LAL assay but without any interference with the assay. In DHR123 corresponding to 200 nM). The fluorescence spectra were brief, PBMCs were isolated from buffy coats obtained from healthy recorded immediately (<1 min) and 10 min after adding DHR123 under human blood donors (Karolinska University Hospital, Stockholm, Swe­ the excitation at 500 nm and were corrected for the baseline and the den) by density gradient centrifugation, as described previously (Fadeel solvent. The fluorescence signal of 0.1% BSA following the addition of et al., 1998). Then, PBMCs were positively selected for CD14 expression DHR123 was also tested at 0 min. A Fenton reaction was performed as a with CD14 MicroBeads (Miltenyi Biotech, Bergisch Gladbach, Ger­ positive control. A 200 µL stock solution (10 mM) of the FeSO4.7 H2O many). For obtaining HMDMs, CD14 + monocytes were cultured in was added to 1.6 mL of water followed by 0.4 µL of DHR123 (final RPMI-1640 cell medium supplemented with 2 mM L-glutamine, 100 concentration of DHR123 corresponding to 200 nM) and 200 µL H2O2. IU/mL penicillin, 100 mg/mL streptomycin, and 10% heat-inactivated 4 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 fetal bovine serum (FBS), supplemented with 50 ng/mL recombinant cytotoxicity assays, cells were seeded in flat bottom 96-well plates at a macrophage colony-stimulating factor for 3 days in 96-well plates. density of 5 × 103 cells/well and exposed to different concentrations of HMDMs were exposed to PA6, PA6-rGO, and rGO (20 μg/mL) or to the tested materials (2.5–40 μg/mL) for 24 h and 48 h. Before each bacterial LPS (0.01 µg/mL) (Sigma-Aldrich) in the presence or absence assay, cells were washed twice with phosphate buffered saline (PBS). of the specific LPS inhibitor, polymyxin B (10 µM) (Sigma-Aldrich), and TNF-a secretion was measured at 24 h of exposure with a Human TNF-α 2.1.3. Intestinal cultures ELISA Kit purchased from MABTECH (Nacka Strand, Sweden). The human colorectal adenocarcinoma cell line Caco-2 was obtained from the German collection of microorganisms and cell cultures (DSMZ, 1.6. Detection of released graphene from PA6-rGO abraded composites Braunschweig, Germany). Cells were maintained in Minimum Essential Medium (MEM; Sigma-Aldrich) supplemented with 10% FBS, 2 mM L- 1.6.1. Freeze fractured samples and SEM visualization glutamine (Sigma-Aldrich), 1% PSN (Sigma-Aldrich) and 1% non- As the release of rGO could be explained by the debonding of the rGO essential amino acids (Sigma-Aldrich) Cell cultures were maintained at from the PA-6 matrix, freeze fractured method was employed to simu­ 37 ◦ C, 5% CO2 and 95% humidity and routinely sub-cultured twice a late the debonding process and to explain the debonding mechanism. In week at 70–80% confluence using 0.5% trypsin–EDTA (Sigma-Aldrich). order to generate the freeze fractured surface, the sample was immersed Differentiation of intestinal cells. Caco-2 cells were seeded at a density into liquid nitrogen for 30 s. After removing the sample from liquid of 250,000 cells/well (corresponds to 2.2 ×105 cells/cm2), in Thin­ nitrogen, the sample was immediately broken by manual bending. Cert™ cell culture inserts for 12-well plates (high pore density trans­ Freeze fractured surface of the samples was analyzed using scanning parent PET membrane, with 3 µm diameter pore size; Greiner bio-one, electron microscope (Nova NanoSEM 230, FEI company, Hillsboro, OR, Kremsmünster, Austria). Cells were grown for 21 days to obtain mature USA). differentiated monolayer and cell culture medium was refreshed three times per week. At day 22, cells were treated with different concentra­ 1.6.2. Transmission electron microscopy of abraded PA6-rGO tions of the tested materials at a range of 2.5–40 μg/mL (100 µL of It was important to evaluate whether abrasion process could induce particle suspension, added apically), for 24 h and 48 h. the release of the rGO from the polymer matrix or not. If the rGO can be released, it is also crucial to assess in which forms rGO was released. 2.1.4. Immune cell cultures Therefore, to understand the characteristics of rGO in the abraded PA-6- The human monocytic THP-1 cell line was purchased from the ATCC rGO, the abraded PA6-rGO was analyzed using TEM (FEI Tecnai G2 F20 (Virginia, USA). Cells were grown in RPMI 1640-GlutamaxTM-I media (FEI Company, OR, USA)). containing HEPES (Gibco, Sweden) and supplemented with 100 U/mL penicillin, 100 mg/mL streptomycin and 10% heat-inactivated FBS 2. In vitro experiments and post-exposure biological response (Sigma). The cells were passaged at a cell density of up to maximum 8.0 × 105/mL every 3–4 days. 2.1. Cell cultures Differentiation of THP-1 cells. For the experiments, the THP-1 cells were plated at the cell density of 60,000 cells/well and pre-incubated for 2.1.1. Lung cultures 24 h in the presence of 0.5 µM PMA, as described previously (Bhatta­ Human A549 epithelial type II cells were obtained from American charya et al., 2017). After 24 h the differentiated THP-1 cells were Type Culture Collection (ATCC; Virginia, USA). After thawing, A549 washed with luke-warm sterile PBS and exposed to rGO, PA-6-rGO and cells were sub-cultured in T75 cell culture flasks in complete Rosewell PA-6 (concentrations 5–40 μg/mL) for 24 h. Cells were tested regularly Memorial Park Institute medium (RPMI-1640, Sigma-Aldrich) with 10% using MycoAlert® mycoplasma detection kit (Lonza, Basel, FBS (Sigma-Aldrich), 1% penicillin, streptomycin, neomycin (PSN; Switzerland). Cells were seeded in 96-well plates in RPMI-1640 medium Sigma-Aldrich) and 1% L-glutamine (Sigma-Aldrich). Cells were main­ at a density of 60000 cells/well and subsequently exposed to the tested tained at 37 ◦ C and 5% CO2 in humidified atmosphere and were materials at concentrations 5–40 µg/mL for 24 h, or were maintained in routinely sub-cultured twice a week at 70–80% confluence. cell medium alone (untreated cells-negative control). Prior to the experiments, cells were seeded apically on transparent The RAW 264.7 macrophage cell line was cultured in DMEM cell polycarbonate 12-well inserts (3 µm pore size, 113.1 mm2 surface area, culture medium supplemented with 10 µg/mL gentamycin, 10 mM N-(2- Greiner Bio-One, Austria) at concentration of 2.5 × 105 cells/insert and hydroxyethyl)-piperazine- N-ethanesulfonic acid, 0.05 mM β-mercap­ were grown for three days on submerged conditions to form a tight toethanol and 10% FBS at 37 ◦ C in a 5% CO2 incubator. Cells were monolayer, Cell cultures were then transferred to air-liquid interface seeded in 96-well plates at a density of 1 × 104 cells per well and (ALI) conditions by removing the apical medium in order to produce cultured for 24 h prior the material exposure. surfactant and better reflect in vivo conditions in human lung (Chortarea The human acute promyelocytic leukemia cell line HL-60 (ATCC – et al., 2015). After 24 h at ALI, cells were then exposed apically to CCL-240) was maintained in phenol red-free RPMI-1640 medium sup­ 100 µL of different concentrations of PA6, PA6-rGO and rGO plemented with 2 mM l-glutamine and 10% heat-inactivated FBS (2.5–40 μg/mL), for 24 h and 48 h (pseudo-ALI approach (Barosova (Sigma) in 5% CO2 at 37 ◦ C. et al., 2018; Endes et al., 2014)). Differentiation of HL-60 cells. In order to allow for neutrophil-like Crystalline quartz DQ (purchased from Dörentrup Quarz GmbH & differentiation, the cells were seeded at 0.5 × 106 cells/mL in the above- Co. (material No. 04; mean particle diameter d(50)= 3.71 µm, d(97)= mentioned cell medium supplemented with 1.25% DMSO for 5 days, as 12.03 µm)) was included as a positive inflammogenic control for A549 described previously (Khandagale et al., 2018). The culture medium was cells. DQ was freshly dispersed in ultrapure ddH2O, and 100 µL were refreshed after 3 days. Cells were tested regularly using MycoAlert® applied apically in A549 cells (final concentration of 100 µg/mL). mycoplasma detection kit (Lonza, Basel, Switzerland). Prior the expo­ sure, experiments the differentiated HL-60 cells were seeded in 96-well 2.1.2. Skin cultures plates in phenol red-free RPMI-1640 medium at a density of 60000 HaCaT cells were cultured in high glucose Dulbecco’s Modified Ea­ cells/well. gle’s medium (DMEM), supplemented with 10% FBS, 1.0 × 10-2 M L- After seeding, the immune cells were treated with the indicating glutamine, 1.0 × 10-4 g/mL penicillin and 1.0 × 10-4 g/mL strepto­ concentrations (5, 10, 20, 40 µg/mL) of PA6, PA6-rGO and rGO and the mycin at 37 ◦ C in a humidified (95%; air: 5% CO2) atmosphere. Cell biological response was evaluated 24 h post-exposure. passage was performed two days post-confluence, once a week. All the experiments were performed between passage 74 and 86. For 5 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 2.2. Cell viability 2.3. Evaluation of intestinal barrier integrity 2.2.1. MTS assay The differentiation process of Caco-2 cultures grown in 12-well in­ The effects of the materials on cell viability were evaluated by the 3- serts was controlled by TEER measurements at day 7, 14 and 21 after (4,5-dimethylthiazol-2-yl)− 5-(3-carboxymethoxyphenyl)− 2-(4- seeding using an Epithelia Voltohmmeter (EVOM) with sterilised STX2 sulfophenyl)− 2 H-tetrazolium (MTS) assay (CellTiter 96® AQueous electrodes (World precision, Instruments, Sarasota Florida, USA). The One Solution Cell Proliferation Assay, Promega, Dübendorf, impact of different concentrations of PA6, PA6-rGO and rGO Switzerland) in all the cell models examined (with the exception of the (2.5–40 μg/mL) on the barrier integrity of differentiated Caco-2 cells neutrophil model, where the ATP assay was performed instead), ac­ was assessed by measuring the TEER before the treatment (0 h) and after cording to ISO 19007:2018 and (Elliott et al., 2017). Briefly, after ma­ 24 h and 48 h of treatment. Three measurements in different parts of the terial exposure, phenol red free fresh medium containing the MTS insert were retrieved from each sample, and TEER values were calcu­ reagent was added to the cells. Then, the cell cultures were incubated for lated using the formula TEER = [Ω (cell inserts) – Ω (cell-free inserts)] an additional 1 h (4 h for the skin model), at 37 ◦ C. Absorbance was × 1.12 cm2. The values for each concentration represent the average of measured at 490 nm by a microplate reader. The chemical positive three independent experiments. control cadmium sulphate (CdSO4, Sigma-Aldrich) was added to the cells at concentration 1000 µM for 24 h. Data are presented as % cell 2.4. Lipid absorption on the intestinal barrier viability as compared to negative control (untreated cells). Cell free interference assay was performed for the rGO and the To investigate the potential effects of rGO and abraded PA6/PA6- abraded particles, at a concentration range of 0–40 μg/mL to exclude rGO particles on lipid uptake, differentiated Caco-2 cells on 12-well potential interference reactions of the tested materials with the MTS inserts were exposed with different concentrations of each tested ma­ assay reagents along that might lead to possible wrong interpretation of terial (5–40 μg/mL) for 24 h and 48 h. Cell cultures treated with 50 µg/ the results (SI 15). mL of the fatty acid synthase inhibitor C75 for 24 h were used as a positive control. Lipid uptake was measured as described in previous 2.2.2. ATP assay work (Hempt et al., 2020). Briefly, after material exposures, the cultures The ATP assay was performed for the viability assessment of the were incubated for 10 min with 20 µM BODIPY™ 500/510 C1, C12 neutrophil model. After the 24 h material exposure, the differentiated (Thermo Fischer, D3823) in 0.1% BSA, according to the manufacturer’s HL-60 cells were lysed, and total cellular ATP content was quantified instructions. Then, the inserts were washed with ice cold 0.1% BSA and with a luminescence-based cell viability assay, CellTiter-Glo (Promega, complete DMEM medium (phenol red free) was added to the apical and Dübendorf, Switzerland). CellTiter-Glo reagent was mixed at a 1:1 ratio basolateral side. After 1 h incubation at standard growth conditions, the with the experimentally treated cells from the treatment plate. The mix fluorescence was measured with a multi-well plate reader (Mithras2 was incubated for 10 min at room temperature on the shaker, followed Plate reader, Berthold Technologies, Germany) at 485/528 nm by luminescence measurement with an Infinite F200 Tecan plate reader (excitation/emission). (Mannendorf, Switzerland). The experiment was performed with at least three biological replicates and three technical replicates for each con­ 2.5. Skin cell mass and proliferation centration of PA-6, PA6-rGO, and rGO. Results were expressed as per­ centage cell viability versus control. Cell mass was evaluated by the sulforhodamine B (SRB; Sigma To control for potential interference of the materials with the assay, Aldrich, Milan, Italy) assay, as a measure of viable adhered cells (Pelin the PA-6, PA6-rGO and rGO were dispersed in cell-free medium (at the et al., 2017). Briefly, after materials exposure, cells were washed twice working concentrations) and mixed with CellTiter-Glo reagent (no with PBS, fixed with 50% trichloroacetic acid for 1 h at 4 ◦ C and stained interference was observed; data not shown). for 30 min with 0.4% SRB in 1% acetic acid. After washings with 1% acetic acid, the protein-bound dye was dissolved in 10 mM TRIZMA base 2.2.3. WST-8 assay solution and the absorbance was read by an Automated Microplate The cell viability in skin cells was additionally evaluated by the 2-(2- Reader EL 311 s (Bio-Tek Instruments, Winooski, VT, USA) at 570 nm. methoxy-4-nitrophenyl)− 3-(4-nitrophenyl)− 5-(2,4-disulfophenyl)− Data are reported as % of cell mass with respect to negative control. 2 H-tetrazolium (WST-8) reduction assay. The WST-8 assay was carried Cell proliferation was evaluated using a colorimetric ELISA based on out using the Cell Counting Kit (CCK)− 8 assay (Sigma Aldrich; Milan, 5-bromo-2′ -deoxyuridine (BrdU) incorporation (Sigma-Aldrich; Milan, Italy), as previously described (Pelin et al., 2018). Briefly, after materials Italy), following the manufacturer’s instructions. Absorbance was read exposure and two washings with PBS, cells were incubated for 4 h with by an Automated Microplate Reader EL 311 s (Bio-Tek Instruments, fresh medium containing 10% WST-8 reagent. Absorbance was deter­ Winooski, VT, USA) at 450 nm. Data are reported as % of cell prolifer­ mined at 450 nm by an Automated Microplate Reader EL 311 s (Bio-Tek ation with respect to negative control. Instruments; Winooski, VT). Data are reported as % cell viability as compared to negative control (untreated cells). 2.6. Inflammatory responses (lung and immune cells) 2.2.4. LDH assay 2.6.1. Cytokine and chemokine release The potential cytotoxicity in RAW 264.7 macrophages was assessed The secretion of human IL-6 and IL-8 in A549 alveolar epithelial cells by the LDH test kit (CytoTox 96® Non-Radioactive Cytotoxicity Assay, (24 h and 48 h post-exposure to rGO and abraded composite materials) Promega, Dübendorf, Switzerland), according to the manufacturer’s was quantitatively detected by enzyme-linked immunosorbent assay instructions. Following the 24 h post-exposure time, the culture medium (ELISA; Human IL-6 and IL-8 Uncoated Elisa Kit, Invitrogen, USA) ac­ was collected for the LDH measurements. Treatment of cells for 30 min cording to manufacturer’s instructions. Cells treated apically with 1 μg/ with lysis solution (included in the assay) was served as positive control mL TNF-α for 24 h were used as positive control. Measurements were to induce cytotoxicity. The OD of each sample was recorded at 490 nm performed on high binding 96-well plates (Coating Costar, New-York, on a Microplate Reader (Thermo, Varioskan Flash). LDH release (% of USA) and at 370 nm optical absorbance using a microplate reader positive) is presented as the percentage of (ODtest-ODblank)/(ODpositive- (Mithras 2 Plate Reader, Berthold Technologies, Germany). ODblank), where ODtest is OD of control cells or cells exposed to materials, Differentiated THP.1 were exposed to rGO (20 μg/mL). The exposed ODpositive is OD of positive control cells and ODblank is OD of well without cell media were collected and stored at − 80 ◦ C for further analysis. IL- cells. 1β release was determined by using a human IL-1β ELISA kit (Invitrogen, 6 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 Sweden) according to the manufacturer’s instruction. Absorbance was image-processing package Fiji. measured at 450 nm using a Tecan Infinite F200 plate reader. Results are RAW 264.7 macrophages were plated in the 96-well plates (1 ×104 expressed as pg/60 000 cells of released cytokine. To assess the role of cells per well) and incubated for 24 h. Abraded materials were added to the NLRP inflammasome, cells were preincubated for 1 h with either the cells at increasing concentrations (5, 10, 20, 40 µg/mL) upon zVAD-fmk (20 µM) and MCC950, an NLRP3 inhibitor (10 µM) (both dispersion in cell culture medium. Untreated cells cultured in cell cul­ from Sigma). ture medium in absence of materials were used as negative control. The The release of TNF-α (BD OptEIA, BD Biosciences) and IL-6 (BD cell morphology was observed under an optical microscopy (Zeiss Axi­ OptEIA, BD Biosciences) in RAW 264.7 macrophages, following 24 h overt 40 CFL, 20 × magnification) at 24 h post-exposure. treatment with the tested materials was determined using ELISA kits (BD Biosciences), following the manufacturer’s instructions. Optical absor­ 3. In vivo exposure experiments and post-exposure tissue bance measurements were performed using a microplate reader (Ther­ analyses moscientific, VARIOSKAN LUX) at 450 nm. LPS (1 µg/mL for 24 h, Sigma) served as the positive control to stimulate inflammatory 3.1. Animal husbandry response. Prior to any experiment, C57BL/6 female mice (Envigo, UK) of 6–8- 2.6.2. Multiplex analysis-based cytokine profiling on immune cells week old were acclimatized for 7 days after arrival. Animals were kept The Meso Scale Discovery (MSD) (Rockville, MD, USA) plate-based randomized in groups of 5 with free access to water and food, at a electrochemiluminescence (ECL) assay was employed to quantify cell temperature of 19–22 ◦ C, relative humidity of 45–65%, under a regular supernatant concentrations of IFN-γ, IL-1β, IL-2, IL-4, IL-6, IL-8, IL-10, 12 h light/dark cycle between 7 am and 7 pm. All procedures were IL-12 p70, IL-13, and TNF-α. As a positive control, cells were exposed conducted in accordance with the ARRIVE guidelines for animal to 0.1 μg/mL LPS for 24 h. The stored cell supernatant was thawed on research and after ethical approval from the UK Home Office, under wet ice just prior to use. Samples were diluted 1:2 using Diluent 2, and Project License no. P089E2E0A. each sample was run in duplicate. The V-PLEX Human Pro-inflammatory Panel 1 Human Biomarker 10-Plex Kit was used. Plates were processed 3.2. Animal exposure according to the manufacturer’s instructions and read using the MSD MESO Sector S 600 instrument. The assay data were analyzed using MSD All materials were prepared on the day of exposure by dispersion in Discovery Workbench 4.0. The cytokine expression data retrieved from water (water for injection, ThermoFisherScientific, Gibco) containing the multiplex assay were analyzed using hierarchical clustering analysis. 0.1% BSA (Merck, Sigma-Aldrich). After anesthesia with 3% of iso­ Complete linkage and Euclidean distances were employed as metrics to flurane, mice were maintained on a slanted board and 30 µL of materials draw association dendrograms between cytokines and the different or vehicle was put at the basis of the tongue, allowing distribution of the treatment conditions, as described previously (Bhattacharya et al., materials into the lungs through physiological breathing. Mice (n = 3) 2017). The cluster analysis and the corresponding heat maps were were exposed by oropharyngeal aspiration to 15 µg (in 30 µL) of PA6- prepared by using R, as previously reported (Bhattacharya et al., 2017). rGO (2.5% of rGO, which represents about 0.3 μg of rGO for 15 μg of PA6-rGO) or controls: neat PA6 (15 µg), rGO-0.3 µg (low dose, repre­ 2.7. Cell morphology senting approximately 2.5% of the high dose) or rGO-15 µg (high dose) and vehicle (0.1% BSA in water). The applied dose of 15 µg per animal Following the 24 h and 48 h material exposure, A549 alveolar equals to 0.75 mg/kg, while 0.3 µg per animal equals to 0.015 mg/kg for epithelial cells were fixed in 4% paraformaldehyde (PFA) and incubated animal weighing on average 20 g at the time of exposure. Outcomes at 4 oC. In order to prepare the cells for staining, cultures were incubated were assessed at 1, 7, and 28 days after single exposure. At each time with 0.1 M glycine in PBS for 15 min. Then, cells were permeabilized point, mice were euthanized by intraperitoneal injection with an over­ with 0.2% Triton X-100 in PBS for another 15 min, prior the application dose of pentobarbitone. of fluorescent dyes. DAPI (4′ ,6-diamidino-2-phenylindole) (Thermo Fischer, USA), and Alexa Fluor 488 (Thermo Fisher, USA) were used to 3.3. Bronchoalveolar lavage (BAL) analysis stain nuclei and cytoskeleton of the cells respectively. DAPI and Alexa Fluor 488 were diluted to ratio of 1:1000 and 1:50 respectively in 0.3% At the end of the post-exposure recovery period, bronchoalveolar Triton X-100 and 1 g BSA in PBS. Following the staining, cells were lavages (BAL) were performed on the right lung (isolated via litigation of incubated with fluorescent dyes for 1.5 h in darkness, at room temper­ the right bronchia) using 3 times 0.8 mL ice cold PBS (Merck, Sigma- ature. Subsequently, cells were washed three times with PBS and Aldrich). BAL fluids were analyzed for composition in cell population mounted on microscopy slides using Mowiol® 4–88 (Sigma-Aldrich) on (differential counting by Kwik-Diff Stains, Epredia, Shandon), LDH (LDH both sides of the membranes. Visualization of the cellular morphology colorimetric assay, Promega) and total protein content (Bi-cinchoninic was achieved using a confocal LSM fluorescence microscope (CLSM, Acid (BCA) assay, Pierce BCA Protein Assays, ThermoFisherScientific). LSM780, 60 × magnification oil, Carl Zeiss AG, Switzerland). IL-6 and TNF-α concentrations in BAL fluids were assessed using Mouse The morphology of keratinocytes was evaluated by confocal micro­ ELISA MAX deluxe Sets (Biolegend). scopy after membranes staining with 1 μM 1,1’-Dioctadecyl-3,3,3’,3’- tetramethylindocarbocyanine perchlorate (DiL) probe (Sigma-Aldrich; 3.4. Whole lung inflammation profile analysis Milan, Italy), as previously described (Pelin et al., 2017). Cells (2 ×105 cells/well) were cultured for 24 h before exposure to each material After BAL, the right lungs were lysed in RIPA buffer (Merck, Sigma- (40 µg/mL) for 48 h. Cells were then washed twice with PBS, probed Aldrich) containing EDTA-free protease inhibitor (cOmplete Mini, with 1 μM DiL for 10 min at room temperature, fixed with 4% PFA for Roche). After homogenization using 5 mm stainless steel beads for 30 min at room temperature and washed twice with PBS. Samples were 10 min at 50 Hz (TissueLyser LT system, Qiagen), cell lysates were mounted on coverslips of 1 mm thickness using the Prolong Gold anti­ centrifuged for 5 min at 2600 g to remove cell debris. Collected super­ fade reagent (Life Technologies; Milan, Italy). Images were taken by a natants were stored at − 80 ◦ C until analysis. Total protein concentra­ confocal microscope (Eclipse C1si, on an inverted microscope TE2000U, tions were measured using a BCA assay (ThermoFisherScientific, Nikon) at 60 × magnification. GBMs presence in each sample was Pierce), and cytokine levels were evaluated using a Mouse Inflammation visualized by the reflection mode property during the confocal acqui­ Panel (13-plex: IL-1α, IL-1b, IL-6, TNF-α, MCP-1, GM-CSF, IL-17A, IL-23, sitions. Reconstructions of the images were performed offline using the IL-12p70, IFN-γ, IFN-β, IL-27 and IL-10, v-plate, Biolegend) following 7 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 manufacturer’s recommendations. A FACSVerse flow cytometer (BD and Si-O-Si bonds, respectively, and indicates the successful chemical biosciences) was used to measure fluorescence intensities. Sample con­ functionalization) and Si-C vibrations (appears at 870 cm-1 as broad centrations were determined using a standard curve and then expressed peak) (SI 1b). Similar patterns were observed by Wan et al. Wan et al. in pg/mg of protein after normalization by total protein concentration. (2014) TGA revealed a mass loss of 3.9 wt% below 525ºC, in agreement IL-12p70 values were below the test sensitivity and were not considered. with the 4.3 wt% loss calculated when assuming a degree of function­ alisation of 0.4% (SI 3). The starting non-functionalized rGO showed a 3.5. Histopathological analysis BET surface area of 736,3 m2/g, while the functionalised rGO incorpo­ rated in the composites, demonstrated a BET surface area of 178,7 m2/g. The left lungs were collected, fixed in formalin, paraffin embedded, The reduction in the surface area is attributed to the non-accessible area sectioned on microtome, and finally stained using Hematoxylin and of the gases due to the stacking during the filtration and drying process Eosin stain for further histopathological analysis. (Guo et al., 2014). Transmission electron microscopy (TEM) observed that the functionalized rGO comprised of folded structures and was 3.6. Statistical analysis agglomerated (SI 4). The functionalized rGO was either embedded in the PA6 composites (2.5 wt%) or used for further characterization and In vitro results are presented as the mean ± standard error of the toxicological analysis (hereafter, rGO refers to functionalized rGO). mean (SEM) of at least 3 independent experiments, and presented as the The morphology of the rGO and the abraded composites (PA6 and mean ± standard deviation for in vivo experiments (n = 3). Statistical PA6-rGO) was studied using scanning electron microscopy (SEM), analysis was performed by one-way ANOVA followed by Bonferroni’s atomic force microscopy (AFM), as well as optical microscopy (Fig. 2a, post-test, Dunnett’s or Sidak’s multiple comparison test analysis, using SI 5a-b). The SEM micrographs of the rGO showed aggregated nano­ the software GraphPad Prism (version 9.0). The results were considered flakes with broad cluster sizes that range from 10 to 100 µm (Fig. 2a). significant if p < 0.05. For each in vivo time point, Kruskall-Wallis fol­ Such an arrangement of graphene sheets is typical of rGO powders (Ali lowed by Dunn’s post-hoc test was used to evaluate statistical differ­ et al., 2017). Moreover, the large SEM magnification allows dis­ ences compared to the negative control; p < 0.05: * . tinguishing crumpled nano-scaled flakes of rGO stacked together. The abraded PA6 particles form a worm-like structure with morphologies 4. Results and discussion comparable to other previous reports (Golchin et al., 2013). The abraded PA6-rGO composites display similar worm-like topography with worm 4.1. Characterization of rGO and abraded particles from PA6-rGO sizes in the range of 10–100 µm. The presence of rGO could not be composites detected by SEM imaging in the abraded PA6 particles. Similarly, Sachse et al., observed that particles generated from PA6-foam glass crystal The rGO samples were produced by thermal reduction from GO (FGC), PA6-SiO2 composites after a drilling process, showed very similar dispersions synthesized using the modified Hummers method (Carosio structures of particles (Irfan et al., 2013). et al., 2018). Subsequently, rGO was functionalized with aminopropyl AFM was employed to evaluate the thickness of the flakes/agglom­ silane. The surface functionalization of rGO was verified by Raman, erates in the tested materials. Height profiles indicated areas of x-ray photoelectron spectroscopy (XPS), thermogravimetric analysis increased thickness of rGO flakes, which are predominantly stacking or (TGA), Brunauer-Emmett-Teller (BET) and Fourier-transform infrared folding on each other (SI 5a). Similarly, in the abraded PA6-rGO, the (FTIR) analysis. SI 1a presents a comparison of the Raman spectra of the stacking of several layers was monitored, which is verified by the vari­ GO, non-functionalized rGO and aminopropyl silane functionalized ation in thickness values. The average thickness for rGO and PA6-rGO rGO). An intense D peak (~1350 cm-1) which confirmed lattice distor­ was similar (6–10 nm and 4–10 nm, respectively). In addition, AFM tions and the G peak (~1585 cm-1) which corresponds to the overlap of showed that lateral dimensions of rGO ranged from 1 to 5 µm, while G and D’ peaks are observed. The 2D peak (~2700 cm-1), and D+D’ and PA6-rGO ranged from 0.75 to 5 µm. Due to difficulties to disperse PA6 to 2D’ peaks which are different overtone and combination peaks of the a non-aggressive solvent, it was not possible to achieve AFM measure­ previous ones, show minimal intensity, corresponding to stage 2 defects. ments for the neat PA6. Non-functionalized rGO shows the highest I(D)/I(G) [I(D)/I(G) 1,15 The presence of rGO in the PA6-rGO composite, after the abrasion (functionalized rGO), 0,70 (non-functionalized rGO), 0,86 (GO)] in process could be confirmed by the presence of the characteristic D and G agreement with the evolution of I(D)/I(G) in stage 2 of defects. A peaks in the Raman spectra. The Raman spectra of the neat abraded PA6 decrease in full width half maximum (FWHM) of D and G peaks were particles presented a strong fluorescence contribution (Fig. 2b). The observed for non-functionalized rGO material FWHMG: 73 (GO), 103 main peaks of this sample and the correspondent vibration modes were (non-functionalized rGO), 70 (functionalized-rGO) cm-1; FWHMD: 107 analysed in a previous report (Ferreiro et al., 2004). The Raman spectra (GO), 186 (non-functionalized rGO), 78 (functionalized-rGO) cm-1, of the rGO and abraded PA6-rGO were analysed through Lorentzian which is associated to a decrease in disorder and in alignment with a fitting. The rGO showed the typical D peak at approximately 1335 cm-1 decrease of the oxygen content due to a reduction process during the and the G at ≈ 1579 cm-1, which is consistent with previous studies functionalization of rGO (Gómez et al., 2017; Eckmann et al., 2013; (Fig. 2b) (Krishnamoorthy et al., 2012; Song et al., 2015). The I(D)/I(G) Beams et al., 2015). The reduction process was further confirmed by XPS ratio was ≈ 1.2. The characteristic 2D peak could be observed between where a decrease in the oxygen content was reported: 30.4% (GO), 2500 and 2800 cm-1. The abraded PA6-rGO composite instead presented 13.5% (non-functionalized rGO), decreased to 6.7% (functionalized a blue shift of approx. 10 cm-1 resulting in a D peak at ≈ 1345 cm-1 and a rGO), during the silanization process (SI 2). It should be noted, that the G at 1587 cm-1 (I(D)/I(G) was ≈ 1.1). rGO is intentionally a partially rGO, because this characteristic (together The presence of surface radicals was examined by EPR and fluores­ with the functionalisation) enables a better integration and improved cence spectra analysis as reactive oxygen species (ROS) account one of mechanical properties in the polymer composite for automobile appli­ the main drivers for particle toxicity (Nel et al., 2006). As reported in the cations. Hence, this rGO is not representative of graphene that would be SI 6a, no radicals were observed in the neat PA6, while low amounts of used in electronics or for conductivity purposes. In the case of func­ radicals were detected in the rGO containing samples (both powder and tionalized rGO, 0.4% of Si and 0.44% of N was observed corresponding dispersed form) with likely no influence on the biological responses in to a 0.4% of functionalization of the rGO. FTIR spectra of the func­ the concentrations tested (SI 6b). tionalized rGO in region 620–1300 cm-1 showed two width and high The particle size distribution of the abraded composites was intensity peaks, which involved different components typically assigned measured online, during the abrasion process, using the aerodynamic to Si-O-Si, Si-O-C (1083 and 1047 cm-1 is typically assigned to the Si-O-C particle sizer (APS) and the scanning mobility particle sizer (SMPS) that 8 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 Fig. 2. Characterization of rGO and abraded particles from PA6-rGO composites. (a) SEM images of rGO, abraded particles from PA6-rGO composite and abraded particles from neat PA6. (b) Raman spectra of the tested materials (rGO, abraded PA6-rGO and abraded PA6). (c) Particle size distributions of the abraded particles in the micrometer range, measured by APS. The results show mean ± S.D. from at least three independent measurements. 9 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 quantify the aerodynamic diameter and the electrical mobility of the monocyte-derived macrophages (HDMD) in the presence or absence of abraded particles, respectively. Online APS measurements of the the specific lipopolysaccharide (LPS) inhibitor, polymyxin B, indicating released particles during the abrasion process revealed a mean aero­ that our tested materials were endotoxin-free. dynamic diameter of 3.16 µm for the neat PA6 and 1.91 µm for the PA6- rGO composite particles (Fig. 2c). For both abraded materials, the ob­ 4.3. Qualitative detection of rGO released from the PA6 matrix tained aerodynamic diameter values were under 4 µm, which falls into the respirable fraction in the alveolar region (Oberdörster et al., 2005). It has been shown that fillers such as GRMs may be released from the More specifically, particles with an aerodynamic diameter smaller than polymer composite matrix as a result of the matrix degradation (e.g., 100 µm can be inhaled, but only particles with an aerodynamic diameter thermal degradation, hydrolysis, UV exposure), during the routine use of smaller than 4 µm can reach the alveolar region of the lung (Oberdörster GRM enriched composites (Duncan and Pillai, 2015; Froggett et al., et al., 2005; Su et al., 2016). Another peak was observed towards the 2014). The released particles can either be GRMs fully embedded, lower detection limit of the instrument (at 0.5 µm) for both abraded partially protruding, or completely released (free-standing) from the materials, indicating the presence of particles with an aerodynamic polymer matrix. Hence, depending on their aerodynamic size and shape, diameter smaller than 0.5 µm. It should be noted that the aerodynamic they may pose a threat to human health. diameter of the abraded particles measured by APS was notably smaller Previous studies have shown that the release of fillers from polymer than the particle size estimated by the SEM. This observation, which was composites could be related to intrinsic failure mechanism. Release of reported previously by Schinwald et al., could be explained by the fact rGO either as protruding or as free-standing from the PA6-rGO com­ that in contrast to the particle size estimated by SEM from the projected posite, during the abrasion process, could be explained by an interfacial area, the aerodynamic diameter measurements depend on both the debonding between the rGO and PA6 molecules or an adhesive failure as shape, the density and the orientation of the particles with respect to its a result of the weak interaction between the rGO and PA6 matrix, as direction of motion (Schinwald et al., 2012). Furthermore, the APS re­ reported by previous studies (Massaq et al., 2019; Zaman et al., 2012; sults suggested that the addition of 2.5 wt% functionalized rGO in the Zheng et al., 2019; Netkueakul et al., 2020). In addition, failure of polymer matrix resulted in a decrease in the average particle size graphene sheets is also highly possible, i.e., graphene layer breakage. compared to the particles released from the neat PA6. Recent studies The failure mechanism of PA6-rGO composites can be examined by have shown that the addition of GRMs at concentrations as low as 0.5 wt visualizing the surface of fractured composites by SEM imaging (Zaman % could significantly improve the mechanical properties of polymer et al., 2012). The fractured surface of PA6-rGO composites showed composites (Netkueakul et al., 2020; Zeng et al., 2015). The presented rougher structure and deeper cracks compared to that of neat PA6, since PA6-rGO composites demonstrated improved mechanical properties as the incorporation of rGO could affect the crack propagation of the PA6 confirmed by the increased Young´s modulus and tensile strength, composite (SI 9a). Moreover, the characteristic layered and platelet-like compared to other GRM-PA6 and neat PA6 composites (Gomez et al., structure of rGO was observed as it is released from the PA6 matrix, due 2020). This increase in toughness could make the composites more to the adhesion failure between rGO and the matrix. Indeed, the exposed challenging to crack, leading to the generation of smaller particles, as rGOs suggested a weak interaction between the rGO and the PA6 matrix, shown in the PA6-rGO. In contrast, the lower hardness of neat PA6 which could also be implied when the composites undergo an abrasion resulted in separation of larger sized particles (Zok and Miserez, 2007). process. In addition, high-resolution TEM analysis of the abraded par­ In addition, a larger number of particles was collected from PA6-rGO ticles from PA6-rGO composites revealed the layered structure of rGO, composite samples than from the neat PA6 samples. This could be appearing as protruding part from the PA6 matrix (SI 9b), thus con­ explained by the observed electrostatic charging in the neat PA6 parti­ firming that rGO can be released from the polymer during the abrasion cles during abrasion and their scattering, prior to their collection by the process. It is worth mentioning, that this layered structure can only be vacuum probe. Jian et al. reported that the electrostatic potential and observed when rGO is wrinkled or when observed from the side of the the tribological behavior of PA6 during a wear process is not a simple stacked layers. Although we observed signs of possible rGO release, it mechanism and can be affected by internal phenomena, such as heat was not possible to draw conclusions on the quantity of rGO released friction accumulation, physical adsorption, wear rate, and surface, as from the polymer. Considering the low percentage of rGO (2.5%) in the well as external conditions (Jian et al., 2009). Even though the under­ PA6 composite and the sensitivity of PA6 polymer even in mild acid lying mechanisms were not defined, the electrostatic charging due to the conditions, it is technically very complex/impossible to determine the abrasion and the subsequent heat production was observed in the neat amount of rGO that would be present at the surface of the abraded PA6 particles, resulting in lower particle concentration. Focusing on the particles of PA6-rGO or totally released from the polymer matrix. SMPS results only background signal was observed up to ~400 nm, which was followed by an increase in particles concentration at 4.4. Impact on lung cells in vitro approximately 700 nm (detection limit of the instrument), indicating the existence of particles in larger dimensions, in accordance with the Numerous studies have proven that among the different routes of APS results (SI 7). engineered nanomaterial (NM) exposure, the pulmonary exposure and consequently the respiratory tract is of the highest concern (Wang et al., 4.2. Detection of endotoxin content in PA6, PA6-rGO and rGO 2015a, 2015b; Oberdörster et al., 2005, 2009). Once inhaled, carbon-based materials can reach the deepest regions of the respiratory An important aspect that needs to be evaluated before any biological tract (alveolar region) where they can accumulate for months and assessment of a (nano)biomaterial is the possible contamination by en­ distribute to other organs (Li et al., 2013; Mercer et al., 2013). Upon dotoxins. It has been shown that endotoxin contamination of carbon- inhalation, respirable particles deposit mainly in the alveolar lung re­ based nanomaterials may generate artifacts, which could distort bio­ gion (Oberdörster et al., 2005; Rodrigues et al., 2020). Potential effects logical outcomes, when investigating effects on immune-competent cells were therefore evaluated in this part of the respiratory tract, using the (Li and Boraschi, 2016). According to the chromogenic endotoxin assay, A549 cell line (human adenocarcinoma derived-alveolar epithelial type the endotoxin content in the tested materials was negligible (below 0.10 II cells), which is the most commonly applied cell culture model of EU/mL). Since it has been reported that GRMs may interfere with human alveolar lung epithelium for hazard assessment. In most studies traditional endotoxin assays, we further evaluated endotoxin, using the A549 cells are grown in 2D conformation and submerged in cell culture tumor necrosis factor-α (TNF-α) expression test (TET) (Mukherjee et al., medium, which is not representative of the conditions in the human 2016). As shown in SI 8, none of the materials (PA6, PA6-rGO and rGO) lung. However, when growing on permeable supports, A549 cells are triggered tumor necrosis factor alpha (TNF-α) secretion in human able to form a polarized tight monolayer and can be maintained at 10 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 air-liquid interface (ALI) conditions, comparatively resembling native approach (where a thin layer of material suspension is applied alveolar epithelia (Wu et al., 2018a 2018b; Chortarea et al., 2019). In apically on the insert), without disturbing the surfactant production. addition, when cultured 24 h at ALI conditions, A549 cells actively Furthermore, pseudo-ALI exposures are more relevant than fully sub­ secrete surfactant proteins, resulting in a surface tension similar to merged conditions since they result in faster particle deposition, lower values measured in vivo, in pulmonary alveoli (Blank et al., 2006; particle agglomeration, and higher particle-cell interactions (Chortarea Chortarea et al., 2015). Taking into consideration these important fea­ et al., 2019; Barosova et al., 2018; Endes et al., 2014). tures, A549 cells were selected to assess the biological effects of pristine An important consideration when performing safety assessment rGO, abraded PA6 and PA6-rGO composite materials on the alveolar studies is the use of realistic exposure concentrations. In case of GRMs, epithelium. Material exposures were performed using a pseudo-ALI human realistic exposure concentrations can only be estimated from the Fig. 3. (a) Cell viability of A549 epithelial cells, measured by the MTS assay after 24 h and 48 h treatment with rGO and the abraded particles from neat PA6 and PA6-rGO composites. 1000 µM CdSO4 was used as a positive control. (b) Confocal LSM images of A549 cells after exposure to the tested materials (scale bars: 10 µm). The green color shows F-actin (cytoskeleton), red color shows DNA (cell nuclei). (c) Pro-inflammatory response of A549 epithelial cells after treatment with the tested materials. Concentrations of IL-6 and IL-8 were measured after 24 h and 48 h of materials exposure. One μg/mL TNF-α was used as the positive control. The results are shown as mean ± standard error of the mean (SEM) from at least three independent experiments. * indicates statistical significance compared to the negative control at 24 h of exposure (p < 0.05). # shows a statistically significant response (p < 0.05) at 48 h of material exposure. 11 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 existing limits for CNT pulmonary exposure, where a full working life- Nakano-Narusawa et al., 2020; Beyeler et al., 2018; Endes et al., 2014). time exposure would result in an alveolar mass retention of 10–50 µg/ Hence, it is frequently used as a positive control for inflammatory re­ cm2 and an acute lung exposure (24–48 h) would be in the range of sponses in hazard assessment studies. As shown in SI 10, 24 h exposure 1 μg/cm2 (Gangwal et al., 2011; Chortarea et al., 2017). The tested doses to DQ (100 µg/mL) elicited significant release of both IL-6 and IL-8 in (2.5–40 µg/mL) in the present study correspond to 1.3–44 µg/cm2 in the A549 cells, thus further confirming the responsiveness of the A549 cells different plate formats used for the different cell types and therefore under the specific culture and exposure conditions. cover the human realistic doses (from low (2.5 μg/mL) to slightly Similar to the abrasion materials, rGO-exposed A549 cells for 24 h overload (40 μg/mL) doses). The selected dose-range is based on human and 48 h did not show evidence of increased pro-inflammatory response, inhalation exposures as it is considered the primary and riskiest expo­ neither in IL-8 nor in IL-6 secretion (Fig. 3c). On the contrary, Reshma sure route for nanomaterials. Considering other exposure routes (e.g. et al., demonstrated that rGO treatment for 24 h resulted in a dose- dermal, ingestion exposure) no human exposure data are available. For dependent reduction in cell viability and NF-κB mediated inflamma­ that reason and for comparability purposes, the same dose range was tory response in A549 2D cultures (Reshma et al., 2016). applied in the different biological systems under investigation. Dose-dependent toxicity in rGO exposed BEAS-2B (bronchial epithelial) To determine the acute impact (24 h and 48 h post-exposure) of the and A549 (alveolar epithelial) cells was also noted by Mittal et al. (2016) materials under investigation on the alveolar cells, cell viability, alter­ These differences in biological response could be ascribed to differences ations in morphology and induction of pro-inflammatory reactions were in the physicochemical properties of the tested materials as well as evaluated, at different concentrations of materials (2.5–40 μg/mL). differences in cell culture conditions and pseudo-ALI versus submerged Adverse effects of particles released from the abrasion wheel on the cell conditions. culture models were not anticipated as the quantified amount of released particles was negligible. Moreover, no effects were noted in 4.5. Impact on gastrointestinal cells in vitro previous studies using a similar experimental setup (Netkueakul et al., 2020; Schlagenhauf et al., 2015). Another important exposure route for NMs is through the GI tract, Acute 24 h exposures to either pristine rGO or abraded PA6 and PA6- where NMs can either directly be ingested or indirectly enter the GI tract rGO composite particles did not significantly alter the mitochondrial by oral breathing and during the clearance process of formerly inhaled activity of alveolar epithelial cells at pseudo-ALI conditions, as shown by airborne NMs (Sohal et al., 2018). In contrast to large number of studies the MTS viability results (Fig. 3a). Similarly, no reduction in cell focusing on GRM inhalation, the impact of GRMs on the GI system has viability was detected in the alveolar barrier, following 48 h exposure to been largely overlooked, despite the potential implications on human the tested materials. In accordance with the viability results, no signs of health in both consumer and occupational settings (Pietroiusti et al., morphological alterations were observed in cells exposed to the highest 2017). Caco-2 cells, a human colon adenocarcinoma-derived cell line, is concentration of abraded particles and rGO (40 μg/mL) for 48 h, when the gold-standard for in vitro studies of the human intestinal epithelium, compared to the negative control cultures (Fig. 3b). Importantly, the with wide-spread application in pharmaceutical and toxicological alveolar epithelial cell layer was maintained, without ruptures or signs research (Kucki et al., 2016, 2017). The existing in vitro studies have of apoptosis (i.e., fragmented cell nuclei or cellular blebbing) to the predominantly been performed with non-polarized undifferentiated nuclei and cytoskeleton. Caco-2 cells (representing pre-enterocytes) for fast and high-throughput Analysis of interleukin-8 (IL-8), a critical airway epithelial-derived screening of potential substances. Nevertheless, differentiated Caco-2 pro-inflammatory chemokine primarily implicated in acute inflamma­ cells (21 days cultivation on microporous inserts) represent a more tion and accumulation of neutrophils in inflammatory diseases, revealed reliable model to mimic the small intestine’s enterocyte barrier as they that both abraded materials did not induce any effect on IL-8 secreted more realistically reflect the mature human enterocytes, both morpho­ levels even after 48 h of exposure (Fig. 3c) (Mukaida, 2003). In addition, logically and functionally (e.g., tight and polarized enterocyte-like interleukin-6 (IL-6) a key inflammatory marker (in both acute and epithelial cell layers with brush border and tight junction formation) chronic inflammation) pivotal for the pathogenesis and exacerbation of (Kucki et al., 2017; Domenech et al., 2020; Carr et al., 2012). Therefore, numerous pulmonary diseases was evaluated (Rincon and Irvin, 2012). a toxicological assessment of pristine rGO and abraded PA6 and Exposure to different concentrations of abraded PA6 and PA6-rGO PA6-rGO, using pseudo-ALI approach to induce the physiologically composite materials for up to 48 h did not elicit an increase in IL-6 relevant, fully differentiated Caco-2 cells in vitro model of the human levels compared to untreated cultures. Since IL-6 and IL-8 are funda­ small intestinal epithelial barrier was performed. mental for the initiation and activation of pivotal inflammatory path­ To evaluate the effects of the tested materials on the intestinal cell ways, evaluation of the secretion of these cytokines provides a valid viability, the mitochondrial activity was assessed after exposure to indication of possible pro-inflammatory reactions. The absence of sig­ material concentrations ranging from 0 to 40 μg/mL for 24 h and 48 h. nificant pro-inflammatory responses is in agreement with Wohlleben As shown in Fig. 4a, the enterocyte epithelial barrier initially demon­ et al. who did not observe any significant toxicity in lung tissue exposed strated high resistance to abraded PA6 and PA6-rGO, as well as, to to polyurethane/3% CNT composites (Wohlleben et al., 2013). Consis­ pristine rGO cytotoxicity. More specifically, cell viability at 24 h was tent with our results, Irfan et al., demonstrated that dust particles unchanged, without significant reduction of mitochondrial activity, released from silicon-PA6 composites and neat PA6 composites after a regardless of materials or concentrations tested. At the 48 h time-point, crash and drilling process, did not exhibit significant cytotoxicity PA6-treated cells remained unaffected, while a slight reduction of following 48 h of submerged exposure in non-polarized A549 cells (Irfan viability was reported in PA6-rGO cultures at high concentrations (non- et al., 2013). Similar observations were also reported in another study significant p > 0.05, 20 μg/mL; 77% viability, 40 μg/mL; 79% viable where no signs of acute adverse effects (e.g., reduction in mitochondrial cells). For rGO, the decline in cell viability was more pronounced even at activity, or increased secretion of pro-inflammatory mediators) were lower material concentrations (10 μg/mL; 72% viability, significant detected in A549 cells, from neat epoxy or from the addition of CNTs p < 0.05, 82% and 66% viability at 20 and 40 μg/mL respectively, non- (1%) to the epoxy matrix (Schlagenhauf et al., 2015). significant p > 0.05) at 48 h. As only a few reports are available on the The reliability and sensitivity of the pseudo-ALI A549 model was effects of GRMs on human enterocytes in vitro, and none of the existing validated using crystalline quartz (DQ). DQ is a well-known inflammo­ studies have investigated the effects of rGO or GRM-embedded com­ genic material, classified as a group 1 human carcinogen (Wilbourn posite materials, a comparison with our in vitro intestinal results was et al., 1997; Clouter et al., 2001; Monteiller et al., 2007). Several studies limited. Nevertheless, the lack of acute effects on mitochondrial activity reported significant inflammatory effects both in vitro and in vivo after even at high concentrations (up to 80 μg/mL) was noted by Kucki et al. DQ administration (Chortarea et al., 2015, 2017; Barosova et al., 2020; who investigated the toxicological impact of four types of GOs on 12 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 To evaluate the functional integrity of the intestinal barrier, the transepithelial electrical resistance (TEER) values before and after ma­ terial exposure were measured. No significant TEER reduction was re­ ported when differentiated Caco-2 cells, were exposed to abraded PA6 or PA6-rGO even after 48 h of exposure (Fig. 4b). Similarly, no effect on barrier integrity was observed following exposure to pristine rGO, sug­ gesting that the observed loss in mitochondrial activity did not have any impact on the barrier function. Consistent with our results, Domenech et al. found that neither GO nor graphene nanoplatelets (GNP) exposures were able to affect the functional/structural integrity of an intestinal co- culture barrier model (Caco-2/HT-29 cells), as reported by TEER and permeability measurements (Domenech et al., 2020). In addition, membrane integrity of cells was maintained even after treatment with digested GOs in a Caco-2 undifferentiated monolayer and a triple co-culture (Caco-2/HT-29/Raji) intestinal model (Guarnieri et al., 2018; Bitounis et al., 2020). The lipid absorption at the brush borders of the enterocytes is a fundamental physiological function of the intestinal barrier (Hansen et al., 2007). To investigate possible alterations in the fatty acid uptake of differentiated Caco-2 cells after treatment with abraded composite materials and rGO, a lipid uptake assay was performed. Cell treatment with fatty acid synthase inhibitor C75 resulted in a significant decrease in lipid uptake at both 24 h and 48 h (Fig. 4c), suggesting a major defect in lipid absorption, in agreement with the observations of Hempt et al., and Accioly et al. Hempt et al. (2020); Accioly et al. (2008) On the contrary, no significant changes in fatty acid absorption were detected after treatment with abraded PA6, PA6-rGO and pristine rGO for 24 h. At 48 h of treatment with abraded PA6 and PA6-rGO a significant reduction was only observed at the concentration of 10 μg/mL (79% and 75% respectively, p < 0.05), however the higher material concentra­ tions did not induce any evident effect in lipid uptake. In the case of rGO, a notable loss of fatty acid uptake was shown at the highest concentra­ tion tested, following 48 h of exposure (40 μg/mL; 64%, p > 0.05). It should be emphasized that no previous studies have analyzed possible alterations in the lipid absorption function of the intestinal barrier after GRM or other carbon-based material exposure. Interestingly, a previous study using synthetic amorphous silica did not observe any decline in the lipid uptake in a triple intestinal culture (Caco-2/HT-29/Raji co-culture) after 48 h of exposure (Hempt et al., 2020). 4.6. Impact on skin cells in vitro The skin as the largest organ in the human body, with primary Fig. 4. (a) Impact of abraded particles (PA6, PA6-rGO) and rGO on cell functions to defend and maintain physiological conditions, is considered viability of differentiated CaCo-2 intestinal epithelial cells after 24 h and 48 h one of the most common routes for GRM exposure, during the life cycle of exposure (MTS assay). CdSO4 served as a chemical positive control. (b) of GRM containing products (Pelin et al., 2018; Zhang et al., 2014; Pelin Assessment of the barrier integrity by TEER measurement before and after et al., 2018). Epidermal cells consist mainly of keratinocytes, structur­ exposure to the tested materials (c) Lipid absorption was determined following ally involved in maintaining the barrier function of the epidermis, but 24 h and 48 h of exposure to the tested materials. Treatment of cells with the also exerting a key role in the initiation and perpetuation of skin in­ fatty acid synthase inhibitor C75 (50 µg/mL) for 24 h served as a chemical positive control for the investigation of lipid uptake. The results are shown as flammatory and immunological responses at the skin level (Hänel et al., mean ± SEM from at least three independent experiments. * indicates statisti­ 2013). Therefore, the potential dermatoxic effects of abraded cal significance compared to the negative control at 24 h of exposure rGO-composite materials and pristine rGO were investigated on human (p < 0.05). # shows a statistically significant response (p < 0.05) at 48 h of HaCaT skin keratinocytes, a non-tumorigenic, spontaneously immor­ material exposure. talized human keratinocyte cell line, already employed to study skin toxicity of pristine GRMs (Frontiñán-Rubio et al., 2018; Pelin et al., undifferentiated Caco-2 cells for 48 h (Kucki et al., 2016). Similarly, the 2017, 2020). viability of undifferentiated Caco-2 cells remained unaffected, following Initially, the acute effects on cell viability were evaluated by means exposure to GO and few layer graphene (FLG) repeatedly for 4 days of mitochondrial activity of HaCaT cells after two exposure times (24 h (Guarnieri et al., 2018). It is pertinent to note that previous studies re­ and 48 h) by the MTS assay (Fig. 5a). The pristine rGO induced a sig­ ported that NM and, in particular, GRM uptake is significantly lower in nificant reduction of cell viability only after 48 h exposure to the highest differentiated Caco-2 barrier models compared to undifferentiated cells concentration (40 µg/mL; 78% cell viability, p < 0.05), in line with the due to the lack of typical structures (e.g., microvilli and tight junctions), weak cytotoxic effects induced by other pristine GRMs such as FLG and which hinder NM internalization (Kucki et al., 2017; Guarnieri et al., GO on the same cell type (Pelin et al., 2017). However, treatment of 2018; Song et al., 2015). Hence, despite higher internalization, no acute keratinocytes with abraded particles from neat or rGO-reinforced PA6 toxicity GRM effects were noted in the existing studies. composites did not induce any decrease in mitochondrial activity. In contrast, a slight increase on cell viability was observed after 24 h 13 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 Fig. 5. Effect of rGO and abraded PA6 and PA6-rGO on the mitochondrial activity of HaCaT cells evaluated by (a) the MTS assay and (b) the WST-8 assay. Potential effects of the tested materials on skin cell proliferation assessed by (c) the SRB incorporation assay and (d) the BrdU assay. Data are the mean ± SEM of 3 independent ex­ periments. * indicates statistical sig­ nificance compared to the negative control at 24 h of exposure (p < 0.05). # shows a statistically significant response (p < 0.05) at 48 h of material exposure. (e) Confocal micrographs of HaCaT cells exposed to 40 μg/mL of rGO, abraded PA6-rGO and abraded neat PA6 for 48 h. Plasma membrane of HaCaT cells was labeled with the fluorescence DiL dye (red, left panel). rGO is visualized by reflection mode acquisition (white, middle panel); confocal reconstruction of red DiL labeled HaCaT cells merged with white reflecting rGO (merged images, right panel). Original magnification: 60×. Scale bar: 50 µm. 14 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 exposure, which was more pronounced after 48 h exposure (significant epidermal cells, HaCaT cells exposed to pristine rGO, abraded PA6-rGO from the concentration of 5 μg/mL, p < 0.01). In particular, after 48 h, and PA6 (40 μg/mL) for 48 h were subjected to confocal microscopy abraded PA6 and PA6-rGO particles increased cell viability at 143% and analysis, in which materials were visualized in white exploiting their 148%, respectively, at the highest concentration. To further validate the light reflection properties during the confocal acquisition. As shown in results from the MTS assay, the WST-8 assay was also performed, as a Fig. 5e, rGO and PA6-rGO were able to interact with cells, where the widely used assay to evaluate the mitochondrial damage caused by presence of rGO, and to a far less extent PA6-rGO, was visualized bound different GRMs on a wide range of in vitro models, including HaCaT to the membranes and inside keratinocytes. In line with this observation, keratinocytes (Pelin et al., 2020, 2017; Chang et al., 2011; Wang et al., using reflection mode acquisition, different GRMs (e.g. GO, FLG) were 2014). WST-8 confirmed the results obtained by the MTS, although more captured to interact and penetrate the plasma membrane of HaCat cells substantial effects were reported (Fig. 5b). More specifically, rGO (Pelin et al., 2017). Reflection mode, however did not allow the detec­ induced a significant reduction of cell viability after 48 h exposure to the tion of the abraded PA6 polymer in the cells, due to low contrast of these highest rGO concentration (40 µg/mL; 64% cell viability, p < 0.001). particles. Similarly, the low signal given by PA6-rGO acquired in Moreover, as observed with the MTS assay, abraded PA6 and PA6-rGO reflection mode by confocal microscopy analysis could be ascribed to a particles significantly increased cell viability, already after 24 h expo­ quenching effect by PA6 polymer bound to rGO. This hypothesis is sure and starting from the concentration of 10 µg/mL (142% cell supported by the lack of light reflection properties of PA6, which could viability, p < 0.05) for PA6 and at the highest concentration of partially mask the signal given by rGO. 40 µg/mL PA6-rGO (156% cell viability, p < 0.05). The effect was more pronounced after 48 h exposure, being significant starting from the 4.7. Impact on immune cells in vitro concentration of 5 μg/mL (p < 0.05), inducing a maximum increase of 167% and 173% for abraded PA6 and PA6-rGO particles, respectively, at The immune system identifies and protects the body from infections the highest concentration. To quantify the viable attached cells, without and other exogenous materials. Therefore, understanding the interac­ relying on metabolic activity measurements that might lead to false tion of GRMs with the pivotal cells of the innate immune system is of positive overestimation of cell viability, the cell mass was determined particular importance (Boraschi et al., 2017; Lin et al., 2021; Mukherjee using the well-established SRB assay (Skehan et al., 1990). A similar et al., 2018; Russier et al., 2013). The key front-line cells of the immune trend to the mitochondrial activity assays was observed. Pristine rGO system assigned to eliminate foreign materials and assist in tissue resulted in a slight but significant reduction of cell mass at 40 µg/mL remodeling after injury across the whole body are macrophages (90%, p < 0.05), while abraded composite particles significantly (Mukherjee et al., 2017). In a pulmonary exposure scenario, macro­ increased cell mass (113%, p < 0.05%, and 117%, p < 0.05, for PA6 and phages will be the first immune cells interacting with GRMs and dealing PA6-rGO, respectively) at the same concentration, and after 48 h of with their clearance or long-term persistence (Schinwald et al., 2012; exposure (Fig. 5c). To further investigate whether abraded PA6 can Rodrigues et al., 2020; Roberts et al., 2016; Bengtson et al., 2017). The promote cell proliferation, the % of proliferating cells was determined as effect of abraded particles and pristine rGO was therefore assessed in shown in Fig. 5d. No alteration of cell proliferation was induced by human macrophage differentiated THP-1 cells. The biological response exposure to pristine rGO for neither 24 h nor 48 h. On the contrary, was also tested in murine macrophage-like cells (RAW 264.7) to deter­ abraded PA6-rGO and PA6 increased cell proliferation already after 24 h mine possible difference in the behavior of the tested materials towards at the highest concentrations (p < 0.05). At 48 h of exposure to abraded human or murine cells. THP-1 and RAW 264.7 macrophage models act composites, cell proliferation was more evident, being significant as proxy for either tissue-recruited macrophages upon inflammation already at the concentration of 10 μg/mL and inducing maximum effects signaling or tissue resident macrophages, such as alveolar macrophages of 147% and 144% for abraded PA6 and PA6-rGO, respectively in the lungs. (p < 0.05). These results suggest that: i) the increased cell viability Cell viability of THP-1 differentiated and RAW 264.7 macrophages induced by abraded PA6 and PA6-rGO may be dependent on a prolif­ exposed to PA6, PA6-rGO and rGO was determined using the MTS assay. erative stimulus and ii) the proliferative stimulus seems to be exerted by As seen in Fig. 6a-b, abraded PA6 and PA6-rGO did not decrease the abraded PA6. Indeed, Dias et al. recently reported the ability of PA6 to mitochondrial activity of either THP-1 differentiated or RAW 264.7 promote cell proliferation in murine fibroblasts when employed as a macrophages even at the highest concentration (40 μg/mL). In agree­ nanostructured scaffold, indicating its potential of modulating skin ment with our findings, Netkueakul et al. did not observe any significant regeneration and wound healing parameters at cellular and histological cytotoxic effects of abraded particles from epoxy/1%GRM (GOs, rGO or levels (Dias et al., 2019). This phenomenon may be explained by the GNPs) composites on THP-1 differentiated macrophages, at the same structural similarities between PA6 and proteins. Indeed, the amide exposure concentrations (Netkueakul et al., 2020). In the same study, groups of the polymer can interact by hydrogen bonds with cellular pristine GRMs, including rGO, did not reduce the cellular mitochondrial molecules, providing a structural environment able to improve cell activity upon exposure to concentrations of up to 40 µg/mL. In addition, growth (Dias et al., 2019). In addition, cell proliferation may be pro­ studies focusing on abraded polymer composites reinforced with other moted and sustained by the hydrophilic wetting behavior of PA6 (Kry­ carbon-based materials, i.e., CNTs, found that the abraded particles from siak et al., 2020). epoxy/1%CNT composites did not cause acute effects on mitochondrial Possible alterations in cell morphology were visualized by confocal activity of THP-1 cells, while pure CNTs resulted in a significant microscopy after staining plasma membranes of HaCaT cells with the decrease in cell viability (Schlagenhauf et al., 2015). In accordance with DiL fluorescence dye. Negative control cultures displayed the typical the above reported cell viability results in the skin and GI in vitro models, cobblestone morphology of skin keratinocytes, which was slightly pristine rGO did induce a dose-dependent loss of cell viability in both condensed in cells treated with 40 μg/mL of each material for 48 h, macrophage cell lines (significant at concentrations of 20–40 μg/mL for particularly for rGO-treated cultures. Moreover, the morphology of rGO- THP-1 cells and 10–40 μg/mL for RAW 264.7 cells, respectively). treated cells showed a moderate cell shrinkage, in line with the observed Moreover, no induction of cytotoxic reactions was reported for RAW cytotoxic effect. 264.7 macrophages after exposure to abraded materials, while a sig­ Fusco et al. recently reported that keratinocytes are able to selec­ nificant release of lactate dehydrogenase (LDH) was observed, following tively sense and interact with GRMs at amounts as low as 1 μg/mL 24 h exposure to 20 and 40 μg/mL of pristine rGO, supporting the MTS (Fusco et al., 2020). In addition, further analysis reported that GRM data (SI 11). Consistent with our results a dose-dependent decline in cell cytotoxic effects on HaCaT cells are only partially reversible, probably viability, especially at higher rGO concentrations (from 20 µg/mL), has because of their ability to be internalized inside keratinocytes (Pelin been observed in bone marrow derived macrophages as well as in the et al., 2020). To investigate the interactions between the materials and J774A.1 macrophage cell line (J. Wu et al., 2018; Y. Wu et al., 2018). 15 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 Fig. 6. Cell viability (MTS assay) of (a) THP-1 and (b) RAW 264.7 macrophages and (c) neutrophil-like HL-60 cells (ATP assay) exposed to increasing concentrations of rGO and abraded PA6 and PA6-rGO composites for 24 h. (d) Cytokine release by RAW264.7 macrophages. IL6 and TNF-α levels were determined after incubation with increasing concentrations of PA6, PA6-rGO and rGO. Cells exposed to 1 μg/mL LPS were used as positive control. Data are the mean ± SEM of 3 independent experiments. The * symbol represents p < 0.05 as compared to the negative control. (e) Cytokine profiling and hierarchical cluster analysis of inflammatory me­ diators released in differentiated THP-1 macrophages exposed to rGO, abraded PA6-rGO and PA6. Cells exposed to 0.1 μg/mL LPS were used as positive control. Each branch in the dendrograms shows the similarity between samples; the shorter the branch, the more similar the samples. This stands in contrast with the work of Netkueakul et al. and Li et al. progression, and resolution of inflammation (Zhou et al., 2012). Hence, where no significant cytotoxicity was shown in rGO-treated THP-1 cells to further evaluate the immunological impact of the tested materials in (Li et al., 2018; Netkueakul et al., 2020). Those differences in biological differentiated THP-1 cells, a multiplex-cytokine array analysis of ten responses may depend on differences in the lateral dimensions, thick­ crucial pro-inflammatory mediators was conducted at a sub-toxic con­ ness and the amphiphilic properties of the materials. centration (20 μg/mL). In parallel, hierarchical clustering was carried Macrophages are also known to be involved in the initiation, out to draw associations between cytokine responses evidenced for 16 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 pristine rGO, the abraded PA6-rGO and PA-6, the LPS positive control, To gain more insights into the interaction of the tested materials with versus the negative control in human macrophages (Fig. 6e). The cyto­ macrophages, the cellular morphology was closely examined. Conven­ kine profiling analysis clearly demonstrated that the effects induced by tional light microscopy images of RAW 264.7 macrophages upon 24 h the tested materials are distinctly different from the effects of LPS. exposure to abraded PA6 and PA6-rGO materials revealed no signs of Moreover, the cluster analysis suggested that the cytokines could be modification in cellular morphology even at the highest concentration of broadly separated into two main clusters: those affected by high secre­ 40 µg/mL (SI 13). Correspondingly, no major alterations were observed tion of cytokines generated by the LPS treatment and those affected by in THP-1 cells after 24 h exposure to abraded GRM epoxy composites rGO, PA-6-rGO, and PA-6 treatment. More specifically, a significant (Netkueakul et al., 2020). In the case of pristine rGO, although increased increase in the secretion of interferon gamma (IFN-γ) was revealed in rGO accumulation was observed on the cells surface, the shape of the THP-1 cells when exposed to rGO and PA-6, comparable to the effect RAW264.7 macrophages was not affected at low concentrations induced by the positive control LPS, a well-known stimulator for cyto­ (5–20 µg/mL). However, macrophages treated with 40 µg/mL of pris­ kine secretion in these cells. Moreover, two classical Th1 tine rGO showed a rounder appearance compared to untreated control pro-inflammatory cytokines, TNF-α and interleukin-1beta (IL-1β), were cells. In good agreement with our observations a loss of protrusions with significantly upregulated following acute exposure to all three materials a subsequent morphological shift towards rounded cells has been pre­ (SI 12a), and IL-1β production was blocked by the pan-caspase inhibitor, viously demonstrated after rGO exposure (20 μg/mL) in primary mac­ zVAD-fmk and the selective NLRP3 inhibitor, MCC950 (SI 12b), sug­ rophages (Wu et al., 2018a, 2018b). The observed morphological gesting that these materials triggered inflammasome activation, as alterations might indicate an increased rGO internalization by the shown for several other NMs (Sun et al., 2013). No significant release of macrophages. Moreover, the significant TNF-α secretion in RAW 264.7 IL-2, IL-4, IL-6, IL-8, IL-10, IL-12 and IL-13 was reported in cells could lead to the activation of macrophages into a material-treated THP-1 cells compared to the untreated cells. Similar to pro-inflammatory state, which can lead to similar alterations in the the results obtained with human macrophages, murine RAW 264.7 cells cellular shape and could be a hint for the intrusion of rGO onto the did not show significant upregulation of the crucial cytokine IL-6 plasma membrane, as shown by Wu et al. (2018a), (2018b) In support to regardless of the material applied, while significant TNF-α release was that Li et al. demonstrated that while GO nanosheets are mostly asso­ noted following exposure to all three materials (20 and 40 μg/mL for ciated with the surface membrane in THP-1 cells, rGO was principally PA6 and PA6-rGO, all tested concentrations for rGO), as shown in internalized by phagocytosis (Li et al., 2018). Fig. 6d. Comparing THP1 and RAW 264.7 cells, no species-to-species Several studies have shown that macrophages may be more suscep­ variations in the investigated biological responses were observed, as tible than other cell types to the biological effects of particles, due to the similar effects were reported in both human and mouse cells. greater dose of internalized particles. A previous study of a panel of Previous reports showed that neat polymer matrix materials such as metal oxide nanoparticles showed that the alveolar macrophage cell line epoxy did not trigger significant pro-inflammatory responses on immune was the most sensitive cell model among the 12 cell models tested cells (Netkueakul et al., 2020; Schlagenhauf et al., 2015). However, the (Farcal et al., 2015). The toxicity of metal/metal oxide particles may be present study demonstrated that abraded neat polymer PA6 did elicit explained by the rapid dissolution of the particles in the acidic envi­ significant release of three pivotal pro-inflammatory proteins in THP-1 ronment of the lysosomes. In addition, high aspect ratio materials may cells. More specifically, upregulation of IFN-γ, a key player in cause lysosomal destabilization, and subsequently lead to host-defense against pathogens, might indicate that abraded particles pro-inflammatory cell activation. However, the materials tested here are from this polymer matrix may be recognized as intracellular parasites by not considered high aspect ratio materials as abraded particles and THP-1 cells, leading to increased IFN-γ secretion. Moreover, both GRMs are not necessarily fiber-like and their biological effects depend abraded materials (neat polymer PA6 and composite PA6-rGO) resulted on different material properties, not only the lateral dimensions (Wick in secretion of TNF-α and IL-1β cytokines that would typically lead to et al., 2014). Nonetheless, the lateral dimensions of graphene oxide (GO) recruiting more immune cells i.e., neutrophils, monocytes/macrophages have been shown to play a role for interleukin-1β (IL-1β) and trans­ to the affected areas, as a hallmark of acute inflammation (Ma et al., forming growth factor-β1 (TGF-β1) production and a more recent study 2015). In addition to the endogenous inflammogenic potential of the has confirmed that the lateral dimensions may impact on cellular uptake polymer matrix, the potential inflammation of PA6-rGO could be and subsequent biological effects in liver cells (Wang et al., 2015a, explained by the transformation of rGO during the production process of 2015b; Li et al., 2021). the composite or due to protruding or released GRMs from the polymer Another important immune cell type often neglected in toxicological matrix. However, the magnitude of the pro-inflammatory effect induced studies is neutrophils, the most abundant circulating leukocytes and by the abraded PA6-rGO particles was notably lower compared to that among the first cells to be recruited in the airways, upon lung exposure induced by pristine rGO or the LPS control, suggesting only a minor to GRMs (Keshavan et al., 2019). These cells also play a key role in the pro-inflammatory response (SI 7b). Since the PA6-rGO composite ma­ initiation of inflammation in several tissues as well as in GRM biodeg­ terial contains only a small percentage of rGO (2.5%), the observed radation processes (Mukherjee et al., 2018a, 2018b; Kurapati et al., limited toxicity of the composite material (containing a small % of rGO) 2015). There is a paucity of data on the biological effect of GRMs on compared to the pristine rGO is probably due to the low amount of rGO neutrophils and the few available studies mainly focused on the degra­ in the composite, the endogenous inflammogenicity of the neat polymer dation of GRMs by a myeloperoxidase-dependent mechanism in neu­ PA6, and not to a different mechanism of toxicity. Previous studies trophils and not on immunotoxicity. Mukherjee et al. recently reported showed that GRM (particularly GO) treatment could provoke inflam­ that GO sheets with differing lateral dimensions triggered a matory responses, including the production of inflammatory cytokines dose-dependent loss of mitochondrial activity in neutrophils as by macrophages (Ma et al., 2015; Russier et al., 2013). The cytokine measured by the ATP assay. At the same time, GO acted directly on the profiling experiments revealed that rGO triggered the cytokine pro­ neutrophil cell membrane, leading to cell activation (Mukherjee et al., duction of the M1 markers TNF-α and IL-1β in THP-1 cells, suggesting 2018b). Therefore, it is important to know if pristine rGO and abraded possible inflammasome activation. These findings support those particles from rGO-reinforced composite materials have an impact on recently reported by Gurunathan et al. who observed significant stim­ neutrophils. Focusing on neutrophils, cell viability was determined on ulation of TNF-α and IL-1β secretion, following exposure to differentiated HL-60 cells using the ATP assay, as the MTS assay was not vanillin-functionalized rGO in THP-1 cells (Gurunathan et al., 2019). A appropriate for this non-adherent cell line (data not shown). Consistent similar trend was found for TNF-α secretion in primary and J774A.1 with the macrophage viability results, no significant effect on the macrophages after treatment with two types of rGO (Wu et al., 2018a, viability of PA6 or PA6-rGO exposed HL-60 cells was observed (Fig. 6c), 2018b). but a 24 h rGO treatment revealed a dose-dependent loss of 17 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 mitochondrial activity, only significant at the highest concentration After a single exposure to abraded PA6-rGO, histopathological (40 μg/mL, 50% viability). In addition to the viability assay, a cytokine analysis of lung sections did not reveal any alteration of the lung tissue array was performed in neutrophils to examine possible induction of at the tested time-points (Fig. 7a). Absence of obvious cell recruitment or pro-inflammatory processes (SI 14). In contrast to THP-1 cells, cluster tissue damages was also noted after exposure to the low dose of pristine analysis in exposed HL-60 cells revealed a large single cluster with low rGO (rGO-0.3 µg), neat polymer alone (PA6) or vehicle control. In cytokine secretion for all tested cytokines, suggesting that neutrophils contrast, exposure to high dose of pristine rGO (rGO-15 µg) induced might not be a good model for cytokine screening. Furthermore, in macrophage-driven granuloma-like formations at all considered time HL-60 cells, no distinct cytokine upregulation was reported, except for points, and a clear infiltration of immune cells at 1 and 7 days. Inter­ LPS-treated cells (SI 1). estingly, the location of those inflammatory structures were overlapping An important, non-trivial matter in relation to the in vitro results with material accumulation sites. Such reactions to foreign materials obtained with respect to pristine rGO, abraded neat PA6 or rGO- occur when materials cannot be easily eliminated from the tissue (Wil­ reinforced PA6 is the in vitro dosimetry. It has been shown that liams and Williams, 1983; Anderson et al., 2008). The key component of cellular uptake of gold nanoparticles depends on their sedimentation granuloma-like formation is the fusion of numerous macrophages into and diffusion velocities, while it is independent of size, shape, density, multinucleated cells, which are able to phagocytose particles that can be surface coating and the initial concentration of the nanoparticles (EC larger than 10 µm (Anderson et al., 2008). Noteworthely, the pro­ et al., 2011). More recent studies of so-called ’’buoyant’’ nanoparticles nounced inflammatory structure formation observed at day 1 and 7 was (e.g. polypropylene) have shown that their biological activities may be followed by an obvious tissue recovery by day 28 (i.e., with notable underestimated when using conventional submerged cell culture reduction in granuloma size, Fig. 7a). These results are not consistent methods (CY et al., 2016). In the present study, A549 and Caco-2 cells with the findings of a previous study in which a prolonged inflammation were exposed using a more physiologically relevant (for the lung and the over 90 day after exposure was associated to the presence of large, black GI) pseudo-ALI approach, which allows faster particle sedimentation and compact rGO agglomerates that persisted in lung tissue for up to 90 and more particle-cell interaction compared to the conventional sub­ days without being phagocytosed (Bengtson et al., 2017). These results merged exposures. Regarding the skin (HaCaT) and immune (THP-1, suggest that bio-persistence of materials was the leading cause for a RAW 264.7, and HL-60) cells, such pseudo-ALI exposures are not rele­ prolonged inflammation, hence material clearence could in contrast be vant; therefore, submerged exposures were applied. The existing models associated to inflammation resolution. In our study, the decrease in to estimate in vitro concentration of nanomaterials in submerged con­ granuloma size and resolution of cell influx and inflammation over time ditions (e.g. the in vitro sedimentation, diffusion, and dosimetry (ISDD) is therefore likely to result from a successful rGO internalization and model, the distorted grid (DG) model, or the 3D-sedimentation-diffu­ elimination by macrophages. Moroever, the in vivo pulmonary impact of sion-dosimetry (3DSDD) model) were developed and are suitable only rGO has been largely overlooked in comparison to other GRMs (Fadeel for spherical particles, such as spherical silica, gold, and metal oxide et al., 2018). In the few existing studies comparing GO and rGO, GO nanoparticles. They have a limited applicability for non-spherical ma­ seemed to provoke more inflammogenic responses compared to rGO terials (e.g. GRMs and CNTs) due to their different form factor and (Bengtson et al., 2017; Poulsen et al., 2021). heterogeneity (DeLoid et al., 2014, 2015; Böhmert et al., 2018). Adverse effects were also assessed by measuring total protein and Consequently, determining the GRM cell delivered dose in submerged LDH release to the airways. These two markers are typically used as conditions is still a challenge due to the analytical difficulties of quan­ hallmarks of lung epithelium tissue damage (Moalli et al., 1987). For the tifying the GRM dose delivered to cells, as it would require several abraded PA6-rGO and PA6 materials, no significant variations of protein analytical techniques and cellular uptake quantification. Particularly, concentration or LDH activity in bronchoalveolar lavage (BAL) fluid for the abraded PA6 and PA6-rGO materials, it would be almost were observed (Fig. 7b-7c), suggesting no disruption of the air-blood impossible as the cellular internalisation cannot be visualised with the barrier and the lung epithelium, in agreement with the absence of re­ existing methodology i.e., TEM, Raman microscopy or optical micro­ sponses found in the A549 cell model (Fig. 2). Despite the observation of scopy and LSM reflection mode (as we have shown for RAW 264.7 and granulomatus formation after exposure to high dose of rGO (rGO-15 µg), HaCaT cells). no significant (p > 0.05) effects on BAL proteins or LDH release were seen at 1, 7, or 28 days post-exposure, suggesting that the response to 4.8. Pulmonary effects in mice rGO was mostly driven by immune cells. In BAL fluid, no variation in the total cell number was observed for The in vitro investigations demonstrated that pristine rGO was the any of the materials tested (Fig. 7d), but there was an increase in material of highest concern amongst the tested materials, and that the neutrophil numbers at day 1 after exposure to either abraded PA6-rGO immune cells and in particular the macrophages were the most or PA6 (Fig. 7e). Nevertheless, this neutrophil recruitment was not responsive in vitro models both with respect to abraded materials (neat stastistically significant compared to the negative control. In contrast, a PA6 and PA6-RGO) and pristine rGO. Amongst all exposure routes, significant increase in neutrophils was found after exposure to rGO- inhalation is the most rapid route of entry into the body and the only 15 µg, whereas no difference was observed for the lowest dose (rGO- route in which macrophages would interact immediately, and without 0.3 µg) compared to the negative control. Interestingly, 7 days after any intervening barrier, with the materials, should these respirable exposure, an increase in eosinophils was found in the alveolar space of materials reach the alveoli. Considering that macrophages and neutro­ mice treated with either PA6-rGO (p = 0.1217) or rGO-15 µg phils along with epithelial cells are the key drivers of the pulmonary (p = 0.0543). However, after 28 days, we did not detect any difference response to respirable materials reaching the alveoli, we sought to study in eosinophils or neutrophils compared to the negative control for any of whether the rGO results obtained in vitro would translate into a the materials tested. This suggested a recovery from the inflammatory macrophage-driven inflammatory response in the lungs of mice. Based stressor and is in agreement with the histological evaluation. Despite on the in vitro results, we anticipated that abraded (neat PA6 and PA6- this recovery, the presence of multinucleated macrophages was still rGO) materials would cause limited inflammation in vivo while the significant at 28 days for rGO-15 µg in comparison to the negative response to pristine rGO would be more pronounced but still moderate. control or other conditions, suggesting that the higher amount of ma­ Moreover, the comparative results obtained using epithelial A549 lung terials was delaying clearance and complete resolution. cells (negative) and THP-1/RAW264.7/HL-60 immune cells (positive) To further characterize the lung inflammatory response, the con­ suggested that the pulmonary response to rGO would be mostly driven centration of 13 inflammatory mediators was evaluated in digested by the resident macrophages, but would not prevent the resolution of lungs at days 1, 7 and 28. As shown in Fig. 8, there was no significant this inflammation with time. change in any of the mediators tested, suggesting a low inflammatory 18 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 Fig. 7. Animals (n = 3) were exposed by oropharyngeal aspiration to abraded polymer (PA6, 15 µg), abraded composite (PA6-rGO, 15 µg; with 2.5% rGO, hence 0.375 µg of rGO in 15 µg of PA6-rGO), reduced graphene oxide (rGO, 0.3 µg or 15 µg; 2.5% of 15 µg equals to about 0.3 µg), or negative control (BSA 0.1% in water). (a) Representative images of H&E-stained lung sections from mice exposed to rGO and abraded composites, following 1, 7 and 28 days after oropharyngeal aspi­ ration. Arrows indicate the formation of granulomas after treatment with rGO. (b) Protein concentration and (c) LDH activity in BAL fluid of exposed mice(d) Total cell number and (e) cell population (%) of BAL samples from mice after 1, 7 and 28 days following material administration. Data are presented as mean ± SD. The * symbol represents p < 0.05 as compared to the negative control. 19 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 Fig. 8. Inflammatory response to rGO and abraded composite materials (neat PA6, PA6-rGO) in the lungs of mice after 1, 7 and 28 days. Individual data points corresponding to each animal are plotted alongside mean values. Data are presented as mean ± SD. profile for all the materials tested. In addition, the concentration of TNF- after exposure to abraded PA6 at 28 days (Figs. 7 and 8) was not α and IL-6 was assessed in the BAL fluid and found to be below the anticipated. The discrepancy could not be attributed to technical or detection limit (7.8 pg/mL) of the assay (SI 16). Taken together, neither instrumental errors as the measurements were repeated and obtained BAL fluid nor whole lung measurements showed significant cytokine the same results. In addition, the health and behaviour of this specific secretion, which is in line with the moderate influx of immune cells in animal before the terminal procedure was not of concern (data not BAL fluids (PMNs, macrophages, or lymphocytes), irrespective of the shown); hence, there was no valid reason for removing the animal from materials considered (Fig. 7e). the study. The observed odd response should therefore be attributed to It should be noted that the unaligned significant response observed animal variability. for one animal (out of 3) for all tested cytokines (and total cell number) In summary, pulmonary exposure to abraded PA6-rGO in mice 20 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 induced a weak but not significant lung inflammation characterized by a after inhalation. With our comprehensive study, we provide clarification transient influx of neutrophils and eosinophils in the alveolar space. In on what is released qualitatively and quantitatively and its impact on contrast, a high dose of pristine rGO induced a persistent presence of human health. immune cells in the alveolar space, in particular macrophages associated with the clearance of materials, without affecting the overall resolution Supporting Information of inflammation with time. Importantly, this inflammation in the form of cell recruitment but without cytokine secretion was not associated with The Supporting Information is available free of charge. any tissue remodelling within the time-frame investigated. Since the Raman spectra, and FTIR patterns of GO, rGO and functionalised administrated bolus dose of 15 µg is considered a high dose, represent­ rGO, XPS spectra and TGA analysis of functionalized rGO, representative ing a worst-case scenario, the limited animal findings underline the TEM micrographs of rGO after the APTES functionalization, represen­ likely low toxicity of the tested abraded composites or the pristine rGO, tative AFM and optical microscopy images of rGO and PA6-rGO, Lateral in any realistic exposure scenario. Overall, the in vivo results are in good dimension and thickness distribution analysis for rGO and PA6-rGO, agreement with the in vitro biological assessment data, in particular the EPR spectra of the powder of PA6, rGO and PA6-rGO and assessment lung epithelial cells (A549), the macrophages (THP-1 and RAW 264.7) of ROS, particle size distributions of PA6-rGO and PA6 particles and neutrophils (HL-60), for which no or limited effects were observed measured by SMPS, TNF-α Expression Test (TET) for endotoxin detec­ in the tested biological systems upon exposure to PA6-rGO or rGO. At tion in HMDMs exposed to PA6, PA6-rGO, and rGO, representative SEM the same time, the in vivo results highlight the limitations of single cell images of the crack surface of the freeze-fractured samples of neat PA6 type in vitro models as alternative models to animals for predictive and PA6-rGO composite, representative TEM image of the abraded toxicology and support the current trend to adopt co-culture in vitro particles from the PA6-rGO composite showing the protruding rGOs models to better replicate in vivo outcomes. Future studies should from PA6 matrix, IL-6 and IL-8 release after DQ exposure in A549 lung therefore investigate the applicability of co-cultures of A549/dTHP-1 cells, LDH release from RAW 264.7 macrophages after exposure to rGO cells, lung-on-a-chip, or whole alveoli organoid models for the and abraded PA6 and PA6-rGO for 24 h, IFN-γ, TNF-α and IL-1β release screening of GRMs. by THP-1 macrophages after incubation with 20 µg/mL of PA6, PA6- rGO and rGO, cell morphology of RAW 264.7 macrophages exposed to 5. Conclusions different concentrations of rGO and abraded PA6 and PA6-rGO com­ posites, cytokine profiling and hierarchical cluster analysis of inflam­ Numerous studies have addressed the hazard potential of GRMs matory mediators released in differentiated HL-60 neutrophils exposed using a variety of model systems (Fadeel et al., 2018). These studies to rGO, abraded PA6-rGO and PA6, interference assessment of the attest to the key importance of defining the properties of the tested abraded particles (PA6 and PA6-rGO) and rGO with the MTS assay. materials including the lateral dimensions, thickness, number of layers, and C:O ratio (Wick et al., 2014). It is noted that most, if not all of these CRediT authorship contribution statement studies have been carried out using pristine (as-produced) materials. However, this may not accurately reflect the actual handling and use of Savvina Chortarea: Writing – original draft, Investigation, Formal GRMs during their life cycle as components of GRM-enabled products analysis, Data curation. Ogul Can Kuru: investigation. Woranan Net­ (Guineé et al., 2017). In a previous study conducted using multi-walled kueakul: Methodology, Investigation. Marco Pelin: Writing – review & carbon nanotubes (MWCNTs), the as-produced MWCNTs were shown to editing, Methodology, Data Curation, Investigation, Validation, Formal elicit inflammation when administered to mice via pharyngeal aspira­ analysis, Data Curation. Sandeep Keshavan: Methodology, Investiga­ tion whereas the aerosols generated from sanding of composites con­ tion, Formal analysis, Data Curation. Zhengmei Song: Methodology, taining polymer-coated MWCNTs, representative of the actual Investigation. Baojin Ma: Methodology, Investigation. Julio Gómes: end-product, did not exert such toxicity (Bishop et al., 2017). Herein, Data curation, Investigation. Elvira Villaro Abalos: Data Curation, we applied similar product life cycle considerations to assess the toxi­ Investigation. Luis Augusto Visani de Luna: Methodology, Investiga­ cological impact of rGO in the context of its embedding in a thermo­ tion. Thomas Loret: investigation. Alexander Fordham: Investigation. plastic PA6 polymer matrix and its potential release during the life cycle Matthew Drummond: investigation. Nikolaos Kontis: Methodology, of rGO-reinforced PA6 composites. A multi-endpoint comparison was Investigation. George Anagnostopoulos: Methodology, Investigation. performed between as-produced rGO, abraded PA6-rGO composite and George Paterakis: Methodology, Investigation. Pietro Cataldi: Inves­ abraded neat PA6 using a panel of robust and commonly used in vitro tigation. Aurelia Tubaro: Investigation. Costas Galiotis: Conceptuali­ models as well as a mouse model of pulmonary exposure. Overall, the zation, Supervision, Writing – review & editing, Resources, Data present findings show a negligible impact of rGO-reinforced PA6 com­ curation, Funding acquisition. Ian Kinloch: Data curation, Supervision, posites on all the models tested, suggesting a likely low risk to human Writing – review & editing. Bengt Fadeel: Conceptualization, Super­ health at acute exposure conditions. Nevertheless, the results found for vision, Writing – review & editing, validation, resources, project rGO alone, in particular in vivo, suggest that long-term effects after administration, data curation, funding acquisition. Cyrill Bussy: Su­ repeated exposure cannot be excluded and further studies are required pervision, Writing – review & editing, Validation, Project administra­ to address the possible chronic impact of rGO-reinforced composite tion. Kostas Kostarelos: Supervision, Writing – review & editing, materials. Resources, Data curation, Funding acquisition. Tina Buerki-Thurn­ herr: Conceptualization, Visualization, Supervision, Validation, Writing Environmental implication statement for the manuscript – review & editing, Data curation. Maurizio Prato: Conceptualization, Supervision, Writing – review & editing, Resources, Project adminis­ ’Hazard assessment of abraded thermoplastic composites reinforced tration, Funding acquisition. Alberto Bianco: Conceptualization, Su­ with reduced graphene oxide’. pervision, Writing – review & editing, Resources, Project The majority of graphene production is used to reinforce polymers. administration, Funding acquisition. Peter Wick: Conceptualization, Despite considerable progress made in recent years in terms of graphene Supervision, Writing – review & editing, resources, validation, project safety assessment, only limited information is available concerning the administration, Funding acquisition. release and hazard potential of graphene-containing products after mechanical treatment. Airborne particles released into the environment Declaration of Competing Interest will get into contact with human beings. There are some indications that these released materials might have a hazardous effects in particular The authors declare that they have no known competing financial 21 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 interests or personal relationships that could have appeared to influence Lupoi, J.S., Lersch, T.L., Stefaniak, A.B., Sager, T., Afshari, A., Schwegler-Berry, D., Friend, S.;, 2017. In Vivo toxicity assessment of occupational components of the the work reported in this paper. carbon nanotube life cycle to provide context to potential health effects. In: ACS Nano, 11, pp. 8849–8863. https://doi.org/10.1021/acsnano.7b03038. Acknowledgments Bitounis, D., Parviz, D., Cao, X., Amadei, C.A., Vecitis, C.D., Sunderland, E.M., Thrall, B. D., Fang, M., Strano, M.S., Demokritou, P., 2020. Synthesis and physicochemical transformations of size-sorted graphene oxide during simulated digestion and its This work was supported by the European Union (EU) 8th Frame­ toxicological assessment against an in Vitro model of the human intestinal work Program for Research and Technological Development, Graphene epithelium. Small 16 (21), 1907640. https://doi.org/10.1002/smll.201907640. Flagship project (H2020-FET- GrapheneCore2 - #785219 and H2020- Blank, F., Rothen-Rutishauser, B.M., Schurch, S., Gehr, P., 2006. An Optimized in Vitro model of the respiratory tract wall to study particle cell interactions. J. Aerosol Med. FET- GrapheneCore3 - #881603). Part of this work was supported by the 19 (3), 392–405. https://doi.org/10.1089/jam.2006.19.392. University of Trieste, the Agence Nationale de la Recherche (ANR) Böhmert, L., König, L., Sieg, H., Lichtenstein, D., Paul, N., Braeuning, A., Voigt, A., through the LabEx project Chemistry of Complex Systems (ANR-10- Lampen, A., 2018. In Vitro nanoparticle dosimetry for adherent growing cell monolayers covering bottom and lateral walls. Part. Fibre Toxicol. 15 (1), 1–20. LABX- 0026_CSC), the Maria de Maeztu Units of Excellence Program https://doi.org/10.1186/S12989-018-0278-9. from the Spanish State Research Agency – Grant No. MDM-2017-0720 Boraschi, D., Italiani, P., Palomba, R., Decuzzi, P., Duschl, A., Fadeel, B., Moghimi, S.M., and the Swiss National Science Foundation (Grant 310030_169207). 2017. Nanoparticles and innate immunity: new perspectives on host defence. Semin. Immun. 33–51. https://doi.org/10.1016/j.smim.2017.08.013. We would like to acknowledge the staff at the Faculty of Biology, Bortz, D.R., Heras, E.G., Martin-Gullon, I., 2012. Impressive fatigue life and fracture Medicine, and Health, the University of Manchester, particularly P. toughness improvements in graphene oxide/epoxy composites. Macromolecules 45 Walker and G.Bako (Histology facility), Dr G. Howell (Flow Cytometry (1), 238–245. https://doi.org/10.1021/ma201563k. Carosio, F., Maddalena, L., Gomez, J., Saracco, G., Fina, A., 2018. Graphene oxide Facility), C. Morrison (Genomic Facility), Dr P. March (Bioimaging Fa­ exoskeleton to produce self-extinguishing, nonignitable, and flame resistant flexible cility), all staff members at the Biological Services Facility for expert foams: a mechanically tough alternative to inorganic aerogels. Adv. Mater. Interfaces advice and assistance in using core facility, and their provision of 5 (23), 1801288. https://doi.org/10.1002/admi.201801288. Carr, K.E., Smyth, S.H., McCullough, M.T., Morris, J.F., Moyes, S.M., 2012. technical support services. We also wish to acknowledge the Centre Morphological aspects of interactions between microparticles and mammalian cells: National de la Recherche Scientifique (CNRS) and the International intestinal uptake and onward movement. Prog. Histochem. Cytochem. 185–252. Center for Frontier Research in Chemistry (icFRC). https://doi.org/10.1016/j.proghi.2011.11.001. The Bioimaging Facility microscopes used in the in vivo part of the Chang, Y., Yang, S.T., Liu, J.H., Dong, E., Wang, Y., Cao, A., Liu, Y., Wang, H., 2011. In Vitro toxicity evaluation of graphene oxide on A549 cells. Toxicol. Lett. 200 (3), study were purchased with grants from the UKRI Biotechnology and 201–210. https://doi.org/10.1016/j.toxlet.2010.11.016. Biological Sciences Research Council (UKRI-BBSRC), the Wellcome Chen, G.Y., Yang, H.J., Lu, C.H., Chao, Y.C., Hwang, S.M., Chen, C.L., Lo, K.W., Sung, L. Trust, and the University of Manchester Strategic Fund. Y., Luo, W.Y., Tuan, H.Y., Hu, Y.C., 2012. Simultaneous induction of autophagy and toll-like receptor signaling pathways by graphene oxide. Biomaterials 33 (27), The graphics for the Table of Contents (ToC) and for Fig. 1 were 6559–6569. https://doi.org/10.1016/j.biomaterials.2012.05.064. created using BioRender.com. Chortarea, S., Barosova, H., Clift, M.J.D., Wick, P., Petri-Fink, A., Rothen-Rutishauser, B., 2017. Human asthmatic bronchial cells are more susceptible to subchronic repeated exposures of aerosolized carbon nanotubes at occupationally relevant doses than Appendix A. Supporting information healthy cells. ACS Nano 11 (8), 7615–7625. https://doi.org/10.1021/ acsnano.7b01992. Supplementary data associated with this article can be found in the Chortarea, S., Barosova, H., Clift, M.J.D., Wick, P., Petri-Fink, A., Rothen-Rutishauser, B., 2017. Human asthmatic bronchial cells are more susceptible to subchronic repeated online version at doi:10.1016/j.jhazmat.2022.129053. exposures of aerosolized carbon nanotubes at occupationally relevant doses than healthy cells. ACS Nano 11 (8), 7615–7625. https://doi.org/10.1021/ References acsnano.7b01992. Chortarea, S., Clift, M.J.D., Vanhecke, D., Endes, C., Wick, P., Petri-Fink, A., Rothen- Rutishauser, B., 2015. Repeated exposure to carbon nanotube-based aerosols does Accioly, M.T., Pacheco, P., Maya-Monteiro, C.M., Carrossini, N., Robbs, B.K., Oliveira, S. not affect the functional properties of a 3D human epithelial airway model. S., Kaufmann, C., Morgado-Diaz, J.A., Bozza, P.T., Viola, J.P.B., 2008. Lipid bodies Nanotoxicology 5390, 1–11. are reservoirs of cyclooxygenase-2 and sites of prostaglandin-E2 synthesis in colon Chortarea, S., Zerimariam, F., Barosova, H., Septiadi, D., Clift, M.J.D., Petri-Fink, A., cancer cells. Cancer Res. 68 (6), 1732–1740. https://doi.org/10.1158/0008-5472. Rothen-Rutishauser, B., 2019. Profibrotic activity of multiwalled carbon nanotubes CAN-07-1999. upon prolonged exposures in different human lung cell types. Appl. Vitr. Toxicol. 5 Ali, G., Mehmood, A., Ha, H.Y., Kim, J., Chung, K.Y., 2017. Reduced graphene oxide as a (1), 47–61. https://doi.org/10.1089/aivt.2017.0033. stable and high-capacity cathode material for na-ion batteries. Sci. Rep. 7 (2), 1–8. Clouter, A., Brown, D., Höhr, D., Borm, P., Donaldson, K., 2001. Inflammatory effects of https://doi.org/10.1038/srep40910. respirable quartz collected in workplaces versus standard DQ12 quartz: particle Anderson, J.M., Rodriguez, A., Chang, D.T., 2008. Foreign body reaction to biomaterials. surface correlates. Toxicol. Sci. 63 (1), 90–98. Sem. Immun. 86–100. https://doi.org/10.1016/j.smim.2007.11.004. CY, W., GM, D., A, P., P, D., 2016. Buoyant nanoparticles: implications for nano- Barosova, H., Chortarea, S., Peikertova, P., Clift, M.J.D., Petri-Fink, A., Kukutschova, J., biointeractions in cellular studies. Small 12 (23), 3172–3180. https://doi.org/ Rothen-Rutishauser, B., 2018. Biological response of an in vitro human 3D lung cell 10.1002/SMLL.201600314. model exposed to brake wear debris varies based on brake pad formulation. Arch. DeLoid, G., Cohen, J.M., Darrah, T., Derk, R., Rojanasakul, L., Pyrgiotakis, G., Toxicol. 92 (7), 2339–2351. https://doi.org/10.1007/s00204-018-2218-8. Wohlleben, W., Demokritou, P., 2014. Estimating the effective density of engineered Barosova, H., Maione, A.G., Septiadi, D., Sharma, M., Haeni, L., Balog, S., O’Connell, O., nanomaterials for in vitro dosimetry. Nat. Commun. 5 (1), 1–10. https://doi.org/ Jackson, G.R., Brown, D., Clippinger, A.J., Hayden, P., Petri-Fink, A., Stone, V., 10.1038/ncomms4514. Rothen-Rutishauser, B., 2020. Use of EpiAlveolar lung model to predict fibrotic DeLoid, G.M., Cohen, J.M., Pyrgiotakis, G., Pirela, S.V., Pal, A., Liu, J., Srebric, J., potential of multiwalled carbon nanotubes. ACS Nano 14 (4), 3941–3956. https:// Demokritou, P., 2015. Advanced computational modeling for in Vitro nanomaterial doi.org/10.1021/acsnano.9b06860. dosimetry. Part. Fibre Toxicol. 12 (1), 1–20. https://doi.org/10.1186/S12989-015- Beams, R., Gustavo Cançado, L., Novotny, L., 2015. Raman characterization of defects 0109-1. and dopants in graphene. J. Phys. Condens. Matter, 083002. https://doi.org/ Dias, F.T.G., Ingracio, A.R., Nicoletti, N.F., Menezes, F.C., Dall Agnol, L., Marinowic, D. 10.1088/0953-8984/27/8/083002. R., Soares, R.M.D., da Costa, J.C., Falavigna, A., Bianchi, O., 2019. Soybean-modified Bengtson, S., Knudsen, K.B., Kyjovska, Z.O., Berthing, T., Skaug, V., Levin, M., polyamide-6 mats as a long-term cutaneous wound covering. Mater. Sci. Eng. C. 99, Koponen, I.K., Shivayogimath, A., Booth, T.J., Alonso, B., Pesquera, A., Zurutuza, A., 957–968. https://doi.org/10.1016/j.msec.2019.02.019. Thomsen, B.L., Troelsen, J.T., Jacobsen, N.R., Vogel, U., 2017. Differences in Domenech, J., Hernández, A., Demir, E., Marcos, R., Cortés, C., 2020. Interactions of inflammation and acute phase response but similar genotoxicity in mice following graphene oxide and graphene nanoplatelets with the in Vitro Caco-2/HT29 model of pulmonary exposure to graphene oxide and reduced graphene oxide. PLoS One 12 intestinal barrier. Sci. Rep. 10 (1), 1–15. https://doi.org/10.1038/s41598-020- (6), e0178355. https://doi.org/10.1371/journal.pone.0178355. 59755-0. Beyeler, S., Chortarea, S., Rothen-Rutishauser, B., Petri-Fink, A., Wick, P., Tschanz, S.A., Drasler, B., Kucki, M., Delhaes, F., Buerki-Thurnherr, T., Vanhecke, D., Korejwo, D., von Garnier, C., Blank, F., 2018. Acute effects of multi-walled carbon nanotubes on Chortarea, S., Barosova, H., Hirsch, C., Petri-Fink, A., Rothen-Rutishauser, B., primary bronchial epithelial cells from COPD Patients. Nanotoxicology 12 (7), Wick, P., 2018. Single exposure to aerosolized graphene oxide and graphene 699–711. https://doi.org/10.1080/17435390.2018.1472310. nanoplatelets did not initiate an acute biological response in a 3D human lung Bhattacharya, K., Kiliç, G., Costa, P.M., Fadeel, B., 2017. Cytotoxicity screening and model. Carbon N. Y 137, 125–135. https://doi.org/10.1016/J. cytokine profiling of nineteen nanomaterials enables hazard ranking and grouping CARBON.2018.05.012. based on inflammogenic potential. Nanotoxicology 11 (6), 809–826. https://doi. Duncan, T.V., Pillai, K., 2015. Release of engineered nanomaterials from polymer org/10.1080/17435390.2017.1363309. nanocomposites: diffusion, dissolution, and desorption. ACS Appl. Mater. Interfaces Bishop, L., Cena, L., Orandle, M., Yanamala, N., Dahm, M.M., Eileen Birch, M., Evans, D. 2–19. https://doi.org/10.1021/am5062745. E., Kodali, V.K., Eye, T., Battelli, L., Zeidler-Erdely, P.C., Casuccio, G., Bunker, K., 22 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 Eckmann, A., Felten, A., Verzhbitskiy, I., Davey, R., Casiraghi, C., 2013. Raman study on Gong, L., Yin, B., Li, L.P., Yang, M.B., 2015. Nylon-6/Graphene composites modified defective graphene: effect of the excitation energy, type, and amount of defects. through polymeric modification of graphene. Compos. Part B Eng. 73, 49–56. Phys. Rev. B Condens. Matter Mater. Phys. 88 (3), 035426 https://doi.org/10.1103/ https://doi.org/10.1016/j.compositesb.2014.12.009. PhysRevB.88.035426. Guarnieri, D., Sánchez-Moreno, P., Del Rio Castillo, A.E., Bonaccorso, F., Gatto, F., EC, C., Q, Z., Y, X., 2011. The effect of sedimentation and diffusion on cellular uptake of Bardi, G., Martín, C., Vázquez, E., Catelani, T., Sabella, S., Pompa, P.P., 2018. gold nanoparticles. Nat. Nanotechnol. 6 (6), 385–391. https://doi.org/10.1038/ Biotransformation and biological interaction of graphene and graphene oxide during NNANO.2011.58. simulated oral ingestion. Small 14 (24), 1800227. https://doi.org/10.1002/ Elliott, J.T., Rösslein, M., Song, N.W., Toman, B., Kinsner-Ovaskainen, A., smll.201800227. Maniratanachote, R., Salit, M.L., Petersen, E.J., Sequeira, F., Romsos, E.L., Kim, S.J., Guineé, J.B., Heijungs, R., Vijver, M.G., Peijnenburg, W.J.G.M., 2017. Setting the stage Lee, J., Von Moos, N.R., Rossi, F., Hirsch, C., Krug, H.F., Suchaoin, W., Wick, P., for debating the roles of risk assessment and life-cycle assessment of engineered 2017. Toward achieving harmonization in a nanocytotoxicity assay measurement nanomaterials. Nat. Nanotech. 727–733. https://doi.org/10.1038/ through an interlaboratory comparison study. ALTEX 34 (2), 201–218. https://doi. NNANO.2017.135. org/10.14573/altex.1605021. Guo, F., Creighton, M., Chen, Y., Hurt, R., Külaots, I., 2014. Porous structures in stacked, Endes, C., Schmid, O., Kinnear, C., Mueller, S., Camarero-Espinosa, S., Vanhecke, D., crumpled and pillared graphene-based 3D materials. Carbon N. Y 66, 476–484. Foster, E.J., Petri-Fink, A., Rothen-Rutishauser, B., Weder, C., Clift, M.J., 2014. An in https://doi.org/10.1016/j.carbon.2013.09.024. Vitro testing strategy towards mimicking the inhalation of high aspect ratio Gurunathan, S., Kang, M.-H., Jeyaraj, M., Kim, J.-H., 2019. Differential nanoparticles. Part. Fibre Toxicol. 11 (1), 40. https://doi.org/10.1186/s12989-014- Immunomodulatory effect of graphene oxide and vanillin-functionalized graphene 0040-x. oxide nanoparticles in human acute monocytic leukemia cell line (THP-1). Int. J. Endes, C., Schmid, O., Kinnear, C., Mueller, S., Camarero-Espinosa, S., Vanhecke, D., Mol. Sci. 20 (2), 247. https://doi.org/10.3390/ijms20020247. Foster, E.J., Petri-Fink, A., Rothen-Rutishauser, B., Weder, C., Clift, M.J.D., 2014. An Hänel, K.H., Cornelissen, C., Lüscher, B., Baron, J.M., 2013. Cytokines and the skin in Vitro testing strategy towards mimicking the inhalation of high aspect ratio barrier. Int. J. Mol. Sci. 6720–6745. https://doi.org/10.3390/ijms14046720. nanoparticles. Part. Fibre Toxicol. 11 (1) https://doi.org/10.1186/s12989-014- Hansen, G.H., Niels-Christiansen, L.-L., Immerdal, L., Nystrøm, B.T., Danielsen, E.M., 0040-x. 2007. Intestinal alkaline phosphatase: selective endocytosis from the enterocyte Fadeel, B., Åhlin, A., Henter, J.I., Orrenius, S., Hampton, M.B., 1998. Involvement of brush border during fat absorption. Am. J. Physiol. Liver Physiol. 293 (6), caspases in neutrophil apoptosis: regulation by reactive oxygen species. Blood 92 G1325–G1332. https://doi.org/10.1152/ajpgi.00379.2007. (12), 4808–4818. https://doi.org/10.1182/blood.v92.12.4808.424k01_4808_4818. Hempt, C., Hirsch, C., Hannig, Y., Rippl, A., Wick, P., Buerki-Thurnherr, T., 2020. Fadeel, B., Bussy, C., Merino, S., Vázquez, E., Flahaut, E., Mouchet, F., Evariste, L., Investigating the effects of differently produced synthetic amorphous silica (E 551) Gauthier, L., Koivisto, A.J., Vogel, U., Martín, C., Delogu, L.G., Buerki-Thurnherr, T., on the integrity and functionality of the human intestinal barrier using an advanced Wick, P., Beloin-Saint-Pierre, D., Hischier, R., Pelin, M., Candotto Carniel, F., in Vitro co-culture model. Arch. Toxicol. 1, 3. https://doi.org/10.1007/s00204-020- Tretiach, M., Cesca, F., 2018. Safety assessment of graphene-based materials: focus 02957-2. on human health and the environment. ACS Nano 10582–10620. https://doi.org/ Irfan, A., Sachse, S., Njuguna, J., Pielichowski, K., Silva, F., Zhu, H., 2013. Assessment of 10.1021/acsnano.8b04758. nanoparticle release from polyamide 6- and polypropylene-silicon composites and Farcal, L., Andón, F.T., Di Cristo, L., Rotoli, B.M., Bussolati, O., Bergamaschi, E., cytotoxicity in human lung A549 cells. J. Inorg. Organomet. Polym. Mater. 23 (4), Mech, A., Hartmann, N.B., Rasmussen, K., Riego-Sintes, J., Ponti, J., Kinsner- 861–870. https://doi.org/10.1007/s10904-013-9856-3. Ovaskainen, A., Rossi, F., Oomen, A., Bos, P., Chen, R., Bai, R., Chen, C., Rocks, L., Jian, L., Ning, L., Yang, S., Wang, J., Hua, M., 2009. Triboelectrification electrostatic Fulton, N., 2015. Comprehensive in Vitro toxicity testing of a panel of representative potential of MC Nylon 6 under point contact dry sliding. Tribol. Lett. 36 (3), oxide nanomaterials: first steps towards an intelligent testing strategy. PLoS One 10 199–208. https://doi.org/10.1007/s11249-009-9473-z. (5). https://doi.org/10.1371/journal.pone.0127174. Keshavan, S., Calligari, P., Stella, L., Fusco, L., Delogu, L.G., Fadeel, B., 2019. Nano-bio Ferrari, A.C., Bonaccorso, F., Fal’ko, V., Novoselov, K.S., Roche, S., Bøggild, P., Borini, S., interactions: a neutrophil-centric view. Cell Death Dis. 1–11. https://doi.org/ Koppens, F.H.L., Palermo, V., Pugno, N., Garrido, J.A., Sordan, R., Bianco, A., 10.1038/s41419-019-1806-8. Ballerini, L., Prato, M., Lidorikis, E., Kivioja, J., Marinelli, C., 2015. Science and Khandagale, A., Lazzaretto, B., Carlsson, G., Sundin, M., Shafeeq, S., Römling, U., technology roadmap for graphene, related two-dimensional crystals, and hybrid Fadeel, B., 2018. JAGN1 Is required for fungal killing in neutrophil extracellular systems. Nanoscale 7 (11), 4598–4810. https://doi.org/10.1039/C4NR01600A. traps: implications for severe congenital neutropenia. J. Leukoc. Biol. 1–15. https:// Ferreiro, V., Depecker, C., Laureyns, J., Coulon, G., 2004. Structures and morphologies of doi.org/10.1002/JLB.4A0118-030RR. cast and plastically strained polyamide 6 films as evidenced by confocal raman Kong, W., Kum, H., Bae, S.H., Shim, J., Kim, H., Kong, L., Meng, Y., Wang, K., Kim, C., microspectroscopy and atomic force microscopy. Polym. (Guildf. ) 45 (17), Kim, J., 2019. Path towards graphene commercialization from lab to market. Nat. 6013–6026. https://doi.org/10.1016/j.polymer.2004.06.018. Nanotechnol. 927–938. https://doi.org/10.1038/s41565-019-0555-2. Froggett, S.J., Clancy, S.F., Boverhof, D.R., Canady, R.A., 2014. A review and perspective Krishnamoorthy, K., Veerapandian, M., Mohan, R., Kim, S.J., 2012. Investigation of of existing research on the release of nanomaterials from solid nanocomposites. Part Raman and photoluminescence studies of reduced graphene oxide sheets. Appl. Fibre Toxicol. 1–28. https://doi.org/10.1186/1743-8977-11-17. Phys. A Mater. Sci. Process. 106 (3), 501–506. https://doi.org/10.1007/s00339-011- Froggett, S.J., Clancy, S.F., Boverhof, D.R., Canady, R.A., 2014. A review and perspective 6720-6. of existing research on the release of nanomaterials from solid nanocomposites. Part. Krysiak, Z.J., Gawlik, M.Z., Knapczyk-Korczak, J., Kaniuk, L., Stachewicz, U., 2020. Fibre Toxicol. 1–28. https://doi.org/10.1186/1743-8977-11-17. Hierarchical composite meshes of electrospun PS microfibers with PA6 nanofibers Frontiñán-Rubio, J., Victoria Gómez, M., Martín, C., González-Domínguez, J.M., Durán- for regenerative medicine. Materials 13 (8), 1974. https://doi.org/10.3390/ Prado, M., Vázquez, E., 2018. Differential effects of graphene materials on the MA13081974. metabolism and Function of Human Skin Cells. Nanoscale 10 (24), 11604–11615. Kucki, M., Aengenheister, L., Diener, L., Rippl, A.V., Vranic, S., Newman, L., Vazquez, E., https://doi.org/10.1039/c8nr00897c. Kostarelos, K., Wick, P., Buerki-Thurnherr, T., 2018. Impact of Graphene Oxide on Fusco, L., Pelin, M., Mukherjee, S., Keshavan, S., Sosa, S., Martín, C., González, V., Human Placental Trophoblast Viability, Functionality and Barrier Integrity. 2D Vázquez, E., Prato, M., Fadeel, B., Tubaro, A., 2020. Keratinocytes are capable of Mater. 5 (3), 035014 https://doi.org/10.1088/2053–1583/aab9e2. selectively sensing low amounts of graphene-based materials: implications for Kucki, M., Diener, L., Bohmer, N., Hirsch, C., Krug, H.F., Palermo, V., Wick, P., 2017. cutaneous applications. Carbon N. Y 159, 598–610. https://doi.org/10.1016/j. Uptake of label-free graphene oxide by caco-2 cells is dependent on the cell carbon.2019.12.064. differentiation status. J. Nanobiotechnol. 15 (1), 46. https://doi.org/10.1186/ Fu, X., Yao, C., Yang, G., 2015. Recent advances in graphene/polyamide 6 composites: a s12951-017-0280-7. review. RSC Adv. 61688–61702. https://doi.org/10.1039/c5ra09312k. Kucki, M., Rupper, P., Sarrieu, C., Melucci, M., Treossi, E., Schwarz, A., León, V., Gangwal, S., Brown, J.S., Wang, A., Houck, K.A., Dix, D.J., Kavlock, R.J., Cohen Hubal, E. Kraegeloh, A., Flahaut, E., Vázquez, E., Palermo, V., Wick, P., 2016. Interaction of A., 2011. Informing selection of nanomaterial concentrations for toxcast in vitro graphene-related materials with human intestinal cells: an in Vitro approach. testing based on occupational exposure potential. Environ. Health Perspect. Nanoscale 8 (16), 8749–8760. https://doi.org/10.1039/c6nr00319b. 1539–1546. https://doi.org/10.1289/ehp.1103750. Kurapati, R., Russier, J., Squillaci, M.A., Treossi, E., Ménard-Moyon, C., Del Rio- Geim, A.K., Novoselov, K.S., 2007. The rise of graphene. Nat. Mater. 6 (3), 183–191. Castillo, A.E., Vazquez, E., Samorì, P., Palermo, V., Bianco, A., 2015. Dispersibility- https://doi.org/10.1038/nmat1849. dependent biodegradation of graphene oxide by myeloperoxidase. Small 11 (32), George, G., Sisupal, S.B., Tomy, T., Kumaran, A., Vadivelu, P., Suvekbala, V., Sivaram, S., 3985–3994. https://doi.org/10.1002/smll.201500038. Ragupathy, L., 2018. Facile, environmentally benign and scalable approach to Lankone, R.S., Ruggiero, E., Goodwin, D.G., Vilsmeier, K., Mueller, P., Pulbere, S., produce pristine few layers graphene suitable for preparing biocompatible polymer Challis, K., Bi, Y., Westerhoff, P., Ranville, J., Fairbrother, D.H., Sung, L.-P., nanocomposites. Sci. Rep. 8 (1), 11228. https://doi.org/10.1038/s41598-018- Wohlleben, W., 2020. Evaluating performance, degradation, and release behavior of 28560-1. a nanoform pigmented coating after natural and accelerated weathering. Golchin, A., Simmons, G., Glavatskih, S., Prakash, B., 2013. Tribological behaviour of NanoImpact 17. polymeric materials in water-lubricated contacts. Proc. Inst. Mech. Eng. Part J. J. Lin, H., Song, Z., Bianco, A., 2021. How macrophages respond to two-dimensional Eng. Tribol. 227 (8), 811–825. https://doi.org/10.1177/1350650113476441. materials: a critical overview focusing on toxicity. J. Environ. Sci. Health 333–356. Gomez, J., Villaro, E., Karagiannidis, P.G., Elmarakbi, A., 2020. Effects of chemical https://doi.org/10.1080/03601234.2021.1885262. structure and morphology of graphene-related materials (GRMs) on melt processing Liu, Y., Luo, Y., Wu, J., Wang, Y., Yang, X., Yang, R., Wang, B., Yang, J., Zhang, N., 2013. and properties of GRM/Polyamide-6 nanocomposites. Results Mater. 7, 100105 Graphene oxide can induce in Vitro and in Vivo mutagenesis. Sci. Rep. 3. https://doi. https://doi.org/10.1016/j.rinma.2020.100105. org/10.1038/srep03469. Gómez, J., Villaro, E., Navas, A., Recio, I., 2017. Testing the influence of the Li, Y., Boraschi, D., 2016. Endotoxin contamination: a key element in the interpretation temperature, RH and filler type and content on the universal power law for new of nanosafety studies. Nanomedicine 269–287. https://doi.org/10.2217/ reduced graphene oxide TPU composites. Mater. Res. Express 4 (10), 105020. nnm.15.196. https://doi.org/10.1088/2053-1591/aa8e11. 23 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 Li, R., Guiney, L.M., Chang, C.H., Mansukhani, N.D., Ji, Z., Wang, X., Liao, Y.P., Ou, L., Song, B., Liang, H., Liu, J., Feng, X., Deng, B., Sun, T., Shao, L., 2016. Toxicity of Jiang, W., Sun, B., Hersam, M.C., Nel, A.E., Xia, T., 2018. Surface oxidation of graphene-family nanoparticles: a general review of the origins and mechanisms. Part graphene oxide determines membrane damage, lipid peroxidation, and cytotoxicity Fibre Toxicol. 1–24. https://doi.org/10.1186/s12989-016-0168-y. in macrophages in a pulmonary toxicity model. ACS Nano 12 (2), 1390–1402. O’Neill, A., Bakirtzis, D., Dixon, D., 2014. Polyamide 6/graphene composites: the effect https://doi.org/10.1021/acsnano.7b07737. of in situ polymerisation on the structure and properties of graphene oxide and Li, J., Wang, X., Mei, K.C., Chang, C.H., Jiang, J., Liu, X., Liu, Q., Guiney, L.M., reduced graphene oxide. Eur. Polym. J. 59, 353–362. https://doi.org/10.1016/j. Hersam, M.C., Liao, Y.P., Meng, H., Xia, T., 2021. Lateral size of graphene oxide eurpolymj.2014.07.038. determines differential cellular uptake and cell death pathways in kupffer cells, Pant, H.R., Pant, B., Park, C.H., Kim, H.J., Lee, D.S., Tijing, L.D., Hwang, B.S., Kim, H.Y., LSECs, and hepatocytes. Nano Today 37, 101061. https://doi.org/10.1016/J. Kim, C.S., 2013. RGO/Nylon-6 composite mat with unique structural features and NANTOD.2020.101061. electrical properties obtained from electrospinning and hydrothermal process. Fibers Li, B., Yang, J., Huang, Q., Zhang, Y., Peng, C., Zhang, Y., He, Y., Shi, J., Li, W., Hu, J., Polym. 14 (6), 970–975. https://doi.org/10.1007/s12221-013-0970-1. Fan, C., 2013. Biodistribution and pulmonary toxicity of intratracheally instilled Pelin, M., Fusco, L., León, V., Martín, C., Criado, A., Sosa, S., Vázquez, E., Tubaro, A., graphene oxide in mice. NPG Asia Mater. 5 (4) https://doi.org/10.1038/am.2013.7. Prato, M., 2017. Differential cytotoxic effects of graphene and graphene oxide on Massaq, A., Rusinek, A., Klosak, M., Bahi, S., Arias, A., 2019. Strain rate effect on the skin keratinocytes. Sci. Rep. 7 https://doi.org/10.1038/srep40572. mechanical behavior of polyamide composites under compression loading. Compos. Pelin, M., Fusco, L., Martín, C., Sosa, S., Frontiñán-Rubio, J., González-Domínguez, J.M., Struct. 214, 114–122. https://doi.org/10.1016/j.compstruct.2019.01.101. Durán-Prado, M., Vázquez, E., Prato, M., Tubaro, A., 2018. Graphene and graphene Ma, J., Liu, R., Wang, X., Liu, Q., Chen, Y., Valle, R.P., Zuo, Y.Y., Xia, T., Liu, S., 2015. oxide induce ROS production in human HaCaT skin keratinocytes: the role of Crucial role of lateral size for graphene oxide in activating macrophages and xanthine oxidase and NADH dehydrogenase. Nanoscale 10 (25), 11820–11830. stimulating pro-inflammatory responses in cells and animals. ACS Nano 9 (10), https://doi.org/10.1039/c8nr02933d. 10498–10515. https://doi.org/10.1021/acsnano.5b04751. Pelin, M., Lin, H., Gazzi, A., Sosa, S., Ponti, C., Ortega, A., Zurutuza, A., Vázquez, E., Mercer, R.R., Scabilloni, J.F., Hubbs, A.F., Wang, L., Battelli, L.A., McKinney, W., Prato, M., Tubaro, A., Bianco, A., 2020. Partial reversibility of the cytotoxic effect Castranova, V., Porter, D.W., 2013. Extrapulmonary transport of MWCNT following induced by graphene-based materials in skin keratinocytes. Nanomaterials 10 (8), inhalation exposure. Part. Fibre Toxicol. 10 (1), 38. https://doi.org/10.1186/1743- 1–14. https://doi.org/10.3390/nano10081602. 8977-10-38. Pelin, M., Sosa, S., Prato, M., Tubaro, A., 2018. Occupational exposure to graphene based Mittal, S., Kumar, V., Dhiman, N., Chauhan, L.K.S., Pasricha, R., Pandey, A.K., 2016. nanomaterials: risk assessment. Nanoscale 15894–15903. https://doi.org/10.1039/ Physico-chemical properties based differential toxicity of graphene oxide/reduced c8nr04950e. graphene oxide in human lung cells mediated through oxidative stress. Sci. Rep. 6 Pietroiusti, A., Bergamaschi, E., Campagna, M., Campagnolo, L., De Palma, G., (1), 1–16. https://doi.org/10.1038/srep39548. Iavicoli, S., Leso, V., Magrini, A., Miragoli, M., Pedata, P., Palombi, L., Iavicoli, I., Moalli, P.A., MacDonald, J.L., Goodglick, L.A., Kane, A.B., 1987. Acute injury and 2017. The unrecognized occupational relevance of the interaction between regeneration of the mesothelium in response to asbestos fibers. Am. J. Pathol. 128 engineered nanomaterials and the gastro-intestinal tract: a consensus paper from a (3), 426–445. multidisciplinary working group. Part. Fibre Toxicol. https://doi.org/10.1186/ Monteiller, C., Tran, L., MacNee, W., Faux, S., Jones, A., Miller, B., Donaldson, K., 2007. s12989-017-0226-0. The Pro-inflammatory effects of low-toxicity low-solubility particles, nanoparticles Poulsen, S.S., Bengtson, S., Williams, A., Jacobsen, N.R., Troelsen, J.T., Halappanavar, S., and fine particles, on epithelial cells in Vitro: the role of surface area. Occup. Environ. Vogel, U., 2021. A transcriptomic overview of lung and liver changes one day after Med. 64 (9), 609–615. https://doi.org/10.1136/oem.2005.024802. pulmonary exposure to graphene and graphene oxide. Toxicol. Appl. Pharmacol. Mouhmid, B., Imad, A., Benseddiq, N., Benmedakhène, S., Maazouz, A., 2006. A Study of 410, 115343 https://doi.org/10.1016/j.taap.2020.115343. the mechanical behaviour of a glass fibre reinforced polyamide 6,6: experimental Reiss, T., Hjelt, K., Ferrari, A.C., 2019. Graphene is on track to deliver on its promises. investigation. Polym. Test. 25 (4), 544–552. https://doi.org/10.1016/j. Nat. Nanotech. 907–910. https://doi.org/10.1038/s41565-019-0557-0. polymertesting.2006.03.008. Reshma, S.C., Syama, S., Mohanan, P.V., 2016. Nano-biointeractions of PEGylated and Mukaida, N., 2003. Pathophysiological roles of interleukin-8/CXCL8 in pulmonary bare reduced graphene oxide on lung alveolar epithelial cells: a comparative in Vitro diseases. Am. J. Physiol. Lung Cell. Mol. Physiol. https://doi.org/10.1152/ study. Colloids Surf. B Biointerfaces 140, 104–116. https://doi.org/10.1016/j. ajplung.00233.2002. colsurfb.2015.12.030. Mukherjee, S.P., Bottini, M., Fadeel, B., 2017. Graphene and the immune system: a Rincon, M., Irvin, C.G., 2012. Role of IL-6 in asthma and other inflammatory pulmonary romance of many dimensions. Fro. Immun. 1. https://doi.org/10.3389/ diseases. Intern. J. Biol. Sci. 1281–1290. https://doi.org/10.7150/ijbs.4874. fimmu.2017.00673. Roberts, J.R., Mercer, R.R., Stefaniak, A.B., Seehra, M.S., Geddam, U.K., Chaudhuri, I.S., Mukherjee, S.P., Gliga, A.R., Lazzaretto, B., Brandner, B., Fielden, M., Vogt, C., Kyrlidis, A., Kodali, V.K., Sager, T., Kenyon, A., Bilgesu, S.A., Eye, T., Scabilloni, J.F., Newman, L., Rodrigues, A.F., Shao, W., Fournier, P.M., Toprak, M.S., Star, A., Leonard, S.S., Fix, N.R., Schwegler-Berry, D., Farris, B.Y., Wolfarth, M.G., Porter, D. Kostarelos, K., Bhattacharya, K., Fadeel, B., 2018. Graphene oxide is degraded by W., Castranova, V., 2016. Evaluation of pulmonary and systemic toxicity following neutrophils and the degradation products are non-genotoxic. Nanoscale 10 (3), lung exposure to graphite nanoplates: a member of the graphene-based nanomaterial 1180–1188. https://doi.org/10.1039/c7nr03552g. family. Part. Fibre Toxicol. 13 (1), 1–22. https://doi.org/10.1186/s12989-016- Mukherjee, S.P., Kostarelos, K., Fadeel, B., 2018. Cytokine profiling of primary human 0145-5. macrophages exposed to endotoxin-free graphene oxide: size-independent NLRP3 Rodrigues, A.F., Newman, L., Jasim, D., Mukherjee, S.P., Wang, J., Vacchi, I.A., Ménard- inflammasome activation. Adv. Healthc. Mater. 7 (4), 1700815 https://doi.org/ Moyon, C., Bianco, A., Fadeel, B., Kostarelos, K., Bussy, C., 2020. Size-dependent 10.1002/adhm.201700815. pulmonary impact of thin graphene oxide sheets in mice: toward safe-by-design. Mukherjee, S.P., Lazzaretto, B., Hultenby, K., Newman, L., Rodrigues, A.F., Lozano, N., Adv. Sci. 7 (12), 1903200 https://doi.org/10.1002/advs.201903200. Kostarelos, K., Malmberg, P., Fadeel, B., 2018. Graphene oxide elicits membrane Russier, J., Treossi, E., Scarsi, A., Perrozzi, F., Dumortier, H., Ottaviano, L., lipid changes and neutrophil extracellular trap formation. Chem 4 (2), 334–358. Meneghetti, M., Palermo, V., Bianco, A., 2013. Evidencing the mask effect of https://doi.org/10.1016/j.chempr.2017.12.017. graphene oxide: a comparative study on primary human and murine phagocytic Mukherjee, S.P., Lozano, N., Kucki, M., Del Rio-Castillo, A.E., Newman, L., Vázquez, E., cells. Nanoscale 5 (22), 11234–11247. https://doi.org/10.1039/c3nr03543c. Kostarelos, K., Wick, P., Fadeel, B., 2016. Detection of endotoxin contamination of Schinwald, A., Murphy, F.A., Jones, A., MacNee, W., Donaldson, K., 2012. Graphene- graphene based materials using the TNF-α expression test and guidelines for based nanoplatelets: a new risk to the respiratory system as a consequence of their endotoxin-free graphene Oxide production. PLoS One 11 (11), e0166816. https:// unusual aerodynamic properties. ACS Nano 6 (1), 736–746. https://doi.org/ doi.org/10.1371/journal.pone.0166816. 10.1021/nn204229f. Nakano-Narusawa, Y., Yokohira, M., Yamakawa, K., Saoo, K., Imaida, K., Matsuda, Y., Schlagenhauf, L., Buerki-Thurnherr, T., Kuo, Y.Y., Wichser, A., Nüesch, F., Wick, P., 2020. Single intratracheal quartz instillation induced chronic inflammation and Wang, J., 2015. Carbon nanotubes released from an epoxy-based nanocomposite: tumourigenesis in rat lungs. Sci. Rep. 10 (1), 1–10. https://doi.org/10.1038/s41598- quantification and particle toxicity. Environ. Sci. Technol. 49 (17), 10616–10623. 020-63667-4. https://doi.org/10.1021/acs.est.5b02750. Nel, A., Xia, T., Mädler, L., Li, N., 2006. Toxic potential of materials at the nanolevel. Shi, Y., Yu, B., Zheng, Y., Yang, J., Duan, Z., Hu, Y., 2018. Design of reduced graphene Science 311 (5761), 622–627. https://doi.org/10.1126/science.1114397. oxide decorated with DOPO-phosphanomidate for enhanced fire safety of epoxy Netkueakul, W., Fischer, B., Walder, C., Nüesch, F., Rees, M., Jovic, M., Gaan, S., resin. J. Colloid Interface Sci. 521, 160–171. https://doi.org/10.1016/j. Jacob, P., Wang, J., 2020. Effects of combining graphene nanoplatelet and jcis.2018.02.054. phosphorous flame retardant as additives on mechanical properties and flame Skehan, P., Storeng, R., Scudiero, D., Monks, A., Mcmahon, J., Vistica, D., Warren, J.T., retardancy of epoxy nanocomposite. Polym. (Basel) 12 (10), 2349. https://doi.org/ Bokesch, H., Kenney, S., Boyd, M.R., 1990. New colorimetric cytotoxicity assay for 10.3390/polym12102349. anticancer-drug screening. J. Natl. Cancer Inst. 82 (13), 1107–1112. https://doi.org/ Netkueakul, W., Korejwo, D., Hammer, T., Chortarea, S., Rupper, P., Braun, O., 10.1093/jnci/82.13.1107. Calame, M., Rothen-Rutishauser, B., Buerki-Thurnherr, T., Wick, P., Wang, J., 2020. Sohal, I.S., O’Fallon, K.S., Gaines, P., Demokritou, P., Bello, D., 2018. Ingested Release of graphene-related materials from epoxy-based composites: engineered nanomaterials: state of science in nanotoxicity testing and future characterization, quantification and hazard assessment: in Vitro. Nanoscale 12 (19), research needs. Part. Fibre Toxicol. 1–31. https://doi.org/10.1186/s12989-018- 10703–10722. https://doi.org/10.1039/c9nr10245k. 0265-1. Novoselov, K.S., Fal’Ko, V.I., Colombo, L., Gellert, P.R., Schwab, M.G., Kim, K., 2012. Song, P., Cao, Z., Cai, Y., Zhao, L., Fang, Z., Fu, S., 2011. Fabrication of exfoliated A roadmap for graphene. Nature 192–200. https://doi.org/10.1038/nature11458. graphene-based polypropylene nanocomposites with enhanced mechanical and Oberdörster, G., Oberdörster, E., Oberdörster, J., 2005. Nanotoxicology: an emerging thermal properties. Polym. (Guildf. ) 52 (18), 4001–4010. https://doi.org/10.1016/ discipline evolving from studies of ultrafine particles. Environ. Health Perspect. 113 j.polymer.2011.06.045. (7), 823–839. https://doi.org/10.1289/ehp.7339. Song, Z.M., Chen, N., Liu, J.H., Tang, H., Deng, X., Xi, W.S., Han, K., Cao, A., Liu, Y., Oberdörster, G., Stone, V., Donaldson, K., 2009. Toxicology of nanoparticles: a historical Wang, H., 2015. Biological effect of food additive titanium dioxide nanoparticles on perspective. Nanotoxicology. 24 S. Chortarea et al. Journal of Hazardous Materials 435 (2022) 129053 intestine: an in Vitro study. J. Appl. Toxicol. 35 (10), 1169–1178. https://doi.org/ Wohlleben, W., Meier, M.W., Vogel, S., Landsiedel, R., Cox, G., Hirth, S., Tomović, Ž., 10.1002/jat.3171. 2013. Elastic CNT-polyurethane nanocomposite: synthesis, performance and Song, N., Yang, J., Ding, P., Tang, S., Shi, L., 2015. Effect of polymer modifier chain assessment of fragments released during use. Nanoscale 5 (1), 369–380. https://doi. length on thermal conductive property of polyamide 6/graphene nanocomposites. org/10.1039/c2nr32711b. Compos. Part A Appl. Sci. Manuf. 73, 232–241. https://doi.org/10.1016/j. Wu, J., Wang, Y., Liu, G., Jia, Y., Yang, J., Shi, J., Dong, J., Wei, J., Liu, X., 2018. compositesa.2015.03.018. Characterization of air-liquid interface culture of A549 alveolar epithelial cells. Braz. Stankovich, S., Dikin, D.A., Dommett, G.H.B., Kohlhaas, K.M., Zimney, E.J., Stach, E.A., J. Med. Biol. Res. 51 (2) https://doi.org/10.1590/1414-431x20176950. Piner, R.D., Nguyen, S.B.T., Ruoff, R.S., 2006. Graphene-based composite materials. Wu, Y., Wang, F., Wang, S., Ma, J., Xu, M., Gao, M., Liu, R., Chen, W., Liu, S., 2018. Nature 442 (7100), 282–286. https://doi.org/10.1038/nature04969. Reduction of graphene oxide alters its cyto-compatibility towards primary and Sun, B., Wang, X., Ji, Z., Li, R., Xia, T., 2013. NLRP3 inflammasome activation induced immortalized macrophages. Nanoscale 10 (30), 14637–14650. https://doi.org/ by engineered nanomaterials. Small 1595–1607. https://doi.org/10.1002/ 10.1039/c8nr02798f. smll.201201962. Xu, Z., Gao, C., 2010. In Situ polymerization approach to graphene-reinforced nylon-6 Su, W.C., Ku, B.K., Kulkarni, P., Cheng, Y.S., 2016. Deposition of graphene nanomaterial composites. Macromolecules 43 (16), 6716–6723. https://doi.org/10.1021/ aerosols in human upper airways. J. Occup. Environ. Hyg. 13 (1), 48–59. https://doi. ma1009337. org/10.1080/15459624.2015.1076162. Zaman, I., Kuan, H.C., Dai, J., Kawashima, N., Michelmore, A., Sovi, A., Dong, S., Wang, X., Duch, M.C., Mansukhani, N., Ji, Z., Liao, Y.-P., Wang, M., Zhang, H., Sun, B., Luong, L., Ma, J., 2012. From carbon nanotubes and silicate layers to graphene Chang, C.H., Li, R., Lin, S., Meng, H., Xia, T., Hersam, M.C., Nel, A.E., 2015. Use of a platelets for polymer nanocomposites. Nanoscale 4 (15), 4578–4586. https://doi. pro-fibrogenic mechanism-based predictive toxicological approach for tiered testing org/10.1039/c2nr30837a. and decision analysis of carbonaceous nanomaterials. ACS Nano 9 (3), 3032–3043. Zeng, C., Lu, S., Xiao, X., Gao, J., Pan, L., He, Z., Yu, J., 2015. Enhanced thermal and https://doi.org/10.1021/nn507243w. mechanical properties of epoxy composites by mixing noncovalently functionalized Wang, Y., Wu, S., Zhao, X., Su, Z., Du, L., Sui, A., 2014. In vitro toxicity evaluation of graphene sheets. Polym. Bull. 72 (3), 453–472. https://doi.org/10.1007/s00289- graphene oxide on human RPMI 8226 Cells. In: Bio-Medical Materials and 014-1280-5. Engineering, Vol. 24. IOS Press,, pp. 2007–2013. https://doi.org/10.3233/BME- Zepp, R., Ruggiero, E., Acrey, B., Acrey, B., Davis, M.J.B., Davis, M.J.B., Han, C., Han, C., 141010. Han, C., Hsieh, H.S., Hsieh, H.S., Vilsmeier, K., Wohlleben, W., Sahle-Demessie, E., Wang, Y., Yu, J., Dai, W., Song, Y., Wang, D., Zeng, L., Jiang, N., 2015. Enhanced thermal 2020. Fragmentation of polymer nanocomposites: modulation by dry and wet and electrical properties of epoxy composites reinforced with graphene weathering, fractionation, and nanomaterial filler. Environ. Sci. Nano 7 (6), nanoplatelets. Polym. Compos. 36 (3), 556–565. https://doi.org/10.1002/pc.22972. 1742–1758. https://doi.org/10.1039/c9en01360a. Wan, Y.J., Gong, L.X., Tang, L.C., Wu, L., Bin, Jiang, J.X., 2014. Mechanical properties of Zhang, Y., Bai, Y., Jia, J., Gao, N., Li, Y., Zhang, R., Jiang, G., Yan, B., 2014. Perturbation epoxy composites filled with silane-functionalized graphene oxide. Compos. Part A of physiological systems by nanoparticles. Chem. Soc. Rev. 3762–3809. https://doi. Appl. Sci. Manuf. 64, 79–89. https://doi.org/10.1016/j.compositesa.2014.04.023. org/10.1039/c3cs60338e. Waquier, L., Myles, B.S., Henrard, L., Vautard, F., Pappas, C.M., Feneon, B., Delaitre, C., Zhao, X., Zhang, Q., Chen, D., Lu, P., 2010. Enhanced mechanical properties of graphene- Mehlem, J.J., Khripin, C.Y., 2020. Quantitative measurement of nanoparticle release based polyvinyl alcohol composites. Macromolecules 43 (5), 2357–2363. https:// from rubber composites during fabrication and testing. J. Nanopart. Res. 22 (9), doi.org/10.1021/ma902862u. 1–10. https://doi.org/10.1007/S11051-020-04977-6/FIGURES/5. Zheng, W., Chen, W.G., zhao, Q., Ren, S.X., Fu, Y.Q., 2019. Interfacial structures and Wick, P., Louw-Gaume, A.E., Kucki, M., Krug, H.F., Kostarelos, K., Fadeel, B., Dawson, K. mechanisms for strengthening and enhanced conductivity of graphene/epoxy A., Salvati, A., Vázquez, E., Ballerini, L., Tretiach, M., Benfenati, F., Flahaut, E., nanocomposites. Polymer 163, 171–177. https://doi.org/10.1016/j. Gauthier, L., Prato, M., Bianco, A., 2014. Classification framework for graphene- polymer.2018.12.055. based materials. Angew. Chem. Int. Ed. 53 (30), 7714–7718. https://doi.org/ Zhou, H., Zhao, K., Li, W., Yang, N., Liu, Y., Chen, C., Wei, T., 2012. The interactions 10.1002/anie.201403335. between pristine graphene and macrophages and the production of cytokines/ Wilbourn, J.D., McGregor, D.B., Partensky, C., Rice, J.M., 1997. IARC reevaluates silica chemokines via TLR- and NF-ΚB-related signaling pathways. Biomaterials 33 (29), and related substances. Environ. Health Perspect. 105 (7), 756–759. https://doi.org/ 6933–6942. https://doi.org/10.1016/j.biomaterials.2012.06.064. 10.1289/ehp.97105756. Zok, F.W., Miserez, A., 2007. Property maps for abrasion resistance of materials. Acta Williams, G.T., Williams, W.J., 1983. Granulomatous inflammation - a review. J. Clin. Mater. 55 (18), 6365–6371. https://doi.org/10.1016/j.actamat.2007.07.042. Pathol. 723–733. https://doi.org/10.1136/jcp.36.7.723. 25