Environmental Pollution 329 (2023) 121656 Contents lists available at ScienceDirect Environmental Pollution journal homepage: www.elsevier.com/locate/envpol The potential effects of in vitro digestion on the physicochemical and biological characteristics of polystyrene nanoplastics☆ Lourdes Vela a, b, Aliro Villacorta a, c, Tom Venus d, Irina Estrela-Lopis d, Susana Pastor a, Alba García-Rodriguez a, Laura Rubio a, e, Ricard Marcos a, *, Alba Hernández a a Group of Mutagenesis, Department of Genetics and Microbiology, Faculty of Biosciences, Universitat Autònoma de Barcelona, Cerdanyola Del Vallès, Spain b Facultad de Ciencias de La Salud, Eugenio Espejo. Universidad UTE, Avenida Occidental y Mariana de Jesús, Quito, Ecuador c Facultad de Recursos Naturales Renovables, Universidad Arturo Prat, Iquique, Chile d Institute of Medical Physics and Biophysics, University of Leipzig, 04107, Leipzig, Germany e Nanobiology Laboratory, Department of Natural and Exact Sciences, Pontificia Universidad Católica Madre y Maestra, PUCMM, Santiago de Los Caballeros, Dominican Republic A R T I C L E I N F O A B S T R A C T Keywords: The presence of plastic waste in our environment has continued growing and become an important environ­ Nanoplastics mental concern. Because of its degradation into micro- and nanoplastics (MNPLs), MNPLs are becoming envi­ Polystyrene ronmental pollutants of special environmental/health concern. Since ingestion is one of the main exposure routes Digestion to MNPLs, the potential effects of digestion on the physicochemical/biological characteristics of polystyrene Characterization Cell uptake nanoplastics (PSNPLs) were determined. The results indicated a high tendency of digested PSNPLs to agglom­ Hazard effects erate and a differential presence of proteins on their surface. Interestingly, digested PSNPLs showed greater cell uptake than undigested PSNPLs in all three tested cell lines (TK6, Raji-B, and THP-1). Despite these differences in cell uptake, no differences in toxicity were observed except for high and assumed unrealistic exposures. When oxidative stress and genotoxicity induction were determined, the low effects observed after exposure to undi­ gested PDNPLs were not observed in the digested ones. This indicated that the greater ability of digested PSNPLs to internalize was not accompanied by a greater hazard. This type of analysis should be performed with other MNPLs of varying sizes and chemical compositions. 1. Introduction inhalation and ingestion as the most frequent exposure routes. After exposure, MNPLs can translocate through physiological barriers, reach Due to the multiple advantages of plastic goods, including low cost, blood vessels, and internalize into different organs and tissues (Dome­ their use has experienced steady growth since the middle of the last nech et al., 2020; Fournier et al., 2020). century. Consequently, their presence in the environment, as waste, has When ingestion is the exposure route of MNPLs, they are exposed to not stopped growing and has become an important environmental the digestion process taking part in the digestive tract, which may concern (Lakshmi Kavya et al., 2020). Unfortunately, this plastic modify some of their physicochemical characteristics. The process starts contamination has led to the introduction of micro/nanoplastics in the mouth where saliva and amylase are present and follows in the (MNPLs), resulting from the degradation of larger plastic materials, stomach with the presence of strong acid gastric juices. Next, the gastric which are possibly more dangerous from a human health point of view fluid is neutralized by the alkaline intestinal fluid, containing a multi­ (Bouwmeester et al., 2015; EFSA, 2016). Although the presence of tude of enzymes required for food digestion. Accordingly, different in MNPLs was initially associated with the marine environment, today, it is vitro approaches have been proposed to mimic this process (Kämpfer clear that MNPLs can be found in any type of environment (Mitrano et al., 2020). Nevertheless, despite the usefulness of the in vitro digestion et al., 2021). Consequently, humans are inescapably exposed, with approaches, formal official validation of the static models is awaited. This paper has been recommended for acceptance by Michael Bank. ☆ * Corresponding author. Group of Mutagenesis, Department of Genetics and Microbiology, Faculty of Biosciences, Universitat Autònoma de Barcelona, Campus of Bellaterra, 08193, Cerdanyola del Vallès, Barcelona, Spain. E-mail address: ricard.marcos@uab.cat (R. Marcos). https://doi.org/10.1016/j.envpol.2023.121656 Received 13 December 2022; Received in revised form 12 April 2023; Accepted 15 April 2023 Available online 17 April 2023 0269-7491/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). L. Vela et al. Environmental Pollution 329 (2023) 121656 Thus, the consensus reached in the COST Action “Infogest” in 2014 has NaHCO3. After that, 0.50 mL of sterile pancreatin-bile solution (1.40 not yet generated validated guidelines (Colombo et al., 2021). Thus, the mg/mL pancreatin and 8 mg/mL bile) was added to the sample, and the formation of a corona on the MNPL surface, induced by the influence of pH was adjusted to 6.9–7 using sterile 0.10 M NaHCO3. The result of the in vitro digestive treatment, has been reported to lead to changes in size, indicated digestion process was weighed and adjusted to 15 g by using a zeta potential, and the ability to adsorb compounds (Liu et al., 2020). sterile (140 mM NaCl, 5 mM KCl, pH = 6.7) solution. All the simulated When tire degradation particles were exposed in vitro to simulated digestive juices were prepared under sterile conditions. The digested gastric/intestinal fluids, the solubilization of heavy metal contents was polystyrene nanoplastics are hereafter referred to as dPSNPLs. greatly enhanced. It should be noted that metals, basically zinc, are used One milliliter of the mix containing all the components used in the as catalysts during the vulcanization process of rubber, and conse­ digestion process (18750 μg/mL) was dispersed in 0.05% sterilized quently, they are components of the MNPLs resulting from tire degra­ bovine serum albumin (BSA) and 30 μL of ethanol (100%) to reach a dation (Masset et al., 2021). Changes in the surface of the PETMPL stock concentration of 3.125 μg/mL and then dispersed by ultra­ samples were observed after in vitro intestinal digestion, as well as the sonication in a cold-water bath. The digested fluorescent polystyrene presence of crystalline and organic deposits. These characteristics were nanoplastics (dfPSNPL), at a concentration of 725 μg/mL, were also considered relevant in the changes observed in their ability to attach to dispersed by ultrasonication. All these steps followed the dispersion different bacterial phyla (Tamargo et al., 2022). procedure generated in the EU project Nanogenotox (Nanogenotox, Once MNPLs cross protective barriers, they can reach the blood and 2011). The dfPSNPLs were used only for the internalization and uptake interact with its components, including cells. Interestingly, it has been assays, whereas dPSNPLs were used for the remaining assays. reported that MNPLs were able to differentially affect several white blood lineages in an ex vivo experiment using blood samples from 2.2. Characterization of digested/undigested PSNPLs different donors. While limited uptake was observed in lymphocytes, very high uptake was observed in monocytes and polymorphonuclear 2.2.1. Transmission electron microscopy (TEM) cells. In addition, significant increases in the levels of DNA damage were The different particle dispersions (PSNPLs, dPSNPLs, fPSNPLs, and observed in monocytes and polymorphonuclear cells but not in lym­ dfPSNPLs) were diluted in distilled water to obtain a 200 μg/mL con­ phocytes (Ballesteros et al., 2020). These results agree with the data centration, mounted on carbon-coated TEM grids, and visualized with a obtained in vitro using different hematopoietic cell lines, such as Raji-B JEOL JEM-1400 instrument (JEOL LTD, Tokyo, Japan). To determine (B-lymphocytes), TK6 (lymphoblasts), and THP-1 (monocytes), where their mean size and distribution, 100 randomly selected particles were exposure to nanopolystyrene produced differential effects on parameters measured by using ImageJ software. such as cytotoxicity, cellular uptake, reactive oxygen species produc­ tion, and genotoxicity (Rubio et al., 2020). In addition, it has been 2.2.2. Scanning electron microscopy (SEM) proven that after ingestion, in the digestive tract, MNPLs meet the in­ The different NPL dispersions were diluted in distilled water to reach testinal mucosa and the intestinal epithelium, which constitute a barrier a concentration of 200 μg/mL, and a drop of each dispersion was placed that helps to avoid the absorption of xenobiotics. Even considering this, over a silica grid and allowed to evaporate overnight. NPL characteris­ several studies have shown that MNPLs can cross this biological barrier, tics were visualized with an SEM Zeiss Merlin device (Zeiss, Oberko­ reaching the circulatory system as well as more distant tissues and or­ chen, Germany) coupled with an X-Max 20 mm EDS system (Oxford gans that include even the human placenta (Wright and Kelly, 2017; Instruments, Oxford, UK). Once the signals were recorded, they were Banerjee and Shelver, 2021; Ragusa et al., 2021; Kumar et al., 2022). analyzed by using INCA Energy software. In this context, our aim has been to identify if the application of in vitro digestion procedures can modulate some of the physicochemical 2.2.3. Dynamic light scattering (DLS) and zeta potential characteristics of nanopolystyrene, used as a model of MNPLs. To this The hydrodynamic size was analyzed using a Zetasizer® Ultra from end, we used TEM, SEM, NTA, and Zetasizer devices to determine the Malvern Panalytical. To proceed, PSNPLs, dPSNPLs, fPSNPLs, and size, morphology, hydrodynamic size, and surface charge. Furthermore, dfPSNPLs at a concentration of 100 μg/mL were prepared in Milli-Q confocal Raman microscopy was applied to visualize and characterize water, and the surface charge was determined by zeta potential proteins on the surface of MNPLs. Finally, changes in different biological values. The hydrodynamic size was determined by DLS. All the param­ characteristics, such as cell uptake, toxicity, intracellular ROS levels, eters for each sample were measured in triplicate. and genotoxicity, were also determined. 2.2.4. Nanotracking analysis (NTA) 2. Materials and methods A NanoSight LM10 (NanoSight, Amesbury, United Kingdom) device equipped with a 532-nm laser was used for NTA measurements. NTA 3.0 2.1. Nanopolystyrene selection, in vitro digestion, and dispersion software was used to capture and analyze the data. Samples were preparation injected in the sample chamber, and a digital CCD camera detected the light scattered off the particles with a 20× microscope objective. The Commercial polystyrene nanoplastics (PSNPLs, PP-008-10) with di­ NTA software allowed the detection of each individual diffusion track ameters of approximately 50–100 nm and their fluorescent yellow and the determination of the hydrodynamic diameter of the particles. counterpart (fPSNPLs, FP-00552-2) with diameters of approximately The particle concentration was set to 108–109 particles/mL by diluting 40–90 nm, obtained from Spherotech (Chicago, IL, USA), were used in the solution in Millipore water. Each sample was measured for 60 s at this study. five different positions for three replicates. All measurements were According to the aim of this study, both PSNPLs and fPSNPLs were performed at a temperature of 25 ◦ C. subjected to an in vitro digestion process following an adaptation of the method published by Mahler et al. (2009). This process mimics the 2.2.5. Confocal Raman microspectroscopy (CRM) particle/digestive fluid interactions that takes place in both the stomach CRM was performed using an Alpha300 R microscope (WITec GmbH, and intestine compartments. Briefly, 0.40 g of PSNPLs (or 0.01 g of Germany). PSNPLs, at a concentration of 312.5 μg/mL, were dried on fPSNPLs) was dispersed in 10 mL of a sterile (140 mM NaCl, 5 mM KCl, glass slides, and Raman spectra were recorded pixelwise in the lateral pH = 2) solution, and the pH was adjusted to 2 with sterile 0.10 M HCl. A plane using a laser power of ca 35 mW, an integration time of 0.15 s, and 0.50 mL sterile porcine pepsin solution (25 mg/mL diluted in 0.10 M a step size of 250 nm. The images and single spectra were analyzed using HCl) was added and incubated for 1 h at 37 ◦ C. Once the incubation was Project FIVE PLUS software (WITec GmbH, Ulm, Germany). A water­ over, the pH of the samples was adjusted to 5.5–6 with sterile 0.10 M shed algorithm was applied to identify single particles and to determine 2 L. Vela et al. Environmental Pollution 329 (2023) 121656 their amount of protein and coverage level. A total of 53 particles, such 2.8. Genotoxic and oxidative DNA damage induction as single particles and their agglomerates, were analyzed. In a previous study carried out by our group, the genotoxic and 2.3. Cells used and culture conditions oxidative DNA damage (ODD) induced by PSNPLs were determined (Rubio et al., 2020). The same methodology was used to evaluate the Three human cell types were selected for this study, namely, TK6 effects of dPSNPL exposure. The induced DNA damage was measured (lymphoblasts), THP-1 (monocytes), and Raji-B (B-lymphocytes). We according to the percentage of DNA in the comet tails obtained using selected lymphoblastic cells because once MNPLs cross the biological Komet 5.5 image analysis software (Kinetic Imaging Ltd., Liverpool, barriers, they reach the blood compartment, interacting with cells UK). Methyl methanesulfonate and potassium bromate were used as equivalent to those used in this study. This is the rationale for using the positive controls for general genotoxic damage and for ODD, selected cells, which were purchased from Sigma Aldrich (MO, USA). respectively. The three different cell lines were cultured at 37 ◦ C in a humidified at­ mosphere of 5% CO2 and 95% air and maintained in 25 cm2 culture 3. Results and discussion flasks with RPMI (Biowest, France) supplemented with 10% FBS, 1% glutamine (Biowest, France), and 2.5 μg/mL Plasmocin™ (InvivoGen, 3.1. Particle characterization CA, USA). The cell density range was kept from 5 × 105 cells/mL to 1 × 106 cells/mL (Rubio et al., 2020). 3.1.1. Morphology, dry state size distribution, and shape PSNPLs, dPSNPLs, fPSNPLs, and dfPSNPLs were determined by TEM 2.4. Cell internalization by flow cytometry and SEM to visualize their size and morphology. The Martin diameter was measured by using TEM, and the size distribution was determined Flow cytometry (FC) was used to measure cell internalization. To by measuring three times 100 images of each PSNPL type. The relative proceed, cell suspensions seeded in 96 U-type well plates were treated percentages of the different particle sizes are represented in Supple­ for 24 h with different concentrations of fPSNPLs and dFPSNPLs. The mentary Fig. 1. Both TEM/SEM images show that all particles were exposed cells were analyzed using a Beckman Coulter CytoFlex S spherically shaped, and the average sizes of the nondigested particles (Indianapolis, USA) device, and the obtained data were analyzed using were approximately 66.07 nm for PSNPLs and 51.71 nm for fPSNPLs. CytExpert software. Three independent experiments using duplicates for These sizes agreed with the manufacturer’s specifications. Once the each concentration were carried out. particles were submitted to the in vitro digestion process, the obtained average TEM sizes were slightly modified to approximately 60.68 nm for dPSNPLs and 54.81 nm for dfPSNPLs. Although TEM and SEM images 2.5. fPSNPL and dfPSNPL intracellular localization by confocal showed that the nondigested and the in vitro digested particles were microscopy similar in appearance and sizes, the images of the digested in vitro par­ ticles showed a relevant tendency to agglomerate, higher than in the Laser confocal microscopy was used to confirm the internalization of nondigested particles, forming large groups of particles. This high the fPSNPLs and dfPSNPLs by the three cell lines, as previously reported agglomeration status was very evident from the SEM figures corre­ (Cortés et al., 2022). Following the published protocol, fPSNPs and sponding to the dfPSNPL condition. TEM/SEM analyses are not usually dfPSNPs were visualized inside cells. The obtained images were pro­ carried out in studies evaluating the effects of in vitro digestion (Liu cessed using ImageJ software. In addition, the images were also et al., 2020); nevertheless, in a very recent study (Chen et al., 2022), no analyzed by using IMARIS 9.5 software for 3D/4D image analysis to changes in size or shape were reported, confirming our findings as well better place PSNPLs inside the cell structure. Unexposed cells were used as those obtained with different plastic materials such as polyethylene, as a negative control for both experimental approaches. polypropylene, polyvinyl chloride, polyethylene terephthalate, and polystyrene (Stock et al., 2020). These in vitro digestion effects would 2.6. Cell viability differ from those observed in vivo in the intestine of Drosophila larvae. When Drosophila larvae were exposed to three sizes of polystyrene Viability assays were performed as previously reported (Rubio et al., MNPLs (50, 200, and 500 nm), differences were observed by TEM be­ 2020). Samples from the three selected cell lines were exposed to tween sizes in the initial dispersion and into the intestine lumen. How­ PSNPLs/dPSNPLs at concentrations up to 200 μg/mL over 24 and 48 h. ever, these differences were observed only for the highest sizes (200 and The treated cells were diluted 1:100 in ISOTON® and counted in a 500 nm) and not for 50 nm, which was the size used in our study Beckman Coulter Counter Z2 (Indianapolis, USA). Three different ex­ (Alaraby et al., 2022). periments were performed, and in each experiment, three replicates were carried out. The average number of cells counted in each treatment 3.1.2. Aqueous solution size distribution and zeta potential was compared with the average number of untreated control cells. The The hydrodynamic size was determined by DLS for the four condi­ results from the viability assay are useful to determine the range of tions. As observed in Supplementary Fig. 2, in general, the digested noncytotoxic concentrations to be used in further experiments. particles had a larger hydrodynamic size than the nondigested particles. In addition, the nondigested particles had a greater number of particles 2.7. Reactive oxygen species (ROS) quantification below 100 nm compared with the digested particles that showed more particles over 100 nm in size. This would agree with a higher ability of The production of intracellular reactive oxygen species (ROS) was digested particles to agglomerate. This was also in accordance with the determined using the dihydroethidium (DHE, Calbiochem, USA) assay polydispersity index (PdI) values, where the nondigested particles were and measured by FC. The three cell lines were exposed for 3 and 24 h to more dispersed than the digested ones. It should be remembered that the selected concentrations of dPSNPLs. Cigarette smoke condensate values closer to 0 indicate that the sample is monodisperse. Changes in (CSC, 100 μg/mL, Murty Pharmaceuticals, USA) was used as a positive the hydrodynamic size after in vitro digestion have also been reported as control. Following a previously published protocol (Rubio et al., 2020), a consequence of the presence of an induced corona on the surface of cells were analyzed by using a FACSCanto Flow Cytometer (BD Biosci­ 100 nm polystyrene NPLs (Liu et al., 2020). The results obtained from ence, USA) to evaluate a total of 10,000 events from living cells. For each the Z-potential measurements also agreed with those reported above, condition, three independent experiments (with duplicate samples) showing that the digested particles had a lower Z-potential than the were run. nondigested particles. All these results indicate that the in vitro digested 3 L. Vela et al. Environmental Pollution 329 (2023) 121656 particles presented a higher tendency to agglomerate than the non­ alpha helices, and beta sheets at 1640 cm− 1, 1660 cm− 1, and 1680 cm− 1, digested particles. respectively (Maiti et al., 2004) (Fig. 2c). The large amount of beta sheets might have originated from pepsin, which has 44% beta sheets 3.1.3. Nanotracking analysis (NTA) and only 14% alpha helices (Sielecki et al., 1990; Berman et al., 2000). Additionally, the number size distribution was determined by means Amylase can also contribute to beta sheet formation with 23% of the of NTA (Fig. 1). NTA is a more reliable tool for the evaluation of poly­ beta fraction (Ragunath et al., 2008). In contrast, BSA has a mainly alpha disperse samples, as it is based on single particle tracking. Nanotracking helix structure (Lin and Koenig, 1976). The fitting showed an additional analysis revealed similar mode values of 75 nm for digested and undi­ band at 1629 cm− 1 that was associated with monomeric PS (Sears et al., gested PS particles. This value was higher than the diameter obtained by 1981; Noda and Sala, 2000). It might be assumed that a small number of TEM, as NTA measures the hydrodynamic radius, which represents not monomers were generated due to the degradation of the polymer caused only the hard core of PS particles but also solvent molecules and ions by digestion. that surround them. On the other hand, DLS measurements showed a Furthermore, CRM images of digested PS particles were analyzed at higher particle diameter. DLS delivers an intensity-weighted distribu­ the single particle level by using a watershed algorithm. The degree of tion of particles that is highly sensitive to large particles. coverage with protein and the semiquantification of the protein amount Similar to the TEM and DLS results, NTA showed a higher poly­ on the particle surface across a particle population are summarized in dispersity of digested PS particles with the presence of various Fig. 3. agglomerated populations at 140 nm and approximately 200 nm, which As observed, most of the particles showed only a small degree of were roughly double and triple the size of the initial particle size. coverage with proteins on their surface. On average, 7.7% of the surface PSNPLs showed only a small additional accumulation at 140 nm (Fig. 1). area of particles was covered by proteins. However, some particles showed coverage of 20%–45%, which indicated strong differences be­ 3.1.4. Visualization and characterization of proteins on the surface of PS tween the particles (Fig. 3a). Approximately 30% of the particles particles showed no protein association. Fig. 3b shows a histogram of semi­ The Raman spectrum of PS particles has characteristic modes, such as quantification of protein on digested particles across the particle pop­ ring breathing at 1001 cm− 1, CH in-plane deformation at 1031 cm− 1, ulation. The semiquantification was based on the calculated ratio of the CH2 deformation at 1450 cm− 1, ring skeletal stretching at 1602 cm− 1, integral intensity of the amide I band to that of the PS mode at 1001 symmetric stretching at 2852 cm− 1 and asymmetric CH2 aliphatic cm− 1 for each particle. Most of the particles showed only small amounts stretching modes at 2904 cm− 1, as well as CH ring stretching at 3054 of associated protein. Similar to the particle coverage distributions, cm− 1 (Hong et al., 1991; Mazilu et al., 2010; Nava et al., 2021). The ring there were large differences between the particles, which indicated a breathing mode of PS has been used for visualization of dPSNPLs heterogeneous distribution of protein amount on the particle surface (Fig. 2a). Fig. 2b shows the average spectra of PSNPLs and dPSNPLs that (Fig. 3). The presence of proteins on the surface of digested particles has were extracted from regions identified as particles in CRM images. This also been found in other studies, although the content was modulated by ensured that only proteins located on the particle surface would be size, being higher in 100 nm PS MNPLs than in 5 μm PS MNPLs (Liu analyzed. Compared to undigested particles, the spectrum of dPSNPLs et al., 2020). revealed additional bands related to proteins. The intensity of bands related to CH2 deformation and stretching region of proteins at 1450 3.2. Cell internalization cm− 1, 2852 cm− 1, and 2904 cm− 1 were increased. Furthermore, addi­ tional bands corresponding to the symmetrical CH3 mode of proteins at The cellular uptake of both fPSNPLs and dfPSNPLs was determined 2936 cm− 1 and to the amide I region of proteins at approximately 1660 by flow cytometry. The study involved three cell lines, an exposure cm− 1 were recorded (Carey, 2006; Rygula et al., 2013). The overlapping period of 24 h, and five concentrations (1, 5, 10, 25, and 50 μg/mL). map image of the amide I distribution with that of digested PS particles Fig. 4 shows that for the three cell lines, the observed fluorescence shows an association of proteins on the particle surface (Fig. 2a). values were higher after dfPSNPL exposure than with the undigested The region of the amide I band was thoroughly analyzed to obtain counterpart. Such an effect was observed at all the tested concentrations information about the secondary protein structure on the particle sur­ and in the three cell lines. Interestingly, at the highest concentration (50 face, as well as PS polymer degradation due to the digestion process. To μg/mL) of dfPSNPLs, almost 100% of the three cell lines showed fluo­ this aim, the spectrum of undigested PS was subtracted from that of rescent signals. Regarding the uptake of fPSNPLs, marked differences dPSNPLs, leaving only the protein signal associated with the particle. were observed between cell lines, with Raji-B cells showing higher up­ Fitting of the amide I band revealed the occurrence of random coils, take, followed by THP1 cells, with TK-6 showing lower uptake. When Fig. 1. Nanotracking analysis of undigested (a) and digested (b) PS particles. 4 L. Vela et al. Environmental Pollution 329 (2023) 121656 Fig. 2. CRM analysis of PSNPLs and dPSNPLs: (a) the color-coded images of dPSNPLs (left), proteins (middle) and their overlapping (right); (b) Raman spectra of digested and undigested particles extracted from region of interest; (c) fitting analysis of amid I band represented a secondary structure of proteins on the surface of digested particles. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) Fig. 3. Histograms of degree of protein coverage (a) and protein amount (b) on the surface of single dPSNPLs across a particle population. the uptake reached after exposure to the highest concentration (50 μg/ white peripheral blood cells (namely, lymphocytes, monocytes, and mL) was compared, in Raji-B cells, almost 80% of the cells showed polymorphonuclear cells) were sorted from whole blood samples of fluorescence, whereas the observed values for THP1 and TK6 cells were different donors, sharp differences in cell uptake between cell types were 35% and 20%, respectively. These results agreed with previously re­ observed. Thus, very limited uptake was observed in lymphocytes, very ported data for the same three cell lines used (Rubio et al., 2020). high uptake in monocytes, and intermediate uptake values for poly­ Interestingly, this different uptake between human hematopoietic cell morphonuclear cells (Ballesteros et al., 2020). This indicates that uptake lines was also detected in vivo studies. Thus, when different subsets of is modulated not only by the physicochemical characteristics of the 5 L. Vela et al. Environmental Pollution 329 (2023) 121656 Fig. 4. Cellular internalization of fPSNPLs and dfPSNPLs by THP-1 cells (a), TK6 cells (b), and Raji-B cells (c), after exposures lasting for 24 h. The graph represents the percentage of fluorescent-positive cells over the total cell population. Data are analyzed by the one-way ANOVA test with a Dunnett’s post-test. Statistical significance indicated in the graph correspond to *P < 0.05, **P < 0.01, ***P < 0.001. MNPLs but also by the cell type used. with the uptake indicated in the previous section. PSNPLs were not Unfortunately, no studies have evaluated differences in cell uptake localized in specific regions, and no relevant differences were observed between digested and undigested MNPLs. The study evaluating cell between digested and nondigested particles. Thus, particles were uptake and translocation of digested MNPLs in an in vitro coculture observed attached to the membrane of the cells, spread into the cyto­ model of the intestinal barrier did not compare digested vs. undigested plasm, or accumulated in the perinuclear areas. Interestingly, particles MNPLs (DeLoid et al., 2021). were also internalized into the nucleus. These results agreed with those previously reported in the same three hematopoietic cell lines (Rubio et al., 2020). The localization of PSNPLs into cells has also been reported 3.3. Intracellular presence of fPSNPLs and dfPSNPLs in undifferentiated human intestinal Caco-2 cells (Cortés et al., 2020), as well as in cocultures of differentiated Caco-2 cells forming an in vitro To determine the cellular location of both fPSNPLs and dfPSNPLs in intestinal barrier (Domenech et al., 2020). the three cell types, confocal microscopy was used at a concentration of 100 μg/mL and exposures lasting for 24 h. As observed in Fig. 5, both types of particles were found inside the three cell lines, in agreement Fig. 5. Selection of representative three-dimensional images of THP1, TK6, and Raji-B cells exposed to 100 μg/mL of dfPSNPLs (left) and fPSNPLs (right) for 24 h. (a) and (b) represent THP1 cells; (c) and (d) represent TK6 cells; and (e) and (f) represent Raji-B cells. dfPSNPL and fPSNPL particles are shown in green, nuclei in blue, and cell membranes are red. ImageJ was used to process the figures, and the Imaris 9.5 software was use for obtaining 3D/4D images. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.) 6 L. Vela et al. Environmental Pollution 329 (2023) 121656 3.4. Cytotoxicity assessment lines were exposed to different concentrations from 0 to 50 μg/mL with exposures lasting for 3 and 24 h. The figure shows an increased pro­ To determine the potential cytotoxic concentrations of PSNPLs and duction of ROS after exposure lasting for 3 h, but only at the highest dPSNPLs, a viability assay was carried out on TK6, THP1, and Raji-B concentration (50 μg/mL) and in TK6 and Raji-B cells. These effects were cells. The range of tested concentrations was from 0 μg/mL (negative increased in exposures lasting for 24 h, but only in the TK6 cell line, control) to 200 μg/mL, and the exposures lasted for 24 and 48 h. The which seemed to be especially sensitive to PSNPL effects. These effects obtained results are indicated in Supplementary Fig. 3. The significant were not observed when cells were exposed to the digested PSNPLs. The effects, when they existed, were always observed at the highest con­ induction of oxidative stress by nanoparticle exposure has been accepted centrations tested (150/200 μg/mL), which were very high concentra­ as a regular mechanism of action (Li et al., 2022). This assumption has tions and, consequently, unrealistic exposures. Consequently, PSNPLs been proposed in the case of MNPLs, but the obtained results are usually could be considered nontoxic compounds, independent of whether they weak and contradictory (Yong et al., 2020). Positive induction, ac­ are digested. Furthermore, the exposure time modulated the effects, cording to the cell type, was demonstrated previously (Rubio et al., which were always higher after 48 h of exposure. Finally, the effects of 2020), and the ability to induce oxidative stress has also been associated digestion were only remarkable on the TK6 cells and at the longer with surface functionalization (He et al., 2020). Unfortunately, no exposure time (48 h). According to the toxicity data and aiming to work studies have compared the ability of in vitro digested MNPLs to induce with noncytotoxic doses, concentrations ≤50 μg/mL were chosen for the oxidative stress, which highlights the interest in the results reported in rest of the assays. The potential toxicity of MNPLs has been recently this study. reviewed (Schröter and Ventura, 2022), and the lack of important cytotoxicity is usually reported, as in human lung and intestinal cells 3.6. Genotoxic and oxidative DNA damage (Zhang et al., 2022), where three different polystyrene nanoplastic sizes were evaluated. The comet assay (with and without FPG) was carried out to measure DNA damage induced by the different exposures of PSNPLs. The three 3.5. Production of intracellular ROS cell lines were exposed to different concentrations of PSNPLs and dPSNPLs at two different time points, 3 and 24 h. The obtained results To determine the intracellular levels of ROS induced by dPSNPL are indicated in Figs. 7 and 8 for exposures lasting for 3 and 24 h, exposure, a DHE assay was used. DHE (dihydroethidium) easily crosses respectively. Fig. 7 shows the results at exposures lasting for 3 h, and as cell membranes reacting with intracellular superoxide anions. This re­ observed, no genotoxic or oxidative damage was observed in any of the action produces a red fluorescent compound (2-hydroxyethidium), three cell lines. This means that under these exposure conditions, no which can intercalate into DNA. Fig. 6 represents the intracellular ROS differences were observed between PSNPLs and dPSNPLs. production measured with the DHE assay by using FC. The three cell The results obtained in exposures lasting for 24 h are presented in Fig. 6. Intracellular ROS production analysis in THP1, TK6 and Raji-B cells exposed to both PSNPLs and dPSNPLs. Exposures lasted for 3 h (a and c) and 24 h (b and d). Exposure to 100 μg/mL CSC was used as positive control (CRT). ROS production was extrapolated from the average of the mean fluorescence intensity from three independent experiments with duplicated samples (10,000 events taken from the living cell population for each sample). Data are represented as the percentage of fluorescence intensity compared to the negative control. One-way ANOVA with Dunnett’s post-test was used for the statistical analysis. Statistical significance was indicated in the graph as *P < 0.05, **P < 0.01, and ***P < 0.001. 7 L. Vela et al. Environmental Pollution 329 (2023) 121656 Fig. 7. Genotoxic and oxidative damage detected by the comet assay after 3 h of exposure to PSNPLs (a, b) and to dPSNPLs (c, d). Oxidative DNA damage (ODD) was detected using the FPG enzyme. The concentration of 200 μM of methyl methanesulfonate (MMS) was used as a positive control (PC) for genotoxic damage induction while 5 mM of potassium bromate (KBrO3) was used as positive control for ODD. Experiments were performed with duplicates and 100 randomly selected cells were analyzed by the Komet 5.5 software. Data are represented as mean ± SEM and analyzed by the one-way ANOVA test with a Dunnett post-test. Statistical significance was indicated in the graph as follow, *P < 0.05, **P < 0.01, ***P < 0.001. Fig. 8. When the genotoxic damage was analyzed (a and c), slight effects 4. Conclusions were observed in both exposures, PSNPLs and dPSNPLs, although with differences between cell lines. Thus, PSNPL exposure induced effects One of the main exposure routes for MNPLs is ingestion. Their long only in Raji-B cells at the highest concentrations (25 and 50 μg/mL). On journey through the different compartments of the intestinal tract can the other hand, dPSNPLs induced statistically significant effects only in modify some of their properties, such as size and surface, before THP1 cells, although a tendency to show high DNA damage was also reaching the intestinal barrier. These changes would be a consequence observed in Raji-B cells. of the accompanying digestion processes. Consequently, modifications When the ability to induce oxidative damage in the DNA bases was in their uptake ability and their distribution through the entire body can determined, differences between exposures were observed. In that case, occur. We explored whether the in vitro digestion of polystyrene nano­ the induced effects observed, mainly in the TK6 cells after exposure to plastics (PSNPLs) of approximately 50 nm can modify some of their PSNPLs, were not observed when the effects of dPSNPLs were evaluated. properties. First, independent of the method used (TEM, DLS, or NTA), Interestingly, these results matched well with the results obtained in the no relevant differences in size were found between undigested and experiments evaluating the induction of intracellular ROS levels. Glob­ digested PDNPLs. Nevertheless, a higher tendency of digested PSNPLs to ally, some genotoxic effects are observed to be associated with exposure, agglomerate was observed, and CRM indicated the differential presence although such effects can be considered mild if they truly exist. Thus, of proteins on the surface of digested particles. Interestingly, great dif­ when the observed effects were low and without a concentration- ferences in cell uptake were observed between nondigested and digested dependent relationship, simple stochastic effects could be responsible cells, with the latter showing greater uptake in the three tested cell lines. for the observed “significant” effects. Despite these differences in cell uptake, no significant differences in The link between ROS induction and genotoxicity is well known toxicity were observed except for high and assumed unrealistic expo­ (Quezada-Maldonado et al., 2021), which supports our findings. sures. It is worth noting the different uptake capabilities observed be­ Nevertheless, the potential genotoxicity of MNPLs is a topic that has tween cell lines. Consequently, in addition to the physicochemical generated conflicting consensus, since contradictory results have been characteristics (including digested vs. undigested PSNPLs), the cell type reported (Xu et al., 2022). Different reasons can be associated with the used was a very relevant factor modulating cell uptake. When oxidative contradictory genotoxicity data. First, under the MNPL umbrella, there stress and genotoxicity induction were determined, the low effects is great variability in size that can affect the genotoxicity outcome. observed after exposure to undigested PDNPLs were not observed in the Alternatively, the cell line used is another potential source of variability digested ones. This was also observed when the different cell types were since MNPL uptake is cell dependent. Finally, there is the genotoxic considered. This would indicate that the greater ability of digested potential (if it exists) of MNPLs, since it is weak, and a large variability is PSNPLs to internalize was not accompanied by a greater hazard if not expected based only on a probabilistic point of view. the opposite. 8 L. Vela et al. Environmental Pollution 329 (2023) 121656 Fig. 8. Genotoxic and oxidative DNA damage detected by the comet assay after 24 h of PSNPLs (a, b) and dPSNPLs (c, d) exposures. The oxidative DNA damage (ODD) was detected using the FPG enzyme. The concentration of 200 μM of methyl methanesulfonate (MMS) was used as a positive control (PC) for genotoxic damage induction while 5 mM of potassium bromate (KBrO3) was used as positive control for ODD. Experiments were performed with duplicates and 100 randomly selected cells were analyzed by the Komet 5.5 software. Data are represented as mean ± SEM and analyzed by the one-way ANOVA test with a Dunnett post-test. Statistical significance was indicated in the graph as follow, *P < 0.05, **P < 0.01 ***P < 0.001. Author contribution statement This project has received funding from the European Union’s Hori­ zon 2020 Research and Innovation Programme under Grant Agreement RM and AH planned the experiments. LV, AV, SP, AGR, and LR No. 965196. This work was also partially supported by the Spanish carried out the experimental part. LV, and AV analyzed the data, carried Ministry of Science and Innovation [PID 2020-116789, RB-C43] and by out the statistical analysis, and prepared tables/figures. TV and IEL were the Generalitat de Catalunya (2021-SGR-00731). involved in the part containing NTA and CRM information. LV, RM, and AH wrote the final manuscript. Appendix A. Supplementary data Declaration of competing interest Supplementary data to this article can be found online at https://doi. org/10.1016/j.envpol.2023.121656. The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence References the work reported in this paper. Alaraby, M., Abass, D., Domenech, J., Hernández, A., Marcos, R., 2022. Hazard assessment of ingested polystyrene nanoplastics in Drosophila larvae. Environ. Sci. Data availability Nano 9, 1845–1857. https://doi.org/10.1039/d1en01199e. Ballesteros, S., Domenech, J., Barguilla, I., Cortés, C., Marcos, R., Hernández, A., 2020. Genotoxic and immunomodulatory effects in human white blood cells after ex vivo Data will be made available on request. exposure to polystyrene nanoplastics. Environ. Sci.: Nano 7, 3431–3446. https://doi. org/10.1039/d0en00748j. Acknowledgments Banerjee, A., Shelver, W.L., 2021. Micro- and nanoplastic induced cellular toxicity in mammals: a review. Sci. Total Environ. 755 (Pt 2), 142518 https://doi.org/10.1016/ j.scitotenv.2020.142518. L. Vela was supported by PhD fellowships from the Fundación Car­ Berman, H.M., Westbrook, J., Feng, Z., Gilliland, G., Bhat, T.N., Weissig, H., olina. A. Villacorta was supported by PhD fellowships from the National Shindyalov, I.N., Bourne, P.E., 2000. The protein data bank. Nucleic Acids Res. 28, 235–242. https://doi.org/10.1093/nar/28.1.235. Agency for Research and Development (ANID), CONICYT PFCHA/ Bouwmeester, H., Hollman, P.C., Peters, R.J., 2015. Potential health impact of DOCTORADO BECAS CHILE/2020–72210237. A. García-Rodríguez environmentally released micro- and nanoplastics in the human food production received funding from the postdoctoral fellowships program Beatriu de chain: experiences from Nanotoxicology. Environ. Sci. Technol. 49 (15), 8932–8947. https://doi.org/10.1021/acs.est.5b01090. Pinós, funded by the Secretary of Universities and Research [Govern­ Carey, P.R., 2006. Raman crystallography and other biochemical applications of Raman ment of Catalonia] and by the Horizon 2020program of research and microscopy. Annu. Rev. Phys. Chem. 57, 527–554. https://doi.org/10.1146/ innovation of the European Union under the Marie Sktodowsdka-Curie annurev.physchem.57.032905.104521. grant agreement No 801370. L. Rubio was supported by a contract Chen, H., Chen, H., Nan, S., Liu, H., Chen, L., Yu, L., 2022. Investigation of microplastics in digestion system: effect on surface microstructures and probiotics. Bull. Environ. Juan de la Cierva [IJC 2020–2686I/AEI/10.13039/501100011033]. A. Contam. Toxicol. https://doi.org/10.1007/s00128-022-03571-x. Online ahead of Hernández was granted an ICREA ACADEMIA award. print. 9 L. Vela et al. Environmental Pollution 329 (2023) 121656 Colombo, R., Ferron, L., Frosi, Papetti A., 2021. Advances in static in vitro digestion Mitrano, D.M., Wick, P., Nowack, B., 2011. Placing nanoplastics in the context of global models after the COST action Infogest consensus protocol. Food Funct. 12 (17), plastic pollution, 2021 Nat. Nanotechnol. 16 (5), 491–500. https://doi.org/10.1038/ 7619–7636. https://doi.org/10.1039/d1fo01089a. s41565-021-00888-2. Nanogenotox. http://www.nanogenotox.eu/files/PDF/Deliv Cortés, C., Domenech, J., Salazar, M., Pastor, S., Marcos, R., Hernández, A., 2020. erables/nanogenotox%20. Nanoplastics as potential environmental health factors. Effects of polystyrene Nanogenotox. http://www.nanogenotox.eu/files/PDF/Deliverables/nanogenotox%20 nanoparticles on the human intestinal epithelial Caco-2 cells. Environ. Sci.: Nano 7, deliverable%203_wp4_%20dispersion%20protocol.pdf. 272–285. https://doi.org/10.1039/c9en00523d. Nava, V., Frezzotti, M.L., Leoni, B., 2021. Raman spectroscopy for the analysis of DeLoid, G.M., Cao, X., Bitounis, D., Singh, D., Llopis, P.M., Buckley, B., Demokritou, P., microplastics in aquatic systems. Appl. Spectrosc. 75, 1341–1357. https://doi.org/ 2021. Toxicity, uptake, and nuclear translocation of ingested micro-nanoplastics in 10.1364/AS.75.001341. an in vitro model of the small intestinal epithelium. Food Chem. Toxicol. 158, Noda, L.K., Sala, O., 2000. A resonance Raman investigation on the interaction of styrene 112609 https://doi.org/10.1016/j.fct.2021.112609. and 4-methyl styrene oligomers on sulphated titanium oxide. Spectrochim. Acta A. Domenech, J., Hernández, A., Rubio, L., Marcos, R., Cortés, C., 2020. Interactions of Mol. Biomol. Spectrosc. 56A, 145–155. https://doi.org/10.1016/S1386-1425(99) polystyrene with in vitro models of the human intestinal barrier. Arch. Toxicol. 94 00128-6. (9), 2997–3012. https://doi.org/10.1007/s00204-020-02805-3. Quezada-Maldonado, E.M., Sánchez-Pérez, Y., Yi, Chirino, García-Cuellar, C.M., 2021. EFSA, 2016. Presence of microplastics and nanoplastics in food, with particular focus on Airborne particulate matter induces oxidative damage, DNA adduct formation and seafood. EFSA J. 14 (6), 1–30. https://doi.org/10.2903/j.efsa.2016.4501. alterations in DNA repair pathways. Environ. Pollut. 287, 117313 https://doi.org/ Fournier, S.B., D’Errico, J.N., Adler, D.S., Kollontzi, S., Goedken, M.J., Fabris, L., 10.1016/j.envpol.2021.117313. Yurkow, E.J., Stapleton, P.A., 2020. Nanopolystyrene translocation and fetal Ragunath, C., Manuel, S.G., Venkataraman, V., Sait, H.B., Kasinathan, C., deposition after acute lung exposure during late-stage pregnancy. Part. Fibre Ramasubbu, N., 2008. Probing the role of aromatic residues at the secondary Toxicol. 17 (1), 55. https://doi.org/10.1186/s12989-020-00385-9. saccharide-binding sites of human salivary alpha-amylase in substrate hydrolysis and He, Y., Li, J., Chen, J., Miao, X., Li, G., He, Q., Xu, H., Li, H., Wei, Y., 2020. Cytotoxic bacterial binding. J. Mol. Biol. 384, 1232–1248. https://doi.org/10.1016/j. effects of polystyrene nanoplastics with different surface functionalization on human jmb.2008.09.089. HepG2 cells. Sci. Total Environ. 723, 138180 https://doi.org/10.1016/j. Ragusa, A., Svelato, A., Santacroce, C., Catalano, P., Notarstefano, V., Carnevali, O., scitotenv.2020.138180. Papa, F., Ciro, M., Rongioletti, A., Baiocco, F., Draghi, S., Amore, E.D., Rinaldo, D., Hong, P.P., Boerio, F.J., Clarson, S.J., Smith, S.D., 1991. An investigation of the Matta, M., Giorgini, E., 2021. Plasticenta : first evidence of microplastics in human interdiffusion of polystyrene and deuterated polystyrene using surface-enhanced placenta. Environ. Int. 146, 106274 https://doi.org/10.1016/j.envint.2020.106274. Raman-Scattering. Macromolecules 24, 4770–4776. https://doi.org/10.1021/ Rygula, A., Majzner, K., Marzec, K.M., Kaczor, A., Pilarczyk, M., Baranska, M., 2013. ma00017a007. Raman spectroscopy of proteins: a review. J. Raman Spectrosc. 44, 1061–1076. Kämpfer, A.A.M., Busch, M., Schins, R.P.F., 2020. Advanced in vitro testing strategies and https://doi.org/10.1002/jrs.4335. models of the intestine for nanosafety research. Chem. Res. Toxicol. 33 (5), Rubio, L., Barguilla, I., Domenech, J., Marcos, R., Hernández, A., 2020. Biological effects, 1163–1178. https://doi.org/10.1021/acs.chemrestox.0c00079. including oxidative stress and genotoxic damage, of polystyrene nanoparticles in Kumar, R., Manna, C., Padha, S., Verma, A., Sharma, P., Dhar, A., Ghosh, A., different human hematopoietic cell lines. J. Hazard Mater. 398, 122900 https://doi. Bhattacharya, P., 2022. Micro(nano)plastics pollution and human health: how org/10.1016/j.jhazmat.2020.122900. plastics can induce carcinogenesis to humans? Chemosphere 298, 134267. https:// Schröter, L., Ventura, N., 2022. Nanoplastic toxicity: insights and challenges from doi.org/10.1016/j.chemosphere.2022.134267. experimental model systems. Small 18 (31), e2201680. https://doi.org/10.1002/ Lakshmi Kavya, A.N.V., Sundarrajan, S., Ramakrishna, S., 2020. Identification and smll.202201680. characterization of micro-plastics in the marine environment: a mini review. Mar. Sears, W.M., Hunt, J.L., Stevens, J.R., 1981. Raman scattering from polymerizing Pollut. Bull. 160, 111704 https://doi.org/10.1016/j.marpolbul.2020.111704. styrene. 1 Vibrational-mode analysis. J. Chem. Phys. 75, 1589–1598. https://doi. Li, B., Zhang, T., Tang, M., 2022. Toxicity mechanism of nanomaterials: focus on org/10.1063/1.442262. endoplasmic reticulum stress. Sci. Total Environ. 834, 155417 https://doi.org/ Sielecki, A.R., Fedorov, A.A., Boodhoo, A., Andreeva, N.S., James, M.N.G., 1990. 10.1016/j.scitotenv.2022.155417. Molecular and crystal-structures of monoclinic porcine pepsin refined at 1.8-a Lin, V.J., Koenig, J.L., 1976. Raman studies of bovine serum albumin. Biopolymers 15, resolution. J. Mol. Biol. 214, 143–170. https://doi.org/10.1016/0022-2836(90) 203–218. https://doi.org/10.1002/bip.1976.360150114. 90153-D. Liu, S., Wu, X., Gu, W., Yu, J., Wu, B., 2020. Influence of the digestive process on Stock, V., Fahrenson, C., Thuenemann, A., Dönmez, M.H., Voss, L., Böhmert, L., intestinal toxicity of polystyrene microplastics as determined by in vitro Caco-2 Braeuning, A., Lampen, A., Sieg, H., 2020. Impact of artificial digestion on the sizes models. Chemosphere 256, 127204. https://doi.org/10.1016/j. and shapes of microplastic particles. Food Chem. Toxicol. 135, 111010 https://doi. chemosphere.2020.127204. org/10.1016/j.fct.2019.111010. Mahler, G.J., Shuler, M.L., Glahn, R.P., 2009. Characterization of Caco-2 and HT29-MTX Tamargo, A., Molinero, N., Reinosa, J.J., Alcolea-Rodriguez, V., Portela, R., Bañares, M. cocultures in an in vitro digestion/cell culture model used to predict iron A., Fernández, J.F., Moreno-Arribas, M.V., 2022. PET microplastics affect human gut bioavailability. J. Nutr. Biochem. 20 (7), 494–502. https://doi.org/10.1016/j. microbiota communities during simulated gastrointestinal digestion, first evidence jnutbio.2008.05.006. of plausible polymer biodegradation during human digestion. Sci. Rep. 12 (1), 528. Maiti, N.C., Apetri, M.M., Zagorski, M.G., Carey, P.R., Anderson, V.E., 2004. Raman https://doi.org/10.1038/s41598-021-04489-w. spectroscopic characterization of secondary structure in natively unfolded proteins: Wright, S.L., Kelly, F.J., 2017. Plastic and human health: a micro issue? Environ. Sci. alpha-synuclein. J. Am. Chem. Soc. 126, 2399–2408. https://doi.org/10.1021/ Technol. 51 (12), 6634–6647. https://doi.org/10.1021/acs.est.7b00423. ja0356176. Xu, J.L., Lin, X., Wang, J.J., Gowen, A.A., 2022. A review of potential human health Masset, T., Ferrari, B.J.D., Oldham, D., Dudefoi, W., Minghetti, M., Schirmer, K., impacts of micro- and nanoplastics exposure. Sci. Total Environ. 851 (Pt-1), 158111 Bergmann, A., Vermeirssen, E., Breider, F., 2021. Vitro digestion of tire particles in a https://doi.org/10.1016/j.scitotenv.2022.158111. fish model (Oncorhynchus mykiss): solubilization kinetics of heavy metals and effects Yong, C.Q.Y., Valiyaveettil, S., Tang, B.L., 2020. Toxicity of microplastics and of food coingestion. Environ. Sci. Technol. 55 (23), 15788–15796. https://doi.org/ nanoplastics in mammalian systems. Int. J. Environ. Res. Publ. Health 17 (5), 1509. 10.1021/acs.est.1c04385. https://doi.org/10.3390/ijerph17051509. Mazilu, M., De Luca, A.C., Riches, A., Herrington, C.S., Dholakia, K., 2010. Optimal Zhang, Y.X., Wang, M., Yang, L., Pan, K., Miao, A.J., 2022. Bioaccumulation of algorithm for fluorescence suppression of modulated Raman spectroscopy. Opt differently-sized polystyrene nanoplastics by human lung and intestine cells. Express 18, 11382–11395. https://doi.org/10.1364/OE.18.011382. J. Hazard Mater. 439, 129585 https://doi.org/10.1016/j.jhazmat.2022.129585. 10