Microplastic particles reduce EROD-induction specifically by highly lipophilic compounds in RTL-W1 cells

Ecotoxicology and Environmental Safety 189 (2020) 110041

Contents lists available at ScienceDirect

Ecotoxicology and Environmental Safety journal homepage: www.elsevier.com/locate/ecoenv

Microplastic particles reduce EROD-induction specifically by highly lipophilic compounds in RTL-W1 cells

T

Patrick Heinrich∗, Thomas Braunbeck Aquatic Ecology and Toxicology Section, Centre for Organismal Studies, University of Heidelberg, Im Neuenheimer Feld 504, D-69120, Heidelberg, Germany

A R T I C LE I N FO

A B S T R A C T

Keywords: Sorption Hydrophobic organic contaminants Surface chemistry Pollutant removal Detoxification Depuration

Microplastic particles (MPs) from lipophilic polymers have been shown to efficiently accumulate hydrophobic organic contaminants (HOCs) in aquatic environments. MPs have, therefore, frequently been discussed as vectors for contaminants, enhancing HOC uptake by various organisms after ingestion followed by pollutant release; however, integrative models of sorption argue against this mechanism and even predict cleansing of pollutants from biological systems under particular circumstances. In order to experimentally investigate such a depuration mechanism, RTL-W1 cells were dosed with three 7-ethoxyresorufin-O-deethylase (EROD) inducers of distinct lipophilicity via the medium before adding both native and hexane-purified polyethylene MPs (20–25 μm) to the medium surface. EROD activity was significantly reduced in the presence of MP, the extent of which correlated with the inducers’ lipophilicity (KOW) and thus affinity to MP. For hexane-purged MPs and TCDD (KOW = 6.8), MPs reduce the bioavailability by up to 79%; the effect was marginally weaker with benzo[k]fluoranthene (KOW = 6.11) and almost absent with β-Naphthoflavone (KOW = 4.68). Compared to hexane-purged MPs, native particles possessed slightly less detoxification potential. These experimental results corroborate theoretically predicted mechanisms of detoxification via MPs. Yet, it is unclear if, under corresponding conditions in the environment, MPs can compete with organismal tissues for highly lipophilic compounds and, if so, to which degree they may act as a sink reducing the amount of bioavailable pollutants in situ. However, the present results suggest that in scenarios where pollutant-free MPs interact with organisms that accumulated HOCs via other routes of uptake, qualitatively the presence of such a mechanism is likely.

1. Introduction Microplastic particles (MPs) are known contaminants in virtually any environmental compartment (Cole et al., 2011; He et al., 2018; Li et al., 2018; Mintenig et al., 2019) and have also been detected in foodstuffs (Smith et al., 2018) and tap water (Eerkes-Medrano et al., 2019). Exposure of both wildlife and humans to MPs is considered very likely, albeit clear detrimental effects as a direct consequence are still unclear and being discussed (Foley et al., 2018). Most MPs are physical degradation products of larger plastic debris (‘secondary MPs’) present in the environment in comparatively large amounts (Browne et al., 2011). MPs originate from polymers with a wide range of anthropogenic uses, among which polyethylene (PE), polypropylene, polystyrene and polyamide are the most commonly detected types in environmental samples (Erni-Cassola et al., 2019). Alternatively, MPs are being produced in the respective sizes (‘primary MPs’) and applied for specific industrial purposes, e.g. as peeling abrasives in cosmetics, matting agents in lacquers or slip-resistance

∗

coatings. While MPs have been shown to be able to physically harm certain organisms after ingestion (Rehse et al., 2016), a major issue associated with toxicity is the hypothesized role of lipophilic polymers as vectors for hydrophobic organic contaminants (HOCs). Since many MP polymers are characterized by a non-polar, hydrophobic surface chemistry, they are likely to accumulate substances of similarly lipophilic solubility properties, which has conclusively been shown in numerous studies (e.g. Teuten et al., 2007). Since the mechanisms involved ultimately depend on relatively simple physisorption (the theoretical basis of which has been well established for over 100 years, Dabrowski, 2001), the major factors involved in sorption of HOCs are well understood, which has allowed the creation of mathematical sorption models for HOCs towards MPs (e.g. Gouin et al., 2011; Koelmans et al., 2013). Early, superficial conceptions postulated that MPs would accumulate minor concentrations of HOCs from the water phase, which – after ingestion by organisms – may be transferred to organismal tissues, where they would elicit toxic effects depending on the compounds

Corresponding author. E-mail address: [email protected] (P. Heinrich).

https://doi.org/10.1016/j.ecoenv.2019.110041 Received 26 March 2019; Received in revised form 29 November 2019; Accepted 30 November 2019 0147-6513/ © 2019 Elsevier Inc. All rights reserved.

Ecotoxicology and Environmental Safety 189 (2020) 110041

P. Heinrich and T. Braunbeck

Fig. 1. a–f: Detoxification of β-naphthoflavone (β-NF; a,b), benzo[k]fluoranthene (BkF; c,d) and 2,3,7,8-tetrachloro-p-dibenzodioxin (TCDD; e,f) by native (a, c, e) and hexane-purged (b, d, f) polyethylene microplastic particles (MPs) assessed via reduced EROD induction in RTL-W1 cells (one representative replicate out of 4 independent runs per inducer/MP combination; means ± SEM; n = 8 wells per treatment or n = 3 for standard curves). Cells were dosed with the concentrations of inducers indicated, which were added via the medium; in designated wells, the cell culture medium surface was saturated with polyethylene MPs. EROD inducers were added at known concentrations to allow estimation of MP-bound, non-bioavailable inducers; datapoints of higher inducer concentrations are not displayed for reasons of clarity of the graph. Treatments without (black columns) and with (white columns) MPs were compared pairwise using two-tailed t-tests (n.s. = not significant; * = p < 0.05; ** = p < 0.01; *** = p < 0.001; # = negative value slightly beyond graph range).

Lohmann, 2017). Rather, they even predicted that under specific circumstances MPs may even detoxify organisms by competition for the pollutants (Gouin et al., 2011) by a mechanism similar to activated charcoal therapy. In this study, our aim was to experimentally investigate the barely discussed role of MPs as a hypothetical remedy for organisms by their ability to bind lipophilic contaminants previously acquired by other routes of uptake. This position may complement other studies, which for the most part only investigate the release of pollutants to organisms, eventually enabling a more comprehensive picture. For this, we utilized

involved (e.g. Teuten et al., 2007). However, this perspective had to be considered inadequate, since it only considered a fraction of the total processes involved in the environment, where MPs are only one single component in a complex network of interdependent distribution equilibria of HOCs between organisms, non-organismal organic phases (e.g. sediments, particulate organic matter (POM), dissolved organic carbon (DOC)) and water. More comprehensive models understanding MPs as only one component of a more complex sorption network argued against the postulated role as an enhancer for bioavailability of HOCs for several reasons (Burns and Boxall, 2018; Koelmans et al., 2016; 2

Ecotoxicology and Environmental Safety 189 (2020) 110041

P. Heinrich and T. Braunbeck

wells (n = 3 per concentration), serial dilutions of the respective inducers were added to establish calibration curves for quantification of bioavailability in presence or absence of MP via EROD induction (see Fig. 1a–f for exact concentrations and illustration), as well as different cell numbers (10,000 to 40,000 cells per well) for quantification of cellular metabolism. After sealing and incubation for 24 h, the plates were processed and EROD activity was measured as previously described in detail (Heinrich and Braunbeck, 2019a): Cells were incubated for 60 min in 300 μl of phenol-red-free, HEPES-buffered DMEM medium containing 2 μM 7-ethoxyresorufin, after which 250 μl were transferred to 100 μl PBS in a black fluorescence detection plate along with resorufin standards (18.75–150 μM). Fluorescence was measured at 544/24 nm excitation and 593/46 nm (Tecan Genios, Crailsheim, Germany). The remaining cell monolayers were incubated in HEPESbuffered DMEM medium containing 0.3 mg/ml MTT for 3 h before dissolving the product in DMSO with 6.25% ammonia and measuring absorbance at 595 nm. All experiments were independently replicated four times (run A-D).

a recently published experimental approach allowing the direct experimental bioanalytical quantification of sorption to and from MPs in a biological system using cultured cells (Heinrich and Braunbeck, 2019a) on basis of the induction of 7-Eethoxyresorufin-O-deethylase (EROD) activity by dioxin-like compounds. After applying native or hexanepurged PE particles to the medium surface of cell cultures previously dosed with model pollutants of distinct lipophilicity, a change in the magnitude of the cells’ responses towards those compounds was measured. If compounds were binding to MP under cell culture conditions, a reduced cellular response is to be expected. 2. Material & methods 2.1. Reagents & materials The microplastic particles employed in the assays were commercially available, non-modified high-density (0.96 g/cm3) polyethylene with a mean diameter of 20–25 μm, mechanically micronized and designated for industrial applications in inks and coatings. Information on the supplier is intentionally undisclosed for potential legal conflicts (but will be made available upon request). Since the particles used had been shown to potentially contain minor amounts of EROD-inducing substances (in the low ng/g range, Heinrich and Braunbeck, 2019a) and possibly other compounds from the manufacturing process, a 10 g batch of the MPs was purged by two consecutive washing steps with 100 ml hexane each, followed by decanting of the solvent and complete drying of the particles under vacuum in a Laborota 4000 rotary evaporator (Heidolph, Schwabach, Germany). Unless specifically noted, all reagents were obtained from SigmaAldrich (Deisenhofen, Germany) at the highest purity available.

2.4. Data analysis Bioavailable concentrations were calculated individually for each test by regression analysis of the standard curve data employing a fourparametric logistic model. EROD activities were compared pairwise using a two-tailed t-test with several thresholds (p < 0.05, 0.01 and 0.001) after ensuring normality by Shapiro-Wilk's test. Factors involved in the detoxification magnitude within the dataset (48 tests) were assessed by three-way mixed ANOVA (3 × 2 × 2 layout), investigating inducers (β-NF, BkF, TCDD), doses (low and high) and MP pre-treatment (native and hexane-purged). Inducers were compared to each other using the Holm-Šidák method with significance thresholds of p < 0.05, 0.01 and 0.001.

2.2. Cell culture 3. Results RTL-W1 cells (Lee et al., 1993) were maintained in 75 cm2 flasks (Greiner) in L-15 medium with 10% (v/v) fetal bovine serum (Biochrom, Berlin, Germany) and penicillin/streptomycin (100 U/ml and 100 μg/ml, respectively) at 20 °C and weekly passaged at a ratio of 1:2 after dislodgement with trypsin/EDTA. Cultures used for experiments were below passage number 110 and tested negative for mycoplasma contamination in screenings at regular intervals.

As exemplified in Fig. 1a, addition of native MP to wells dosed with β-NF did not cause a significant reduction in EROD activity vs. controls (with inducer but without MP) regardless of the particular concentration (p > 0.05). In contrast, with hexane-purged MPs (Fig. 1b), there was a significant difference (p < 0.01) at 10 nM. For BkF (Fig. 1c), a significant reduction of EROD induction was observed for native MP at both 200 pM (p < 0.05) and 1 nM (p < 0.001), whereas the difference was highly significant (p < 0.001) for hexane-purged MPs (Fig. 1d). EROD activity induced by both concentrations of TCDD was highly significantly (p < 0.001) reduced regardless of MP-pretreatment or concentration (Fig. 1e and f). Using the internal inducer standard calibration curves, absolute amounts of bioavailable inducer per well were calculated, from which comprehensive percentile changes were derived, which are shown in detail in Table 1 and illustrated in Fig. 2. According to ANOVA, the choice of inducer had a prominent effect on the reduction of its bioavailability in presence of MPs (F (2,7) = 38.179, p < 0.001); differences for β-NF vs. BkF and TCDD were highly significant (p < 0.001) but non-significant (p = 0.138) for BkF vs. TCDD. The concentration levels of the inducers did not have an effect (F(1,7) = 0.004, p = 0.950), while purging the MPs with hexane resulted in significantly enhanced reduction of bioavailable inducers (F(1,7) = 9.127, p = 0.019). For β-NF, the supplementation of native MPs did not change bioavailability of the inducer for cells (Table 1), since percentile changes fluctuated around 0% and only occasionally indicated an increase in bioavailability. In the presence of hexane-purged MPs, a slight detoxification of β-NF by approx. 20% could be documented. When BkF was co-incubated with MPs (Table 1), the reduction in bioavailable inducer was considerably higher averaging 45–62%, while a clear effect of hexane-purging was observed only at the lower concentration (200

2.3. Quantification of cellular bioavailable inducers in the presence or absence of surface-dosed microplastic Stock solutions of the test substances were prepared in dimethyl sulfoxide (DMSO), whose final concentration was later adjusted to 0.01% (v/v) in all treatments including controls and calibration curves. Bioavailable concentrations of the EROD-inducers β-naphthoflavone (βNF), benzo[k]fluoranthene (BkF) and 2,3,7,8-tetracholo-p-dibenzodioxin (TCDD) were quantified using a concept recently presented in detail (Heinrich and Braunbeck, 2019a) with only minor modifications for treatment of the cells: 24 h after seeding, 100 μl of medium containing the treatments at twofold concentrations were added to 100 μl medium already present in the wells to establish the targeted final concentrations. β-NF, BkF and TCDD were diluted from 10,000x stock solutions in DMSO, setting up two distinct concentrations readily detectable via their EROD induction in the assay (2 nM/10 nM, 200 pM/ 1 nM and 2 pM/10 pM, respectively; cf. Fig. 1a–f). DMSO was added to a final concentration of 0.01% (v/v) to the inducer-free treatments (negative controls). Immediately after inducer dosage, MPs were added using a plastic spatula to completely cover the medium surface in each well after gentle shaking of the plate (n = 8 wells per MP-pre-treatment, native or hexane-purged, inducer combination and control treatments), resulting in approx. 275 μg resp. 96,000 particles interacting with medium (Heinrich and Braunbeck, 2019a). Into separate 3

Ecotoxicology and Environmental Safety 189 (2020) 110041

P. Heinrich and T. Braunbeck

4. Discussion

Table 1 Detoxification data of β-naphthoflavone, benzo[k]fluoranthene and 2,3,7,8tetrachloro-p-dibenzodioxin by addition of native or hexane-cleaned polyethylene microplastic particles (MPs). Data is given as percentile change in bioavailable inducer concentration (n = 8 per treatment and test) calculated from the individual inducer standard curves (n = 3 per concentration and test, cf. Fig. 1a–c). For illustration and comprehensive statistical analysis, see Fig. 2. β-Naphthoflavone Native MPs Hexane-purged MPs

The experimental findings described above agree with theoretical models on sorption of pollutants to MP; Gouin et al. (2011) predicted that only compounds with high lipophilicity (e.g. TCDD) are able efficiently bind to PE in an organismic resp. environmentally relevant environment, while those with only intermediate log KOW-values of < 5 would not interact under such conditions to a substantial degree. Indeed, the EROD-induction of TCDD and BkF was greatly reduced by addition of MP to the medium surface, however the effect towards the only intermediately lipophilic inducer β-NF was barely noticeable. In the presence of EROD-inducing substances, addition of polyethylene MPs to the surface of the cell culture medium resulted in most cases in a decrease of EROD induction (Fig. 1a–f). This decline in EROD induction is the result of reduced bioavailability of the inducers due to an interaction between the inducers and MPs, therefore becoming unavailable for cellular uptake, ultimately resulting in lower intracellular concentration and consequently weaker induction of the biotransformation enzyme EROD (Whyte et al., 2000). Similar indirect approaches for quantification of inducer concentrations on the basis of toxicological effects have frequently been applied for the quantitative bioanalysis of both single PAHs and complex environmental samples (Brack, 2003). In fact, the magnitude of detoxification was strictly dependent on the lipophilicity of the inducers: For β-NF (log KOW = 4.68), cellular concentrations were only reduced by 8% (disregarding the effects of hexane-purging and inducer dose, which are discussed later), while for the more lipophilic BkF (log KOW = 6.11) and TCDD (log KOW = 6.80), this effect was much more pronounced (51 and 62%, respectively). Given that the affinity of β-NF to bind to polyethylene is theoretically lower than that of the other compounds (however, based on lipophilicity only, as specific sorption data are not available), under the prevalent conditions β-NF appears to only loosely associate with MPs and is readily available for cellular uptake under the conditions described, eventually triggering effects. The more lipophilic inducers, however, apparently form a stronger association with the polymer, resulting in a more efficient detoxification following addition of MPs. Although there

Independent experimental runs 2 nM 10 nM 2 nM 10 nM

+16.67% - 10.98% - 11.71% - 20.03%

- 4.04% - 4.05% +13.38% -15.94%

+9.61% - 8.39% - 57.94% - 0.98%

Benzo[k]fluoranthene

Independent experimental runs

Native MPs Hexane-purged MPs

-

200 pM 1 nM 200 pM 1 nM

2,3,7,8-TCDD Native MPs Hexane-purged MPs

4.1. Polyethylene competes with tissue for pollutants, leading to differential detoxification depending on the compound's lipophilicity in vitro

44.39% 51.64% 90.25% 53.31%

-

26.59% 45.47% 71.73% 48.75%

-

48.40% 53.16% 45.70% 46.35%

+24.27% - 1.29% - 37.30% - 24.50%

-

62.55% 47.03% 41.40% 47.16%

-

68.67% 72.86% 97.73% 78.36%

Independent experimental runs 2 pM 10 pM 2 pM 10 pM

-

29.28% 40.94% 86.50% 58.22%

-

53.40% 59.25% 47.41% 54.36%

-

35.52% 66.99% 85.85% 64.10%

pM). For TCDD (Table 1), the reduction of bioavailability in the presence of MP was more conspicuous, showing reductions by 47–79%. Again, the reduction in bioavailable inducer was stronger for hexanepurged MPs, especially at the lower concentration of 2 pM. The choice of either a low or high concentration of inducer across the other factors did not affect mean bioavailability (40.9% vs. 40.6%), but variability was significantly higher for the lower concentrations (22.0% CV vs. 7.6% CV, p = 0.005).

Fig. 2. Detoxification of three EROD-inducers employed at two concentrations with distinct lipophilicity by either native (N, black) or hexane-purged (Hex, white) polyethylene microplastic particles (MPs) in RTL-W1 cells. Addition of MPs to the medium surface reduced the amounts of bioavailable inducer by the indicated percentages, means ± SEM from four independent experimental runs (cf. Table 1). Three-way ANOVA analysis revealed strong, significant effects of inducers (F (2,7) = 38.179, p < 0.001); differences between βnaphthoflavone and benzo[k]fluoranthene (BkF) or 2,3,7,8-tetrachloro-p-dibenzodioxin (TCDD) were highly significant (Holm- Šidák test; p < 0.001), while those of BkF vs. TCDD were not (p = 0.138). The inducer concentration did not have an effect on detoxification (F(1,7) = 0.004, p = 0.95), whereas the use of hexane-purged MP resulted in a stronger reduction of bioavailable inducer (F(1,7) = 9.127, p = 0.019) than with native MP.

4

Ecotoxicology and Environmental Safety 189 (2020) 110041

P. Heinrich and T. Braunbeck

less lipophilic, naturally occurring compounds (e.g. DOC, surfactants, …). Thereby, binding sites are blocked and consequently, potential for accumulation of lipophilic pollutants is reduced. Consequently, the results incorporating hexane-purged MPs should rather be considered a mechanistic control, indicating that the (putative) presence of surfacebound compounds can have a profound effect on the binding of additional substances to MP.

is a modest difference in lipophilicity between BkF and TCDD, statistical analysis failed to prove a stronger detoxification of TCDD (p = 0.138), however there is an obvious relationship between lipophilicity of compounds and their detoxification by MPs. As reported repeatedly (e.g. Fries and Zarfl, 2012; Karapanagioti et al., 2010; Rochman et al., 2013), highly lipophilic PE particles as well as those of a similar surface chemistry (e.g. polypropylene) are able to bind lipophilic substances with high affinity, which is a simple physiosorption mechanism: In a solvent system, substances will have the tendency to associate with phases of similar solution properties (in this case, non-polar environments). Driven by the tendency to establish a stable equilibrium, compounds will accumulate in such phases. The test system presented can be interpreted as a three-phase sorption system (Peng and Robinson, 1976): Pollutants will establish equilibria between the medium and MPs as well as between the medium and the cells. Under the experimental conditions given (isothermal and isobaric), equilibria are mainly dictated by affinities towards the respective phases. Under these assumptions, the present results indicate that for the more lipophilic compounds (approximately log KOW > 5–6), polyethylene MPs have a similar or even higher binding potential, if compared to living tissues, and are able to compete for the pollutant with lipophilicities frequently discussed in the context of MP's pollutant vector potential and toxicity. In a parallel study (Heinrich and Braunbeck, 2019b) investigating desorption of pollutants from MPs using the same material, where inducers were loaded to the MPs and then exposed to cultured cells (i.e. inverting the exposure scenario), results were in line with this hypothesis: β-NF, which is anticipated to only weakly bind polyethylene, was released completely from MPs, whereas BkF and TCDD largely resp. completely remained bound to the polymer.

4.3. In vitro detoxification did not depend on pollutant levels Comparisons between high and low EROD inducer concentrations across inducers and MP-pre-treatment did not reveal differences in the extent of detoxification. Common sorption models, where the relationship in concentration ratios of adsorbate on the adsorbent and in the liquid phase at different total concentration (e.g. Langmuir) is often explained by saturation of binding sites. This allows the hypothesis that higher concentrations may have a saturation effect, which would result in higher concentrations of free inducer in the liquid phase and, thus, increased bioavailability to the cells. This was apparently not the case, indicating that saturation effects did not play a role at the concentrations and conditions used in this study. However, the concentrations of the inducer had an obvious effect on the variability of results: Mean coefficients of variation across all inducers and MP-pretreatments were higher for the lower concentrations, which is mainly owing a certain bioanalytical uncertainty and measurement variation at lower concentrations. 4.4. Limitations for the discussion on microplastic as pollutant vectors in the environment Polyethylene MPs were shown to efficiently bind lipophilic compounds both in the in vitro system employed in this study as well as in general, which, under particular circumstances, may result in their detoxification by comparatively strong binding preferentially to the polymer rather than to organismal tissues. While the main intention of this study was to qualitatively confirm a putative mechanism for detoxification of MPs via HOCs in biological systems, it may also provide a more general insight into the role of MPs for the transfer of pollutants to and from intact organisms in the environment as well as confirm theoretically predicted mechanisms. Still, it is important to carefully consider the differences of this in vitro model and in vivo exposure scenarios. The in vivo analogues of the cell culture medium would be fluid compartments (blood, extracellular fluid …), whereas the cells represent cellular tissues. If the exposure scenario in this study (pollutantfree MP and cells, inducer added to the medium) is directly transferred to the in vivo situation, pollutants would exclusively be restricted to the circulatory system, which is rather unlikely. In pre-tests, pre-exposure of cells with pollutants for 24 h before addition of MPs was carried out to make the in vitro scenario more realistic, which would allow estimating the ability of MP to actually “pull” pollutants out of cells rather than only competing with cells for pollutants present in the medium. This, however, failed to arrive at appropriate sensitivity due to background fluctuations in the test system based on EROD induction (details not shown), thus disallowing adequately accurate quantification of bioavailable inducers. When lipophilic inducers were added to the relatively polar cell culture medium, these will rapidly accumulate within phases of similar solvent properties, i.e. within both cells and MPs. In fact, the in vitro exposure scenario of the present study would simulate injection of pollutants into organisms rather than gradual accumulation from the environment. Nevertheless, according to the universally accepted concept of sorption equilibria, there is always a certain fraction of the pollutant in less polar extracellular compartments, i.e. mainly in fatty tissues (Ossiander et al., 2008). Since those mechanisms also apply to intact organisms and aquatic systems in general, detoxification processes

4.2. Purging MPs with hexane enhances detoxification of pollutants in vitro In a previous study with the same test materials (Heinrich and Braunbeck, 2019a), results indicated that these may contain a certain amount of unidentified contaminants. These substances may possibly occupy specific binding sites and might, thus, interfere with binding of further substances. Therefore, in parallel to the tests with native particles, a set of experiments was conducted with a batch of MPs that had been extracted with hexane. Purging indeed significantly improved detoxification (across all other factors), if compared to native MPs (−49% vs. −33%), most probably due to more binding sites becoming available after removal of the adsorbates. Since a comprehensive chemical analysis of the MP extracts is not available, this can be neither conclusively proven, nor is there information on the chemical identity of the adsorbates. While the removal of adsorbate is the most attractive explanation for enhanced detoxification by hexane-purged MPs, other factors may play a role: Polymers are known to be modified in the presence of solvents by a process called swelling, which has been shown to impact sorption (Podivinská et al., 2017). Solvent molecules enter the solid phase of the polymer, increasing both its volume and accessibility to adsorbates. Although the purged MPs applied in this study had been vacuum-dried before use to remove any solvent from the particles’ solid phase, it cannot be excluded that exposure to hexane resulted in permanent structural alterations in the polymer, possibly enhancing absorption to the particles, similar to what has been postulated in context of pollutant-binding to MP (Goedecke et al., 2017). Regardless of the mechanism(s), pre-treatment of MPs with organic solvents could be shown to have an effect on the sorption properties, an effect that should be considered in experimental studies on the toxicity of MPs and associated substances. However, the limited environmental relevance of the results using hexane-purged MP should be underlined. Even if particles without any adhering substances were introduced into the environment at considerable amounts, they can be believed to quickly accumulate more or 5

Ecotoxicology and Environmental Safety 189 (2020) 110041

P. Heinrich and T. Braunbeck

highly lipophilic pollutants, resulting in only barely reversible binding, thus by trend reducing waterborne and therefore ultimately decrementing pollutant concentrations in organisms. Actually, very similar approaches have been discussed for the purposeful remediation of lipophilic contaminants in aquatic systems (Guerra et al., 2018; Matsuzawa et al., 2010). However, considering the vanishingly small amount of MPs in the environment (significantly contradicting the public perception in this issue) if compared to other natural compartments (Koelmans et al., 2016), the effects of MPs in the marine environment can been assumed to be insignificant, regardless if they act as vectors or sinks for pollutants.

similar to those observed in cell cultures are plausible to occur in complete organisms, albeit to a different, yet largely unknown extent. However, the present results suggest that – if by now only proven mechanistically – extracellular concentrations may be lowered by pollutant-free MPs, which may then lead to remobilization of pollutants stored in tissues (Merrill et al., 2013) due to a shift in equilibrium. For a number of reasons, sorption of HOCs to MPs in organisms can be expected to be much slower than under in vitro conditions. First, for HOCs stored in fatty tissues, several sorption transitions across epithelia have to take place, and distribution equilibria are known to require time to establish. This is very much dependent on both concentration gradients and affinity towards the involved phases (tissue, circulatory fluids and MP); regardless of the particular equilibrium concentration, sorption of HOCs from a polar to a non-polar phase (as it is the case in this study's scenario) can be expected to be much faster than vice versa (as would be the case in organisms, Plazinski et al., 2009). Given that MPs can be considered to only transiently interact with organisms (e.g. during passage though the gut, Hämer et al., 2014), the duration for taking up pollutants and consequently the total amount of adsorbed pollutants is limited (Rehse et al., 2018). In the present study, MPs were exposed for 24 h, which can be considered quite long given that gut passage times in organisms range from a few minutes in branchiopods (Murtaugh, 1985) to a few hours in fish and cetaceans (Mathiesen et al., 1995; Opuszynski and Shireman, 1991), which might lead to an overestimation of the detoxification capacity in this study. The second major factor to consider is the use of adsorbate-free (“naked”) particles in this study. When adhering substances are absent, due to a large number of available binding sites and steep concentration gradients, greatly accelerated sorption kinetics can be expected, since at the start of the sorption process, binding of adsorbate is statistically greatly favored over its (re-)release. In the environment, a multitude of natural substances (frequently summarized as dissolved organic carbon, DOC) are present, which will readily associate with MPs, blocking binding sites and reducing the rate and concentration of uptake of other substances via competitive sorption. While this highly complex process dependents on various factors (differential affinities, concentrations, kinetics …), detoxification by depuration in vivo is most probably slower and less extensive. Actually, assuming that the native test material used in this study actually harbors adsorbates as discussed earlier, the results with hexane-purged MPs (showing higher detoxification potential) support this hypothesis. It should be noted that the experimental model described presumes exposure of organisms to pollutant-free MPs; in the field, where virtually all MPs encountered by organisms are comparatively old (weathered), most probably at equilibrium with the surrounding water and therefore has already accumulated pollutants (Koelmans et al., 2016). After ingestion, in sum sorption will depend on the distribution equilibrium between the organism and MPs (Gouin et al., 2011; Koelmans et al., 2016). If pollutant-free organisms are exposed to polluted MPs, pollutants can be expected to be taken up via establishment of the distribution equilibrium; if the scenario is inverted (like in this study), progression towards the equilibrium conditions may lead to equilibrium-driven shifts of lipophilic pollutants from organismal tissue to MP, eventually resulting in detoxification of the organism as hypothesized earlier. Further, detoxification can be expected to be weaker for more hydrophilic polymers; however, PE (next to polypropylene) represents one of the most abundant and consequently relevant non-polar polymers known in the environment. Analogously to the weaker accumulation of lipophilic pollutants from the environment for such polymers (Šimko et al., 2006), equilibria of HOCs by tendency will probably be shifted towards organismal tissues. However, sorption processes involving MPs are interwoven, occur simultaneously and interact with each other. Emission of MPs to the environment can at least qualitatively be interpreted as an increase in the amount of abiotic lipophilic phase competing with organisms for

5. Conclusions Addition of plain polyethylene MPs was shown to reduce EROD induction by different model compounds in cell cultures. The magnitude of reduction, however, was strongly dependent on the lipophilicity of the inducers, as significant effects were only observed for BkF and TCDD, whereas the decrease with only moderately lipophilic β-NF was much less pronounced. Pre-treatment of MPs with an organic solvent before exposure to remove putative adsorbates from the material resulted in stronger reductions, presumably via increasing the number of sites available for binding of pollutants. With respect to a putative role of MPs as a vector for pollutants in the environment, the present in vitro study suggests the presence of a mechanism that allows detoxification of lipophilic pollutants by lipophilic polymers, which may be able to reduce body-burdens of contaminated organisms by a mechanism similar to that of activated charcoal treatment. However, the relevance of this mechanism still has to be verified for intact organisms and under more environmentally representative conditions, to allow for comparisons with appropriate in vivo experimental models to estimate the extent of a possible detoxification mechanism associated with MPs for aquatic organisms in the environment. Author contributions Patrick Heinrich: Conceptualization, Methodology, Validation, Formal analysis, Investigation, Writing – original draft, Writing – review & editing, Visualization, Thomas Braunbeck: Resources, Writing – review & editing, Supervision, Project administration, Funding acquisition. Declaration of competing interest The authors declare no competing financial interests. Acknowledgements This study was funded by the German Federal Ministry for Science and Research (BMBF) under contracts no. 03F0735A within the “Joint Programming Initiative Healthy and Productive Seas and Oceans” (JPI Oceans) project EPHEMARE (“Ecotoxicological Effects of Microplastics in Marine Ecosystems”) and no. 02WRS1378J for the joint project MiWa (“Microplastics in the Water Cycle – Sampling, Specimen Treatment, Chemical Analyses, Distribution, Removal and Risk Assessment”) within the scope of the RiSKWa (“Risk Management of Novel Contaminants and Pathogens in the Water Circle”) program. References Brack, W., 2003. Effect-directed analysis: a promising tool for the identification of organic toxicants in complex mixtures? Anal. Bioanal. Chem. 397–407. https://doi.org/10. 1007/s00216-003-2139-z. Browne, M.A., Crump, P., Niven, S.J., Teuten, E., Tonkin, A., Galloway, T., Thompson, R., 2011. Accumulation of microplastic on shorelines woldwide: sources and sinks. Environ. Sci. Technol. 45, 9175–9179. https://doi.org/10.1021/es201811s.

6

Ecotoxicology and Environmental Safety 189 (2020) 110041

P. Heinrich and T. Braunbeck

Biol. Toxicol. 9, 279–294. Li, J., Liu, H., Chen, J.P., 2018. Microplastics in freshwater systems : a review on occurrence, environmental effects, and methods for microplastics detection. Water Res. 137, 362–374. https://doi.org/10.1016/j.watres.2017.12.056. Lohmann, R., 2017. Microplastics are not important for the cycling and bioaccumulation of organic pollutants in the oceans — but should microplastics Be considered POPs themselves? Integr. Environ. Assess. Manag. 13, 460–465. https://doi.org/10.1002/ ieam.1914. Mathiesen, S.D., Aagnes, T.H., Sørmo, W., Nordøy, E.S., Blix, A.S., Olsen, M.A., 1995. Digestive physiology of minke whales. Dev. Mar. Biol. 4, 351–359. https://doi.org/ 10.1016/S0163-6995(06)80036-3. Matsuzawa, Y., Kimura, Z., Nishimura, Y., Shibayama, M., Hiraishi, A., 2010. Removal of hydrophobic organic contaminants from aqueous solutions by sorption onto biodegradable polyesters. J. Water Resour. Prot. 2, 214–221. https://doi.org/10.4236/ jwarp.2010.23024. Merrill, M. La, Emond, C., Kim, M.J., Antignac, J., Bizec, B., Le Birnbaum, L.S., Barouki, R., 2013. Toxicological function of adipose tissue: focus on persistent organic pollutants. Environ. Heath Perspect. 162, 162–169. Mintenig, S.M., Löder, M.G.J., Primpke, S., Gerdts, G., 2019. Low numbers of microplastics detected in drinking water from ground water sources. Sci. Total Environ. 648, 631–635. https://doi.org/10.1016/j.scitotenv.2018.08.178. Murtaugh, P.A., 1985. The influence of food concentration and feeding rate on the gut residence tune of Daphnia. J. Plankton Res. 7, 415–420. Opuszynski, K., Shireman, J.V., 1991. Food passage time and daily ration of bighead carp, Aristichthys nobilis, kept in cages. Environ. Biol. Fish. 30, 387–393. Ossiander, L., Reichenberg, F., Mclachlan, M.S., Mayer, P., 2008. Immersed solid phase microextraction to measure chemical activity of lipophilic organic contaminants in fatty tissue samples. Chemosphere 71, 1502–1510. https://doi.org/10.1016/j. chemosphere.2007.11.060. Peng, D.-Y., Robinson, D.B., 1976. Two and three phase equilibrium calculations for systems containing water. Can. J. Chem. Eng. 54, 595–599. https://doi.org/10.1002/ cjce.5450540541. Plazinski, W., Rudzinski, W., Plazinska, A., 2009. Theoretical models of sorption kinetics including a surface reaction mechanism: a review. Adv. Colloid Interface Sci. 152, 2–13. https://doi.org/10.1016/j.cis.2009.07.009. Podivinská, M., Jindrová, K., Chmelar, J., Kosek, J., 2017. Swelling of polyethylene particles and its relation to sorption equilibria under gas-phase polymerization conditions. J. Appl. Polym. Sci. 45035, 1–7. https://doi.org/10.1002/app.45035. Rehse, S., Kloas, W., Zar, C., 2016. Short-term exposure with high concentrations of pristine microplastic particles leads to immobilisation of Daphnia magna. Chemosphere 153, 91–99. https://doi.org/10.1016/j.chemosphere.2016.02.133. Rehse, S., Kloas, W., Zarfl, C., 2018. Microplastics reduce short-term effects of environmental contaminants. Part I: effects of bisphenol A on freshwater zooplankton are lower in presence of polyamide particles. Int. J. Environ. Res. Public Health 15, 280. https://doi.org/10.3390/ijerph15020280. Rochman, C.M., Hoh, E., Hentschel, B.T., Kaye, S., 2013. Long-Term field measurement of sorption of organic contaminants to five types of plastic pellets: implications for plastic marine debris. Environ. Sci. Technol. 47, 1646–1654. Šimko, P., Kolek, E., Skláršová, B., Šimon, P., 2006. Removal of benzene from water by adsorption on polymer materials. J. Food Nutr. Res. 45, 179–182. Smith, M., Love, D.C., Rochman, C.M., Neff, R.A., 2018. Microplastics in seafood and the implications for human health. Curr. Environ. Heal. Reports 5, 375–386. Teuten, E.L., Rowland, S.J., Galloway, T.S., Thompson, R.C., 2007. Potential for plastics to transport hydrophobic contaminants. Environ. Sci. Technol. 41, 7759–7764. Whyte, J.J., Jung, R.E., Schmitt, C.J., Tillitt, D.E., 2000. Ethoxyresorufin-O-deethylase (EROD) activity in fish as a biomarker of chemical exposure. Crit. Rev. Toxicol. 30, 347–570. https://doi.org/10.1080/10408440091159239.

Burns, E.E., Boxall, A.B.A., 2018. Microplastics in the aquatic environment: evidence for or against adverse impacts and major knowledge gaps. Environ. Toxicol. Chem. 37, 2776–2796. https://doi.org/10.1002/etc.4268. Cole, M., Lindeque, P., Halsband, C., Galloway, T.S., 2011. Microplastics as contaminants in the marine environment: a review. Mar. Pollut. Bull. 62, 2588–2597. https://doi. org/10.1016/j.marpolbul.2011.09.025. Dabrowski, A., 2001. Adsorption - from theory to practice. Adv. Colloid Interface Sci. 93, 135–224. Eerkes-Medrano, D., Leslie, H.A., Quinn, B., 2019. Microplastics in drinking water : a review and assessment. Curr. Opin. Environ. Sci. Heal. https://doi.org/10.1016/j. coesh.2018.12.001. Erni-Cassola, G., Zadjelovic, V., Gibson, M.I., Christie-Oleza, J.A., 2019. Distribution of plastic polymer types in the marine environment; A meta-analysis. J. Hazard Mater. 369, 691–698. https://doi.org/10.1016/j.jhazmat.2019.02.067. Foley, C.J., Feiner, Z.S., Malinich, T.D., Höök, T.O., 2018. A meta-analysis of the effects of exposure to microplastics on fish and aquatic invertebrates. Sci. Total Environ. 631–632, 550–559. https://doi.org/10.1016/j.scitotenv.2018.03.046. Fries, E., Zarfl, C., 2012. Sorption of polycyclic aromatic hydrocarbons (PAHs) to low and high density polyethylene (PE). Environ. Sci. Pollut. Res. 19, 1296–1304. https://doi. org/10.1007/s11356-011-0655-5. Goedecke, C., Mülow-Stollin, U., Hering, S., Richter, J., Piechotta, C., Paul, A., Braun, U., 2017. A first pilot study on the sorption of environmental pollutants on various microplastic materials. J. Environ. Anal. Chem. 4, 1–8. https://doi.org/10.4172/23802391.1000191. Gouin, T., Roche, N., Lohmann, R., Hodges, G., 2011. A thermodynamic approach for assessing the environmental exposure of chemicals absorbed to microplastic. Environ. Sci. Technol. 45, 1466–1472. https://doi.org/10.1021/es1032025. Guerra, F.D., Attia, M.F., Whitehead, D.C., Alexis, F., 2018. Nanotechnology for Environmental Remediation : Materials and Applications 1–23. https://doi.org/10. 3390/molecules23071760. Hämer, J., Gutow, L., Köhler, A., Saborowski, R., 2014. Fate of microplastics in the marine isopod idotea emarginata. Environ. Sci. Technol. 48, 13451–13458. https:// doi.org/10.1021/es501385y. He, D., Luo, Y., Lu, S., Liu, M., Song, Y., Lei, L., 2018. Trends in Analytical Chemistry Microplastics in soils: analytical methods, pollution characteristics and ecological risks. Trends Anal. Chem. 109, 163–172. https://doi.org/10.1016/j.trac.2018.10. 006. Heinrich, P., Braunbeck, T., 2019a. Microplastic testing in vitro: realistic loading of pollutants, surfactant-free solid surface-dosing and bioanalytical detection using a sensitivity- optimized EROD assay. Toxicol. In Vitro 54, 194–201. https://doi.org/10. 1016/j.tiv.2018.10.002. Heinrich, P., Braunbeck, T., 2019b. Bioavailability of microplastic-bound pollutants in vitro: the role of adsorbate lipophilicity and surfactants. Comp. Biochem. Physiol. C Toxicol. Pharmacol. 221, 59–67. https://doi.org/10.1016/j.cbpc.2019.03.012. Karapanagioti, H.K., Ogata, Y., Takada, H., 2010. Eroded plastic pellets as monitoring tools for polycyclic aromatic hydrocarbons (PAH): laboratory and field studies. Glob. NEST J. 12, 327–334. Koelmans, A.A., Bakir, A., Burton, G.A., Janssen, C.R., 2016. Microplastic as a vector for chemicals in the aquatic environment: critical review and model-supported reinterpretation of empirical studies. Environ. Sci. Technol. 50, 3315–3326. https:// doi.org/10.1021/acs.est.5b06069. Koelmans, A.A., Besseling, E., Wegner, A., Foekema, E.M., 2013. Plastic as a carrier of POPs to aquatic organisms: a model analysis. Environ. Sci. Technol. 47, 7812–7820. https://doi.org/10.1021/es401169n. Lee, L.E., Clemons, J.H., Bechtel, D.G., Caldwell, S.J., Han, K.B., Pasitschniak-Arts, M., Mosser, D.D., Bols, N.C., 1993. Development and characterization of a rainbow trout liver cell line expressing cytochrome P450-dependent monooxygenase activity. Cell

7