Toxicology in Vitro 29 (2015) 345–351

Contents lists available at ScienceDirect

Toxicology in Vitro journal homepage: www.elsevier.com/locate/toxinvit

Differential immunomodulatory responses to nine polycyclic aromatic hydrocarbons applied by passive dosing Gertie J. Oostingh a,⇑, Kilian E.C. Smith b,1, Ulrike Tischler a, Isabella Radauer-Preiml a, Philipp Mayer b,2 a b

Department of Molecular Biology, University of Salzburg, Hellbrunnerstrasse 34, 5020 Salzburg, Austria Department of Environmental Science, Aarhus University, Frederiksborgvej 399, DK-4000 Roskilde, Denmark

a r t i c l e

i n f o

Article history: Received 3 March 2014 Accepted 18 November 2014 Available online 27 November 2014 Keywords: Polycyclic aromatic hydrocarbons Passive dosing Cell culture Immunomodulation

a b s t r a c t Studying the effects of hydrophobic chemicals using in vitro cell based methods is hindered by the difficulty in bringing and keeping these chemicals in solution. Their effective concentrations are often lower than their nominal concentrations. Passive dosing is one approach that provides defined and stable dissolved concentrations during in vitro testing, and was applied to control and maintain freely dissolved concentrations of polycyclic aromatic hydrocarbons (PAHs) at levels up to their aqueous solubility limit. The immunomodulatory effects of 9 different PAHs at aqueous solubility on human bronchial epithelial cells were determined by analysing the cytokine promoter expression of 4 different inflammatory cytokines using stably transfected recombinant A549 cell lines. Diverse immunomodulatory responses were found with the highest induction observed for the most hydrophobic PAHs chrysene, benzo(a)antracene and benzo(a)pyrene. Cytokine promoter expression was then studied in dose response experiments with acenaphthene, phenanthrene and benzo(a)anthracene. The strongest induction was observed for benzo(a)anthracene. Cell viability analysis was performed and showed that none of the PAHs induced cytotoxicity at any of the concentrations tested. Overall, this study shows that (1) immunomodulatory effects of PAHs can be studied in vitro at controlled freely dissolved concentrations, (2) the most hydrophobic PAHs were the strongest inducers and (3) induction was often higher at lower exposure levels and decreased then with concentration despite the apparent absence of cytotoxicity. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction More than 100 different polyaromatic hydrocarbons (PAHs) have been identified in the environment and working place (Srogi, 2007). Low level exposure to PAHs mainly occurs via food ingestion, whereas increased exposure levels are observed when people are occupationally exposed (Srogi, 2007). The latter group includes asphalt workers, coke oven workers and professional drivers, with exposure mainly taking place via inhalation (Karakaya et al., 2004, 1999; Srogi, 2007). Exposure to PAHs has been linked to cancer development, and the mechanisms behind the carcinogenicity of these compounds are well-described. Benzo(a)pyrene is ⇑ Corresponding author at: Biomedical Sciences, Salzburg University of Applied Sciences, Urstein Sued 1, 5412 Puch/Salzburg, Austria. Tel.: +43 50 2211 1410; fax: +43 50 2211 1499. E-mail address: [email protected] (G.J. Oostingh). 1 Current address: Convergence Environment Team, Korea Institute of Science and Technology Europe Forschungsgesellschaft mbH, Campus E7 1, University of Saarland, 66123 Saarbrücken, Germany. 2 Current address: Department of Environmental Engineering, Technical University of Denmark, Bygningstorvet 115, 2800 Kongens Lyngby, Denmark. http://dx.doi.org/10.1016/j.tiv.2014.11.007 0887-2333/Ó 2014 Elsevier Ltd. All rights reserved.

the most carcinogenic PAH, particularly when taking into account the potentially high exposure to this compound (Petry et al., 1996). A meta-analysis has shown that the risk of lung cancer is also increased upon inhalation of PAHs (Armstrong et al., 2004). The potential of PAHs to induce carcinogenesis in the lung has also been shown using in vitro systems (Mollerup et al., 2001). Furthermore, a recent study has shown that the exposure of human airway cells to benzo(a)pyrene results in mitochondrial dysfunction and an altered oxidation status, leading to cellular toxicity (Min et al., 2011). Oxidative cellular stress induced by PAHs can lead to a number of different responses, including those related to the immune system (Sorensen et al., 2003). For example, exposure of the human alveolar cell line A549 to benzo(a)pyrene or 1-nitropyrene resulted in the activation of the NF-jB pathway, which led to an increased IL-8 expression (Pei et al., 2002). Immunomodulatory effects due to cellular stress were also found after exposure of A549 cells to fluoranthene when directly added to the cell culture (Oostingh et al., 2008). Although the lung epithelium has long been regarded to function only as a physical barrier for foreign compounds, more recently the role of the lung epithelium in innate and adaptive

346

G.J. Oostingh et al. / Toxicology in Vitro 29 (2015) 345–351

immunological responses has become increasingly recognised. Alveolar and bronchial epithelial cells can secrete a range of different cytokines upon exposure to foreign compounds that can promote an innate immune response, and in certain situations also specifically direct adaptive immune responses (Lambrecht and Hammad, 2008, 2010). Therefore, the current study investigated the immunomodulatory response of lung epithelial cells to different PAHs using the alveolar epithelial cell line A549. The immunomodulation of several different environmental chemicals, single walled carbon nanotubes and nanoparticles have already been described in previous publications, which have shown the suitability of this approach for screening for immunomodulatory effects (Herzog et al., 2009; Oostingh et al., 2008; Pfaller et al., 2010). PAHs are poorly soluble and hydrophobic compounds that are susceptible to sorptive and in some cases evaporative losses. Such losses can be dramatic, particularly when testing in open plastic cell culture plates at elevated temperatures, which are the typical test conditions for in vitro assays. To date, most human immunomodulation studies have involved first dissolving the PAHs in dimethyl sulfoxide (DMSO), followed by direct or indirect dosing of the aqueous solutions used for the in vitro tests (Davila et al., 1996; Machala et al., 2001; Min et al., 2011; Murahashi et al., 2007; Oostingh et al., 2008; Pei et al., 2002). This approach is practical, but the dissolved concentrations of the PAHs are poorly defined since sorption to the plastic or serum components, evaporative losses and precipitation are unaccounted for. This often leads to freely dissolved concentrations that are much lower than the nominal added concentrations and that also decrease during the test (Schreiber et al., 2008; Riedl and Altenburger, 2007). Freely dissolved concentrations are generally perceived as the effective exposure concentration in in vitro tests, and reductions in freely dissolved concentrations can thus directly affect both the in vitro response and furthermore the apparent sensitivity of the test (Heringa et al., 2004; Gülden and Seibert, 2005). In addition, protein-chemical interactions are also an issue for in vivo exposures, since pulmonary surfactants lining the lung epithelium bind PAHs and interact with other binding constituents (Zhao et al., 2012). Indirect dosing involves the initial dissolution of the test chemical in DMSO in the cell culture medium, followed by addition of this medium to the cells. This has been shown to give more reliable data compared to the direct spiking of test chemical in DMSO to the cells (Tanneberger et al., 2010). However, even with indirect spiking, losses are not compensated for, and in any case the effective freely dissolved concentrations are still unknown. Moreover, direct or indirect spiking also result in the addition of DMSO to the cells in culture. Although DMSO might not affect the cells when administered alone, spiking inevitably results in the analysis of mixture toxicity since the DMSO could interact or modify the effect of the test chemicals. Therefore, in order to produce reliable, defined and constant exposure of hydrophobic compounds in cell culture systems new dosing approaches have been developed; with one being passive dosing (Booij et al., 2011; Kramer et al., 2010; Kwon et al., 2009; Smith et al., 2010). Therefore, in the presented study the PAHs were administered via passive dosing using silicone, which ensured a well-controlled dosing regimen (Smith et al., 2010, 2013). Passive dosing involves a reservoir of test chemical dissolved in a biocompatible polymer, such as silicone, acting as a partitioning source to the cell culture medium, giving stable dissolved concentrations at well-defined levels up to and including aqueous solubility. Stably transfected promoter cell lines, containing the interleukin (IL)-8 promoter sequence or NF-jB binding sequences, were used to determine the effects of the different PAHs at the promoter level. The advantages when using these cell lines are multiple. The luciferase reporter system is very sensitive and small changes in promoter activation can be

detected. Moreover, previous studies performed with these cell lines have shown that alterations in cytokine and chemokine promoter induction occur at earlier stages than apoptosis or cell death (Oostingh et al., 2008; Röder-Stolinski et al., 2008). Initially, the immunomodulatory responses upon exposure to 9 PAHs at their respective aqueous solubilities were determined, and for 3 of these PAHs full concentration–response testing was additionally performed. Cell viability was always analysed in parallel, to determine whether the observed responses were due to cell death as such. The results were analysed in relation to a number of different factors, including the properties of the PAHs and their concentrations. 2. Materials and methods 2.1. Chemicals and materials Passive dosing was performed using food-grade silicone O-rings (outer diameter of 14.4 mm, inner diameter of 9.6 mm, mass of 231 mg (C.V. 1%, n = 10), calculated volume of 0.171 ml (Order no. ORS-0096-24, Altec, Cornwall, United Kingdom). Cell culture was performed in Costar 24-well flat bottom cell culture-treated polystyrene plates (Corning Inc., Corning, NY). Nine PAHs were selected as model hydrophobic compounds as listed in Table 1. The chemicals were obtained from the following companies: acenaphthene (99%, Sigma), fluorene (99%, Aldrich, Germany), phenanthrene (99.5%, Aldrich), anthracene (99%, Acros, Belgium), fluoranthene (99%, Aldrich), pyrene (>99% Fluka, Germany), benz(a)anthracene (99%, Aldrich), chrysene (99% Cerilliant, TX, USA) and benzo(a)pyrene (98%, Cerilliant). Ethylacetate (p.a. grade) and methanol (HPLC grade) were used as solvents (Merck, Darmstadt, Germany), and super Q treated Milli-Q water was used for cleaning (Millipore, MA). Unless stated differently, the cells were cultured in RPMI 1640 medium supplemented with 10% v/v foetal calf serum (FCS), penicillin (end concentration 100 U ml 1), streptomycin (end concentration 100 lg ml 1) and L-glutamine (end concentration 2 mM) as previously described (Oostingh et al., 2008). Cell culture medium reagents were obtained from PAA Laboratories (Pasching, Austria). IL-8 -specific antibodies and the IL-8 standard were obtained from BD Biosciences (Schwechat, Austria). The CellTiterBlue test was purchased from Promega (Madison, WI). D-Luciferin was obtained from Sigma–Aldrich Co. (St. Louis, USA, Order Number L9504) and was used at a concentration of 66.7 mg l 1. 2.2. Loading of the silicone O-rings with PAHs O-rings were cleaned and loaded with PAHs as previously described (Smith et al., 2010). In brief, O-rings were first

Table 1 Relevant characteristics of the PAHs used. PAH

MW

Aqueous solubility at 37 °C (lg L 1)a

AhR agonistb

Acenaphthene Fluorene Phenanthrene Anthracene Fluoranthene Pyrene Chrysene Benzo(a)anthracene Benzo(a)pyrene

154.2 166.2 178.2 178.2 202.3 202.3 228.3 228.3 252.3

7315 3662 1617 85.8 381.6 223.9 n.a. 18.6 3.5

Poor Poor Weak Poor Weak Weak Strong Strong Strong

a The aqueous solubility was calculated using the aqueous solubility versus temperature regressions given in Mackay et al. (2006). b Capacity of the PAH to function as an AhR agonist is taken from Machala et al. (2001) and Murahashi et al. (2007).

G.J. Oostingh et al. / Toxicology in Vitro 29 (2015) 345–351

thoroughly cleaned by soaking once overnight in ethylacetate, followed by three overnight washes in methanol and three overnight washes in MilliQ water. Cleaned O-rings were stored in a sealed bottle in Milli-Q water until use. Before the PAH loading step, the O-rings were dried by wiping the surfaces using lint-free tissue. O-rings were loaded by partitioning at 21 °C from methanol solutions of the individual PAHs for a minimum period of 72 h. For testing the PAHs at aqueous solubility, O-rings were loaded to saturation using methanol suspensions of the respective PAHs. The PAH crystals served to keep the methanol saturated, which in turn saturated the silicone by equilibrium partitioning. During the in vitro assays, the saturated O-rings provided exposure at the solubility level of the medium, which is characterised by freely dissolved concentrations (Cfree) that are at aqueous solubility. For acenaphthene, phenanthrene and benzo(a)anthracene, the O-rings were additionally loaded using dilutions of their respective saturated methanol solutions in order to achieve loading levels following a geometric concentration series starting at saturation S and with a spacing of factor 2, i.e., S, S/2, S/4, S/8 and S/16 and S/32. The volume of loading solution was sufficient to avoid depletion due to compound partitioning into the silicone O-rings. During the in vitro assays, these O-rings provided well defined exposure at the solubility level of the medium and factor 2 dilutions of this, which means Cfree from S down to S/32. The aqueous solubilities of the PAHs at 37 °C are given in Table 1. Exposure levels in these tests were expressed as freely dissolved concentration, which are considered the effective concentration for cell-uptake and thus for the eventual response (Heringa et al., 2004). Exposure levels were additionally expressed as a percentage of the solubility limit, which can be easily converted into the corresponding dissolved concentrations using the aqueous solubilities given in Table 1. Furthermore, in some cases this can additionally serve as a useful conceptual and thermodynamic reference for exposure and toxicity (Schmidt and Mayer, 2015). After loading, the O-rings were removed from the loading solution, and the surfaces were thoroughly wiped using lint-free tissue to remove any adhering suspension or solution. Finally, the O-rings were rinsed three times for 1 h each with a small volume of Milli-Q water to remove any residual methanol. As a control, O-rings were ‘‘loaded’’ using pure methanol without any PAH.

2.3. Cell culture and exposure A549 cells, a human bronchial epithelial cell line, were stably transfected with the promoter elements of different cytokines or with 4 copies of the NF-jB binding sequence attached to luciferase. The establishment of these cell lines as well as their response to different biological stimuli has been previously described (Oostingh et al., 2008). For the current study, the promoter (p)IL8, pIL-6, and pTNF-a A549 cell lines established in our laboratory and the binding sequence (bs)NF-jB A549 cell line from Panomics (Fremont, USA) were used. The pIL-8, pIL-6, and pTNF-a cell lines were cultured in the presence of the selection antibiotic G418 (0.5 mg/ml final concentration). The bsNF-jB A549 cell line was cultured as described by the distributor. A549 cells were plated out on day one in tissue culture treated 24-well plates at a density of 5  103 cells/well and left overnight to adhere and obtain their normal morphology (day 2). On day 2, all the medium was removed and 1 ml fresh medium with or without 20 ng/ml rhTNF-a was added to each well. Thereafter, the O-rings were added to the cells in culture using sterile single wrapped plastic one-way tweezers. Initially, 9 different PAHs were tested at aqueous solubility (see Table 1), and subsequently the responses to acenaphthene, phenanthrene and benzo(a)anthracene were tested at different freely dissolved concentrations. All exposure

347

treatments were prepared in triplicate, and all experiments were repeated at least twice. 2.4. Luciferase-test and analysis of the cell viability Cytokine promoter induction were analysed after 24 and 48 h and NF-jB binding sequence activation after 6 and 24 h. After an incubation period of 24–48 h depending on the assay, two separate tests were performed on the different PAH-exposed cell cultures; a luciferase assay to determine promoter induction and a cell viability test to determine whether the specific PAH treatment was cytotoxic. For the luciferase assay, the supernatant was removed from all wells and 50 ll passive lysis buffer (Promega, Madison, USA) was added to each well according to the manufacturers’ instructions. After a 10-min incubation time, the lyzed cells (total 50 ll) were transferred to white 96-well microtiter plates. The luciferase signal was detected using a Tecan luminometer (Tecan, Grödig, Austria), whereby a freshly prepared luciferin substrate was added to the cell lysate and the luminescence was determined immediately thereafter. Cell viability was analysed using the CellTiterBlueÒ cell viability assay from Promega (Madison, USA) according to the manufacturers’ instructions. The CellTiterBlueÒ assay relies on the conversion of rezasurin to resorufin by viable cells. For this analysis, separate plates were used that were treated exactly the same as those used for the luciferase assay. In brief, the reagent was directly added to the 96-well plates that were compatible with fluorescent plate reader assays, with the wells containing the A549 cells in culture after the appropriate incubation time with the different PAHs. Thereafter, the plate was incubated for 1 h at 37 °C and 5% CO2, and the fluorescence determined at an excitation wavelength of 560 nm and an emission wavelength of 590 nm using a Tecan plate reader. 2.5. Statistical analysis The data were plotted as the mean ± the standard error of the mean using GraphPad Prism version 6.03 for Windows (GraphPad Software, La Jolla California USA, www.graphpad.com). GraphPad Prism was also used to perform a one-way ANOVA followed by a Dunnett’s multiple comparisons test to specifically compare the various treatments (i.e., different PAHs in Fig. 1 and different concentrations of the individual PAHs in Figs. 2 and 3) to the control. 3. Results 3.1. Cell viability Results of the CellTiterBlueÒ assay showed that A549 exposure to the nine different PAHs at aqueous solubility did not induce any loss in cell viability during the complete incubation time of 48 h (see Fig. 1). Similarly, for the concentration–response testing of acenaphthene, phenanthrene and benzo(a)anthracene, there was no loss in cell viability even at the highest concentration tested (i.e., aqueous solubility). Therefore, it can be excluded that any of the observed effects on the promoter level are due to cell death or decreased cell viability. 3.2. Cytokine promoter induction and NF-jB binding sequence activation upon exposure to 9 different PAHs at aqueous solubility Here, silicone O-rings were loaded to saturation with 9 different PAHs and then added to plated cells in 24-well cell culture trays using sterile tweezers with exposures taking place for a maximum time of 48 h. As a control, O-rings were ‘‘loaded’’ using pure methanol without any PAH. After addition of the O-rings, PAH

348

G.J. Oostingh et al. / Toxicology in Vitro 29 (2015) 345–351

Fig. 1. Immunomodulatory effect of 9 different PAHs on 4 different stably transfected cell lines. A549 cells transfected with the IL-8 promoter (A), IL-6 promoter (B), TNF-a promoter (C) or NF-jB binding sequence (D) were exposed to 9 different PAHs at 100% saturation. Luciferase production was analysed in the cells after 48 h (A–C) or after 24 h (D). In addition, cytotoxicity after 48 h was measured (E). Data are shown as mean (±SEM) and ⁄ indicates a statistically significant difference (p < 0.05) from the control value (empty ring).

Fig. 2. Effects on IL-8 promoter induction of three PAHs at varying percentages of saturation or dissolved concentrations. The effects of acenaphthene, phenanthrene and benzo(a)anthracene on the IL-8 promoter induction in stably transfected A549 cell lines are shown. The dissolved concentrations were calculated using the values provided in Table 1. Data are presented as mean (±SEM) and the arrows indicate a statistically significant difference (p < 0.05) from the control value (empty ring).

Fig. 3. Effects on NF-jB binding sequence activation of three PAHs at varying percentages of saturation or dissolved concentrations. The effects of acenaphthene, phenanthrene and benzo(a)anthracene on NF-jB binding sequence activation in stably transfected A549 cell lines are shown. The dissolved concentrations were calculated using the values provided in Table 1. Data are presented as mean (±SEM) and the arrows indicate a statistically significant difference (p < 0.05) from the control value (empty ring).

G.J. Oostingh et al. / Toxicology in Vitro 29 (2015) 345–351

concentrations in the medium increase until a partitioning equilibrium between the silicone and medium are reached after which the dissolved concentrations remain stable. Smith et al. (2010) analytically determined the O-ring release of the PAHs into the same culture medium as used in this study. A partitioning equilibrium was attained within 3 h or less, with the exception of benzo(a)pyrene which required just over 5 h. Cytokine promoter induction and NF-jB binding sequence activation were analysed after 24 and 48 h or 6 and 24 h, respectively. Therefore, for incubation times of 24 h or more it can be certain that the PAHs had reached equilibrium dissolved concentrations within 5 h or less and that these were subsequently maintained for the duration. The exception was the 6 h measurements of NFjB binding sequence activation where equilibrium was only reached towards the end of the incubation period. In addition, although similar responses were found at both time points the effects were more pronounced at the later time points (i.e., 24 or 48 h depending on the assay), and these were therefore chosen for the below discussion. As has been previously reported (Oostingh et al., 2008), addition of rhTNF-a consistently resulted in a marked increase in reporter gene induction for all 4 cell lines, as evidenced by the increased luciferase signals in the rhTNF-a treated samples when compared to the samples without rhTNF-a (data not shown). Therefore, to compare the relative changes in cell activation specifically attributable to the PAHs, the data were normalised to the respective controls with or without rhTNF-a as shown in Fig. 1. Some differences due to added rhTNF-a were observed in the relative change in induction or activation compared to the respective blank O-ring controls. Even though rhTNF-a did strongly increase the reporter induction in all cases, as soon as the blank O-ring values of treated and untreated cells were set to 1, the differences between rhTNF-a treated and untreated cells was relatively minor and did not influence the direction of change (i.e., induction versus suppression). Thus, it appears that irrespective of any additional induction due to the exogenously added rhTNF-a, the PAHs always acted in a consistent manner to either increase or decrease the induction of reporter gene expression. When presented at their respective aqueous solubilities, the different PAHs affected A549 cells to different degrees (Fig. 1). For example, the results of the IL-8-promoter transfected cell line show that incubation with chrysene, benzo(a)anthracene or benzo(a)pyrene resulted in a statistically significant increased cytokine promoter activity by a factor of around two. Anthracene and fluoranthene resulted in a more modest increase in IL-8 promoter induction, whereas fluorene, phenanthrene and pyrene did not have an appreciable effect. In contrast, acenaphthene resulted in a down-regulation of the IL-8 promoter induction (although this was not statistically significant). The same trends between the different PAHs were also consistently observed for IL-6 and TNF-a promoter induction, as well as for NF-jB binding sequence activation (Fig. 1). In summary, the immunomodulatory responses towards PAHs present at their aqueous solubilities can be divided into three distinct groups: (1) acenaphthene which resulted in a low, but consistent, down-regulation of the promoter induction; (2) fluorene, phenanthrene, anthracene, fluoranthene and pyrene which all showed either no effect or only a moderate increase in promoter induction and (3) chrysene, benzo(a)anthracene and benzo(a)pyrene which all showed an increase in the promoter induction which in the main were also statistically significant. Therefore, particularly these latter 3 PAHs might be regarded as potential activators of the immune system and might therefore aggravate any on-going immune responses in human lung epithelial cells.

349

3.3. IL-8 promoter induction and NF-jB binding sequence activation upon exposure to different freely dissolved concentrations of acenaphthene, phenanthrene and benzo(a)anthracene Aqueous solubility is PAH specific, ranging from 3.5 lg L 1 for benzo(a)pyrene to 7315 lg L 1 for acenaphthene at 37 °C. Therefore, part of the explanation for the above differences in responses could be attributable to the different Cfree exposure levels in the cell culture medium, resulting in different positions on the underlying concentration–response curve. Note, however, that these differences in response are likely not due to cytotoxic effects since cell viability was not impaired by any of the PAHs at their respective aqueous solubilities (see Fig. 1). To investigate this further, three PAHs were selected for investigating the immunomodulatory responses as a function of the Cfree in the medium. The three PAHs were selected based on the contrasting responses observed in Fig. 1. Acenaphthene consistently resulted in a reduced induction of the promoter or binding sequence activation, phenanthrene had no appreciable effect and benzo(a)anthracene exposure resulted in an increased induction. Figs. 2 and 3 respectively show the relative changes in IL-8promoter induction and NF-jB binding sequence activation in response to different Cfree levels of acenaphthene, phenanthrene and benzo(a)anthracene. Similar to the situation observed in Fig. 1, addition of rhTNF-a resulted in increased induction/activation, but the relative change in response attributable to the different PAHs was similar with or without rhTNF-a (Figs. 2 and 3). Exceptions were phenanthrene and benzo(a)anthracene, where higher pIL8 induction was observed in the presence of rhTNF-a. Acenaphthene resulted in a concentration-dependent suppression in IL-8 promoter induction or NF-jB binding sequence activation with increasing Cfree levels (Figs. 2 and 3). This could theoretically result from two different events. Acenaphthene might still induce an immunomodulatory response in the cells when applied at lower concentrations, but a loss in cell viability at higher concentrations might be reflected in a reduction of the immunomodulatory response. However, this is relatively unlikely since a reduction in the cell viability did not occur even after 48 h exposure to acenaphthene at aqueous solubility (or was below the detection limit of the CellTiterBlueÒ test). A second option is that acenaphthene acts differently on the A549 cells compared to all other PAHs and directly inhibits immune responses, albeit to a low degree. Phenanthrene induced IL-8 promoter induction (Fig. 2) and NF-jB binding sequence activation (Fig. 3) at lower concentrations, but this effect was reduced at higher concentrations. Similarly, exposure to increasing Cfree levels of benzo(a)anthracene initially led to a strong increase in IL-8 promoter induction and NF-jB binding sequence activation, but this was reduced at the highest concentrations. This data therefore confirms that the PAHs affect the A549 cells to different degrees. Furthermore, the results showed that all responses were reduced at higher concentrations, indicating that the responsiveness of the cells is reduced, even when a decrease in cell viability could not be detected using a classical cell viability assay. This illustrates the importance of performing concentration–response testing in immunomodulation studies, both with regards to assessing and understanding the effect of a single chemical and even more so when comparing responses to different chemicals.

4. Discussion The results of the current study show the immunomodulatory responses of human lung epithelial cells exposed to specific PAHs, and indicate that different mechanisms are involved for the various

350

G.J. Oostingh et al. / Toxicology in Vitro 29 (2015) 345–351

PAHs tested. The reason for the different responses might in part be explained by interactions of the PAHs with a range of different pathways that are involved in PAH-toxicity. For example, a study in which human peripheral blood lymphocytes were exposed to different PAHs has shown that the responses differed between PAHs and that different pathways are involved, including induction via Ah-receptor binding, alterations in P450 metabolism and interference with signal transduction mechanisms (Davila et al., 1996). The agonistic effects of different PAHs on the AhR differ greatly; for example chrysene, benzo(a)pyrene and benzo(a)anthracene are strong agonists for AhR, whereas fluorene, acenaphthene and anthracene do not have any agonistic effects (Machala et al., 2001; Murahashi et al., 2007). Interestingly, those PAHs that induced a strongly increased immunomodulatory response are exactly those with a strong agonistic AhR effect, i.e., chrysene, benzo(a)pyrene and benzo(a)anthracene (Table 1). Therefore, binding to the AhR might result in activation of the inflammatory immune response. In Pei et al. (2002), exposure of the human alveolar cell line A549 to benzo(a)pyrene resulted in the activation of the NFjB pathway, which led to an increased IL-8 expression similar to that observed here. However, the difference between the immunomodulatory effects of PAHs with and without AhR agonistic effects on human lung cells has not been studied in detail, particularly not by using human material. This latter aspect is of main importance, since it has been shown that the immune response towards PAHs in mice differs from those observed in humans (Davila et al., 1996). Currently, the exact role of the AhR on immune responses in general is not known. AhR agonists act in some cases as inhibitors of the immune system, for example in relation to allergic responses. In contrast, diseases such as chronic obstructive pulmonary diseases are increased after exposure to AhR agonists, indicating a stimulation of the pro-inflammatory pathway (Beamer and Shepherd, 2013). The exact mechanisms behind the role of the AhR in immune responses, and the potential of AhR agonists to inhibit these responses need to be studied in further detail. When comparing the immunomodulation of different compounds, it is important to include concentration–response testing. This enables a better differentiation between specific responses, such as the immunomodulatory responses detected here, and unspecific responses, such as a decrease in cell viability or a general reduction in the cell metabolism. The study of PAHs in cell culture systems is technically challenging because of their hydrophobicity. Recently, passive dosing was applied to expose human cells in culture to PAHs at aqueous solubility (Smith et al., 2010). As was done in this study, silicone O-rings were loaded to saturation and added to the cells in a 24well format. The PAHs were rapidly released into the cell culture medium and the concentrations stayed constant for at least 72 h after adding the O-rings, resulting in highly reproducible concentrations of PAHs in the cell culture medium. Similar systems have also been applied in several different, mostly non-human, in vitro assays (Booij et al., 2011; Bougeard et al., 2011; Kramer et al., 2010; Smith et al., 2010, 2013). It is difficult to compare the results from this study where the dissolved concentrations are defined by passive dosing with those from spiking studies where only the initial nominal concentrations are defined. The differences between the spiked nominal concentrations and the corresponding dissolved concentrations due to sorption to the medium and well plate are generally not known. Furthermore, concentrations decrease after spiking, whereas with passive dosing these are buffered and remain constant. This means a direct comparison between the spiked nominal and passive dosing dissolved concentrations is incorrect since these both represent different exposure concentrations. This is particularly problematic for those compounds that show marked trends in their concentration–response curves.

Despite the important advantages, the applied passive dosing procedure also had some limitations. The time and work related to the preparation of the passive dosing polymer was significant. In the present study we applied commercially available silicone O-rings to avoid the initial steps of casting or coating the polymer, and this already saved a considerable amount of time. However, the loading and cleaning of the O-rings still constituted a significant part of the experimental work, and further development is needed in order to simplify the cleaning and loading procedures. Furthermore, it was more practical to use O-rings that fitted in 24-well plates not the more commonly used 96-well plates. Passive dosing formats that are better suited for 96-well or even 384-well plates are thus still needed. Nevertheless, passive dosing is currently the only available method that can provide a defined and constant exposure when direct chemical analysis cannot be performed, and thus will continue to play an important role in reliably testing the immunomodulation of such hydrophobic chemicals. Particularly interesting in this regard would be testing complex mixtures of such chemicals, these being more representative real-life exposure scenarios. In summary, the results of this study indicate that at least some of the PAHs (e.g., chrysene, benzo(a)anthracene and benzo(a)pyrene) have the very real potential to initiate or aggravate an inflammatory immune response. Furthermore, the concentration– response testing indicates that this effect might in fact be more pronounced at the lower exposure concentrations that are likely more representative of real-life exposure.

Conflict of Interest The authors declare that there are no conflicts of interest.

Transparency Document The Transparency document associated with this article can be found in the online version.

Acknowledgements We acknowledge Margit Fernqvist (Aarhus University, Denmark) for her skilled passive dosing work. In addition, we acknowledge Albert Duschl (University of Salzburg, Austria) for providing access to his laboratory facilities to perform these experiments. The applied passive dosing approach has been developed within the EU Integrated Project OSIRIS (Strategies for Risk Assessment of Industrial Chemicals through Integration of Non-Test and Test Information, COGE-037017, http://www. osirisreach.eu). References Armstrong, B., Hutchinson, E., Unwin, J., Fletcher, T., 2004. Lung cancer risk after exposure to polycyclic aromatic hydrocarbons: a review and meta-analysis. Environ. Health Perspect. 112, 970–978. Beamer, C.A., Shepherd, D.M., 2013. Role of the aryl hydrocarbon receptor (AhR) in lung inflammation. Semin. Immunopathol. 35, 693–704. Booij, P., Lamoree, M.H., Leonards, P.E.G., Cenijn, P.H., Klamer, H.J.C., Van Vliet, L.A., Akerman, J., Legler, J., 2011. Development of a polydimethylsiloxane film-based passive dosing method in the in vitro DR-CALUXÒ assay. Environ. Toxicol. Chem. 30, 898–904. Bougeard, C., Gallampois, C., Brack, W., 2011. Passive dosing: an approach to control mutagen exposure in the Ames fluctuation test. Chemosphere 83, 409–414. Davila, D.R., Romero, D.L., Burchiel, S.W., 1996. Human T cells are highly sensitive to suppression of mitogenesis by polycyclic aromatic hydrocarbons and this effect is differentially reversed by alpha-naphthoflavone. Toxicol. Appl. Pharmacol. 139, 333–341.

G.J. Oostingh et al. / Toxicology in Vitro 29 (2015) 345–351 Gülden, M., Seibert, H., 2005. Impact of bioavailability on the correlation between in vitro cytotoxic and in vivo acute fish toxic concentrations of chemicals. Aquat. Toxicol. 72, 327–337. Heringa, M.B., Schreurs, R., Busser, F., Van der Saag, P.T., Van der Burg, B., Hermens, J.L.M., 2004. Toward more useful in vitro toxicity data with measured free concentrations. Environ. Sci. Technol. 38, 6263–6270. Herzog, E., Byrne, H.J., Casey, A., Davoren, M., Lenz, A.G., Maier, K.L., Duschl, A., Oostingh, G.J., 2009. SWCNT suppress inflammatory mediator responses in human lung epithelium in vitro. Toxicol. Appl. Pharmacol. 234, 378–390. Karakaya, A., Yucesoy, B., Turhan, A., Erdem, O., Burgaz, S., Karakaya, A.E., 1999. Investigation of some immunological functions in a group of asphalt workers exposed to polycyclic aromatic hydrocarbons. Toxicology 135, 43–47. Karakaya, A., Ates, I., Yucesoy, B., 2004. Effects of occupational polycyclic aromatic hydrocarbon exposure on T-lymphocyte functions and natural killer cell activity in asphalt and coke oven workers. Hum. Exp. Toxicol. 23, 317–322. Kramer, N.I., Busser, F.J.M., Oosterwijk, M.T.T., Schirmer, K., Escher, B.I., Hermens, J.L.M., 2010. Development of a partition-controlled dosing system for cell assays. Chem. Res. Toxicol. 23, 1806–1814. Kwon, J.H., Wuethrich, T., Mayer, P., Escher, B.I., 2009. Development of a dynamic delivery method for in vitro bioassays. Chemosphere 76, 83–90. Lambrecht, B.N., Hammad, H., 2008. Dendritic cells and epithelial cells: linking innate and adaptive immunity in asthma. Nat. Rev., Immunol. 8, 193–204. Lambrecht, B.N., Hammad, H., 2010. The role of dendritic and epithelial cells as master regulators of allergic airway inflammation. Lancet 376, 835–843. Machala, M., Vondracek, J., Blaha, L., Ciganek, M., Neca, J., 2001. Aryl hydrocarbon receptor-mediated activity of mutagenic polycyclic aromatic hydrocarbons determined using in vitro reporter gene assay. Mutat. Res. 497, 49–62. Mackay, D., Shiu, W.-Y., Ma, K.-C., Lee, S.C., 2006. Handbook of Physical-Chemical Properties and Environmental Fate for Organic Chemicals, second ed. CRC Press, Taylor & Francis Group, Boca Raton, FL. Min, L.F., He, S.Y., Chen, Q.O., Peng, F., Peng, H.B., Xie, M.X., 2011. Comparative proteomic analysis of cellular response of human airway epithelial cells (A549) to benzo(a)pyrene. Toxicol. Mech. Methods 21, 374–382. Mollerup, S., Ovrebo, S., Haugen, A., 2001. Lung carcinogenesis: resveratrol modulates the expression of genes involved in the metabolism of PAH in human bronchial epithelial cells. Int. J. Cancer 92, 18–25. Murahashi, T., Watanabe, T., Kanayama, M., Kubo, T., Hirayama, T., 2007. Human aryl hydrocarbon receptor ligand activity of 31 non-substituted polycyclic aromatic hydrocarbons as soil contaminants. J. Health Sci. 53, 715–721. Oostingh, G.J., Schmittner, M., Ehart, A.K., Tischler, U., Duschl, A., 2008. A highthroughput screening method based on stably transformed human cells was used to determine the immunotoxic effects of fluoranthene and other PAHs. Toxicol. In Vitro 22, 1301–1310.

351

Pei, X.H., Nakanishi, Y., Inoue, H., Takayama, K., Bai, F., Hara, N., 2002. Polycyclic aromatic hydrocarbons induce IL-8 expression through nuclear factor kappa b activation in A549 cell line. Cytokine 19, 236–241. Petry, T., Schmid, P., Schlatter, C., 1996. The use of toxic equivalency factors in assessing occupational and environmental health risk associated with exposure to airborne mixtures of polycyclic aromatic hydrocarbons (PAHs). Chemosphere 32, 639–648. Pfaller, T., Colognato, R., Nelissen, I., Favilli, F., Casals, E., Ooms, D., Leppens, H., Ponti, J., Stritzinger, R., Puntes, V., Boraschi, D., Duschl, A., Oostingh, G.J., 2010. The suitability of different cellular in vitro immunotoxicity and genotoxicity methods for the analysis of nanoparticle-induced events. Nanotoxicology 4, 52–72. Riedl, J., Altenburger, R., 2007. Physicochemical substance properties as indicators for unreliable exposure in microplate-based bioassays. Chemosphere 67, 2210– 2220. Röder-Stolinski, C., Fischäder, G., Oostingh, G.J., Eder, K., Duschl, A., Lehmann, I., 2008. Chlorobenzene induces the NF-kappa B and p38 MAP kinase pathways in lung epithelial cells. Inhal. Toxicol. 20, 813–820. Schmidt, S.N., Mayer, P., 2015. Linking algal growth inhibition to chemical activity: baseline toxicity required 1% of saturation. Chemosphere 120, 305–308. Schreiber, R., Altenburger, R., Paschke, A., Kuster, E., 2008. How to deal with lipophilic and volatile organic substances in microtiter plate assays. Environ. Toxicol. Chem. 27, 1676–1682. Smith, K.E.C., Oostingh, G.J., Mayer, P., 2010. Passive dosing for producing defined and constant exposure of hydrophobic organic compounds during in vitro toxicity tests. Chem. Res. Toxicol. 23, 55–65. Smith, K.E.C., Heringa, M., Uytewaal, M., Mayer, P., 2013. The dosing determines mutagenicity of hydrophobic compounds in the Ames II assay with metabolic transformation: passive dosing versus solvent spiking. Mutat. Res. 750, 12–18. Sorensen, M., Autrup, H., Moller, P., Hertel, O., Jensen, S.S., Vinzents, P., Knudsen, L.E., Loft, S., 2003. Linking exposure to environmental pollutants with biological effects. Mutat. Res. 544, 255–271. Srogi, K., 2007. Monitoring of environmental exposure to polycyclic aromatic hydrocarbons: a review. Environ. Chem. Lett. 5, 169–195. Tanneberger, K., Rico-Rico, A., Kramer, N.I., Busser, F.J.M., Hermens, J.L.M., Schirmer, K., 2010. Effects of solvents and dosing procedure on chemical toxicity in cellbased in vitro assays. Environ. Sci. Technol. 44, 4775–4781. Zhao, J., Wang, Z., Mashayekhi, H., Mayer, P., Chefetz, B., Xing, B., 2012. Pulmonary surfactant suppressed phenanthrene adsorption on carbon nanotubes through solubilization and competition as examined by a passive dosing technique. Environ. Sci. Technol. 46, 5369–5377.