Usefulness of Toxicological Validation of Vocs Catalytic Degradation by Air- Liquid Interface Exposure System Crossmark

Environmental Research 152 (2017) 328–335

Contents lists available at ScienceDirect

Environmental Research

journal homepage: www.elsevier.com/locate/envres

Usefulness of toxicological validation of VOCs catalytic degradation by air- liquid interface exposure system crossmark

Margueritta Al Zallouha, Yann Landkocz, Julien Brunet, Renaud Cousin, Eric Genty, ⁎ Dominique Courcot, Stéphane Siﬀert, Pirouz Shirali, Sylvain Billet

Unité de Chimie Environnementale et Interactions sur le Vivant EA4492, Université du Littoral Côte d′Opale, 189 A Avenue Maurice Schumann, 59140 Dunkerque, France

ARTICLE INFO ABSTRACT

Keywords: Toluene is one of the most used Volatile Organic Compounds (VOCs) in the industry despite its major health Catalytic oxidation impacts. Catalytic oxidation represents an efficient remediation technique in order to reduce its emission Air-liquid interface exposure directly at the source, but it can release by-products. To complete the classical performance assessment using Toluene dedicated analytical chemistry methods, we propose to perform an untargeted toxicological validation on two Toxicological validation efficient catalysts. Using biological system allows integrating synergy and antagonism in toxic effects of emitted By-products identification VOCs and by-products, often described in case of multi-exposure condition. Catalysts Pd/α-Al2O3 and Pd/γ- ® Al2O3 developed for the oxidation of toluene were both coupled to a Vitrocell Air-Liquid Interface (ALI) system, for exposure of human A549 lung cells during 1 h to toluene or to catalysts exhaust before quantification of xenobiotics metabolizing enzymes. This study validated initially the Vitrocell® as an innovative, direct and dynamic model of ALI exposure in the assessment of the performances of new catalysts, showing the presence of chemically undetected by-products. The comparison of the two catalysts showed then that fewer organic compounds metabolizing genes were

induced by Pd/γ-Al2O3 in comparison to Pd/α-Al2O3, suggesting that Pd/γ-Al2O3 is more eﬃcient for toluene total oxidation from a toxicological point of view.

1. Introduction where 80% of the absorbed dose undergoes a chain of oxidation reactions by the intervention of different enzymes which make the Volatile Organic Compounds (VOCs) represent a variety of sub- molecule more hydrophilic and allow urine excretion of metabolites. stances belonging to different chemical families (aromatic hydrocar- Between 10% and 20% of inhaled toluene are excreted in the expired bons, ketones, alcohols, alkanes, aldehydes, etc.) and known for being air with a half-life of about 25 min (Benoit et al., 1985). Toluene is important contributors to air pollution (Khan and Ghoshal, 2000). known for both acute and chronic toxic effects (Tormoehlen et al., Among the major VOCs, Benzene, Toluene, Ethylbenzene and Xylenes 2014). It is classified as a Carcinogenic, Mutagenic and Reprotoxic (BTEX) have major and direct impact on human health. Toluene is (CMR) category 3 due to its toxicity for reproduction. widely used in many industrial sectors where 32% of commercial Because of toluene toxicity, it is necessary to reduce emissions toluene enter in the process of benzene synthesis and 19% is used as directly at source. When substitution is not possible, catalytic oxidation solvents with almost 90 kt/year as paint solvents in European Union represents an economical and environmental alternative to the thermal (Hansen et al., 2002). Toluene has a relatively well known toxicity with oxidation of VOCs. The choice of suitable catalysts for the total the respiratory tract as major route of absorption. Limit values for oxidation of toluene should usually take into account the activity, occupational exposure to toluene were established in France, in 2012, stability, cost and feasibility of large scaling-up of laboratory systems. with an exposure limit value of 20 ppm equivalent to 76.8 mg/m3 for A catalyst is generally formed by a support and an active surface. 3 8 h and 100 ppm (384 mg/m ) on a short-term (under 15 min). The Aluminium oxides, especially α-Al2O3 and γ-Al2O3, present good same values were set for the 8 h exposure in the USA since 2007 (INRS, performances as supports for active phases for BTEX oxidation. 2012). Aluminas have some interesting properties like a high thermal stability

After absorption, toluene is primarily metabolized in the liver, for α-Al2O3 or a high speciﬁc surface area with acidic properties for γ-

⁎ Corresponding author. E-mail address: [email protected] (S. Billet). http://dx.doi.org/10.1016/j.envres.2016.10.027 Received 22 July 2016; Received in revised form 24 October 2016; Accepted 27 October 2016 0013-9351/ © 2016 Published by Elsevier Inc. M. Al Zallouha et al. Environmental Research 152 (2017) 328–335

Al2O3 (Burgos et al., 2002; Kim and Shim, 2009; Ordóñez et al., 2002; 2.2.1. Palladium impregnation of support Papaefthimiou et al., 1997). In order to increase the activity of these Palladium was deposed on α-Al2O3 by an aqueous impregnation oxides, palladium and platinum can be used as actives phases since method with a content of 0.5 wt%. Alpha alumina was suspended in an they showed very good performances although palladium constitutes a appropriate volume of a palladium nitrate solution (0.25 g/L). The good compromise between performance and cost (Papaefthimiou et al., suspension was maintained for 18 h at 60°C. Then, water was removed 1998; Rusu and Dumitriu, 2003). by rotary evaporator. The solid was dried one night at 100°C before to Total oxidation of VOCs should produce only water and carbon be calcined 4 h at 400°C (1°C/min) under air flow (2 L/h) (Brunet dioxide. However, during operation, the catalytic process can lead to et al., 2015). the release of by-products, sometimes more toxic than the original VOCs. In order to ensure an evaluation procedure of catalyst perfor- 2.2.2. Characterization of catalysts mances from an environmental and health-safety approach, here we Specific surface areas (SBET) were measured by the Brunauer- propose to perform an untargeted toxicological validation of two Emmet-Teller method with a “ThermoElectron Qsurf series Surface catalysts preliminary selected on chemical criteria. The validation we Area Analyzer” apparatus. Prior to measurement, calcined catalyst propose corresponds to the evaluation the response of biological samples were heated at 130°C for 1 h under a helium flow. Adsorption system exposed to catalytic exhaust. This method integrates potential was made with a N2/He mixture (30/70) at -196°C. Desorption of synergy and antagonism existing in toxic effects of emitted VOCs and gaseous N2 was quantized with a thermal conductivity detector. by-products, often described in case of multi-exposure condition. Elementary analysis was performed with an ICP-OES (Thermo, model In case of gaseous or very volatile compounds, exposure of in vitro ICAP 6300 Duo) after acidic dissolution of 50 mg of catalyst sample. systems becomes a challenge. According to Rasmussen, in vitro methods of exposure to gaseous mixtures have to follow some funda- 2.2.3. Catalytic test mental requirements: (i) the contact between the cells and the test Total oxidation of toluene was studied in a conventional fixed bed compounds should be as close as possible to avoid interactions, and to reactor loaded with 100 mg of catalyst. A Toluene/Air mixture was realize direct contact of cells and gas phase components and (ii) generated with a saturator CAL-PC-5 from Calibrage in order to obtain methods have to be developed to maintain a humidified atmosphere 1000 ppm in a flow of 100 mL/min. The tests were made between 50 in order to avoid cells drying (Rasmussen, 1984). Recent methodolo- and 400°C with a temperature ramp of 1.5°C/min. The catalysts were gical and technical breakthroughs of in vitro methods have the pre-treated 2 h at 200°C (1°C/min) under air flow (2 L/h−1) and then, potential to fulfill these essential requirements. This pattern of cell for impregnated catalysts, were reduced 2 h at 200°C (1°C/min) under exposure was initially performed culturing cells on collagen-coated H2 (5.0) flow (2 L/h). The organic compounds were analysed by gas membranes located on special platforms (Chen et al., 1993) and more chromatography with a CP-4900 microGC (Agilent) coupled to a recently on porous membranes located in Transwell® inserts Pfeiffer-Vacuum Omnistar Quadrupole Mass Spectrometer (QMS- ® (Aufderheide et al., 2003). Cells culture exposed on Transwell inserts 200). CO and CO2 were analysed with an infrared analyzer (ADEV indeed provides a perfect system to study the cellular responses caused 4410 IR). by very close exposure to airborne chemical, avoiding the possible Toluene conversion was calculated considering products, by-pro- interference of culture medium. The dynamic direct exposure of human ducts and carbon number for each compound: cells to chemicals in the atmosphere can be achieved using the ⎛ ⎞ ® [CO2T ] + [CO] T + [C 6 H 6T ] *6 Vitrocell model. This system provides the ability to grow cells on X=T ⎜ ⎟*100 ⎝ [CO ] + [CO] + [C H ] *6 + [C H ] *7 ⎠ permeable membranes in the presence of culture medium and to 2T T 6 6T 7 8T expose them directly to a continuous and dynamic gas flow at the Air- Where: Liquid Interface (ALI) (Aufderheide and Mohr, 1999). Target cells can thus be continuously exposed to airborne chemicals on their apical – XT was the toluene conversion at the T temperature (%); side, while being nourished and hydrated with culture medium from – [I]T was the concentration of the compounds I at the T temperature their basolateral side. (ppm); In this study, two catalysts Pd/α-Al2O3 and Pd/γ-Al2O3 were both chemically and toxicologically tested in order to determine the most Catalytic intrinsic activity was calculated at a toluene conversion of efficient one for toluene oxidation. Thereby human lung cells were 20% and considering a plug-reactor: exposed at the ALI for 1 h to a 1000 ppm stream of toluene and to gas Q 273.15 [C780 H ] X 1 mixtures derived from its oxidation. A549 cells were selected since type Ai = . . . . V T 106 m S II pneumocytes are the main cells in contact with airborne pollutants in M20 BET the lung. Following the exposure, toxicological tests were conducted to Where: validate the best performing catalyst for toluene remediation. 2 1 – Ai was the catalytic intrinsic activity (mol/m h ); – Q was the volume flow (L/h−1); 2. Materials and methods – VM was the molar volume (L/mol); – T was the catalyst temperature for X% toluene conversion (K); 2.1. Chemicals X – [C7H8]0 was the toluene initial concentration (ppm); – X was the toluene conversion (%); Toluene C H CH is the VOC chosen as a model for the oxidation of 6 5 3 – α m was the catalyst mass (g); BTEX. This monocyclic aromatic compound and the catalyst support - 2 – SBET was the specific surface area of the catalyst (m /g). Al2O3 were provided by Acros Organics. In a previous study, in order to compare the catalysts and taken into 2.2. Catalysts account the by-products formed, several parameters have been char- acterized to screen efficient catalysts (Brunet et al., 2015): Two catalysts were studied for toluene total oxidation. The first was γ a commercial formulation Pd/ -Al2O3 (Acros Organics, 0.5 wt%Pd). – Ti: temperature at which the by-product appears; α The second was a Pd/ -Al2O3 (0.5 wt%Pd) prepared by a Pd impreg- – Tmax: temperature at which the amount of the by-product is the nation according to the following procedure.

329 M. Al Zallouha et al. Environmental Research 152 (2017) 328–335

largest; exposure in terms of by-products. For example, concerning the Pd/γ-

– Qmax: maximum observed amount of the considered by-product; Al2O3 catalyst, when the catalytic test was stabilized at 200°C, the – Tf: temperature at which the by-product is totally oxidized; amount of benzene produced is quite stable at about 9 ppm. Moreover, – R: range emission of benzene, corresponding to Tf−Ti; the exhausts of this last catalyst were also evaluated at the temperature – P: persistence of benzene, corresponding to Tf−T100; at which no known by-product was formed anymore (i.e. T > Tf). At the – X: toluene conversion value corresponding to the Qmax value. cell level, the temperature reached 37°C. After exposure, cells were rinsed with PBS and collected by trypsination. After centrifugation at 2.3. Cell line and culture conditions 500g, cell pellets and supernatants were frozen until use at -80°C.

The human alveolar type II pneumocytes A549 (ATCC CCL-185) 2.6. Toxicological analysis were used for their known metabolic capacities (Courcot et al., 2012). ® The cells were cultured in sterile plastic flasks (Corning , Fisher 2.6.1. Cytotoxicity Scientific), in a minimum essential medium (MEM) supplemented LDH assay was used as a viability test where cell membrane with 5% (v/v) of foetal bovine serum, 1% (v/v) L-glutamin and 1% (v/v) integrity of A549 cells was evaluated by measurement of extracellular penicillin and streptomycin (Invitrogen-Life Technologies). Cells were released lactate dehydrogenase (LDH) in the culture medium, accord- incubated at 37°C, in a 5% CO2 and 100% humidity atmosphere until ing to the manufacturer's instructions (Roche). The LDH released by 80% of confluency. Then, cells were washed twice with PBS and the cells upon exposure to gaseous mixtures catalyzes the conversion of detached with trypsin + EDTA (Gibco Invitrogen) at 37°C during lactate to pyruvate, causing at the same time the reduction of NAD+ to 3 min. Cells were collected by centrifugation and counted with a NADH/H+. Briefly, 100 µL of each cell-free culture medium was ® Cellometer . Four hours before ALI exposure experiments, 5×105 cells transferred in triplicate in a 96-well plate. Then 100 µL of LDH-assay ® ® were seeded in Corning Transwell inserts (Sigma-Aldrich) with a reaction mixture was added to each well. After 30 min of incubation at 2 surface area of 4.7 cm in MEM supplemented with 1% (v/v) of L- 37°C in the dark, the optical density at 490 nm was measured using a glutamin and 1% (v/v) penicillin and streptomycin. microplate reader. The intensity of the color was proportional to the amount of LDH released, in connection with the cytotoxic disturbances 2.4. Cell exposure device caused by exposure to VOCs.

® The Vitrocell system (Vitrocell GmbH, Germany) is a direct and 2.6.2. Gene expression dynamic system for ALI exposure. This technique allows simulating the TaqMan® Gene expression Cells-to-Ct kit (Ambion, Invitrogen-Life in vivo inhalation exposure to VOCs. The used system was composed of Technologies) was used to extract RNA from collected cells and to three modules 6/3 CF Stainless thermoregulated at 37°C using a water synthesis cDNA by reverse transcription. Real-time quantitative ® bath. Each module consisted of three Transwell inserts (replicates of Polymerase Chain Reaction (qPCR) using TaqMan® Gene Expression the same exposure condition) exposed in parallel to different gaseous Assay (Applied Biosystems) was used to quantify the gene expression of mixtures: (i) gas directly from the catalytic exhaust without dilution, (i) Xenobiotic Metabolizing Enzymes (XMEs): Cytochromes P-450 (ii) gas from the catalytic exhaust with a dilution of 10% in purified air, (CYP) 1A1, CYP1B1, CYP2E1, CYP2B6, CYP2S1, CYP2F1, CYP3A4, (iii) purified air (negative control), (iv) 1000 ppm of toluene, and (v) Epoxyde hydrolase 1 (EPHX1), NADPH quinone oxydoreductase 1 100 ppm of toluene. These two last exposure conditions served as (NQO1); (ii) xenobiotics receptors: Aryl hydrocarbon receptor (AhR), positive controls to confirm cells metabolism. The supernatant of cell Pregnane X receptor (PXR), Constitutive androstane receptor culture was removed just before exposure. A549 cells seeded on the (CAR);and (iii) 18S rRNA, a housekeeping gene, using TaqMan® insert were then directly exposed on their apical side to the gaseous specific primers (Table 1) sets on a 7500 Fast Real-Time PCR System mixtures. (Applied Biosystems). After 40 cycles of amplification, Cycle threshold (Ct) values were determined by Sequence Detection Software v2.0.3 ® 2.5. Coupling catalyst-Vitrocell (Life Technologies) allowing the calculation of the Relative Quantification (RQ) between cells exposed to VOCs mixtures and cells In order to study the toxicological impacts of the by-products exposed to purified air according to the following formulas: RQ=2–ΔΔCt formed during toluene oxidation, A549 cells were exposed to gas where ΔΔCt=ΔCt −ΔCt and ΔCt=Ct −Ct . ® VOC-exposed Air exposed target gene 18S released by the catalytic conversion procedure. The Vitrocell system Obtained results were expressed as median values of triplicates. was coupled with the two VOCs oxidation experimental setups and the ffl α γ e uent gas from the two catalysts Pd/ -Al2O3 and Pd/ -Al2O3 were 2.7. Statistical analyses directed on cells for 1 h of exposure (Fig. 1). Firstly, to compare the exhausts of the two tested catalysts, the working temperature used for Results were expressed as median values and interquartile range. coupling corresponded to the temperature at which toluene total Data from catalytic exhaust were compared to purified air exposed α conversion (T100) was obtained (i.e. 233°C and 200°C for Pd/ -Al2O3 cells. Statistical analyses were performed using the non-parametric γ and Pd/ -Al2O3, respectively). Prior to the exposure of cells, the Mann-Whitney U-test (SPSS for Windows, v18.0). Statistically signifi- catalytic system was stabilized during 8 h in order to have a constant cant differences were reported with p < 0.05.

Air/Liquid Interface Vitrocell® Table 1 ® Toluene/Air Thermally TaqMan Gene Expression Assays used in RT-qPCR. 1000 ppm regulated 100% exhaust 100 mL.min-1 catalyc test Genes Primers references Genes Primers references

10% exhaust CYP1A1 Hs01054797_g1 AhR Hs00169233_m1 CYP1B1 Hs02382916_s1 CAR Hs00901571_m1 CYP2B6 Hs04183483_g1 EPHX1 Hs01116806_m1 Air Control CYP2E1 Hs00559368_m1 NQO1 Hs02512143_s1 CYP2F1 Hs01035018_g1 PXR Hs01114267_m1 CYP2S1 Hs00258076_m1 18S Hs99999901_s1 ® Fig. 1. Experimental setup of the coupling of the catalyst to the ALI Vitrocell exposure CYP3A4 Hs00604506_m1 system.

330 M. Al Zallouha et al. Environmental Research 152 (2017) 328–335

Table 2 results showing no cytotoxicity in all exposure conditions allowed the α γ Compared catalytic parameters for Pd/ -Al2O3 and Pd/ -Al2O3 in the toluene total gene expression determination of exposed cells. The levels of mRNA for oxidation. 9 genes coding for XMEs and 3 genes coding for receptors were obtained by RT-qPCR after 1 h of exposure to (i) toluene, (ii) catalytic Pd/α-Al2O3 Pd/γ-Al2O3 exhaust using Pd/γ-Al2O3, (iii) catalytic exhaust using Pd/α-Al2O3, and 2 −1 SBET (m g ) 1 252 (iv) catalytic exhaust using the most eﬃcient catalyst at a higher Pd (wt%) 0.48 0.40 temperature. The results were represented in the form of relative ﬁ Toluene oxidation quanti cation (RQ). These quantitative results are expressed as the

T50 (°C) 218 179 level of induction of a given gene in cells exposed to 10% or 100% of the T100 (°C) 233 200 catalytic emissions or toluene compared to the same gene expressed in −2 −1 −7 −9 Ai (mol.m h ) 2.84 10 1.27 10 cells exposed to puriﬁed air. The results of RT-qPCR showed that the receptors PXR, CAR and the cytochromes CYP2B6 and CYP3A4 were Benzene emission not expressed in all culture conditions and with the 4 performed Ti (°C) 176 129

Tmax (°C) 220 193 exposures (data not shown). Qmax (ppm) 814Cells exposed to 1000 ppm of toluene for 1 h did not show any Tf (°C) 300 244 significant induction in the gene expression of CYP1A1, CYP1B1, R (°C) 124 115 P (°C) 67 44 CYP2S1 and EPHX1. On the contrary, CYP2E1, CYP2F1, NQO1 and X (%) 86 95 AhR were up-regulated when compared to control cells. CYP2E1 and NQO1 expression showed a dose-response increase with RQ values of 26 and 27 respectively at 100 ppm of toluene, and RQ values of 89 and 3. Results 27 respectively at 1000 ppm of toluene. However 100 ppm of toluene did not induce significant gene expression of CYP2F1 and AhR in 3.1. Catalysts exposed cells (1.8 and 1.6 respectively). These genes were only induced when the cells were exposed to 1000 ppm of toluene with a value of 3.2 The toluene light-off curves are reported on the Fig. 1. Table 2 for CYP2F1 and 2.3 for AhR (Fig. 3). fi reports speci c surface areas and palladium content of each catalyst, as When Pd/α-Al2O3 was used for the oxidation of toluene, the well as the values of T50,T100 and intrinsic activity. The results show exposed cells showed an increased gene expression of CYP1B1, relatively close performances, but the catalysts can be discriminated NQO1, EPHX1 and AhR at 10% and 100% of oxidation exhausts, γ according to T50 and T100 values. The commercial catalyst, Pd/ -Al2O3, compared to the control cells. CYP1A1, CYP2F1 and CYP2S1 were up- shows the best performance with a T50 of 179°C and a total conversion regulated only when the cells were exposed to 100% of the catalytic α at 200°C. However, according to the intrinsic activity, Pd/ -Al2O3 was exhaust with RQ values of 2.1, 10.4 and 6.3 respectively. However, the best catalyst. CYP2E1 did not show any significant induction (Fig. 4A).

In dynamic conditions and in order to carry out a total oxidation, The cells exposed for 1 h to catalytic exhaust with the Pd/γ-Al2O3 the knowledge of the generated by-products is limited. In order to revealed an up-regulation in the gene expression of CYP1A1, NQO1 define the nature of by-products, analysis by microGC was performed. and AhR at 10% and 100% with RQ values of 3.2, 5.1 and 5.2 The microGC technique, adapted for quick (less 4 min) and sensitive respectively at 100%. EPHX1 showed a significant induction (2.5) only analysis, has allowed detection of the majority of oxidation intermedi- at 100% whereas CYP1B1, CYP2E1, CYP2F1 and CYP2S1 were not ates (eg. benzene, benzaldehyde, methyldiphenylmethane) (Brunet induced in these cells (Fig. 4B). To evaluate the efficiency of this et al., 2015). Benzene was thus identified as the main by-product catalyst in total degradation of VOCs, A549 cells were exposed for 1 h resulting from toluene total oxidation (Andersson, 1986). Its toxic to exhaust of the Pd/γ-Al2O3 catalyst at a temperature higher than the properties, particularly its carcinogenic potential, are known since final temperature Tf at which 100% of toluene was converted with no many years. Consequently, it is important to know under what benzene formation. The gene expression of the XMEs (CYP2E1, conditions and in what amounts this compound may be issued by the CYP1A1, CYP1B1, CYP2S1, NQO1, and EPHX1) and the xenobiotic catalytic process. Benzene could be quantified by microGC. Emission receptor AhR was evaluated (Fig. 5). The exposed cells did not show profiles have been determined according to the toluene conversion and any significant change in the expression of these genes when compared the working temperature for each catalyst. Benzene production began to the cells exposed to purified air. at the same temperature as the toluene conversion (Fig. 1). Its quantity increased rapidly until reaching a maximum emission, after which the 4. Discussion issued amount of benzene decreased slowly. The results showed quite fi α similar benzene emission pro les and Qmax for the catalysts Pd/ - Toluene is widely used in industries despite its major and direct γ Al2O3 and Pd/ -Al2O3. Regarding the emission range (R), this value impacts on human health. It is therefore important to reduce its was relatively large with an average of 120°C. Concerning the persis- emissions directly at the source. Catalytic oxidation is an efficient tence, P values clearly showed that benzene was not removed when the remediation technique for the treatment of these organic compounds. toluene conversion was complete (T100). This implies that it is Catalysts developed in this project to reduce toluene emissions must be necessary to increase the working temperature for both catalysts in validated, before thinking about their potential scaling-up, and appli- order to completely remove the initial VOC and its by-products. This cation on site. The originality of this research is that the two catalysts point is particularly important because it will have a direct impact on were not only chemically validated by microGC, but also toxicologically the treatment unit operating cost at industrial scale. With a R value of evaluated using an ALI exposure system Vitrocell®. Pd based catalysts, γ 115°C and a P value of 44°C, Pd/ -Al2O3 was the lowest emitter of such as Pd/α-Al2O3 and Pd/γ-Al2O3, are indeed well known for the benzene catalyst ( Fig. 2). total oxidation of VOCs. However, it's relatively unknown that these catalysts can emit several toxic by-products in operating test conditions 3.2. Toxicological analysis especially for BTEX total oxidation. We thus chose particularly these catalysts in order to compare the by-product formation during the The measure of extracellular release of LDH did not show any toluene oxidation. significant differences between cells exposed to toluene or to VOCs In order to assess the toxicity of catalysts exhausts and to validate mixtures and cells exposed to purified air (data not shown). These the choice of the best catalyst, the cells had to be exposed to gaseous

331 M. Al Zallouha et al. Environmental Research 152 (2017) 328–335

Fig. 2. Light-oﬀ curves of impregnated catalysts and production of benzene versus temperature. The dashed lines correspond to the benzene production as function of toluene light-oﬀ curve. (Qmax: maximum observed amount of benzene, R: range emission of benzene, and P: persistence of benzene).

Toluene 1000 ppm 128 * Pd/γ-Al2O3 32 64 * * * 32 16

16 8 8

4 * 4 Relative expression * 2 Relative expression 2 1 CYP1A1 CYP1B1 CYP2E1 CYP2F1 CYP2S1 NQO1 EPHX1 AhR 1 Fig. 3. Gene expression of xenobiotic metabolizing enzymes (XMEs) (CYP1A1, CYP1B1, CYP1A1 CYP1B1 CYP2E1 CYP2S1 NQO1 EPHX1 AhR CYP2E1, CYP2F1, CYP2S1, NQO1, EPHX1) and AhR receptor in A549 cells exposed for Fig. 5. Gene expression of xenobiotic metabolizing enzymes (XMEs) (CYP1A1, CYP1B1, 1 h to puriﬁed air (controls, RQ=1) and to toluene at 10% (100 ppm, light grey) or 100% CYP2E1, CYP2S1, NQO1, EPHX1) and AhR receptor in A549 cells exposed for 1 h to (1000 ppm, dark grey). Data are shown as median RQ values vs controls and puriﬁed air (controls, RQ=1) and to the gas from the catalytic degradation of toluene by interquartile range. (non-parametric Mann-Whitney U-test, *: p < 0.05). Pd/γ-Al2O3 at 256°C at 10% (100 ppm, light grey) or 100% (1000 ppm, dark grey). Data are shown as median RQ values vs controls and interquartile range. (non-parametric mixtures. Several exposure techniques can be used with direct dilution Mann-Whitney U-test, vs controls, *: p < 0.05). of the compounds in the liquid culture medium under submerged conditions, or cell exposure using roller bottles and rocking or rotating

Fig. 4. Gene expression of xenobiotic metabolizing enzymes (XMEs) (CYP1A1, CYP1B1, CYP2E1, CYP2F1, CYP2S1, NQO1, EPHX1) and AhR receptor in A549 cells exposed for 1 h to puriﬁed air (controls, RQ=1) and to the gas from the catalytic degradation of toluene by Pd/α-Al2O3 (T=233°C) (Fig. 4A) and Pd/γ-Al2O3 (T=200°C) (Fig. 4B) at 10% (100 ppm, light grey) or 100% (1000 ppm, dark grey). Data are shown as median RQ values vs controls and interquartile range. (non-parametric Mann-Whitney U-test, vs controls, *: p < 0.05).

332 M. Al Zallouha et al. Environmental Research 152 (2017) 328–335 platforms, or exposure on collagen gel or microporous membranes XME in cells (Mendoza-Cantú et al., 2006). CYP2E1 can thereby help (Aufderheide, 2005). The Vitrocell® system is a unique ALI exposure in the toxicological validation phase, being helpful to detect the system presenting many advantages against other exposure techniques. presence of toluene and therefore to evaluate the efficiency of the Firstly, it avoids any loss of volatile pollutants out of the system. catalyst tested in toluene degradation. Finally, CYP1A1, CYP1B1, Secondly, and unlike other methods of exposure, no dilution of the CYP2S1 and EPHX1 were not induced after toluene exposure which pollutant in culture medium is needed during exposure. The exposure was in accordance with Billet et al., showing that no significant concentration stays constant. Thirdly, this technique allows a perma- induction of CYP1A1 gene took place after A549 exposure to 7 µM of nent contact between the chemical substances and the exposed cells toluene (Billet et al., 2007). which simulates quite good the in vivo exposure techniques by Secondly, toluene catalytic oxidation was performed using two inhalation. Fourthly, this exposure system heated by a water bath can catalysts Pd/α-Al2O3 and Pd/γ-Al2O3. The comparison of the level of operate independently of cell culture incubators. That could become an induction of genes involved in toluene metabolism CYP2E1 and NQO1, advantage for future toxicological study of airborne chemicals and real revealed that these genes were less expressed in cells exposed to mixtures generated in chemical processes or sampled in the environ- catalytic exhausts than in cells exposed to toluene. CYP2E1 was more ment. As the other in vitro exposure models, this system cannot than 40 fold less induced. This means that toluene should be totally reproduce the toxicodynamic and toxicokinetic of a xenobiotic in the oxidized at the end of the catalytic reaction. NQO1 showed a 3- and 5- organism, but it allows to measure directly on cells the local toxicity, fold decrease in induction with Pd/α-Al2O3 and Pd/γ-Al2O3 respec- cytotoxicity and mutagenicity. In addition to that, the Vitrocell® system tively. This latter result suggests that other organic compounds than artificially reproduces the physical interactions involved in the pul- toluene still were always present at the end of the catalytic reaction, monary respiration, therefore, the transferability of results is more making Pd/γ-Al2O3 probably more efficient than Pd/α-Al2O3 in toluene reliable than classical submerged in vitro method of exposure to gas oxidation. These two catalysts actually decreased significantly toluene (Pariselli et al., 2009). Because of these properties, the Vitrocell® concentration in exhaust gases (Fig. 1). This outcome allowed us to system is more and more used in the assessment of gaseous chemicals validate the performed coupling. toxicity. Human lung cells A549 were exposed through air-liquid Even if catalytic oxidation decreases toluene concentration, it can interface to low concentrations of toluene, benzene and formaldehyde generate by-products that can be more toxic than the initial treated in order to study the biological effects of the VOCs (Pariselli et al., toluene. Thus, the choice of the most suitable catalyst should not 2009). neglect the toxicological effects of these secondary compounds. Human lung cells A549 were exposed for 1 h to a stream of Analysis by microGC identified benzene among the main by-product 1000 ppm of toluene with or without catalytic treatment and with or of toluene catalytic oxidation. Toluene light-off curve reveals that after without 10-fold dilution of gases. Cytotoxicity study performed with achieving the total degradation, the amount of benzene emitted is

LDH test did not show any significant changes in the membrane around 5 ppm for Pd/α-Al2O3 and Pd/γ-Al2O3 (Fig. 1). integrity of cells exposed to catalytic emissions or toluene. The absence Cell exposure to gases from catalytic oxidation of toluene using Pd/ of significant cell death allowed us to consider the underlying patho- α-Al2O3 enhanced mRNA expression of CYP2F1 gene whereas with Pd/ physiological mechanisms of toxicity. When specific, these mechanisms γ-Al2O3 no significant induction was observed. However, the higher could be considered as biomarkers of exposure to VOCs or to their by- induction level of CYP2F1 observed with Pd/α-Al2O3 when compared products, and could be used as biological sensors of their presence. One to control cells can be explained by the fact that this gene is induced by of the first biological responses activated by the contact between the toluene by-products, and first of all, by benzene (Powley and Carlson, cells and xenobiotics is the activation via specific receptors of proteins 2000). This latter result means that less benzene was present at the end involved in their biotransformation. Two phases facilitate the excretion of the catalytic reaction making Pd/γ-Al2O3 more efficient than Pd/α- of lipophilic xenobiotics out of the body by increasing their hydro- Al2O3, emitting less benzene. philicity. In this study, the level of gene expression of 6 XMEs In the two catalytic processes, the cells exposed to catalytic exhaust (CYP2E1, CYP1A1, CYP1B1, CYP2S1, EPHX1 and NQO1) and 3 showed also an induction of many genes not involved in toluene receptors (CAR, PXR and AhR) were considered in A549 cells exposed metabolism such as CYP1A1, CYP1B1, CYP2S1 and EPHX1. Even if for 1 h to 1000 ppm of toluene or to gases from the catalytic EPHX1 is involved in benzene detoxification, it is also linked to more degradation of a stream of 1000 ppm toluene by Pd/α-Al2O3 and Pd/ complex compounds of higher molecular weight. EPHX1 was 10 fold γ-Al2O3. more expressed in cells treated with exhaust using Pd/α-Al2O3 than Firstly, to verify the classical activation pathway of toluene, A549 with Pd/γ-Al2O3. The more significant induction observed revealed cells were exposed for 1 h to 1000 ppm of toluene. These cells revealed that a higher concentration of by-products was generated with Pd/α- an increase in the expression of CYP2F1 and AhR genes only at Al2O3 catalyst. 1000 ppm whereas NQO1 and CYP2E1 were significantly and dose- The results showed an increase in gene induction of CYP1A1, dependently induced at both concentrations. NQO1 is not only CYP1B1 and CYP2E1. CYP1 subfamily members are the major CYP involved in toluene metabolism, its induction can also be due to involved in PAH bioactivation with numerous studies emphasizing the substrates belonging to different families like BTEX, Polycyclic capacity of PAHs in CYP1A1 and CYP1B1 induction (Garçon et al., Aromatic Hydrocarbons (PAHs) and phenolic oxidants (Ross et al., 2004; Shimada et al., 1996; Shimada and Fujii-Kuriyama, 2004). Even 2000). In lungs, CYP2E1 and CYP2F1 are the most involved enzymes in though CYP2S1 belongs to the CYP2 family, it is involved in PAH the metabolism of VOCs (Rossi et al., 1999). CYP2E1 participates in metabolism as in the case of naphthalene exposure (Karlgren et al., the main metabolic pathway of toluene by oxidizing it into benzyl 2005; Preuss et al., 2003). In addition, our results reported an increase alcohol. In addition, a significant increase in CYP2E1 activity was in AhR gene expression of 2- and 3-fold using Pd/γ-Al2O3 and Pd/α- observed in cells treated with 5 mM of toluene for 24 h (Al-Ghamdi Al2O3 respectively when compared to control. This nuclear receptor is a et al., 2003). CYP2F1 was shown induced by benzene in A549 cells cytochrome inducer (CYP1A1, CYP1B1 and CYP2E1) regulated by (Sheets et al., 2004). However, this XME seemed to be slightly induced PAHs (Mimura and Fujii-Kuriyama, 2003). The increase in the by toluene only at a high concentration of exposure. Moreover, results expression of cytochromes and their regulatory receptor confirmed also revealed that cells exposed to 100% of toluene showed a higher the presence of PAHs in the catalytic exhausts after oxidation of induction of CYP2E1 and CYP2F1 than those exposed to 10% dilution toluene using both catalysts. However with Pd/α-Al2O3 a more which suggests the presence of a dose-effect relationship. In agreement significant induction is observed suggesting that more PAHs were with Mendoza et al., CYP2E1 can be considered as a useful biomarker generated. Further chemical analyses by GC-MS finally confirmed the of toluene exposure since this VOC increases mRNA content of this presence of traces of PAH (C18H12) in both Pd/α-Al2O3 and Pd/γ-Al2O3

333 M. Al Zallouha et al. Environmental Research 152 (2017) 328–335 exhausts. This PAH could be chrysene, benzanthracene or benzo[c] cultivation and exposure of cells at the air/liquid interface. Exp. Toxicol. Pathol. Off. J. Ges. Für Toxikol. Pathol. 51, 489–490. http://dx.doi.org/10.1016/S0940- phenanthrene, since their respective bay or fjord region is known to 2993(99)80121-3. interact with AhR (Zhang et al., 2012). Benoit, F.M., Davidson, W.R., Lovett, A.M., Nacson, S., Ngo, A., 1985. Breath analysis by The purpose of the toxicological validation was to measure the API/MS–human exposure to volatile organic solvents. Int. Arch. Occup. Environ. ffi Health 55, 113–120. e ciency of toluene degradation by comparing gene induction in Billet, S., Garçon, G., Dagher, Z., Verdin, A., Ledoux, F., Cazier, F., Courcot, D., Aboukais, toluene exposure and after catalytic oxidation taking into account also A., Shirali, P., 2007. Ambient particulate matter (PM2.5): physicochemical the composition of the generated by-products. The comparison of the characterization and metabolic activation of the organic fraction in human lung – two catalysts shows that less toluene metabolizing genes and less PAH epithelial cells (A549). Environ. Res. 105, 212 223. http://dx.doi.org/10.1016/ j.envres.2007.03.001. genes are induced with Pd/γ-Al2O3 making it more favourable than Pd/ Brunet, J., Genty, E., Landkocz, Y., Al Zallouha, M., Billet, S., Courcot, D., Siffert, S., fi α-Al2O3 for toluene oxidation. The exposure performed using Pd/γ- Thomas, D., De Weireld, G., and Cousin, R., 2015. Identi cation of by-products Al O at 256°C (T > T ) did not cause any significant change in the issued from the catalytic oxidation of toluene by chemical and biological methods. 2 3 f Comptes Rendus Chim., International Symposium on Air and Water Pollution expression of the genes coding for the enzymes involved in benzene and Abatement Catalysis (AWPAC)Voume 1 – Catalytic pollution control for stationary PAHs metabolism. This result revealed that these toxic by-products and mobile sources 18, 1084–1093. doi:http://dx.doi.org/10.1016/j.crci.2015.09. 001 were not formed anymore when Pd/γ-Al2O3 is used for toluene γ Burgos, N., Paulis, M., Mirari Antxustegi, M., Montes, M., 2002. Deep oxidation of VOC oxidation at this temperature, which validates the choice of Pd/ - mixtures with platinum supported on Al2O3/Al monoliths. Appl. Catal. B Environ. Al2O3 as the better performing catalyst. 38, 251–258. http://dx.doi.org/10.1016/S0926-3373(01)00294-6. This study highlighted the importance of reducing toluene, due to Chen, L.C., Fang, C.P., Qu, Q.S., Fine, J.M., Schlesinger, R.B., 1993. A novel system for ff the in vitro exposure of pulmonary cells to acid sulfate aerosols. Fundam. Appl. its direct e ects on human health. To increase both the relevance and Toxicol. 20, 170–176. http://dx.doi.org/10.1006/faat.1993.1023. efficiency of catalytic degradation, we proposed to assess the toxicity of Courcot, E., Leclerc, J., Lafitte, J.-J., Mensier, E., Jaillard, S., Gosset, P., Shirali, P., catalyst exhaust during toluene degradation. In a first step, we Pottier, N., Broly, F., Lo-Guidice, J.-M., 2012. Xenobiotic metabolism and ® disposition in human lung cell models: comparison with in vivo expression profiles. evaluated the adaptation of the Vitrocell system as a new, direct and Drug Metab. Dispos. Biol. Fate Chem. 40, 1953–1965. http://dx.doi.org/10.1124/ dynamic model of ALI continuous exposure. This device appeared dmd.112.046896. complementary to the classical chemical approach of characterization Garçon, G., Gosset, P., Maunit, B., Zerimech, F., Creusy, C., Muller, J.-F., Shirali, P., fl 56 54 of emissions, since exposed cells could be considered as sensors of toxic 2004. In uence of iron ( Fe2O3 or Fe2O3) in the upregulation of cytochrome P4501A1 by benzo[a]pyrene in the respiratory tract of Sprague-Dawley rats. J. Appl. products. Exposed cells could indeed be sensitive to all classes of Toxicol. JAT 24, 249–256. http://dx.doi.org/10.1002/jat.979. xenobiotics. Moreover they integrate the potential interactions between Hansen, B. G., Munn, S. J., Pakalin, S., Luotamo, M., Allanou, R., Pellegrini, G., De toxics that could trigger synergy or antagonism. Finally, using biolo- Bruijn, J., Berthault, F., Vegro, S., Scheer, S., 2002. European Union Risk Assessment Report Toluene. CAS No. 108-88-3. EINECS No. 203-625-9 (EUR - gical systems could be considered as a powerful untargeted approach Scientific and Technical Research Reports) since cells react without the need of knowing the chemical nature of the INRS, 2012. Fiche toxicologique: Toluène.Institut National de Recherche et de Sécurité exhaust. This finding explained the interest of performing a toxicolo- pour la prévention des maladies professionnelles et des accidents du travail. Karlgren, M., Miura, S., Ingelman-Sundberg, M., 2005. Novel extrahepatic cytochrome gical validation in order to determine the more suitable catalyst for P450s. Toxicol. Appl. Pharmacol. 207, 57–61. http://dx.doi.org/10.1016/ toluene oxidation. The increase in gene expression suggested the j.taap.2004.12.022. presence of by-products whose toxicity must be assessed before the Khan, F.I., Ghoshal, A, Kr, 2000. Removal of Volatile Organic Compounds from polluted air. J. Loss Prev. Process Ind. 13, 527–545. http://dx.doi.org/10.1016/S0950- adoption of new catalysts. The application of toxicological validation 4230(00)00007-3. here confirmed that Pd/γ-Al2O3 is more efficient than Pd/α-Al2O3 in Kim, S.C., Shim, W.G., 2009. Properties and performance of Pd based catalysts for toluene remediation. catalytic oxidation of volatile organic compounds. Appl. Catal. B Environ. 92, 429–436. http://dx.doi.org/10.1016/j.apcatb.2009.09.001. Mendoza-Cantú, A., Castorena-Torres, F., Bermúdez de León, M., Cisneros, B., López- Acknowledgements Carrillo, L., Rojas-García, A.E., Aguilar-Salinas, A., Manno, M., Albores, A., 2006. Occupational toluene exposure induces cytochrome P450 2E1 mRNA expression in – This work was supported by the European Interreg IV REDUGAZ peripheral lymphocytes. Environ. Health Perspect. 114, 494 499. Mimura, J., Fujii-Kuriyama, Y., 2003. Functional role of AhR in the expression of toxic (Convention no. FW4.1.5), and the French Agency of the Environment effects by TCDD. Biochim. Biophys. Acta 1619, 263–268. and Energy (ADEME CORTEA ProBTEX; Convention no. 1281 Ordóñez, S., Bello, L., Sastre, H., Rosal, R., D́ıez, F.V., 2002. Kinetics of the deep fi oxidation of benzene, toluene, n-hexane and their binary mixtures over a platinum C0095). MAZ bene ts from grants from the Hauts-de-France Region γ – ′ on -alumina catalyst. Appl. Catal. B Environ. 38, 139 149. http://dx.doi.org/ and the Pôle métropolitain de la Côte d Opale. This work is a 10.1016/S0926-3373(02)00036-X. contribution to the CPER research project CLIMIBIO. The authors Papaefthimiou, P., Ioannides, T., Verykios, X.E., 1997. Combustion of non-halogenated thank the French Ministère de l′Enseignement Supérieur et de la volatile organic compounds over group VIII metal catalysts. Appl. Catal. B Environ. 13, 175–184. http://dx.doi.org/10.1016/S0926-3373(96)00103-8. Recherche, the Hauts-de-France Region and the European Funds for Papaefthimiou, P., Ioannides, T., Verykios, X.E., 1998. Catalytic incineration of volatile Regional Economical Development for their financial support to this organic compounds Present in industrial waste streams. Appl. Therm. Eng. 18, project. The authors thank Pr Maurizio Gualtieri for the preparation of 1005–1012. http://dx.doi.org/10.1016/S1359-4311(98)00021-0. fi Pariselli, F., Sacco, M.G., Rembges, D., 2009. An optimized method for in vitro exposure the manuscript. They declare no actual or potential competing nancial of human derived lung cells to volatile chemicals. Exp. Toxicol. Pathol. 61, 33–39. interests. http://dx.doi.org/10.1016/j.etp.2008.05.008. Powley, M.W., Carlson, G.P., 2000. Cytochromes P450 involved with benzene metabolism in hepatic and pulmonary microsomes. J. Biochem. Mol. Toxicol. 14, References 303–309. http://dx.doi.org/10.1002/1099-0461(2000)14:6 < 303::AID-JBT2 > 3.0.CO;2-8. Al-Ghamdi, S.S., Raftery, M.J., Yaqoob, M.M., 2003. Acute solvent exposure induced Preuss, R., Angerer, J., Drexler, H., 2003. Naphthalene–an environmental and activation of cytochrome P4502E1 causes proximal tubular cell necrosis by oxidative occupational toxicant. Int. Arch. Occup. Environ. Health 76, 556–576. http:// stress. Toxicol. Vitr. 17, 335–341. http://dx.doi.org/10.1016/S0887-2333(03) dx.doi.org/10.1007/s00420-003-0458-1. 00021-3. Rasmussen, R.E., 1984. In vitro systems for exposure of lung cells to NO2 and O3. J. Andersson, S.L.T., 1986. Reaction networks in the catalytic vapor-phase oxidation of Toxicol. Environ. Health 13, 397–411. http://dx.doi.org/10.1080/ toluene and xylenes. J. Catal. 98, 138–149. http://dx.doi.org/10.1016/0021- 15287398409530506. 9517(86)90304-0. Ross, D., Kepa, J.K., Winski, S.L., Beall, H.D., Anwar, A., Siegel, D., 2000. NAD(P) Aufderheide M., 2005. Direct exposure methods for testing native atmospheres. Exp. H:quinone oxidoreductase 1 (NQO1): chemoprotection, bioactivation, gene Toxicol. Pathol. 57 (Suppl. 1), S213–S226. http://dx.doi.org/10.1016/ regulation and genetic polymorphisms. Chem. Biol. Interact. 129, 77–97. http:// j.etp.2005.05.019. dx.doi.org/10.1016/S0009-2797(00)00199-X. Aufderheide M., Knebel J.W., Ritter D., 2003. An improved in vitro model for testing the Rossi, A.M., Guarnieri, C., Rovesti, S., Gobba, F., Ghittori, S., Vivoli, G., Barale, R., 1999. pulmonary toxicity of complex mixtures such as cigarette smoke. Exp. Toxicol. Genetic polymorphisms influence variability in benzene metabolism in humans. Pathol. Off. J. Ges. Für Toxikol. Pathol. 55, 51–57. http://dx.doi.org/10.1078/0940- Pharmacogenetics 9, 445–451. 2993-00298. Rusu, O.-A., Dumitriu, E., 2003. Destruction of volatile organic compounds by catalytic Aufderheide M., Mohr, U., 1999. CULTEX–a new system and technique for the oxidation. Environ. Eng. Manag. J. 2, 243–254.

334 M. Al Zallouha et al. Environmental Research 152 (2017) 328–335

Sheets, P.L., Yost, G.S., Carlson, G.P., 2004. Benzene metabolism in human lung cell lines 450 1B1. Cancer Res. 56, 2979–2984. BEAS-2B and A549 and cells overexpressing CYP2F1. J. Biochem. Mol. Toxicol. 18, Tormoehlen, L.M., Tekulve, K.J., Nañagas, K.A., 2014. Hydrocarbon toxicity: a review. 92–99. http://dx.doi.org/10.1002/jbt.20010. Clin. Toxicol. Phila. Pa 52, 479–489. http://dx.doi.org/10.3109/ Shimada, T., Fujii-Kuriyama, Y., 2004. Metabolic activation of polycyclic aromatic 15563650.2014.923904. hydrocarbons to carcinogens by cytochromes P450 1A1 and 1B1. Cancer Sci. 95, Zhang, L., Jin, Y., Huang, M., Penning, T.M., 2012. The role of human aldo-keto 1–6. reductases in the metabolic activation and detoxication of polycyclic aromatic Shimada, T., Hayes, C.L., Yamazaki, H., Amin, S., Hecht, S.S., Guengerich, F.P., Sutter, hydrocarbons: interconversion of PAH catechols and PAH o-quinones. Exp. T.R., 1996. Activation of chemically diverse procarcinogens by human cytochrome P- Pharmacol. Drug Disco. 3, 193. http://dx.doi.org/10.3389/fphar.2012.00193.

335