Non-synergistic cytotoxic effects of Fusarium and Alternaria toxin combinations in Caco-2 cells

Toxicology Letters 241 (2016) 1–8

Contents lists available at ScienceDirect

Toxicology Letters journal homepage: www.elsevier.com/locate/toxlet

Non-synergistic cytotoxic effects of Fusarium and Alternaria toxin combinations in Caco-2 cells Katharina Vejdovszkya , Benedikt Wartha , Michael Sulyokb , Doris Markoa,* a

University of Vienna, Department of Food Chemistry and Toxicology, Waehringer Str. 38, A-1090 Vienna, Austria Center for Analytical Chemistry, Department for Agrobiotechnology (IFA-Tulln), University of Natural Resources and Life Sciences, Vienna (BOKU), KonradLorenz-Str. 20, Vienna, A-3430 Tulln, Austria b

H I G H L I G H T S








Antagonistic effects of combined Fusarium mycotoxins at high concentrations. Tenuazonic acid seems to decrease the cytotoxicity of deoxynivalenol. Enniatin B showed stronger cytotoxicity in Caco-2 than deoxynivalenol or nivalenol. Aurofusarin mediated pronounced cytotoxicity—first toxicological characterization. Aurofusarin lead to additive effects on cytotoxicity with other mycotoxins.

A R T I C L E I N F O

A B S T R A C T

Article history: Received 11 September 2015 Received in revised form 23 October 2015 Accepted 26 October 2015 Available online 31 October 2015

Exposure of humans and animals to mycotoxins via food and feed generally involves a conglomeration of compounds contaminating the consumed products. Investigations on combinatory effects of mycotoxins are therefore of great importance. In this study, cytotoxic effects of binary mixtures of the Fusarium toxins enniatin B, aurofusarin, deoxynivalenol, nivalenol and zearalenone, and tenuazonic acid produced by Alternaria spp., were evaluated by the WST-1 assay in the colorectal carcinoma cell-line Caco-2 after 24 h of incubation. The selection of these mycotoxins was based on typically occurring natural contamination patterns in grains. Aurofusarin, which can be found abundantly in contaminated foodstuff and has not been toxicologically characterized properly so far, showed pronounced cytotoxicity, decreasing the mitochondrial activity at 10 mM to 51% compared to a solvent control. Combinations of other mycotoxins with aurofusarin showed additive effects. In contrast, binary mixtures of enniatin B, deoxynivalenol, nivalenol and zearalenone at cytotoxic concentrations, predominantly resulted in antagonistic effects. Binary combinations of these four Fusarium toxins with tenuazonic acid also revealed interacting effects leading to a decrease in cytotoxicity, compared to expected combinatory effects. Especially in combination with deoxynivalenol, tenuazonic acid was found to significantly reduce the cytotoxicity of this mycotoxin in Caco-2 cells. Synergistic effects were not observed for any toxin combination under the chosen conditions. ã 2015 Elsevier Ireland Ltd. All rights reserved.

Keywords: Food contaminants Combinatory effects Mycotoxins Antagonism Caco-2 WST-1 assay

1. Introduction Infestation of cereals and fruits by mycotoxin producing fungi and, consequently, the entry of mycotoxins into the food chain is a

Abbreviations: DON, deoxynivalenol; NIV, nivalenol; ZEN, zearalenone; ENN B, Enniatin B; AURO, aurofusarin; TeA, tenuazonic acid; CI, Combination Index; DMSO, dimethyl sulfoxide; SEM, standard error of the mean. * Corresponding author. E-mail address: [email protected] (D. Marko). http://dx.doi.org/10.1016/j.toxlet.2015.10.024 0378-4274/ ã 2015 Elsevier Ireland Ltd. All rights reserved.

worldwide issue. At present, risk assessment of mycotoxins is generally based on the evaluation of single compounds (EFSA, 2004, 2011a,b, 2013, 2014). Several studies, which focused on the analysis of mycotoxin levels in feed and foodstuff, indicated that contamination of a product by just one mycotoxin does hardly occur in contrast to co-contaminations with several compounds, which are observed in most of the tested samples. Since one given fungal species may be capable of producing several secondary metabolites and, additionally, several fungal species may possibly be present in one sample, such co-contaminations are plausible

2

K. Vejdovszky et al. / Toxicology Letters 241 (2016) 1–8

(Streit et al., 2013; Uhlig et al., 2013). Evidence for the co-exposure of humans through contaminated foodstuff was recently also confirmed directly in human samples by analysis of multibiomarkers in urine (Abia et al., 2013; Warth et al., 2013). Therefore, exposure of humans and animals is not limited to one mycotoxin at a time, which highlights the great need of testing combinatory effects of mycotoxins. So far, the knowledge in this field is still limited and corresponding studies mainly focused on combinations of different Fusarium toxins. What has not been investigated so far, are combinatory effects of fusarotoxins with Alternaria toxins, although analytical data repeatedly demonstrated their natural co-occurrence (Domijan et al., 2005; Ezekiel et al., 2013, 2012; Streit et al., 2013; Uhlig et al., 2013; Warth et al., 2012). At the current state of knowledge, toxicological effects of some individual mycotoxins are well understood, as it is the case for

deoxynivalenol (DON), nivalenol (NIV) and zearalenone (ZEN), whilst other mycotoxins still need to be characterized in more detail. DON as a type-B trichothecene for example, is known to inhibit protein synthesis by binding to ribosomes as main effect (Pestka, 2010). Furthermore, some studies indicate a genotoxic potential of DON (Yang et al., 2014). NIV, another type-B trichothecene, also inhibits protein synthesis as well as DNA synthesis. In addition, DNA-damaging properties and induction of apoptosis seem to be likely (Minervini et al., 2004; Tatsuno, 1968). ZEN, and especially its reduced metabolite a-zearalenol, is known to possess estrogenic activity due to structural similarities to the female sexual hormone 17b-estradiol (Shier et al., 2001). Other studies indicate, that ZEN is capable of activating the pregnane X receptor (PXR) and thereby may interfere with the regulation of genes involved in endo- and xenobiotic metabolism (Ding et al.,

Fig. 1. Effects of 24 h incubation of Caco-2 cells with individual mycotoxins on mitochondrial activity determined by the WST-1 assay and calculated as % of solvent control (mean + SEM). Triton X was used as positive control. Statistical significance to the NOEL was evaluated by one-way ANOVA and Bonferroni post-hoc test (*p > 0.05; **p > 0.01; ***p > 0.001).

K. Vejdovszky et al. / Toxicology Letters 241 (2016) 1–8

2006). Nevertheless, ZEN was also shown to generate reactive oxygen species and to cause DNA damage (Abid-Essefi et al., 2003, 2004). Enniatin B (ENN B) is known to interfere with biological membrane integrity by forming cation-selective channels, which may cause disturbance of ionic homeostasis (Kamyar et al., 2004). The toxicity of aurofusarin (AURO), also a fusarotoxin, is widely unexplored yet, but one study indicated that it may have impact on the oxidative status and fatty acid metabolism (Dvorska et al., 2002). The ability of Alternaria species to produce toxins is known for more than four decades. However, relatively little attention has been paid to the specific toxic effects of Alternaria toxins. The present data on the occurrence and toxicity of Alternaria toxins were assessed by the CONTAM panel of EFSA with the conclusion, that a risk to human health by these mycotoxins cannot be excluded (EFSA, 2011a). In vivo studies in chicken embryos with Alternaria toxins describe pronounced toxic effects of tenuazonic acid (TeA), exceeding those of other Alternaria toxins like alternariol or its monomethyl ether (Griffin and Chu, 1983). Although bacterial tests did not reveal mutagenic effects, TeA was associated with precancerous changes in the esophageal mucosa of mice (Schrader et al., 2001; Scott and Stoltz, 1980; Yekeler et al., 2001). In vitro studies however, repeatedly showed that TeA mediates less cytotoxicity and genotoxicity compared to other Alternaria toxins such as alternariol and altertoxin II (Pero et al., 1973; Schwarz et al., 2012; Zhou and Qiang, 2008). With respect to the mode of action, TeA has been described to inhibit protein synthesis (Shigeura and Gordon, 1963), similar to the effects of DON and NIV which might lead to potential combinatory effects of these substances. Available occurrence data in food and feed demonstrate that TeA and the Fusarium toxin AURO are frequently found in high concentrations compared to other mycotoxins, underlining the importance to evaluate combinatory effects of these substances (Asam et al., 2012; Ezekiel et al., 2012; Uhlig et al., 2013; Warth et al., 2012). The present study examined combinatory effects on cytotoxicity of the Alternaria toxin TeA and Fusarium toxins AURO, ENN B, DON, NIV and ZEN. Irrespective of the toxicity of the single compounds, there is an urgent need for analysis of interactions

3

between these mycotoxins, since humans and animals are incessantly exposed to co-contaminated products. 2. Materials and methods 2.1. Chemicals and reagents Cell culture media and supplements were purchased from GIBCO Invitrogen (Karlsruhe, Germany), Sigma–Aldrich (Schnelldorf, Germany) and SARSTEDT AG & CO. (Nuembrecht, Germany). Aurofusarin was purchased from Adipogen AG (Liestal, Switzerland). Enniatin B, zearalenone, deoxynivalenol, nivalenol, tenuazonic acid copper salt and other substances used in these studies were purchased from Sigma–Aldrich (Schnelldorf, Germany) and Roth (Karlsruhe, Germany). Previous to cell culture tests, copper was removed from tenuazonic acid copper salt by ion exchange chromatography according to Solfrizzo et al. (2004). 2.2. Cell culture The cell clone C2BBe1 of the human colorectal adenocarcinoma cell line Caco-2 was chosen to serve as a simplified model system for the human intestine in the experiments. It was purchased from LGC Standards (Wesel, Germany). Cells were cultured in humidified incubators (37  C, 5% CO2) in Dulbecco’s Modified Eagle’s Medium (DMEM) GlutaMAXTM supplemented with 10% fetal calf serum (FCS) and 1% penicillin/streptomycin. 2.3. Cell proliferation assay (WST-1) Caco-2 cells were seeded into 96-well plates at a density of 5000 cells per well and cultured for 48 h. Incubation was conducted in serum-free medium containing the corresponding mycotoxin concentration and a final concentration of 1% of the solvent dimethyl sulfoxide (DMSO). Solutions with 1% DMSO and 1% Triton-X 100 were used as negative and positive control, respectively. After 24 h the incubation solution was replaced by a detection solution containing serum-free medium and 10% WST1 reagent (Roche Applied Science, Mannheim, Germany; WST1 Test Kit). After 2 h, absorbance was measured at 450 nm and

Fig. 2. Results of the WST-1 assay of combinations at “low” concentrations represented as % of solvent control (mean + SEM). Bars for single mycotoxins (black bars and squared bars), expected effects of binary combinations calculated on basis of Independent Joint Action (angular striped bars, C), and measured effects of respective combinations (angular squared bars) are shown as indicated in the legend beneath the graph. No statistical significances between expected and measured effects were found by unpaired Student’s t-test.

4

K. Vejdovszky et al. / Toxicology Letters 241 (2016) 1–8

650 nm as reference wavelength (Wallac 1420 plate reader, PerkinElmer, Turku, Finland). 2.4. Experimental design and evaluation of combinatory effects The selection of mycotoxins to be tested in combinations was based on naturally occurring contamination patterns found in temperate climate regions during the last decade, in order to reflect a realistic scenario of co-exposure (Streit et al., 2013; Uhlig et al., 2013). The aim of the presented study was to gain information on the effects of binary combinations of the selected mycotoxins. Initially, a dose-response relation for each individual mycotoxin was determined by concentration range tests in the WST-1 assay. For details on tested concentrations see Fig. 1. In order to optimize the experimental setup, binary mixtures were tested in two concentration sets, one consisting of mycotoxins in “low” and another in “high” concentrations. Respectively, two concentrations have been defined for each individual mycotoxin. According to the initial dose–response tests, defined “low” concentrations showed no or negligible cytotoxic effects and “high” concentrations mediated pronounced cytotoxicity. Applied concentrations of each mycotoxin are indicated in Figs. 2–4 . Tests with mycotoxin combinations and the respective concentration of each single mycotoxin were conducted in parallel to ensure comparability of the data. Effects of single toxins and combinations were calculated by relating the results to the solvent control as described by following formula: % mitochondial activity of DMSO ¼ ð1  effectÞ100 For evaluation of combinatory effects two different mathematical models have been applied. A general principle describes a combinatory effect of two substances that do not interact as simply additive. Accordingly, interactions between compounds are indicated when the combinatory effect is not additive. The Combination Index Theorem is the model of first choice, which allows quantitative determination of synergistic or antagonistic effects (Chou, 2006). For tests of which no Combination Index (CI) could be calculated, due to reasons discussed below, the model of Independent Joint Action was applied (Groten et al., 2013). For the Combination Index, the dataset must provide several prerequisites,

Fig. 4. Results of WST-1 assay represented as % of solvent control (mean + SEM) of single mycotoxins and binary combinations. Combinatory effects were evaluated according to Chou and Talalay (1984) by the Combinatory Index (CI). “- - -” indicate antagonistic effects (CI = 1.45–3.3) and “- - - -” indicate strong antagonistic effects (CI = 3.3–10).

which may not always be the case. Most importantly, for both combined mycotoxins a dose–response relation must be detectable. In addition, if measured values exceed 100% mitochondrial activity, related to the DMSO control, which applies to some data of low concentrated mycotoxins, they cannot be expressed as an effect between 0 and 1, making the calculation of the CI inapplicable. Since TeA did not show a pronounced cytotoxic effect in the concentration range tests, it was not possible to establish a reliable dose–response relation. In case of AURO, its low solubility limited the determination of a complete dose–response curve exceeding 50% effect. For combinations that could not be evaluated by CI, which included all combinations of low concentrations and combinations with TeA and AURO, the model of Independent Joint Action was used to determine a potential deviation from additive effects.

Fig. 3. Results of the WST-1 assay represented as % of solvent control (mean + SEM) of single mycotoxins (black bars), expected effects of binary combinations calculated on basis of Independent Joint Action (striped bars; C), and measured effects of respective combinations (squared bars). (A) Combinations with TeA. (B) Combinations with AURO. Statistical significances were evaluated between expected and measured effects by unpaired Student’s t-test (*p > 0.05; **p > 0.01; ***p > 0.001).

K. Vejdovszky et al. / Toxicology Letters 241 (2016) 1–8

2.4.1. Combination Index Theorem For mycotoxins, for which a reliable dose–response relation could be determined, evaluation of synergistic or antagonistic effects of combinations in high concentrations was based on the Medium Effect Equation and the Combination Index Theorem (Chou, 1975; Chou and Talalay, 1983, 1984). The equation for the Combination Index Theorem can be found below, where D1 and D2 are the concentrations of the two mycotoxins applied in the combination, Dm and m are both determined via the medium effect equation and are parameters describing the potency (equivalent to ED50) and the shape of the dose–response curves respectively of each mycotoxin, and fa is the effected fraction by the combination: CI ¼

D1 Dm1 ½f a=1  f a1=m1

þ

D2 Dm2 ½f a=1  f a1=m2

A Combination Index (CI) value of one indicates additive effects whereas a CI < 1 indicates synergism and CI > 1 antagonism. 2.4.2. Independent Joint Action In this model an expected effect of the combination is calculated based on the effects of single toxins by the following equation: Expected ef f ect of combination ¼ ef f ect mycotoxin 1 þ ef f ect mycotoxin 2 ðef f ect mycotoxin 1  ef f ect mycotoxin 2Þ In order to evaluate effects below or above additivity, expected additive values were compared to actually measured values. 2.4.3. Additional tests on combinations with TeA In order to examine if interactions with TeA lead to reduced toxicity of other mycotoxins, as indicated in our preliminary tests, “high” concentrations of ENN B (5 mM), DON (10 mM), NIV (10 mM) and ZEN (50 mM) were combined with 1 mM, 10 mM, 100 mM, 200 mM and 250 mM of TeA. For evaluation, effects of binary mixtures with TeA were compared to the effects of the cytotoxic mycotoxins alone. This method was chosen to assess the combinatory effect of TeA independently of any mathematical model for additive effects. 2.5. Statistics All experiments were conducted in technical triplicates and a minimum of three independent biological replicates were tested. Statistical analyses were performed with Origin Pro 9.1G. Significance levels were set to 5% (*p > 0.05; **p > 0.01; ***p > 0.001). All data were analyzed for normality by the Shapiro–Wilk test. For results of concentration series of single mycotoxins, one-way ANOVA and Bonferroni post-hoc test were used to identify significant differences of all tested concentrations of one mycotoxin to the lowest concentration corresponding to the no observed effect level (NOEL). Significant differences between NOEL and the positive control, conducted with 1% Triton X, were analyzed by unpaired t-test. Data on combination tests evaluated by the Independent Joint Action were tested for normality. Significant differences between the expected values and the experimentally obtained values were analyzed using the unpaired t-test. Results of additional measurements of combinations with TeA have also been tested for normality. One-way ANOVA and Bonferroni post-hoc test were applied to identify significant differences between effects of single mycotoxins and combinations with TeA.

5

3. Results 3.1. Determination of cytotoxic effects in Caco-2 cells Concentration range tests for the five Fusarium toxins ENN B, DON, NIV, ZEN, AURO and the Alternaria toxin TeA have been conducted in the WST-1 assay in order to indirectly determine their cytotoxicity by measuring mitochondrial activity (Fig. 1). ENN B, DON, NIV and ZEN showed clear dose-dependent effects. Regarding the determined concentrations that give the halfmaximum response of each toxin, generally known as IC50 values (inhibitory concentration; 50%) (Table 1), ENN B (IC50 = 6.3 mM) revealed to be the most cytotoxic compound followed by NIV (IC50 = 6.9 mM) and DON (IC50 = 13.0 mM), whereas ZEN (IC50 = 49.5 mM) showed the least cytotoxic effect in Caco-2 cells after 24 h of incubation. AURO was found to be limited in its solubility in DMSO, although DMSO is described as the most suitable solvent by the manufacturer (Adipogen AG, Liestal, Switzerland). A maximal concentration of 1 mM was dissolvable in pure DMSO. When complying with a maximal final amount of DMSO of 1% in the cell culture medium, a maximal concentration of 10 mM AURO was achievable in the tests, mediating a mean reduction of mitochondrial activity to 51%. Tests with TeA showed limited cytotoxicity even in comparably high concentrations. At 250 mM TeA mitochondrial activity was decreased to 68%. In conclusion, the evaluated mycotoxins can be ranked according to their cytotoxicity in the following order: ENN B > NIV > DON  AURO > ZEN > TeA. 3.2. Cytotoxic effects of binary mycotoxin combinations Binary combinations of mycotoxins at “low” concentrations did not reveal effects significantly different to the calculated expectation for additivity (Fig. 2). Binary combinations of TeA and AURO with ENN B, DON, NIV or ZEN at “high” concentrations were evaluated via the model of Independent Joint Action. TeA combined with ENN B, DON, NIV or ZEN showed significantly lower cytotoxicity than the calculated expected additive effect (Fig. 3A). None of the combinations with AURO showed significant differences to the hypothetical additive effect (Fig. 3B). Combinatory effects of ENN B, DON, NIV and ZEN in “high” concentrations were determined via the Combination Index. All combinations were found to have antagonistic, or strong antagonistic effects, according to the categorization by Chou (2006) (Fig. 4,Table 2). 3.3. Additional tests on combinations with TeA In order to examine in more detail if interactions with TeA lead to reduced toxicity of other mycotoxins, as indicated in our initial tests (Fig. 3A), “high” concentrations of ENN B (5 mM), DON (10 mM), NIV (10 mM) and ZEN (50 mM) were combined with 1 mM, 10 mM, 100 mM, 200 mM and 250 mM of TeA. The observed cytotoxicity of 5 mM ENN B or 50 mM ZEN in binary combinations with TeA in the respective concentration range was equivalent to the toxicity of the single toxins (data not shown). On the contrary, the toxic effect of 10 mM DON was found to be slightly but Table 1 Results of dose-response measurements evaluated by Medium Effect Equation. Mycotoxin

Dm (, IC50)

m

ENN B ZEA DON NIV

6.3 49.5 13.1 6.9

2.6 2.8 0.6 0.2

6

K. Vejdovszky et al. / Toxicology Letters 241 (2016) 1–8

Table 2 Combinatory effects of binary mixtures evaluated by the Combination Index Theorem. Mycotoxin combination

CI

Indication

5 mM ENN B + 50 mM ZEA 5 mM ENN B + 10 mM DON 5 mM ENN B + 10 mM NIV 50 mM ZEA + 10 mM DON 50 mM ZEA + 10 mM NIV 10 mM DON + 10 mM NIV

1.6 1.8 5.2 2.6 4.9 6.0

Antagonism Antagonism Strong antagonism Antagonism Strong antagonism Strong antagonism

exerts lower cytotoxicity than the trichothecenes DON and NIV in Caco-2 cells. In this study ENN B, with an IC50 of 6.3 mM, was found to be the most cytotoxic mycotoxin when compared to NIV, DON, ZEN, AURO and TeA. Available data on ENN B also describe strong cytotoxic effects in Caco-2 cells (Ivanova et al., 2012). With respect to toxicity, AURO is a widely unexplored compound produced by Fusarium spp. To the best of our knowledge no in vitro toxicity tests have been performed on AURO so far. The limited solubility of this compound complicated a detailed characterization of its toxicity. A maximum of 10 mM AURO could be applied in the tests. Considering a mean reduction of mitochondrial activity to 51% by 10 mM, AURO shows similar cytotoxic potency as DON, although the results obtained with 1 and 5 mM indicate that AURO may show a steeper dose–response curve compared to DON. Our results on the impairment of mitochondrial activity in Caco-2 cells by TeA confirm previous studies which also indicated that TeA is less toxic compared to other Alternaria toxins (Pero et al., 1973; Schwarz et al., 2012; Zhou and Qiang, 2008). Only at comparably high concentrations of 250 mM cytotoxic effects could be detected, decreasing mitochondrial activity to 68%. 4.2. Cytotoxic effects of binary mycotoxin combinations

Fig. 5. Results of WST-1 assay represented as % of solvent control (mean + SEM) of concentration series of TeA (dotted line), 10 mM DON (A) or 10 mM NIV (B) (dashed lines) and the combination of 10 mM DON (A) or 10 mM NIV (B) with concentration series of TeA (solid line). Statistical significances were evaluated between values of 10 mM DON or 10 mM NIV and the respective combinations with TeA by one-way ANOVA (*p > 0.05; **p > 0.01; ***p > 0.001).

significantly decreased by TeA at concentrations between 10 mM and 200 mM, suggesting a concentration-independent interaction between TeA and DON, leading to antagonistic effects (Fig. 5A). Combinations of TeA with 10 mM NIV showed tendencies towards similar effects, however they were less pronounced than mixtures with DON and not found to be significant (Fig. 5B). 4. Discussion 4.1. Determination of cytotoxic effects in Caco-2 cells The results on the cytotoxicity of single mycotoxins investigated in this study are largely in accordance with data found in literature. In line with Wan et al. (2013) and Alassane-Kpembi et al. (2013), we confirmed that NIV is slightly more cytotoxic than DON. ZEN

The objective of this study was to perform a first characterization of possible interactions between the tested mycotoxins ENN B, DON, NIV, ZEN, AURO and TeA in human colon cells. Two concentrations of each mycotoxin, one defined as “low” and another as “high” concentration, depending on their cytotoxic effects, were chosen for binary combinations in Caco-2 cells. First, concentrations of no or not significant cytotoxic effects of single substances were applied in binary mixtures. None of these tested combinations revealed any effects deriving from additivity, indicating no interactions. Similar tests of combinations of DON, NIV ZEN and fumonisin B1 have been reported in swine jejunal epithelial cells after toxin exposure for 48 h (Wan et al., 2013). In the cited study, binary combinations of DON, NIV and ZEN, in noncytotoxic concentrations, were found to show cytotoxicity when combined. These interactions may possibly not be pronounced to this extent after only 24 h of incubation as it was tested in our study. In addition, combinations of “high” concentrations were tested. Since it was not possible to establish reliable dose–response curves for TeA and AURO, due to low toxicity or low solubility, respectively, combinatory effects were evaluated based on the model of Independent Joint Action. The present study represents the first investigation of combinatory effects of these fungal metabolites with other mycotoxins. Measurements on mixtures containing AURO showed additive effects with all tested substances. It should be considered, however, that the low solubility of AURO limited the concentration that could be applied in combinatory tests to 5 mM, which did not show pronounced cytotoxicity when applied as a single compound. Binary combinations of TeA with ENN B, DON, NIV or ZEN were found to be nonadditive and showed significantly decreased cytotoxicity compared to the expected effect. Since TeA is known to inhibit protein synthesis like DON and NIV, the evaluated non-additive effects indicate that there are different modes of action underlying these effects which may cause interactions (Pestka, 2010; Shigeura and Gordon, 1963; Tatsuno, 1968). Evaluation of combinatory effects of concentration series of TeA in combination with 10 mM DON showed that TeA slightly, but yet significantly, decreased the toxicity of DON in a concentration-independent manner. These findings suggest possible antagonistic effects of TeA on DON cytotoxicity in Caco-2 cells. Interestingly, this effect was found in similar tests of combinations of TeA and NIV to a lesser extent. However, DON and NIV differ in their modes of protein synthesis

K. Vejdovszky et al. / Toxicology Letters 241 (2016) 1–8

inhibition. DON inhibits translational initiation whereas NIV interferes with elongation and termination steps (Rocha et al., 2005). This may also lead to different interactions of DON and NIV with TeA. Possible mechanisms underlying the interactions between TeA and ENN B or ZEN, respectively, are not obvious and further investigations on more specific targets are required. However considering the possibility that the toxicity of one substance, for example ENN B or ZEN, may depend on the expression of specific factors, like metabolizing enzymes, the inhibition of protein synthesis by TeA may potentially lead to interactions lowering the toxicity of these substances, as it has been found here. Generally it cannot be excluded that not the mycotoxins themselves but maybe their metabolites, formed in the cells, play a role in any of the observed interacting effects. For mycotoxins of which a dose–response curve could be calculated, combinatory effects of binary mixtures of “high” concentrations were evaluated by calculation of the Combination Index. The applied “high” concentrations decreased mitochondrial activity in case of ENN B to 44 %, of DON to 48%, of NIV to 46% and of ZEA to 57% when tested alone in the initial dose–response experiments. All combinations of these concentrations (ENN B + DON, ENN B + NIV, ENN B + ZEN, DON + NIV, DON + ZEN and NIV + ZEN) showed antagonistic effects on cytotoxicity in Caco2 cells. For mixtures containing NIV even strong antagonistic effects were indicated, according to the categorization by Chou (2006). Our data are in line with the results of a previous study on combinatory effects of DON, NIV and ZEN, although Wan et al. (2013) described effects as just being non-additive and a precise calculation of synergism or antagonism was not conducted (Wan et al., 2013). Another study also found antagonistic effects between DON and ZEN (Bensassi et al., 2014). Interestingly, the combination of DON and NIV has previously been described as being synergistic in Caco2-cells after exposure for 48 h, or in IPEC-1 cells after 24 h of exposure (Alassane-Kpembi et al., 2013, 2015). Thus, combinatory effects may be very specific to the system in which they are investigated, concerning the chosen cell model, duration of incubation and probably also applied concentration as our data suggest. To the best of our knowledge, cytotoxic effects of ENN B in combination with DON, NIV, ZEN, AURO or TeA have not been tested so far. Binary mixtures of ENN B with DON, NIV or ZEN in “high” concentrations showed antagonistic effects. ENN B is known to form cation-selective channels, thereby disturbing the integrity of membranes, which is thought to contribute to its cytotoxicity. It has also been described that ABC-transporters, which play a key role as cellular efflux pumps for xenobiotics in phase III metabolism, may influence the toxicity of ENN B. Specifically, ABCB1 and ABCG2 were found to decrease the cytotoxicity of ENN B in chemoresistent sublines of human cell lines, suggesting that ENN B may be exported out of the cell by these two transporters (Dornetshuber et al., 2009). A mechanism like this is vulnerable to be influenced by the presence of other xenobiotics, which may bind to the same ABC-transporters or regulate their expression, leading to interactive effects with ENN B. Several studies indicate that the pregnane X receptor (PXR) not only induces the expression of phase I and II enzymes, but also of some phase III transporters like ABCB1 and ABCG2 (Hariparsad et al., 2009; Lemmen et al., 2013). ZEN, interestingly, has been described to activate the PXR and up-regulate ABCB1 and ABCG2 (Ding et al., 2006; Koraichi et al., 2013). Our measurements of the combination of ENN B and ZEN in high concentrations showed antagonistic effects. As a potential underlying mechanism it might be speculated that the activation of PXR by ZEN leads to the up-regulation of ABCB1 and ABCG2 and consequently to increased export of ENN B out of the cell, which finally might result in this antagonistic interaction between ENN B and ZEN concerning cytotoxicity. Similar mechanisms may potentially be involved in all other binary mixtures in

7

“high” concentrations with ENN B showing non-additive effects like ENN B + DON, ENN B + NIV and ENN B + TeA. However, there are no data available so far, suggesting that ABC transporters interact with DON, NIV or TeA. On the other hand, it is well known that the trichothecenes DON and NIV inhibit protein synthesis (Pestka, 2010; Tatsuno, 1968) and as already discussed in context with similar effects of TeA, this inhibition may at least contribute to antagonistic interactions between trichothecenes and other mycotoxins, as described here. 4.3. General aspects and potential relevance for risk assessment of mycotoxin mixtures The selection of the tested mycotoxins in this study was based on profound knowledge on typical contamination pattern in foodstuffs from temperate climate regions gained through the analysis of thousands of food samples during the last decade (Streit et al., 2013; Uhlig et al., 2013). Evaluation of cytotoxic effects of mycotoxins, individually and in combinations was based on measurements of mitochondrial activity in Caco-2 cells as a simplified model system for the human intestine. The cells have been incubated with the toxins for 24 h, which reflects a realistic estimate of the human gut being exposed to mycotoxins by food consumption. Therefore, this study design represents a scenario of exposure as realistic as possible. Interestingly and in contrast with studies utilizing different toxin combinations, time points or cell types (Alassane-Kpembi et al., 2013, 2015; Bensassi et al., 2014; Wan et al., 2013), our results indicate only additive or even antagonistic effects. No case of synergism was found in any setting. The Alternaria toxin TeA, of which combinations have been tested in more detail, was even found to reduce toxicity in combination with NIV and especially DON. Although in vitro data must always be interpreted with caution and the experimental settings greatly influence the outcome of a certain experiment, this might indicate a more positive co-exposure scenario for consumer safety than generally assumed in the past. However, in vivo studies on combinatory effects are urgently required to support eventual conclusions for risk assessment. In order to reveal a more detailed understanding of combinatory effects of mycotoxins, research on potentially involved metabolites and on combinations of more than two mycotoxins is needed in the future. Besides, it will be indispensable to focus investigations on more specific mechanisms, which are known to be involved in the manifestation of the toxicity of single mycotoxins. Evaluation of combinatory effects on genotoxicity or the induction of oxidative stress for instance, may reveal different interactions between the investigated mycotoxins, compared to the cytotoxicity screening performed in this study. Even potential synergistic effects of the toxins on other mechanisms cannot be excluded so far. 5. Conclusion and outlook This study gives an overview on combinatory effects on cytotoxicity of some naturally co-occurring mycotoxins. We demonstrated that highly concentrated binary combinations of selected mycotoxins by trend exhibited antagonistic effects in Caco-2 cells after 24 h of incubation. These findings are substantial, especially in the case of TeA, which is frequently found in comparably high concentrations in contaminated products. Such tests on combinatory effects of binary mixtures are just the first step towards the analysis of more complex mixtures, which reflect a notably realistic exposure scenario. In case such antagonistic effects may be verified by mechanistic and metabolic elucidation and likewise by in vivo studies, the question may arise whether cocontaminated food represent a lower risk than formerly thought.

8

K. Vejdovszky et al. / Toxicology Letters 241 (2016) 1–8

However, since differing combinatory interactions, also synergism, of these mycotoxins at more specific mechanisms of action cannot be excluded so far, further research in this field is urgently needed. Conflict of interest The authors declare to have no conflict of interest. Acknowledgement This study was funded by the University of Vienna. References Abia, W.A., Warth, B., Sulyok, M., Krska, R., Tchana, A., Njobeh, P.B., Turner, P.C., Kouanfack, C., Eyongetah, M., Dutton, M., Moundipa, P.F., 2013. Bio-monitoring of mycotoxin exposure in Cameroon using a urinary multi-biomarker approach. Food Chem. Toxicol. 62, 927–934. Abid-Essefi, S., Baudrimont, I., Hassen, W., Ouanes, Z., Mobio, T.A., Anane, R., Creppy, E.E., Bacha, H., 2003. DNA fragmentation, apoptosis and cell cycle arrest induced by zearalenone in cultured DOK, Vero and Caco-2 cells: prevention by Vitamin E. Toxicology 192, 237–248. Abid-Essefi, S., Ouanes, Z., Hassen, W., Baudrimont, I., Creppy, E., Bacha, H., 2004. Cytotoxicity, inhibition of DNA and protein syntheses and oxidative damage in cultured cells exposed to zearalenone. Toxicol. In Vitro 18, 467–474. Alassane-Kpembi, I., Kolf-Clauw, M., Gauthier, T., Abrami, R., Abiola, F.A., Oswald, I.P., Puel, O., 2013. New insights into mycotoxin mixtures: the toxicity of low doses of Type B trichothecenes on intestinal epithelial cells is synergistic. Toxicol. Appl. Pharmacol. 272, 191–198. Alassane-Kpembi, I., Puel, O., Oswald, I.P., 2015. Toxicological interactions between the mycotoxins deoxynivalenol, nivalenol and their acetylated derivatives in intestinal epithelial cells. Arch Toxicol. 89, 1337–1346. Asam, S., Lichtenegger, M., Liu, Y., Rychlik, M., 2012. Content of the Alternaria mycotoxin tenuazonic acid in food commodities determined by a stable isotope dilution assay. Mycotoxin Res. 28, 9–15. Bensassi, F., Gallerne, C., Sharaf el Dein, O., Hajlaoui, M.R., Lemaire, C., Bacha, H., 2014. In vitro investigation of toxicological interactions between the fusariotoxins deoxynivalenol and zearalenone. Toxicon 84, 1–6. Chou, T.C., Talalay, P., 1983. Analysis of combined drug effects – a new look at a very old problem. Trends Pharmacol. Sci. 4, 450–454. Chou, T.C., Talalay, P., 1984. Quantitative analysis of dose–effect relationships: the combined effects of multiple drugs or enzyme inhibitors. Adv. Enzyme Regul. 22, 27–55. Chou, T.C., 1975. General procedure for determination of median-effect doses by a double logarithmic transformation of dose–response relationships. Fed. Proc. 34, 228. Chou, T.C., 2006. Theoretical basis, experimental design, and computerized simulation of synergism and antagonism in drug combination studies. Pharmacol. Rev. 58, 621–681. Ding, X., Lichti, K., Staudinger, J.L., 2006. The mycoestrogen zearalenone induces CYP3A through activation of the pregnane X receptor. Toxicol. Sci. 91, 448–455. Domijan, A.M., Peraica, M., Cvjetkovic, B., Turcin, S., Jurjevic, Z., Ivic, D., 2005. Mould contamination and co-occurrence of mycotoxins in maize grain in Croatia. Acta Pharm. 55, 349–356. Dornetshuber, R., Heffeter, P., Sulyok, M., Schumacher, R., Chiba, P., Kopp, S., Koellensperger, G., Micksche, M., Lemmens-Gruber, R., Berger, W., 2009. Interactions between ABC-transport proteins and the secondary Fusarium metabolites enniatin and beauvericin. Mol. Nutr. Food Res. 53, 904–920. Dvorska, J.E., Surai, P.F., Speake, B.K., Sparks, N.H.C., 2002. Antioxidant systems of the developing quail embryo are compromised by mycotoxin aurofusarin. Comp. Biochem. Phys. C 131, 197–205. EFSA, 2004. Opinion of the Scientific Panel on contaminants in the food chain [CONTAM] related to deoxynivalenol (DON) as undesirable substance in animal feed. EFSA J. 73, 1–42. EFSA, 2011a. Scientific ppinion on the risks for animal and public health related to the presence of Alternaria toxins in feed and food. EFSA J. 9 (10), 2407–2504. EFSA, 2011b. Scientific opinion on the risks for public health related to the presence of zearalenone in food. EFSA J. 9 (6), 2197 124 pp.. EFSA, 2013. Scientific opinion on risks for animal and public health related to the presence of nivalenol in food and feed. EFSA J. 11 (6), 3262 119 pp.. EFSA, 2014. Scientific opinion on the risks to human and animal health related to the presence of beauvericin and enniatins in food and feed. EFSA J. 12 (8):3802) 174 pp.. Ezekiel, C.N., Sulyok, M., Warth, B., Krska, R., 2012. Multi-microbial metabolites in fonio millet (acha) and sesame seeds in Plateau State, Nigeria. Eur. Food Res. Technol. 235, 285–293.

Ezekiel, C.N., Sulyok, M., Frisvad, J.C., Somorin, Y.M., Warth, B., Houbraken, J., Samson, R.A., Krska, R., Odebode, A.C., 2013. Fungal and mycotoxin assessment of dried edible mushroom in Nigeria. Int. J. Food Microbiol. 162, 231–236. Griffin, G.F., Chu, F.S., 1983. Toxicity of the Alternaria metabolites alternariol, alternariol methyl ether, altenuene, and tenuazonic acid in the chicken embryo assay. Appl. Environ. Microbiol. 46, 1420–1422. Groten, J.P., Cassee, F.R., Van Bladeren, P.J., De Rosa, C.T., Feron, V.J., Sühnel, J., 2013. Mischungen chemischer Stoffe. In: Marquardt, H., Schäfer, S.G., Barth, H. (Eds.), Toxikologie, vol. 3. Wissenschaftliche Verlagsgesellschaft, Stuttgart, pp. 283– 286. Hariparsad, N., Chu, X., Yabut, J., Labhart, P., Hartley, D.P., Dai, X., Evers, R., 2009. Identification of pregnane-X receptor target genes and coactivator and corepressor binding to promoter elements in human hepatocytes. Nucleic Acids Res. 37, 1160–1173. Ivanova, L., Egge-Jacobsen, W.M., Solhaug, A., Thoen, E., Faeste, C.K., 2012. Lysosomes as a possible target of enniatin B-induced toxicity in Caco-2 cells. Chem. Res. Toxicol. 25, 1662–1674. Kamyar, M., Rawnduzi, P., Studenik, C.R., Kouri, K., Lemmens-Gruber, R., 2004. Investigation of the electrophysiological properties of enniatins. Arch. Biochem. Biophys. 429, 215–223. Koraichi, F., Inoubli, L., Lakhdari, N., Meunier, L., Vega, A., Mauduit, C., Benahmed, M., Prouillac, C., Lecoeur, S., 2013. Neonatal exposure to zearalenone induces long term modulation of ABC transporter expression in testis. Toxicology 310, 29–38. Lemmen, J., Tozakidis, I.E.P., Galla, H.-J., 2013. Pregnane X receptor upregulates ABCtransporter Abcg2 and Abcb1 at the blood-brain barrier. Brain Res. 1491, 1–13. Minervini, F., Fornelli, F., Flynn, K.M., 2004. Toxicity and apoptosis induced by the mycotoxins nivalenol, deoxynivalenol and fumonisin B1 in a human erythroleukemia cell line. Toxicol. In Vitro 18, 21–28. Pero, R.W., Posner, H., Blois, M., Harvan, D., Spalding, J.W., 1973. Toxicity of metabolites produced by the Alternaria. Environ. Health Perspect. 4, 87–94. Pestka, J.J., 2010. Deoxynivalenol: mechanisms of action, human exposure, and toxicological relevance. Arch. Toxicol. 84, 663–679. Rocha, O., Ansari, K., Doohan, F.M., 2005. Effects of trichothecene mycotoxins on eukaryotic cells: a review. Food Addit. Contam. 22, 369–378. Schrader, T.J., Cherry, W., Soper, K., Langlois, I., Vijay, H.M., 2001. Examination of Alternaria alternata mutagenicity and effects of nitrosylation using the Ames Salmonella test. Teratog. Carcinog. Mutagen. 21, 261–274. Schwarz, C., Kreutzer, M., Marko, D., 2012. Minor contribution of alternariol, alternariol monomethyl ether and tenuazonic acid to the genotoxic properties of extracts from Alternaria alternata infested rice. Toxicol. Lett. 214, 46–52. Scott, P.M., Stoltz, D.R., 1980. Mutagens produced by Alternaria alternata. Mutat. Res. 78, 33–40. Shier, W.T., Shier, A.C., Xie, W., Mirocha, C.J., 2001. Structure–activity relationships for human estrogenic activity in zearalenone mycotoxins. Toxicon 39, 1435– 1438. Shigeura, H.T., Gordon, C.N., 1963. The biological activity of tenuazonic acid. Biochemistry 2, 1132–1137. Solfrizzo, M., De Girolamo, A., Vitti, C., Visconti, A., van den Bulk, R., 2004. Liquid chromatographic determination of Alternaria toxins in carrots. J. AOAC Int. 87, 101–106. Streit, E., Schwab, C., Sulyok, M., Naehrer, K., Krska, R., Schatzmayr, G., 2013. Multimycotoxin screening reveals the occurrence of 139 different secondary metabolites in feed and feed ingredients. Toxins (Basel) 5, 504–523. Tatsuno, T., 1968. Toxicologic research on substances from Fusarium Nivale. Cancer Res. 28, 2393. Uhlig, S., Eriksen, G.S., Hofgaard, I.S., Krska, R., Beltran, E., Sulyok, M., 2013. Faces of a changing climate: semi-quantitative multi-mycotoxin analysis of grain grown in exceptional climatic conditions in Norway. Toxins (Basel) 5, 1682–1697. Wan, L.Y., Turner, P.C., El-Nezami, H., 2013. Individual and combined cytotoxic effects of Fusarium toxins (deoxynivalenol, nivalenol, zearalenone and fumonisins B1) on swine jejunal epithelial cells. Food Chem. Toxicol. 57, 276–283. Warth, B., Parich, A., Atehnkeng, J., Bandyopadhyay, R., Schuhmacher, R., Sulyok, M., Krska, R., 2012. Quantitation of mycotoxins in food and feed from Burkina Faso and Mozambique using a modern LC-MS/MS multitoxin method. J. Agric. Food Chem. 60, 9352–9363. Warth, B., Sulyok, M., Krska, R., 2013. LC-MS/MS-based multibiomarker approaches for the assessment of human exposure to mycotoxins. Anal. Bioanal. Chem. 405, 5687–5695. Yang, W., Yu, M., Fu, J., Bao, W., Wang, D., Hao, L., Yao, P., Nussler, A.K., Yan, H., Liu, L., 2014. Deoxynivalenol induced oxidative stress and genotoxicity in human peripheral blood lymphocytes. Food Chem. Toxicol. 64, 383–396. Yekeler, H., Bitmis, K., Ozcelik, N., Doymaz, M.Z., Calta, M., 2001. Analysis of toxic effects of Alternaria toxins on esophagus of mice by light and electron microscopy. Toxicol. Pathol. 29, 492–497. Zhou, B., Qiang, S., 2008. Environmental, genetic and cellular toxicity of tenuazonic acid isolated from Alternaria alternata. Afr. J. Biotechnol. 7, 1151–1156.