Porcine intestinal epithelial barrier disruption by the Fusarium mycotoxins deoxynivalenol and T-2 toxin promotes transepithelial passage of doxycycline and paromomycin

Background The gastrointestinal tract is the first target for the potentially harmful effects of mycotoxins after intake of mycotoxin contaminated food or feed. With deoxynivalenol (DON), T-2 toxin (T-2), fumonisin B1 (FB1) and zearalenone (ZEA) being important Fusarium toxins in the northern hemisphere, this study aimed to investigate in vitro the toxic effect of these mycotoxins on intestinal porcine epithelial cells derived from the jejunum (IPEC-J2 cells). Viability of IPEC-J2 cells as well as the proportion of apoptotic and necrotic IPEC-J2 cells was determined by flow cytometry after 72 h of exposure to the toxins. Correlatively, the integrity of the intestinal epithelial cell monolayer was studied using Transwell® inserts, in which the trans-epithelial electrical resistance (TEER) and passage of the antibiotics doxycycline and paromomycin were used as endpoints. Results We demonstrated that the percentage of Annexin-V-FITC and PI negative (viable) cells, Annexin-V-FITC positive and PI negative (apoptotic) cells and Annexin-V-FITC and PI positive (necrotic) IPEC-J2 cells showed a mycotoxin concentration-dependent relationship with T-2 toxin being the most toxic. Moreover, the ratio between Annexin-V-FITC positive and PI negative cells and Annexin-V-FITC and PI positive cells varied depending on the type of toxin. More Annexin-V-FITC and PI positive cells could be found after treatment with T-2 toxin, while more Annexin-V-FITC positive and PI negative cells were found after exposure to DON. Consistent with the cytotoxicity results, both DON and T-2 decreased TEER and increased cellular permeability to doxycycline and paromomycin in a time- and concentration-dependent manner. Conclusions It was concluded that Fusarium mycotoxins may severely disturb the intestinal epithelial barrier and promote passage of antibiotics.


Background
Mycotoxins are naturally occurring secondary metabolites produced by fungi. They can be formed in the field on crops as well as during storage and are chemically very stable, enabling them to survive processing, and thereby end up in the feed and food chain. In the northern temperate regions Fusarium species are the most prevalent toxigenic fungi [1,2] with trichothecenes, fumonisins and zearalenone being the most important toxins produced [3]. If ingested, these mycotoxins can cause a variety of adverse health effects, both on humans and animals [4]. Since the gastrointestinal epithelium is the first barrier to come in contact with mycotoxins after ingestion, several studies investigated the effects of mycotoxins on intestinal cells. Most of these were conducted on human intestinal epithelial Caco-2 cells [5][6][7][8][9][10][11][12][13][14][15][16], but also the porcine intestinal epithelial cell lines IPEC-1 [17] and IPEC-J2 [18][19][20] have been used to demonstrate that mycotoxins can alter the integrity of the gut barrier. Moreover, Pinton et al. [12,13] showed that the trichothecene deoxynivalenol (DON) selectively affects the expression of tight junction proteins in porcine intestinal epithelial cell monolayers (IPEC-1) and Caco-2 cell monolayers resulting in increased paracellular passage of fluorescein isothiocyanate (FITC)dextran and Escherichia coli. It was also demonstrated that fumonisin B 1 (FB1) affects the membrane composition of IPEC-1 cells resulting in an increased passage of this toxin [21]. Intestinal epithelial cells are thus important targets for the toxic effects of mycotoxins and it is clear that an altered barrier could result in an altered passage of mycotoxin co-contaminants, xenobiotics and pathogens.
The aim of the present study was to evaluate the cytotoxic effect of the four most important Fusarium mycotoxins i.e. DON, T-2 toxin (T-2), zearalenone (ZEA) and FB1 in intestinal epithelial cells derived from the jejunum of pigs (IPEC-J2 cells). Therefore the percentage viable, apoptotic and necrotic IPEC-J2 cells was determined using flow cytometry. Correlatively, it was investigated whether differences in cell viability result in a changed transepithelial passage of the antibiotics doxycycline and paromomycin. This was studied using Transwell W devices which are commonly applied and widely accepted to perform permeability studies. The hypothesis tested in this study was that an altered integrity of the gut barrier provoked by mycotoxins, could result in an increased passage of antibiotics through the epithelium. Increased passage of these antibiotics can indeed have toxic consequences for the animal, withdrawal time, and consequently public health with respect to possible residues of drugs in edible tissues. To our knowledge, this is the first time that transport of antibiotics is used as a marker to study possible damage of mycotoxins to intestinal epithelial cells.

Chemicals
DON, T-2, FB1 and ZEA (Sigma-Aldrich, Bornem, Belgium) stock solutions of 2000 μg/ml were prepared in methanol for DON, ethanol for T-2 and acetonitrile for FB1 and ZEA (Merck, Darmstadt, Germany). Doxycycline and paromomycin stock solutions (Sigma-Aldrich, 2000 μg/ml) were prepared in high-performance liquid chromatography (HPLC) grade water. All stock solutions were stored at −20°C. Serial dilutions were prepared in culture medium in order to obtain non-cytotoxic concentrations of organic solvents and antibiotics, and allowing the addition of a similar volume of vehicle in all experiments.
Determination of non-cytotoxic concentrations of the organic solvents IPEC-J2 cells were seeded into a 96-well plate at a concentration of 1.5 × 10 5 cells/ml in a total volume of 200 μl. The cells were maintained in an atmosphere of 5% CO 2 at 37°C. After overnight incubation, the cells were treated for 72 h with a final concentration of 1, 5, 10 or 20% of the organic solvents (methanol, ethanol or acetonitrile) in the total volume of culture medium. Next, to assess cytotoxicity, 10 μl of a water soluble tetrazolium salt (2-(4-iodophenyl)-3-(4-nitrophenyl)-5-(2,4-disulfophenyl)-2H-tetrazolium) (WST-1) was added to each well and the plate was incubated at 37°C for an additional 3 h. The absorbance was determined at 450 nm using a microplate ELISA reader (Multiscan MS, Thermo Labsystems, Helsinki, Finland). Each experiment was conducted in triplicate. The percentage of viable cells was calculated using the following formula: In this formula a = OD 450nm derived from the wells incubated with solvent, b = OD 450nm derived from blank wells, c = OD 450nm derived from untreated control wells.

Assessment of cytotoxicity after exposure of IPEC-J2 cells to the investigated antibiotics doxycycline and paromomycin
A flow cytometric technique was used to assess the direct cytotoxic effect of the antibiotics doxycycline and paromomycin. IPEC-J2 cells were seeded at a concentration of 5 × 10 5 cells/ml in 24-well plates in 1 ml of culture medium and allowed to grow for 24 h. Next, cells were treated for 24 h with 0, 5, 10, 20 and 40 μg/ml of doxycycline or paromomycin. Each experiment was conducted in triplicate. After the incubation period, cells were trypsinized, culture medium and cells were collected, centrifuged for 5 min at 524 × g and 4°C and resuspended in 0.5 ml Hank's Buffered Salt Solution (HBSS, Gibco). This procedure was repeated three times in order to remove cellular debris and antibiotics. Cell death was assessed using dual staining with Annexin-Vfluorescein isothiocyanate (Annexin-V-FITC, Roche Diagnostics, Belgium) and propidium iodide (PI, Sigma-Aldrich) which allows the discrimination of viable cells (FITC -/PI -), apoptotic (FITC + /PI -) and necrotic cells (FITC + /PI + ). Cells were incubated for 20 min in the dark on ice with 100 μl of a solution containing 20 μl Annexin-V-FITC and 20 μl PI dissolved in 1000 μl incubation buffer containing 10 mM 4-(2-hydroxyethyl)-1piperazineethanesulfonic acid (HEPES), 140 mM NaCl and 5 mM CaCl 2 . Cells were assessed by a FACSCanto flow cytometer (Becton, Dickinson and Company, Erembodegem, Belgium) and the percentage of viable, apoptotic and necrotic cells was obtained using the FACSDiva Software (Becton, Dickinson and Company, Franklin Lakes, NJ, USA).

Assessment of cytotoxicity after exposure of IPEC-J2 cells to Fusarium toxins
The flow cytometric technique as described above was also used to assess the direct cytotoxic effect of mycotoxins. IPEC-J2 cells were seeded at a concentration of 5 × 10 5 cells/ml in 24-well plates in 1 ml of culture medium and allowed to grow for 24 h. Next, cells were treated for 72 h with increasing concentrations of T-2 (0-10 ng/ml), DON (0-10 μg/ml), ZEA (0-20 μg/ml) and FB1 (0-15 μg/ml). The molar concentration corresponding with 1 μg/ml DON, FB1 and ZEA and 1 ng/ml T-2 is 3.37 μM, 1.39 μM, 3.14 μM and 2.14 nM, respectively. Each experiment was conducted in triplicate. After the incubation period, cells were processed as described above before analysis on the FACSCanto flow cytometer. The concentration corresponding to 50% reduction in viable cells (IC50) was calculated with linear regression curves. A positive control was provided by cells single stained with Annexin-V-FITC and cells single stained with PI, also used to determine spectral overlap, after treatment for 72 h with 10 μg/ml DON. Furthermore, unstained cells (no Annexin-V-FITC and no PI) were also included to determine autofluorescence of IPEC-J2 cells.

Assessment of TEER after exposure of IPEC-J2 cells to DON and T-2
The effect of DON and T-2 on barrier integrity was determined by measuring the transepithelial electrical resistance (TEER) using a Millicell Electrical Resistance System (Millipore, World Precision Instruments, Sarasota, FL, USA). Cells were grown and differentiated on Transwell W collagen-coated polytetrafluoroethylene (PFTE) membrane inserts (0.4 μm pore diameter, 6.5 mm diameter, Corning Inc., NY, USA) at a density of 1.5 × 10 5 cells/ml. Culture medium was replaced three times a week and IPEC-J2 cells were used for TEER experiments at day 21 post-seeding. Next, cells were incubated for 72 h with different concentrations of T-2 (0-100 ng/ ml) or DON (0-10 μg/ml). TEER was measured every 24 h from the start of the toxin exposure until 3 days. TEER values were expressed as kΩ × cm 2 . Each single experiment was conducted in threefold.

Passage of doxycycline and paromomycin after exposure of IPEC-J2 cells to DON and T-2
Cells were seeded on Transwell W collagen-coated PFTE membrane inserts as described above for the TEER experiment. Only cells which are cultured for 21 days show stable TEER values indicating cells have reached confluence and form a tight monolayer (data not shown). Next, cells were incubated for 72 h with different concentrations of T-2 (0-100 ng/ml) or DON (0-10 μg/ml), during which the last 24 h a non-cytotoxic concentration of both paromomycin (30 μg/ml) and doxycycline (10 μg/ml) was added to the apical (AP) compartment of the transwell. To evaluate the permeability of the monolayers, 50 μl of transport medium was withdrawn from the basolateral (BL) compartment at time 0 (before) and at 0.5, 1, 2, 3, 4, 6, 8, 12 and 24 h after addition of the drugs and replaced immediately by 50 μl of fresh culture medium. Each experiment was conducted in triplicate.
Acceptor media were analyzed for the paromomycin content by high-performance liquid chromatography tandem mass spectrometry (LC-MS/MS). The LC-MS/ MS system consisted of a Surveyor Autosampler Plus and MS Pump Plus, coupled to a TSQ Quantum Ultra W triple quadrupole mass spectrometer (all from Thermo-Fisher Scientific, Zellik, Belgium). Chromatographic separation was performed on a PLRP-S column (150 mm x 2.1 mm i.d., 100 Å, dp: 5 μm, Varian, Middelburg, The Netherlands) using a gradient programme (flow-rate: 200 μl/min). The mobile phase consisted of (A) 20 mM pentafluoropropionic acid (PFPA) (Sigma-Aldrich) in water and (B) 20 mM PFPA in water/acetonitrile (50/50, v/v) (VWR International, Leuven, Belgium). The MS instrument was operated in the selected reaction monitoring (SRM) mode, using two precursor ion > product ion transitions. Tobramycin (Sigma-Aldrich) was used as internal standard (IS). Samples were prepared by pipetting 50 μl culture medium into an Eppendorf cup (Novolab, Geraardsbergen, Belgium). Each sample was spiked with 10 μl of the IS working solution of 1 μg/ml in HPLC grade water (VWR International) and diluted with 65 μl methanol, 100 μl HPLC grade water and 25 μl 200 mM PFPA. After vortexing, samples were centrifuged at 7 800 × g for 10 min to remove possible solid particles. The sample was transferred into screw-capped conical vials and 5 μl were injected onto the LC-MS/MS system. Doxycycline concentrations were determined using the same LC-MS/MS configuration and HPLC column as for paromomycin analysis. The mobile phase consisted of (A) 0.5% acetic acid in water and (B) acetonitrile. A gradient programme was applied and the flow-rate was 200 μl/min. The MS instrument was operated in the SRM mode, using two precursor ion > product ion transitions. Demethylchlortetracycline (DMCTC, Sigma-Aldrich) was used as IS. Samples were prepared by pipetting 25 μl culture medium into an Eppendorf cup. Each sample was spiked with 12.5 μl of the IS working solution of 10 μg/ml in HPLC grade water, followed by the addition of 75 μl 0.5% formic acid in HPLC grade water. After vortexing briefly and centrifugation, 5 μl were injected onto the LC-MS/MS system.
The limit of quantification (LOQ) was 10 ng/ml for both doxycycline and paromomycin. A detailed description of both methods can be found in Goossens et al. [23].

Statistical analysis
Data were analyzed using one-way analysis of variance (ANOVA) to address the significance of difference between mean values with significance set at P < 0.05.

Results
A concentration of 1% ethanol, 5% methanol or 5% acetonitrile of the total volume of culture medium was determined to be non-cytotoxic for IPEC-J2 cells Figure 1 represents the percentage of viable cells after exposure for 72 h to 1, 5, 10 or 20% ethanol, methanol or acetonitrile in the total volume of culture medium.
A final concentration of 5% acetonitrile, 5% methanol and 1% ethanol was accepted to use in the experiment as the viability of IPEC-J2 cells was on average > 95%. Non-cytotoxic concentrations of the solvents were ensured by appropriate dilution of the toxins stock solutions in culture medium.
A concentration of 10 μg/ml doxycycline and 30 μg/ml paromomycin was determined to be non-cytotoxic for IPEC-J2 cells Figure 2 represents the percentage of viable cells after exposure for 24 h to 5, 10, 20 or 40 μg/ml doxycycline (a) or paromomycin (b), respectively.
A concentration up to 40 μg/ml doxycycline or paromomycin caused no significant reduction in cell viability of IPEC-J2 cells. It was concluded that a concentration of 10 μg/ml doxycycline and 30 μg/ml paromomycin could be used in the experiments. The concentration of the antibiotics was based on, on the one hand, the dose of the antibiotic that is usually administered to pigs, and on the other hand, the capacity of the intestine of an adult pig. When 10 mg/kg body weight doxycycline and 40 mg/kg body weight paromomycin is administered to a pig of 20 kg, and this dose ends up in an intestinal volume of 27.5 l, this corresponds with a concentration of about 8 and 29 μg/ml respectively. Exposure of IPEC-J2 cells for 72 h to different mycotoxin concentrations showed a concentration-dependent toxicity. T-2 was clearly the most toxic one with a significantly lower percentage of Annexin-V-FITC and PI negative cells in the ng/ml range compared to DON and ZEA which were toxic in the μg/ml range.    In both the T-2 and DON experiment, increased passage of doxycycline and paromomycin was seen after incubation of IPEC-J2 cells with cytotoxic concentrations of the tested mycotoxins.

Discussion
The gastrointestinal tract can be exposed to high concentrations of mycotoxins after ingestion of mycotoxin contaminated food or feed. Since swine are the most sensitive domestic animal species to mycotoxins [24], this study aimed to investigate the toxic effect of four common Fusarium toxins, T-2, DON, ZEA and FB1 on IPEC-J2 cells. Studies concerning the effects of mycotoxins on IPEC-J2 cells are limited. The concentrations of the mycotoxins tested in this study are the result of several optimization experiments performed to refine the concentration range to be used for these in vitro experiments. According to Sergent et al. [14] DON concentrations in human intestine are estimated between 0.16 μg/ ml and 2 μg/ml depending on the contamination level of the feed, including thus our tested range. Mycotoxin concentrations in the test solutions are not based on levels known to cause problems in pigs, but are based on the cytotoxicity results of the flow cytometer experiments, the solubility of the mycotoxin in aqueous solvents, and cost of the mycotoxin.

cells and Annexin-V-FITC and PI positive (necrotic) cells.
After 72 h of exposure with T-2, DON, FB1 or ZEA, T-2 clearly was the most cytotoxic with a toxicity in the ng/ml range, while the other mycotoxins are only toxic at concentrations in the μg/ml range. This finding is in accordance with the study of Calvert et al. [7], who demonstrated that T-2 was most cytotoxic against three human cell lines tested as determined by the MTT assay.
When comparing the percentage Annexin-V-FITC positive and PI negative cells and Annexin-V-FITC and PI positive cells, it is striking that more Annexin-V-FITC and PI positive cells than Annexin-V-FITC positive and PI negative cells were found after treatment of the cells with T-2. This corresponds, however, with findings in practice where T-2, present in feed, causes skin necrosis [25,26]. Also necrosis of the lymphoid cells of the intestinal mucosa was seen after intake of T-2 contaminated feed [25,27]. DON, on the other hand, causes more Annexin-V-FITC positive and PI negative cells than Annexin-V-FITC and PI positive cells. The discrepancy between type A (T-2) and type B (DON) trichothecenes was also demonstrated by Nasri et al. [28] who showed that DON causes rather apoptosis and T-2 rather necrosis in Jurkat T-lymphocytes. As it is known that type B trichothecenes activate the MAPK signaling pathway resulting in apoptosis, while type A trichothecenes are not able to activate the MAPK families [29][30][31], the authors suggested this as an explanation for the results. For ZEA and FB1, although FB1 was not cytotoxic at the tested concentrations, an approximately equal percentage of apoptotic and necrotic cells were seen after treatment of IPEC-J2 cells for 72 h with those toxins. Our results suggest that different toxins may induce different types of cell death in IPEC-J2 cells. To our knowledge, this is the first report of DON and T-2 induced different mechanisms of cell death in IPEC-J2 cells. Since it was not the principal aim of this study to elaborate further the differences in the mode of cell death induced by different mycotoxins, no further experiments were conducted. Nevertheless, it could be interesting to investigate these differences in more detail, including the investigation of associated pathway molecules, the analysis of potential protective effects of caspase inhibitors and/or necrose inhibitors, etc. Also other types of cell death can be taken  Next, at day 0, cells were incubated for 72 h with different concentrations of DON (untreated inserts, inserts treated with 0.5, 1, 5 or 10 μg/ml), during which the last 24 h a non-cytotoxic concentration of doxycycline and paromomycin was added to the apical compartment (n = 3, mean + stdev). Significant differences in passage have been marked with * (P < 0.05), ** (P < 0.01) or *** (P < 0.001) according to the significance level.
into account since there are more than two ways for a cell to die. Cell death can be classified according to its morphological appearance (apoptotic, necrotic, autophagic or associated with mitosis), enzymological criteria (with and without the involvement of nucleases or distinct classes of proteases, such as caspases, calpains, cathepsins and transglutaminases), functional aspects (programmed or accidental, physiological or pathological) or immunological characteristics (immunogenic or non-immunogenic) [32]. It could be interesting exploring these differences in relation with mycotoxins. Correlatively with direct cytotoxicity, the consequences on barrier integrity (TEER and passage of drugs) were also assessed. The effect on TEER and passage was only studied for DON and T-2 because, of the four mycotoxins tested in this study, these are the most toxic ones and their ratio of Annexin-V-FITC positive and PI negative cells and Annexin-V-FITC and PI positive cells is antithetic. Notwithstanding the assertion of Geens and Niewold [33] that IPEC-J2 cells cultured on Transwell W -collagen coated inserts with a surface area of 0.33 cm 2 are not appropriate to measure TEER values, our IPEC-J2 monolayer showed TEER values above 10 kΩ cm 2 after 3 weeks of culturing. According to Fromter and Diamond [34], IPEC-J2 cell lines with TEER values of approximately 2 kΩ cm 2 can be considered to be tight epithelia. When IPEC-J2 cells were incubated with cytotoxic concentrations of DON (5 and 10 μg/ml) and T-2 (10 and 100 ng/ml) TEER values dropped to nearly 0 kΩ cm 2 after 24 h suggesting a complete disruption of the epithelial monolayer within 24 h. In contrast, TEER values of IPEC-J2 cells incubated with non-toxic concentrations (0.5 μg/ml DON and 0.1 and 1 ng/ml T-2) were comparable to the TEER of control cells. Treatment with 1 μg/ml DON on the other hand shows a time-dependent reduction of the TEER. This was also reported by Pinton et al. [12] who showed a time-and dose-dependent reduction of the TEER after exposure of IPEC-1 and Caco-2 cells to different concentrations (0, 3, 6, 15, 30 μg/ml) of DON. A decreased barrier function in vitro is reflected in both a decreased TEER and an increased tracer flux [17,35]. In our study, the decrease in TEER is related to the passage of the used tracers, doxycycline (molecular weight (mw) 513 Da) and paromomycin (mw 714 Da), two commonly used antibiotics in pig industry. The higher the toxin concentration, the higher the marker passage. Most studies use tracers such as FITC-dextran, while in this study and for the first time the antibiotics doxycycline and paromomycin were used as markers. We were, however, not capable of demonstrating that induction of more apoptosis or more necrosis results in differences in transepithelial passage of antibiotics. Both doxycycline and paromomycin pass the epithelial monolayer, regardless the induction of a different cell death mechanism.
It is very interesting to study the passage of these therapeutic compounds during in vivo trials, as veterinary drugs are often used in pig mass medication and frequently administered by the oral route, i.e. mixed in the feed or drinking water.

Conclusions
When discussing the relevance of the in vitro results for the in vivo situation it needs to be taken into account (a) (b) Figure 7 Passage of doxycycline (a) and paromomycin (b) across a differentiated IPEC-J2 cell monolayer, measured as the concentration in the basolateral compartment. Cells were grown for 21 days on Transwell W collagen-coated PTFE filters (0.33 cm 2 , 0.4 μm).
Next, at day 0, cells were incubated for 72 h with different concentrations of T-2 (untreated inserts, inserts treated with 0.1, 1, 10 or 100 ng/ml of T-2), during which the last 24 h a non-cytotoxic concentration of doxycycline and paromomycin was added to the apical compartment (n = 3, mean + stdev). Significant differences in passage have been marked with * (P < 0.05), ** (P < 0.01) or *** (P < 0.001) according to the significance level.
that enterocytes are not continuously exposed to mycotoxins. Rather a time-dependent fluctuation in mycotoxin concentrations exists in vivo, whereas in vitro cells are exposed continuously at constant concentrations. Nevertheless, the current in vitro findings warrant further investigation in vivo because an increased passage of antibiotics may have important implications for both animal and human health. Increased absorption of antibiotics may lead to higher plasma levels, including the risk of exceeding the toxic threshold. Moreover, increased antibiotic plasma levels may also lead to a tissue residue problem in slaughter animals and inherent food safety concerns. It could thus be interesting to investigate in vivo the effect of mycotoxins on the intestinal absorption of antibiotics.