Toxicology Letters 232 (2015) 106–112

Assessment of toxic potential of mycotoxin contaminated bread duringin vitro human digestion on human B lymphoid cell line

Linda Monaci *, Antonella Garbetta, Elisabetta De Angelis, Angelo Visconti,Fiorenza MinerviniInstitute of Sciences of Food Production, National Research Council of Italy (ISPA–CNR), Via Amendola 122/O, 70126, Bari, Italy

H I G H L I G H T S G R A P H I C A L A B S T R A C T

� Mycotoxin contaminated in vitrodigests were assayed on RPMI lym-phoid B cells.

� Findings proved that digestion islikely to affect the final levels ofHT-2 detected.

� A more persistent toxic activity wasrecorded in the later duodenal phase.

� Food matrix modulated HT-2 toxicitywhen assayed on lymphoid B cells.

� A different trend displayed betweentoxicological tests and mycotoxinlevels in digests.

A R T I C L E I N F O

Article history:Received 18 July 2014Received in revised form 25 September 2014Accepted 26 September 2014Available online 28 September 2014

Keywords:MycotoxinTrichothecenesHT-2 ?toxinIn vitro digestionToxicityLymphoid cell lineBioassay

A B S T R A C T

Ingestion of food is considered a major route of exposure to many contaminants including mycotoxins.The amount of mycotoxin resisting to the digestion process and potentially absorbable by the systemiccirculation is only a smaller part of that ingested. In vitro digestion models turn useful for evaluatingmycotoxins bioaccessibility during the intestinal transit and can be intended as a valuable tool for theassessment of mycotoxin bioavailability in food. In this paper we describe a study aimed at investigatingtoxicity of in vitro gastro-duodenal digests of mycotoxin contaminated bread collected along thedigestion time-course. Toxicity tests were carried out on a sensitive RPMI lymphoid B cell line chosen asthe most suitable lineage to assess toxicity retained by gastro-duodenal digests. In parallel, a chemicalquantification of T-2 and HT-2 toxins contaminating the bread digests was accomplished during thegastric and duodenal transit. The digestive fluids undergoing chemical and toxicological analysis werecollected at the beginning and end of gastric phase, and after completion of the duodenal phase. Resultsproved that a correlation between HT-2 content and toxicity did exist although a more persistent toxicactivity was displayed in the later stage of the duodenal phase. This persistent toxicity might be explainedby the co-occurrence of unknown HT-2-related conjugates or metabolites formed during digestion.

ã 2014 Elsevier Ireland Ltd. All rights reserved.

Contents lists available at ScienceDirect

Toxicology Letters

journal homepage: www.elsev ier .com/locate / tox let

* Corresponding author. Tel.: +39 080 5929343; fax: +39 080 5929374.E-mail address: linda.monaci@ispa.cnr.it (L. Monaci).

http://dx.doi.org/10.1016/j.toxlet.2014.09.0210378-4274/ã 2014 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Mycotoxin contamination of agricultural commodities repre-sents an important safety issue due to the toxicological implicationthat can occur on human health (Foroud and Eudes, 2009). In

L. Monaci et al. / Toxicology Letters 232 (2015) 106–112 107

particular, several species belonging to Fusarium are renowntrichothecenes producers. Trichothecene mycotoxins are classifiedinto subgroups with T-2/HT-2 toxins and deoxynivalenol (DON)being the main representatives of the type A and B trichothecenesrespectively. DON, T-2 and HT-2 mycotoxins are commonlyisolated from cereals such as wheat, rye, barley, oats and cornand other cereal based products due to the extreme stability shownupon food processing (Canady et al., 2001; Placinta et al., 1999;Sudakin, 2003). Trichotecenes are responsible for a wide variety oftoxic effects in humans and animals and in particular among thetype class A (Miller, 2002), T-2 and HT-2 toxins are considered themost acutely toxic compounds (EFSA, 2011). The effects of T-2 toxinon human and animal systems have been widely investigated anddeeply reviewed by EFSA (EFSA, 2011). In particular, T-2 toxin is awell described inhibitor of eukaryotic protein synthesis affecting,at higher doses, also that of RNA and DNA. It was reported to causelipid peroxidation affecting cell membrane integrity and apoptoticeffects in different tissues and cell culture models (Battilani et al.,2009; Rocha et al., 2005). In addition, acute toxic effects, likevomiting, leukocytosis, skin irritation, hemorrhage and immunemodulation after T-2 toxin exposure, have been also described(EFSA, 2011). Concerning HT-2 toxin, very little is known about itstoxicity. Data from the literature indicate that T-2- and HT-2 toxinsinduce adverse effects with similar potency (EFSA, 2011). Acuteexposure to trichothecenes resulted in severe damage to activelydividing cells in tissue such as bone marrow, lymph nodes, spleen,thymus and intestinal mucosa and produced their toxic effects bybinding to 60 s ribosomes and interrupting protein synthesis ineukaryotic cells (Yang et al., 2000). Lymphocytes are moresensitive to T-2 than other cultured cell lines correlating wellwith data from in vivo experiments showing that trichothecenesact as immunosuppressive agents (Gutleb et al., 2002).

Furthermore, T-2 is rapidly converted to HT-2 in the gut, then invivo toxicity of T-2 can be considered comparable to that of HT-2(WHO, 2001). The risk that T-2 and HT-2 toxins pose on human andanimals health, prompted EFSA Panel on contaminants in the foodchain to set a group tolerable daily intake (TDI) for the sum of thetwo toxins as low as 100 ng/kg body weight per day (EFSA, 2011) inorder to minimize the exposure in diet. More recently someindicative levels to be respected for the sum of the two toxins inraw and processed cereals have been clearly highlighted in the lastrecommendation issued by the European Commission within theEuropean Union (Commission Recommendation, 2013). Recom-mended levels range from 25 ng/g for bakery products up to1000 ng/g for unprocessed oat products.

In human health risk assessment, ingestion of food isconsidered a major route of exposure to many contaminantsincluding mycotoxins. The total amount of an ingested contami-nant does not always reflect the amount available to the bodybecause only a smaller part will be available for further absorption(Versantvoort et al., 2005). As a consequence, bioaccessibility,defined as the amount of contaminant released through thegastrointestinal tract from the food matrix and then potentiallyabsorbable, can be considered a measure for the assessment ofmycotoxin bioavailability in food. In vitro digestion models couldbe successfully utilized to address this issue. Several studies havebeen accomplished with this aim and static or dynamic digestivemodels have been developed and implemented to evaluatemycotoxins bioaccessibility during the intestinal transit (Versant-voort et al., 2005; Avantaggiato et al., 2003, 2004; Kabak et al.,2009). Studies were aimed at investigating the stability of severalmycotoxins by using a biochemical digestion model (Avantaggiatoet al., 2003, 2004, 2007; Dall’Erta et al., 2013; Raiola et al., 2012). Ina recent paper we investigated the fate of T-2 and HT-2 in in vitrodigests and calculated the relevant bioaccessibility (De Angeliset al., 2014b). Findings proved that digestion is likely to affect the

final amount of mycotoxins detected in gastro-duodenal digests ofcontaminated bread samples, with higher values found for HT-2 and lower for T-2 toxin. As a step forward, in the present paper weinvestigated immunotoxicity of the digestive fluids collected alongsimulated gastro-duodenal digestion of contaminated breadsamples on a human B lymphoid cell line. Such lymphoid cellline was selected since proved to be a more sensitive lineage to T-2 than T lymphoid lineage as already assessed in a previous study(Minervini et al., 2005). The use of this sensitive cellular lineageaimed to give more insights on the eventual toxic activity shown bythe mycotoxin-containing bread digests. Results indicate thattoxicity is retained along the whole digestion process with anapparent increase of the inhibition of cellular proliferationespecially referring to the last phase of gastro-duodenal digestion.This suggests that other molecules than the solely HT-2, probablyoriginating during duodenal digestion, might account for theincreased toxicity observed along the digestion time-course.

2. Materials and method

2.1. Materials

Solid standards of T-2 toxin (T-2) and HT-2 toxin (HT-2)(purity = 98%), ammonium acetate (MS grade) were purchasedfrom Sigma–Aldrich (Milan, Italy). Acetonitrile (CHROMASOLV1

Plus for HPLC, purity �99.9%) and methanol (CHROMASOLV1

gradient grade for HPLC, purity �99.9%) were obtained fromSigma–Aldrich (Milan, Italy) while formic acid (MS grade) andmethanol were provided by Fluka (Milan, Italy). C18 SPE columnswere purchased from Supelco (Discovery DSC-18 50 mg/ml,Supelco, Bellafonte PA, USA) and PTFE syringe filters (4 mm,0.2 mm) from Sartorius Stedim Biotech GmbH (Göettingen,Germany). Ultrapure water was produced by a Millipore Milli-Qsystem (Millipore, Bedford, MA, USA). All reagents for in vitrodigestion experiments (Sodium chloride-BioXtra �99.5%, Ammo-nium Bicarbonate, Calcium chloride, Bis-Tris, Phenylmethansu-fonyl fluoride (PMSF), Egg lechtin-PC, Sodium taurocholate andSodium glicodeoxycholate) as well as the enzymes (a-amylasefrom human saliva Type XIII-A, Pepsin from porcine gastric,Trypsin from porcine pancreas Type IX-S, a-chymotrypsin frombovine pancreas Type II and a-amylase from Bacillus sp. Type II-A)were purchased from Sigma–Aldrich (Milan, Italy).

RPMI 1640 medium, Trypsin-EDTA solution, L-glutamine200 mM, antibiotic and antimycotic solution and Thiazolyl BlueTetrazolium Bromide (MTT) were purchased from Sigma–Aldrich(Milan, Italy). Foetal bovine serum (FBS) was purchased from Gibco(Milan, Italy).

2.2. HPLC/MS analysis

2.2.1. EquipmentFor HPLC-high resolution-mass spectrometry (HR-MS) meas-

urements a system consisting of a (U)HPLC pump (AccelaTM,Thermo Fisher Scientific, San Jose, CA, USA) coupled with a highresolution mass spectrometer ExactiveTM (Thermo Fisher Scientif-ic, Bremen, Germany) through an HESI II interface was used.Mycotoxin separation was accomplished by gradient elution on aKinetexTM PentaFluoroPhenyl column (i.d. 2.1 mm � 100 mm;particle size 2.6 mm; porosity 100 Å; Phenomenex, CA, USA) byapplying the chromatographic conditions detailed elsewhere (DeAngelis et al., 2013). The volume injected was set at 20 ml and theflow-rate kept at 250 ml/min.

MS and HCD-MS (collision energy at 25 eV) analyses wereperformed in positive ion mode by setting two scan events for eachanalysis (full scan MS and full ion fragmentation) and the systemwas operated in positive ion mode at the maximum resolving

108 L. Monaci et al. / Toxicology Letters 232 (2015) 106–112

power (=100.000 FWHM). MS and HCD-MS scan range was in therange 70–1000 Th. Employing the mass correction function, a massaccuracy better than 0.5 ppm was attained.

2.3. Sample preparation

2.3.1. Mycotoxin standard solutionsIndividual stock solutions of T-2 and HT-2 were prepared by

dissolving the solid commercial toxin in acetonitrile to reach thefinal concentration of 100 mg/ml. Then a standard mixture of T-2and HT-2 toxins, at the final concentration of 10 mg/ml each, wasprepared by combining aliquots of both standard stock solutionswith appropriate volumes of acetonitrile. For toxicological experi-ments 10 mg/ml working standard solutions were preparedseparately for T-2 and HT-2 toxins by drying an appropriatealiquot of the respective stock solution under a gentle air streamand re-suspending it in an equivalent volume of methanol.

2.3.2. Production of mycotoxin free and naturally contaminated breadsamples

Bread model foods to be submitted to in vitro digestionexperiments were prepared at laboratory-scale according to arecipe standardized within the EU project DREAM (Design anddevelopment of realistic food models with well characterizedmicro- and macro-structure and composition) described elsewhere(De Angelis et al., 2013). Recipients were: 350 g of wholemeal wheatflour, 6.3 g of salt,10.5 of margarine, 5.25 g of dry yeast and 210 ml oftap water. Each loaf produced (of approximately 500 g weight) wasdried in a stove (55 �C) for approximately 40 h until a weightstabilization was observed. Each dried sample was milled coarselyand successively more finely using a sieve of 1 mm (Cyclone SampleMill, PBI International). Samples were stocked under vacuum atroom temperature until their use.

Bread samples contaminated with mycotoxins were producedby starting from wholemeal wheat flours found contaminated withT-2 and HT-2 at different levels. In total three different types ofbread samples were prepared (three loafs produced for each typeof bread): (A) mycotoxin-free bread (B) bread contaminated athigher level and (C) bread contaminated at medium level.

Fig. 1. Experimental workflow describing

2.4. In vitro digestion model

Digestive fluids to be tested on the human B lymphoid cell linewere obtained by submitting bread samples (blank and naturallycontaminated breads) to a “static-biochemical” in vitro digestionmodel thoroughly described in another paper (De Angelis et al.,2014a,b). The model in use mimics the biochemical conditionsoccurring in the upper GI tract of humans consisting of threephases: chewing, gastric and duodenal digestion.

HPLC/MS experiments were carried out in gastric and duodenalfluids collected at G = 0, G = 120 and D = 180 mins. In particular, twodifferent digestion experiments were performed for each breadsample analyzed, and three different aliquots were collected ateach time-point investigated along each digestion experiment andanalyzed (for the mycotoxin-free bread and the contaminatedbreads). Before HPLC/MS analysis, gastric and duodenal breaddigests were pre-treated according to the protocol alreadyoptimized in a previous work (De Angelis et al., 2014b). Briefly,each collected digested aliquot was first centrifuged at 770 � g for15 min and the supernatants treated differently depending on thefollowing analysis. For T-2/HT-2 analysis the aliquots collectedwere loaded on a C18 cartridge after acidification to pH 2.5. After awashing step with acidified water, mycotoxins were eluted with500 ml of methanol. The volume eluted was dried under gentlestream of air and dissolved in 500 ml of LC mobile phase beforeinjection in the LC/MS system. A scheme describing the workflowfollowed is pictured in Fig. 1.

2.5. Toxicological experiments

2.5.1. Human B lymphoid cell lineThe human lymphoid cell line RPMI 8226 (ECACC, Sigma–

Aldrich) was grown in 25 cm2flasks at a starting density of

250,000 cells/ml in RPMI1640 medium with 2 g/L glucose,supplemented with 10% FBS, 1% L-Glutamine, 1% antibiotic andantimycotic solution. Cells were harvested twice a week. Celldensity and viability were determined by Scepter 2.0 automatedhandheld cell counter (Merk Millipore). The cells used forexperimental protocols showed a mean viability of 90%.

the main steps of the current work.

L. Monaci et al. / Toxicology Letters 232 (2015) 106–112 109

2.5.2. Proliferation testThe toxic effect of trichothecenes (T-2, HT-2) and bread digests

was assessed on proliferation of RPMI 8226 cells by using acolorimetric MTT assay as previously reported by Minervini et al.(2005). Briefly, RPMI 8226 cells were seeded at a cell density of250,000 cells/ml in 96-well plates, exposed to trichothecenesmethanol standard solutions or to bread digests contaminated atdifferent levels and incubated for 24 h. After incubation, cells wereloaded with MTT (5 mg/ml) and incubated for 4 h at 37�C.Absorbance was measured at 580 nm in a spectrophotometricplate reader (ELISA Reader Multiskan MS Plus MK II Labsystem,Finland).

2.5.3. Cytotoxicity of standard mycotoxinsBefore evaluating the effects that bread digests contaminated

with T-2 and HT-2 toxins induced on human B lymphoid cellproliferation, dose-response experiments were carried out on T-2 and HT-2 standard solutions in order to assess their toxicity inabsence of matrix. To this aim, standard solutions at the fixedconcentration of 10 mg/ml for T-2 and HT-2, were prepared andappropriately diluted in order to eliminate cytotoxicity induced bythe solvent (e.g. methanol). Serial dilutions (1:10) for T-2 and HT-2 were performed in order to test a concentrations ranging from5 �10�5 to 500 ng/ml for T-2 and HT-2 toxins. Each concentrationwas tested in 8 wells and three independent experiments wereperformed for each mycotoxin.

2.5.4. Cytotoxicity of digestive fluids from in vitro digestion ofcontaminated bread

For toxicity evaluation, gastric and duodenal digestive fluidswithdrawn at G0, G120 and D180 min along the gastro-duodenal invitro digestion of bread, were submitted to a minimal pre-treatment before toxicity testing. Each digestive fluid (n = 3) wascentrifuged for 15 min at 770 � g to allow sedimentation and thefollowing removal of the particulate precipitate within the digest.The supernatants were collected and each fluid diluted 16 and32 times with cell medium before performing the test. Each digestdilution was tested in 8 wells with two independent experimentsperformed.

2.6. Statistical analysis

Statistical analysis were performed by using SigmaPlotTM

software v.12 (Systat Software, Inc., SigmaPlot for Windows). Allpairwise Multiple Comparisons Dunn’s method was used toevaluate significant differences between cells treated with stocksolution of mycotoxins or bread digests and control samples.Values of p < 0.05 were considered statistically different.

3. Results

3.1. Mycotoxin quantification in bread digests

Digestive fluids were collected at the different time-pointsalong digestion occurring in the upper part of the human digestivetract and were further tested on B lymphoid cell line, according tothe experimental workflow schematized in Fig. 1. Then, a chemical

Table 1Concentrations of T-2 and HT-2 calculated in gastric and duodenal fluids collected alon

Time point Sample A (ng/ml) (mean � s.d.) Sample B (n

HT-2 T-2 HT-2

G0 – 31 � 1 811 � 96

G120 – 31 � 1 654 � 69

D180 – – 614 � 72

quantification of T-2 and HT-2 mycotoxins in each digestive fluid(G0, G120 and D180) was performed by using a method based onHPLC separation and high resolution mass spectrometric detection(De Angelis et al., 2014b; Monaci et al., 2011). This information wasan essential pre-requisite to successively correlate the mycotoxinlevels with the observed toxicity upon assaying the digests on Blymphoid cells. In Table 1 are reported T-2 and HT-2 concentrationscalculated in gastric and duodenal fluids referred to naturallycontaminated bread samples digested and expressed as nano-grams of mycotoxin per milliliter of liquid extract. As appearingfrom the table, gastro-duodenal fluids obtained upon digestion ofsamples B and C, were found to be contaminated by HT-2 at high(�500 ng/ml) and medium level (�500 ng/ml) respectively,whereas lower levels of T-2 (�52 ng/ml) were found for bothtype of samples. In stark contrast with those, sample A was foundonly slightly contaminated by T-2. These levels are in agreementwith what expected in cereal based products, where the presenceof an endogenous carboxylestherases catalyzes the enzymaticconversion of T-2 into HT-2 (Lattanzio et al., 2009).

3.2. Toxicity of single mycotoxin standard solutions on human RPMI8226 lymphoid cell line

As appearing from Table 1, both samples showed a heavycontamination by HT-2 despite the lower contamination observedby T-2. However, since both mycotoxins possess a renown toxicityon biological systems, our investigation was aimed to assess thetoxic potential of these mycotoxins on B lymphoid cells. At thispurpose, mycotoxin standard solutions were preliminary assayedon the B lymphoid cells followed by mycotoxin containingdigestive fluids. The inhibition of cell proliferation induced bymycotoxins singly tested in methanol standard solutions wasperformed by using a colorimetric MTT cleavage test and wasassessed as inhibiting concentration 50% (IC50) value by interpola-tion of the dose-response curve. T-2 and HT-2 toxins induced, after24 h of incubation, high toxicity and showed similar IC50 values of4.1 �0.63 (mean � standard deviation) and 5.5 �1.2 ng/ml, respec-tively. By extending the incubation time to 48 h, the IC50 valuesobtained were 3.1 �0.8 and 3.8 � 0.8 ng/ml for T-2 and HT-2 toxin,respectively.

3.3. Toxicity of gastro-duodenal fluids from in vitro digestion ofcontaminated bread assessed on human RPMI 8226 lymphoid cell line

3.3.1. Optimization of the best dilution for toxicity studies ofcontaminated digests

The first issue to be overcome during toxicity assessment wasthe evaluation of the toxic contribution given by the food matrixwhen assaying lymphoid B cells; this translated into identificationof the best dilution factor, to apply to digests collected at the end ofduodenal phase, able to neutralize an eventual toxic effectdisplayed by the matrix. According to our data, 1:16 dilutionwas the minimum dilution factor requested to have a negligibletoxic contribution from the food matrix on RPMI 8226 cell line. Thisvalue was obtained by comparing the cell proliferation % of a noncontaminated bread digest diluted several times (1:8, 1:16; 1:32)with the cell medium.

g the in vitro digestion time-course and referred to the samples A, B, C.

g/ml) (mean � s.d.) Sample C (ng/ml) (mean � s.d.)

T-2 HT-2 T-2

52 � 1 291 � 63 33 � 150 � 4 176 � 44 33 � 111 � 1 273 � 27 –

G0 G120 D180

Pro

life

ratio

n %

re

spe

ct to

the

co

ntro

l

0

20

40

60

80

100

120 D64 SA MPLE B D16 SA MPLE C

a

bb

A

B B

DIGES TION TIME POINTS

G0 G12 0 D18 0

HT

2 (

ng/m

l)

0

5

10

15

20

25B

A

a

b b

A

B

A

Fig. 3. Cell proliferation of lymphoid cell line (RPMI) after exposure to HT-2 containing gastro-duodenal fluids and respective HT-2 levels calculated. A:Toxicity induced by digestive fluids, calculated by MTT test, and referred to the high(samples B) and low (sample C) contaminated bread sample differently diluted(1:64 and 1:16, respectively) and collected at different time-points. B: Levels of HT-2 calculated by HPLC–MS analysis in the final diluted gastro-duodenal fluids

110 L. Monaci et al. / Toxicology Letters 232 (2015) 106–112

3.3.2. Immunotoxicity of bread digests on RPMI lymphoid B cell lineA preliminary investigation was carried out on non contami-

nated bread digests (also termed blank digests) that were assessedfor their toxicity along the digestion time-course to check for anyeventual matrix effect. As previously reported, blank digestsdiluted 1:16 showed a negligible toxic contribution on RPMI8226 cell line proliferation. Moreover, according to the statisticaltest performed, no significant difference among the means of thethree digestive fluids (collected at G0, G120 and D180) analysedand the medium (CTR), was found as shown in Fig. 2. Thereforethese values were pooled together to generate a unique populationof data for each experiment. As a result, for further investigationsthe value of pooled controls was used as solely control for thetoxicity assessment of the mycotoxin containing digests. Accordingto the chemical quantification previously shown, sample Bappeared to be the sample worthy to be deeply investigated, ascontaining the highest levels of HT-2. Sample C had showed levelsof HT-2, whereas sample A did not show any significant inhibitionof cell proliferation and then was not used for further analysis. Dueto the high concentration of HT-2 recorded in sample B digests, wetested different dilution factors to better spot differences along thedigestion time-course. Concerning sample B digest, we found that16 times dilution did not produce significant differences on cellularproliferation at all time-points analyzed with an inhibition ofcellular proliferation of approximately 75% (data not shown). Thiscould be probably attributed to the high toxicity shown by thismycotoxin whose content remained constant along the digestiontime-course. Moreover, a still high toxicity (nearly 45%) was alsoobserved when the initial and final phases of digestion werecompared by assaying 1:32 dilution factor (data not shown). As aresult, due to the toxic activity evidenced along the wholedigestion process, we decided to stretch potentials of the assayby testing 1:64 dilution on sample B digests as the highest dilutionfactor able to discriminate the toxicity among the different pointsduring digestion time-course. In case of 1:64 digest dilution, asignificant difference was highlighted along digestion, as reportedin panel A of Fig. 3, by comparing time-points G0 versus G120 andD180 both for sample B and C (differences are indicated by theletters). Anyway the remarked reduction observed on cellularproliferation during digestion cannot be explained on the basis ofthe content of mycotoxins calculated in the final diluted digests(see panel B of Fig. 3). As shown in Fig. 3A our results indicate a

�

DIGES TION TIME POINTS

CTR G0 G12 0 D18 0

OD

58

0 n

m

0,0

0,2

0,4

0,6

0,8

1,0

1,2

1,4

1,6

1,8

Fig. 2. Comparison among absorbance values of three digestive fluids referred to ablank bread collected at different time points along digestion and the cell medium(CTR) on RPMI lymphoid cell line by using MTT test. Data are expressed asmean � standard deviation (n = 8) from a representative experiment. No statisticallysignificant differences were found among groups (p > 0.05).

collected along the digestion process of B (after 1:64 dilution) and C (after1:16 dilution) bread samples.*Control represents pools of digestive fluids from non-contaminated bread samplecollected at all time-points. Data are expressed as mean � S.D. (n = 16). Values withdifferent letters (lowercase letters for sample B and uppercase letters for sample C)showed to be significantly different (p < 0.05) according to the Dunn’s testperformed.

significant (p < 0.05) increase in the toxic values between G0 andD180 for sample B. A similar trend was also observed for bread Cdigests. Also in this case a significant (p < 0.05) toxicity was foundby comparing G0 versus G120 and D180. Such a general increase inthe toxicity recorded, between beginning and end of digestion,might be due to the generation of some degradation products oreventual conversion in other mycotoxin derivatives rather than thenative mycotoxin itself, whose content apparently remained quitestable until completion of the gastro-duodenal phase (see Fig. 3B).

3.3.3. Evaluation of the matrix effect on HT-2 toxicityWith the aim to confirm the influence of the food matrix on the

final toxicity shown by HT-2 in digested bread, sample B (thehighest contaminated sample) was progressively diluted with themycotoxin-free (blank) bread digest by assaying several dilutionfactors. The last digestion time-point (D180) of sample B was thenchosen for such investigation, since representing the final point of

Fig. 4. Correlation between cell proliferation % and HT-2 concentration calculatedin sample B digest (at D180) obtained after appropriate dilutions with the blankbread digest (from D 1:16 to D 1:512).

L. Monaci et al. / Toxicology Letters 232 (2015) 106–112 111

gastro-duodenal digestion. By diluting the sample with blankdigest, we could then evaluate the toxic effect induced by thelowest HT-2 concentration and the eventual modulation played byother food components that might contribute to enhancing orquenching HT-2 toxicity.

In this way the contribution of the digestive matrix wasalways taken into account along the step-wise dilution. SampleB duodenal digest at 1:16 times dilution (time point D180) wasfurther diluted 1:32, 1:64; 1:128; 1:256 and 1:512 in order toestimate the lowest level able to significantly reduce cellularproliferation. Toxicity values, obtained after exposure of differentdilutions of contaminated bread digests, were plotted against theestimated concentrations of HT-2 (after appropriate dilutions)and points were interpolated by a polynomial curve as depictedin Fig. 4. In a parallel experiment, HT-2 standard solution wastested by progressive dilutions with culture medium to cover awide concentration range and an IC50 of 5.5 ng/ml wascalculated. By contrast, according to our data shown in Fig. 4,a concentration of approximately 40 ng/ml of HT-2 in the digestproved to induce 50% of cell inhibition. This suggests that thefood matrix involved in this study, namely bread, can modulatethe toxic potential of HT-2.

4. Discussion

Toxicity of T-2 and HT-2 toxins proved to be similar when testedon RPMI 8226 and the different incubation times investigated didnot significantly alter this effect. Immunotoxicity of bothmycotoxins demonstrated in this study was in agreement withdata obtained by Nielsen et al. (2009) by assaying the same Bhuman lymphoid cell line. On other human lymphoid cell linesdifferent IC50 values were calculated, and values obtained for HT-2 were 10 times lower than those calculated for T-2 (Visconti et al.,1991). A different toxicity and toxic mechanism were described byMinervini et al. (2005) when lymphoid human cell of type T(MOLT-4) and B (IM-9) were investigated. The authors alsoreported that the cytotoxicity induced by T-2 appeared due toearly apoptosis in MOLT-4 and to direct membrane damage in IM-9 cells, underlining that T-2 toxic effects may be cell type-specific(Minervini et al., 2005).

Based on the quantification carried out by LC/MS analysis,gastric and duodenal digestive fluids were found to be mainlycontaminated by HT-2. In order to correlate the quantitative datawith the outcome of the toxicological testing, digests preliminaryanalyzed for mycotoxin content were submitted to the bioassaywithout sample pre-treatment, except a mild centrifugation step

with the only aim to remove particulate that might interfere withthe outcome of the biological testing. Such alternative fastapproach avoiding extract depletion aimed at retaining all themultitude of compounds in the extracts that might have an effecton RPMI lymphoid B cell line proliferation.

Considering the toxicity of both mycotoxins and the dilution ofdigests (to eliminate toxic effect due to the matrix), only B and Csamples displayed a certain toxicity on RPMI 8226 cells. On theother hand, due to the low levels of T-2 detected in the digestivefluids (considered to be below the minimum level able to exert atoxic effect) the final toxicity observed in the assayed fluids wasmainly attributed to the higher amount of HT-2 present.

By taking into consideration the different HT-2 levels found in Band C bread samples and the high toxicity shown by the respectivedigests, different dilution factors were tested in order to assess theextent of toxicity at all time-points investigated. In the highlycontaminated bread sample (sample B), dilution 1:64 allowed todiscriminate from a toxicological point of view the different phasesalong gastro-duodenal digestion. However, for both samples (B andC) our results showed an increase in the toxicity especiallyrecorded in the later phase of duodenal digestion. These findingsclearly indicate a different trend between the toxicological testsand the levels of the native mycotoxin in the digests, thussuggesting that other factors than the solely HT-2 presence canaccount for the observed toxicity of the gastro-duodenal digests. Asalready reported (Hazel, 2004; Dall’Erta et al., 2013), mycotoxintransformation and modification is likely to take place duringdigestion and might consequently originate new conjugates ormetabolites still possessing a toxic effect.

As a further step, the lowest concentration of mycotoxin in thedigest able to display a toxic effect was investigated with the aim toprovide more insights on the influence of the digested breadmatrix on the final HT-2 toxicity. By progressively diluting thecontaminated bread digest with the blank digest, we assessed themodulation played by the food matrix on HT-2 toxicity. This mightbe probably due to the release of antioxidant compounds in thedigested bread. As well known cereals, especially wheat, corn andsorghum are rich in antioxidants of various origins that might bereleased in the digestive fluids during digestion (Fardet et al.,2008). On this regard, it has been reported that digestion increasesthe antioxidant capacity of cereal products thanks to the favorableenvironment conditions created (acidic conditions in the stomachand enzymatic hydrolysis in the duodenum) that contribute toincrease the solubility and activity of polyphenols in the finaldigest (Nagah and Seal, 2005; Fardet et al., 2008). The antioxidantaction promoted by these compounds against the oxidative stressmight account for the protective mechanism observed on in vivorat experiments (Gianotti et al., 2011). As a consequence of the co-occurrence of HT-2 toxin or derivatives, notoriously inducingoxidation stress as one of the main toxic mechanism on cell lines(Wu et al., 2014), along with antioxidant compounds, the overalltoxicity might result modulated in the contaminated bread digestsanalyzed. In conclusion, the presence of a food matrix like bread inthe digestive fluids should not be under-estimated in theevaluation of the final toxicological risk associated. According tothe interpolation curve reported in Fig. 4 describing the correlationbetween HT-2 content in the digest and its toxicity, we found that40 ng/ml of HT-2 in the duodenal digest enabled 50% of cell growthon human B lymphoid cells. By contrast, the IC50 calculated for HT-2 in standard solutions by employing the same cell line (5.5 ng/ml)resulted nearly 10 times lower than the value obtained in HT-2 contaminated bread digests. Finally, in presence of duodenalbread matrix rich in antioxidant compounds, HT-2 could display alower cytotoxic activity. In general, these results turn veryinteresting from a toxicological point of view because wedemonstrated that digestive matrix can modulate the toxicity

112 L. Monaci et al. / Toxicology Letters 232 (2015) 106–112

toward RPMI human B lymphoid cells especially attributed to HT-2 mycotoxin or related derivatives.

5. Conclusions

In this study the correlation between HT-2 content and toxicity,in gastro-duodenal digests collected at beginning and end ofduodenal digestion, was investigated. At a deeper analysis, a morepersistent toxic activity was assessed especially upon completionof the duodenal phase. Such toxicity might be attributed to thepresence of mycotoxin-related metabolites or conjugates formedthroughout the digestive process. In addition, the contribution ofthe cereal food matrix in the estimation of the final toxicity wasinvestigated and found to quench the toxic activity of HT-2. Aremarked reduction of toxicity was displayed in the bread digestand was found to be approximately one order of magnitude lowerthan that calculated for HT-2 standard solutions. Future efforts willbe aimed at the identification and structural characterization ofconjugated mycotoxins or eventual modification products thatmight account for the observed toxicity in the later duodenalphase.

Conflict of interest

The authors declare that there are no conflicts of interest.

Transparency document

The Transparency document associated with this article can befound in the online version.

Acknowledgments

European Commission’s Seventh Framework Programme (FP7/2007–2013) funded this research under grant agreement no. FP7-222-654-DREAM. The authors acknowledge the project L.A.I.F.F. –

Rete Di Laboratori per L’innovazione nel Campo Degli AlimentiFunzionali (codice 47) “PO Puglia FESR 2007–2013”, Asse I, Linea1.2. Accordo Quadro della ricerca scientifica, intervento “Reti dilaboratorio di ricerca” that funded the purchase of HR-MSinstrument.

References

Avantaggiato, G., Havenaar, R., Visconti, A., 2003. Assessing the zearalenone-bindingactivity of adsorbent materials during passage through a dynamic in vitrogastrointestinal model. Food Chem. Toxicol. 41, 1283–1290.

Avantaggiato, G., Havenaar, R., Visconti, A., 2004. Evaluation of the intestinalabsorption of deoxynivalenol and nivalenol by an in vitro gastrointestinal modeland the binding efficacy of activated carbon and other adsorbent materials.Food Chem. Toxicol. 42, 817–824.

Avantaggiato, G., Havenaar, R., Visconti, A., 2007. Assessment of the multi-mycotoxin-binding efficacy of a carbon/aluminosilicate-based product in an invitro gastrointestinal model. J. Agr. Food Chem. 55, 4810–4819.

Battilani, P., Costa, L.G., Dossena, A., Gullino, M.L., Marchelli, R., Galaverna, G., Pietri,A., Dall’Asta, C., Giorni, P., Spadaro, D., Gualla, A., 2009. Scientific information onmycotoxins and natural plant toxicants. CFP/EFSA/CONTAM/2008/01. Scientific/Technical Report submitted to EFSA. Available at http://www.efsa.europa.eu/de/scdocs/doc/24e.pdf.

Canady, R.A., Coker, R.D., Egan, S.K., Krska, R., Kuiper-Goodman, T., Olsen, M., Pestka,J., Resnik, S., Schlatter, J., 2001. Deoxynivalenol. Safety Evaluation of CertainMycotoxins in Food. WHO Food Additives Series 47, FAO Food and NutritionPaper 74. WHO, Geneva, Switzerland, pp. 419–555.

Commission Recommendation of 27 March 2013 on the presence of T-2 and HT-2 toxin in cereals and cereal products. Off. J. Eur. Union. L 91/12.

Dall’Erta, A., Cirlini, M., Dall’Asta, M., Del Rio, D., Galaverna, G., Dall’Asta, C., 2013.Masked mycotoxins are efficiently hydrolized by human colonic microbiotareleasing aglycones. Chem. Res. Toxicol. 26, 305–312.

De Angelis, E., Monaci, L., Pascale, M., Visconti, A., 2013. Fate of deoxynivalenol: T-2 and HT-2 toxins and their glucoside conjugates from flour to bread: aninvestigation by high-performance liquid chromatography high-resolutionmass spectrometry. Food Addit. Contam. A 30, 345–355.

De Angelis, E., Monaci, L., Visconti, A., 2014a. Investigation on the stability ofdeoxynivalenol and DON-3 glucoside during gastro-duodenal in vitro digestionof a naturally contaminated bread model food. Food Control 43, 270–275.

De Angelis, E., Monaci, L., Mackie, A., Salt, L., Visconti, A., 2014b. Bioaccessibility of T-2 and HT-2 toxins in mycotoxin contaminated bread models submitted to invitro human digestion. Innov. Food Sci. Emerg. 22, 248–256.

European Food Safety Authority Panel on Contaminants in the Food Chain, 2011.Scientific opinion on the risks for animal and public health related to thepresence of T-2 and HT-2 toxin in food and feed. EFSA J. 9, 2481.

Fardet, A., Rock, E., Remesy, C., 2008. Is the in vitro antioxidant potential of whole-grain cereals and cereal products well reflected in vivo? J. Cereal Sci. 48, 258–276.

Foroud, N.A., Eudes, F., 2009. Trichothecenes in cereal grains. Int. J. Mol. Sci. 10, 147–173.

Gianotti, A., Danesi, F., Verardo, V., Serrazanetti, D.I., Valli, V., Russo, A., Riciputi, Y.,Tossani, N., Caboni, M.F., Guerzoni, M.E., Bordoni, A., 2011. Role of cereal typeand processing in whole grain in vivo protection from oxidative stress. Front.Biosci. 16, 1609–1618.

Gutleb, A.C., Morrison, E., Murk, A.J., 2002. Cytotoxicity assays for mycotoxinsproduced by Fusarium strains: a review. Environ. Toxicol. Pharmacol. 11, 309–320.

Kabak, B., Brandon, E.F.A., Var, I., Blokland, M., Sips, A.J.A.M., 2009. Effects ofprobiotic bacteria on the bioaccessibility of aflatoxin B1 and ochratoxin A usingan in vitro digestion model under fed conditions. J. Environ. Sci. Health B 44,472–480.

Lattanzio, V.M.T., Solfrizzo, M., Visconti, A., 2009. Enzymatic hydrolysis of of T-2 toxin for the quantitative determination of total T-2 and HT-2 toxins in cereals.Anal. Bioanal. Chem. 395, 1325–1334.

Miller, J.D., 2002. Aspects of the ecology of fusarium toxins in cereals. In: De Vries, J.,Trucksess, M., Jackson, L. (Eds.), Mycotoxins and Food Safety. Kluwer Aademic/Plenum Publishers, New York.

Minervini, Fornelli, F., Lucivero, G., Romano, C., Visconti, A., 2005. T-2 toxinimmunotoxicity on human B and T lymphoid cell lines. Toxicology 210, 81–91.

Monaci, L., De Angelis, E., Visconti, A., 2011. Determination of deoxynivalenol: T-2 and HT-2 toxins in a bread model food by liquid chromatography-highresolution-Orbitrap-mass spectrometry equipped with a high-energy collisiondissociation cell. J. Chromatogr. A 1218, 8646–8654.

Nagah, A.M., Seal, C.J., 2005. In vitro procedure to predict apparent antioxidantrelease from wholegrain foods measured using three different analyticalmethods. J. Sci. Food Agri. 85, 1177–1185.

Nielsen, C., Casteel, M., Dietrich, R., Märtlbauer, E., 2009. Trichothecene-inducedcytotoxicity on human cell line. Mycotoxin Res. 25 (2), 77–84.

Placinta, C.M., D’Mello, J.P.F., Macdonald, A.M.C., 1999. A review of worldwidecontamination of cereal grains and animal feed with Fusarium mycotoxins.Anim. Feed Sci. Technol. 78, 21–37.

Raiola, A., Meca, G., Manes, J., Ritieni, A., 2012. Bioaccessibility of deoxynivalenol andits natural co-occurrence with ochratoxin A and aflatoxin B1 in Italiancommercial pasta. Food Chem. Toxicol. 50, 280–287.

Rocha, O., Ansari, K., Doohan, F.M., 2005. Effects of trichothecene mycotoxins oneukaryotic cells: a review. Food Addit. Contam. 22, 369–378.

Sudakin, D.L., 2003. Trichothecenes in the environment: relevance to human health.Toxicol. Lett. 143, 97–107.

Versantvoort, C., Oomen, A., Van de Kamp, E., Rompelberg, C., Sips, A., 2005.Applicability of an in vitro digestion model in assessing the bioaccessibility ofmycotoxins from food. Food Chem. Toxicol. 43, 31–40.

Visconti, A., Minervini, F., Lucivero, G., Gambatesa, V., 1991. Cytotoxic andimmunotoxic effects of Fusarium mycotoxins using a rapid colorimetricbioassay. Mycopathologia 113, 181–186.

World Health Organization, 2001. WHO Food Additives, Series: 47. Safety evaluationof certain mycotoxins in food. Prepared by the fifty-sixth meeting of the JointFAO/WHO Expert Committee on Food Additives (JECFA). Available at http://www.inchem.org/documents/jecfa/jecmono/v47je01.htm.

Wu, Q.-H., Wang, X., Yang, W., Nüssler, A.K., Xiong, L.-Y., Ku9ca, K., Dohnal, V., Zhang,X.-J., Yuan, Z.-H., 2014. Oxidative stress-mediated cytotoxicity and metabolis ofT-2 toxin and deoxynivalenol in animals and human: an update. Arch. Toxicol.88, 1309–1326.

Yang, G.H., Jarvis, B.B., Chung, Y.L., Pestka, J.J., 2000. Apoptosis induction by thesatratoxins and other trichothecene mycotoxins: relationship to ERK, p38 MAPKand SAPK/JNK activation. Toxicol. Appl. Pharmacol. 164, 149–160.