Toxicokinetics of Hydrolyzed Fumonisin B1 after Single Oral or Intravenous Bolus to Broiler Chickens Fed a Control or a Fumonisins-Contaminated Diet

The toxicokinetics (TK) of hydrolyzed fumonisin B1 (HFB1) were evaluated in 16 broiler chickens after being fed either a control or a fumonisins-contaminated diet (10.8 mg fumonisin B1, 3.3 mg B2 and 1.5 mg B3/kg feed) for two weeks, followed by a single oral (PO) or intravenous (IV) dose of 1.25 mg/kg bodyweight (BW) of HFB1. Fumonisin B1 (FB1), its partially hydrolyzed metabolites pHFB1a and pHFB1b, and fully hydrolyzed metabolite HFB1, were determined in chicken plasma using a validated ultra-performance liquid chromatography–tandem mass spectrometry method. None of the broiler chicken showed clinical symptoms of fumonisins (FBs) or HFB1 toxicity during the trial, nor was an aberration in body weight observed between the animals fed the FBs-contaminated diet and those fed the control diet. HFB1 was shown to follow a two-compartmental pharmacokinetic model with first order elimination in broiler chickens after IV administration. Toxicokinetic parameters of HFB1 demonstrated a total body clearance of 16.39 L/kg·h and an intercompartmental flow of 8.34 L/kg·h. Low levels of FB1 and traces of pHFB1b were found in plasma of chickens fed the FBs-contaminated diet. Due to plasma concentrations being under the limit of quantification (LOQ) after oral administration of HFB1, no toxicokinetic modelling could be performed in broiler chickens after oral administration of HFB1. Moreover, no phase II metabolites, nor N-acyl-metabolites of HFB1 could be detected in this study.


Introduction
Fumonisins (FBs) are secondary fungal metabolites of Fusarium verticillioides, F. proliferatum or other Fusarium species, that resemble each other structurally [1,2]. Fumonisin B 1 (FB 1 ), B 2 (FB 2 ) and B 3 (FB 3 ) represent the most important of the more than 15 FBs analogues that have been described [3]. FBs can be divided into four main categories, namely A, B, C and P [4]. Since FB 1 is the most prevalent and toxic analogue, toxicological assessment of FBs has mostly been done for FB 1. FB 1 is a diester of propane-1,2,3-tricarboxylic acid (TCA), which has a long aminopolyol side-chain, namely 2-amino-12,16-dimethyl-3,5,10,14,15-pentahydroxyeicosane [3]. FB 2 and FB 3 differ structurally from FB 1 in the number and position of hydroxyl groups, since a free hydroxyl group is lacking on either the C-10 or the C-5 position, as shown in Figure 1 [2]. The primary amino group is known for the biological activity of FBs, as acetylation of FB 1 to fumonisin A 1 hampers the cytotoxicity as well as the inhibition of ceramide synthase [3].
Toxins 2020, 12, 413 2 of 13 and toxic analogue, toxicological assessment of FBs has mostly been done for FB1. FB1 is a diester of propane-1,2,3-tricarboxylic acid (TCA), which has a long aminopolyol side-chain, namely 2-amino-12,16-dimethyl-3,5,10,14,15-pentahydroxyeicosane [3]. FB2 and FB3 differ structurally from FB1 in the number and position of hydroxyl groups, since a free hydroxyl group is lacking on either the C-10 or the C-5 position, as shown in Figure 1 [2]. The primary amino group is known for the biological activity of FBs, as acetylation of FB1 to fumonisin A1 hampers the cytotoxicity as well as the inhibition of ceramide synthase [3]. FBs reveal distinct structural similarities to sphingolipids, due to their long-chain base backbones. Hence, FBs have the ability to completely inhibit sphinganine N-acyl transferase (ceramide synthase), resulting in a disruption of the ceramide and sphingolipid metabolism [6][7][8]. Blockage of the ceramide synthase enzyme not only leads to the inhibition of sphingolipids synthesis, but also induces an accumulation of free sphinganine (Sa) and sphingosine (So) in tissues, serum and urine [2,4]. This associated increase of the Sa:So ratio, observed in tissues and body fluids after exposure to FBs, is a suitable biomarker of effect in animals and humans [2,9,10]. In most animal species, the liver, as well as the kidneys and the intestinal tract, are the main target organs of FBs toxicity [2,11,12]. Generally, poultry are rather resistant to FBs in comparison to mammals, especially pigs and horses [13][14][15][16][17][18]. Only during the first three days of life (≥ 125 mg/kg feed) was an increase in mortality after FB1 exposure observed in broiler chickens [13], as well as in growing ducks aged 12-14 weeks (20 mg/kg feed) [16]. Feeding a FB1-contaminated diet (100-400 mg/kg feed) to broilers resulted in black, sticky diarrhea in the first 2 weeks of life [15]. Besides, broiler chickens fed an FBscontaminated diet in doses from 100 to 400 mg FB1/kg feed for 2 to 3 weeks showed a dose-dependent decrease in feed intake and bodyweight gain [13,15]. No negative effects on performance were observed under experimental conditions, at concentration levels which were below or approached Figure 1. Structures of the fumonisins FB 1 , FB 2 and FB 3 with tricarballylic acid side-chains (TCA), and of the hydrolyzed fumonisins HFB 1 and HFB 2 (or aminopentols, AP 1 and AP 2 , respectively) and the corresponding N-acyl-derivatives [5]. Copyright © 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.
FBs reveal distinct structural similarities to sphingolipids, due to their long-chain base backbones. Hence, FBs have the ability to completely inhibit sphinganine N-acyl transferase (ceramide synthase), resulting in a disruption of the ceramide and sphingolipid metabolism [6][7][8]. Blockage of the ceramide synthase enzyme not only leads to the inhibition of sphingolipids synthesis, but also induces an accumulation of free sphinganine (Sa) and sphingosine (So) in tissues, serum and urine [2,4]. This associated increase of the Sa:So ratio, observed in tissues and body fluids after exposure to FBs, is a suitable biomarker of effect in animals and humans [2,9,10]. In most animal species, the liver, as well as the kidneys and the intestinal tract, are the main target organs of FBs toxicity [2,11,12]. Generally, poultry are rather resistant to FBs in comparison to mammals, especially pigs and horses [13][14][15][16][17][18]. Only during the first three days of life (≥125 mg/kg feed) was an increase in mortality after FB 1 exposure observed in broiler chickens [13], as well as in growing ducks aged 12-14 weeks (20 mg/kg feed) [16]. Feeding a FB 1 -contaminated diet (100-400 mg/kg feed) to broilers resulted in black, sticky diarrhea in the first 2 weeks of life [15]. Besides, broiler chickens fed an FBs-contaminated diet in doses from 100 to 400 mg FB 1 /kg feed for 2 to 3 weeks showed a dose-dependent decrease in feed intake and bodyweight gain [13,15]. No negative effects on performance were observed under experimental conditions, at concentration levels which were below or approached the European maximum guidance level in feed (20 mg FB 1 + FB 2 /kg feed) [19]. However, it was recently demonstrated that even low to moderate dietary levels of FBs negatively affect enterocyte viability and proliferation, and the production of pro-inflammatory cytokines, and alter the intestinal barrier function, and therefore increase the susceptibility of avian species to important enteric infectious diseases, such as coccidiosis and necrotic enteritis [20][21][22][23]. Furthermore, the mRNA expression of genes encoding for the intestinal cytochrome P450 (CYP450), the drug-metabolizing enzyme CYP1A4, and the drug transporter mechanisms, such as multiple drug resistance protein 1 (MDR1 or P-glycoprotein, P-gp), in broiler chickens were upregulated following exposure to 25 mg FB 1 + FB 2 /kg feed for 15 days, which can affect the pharmacokinetic (PK) or toxicokinetic (TK) properties of other xenobiotics substrates [24].
Therefore, taking into account that mycotoxin contamination of feed is a continuous feed safety issue, resulting in economic losses during animal production [25], methods for the detoxification of feed have been developed. For FBs, these imply feed processing techniques, such as alkaline cooking (nixtamalization) used during the production of masa and tortillas for human consumption, and the application of mycotoxin binders/modifiers in animal feed. Mycotoxin binders, such as clays, are highly effective against aflatoxins, however, their detoxification capacity wit regards to FBs is rather limited. In contrast, following the application of FB esterase FumD (EC 3.1.1.87), enzymatic degradation results in conversion of FBs to the less toxic partially hydrolyzed FB 1 (pHFB 1 ), or fully hydrolyzed HFB 1 , by cleavage of the tricarballylic acid esters at the C-14 and/or C-15 position. FB esterase FumD is an enzyme of bacterial FB catabolism [26,27], which has recently been commercialized as FUMzyme ® (BIOMIN GmbH, Getzersdorf, Austria) and authorized by the EU for usage in poultry and pigs. Enzymatic hydrolysis has been shown to be an effective detoxification method in broilers, since it prevents the FB-associated disruption of the sphingolipid metabolism, characterized by a reduced serum and hepatic Sa:So ratio and a counteracting FB-induced intestinal up-regulation of cytokine gene expression [interleukin-8 (IL-8) and IL-10] [28]. Besides, in accordance with different mammalian species [29][30][31], it has been demonstrated that the intestinal microbiota of broiler chickens and turkeys also has a limited capacity to hydrolyze FBs [28,32].
The only known metabolic pathway of FB 1 is the hydrolyzation (hydrolysis) of its side-chains in vertebrates [11]. HFB 1 is not only a 10-fold less potent inhibitor of ceramide synthase than FB 1 , but it is also a substrate for ceramide synthase in rat liver microsomes [5,26,31,33,34]. Acetylation of HFB 1 at the primary amino group, with fatty acids of variable chain lengths, results in the formation of N-acyl-HFB 1 (NAHFB 1 ) metabolites, which are also known to be ceramide analogues. The in vivo formation of NAHFB 1 has been demonstrated in rats [5]. Furthermore, phase II conjugation processes of FB 1 , e.g., sulphatation as proposed by Hopmans et al. (1997) [35], or glucuronidation, have not yet been identified. However, no data are available on the in vivo TK of HFB 1 , nor on the formation of NAHFB 1 in chickens [36]. Furthermore, FB 1 is not absorbed in vitro by human colon adenocarcinoma Caco-2 cells, also suggesting a low oral bioavailability. The oral bioavailability of FB 1 in chickens is only 0.7% [36]. In contrast, for HFB 1 a high transcellular passage has been observed in vitro, particularly from the basolateral to the apical side, suggesting that HFB 1 is effluxed by P-gp [33,37]. Remarkably, this P-gp transport mechanism was upregulated following exposure to the toxic FBs parent molecules, which could affect the TK properties of the HFB 1 metabolite, as it is a P-gp substrate. Since exposure to HFB 1 is linked with the hydrolysis of FB 1 by intestinal enzymes or by a feed additive enzyme, the aim of this study was to investigate not only the TK parameters of HFB 1 alone, but also the impact of prior exposure to FBs on these parameters.

Results and Discussion
During the animal experiment, no signs of illness due to FBs or HFB 1 toxicity were observed in any of the broiler chickens. All birds were alert, showed a normal feed intake and normal droppings, and no regurgitation was observed. The average bodyweight (BW) was similar (p = 0.820) in both groups (control and FBs diet), 974 ± 109 and 1036 ± 110 g at day 21, respectively. This matched expectations, since the contaminated feed contained 10.8 mg FB 1 , 3.3 mg FB 2 and 1.5 mg FB 3 /kg feed, and therefore approached, but was below, the EU guidance level of 20 mg FB 1 + FB 2 /kg [19]. Additionally, HFB 1 is generally known to be less toxic than its parent toxin FB 1 [26,27], and poultry species are known to be able to tolerate high doses of FB 1 , especially regarding performance response, with no effect on growth of broiler chickens up to 75-100 mg FB 1 /kg feed [15].
To the authors' knowledge, for the first time, the TK properties of HFB 1 were determined in broiler chickens, both after feeding a control and FBs-contaminated diet. In pigs, biotransformation of FB 1 into pHFB 1 , and further to HFB 1 , by their digestive microbiota and liver has been observed after being fed an FBs-contaminated diet containing 45 mg FB 1 , 8.6 mg FB 2 and 4.6 mg FB 3 /kg for 10 days, together with a persistence of low levels of pHFB 1 in most organs for several days [30]. However, only low concentrations of pHFB 1a and pHFB 1b have been detected in the plasma of broiler chickens, fed an FBs-contaminated diet with 10 mg FBs/kg for 14 days [28]. The formation of pHFB 1a , pHFb 1b and HFB 1 after FB 1 exposure has been described in turkeys [32], monkeys [29] and rats [31] as well. Accordingly, in this study, only traces (>limit of quantification; LOQ) of pHFB 1b (0.94-1.34 ng/mL and 0.93-2.43 ng/mL, respectively) were detected in a few samples of broiler chickens fed the FBs-contaminated diet, and subsequently administered an IV or PO bolus of HFB 1 , respectively. Although, the presence of pHFB 1a was detected in some samples of broilers fed the FBs-contaminated diet (>limit of detection; LOD), it could not be quantified (<LOQ). No pHFB 1a or pHFB 1b was detected in chickens fed the control diet. Low levels of FB 1 were observed in chickens fed the FBs-contaminated diet, with average background levels at all timepoints of 4.89 ± 5.91 ng/mL and 2.67 ± 3.70 ng/mL, in broiler chickens administered HFB 1 IV and PO, respectively. No FB 1 was detected in the plasma of animals fed the control diet. Consequently, it can be concluded that the observed pHFB 1 in the plasma of animals fed the FBs-contaminated diet resulted from hydrolysis of the FB 1 present in the diet. However, the intestinal microbiota of broiler chickens and turkeys seems to have a rather limited capacity to hydrolyze FBs [28,32]. HFB 1 initiated lesser inflammatory responses than FB 1 in an in vitro co-culture model of porcine intestinal epithelial and immune cells [38]. In vivo toxicity of the intermediate, pHFB 1 , has only been investigated in rats so far, resulting in no observed in vivo toxicity [31]. On the other hand, hepatic and renal lesions, in terms of single-cell necrosis, mitosis, apparent collapse of the centrilobular parenchyma of the liver and single-cell necrosis of the outer medulla, and cytoplasmic vacuolation of the kidney, have been reported in rats fed nixtamalized feed containing HFB 1 [39], while no hepatotoxicity or pathological changes of the liver, brain, heart, kidneys, thymus or mesenteric lymph nodes examined in mice fed purified HFB 1 have been detected [40]. This different outcome could probably be explained by the presence of residual, partially hydrolyzed or masked FB 1 , following the process of nixtamalization [5]. However, no significant influence of HFB 1 on the intestinal lesion score of the villi morphology in piglets exposed to HFB 1 has been observed [41].
No N-acyl or phase II metabolites of HFB 1 were detected in any of the samples in this study. N-acyl-metabolites of HFB 1 were highly cytotoxic in human colonic HT29 cells, and were also characterized as a potent inhibitor of ceramide synthase [34]. HFB 1 acylation has been evaluated in rats, and although N-acyl-metabolites were detected, no toxicity was observed in vivo [5]. While N-acylation of FB 1 and HFB 1 occurred in human cell lines and in rats [33,34], the occurrence of these metabolites in avian species has not yet been observed, matching the absence of these metabolites in this study.
Comparative HFB 1 plasma concentration-time profiles, following IV HFB 1 administration to broiler chickens fed either a control diet or a FBs-contaminated diet, are shown in Figure 2. Visual inspection of the goodness-of-fit plots of the individual, model-predicted concentrations (IPRED) versus the observed HFB 1 plasma concentrations (C obs ) revealed an appropriate structural model for most individuals ( Figure S1A). The QQ-plots of the conditionally weighted residuals of C obs demonstrated the normal distribution of the weighted residuals ( Figure S1B). The addition of the covariate experimental diet (control versus FBs-contaminated) did not significantly improve the -2 log likelihood-ratio (2LL) of any of the fixed effect parameters' volume of distribution of the central compartment (V c ), volume of distribution of the peripheral compartment (V p ), total body clearance (Cl), or intercompartmental flow (Q), and was therefore not retained in the final model. The toxicokinetic parameters of HFB1 following IV administration are shown in Table 1, and are described by a two-compartment model. The mean plasma concentration of HFB1 in broiler chickens fed an FBs-contaminated diet for 2 weeks following IV administration (plasma concentration at time 0; C0) was 339.59 ng/mL. The total exposure of HFB1 in broiler chickens after IV administration, from timepoint 0 to infinity (area under the plasma-concentration time curve from time 0 to infinity; AUC0inf), was 76.26 ng·h/mL. Regarding volume of distribution (Vd), values of 3.68 L/kg and 5.04 L/kg have been determined for the central and peripheral compartments, respectively, thus resulting in a Vd of 8.7 L/kg at steady state (Vss). Accordingly, pigs given a single HFB1 IV bolus showed a Vd of 11.0 L/kg, in a one-compartmental model [42]. Furthermore, a short distribution half-life (T1/2ɑ) of 0.09 h, followed by a longer elimination half-life (T1/2β) of 0.69 h, was revealed. In barrows given a single IV bolus, of 0.056 mg/kg BW, of HFB1, a T1/2ɑ of 0.05 h and a T1/2β of 1.01 ± 0.80 h was observed [42]. The mean residence time (MRT) of HFB1 after IV administration to broiler chickens was 0.53 h, and the elimination rate constant (ke) was 4.4 h -1 . In comparison to pigs, with a total body clearance of 6.7 L·kg/h in barrows after IV HFB1 administration, broilers seem to have a higher total body clearance (16.39 L/kg·h). The latter could concur with the oral bioavailability of parental FB1 being higher in swine than in poultry, with values of 3.1% [42] and 0.7% [36], respectively, and with poultry being rather resistant to FBs toxicity when compared to pigs or other mammals [13][14][15][16][17][18]. The toxicokinetic parameters of HFB 1 following IV administration are shown in Table 1, and are described by a two-compartment model. The mean plasma concentration of HFB 1 in broiler chickens fed an FBs-contaminated diet for 2 weeks following IV administration (plasma concentration at time 0; C 0 ) was 339.59 ng/mL. The total exposure of HFB 1 in broiler chickens after IV administration, from timepoint 0 to infinity (area under the plasma-concentration time curve from time 0 to infinity; AUC 0-inf ), was 76.26 ng·h/mL. Regarding volume of distribution (Vd), values of 3.68 L/kg and 5.04 L/kg have been determined for the central and peripheral compartments, respectively, thus resulting in a Vd of 8.7 L/kg at steady state (V ss ). Accordingly, pigs given a single HFB 1 IV bolus showed a Vd of 11.0 L/kg, in a one-compartmental model [42]. Furthermore, a short distribution half-life (T 1/2α ) of 0.09 h, followed by a longer elimination half-life (T 1/2β ) of 0.69 h, was revealed. In barrows given a single IV bolus, of 0.056 mg/kg BW, of HFB 1 , a T 1/2α of 0.05 h and a T 1/2β of 1.01 ± 0.80 h was observed [42]. The mean residence time (MRT) of HFB 1 after IV administration to broiler chickens was 0.53 h, and the elimination rate constant (k e ) was 4.4 h −1 . In comparison to pigs, with a total body clearance of 6.7 L·kg/h in barrows after IV HFB 1 administration, broilers seem to have a higher total body clearance (16.39 L/kg·h). The latter could concur with the oral bioavailability of parental FB 1 being higher in swine than in poultry, with values of 3.1% [42] and 0.7% [36], respectively, and with poultry being rather resistant to FBs toxicity when compared to pigs or other mammals [13][14][15][16][17][18].  Plasma concentrations of HFB 1 after oral administration were only above the LOQ in a limited number of samples, and are shown in Figure 3. Consequently, no TK modelling could be performed on these data. Still, the rather fast appearance of HFB 1 in the systemic circulation after oral administration indicates that the ingested mycotoxin is mainly absorbed in the proximal part of the small intestines. Plasma concentrations of HFB1 after oral administration were only above the LOQ in a limited number of samples, and are shown in Figure 3. Consequently, no TK modelling could be performed on these data. Still, the rather fast appearance of HFB1 in the systemic circulation after oral administration indicates that the ingested mycotoxin is mainly absorbed in the proximal part of the small intestines.  The T max of HFB 1 (10 min) after oral HFB 1 administration in the FBs-fed group was reached remarkably fast, when compared to studies of oral FB 1 intake in laying hens, turkeys and ducks, where respective T max values of 60 min, 180 min and 60-120 min were observed [43].
Previously, a disturbance of intestinal homeostasis of chickens fed an FBs-contaminated diet (18.6 mg FB 1 + FB 2 /kg feed) was demonstrated [21]. In another study, broiler chickens were either Toxins 2020, 12, 413 7 of 13 fed an FBs-contaminated (8.70 mg FB 1 /kg, 4.15 mg FB 2 / kg and 1.44 mg FB 3 /kg feed) or a control diet, resulting in FB 1 influencing the expression of intestinal P-gp [24]. With HFB 1 being a substrate of this receptor [24], another aim of this research was to examine whether prolonged FB 1 exposure would influence the uptake of its metabolite HFB 1 in the intestines of broiler chickens. Regarding the difference in plasma concentrations between the FBs and the control diet fed groups, after oral administration of HFB 1 , in this study, one could presume that a prolonged FBs exposure actually does influence the intestinal uptake of HFB 1 . But as mentioned before, the quantity of plasma concentration measurements above the LOQ, after oral HFB 1 administration, was insufficient to allow the drawing of conclusions from those results. The latter is also held responsible for the fact that no calculation of oral bioavailability of HFB 1 could be made. Nevertheless, a high transcellular passage, particularly from the basolateral to the apical side, was revealed for HFB 1 in human colon adenocarcinoma Caco-2 cells, suggesting that HFB 1 is effluxed by P-gp [33,37]. Since an increased expression of multi-drug resistance protein 1 (MDR1), encoding for P-gp, could be observed in the intestine of broiler chickens fed an FBs-contaminated diet [24], oral bioavailability of HFB 1 could vary, leaving one with the need for more research on the oral bioavailability of HFB 1 in avian species. In broiler chickens, the oral bioavailability of FB 1 , being mainly negatively charged in the duodenum and jejunum due to the pH which therefore limits oral absorption by passive non-ionic transcellular diffusion [43], is known to be rather low [44]. Similarly, in this study only very low plasma levels of HFB 1 were detected in broiler chickens following oral administration, suggesting a very low oral bioavailability. That said, in rats, the bioavailability of HFB 1 was thought to be greater than that of FB 1 , since about 2.5-fold greater amounts of 14 C-HFB 1 , compared to 14 C-FB 1 , were excreted in urine after oral administration of 0.69 µmol 14 C-FB 1 and 14 C-hydrolyzed FB 1 /kg BW [45], increasing the likelihood that HFB 1 can be absorbed by the gut, and in the meantime allowing the subsequent metabolism of N-acyl-HFB 1 and the enhancing of its possible toxicological effects.

Conclusions
While HFB 1 is shown to be less toxic than its parental toxin FB 1 , its toxicokinetic pathways in avian species are still widely unexplored. In this study, toxicokinetic parameters following administration of a single IV bolus of HFB 1 to broiler chickens have been determined (V c , V p , V ss , Cl, Q, AUC 0-inf , C 0 , k e , MRT, T 1/2α and T 1/2β ). The plasma concentrations of HFB 1 after oral administration were only above the LOQ in a limited number of samples. Consequently, no TK modelling could be performed on these data. Traces of pHFB 1b (0.94-1.34 ng/mL and 0.93-2.43 ng/mL, respectively) have been detected in a few samples of broiler chickens fed the FBs-contaminated diet, and subsequently administered an IV bolus or PO bolus of HFB 1 , respectively. Using UPLC-HRMS, no phase II metabolites or N-acyl metabolites of HFB 1 could be detected in any of the samples.

HFB 1 for Animal Trial
HFB 1 for oral (PO) and intravenous (IV) administration was prepared as previously described [31] with some minor changes. HFB 1 for PO administration was obtained by enzymatic treatment of F. verticillioides culture material with the FB carboxylesterase FumD (80 U enzyme/g culture material) for 30 min, and subsequently purified and lyophilized [31]. Per 250 g of culture material, 2550 µmol HFB 1 was obtained. HFB 1 for IV administration was prepared by alkaline hydrolysis of FB 1 analytical standard (Biopure, Tulln, Austria). Therefore, 5 mL of FB 1 solution (7.4 mg/mL water) was hydrolyzed with 4 mL of 2.5M NaOH by shaking for 7.5 h. After completing the conversion, the formed HFB 1 was recovered by liquid-liquid extraction (eight times with acetonitrile (ACN)). The achieved HFB 1 solution was purified over an SPE C18 column to remove salts and tricarballylic acids, and subsequently diluted in a Toxins 2020, 12, 413 8 of 13 0.9% NaCl solution up to a concentration of 4.18 mg HFB 1 /mL, with traces of pHFB 1 a (0.0045 mg/mL). HFB 1 for PO and IV administration was stored at 2-8 • C.

Feed Preparation and Experimental Diets
A commercial broiler starter diet was fed to all chickens (Vanden Avenne, Ooigem, Belgium) during the first week of the experiment. Later, this diet is referred to as control diet. From day 8 until day 24 chickens were fed either the control diet without mycotoxins [commercial broiler grower diet (Vanden Avenne)], or a grower diet experimentally contaminated with FBs. None of the diets contained a mycotoxin detoxifier. Screening of the control starter and grower feeds for mycotoxin contamination was executed by a liquid chromatography tandem mass spectrometer (LC-MS/MS) method, as described by Monbaliu et al. [46]. The levels of all tested mycotoxins were below the decision limit (CCα) in the control starter and grower diets. More specifically, the CCα values of FB 1 , FB 2 and FB 3 were 58, 45 and 42 µg/kg feed, respectively.
To produce the grower diet experimentally contaminated with FBs, FBs culture material was mixed with 500 g of control feed. Lyophilized FBs culture material of F. verticillioides (M-3125) [47] (7.37 g FB 1 /g and 2.93 g FB 2 /g) was obtained from Biopure, Romer Labs Diagnostic GmbH (Tulln, Austria) and stored at 2-8 • C. This premix was then blended with 5 kg of control feed to ensure a homogeneous distribution of the mycotoxins. The premix was blended for 20 min into the total quantity of feed necessary for the trial. To test the homogeneity of FBs in feed, samples were taken at three different points in the batch and analyzed for FBs as described for the control diets. The FBs-contaminated diet contained 10.8 mg FB 1 , 3.3 mg FB 2 and 1.5 mg FB 3 /kg feed.

Animal Experiment
A total of 16 one-day-old broiler chickens (Ross 308) were obtained from a commercial hatchery (Vervaeke-Belavi, Tielt, Belgium), and were fed a starter diet without mycotoxins during the first week. Subsequently, animals were randomly divided into two experimental groups of eight animals (4 male/4 female). While one group was being fed a control grower diet from day 8 onwards, the other group was given a FBs-contaminated grower diet. An 18 h/6 h light/darkness program was applied. Feed and drinking water were provided ad libitum.
A toxicokinetic (TK) study of HFB 1 was performed following a two-way cross-over design. Each broiler chicken was administered a bolus of HFB 1 (1.25 mg/kg bodyweight) either orally (PO) or intravenously (IV) (vena cutanea ulnaris superficialis or wing vein). At day 21, four chickens of each group were administered HFB 1 IV, and the four other animals received HFB 1 PO. After a wash out and recovery period of two days, the protocol was repeated at day 24 in a cross-over design. Animals were feed-deprived overnight (8 h) prior to, and until 3 h after, HFB 1 administration. After each bolus dosing, blood was collected into heparinized tubes by direct venipuncture of the leg vein (vena metatarsalis plantaris superficialis) before (0 h) and at different time points after HFB 1 administration: 5 min, 10 min, 20 min, 30

Preparation of Standard Solutions
Stock solutions of FB 1 , FB 2 and FB 3 (1 mg/mL) were made in water/ACN (50/50, v/v) and kept at 2-8 • C. Working solutions of 10 µg/mL, 1 µg/mL and 0.1 µg/mL were made by adequate dilution of the stock solution in water/ACN (50/50, v/v). The standard mixture solution contained FB 1 , HFB 1 , pHFB 1 a and pHFB 1 b and was adequately diluted in water/ACN (50/50, v/v) in order to prepare working solutions necessary for the preparation of calibrator and quality control (QC) samples. A working solution of 1 µg/mL of 13 C 34 -FB1 was made in water/ACN (50/50, v/v) to obtain the internal standard (IS). All working solutions were kept at 2-8 • C.

Plasma Sample Pre-Treatment
To 100 µL of plasma 12.5 µL of the IS working solution (1 µg/mL) was added. Then, after vortex mixing the sample was transfered onto an Ostro TM 96-well plate. Next, 300 µL of 1% formic acid (FA) in ACN was added, after which the sample was aspirated 3 times in order to stimulate protein precipitation. By the application of a vacuum (67.7 kPa) for 10 min, the sample was run through the 96-well plate. An aliquot of 2.5-µL was injected onto the LC-MS/MS instrument.

UPLC-MS/MS Method for Quantification
Briefly, the LC system was composed of an Acquity UPLC H-Class Quaternary Solvent Manager and Flow-Through-Needle Sample Manager with temperature-controlled tray and column oven from Waters (Zellik, Belgium). Chromatographic separation was accomplished on an Acquity UPLC HSS T3 column (100 mm × 2.1 mm i.d., dp: 1.8 µm) (Waters, Zellik, Belgium) combined with an Acquity HSS T3 1.8 µm Vanguard pre-column (Waters, Zellik, Belgium). The mobile phase A was composed of 0.3% FA and 10 mM NH 4 FA in water, and the mobile phase B was ACN. A gradient elution of 0-0.5 min (90% A, 10% B), 5.5 min (linear gradient to 90% B), 5.5-7.5 min (10% A, 90% B), 7.7 min (linear gradient to 90% A) and 7.7-10.0 min (90% A, 10% B) was performed. The flow-rate was set at 0.4 mL/min. The temperatures of the column oven and autosampler tray were adjusted to 40 • C and 8 • C, respectively. The UPLC column effluent was coupled to a Xevo TQ-S ® MS/MS system, supplied with a positive electrospray ionization (ESI) probe (all from Waters). A divert valve was utilized and the UPLC effluent was routed to the mass spectrometer from 2.5 to 4.5 min. Optimization of instrument parameters was obtained by directly infusing working solutions of 1 µg/mL of FB 1 , FB 2 , FB 3 , the IS and a diluted standard mixture solution, which contained FB 1 , HFB 1 , pHFB 1 a and pHFB 1 b at concentrations of 0.5, 0.86, 1.43 and 2.5 µg/mL, respectively, combined with the mobile phase (50% A, 50% B, flow-rate: 200 µL/min). The flow-rate was set at 10 µL/min. MS/MS acquisition was operated in the multiple reaction monitoring (MRM) mode [48].

UPLC-HR-MS Analysis for Identification
To determine potential HFB 1 phase II and N-acyl metabolites in plasma samples of broiler chickens, an Acquity I-Class UPLC interfaced to a Synapt G2-Si HDMS instrument (Waters, Zellik, Belgium) was used [48].

Toxicokinetic Modelling
Plasma concentration-time data were analyzed with a nonlinear mixed-effects modelling approach using quasi-random parametric expectation maximization (QRPEM) as an estimation method in Phoenix NLME ® (Certara, Cary, NC, USA). The structural toxicokinetic model for the IV data was a two-compartmental model with first order elimination as shown in Equation (1).
where dA 1 /dt is the rate of the decrease of the amount of mycotoxin in the plasma or central compartment, dA 2 /dt is the rate of decrease of the amount of mycotoxin in the peripheral compartment, V c is the volume of distribution of the central compartment, V p is the volume of distribution of the peripheral compartment, Cl is total body clearance, Q is the intercompartmental flow and C and C2 are the concentrations of the toxin in the central and peripheral compartments, respectively. Interindividual variability was expressed using an exponential error model according to Equation (2): where P i is the parameter in the ith bird, θ P is the typical value of the parameter in the population, and η Pi is a random variable in the ith bird, with a mean of zero and a variance of ω 2 . Interindividual variability is reported as ω. Residual variability (ε), with a mean of zero and a variance of σ 2 , was best described with a multiplicative error model in Equation (3).
where C obs is the observed concentration for the individual and C pred is the model-predicted concentration plus the error value (ε). Structural and error model selection was guided by visual inspection of goodness-of-fit plots (e.g., observed vs predicted plasma concentrations, weighted residuals versus predicted concentrations, and weighted residuals versus time), -2LL, Akaike information criterion (AIC) and Bayesian information criterion (BIC) as well as precision of the parameter estimates. The models were chosen based on the smaller values of -2LL, AIC and BIC, the better precision of estimates, and the superior goodness-of-fit plots.
The evaluated covariate was the prior feeding with FBs-contaminated feed versus feeding with uncontaminated control feed (categorical variable). A stepwise forward-backward process was used to evaluate whether inclusion of the covariates significantly improved the model fit using a -2LL test. A decrease in -2LL with a p-value < 0.01 was considered significant for addition, and p < 0.001 for exclusion, of the covariate.
The following fixed effect parameters were determined: V c, V p , Cl and Q. The computed secondary parameters were: C 0 , AUC 0-inf, K e , MRT, T 1/2α and T 1/2β .