Toxicokinetics of citreoviridin in vivo and in vitro

Citreoviridin (CIT) produced by Penicillium citreonigrum as a secondary metabolite is a yellow rice toxin that has been reported to be related to acute cardiac beriberi; however, its toxicokinetics remain unclear. The present study elucidated the toxicokinetics through swine in vivo experiments and predicted the human toxicokinetics by a comparison with findings from in vitro experiments. Swine in vivo experiments revealed that CIT had a high bioavailability of more than 90%. In addition, it showed a large volume of distribution (1.005 ± 0.195 L/kg) and long elimination half-life (17.7 ± 3.3 h) in intravenous. These results suggested the possibility of a slow metabolism of CIT. An intestinal permeability study using the human cell line Caco-2 showed that CIT had a high permeability coefficient, suggesting it would be easily absorbed in human intestine, similar to its absorption in swine. The metabolite profiles were investigated by incubating CIT with S9 obtained from swine and humans. Hydroxylation, methylation, desaturation and dihydroxylation derivatives were detected as the predominant metabolites, and CIT glucuronide was produced slowly compared with above metabolites. A comparison of the peak area ratios obtained using quadrupole time-of-flight mass spectrometer showed that the rates of all of the main metabolites except for glucuronide produced using human S9 were three-fold higher than those obtained using swine S9. Furthermore, the elimination of CIT using human S9 was more rapid than when using swine S9, indicating that CIT would be metabolized faster in humans than in swine. These in vivo results suggested that CIT is easily absorbed in swine and persists in the body for a long duration. Furthermore, the CIT metabolism appeared to be faster in human liver than in swine liver in vitro, although the bioavailability of CIT was predicted to be similarly high in humans as in swine.

10 140 mg/mL of CIT-ethanol solution and a small amount of water were added to feed (10 g) in 141 order to make 0.1 mg/kg·BW. This was then fashioned into a sphere and fed to the animals. It 142 was visually confirmed that the animals had eaten the CIT-contaminated feed.

143
Blood was sampled from the jugular vein at 0 min (before administration), and 5, 10,

151
In order to frequently sample blood in a short period of time, up to 1 h after 152 administration, CIT was administered intravenously and orally, followed by immediately 177 was transferred to a micro tube, a three-fold volume of acetonitrile was added. Samples were 178 mixed with a vortex mixed, followed by centrifugation at 8500 g for 10 min at 4 °C. The    Fig. 2A. The two-compartment model was described as follows: where   278 Values are presented as the mean ± SD.

280
The CIT concentration profile in plasma after PO administration is shown in Fig. 2B.

281
The toxicokinetic parameters of PO administration were analyzed by a non-compartmental 282 analysis [16]. The peak plasma concentration (Cmax, 38.2 ± 6.7 ng/mL) was observed 283 between 15 ± 6 h (Tmax) after administration ( Table 2). The mean residence time (MRT) 284 was obtained from equation (2) as follows: The area under the first moment curve (AUMC) and AUC of PO and IV data were 287 calculated using the trapezoidal rule. The MRT obtained from those values was relatively 20 288 long (32.8 ± 11.8 h) ( Permeability study using Caco-2 cells 300 The results of the administration study showed that CIT had a high bioavailability in swine.

305
The apparent permeability coefficient (Papp) estimated from a Caco-2 permeability 306 assay has been reported to correlate well with the human in vivo absorption data for many  (Table 3). These findings indicated that human intestine cells were highly permeable to CIT

344
The main metabolites could not be quantified because standard substances of the 345 metabolites detected were not commercially available. Therefore, comparison of metabolites 346 in humans and swine at each time was conducted with the mean area of each metabolite in 24 347 swine as 100% at 240 minutes. The rate of metabolite peak area by incubation with the 348 human hepatic S9 fraction was two to three times higher than with the swine hepatic S9 349 fraction for all metabolites (Fig. 3B-D). These results indicate that human hepatic S9 has a 350 greater ability to metabolize CIT than swine hepatic S9. 362 Asterisks indicate a significant difference (p<0.05).

25 364
The concentration of CIT hardly decrease up to 240 min after incubation with the 365 human intestinal S9 fraction, whereas it increased from 30 to 60 min after incubation with the 366 swine intestinal S9 fraction, followed by almost no change from 60 to 240 min (Fig. 4A).
367 Although the reason that the increase observed in incubation with swine intestinal S9 fraction 368 was unclear, these indicated that CIT was hardly metabolized either intestinal S9 fractions.
369 For the comparison, the same method as the comparison in S9 fractions supplemented with 370 NADP was used. The rate of metabolite peak area by incubation with the human intestinal S9 371 fraction was roughly three times higher than with the swine intestinal S9 fraction for all 372 metabolites (Fig. 4B-D). These were almost the same rates as in the above incubation using 373 hepatic S9.

374
The peak areas of CIT metabolites produced by incubation with intestinal S9 were 375 about one-third the size of those with hepatic S9 in both swine and humans (data not shown).
376 However, these findings showed that CIT was hardly metabolized at all in the intestine of 377 swine and humans.