**4. Discussion**

Mycotoxin co-contamination of finished pig feed is more common than single mycotoxin contamination [13], emphasizing the importance of investigating the effects of such naturally contaminated feeds, even when mycotoxins are detected below the accepted tolerated concentrations for individual mycotoxins. A review from Escrivá et al. [17] showed that in vivo toxicity studies of *Fusarium* mycotoxins had become much more frequent in the decade between 2003 and 2014, thus highlighting their importance. These mycotoxicoses can manifest as acute diseases with high morbidity and death or as chronic diseases, reduced animal productivity, and decreased resistance to pathogens [45].

In the present study, gilts were fed a diet containing maize naturally contaminated with *Fusarium* mycotoxins from roughly the beginning of the last quarter of their pregnancy until a week after the weaning of their piglets (a total of 54 ± 1 day). Our experimental diet contained three *Fusarium* mycotoxins, namely DON (5.08 mg per kg of feed), ZEN (0.09 mg per kg of feed), and FA (21.6 mg per kg of feed). Undertaking this approach, we assumed that the presenting liver pathology of both gilts and their piglets would mimic chronic exposure to mycotoxins in a typical pig production setting.

A detailed investigation into liver histopathology revealed a statistically significant difference in hepatocellular necrosis and apoptosis as well as sinusoidal leukocytosis with inflammatory infiltrates of hepatic lobules. Additionally, these two histopathological changes were strongly correlated, and they likely represented recruitment of inflammatory cells to sites of hepatocellular necrosis. Since the inflammatory cells were composed of a mixed population of neutrophils, eosinophils, lymphocytes, and, occasionally, macrophages, the observed rise in sinusoidal leukocytes could also be an indicator of a concurrent inflammatory process, knowing that DON is suspected to raise the susceptibility to infection and chronic diseases [46].

In humans and animals, the toxic effects of DON include emesis and anorexia, alteration of intestinal and immune functions, reduced absorption of nutrients, and elevated susceptibility to infection and chronic diseases [46]. As DON can induce such a variety of toxic effects, it is difficult to interpret whether liver pathology observed in the in vivo experiments was mostly due to direct hepatotoxic effects or significantly aggravated by indirect effects related to reduced nutrient absorption and overall decreased food intake.

Some studies have shown no changes in liver morphology when DON was either the sole potential toxic factor or in combination with another mycotoxin [30,31,40,42,43], but that was not the case in all studies. When piglets received diets either mono-contaminated with DON (1.5 mg per kg of feed) or multi-contaminated with DON (2 or 3 mg per kg of feed), ZEN (1.5 mg per kg of feed), and NIV (1.3 mg per kg of feed), the most prominent histopathological features were disorganization of hepatic cords, cytoplasmic vacuolization of hepatocytes, and megalocytosis. Piglets fed the co-contaminated diet with the higher dose of DON also exhibited focal necrosis in the liver [33]. In previous studies, feed co-contaminated with DON and ZEN often elicited microscopic changes in the liver. Chen et al. [28] fed pigs a diet containing 1 mg of DON and 0.250 mg of ZEN per kg of feed and mentioned blood vessel thickening and dilation as the only histopathological finding in the liver but did not provide a clear grading scheme or statistical analysis. When feeding prepuberal gilts for 35 days with wheat containing increasing concentrations of DON and ZEN, the amount of intracytoplasmic glycogen decreased in a dose-dependent manner, whereas hemosiderin deposition increased. Wheat containing the highest doses of DON (6.1 mg or 9.57 mg per kg of feed) and ZEN (0.235 mg or 0.358 mg per kg of feed) also elicited a statistically significant increase in the thickness of interlobular connective tissue [27]. On the other hand, a similar study, where pregnan<sup>t</sup> sows were exposed to a concentration of 4.42 mg DON and 0.048 mg ZEN per kg of feed for 35 days, did not show any difference in thickness of interlobular connective tissue [29], whereas a slight difference was observed in pregnan<sup>t</sup> sows receiving a concentration of 9.57 mg DON and 0.358 mg ZEN per kg of feed for 35 days [30]. Another study assessed the effects of feeding gilts for 1, 3, or 6 weeks with either DON at a dose of 0.012 mg/kg body weight (BW) per day, ZEN at 0.04 mg/kg BW, or a mixture of DON and ZEN. Histologically, several changes were observed in the liver, especially in gilts receiving DON and ZEN, such as increased thickness of perilobular connective tissue, increased total microscopic liver score,

dilation of hepatic sinusoids, temporary changes in glycogen content, and increased iron accumulation in hepatocytes [34].

Interestingly, a statistically significant decrease in interlobular connective tissue was observed in our experimental group of gilts when automated detection was used on slides stained with Goldner's Masson trichrome stain. Our finding is in contrast with previous studies because in those cases, experimental animals had an increased amount of interlobular connective tissue in the liver [27,30,34]. A recent study on collagen and elastin content in skin of mink receiving DON-contaminated feed showed a decrease in type III (immature) collagen when mink received DON at a concentration of 1.1 mg/kg of feed and complete absence of type III (immature) collagen when mink received a dose of 3.7 mg/kg DON in feed with or without 0.05% bentonite [47]. As interlobular connective tissue in pig liver contains both type I and type III collagen [48], the observed decrease in interlobular connective tissue in our study may have been due to DON influencing the expression of fibrous collagens. Further investigation would be needed to confirm or refute this assumption.

Cell culture experiments on porcine hepatocytes have shown that DON causes morphological and functional disorders in hepatocytes. Cell death of hepatocytes occurred in a dose-dependent manner and exhibited morphological changes characteristic of apoptosis. Apoptosis was further confirmed by consistently TUNEL-positive nuclei and increased activity of caspase-3, a key enzyme in apoptotic cell death [49]. Our study showed a significantly increased number of apoptotic cells in the liver of experimental gilts; an increase was observed for hepatocytes and even more so for cells in hepatic sinusoids. The apoptotic cells in hepatic sinusoids were most likely lymphocytes, based on their morphology, but we did not attempt double immunolabelling to confirm this. Similarly, piglets intravenously injected with DON at a concentration of 1 mg/kg body weight displayed systemic apoptosis of lymphocytes in lymphoid tissues as well as hepatocytes, thereby proposing a hepatotoxic potential of DON next to its already known immunotoxic effect [41]. Based on these findings, we suspect that both oral and intravenous administration of DON can cause apoptosis of hepatocytes as well as circulating leukocytes.

As DON has also been associated with antiproliferative activity [50], we assessed the proliferation activity by determining the number of Ki-67-positive hepatocytes, but found no significant differences, neither in the liver of gilts nor their piglets. The same was observed in the liver of porcine fetuses when pregnan<sup>t</sup> sows were exposed to DON and ZEN for 35 days during pregnancy [40], whereas an increased proliferation index was observed when piglets received feed mono-contaminated with FBs (6 mg per kg of feed) or DON (3 mg per kg of feed) and especially when co-contaminated with both FBs and DON [14].

*Fusarium* mycotoxins clearly have direct and indirect effects on the liver of pigs that are fed contaminated diets, thereby some studies have examined possible placentar transfer from sows to their offspring [20,21,29–31,39,40]. Some did not identify any changes in the fetuses [40] or piglets [39]. Dänicke et al. [20] suggested that the developing fetus is exposed to DON, ZEN, and their metabolites when sows are fed contaminated feed, but did not observe any histopathological changes, and Goyarts et al. [21] also found that DON and de-epoxy-DON pass the placental barrier to a significant extent. Fetuses that were exposed to DON and ZEN between the 35th and 70th day of gestation exhibited increased glycogen content and changes in the architecture of hepatocellular mitochondria, likely caused by diaplacentar toxin transfer of ingested toxins from the mother to the developing fetuses. Feed consumption did not play a role in that experiment because both the control and experimental groups received the same amount of feed per day [29]. When sows were fed a diet contaminated with DON (9.57 mg per kg of feed) and ZEN (0.358 mg per kg of feed) between the 75th and 110th day of gestation, there were no histological changes in the livers of their piglets [30]. When sows received DON and ZEN co-contaminated diet between the 63rd and 70th day of gestation, DON was detected in fetus plasma and there was a change in the proportion of their white blood cells [31]. In our study, there were no significant

differences in the assessed histopathological changes, apoptosis, and proliferation index in the livers of piglets. This suggests that a decreased feed consumption by the gilts leading to a lower energy and nutrient intake or a direct effect of diaplacentar transfer of ingested mycotoxins did not cause morphologically apparent changes in the developing fetuses, possibly due to the relatively low concentrations of *Fusarium* mycotoxins in the diet fed to the pregnan<sup>t</sup> gilts.
