Several bacterial functional pathways were

Several bacterial functional pathways were observed after DON administration, of which signal transduction, metabolism and genetic information processing displayed the highest levels of enrichment. Importantly, these three pathways were reported to be closely associated with DON's toxicity. Many evidences showed that DON impaired body's health via changing expressions of different signal transduction pathways. For instance, DON inhibited the TNF-α and TLR-dependent NF-κB activation (Hirano and Kataoka, 2013; Sugiyama et al., 2016), triggered mouse skin cell proliferation and inflammation through mitogen-activated protein kinases (MAPKs) pathway (Mishra et al., 2014b) and caused embryo-toxicity via nuclear factor erythroid 2-related factor 2 (Nrf2)/HO-1 pathway (Yu et al., 2017). Recent studies not only revealed that microbiota participated in the detoxification for DON that could be transformed to de-epoxy forms or other compounds by a highly enriched microbial consortium (Sass et al., 2012; Gratz et al., 2017; Mayer et al., 2017), our present study also indicated that gut microbiota might play an important role in signaling transforming process: DON administration significantly reduced Odoribacter that potentially associated with the metabolism of carbohydrates, which matched by PICRUSt predictions. HO-1 plays a protective role in various liver injuries (Wang et al., 2011; Sass et al., 2012; Origassa and Camara, 2013). However, there are few studies exploring the association between HO-1 and DON-induced hepatotoxicity. HO-1 was also found to be induced by gut microbiota (Onyiah et al., 2013), which suggest a potential connection between gut microbiota, HO-1 and DON-induced hepatotoxicity. Several studies have shown that AAV8 had the best performance with gene editing in liver-directed PF-4708671 synthesis in macaques as well as murine models (Wang et al., 2010a; Wang et al., 2010b), therefore, we performed HO-1 gene-editing in mice liver via AAV8 transfection to explore the potential correlation among them. In the current study, Western blotting results indicated a high efficiency of 3-week virus transfection for both HO-1shRNA and HO-1OE. According to histological examination with H&E-staining, the overexpression of HO-1 attenuated DON-induced inflammatory cell infiltration to a level similar to control group, while the knockdown of HO-1 caused more severe inflammatory changes compared to DON group, which confirmed that HO-1 could protect from DON-induced liver damages. ALT activity in HO-1OE groups manifested no statistical difference compared with control group probably due to a protective role of HO-1. Moreover, the highest mean activity of ALT in HO-1shRNA group confirmed the anti-inflammatory nature of HO-1 from the other side. Surprisingly, AST activity was significantly increased in HO-1shRNA group in comparison with control and HO-1OE groups, which indicated that 25 μg/kg bw/day DON administration could cause damage to hepatocellular mitochondria in the absence of HO-1. Summing up, our experiment data basically demonstrated that HO-1 could protect against low-dose of DON-induced liver damage. To explore a possible influence of HO-1 on intestinal flora homeostasis in DON-induced hepatotoxicity, we primarily examined whether AAV8 would affect gut microbiota independently. Fecal 16S rRNA data suggested that the composition of gut microbiota in both HO-1OE and HO-1shRNA group was not markedly different before and after AAV8 transfection. Namely, the above results excluded the bacteria fluctuation caused by HO-1 editing alone. Remarkably, at the end of DON administration, gut microbiota in HO-1OE and HO-1shRNA groups exhibited different patterns of alteration compared with that of DON group. DON group and HO-1shRNA group shared a similar increase in Parabacteroides. Besides, Alloprevotella and Ruminococcus were especially reduced in HO-1shRNA group compared with the DON group. It is very interesting that DON reduces the number of Ruminococcus in HO-1shRNA group since the dominant Ruminococcus strain in the gut is R. navus that is known to stimulate mucin expression and glycosylation (Graziani et al., 2016). The fact that DON reduces the number of Ruminococcus in absence of HO-1 is important since it may mean that, in addition to direct inhibitory action of DON on mucus expression, DON may have an indirect effect through the suppression/reduction of commensal bacteria able to stimulate mucus expression/glycosylation such as Ruminococcus. Additionally, Alloprevotella was observed to enrich in various diseases (Downes et al., 2013). Overall, gut ecosystem is complicated because the dynamic and delicate balance can be disturbed by external challenges or internal interactions among microbiome. In consistent with the changes of microbiota, the PICRUSt predicted more altered pathways under influence of DON in HO-1shRNA group, compared with DON group.