Fusarium graminearum is the main fungi species that produces tricothecenes. All tricothecenes contain an epoxide at the C12, C13 positions, which is responsible for their toxicological activity (Nathanail et al., 2015). T-2 toxin (Type A) and DON (Type B) are the major mycotoxins that cause toxicity to humans and animals via oral ingestion (Nathanail et al., 2015).

During World War II, a biological weapon caused an acute syndrome consists of cough, sore throat, dyspnea, bloody nasal discharge, and fever was reported by Soviet scientists (Pitt and Miller, 2016). Twenty years later, T-2 mycotoxin was discovered when civilians consumed wheat that was unintentionally contaminated with Fusarium fungi (Pitt and Miller, 2016). A human toxicosis due to ingestion of moldy rice contaminated with T-2 toxin has been reported in China. According to Wang Z. et al. (1993), 65% of patients developed food poisoning symptoms such as chills, nausea, abdominal distension, dizziness, vomiting, thoracic stuffiness, abdominal pain, and diarrhea. Similar to T-2 toxicity, victims of DON outbreak suffered from vomiting syndromes (Etzel, 2014). Several outbreaks of acute DON toxicity in human have been reported in India, China, and the USA (Etzel, 2014).

Trichothecenes toxic effects in animals (dairy cattle, swines, broilers, and rats) include decreased plasma glucose, reduced blood cell and leukocyte count, weight loss, alimentary toxic aleukia, as well as pathological changes in the liver and stomach (Adhikari et al., 2017). The mechanism involved in T-2 and DON toxicity is generally via oxidative stress-mediated deoxyribonucleic acid (DNA) damage and apoptosis (Wu et al., 2014). Furthermore, T-2 and DON are well-known inhibitors of protein synthesis resulting from the binding of peptidyl-transferase, which is located in the 60s ribosomal subunit (Yang et al., 2017).

In the GI tract, a decreased absorption of glucose was observed following T-2 and DON intoxication resulted from suppressed SGLT1 (glucose transporter) mRNA expression. Apart from the glucose absorption, SGLT1 also responsible for water reabsorption, thus reduction of SGLT1 transporter induces diarrhea as well (Grenier and Applegate, 2013).

The weight loss effect of trichothecenes involved neuroendocrine factors and cytokines. DON and T-2 elevated concentrations of the indoleamines, serotonin and 5-hydroxy-3-indoleacetic acid (HIAA) in all brain regions (Wang J. et al., 1993). These neuroendocrine factors can affect the secretion of both anorexigenic and/or orexigenic hormones (Maresca, 2013). Through increasing gene expression of anorexia-inducing proinflammatory cytokines such as interleukin-1β (IL-1β), interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α), trichothecenes exacerbate the condition of anorexia (Wu et al., 2015). In addition, DON and T-2 also induced the release of the satiety hormones, peptide YY (PYY) and cholecystokinin (CCK), which are critical mediators of anorexia (Wu et al., 2015).

Using animal models, trichothecenes was found to induce necrotic lesions in the GI tract (Kolf-Clauw et al., 2013). A shortening of villi height was also observed in trichothecenes-treated animals (swine, poultry, and rat model). The changes on villi were due to activation of the apoptotic pathway by trichothecenes, which in turn leads to nutrition malabsorption (Alizadeh et al., 2015). Furthermore, results obtained from in vivo and in vitro studies showed that trichothecenes increased intestinal permeability. Using porcine epithelial cell, trichothecenes increased the intestinal permeability by lowering tight junction proteins expression (Osselaere et al., 2013). In addition, previous studies revealed a significant (P < 0.05) decreased in the number of goblet cells that secrete mucin in trichothecenes-treated animals. Mucin is primarily involved in the gut barrier function (Pinton and Oswald, 2014). The disruption in the integrity of intestinal epithelium allows the entry of the pathogen into the gut lumen (Lessard et al., 2015). Besides, trichothecenes have been linked with a decreased level of IL-8 in the intestine, which is responsible for pathogen removal (Kadota et al., 2013). Overall, trichothecenes exert negative impacts on GI tracts specifically on the gut absorption, integrity, and immunity.

Zearalenone (ZEA) is a mycotoxin that primarily produced by Fusarium graminearum and Fusarium culmorum in foods and feeds. The high rate of co-occurrence of ZEA with FBs and DON indicates that these mycotoxins might be involved in a wide range of synergistic and additive interactions. ZEA has been linked to scabby grain toxicosis occurred in Japan, China, Australia, and the USA, with symptoms including nausea, vomiting, and diarrhea (Liao et al., 2009).

It is well recognized that ZEA is a non-steroidal estrogenic mycotoxin that is implicated in the reproductive disorders of farm animals (swines, cattle, and sheep) and hyperoestrogenic syndromes in humans (Kotowicz et al., 2014). Toxicological studies of ZEA revealed its effects on the reproductive system, including enlarged uterus, altered reproductive tract, decreased fertility, as well as abnormal level of progesterone and estradiol. Besides, the ingestion of ZEA during pregnancy reduced fetal weight and survival rate of embryo (Zhang et al., 2014). This phenomenon can be explained through the structure of ZEA. ZEA has a structure which allows it to bind to the mammalian estrogen receptor, although with lower affinity compared to the natural-occurring estrogens (Hueza et al., 2014). Besides, ZEA has also been shown to be hepatotoxic, haematotoxic, immunotoxic and genotoxic (Zhou et al., 2017).

Although the reproductive organ is the main target for ZEA to induce toxicity, the adverse effects of ZEA on GI tracts have been reported. The effects of ZEA ingestion on the GI tract are not as detrimental compared to the other mycotoxins. Studies using intestinal epithelial cells showed that ZEA induced cell death without altering the cell integrity as indicated by transepithelial electrical resistance (Marin et al., 2015). In contrast, it was discovered that the metabolites of ZEA (α- and β-zearalenol) significantly (P < 0.05) decreased the cell integrity. The study showed that ZEA and its metabolites acted differently in the gut (Marin et al., 2015). Abassi et al. (2016) demonstrated that ZEA enhanced cell proliferation, increased colony formation and fastened cell migration of colon carcinoma cell line HCT116. Another study also showed that ZEA down-regulated the expression of tumor-suppressor genes (PCDH11X, DKK1, and TC5313860) in intestinal cells (Taranu et al., 2015). In fact, the modulation of gene expression was responsible for the carcinogenic effects of ZEA. Nevertheless, swines ingested ZEA did not showed changes in the height of villi, the thickness of the mucosa, and number of goblet cells (Gajecka et al., 2016a,b; Lewczuk et al., 2016). In brief, ZEA plays a negative role in gut health although no apparent histological changes have been observed.

Fusarium verticillioides is the major producer of fumonisins, where fumonisin B1 (FB1) is the most abundant in nature (Lerda, 2017). In contrast to most mycotoxins, which are hydrophobic in nature, fumonisins are hydrophilic compounds, which hinders its discovery until 1988 (Gelderblom et al., 1988). Human epidemiological studies in South Africa, Italy, and China revealed that the esophageal cancer is related to the intake of corn grains containing fumonisins (Chilaka et al., 2017). Another epidemic of neural tube defects (birth defects of the brain, spine, or spinal cord) occurred along the Texas-Mexico border, China and South Africa were also found to be associated with fumonisins-contaminated corn consumption (Ortiz et al., 2015). In animals, fumonisins have been found to cause pulmonary edema and hydrothorax in swines; leukoencephalomalacia in equine; and HCC in rats (da Rocha et al., 2014).

FB1 shares the same structure as cellular sphingolipids (Masching et al., 2016). Sphingolipids are responsible for neurological and immunological diseases, as well as cancer. Normal degradation of sphingolipids to ceramide requires sphingomyelinase and ceramidase (Boini et al., 2017). However, FB1 disturbs sphingolipids metabolism via ceramide synthase inhibition which leads to sphingosine accumulation in cells (Masching et al., 2016). FB1 elevates sphingosine levels in urine, serum, kidney, liver, and small intestine. The abnormal turnover of sphingosine induced cytotoxicity, oxidative stress, apoptosis in cells (Hahn et al., 2015).

Using intestinal cell lines (IPEC-1, Caco-2, and HT29), Minervini et al. (2014) found FB1 decreased the cell viability and proliferation in a concentration-dependent manner. A possible mechanism has been suggested through the accumulation of sphinganine by FB1. In the intestinal epithelial cells, sphinganine accumulation blocked G0/G1 phase in cell and resulted in growth inhibition and apoptosis (Angius et al., 2015). The accumulation of sphinganine also altered glycoprotein distribution in the jejunum and caused an increase in transepithelial passage of FB1 (Yamazoe et al., 2017). In addition, FB1 altered the integrity of intestinal barrier by suppressing tight junction (TJ) protein expression level (Romero et al., 2016). The increase in intestinal permeability, in turn, promotes translocation of bacteria (Kelly et al., 2015).

Besides, high levels of FB1 also induced an overgrowth in intestinal goblet cell of broiler and swine (Alassane-Kpembi and Oswald, 2015). Goblet-cell hyperplasia is associated with increased mucin secretion. However, continuous hypersecretion of mucins might deplete the number of goblet cells, resulting in devastation of mucus barrier (Johansson and Hansson, 2016). Previous studies conducted using intestinal cell lines (IPEC-1, Caco-2, and HT29) showed that FB1 was able to regulate immune responses. Upon LPS exposure to the FB1-treated cell line, a reduction in IL-8 synthesis was detected (Minervini et al., 2014). Such reduction could be responsible for a low number of polymorphonuclear leukocytes (PMNs) recruited to infection sites, thus leading to the ineffective elimination of pathogen from the gut (Brazil et al., 2013). Generally, in the gut, FB1 increased intestinal cell apoptosis, reduced intestinal barrier and caused immune dysfunction.

Ochratoxin is mainly produced by Aspergillus species and Penicillium species. Ochratoxin A (OTA) is the most prevalent and relevant fungal toxin of this group (Liuzzi et al., 2017). The main target site of OTA is kidney. Previous findings from animals showed OTA is a potent renal carcinogen (Russo et al., 2016). The International Agency for Research on Cancer (IARC) categorized OTA as possibly carcinogenic to humans under Group 2B carcinogen. Apart from that, OTA is an immunosuppressive, teratogenic, and nephrotoxic compound (Ladeira et al., 2017).

In human studies, OTA is associated with kidney diseases, such as Balkan endemic nephropathy (BEN). BEN is a chronic tubulointerstitial disease which slowly progressed into terminal renal failure. Indeed, a 15 years study confirmed that BEN is correlated with upper urothelial tract cancer (Rouprêt et al., 2015). Furthermore, OTA has been associated with the occurrence of upper urothelial tract cancer (Fahmy et al., 2014). However, a systemic review by Bui-Klimke and Wu (2015) revealed that there is no significant evidence for human health risks associated with OTA exposure based on the epidemiological data. The modes of toxic action of OTA are identified through the blockage of protein synthesis and energy production, the formation of DNA adduct formation, apoptosis, as well as the induction of oxidative stress (Kőszegi and Poór, 2016). Moreover, recent studies showed OTA triggered autism via epigenetic mechanism (Mezzelani et al., 2016).

Other than its adverse effects on the kidney, previous studies also revealed the gut changes induced by OTA. OTA altered nutrition absorption in the intestine. In vitro studies demonstrated that OTA decreased glucose absorption via SGLT1 transporter (Peraica et al., 2011). In addition, OTA-treated animals experienced faster and more harmful parasite infections (provoked by Eimeria acervulina and E. adenoeides) in chicks and turkey compared to control. The results showed that animals fed with OTA had higher lesion and oocyst indexes in the intestine and more damage at mucosa (Manafi et al., 2011). This can be explained by the increased intestinal permeability in the presence of OTA. In addition, results from immunoblotting and immunofluorescence showed that the expression of TJ proteins responsible for intestinal integrity was significantly (P < 0.05) suppressed by OTA (McLaughlin et al., 2004). Besides, OTA-induced oxidative stress also can alter intestinal permeability (Anderson et al., 2016). The oxidative stress induced by OTA has been found to be associated with the apoptosis in the intestinal IPEC-J2 cells (Wang et al., 2017). A study by Solcan et al. (2015) on OTA-fed broilers revealed there is a decrease in villi height and increase in apoptosis of intestinal epithelial cells. Similar results were obtained from another study using broiler model (Qu et al., 2017). Inflammation pathway in the intestine was also affected by OTA. The expression of inflammation-related cytokines (IL-8, IL-6, IL-17A, IL-12, and IL-18) was significantly (P < 0.05) decreased in the intestine of the piglets exposed to the toxin (Marin et al., 2017). The alteration of immune system renders the gut vulnerability to infection. OTA exerted its effect on gut via the reduction of nutrient absorption, disruption of intestinal permeability, cell apoptosis, and modulation of immune system.

AF is a mycotoxin produced by Aspergillus flavus and Aspergillus parasiticus. The most common mycotoxin found in human food and animal feed is AFB1. In fact, AFB1 is the most potent hepatocarcinogen recognized in mammals and listed as Group I carcinogen by IARC (Muhammad et al., 2017). Liver is the main target site of AFB1. Cumulative evidences from human and animals revealed a strong linkage occurs between AFB1 and HCC. While, acute aflatoxicosis induced abdominal pain, vomiting, edema, and death (Mohd-Redzwan et al., 2013).

Aflatoxicosis outbreak has been recorded four times in Kenya from 2004 to 2014, with near to 600 individuals were affected and 211 deaths were reported from the tragic outbreak (Awuor et al., 2017). As discussed earlier, p450 enzymes in the liver metabolize AFB1 into AFB1-8,9-exo-epoxide. The highly reactive exo-epoxides form derivatives with DNA, RNA and proteins which subsequently react with the p53 tumor suppressor gene. The reaction generates AFB1-N7-Gua which is then converted to its stabilized form, AFB1-formamidopyrimidine (AFB1-FABY) adduct. AFB1-FABY causes transversion of guanine (G) to thymine (T), which leads to mutation and malignant transformation (Kew, 2013). In addition to the hepatotoxicity of AFB1 mentioned above, other adverse effects include growth retardation, immunosuppression, and genotoxicity have been reported (Kumar et al., 2017).

Like most of the mycotoxins, AFB1 compromised the health of GI tracts. Colon cell line (Caco-2) was used in in vitro experiment to determine the AFB1 toxicity in the intestine. AFB1 significantly (P < 0.05) inhibited cell growth, increased lactate dehydrogenase activity and caused genetic damage. It is found that the mechanism of AFB1 cytotoxicity are associated with reactive oxygen species (ROS) generation, which leads to the damage of cell membrane and DNA (Zhang et al., 2015). Besides, transepithelial electrical resistance assay showed a reduction in intestinal Caco-2 cells' integrity after AFB1 treatment (Romero et al., 2016).

Similar result has been observed in in vivo study where AFB1 affects intestinal barrier function in broiler model as indicated by an increased ratio of lactulose to rhamnose ratio in the plasma (Chen et al., 2016). Several studies on broiler exposed to AFB1 showed that the density (weight/length) of intestine was reduced (Hossein and Gurbuz, 2015). An increase of apoptotic events was found in the jejunum, accompanied with elevated apoptotic markers (Bax and caspase-3) mRNA expression level. Moreover, the increased apoptosis was corresponded to a lower jejunal villi height as found in the other studies (Peng et al., 2014; Zheng et al., 2017). Another study by Akinrinmade et al. (2016) demonstrated intestinal injuries induced by AFB1 in rats. In the AFB1-fed rat, leucocyte and lymphocyte infiltration were observed at lamina propria of the intestinal mucosa. In the duodenum and ileum, AFB1 exposure caused intestinal lesions such as the development of sub-epithelial space and villi degeneration. The adverse effects on the gut from AFB1 exposure include the disruption of intestinal barrier, cell proliferation, cell apoptosis, and immune system. Although AFB1 is the most life-threatening mycotoxin, yet its toxicity on the gut is comparable to the other mycotoxins.

A study has demonstrated the ability of gut microbiota to remove deoxynivalenol (DON) using an in vitro study (He et al., 1992). It was reported that the microorganisms in large intestines of broilers were able to transform DON into de-epoxy-DON. The gut microbes of broiler were further shown to transform DON to the less toxic metabolite, de-epoxy-DON via epoxide reductase. Similar findings have been observed in other studies using microbial content from the broiler intestine (Lun et al., 1988; Young et al., 2007). Some other studies have also shown that intestinal microorganisms of other animal species including rat (Worrell et al., 1989) and swine (Kollarczik et al., 1994) possess the same ability. However, no alteration was found when swines' intestines content was used in the study conducted by He et al. (1992).

Apart from these, mycotoxin degrading bacteria had been isolated from intestinal content for extensive studies. Microbiological selection strategies guided by PCR DGGE (denaturing gradient gel electrophoresis) had been employed to isolate DON-transforming bacteria and the isolates obtained belong to four different bacterial groups; Anaerofilum, Bacillus Clostridiales, and Collinsella (Yu H. et al., 2010). In addition, a microbial community, namely microbial culture C133 from catfish digesta was screened, and capable to completely transformed DON to de-epoxy-DON after 96 h incubation (Guan et al., 2009). Several intestinal bacterial strains have been identified as biological trichothecene detoxification agent via in vitro screening and microbial analyses in other recent reviews (Hathout and Aly, 2014). Interestingly, a study showed the ability to metabolize DON can be obtained via gut microbiota transfer in swine. However, the transfer of gut microbiota revealed no changes in the DNA-profiles of the gut bacterial composition (Eriksen et al., 2002).

Several studies have been done on the interaction of DON toward the gut microbiota. Consumption of feed contaminated with DON for a time duration of 4 weeks has been shown to exert minor effect on the total number of fecal aerobic mesophilic bacteria and anaerobic sulfite-reducing bacteria in swine. Although there was no effect of DON on microbial diversity, the richness index was significantly (P < 0.05) increased by DON exposure (Waché et al., 2008). In another study by Saint-Cyr et al. (2013), GF male rats were inoculated with fecal flora from healthy human, in order to investigate the human gut microbial changes induced by DON. By using real-time PCR quantification, a significant increase of Bacteroides/Prevotella group and decreased concentration levels of Escherichia coli were observed after feeding the rats with DON for 4 weeks (P < 0.05). In human, the shift in the proportion of Bacteroides is highly associated with diseases as individuals with Crohn's disease or celiac diseases often exhibit a higher abundance of Bacteroides than healthy individuals (Kamada, 2016). A recent study, however, revealed that DON has no significant changes in the diversity and relative abundance of gut microbiota based on 16S rRNA microbiota analysis (Payros et al., 2017).

Gratz et al. (2017) demonstrated that masked T-2 toxin was released as a parent mycotoxin by human gut microbiota, and thereby contribute to mycotoxin exposure. Trichothecene mycotoxins are generally known as ribotoxic stress inducer which effectively blocks eukaryotic 28S rRNA. Thus, in theoretical aspect, the T-2 toxin would not interfere with bacterial protein translation and growth as suggested by Schmeits et al. (2014). Nevertheless, this is in contrast to the observation seen in a study conducted by Tenk et al. (1982). It was shown that the administration of T-2 toxin for 1 week was sufficient to induce a substantial increase in the aerobic bacteria count in the intestine of swines and rats (Tenk et al., 1982). While the bacterial populations have been shown to be greatly affected by trichothecene, the mechanism which causes the perturbation of bacterial population remains to be elucidated.

An in vitro experiment demonstrated that zearalenone compounds were converted into unknown metabolites by human gut microbiota (Gratz et al., 2017). The first study on the effect of ZEA on gut microbiota has been carried out by Piotrowska et al. (2014). The changes in gut microbiota were evaluated using Biolog EcoPlate method which only allows the quantification of culturable bacteria. After 6 weeks of ZEA ingestion, the data showed the concentration of Clostridium perfringens, Enterobacteriaceae, and E. coli was significantly reduced (P < 0.05).

Using capillary electrophoresis single-stranded conformation polymorphism (CE-SSCP), it is shown that fumonisins decreased the fecal microbiota SSCP-profiles similarity of the fumonisins-treated swines, compared to the untreated control group. The results indicated that there is an increase in the diversity of microbiota. The balance of the digestive microbiota was transiently but markedly affected after 63 days of chronic exposure to fumonisin, which consists of a mixture of FB1 and FB2 (Burel et al., 2013).

A study investigating the effect of OTA on gut microbiota using bioreactors has been carried out by Ouethrani et al. (2013), in which each bioreactor represents different parts of the adult human gut (Ouethrani et al., 2013). Based on the study, the gut microbiota degradation of OTA and microbiota diversity alteration were observed only at the descending colon after 1-week exposure to OTA. PCR-TTGE targeting Lactobacilli populations showed that Lactobacillus reuteri present during the start-up period, was permanently disappeared at the end of the OTA treatment period accompanied by some minor changes in the bifidobacteria population. The alteration was explained by a significant reduction in acetic, butyric and total SCFA concentration (P < 0.05). The reduction of beneficial microbes, lactobacillus and bifidobacteria indicated that the OTA shifted the microbiota balance and possibly led to impaired immunity.

In vivo study in rat demonstrated that OTA treatment decreased the diversity of the gut microbiota (Guo et al., 2014). The authors also reported that the relative abundance of Lactobacillaceae was increased whereas the Bacteroidaceae was decreased. Moreover, at the genus level, the OTA decreased the population of Bacteroides, Dorea, Escherichia, Oribacterium, Ruminococcus, and Syntrophococcus, while increased the number of Lactobacillus. The results showed Lactobacillus was more resistant to OTA and Lactobacillus may play a role in OTA detoxification process. Apart from that, it was also reported that the total facultative anaerobes were increased by the OTA treatment. In fact, the increase in facultative anaerobes was often observed in individuals with health complication (Shimizu et al., 2011) as well as in the elderly (Rudi and Avershina, 2015). This may further suggest that the OTA may cause negative effects on the host health via gut microbiota modulation.

In contrast to the intense research on AFB1 untoward effects, little information is available in regard to the outcomes of AFB1 on the gut microbiota. The findings from Wang et al. (2015) suggested that AFB1 could alter the gut microbiota in a dose-dependent manner. AFB1 decreased phylogenetical diversity but increased evenness of community composition. Although there was no changes at the phylum level, some lactic acid bacteria were significantly (P < 0.05) reduced in the presence of AFB1 (Wang et al., 2015). The reduction of LAB in animals treated with a lower dosage of AFB1 explains the severe immune malfunction induced by lower dosage of AFB1 (Qian et al., 2014).

A recent study showed that AFB1 at a dosage level of 1 ppm significantly (P < 0.05) reduced total LAB in the broiler. In contrast, the total number of Gram-negative bacteria and LAB were significantly (P < 0.05) increased in the group of broiler fed with 1.5 and 2 ppm of AFB1 (Galarza-Seeber et al., 2016). Besides, it has also been reported that 2.5 ppm of AFB1 increased the production of total volatile fatty acids in broilers, which suggested the association of higher dosage of AFB1 with a higher prevalence of LAB in the intestine (Kubena et al., 2001). Interestingly, a separate study showed greater numbers of bacterial mutants were recovered from mice exposed to AFB1 (Rowland, 1988). The data implies that the genotoxic effects of AFB1 not only affecting the host, but the gut microbiota as well.

On the other hand, the effect of a combination of mycotoxins on the modulation of gut microbiota has also been investigated. The exposure of gilts to ZEA and DON was found to pose an adverse impact on mesophilic aerobic bacteria. In particular, the amounts of C. perfringens, E. coli, and other bacteria in the family Enterobacteriaceae were reduced significantly after the 6th week of the experiment (P < 0.05). Nevertheless, the biodiversity of microorganisms in the gut was increased. Apart from that, an increase in the metabolism of amino acid by the gut microbiota was also observed. It was suggested that the increased metabolism of amino acid may be detrimental due to the formation of biogenic amines and procarcinogenic compounds (Piotrowska et al., 2014). Besides, a study confirmed AF and fumonisin mixture increased Shiga Toxin-producing E. coli (STEC) level in fecal (Baines et al., 2013). The composition of STEC-secreted cytotoxin was affected as reflected in the elevated concentration of intracellular Ca²⁺ with a corresponding increase in cytotoxicity. Mycotoxins are capable of altering the microbial balance of the intestine. Furthermore, the possible pathway proposed is via oxidative stress induced by mycotoxins (Vinderola and Ritieni, 2014). Nonetheless, the mechanisms by which mycotoxins affect the gut bacterial composition however remain unclear.