Toxins | Free Full-Text | Metagenome Analysis Identifies Microbial Shifts upon Deoxynivalenol Exposure and Post-Exposure Recovery in the Mouse Gut

1. Introduction

Fusarium fungi are world-wide producers of a range of mycotoxins. Deoxynivalenol (DON) belonging to the group B trichothecenes, is one of the most prevalent food-associated mycotoxins mainly produced by Fusarium graminearum and Fusarium culmorum, and frequently contaminates cereals and cereal products [1,2]. Almost half of a total of 26,613 cereal samples collected from 21 European countries were found to be contaminated by DON, with the highest levels observed in wheat, maize, and oat grains [2]. A survey conducted by Mishra et al. [3] summarized the occurrence of DON in food and related human risk assessments in the past 10 years. Their report shows that current DON exposure levels may pose a health risk to consumers, especially children. The concentration of urinary DON in children was higher than that in adults and the elderly, and researchers suggested that necessary measures must be taken to ensure health [1,4]. The most common route of exposure to DON is through dietary intake [3]. DON is known to cause a variety of adverse effects, of which many are related to the gastrointestinal tract and its functioning. The intestinal tract serves as the main target organ [5,6]. The acute effects are, among others, nausea, vomiting, gastro-intestinal (GI) upset, dizziness, diarrhea, and headache [7,8]. Previous studies reported that chronic exposure to DON can cause intestinal toxicity, including induction of intestinal lesions, affecting cell proliferation and differentiation, altering the intestinal barrier function, and it also can cause immunotoxicity, hematotoxicity, and myelotoxicity [7,9,10,11,12,13].
Trillions of microbes with more than 700–1000 different bacterial species reside in the gut, representing the gut microbiota [14]. Many factors such as pH variation, diet, mucus, host immunity, and environmental factors have been shown to influence the biogeography and composition (both abundance and diversity) of bacteria along the GI tract [15]. Xenobiotics are important environmental factors which can interact with gut microbiota [16]. Previous studies reported different effects of DON on the microbiota in different animal models and at different dose levels (0, 2.5, 5, and 10 mg per kg diet) [17]. DON was reported to cause a decrease in total fecal bacterial numbers [11], and changes in the richness (Chao1) and evenness index (Shannon) [11,12]. Notably, DON treatment altered intestinal microbiota, resulting in increases and decreases at the phylum, family, or genus level [13,17]. At the phylum level, the relative abundance of Firmicutes and Proteobacteria in chicken cecal digesta were reported to be increased upon DON exposure [12]. At the family level, in human fecal microbiota-inoculated germ-free rats, the relative abundance of the Clostridiaceae family increased, Ruminococcaceae remained unaffected, while Enterobacteriaceae decreased upon exposure of the rats to DON. At the genus level, the abundance of the Bacteroides/Prevotella group increased, while the Escherichia coli group decreased after feeding rats with DON [13]. In another study, the fecal microbiota of Wistar rats showed the genus Coprococcus to be increased upon exposure of the rats to DON [17].
Despite above-mentioned evidence on the effect of DON on gut microbiome composition, our understanding of the effects of DON exposure on gut microbiome function and of the post-exposure recovery process is limited. In this study, we employed a metagenome-wide association approach to identify microbial species and functional shifts that follow DON exposure, and to characterize natural recovery of the microbiome after termination of toxin exposure in the mouse gut microbiota. For the recovery, an intervention with the prebiotic inulin was also included in the studies. Interventions with inulin have been reported to modulate gut microbiota composition and function to promote gut health [18]. However, the impacts of inulin on post-DON exposure recovery of the gut microbiome have not been defined. Therefore, in this study, inulin was added as a dietary supplement to the mouse diet during the recovery period after DON exposure, in order to investigate whether this dietary intervention would facilitate the recovery of the DON-induced disturbances in the mouse gut microbiota as compared to a recovery phase without an extra intervention. The results obtained help to better understand the effect of DON on the gut microbiome, and the gut microbiota’s recovery upon termination of the exposure. The study also aims to understand the role of inulin supplementation in the recovery of the gut microbiome after DON exposure.

3. Discussion

We used whole genome gene sequencing to study the impact of DON exposure on the gut microbiome. The data obtained clearly show that DON exposure induced an obvious change in the gut microbiome composition of mice. Our observation and taxonomic annotation of the gut microbiota in mice were not fully consistent with a previous report by Wang et al. [19]. Wang et al. reported viruses as being the most dominant phylum across all samples in the control group, Firmicutes being the most abundant in the low-dose group (2.0 mg/kg body weight of DON), and Bacteroidetes, Firmicutes, and Deferribacteres being the most abundant in the high-dose group (5.0 mg/kg body weight of DON) in a mouse model. In our study, Firmicutes, Bacteroidetes, and Verrucomicrobia were predominant in the gut bacteria of mice at the phylum level, followed by Proteobacteria, Deferribacteres, and Actinobacteria. Only 0.1% of the detected complete genes were annotated as coming from viruses, pointing at a substantial difference compared to the study from Wang et al. [19]. The high prevalence of Firmicutes and Bacteroidetes was consistent between the two studies. A possible assumed reason for the discrepancies between the two metagenomic studies might be related to the fact that the mice used in the Wang et al. study were virus-infected, leading to different outcomes in metagenome composition. In addition to this, feed, water, and living environment exposure of the mice during the study may also cause differences in gut microbiome composition.
In this study, we observed that the relative abundance of Verrucomicrobia increased with DON exposure in a dose-dependent way. Verrucomicrobia is a phylum of Gram-negative bacteria, that is occasionally observed in humans. Broad-spectrum antibiotic therapy can cause high-level colonization of the human gut by Verrucomicrobia [20]. However, the mechanism underlying this colonization by Verrucomicrobia was not clear. A human progeria patient exhibited a similar tendency in Verrucomicrobia abundance, and long-lived humans such as centenarians were shown to display a substantial increase in Verrucomicrobia and a reduction in Proteobacteria [21].
DON exposure increased the relative abundance of species A. muciniphila, L.bacterium 28-4, and H. hathewayi. A. muciniphila belongs to the phylum Verrucomicrobia, and is a strict anaerobic, Gram-negative species of bacteria. A. muciniphila is one of the most abundant single species in intestinal microbiota, and is considered as a promising next-generation beneficial microbe [22]. It is a highly specialized bacteria because of its ability to use mucins as a sole source of carbon and nitrogen [23]. Meanwhile, A. muciniphila can stimulate mucin expression and mucus secretion by positive feedback [23]. This property reveals a competitive advantage under conditions of nutrient deprivation, such as, for example, fasting, malnutrition, or total parenteral nutrition. A. muciniphila releases easily available short-chain fatty acids, acetate, and propionate for the host as a result of mucin degradation. A. muciniphila was inversely associated with obesity, diabetes, cardiometabolic diseases, and low-grade inflammation [22]. A previous study observed that mice treated with A. muciniphila gained less weight, improved glucose tolerance, and exhibited insulin resistance under hyperlipidic diet conditions [24]. The administration of specific dietary components or pharmaceutical treatments affected the level of A. muciniphila, such as polyphenols, fructo-oligosaccharide, conjugated linoleic acid, oat bran, and metformin intake [25]. Our study shows that DON exposure can also increase the abundance of A. muciniphila. The reason may be due to the fact that DON causes damages to the intestinal barrier function, which may simulate A. muciniphila growth to repair the damages. The mode of action underlying this observation is of interest for future exploration.
L. bacterium 28-4 was enriched upon DON exposure, which was previously shown to be enriched in pigs with low residual feed intake (high feed efficiency) [26]. L. bacterium 28-4 was also investigated to be the dominant species in mice with resistance to high-fat diet-induced obesity [27]. A study conducted in pathogen-free C57BL/6J mice showed that pomegranate fruit polyphenols enriched A. muciniphila, L. bacterium 28-4, and three other species. These enriched species were negatively correlated with body weight, glucose, triglycerides, and total cholesterol levels in serum [28].
H. hathewayi was also enriched upon DON exposure, belonging to Lachnospiraceae, which was less likely to cause human diseases, and was proven to be a common fecal flora commensal [29]. H. hathewayi was reported to be involved in the progress of some diseases, such as unruptured intracranial aneurysm [30]. The abundance of H. hathewayi was positively associated with taurine concentration in mice and human circulation, and oral gavage with it can normalize taurine levels in serum. The taurine could protect mice from the formation and rupture of intracranial aneurysms, and H. hathewayi was negatively associated with the pathogenesis of the disease [30].
We noticed that the relative abundance of the species Oscillibacter sp. 1-3 decreased with DON exposure, but increased in the recovery groups (LD2 and HD2) compared to the levels of the control group (CK2) after 14 days of natural recovery in this study. Oscillibacter was proven to be causally linked to decreased triglyceride in the blood (Liu et al., 2022) [31]. Oscillibacter was also associated with obesity; the relative abundance of Oscillibacter was reduced in obese individuals [32]. Oscillibacter spp. and Akkermansia spp. showed a significantly negative correlation with lipopolysaccharide (LPS) levels in plasma [33]. Speculatively, the decreased abundance of Oscillibacter sp. 1-3 over time upon DON exposure could give rise to high levels of lipopolysaccharide (LPS) concentration in plasma. Oscillibacter belongs to butyrate-producing bacterial families, and was decreased in patients with early hepatocellular carcinoma (HCC) [34]. The genome of Oscillibacter sp. 1-3 encodes tryptophanase, capable of synthesizing indole from tryptophan. Oscillibacter sp. 1-3 and Firmicutes bacterium ASF500 have important characteristics in enhancing intestinal epithelial barrier functions and immunity [35]. Accordingly, the restructuring of microbial populations responding to DON exposure is likely the cause of perturbation of the intestinal barrier function. In addition, Oscillibacter was considered a harmful bacteria [36]. It was reported that Oscillibacter was associated with trimethylamine oxide (TMAO), which is considered as a risk factor for cardiovascular and cerebrovascular disease [37].
Gut microbiota resilience and natural recovery were reported after antibiotic administration [38]. Natural recovery of gut microbiota changes after mycotoxin exposure has not been investigated before. In this research, we showed that upon 2 weeks of natural recovery after DON exposure, mice gut microbiota communities returned to their initial state and fully recovered, especially in the low-dose exposure group. The Low-DON + Recover group (LD2) recovered better than the High-DON + Recover group (HD2) based on both the composition and function of the gut microbiome. In contrast to the low-dose DON exposure group, the changes observed upon high-dose DON exposure did not fully return to control levels in the 2-week spontaneous recovery period.
Dynamics of the microbial community are strictly correlated with disease development by altering the metabolic processes and/or the immune responses of the host [39]. More and more prebiotics have been characterized to enhance the beneficiary effects of gut microbiota substantially. One of the most commonly applied prebiotics in the restoration of intestinal microbiota is inulin. Inulin is extracted from chicory root, and is normally commercially used to stimulate the growth of Bifidobacterium and Lactobacillus [40]. In an earlier study conducted by Lin et al. [41], inulin showed the greatest elimination ability on a single-course amoxicillin-induced disruption of mice gut microbiota abundance and diversity enrichment. In our study, we observed a positive effect of inulin on reviving the destructed DON-induced gut microbiota disruption for the low-dose exposure group. However, in the high-dose group, the inulin supplementation showed a negative effect on gut microbiome recovery. It even further decreased the diversity of the mice gut microbiome, and led to dysbiosis compared with spontaneous recovery. After high-dose DON exposure, the mice gut microbiome composition and function were changed, and the supplementation of inulin appeared to amplify the differences in the gut microbiome by an as-yet unknown mode of action. Additionally, immediately following the treatment, the phylum Verrucomicrobia was dose-dependently increased in the DON groups (LD1 and HD1) compared to the corresponding control (CK1) at the cost of the Fermicutes. Furthermore, upon 14 days of natural recovery, this increase in Verrucomicrobia was not fully annihilated in the High-DON + Recover group (HD2) compared to the corresponding control (CK2).
While main research efforts to elucidate the impact of DON treatment on gut microbiota have mainly focused on community shifts [17], little is known about the functional consequences of these shifts for the cross-talk between gut microbial metabolism and host responses. In this study, we characterized the genes’ functional annotation in all experimental groups. All of the six functional gene clusters (environmental information processing, cellular process, human diseases, metabolism, organismal systems, genetic information processing) involved in the response to DON exposure exhibited differently in terms of their abundance. It is apparent to see that natural recovery can repair the gene function loss under low-dose DON exposure after 2 weeks. Unexpectedly, inulin supplements showed no obvious beneficial effects on the recovery of gene functions after DON perturbation.

5. Materials and Methods

5.1. Chemicals and Solutions

DON was purchased from Beijing Meizheng Testing Co., Ltd. (Beijing, China), purity ≥ 98%. Inulin was purchased from Shanghai Sangon Company (Shanghai, China). Mice daily feed was prepared by Xiaoshuyoutai Co., Ltd. (Beijing, China), according to AIN-93M standard.

5.2. Animals

A number of 72 BALB/c female mice, aged 6–7 weeks, weight 20–22 g (Vital River, Beijing, China), were housed at room temperature (25 ± 2) °C, under 12 h light and dark cycles. The mice were allowed access to food and water ad libitum, and were maintained with 4 animals per cage. The mice were housed in a standard SPF facility of the Institute of Food Science and Technology (IFST), Chinese Academy of Agricultural Sciences (CAAS). All the animal experiments were carried out under the approval and supervision of the ethics committee of IFST, CAAS (No. JGS-20181005). All the animal experiments were in accordance with the NIH Guide for the Care and Use of Laboratory Animals.

5.3. Animal Groups and Treatments

After 1 week of acclimation, 72 female mice were assigned to 9 different treatments randomly, 8 animals per group. 200 μL purified water with or without 2 or 5 mg/kg bw/day DON was administered to animals via intragastric infusion (IG) once daily. Chemical DON was suspended in purified water using ultrasound for 15 min.

To study the effects of DON on mice gut microbiota, the treatment groups were as follows: (1) CON group (CK1): purified water for 14 days; (2) Low-DON group (LD1): 2 mg/kg bw/day DON for 14 days; (3) High-DON group (HD1): 5 mg/kg bw/day DON for 14 days.

To study the 2-week recovery period after DON exposure, additional groups were treated as follows: (4) CON + Recover group (CK2): purified water for 14 days, followed by natural recovery for 14 days; (5) Low-DON + Recover group (LD2): 2 mg/kg bw/day DON for 14 days, followed by purified water and regular diet for 14 days; (6) High-DON + Recover group (HD2): 5 mg/kg bw/day DON for 14 days, followed by purified water and regular diet for 14 days. (7) CON + Inulin group (CK3): purified water for 14 days, followed by purified water and an inulin diet (5% inulin addition to AIN-93M) for 14 days; (8) Low-DON + Inulin group (LD3): 2 mg/kg bw/day DON for 14 days, followed by purified water and an inulin diet for 14 days; (9) High-DON + Inulin group (HD3): 5 mg/kg bw/day DON for 14 days, followed by purified water and an inulin diet for 14 days.

Mice body weight was measured every four days over the whole duration of the study. The mice were sacrificed after anesthesia on day 15 (groups 1–3) or day 29 (groups 4–9). Whole blood was collected from the mice orbit after sacrifice. Plasma was collected by centrifugation (3000 rpm, 20 min, 4 °C), and stored at −80 °C. Intestinal content (cecum) was collected after sacrifice, and stored at −80 °C until further analysis.

5.4. DNA Extraction, Library Construction and Sequencing

Frozen cecum contents from each group were used for metagenomics study. Genomic DNA was extracted with the QIAamp DNA stool mini kit (Qiagen, Valencia, CA, USA) following the protocol provided by the supplier. Extracted genomic DNA (2 ng/μL) was used for library preparation. The purity and integrity of the DNA was determined with a nanodrop (ND-1000) spectrophotometer (Nanodrop Technologies, Wilmington, DE, USA) through 1% agarose gel electrophoresis (AGE). DNA concentration was measured using a Qubit® dsDNA Assay Kit in a Qubit® 2.0 Fluorometer (Life Technologies, Carlsbad, CA, USA). Samples with A260/A280 values between 1.8–2.0 and a total mass of DNA above 1 μg were collected for metagenomic sequencing and used to construct the library. Sequencing libraries were generated using a NEBNext®Ultra™ DNA library Prep Kit for Illumina (NEB, lpswich, MA, USA) following the manufacturer’s recommendations and index codes were added to attribute sequences to each sample. Briefly, the DNA sample was sonicated into fragments of 350 bp on average, then DNA fragments were end-polished, A-tailed, and ligated with the full-length adaptor for Illumina sequencing with further PCR application. Finally, PCR products were purified (AMPure XP system) and libraries were analyzed for size distribution with an Agilent 2100 Bioanalyzer and quantified using real-time PCR. The clustering of the index-coded samples was performed on a cBot Cluster Generation System according to the manufacturer’s instructions. After cluster generation, the library preparations were sequenced on an Illumina HiSeq platform and paired-end reads were generated.

5.5. Sequencing Data Pretreatment and Metagenome Assembly

The raw data obtained from the Illumina HiSeq sequencing platform using Readfq were processed to acquire the clean data for subsequent analysis. Considering that the possibility of host pollution of the samples may exist, clean data were blasted to the host database which, by default, uses Bowtie 2.2.4 software to filter out the reads that are of host origin. The clean data were assembled and analyzed by SOAP de novo software (V2.04).

5.6. Gene Prediction and Abundance Analysis

The Scaftigs (≥500 bp) assembled from both single and mixed assemblies were all predicted by the ORF by MetaGeneMark (V 2.10) software, and the length information for fragments shorter than 100 nt was filtered from the predicted result with default parameters. CD-HIT software (V4.5.8) was adopted for redundancy and to obtain a unique initial gene catalogue. The clean data of each sample were mapped to an initial gene catalogue using Bowtie 2.2.4. The basic information statistics, core-pan gene analysis, correlation analysis of samples, and Venn figure analysis of the number of genes were all based on the abundance of each gene in the respective sample.

5.7. Taxonomic Assignment of Genes

DIAMOND software (V0.9.9) was used to analyze the unigenes via BLAST against the sequences of bacteria, fungi, archaea, and viruses which were all extracted from the NR database (Version: 20180102, www.ncbi.nlm.nih.gov/, accessed on 10 February 2019). For each sequence’s blast result, the best Blast Hit was used for subsequent analysis. A functional database including the KEGG database, eggNOG database, and CAZy database was used in this study. The least common ancestors (LCA) algorithm was applied to the system classification using MEGAN software to characterize the species annotation information of sequences. PCA (R ade4 package, Version 2.15.3) and NMDS (R vegan package, Version: 2.15.3) decrease dimension analyses were based on the abundance table of each taxonomic hierarchy.

5.8. Statistical Analysis

Results were expressed as mean ± SEM. Significances of differences between two or multiple groups were determined using a two-sided unpaired Student’s t-test or one-way analysis of variance (ANOVA). All analyses were performed at least in triplicate. Statistical analyses were performed using GraphPad Prism v9.0. p < 0.05 was considered to be statistically significant. Metastats and LEfSe analysis were used in Metastats analysis for each taxonomy and to obtain the p value, then the Benjamini and Hochberg false discovery rate procedure was used to correct the p value and acquire the q value. LefSe analysis was conducted by using LEfSe software. Random forest (RandoForest) (P pROC and randomForest packages, Version 2.15.3) was used to construct a random forest model. Important species were screened out by MeanDecreaseAccuracy and MeanDecreaseGin, and the receiver operating characteristic curve was plotted for cross-analysis validation of each model. The heat maps were generated using R language to visualize the gut microbiome differences between treatment groups.

Read more here: Source link