Ex Vivo and In Vitro Studies Revealed Underlying Mechanisms of Immature Intestinal Inflammatory Responses Caused by Aflatoxin M1 Together with Ochratoxin A

Abstract

Aflatoxin M1 (AFM1) and ochratoxin A (OTA), which are occasionally detected in milk and commercial baby foods, could easily enter and reach the gastrointestinal tract, posing impairment to the first line of defense and causing dysfunction of the tissue. The objective of this study was to investigate the immunostimulatory roles of individual and combined AFM1 and OTA on the immature intestine. Thus, we used ELISA assays to evaluate the generation of cytokines from ex vivo CD-1 fetal mouse jejunum induced by AFM1 and OTA and explored the related regulatory pathways and pivot genes using RNA-seq analysis. It was found that OTA exhibited much stronger ability in stimulating pro-inflammatory cytokine IL-6 from jejunum tissues than AFM1 (OTA of 4 μM versus AFM1 of 50 μM), whereas the combination of the two toxins seemed to exert antagonistic actions. In addition, transcriptomics also showed that most gene members in the enriched pathway 'cytokine-cytokine receptor interaction' were more highly expressed in OTA than the AFM1 group. By means of PPI network analysis, NFKB1 and RelB were regarded as hub genes in response to OTA but not AFM1. In the human FHs 74 Int cell line, both AFM1 and OTA enhanced the content of reactive oxygen species, and the oxidative response was more apparent in OTA-treated cells in comparison with AFM1. Furthermore, OTA and AFM1 + OTA raised the protein abundance of p50/RelB, and triggered the translocation of the dimer from cytosol to nucleus. Therefore, the experimental data ex vivo and in vitro showed that OTA-induced inflammation was thought to be bound up with the up-regulation and translocation of NF-κB, though AFM1 seemed to have no obvious impact. Since it was the first attempt to uncover the appearances and inner mechanisms regarding inflammation provoked by AFM1 and OTA on immature intestinal models, further efforts are needed to understand the detailed metabolic steps of the toxin in cells and to clarify their causal relationship with the signals proposed from current research.

Keywords: RNA-seq; aflatoxin M1; immature intestine; inflammation; ochratoxin A.

Figure 1

Effects of AFM1, OTA and AFM1 + OTA treatment for 24 h on the release of (A–C) LDH, (D–F) IL-6, and (G–I) TNF-α from the isolated jejunal tissues. In panel (C,F,I), the concentration of AFM1 and OTA was 50 μM and 4 μM, respectively. Results were shown as mean ± SEM (n ≥ 6). * p < 0.05 statistically significantly compared with control.

Figure 2

(A) Principal component analysis (PCA) plot, (B) Number of differentially expressed genes (DEGs) screened out by comparison between groups, and (C) Venn diagram plot for transcripts among there toxin treatment compared to DMSO vehicle group.

Figure 3

WGCNA analysis to identify groups of genes. (A) Clustering diagram showing the co-expression modules recognized by WGCNA. Different colors denote different modules. The longitudinal distance indicates the distance between genes, while the horizontal distance is meaningless. (B) The number of genes contained in each module. (C) Heatmap of gene expression pattern of blue module. (D) Heatmap of gene expression pattern of yellow module.

Figure 4

(A) Protein–protein interaction (PPI) network analysis of 1788 up-regulated genes in ‘blue’ module plus 16 up-regulated genes in ‘yellow’ module. (B) The network was divided into 9 small blocks using ‘Reactome FIviz’ plugin in Cytoscape. Colors represent different blocks. (C) PPI network analysis of genes in block 1. In the network, the edges represent gene interactions, and the size and color depth of every node is proportional to the degree of connexity.

Figure 4

(A) Protein–protein interaction (PPI) network analysis of 1788 up-regulated genes in ‘blue’ module plus 16 up-regulated genes in ‘yellow’ module. (B) The network was divided into 9 small blocks using ‘Reactome FIviz’ plugin in Cytoscape. Colors represent different blocks. (C) PPI network analysis of genes in block 1. In the network, the edges represent gene interactions, and the size and color depth of every node is proportional to the degree of connexity.

Figure 5

Effects of AFM1 and OTA alone and in combination on the viability and cytokines expression of FHs 74 Int Cells. Cellular viability of FHs 74 Int cells which were treated with (A) various doses of AFM1, (B) various doses of OTA, (C) selected lower and higher doses of AFM1 and OTA alone and in combination for 48 h using CCK-8 analysis. The data of treatment groups are normalized to control (untreated) as the basal 100%. (D) IL-6 and (E) TNF-α production in the cells treated with toxins were detected by ELISA tests. (F) IL-6, (G) TNF-α expression levels in FHs 74 cells detected by RT-qPCR. Gene expression were calculated relative to the internal control (GAPDH) (set as 1.0). Results were shown as mean of three separate experiments ± SEM (n ≥ 6). * p < 0.05 significantly compared with control, # p < 0.05 significantly compared with combined AFM1 and OTA treatment.

Figure 6

Effect of single and combined AFM1 and OTA on IκBα degradation and NF-κB activation in FHs 74 Int cells. (A) Relative protein expression levels of IκBα (n = 3). (B) NFKB1, and (C) RelB mRNA expression levels in FHs 74 Int cells detected by RT-qPCR (n = 6). Relative protein expression of p50 and RelB in (D) total protein, (E) nuclear protein, and (F) cytoplasmic protein of cells (n = 3). The bands of IκBα, p50 and RelB were detected by western blotting and quantitated by density analysis tool of Image J software. The data was represented as mean ± SEM. All values in treatment group were compared and normalized to control (set as 1.0). Different lowercase indicates statistical differences between groups (p < 0.05).

Figure 7

(A) Effects of individual and combined AFM1 and OTA on ROS production in FHs 74 Int cells for 48 h (n = 3). The data of treatment are normalized to control (untreated) as the basal 100%. (B) Effects of ROS inhibitor NAC on the IL-6 production in cells treated with toxins for 48 h (n = 6). (C) Effects of ROS inhibitor NAC on the TNF-α production in cells treated with toxins for 48 h (n = 6). The data was represented as mean ± SEM. Different lowercase indicates statistical differences between groups (p < 0.05). Con represents the untreated group, AFM1 represents AFM1 at 50 μM, OTA represents OTA at 2 μM, and M + O represents 50 μM AFM1 + 2 μM OTA, NAC represents NAC at 5 μM.