Responses of Aspergillus flavus to Oxidative Stress Are Related to Fungal Development Regulator, Antioxidant Enzyme, and Secondary Metabolite Biosynthetic Gene Expression

Abstract

The infection of maize and peanut with Aspergillus flavus and subsequent contamination with aflatoxin pose a threat to global food safety and human health, and is exacerbated by drought stress. Drought stress-responding compounds such as reactive oxygen species (ROS) are associated with fungal stress responsive signaling and secondary metabolite production, and can stimulate the production of aflatoxin by A. flavus in vitro. These secondary metabolites have been shown to possess diverse functions in soil-borne fungi including antibiosis, competitive inhibition of other microbes, and abiotic stress alleviation. Previously, we observed that isolates of A. flavus showed differences in oxidative stress tolerance which correlated with their aflatoxin production capabilities. In order to better understand these isolate-specific oxidative stress responses, we examined the transcriptional responses of field isolates of A. flavus with varying levels of aflatoxin production (NRRL3357, AF13, and Tox4) to H₂O₂-induced oxidative stress using an RNA sequencing approach. These isolates were cultured in an aflatoxin-production conducive medium amended with various levels of H₂O₂. Whole transcriptomes were sequenced using an Illumina HiSeq platform with an average of 40.43 million filtered paired-end reads generated for each sample. The obtained transcriptomes were then used for differential expression, gene ontology, pathway, and co-expression analyses. Isolates which produced higher levels of aflatoxin tended to exhibit fewer differentially expressed genes than isolates with lower levels of production. Genes found to be differentially expressed in response to increasing oxidative stress included antioxidant enzymes, primary metabolism components, antibiosis-related genes, and secondary metabolite biosynthetic components specifically for aflatoxin, aflatrem, and kojic acid. The expression of fungal development-related genes including aminobenzoate degradation genes and conidiation regulators were found to be regulated in response to increasing stress. Aflatoxin biosynthetic genes and antioxidant enzyme genes were also found to be co-expressed and highly correlated with fungal biomass under stress. This suggests that these secondary metabolites may be produced as part of coordinated oxidative stress responses in A. flavus along with antioxidant enzyme gene expression and developmental regulation.

Keywords: Aspergillus flavus; aflatoxin; aflatrem; kojic acid; oxidative stress.

Figure 1

Venn diagrams of genes expressed with an FPKM ≥ 2. Genes exhibiting an FPKM ≥ 2 in at least one isolates and treatment were considered to be expressed. Genes expressed within all isolates and treatments (A), and between treatments within the NRRL3357 (B), AF13 (C), and Tox4 (D) isolates were compared using Venn diagrams generated with Venny 2.1.0.

Figure 2

Principal component analysis (PCA) and hierarchical clustering analysis (HCA) of the isolate gene expression profiles. (A) Principal component analysis (PCA) of the isolate expression patterns with PC1 and PC2 contributing 33.57 and 25.98% of overall variance, respectively. Similar clusters are indicated by the colored circles. (B) Hierarchical clustering analysis (HCA) of the isolate expression patterns. Both analyses indicate a similarity in the expression profiles of AF13 and highly stressed Tox4 with NRRL3357 exhibiting a more distinct profile.

Figure 3

Heatmap of aflatoxin, aflatrem, and kojic acid biosynthetic gene expression with increasing oxidative stress. The expression of aflatoxin (A), aflatrem (B), and kojic acid (C) biosynthetic genes are plotted with colors corresponding to the level of expression of the genes in each condition. Here, FPKM expression values were log₁₀ transformed and plotted based on the color key increasing from blue to red using MeV 4.9. Isolate and treatment combinations for each column are consistent for each heatmap.

Figure 4

Relative expression and FPKM associations for select secondary metabolite genes. Real-time PCR was used to determine the relative expression of six select secondary metabolite genes: pksA and vbs (aflatoxin), atmG and atmQ (aflatrem), and kojA and SVOP (kojic acid). Relative expression is indicated by the blue bars scaled to the left y-axis. The RNA sequencing-derived FPKM data is indicated by the red line scaled to the right y-axis. The plots are organized into three columns corresponding to data obtained from NRRL3357, AF13, Tox4, respectively, with each plot containing data from the 0, 10, and 20/25 mM H₂O₂ treatments in increasing order.

Figure 5

Gene ontology (GO) analysis of genes differentially expressed between 0 and 20/25 mM H₂O₂. Gene ontology (GO) analysis of significant DEGs between 0 and 20/25 mM H₂O₂ were determined using Blast2GO. Bars indicate the number of genes differentially expressed within each gene ontology group (biological process). The light blue bar represents Tox4, the gray bar AF13, and the dark blue bar NRRL3357.

Figure 6

Weighted gene co-expression network analysis of isolate transcriptional responses to oxidative stress. (A) Topographical overlap matrix (TOM) of the co-expression network of 500 randomly selected genes in the dataset. Genes are sorted by hierarchical clustering, and represented by the rows and columns with the colors showing the strength of the associations between genes increasing from white to red. Genes clustered tightly together are assigned to color-coded modules shown below the dendrograms. (B) Dendrogram produced by hierarchical clustering of the eigengene network showing the relationship of the modules and isolate dry weight under oxidative stress. (C) Eigengene adjacency heatmap showing the correlation between the modules and isolate dry weight. Mutual correlations between corresponding modules is higher than most module associations with dry weight, however the purple and turquoise modules exhibit a strong positive correlation with dry weight. In contrast, the blue module shows a strong negative correlation. Eigengene adjacency, which reflects correlation in the heatmap, increases from blue to red in the heatmap.