Fig 1.
The transcription factor encoded by Afu1g09190 is important for resistance against oxidative stress.
a-d, Strains were grown for 5 days from 105 spores at 37°C on glucose minimal media supplemented with increasing concentrations of the oxidative stress-inducing compounds (a) ally alcohol, (b) acrolein, (c) menadione and (d) t-butyl hydroperoxide (t-butyl). Graphs indicate the % of growth in the presence of the respective drug with respect to the control condition (without drug). Graphs are the quantitation of radial growth of the pictures shown in the same panel, with standard deviations representing three biological replicates (*P-value < 0.01; **P-value < 0.001; ***P-value < 0.0001 in a two-way ANOVA test).
Fig 2.
The Afu1g09190-encoded TF regulates the transcription of gliotoxin (GT) biosynthetic genes and of gtmA through directly binding to the promoter regions in the presence of allyl alcohol (AA).
a, AA induces transcriptional up-regulation of genes required for the oxidative stress response whilst down-regulating genes encoding proteins involved in growth in the A. fumigatus wild-type (WT) strain. Functional categorisation (FunCat) description of the 5 most significant enriched categories (p-value < 0.0005), that contain up- (UP) or down (DOWN)-regulated genes with a -1 > Log2 fold-change > 1 (p-value < 0.05), as identified by RNA-sequencing. The WT strain was exposed to 10 mM AA for 30 min after 24 h growth in glucose minimal medium (GMM, control) and gene expression was compared between both conditions. b, Fumagillin and GT biosynthetic genes are severely repressed in the ΔAfu1g09190 strain. Heatmap of the Log2FC values, as determined by RNA-sequencing, of genes that were significantly (p-value < 0.05) differently (-1 > Log2FC < 1) expressed between the WT and ΔrglT in the presence of 10 mM AA for 30 min. Genes were classified into virulence-relevant categories, such as oxidative stress-related proteins, transcription factors, secondary metabolite biosynthesis and transporters. Log2FC gene expression values are also shown for the WT strain, when comparing gene expression in the presence of AA to the control condition. Hierarchical clustering was performed in MeV (http://mev.tm4.org/), using Pearson correlation with complete linkage clustering. c, Validation of the RNA-sequencing data by RT-qPCR. RNA was extracted from strains grown for 24 h in GMM (control) before 10 mM AA was added for 30 min and RT-qPCR was performed. Gene expression values were normalized by gene Afu1g05340, whose expression remained constant across all conditions, as confirmed by RNA-seq. Standard deviations represent three biological replicates (**P-value < 0.001; ***P-value < 0.0001 in a one-way ANOVA test). d, Chromatin immunoprecipitation coupled to DNA sequencing (ChIP-seq) of the WT and Afu1g09190::HA strains, when grown for 24 h in GMM (control) and then exposed to 10 mM AA for 30 min, identified 538 RglT::HA binding peaks that could be classified into three main groups: i) unique (150) or enriched binding (122) in the control condition, ii) consistent binding (160) in both the control and AA conditions and iii) unique (67) or enriched (39) binding in the AA condition. e, Afu1g09190 binds to the promoter regions of gliZ, gliM, gliA, gliF and gliT and was subsequently named RglT (regulator of GT). Heat map depicting ChIP-seq signal for all genes of the GT biosynthetic gene cluster in the control and AA conditions. Also shown is the differential binding analysis fold change (FC) between both conditions. f, RNA-seq results are in agreement with the ChIP-seq results, confirming that GT biosynthetic genes are under the direct transcriptional control of RglT. Heat maps depicting RNA-seq FPKM values for the WT and ΔrglT strains as well as Log2FC (fold change) values between the WT to the ΔrglT strain of all gli genes in the control and AA conditions.
Table 1.
Minimal inhibitory concentrations (MIC) of fumagillin and gliotoxin on the growth of the wild-type (WT) and ΔAfu1g09190 strains.
Standard deviations are shown for 3 independent replicates for the two strains in the presence of each mycotoxin.
Fig 3.
RglT is crucial for gliotoxin (GT) biosynthesis.
a, HPLC (high performance liquid chromatography) analysis for GT of supernatants of different strains, that were grown for 72 h in GT-inducing conditions, showed similar chromatographs for all strains except for the ΔgliT strain (negative control). Standard GT (positive control) eluted after 8.45 min at 254 nm. All experiments were performed in biological triplicates. b, UV-vis spectra of the ΔrglT and ΔgliT strains differ from the spectra of the wild-type (WT) and ΔrglT::rglT+ strains, indicating the absence of GT. c, LC-MS (Liquid Chromatography-Mass Spectrometry) of biological triplicates, confirms the absence of GT from the supernatant of the ΔrglT strain. Shown are the extended chromatograms of organic extracts from: (1, brown) standard gliotoxin; (2, orange) WT; (3, dark green) ΔrglT::rglT; (4, pale green) ΔrglT; (5, blue) ΔgliT. d, The ΔrglT strain does not produce GT. Diagram summarizing in which strains GT or BmGT [bisdethiobis(methylthio)gliotoxin] was detected. e, RglT is essential for gliZ, gliT, gliF and gtmA expression in GT-inducing conditions. RT-qPCR was performed after strains were grown for 72 h in GT-inducing conditions. Gene expression was normalized by the β-tubulin-encoding gene. Standard deviations represent three biological replicates (**P-value < 0.001, ***P-value < 0.0001 in a one-way ANOVA test). f, RglT binds to the promoter regions of gliZ, gliT and gliF in GT-inducing conditions. ChIP (chromatin-immunoprecipitation)-qPCR was performed for the WT and RglT::HA strains when grown for 48 h in GT-inducing conditions. Binding motifs were identified by ChIP-sequencing and binding was calculated using the % input method, normalizing the RglT::HA-specific binding by the total cellular DNA (input). Standard deviations represent three biological replicates (**P-value < 0.001, ***P-value < 0.0001 in a paired, one-tailed student t-test). g-i, Endogenous GT resistance is completely abolished in the ΔrglT strain. Strains were grown for 5 days at 37°C on glucose minimal medium supplemented with or without GT (g, h). Graphs (i) show the percentage of growth in the presence of GT when compared to the control, no GT condition and standard deviations represent biological triplicates (***P-value < 0.0001 in a two-way ANOVA test).
Fig 4.
RglT is essential for A. fumigatus virulence.
a, b Bone marrow-derived murine C57BL/6 (wild-type) and CGD (chronic granulomatous disease) macrophages phagocytize (a) and kill (b) a significant higher number of ΔrglT conidia in vitro than when compared to the wild-type (WT) and ΔrglT::rglT strains. Standard deviations represent three biological replicates (*P-value < 0.05 in a two-tailed, unpaired student t-test). c, Survival curve of chemotherapeutic BALB/c mice, which were immunosuppressed with a combination of cyclophosphamide and hydrocortisone, infected intra-nasally with the WT, ΔrglT and ΔrglT::rglT strains show that the ΔrglT strain is hypovirulent (***p-value < 0.0001 in the Mantel-Cox and Gehan-Brestow-Wilcoxon tests, comparing the ΔrglT to the WT and ΔrglT::rglT strains). d, Fungal burden of the naïve control (CTRL = phosphate buffered saline), WT, ΔrglT and ΔrglT::rglT strains show a significant (***p-value < 0.0001 in a two-tailed, unpaired student t-test) reduction of the ΔrglT growth in chemotherapeutic murine lung tissue at 3 days post-infection. e, Histopathology of murine lungs at three days post-infection with the WT, ΔrglT and ΔrglT::rglT strains indicates the absence of ΔrglT proliferation in the lung tissue. 5 μm lung tissue sections were stained with haematoxylin and eosin (HE) or with Grocott’s Methenamine Silver (GMS) before slides were viewed and pictures taken. f, Tumor necrosis factor alpha (TNF-α) and g, interleukin (IL)-12p40 concentrations in macrophages infected with heat killed (HK) conidia or live-resting (LR) conidia.
Fig 5.
The homologue of A. fumigatus RglT is important for gliotoxin (GT) self-protection, through regulating the expression of the gliT homologue in the non-GT-producing fungus A. nidulans.
a, The phylogenetic distribution of RglT, GliT and GT biosynthetic gene cluster homologs across 458 fungal genomes. A 3-gene phylogeny of genomes from Eurotiomycetes (shown by the red branches) and Sordariomycetes (shown by the blue branches). For every tip in the phylogeny, the presence of RglT, GliT and gliotoxin BGC homologs is depicted using orange, purple, and green bars, respectively; absences are depicted in white. The grey bar plots depict how many of the 13 Gli genes are present in the gliotoxin BGC homolog. The tip corresponding to the A. fumigatus Af293 genome is indicated by a red dot, and the tip corresponding to the A. nidulans A4 genome is indicated by a maroon dot. b, Clustal W alignment between AfRglT and AnRglT (An1368) shows 54% identity at the protein level (e-value 8e-137). c-f, AnRglT (AN1368) is important for resistance against oxidative stress and GT self-protection. Strains were grown for 5 days from 105 spores at 37°C on glucose minimal media supplemented with increasing concentrations of (c) gliotoxin, (d) allyl alcohol, (e) t-butyl hydro-peroxide (t-butyl) and (f) menadione. Legend for all graphs is displayed. Graphs indicate the % of growth in the presence of the respective drug with respect to the control condition (without drug). Standard deviations represent three biological replicates (*P-value < 0.01; **P-value < 0.001; ***P-value < 0.0001 in a two-way ANOVA test). g-i, The expression of (g) AngliT (AN3969) and (h) AfgliT, as determined by qRT-PCR, is transcriptionally dependent on RglT when GT is added exogenously for 3 h at a final concentration of 5 μg/ml. As a control, (i) AfgliT gene expression was confirmed to be independent of GliZ. Gene expression was normalised by tubulin gene expression in both fungi. Standard deviations represent three biological replicates (***P-value < 0.0001 in a two-way ANOVA test).