Comparative transcriptomics of Aspergillus fumigatus strains upon exposure to human airway epithelial cells

Abstract

Aspergillus fumigatus is an opportunistic, ubiquitous, saprophytic mould that can cause severe allergic responses in atopic individuals as well as life-threatening infections in immunocompromised patients. A critical step in the establishment of infection is the invasion of airway epithelial cells by the inhaled fungi. Understanding how A. fumigatus senses and responds to airway cells is important to understand the pathogenesis of invasive pulmonary aspergillosis. Here, we analysed the transcriptomes of two commonly used clinical isolates, Af293 and CEA10, during infection of the A549 type II pneumocyte cell line in vitro. We focused our RNA-seq analysis on the core set of genes that are present in the genomes of the two strains. Our results suggest that: (a) A. fumigatus does not mount a conserved transcriptional response to airway epithelial cells in our in vitro model and (b) strain background and time spent in the tissue culture media have a greater impact on the transcriptome than the presence of host cells. Our analyses reveal both common and strain-specific transcriptional programmes that allow for the generation of hypotheses about gene function as it pertains to pathogenesis and the significant phenotypic heterogeneity that is observed among A. fumigatus isolates.

Keywords: Af293; Aspergillus fumigatus; CEA10; RNA-seq; airway epithelial cells.

Fig. 1.

Global analysis of differential gene expression across all samples. (a) Hierarchical cluster analysis, based on log₂-transformed TPM, of all 7888 differentially expressed A. fumigatus core genes after growth in tissue culture medium with or without host cells. Each column represents an individual sample (n=24) as designated by the colour code above the dendrogram. Each row represents a different gene. Red indicates high gene expression. Blue indicates low gene expression. (b) PCA of TPM values from all differentially expressed genes in all 24 samples.

Fig. 2.

Heatmaps of clusters of differentially expressed genes. Groups of co-expressed genes revealed by k-means cluster analysis. Each column represents an individual sample (n=24). Values represent log₂-transformed TPM values. Red indicates high gene expression. Blue indicates low gene expression. Clusters 12, 14 and 5 represent common responses to time in the culture model independent of the presence of host cells. Clusters 15, 3 and 16 represent groups of genes with strain-specific patterns of gene expression.

Fig. 3.

Differential expression of genes in response to host cells. (a) Circos plot comparing the host-cell-induced differential expression (FDR<0.01; LFC≥|1.0|) of the two A. fumigatus strains at each time-point of infection. The four numbered tracks represent individual comparisons of infection sample groups with time-matched negative controls: Track 1, CEA10 at 16 h post-infection (h p.i.); track 2; Af293 at 16 h p.i.; track 3, CEA10 at 6 h p.i.; track 4, Af293 at 6 h p.i. Red represents up-regulation in response to incubation with host cells. Green represents down-regulation under the same conditions. Genes that did not exhibit significant differential expression are absent from this plot. Chromosomal location is indicated by the outer black line and the adjacent numbers (1–8). (b) Venn diagrams representing the overlap in gene regulation between the two strains.

Fig. 4.

Fusarinine C biosynthetic genes are induced during exposure to A549 cells. (a) Simplified schematic of FSC and TAFC biosynthesis by the enzymatic activities of the sid genes (adapted from [32]). (b) Expression of FSC biosynthetic genes in each of the eight experimental groups. Values represent the average of the log₂-transformed normalized TPM values for each group. Red indicates high gene expression. Blue indicates low gene expression. The expression of sidG, required for conversion of FSC to TAFC, was not induced by exposure to host cells.