Functional genomics of human bronchial epithelial cells directly interacting with conidia of Aspergillus fumigatus

Abstract

Background: Aspergillus fumigatus (A. fumigatus) is a ubiquitous fungus which reproduces asexually by releasing abundant airborne conidia (spores), which are easily respirable. In allergic and immunocompromised individuals A. fumigatus can cause a wide spectrum of diseases, including allergic bronchopulmonary aspergillosis, aspergilloma and invasive aspergillosis. Previous studies have demonstrated that A. fumigatus conidia are internalized by macrophages and lung epithelial cells; however the exact transcriptional responses of airway epithelial cells to conidia are currently unknown. Thus, the aim of this study was to determine the transcriptomic response of the human bronchial epithelial cell line (16HBE14o-) following interaction with A. fumigatus conidia. We used fluorescence-activated cell sorting (FACS) to separate 16HBE14o- cells having bound and/or internalized A. fumigatus conidia expressing green fluorescent protein from cells without spores. Total RNA was then isolated and the transcriptome of 16HBE14o- cells was evaluated using Agilent Whole Human Genome microarrays.

Results: Immunofluorescent staining and nystatin protection assays demonstrated that 16HBE14o- cells internalized 30-50% of bound conidia within six hrs of co-incubation. After FAC-sorting of the same cell culture to separate cells associated with conidia from those without conidia, genome-wide analysis revealed a set of 889 genes showing differential expression in cells with conidia. Specifically, these 16HBE14o- cells had increased levels of transcripts from genes associated with repair and inflammatory processes (e.g., matrix metalloproteinases, chemokines, and glutathione S-transferase). In addition, the differentially expressed genes were significantly enriched for Gene Ontology terms including: chromatin assembly, G-protein-coupled receptor binding, chemokine activity, and glutathione metabolic process (up-regulated); cell cycle phase, mitosis, and intracellular organelle (down-regulated).

Conclusions: We demonstrate a methodology using FACs for analyzing the transcriptome of infected and uninfected cells from the same cell population that will provide a framework for future characterization of the specific interactions between pathogens such as A. fumigatus with human cells derived from individuals with or without underlying disease susceptibility.

Figure 1

Localization of A. fumigatus conidia within the epithelial cell monolayer. GFP-expressing A. fumigatus conidia and 16HBE14o- cells were co-incubated for 6 hrs, fixed and then stained with DAPI to localize nuclei (blue) and an antibody to adherens junction protein, E-cadherin, to localize cell membranes (red), which were visualized using confocal microscopy. GFP-expressing A. fumigatus conidia (green) found within 16HBE14o- cells, are highlighted by the white arrows in the XY plane and white circles in the XZ and YZ planes.

Figure 2

Differential staining of extracellular and internalized conidia by anti-A. fumigatus antibody. GFP-expressing A. fumigatus conidia and 16HBE14o- cells were co-incubated for 6 hrs, fixed and then stained with DAPI, to label cell nuclei, and a polyclonal anti-A. fumigatus antibody, to label extracellular conidia, before visualization by confocal microscopy. One representative field is shown in the following channels: A) wavelength 495 nm for GFP (green); B) wavelength 594 nm for anti-A. fumigatus antibody (red); C) wavelength 405 nm for DAPI (blue); D) merged GFP, anti-A. fumigatus antibody and DAPI image. Conidia not labeled by the anti-A. fumigatus antibody, and therefore only visible in the green but not red channel, were considered to be internalized by 16HBE14o- cells, allowing quantification of internalization.

Figure 3

Extent of internalization of A. fumigatus conidia by epithelial cells determined by nystatin protection assay. A. fumigatus conidia and 16HBE14o- cells were co-incubated for the indicated times, then treated with nystatin-supplemented or nystatin-free media for 3 hrs. Cells were lysed, and recovered conidia were plated to count colony forming units. At each time point, the rate of internalization was determined as the number of colony forming units recovered from nystatin-treated wells divided by the number from control-treated wells. The mean % of internalization ± SD from three biological replicates is shown for each time point.

Figure 4

FACS analysis of A. fumigatus conidia and epithelial cells incubated alone or together. Cultures of GFP-expressing A. fumigatus conidia alone (A), 16HBE14o- cells alone (B), or cells and conidia incubated together for 6 hrs (C) were analyzed by flow cytometry, using forward scatter (FSC) and side scatter (SSC). The gates defining conidia (labeled 'spores', coloured green) and 16HBE14o- cells (labeled 'cells', coloured blue and purple) were set manually based on the observed distributions by FSC and SSC. Histograms showing the FITC channel signal intensity from the selected gates for (D) conidia alone, (E) 16HBE14o- cells alone, and (F) co-cultures of 16HBE14o- cells and conidia. The bimodal distribution of FITC intensity allowed the sorting of 16HBE14o- cells positive for GFP from negative cells (Figure 4G).

Figure 5

Re-analysis of positive and negative sorted samples to determine the accuracy of sorting. FACS-sorted negative (A) and positive (B) 16HBE14o- cell samples were re-analyzed to determine the accuracy of sorting based on fluorescence intensity signals. Histograms showing the FITC channel signal intensity from the selected gates for negative (C) and positive (D) 16HBE14o- cells for GFP-expressing conidia. Dot plots showing fluorescence intensity in FITC and PE-Texas Red channels associated with negative (E) and positive (F) 16HBE14o- cells for GFP-expressing conidia.

Figure 6

Microscopic visualization of negative and positive sorted samples to determine the accuracy of sorting. FAC-sorted negative (A, B) and positive (C, D) cell samples were visualized by differential interference contrast (DIC) and fluorescence microscopy to verify the accuracy of sorting based on fluorescence intensity signals. Shown for each sample are DIC images (A, C) and green fluorescence images (B, D). Individual 16HBE14o- cells are clearly visible in the DIC images, while GFP-expressing A. fumigatus conidia present as bright green spots in the green fluorescence channel. The majority of cells in the negative samples were free of conidia, while most cells in positive samples were associated with at least one conidia spore.