Improved characterization of medically relevant fungi in the human respiratory tract using next-generation sequencing

Abstract

Background: Fungi are important pathogens but challenging to enumerate using next-generation sequencing because of low absolute abundance in many samples and high levels of fungal DNA from contaminating sources.

Results: Here, we analyze fungal lineages present in the human airway using an improved method for contamination filtering. We use DNA quantification data, which are routinely acquired during DNA library preparation, to annotate output sequence data, and improve the identification and filtering of contaminants. We compare fungal communities and bacterial communities from healthy subjects, HIV+ subjects, and lung transplant recipients, providing a gradient of increasing lung impairment for comparison. We use deep sequencing to characterize ribosomal rRNA gene segments from fungi and bacteria in DNA extracted from bronchiolar lavage samples and oropharyngeal wash. Comparison to clinical culture data documents improved detection after applying the filtering procedure.

Conclusions: We find increased representation of medically relevant organisms, including Candida, Cryptococcus, and Aspergillus, in subjects with increasingly severe pulmonary and immunologic deficits. We analyze covariation of fungal and bacterial taxa, and find that oropharyngeal communities rich in Candida are also rich in mitis group Streptococci,a community pattern associated with pathogenic polymicrobial biofilms. Thus, using this approach, it is possible to characterize fungal communities in the human respiratory tract more accurately and explore their interactions with bacterial communities in health and disease.

Figure 1

Proportional abundance of fungal genera in oropharyngeal wash, bronchoalveolar lavage, and contamination control samples. Many genera with high proportional abundance appear in only a few samples. For simplicity, only genera that were detected in 10 OW or 10 BAL samples are identified by name (full data in Additional file 2).

Figure 2

Quantification of post-PCR ITS DNA is linear in the range of oropharyngeal, lung, and contamination control samples. (A) Two serial dilutions of a standard sample of S. cerevisiae showed a linear response between input and post-PCR PicoGreen-quantification in the range yielding 0 to 60 ng/μL of ITS product. (B) Serial dilutions of two oropharyngeal wash samples from lung transplant recipients also showed a linear response in this range. (C) Genomic DNA from S. cerevisiae spiked into an oropharyngeal wash sample resulted in a linear increase in total DNA concentration within the 0 to 60 ng/μL range, as measured by post-PCR PicoGreen quantification. (D) Although the total concentration of post-PCR ITS DNA differed between sample types, the median concentration of contamination control samples was about 10% of the concentration in oropharyngeal wash. The median concentration of lung samples was only a few times that of contamination controls.

Figure 3

A global threshold for PicoGreen-corrected OTU abundance identifies genera in experimental samples that are extremely unlikely to arise from contamination sources. (A) Histogram of PicoGreen-corrected OTU abundances in control samples. The inset plot shows abundances above the 95% limit of the distribution, colored by contamination sample type. (B) Agreement between the PicoGreen-corrected abundance of fungi computed from ITS sequencing and clinical culture results in BAL samples. (C) ROC curve of post-PCR ITS abundance vs. culture results for all cultured fungi. The sequencing method is a good predictor of culture results (AUC = 0.93).

Figure 4

Heatmap of fungal genera identified in a matched set of oropharyngeal and lung samples following curation by ITS abundance.

Figure 5

Presence-absence analysis identifies fungal genera present more often in experimental samples relative to contamination controls. Aspergillus and Penicillium are present significantly more often in oropharyngeal and lung samples relative to controls, while Pichia and Saccharomyces are present more often only in OW relative to controls. Conversely, Wallemia likely derives solely from contamination sources.

Figure 6

Bacterial covariation with Candida in OW samples. (A) Principal coordinates analysis of weighted UniFrac distance between bacterial communities for oropharyngeal wash samples. Group centroids are shown with open triangles. The PicoGreen-corrected abundance of Candida has a significant effect on bacterial community composition (PERMANOVA P = 0.004, F = 3.8). (B) The QPCR-corrected abundance of Streptococcus, Rothia, and Veillonella increased with the abundance of Candida (Spearman correlation, P <0.05 after FDR correction). (C) Ten of the top 20 most abundant Streptococcus OTUs were found to increase with Candida abundance (Spearman correlation, P <0.05 after FDR correction).