Culture enriched molecular profiling of the cystic fibrosis airway microbiome

Abstract

The microbiome of the respiratory tract, including the nasopharyngeal and oropharyngeal microbiota, is a dynamic community of microorganisms that is highly diverse. The cystic fibrosis (CF) airway microbiome refers to the polymicrobial communities present in the lower airways of CF patients. It is comprised of chronic opportunistic pathogens (such as Pseudomonas aeruginosa) and a variety of organisms derived mostly from the normal microbiota of the upper respiratory tract. The complexity of these communities has been inferred primarily from culture independent molecular profiling. As with most microbial communities it is generally assumed that most of the organisms present are not readily cultured. Our culture collection generated using more extensive cultivation approaches, reveals a more complex microbial community than that obtained by conventional CF culture methods. To directly evaluate the cultivability of the airway microbiome, we examined six samples in depth using culture-enriched molecular profiling which combines culture-based methods with the molecular profiling methods of terminal restriction fragment length polymorphisms and 16S rRNA gene sequencing. We demonstrate that combining culture-dependent and culture-independent approaches enhances the sensitivity of either approach alone. Our techniques were able to cultivate 43 of the 48 families detected by deep sequencing; the five families recovered solely by culture-independent approaches were all present at very low abundance (<0.002% total reads). 46% of the molecular signatures detected by culture from the six patients were only identified in an anaerobic environment, suggesting that a large proportion of the cultured airway community is composed of obligate anaerobes. Most significantly, using 20 growth conditions per specimen, half of which included anaerobic cultivation and extended incubation times we demonstrate that the majority of bacteria present can be cultured.

Figure 1. Bacterial genera recovered from CF sputum during 28 years of using conventional cultivation approaches.

The proportional abundance of each genus in the culture collection (19 250 isolates) is depicted with solid circles.

Figure 2. The abundance of 3% OTUs (percent of the total isolates in the culture collection) generated by using non-conventional approaches for microbial cultivation from CF sputum and the phylogenetic relationship between recovered isolates.

OTUs that are represented in greater than 1% of the entire culture collection (2 015 isolates) are highlighted on the phylogenetic tree with proportionally sized solid circles according to the legend above the abundance plot. OTUs with less than 97% identity to any 16S rRNA sequence in public databases are indicated with red branches.

Figure 3. The aerotolerance of the airway microbiome was evaluated by assigning T-RFs to three broad categories representing obligate anaerobes, facultative anaerobes and obligate aerobes.

An example of a T-RFLP profile generated directly from sputum is shown (A). The three T-RFs highlighted (size in bp is shown) in the sputum profile provide representative examples of each category, which were determined based on the culture-enrichment conditions under which each T-RF was recovered (B). The overall proportion of each organism category in the six patients investigated is illustrated (C).

Figure 4. Culture-enrichment increases the number of discernable community members as compared to direct molecular detection from sputum (conventional) by using T-RFLP.

Figure 5. Each patient's sputum required a unique set of culture conditions to represent the complete collection of T-RFs recovered by enrichment.

The culture conditions required for inclusive representation of cultivated richness from each sample (shown horizontally) are shaded according to the number of unique T-RFs they recovered according to the legend provided. Culture conditions that are not shaded did not recover unique T-RFs. Culture conditions currently recommended for CF microbiology are indicated with an asterisk.

Figure 6. The vast majority of bacterial families detected in this study were detected with a culture-dependent approach.

Culture-dependent techniques included culture-enrichment (from four patients) and using expanded culture conditions and picking representative colonies (culture collection from 117 patients). Five families were detected only with a culture-independent approach (deep 16S rRNA sequencing). The families in each sector of the Venn diagram are listed in Table S7.

Figure 7. The abundance of each species detected by 16S rRNA sequencing is shown as a rank order by percentage of the total sequences recovered from each patient (A).

16S rRNA sequencing from the enrichment pools was used to determine the percentage of organisms detected by direct molecular detection from sputum that were recovered by culture-enrichment (B). The percentage of cultured species from the four patients is shown as cumulative total as sequencing depth is increased (solid bars) and as a percentage of only the organisms represented at a specific concentration range (open bars).