Reducing bias in bacterial community analysis of lower respiratory infections

Abstract

High-throughput pyrosequencing and quantitative PCR (Q-PCR) analysis offer greatly improved accuracy and depth of characterisation of lower respiratory infections. However, such approaches suffer from an inability to distinguish between DNA derived from viable and non-viable bacteria. This discrimination represents an important step in characterising microbial communities, particularly in contexts with poor clearance of material or high antimicrobial stress, as non-viable bacteria and extracellular DNA can contribute significantly to analyses. Pre-treatment of samples with propidium monoazide (PMA) is an effective approach to non-viable cell exclusion (NVCE). However, the impact of NVCE on microbial community characteristics (abundance, diversity, composition and structure) is not known. Here, adult cystic fibrosis (CF) sputum samples were used as a paradigm. The effects of PMA treatment on CF sputum bacterial community characteristics, as analysed by pyrosequencing and enumeration by species-specific (Pseudomonas aeruginosa) and total bacterial Q-PCR, were assessed. At the local community level, abundances of both total bacteria and of P. aeruginosa were significantly lower in PMA-treated sample portions. Meta-analysis indicated no overall significant differences in diversity; however, PMA treatment resulted in a significant alteration in local community membership in all cases. In contrast, at the metacommunity level, PMA treatment resulted in an increase in community evenness, driven by an increase in diversity, predominately representing rare community members. Importantly, PMA treatment facilitated the detection of both recognised and emerging CF pathogens, significantly influencing ‘core’ and ‘satellite’ taxa group membership. Our findings suggest failure to implement NVCE may result in skewed bacterial community analyses.

Changes in Cystic Fibrosis Airway Microbiota at Pulmonary Exacerbation

“…The approach that has been used for deep-sequencing in most human microbiomic studies does not differentiate DNA derived from live versus dead bacteria. Although the exclusion of DNA from nonviable bacterial cells with the use of reagents such as propidium monoazide has been proposed (32)(33)(34), there is no clear consensus among Human Microbiome Project investigators as to the best approach to deal with this issue. There are reasonable concerns that essentially any step in DNA processing presents an opportunity for introducing bias (e.g., propidium reagents do not interact with all cell types in the same way).…”

Complexity, Temporal Stability, and Clinical Correlates of Airway Bacterial Community Composition in Primary Ciliary Dyskinesia

“…Such analyses have typically revealed much greater bacterial diversity than is reported through standard diagnostic microbiology (7)(8)(9)(10)(11)(12), with a substantial contribution often made by species requiring anaerobic conditions for growth, including members of the genera Prevotella, Veillonella, Propionibacterium, and Actinomyces (8,13).…”

Directly Sampling the Lung of a Young Child with Cystic Fibrosis Reveals Diverse Microbiota

“…The predominant bacterial species identified were Ralstonia pickettii and P. acnes (Figure 2b and Tables E2 and E3). These two organisms, both identified as common in prior CF microbiota studies (3,(29)(30)(31), were not detected in any of the cultures. Taxa of the genus Anaerococcus, which, like Propionibacterium, includes anaerobes commonly found in CF sputum (31), were also detected in each section.…”