Study of the viral and bacterial communities associated with asthma: a metagenomic approach

Introduction: An important unexplored source of interpersonal variation in asthma presentation is the complex and dynamic community of microorganisms located on different sites of the respiratory tract. Our aim was to identify and explore the bacterial and viral diversity in a small pilot study of asthmatic children using metagenomics.

Method: Nasopharyngeal samples were obtained from five asthmatic children (2 males) at a baseline asymptomatic visit. Mean age was 5.2 years (range 4.6–5.8). Communities of viruses (DNA and RNA genomes) and bacteria were evaluated using high depth paired-end shotgun sequencing. We performed taxonomic classification of sequences and calculated environmental indexes (alpha and beta) to estimate the microbial diversity both within (Shannon, Chao1, Jack1) and between (Jaccard) patients. Rarefaction curves were produced to study the relationship between microbial diversity and the number of generated sequences and/or number of samples. A correlation matrix was used to identify positive and negative correlations between bacterial and viral families and the results were superimposed in a network graph.

Results: Taxon classification revealed 198 bacterial and 20 viral families. Unsupervised hierarchical clustering based on beta diversity (Jaccard matrix) clustered controlled, partly controlled and uncontrolled asthma separately. Topographical analysis of the metagenomic interaction map identified the phage family of Myoviridae as a central component. Further analysis of the Myoviridae family and its direct interactions with bacterial families (17 families) revealed that patients with partly controlled and uncontrolled asthma have reduced quantity of Myoviridae phages and reduced diversity and quantity of their bacterial interactors.

Conclusion: Based on the Rarefaction curves, we estimate that, even with 5 samples a high percentage of the microbial diversity can be robustly represented. Asthma patients were efficiently clustered based on their microbial similarity which seems to be representative of asthma control. The phage family Myoviridae is predicted to hold a central role in the viral-bacterial interaction on the specific environmental site. The reduced number of reads of the Myoviridae phages along with the reduced diversity and quantity of 17 bacterial families in uncontrolled asthma could be evident of an ongoing ‘fight’ between components of the upper respiratory metagenome.