Metagenomic analysis of respiratory tract DNA viral communities in cystic fibrosis and non-cystic fibrosis individuals

Abstract

The human respiratory tract is constantly exposed to a wide variety of viruses, microbes and inorganic particulates from environmental air, water and food. Physical characteristics of inhaled particles and airway mucosal immunity determine which viruses and microbes will persist in the airways. Here we present the first metagenomic study of DNA viral communities in the airways of diseased and non-diseased individuals. We obtained sequences from sputum DNA viral communities in 5 individuals with cystic fibrosis (CF) and 5 individuals without the disease. Overall, diversity of viruses in the airways was low, with an average richness of 175 distinct viral genotypes. The majority of viral diversity was uncharacterized. CF phage communities were highly similar to each other, whereas Non-CF individuals had more distinct phage communities, which may reflect organisms in inhaled air. CF eukaryotic viral communities were dominated by a few viruses, including human herpesviruses and retroviruses. Functional metagenomics showed that all Non-CF viromes were similar, and that CF viromes were enriched in aromatic amino acid metabolism. The CF metagenomes occupied two different metabolic states, probably reflecting different disease states. There was one outlying CF virome which was characterized by an over-representation of Guanosine-5'-triphosphate,3'-diphosphate pyrophosphatase, an enzyme involved in the bacterial stringent response. Unique environments like the CF airway can drive functional adaptations, leading to shifts in metabolic profiles. These results have important clinical implications for CF, indicating that therapeutic measures may be more effective if used to change the respiratory environment, as opposed to shifting the taxonomic composition of resident microbiota.

Figure 1. Mapping of best tBLASTx hits to the phage proteomic tree by percentage for Non-CF (A) and CF (B) viromes.

The phage genome with the highest percentage of hits (normalized to the length of the genome) is labeled for each virome.

Figure 2. Principal components analysis (PCA) of respiratory tract viromes based on phage taxonomic composition.

Non-CF metagenomes are shown in blue and CF metagenomes are shown in red. Inputs to PCA were normalized percentages of best tBLASTx hits to completely sequenced phage genomes. Non-CF1Asthma and Non-CF4Spouse cluster with the CF metagenomes.

Figure 3. Putative host range for phage communities in respiratory tract viromes.

Host range was inferred from normalized best tBLASTx hits to phage genomes. Host ranges for CF viromes and for Non-CF1Asthma and Non-CF4Spouse were not statistically significantly different as determined by XIPE and were combined.

Figure 4. Distribution of normalized best tBLASTx hits to DNA and Retro-transcribing eukaryotic viruses in Non-CF(A) and CF(B) individuals.

Reticuloendotheliosis virus is indicated by the red rectangle in (B).

Figure 5. Principal components analysis based on best tBLASTx hits to 3074 eukaryotic viruses.

Non-CF viromes are shown in blue and CF viromes are shown in red.

Figure 6. Non-metric multidimensional (NM-MDS) scaling of top-level SEED metabolic subsystems.

All Non-CF metagenomes are shown in blue. CF1-5 are shown in red. The inputs to NM-MDS were the number of hits to subsystems in the highest level of the SEED hierarchy.

Figure 7. Distribution of similarities to metabolic subsystems in respiratory viromes.

(A) Distribution of top-level subsystems in respiratory tract viromes. (B) Second-level subsystems from the SEED hierarchy for aromatic metabolism. Non-CF viromes are shown in blue, CF viromes 1–4 are shown in red, and CF5 is in gray. Subsystems determined by XIPE to be over-represented in a particular group are marked with a (+) while those that are under-represented are marked with an asterisk (*).