What is already known on this topic
Bronchiectasis is a progressive inflammatory disease characterized by abnormal widening of the airway and diverse symptoms including heavy mucus production, diminished ability to clear mucus, chronic cough, and shortness of breath. Previous bacterial microbiome analyses have shown clinical correlations between the abundance of certain bacterial genera and exacerbation—a period of heightened inflammation in the course of the disease. However, longitudinal studies have not shown significant changes in bacterial communities associated with exacerbation.
What this research adds
Researchers used a novel Integrative Microbiomics tool to merge bacteriome, virome, and mycobiome data from airways of bronchiectasis patients to capture more authentic in vivo dynamics among microbial communities. They further incorporated longitudinal samples to study patterns associated with exacerbation and resolution. Contrary to the previous bacterial microbiome studies, patients with higher rates of exacerbation showed significantly lower abundance of the bacterial genus, Pseudomonas, and further exhibited exclusionary interactions toward other microbes.
The Integrative Microbiomics tool utilizes a multi-biome approach to offer more robust pattern discovery and a view of the interactome—how the bacterial, fungal, and viral microbial groups influence one another. It could potentially delineate subtypes of bronchiectasis and other heterogeneous respiratory diseases to create more coherent exacerbation risk and treatment response prediction models.
A study published in Nature Medicine revealed that microbial interactions, rather than the presence, absence, or relative abundance of certain genera, are associated with bronchiectasis exacerbation.
Bronchiectasis is a progressive inflammatory disease characterized by abnormal widening of the airway and diverse symptoms including heavy mucus production, chronic cough, frequent respiratory tract infections, and shortness of breath. A heterogeneous etiology ranging from cystic fibrosis and pneumonia to foreign substances in the airway and unmanaged asthma make it complicated to study and treat. Previous bacterial microbiome analyses have correlated exacerbation—periods of heightened inflammation in the course of disease progression—with a less diverse airway bacterial microbiome and higher prevalence of a particular genus of gram-negative bacteria, Pseudomonas. However, longitudinal studies have not shown significant changes in bacterial communities associated with exacerbation.
In this study, the researchers, led by Micheál Mac Aogáin and Jayanth Kumar Narayana of Lee Kong Chian School of Medicine of Nanyang Technological University in Singapore, used Integrative Microbiomics, which is available to the public as a web tool hosted by Nanyang Technological University in Singapore. It employs a computational method known as similarity network fusion (SNF) which creates networks of data for each of the available data types (e.g. bacteriome, virome, mycobiome) and then iteratively merges them and weights the data, in this case, by the relative number of taxa in each microbial biome. The tool enables exploratory data analysis called spectral clustering—a form of multivariate statistical analysis. The objective is to assemble the complex, merged dataset into a few groups, or clusters, based upon similarities in any of the parameters to discover predictor attributes for a defined outcome.
When applied to the bronchiectasis data, the weighted SNF clearly revealed two clusters of study subjects, and those who had a higher frequency of exacerbation episodes exhibited lower airway alpha diversity, lower total numbers of microbes within the network, higher body mass index (BMI), greater use of inhaled corticosteroids, were more likely to be of European origin, and were more likely to report a history of smoking. Additionally, in contrast to the previous bacterial microbiome studies, patients with higher rates of exacerbation showed significantly lower levels of Pseudomonas abundance. Whereas forced expiratory volume test of lung function and bronchiectasis disease severity indices were not significantly different between the two clusters of study subjects, the Integrative Microbiomics tool distinguished the study subjects with more robust precision than the bacterial microbiome studies or clinical markers alone.
The investigators collected sputum samples from 217 patients from Scotland, Malaysia, Kuala Lumpur, and Singapore with moderate to severe bronchiectasis as well as from 40 healthy control subjects and established bacterial, viral, and fungal microbial records for each subject using 16S rRNA gene sequencing. The Interactome describes interactions among microbes which were defined as co-occurrence vs. co-exclusion. The patients who fell into the spectral cluster characterized by higher frequency of exacerbation episodes exhibited lower ecological diversity in their airway microbiomes and more negative—or exclusionary—interactions toward other microbes in the community.
The researchers have elucidated a disrupted interactome associated with exacerbation frequency, which suggests that beyond the mere presence or absence of specific microbes, the interactions among microbes may inform clinical exacerbation. The authors theorize that antibiotic treatment of exacerbations may be disrupting the interactome rather than targeting specific bacterial organisms.
The team conducted two separate, smaller scale studies to validate their observations. The first was a prospective cohort study with patients in Scotland to track the interactome longitudinally before, during, and after antibiotic treatment for exacerbation. This investigation supported the findings that microbial interactions were more potent predictors of time to next exacerbation than microbial abundance. The second study in collaboration with a team in Italy applied metagenomic sequencing to a separate cohort of bronchiectasis patients. Here, interactome analysis again revealed two clearly delineated clusters of patients distinguished by exacerbation frequency. The patients in the higher exacerbation frequency cluster exhibited a significantly different virome profile, including that of bacteriophages, and greater abundance of multidrug resistance genes than patients in the low frequency exacerbation cluster. This provided further support for the clinical predictive value and reproducibility of interactome analysis.
The scientists utilized a merged multi-biomics dataset that incorporated bacterial, viral, and fungal microbiome data.
Drilling down for further detail within the high frequency exacerbation cluster, the analysis demonstrated that species in the Pseudomonas genus behaved in a significantly different manner toward other microbes than in the low frequency exacerbation cluster.
By elucidating the interactions among the microbial taxa, the Integrative Microbiomics tool could potentially define subtypes of bronchiectasis and other respiratory diseases to create more coherent exacerbation risk prediction models and analyze interventions. This approach was able to identify the shift toward more competitive microbial relationships during periods of bronchiectasis exacerbation. This had not been previously captured by studies involving genetic sequencing to identify microbes and determine abundance.
The authors indicated that future studies should consider competition among microbes for substrate, the influence of bacteriophages, and host immune response to shifts in the microbial interactome.