View Abstract

A6200 - Integrated Microbiome and Metabolome Analysis of Lung Sputum as a Novel Approach to Precision Medicine
Author Block: R. Quinn1, L. Deright-Goldasich1, G. Humphrey1, R. Knight1, D. J. Conrad2, P. C. Dorrestein1; 1University of California at San Diego, La Jolla, CA, United States, 2Univ of California At San Diego Med Ctr, La Jolla, CA, United States.
Rationale: Chronic lung diseases, such as cystic fibrosis, COPD and others, involve complex polymicrobial infections causing severe inflammation over decades. Exacerbations of these chronic diseases contribute to lung function decline and poor patient outcomes, but the cause of these events are poorly understood. Longitudinal analysis of sputum samples with integrated omics methods can provide a big-data science view of the dynamics of microbial infection, host inflammation and antibiotic therapy. Methods: Longitudinal sputum samples were collected from 6 cystic fibrosis patients in their home freezers for 1.5 years. Over 10 exacerbation events were captured in this data set including one that led to the death of a patient. A total of 597 sputum samples were collected and analyzed with 16S rRNA gene amplicon sequencing and LC-MS/MS mass spectrometry data for integrated microbiome and metabolome analysis. Results: Beta-diversity analysis of the metabolomic data revealed highly personalized sputum chemistry, where there was little overlap metabolome between patients through time. Dynamic changes through exacerbation events were observed in the metabolome including the detection of unique metabolites during exacerbation. Oral and intravenous antibiotics were detected in the data set enabling an integrated look at the presence and abundance of antibiotics in the lung and its impact on the lung microbiome. Metabolites from the bacterial pathogen Pseudomonas aeruginosa were detected in three of six patients, including small molecule virulence factors. These metabolites were only intermittently detected in the data and did not correspond to the abundance of the pathogen in microbiome data. P. aeruginosa metabolites were not detected in the days leading up to the fatal exacerbation event, but were detected months prior, indicating that this pathogen was not actively growing during this severe acute event. Conclusion: Integrated microbiome and metabolome data analysis on longitudinal samples of chronic lung disease provides and in depth look at the interactions between pathogens, host and antibiotic treatment. The patients in this study had personalized signatures in their omics data supporting the need for a personalized approach to the treatment of chronic lung infections. Co-analysis of metabolites from a pathogen with DNA based detection methods provides further information about the actual growth of the bacterium in a clinical sample.