| Application of culture-independent techniques have revealed the presence of more types of bacteria than were previously thought, which led to the current description of Cystic Fibrosis (CF) being a polymicrobial disease. We know this polymicrobial community changes over time and during exacerbation events, and that interactions with non-pathogenic taxa can influence pathogen gene expression. The polymicrobial nature of infection may explain why in vitro responses and susceptibility of bacteria such as Pseudomonas aeruginosa to antibiotics do not always correlate with in vivo outcomes. The ability of bacteria to adapt to the CF lung complicates long-term treatment strategies. Much is known about the involvement of P. aeruginosa in lung colonization and deterioration, and the genetic adaptations it undergoes over time. Less is known about the adaptations that enable another CF pathogen, Burkholderia multivorans, to become resistant to antibiotics and persist in the lung environment. We identified a B. multivorans strain that acquired resistance in vivo to an antibiotic and became the dominant strain within a period of four days.;Expectorated sputum samples are the gold standard for identifying the pathogens present in the CF lungs. Sputum in CF is primarily composed of free DNA from host immune cells and bacterial cells which is markedly different from the normal mucus that lines the lung epithelia. This composition, along with the dehydrated nature of sputum, increases viscosity and heterogeneity of bacterial distribution. Culture-independent assays which examine bacterial diversity and abundance in sputum rely on bacterial DNA extracted from aliquots which may not be representative of the whole sample. Sputum is typically homogenized through chemical means prior to DNA extraction but we have shown that adding a mechanical homogenization step significantly increases bacterial distribution within a sputum sample.;Acute bacterial infections are the major cause for pulmonary exacerbations (PE) in Cystic Fibrosis. PEs are connected to increased mortality and may result in a permanent impairment in lung function. Attempts at developing tools to predict an oncoming PE have been met with limited success due to the heterogeneity of patient characteristics. We analyzed bacterial DNA from 130 sputum samples collected weekly for three years to identify changes in bacterial diversity and abundance by combining frequent patient sampling, next generation sequencing, and quantitative PCR (qPCR). Approximately 81,000,000 sequences containing 150 taxa were identified. Changes in microbial diversity and abundance did not correlate to antibiotic treatment for a PE. A gradual increase in abundance of all bacteria, Pseudomonas, and Burkholderia was shown over the sampling period along with a gradual decline in lung function. Ours is the first to demonstrate a stable microbial diversity coupled with a gradual change in abundance of all bacteria, Pseudomonas, and Burkholderia over a multi-year period.;Regardless of the specific goal, it is clear that to understand CF infections requires knowledge of more than the dominant pathogen. The data described in this dissertation demonstrate the importance of repeated, longitudinal sampling for studying microbial communities in human subjects where some variation in microbial community composition can occur, even between sequential samples from a single clinically stable patient. |
