Abstract
OBJECTIVES: Legionella pneumophila is the leading cause of Legionnaires' disease, a severe form of pneumonia acquired from environmental sources. Investigations of both sporadic cases and outbreaks rely mostly on analysis of a single to a few colony pick(s) isolated from each patient. However, because of the lack of data describing diversity within single patients, the optimal number of picks is unknown. Here, we investigated diversity within individual patients using sequence-based typing (SBT) and whole-genome sequencing (WGS). METHODS: Ten isolates of L. pneumophila were obtained from each of ten epidemiologically unrelated patients. SBT and WGS were undertaken, and single-nucleotide polymorphisms (SNPs) were identified between isolates from the same patient. RESULTS: The same sequence type (ST) was obtained for each set of ten isolates. Using genomic analysis, zero SNPs were identified between isolates from seven patients, a maximum of one SNP was found between isolates from two patients, and a maximum of two SNPs was found amongst isolates from one patient. Assuming that the full within-host diversity has been captured with ten isolates, statistical analyses showed that, on average, analysis of one isolate would yield a 70% chance of capturing all observed genotypes, and seven isolates would yield a 90% chance. CONCLUSIONS: SBT and WGS analyses of multiple colony picks obtained from ten patients showed no, or very low, within-host genomic diversity in L. pneumophila, suggesting that analysis of one colony pick per patient will often be sufficient to obtain reliable typing data to aid investigation of cases of Legionnaires' disease.