Assessing diversity of the female urine microbiota by high throughput sequencing of 16S rDNA amplicons

Abstract

Background: Urine within the urinary tract is commonly regarded as "sterile" in cultivation terms. Here, we present a comprehensive in-depth study of bacterial 16S rDNA sequences associated with urine from healthy females by means of culture-independent high-throughput sequencing techniques.

Results: Sequencing of the V1V2 and V6 regions of the 16S ribosomal RNA gene using the 454 GS FLX system was performed to characterize the possible bacterial composition in 8 culture-negative (<100,000 CFU/ml) healthy female urine specimens. Sequences were compared to 16S rRNA databases and showed significant diversity, with the predominant genera detected being Lactobacillus, Prevotella and Gardnerella. The bacterial profiles in the female urine samples studied were complex; considerable variation between individuals was observed and a common microbial signature was not evident. Notably, a significant amount of sequences belonging to bacteria with a known pathogenic potential was observed. The number of operational taxonomic units (OTUs) for individual samples varied substantially and was in the range of 20-500.

Conclusions: Normal female urine displays a noticeable and variable bacterial 16S rDNA sequence richness, which includes fastidious and anaerobic bacteria previously shown to be associated with female urogenital pathology.

Figure 1

Summary of the microbial phyla and orders detected in human female urine. A: An overview of the taxonomy at the phylum level as computed using MEGAN V3.4, using normalized counts by pooling together the V1V2 and V6 16S rDNA reads. The size of the circles is scaled logarithmically to the number of reads assigned to the taxon. Nodes denoted as "Not assigned" and "No hits" are the number of reads that were assigned to a taxon with fewer than 5 hits, or did not match to any sequence when compared to the SSUrdp database, respectively. B and C: Comparison of taxonomic assignments for human female urine sequences at the order level. Reads obtained using the V1V2 hypervariable 16S rDNA region were predominantly assigned to Lacobacillales, and identified in total 18 different orders where Desulfuromonadales and Spirochaetales are unique to this V1V2 dataset. V6 reads revealed a slightly higher diversity with 20 different orders; Bdellovibrionales, Myxococcales, Rhizobiales and Enterobacteriales are only identified by this V6 method.

Figure 2

Bacterial genera detected in healthy female urine. A: Comparison of healthy female urine bacterial genera abundance determined by sequencing 2 different hypervariable 16S rDNA regions, V1V2 and V6. Relative abundance of 18 major bacterial genera found in the sequence pool of eight different urine samples are shown for the two 16S rDNA regions. Groups denoted "other" represent minor groups classified. Y-axis represents relative abundance. B: Heat map showing the relative abundance of bacterial genera across urine samples of eight healthy females. Genera denoted as phylum_genus, samples denoted as samplenumber_V1V2 or V6. Taxa marked with asterisk (*) could not be assigned to any genera, and are shown at the lowest common taxon: family and order. Color intensity of the heat map is directly proportional to log 10 scale of the abundance normalized sequence data as done by MEGAN.

Figure 3

Number of OTUs as function of the total number of sequences. A and B: Rarefaction curves of individual samples for the V1V2 (A) and the V6 datasets (B). Curves were generated at 3% genetic difference using MOTHUR v1.17.0 [39]. C and D: Rarefaction curves of the pooled dataset for both V1V2 reads (C) and V6 reads (D). OTUs with ≤3%, ≤6% and ≤10% pairwise sequence difference generated using MOTHUR v1.17.0 [39] are assumed to belong to the same species, genus and family, respectively.