Figures
Abstract
Background
The microbiome of the male urogenital tract is poorly described but it has been suggested that bacterial colonization of the male urethra might impact risk of sexually transmitted infection (STI). Previous cultivation-dependent studies showed that a variety of non-pathogenic bacteria colonize the urethra but did not thoroughly characterize these microbiomes or establish links between the compositions of urethral microbiomes and STI.
Methodology/Findings
Here, we used 16S rRNA PCR and sequencing to identify bacteria in urine specimens collected from men who lacked symptoms of urethral inflammation but who differed in status for STI. All of the urine samples contained multiple bacterial genera and many contained taxa that colonize the human vagina. Uncultivated bacteria associated with female genital tract pathology were abundant in specimens from men who had STI.
Citation: Nelson DE, Van Der Pol B, Dong Q, Revanna KV, Fan B, Easwaran S, et al. (2010) Characteristic Male Urine Microbiomes Associate with Asymptomatic Sexually Transmitted Infection. PLoS ONE 5(11): e14116. https://doi.org/10.1371/journal.pone.0014116
Editor: Raphael H. Valdivia, Duke University, United States of America
Received: July 23, 2010; Accepted: November 2, 2010; Published: November 24, 2010
Copyright: © 2010 Nelson et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: JD Fortenberry and GM Weinstock were supported by National Institutes of Health grants UH2KD083980 and U54HG004968, respectively. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Sequencing of bacterial 16S rRNA alleles from the human skin, mouth, gastrointestinal tract and vagina has revealed that each of these sites supports surprisingly diverse microbial communities, including substantial numbers of uncharacterized and uncultivated species [1], [2]. These microbial communities are believed to play a key role in maintaining health, and altered microbiomes are associated with a variety of conditions including obesity [3], [4], inflammatory bowel disease [5], [6], Crohn's disease [6] and bacterial vaginosis (BV) [7], [8], [9], [10].
Use of non-culture based methods to characterize the male urethra in health and disease has lagged behind investigations in other body sites. The hypothesis that male urethral microbiomes could impact STI has been tested using primarily cultivation dependent microbial identification methods [11], [12], [13], [14], [15], [16], [17]. A culture-based study of distal urethral swabs identified diverse bacterial populations in specimens from healthy men, men with chlamydial urethritis, and men with non-chlamydial, non-gonococcal urethritis (NGU) [17]. Another study found that Lactobacilli spp. were more frequent in specimens from men who did not have NGU as opposed to NGU positive men [11]. Willen reported that first-catch urine specimens from healthy adult men contained aerobic and anaerobic bacteria, and many of these organisms were absent from corresponding prostrate secretion samples [16]. A study using 16S rRNA allele restriction fragment length polymorphisms (RFLP) identified diverse microbial communities and substantial intra- and inter-person variability in distal urethral swabs from men with urethritis compared to those from healthy men [15]. Thus, it is clear that diverse bacteria commonly colonize the male urethra but there are few data about the normal composition of male urethral microbiomes or if these microbiomes are relevant to male STI.
Here we used 16S rRNA PCR and DNA sequencing to characterize microbial communities in first catch urine, as a proxy for urethral swabs, collected from a cohort of sexually active adult men at high risk for STI. A high degree of inter-subject variability in urine microbiomes was evident but clustering analyses indicated striking differences in the microbiomes of urines from groups of STI positive and negative individuals. In summary, our results indicate that urine from sexually active men often contains complex microbial communities and that the composition of these urine communities is relevant to STI.
Materials and Methods
Ethics Statement
All procedures for collection and handling of patient samples and data were approved by the Indiana University-Clarian Institutional Review Board study 0710-80, “Genomic analysis of male urethral micro-environment”. All study subjects provided written informed consent to participate in the study.
Specimens
Urine specimens were obtained from the Indiana University Infectious Disease Laboratory. Convenience urine specimens were collected from 19 men over the age of 18, at an STD clinic and were frozen at −80°C on the day of collection. Criteria for inclusion in the study were no overt symptoms of urethral infection (e.g. urethral discharge or dysuria) and negative for urethritis (defined as <5 neutrophils per high power field in first catch urine). All men had refrained from taking antibiotics in at least the previous week. Clinical histories and demographic data were available for all subjects and sexual histories were provided by 10/19 (Tables S1 and S2).
STI testing
Urine specimens were tested for C. trachomatis and N. gonorrhoeae using the Aptima Combo 2 CT/GC assay (Gen-Probe, San Diego CA), following the manufacturer's instructions. Immediately after removing aliquots for STI testing, residual urine samples were aliquoted and frozen without additives at −80°C.
DNA isolation
Urine aliquots (up to 50 ml) were thawed and 1.5 ml of each specimen was pelleted by centrifugation at 12,000×g for 10 min at 4°C. DNA was harvested from pellets using a Qiagen DNeasy (Qiagen Inc.) tissue extraction kit using the Gram-positive bacteria protocol, according to the manufacturer's protocol. DNA was stored at 4°C until subsequent 16S PCR amplification and sequencing. Mock specimens (molecular grade water) were processed in parallel with patient samples to monitor reagent purity. Individuals were considered STI positive if Aptima testing revealed pathogens and or if one or more sequences 99% identical to a known urethral pathogen was detected by sequence analysis (N. gonorrhoeae, M. hominis, M. genitalium, U. urealyticum.) (Table S3).
16S rRNA allele PCR, cloning and sequencing
Urine gDNAs were PCR amplified using the broad range primers 27F 5-AGAGTTTGATCCTGGCTCAG-3′ and 1391R 5′-GACGGGCGGTGWGTRCA-3′ using Takara KA-HS Taq Polymerase (Takara Bio Inc.) in a total reaction volume of 15 µl. PCR cycling conditions were 95°C for 5 min followed by 30 cycles of 95°C for 30s, 55°C for 30s, 72°C for 1.5 min followed by a final cycle of 72°C for 10 min. All urine gDNAs yielded robust amplicons after 30 PCR cycles, whereas identically processed mock specimens (molecular grade water) failed to yield amplicons after 40 cycles. Products from each PCR were gel extracted, tailed with GO TAQ (Promega Inc.) and then TA-cloned into pCRII-TOPO (Invitrogen Inc.). Insert 16S rRNA alleles were sequenced from the ends and middle of the insert using primers M13F 5′-TGTAAAACGACGGCCAGT-3′, M13R 5′-CAGGAAACAGCTATGACC-3 and 907R 5′-CCGTCAATTCCTTTRAGTTT-3′.
The three reads (average length of 769bp) were assembled into a single consensus sequence. 6,877 unique 16S rDNA contigs were assembled with Phrap [18], [19]. Of these, 942 were removed from the subsequent analysis because more than 5% of their average Phred quality scores were <19.
Bioinformatics and data processing
154 contigs matched human genome sequences, NCBI BLASTn [20] (1) (E-value cutoff 10–20), and were eliminated as likely contaminants. An additional 1,185 contigs were parsed because they failed to pass Bellerophon chimera checking (default parameters) at GreenGenes [21]. The remaining 4,606 contigs (each at least 1,250 bp in length) were considered to represent full length 16S rDNA alleles and were compared to RDP II database [22] sequences using the RDP Classifier program v2.2 [23]. For phylogenetic comparisons of microbiomes, multiple sequence alignments (MLS) of the 16S contig sequences were produced using ClustalW [24] with default parameters, and the MLS were used to construct a neighbor-joining phylogenic tree using PHYLIP [25]. Principal coordinate analyses (PCA) of microbial communities were performed using Unifrac [26], [27]. Hierarchical clustering was performed and heat maps were generated using a Spearman's rank correlation coefficient [28], [29] as a distance measure and a customized script developed in the R statistical package (http://www.r-project.org/). All sequences from this study are available at the Urethral Microbiome Project website (http://www.microbiota.org/cgi-bin/mum/pilot/analysis.cgi).
Statistics
Chi-square and Fisher-exact tests were used to test if bacterial taxa were biased in urine with or without STI. Wilcoxon rank-sum tests were also performed after reads in each specimen were normalized (i.e., ratios were compared). Pair-wise Spearman's rank correlation coefficients were computed to identify taxa, which were either correlated or anti-correlated in urine specimens. All statistical tests were performed using a customized R script. To correct for multiple tests, false discovery rates were computed with the R package function q value.
Results
First-catch urine from adult men contains diverse bacterial taxa
We used urines as a proxy for sampling of the male urethra based upon reports that urine and urethral swabs are similarly suitable specimens for detection of diverse urethral pathogens [30], [31], [32], [33]. All 19 urine specimens yielded robust 16S rRNA PCR amplicons following 30 reaction cycles whereas mock control specimens, processed identically to urine, failed to yield amplicons following 40 cycles. All 16S rRNA contigs larger than 1250 bp and Phred quality scores ≥20 were classified using RDP II Classifier Version 2.2 [22], [23]. Total numbers of sequences from individual urine fitting these criteria ranged from 35–352 (SD) 86).
On average, more than 95% of these sequences could be assigned to the genera level using RDP taxonomy with ≥90% confidence and the proportions of sequences in different urines which met this threshold were similar (range 85.8%–98.8%, mean 95.1%, SD 3.6%). The remaining sequences could be classified with 90% confidence to at least the order level, with the exception of 6 Archaeal sequences. Rarefaction analyses, where terminal taxa were classified based on sequence similarity, indicated that collector's curves for most urine samples had entered the plateau phase (Figure S1). Consistent with this, analysis of all sequences from the three specimens which yielded the fewest high quality sequences (U12, U13 and U18) using classifier and a 60% confidence threshold failed to significantly alter the overall composition of the urine communities by Unifrac or to reveal pathogen sequences in the STI negative specimens (U12 and U16) (Data not shown).
Phylum level composition of urine microbiomes
To identify bacterial taxa, which could represent the “core” urine microbiome 16S rRNA sequences, were classified to the phylum level, according to RDP II taxonomy, and compared. Seven phyla were identified in total. Sequences corresponding to five bacterial phyla including Firmicutes (52.6%), Actinobacteria (18.7%), Fusobacteria (10.0%), Proteobacteria (9.4%) and Bacteroidetes (7.4%) were frequently detected, whereas sequences corresponding to Tenericutes (1.8%) and TM7 (<0.1%) were less abundant. The distribution of these phyla differed markedly among the urine specimens (Figure 1). Firmicutes were present in all specimens and Actinobacteria and Proteobacteria were each present in 18/19. Proportions of total sequences in urine corresponding to these phyla (where present) ranged from 3.5–93.1%, 0.4–90.3% and 1.3–62.2% for Firmicutes, Actinobacteria and Proteobacteria, respectively. Bacteroidetes and Tenericutes were identified in 15/19 and 9/19 urine specimens but never constituted more than 25% of total sequences in any urine. In contrast, Fusobacteria were only detected in 9/19 urine specimens but accounted for a significant proportion of total sequences (0.4–47.8%, mean 18.9%, SD 15.4%) when present. TM7 was detected in 3 urine in which it accounted for 0.3–1.5% of the total sequences. These results indicated that there is substantial intra-individual variation in urine microbiomes even at the phylum level.
4,386-16S rRNA Urine sequences were sorted to the Phyla level using RDP Classifier at 90% confidence cutoff [23]. Subjects are indicated below bars on the X-axis and the percents of clones corresponding to specific phyla are indicated on the Y-axis. STI status of the subjects is indicated below the X-axis.
Many bacteria in urine are similar to residents of the human vagina, colon and skin
To better characterize the urine microbiomes the 16S rRNA sequences were sorted to the genera level. Lactobacillus, Corynebacterium, Streptococcus and Sneathia spp. accounted for approximately 50% of the total urine sequences (Table 1) and the 10 most common genera represented almost 75% of urine sequences, although 72 genera were detected in total (Table S4). This result showed that the overwhelming majority of the urine sequences corresponded to a few abundant genera.
The most similar sequence in the RDP database corresponding to each urine sequence and the original source of the RDP sequence were identified. More than 72% of the urine sequences best-matched only 25 RDP sequences. Most of these RDP sequences were originally from human female urogenital tract specimens [7], [9], [34], [35] and smaller proportions were from the mouth and skin (Table 1) [36]. The most common urine sequences corresponded to Lactobacillus iners. Other abundant taxa included various species of Aerococcus, Anaerococcus, Prevotella, Gemella, Veillonella and Sneathia (Table 1 and Table S5).
Clustering analysis can separate STI and non-STI urine microbiomes
To identify potential associations between bacterial taxa the urine microbiomes were hierarchically clustered and genera were weighted to reflect their abundance. Figure 2 depicts clustering results produced using pair-wise Spearman's rank correlation coefficients among all the genera from urine. The first cluster (Fig. 2, from the left end) contains eight urine specimens, all of which were positive for one or more STI by specific nucleic acid amplification tests and/or by the presence of sequences corresponding to known sexually transmitted pathogens (Table S3). All eight specimens in this cluster contained Prevotella and Sneathia spp., Prevotella was only detected in 2 specimens outside this group, and Sneathia spp. were unique to this group. Other frequent and or abundant genera in this group included Dialister (7/8), Gemella (7/8), and Atopobium (5/8). The second cluster contained two specimens and the majority of sequences in both of these were Lactobacillus iners. One of these was also positive for Mycoplasma spp. There are nine specimens in the third cluster. Although there was substantial diversity comparing these urine, one or more of Corynebacteria, Propionobacteria, Staphylococcus and or Streptococci spp. were dominant. Approximately 5% of the sequences from one urine (Urine 13) corresponded to a reference strain of N. gonorrhoeae (>99% identical), but this was the only STI positive urine in this cluster (Table S3).
Genus names corresponding to terminal taxa depicted in the heat map are listed to the right of the figure. Subjects are listed at the bottom. Intensity of the coloration of cells in the heat map indicates the extent of correlation of different urine samples. Dendograms at the top and right of the heat map indicate relationships of microbiomes and genera, respectively. The top infection color bar indicates STI (labeled as YES) or non-STI (labeled as NONE) participants.
To independently assess associations between STI and specific microbiomes the data were analyzed by un-weighted Unifrac PCA [26], [37]. This analysis provides a quantitative measure of β-diversity which reflects the evolutionary distance between the component taxa in the communities [38]. Unifrac reproduced the groups revealed by hierarchical clustering (Figure 3). Specifically, all eight specimens from cluster 1 grouped together and separately from specimens in clusters 2 and 3.
Clustering of urine microbiomes was performed using un-weighted Unifrac [37]. The percent variations explained by each principle component are indicated on the axis. Blue circles indicate urine from individuals positive for STI; red circles correspond to urine negative for STI.
To identify organisms driving the cluster results, genera that were significantly enriched in urines from men with STI were identified using Fisher's exact, Chi-square, and Wilcoxons tests (Table S6). Many of the abundant genera in cluster 3 were enriched in men with STI. To detect if associations between any urine taxa existed, Spearman's rank correlation coefficients were calculated for all pair-wise combinations of the genera identified in this study (Table S7). The abundant genera in cluster one discussed above were strongly correlated with one another and anti-correlated with abundant genera from the other clusters. For example, Sneathia was strongly positively correlated with Gemella, Dialister, Aerococcus, Prevotella and Veillonella, and negatively correlated with Staphylococcus, Propionibacterium and Streptococcus.
Discussion
We report a detailed cultivation independent comparison of urine microbiomes in men with and without STI. Large-scale Sanger sequencing of near full-length 16S rRNA alleles was utilized so the data set would be a robust resource for future comparisons.
Our data support an expanded view of the association of the urogenital microbial communities in health and disease. First, the bacterial communities in urine are complex and a single, characteristic microbial community was not apparent. Urogenital genera overlap to some degree microbial communities found in the superficial skin [36], colon [39], vagina [7], [9], [34], [35]. A variety of taxa identified on the penile coronal sulcus in a recent study were also observed [40]. However, our data suggest the overall composition of urine and coronal sulcus microbiomes differ. The sulcus is dominated by Pseudomonadaceae spp. [40] whereas sequences corresponding to this family were rare in urine. Second multiple BV-associated anaerobic taxa, such as Sneathia spp., which appear to be minor components of sulcus microbiomes in uncircumcised men were dominant components of urine microbiomes. Finally, Lactobacillus spp. were abundant in some urines but were not reported in sulcus specimens. Caveats of these comparisons include subject populations (African men versus an ethnically diverse population of American men). The power of our study was also too low to assess differences in microbiomes associated with race or ethnicity. Finally this study and the survey in question employed different PCR primers, PCR amplification conditions and 16S rRNA sequencing, all of which limit the utility of direct comparisons of the data sets.
A second key finding is the association of STI with a cluster of organisms not previously linked to the male urogenital tract: Sneathia, Gemella, Aerococcus, Anaerococcus, Prevotella and Veillonella, to name a few [9], [35], [41]. These organisms have previously been identified in the coronal sulcus of uncircumcised men, and became less frequent following circumcision to reduce risk of sexually transmitted human immunodeficiency virus infection [40]. Some of these organisms are also commonly observed in the vaginal microbial communities of asymptomatic women [7], as well as women with bacterial vaginosis [9], [41] and upper genital tract pathology [42], [43]. Similar anaerobic communities in female genital tract are associated with increased risk for STI [44]. These observations suggest the hypothesis that establishment of BV-like communities in the male urethra is similarly associated with STI risk. However, present data are clearly inadequate to differentiate if the STI-associated communities precede, are co-transmitted with or are established subsequent to STI.
An limitation of this study is that urine can sample multiple body compartments and that these specimens were used as a proxy for urethral swabs because the latter technique is contraindicated in STI negative men. Urethral swabs and urine are equally suitable specimens for PCR-based detection of a variety of sexually transmitted pathogens [30], [31], [32], [33]. Comparisons of paired first-catch urine and urethral swabs support that interpretation that the microbiomes sampled by with these specimen types are highly similar (Q Dong et al., manuscript in preparation). Thus we assume here that urine specimens primarily contain bacteria associated with the urethral epithelium, and lower numbers of organisms from the bladder and urethral meatus. Incidence of upper urinary tract infection is low in our study population which suggests bladder flora were infrequent. Thus, although male urine may not exclusively sample urethral flora, we believe it provides as good of a measure of this flora as is practical and ethical in individuals not suspected to have STI.
In conclusion, our data provide the first glimpse into the urogenital microbiome in men with STI. Caution is clearly warranted in interpreting associations between characteristic microbiomes and STI; we cannot exclude other possible explanations for the patterns observed, such as long-term effects of undocumented antibiotic use [45]. Considering these and similar caveats we believe the data suggest the hypothesis that establishment characteristic microbiomes might be related to risk for subsequent STI. Longitudinal studies of urine microbiomes prior to and post STI will be necessary to test this and we believe such studies could elucidate inter-species microbial interactions relevant to a variety of poorly understood urogenital syndromes.
Supporting Information
Table S2.
Demographics and recent sexual history.
https://doi.org/10.1371/journal.pone.0014116.s002
(0.08 MB DOC)
Table S3.
Sequences corresponding to pathogens.
https://doi.org/10.1371/journal.pone.0014116.s003
(0.11 MB DOC)
Table S4.
Complete list of urine genera sorted by RDP classifier.
https://doi.org/10.1371/journal.pone.0014116.s004
(0.03 MB XLS)
Table S5.
Complete list of urine taxa by RDP best match.
https://doi.org/10.1371/journal.pone.0014116.s005
(0.48 MB XLS)
Table S6.
Statistical comparisons of correlated urine taxa.
https://doi.org/10.1371/journal.pone.0014116.s006
(0.03 MB XLS)
Table S7.
Spearman's rank correlation coefficients for all urine taxa.
https://doi.org/10.1371/journal.pone.0014116.s007
(0.61 MB XLS)
Figure S1.
Rarefaction curves of individual urine microbiomes. Rarefaction curves were generated using the MOTHUR package after clustering sequences at the genetic distance cutoffs of 0.03, 0.05, and 0.10.
https://doi.org/10.1371/journal.pone.0014116.s008
(0.48 MB PDF)
Acknowledgments
We would like to thank Xiang Gao, Ian Chen, Shawnwei Chen, Winston Ha and Eugene Pak for their assistance in data analysis, and the production-sequencing group at the Washington University Genome Center. Finally, we would like to thank Evelyn Toh for review of this manuscript.
Author Contributions
Conceived and designed the experiments: DEN BVDP GMW JDF. Performed the experiments: DEN BVDP QD BF SE ES LD. Analyzed the data: DEN BVDP QD KVR BF SE ES GMW LD JDF. Contributed reagents/materials/analysis tools: DEN BVDP BF SE ES GMW. Wrote the paper: DEN BVDP QD KVR BF SE ES GMW LD JDF.
References
- 1. Robinson CJ, Bohannan BJ, Young VB (2010) From structure to function: the ecology of host-associated microbial communities. Microbiol Mol Biol Rev 74: 453–476.
- 2. Frank DN, Pace NR (2001) Molecular-phylogenetic analyses of human gastrointestinal microbiota. Curr Opin Gastroenterol 17: 52–57.
- 3. Turnbaugh PJ, Hamady M, Yatsunenko T, Cantarel BL, Duncan A, et al. (2009) A core gut microbiome in obese and lean twins. Nature 457: 480–484.
- 4. Turnbaugh PJ, Ley RE, Mahowald MA, Magrini V, Mardis ER, et al. (2006) An obesity-associated gut microbiome with increased capacity for energy harvest. Nature 444: 1027–1031.
- 5. Peterson DA, Frank DN, Pace NR, Gordon JI (2008) Metagenomic approaches for defining the pathogenesis of inflammatory bowel diseases. Cell Host Microbe 3: 417–427.
- 6. Frank DN, St Amand AL, Feldman RA, Boedeker EC, Harpaz N, et al. (2007) Molecular-phylogenetic characterization of microbial community imbalances in human inflammatory bowel diseases. Proc Natl Acad Sci U S A 104: 13780–13785.
- 7. Ravel J, Gajer P, Abdo Z, Schneider GM, Koenig SS, et al. (2010) Microbes and Health Sackler Colloquium: Vaginal microbiome of reproductive-age women. Proc Natl Acad Sci U S A.
- 8. Forney LJ, Gajer P, Williams CJ, Schneider GM, Koenig SS, et al. (2010) Comparison of self-collected and physician-collected vaginal swabs for microbiome analysis. J Clin Microbiol 48: 1741–1748.
- 9. Fredricks DN, Fiedler TL, Marrazzo JM (2005) Molecular identification of bacteria associated with bacterial vaginosis. N Engl J Med 353: 1899–1911.
- 10. Schwebke JR, Desmond R (2005) Risk factors for bacterial vaginosis in women at high risk for sexually transmitted diseases. Sex Transm Dis 32: 654–658.
- 11. Bowie WR, Pollock HM, Forsyth PS, Floyd JF, Alexander ER, et al. (1977) Bacteriology of the urethra in normal men and men with nongonococcal urethritis. J Clin Microbiol 6: 482–488.
- 12. Bowie WR, Wang SP, Alexander ER, Floyd J, Forsyth PS, et al. (1977) Etiology of nongonococcal urethritis. Evidence for Chlamydia trachomatis and Ureaplasma urealyticum. J Clin Invest 59: 735–742.
- 13. Bradshaw CS, Tabrizi SN, Read TR, Garland SM, Hopkins CA, et al. (2006) Etiologies of nongonococcal urethritis: bacteria, viruses, and the association with orogenital exposure. J Infect Dis 193: 336–345.
- 14. Ivanov YB (2007) Microbiological features of persistent nonspecific urethritis in men. J Microbiol Immunol Infect 40: 157–161.
- 15. Riemersma WA, van der Schee CJ, van der Meijden WI, Verbrugh HA, van Belkum A (2003) Microbial population diversity in the urethras of healthy males and males suffering from nonchlamydial, nongonococcal urethritis. J Clin Microbiol 41: 1977–1986.
- 16. Willen M, Holst E, Myhre EB, Olsson AM (1996) The bacterial flora of the genitourinary tract in healthy fertile men. Scand J Urol Nephrol 30: 387–393.
- 17. Mazuecos J, Aznar J, Rodriguez-Pichardo A, Marmesat F, Borobio MV, et al. (1998) Anaerobic bacteria in men with urethritis. J Eur Acad Dermatol Venereol 10: 237–242.
- 18. Ewing B, Hillier L, Wendl MC, Green P (1998) Base-calling of automated sequencer traces using phred. I. Accuracy assessment. Genome Res 8: 175–185.
- 19. Ewing B, Green P (1998) Base-calling of automated sequencer traces using phred. II. Error probabilities. Genome Res 8: 186–194.
- 20. Altschul SF, Madden TL, Schaffer AA, Zhang J, Zhang Z, et al. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25: 3389–3402.
- 21. DeSantis TZ, Hugenholtz P, Larsen N, Rojas M, Brodie EL, et al. (2006) Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB. Appl Environ Microbiol 72: 5069–5072.
- 22. Cole JR, Chai B, Farris RJ, Wang Q, Kulam-Syed-Mohideen AS, et al. (2007) The ribosomal database project (RDP-II): introducing myRDP space and quality controlled public data. Nucleic Acids Res 35: D169–172.
- 23. Wang Q, Garrity GM, Tiedje JM, Cole JR (2007) Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy. Appl Environ Microbiol 73: 5261–5267.
- 24. Chenna R, Sugawara H, Koike T, Lopez R, Gibson TJ, et al. (2003) Multiple sequence alignment with the Clustal series of programs. Nucleic Acids Res 31: 3497–3500.
- 25. Felsenstein J (1989) PHYLIP- Phylogeny ingerence package. Cladistics 164–166.
- 26. Hamady M, Lozupone C, Knight R (2010) Fast UniFrac: facilitating high-throughput phylogenetic analyses of microbial communities including analysis of pyrosequencing and PhyloChip data. ISME J 4: 17–27.
- 27. Lozupone C, Hamady M, Knight R (2006) UniFrac–an online tool for comparing microbial community diversity in a phylogenetic context. BMC Bioinformatics 7: 371.
- 28.
Freund JE, editor. (1992) Mathematical statistics 5th ed. Prentice Hall. pp. 494–546.
- 29.
Goldman RNWJS, editor. (1985) Statistics: an introduction. Prentice Hall. pp. 72–98.
- 30. Bianchi A, Scieux C, Brunat N, Vexiau D, Kermanach M, et al. (1994) An evaluation of the polymerase chain reaction amplicor Chlamydia trachomatis in male urine and female urogenital specimens. Sex Transm Dis 21: 196–200.
- 31. Chernesky MA, Jang D, Lee H, Burczak JD, Hu H, et al. (1994) Diagnosis of Chlamydia trachomatis infections in men and women by testing first-void urine by ligase chain reaction. J Clin Microbiol 32: 2682–2685.
- 32. Chernesky MA, Lee H, Schachter J, Burczak JD, Stamm WE, et al. (1994) Diagnosis of Chlamydia trachomatis urethral infection in symptomatic and asymptomatic men by testing first-void urine in a ligase chain reaction assay. J Infect Dis 170: 1308–1311.
- 33. Crotchfelt KA, Welsh LE, DeBonville D, Rosenstraus M, Quinn TC (1997) Detection of Neisseria gonorrhoeae and Chlamydia trachomatis in genitourinary specimens from men and women by a coamplification PCR assay. J Clin Microbiol 35: 1536–1540.
- 34. Hyman RW, Fukushima M, Diamond L, Kumm J, Giudice LC, et al. (2005) Microbes on the human vaginal epithelium. Proc Natl Acad Sci U S A 102: 7952–7957.
- 35. Zhou X, Bent SJ, Schneider MG, Davis CC, Islam MR, et al. (2004) Characterization of vaginal microbial communities in adult healthy women using cultivation-independent methods. Microbiology 150: 2565–2573.
- 36. Gao Z, Tseng CH, Pei Z, Blaser MJ (2007) Molecular analysis of human forearm superficial skin bacterial biota. Proc Natl Acad Sci U S A 104: 2927–2932.
- 37. Lozupone C, Knight R (2005) UniFrac: a new phylogenetic method for comparing microbial communities. Appl Environ Microbiol 71: 8228–8235.
- 38. Lozupone CA, Hamady M, Kelley ST, Knight R (2007) Quantitative and qualitative beta diversity measures lead to different insights into factors that structure microbial communities. Appl Environ Microbiol 73: 1576–1585.
- 39. Eckburg PB, Bik EM, Bernstein CN, Purdom E, Dethlefsen L, et al. (2005) Diversity of the human intestinal microbial flora. Science 308: 1635–1638.
- 40. Price LB, Liu CM, Johnson KE, Aziz M, Lau MK, et al. (2010) The effects of circumcision on the penis microbiome. PLoS One 5: e8422.
- 41. Marrazzo JM, Thomas KK, Fiedler TL, Ringwood K, Fredricks DN (2008) Relationship of specific vaginal bacteria and bacterial vaginosis treatment failure in women who have sex with women. Ann Intern Med 149: 20–28.
- 42. DiGiulio DB, Romero R, Kusanovic JP, Gomez R, Kim CJ, et al. (2010) Prevalence and diversity of microbes in the amniotic fluid, the fetal inflammatory response, and pregnancy outcome in women with preterm pre-labor rupture of membranes. Am J Reprod Immunol 64: 38–57.
- 43. DiGiulio DB, Romero R, Amogan HP, Kusanovic JP, Bik EM, et al. (2008) Microbial prevalence, diversity and abundance in amniotic fluid during preterm labor: a molecular and culture-based investigation. PLoS One 3: e3056.
- 44. Ness RB, Kip KE, Soper DE, Hillier S, Stamm CA, et al. (2005) Bacterial vaginosis (BV) and the risk of incident gonococcal or chlamydial genital infection in a predominantly black population. Sex Transm Dis 32: 413–417.
- 45. Dethlefsen L, Huse S, Sogin ML, Relman DA (2008) The pervasive effects of an antibiotic on the human gut microbiota, as revealed by deep 16S rRNA sequencing. PLoS Biol 6: e280.