Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
Fouts et al. Journal of Translational Medicine 2012, 10:174http://www.translational-medicine.com/content/10/1/174
RESEARCH Open Access
Integrated next-generation sequencing of 16SrDNA and metaproteomics differentiate thehealthy urine microbiome from asymptomaticbacteriuria in neuropathic bladder associatedwith spinal cord injuryDerrick E Fouts1*, Rembert Pieper1, Sebastian Szpakowski1, Hans Pohl2, Susan Knoblach2, Moo-Jin Suh1,Shih-Ting Huang1, Inger Ljungberg3, Bruce M Sprague2, Sarah K Lucas1, Manolito Torralba1, Karen E Nelson1
and Suzanne L Groah3,4
Abstract
Background: Clinical dogma is that healthy urine is sterile and the presence of bacteria with an inflammatoryresponse is indicative of urinary tract infection (UTI). Asymptomatic bacteriuria (ABU) represents the state in whichbacteria are present but the inflammatory response is negligible. Differentiating ABU from UTI is diagnosticallychallenging, but critical because overtreatment of ABU can perpetuate antimicrobial resistance whileundertreatment of UTI can result in increased morbidity and mortality. In this study, we describe key characteristicsof the healthy and ABU urine microbiomes utilizing 16S rRNA gene (16S rDNA) sequencing and metaproteomics,with the future goal of utilizing this information to personalize the treatment of UTI based on key individualcharacteristics.
Methods: A cross-sectional study of 26 healthy controls and 27 healthy subjects at risk for ABU due to spinal cordinjury-related neuropathic bladder (NB) was conducted. Of the 27 subjects with NB, 8 voided normally, 8 utilizedintermittent catheterization, and 11 utilized indwelling Foley urethral catheterization for bladder drainage. Urine wasobtained by clean catch in voiders, or directly from the catheter in subjects utilizing catheters. Urinalysis, urineculture and 16S rDNA sequencing were performed on all samples, with metaproteomic analysis performed on asubsample.
Results: A total of 589454 quality-filtered 16S rDNA sequence reads were processed through a NextGen 16S rDNAanalysis pipeline. Urine microbiomes differ by normal bladder function vs. NB, gender, type of bladder catheterutilized, and duration of NB. The top ten bacterial taxa showing the most relative abundance and change amongsamples were Lactobacillales, Enterobacteriales, Actinomycetales, Bacillales, Clostridiales, Bacteroidales,Burkholderiales, Pseudomonadales, Bifidobacteriales and Coriobacteriales. Metaproteomics confirmed the 16S rDNAresults, and functional human protein-pathogen interactions were noted in subjects where host defenses wereinitiated.(Continued on next page)
* Correspondence: [email protected] Craig Venter Institute, 9704 Medical Center Drive, Rockville, MD 20850,USAFull list of author information is available at the end of the article
© 2012 Fouts et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the CreativeCommons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, andreproduction in any medium, provided the original work is properly cited.
Fouts et al. Journal of Translational Medicine 2012, 10:174 Page 2 of 17http://www.translational-medicine.com/content/10/1/174
(Continued from previous page)
Conclusions: Counter to clinical belief, healthy urine is not sterile. The healthy urine microbiome is characterized bya preponderance of Lactobacillales in women and Corynebacterium in men. The presence and duration of NB andmethod of urinary catheterization alter the healthy urine microbiome. An integrated approach of 16S rDNAsequencing with metaproteomics improves our understanding of healthy urine and facilitates a more personalizedapproach to prevention and treatment of infection.
Keywords: Bacteriuria, Urine, Catheter, Neuropathic, Bladder, Microbiome, Metaproteome, Next-generation,Personalized, rRNA
BackgroundAffecting nearly one half of all Americans over thecourse of a lifetime [1] and with costs exceeding $1 bil-lion annually [2], urinary tract infection (UTI) is a majorpublic health problem [3,4]. It is the most common uro-logic disorder in the outpatient setting [3,4] and themost common health care associated infection [3,5,6]. Inthe health care setting alone, approximately 561777UTIs occur annually, costing an estimated $1006 per in-fection, totaling more than $500 million, and being re-sponsible for 8205 deaths [5]. This does not include thepersonal suffering or time lost from gainful employment.Often considered an antecedent to UTI, asymptomatic
bacteriuria (ABU) represents an asymptomatic carrierstate recognized to have little impact on health or qual-ity of life. This is in contrast to healthy urine, consideredto be sterile until reaching the urethra, which is colo-nized by facultative anaerobic Gram-negative rods andcocci. The most common risk factor for the develop-ment of ABU and UTI is use of a urinary catheter [5-7],as catheters provide a conduit for bacterial colonizationand symptomatic infection. Resolution of ABU typicallyoccurs with removal of the urinary catheter. This is notpossible in many cases, however, as the urinary catheterfacilitates function and emptying in cases of bladderimpairment.The clinical distinction between symptomatic UTI and
ABU is not trivial since symptomatic UTI requires treat-ment and perhaps further evaluation irrespective of thecircumstances in which the UTI occurred, while ABUdoes not. Exceptions to this rule include the treatmentof ABU in select at-risk populations, such as pregnantwomen, which has been shown to be associated withimproved outcomes [8]. Distinction between these statesis particularly relevant, as renewed urgency and heigh-tened focus have been placed on UTI occurrence by na-tional policy-makers and payers. The Centers forMedicare & Medicaid Services (CMS) has identifiedcatheter-associated UTI, the most common hospitalacquired infection, as a “never event”. Effective in 2008,this has resulted in non-reimbursement for catheter-associated UTIs that were not present on admission toacute care hospitals [9,10]. An unintended consequence
of designating catheter-associated UTI as a “never event”is more aggressive screening for ABU and UTI upon ad-mission of patients to hospitals, a strategy that may leadto increased unnecessary antibiotic treatment and emer-gence of antimicrobial resistance.To achieve improved outcomes in the care of patients
with ABU and UTI, improved information distinguishingstates of urine in health and disease is needed. To thisend, we sought to first describe states of urinary healthutilizing a highly sensitive, culture independent approachto determine whether the urine microbiome of healthypeople who are at risk for ABU because they utilizeurinary catheters differs from that of healthy controls,and if so, to identify key factors or bacterial signaturesthat might ultimately lead to UTI requiring antimicro-bial treatment. Subjects with neuropathic bladder due tospinal cord injury who are known to be at highest riskfor ABU and UTI due to their need for catheter-assistedbladder management [4,11,12] were assessed and com-pared with healthy controls to achieve our goal.
MethodsThe study was approved by the MedStar InstitutionalReview Board (IRB). All study personnel were certifiedin and the study protocol conformed to the ethicalguidelines of the 1975 Declaration of Helsinki asreflected in approval by the MedStar IRB.
Sample acquisition and clinical urinalysisPatients and healthy controls were recruited into thisIRB-approved study (NRH IRB# 2011–019) from the out-patient clinic and inpatient ward at National Rehabilita-tion Hospital (Washington, DC). Following writtenconsent, urine samples were obtained from 26 healthy,non-SCI controls and 27 with neuropathic bladder (NB)due to spinal cord injury (SCI). Patients provided urinesamples by sterile collection using the means by whichthey customarily empty their bladder (i.e. midstreamcollection during voiding, or sterile catheterization if un-able to void). (Table 1) The samples were coded with ananonymous research identification number and separatedinto two aliquots: one, for standard urine analysis andculture (Quest Diagnostics) and another for microbiome
Table 1 Patient demographics
Group Gender Age Race/Ethnicity Months withNB
Urinanalysis Urine culture with >50000 cfu
Leukocyteesterase
WBC(no./hpf)
Healthycontrols
Female(57.7%)
Mean35.6
S01 Female 40 Asian n/a NEG 0 NEG
S02 Male 24 Asian n/a NEG 0 NEG
S03 Female 32 Caucasian n/a NEG 0 NEG
S04 Female 35 Caucasian n/a NEG 0 Streptococcus (beta-hemolytic)
S05 Male 32 Caucasian n/a NEG 0-1 NEG
S06 Female 57 Caucasian n/a NEG 0-1 NEG
S07 Male 35 Caucasian n/a NEG 0-1 NEG
S08 Female 43 African American n/a NEG 0-1 NEG
S09 Female 25 Caucasian n/a NEG 0-1 NEG
S10 Female 34 Caucasian n/a NEG 0-1 NEG
S11 Male 33 Caucasian n/a NEG 0-1 NEG
S12 Male 29 Asian n/a NEG 0-1 NEG
S13 Male 35 Caucasian n/a NEG 0-1 NEG
S14 Female 22 Caucasian n/a NEG 0-1 Staphylococcus
S15 Female 34 Asian n/a NEG 0-1 Lactobacillus
S16 Female 45 Caucasian n/a NEG 0-1 Lactobacillus
S17 Female 46 African American n/a TRA 0-1 Escherichia coli
S18 Female 51 Asian n/a NEG 1-2 Escherichia coli,Staphylococcus aureus
S19 Female 40 Caucasian n/a NEG 1-2 Lactobacillus
S20 Male 50 Caucasian n/a NEG 1-2 NEG
S21 Male 25 Caucasian n/a NEG 5-9 NEG
S22 Male 29 Caucasian n/a 1+ 1-2 NEG
S23 Female 30 Caucasian n/a 1+ 3-4 NEG
S24 Male 39 Caucasian n/a 2+ 5-9 NEG
S25 Female 27 Asian n/a 2+ 100+ Escherichia coli
S26 Male 33 Asian n/a 3+ 30-49 NEG
NB - void Female(37.5%)
Mean37.3
Mean 41.5
S27 Male 20 African American 9 ND ND NEG
S28 Male 31 African American 157 ND ND NEG
S29 Male 19 Caucasian 3 NEG 1-2 Klebsiella pneumoniae
S30 Female 41 African American 1 NEG 1-2 Enterococcus faecalis
S31 Female 54 Caucasian 1 TRA 1-2 NEG
S32 Male 31 African American 158 TRA 3-4 Enterococcus faecalis
S33 Male 48 Hispanic 1 2+ 5-9 Escherichia coli,Enterococcus faecalis
S34 Female 54 Caucasian 2 2+ 5-9 Klebsiella oxytoca
NB-IC Female(50.0%)
Mean44.1
Mean 140.1
S35 Male 40 African American 84 NEG 0 Escherichia coli
S36 Female 36 Caucasian 7 NEG 0-1 Escherichia coli
S37 Female 55 Caucasian 3 NEG 1-2 NEG
S38 Female 55 Caucasian 442 NEG 1-2 NEG
Fouts et al. Journal of Translational Medicine 2012, 10:174 Page 3 of 17http://www.translational-medicine.com/content/10/1/174
Table 1 Patient demographics (Continued)
S39 Male 48 Native American 261 NEG 3-4 Klebsiella pneumoniae
S40 Male 48 Native American 260 1+ 3-4 NEG
S41 Female 50 Caucasian 2 2+ 3-4 NEG
S42 Male 21 African American 62 2+ 5-9 Proteus
NB-FC Female(54.5%)
Mean37.6
Mean 136
S43 Female 47 African American 79 ND ND NEG
S44 Female 47 African American 80 NEG 0 Enterococcus faecalis,Gram Negative Rods
S45 Male 23 African American 20 NEG 0-1 Escherichia coli (ESBL),Klebsiella pneumoniae,Providencia stuartii,Pseudomonas aeruginosa,Enterococcus faecalis
S46 Female 40 African American 236 NEG 1-2 Escherichia coli, Citrobacterkoseri (diversus),Enterococcus faecalis
S47 Female 40 African American 235 NEG 3-4 NEG
S48 Female 61 Caucasian 469 TRA 1-2 Escherichia coli
S49 Male 27 African American 92 1+ 3-4 Pseudomonas aeruginosa
S50 Female 40 African American 235 1+ 15-19 NEG
S51 Male 48 African American 18 2+ 5-9 Escherichia coli,Escherichia coli (ESBL),Pseudamonas aeruginosa,Enterococcus faecalis
S52 Male 20 African American 25 2+ 10-14 NEG
S53 Male 21 African American 7 2+ 50+ NEG
No.= number, WBC = white blood cells, hpf = high power field, HC = healthy control, IC = intermittent catheter, FC = Foley catheter, NEG = negative, TRA = trace,n/a = not applicable, ND = not determined, ESBL = Extended-Spectrum-Beta-Lactamases.
Fouts et al. Journal of Translational Medicine 2012, 10:174 Page 4 of 17http://www.translational-medicine.com/content/10/1/174
analysis. Quest Diagnostics performed analysis of urinesamples for nitrite formation, leukocyte esterase, andmicroscopic examination for the presence and quantityof leukocytes and erythrocytes in each sample. Bacterialcultures were performed by inoculation of blood agarplates and incubation at 37°C for 48 hours.
Sample preparation and PCRThawed urine samples were clarified by low-speed centri-fugation and bacterial genomic DNA was extracted fromurine pellets by enzymatic digestion using a final concen-tration of 20 mg/ml Lysozyme (Invitrogen) followed byphysical lysis using Lysing Matrix B tubes (QBiogene). Aprevious study comparing mechanical and enzymaticmethods for extracting microbial genomic DNA showedthat mechanical cell disruption by bead beating producedthe highest bacterial diversity [13]. The samples were vor-texed at maximum speed for 45 seconds using a Fastprepfp120 (MP Biomedicals) then cooled on ice. DNA wasextracted from the lysate using phenol chloroform isoamylalcohol extraction and ethanol precipitation. 16S rDNAsequences were generated by amplifying the V1-V3 regionof the bacterial 16S rRNA gene using primers 27Fand 534R fused with 454 adaptors and barcodes for
multiplexing. Primers targeting V2 and V3 were shown toperform as well as full-length 16S rDNA sequence forcommunity clustering and taxonomic assignments [14].The amplicons were normalized and pooled prior to
emulsion PCR and 454 sequencing (Roche, Inc.) usingtitanium chemistry.
DNA sequence processingThe 16S rDNA sequence-processing pipeline used forthis study is composed of a selection of bioinformaticstools proven to be accurate, robust and fast.. A supple-mentary archive contains the dot language representationof a graph depicting the entire workflow executed, out-lining the specific parameters used for each command.Initially, the SFF file, output from the sequencer, was
converted into fasta and qual files using the sffinfo pro-gram included as a part of 454/Roche software package.Subsequently, the trim.seqs function in mothur [15](version v.1.22.2) was used to de-multiplex sequencerreads. No barcode mismatches, and up to one primer mis-match were allowed past this step. The de-multiplexedreads were processed using LUCY [16-18] to filter outreads with low quality segments. At this point, the sub.sample function of mothur was used to select an equal
Fouts et al. Journal of Translational Medicine 2012, 10:174 Page 5 of 17http://www.translational-medicine.com/content/10/1/174
number of reads per biological sample (n=3671 based onthe biological sample with the fewest number of reads).Subsequent, the screen.seqs function of mothur was usedto remove sequences shorter than 220 bases [17]. Further-more CD-HIT-454 [19,20] was used to collapse duplicatereads, while retaining their count for subsequent enrich-ment statistics, (analogous to the functionality of theunique.seqs function of mothur, but orders of magnitudefaster and less demanding on the computer hardware).The sequences were aligned against the SILVA database of16S rDNA sequences [15,21] to verify 1) the orientation ofnoise-filtered sequences; and 2) the correct positioning ofthe reads with respect to the expectation of which variableregions should have been amplified and sequenced. There-after, the remaining sequences were subjected to mothur’simplementation of chimera slayer [15,22] to filter outchimeric reads. The processed 16S rDNA data from thisstudy can be obtained at NCBI under BioProject ID 97505.
Taxonomical classification of OTU representative readsTaxonomical classification of the final set of 82160 oper-ational taxonomic unit (OTU)-representative readsdown to the genus level was performed using mothur’sversion of the RDP Bayesian classifier using a normal-ized RDP training dataset [23]. The final step of thepipeline clustered the sequences based on their similarityto produce OTUs. Customarily, a similarity threshold of97% has been used to define OTUs at approximately thespecies level [24]. A module of CD-HIT suite [19] calledCD-HIT-EST was employed to perform species-levelread-clustering for subsequent analyses.
Statistical analysesThe orchestration and automation of steps has beenachieved using a custom set of in-house utilities written inpython and R [25] programming languages. These utilitiesare available online as a part of the YAP package on github[26]. JCVI grid infrastructure based on the Oracle GridEngine (OGE) was used for all steps described. Relativeabundance and diversity statistics were calculated withinmothur [15]. Further statistical analyses were accom-plished using R. Heat maps were generated using theheatmap.2 function of the gplots package available onCRAN [25]. OTU counts have been normalized to 100%per individual, to facilitate comparability. Only taxa (orderor genus) with a standard deviation greater than 5% acrossall 52 individuals were used to generate the heat maps.Differences between subject OTU communities wereassessed using the Bray-Curtis beta diversity statisticimplemented in the vegdist function, a part of vegan Rpackage available on CRAN. Clustering was accomplishedusing average-neighbor-joining method implemented inhclust function in the default installation of R. PCA ana-lysis was performed using ade4 [27]. P-values used to
determine statistical significance in relative OTU differ-ence plots were established using the default installationof R and kruskal.test functions implementing the Kruskal-Wallis rank-sum statistic test [28].
Phylogenetic tree buildingOTU-representative sequences classified as either Lacto-bacillales or Enterobacteriales were aligned using toolsavailable from Release 10 of the RDP web site [29]. Specif-ically, sequences from OTUs composed of reads frommore than one individual were aligned to the RDP refer-ence 16S rRNA sequence, taking into account secondarystructure. At most one nearest neighbor sequence fromRDP was recruited into the alignment per input sequence.The alignment was downloaded and trimmed to removecolumns whose gap fractions were greater than 50%, usingBelvu [30]. Based on the alignment, a bootstrappedNeighbor-Joining (NJ) tree was subsequently inferredusing paupFasta, an in-house wrapper script around thePAUP* program as described [31], and edited usingFig Tree [32]. In combination with nearest neighbortaxonomy, BLASTN was used against the NCBI referenceRNA database, which lacks uncultured organisms, to iden-tify certain Lactobacillales OTU-representative branchesthat lacked classified RDP top matches, to the specieslevel, using a cut-off of 97% identity.
ProteomicsA urinary pellet specimen equivalent to 5–10 ml voidedurine washed two times with ~10 ml ice-cold PBS wasre-suspended in 1 ml of 10 mM ammonium bicarbonatecontaining 0.1% Triton-X100, 0.5% octylglucoside, 5 μg/mlleupeptin, 10 mM EDTA and 2 mM benzamidine. Thesuspension was heated to 85°C for 5 min followed bysonication (amplitude 4, Misonex 3000 sonicator) in 30son/15s off cycles 10 times on ice. The suspension wascentrifuged for 15 min at 16,100 x g and the supernatantrecovered. Following protein quantity estimates based onSDS-PAGE analysis, an aliquot with ~10 μg protein wasdigested at a 1:50 ratio (trypsin/protein) using Filter-Aided Sample Preparation (FASP) [33]. The digestionmixture was reconstituted in 50 μl 0.1% formic acid. Forshotgun proteomic analysis, peptides in a 20 μl samplealiquot were separated on a capillary C18 LC column in122 min binary gradient runs from 97% solvent A (0.1%formic acid) to 80% solvent B (0.1% formic acid, 90%acetonitrile) at a flow rate of 350 nl/min. Nano-electrospray into the source was followed by mass analysisand spectral acquisition in automated MS/MS mode, withthe top five parent ions selected for fragmentation in scansof the m/z range 350–2,000 and a dynamic exclusionsetting of 90 sec. The LC-MS/MS workflow using theLTQ-XL ion trap system (Thermo Fisher Scientific) waspreviously described in more detail [34]. The instrument
Fouts et al. Journal of Translational Medicine 2012, 10:174 Page 6 of 17http://www.translational-medicine.com/content/10/1/174
was calibrated prior to performance of LC-MS/MS experi-ments with 200 nmol human [Glu1]-fibrinopeptide B (M.W. 1570.57), verifying that peaks representing ion countshad widths at half-height of <0.25 min, signal/noiseratios >200 and peak heights >107. The LTQ search para-meters (+1 to +3 ions) included mass error tolerances of± 1.4 Da for peptide precursor ions and ± 0.5 Da for pep-tide fragment ions, selecting monoisotopic m/z values.The search engine used to select these parameters andidentify peptides and proteins was Mascot v.2.3 (MatrixScience). The protein sequence database was comprised of19 bacterial genomes (details in Additional file 1) andUniref90’s human database subset [35]. We limited thisdatabase to the human Uniref90 subset and 19 bacterialgenomes frequently associated with ABU and UTI, be-cause significant computational challenges for peptide-spectral match (PSM) assignments are encountered whenvery large protein databases are used [36]. As describedin Additional file 1, we used stringent criteria for PSMs(q-value <=0.01; PEP-value <=10-4) using the MascotPercolator algorithm that improves discrimination be-tween correct and incorrect PSMs, particularly when thesearched sequence space in the database is large [37].
ResultsUrine samples were obtained from 26 healthy, non-SCIcontrols and 27 individuals with NB due to SCI at theNational Rehabilitation Hospital in Washington, DC.Among those with SCI and NB, 8 voided spontaneouslywithout a catheter, 8 emptied by clean intermittent cathe-terization (IC), and 11 used indwelling urethral Foley cathe-terization (FC) (Table 1). Frozen, resuspended pellets fromthe urine samples were analyzed with culture-independentsurveys of bacterial 16S rDNA and urinary proteins.A total of 589454 quality-filtered 454 16S rDNA se-
quence reads were processed through a NextGen 16SrDNA analysis pipeline [26] (Additional file 2) wheretaxonomy of species-level operational taxonomic units(OTUs) was determined using the RDP classifier [38].Analysis of the bacterial community in these urine sam-ples revealed between 5 and 236 species-level (97% iden-tity) OTUs per individual (Additional file 3). Per sampletaxonomic profiles were generated, showing considerablesample-to-sample variation (Additional file 4). To bettervisualize sample-to-sample taxonomic profiles, a heatmap was generated, clustering the distribution of OTUtaxonomy at the level of bacterial order using the Bray-Curtis index (Figure 1A). The top ten bacterial taxashowing the most relative abundance and change amongsamples were Lactobacillales, Enterobacteriales, Actino-mycetales, Bacillales, Clostridiales, Bacteroidales, Bur-kholderiales, Pseudomonadales, Bifidobacteriales andCoriobacteriales (Figure 1A). The Lactobacillales and
Enterobacteriales were the two most relative abundantand changing taxonomic groups.
Urinary microbiome differed by gender and bladderfunctionTo determine whether distinct microbial signatures wereassociated with gender (male versus female) and/or blad-der function (healthy control versus NB), samples weregrouped by these variables. Taxonomic counts were thennormalized by total number of OTUs per sample, andvisualized by a heatmap, clustering the distribution ofOTUs at the level of bacterial genus using the Bray-Curtis index (Figure 1B). The samples were colored byrelative abundance (red/warm most abundant to blue/cool least abundant). Six main taxonomic profile clustersemerged with distinct patterns when grouped by genderand bladder function. Cluster 1 was composed of almostan equal proportion of healthy and NB samples, clusters2 and 3 were dominated by healthy controls, whereasclusters 4–6 were entirely composed of patients with NB(Figure 1B). The 2 “healthy” clusters, (2 and 3) were dis-tinguished by gender, with females in cluster 2 and malesin cluster 3, and by bacterial genus, Lactobacillus group-ing with females in cluster 2 and completely absent inthe male-dominated cluster 3. Cluster 3 was composedof different Gram-positive organisms. Cluster 1 had themost diverse bacterial genus profile, composed largely ofLactobacillus, but not as abundant as cluster 2 with ele-ments of cluster 3 and a few potentially pathogenic gen-era (Enterococcus, Salmonella, and Peptoniphilus).Closer inspection of clusters 4–6 showed a very differentpattern of bacteria, with known UTI pathogens dominat-ing the profiles (Figure 1B). Cluster 4 was primarilycomposed of Enterococcus, Escherichia and Salmonella.Cluster 5 was dominated by Klebsiella sp. and was theonly cluster comprised of all males. Cluster 6 had themost Enterococcus counts of any cluster and also con-tained Aerococcus and Proteus sp.By taking the difference in normalized relative abundance
between controls and SCI groups, the most relative abun-dant bacterial taxa per group as confirmed (Figure 1C).Statistical significance was established using the Kruskal-Wallis test. Significant differences (P < 0.05) betweenthe top relative OTU counts suggest that Lactobacillus,Corynebacterium, Gardnerella, Prevotella and Enterococcusdefine gender differences (“+” in Figure 1C). Lactobacillus,Klebsiella, Corynebacterium, Staphylococcus, Streptococcus,Gardnerella, Prevotella, Escherichia and Enterococcusdefined statistically significant differences between healthybladder controls and NB (“*” in Figure 1C).
Urinary microbiome differed with duration of NBTo further investigate the dynamics of the urinary micro-biome in NB, the NB group was divided into bins based
Healthy − NB
5
0 %
4
0 %
3
0 %
2
0 %
10
%
0 %
10
%
20
%
30
%
40
%
50
%
[remaining 184 ]
Atopobium
Veillonella
Enterococcus
Proteus
Salmonella
Brevibacterium
Escherichia
Prevotella
Aerococcus
Gardnerella
Streptococcus
Staphylococcus
Corynebacterium
Klebsiella
Lactobacillus
Ge
nd
er
Female
Male
P < 0.05
HealthyNB
n=13 n=15
n=14 n=10
+ *Bladder function
*
*
*
**
***
**
*
+
+
+
+
+
+
predominantly in healthypredominantly in NB
A
0.0
0.2
0.4
0.6
0.8
1.0
Normalized# OTUs
C
*
Nei
sser
iale
sU
ncla
ssifi
ed T
M7
Unc
lass
ified
Bac
teria
Syn
ergi
stal
esE
rysi
pelo
tric
hale
sR
hodo
spiri
llale
sC
yano
bact
eria
_ord
er_i
ncer
tae_
sedi
sP
aste
urel
lale
sC
ampy
loba
cter
ales
Fus
obac
teria
les
Rho
doba
cter
ales
Met
hylo
phila
les
Sol
irubr
obac
tera
les
Cau
loba
cter
ales
Unc
lass
ified
Aci
doba
cter
ia_G
p4R
hizo
bial
esA
erom
onad
ales
Sph
ingo
mon
adal
esF
lavo
bact
eria
les
Alte
rom
onad
ales
Oce
anos
piril
lale
sR
hodo
cycl
ales
Xan
thom
onad
ales
Chr
omat
iale
sV
ibrio
nale
sM
ycop
lasm
atal
esU
ncla
ssifi
ed F
irm
icut
esC
ori
ob
acte
rial
esB
ifid
ob
acte
rial
esP
seu
do
mo
nad
ales
Bu
rkh
old
eria
les
Bac
tero
idal
esC
lost
rid
iale
sB
acill
ales
Act
ino
myc
etal
esE
nte
rob
acte
rial
esL
acto
bac
illal
es
Healthy Control
Bre
vib
acte
riu
mP
rote
us
Kle
bsi
ella
Aer
oco
ccu
sE
nte
roco
ccu
sS
alm
on
ella
Esc
her
ich
iaL
acto
bac
illu
sP
epto
nip
hilu
s
Str
epto
cocc
us
Sta
ph
yoco
ccu
sC
ory
neb
acte
riu
m
B
NBFemaleMale
0.0
0.2
0.4
0.6
0.8
1.0
Normalized# OTUs
1
2
3
4
5
6
Healthy ControlNB
Figure 1 Differences in relative bacterial OTU counts between neuropathic and healthy bladder in males and females. For everyindividual, the OTU counts were normalized to the individual's total OTU count. A heat map of the clustered distribution of OTU taxonomy at thelevel of bacterial order (A) and genus (B) was constructed using the Bray-Curtis index. For panels A and B, only the taxa with a standard deviation> 5% across all individuals are shown. Differences in the average OTU count between females and males are plotted in light and dark gray,respectively (C). In panel (C), the top 15 (8%) most abundant bacterial OTUs are represented. The X-axis indicates the difference in relative OTUcounts per bacterial genus indicated on the Y-axis. Statistical significance was established using Kruskal-Wallis test. Significant differences(P < 0.05) between the relative OTU counts are indicated by an asterisk (*) for bladder function, and a plus sign (+) for gender.
Fouts et al. Journal of Translational Medicine 2012, 10:174 Page 7 of 17http://www.translational-medicine.com/content/10/1/174
on the time post SCI. The bins consisted of 0–2 months(n=5), 3–12 months (n=5), 13–48 months (n=3), and 49+months (n=14). Principal component analysis (PCA)revealed that the healthy control group and the 0–2month duration were very similar (Figure 2). Likewise,the 13–48 and 49+ month groups were similar to eachother, while the 3–12 month group was too distortedby one outlier to provide meaningful information.
Enterococcus and Escherichia emerged as major contribu-tors to these profiles as illustrated in the vector diagram(inset, Figure 2).
Urinary microbiome differed with bladder managementTo get an overview of bacterial species variation withinthe NB group, differences in microbiome taxonomicprofiles between males and females and among bladder
0 ~ 2
13 ~ 483 ~ 12
49 +Healthy
Corynebacterium
Enterococcus
Escherichia
Lactobacillus
Peptoniphilus
Prevotella
Streptococcus
Figure 2 OTU differences among individuals by duration of neuropathic bladder. A PCA analysis of the OTU counts of 52 individuals. Thepoints are circled and colored based on the duration (in months) of neuropathic bladder (see key). The inset depicts a vector plot indicating themost influential principal component (bacterial genus).
Fouts et al. Journal of Translational Medicine 2012, 10:174 Page 8 of 17http://www.translational-medicine.com/content/10/1/174
management groups were plotted (Figure 3). Lactobacil-lus species were significant (P < 0.05) contributors to fe-male healthy controls and the void NB groups, whileCorynebacterium sp. defined the healthy bladder maleurinary microbiome (Figure 3). The preponderance ofEnterococcus sp. in male NB observed in Figure 1C wasprimarily within the patients with NB who emptied byspontaneous voiding (Figure 3).
Phylogenetic tree of the urinary Lactobacillales revealedpotentially pathogenic speciesA phylogenetic tree was constructed with OTU repre-sentatives of all OTUs with matches to Lactobacillalesand nearest neighbors in the RDP database in order toprovide an environmental context and determinespecies-level taxonomy (Figure 4). Labels of the treewere colored based on NB composition. Red indicatesOTUs composed only of samples from NB, light redindicates OTUs with a majority composition of NB sam-ples, while dark blue denotes OTUs only solely ofhealthy controls, and light blue illustrates OTUs with amajority of reads from healthy controls. There is muchgreater phylogenetic diversity among the Lactobacillusand Streptococcus branches (green and gray shading, re-spectively, Figure 4). Aerococcus and Enterococcusbranches are dominated by OTUs (red leaves) fromurine of NB patients, suggesting qualitatively that these
two genera may be important indicators of bacteriuria,or of catheter usage. In contrast to these two branches,the Lactobacillus and Streptococcus branches are largelycomprised of OTUs from healthy controls with a fewexceptions (red leaves). One such exception was Lacto-bacillus iners (Figure 4), which was previously shown tosignificantly contribute to UTI [39,40]. Another excep-tion was determined to be Streptococcus salivarius, a lac-tic acid-producing Gram-positive organism typicallyfound in the oral cavity, and an opportunistic pathogenimplicated in bacteremia [41-43] and septicemia [44].
Phylogenetic diversity of the urinary enterobacterialesIn contrast to the Lactobacillales, the Enterobacterialeshad no OTUs composed solely of healthy controls; indi-cating this group of bacteria may be a potential indicatorof future UTI (Figure 5). Shaded regions of the treenoted unambiguous genus taxonomy as follows: Escheri-chia, Enterobacter, Klebsiella, Proteus, Morganella, andProvidencia. All of these genera have been associatedwith UTI, and dominated the tree. Klebsiella stood outfrom the others, having been divided into three distinctbranches. Top matches to N2-fixing, plant-associatedKlebsiella pneumoniae and K. variicola isolates ratherthan known K. pneumoniae clinical strains suggests thatN2-fixing, plant-associated Klebsiella may exist in urine.Further work is needed to confirm this result.
... [remaining 189]
Corynebacterium
Salmonella
Aerococcus
Streptococcus
Staphylococcus
Klebsiella
Lactobacillus
Escherichia/Shigella
Proteus
Enterococcus
50 %
40 %
30 %
20 %
10 % 0 %
10 %
20 %
30 %
40 %
50 %
taxon: Genuspredominantly in Male
*
*
*
*
*
+
+
++
+
+
Bla
dd
er f
un
ctio
n
Healthy
Void
P < 0.05
FemaleMale
n=10 n=15
n=5 n=8
+*Gender
IC
FC
n=4 n=4
n=5 n=6
predominantly in Female
Female - Male
100%
0
M F
100%
0
M F
Figure 3 Differences in the relative OTU counts between males and females stratified by bladder management. For every individual, theOTU counts were normalized to the individual’s total OTU count. Differences between relative OTU counts were calculated by subtracting theaverage OTU counts from females and males per bladder management category; healthy control, void (SCI patient with no catheter usage),indwelling catheter (IC), and Foley catheter (FC) (see key for color coding and sample sizes). The X-axis indicates the difference in relative OTUcounts per bacterial genus indicated on the Y-axis. Significant differences (P < 0.05) between the relative OTU counts are indicated by plus sign(+) for gender, and an asterisk (*) for bladder management. The inset depicts the mean and standard deviation of OTU counts of the indicatedgenera for each catheter management group.
Fouts et al. Journal of Translational Medicine 2012, 10:174 Page 9 of 17http://www.translational-medicine.com/content/10/1/174
Metaproteomics reveals anti-microbial and pro-inflammatory responses in the absence of diagnosed UTIShotgun proteomic data for a subset of nine urinarysamples searching a protein sequence databases of 19bacterial genomes were compared with the OTU ana-lysis derived from 16S rDNA sequencing. In addition toprotein identifications targeting the 19 species frequently
associated with UTIs or urethra colonization, a non-redundant human protein database was searched to as-sess the detection of host responses towards bacterialcolonization. While proteomic analysis identified morespecies than those diagnosed by culture methods, it didnot identify the fastidious anaerobic or microaerophilicbacteria frequently assigned by the RDP classifier in the
L. fornicalis
B9PWM_0/2_2/0
BATVQ_2/9_4/7
CH
YTM
_0/2_0/2E
Q9K
3_1/6_1/6
DQ97
6175
_unc
ultu
red_
Lact
obac
illus_
hum
an_v
ulva
AY
959115_uncultured_bacterium_hum
an_vaginal_epithelium
ELG60_0/2_0/2
9/7_
31/3
_5X9
0C
CJL94_0/6_2/4
AY959007_uncultured_bacterium_human_vaginal_epithelium
U64
458_
Aer
ococ
cus_
urin
ae_1
656-
92
AY830405_uncultured_Enterococcus_sp_insect_gut
AY
959163_uncultured_bacterium_hum
an_vaginal_epithelium
DI99W
_1/1_0/2
BUCQ1_0/4_1/3
DRUV7_1/1_1/1
AOOHM_1/1_2/0
FM874143_uncultured_bacterium_mattress_dust
E4HA6_1/1_1/1
AV6VP_1/5_1/5
A67P0_1/5_2/4
AF477496_Enterococcus_faecalis_PL9003_anti_Hpylori
EF61
9539
_Aer
ococ
cus_
urin
ae_4
288_
pedi
atric
_inf
ectio
n
CP
KJ4
_1/1
_0/2
GU400239_Streptococcus_sp_oral_taxon_071_human_oral_cavity
AY959057_uncultured_bacterium_hum
an_vaginal_epithelium
B98D5_
0/2_2
/0
CKRR3_1/6_3/4
ET
MN
B_0
/2_0
/2
HM812060_uncultured_bacterium_mouse_skin
GU4171
95_E
_faec
alis_
CMC1260
_hum
an_o
ral_c
avity
AS
OP
6_7/
2_0/
9
A6WC5_2/5_1/6
AJ580649_uncultured_bacterium_Indonesia_sediment_of_black_water_ecosystem
DNVDQ_2/4_4/2
A1I42_4/5_9/0
AF215878_Lactobacillus_sp_KLB46_human_vagina
A1H
3Q_0/4_0/4
EXBUD_1/2_2/1
FJ836224_uncultured_bacterium_streptomycin-treated_mouse_ileum
DPK50_1/1_2/0
CNPMV_1/1_2/0
DQJ1A_1/1_1/1
D04U
3_0/4_2/2
EE
H8E
_2/0_0/2H
M31
2111
_unc
ultu
red_
bact
eriu
m_h
uman
_ski
n_an
tecu
bita
l_fo
ssa
EF7034
75_u
ncult
ured
_Bac
illi_b
acte
rium
_hum
an_g
ut_I
BD
CB5SD_1/2_0/3
CNSAR_3/14_5/12
EEDFZ_1/5_1/5
DJP
1P_1/1_0/2
Y1731
8_Aer
ococ
cus_
chris
tens
enii_
nov_
hum
an_v
agina
GQ
925858_Streptococcus_agalactiae_cattle_subclinical_mastitis
CX
YU
D_2
/1_0
/3
CF
8MT
_1/1_1/1
EF3
6559
2_un
cultu
red_
bact
eriu
m_h
uman
_vag
ina_
stre
ss
EF473998_Streptococcus_mitis_human_sputum_cystic_fibrosis
AF243177_Lactobacillus_vaginalis_ATCC
49540_human_vagina_T
C6P
E2_
1/1_
1/1
BT
96A
_1/2
_1/2
BN4G4_1/5_0/6
BV1G
U_2
/2_3
/1
C5B0T_2/14_6/10
AY959006_uncultured_bacterium_human_vaginal_epithelium
JF227975_uncultured_bacterium
_human_skin_volar_forearm
D4EYK_3/10_5/8
BUA8N_0/2_2/0
A15IK_2/0_2/0
DQ975893_uncultured_Lactobacillus_human_vulva
AJ968601_Enterococcus_faecium_uc1402_human_healthy_umbilical_cord_blood
EJ0W
Z_3/14_3/14
AY
9591
79_u
ncul
ture
d_ba
cter
ium
_hum
an_v
agin
al_e
pith
eliu
m
CB18Z_7/7_12/2
GU418275_Streptococcus_gordonii_human_oral_cavity
EF
2058
39_u
ncul
ture
d_ba
cter
ium
_hum
an_i
leum
_Cro
hns
2/0_2/0_V
BS7
B
DZHMC_0/3_0/3
GQ
019435_uncultured_bacterium_hum
an_skin_umbilicus
CHHEL_1/1_2/0
C9FXZ_2/5_0/7
ERG8K_1
/10_
4/7
EU452467_uncultured_bacterium_mouse_cecum
DP0JB_2/2_4/0BJK07
_1/2
_1/2
B3LDP_0/2_1/1
DQ
9759
85_u
ncul
ture
d_La
ctob
acill
us_h
uman
_vul
va
EU630086_uncultured_Enterococcaceae_bacterium_showerhead
B3PI
U_2
/17_
6/13
AY958963_uncultured_bacterium_human_vaginal_epithelium
AR
JJX
_0/2
_1/1
C7M
SW
_0/3
_2/1
FN826137_uncultured_bacterium_hum
an_sputum_cystic_fibrosis
AX21Q_8/3_0/11
GU430341_Streptococcus_human_oral_cavity
GU
4288
52_A
naer
ococ
cus_
lact
olyt
icus
_hum
an_o
ral_
cavi
ty
B9ISU_2/6_1/7
GU417978_Lactobacillus_gasseri_human_oral_cavity
DY91I_2/0_0/2
AX
7B5_
2/4_
0/6
CAABG_1/2_0/3
AY256595_uncultured_bacterium_natural_gas_pipelines
DQ346421_uncultured_Streptococcus_sp_human_oral_enamel
GU417260_Enterococcus_faecalis_human_oral_cavity
AY
958802_uncultured_bacterium_hum
an_vaginal_epithelium
A2M
KT_2
/0_2
/0
A5WAM_2/10_1/11
CH
2XB
_1/1_1/1
CB6BM_1/1_2/0
AB
441039_Streptococcus_anginosus_hum
an_bacteremia
GU
415145_Streptococcus_anginosus_hum
an_oral_cavity
BU26R_2
/2_3/1
BA
ICN
_0/5
_2/3
GU173844_Lactobacillus_gasseri_Kx356C1_human_stomach
CGB6C_3/13_4/12
B1K79_1/3_2/2
D0AQO_1/1_1/1
AB023574_Streptococcus_agalactiae_JCM
5671_ATCC
13813_raw_m
ilk_T
C54MF_3/1_0/4
GQ087229_uncultured_bacterium_human_skin_gluteal_crease
ERBTK_1/5_1/5
DPX1R_4/1_0/5
DS
6GE
_3/0_0/3
E14EV_1/1_1/1
DNYG7_2/7_2/7
GQ153750_uncultured_bacterium_laying_hen_crop
DQ
9759
32_u
ncul
ture
d_La
ctob
acillu
s_hu
man
_vul
va
BUSG
H_2
/1_3
/0
U64
459_
Aer
ococ
cus_
urin
ae_9
44-9
4
AF243143_Lactobacillus_jensenii_BLB1a_human_vagina
GQ
105899_uncultured_bacterium_hum
an_skin_gluteal_crease
CMO9V_2/7_2/7
C2D35_2/0_0/2
B1S
LS_6
/3_1
/8
BX
Y4Z
_1/2_1/2
DT22J_9/3_2/10HM822091_uncultured_bacterium_mouse_skin
CX626_6/1_0/7
B8JV
B_3/
13_7
/9
GQ153752_uncultured_bacterium_laying_hen_crop
HM
3357
20_u
ncul
ture
d_ba
cter
ium
_hum
an_s
kin_
popl
iteal
_fos
sa
CK
XW
K_2/8_5/5
DLFP6_0/2_0/2
AB
002489_Streptococcus_dysgalactiae_subsp_dysgalactiae_sw
ine
EF151144_Streptococcus_sp_10aMclG2_hum
an_blood_alpha-hemolytic
GQ
0014
83_u
ncul
ture
d_ba
cter
ium
_hum
an_s
kin_
butto
ck
B55HL_9/9_16/2
EO
AFF
_2/3
_1/4
EA8WJ_2/3_2/3
D7J3W_1/2_2/1
ELMEY_0/7_2/5
DW
SF
K_2
/1_1
/2
DW
LRT
_0/2
_2/0
Streptococcus
Fack
lam
ia/G
lobi
cate
lla
Aeroco
ccus
Enterococcus
Lactobacillus
L. gasseri
L. iners
L. fornicalisL. crispatus
L. fornicalis
S. salivarius
Figure 4 Phylogenetic diversity of Lactobacillales 16S rDNA sequences in human urine. NJ tree clustering of Lactobacillales OTUrepresentatives labeled based on similarity to known RDP database sequences (gray), and OTU composition. Leaves are colored as follows: OTUsconsisting of only healthy individuals (dark blue), mostly healthy (light blue), only NB (red), mostly NB (pink/salmon). Branches were highlightedand labeled by identifiable bacterial genera. Genus-level classification was based on the OTU representative RDP classification and theclassification of nearest neighbors the RDP alignment. The nodes show SequenceID_#male/#female_#SCI/#healthy subjects.
Fouts et al. Journal of Translational Medicine 2012, 10:174 Page 10 of 17http://www.translational-medicine.com/content/10/1/174
16S rDNA analysis with the exception of Lactobacillus.We assume this to be due to species abundance issues.Proteins encoded by Escherichia, Klebsiella and Entero-bacter species were identified in half of the examinedsamples, generally in agreement with 16S rDNA data(Table 2). Metaproteomic profiles allowed preliminaryinsights into the production of bacterial factors interact-ing with the urinary tract environment. Identifications ofsuch proteins as limited to those urine donor samples
with evidence for the initiation of pro-inflammatory andmicrobicidal host defenses (high spectral counts for cal-protectin subunits; identification of lactotransferrin, mye-loperoxidase and eosinophil peroxidase). We emphasizethat the sample sizes are too small to predict whetherthese observations will be reproducible using larger-scaleproteomic surveys. Pseudomonas aeruginosa, Enterobac-ter hormaechei and E. coli (known opportunistic patho-gens in the urinary tract) produced proteins for iron/
C9S
JN_9
/5_1
2/2
A8HW8_8/10_14/4
AP2AT_2/0_2/0
AQ
TX
7_6/
1_7/
0
D9P9Z_2/2_3/1
EE
ISB
_1/2_3/0
AM942759_Proteus_mirabilis_HI4320
D9F
XS
_1/3_4/0
CSSOF_5/1_6/0
EQKAQ_3/0_3/0
DKE1J_3/1_3/1
DGWP6_9/9_14/4
CNWY5_2/0_2/0
BCOTZ_2/0
_2/0
GU477712_Proteus_mirabilis_Qy
B2V6H_11/8_15/4
CDHKM_3/1_3/1
CU1T4_3/0_3/0
D9O
GA
_2/0
_2/0
AY941834_Pantoea_agglomerans_XW112
D1W
60_4/1_4/1
BD
DN
D_3
/0_3
/0
EF5
0982
4_un
cultu
red_
bact
eriu
m_h
uman
_end
otra
chea
l_as
pira
te
CC
G22
_2/0
_2/0
CU
QU
S_3
/0_3
/0
AJ005541_uncultured_gamma_proteobacterium_groundwater
EF603446_uncultured_bacterium_mouse_cecum
COV7Y_0/2_2/0
AMTFI_2/0_2/0
E25J6_3/4_7/0
D4EAY
_9/6
_13/
2
BT1SP_5/2_6/1
JF144604_uncultured_bacterium
1_human_skin_antecubital_fossa
EEE3I_4/1_5/0
C36QD_5/1_6/0
EC
Q1K
_2/0
_2/0
A3G0Z_2/0_2/0
BB
V55_3/0_3/0
GQ
4166
40_u
ncul
ture
d_K
lebs
iella
_sp_
biol
ogic
al_d
egre
asin
g_sy
stem
s
EDK5T_2/0_2/0
BFZJQ_1/1_2/0
B6CL9
_2/0
_2/0
DHG3H_2
/0_2/0
AITJD_6/8_12/2
D7AWL_3/0_3/0
CAL
J2_3
/0_3
/0
EJUAG_8/8_15/1
GQ
097454_uncultured_bacterium_hum
an_skin_skin_popliteal_fossa
HQ
2042
83_b
acte
rium
_N2_
fixin
g_su
garc
ane
A0F
Q5_3/0_3/0
ESMIP_4/1_5/0
C6PC4_8/5_9/4
DQ
816193_uncultured_bacterium_cecum
_exgermfree_m
ouse_zebrafish
GU420979_Proteus_mirabilis_human_oral_cavity
CN
JCE
_2/0_2/0
DY
ZE
U_3
/1_3
/1
AIHBI_2/0_2/0
BPEKN_3/2_4/1
B00H
B_3/0_3/0
EIN2R
_3/3
_6/0
B7O2F_3/0_3/0
DN
P1V
_3/1
_3/1
C10FY_2/0_2/0
CV
E4E
_1/1_1/1
AJ417471_Enterobacter_cloacae_subsp_cloacae_30
CP0
0064
7_Kl
ebsi
ella
_pne
umon
iae_
subs
p_pn
eum
onia
e_M
GH
7857
8
EK0QY_10/2_11/1
EZ
LKX
_2/1_2/1
GQ157710_uncultured_bacterium_human_intestinal_pouch_pouchitis
C7PC4_3/1_4/0
D6OB0_3/0_3/0
AR
N6V
_4/4_6/2
CW
AF
Z_7
/1_8
/0
CMXYJ_
3/0_
3/0
BO
B9Y
_4/1
_5/0
B4T57_4/1_5/0
EY01E_12/11_19/4
C6L5C
_5/0_5/0
EUZTE_4/1_5/0
CFG1X_1/1_1/1
AH
IUK
_4/0
_4/0
EF
95S_4/0_4/0
AY
2566
02_u
ncul
ture
d_ba
cter
ium
_nat
ural
_gas
_pip
elin
e
B35IB_3/0_3/0
GQ293382_uncultured_bacterium_Cuiaba_River
BKI42_2/0_2/0
DI9
JI_2
/1_2
/1
COI8B_1/1_2/0
AF1
8969
7_ac
tivat
ed_s
ludg
e_ba
cter
ium
AVCYZ_6/6_11/1
BL8ST_1/2_2/1
DXIZJ_3/0_3/0
BCD89_1/3_3/1
GU414602_Pseudomonas_aeruginosa_human_oral_cavity
EIBOZ_2/0_2/0
DQ
432287_uncultured_gamm
a_proteobacterium_Lake_H
amra_sedim
ent
secef_namuh_
muiret cab_derutlucnu_744891U
ED
RM
CP
_6/2
_7/1
CBKNW_3/1_3/1
BG088_1/2_3/0
AG
UX
B_2/0_2/0
AF294410_Pasteurella_multocida_CCUG17976_T
DQ364575_Proteus_mirabilis_buffalo_hide
GQ
416828_uncultured_Providencia_sp_biological_degreasing_systems
CV5BB_3/1_4/0
BAI1L_1/1_2/0
CSPP2_6/2_6/2
AY664596_Escherichia_coli_PETROMIC_H01_injection_water_oilfield
EU1F9_
2/0_2
/0EU6H
8_4/
0_4/
0
D0ZN9_3/0_3/0
EN
T4K
_2/0_2/0
EJE5T_3/0_3/0
C1Y
3J_8/2_10/0
AJ580652_uncultured_bacterium_sediment_black_water
C41F5_4/7_9/2
CLXC8_3/0_3/0
HM827132_uncultured_bacterium_mouse_skin
DQ
8042
08_u
ncul
ture
d_ba
cter
ium
_hum
an_f
eces
_obe
sity
CICVW_2/3_5/0
EF52
5539
_Mor
gane
lla_s
p_R
P4
D5G
KL_2/0_2/0
CN0FI_7/8_9/6
EFZA
9_2/1_2/1
BQL2P_3/0_3/0
FJ454090_uncultured_bacterium_hum
an_stool_obesity
DYACK_7/2_8/1
DKGM2_3/0_3/0
EQ
CU
X_2/0_2/0
EY
TF
Y_7
/3_9
/1
AU
3YU
_3/0
_3/0
AJ005653_uncultured_gamma_proteobacterium_groundwater
ET
8BL_
2/0_
2/0
AT1LI_2/1_3/0
BUGIL_11/12_16/7
CH67P_1/3_4/0
AY941085_biocide_degrading_bacte
rium
EOG88_3/0_3/0
AP009378_Escherichia_coli_SE15CWUJI_10/10_16/4
Klebs
iella
3
Kle
bsie
lla 2
Escherichia
ProteusProteus
Klebsiella 1
Enterobacter
Morganella/Providencia
Figure 5 Phylogenetic diversity of Enterobacteriales 16S rDNA sequences in human urine. NJ tree clustering of Enterobacteriales OTUrepresentatives labeled based on similarity to known RDP database sequences (gray), and OTU composition. Leaves are colored as follows: OTUsconsisting of only healthy individuals (dark blue), mostly healthy (light blue), only NB (red), mostly NB (pink/salmon). Branches were highlightedand labeled by identifiable bacterial genera. Genus-level classification was based on the OTU representative RDP classification and theclassification of nearest neighbors the RDP alignment. The nodes show SequenceID_#male/#female_#SCI/#healthy subjects.
Fouts et al. Journal of Translational Medicine 2012, 10:174 Page 11 of 17http://www.translational-medicine.com/content/10/1/174
siderophore acquisition and high mobility (flagellins) insubjects 36 and 51 (Table 3). Flagellins are important forswarming and spread in the urogenital tract during infec-tions [45]. The iron-sequestering protein lactotransferrinmay be expressed and released into the urinary tract byneutrophils to sequester iron, which in turn, E. coli andE. hormaechei may counteract by addition of copiousamounts of iron/siderophore receptors to the outermembrane protein repertoire (Table 3). While the S100-A8 and S100-A9 calprotectins appear to have numerousphysiological functions, one of them is to sequester zincin response to sensing the presence of bacterial patho-gens. This sequestration can inactivate metalloprotei-nases that the bacteria produce and secrete to invade thehost tissue [46]. Such proteases were not identified inour datasets.
Several other proteins implicated in the innate im-mune response were detected (Table 3). Calprotectin hasacute and chronic pro-inflammatory functions andrecruits immune cells to the site of inflammation [47].Antimicrobial peroxidases, such as eosinophil and mye-loperoxidase, produce reactive oxygen species during therespiratory burst of neutrophils and are directly microbi-cidal. Like calprotectin, annexin A1 is a calcium-bindingprotein that was detected in high abundance primarilyin those donors where respiratory burst of neutrophilsdid seem to be muted (subjects 1, 16, 31, 34, 37, 39).These proteins are implicated in innate immunity, influ-ence cell apoptosis and function as damage-associatedmolecular patterns (DAMPs) [48,49]. For a complete listof bacterial and human proteins profiled in this study,see Additional file 1.
Table 2 Bacterial profiles of urinary samples
Urinedonor
Species(proteome)
Species(urine culture)
Species (16SrRNA profile)*
LeukocyteEsterase
WBC(no./hpf)
1(HC) - - Lactobacillus NEG 0
16(HC) Lj Lactobacillus spp Lactobacillus NEG 0-1
31(void) Ec, Eh, Kp - Enterobacter, Enterococcus, Escherichia,Klebsiella, Lactobacillus, Streptococcus
TRA 1-2
33(void) - Ec, Ef Enterobacter, Enterococcus,Escherichia, Klebsiella
2+ 5-9
34(void) Ec, Eh, Kp Klebsiella oxytoca Enterobacter, Escherichia,Lactobacillus, Streptococcus
2+ 5-9
36(IC) Ec, Eh Ec Enterobacter, Escherichia, Klebsiella,Lactobacillus, Streptococcus
NEG 0-1
37(IC) - - Enterobacter, Enterococcus, Escherichia, Klebsiella,Lactobacillus, Pseudomonas, Streptococcus
NEG 1-2
39(IC) Ec, Kp Kp Enterobacter, Klebsiella,Lactobacillus, Streptococcus
NEG 3-4
45(FC) Eh, Kp, Pa, Pm# Ec, Ef, Kp, Pa, Ps Enterobacter, Enterococcus, Escherichia, Klebsiella,Lactobacillus, Proteus, Pseudomonas, Streptococcus
NEG 0-1
51(FC) Ec, Eh, Pa, Sp Ec, Ef, Pa Enterobacter, Enterococcus, Escherichia, Klebsiella,Pseudomonas, Streptococcus
2+ 5-9
*Not comprehensive. 16S rDNA data is at the genus level.#The species for this sample were determined from urine cultures, but based on shotgun proteomic analysis after protein extraction from colonies.Ec Escherichia coli, Eh Enterobacter hormaechei, Ef Enterococcus faecalis, Kp Klebsiella pneumoniae, Pa Pseudomonas aeruginosa, Pm Proteus mirabilis, Ps Providenciastuartii, Sp Streptococcus pneumoniae.WBC = white blood cells, HC = healthy control, IC = intermittent catheter, FC = Foley catheter, NEG = negative, TRA = trace.
Fouts et al. Journal of Translational Medicine 2012, 10:174 Page 12 of 17http://www.translational-medicine.com/content/10/1/174
DiscussionIn this study we describe the healthy urine microbiomein a heterogeneous population of men and women withand without NB, using both 16S rDNA sequencing andmetaproteomic analysis. Based on other reports [50-53]and including our data, this is further confirmation thatthe commonly held clinical belief that healthy urineshould be sterile is false. Specifically, our data indicatethat (1) when collected by the routine method of “clean,midstream catch”, healthy urine is not aseptic; (2) thehealthy and NB urine microbiomes differs by gender; (3)the asymptomatic bacteriuria urine microbiome ofpeople with NB differs from that of healthy controls;and (4) the asymptomatic bacteriuria urine microbiomeof people with NB differs depending on duration of ex-posure to and type of urinary catheter.This is the first report comparing the healthy urine
microbiome in male and female subjects. Historically,and based on cultivation results, clinicians have assumedurine to be ‘sterile’. However, Wolfe et al. recently used16S rDNA sequencing to identify uncultivated bacteriain the urine of healthy women [53]. Our data confirmthese results in women, and further show that unculti-vated bacteria are also present in the urine of healthymen. Moreover, we demonstrate that the healthy urinemicrobiome of males and females differs, with men hav-ing significantly more relative abundance of Corynebac-terium, a common component of the superficial skinflora, and women having significantly greater relative
abundance of Lactobacillus. Critical to this discussion isan understanding that because all samples from healthysubjects and those of subjects with NB who voided werecollected by midstream clean catch, it is not possible todistinguish whether the microbes identified originated inthe bladder, urethra, or both. Therefore, the possibilityexists that the identified urine microbiomes are popu-lated by species that exist in the bladder, urethra, orboth.Our finding of predominance of Corynebacterium in
healthy males suggests that this microbe may contributeto the healthy urine microbiome. Not only was Coryne-bacterium identified to a significantly greater degree inhealthy males compared to those with NB, males with NBwho voided and were sampled by clean catch had the low-est abundance of this species. Dong et al. compared cleancatch urine and distal urethral swabs in healthy volunteersand similarly found a predominance of Corynebacteriumin both types of sampling but in significantly greateramounts in the clean catch urine samples [51]. Takentogether, the data suggest that Corynebacterium mayreside in the proximal urethra and/or bladder in additionto the distal urethra, and may play a role in the healthyurine microbiome.While it is well established that in most healthy
women of childbearing age the vaginal tract is colonizedby Lactobacillus species [54-56], this has not been inves-tigated in women with NB. We too found a clear pre-ponderance of Lactobacillus in healthy control females;
Table 3 Human and bacterial proteins potentially contributing directly or indirectly to host-pathogen interactions inthe urinary tract
Human protein Gene name PSMs Functional role Urine donor
Protein S100-A9 Calprotectin L1H subunit S10A9 60 Pro-inflammatory, metal ion-chelating 1,16,33,36,51
Protein S100-A8, Calprotectin L1L subunit S10A8 10 Pro-inflammatory, metal ion-chelating 1,16,36,51
Protein S100-A12 S10AC 14 Pro-inflammatory 36
Myeloperoxidase PERM 13 Microbicidal 36,51
Eosinophil peroxidase PERE 4 Microbicidal 51
Lactotransferrin TRFL 6 Pro-inflammatory, iron-chelating 36
14-3-3 protein sigma SFN 5 DNA damage response, cell proliferation 16
SNC66 protein - 12 Secreted, Ig-like domain 51
Heat shock protein beta-1 HSPB1 20 Anti-inflammatory, cell proliferation 1
Annexin A2 ANXA2 8 Cell proliferation, cell adhesion 16
Uromodulin UMOD 78 Cell protection, inhibitor of Ca crystallization, 1,16,31,36,37,39,51
Cystatin-B CYTB 8 Immunomodulatory, cathepsin inhibitor 1,16
14-3-3 protein zeta/delta YWHAZ 2 Adaptor protein, tyrosine phosphorylation pathways 1,16
Serpin B3 SPB3 7 Immunomodulatory, serine protease inhibitor 16
Small proline-rich protein 3 SPRR3 7 Cell repair and proliferation 16
Annexin A1 ANXA1 28 Anti-apoptotic, T-cell differen-tiation, signaling pathways 1,16,31,34,37,39
Glutathione S-transferase P GSTP1 3 Anti-apoptotic, tyrosine phosphorylation pathways 1,16,31,34,37
Bacterial protein Species*
Colicin receptor CirA Ec 2 Iron/colicin-binding 36
OM heme/hemoglobin receptor ChuA Ec 12 Iron-binding 51
Putative pesticin receptor Psn Ec 23 Iron-binding 51
Ferrienterobactin receptor FepA Eh 9 Iron-binding 51
Putative fimbrillin MatB Ec 2 Adhesion 51
Flagellin protein type B FliC Pa 13 Mobility and adhesion 51
Flagellin protein FliC Eh 8 Mobility and adhesion 51
Flagellin Ec 7 Mobility and adhesion 51
Ferrienterobactin receptor FepA Ec 22 Iron-binding 36, 51
Section one of the table lists proteins with potential pro/anti-inflammatory, immune-modulatory and microbicidal activities. Section two lists identified bacterialproteins implicated in virulence/survival or serving as a target for host defensive mechanisms.PSMs: the highest number of peptide-spectral matches (PSMs) is provided for each listed protein * for species abbreviations, see Table 2 (minimal MascotPercolator PEP value: 10-4).
Fouts et al. Journal of Translational Medicine 2012, 10:174 Page 13 of 17http://www.translational-medicine.com/content/10/1/174
however, in addition to the greatest relative abundanceof Lactobacillus in healthy control females, there was aprogressive reduction in abundance of Lactobacillus infemales with NB who void (clean catch sample) or whouse intermittent catheterization, and females with NBwho use indwelling (Foley) catheters (Figure 3 inset).This may suggest that either increasing exposure to aurinary catheter and/or increasing severity of NB can in-fluence the ability of Lactobacillus to colonize the urinarytract. Alternatively, Lactobacillus may merely be a con-taminant of the external urethra that arises from proxim-ity to the vaginal flora. However, this appears less likelysince Wolfe et al. showed the presence of Lactobacillusin urine collected by transurethral catheters and supra-pubic aspirate, which samples the bladder directly [53].Because lactic acid-producing Lactobacillus speciescontribute to controlling the growth of more virulent
bacteria that cannot survive in a more acidic environ-ment, the presence of Lactobacillus within the urethraand/or bladder may be protective. This has beenshown to be the case in infants [57] and may also betrue for males as Lactobacillus has been shown to bepresent in clean catch urine samples of healthy malesby Dong et al. [51] and in our study. Our findingssuggest that Lactobacillus may be a commensal organismpresent during states of health, more in females than inmales, and that the microbiome of at-risk populationsmay be characterized by a distinct lack of Lactobacillus,which perhaps creates a better environment for thegrowth of pathogenic microorganisms. Together, thesefindings suggest that the clinical objective of ‘sterile’,microbe-free urine may not be optimal for the patient.This is the first report of 16S rDNA sequencing of
urine in people with NB, providing much more detail
Fouts et al. Journal of Translational Medicine 2012, 10:174 Page 14 of 17http://www.translational-medicine.com/content/10/1/174
about the ABU state than has previously been availablethrough cultivation-based evidence. Standard cultivationdiagnostics show populations vulnerable to bacteriuriainclude nursing home residents utilizing long-termindwelling catheterization [58], institutionalized Veter-ans utilizing intermittent catheterization [59], andpatients with SCI who utilize urinary catheters [60-62].Our analysis not only confirms the cultivation-based evi-dence, but also shows that a significantly differentmicrobiome was present in the NB group, and that 16SrDNA sequencing specifically identifies microorganisms,such as Enterobacteriales, as potential major contribu-tors to a pathogenic microbiome. These results supple-ment those from Bank et al., where urine specimens of142 consecutive patients with varied genitourinary path-ology (kidney stones, indwelling catheters, or suspectedUTI) were analyzed with standard cultivation andscreened for Actinobaculum schaalii using PCR [50].Those authors found that the most heavily colonizedpatients were those who were older and who utilizedindwelling urinary catheters, while the younger patientswho typically use intermittent catheterization (and mayhave utilized urinary catheters for a shorter amount oftime) were colonized with bacteria to a lesser degree.The Enterobacteriaceae was most commonly isolated inthe catheterized specimens in that study. Significantvariance in medical comorbidities, underlying genitour-inary pathology, and method of urine collection limitany further comparisons to that study. Moreover, be-cause our NB population was reportedly asymptomatic(i.e. infection free), our findings demonstrate that in thecatheterized population, the microbiome is intrinsicallydifferent than in controls, even in the absence of illness.This distinction is important, since in the clinical settingABU is often inappropriately treated with antibiotics,which may further disrupt the NB microbiome.The present data also indicate that the urine micro-
biome of healthy subjects with NB became increasinglyabundant with Enterobacteriales with increasing dur-ation of NB, whereas Lactobacillus decreased over time,both in men and women. While the urinary microbiomeof men and women with NB remained similar to that ofhealthy controls during the first several months after NBdiagnosis, by one year the urine microbiome was nearlydevoid of Lactobacillus and dominated by Enterococcus.This further suggests a change in the microbiome withduration of NB that may place patients at increased riskof UTI.Fundamental to these discoveries is the diverse sample
population and our novel analytic approach of utilizing acombination of 16S rDNA sequencing in all subjectsand metaproteomics in a subsample. Clinical gold stand-ard diagnostic testing when a patient presents with signsand/or symptoms of UTI includes (1) urinalysis to
confirm urinary tract inflammation and (2) urine cultureto identify, quantify and predict antimicrobial resistanceto a given pathogen(s). We believe that 16S rDNA se-quencing has the potential for translation to the clinic,offering significant clinical advancement over diagnosticurine culture because it provides a greater depth ofunderstanding and sensitivity pertaining to the compos-ition of commensal and potentially pathogenic microbespresent in urine. Furthermore, prospective assessmentsduring varying periods of health and disease may allowpersonalization of care that has not been possible to datewith our current diagnostic methods. Urinary metapro-teomic profiles in parallel may contribute to the identifi-cation of a host inflammatory response utilizing urinarybiomarkers with greater sensitivity and specificity forUTI than traditional measures of urinary leukocyte es-terase production or white blood cell count detected byurinalysis. We hypothesize that protein profiles withdistinct abundance ratios of immunomodulatory versuspro-inflammatory and microbiocidal molecules, are indi-cative of either UTI or reflect asymptomatic colonization.For instance, the presence of lactotransferrin in urine hasbeen used to support the notion of a “battle for iron”being waged between the host and the pathogen, involv-ing E. coli, particularly in the case of UTI by E. coli giventhe abundance of its triad of iron acquisition receptors[63]. While metaproteomics holds promise as a diagnos-tic method to discriminate between symptomatic UTIand ABU, more in-depth fractionation of samples isneeded to evaluate whether this method can reach thesensitivity of 16S rDNA profiling methods. Larger patientcohorts, including those diagnosed with UTI symptoms,are required to assess if such ‘omics methods’ more ac-curately differentiate UTI from ABU than current diag-nostic standards. In addition, they suggest newconsiderations that may impact preventive and treatmentoptions for people at-risk for UTI.Several limitations of this study are to be considered
when interpreting the results. The major limiting factoris that some subjects (healthy controls and NB subjectswho voided spontaneously) had urine collection viamidstream clean catch sampling while the subjects withNB who managed their bladder with intermittentcatheterization or indwelling Foley (urethral) catheterswere sampled directly from the catheter. Therefore,microbes identified in the former groups could be repre-sentatives from the bladder, urethra, skin, or a combin-ation of these microbiomes, whereas urine from subjectsin the latter groups represents bladder microbiota. Thisdistinction raises other questions, such as to what degreedo differing microbiomes in the urethra and bladder in-fluence each other, and does a changing urethral micro-biome represent a potential antecedent UTI state.Further, given that clinical urine sampling is unlikely to
Fouts et al. Journal of Translational Medicine 2012, 10:174 Page 15 of 17http://www.translational-medicine.com/content/10/1/174
become more invasive (via direct sampling from thebladder), how do clinicians interpret the presence of bac-teria in clean catch midstream urine samples that couldpotentially originate from multiple anatomic locations?Lastly, future studies will employ a larger sample size asour results indicate that the urine microbiome differs bya number of clinical factors requiring multiple stratifica-tion points.
ConclusionsContrary to clinical dogma that healthy urine is sterile;these results suggest that the state of healthy urine is, infact, one of ‘asymptomatic bacteriuria’. Utilizing highthroughput sequencing and metaproteomics, we havedescribed the healthy urine microbiome of a number ofpopulations: male and female healthy controls and healthysubjects with NB. Differing urinary microbiomes for malesand females were described. We have demonstrated thatNB and/or urinary catheterization impacts the healthyurine microbiome in both genders and this varies by typeof bladder management and duration of NB. Furthermore,the presence of a variety of urine microbiomes differingon key, clinical characteristics suggests the benefit of amore personalized approach to UTI care. Clearly, DNAsequencing techniques allow for more specific assessmentof the contributing microorganisms than do current clin-ical diagnostic standards, offering the potential for signifi-cant clinical advancement of diagnostic methods for UTI,which have otherwise remained relatively unchanged fordecades. Longitudinal differentiation of the urine micro-biome at the time of, prior to, and after infection also willbe necessary to fully describe the course of disease and itsantecedents. These findings advance clinical translationalscience toward improved diagnostics and more targeteduse of therapeutics.
Additional files
Additional file 1: Proteomic analyses of urinary pellet samplesusing the LTQ XL ion trap instrument (Thermo-Electron) and theMascot search engine version 2.3 (Matrix Science) for spectralmatches with a 19-species database.
Additional file 2: YAP 16S rDNA sequence-processing and analysispipeline workflow diagram.
Additional file 3: A table summarizing the results of bacterial OTU-based microbiome analysis.
Additional file 4: Bacterial OTU taxonomic profiles at the genuslevel per individual.
AbbreviationsABU: Asymptomatic Bacteriuria; CAUTI: Catheter Associated Urinary TractInfections; CNMC: Children’s National Medical Center; CRAN: ComprehensiveR Archive Network; CTSA: Clinical and Translational Science Awards;DAMPs: Damage-Associated Molecular Patterns; DHHS: Department of Healthand Human Services; DNA: Deoxyribonucleic Acid; Ec: Escherichia coli;EDTA: Ethylenediaminetetraacetic Acid; Ef: Enterococcus faecalis;Eh: Enterobacter hormaechei; ESBL: Extended-Spectrum-Beta-Lactamases;
FASP: Filter-Aided Sample Preparation; FC: Foley Catheter; HC: HealthyControl; Hpf: High power field; IC: Intermittent Catheter; IRB: InstitutionalReview Board; JCVI: J. Craig Venter Institute; Kp: Klebsiella pneumoniae;LCA: Least Common Ancestor; Lj: Lactobacillus jensenii; LTQ: Linear TrapQuadrupole; NB: Neuropathic Bladder; NCATS: National Center for AdvancingTranslational Sciences; NCBI: National Center for Biotechnology Informatio;NCMRR: National Center for Medical Rehabilitation Research; NCRR: NationalCenter for Research Resources; NEG: Negative; NIH: National Institutes ofHealth; NINDS: National Institute for Neurological Disorders and Stroke;NJ: Neighbor-Joining; No: Number; NRH: National Rehabilitation Hospital;NT: Nucleotide; OGE: Oracle Grid Engine; OTU: Operational Taxonomic Unit;Pa: Pseudomonas aeruginosa; PBS: Phosphate Buffered Saline; PCA: PrincipalComponent Analysis; PCR: Polymerase Chain Reaction; Pm: Proteus mirabilis;Ps: Providencia stuartii; RDP: Ribosomal Database Project; RNA: RibonucleicAcid; rRNA: Ribosomal Ribonucleic Acid; SCI: Spinal Cord Injury; SDS-PAGE: Sodium Dodecyl Sulfate Polyacrylamide Gel Electrophoresis;Sp: Streptococcus pneumoniae; SSU: Small Subunit; TRA: Trace; UTI: UrinaryTract Infection; WBC: White Blood Cells; YAP: Yet Another Pipeline.
Competing interestsThe authors declare that they have no competing interests.
Authors' contributionsDEF, RP, SLG conceived and organized the study; MS, BS, SKL, MT performedlaboratory experiments; DEF, RP, SS, HP, SK, MS, IL, BS, SLG processed and/oranalyzed data; and DEF, RP, SS, HP, SK, MT, KEN, SLG wrote the manuscript.All authors read and approved the final manuscript.
AcknowledgementsWe thank Shih-Ting Huang for her contributions to the MS databaseassembly and computational MS data processing and Anthony Scott Durkinfor registering the NCBI project and for assistance with data submission. Weare grateful to J. Craig Venter at JCVI and Eric Hoffman at CNMC for internalsupport of this project. This work was partially supported by NIH NCMRR/NINDS 2R24HD050846-06 (NCMRR-DC Core Molecular and FunctionalOutcome Measures in Rehabilitation Medicine). In addition, this project hasbeen funded in whole or in part with Federal funds (Grant # UL1TR000101)from the National Center for Advancing Translational Sciences (NCATS),National Institutes of Health (NIH), through the Clinical and TranslationalScience Awards Program (CTSA), a trademark of DHHS, part of the RoadmapInitiative, “Re-Engineering the Clinical Research Enterprise.
Author details1J Craig Venter Institute, 9704 Medical Center Drive, Rockville, MD 20850,USA. 2Childrens National Medical Center, 111 Michigan Avenue NW,Washington, DC 20010, USA. 3MedStar National Rehabilitation Hospital, 102Irving Street NW, Washington, DC 20010, USA. 4Georgetown UniversityHospital, 3800 Reservoir Road, Washington, DC 20007, USA.
Received: 3 May 2012 Accepted: 2 August 2012Published: 28 August 2012
References1. Infectious disease, chapter 7, urinary tract infections. http://pathmicro.med.sc.
edu/infectiousdisease/UrinaryTractInfections.htm.2. Foxman B, Barlow R, D'Arcy H, Gillespie B, Sobel JD: Urinary tract infection:
self-reported incidence and associated costs. Ann Epidemiol 2000,10:509–515.
3. Griebling TL: Urologic diseases in america project: trends in resource usefor urinary tract infections in men. J Urol 2005, 173:1288–1294.
4. Litza JA, Brill JR: Urinary tract infections. Prim Care 2010, 37:491–507. viii.5. HHS action plan to prevent healthcare-sssociated infections. http://www.hhs.
gov/ash/initiatives/hai/actionplan.6. Maki DG, Tambyah PA: Engineering out the risk for infection with urinary
catheters. Emerg Infect Dis 2001, 7:342–347.7. Salgado CD, Karchmer RB, Farr BM: Prevention of catheter-sssociated urinary
tract infections. 4th edition. Philadelphia, PA: Lippincott Williams & Wilkins;2003.
8. Macejko AM, Schaeffer AJ: Asymptomatic bacteriuria and symptomaticurinary tract infections during pregnancy. Urol Clin North Am 2007,34:35–42.
Fouts et al. Journal of Translational Medicine 2012, 10:174 Page 16 of 17http://www.translational-medicine.com/content/10/1/174
9. Wald HL, Kramer AM: Nonpayment for harms resulting from medical care:catheter-associated urinary tract infections. JAMA 2007, 298:2782–2784.
10. Saint S, Meddings JA, Calfee D, Kowalski CP, Krein SL: Catheter-associatedurinary tract infection and the medicare rule changes. Ann Intern Med2009, 150:877–884.
11. Dicianno BE, Wilson R: Hospitalizations of adults with spina bifida andcongenital spinal cord anomalies. Arch Phys Med Rehabil 2010, 91:529–535.
12. Filler G, Gharib M, Casier S, Lodige P: Ehrich JH. Dave S: Prevention of chronickidney disease in spina bifida. International urology and nephrology; 2011.
13. Salonen A, Nikkila J, Jalanka-Tuovinen J, Immonen O, Rajilic-Stojanovic M,Kekkonen RA, Palva A, de Vos WM: Comparative analysis of fecal DNAextraction methods with phylogenetic microarray: effective recovery ofbacterial and archaeal DNA using mechanical cell lysis. J MicrobiolMethods 2010, 81:127–134.
14. Liu Z, DeSantis TZ, Andersen GL, Knight R: Accurate taxonomyassignments from 16S rRNA sequences produced by highly parallelpyrosequencers. Nucleic Acids Res 2008, 36:e120.
15. Schloss PD, Westcott SL, Ryabin T, Hall JR, Hartmann M, Hollister EB,Lesniewski RA, Oakley BB, Parks DH, Robinson CJ, et al: Introducing mothur:open-source, platform-independent, community-supported software fordescribing and comparing microbial communities. Appl Environ Microbiol2009, 75:7537–7541.
16. Chou HH, Holmes MH: DNA sequence quality trimming and vectorremoval. Bioinformatics 2001, 17:1093–1104.
17. Engelbrektson A, Kunin V, Wrighton KC, Zvenigorodsky N, Chen F, OchmanH, Hugenholtz P: Experimental factors affecting PCR-based estimates ofmicrobial species richness and evenness. ISME J 2010, 4:642–647.
18. Kunin V, Engelbrektson A, Ochman H, Hugenholtz P: Wrinkles in the rarebiosphere: pyrosequencing errors can lead to artificial inflation ofdiversity estimates. Environ Microbiol 2010, 12:118–123.
19. Huang Y, Niu B, Gao Y, Fu L, Li W: CD-HIT Suite: a web server forclustering and comparing biological sequences. Bioinformatics 2010,26:680–682.
20. Niu B, Fu L, Sun S, Li W: Artificial and natural duplicates inpyrosequencing reads of metagenomic data. BMC Bioinformatics 2010,11:187.
21. Pruesse E, Quast C, Knittel K, Fuchs BM, Ludwig W, Peplies J, Glockner FO:SILVA: a comprehensive online resource for quality checked and alignedribosomal RNA sequence data compatible with ARB. Nucleic Acids Res2007, 35:7188–7196.
22. Haas BJ, Gevers D, Earl AM, Feldgarden M, Ward DV, Giannoukos G, Ciulla D,Tabbaa D, Highlander SK, Sodergren E, et al: Chimeric 16S rRNA sequenceformation and detection in Sanger and 454-pyrosequenced PCRamplicons. Genome Res 2011, 21:494–504.
23. Cole JR, Wang Q, Cardenas E, Fish J, Chai B, Farris RJ, Kulam-Syed-MohideenAS, McGarrell DM, Marsh T, Garrity GM, Tiedje JM: The Ribosomal DatabaseProject: improved alignments and new tools for rRNA analysis. NucleicAcids Res 2009, 37:D141–D145.
24. Hamady M, Knight R: Microbial community profiling for humanmicrobiome projects: Tools, techniques, and challenges. Genome Res2009, 19:1141–1152.
25. The Comprehensive R Archive Network. http://cran.r-project.org/.26. YAP 16S sequence-processing pipeline. http://github.com/shpakoo/YAP.27. Dray S, Dufour AB: The ade4 package: Implementing the duality diagram
for ecologists. J Stat Softw 2007, 22:1–20.28. Hollander M, Wolfe DA: Wiley series in probability and statistics. In
Nonparametric Statistical Methods. 2nd edition. New York: John Wiley &Sons, Inc; 1999:36–51.
29. Ribosome Database Project. http://rdp.cme.msu.edu.30. Sonnhammer EL, Hollich V: Scoredist: a simple and robust protein
sequence distance estimator. BMC Bioinformatics 2005, 6:108.31. Seal BS, Fouts DE, Simmons M, Garrish JK, Kuntz RL, Woolsey R, Schegg KM,
Kropinski AM, Ackermann HW, Siragusa GR: Clostridium perfringensbacteriophages PhiCP39O and PhiCP26F: genomic organization andproteomic analysis of the virions. Arch Virol 2011, 156:25–35.
32. FigTree Grapical Viewer of Phylogenetic Trees. http://tree.bio.ed.ac.uk/software/figtree/.
33. Wisniewski JR, Zougman A, Nagaraj N, Mann M: Universal samplepreparation method for proteome analysis. Nat Methods 2009, 6:359–362.
34. Pieper R, Zhang Q, Clark DJ, Huang ST, Suh MJ, Braisted JC, Payne SH,Fleischmann RD, Peterson SN, Tzipori S: Characterizing the Escherichia coli
O157:H7 Proteome Including Protein Associations with Higher OrderAssemblies. PLoS One 2011, 6:e26554.
35. Uniref90 Protein Sequence Database. http://www.ebi.ac.uk/uniprot/database/download.html.
36. Verberkmoes NC, Russell AL, Shah M, Godzik A, Rosenquist M, Halfvarson J,Lefsrud MG, Apajalahti J, Tysk C, Hettich RL, Jansson JK: Shotgunmetaproteomics of the human distal gut microbiota. ISME J 2009, 3:179–189.
37. Mascot Percolator online help. http://www.matrixscience.com/help/percolator_help.html.
38. Wang Q, Garrity GM, Tiedje JM, Cole JR: Naïve Bayesian classifier for rapidassignment of rRNA sequences into the new bacterial taxonomy. ApplEnviron Microb 2007, 73:5261–5267.
39. Domann E, Hong G, Imirzalioglu C, Turschner S, Kuhle J, Watzel C, Hain T,Hossain H, Chakraborty T: Culture-independent identification ofpathogenic bacteria and polymicrobial infections in the genitourinarytract of renal transplant recipients. J Clin Microbiol 2003, 41:5500–5510.
40. Imirzalioglu C, Hain T, Chakraborty T, Domann E: Hidden pathogensuncovered: metagenomic analysis of urinary tract infections. Andrologia2008, 40:66–71.
41. Beighton D, Carr AD, Oppenheim BA: Identification of viridansstreptococci associated with bacteraemia in neutropenic cancer patients.J Med Microbiol 1994, 40:202–204.
42. Corredoira JC, Alonso MP, Garcia JF, Casariego E, Coira A, Rodriguez A, PitaJ, Louzao C, Pombo B, Lopez MJ, Varela J: Clinical characteristics andsignificance of Streptococcus salivarius bacteremia and Streptococcusbovis bacteremia: a prospective 16-year study. Eur J Clin Microbiol InfectDis 2005, 24:250–255.
43. Ruoff KL, Miller SI, Garner CV, Ferraro MJ, Calderwood SB: Bacteremia withStreptococcus bovis and Streptococcus salivarius: clinical correlates ofmore accurate identification of isolates. J Clin Microbiol 1989, 27:305–308.
44. Hoecker JL, Pickering LK, Groschel D, Kohl S: Streptococcus salivarius sepsisin children with malignancies. J Pediatr 1978, 92:337–338.
45. Lane MC, Alteri CJ, Smith SN, Mobley HL: Expression of flagella iscoincident with uropathogenic Escherichia coli ascension to the upperurinary tract. Proc Natl Acad Sci USA 2007, 104:16669–16674.
46. Isaksen B, Fagerhol MK: Calprotectin inhibits matrix metalloproteinases bysequestration of zinc. Mol Pathol 2001, 54:289–292.
47. Gebhardt C, Nemeth J, Angel P, Hess J: S100A8 and S100A9 ininflammation and cancer. Biochem Pharmacol 2006, 72:1622–1631.
48. Bianchi ME: DAMPs, PAMPs and alarmins: all we need to know aboutdanger. J Leukoc Biol 2007, 81:1–5.
49. Sato S, St-Pierre C, Bhaumik P, Nieminen J: Galectins in innate immunity:dual functions of host soluble beta-galactoside-binding lectins asdamage-associated molecular patterns (DAMPs) and as receptors forpathogen-associated molecular patterns (PAMPs). Immunol Rev 2009,230:172–187.
50. Bank S, Hansen TM, KM SB, Lund L, Prag J: Actinobaculum schaalii inurological patients, screened with real-time polymerase chain reaction.Scand J Urol Nephrol 2011, 45:406–410.
51. Dong Q, Nelson DE, Toh E, Diao L, Gao X, Fortenberry JD, Van der Pol B:The microbial communities in male first catch urine are highly similar tothose in paired urethral swab specimens. PLoS One 2011, 6:e19709.
52. Siddiqui H, Nederbragt AJ, Lagesen K, Jeansson SL, Jakobsen KS: Assessingdiversity of the female urine microbiota by high throughput sequencingof 16S rDNA amplicons. BMC Microbiol 2011, 11:244.
53. Wolfe AJ, Toh E, Shibata N, Rong R, Kenton K, Fitzgerald M, Mueller ER,Schreckenberger P, Dong Q, Nelson DE, Brubaker L: Evidence ofUncultivated Bacteria in the Adult Female Bladder. J Clin Microbiol 2012,50:1376–1383.
54. Pfau A, Sacks T: The bacterial flora of the vaginal vestibule, urethra andvagina in the normal premenopausal woman. J Urol 1977, 118:292–295.
55. Marrie TJ, Harding GK, Ronald AR: Anaerobic and aerobic urethral flora inhealthy females. J Clin Microbiol 1978, 8:67–72.
56. Zhang R, Daroczy K, Xiao B, Yu L, Chen R, Liao Q: Qualitative and semi-quantitative analysis of Lactobacillus species in vagina of healthy fertile andpostmenopausal Chinese women. J Med Microbiol 2012, 61:729–739.
57. Lee JW, Shim YH, Lee SJ: Lactobacillus colonization status in infants withurinary tract infection. Pediatr Nephrol 2009, 24:135–139.
58. Warren JW, Tenney JH, Hoopes JM, Muncie HL, Anthony WC: A prospectivemicrobiologic study of bacteriuria in patients with chronic indwellingurethral catheters. J Infect Dis 1982, 146:719–723.
Fouts et al. Journal of Translational Medicine 2012, 10:174 Page 17 of 17http://www.translational-medicine.com/content/10/1/174
59. Terpenning MS, Allada R, Kauffman CA: Intermittent urethralcatheterization in the elderly. J Am Geriatr Soc 1989, 37:411–416.
60. Cardenas DD, Moore KN, Dannels-McClure A, Scelza WM, Graves DE, BrooksM, Busch AK: Intermittent catheterization with a hydrophilic-coatedcatheter delays urinary tract infections in acute spinal cord injury: aprospective, randomized, multicenter trial. Pm R 2011, 3:408–417.
61. Esclarin De Ruz A, Garcia Leoni E, Herruzo Cabrera R: Epidemiology and riskfactors for urinary tract infection in patients with spinal cord injury.J Urol 2000, 164:1285–1289.
62. Wyndaele JJ: Complications of intermittent catheterization: theirprevention and treatment. Spinal Cord 2002, 40:536–541.
63. Skaar EP: The battle for iron between bacterial pathogens and theirvertebrate hosts. PLoS Pathog 2010, 6:e1000949.
doi:10.1186/1479-5876-10-174Cite this article as: Fouts et al.: Integrated next-generation sequencingof 16S rDNA and metaproteomics differentiate the healthy urinemicrobiome from asymptomatic bacteriuria in neuropathic bladderassociated with spinal cord injury. Journal of Translational Medicine 201210:174.
Submit your next manuscript to BioMed Centraland take full advantage of:
• Convenient online submission
• Thorough peer review
• No space constraints or color figure charges
• Immediate publication on acceptance
• Inclusion in PubMed, CAS, Scopus and Google Scholar
• Research which is freely available for redistribution
Submit your manuscript at www.biomedcentral.com/submit