of July 9, 2018.This information is current as

InfectionDefense against Bacterial Urinary TractPathways and Rapid Synthesis of IL-10 for

Drive Diverse BiologicalEscherichia coliBladder in Response to Uropathogenic Innate Transcriptional Networks Activated in

Allan W. Cripps, Michael Crowley and Glen C. UlettPetra Derrington, Helen Irving-Rodgers, Andrew J. Brooks,Cui, Richard I. Webb, Makrina Totsika, Mark A. Schembri, Benjamin L. Duell, Alison J. Carey, Chee K. Tan, Xiangqin

http://www.jimmunol.org/content/188/2/781doi: 10.4049/jimmunol.1101231December 2011;

2012; 188:781-792; Prepublished online 19J Immunol 

MaterialSupplementary

1.DC1http://www.jimmunol.org/content/suppl/2011/12/19/jimmunol.110123

Referenceshttp://www.jimmunol.org/content/188/2/781.full#ref-list-1

, 28 of which you can access for free at: cites 82 articlesThis article

        average*  

4 weeks from acceptance to publicationFast Publication! •    

Every submission reviewed by practicing scientistsNo Triage! •    

from submission to initial decisionRapid Reviews! 30 days* •    

Submit online. ?The JIWhy

Subscriptionhttp://jimmunol.org/subscription

is online at: The Journal of ImmunologyInformation about subscribing to

Permissionshttp://www.aai.org/About/Publications/JI/copyright.htmlSubmit copyright permission requests at:

Email Alertshttp://jimmunol.org/alertsReceive free email-alerts when new articles cite this article. Sign up at:

Print ISSN: 0022-1767 Online ISSN: 1550-6606. Immunologists, Inc. All rights reserved.Copyright © 2012 by The American Association of1451 Rockville Pike, Suite 650, Rockville, MD 20852The American Association of Immunologists, Inc.,

is published twice each month byThe Journal of Immunology

by guest on July 9, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

by guest on July 9, 2018

http://ww

w.jim

munol.org/

Dow

nloaded from

The Journal of Immunology

Innate Transcriptional Networks Activated in Bladder inResponse to Uropathogenic Escherichia coli Drive DiverseBiological Pathways and Rapid Synthesis of IL-10 for Defenseagainst Bacterial Urinary Tract Infection

Benjamin L. Duell,*,1 Alison J. Carey,*,1 Chee K. Tan,* Xiangqin Cui,†,‡ Richard I. Webb,x

Makrina Totsika,{ Mark A. Schembri,{ Petra Derrington,‖ Helen Irving-Rodgers,#

Andrew J. Brooks,** Allan W. Cripps,* Michael Crowley,†† and Glen C. Ulett*

Early transcriptional activation events that occur in bladder immediately following bacterial urinary tract infection (UTI) are not

well defined. In this study, we describe the whole bladder transcriptome of uropathogenic Escherichia coli (UPEC) cystitis in mice

using genome-wide expression profiling to define the transcriptome of innate immune activation stemming from UPEC coloni-

zation of the bladder. Bladder RNA from female C57BL/6 mice, analyzed using 1.0 ST-Affymetrix microarrays, revealed extensive

activation of diverse sets of innate immune response genes, including those that encode multiple IL-family members, receptors,

metabolic regulators, MAPK activators, and lymphocyte signaling molecules. These were among 1564 genes differentially regu-

lated at 2 h postinfection, highlighting a rapid and broad innate immune response to bladder colonization. Integrative systems-

level analyses using InnateDB (http://www.innatedb.com) bioinformatics and ingenuity pathway analysis identified multiple

distinct biological pathways in the bladder transcriptome with extensive involvement of lymphocyte signaling, cell cycle alter-

ations, cytoskeletal, and metabolic changes. A key regulator of IL activity identified in the transcriptome was IL-10, which was

analyzed functionally to reveal marked exacerbation of cystitis in IL-10–deficient mice. Studies of clinical UTI revealed signif-

icantly elevated urinary IL-10 in patients with UPEC cystitis, indicating a role for IL-10 in the innate response to human UTI. The

whole bladder transcriptome presented in this work provides new insight into the diversity of innate factors that determine UTI

on a genome-wide scale and will be valuable for further data mining. Identification of protective roles for other elements in the

transcriptome will provide critical new insight into the complex cascade of events that underpin UTI. The Journal of Immu-

nology, 2012, 188: 781–792.

Urinary tract infections (UTI) are among the most commoninfectious diseases of humans. Up to 40% of healthy

adult women experience at least one UTI episode in their

lifetime (1). Escherichia coli is the most common cause of UTI,

and the spectrum of UTI caused by uropathogenic E. coli (UPEC)

includes asymptomatic bacteriuria, cystitis, pyelonephritis, and

urosepsis. UPEC expresses multiple adhesins and virulence fac-

tors that provoke inflammation and enable bacterial colonization

of the bladder as the first step in UTI pathogenesis (2–5). UPEC

may also suppress innate immune responses (6–8).The immediate steps that form the host response to UPEC in the

bladder and mediate the pathogenesis of UPEC UTI in patients are

not well defined. Studies have shown that UPEC triggers inflam-mation that consists of immune mediators, including ILs (9–12),

although some of these responses, including, for example, IL-1a,

occur in a pathogen-specific manner (13). Roles for TNF, NO (14),

CXCR 2 (15), Tamm-Horsfall protein (16, 17), and CD44 (18)

have been shown in models of human UPEC UTI, although, in

most cases, the functions of these host factors at the cellular and

molecular levels remain unclear. UPEC induces host cell apoptosis

in cystitis patients (19–21) and macrophage inflammatory peptide

2 that is induced in response to neutrophils attracted to the bladder

to counteract colonization by the bacterium (6). Specific UPEC

adhesins, including type 1 and P fimbriae, trigger some of these

*School of Medical Sciences, Centre for Medicine and Oral Health, Griffith Univer-sity Gold Coast Campus, Queensland 4222, Australia; †Department of Biostatistics,University of Alabama at Birmingham, Birmingham, AL 35294; ‡Department ofGenetics, University of Alabama at Birmingham, Birmingham, AL 35294; xCentrefor Microscopy and Microanalysis, University of Queensland, Brisbane, Queensland4072, Australia; {School of Chemistry and Molecular Biosciences, University ofQueensland, Brisbane, Queensland 4072, Australia; ‖Pathology Queensland, GoldCoast Hospital, Southport, Queensland 4215, Australia; #School of Obstetrics andGynaecology, University of Adelaide, Adelaide, South Australia 5055, Australia;**Institute for Molecular Biosciences, University of Queensland, Brisbane, Queens-land 4072, Australia; and ††Heflin Center for Human Genetics, University of Ala-bama at Birmingham, Birmingham, AL 35294

1B.L.D. and A.J.C. contributed equally to this work.

Received for publication May 2, 2011. Accepted for publication November 15, 2011.

This work was supported by National Health and Medical Research Council ProjectGrant 569674, a Griffith University Research Infrastructure Fund grant, and a GoldCoast Hospital Collaborative grant.

The sequences presented in this article have been submitted to the Gene ExpressionOmnibus (http://www.ncbi.nlm.nih.gov/geo/) under accession numbers GSE26509and GSE33210.

Address correspondence and reprint requests to Dr. Glen C. Ulett, School of MedicalSciences, Centre for Medicine and Oral Health, Griffith University, Gold CoastCampus, QLD 4222, Australia. E-mail address: g.ulett@griffith.edu.au

The online version of this article contains supplemental material.

Abbreviations used in this article: CT, cycle threshold; GO, gene ontology; IHC,immunohistochemistry; IPA, ingenuity pathway analysis; KEGG, Kyoto Encyclope-dia of Genes and Genomes; qRT-PCR, quantitative RT-PCR; RT, room temperature;SEM, scanning electron microscope; UPEC, uropathogenic Escherichia coli; UTI,urinary tract infection; WT, wild-type.

Copyright� 2012 by TheAmerican Association of Immunologists, Inc. 0022-1767/12/$16.00

www.jimmunol.org/cgi/doi/10.4049/jimmunol.1101231

by guest on July 9, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

processes (22, 23). Host responses that arbitrate these very earlysteps in the pathogenesis of UPEC UTI are largely unknown, andthe overall contributions of innate immune elements beyond pre-viously identified mechanisms of attachment are not well defined.To better understand the nature of innate immunity to UPEC in

the bladder, we performed detailed genome-wide gene expressionmicroarray profiling using the murine model of UPEC UTI toestablish the global host transcriptome of mouse bladder to UPECcystitis. The overall results of the transcriptome were interrogatedusing several cutting-edge bioinformatics platforms and revealcollections of innate transcriptional networks not previously rec-ognized in bacterial UTI. Systems-level bioinformatic analyses ofthe transcriptional networks show that the early response to in-fection in the bladder is broad based and rapid and incorporatesdiverse biological pathways. Functional analysis of a prominentIL identified within the transcriptome, IL-10, which was sharedamong multiple biological networks as well as the most highlyactivated canonical pathway, unveiled a new appreciation of func-tion for this cytokine in the early control of acute bacterial UTI. Theinnate transcriptome of UPEC cystitis defined in this work willprompt renewed examination of early host determinants of acuteUTI aimed at uncovering novel control and preventative strategiesfor bacterial UTI.

Materials and MethodsMurine model of UPEC cystitis

The murine model of UTI utilizing female C57BL/6 mice (8–10 wk; TheJackson Laboratory [Bar Harbor, ME], Animal Resources Centre [CanningVale, WA]) based on transurethral inoculation is described elsewhere (24–27). IL-10–deficient B6.129P2-Il10tm1Cgn/J mice (002251) were purchasedwith appropriate control mice (000664, C57BL/6J) from The JacksonLaboratory. In addition to C57BL/6 mice, CBA mice are also used for UTIstudies (28, 29), and we performed microarray assays to compare strains.For infection, mice were anesthetized by brief inhalation exposure toisoflurane, and the periurethral area was sterilized by swabbing with 10%povidone-iodine, which was subsequently removed with PBS. Mice werecatheterized using a sterile Teflon catheter (0.28 mm internal diameter,0.61 mm outer diameter, and 25 mm length; Terumo, Somerset, NJ) byinserting the device directly into the bladder, and 40 ml challenge inoculumwas instilled transurethrally. The catheter was removed, and mice werereturned to their cages for recovery. Generally, the inoculum was retainedwithin the bladder for ,2 h, which was monitored by inclusion of 0.1%india ink in the bacterial suspension. Mice were infected with 109 CFUUPEC CFT073, and urine was collected at 2 h, 24 h, and 5 d postchallengefor colony counts on lysogeny broth agar (30). Bladders and kidneyscollected from euthanized mice at these time points were homogenized forcolony counts or processed for RNA collection, immunohistochemistry(IHC), or microscopy. For Ab depletion studies, female C57BL/6 micewere administered 300 mg rat anti-mouse IL-10-LE/AF–neutralizing Ab(JES5-2A5; Southern Biotechnology, Birmingham, AL) i.v. at 18 and 2 hprior to challenge.

Fluorescence and electron microscopy

mCherry fluorescent protein-tagged UPEC CFT073 was analyzed in tissuesusing an Olympus SZX16 stereo microscope fitted with a DP71 CCDcamera. Bladders were infected for 2 or 24 h, and UPEC was visualized insitu, which enabled resection of infected areas for subsequent processing forscanning electron microscope (SEM). Resected tissues were prepared byfixation in 3% glutaraldehyde in 0.1M cacodylate buffer (pH 7.4) and storedat 4˚C. After washing with fresh buffer, tissues were postfixed in 1% os-mium tetroxide, dehydrated through a graded ethanol series, and criticalpoint dried. Samples were mounted on SEM stubs and sputter coated withplatinum. Images were acquired with a JSM-6300F high resolution fieldemission SEM (JEOL, Tokyo, Japan) operated at 8 kV.

RNA isolation and microarrays

Total RNA was isolated from whole bladders of mice at 2 h postinfectionusing TRIzol (Life Technologies, Mulgrave, VIC, Australia), accordingto the manufacturer’s instructions. RNase-free DNase-treated RNA thatpassed Bioanalyzer 2100 (Agilent) analysis was quantified, and 100 ng wasamplified into double-stranded cDNA with random hexamers tagged with

a T7 promoter sequence. Gene 1.0 ST Mouse Microarrays (Affymetrix)were performed, according to the manufacturer’s instructions, in quintu-plicate for each treatment group using one microarray per bladder.

Quantitative RT-PCR

Quantitative RT-PCR (qRT-PCR) was carried out for groups of genesidentified as differentially regulated by microarray analysis. Amplificationof cDNA was performed using a GeneAmp 7700 System (Applied Bio-systems). Target genes were amplified using thermal cycling conditionspreviously described (13, 31). Primer sequences are listed in SupplementalTable I. GAPDH and b-actin were used as reference genes. Separatereactions were carried out to ensure that the efficiency of amplification ofthe reference gene was approximately equal to that of the target gene.Relative expression levels of target genes were determined by normalizingreaction cycle thresholds (CTs) to the housekeeper genes, GAPDH andb-actin. DCT values were used in the Equation 2.02[DCt] for calculation ofthe relative mRNA expression levels of target genes, as previously de-scribed using PCR efficiencies and mean crossing point deviations betweensamples and controls (13, 31).

ELISA and IL-10 IHC

Whole bladders from infected and control (PBS-challenged) mice at 2 hwere collected and analyzed for total levels of IL-10 in the tissues by ELISAand assessed by IHC. For ELISA, tissues were immersed in 70 ml proteaseinhibitor mixture, homogenized, and clarified at 12,000 3 g for 20 min at4˚C. Supernatants were stored at 280˚C until assay, which was performedusing quintuplicate samples in a commercial IL-10 ELISA (Pierce Endo-gen, Scoresby, VIC, Australia). For IHC, bladders were fixed in 10%buffered formalin, and sections (4 mm) cut from paraffin blocks wereaffixed to slides and air dried overnight at 37˚C. Sections were dewaxedand rehydrated, underwent heat retrieval using Antigen Retrieval solution(pH 9.0; Dako, Campbellfield, VIC, Australia), and transferred to TBS.Background Sniper (Biocare Medical, Concord, CA) was applied and thenrat IgG1 anti-mouse IL-10 Ab (JES5-2A5, 200 mg/ml, Santa Cruz Bio-technology, Santa Cruz, CA; confirmation assays used ab33471, SapphireBioscience, Waterloo, NSW, Australia). Following immersion in 1% H2O2/TBS, an anti-rat probe (Biocare Medical) was used, and sections weredeveloped using 3,39-diaminbenzidine (Pierce/Progen, Richlands, NSW,Australia). Sections were counterstained in Mayer’s hematoxylin, dehy-drated, cleared in xylene, and mounted. Controls included Ab isotypes andstaining of bladders from control (PBS-treated) mice. Serial sections werestained with H&E. For fluorescent imaging, sections were mounted influorescence medium (Dako). Images were acquired using an OlympusBX50 microscope, and, for whole bladder, an Olympus SZX16 fitted witha DP71 CCD camera, and CellF acquisition software.

Histotypic human cell coculture model

We established a histotypic human cell coculture model of whole bladderin vitro that would facilitate study of intercellular interactions (32). Human5637 bladder urothelial cells (ATCC HTB-9), U937 monocyte-derivedmacrophages (ATCC CRL-1593.2), MC116 B cells (ATCC CRL-1649),and Jurkat T cells (ATCC TIB-152; clone E6-1) were grown in RPMI 1640supplemented with 10% FCS and HEPES, as previously described (33–35). A total of 150,000 5637 cells, 15,000 U937 cells, and 1,500 MC116and Jurkat cells each was seeded into wells of a 96-well cell culture platefor infection. Equivalent cultures were seeded into wells of tissue culture-treated multiwell chamber slides (Lab-Tek II 155409; Nalge Nunc Inter-national, Rochester, NY) for confocal microscopy. UPEC (multiplicity ofinfection 10 and 100 CFU cell21) was added to cocultures and incubated at37˚C. At 2 and 5 h postinfection, supernatants were collected by centri-fuging 96-well plates at 500 3 g and frozen at 280˚C for subsequentIL-10 ELISA (88-7106-88; eBiosciences, San Diego, CA). For confocalmicroscopy, chamber slides were centrifuged (500 3 g for 10 min) at2 h, rinsed with PBS, and fixed using 4% paraformaldehyde for 45 min atroom temperature (RT). Cells were stained with Alexa Fluor 594-phalloidinand Hoechst 33258 (Molecular Probes, Mulgrave, VIC, Australia), andmounted in 0.2 n-propyl gallate. Slides were viewed using an OlympusFluoview FV1000-IX81 confocal microscope using version 1 applicationacquisition software.

mCherry-tagged UPEC strains

Cherry-tagged UPEC CFT073 was used to visualize the interactions be-tween UPEC and human cells in the histotypic coculture model. Theplasmid pmCherry (Clontech) was modified by introducing a tetracyclineresistance cassette into the EcoRI restriction site and then used to transformCFT073 and CFT073fimA. The fimA mutation in CFT073 was constructed

782 EARLY BLADDER TRANSCRIPTOME OF UPEC UTI AND IL-10

by guest on July 9, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

using l-Red–mediated homologous recombination, as previously de-scribed (36). The mCherry-tagged strains were then applied to the histo-typic coculture model of human bladder cells, as described above, formeasurement of IL-10 by ELISA. Experiments using the bacteria inmultiwell chamber slides were performed for confocal microscopy to studytype 1 fimbriae-mediated adherence in relation to UPEC-triggered IL-10synthesis in bladder.

Flow cytometry for identification of IL-10–secreting cells

Mixed human cell cocultures incorporating 5637 bladder urothelial cellsand U937monocytes were infected with UPEC, as above, and separated intononadherent (monocyte-enriched) and adherent populations (urothelialenriched) after 5 h. Nonadherent cells, inclusive of several PBS rinses, werecollected by centrifugation (500 3 g for 5 min); adherent cells weresubsequently collected by trypsinization. Cells were adjusted to 106 cellsper sample and used in FACS analysis using CD45 as a U937 monocytemarker, or RNA isolation and subsequent qRT-PCR for human IL-10 usingspecific primers (Supplemental Table I). For FACS analysis, cells wereresuspended in wash buffer (1% FBS in PBS), blocked (2% BSA for 1 h,RT), and centrifuged, and FcRs were blocked with 1% normal humanserum in wash buffer for 1 h at RT. Cells were centrifuged and resuspendedin wash buffer with mouse anti-human CD45 IgG conjugated to R-PE(NBP1-28512 [F10-89-4]; Novus Biologicals) at the concentration rec-ommended by the manufacturer and incubated for 30 min in the dark atRT. Cells were washed twice in wash buffer, fixed in 4% paraformaldehydeovernight at 4˚C, and stored at 4˚C in the dark until analysis, which wasperformed using a LSRII Fortessa flow cytometer (BD Biosciences, NorthRyde, NSW, Australia) the following day. Data were analyzed using theFACSDiva version 6 software package.

Generation of IL-10–targeting knockdown cells in U937monocytes

We made several stable U937 cell lines containing chromosomally inte-grated constructs from lentiviral vectors to knock down UPEC-inducedIL-10 respones and complement qRT-PCR analysis of CD45-positivecells. Lentiviruses were produced using Human Expression Arrest GIPZlentiviral short hairpin (sh)-microRNA vectors targeting IL-10 (cloneIDsV2LHS_111493, V2LHS_111488, V2LHS_226698) and a nontargetingclone (RHS4346) in conjunction with the TransLenti Viral GIPZ PackagingSystem (Thermo Scientific Open Biosystems), according to the manu-facturer’s protocol. Normal U937 cells harvested between passages 5 and 8were transduced in the presence of 8 mg/ml polybrene. Transduced cellswere selected with 0.35 mg/ml puromycin, and expression of TurboGFPwas confirmed by fluorescence microscopy. The newly contructed U937cell lines were designated U937-IL10-KD1, -KD3, and -KD4, and thenontargeting control line U937-IL10-NTC, and were used in culture com-bination, as described above.

Measurement of IL-10 in patients with UPEC UTI

Analysis of IL-10 in urine of human UTI was performed with adult patientstudy subjects (.18 y) encountered at the Gold Coast Hospital betweenAugust and October 2010. Patients underwent assessment for UTI becauseof symptoms indicating infection or as part of routine patient screening.All patients presenting with cystitis (dysuria, frequency, and/or urgencywith a concentration of a uropathogen in urine .1 3 105 CFU/l) wererecruited into the study. Urine samples were obtained as clean-catchvoided or catheterized samples and were frozen at 280˚C until assay.Control subjects were selected on the basis of not having a current UTI(i.e., no bacteriuria or symptoms indicating UTI). Urinary IL-10 wasmeasured using ELISA. IL-10 was also measured in plasma collected from57 patients diagnosed as having urosepsis due to UPEC, and these werecompared with IL-10 levels in plasma of 60 healthy control subjects.

Statistical analysis

Most of the analysis of the microarray data was in R.2.11 using Bio-conductor packages (http://www.bioconductor.org).

Data preprocessing

The rawmicroarray data were normalized using quantile normalization (37)and summarized using Robust Multi-array Average (38) as implemented inthe affy package in Bioconductor. Differentially expressed genes wereidentified by analyzing the data using the MAANOVA package (39) (http://www.bioconductor.org/packages/bioc/1.8/html/maanova.html). To identifysignificant changes in gene expression induced by UPEC, we used ashrinkage-based t test (40) together with permutation (41) to compare thetreatment group and the control group in each strain. A false discovery rate

(42) of 0.05 was used to control the multiple testing issue and generate thesignificant gene lists. We also generated heat maps of differential genesusing hiarchical-clustering algorithms implemented in the R packagegplots. Gene class testing was performed using the safe package (43) inBioconductor using default settings. Both Kyoto Encyclopedia of Genesand Genomes (KEGG) pathway and gene ontology (GO) used as geneclasses were analyzed. False discovery rate of 0.1 (44) was used as a cut-off. GoMiner and InnateDB were also used to analyze functional groups ofgenes activated in biological pathways, and overrepresentation analysisusing Over Representation Analysis was based on fold-change and p valuecutoffs of 2.0 and 0.01, respectively.

ELISA data

Unpaired Student t test was used with significance level set as a p value,0.05.

Tissue bacterial load data. Mann–Whitney U test was used to comparemean values of bacterial CFU/ml that were not normally distributed, withsignificance level set as a p value ,0.05.

Patient data. Pearson’s x2 test was used to compare the frequency of el-evated urinary IL-10 among patients and controls. Unpaired Student t testwas used to compare plasma levels of IL-10 among patients and controls,with significance level set as a p value ,0.05.

Ethics statement

Animals. This study was carried out in strict accordance with the nationalguidelines of the Australian National Health and Medical Research Coun-cil. The Institutional Animal Care and Use Committee of the Universityof Alabama at Birmingham and the Animal Ethics Committee of GriffithUniversity reviewed and approved all animal experimentation protocolsused in this study (permits: University of Alabama at Birmingham AnimalProject 080708186 and Griffith Approval MSC/14/08AEC). The Universityof Alabama at Birmingham institution has an Animal Welfare Assurance onfile with the Office of Laboratory Animal Welfare (Assurance A3255-01)and is registered as a Research Facility with the United States Department ofAgriculture. The animal care and use program is accredited by the Asso-ciation for Assessment and Accreditation of Laboratory Animal Care(American Association for the Accreditation of Laboratory Animal CareInternational). Transurethral infections were performed under temporaryisoflurane inhalation anesthesia.

Human subjects. This study was performed in accordance with the ethicalstandards of the Gold Coast Hospital, Queensland Health, Griffith Uni-versity, University of Queensland, Princess Alexandra Hospital, and Hel-sinki Declaration. The study was approved, and the need for informedconsent was waived by the institutional review boards of Queensland Healthand Griffith University (MSC/18/10/HREC) and the board of the PrincessAlexandra Hospital (research protocol 2008/264).

ResultsWhole bladder transcriptome of UPEC cystitis

Microarray analysis of early global gene expression in bladdertriggered by UPEC after transurethral challenge of C57BL/6 miceidentified 1,564 genes with significantly altered expression, asfollows: 914 (58.44%) were upregulated and 650 (41.56%) down-regulated. These 1,564 genes represent 3.8 and 2.7% of the entire24,000 gene set significantly upregulated and downregulated,respectively. The complete significant gene list is shown as a heatmap and volcano plot in Supplemental Fig. 1 and listed accordingto statistical significance values in Supplemental Table II. Raw dataare deposited under accession number GSE26509 in the Gene Ex-pression Omnibus (http://www.ncbi.nlm.nih.gov/geo/). The list ofgenes encompassed a diverse set of regulators of metabolism, cellcycle, growth and survival, adaptive and innate immune responses,and genes central to immune regulation, such as NF-kB and ILs.During our analysis of transcriptional activation due to UPEC inthe bladder, we noted 78 significant genes that encode IL familymembers and their receptors, including (fold change) the follow-ing: IL-1b (46.9), IL-6 (46.2), IL-1a (13.9), IL-1r2 (10.28), IL-10ra (3.00), IL-10 (5.05), IL-18rap (3.68), and IL-23 (2.2). Otherhighly expressed genes were the chemokines Cxcl-10 (38.26),Cxcl-1 (22.15), Cxcl-2 (66.72), Cxcl-16 (3.19), Cxcl-3 (9.07),Cxcl-5 (3.46), and Cxcl-9 (2.67).

The Journal of Immunology 783

by guest on July 9, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

We next measured the expression of several IL and chemokinetarget genes using qRT-PCR to validate array datasets. This showedlevels of gene expression of specific target genes that were highlyconsistent with microarray data. For example, IL-1b (195.49-foldincrease), IL-1a (134.57-fold increase), IL-10 (23.67-fold in-crease), IL-6 (132.87-fold increase), and TNF-a (11.23-fold in-crease) were highly expressed according to qRT-PCR (Table I),validating array analysis. For the majority of the genes tested byqRT-PCR, including IL-1b, IL-1a, IL-10, IL-6, TNF-a, CCL2,and Cxcl-9, the responses peaked at 2 h (Table I). Various genesthat were identified in microarray assays as unchanged in ex-pression at 2 h, including Msr-1, IL-10rb, and the housekeepergenes b-actin and GAPDH, also showed no change at this timepoint according to qRT-PCR. Some of these, however, includingMsr-1 and IL-10rb, showed significantly increased expression af-ter 24 h, indicating a slower response to infection.

Transcriptional networks in UPEC cystitis

Gene class testing using KEGG identified 20 significant biologi-cally relevant pathway interactions in C57BL/6 mice (Table II).These included several with significant associations with infec-tion, including metabolism of cofactors (00670), phosphati-dylinositol signaling (04070), TLR (04620), and TCR signaling(04660). Next, we analyzed expression data using GO, whichidentified 223 biologically relevant pathways triggered by UPECUTI (Supplemental Table III). Among the most significant werethose for MAPK activation (187), the response to bacteria-associated molecules (2237), adaptive immune responses (2250),regulation of B cell-mediated immunity (2712), chemotaxis(6935), and leukocyte adhesion (7159). A summary of gene classtesting results using KEGG and GO is illustrated in Fig. 1A.GoMiner analysis also identified extensive gene activation among

pathways for immune defense against bacteria, cytokine inter-actions, and B lymphocyte activation (Fig. 1B).We next sought to examine the microarray dataset derived from

C57BL/6 mice using an integrative approach that incorporatesknown biological interactions from prior studies. InnateDB bio-informatics (45) revealed a suite of 771 additional unique andsignificant biological pathways from REACTOME and IntegratingNetwork Objects with Hierarchies databases (data not shown).Among these there were 68 that were overrepresented accordingto Over Representation Analysis. These included the TLR4 cas-cade (477830), biological oxidations (477446), and hemostasis(477598) (Supplemental Table III). Many of these were confirmedTable I. qRT-PCR of select gene targets identified by microarrays as

significant in the bladder transcriptome triggered by UPEC at 2 hfollowing infection

Gene Infection Period (h) ΔFold UPEC p Value

Identified by Arrays as Significant

IL-1a 2 134.57 0.00124 6.03 0.001

IL-1b 2 195.49 0.00124 36.92 0.001

IL-6 2 132.87 0.000124 10.56 0.001

IL-10 2 23.67 0.00124 2.97 0.001

IL-10ra 2 1.74 0.001524 3.26 0.001

TNF-a 2 11.23 0.00124 4.90 0.001

CCL2 2 27.89 0.00124 4.01 0.001

CXCL5 2 2.32 0.002524 5.37 0.001

CXCL9 2 5.08 0.00124 4.57 0.027

Identified by Arrays as Insignificant

Msr-1 2 1.24 0.3324 3.3 0.001

INFgr1 2 1.033 0.9224 1.96 0.0075

IL-10rb 2 0.941 0.524 1.33 0.029

b-actin 2 0.815 0.11424 0.937 0.128

GAPDH 2 21.18 0.524 1.13 0.66

Table II. Innate transcriptional networks triggered in the bladder byUPEC at 2 h postinfection as identified using KEGG

KEGG No. Pathway Name p Value

00670 Metabolism of cofactors and vitamins 0.01204070 Phosphatidylinositol signaling system 0.01204620 TLR signaling 0.01204660 TCR signaling 0.01204710 Circadian rhythm 0.01205210 Colorectal cancer 0.01205212 Pancreatic cancer 0.01204630 JAK–STAT signaling 0.01705217 Basal cell carcinoma 0.01904060 Cytokine–cytokine receptor interaction 0.02403450 Nonhomologous end joining 0.02705340 Primary immunodeficiency 0.02704210 Apoptosis 0.02800760 Nicotinate and nicotinamide metabolism 0.03304662 BCR signaling 0.03300510 N-glycan biosynthesis 0.03500532 Chondroitin sulfate biosynthesis 0.03504320 Dorso-ventral axis formation 0.03704520 Adherens junction 0.03805220 Chronic myeloid leukemia 0.039

FIGURE 1. Transcriptional networks of gene activation in UPEC-

infected bladder. Diverse gene networks activated in bladder were identi-

fied using KEGG and GO (A), with Venn diagrams summarizing numbers

of pathways altered by infection according to each tool. GoMiner was used

to identify groups of infection/immune-related genes in pathways in which

IL-10 (*) was identified as a central component to several networks (B).

784 EARLY BLADDER TRANSCRIPTOME OF UPEC UTI AND IL-10

by guest on July 9, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

using ingenuity pathway analysis (IPA), the overall results ofwhich are shown in Supplemental Table III.CBAmice are occasionally used as an alternate strain to C57BL/

6 mice, and we analyzed gene expression in these mice to delineateconserved responses, the results of which are shown as a completelist of significant genes in Supplemental Table II. Raw data for CBAmice are deposited under accession number GSE33210 on GEO.Overall, many transcriptional network responses observed inC57BL/6 and CBA mice were shared between the mouse strains.For example, signaling pathways representing JAK–STAT (04630),T (04660) and B (04662) cell receptors, TLRs (04620), phos-phatidylinositol (04070), and apoptosis (04210), identified usingKEGG, were common, with highly significant p values of acti-vation (Supplemental Table III). Collections of distinct responseswere also noted, and CBA mice exhibited a larger number ofactivated genes. Direct comparisons illustrating the overall re-sponse of CBA mice and conserved, CBA-unique, and C57BL/6-unique gene responses are shown in Supplemental Fig. 1.In the course of our bioinformatics analysis, we noted many IL

responses that were overrepresented in activated biological path-ways in response to UPEC. This included, for example, pathwaysfor immune defense against bacteria, cytokine activity, and B cellsignaling, as identified using GoMiner (Fig. 1B). Unlike IL-1b andIL-6 (46–50), many of the ILs and IL receptors in these responses(IL-2ra/b, -8ra, -10, -11, -15ra, -17c, -18rap, and -23a) have notpreviously been associated with UTI. Of note was IL-10, whichwas present in all three GoMiner pathways related to infection(Fig. 1B) and was shared in the activated T lymphocyte activationpathway, as well as eight KEGG (8107, 2794, 2791, 692, 2816,651, 2813, 604) and eight Integrating Network Objects with Hi-erarchies (187, 195, 11, 51, 185, 393, 63, 121) pathways (all p =0.005). IPA also revealed IL-10 signaling as the top hit in ca-nonical pathway recognition and revealed the multiple genes re-lated to IL-10 activation present within the C57BL/6 trans-criptome (Supplemental Table III). As a result of these findings,we chose to study the IL-10 response further.

Induction of bladder IL-10 in response to UPEC

IL-10 was upregulated 5.051-fold in C57BL/6 mice at 2 h(p = 0.005), according to microarrays, and 23.67- and 2.97-fold(p = 0.001) at 2 and 24 h, respectively, according to qRT-PCR.The cognate receptor, IL-10ra, was also upregulated and waspresent in several biological pathways identified by InnateDB,including the cytokine–cytokine receptor pathway (Fig. 2). BothIL-10ra and IL-10rb were also significantly upregulated in CBAmice according to microarrays (Supplemental Table II). To betterunderstand this response, we measured total IL-10 protein inbladders of C57BL/6 mice. Compared with mice that receivedonly PBS, there was a significant increase in IL-10 protein inbladders of infected mice (492 6 33 versus 152 6 25 pg/ml, p ,0.0001) and a parallel increase in kidneys (Fig. 3). Urinary IL-10was also measured, but the data were not significant (data notshown).IHC was then performed on UPEC-colonized uroepithelium

(Fig. 4A–C) and showed IL-10–secreting cells lining the luminalsurface of the bladder in situ (Fig. 4D). IL-10 was not detected insections stained with an isotype-matched control Ab (Fig. 4E) orin PBS-treated noninfected mice (data not shown). The presenceof UPEC within these tissues expressing IL-10 was studied usingmCherry-tagged CFT073. IL-10 expression in colonized uroepi-thelia was not consistently associated with the presence of UPEC(Fig. 4F–H). Thus, IL-10 expression forms part of the innate re-sponse to UPEC bladder colonization, but is not consistentlycolocalized within areas of infection.

Early IL-10 synthesis correlates with UPEC load

To investigate the relationship between IL-10 and bacterial burden,we analyzed the distribution of UPEC on bladder uroepitheliumusing mCherry-tagged UPEC. Binding of UPEC occurred in focalareas (Fig. 5A, 5B, arrows), and early IL-10 expression correlatedwith the highest numbers of bacteria at 2 h (Fig. 5C, 5D). Lowerexpression of IL-10 at 24 h (qRT-PCR data, Table I) correlatedwith fewer UPEC at this time point, as well as urothelial cellinvasion and destruction as observed by SEM (Fig. 5E, 5F). Lowerbacterial burdens were observed at 5 d (Fig. 5C). Extrusions fromurothelial cells at 2 h (Fig. 5D) resembled intracellular bacterialcommunities (51), and elongated UPEC cells at 24 h (Fig. 5E,5F) mirrored unique filamentous UPEC, as noted elsewhere (52).Thus, IL-10 synthesis in bladder peaks early with high UPECloads, which are subsequently reduced, and correlate with lowerIL-10 expression and urothelial cell destruction.

Early IL-10 contributes to control of UPEC UTI

We next analyzed whether early IL-10 in bladder has any functionalrole in UTI by comparing disease in immunodeficient gene-knockout mice and wild-type (WT) mice, as previously re-ported, for other host factors (53). Infection of IL-102/2 mice andassessment of the impact of a lack of IL-10 on disease comparedwith WT mice at 24 h postinfection showed that IL-102/2 micehad significantly more UPEC in the bladders, urine, and kidneyscompared with WT mice (Fig. 6A–C; p = 0.013; p = 0.001; p =0.007). Bacterial loads in IL-102/2 mice were 10-fold greater inbladder and 30-fold greater in urine and kidneys at this time pointcompared with WT mice. Ab depletion studies using an anti–IL-10–neutralizing Ab in WT C57BL/6 mice also demonstratedhigher bacterial burdens in urine and bladders of mice followingpretreatment with Ab and subsequent infection with UPEC (Fig.6D, 6E). Mice that received anti–IL-10 Ab prior to UPEC chal-lenge had 4.50 6 0.20 CFU/ml in urine 24 h postchallengecompared with 3.89 6 0.19 CFU/ml in controls, a statisticallysignificant difference. Bladder bacterial counts in anti–IL-10 Ab-treated WT mice were also higher than controls (5.14 6 0.26CFU/ml versus 4.84 6 0.24 CFU/ml), but these differences werenot statistically significant based on group sizes of 12 mice. To-gether, these data show that early bladder IL-10 is protective in theresponse to UPEC and contributes to bacterial control in UTI,particularly in terms of bacteriuria.

IL-10 is produced in a coculture model of human bladder

We next sought to investigate whether IL-10 is produced by humancells in response to UPEC using a novel histotypic four-cell co-culture model of bladder. In cocultures of urothelial cells,monocytes, T cells, and B cells (ratio 88:10:1:1), there was nodetectable increase in IL-10 in response to UPEC after 2 h (data notshown). However, IL-10 was produced in UPEC-infected culturesafter 5 h, with 3-fold more IL-10 secreted by infected cellscompared with controls (p, 0.001; Fig. 7A). We also analyzed theeffect of the number of T cells on IL-10 production becauseT cells often amplify IL-10 production (54). In cocultures thatcontained more T cells (79:10:10:1), the IL-10 response at 5 h wasslightly higher compared with cultures containing fewer T cells,but these differences were not significant (data not shown). Type 1fimbriae are the major adhesin of UPEC responsible for binding tourothelial cells, and we hypothesized that type 1 fimbriae adher-ence might trigger IL-10 production. When we tested a mCherry-tagged type 1 fimbriae-deficient mutant of UPEC, CFT073fimA,the mutant stimulated levels of IL-10 slightly higher than the WTUPEC (Fig. 7B). Confocal microscopy confirmed less binding ofthe mutant to bladder cells compared with the WT bacteria (Fig.

The Journal of Immunology 785

by guest on July 9, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

7C, 7D). Thus, type 1 fimbriae-mediated attachment to humanbladder cells is not required for UPEC-triggered IL-10 production.

Identification of the IL-10–producing cells

To determine the cell type that produces IL-10 in response toUPEC in the coculture model of human bladder, we used CD45staining as a marker of monocytes and analyzed cocultures after5 h of infection (Fig. 8A) or separated populations of nonadherent(enriched for U937 monocytes) and adherent cells (enriched forurothelial cells). Utilizing qRT-PCR, we observed a 4.1-fold in-crease in IL-10 mRNA in monocyte-enriched (CD45-positive) cellpopulations compared with controls (Fig. 8B). In contrast, IL-10mRNA levels in urothelial-enriched (CD45-negative) cells re-mained unchanged after 5 h of infection (Fig. 8C).We then measured IL-10 protein in cocultures that used various

IL-10–targeting knockdown versions of U937 cells to confirm datafrom the qRT-PCR assays above. In infected cocultures that usedany of three IL-10 knockdown U937 cells (targeting exons 1, 3,and 4), the IL-10 response was abolished compared with controlcocultures that used either normal U937 cells, or a nontargetingcontrol U937 line that was transduced with a nontargeting controlvector (Fig. 9). In a fourth IL-10–targeting U937 cell line, how-ever, the response was not changed (data not shown). Together,

these data confirm that the IL-10 produced in the coculture modelof human bladder in response to UPEC is derived mostly fromU937 monocytes in this model.

IL-10 is produced in human UTI

Measurement of IL-10 in patients with UPEC cystitis by ELISAshowed significantly higher urinary IL-10 levels (.10 pg/ml)among patients compared with control individuals without UTI(Fig. 10A). Low baseline levels of urinary IL-10 (3–5 pg/ml) weremore common among controls compared with UTI patients, butsignificantly more patients exhibited high levels of urinary IL-10(.40 pg/ml) compared with controls (Fig. 10A). We also mea-sured IL-10 in plasma of 57 patients with UPEC urosepsis and 61healthy subjects without UTI to determine whether the urinarytract might be influenced by systemic IL-10 in urosepsis. Patientswith urosepsis had significantly higher levels of IL-10 in plasmacompared with controls (Fig. 10B).

DiscussionThe very early inflammatory signaling events triggered in thebladder immediately following colonization with UPEC define thecascade of events that culminate in UTI disease or resolution ofinfection. Recently, innate resistance to UPEC UTI was demon-

FIGURE 2. Cytokine signaling pathways in UPEC UTI. InnateDB analysis of gene expression networks triggered by UPEC in bladder identified the

cytokine–cytokine receptor pathway as highly significant and uncovered broad activation of genes encoding various ILs and their receptors (red circles,

inset). Included among the the most activated cytokine genes were IL-10 and its receptor IL-10ra (boxed).

786 EARLY BLADDER TRANSCRIPTOME OF UPEC UTI AND IL-10

by guest on July 9, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

strated in the context of pathogen-specific signaling in murinepyelonephritis (55), and the danger-associated molecular patternS100A8/A9 was shown to be unnecessary for effective host de-fense against UTI (56). The overall innate response of bladder inUPEC cystitis in the murine UTI model, however, remains largelyuncharacterized. In this study, we provide a detailed genome-widemicroarray analysis of whole bladder to UPEC UTI in mice usinglarge group sizes and a nonpooled sampling approach for in-creased statistical power to offer a revealing new perspective onthe innate response to infection. Our approach has uncovered abroad range of host factors not previously known to regulate

innate defense to bacterial UTI. Several genes identified in thisstudy are consistent with prior reports that profiled mouse bladder

previously and mapped responses in foci of uroepithelium in-

fected with intracellular UPEC pods, which displayed localized re-

sponses (50, 57). The whole tissue approach we have used gives

broader definition to the total bladder transcriptome of UTI. This

provides crucial new insight into the dynamic interactions of in-

fected cells as well as adjacent bystander cells, the inflammatory

infiltrate associated with UTI (46), and the local vasculature that

contributes to inflammation in the bladder after bacterial coloni-

zation (20, 47, 58). Notably, many of the responses identified in

the whole bladder transcriptome presented in this work are con-

sistent with those observed in localized responses to intracellular

bacterial communities in bladder urothelial cells, as previously re-

ported (57). Of particular note were localized urothelial responses

to intracellular bacterial communities such as those directed at

opposing UPEC salvage of host cell iron and facilitation of glu-

cose import and epithelial structural integrity (57). Many of these

responses are consistent with the genes, such as lipocalin 2 (Lcn2)

and hexokinase 2 (Hk2), identified as highly upregulated in the

whole bladder transcriptome of UPEC cystitis in this study. To-

gether, these findings emphasize the importance of iron scaveng-

ing in the lifestyle of E. coli in the bladder, as recently described

(59–61), and, from the hosts’ perspective, the key role of iron se-

questration, sugar metabolism, and cell integrity in the urinary

tract in response to bacterial uropathogens.The whole bladder transcriptome of UPEC cystitis reveals

a complex network of interactions that encompass a broad func-

tional variety of genes involved in metabolism, cell cycle, growth,

adaptive and innate immune responses, cofactors, and signaling.

Many of the identified biological pathways have not previously

been associated with UTI. Leukocyte signaling triggered in the

bladder appears to reflect a collection of pathways linking TLR,

TCR, cytokines, JAK–STAT signaling, and BCR signaling. That

FIGURE 4. In situ detection of IL-10 in bladder. UPEC colonization of uroepithelium was assessed by H&E (A, B; boxed areas show the same region in

one section at different magnifications) to demonstrate UPEC binding (inset), and fluorescence microscopy was used to visualize mCherry-tagged UPEC

bound to tissue (C), with IHC used to detect IL-10 expression. C, Different regions of the same section are shown. The boxed sections indicate equivalent

regions within serial sections of the same tissue. IL-10–secreting cells are shown in D on the luminal surface of colonized uroepithelium (arrowheads), but

not in controls (E). Significant staining was absent in sections stained with an isotype-matched control Ab. Colocalization of UPEC and IL-10 was detected

in some areas (G, H), but not in others that were not colonized (H, upper arrows). Boxed areas of F and G show same region in different serial sections.

Scale bars, 750 mm (A); 20–55 mm (B–F); 185 mM (G, panel width); 130 mm (H, panel width). Boxed areas, 160 mm (A); 19 mm (B).

FIGURE 3. Protein levels of IL-10 in response to UPEC. Concentrations

of IL-10 were measured by ELISA in bladder (A) and kidney (B) homo-

genates (n = 9 per group, combined from two separate experiments) 2 h

after challenge and demonstrate rapid production of IL-10 protein in

bladder in response to UPEC infection.

The Journal of Immunology 787

by guest on July 9, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

many of these signaling pathways were activated by UPEC not onlyin the universal C57BL/6 mouse strain, but also the alternate CBAstrain, demonstrates a high degree of conservation in many of the

networks related to leukocyte activation and lymphocyte signalingduring acute UTI. These observations underscore the prominenceof these signaling mechanisms in the innate response to UPEC in

FIGURE 5. IL-10 expression correlates with bladder

UPEC load. Synthesis of early IL-10 correlates with

focal UPEC adhesion to bladder uroepithelium at 2 h,

as shown in whole tissue using mCherry-tagged UPEC

(A, B) and higher bacterial loads at 2 h compared with

later time points (C; n = 14–18 per time point, com-

bined from three separate experiments). Early coloni-

zation reflects UPEC bound in focal areas of adhesion

around bulbous extrusions (D), which is followed by

decreasing bacterial loads over time and urothelial cell

destruction (E, F). Scale bars, 800 mm (A); 160 mm

(B); 10 mm (F).

FIGURE 6. Early IL-10 protects against UPEC

UTI. Bladders, urine, and kidneys of C57BL/6 mice

and age-matched IL-10–deficient mice (B6.129P2-

Il10tm1Cgn/J) challenged with UPEC were cultured after

1 d of infection (n = 12 per group, combined from two

separate experiments) to compare tissue bacterial

loads. Higher UPEC numbers detected in IL-10–defi-

cient mice compared with control mice demonstrate

a protective role of early bladder IL-10 expression in

cystitis (A), bacteriuria (B), and pyelonephritis (C).

C57BL/6 mice treated with anti–IL-10–neutralizing

Ab also exhibited higher mean UPEC burdens in

bladders (D) and urine (E) compared with controls,

although differences in bladders were not statistically

significant. Mann–Whitney U test was used for com-

parisons, with significance set at p , 0.05.

788 EARLY BLADDER TRANSCRIPTOME OF UPEC UTI AND IL-10

by guest on July 9, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

the bladder. The short infection window used in our study high-lights the integral role for early lymphocyte signaling immediatelyfollowing UPEC infection. Prior studies with urothelial cells (58,62, 63) and animals (50) uncovered several genes triggered by

UPEC described in this work, including MIP-2, IL-1b, IL-6 (4, 13,46, 50), and CXCR1 (58). Interestingly, our data reveal that theinduction of these mediators, which occurs amid extensive inflam-matory responses, takes place with concomitant UPEC-triggeredtranscriptomic activation of key immunosuppressive genes. Theseinclude a suppressor of cytokine signaling socs3 and rapid induc-tion of the gene for the master regulator of innate immunity, IL-10.We suggest that this induction of suppression-related responsesmay represent a proactive attempt by the host to moderate exces-sive inflammation, which could damage tissue locally and causeimmunopathology in the bladder. Thus, the early bladder tran-scriptome mounted in response to UPEC appears to reflect a bal-ance of host defense strategies designed to drive rapid and robustimmune activation, but also moderate inflammatory responses bystimulating the actions of immune modulators.A notable element within the UPEC cystitis transcriptome was

broad involvement of multiple IL family members and their re-ceptors, including IL-1, IL-2, IL-6, IL-8ra, IL-10, IL-11, IL-17c/ra,IL-18rap, and IL-23a. Considering the role of IL-10 as a masterregulator of innate immunity, we focused on IL-10 in functionaland clinical studies in patients with UPEC cystitis. IL-10 works atthe crossroads of Th1-centric inflammation (54) and immuno-suppression (64). Smith et al. (65) described urinary IL-10 in renaltransplant patients with transplant rejection; however, the func-tion of IL-10 in the infected human urinary tract is hitherto un-known. In our study, patients with UPEC cystitis secreted signif-icant urinary IL-10 and produced high levels of IL-10 systemicallyduring UPEC urosepsis. Whether the systemic IL-10 present inpatients with UPEC urosepsis has coordinated effects in immuneregulation, as described in other clinical conditions, is unknown(66). However, our studies in mice point to the fact that IL-10produced in the bladder plays a functional role early in UTI dis-ease pathogenesis. Infection of IL-102/2 mice showed an un-expected protective role for early IL-10 in UPEC UTI because thesemice had more UPEC and more severe bacteriuria. Bacteriuriawas also more severe in mice undergoing Ab depletion using ananti–IL-10–neutralizing Ab. These findings highlight a functionalimpact of early IL-10 in UPEC UTI and suggest that future in-vestigation of novel anti-inflammatory treatments in the mouseUTI model may be worth investigating as a means of under-standing the inflammation-immunosuppression nexus in acutebacterial UTI. In addition, several IL-102/2 mice also progressedto pyelonephritis, which failed to develop in control mice, sug-gesting additional unknown effects in the upper urinary tract.The apparent protective role for early IL-10 in UPEC UTI in

mice may reflect a moderation of pathological inflammatory re-sponses shown to be central in many of the active biologicalpathways in the whole bladder transcriptome. As a comparison, forexample, in protozoan and helminthic infections, IL-10 aids indisease control by reducing pathological sequelae and developmentof lesions in acute infections (67). It is noteworthy that IL-10 alsohas well-defined roles in mediating pathogenesis in other bacte-rial infections due to broad immunosuppressive effects; IL-10 isknown to contribute to persistent infections such as tuberculosis(67, 68). These contradictory effects of IL-10 in different diseasestates are explained by the well-known pluripotent nature of IL-10and the host requirement for tightly controlled inflammatoryresponses to defend against a diversity of microbial pathogenswhile concurrently minimizing host damage.Our in vitro analysis of the human IL-10 response to UPEC using

a novel coculture model of bladder incorporating urothelial cells,monocytes, and lymphocytes, which comprise a major element ofthe inflammatory infiltrate in the bladder response to E. coli (69),identified the source of the IL-10 as monocytes. The urothelial cell

FIGURE 7. UPEC-triggered IL-10 synthesis in an in vitro model of

human bladder. UPEC triggers IL-10 synthesis in a histotypic coculture

model of human bladder (5637 urothelial cells, monocytes, T cells, and B

cells [ratio 88:10:1:1]) (A). UPEC-induced IL-10 occurs independently of

type 1 fimbriae attachment because levels of IL-10 triggered by infection

(B) were similar in cocultures infected with either mCherry-tagged WT

UPEC (C) or CFT073fimA despite impaired binding of the mutant (D).

Data represent mean 6 SD of quadruplicate samples from three separate

experiments. Unpaired Student t test was used with significance at p ,0.05. Red, mCherry-tagged UPEC; green, Alexa Fluor 488 phalloidin for

F-actin; blue, Hoechst 33258 for nuclei. Scale bars, 10 mm.

FIGURE 8. UPEC-triggered IL-10 in a mixed coculture model of human

bladder is monocyte derived. In cocultures of U937 monocytes and 5637

urothelial cells infected with UPEC for 5 h, then analyzed for CD45 by

FACS (A) (percentage of CD45-positive cells in inset), isolated cell pop-

ulations enriched for nonadherent monocytes (B) exhibited a 4.1-fold in-

crease in IL-10 mRNA according to qRT-PCR (normalized to noninfected

control cocultures). IL-10 mRNA in adherent cell populations, composed

predominantly of urothelial cells, was unchanged after 5 h (C). GAPDH

served as a reference gene.

The Journal of Immunology 789

by guest on July 9, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

coculture system used in this work has inherent limitations, suchas the exclusion of polymorphonuclear leukocytes, but epithelialcell types have been cocultured with lymphocytes and monocytes

previously (70, 71) and have proven useful for assessing IL-10effects (72). The advantages of coculture models incorporating

epithelial cells for the study of host–pathogen interactions are

reviewed elsewhere (32). Epithelial cells in mucosal tissues are

anatomically positioned in close proximity to various subepithe-

lial cell types, including lymphocytes (73), which can contribute

to cell signaling function via paracrine pathways (74). Signaling

between cells may aid these responses because lymphocytes pro-

duce IL-10 in response to various bacteria such as Mycobacteria

spp. (75, 76) and Helicobacter spp. (77), and intercellular inter-

actions can be essential to drive antibacterial responses (78). a-b

T cell activation, mononuclear cell and lymphocyte proliferation,

and receptor signaling, as well as leukocyte cell-cell adhesion and

JAK–STAT signaling for cytokine metabolic processes, were

identified as major conserved mechanisms within the bladder

transcriptome of UPEC UTI underscoring extensive cellular inter-

actions. The precise mechanisms of UPEC-induced IL-10 in this

human cell coculture model, however, and the identity of the

IL-10–producing cells in situ in the murine model of UPEC UTI

will require further study.Type 1 fimbriae of UPEC induce secretion of various cytokines

in the mouse UTI model (48). Surprisingly, IL-10 synthesis in

human cell in vitro occurred independently of fimbriae-mediated

binding in this study. This indicates that the presence of neither

type 1 fimbriae nor bacterial attachment is required for IL-10

production in the mixed cell coculture model of human bladder.

Other components of bacteria that may trigger IL-10 include LPS

(79) and species-specific molecules such as M protein of Strep-

tococcus pyogenes, which induces IL-10 via Tr1 cells (80). LPS

triggers the production of IL-10 in noninfectious experimental

models that incorporate dendritic cells and macrophage stimula-

tion, for example (81, 82). LPS also activates key innate immune

mechanisms, including TLR-4, CD14, and NF-kB (83), which

were shown to be activated by UPEC within the KEGG TLR

signaling pathway and the strongly induced IPA network for IL-10

signaling. In addition to UPEC-triggered synthesis of IL-10, the

ligand-binding chain of the IL-10R, IL-10ra, was also triggered at

the transcriptional level. Bladder cells are not known to express

IL-10R, and, to our knowledge, this is the first report of IL-10ra

responses in bladder tissue. Thus, this may represent an additional

FIGURE 9. Knockdown of UPEC-induced IL-10 in U937 monocytes in a dual coculture model of human bladder. sh-microRNA–based targeting of

IL-10 for knockdown in U937 monocytes confirmed that UPEC-induced IL-10 in cocultures was derived from monocytes. The IL-10 response in UPEC-

infected (UPEC multiplicity of infection 10) cocultures of 5637 urothelial cells in combination with various stable U937 monocyte cell lines harboring

contructs targeting exons 1, 3, and 4 of IL-10 or a nontargeting control (NTC) is shown and compared with the response of cocultures that incorporated

normal U937 cells. IL-10 responses in U937 IL-10KD cocultures were not significantly different in any of three KD cell lines compared with noninfected

control cultures containing normal U937 cells. An independent-samples t test was used to compare replicates (n = 12) of 5637–U937 control cocultures

(containing normal U937 cells) against infected cocultures with U937 shRNA KD cell lines.

FIGURE 10. IL-10 forms part of the human immune response to UPEC.

In patients with UPEC UTI elevated urinary IL-10 (.10 pg/ml) was more

common in individuals with UTI compared with controls (A) (n = 10/154;

6.5% versus n = 3/263; 1.1%; p = 0.006, Pearson’s x2 test). Plasma IL-10

was also elevated in patients with UPEC urosepsis (n = 57) compared with

controls (n = 60) (B) (p = 0.001, unpaired Student t test, independent

samples) illustrating IL-10 in the human immune response to UPEC.

790 EARLY BLADDER TRANSCRIPTOME OF UPEC UTI AND IL-10

by guest on July 9, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

mechanism of immune regulation at the bladder-bacterial interfacegiven the key role of IL-10R–dependent signaling in innate im-munity (84). IL-10R is present on the surface of a variety of otherhuman lymphoid and myeloid cells (85) and is upregulated duringinfection-induced inflammatory conditions such as sepsis (86).Moreover, human epithelial cells lining the colonic mucosa ex-press IL-10 to control inflammation and maintain mucosal ho-meostasis (66). It is important to note, however, that our findingson IL-10R transcripts were not extended to detection of the re-ceptor protein, which is a limitation of the current work. The rolefor IL-10R in addition to IL-10 in maintaining urothelial integrityis an appealing area for further investigation.The UPEC UTI whole bladder transcriptome described in this

study provides a wealth of new information for detailed study andwill be an invaluable tool for further data mining and biologicalpathway exploration. Further molecular analysis of the role ofIL-10 in the early transcriptome of UPEC UTI and investigation ofthe multitude of other signaling events defined to our knowledge forthe first time in this study will provide opportunity for improvedunderstanding of the innate host responses to E. coli in the contextof these important human infections.

AcknowledgmentsWe thank Prof. John Gerrard for facilitation of clinical aspects of this

study, Janice King and Yvette Hale for excellent technical assistance, and

Dr. Michael Hanzal-Bayer at the Facility for Life Science Automation at In-

stitute for Molecular Biosciences, University of Queensland, for performing

the lentivirus transductions. We thank Dr. Joan Faoagali and other members

of the microbiology laboratory at the Princess Alexandra Hospital for help

with the collection of patient plasma samples. We also thank Flora Gathof

for administrative assistance and Prof. David Briles for helpful discussions.

DisclosuresThe authors have no financial conflicts of interest.

References1. Foxman, B. 2002. Epidemiology of urinary tract infections: incidence, mor-

bidity, and economic costs. Am. J. Med. 113(Suppl. 1A): 5S–13S.2. Oelschlaeger, T. A., U. Dobrindt, and J. Hacker. 2002. Virulence factors of

uropathogens. Curr. Opin. Urol. 12: 33–38.3. Wright, K. J., and S. J. Hultgren. 2006. Sticky fibers and uropathogenesis:

bacterial adhesins in the urinary tract. Future Microbiol. 1: 75–87.4. Sivick, K. E., and H. L. Mobley. 2010. Waging war against uropathogenic

Escherichia coli: winning back the urinary tract. Infect. Immun. 78: 568–585.5. Klemm, P., and M. A. Schembri. 2000. Bacterial adhesins: function and struc-

ture. Int. J. Med. Microbiol. 290: 27–35.6. Billips, B. K., S. G. Forrestal, M. T. Rycyk, J. R. Johnson, D. J. Klumpp, and

A. J. Schaeffer. 2007. Modulation of host innate immune response in the bladderby uropathogenic Escherichia coli. Infect. Immun. 75: 5353–5360.

7. Svanborg, C., G. Bergsten, H. Fischer, G. Godaly, M. Gustafsson, D. Karpman,A. C. Lundstedt, B. Ragnarsdottir, M. Svensson, and B. Wullt. 2006. Uropa-thogenic Escherichia coli as a model of host-parasite interaction. Curr. Opin.Microbiol. 9: 33–39.

8. Mabbett, A. N., G. C. Ulett, R. E. Watts, J. J. Tree, M. Totsika, C. L. Ong,J. M. Wood, W. Monaghan, D. F. Looke, G. R. Nimmo, et al. 2009. Virulenceproperties of asymptomatic bacteriuria Escherichia coli. Int. J. Med. Microbiol.299: 53–63.

9. Sheu, J. N., M. C. Chen, K. H. Lue, S. L. Cheng, I. C. Lee, S. M. Chen, andG. J. Tsay. 2006. Serum and urine levels of interleukin-6 and interleukin-8 in

children with acute pyelonephritis. Cytokine 36: 276–282.10. Oregioni, O., P. Delaunay, P. Bruna, A. Gaudart, E. Lemichez, P. Boquet, and

L. Landraud. 2005. Urinary interleukin-8 is elevated in urinary tract infectionsindependently of the causative germs. Cytokine 31: 415–418.

11. Otto, G., J. Braconier, A. Andreasson, and C. Svanborg. 1999. Interleukin-6 anddisease severity in patients with bacteremic and nonbacteremic febrile urinarytract infection. J. Infect. Dis. 179: 172–179.

12. Candela, J. V., E. Park, J. M. Gerspach, R. Davidoff, L. Stout, S. M. Levy,G. E. Leach, G. C. Bellman, and P. M. Lad. 1998. Evaluation of urinary IL-1alpha and IL-1beta in gravid females and patients with bacterial cystitis andmicroscopic hematuria. Urol. Res. 26: 175–180.

13. Ulett, G. C., R. I. Webb, K. B. Ulett, X. Cui, W. H. Benjamin, M. Crowley, andM. A. Schembri. 2010. Group B Streptococcus (GBS) urinary tract infection

involves binding of GBS to bladder uroepithelium and potent but GBS-specificinduction of interleukin 1alpha. J. Infect. Dis. 201: 866–870.

14. Engel, D., U. Dobrindt, A. Tittel, P. Peters, J. Maurer, I. Gutgemann,B. Kaissling, W. Kuziel, S. Jung, and C. Kurts. 2006. Tumor necrosis factoralpha- and inducible nitric oxide synthase-producing dendritic cells are rapidlyrecruited to the bladder in urinary tract infection but are dispensable for bacterialclearance. Infect. Immun. 74: 6100–6107.

15. Olszyna, D. P., S. Florquin, M. Sewnath, J. Branger, P. Speelman, S. J. vanDeventer, R. M. Strieter, and T. van der Poll. 2001. CXC chemokine receptor 2contributes to host defense in murine urinary tract infection. J. Infect. Dis. 184:301–307.

16. Saemann, M. D., T. Weichhart, W. H. Horl, and G. J. Zlabinger. 2005. Tamm-Horsfall protein: a multilayered defence molecule against urinary tract infection.Eur. J. Clin. Invest. 35: 227–235.

17. Saemann, M. D., T. Weichhart, M. Zeyda, G. Staffler, M. Schunn, K. M. Stuhlmeier,Y. Sobanov, T. M. Stulnig, S. Akira, A. von Gabain, et al. 2005. Tamm-Horsfallglycoprotein links innate immune cell activation with adaptive immunity viaa Toll-like receptor-4-dependent mechanism. J. Clin. Invest. 115: 468–475.

18. Rouschop, K. M., M. Sylva, G. J. Teske, I. Hoedemaeker, S. T. Pals,J. J. Weening, T. van der Poll, and S. Florquin. 2006. Urothelial CD44 facilitatesEscherichia coli infection of the murine urinary tract. J. Immunol. 177: 7225–7232.

19. Klumpp, D. J., M. T. Rycyk, M. C. Chen, P. Thumbikat, S. Sengupta, andA. J. Schaeffer. 2006. Uropathogenic Escherichia coli induces extrinsic andintrinsic cascades to initiate urothelial apoptosis. Infect. Immun. 74: 5106–5113.

20. Mulvey, M. A., J. D. Schilling, J. J. Martinez, and S. J. Hultgren. 2000. Bad bugsand beleaguered bladders: interplay between uropathogenic Escherichia coli andinnate host defenses. Proc. Natl. Acad. Sci. USA 97: 8829–8835.

21. Hacker, H., C. Furmann, H. Wagner, and G. Hacker. 2002. Caspase-9/-3 acti-vation and apoptosis are induced in mouse macrophages upon ingestion anddigestion of Escherichia coli bacteria. J. Immunol. 169: 3172–3179.

22. Bergsten, G., B. Wullt, M. A. Schembri, I. Leijonhufvud, and C. Svanborg. 2007.Do type 1 fimbriae promote inflammation in the human urinary tract? Cell.Microbiol. 9: 1766–1781.

23. Wullt, B., G. Bergsten, H. Connell, P. Rollano, N. Gebratsedik, L. Hang, andC. Svanborg. 2001. P-fimbriae trigger mucosal responses to Escherichia coli inthe human urinary tract. Cell. Microbiol. 3: 255–264.

24. Allsopp, L. P., M. Totsika, J. J. Tree, G. C. Ulett, A. N. Mabbett, T. J. Wells,B. Kobe, S. A. Beatson, and M. A. Schembri. 2010. UpaH is a newly identifiedautotransporter protein that contributes to biofilm formation and bladder colo-nization by uropathogenic Escherichia coli CFT073. Infect. Immun. 78: 1659–1669.

25. Valle, J., A. N. Mabbett, G. C. Ulett, A. Toledo-Arana, K. Wecker, M. Totsika,M. A. Schembri, J. M. Ghigo, and C. Beloin. 2008. UpaG, a new member of thetrimeric autotransporter family of adhesins in uropathogenic Escherichia coli. J.Bacteriol. 190: 4147–4161.

26. Tree, J. J., G. C. Ulett, C. L. Ong, D. J. Trott, A. G. McEwan, and M. A. Schembri.2008. Trade-off between iron uptake and protection against oxidative stress:deletion of cueO promotes uropathogenic Escherichia coli virulence in a mousemodel of urinary tract infection. J. Bacteriol. 190: 6909–6912.

27. Roos, V., G. C. Ulett, M. A. Schembri, and P. Klemm. 2006. The asymptomaticbacteriuria Escherichia coli strain 83972 outcompetes uropathogenic E. colistrains in human urine. Infect. Immun. 74: 615–624.

28. Buckles, E. L., X. Wang, M. C. Lane, C. V. Lockatell, D. E. Johnson,D. A. Rasko, H. L. Mobley, and M. S. Donnenberg. 2009. Role of the K2 capsulein Escherichia coli urinary tract infection and serum resistance. J. Infect. Dis.199: 1689–1697.

29. Bahrani-Mougeot, F. K., E. L. Buckles, C. V. Lockatell, J. R. Hebel, D. E. Johnson,C. M. Tang, and M. S. Donnenberg. 2002. Type 1 fimbriae and extracellularpolysaccharides are preeminent uropathogenic Escherichia coli virulencedeterminants in the murine urinary tract. Mol. Microbiol. 45: 1079–1093.

30. Bertani, G. 2004. Lysogeny at mid-twentieth century: P1, P2, and other exper-imental systems. J. Bacteriol. 186: 595–600.

31. Pfaffl, M. W., G. W. Horgan, and L. Dempfle. 2002. Relative expression softwaretool (REST) for group-wise comparison and statistical analysis of relative ex-pression results in real-time PCR. Nucleic Acids Res. 30: e36.

32. Duell, B. L., A. W. Cripps, M. A. Schembri, and G. C. Ulett. 2011. Epithelial cellcoculture models for studying infectious diseases: benefits and limitations. J.Biomed. Biotechnol. 2011: 852419.

33. Ulett, G. C., K. H. Maclean, S. Nekkalapu, J. L. Cleveland, and E. E. Adderson.2005. Mechanisms of group B streptococcal-induced apoptosis of murine mac-rophages. J. Immunol. 175: 2555–2562.

34. Ulett, G. C., and E. E. Adderson. 2005. Nitric oxide is a key determinant ofgroup B streptococcus-induced murine macrophage apoptosis. J. Infect. Dis.191: 1761–1770.

35. Ulett, G. C., J. F. Bohnsack, J. Armstrong, and E. E. Adderson. 2003. Beta-hemolysin-independent induction of apoptosis of macrophages infected withserotype III group B streptococcus. J. Infect. Dis. 188: 1049–1053.

36. Datsenko, K. A., and B. L. Wanner. 2000. One-step inactivation of chromosomalgenes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA97: 6640–6645.

37. Bolstad, B. M., R. A. Irizarry, M. Astrand, and T. P. Speed. 2003. A comparisonof normalization methods for high density oligonucleotide array data based onvariance and bias. Bioinformatics 19: 185–193.

38. Irizarry, R. A., B. Hobbs, F. Collin, Y. D. Beazer-Barclay, K. J. Antonellis,U. Scherf, and T. P. Speed. 2003. Exploration, normalization, and summaries ofhigh density oligonucleotide array probe level data. Biostatistics 4: 249–264.

The Journal of Immunology 791

by guest on July 9, 2018http://w

ww

.jimm

unol.org/D

ownloaded from

39. Wu, W., G. Churchill, M. K. Kerr, and X. Cui. 2003. Analysis of N-dyemicroarray experiments using mixed model effects containing analysis of vari-ance, permutation and bootstrap, cluster and consensus tree. Available at: http://www.bioconductor.org/packages/bioc/1.8/html/maanova.html. Accessed:

40. Cui, X., J. T. Hwang, J. Qiu, N. J. Blades, and G. A. Churchill. 2005. Improvedstatistical tests for differential gene expression by shrinking variance compo-nents estimates. Biostatistics 6: 59–75.

41. Yang, H., and G. Churchill. 2007. Estimating p-values in small microarrayexperiments. Bioinformatics 23: 38–43.

42. Benjamini, Y., and Y. Hochberg. 1995. Controlling the false discovery rate:a practical and powerful approach to multiple testing. J. Roy. Statist. Soc. Ser. B57: 289–300.

43. Barry, W. T., A. B. Nobel, and F. A. Wright. 2005. Significance analysis offunctional categories in gene expression studies: a structured permutation ap-proach. Bioinformatics 21: 1943–1949.

44. Yekutieli, D., and Y. Benjamini. 1999. Resampling-based false discovery ratecontrolling multiple test procedures for correlated test statistics. J. Statist. Plann.Inference 82: 171–196.

45. Lynn, D. J., G. L. Winsor, C. Chan, N. Richard, M. R. Laird, A. Barsky,J. L. Gardy, F. M. Roche, T. H. Chan, N. Shah, et al. 2008. InnateDB: facilitatingsystems-level analyses of the mammalian innate immune response. Mol. Syst.Biol. 4: 218.

46. Ingersoll, M. A., K. A. Kline, H. V. Nielsen, and S. J. Hultgren. 2008. G-CSFinduction early in uropathogenic Escherichia coli infection of the urinary tractmodulates host immunity. Cell. Microbiol. 10: 2568–2578.

47. Sivick, K. E., M. A. Schaller, S. N. Smith, and H. L. Mobley. 2010. The innateimmune response to uropathogenic Escherichia coli involves IL-17A in a murinemodel of urinary tract infection. J. Immunol. 184: 2065–2075.

48. Carbone, M., D. L. Hasty, K. C. Yi, J. Rue, M. T. Fera, F. La Torre, M. Giannone,and E. Losi. 2002. Cytokine induction in murine bladder tissue by type 1 fim-briated Escherichia coli. Ann. N. Y. Acad. Sci. 963: 332–335.

49. Hannan, T. J., I. U. Mysorekar, C. S. Hung, M. L. Isaacson-Schmid, andS. J. Hultgren. 2010. Early severe inflammatory responses to uropathogenic E.coli predispose to chronic and recurrent urinary tract infection. PLoS PathogensDOI:10.1371/journal.ppat.1001042.

50. Mysorekar, I. U., M. A. Mulvey, S. J. Hultgren, and J. I. Gordon. 2002. Mo-lecular regulation of urothelial renewal and host defenses during infection withuropathogenic Escherichia coli. J. Biol. Chem. 277: 7412–7419.

51. Anderson, G. G., J. J. Palermo, J. D. Schilling, R. Roth, J. Heuser, andS. J. Hultgren. 2003. Intracellular bacterial biofilm-like pods in urinary tractinfections. Science 301: 105–107.

52. Justice, S. S., D. A. Hunstad, P. C. Seed, and S. J. Hultgren. 2006. Filamentationby Escherichia coli subverts innate defenses during urinary tract infection. Proc.Natl. Acad. Sci. USA 103: 19884–19889.

53. Jones-Carson, J., E. Balish, and D. T. Uehling. 1999. Susceptibility of immu-nodeficient gene-knockout mice to urinary tract infection. J. Urol. 161: 338–341.

54. Couper, K. N., D. G. Blount, and E. M. Riley. 2008. IL-10: the master regulatorof immunity to infection. J. Immunol. 180: 5771–5777.

55. Fischer, H., N. Lutay, B. Ragnarsdottir, M. Yadav, K. Jonsson, A. Urbano, A. AlHadad, S. Ramisch, P. Storm, U. Dobrindt, et al. 2010. Pathogen specific, IRF3-dependent signaling and innate resistance to human kidney infection. PLoSPathogens DOI:10.1371/journal.ppat.1001109.

56. Dessing, M. C., L. M. Butter, G. J. Teske, N. Claessen, C. M. van der Loos,T. Vogl, J. Roth, T. van der Poll, S. Florquin, and J. C. Leemans. 2010. S100A8/A9 is not involved in host defense against murine urinary tract infection. PLoSOne 5: e13394.

57. Reigstad, C. S., S. J. Hultgren, and J. I. Gordon. 2007. Functional genomicstudies of uropathogenic Escherichia coli and host urothelial cells when intra-cellular bacterial communities are assembled. J. Biol. Chem. 282: 21259–21267.

58. Godaly, G., G. Bergsten, L. Hang, H. Fischer, B. Frendeus, A. C. Lundstedt,M. Samuelsson, P. Samuelsson, and C. Svanborg. 2001. Neutrophil recruitment,chemokine receptors, and resistance to mucosal infection. J. Leukoc. Biol. 69:899–906.

59. Garcia, E. C., A. R. Brumbaugh, and H. L. Mobley. 2011. Redundancy andspecificity of Escherichia coli iron acquisition systems during urinary tract in-fection. Infect. Immun. 79: 1225–1235.

60. Hagan, E. C., and H. L. Mobley. 2009. Haem acquisition is facilitated by a novelreceptor Hma and required by uropathogenic Escherichia coli for kidney in-fection. Mol. Microbiol. 71: 79–91.

61. Watts, R. E., M. Totsika, V. L. Challinor, A. N. Mabbett, G. C. Ulett, J. J. DeVoss, and M. A. Schembri. 2011. Contribution of siderophore systems to growthand urinary tract colonization of asymptomatic bacteriuria Escherichia coli.Infect. Immun. In press.

62. Godaly, G., G. Otto, M. D. Burdick, R. M. Strieter, and C. Svanborg. 2007.Fimbrial lectins influence the chemokine repertoire in the urinary tract mucosa.Kidney Int. 71: 778–786.

63. Tsai, K. W., H. T. Lai, T. C. Tsai, Y. C. Wu, Y. T. Yang, K. Y. Chen, C. M. Chen,Y. S. Li, and C. N. Chen. 2009. Difference in the regulation of IL-8 expression

induced by uropathogenic E. coli between two kinds of urinary tract epithelialcells. J. Biomed. Sci. 16: 91.

64. O’Garra, A., P. L. Vieira, P. Vieira, and A. E. Goldfeld. 2004. IL-10-producingand naturally occurring CD4+ Tregs: limiting collateral damage. J. Clin. Invest.114: 1372–1378.

65. Smith, S. D., M. A. Wheeler, M. I. Lorber, and R. M. Weiss. 2000. Temporalchanges of cytokines and nitric oxide products in urine from renal transplantpatients. Kidney Int. 58: 829–837.

66. Jarry, A., C. Bossard, C. Bou-Hanna, D. Masson, E. Espaze, M. G. Denis, andC. L. Laboisse. 2008. Mucosal IL-10 and TGF-beta play crucial roles in pre-venting LPS-driven, IFN-gamma-mediated epithelial damage in human colonexplants. J. Clin. Invest. 118: 1132–1142.

67. Mege, J. L., S. Meghari, A. Honstettre, C. Capo, and D. Raoult. 2006. The twofaces of interleukin 10 in human infectious diseases. Lancet Infect. Dis. 6: 557–569.

68. Turner, J., M. Gonzalez-Juarrero, D. L. Ellis, R. J. Basaraba, A. Kipnis,I. M. Orme, and A. M. Cooper. 2002. In vivo IL-10 production reactivateschronic pulmonary tuberculosis in C57BL/6 mice. J. Immunol. 169: 6343–6351.

69. Sivan, Y., B. Griffel, O. Medalia, and M. Aronson. 1982. Comparative histologyof the mouse bladder following initial infection and re-infection with Escher-ichia coli. J. Pathol. 138: 353–364.

70. Kaoutzani, P., S. P. Colgan, K. L. Cepek, P. G. Burkard, S. Carlson, C. Delp-Archer, M. B. Brenner, and J. L. Madara. 1994. Reconstitution of cultured in-testinal epithelial monolayers with a mucosal-derived T lymphocyte cell line:modulation of epithelial phenotype dependent on lymphocyte-basolateral mem-brane apposition. J. Clin. Invest. 94: 788–796.

71. McKay, D. M., K. Croitoru, and M. H. Perdue. 1996. T cell-monocyte inter-actions regulate epithelial physiology in a coculture model of inflammation. Am.J. Physiol. 270: C418–C428.

72. Tuettenberg, A., S. Fondel, K. Steinbrink, A. H. Enk, and H. Jonuleit. 2010.CD40 signalling induces IL-10-producing, tolerogenic dendritic cells. Exp. Der-matol. 19: 44–53.

73. Louis, N. A., D. Daniels, and S. P. Colgan. 2006. Cell-cell interactions on solidmatrices. Methods Mol. Biol. 341: 51–59.

74. Blume, E. D., C. T. Taylor, P. F. Lennon, G. L. Stahl, and S. P. Colgan. 1998.Activated endothelial cells elicit paracrine induction of epithelial chloride se-cretion: 6-Keto-PGF1alpha is an epithelial secretagogue. J. Clin. Invest. 102:1161–1172.

75. Roncarolo, M. G., S. Gregori, M. Battaglia, R. Bacchetta, K. Fleischhauer, andM. K. Levings. 2006. Interleukin-10-secreting type 1 regulatory T cells in ro-dents and humans. Immunol. Rev. 212: 28–50.

76. Madura Larsen, J., C. S. Benn, Y. Fillie, D. van der Kleij, P. Aaby, andM. Yazdanbakhsh. 2007. BCG stimulated dendritic cells induce an interleukin-10 producing T-cell population with no T helper 1 or T helper 2 bias in vitro.Immunology 121: 276–282.

77. Kullberg, M. C., D. Jankovic, P. L. Gorelick, P. Caspar, J. J. Letterio,A. W. Cheever, and A. Sher. 2002. Bacteria-triggered CD4(+) T regulatory cellssuppress Helicobacter hepaticus-induced colitis. J. Exp. Med. 196: 505–515.

78. Ulett, G. C., N. Ketheesan, and R. G. Hirst. 1998. Macrophage-lymphocyteinteractions mediate anti-Burkholderia pseudomallei activity. FEMS Immunol.Med. Microbiol. 21: 283–286.

79. Chanteux, H., A. C. Guisset, C. Pilette, and Y. Sibille. 2007. LPS inducesIL-10 production by human alveolar macrophages via MAPKinases- and Sp1-dependent mechanisms. Respir. Res. 8: 71.

80. Price, J. D., J. Schaumburg, C. Sandin, J. P. Atkinson, G. Lindahl, and C. Kemper.2005. Induction of a regulatory phenotype in human CD4+ T cells by strepto-coccal M protein. J. Immunol. 175: 677–684.

81. Pengal, R. A., L. P. Ganesan, G. Wei, H. Fang, M. C. Ostrowski, andS. Tridandapani. 2006. Lipopolysaccharide-induced production of interleukin-10is promoted by the serine/threonine kinase Akt. Mol. Immunol. 43: 1557–1564.

82. Fujita, S., K. Seino, K. Sato, Y. Sato, K. Eizumi, N. Yamashita, M. Taniguchi,and K. Sato. 2006. Regulatory dendritic cells act as regulators of acute lethalsystemic inflammatory response. Blood 107: 3656–3664.

83. Akira, S., S. Uematsu, and O. Takeuchi. 2006. Pathogen recognition and innateimmunity. Cell 124: 783–801.

84. Pils, M. C., F. Pisano, N. Fasnacht, J. M. Heinrich, L. Groebe, A. Schippers,B. Rozell, R. S. Jack, and W. Muller. 2010. Monocytes/macrophages and/orneutrophils are the target of IL-10 in the LPS endotoxemia model. Eur. J.Immunol. 40: 443–448.

85. Elbim, C., H. Reglier, M. Fay, C. Delarche, V. Andrieu, J. El Benna, andM. A. Gougerot-Pocidalo. 2001. Intracellular pool of IL-10 receptors in specificgranules of human neutrophils: differential mobilization by proinflammatorymediators. J. Immunol. 166: 5201–5207.

86. Tamassia, N., F. Calzetti, N. Menestrina, M. Rossato, F. Bazzoni, L. Gottin, andM. A. Cassatella. 2008. Circulating neutrophils of septic patients constitutivelyexpress IL-10R1 and are promptly responsive to IL-10. Int. Immunol. 20: 535–541.

792 EARLY BLADDER TRANSCRIPTOME OF UPEC UTI AND IL-10

by guest on July 9, 2018http://w

ww

.jimm

unol.org/D

ownloaded from