Table Of ContentKarlssonetal.BMCMicrobiology2012,12:15
http://www.biomedcentral.com/1471-2180/12/15
RESEARCH ARTICLE Open Access
Lactobacillus rhamnosus GR-1 enhances
NF-kappaB activation in Escherichia coli-
stimulated urinary bladder cells through TLR4
Mattias Karlsson1, Nikolai Scherbak1, Gregor Reid2,3 and Jana Jass1*
Abstract
Background: Epithelial cells of the urinary tract recognize pathogenic bacteria through pattern recognition
receptors on their surface, such as toll-like receptors (TLRs), and mount an immune response through the
activation of the NF-kappaB pathway. Some uropathogenic bacteria can subvert these cellular responses, creating
problems with how the host eliminates pathogens. Lactobacillus is a genus of lactic acid bacteria that are part of
the microbiota and consist of many probiotic strains, some specifically for urogenital infections.
Immunomodulation has emerged as an important mode of action of probiotic and commensal lactobacilli and
given the importance of epithelial cells, we evaluated the effect of the urogenital probiotic Lactobacillus rhamnosus
GR-1 on epithelial immune activation.
Results: ImmuneactivationthroughtheNF-kappaBpathwaywasinitiatedbystimulationofT24urothelialcellswith
heat-killedEscherichiacoliandthiswasfurtherpotentiatedwhencellswereco-culturedwithliveL.rhamnosusGR-1.
Heat-killedlactobacilliwerepooractivatorsofNF-kappaB.ConcomitantstimulationofbladdercellswithE.coliandL.
rhamnosusGR-1increasedthelevelsofthepro-inflammatorycytokineTNF,whereasIL-6andCXCL8levelswerereduced.
Anotherprobiotic,L.rhamnosusGG,wasalsoabletopotentiateNF-kappaBinthesecellsalthoughatasignificantly
reducedlevelcomparedtotheGR-1strain.Thetranscriptnumbersandproteinlevelsofthelipopolysaccharidereceptor
TLR4weresignificantlyincreasedafterco-stimulationwithE.coliandlactobacillicomparedtocontrols.Furthermore,
inhibitionofTLR4activationbypolymixinBcompletelyblockedthelactobacillipotentiationofNF-kappaB.
Conclusions: The immunological outcome of E. coli challenge of bladder cells was influenced by probiotic L.
rhamnosus GR-1, by enhancing the activation of NF-kappaB and TNF release. Thus the urogenital probiotic L.
rhamnosus GR-1 modulated the activation of the NF-kappaB through increased levels of TLR4 on the bladder cells
and altered subsequent release of cytokines from urothelial cells. By influencing immunological factors such as
TLR4, important in the process of fighting pathogens, lactobacilli could facilitate pathogen recognition and
infection clearance.
Background recognition receptors such as toll-like receptors (TLRs)
Manybacterialdiseases,includingurinarytractinfections that are able to trigger the expression of inflammatory
(UTIs) are initiated by microorganisms adhering to and mediatorsandsubsequentinflammationinthepresenceof
colonizing the epithelium. Epithelial cells of the urinary pathogenic microbes. One of the most studied TLRs is
tract(urothelialcells)respondtopathogensbyproducing TLR4,whichbindslipopolysaccharides(LPS)foundonthe
various immune activating substances including com- cell wall of Gram-negative bacteria [1]. Key proteins
pounds that recruit immune cells such as macrophages. involvedininflammationaretheRel/NuclearFactor(NF)-
Epithelial cells express a number of different pattern (cid:1)Bproteins,whichonceactivatedcaninducethetranscrip-
tionofseveralimmunologicallyessentialmolecules,suchas
tumornecrosisfactor(TNF),interleukin(IL)-6andCXCL8
*Correspondence:[email protected]
1SchoolofScienceandTechnology,LifeScience,ÖrebroUniversity,70182 [2-4].Thesecytokinesareveryimportantintheantimicro-
Örebro,Sweden bialandinflammatoryprocessandtheyeffectivelyrecruit
Fulllistofauthorinformationisavailableattheendofthearticle
©2012Karlssonetal;licenseeBioMedCentralLtd..ThisisanOpenAccessarticledistributedunderthetermsoftheCreative
CommonsAttributionLicense(http://creativecommons.org/licenses/by/2.0),whichpermitsunrestricteduse,distribution,and
reproductioninanymedium,providedtheoriginalworkisproperlycited.
Karlssonetal.BMCMicrobiology2012,12:15 Page2of10
http://www.biomedcentral.com/1471-2180/12/15
immunecellstotheinfectedsite.Initsinactiveform,the of NF-(cid:1)B and cytokine release from bladder cells. After
NF-(cid:1)B transcription factor is located within the cytosol, 24hofchallengewithheat-killedE.coli,cellsresponded
whereinhibitoryproteinsmaskingthenuclearlocalization with more than 10-fold increase in NF-(cid:1)B activation
signalimpairitsnuclearmigration.DuringNF-(cid:1)Bactiva- compared to resting cells, as measured by the luciferase
tion, the inhibitory proteins are disassociated from the reporterassay(Figure1A).Furthermore,challengegavea
transcription factor dimer, which is subsequently trans- substantialincrease inpro-inflammatoryTNF,IL-6, and
portedinto the nucleus[5].Nuclear translocation ofNF- CXCL8levels(Figure1B,C,and1D).Ontheotherhand,
(cid:1)Bduringinfectiousprocessesisimportantforthesubse- L. rhamnosus GR-1 was a poor activator of NF-(cid:1)B. Sti-
quentactivationofimmuneresponses. mulation with viable lactobacilli led to a minor increase
The most prevalent cause of UTI is uropathogenic intheactivationofNF-(cid:1)Bwhileheat-killedbacteriahad
Escherichiacoli(UPEC),whichexpressesnumerousviru- nosignificanteffect(Figure2A).Althoughviablelactoba-
lencefactorsincludingtoxinsandfimbriaeusedforadhe- cilli could marginally increase NF-(cid:1)B activation com-
sion.Eukaryoticcellscanidentifypathogens,forexample pared to resting cells, stimulation did not promote
whentype1fimbriae,P-pili,orLPSbindtoTLR4andeli- release of any of the tested cytokines (TNF, IL-6 and
citaninflammatoryresponse,albeitviadifferentintracel- CXCL8).Incontrast,itresultedinasmallbutsignificant
lular pathways [6]. However, some UPEC are equipped reduction of CXCL8, compared to resting cells, while
with virulence factors that can block immune responses TNFandIL-6levelswereunaffected(Figure2B).
allowing the organisms to freely multiply. These UPEC Lactobacilli do not normally come into contact with
canactdirectlyonTLRactivation,inhibitingNF-(cid:1)Batan bladder cells, therefore we determined the cytotoxicity
earlystageafterligandbindingandconsequently,thereis caused by lactobacilli exposure. However, we did not
no transcription of the integral components needed to observe decreased epithelial cell viability compared to
mount an immune response [7]. Similarly, a low TLR4
expressionoractivityisassociatedwithincreasedUTIsus-
ceptibility. Suchstrategieswouldimpedepathogenclear-
anceinvivoandcauserecurrentUTIs[8,9]. A B
LactobacillusisagenusofGram-positivebacterianatu-
rallyfoundinthehealthyhumanvagina[10]andurethra * *
[11]. Moreover, a low Lactobacillus count is inversely
related to high numbers of E. coli in the vagina and a
history ofrecurrent UTI[12]. Severallactobacillistrains
are used as probiotics to prevent infections within the
gastrointestinalandurogenitaltractsaswellastoamelio-
rate allergic and inflammatory conditions [13-15]. The
probioticmechanismsarebelievedto includetherelease
of antibacterial substances, biosurfactant production,
disruption of biofilms and competitive exclusion [16].
C D
Furthermore,the ability ofprobioticstrainstomodulate
immunitythroughNF-(cid:1)Bandmitogenactivatedprotein *
(MAP) kinase pathways, both important in the develop-
*
mentofinnateandadaptiveimmunity,hasbeenreported
[17,18].LactobacillusrhamnosusGR-1isaprobioticiso-
latedfromafemaleurethra[19]usedtopreventUTIand
bacterial vaginosis, and it has both immunomodulatory
and antimicrobial activity [20,21]. Currently, the immu-
nological effects of lactobacilli on urothelial cells are in
largepartunexplored.Theaimofthiscurrentstudywas
toinvestigatehowL.rhamnosusGR-1canaffecturothe-
lialimmuneresponsestoE.coli. Figure 1 NF-(cid:1)B activation and expression of cytokines in
bladdercellsafterE.colichallenge.Bladdercellswerestimulated
withheat-killedE.colifor24hataconcentrationcorrespondingto
Results 108cfu/ml.(A)RelativeNF-(cid:1)Bactivationwasmeasuredbyluciferase
Bladder cells responded poorly to lactobacilli compared activity(n=4)andtheproteinlevelsof(B)TNF,(C)IL-6,and(D)
to heat-killed E. coli CXCL8weremeasuredbyELISA(n=3).Errorbarsrepresentthe
E. coli are potent activators of epithelial immune standarderrorsofthemeans.Barslabeledwithanasterisk
significantlydifferfromthecontrol(p-values<0.05).
responsesandwerethereforeusedtostimulateactivation
Karlssonetal.BMCMicrobiology2012,12:15 Page3of10
http://www.biomedcentral.com/1471-2180/12/15
A
*
*
E. coli - + - -
L. rhamnosus - - V HK
B
*
L. rhamnosus - V - V - V
Figure2NF-(cid:1)BactivationandexpressionofcytokinesinbladdercellsafterstimulationwithL.rhamnosusGR-1.Viable(V)orheat-killed
(HK)L.rhamnosusGR-1ataconcentrationof2×107cfu/mlwereusedtochallengebladdercellsfor24h.(A)RelativeNF-(cid:1)Bactivation(n=4)
and(B)TNF,IL-6,andCXCL8levels(n=3)weremeasuredusingluciferaseassayandELISA,respectively.Errorbarsrepresentthestandarderrors
ofthemeans.Barslabeledwithanasterisksignificantlydifferfromthecontrol(p-values<0.05).
resting cells, as determined using propidium iodide IL-6 and CXCL8 levels were reduced compared to those
stainedcellsandflowcytometry(datanotshown). found during E. coli challenge alone (Figure 3B).
NF-(cid:1)B activation was significantly reduced when blad-
Viable lactobacilli potentiated NF-(cid:1)B activation and der cells were exposed to heat-stable cell wall compo-
cytokine response in E. coli-stimulated cells nents of lactobacilli (Figure 3A), indicating that
Bladder cells were relatively indifferent towards stimula- potentiation was mediated by compound(s) released
tion with both viable and heat-killed lactobacilli, during the growth of L. rhamnosus GR-1.
whereas the cells responded appropriately towards sti-
mulation with E. coli, leading to increased NF-(cid:1)B activa- L. rhamnosus GR-1 and GG augmented NF-(cid:1)B to different
tion and release of inflammatory mediators. Co- levels
stimulation with viable lactobacilli and heat-killed E. coli LactobacillusrhamnosusGG,awell-studiedimmunomo-
did however result in increased NF-(cid:1)B activation com- dulatory strain used for gastrointestinal disorders, was
pared to cells challenged with E. coli alone (Figure 3A). chosen to compare NF-(cid:1)B augmenting abilities. Both L.
This NF-(cid:1)B induction was beyond an eventual additive rhamnosusGR-1andGGhad theabilitytopotentiateE.
effect, representing a synergistic action on NF-(cid:1)B activa- coliinducedNF-(cid:1)Bactivation(Figure4).WhileL.rham-
tion. On the protein level, co-stimulation influenced the nosus GG induced NF-(cid:1)B twofold, L. rhamnosus GR-1
release of all studied inflammatory mediators. The TNF showed a three- to fourfold induction of NF-(cid:1)B com-
release was increased by a factor of two to three, while paredtocellsthathadnolactobacilliadded.
Karlssonetal.BMCMicrobiology2012,12:15 Page4of10
http://www.biomedcentral.com/1471-2180/12/15
A
b
b
a
E. coli - + + +
L. rhamnosus - - V HK
B
a
b a
b
b
a
E. coli - + + - + + - + +
L. rhamnosus - - V - - V - - V
Figure3NF-(cid:1)BactivationandcytokinesecretionafterconcomitantstimulationwithE.coliandL.rhamnosusGR-1.Bladdercellswere
challengedfor24hwithheat-killedE.colialoneortogetherwithviable(V)orheat-killed(HK)L.rhamnosusGR-1.(A)RelativeNF-(cid:1)Bactivation
(n=4).(B)TNF,IL-6andCXCL8levels(n=3)weremeasured.Barslabeled“a”aresignificantlydifferentfromcontroland“b”significantly
differentfromcellsstimulatedwithE.coli(p-values<0.05).
*
L. rhamnosus GR-1 modified TLR4 expression on bladder
cells
TLR4 is a crucial protein in the detection ofE. coli by
epithelial cells, therefore we proceeded by analyzing the
levels of TLR4 in bladder cells treated with heat-killed
E. coli and L. rhamnosus GR-1. Co-stimulated bladder
cells showed increased expression of TLR4 mRNA com-
pared to cells stimulated with E. coli or lactobacilli
alone (Figure 5A). Furthermore, immunoblotting using
native proteins showed high band intensity in co-stimu-
E. coli - + + +
lated cells suggesting higher TLR4 protein content com-
- - GR-1 GG pared to all other groups (Figure 5B). The effect on
L. rhamnosus
TLR4 protein levels was further characterized using con-
Figure 4 NF-(cid:1)B augmentation by two different L. rhamnosus focal laser microscopy. Control cells and cells stimulated
strains.Bladdercellswereco-stimulatedwithheat-killedE.coliand with only E. coli or lactobacilli showed no or low TLR
viableL.rhamnosusGR-1orL.rhamnosusGGfor24h(n=4).An expression, whereas cells co-stimulated with both E. coli
asteriskdenotessignificantdifferencebetweenthetwogroups(p- and L. rhamnosus GR-1 demonstrated a substantial
values<0.05).
increase in the amount of TLR4 protein (Figure 5C).
Karlssonetal.BMCMicrobiology2012,12:15 Page5of10
http://www.biomedcentral.com/1471-2180/12/15
A B
TLR4 β-actin
*
Control
E. coli
L. rhamnosus GR-1
E. coli +
E. c oli L. rh a mG nRo-s1 u s E. mc nolio s+u s G R-1 L. rhamnosus GR-1
C L. rh a
Control E. coli
Lactobacillus E. coli + Lactobacillus
20 μm
Figure5TLR4expressioninbladdercellsafterLactobacillusstimulation.(A)TLR4qPCRfromcellsco-stimulatedfor3hwithE.coliandL.
rhamnosusGR-1(n=3).(B)AnativeimmunoblotofTLR4proteinafter24hstimulation.(C)ConfocalmicroscopyofTLR4protein(greenpixels)
afterstimulationofT24cells.ThecellswerealsostainedwithDAPI(DNAstain)andAlexa555phalloidin(actinstain)andpseudo-coloredblue
andred,respectively.Immunoblotandconfocalimagesarerepresentativedatafromtwoormoreseparateexperiments.Barslabeledwithan
asteriskaresignificantlydifferentfromcontrolcells(p-values<0.05).
Karlssonetal.BMCMicrobiology2012,12:15 Page6of10
http://www.biomedcentral.com/1471-2180/12/15
Polymyxin B suppressed NF-(cid:1)B augmentation signaling present between viable bacteria as well as the
WecontinuedtocharacterizetheroleofTLR4inNF-(cid:1)B effects of E. coli metabolites on cell cultures [24]. Our
activation by co-stimulation with heat-killed E. coli and results showed that bladder cells challenged with heat-
lactobacilli. The TLR4 activation in bladder cells was killed E. coli and subjected to stimulation with L. rham-
inhibited by pretreatment with polymyxin B, a known nosus GR-1 exhibited increased NF-(cid:1)B activation and
inhibitorofLPS-inducedTLR4activation,andthereafter TNF release.
stimulated by E. coli and L. rhamnosus GR-1 (Figure 6). ThefindingthatL.rhamnosusdoesindeedhaveimmu-
Polymyxin B significantly inhibited NF-(cid:1)B activation in nomodulatorypropertiesisnotnewperse,butmostpre-
cells challenged with both E. coli and lactobacilli vious experiments have been done using immune cells
althoughithadnosignificanteffectonNF-(cid:1)Bactivation [20,25].AdjuvantpropertiesofLactobacillusspecieshave
in resting cells and on lactobacilli treated cells. The beendemonstratedinseveralinvivomodels.AnL.casei
increasedNF-(cid:1)Bactivationobservedduring co-stimula- strain boosted immunoglobulin (Ig)A secretion in a
tion was completely lost after polymyxin B treatment, mousemodelofSalmonellatyphimuriuminfection[26].
demonstratingtheinvolvementofLPSandTLR4. Another effectively potentiatedIgGresponsesaftersub-
cutaneous vaccination of chickens towards Newcastle
Discussion diseasevirusandinfectiousbronchitisvirus[27].Collec-
Activation of NF-(cid:1)B during infection has a profound tively,thesestudiesprovideevidencethatlactobacillican
effect on the expression of multiple targets which guide be used for potentiating immune responses in vivo.
the maturation of immune responses against invading Nevertheless,althoughTNFwasupregulatedbyL.rham-
pathogens [22]. Recently, much attention has been given nosus GR-1 treatment, anti-inflammatory properties of
to the immunomodulatory activities of the microbiota lactobacilli are well established [25]. In our study, both
and various probiotic organisms. Studies have shown a IL-6 and CXCL8 were modulated differently from TNF,
L. plantarum probiotic to be effective at modulating where both were down-regulated after lactobacilli treat-
immunity through NF-(cid:1)B and MAP kinase signaling in ment of E. coli-challenged cells. These effects might
a number of cell types including mucosal epithelial cells representan alternative influence of L.rhamnosus GR-1
[23]. In this study we showed the immunomodulatory on epithelial immune function, guided by transcription
effects of a urogenital probiotic, L. rhamnosus GR-1 on factors other than NF-(cid:1)B, such as MAP kinase/AP-1
human bladder cells. In order to activate the urothelial pathways or post-transcriptional regulation of NF-(cid:1)B-
cell defense mechanisms in a way that resembles the regulatedgenes.AnotherpossibilityisthatL.rhamnosus
response during a UTI, including NF-(cid:1)B and cytokine GR-1 produces substances that can interfere with cyto-
release, we challenged the cells with heat-killed E. coli. kine release from the cell or cytokine stability in the
Although only live bacteria are active in the infection extracellularspace.
process, we wanted to reduce the microbe-to-microbe Probiotichealthbenefitshavebeenshownto be some-
what strain specific. In this study, we showed that two
strains exhibit different abilities to increase activation of
b NF-(cid:1)B.L.rhamnosusGGelicitedaweakerpotentiationof
E.coli-inducedNF-(cid:1)BactivationthanL.rhamnosusGR-1.
Heat-killedpreparationsofL.rhamnosusGR-1marginally
augmented NF-(cid:1)B, in a manner similar to using viable
a
L.rhamnosusGG(belowtwofold).Itispossiblethisaug-
mentation is due to surface-associated structures shared
a bybothstrains.Lactobacillisurfacecomponentshavepre-
viously been shown to modulate NF-(cid:1)B in a contact-
dependentmanner[17].T24cellsexpressTLR2,andcan
recognizelipoteichoicacid(LTA)foundonthesurfaceof
E. coli - + - + lactobacilli with increased NF-(cid:1)B activation as a conse-
quence [28]. However, since heat-killed lactobacilli only
L. rhamnosus - - + +
slightly induced NF-(cid:1)B activation that is not a likely
mechanismgiventhatLTAisanchoredtotheGram-posi-
Figure 6 NF-(cid:1)B potentiation is TLR4 dependant. Polymyxin B
tivecellwall.Amoreprobablemechanismisthatproducts
(PMB)wasaddedtocellculturebeforestimulationtoinhibitTLR4
activation.Cellswerestimulatedwithheat-killedE.coliandviableL. released during bacterial growth are responsible for the
rhamnosusGR-1for24h(n=3).Barslabeled“a”aresignificantly NF-(cid:1)B augmentation by L. rhamnosus GR-1. We have
differentfromcontroland“b”significantlydifferentfromE.coli previously shown that spent culture supernatant from
stimulatedcells(p-values<0.05). L. rhamnosus GR-1 can augment NF-(cid:1)B activation in
Karlssonetal.BMCMicrobiology2012,12:15 Page7of10
http://www.biomedcentral.com/1471-2180/12/15
E. coli-challenged T24 cells[29]. There are no published carry specific receptors, such as TLR4 that can recog-
studies on the identity of the secreted proteins from nize the most common Gram-negative species. Once
L.rhamnosusGR-1.HoweverL.rhamnosusGGisknown these receptors bind the cognate bacterial ligand, the
toreleaseasmallnumberofproteinsduringgrowth,none epithelial cells respond by producing a range of com-
of which have an established immunomodulatory effect pounds including cytokines that are strongly regulated
[30]. A comparison of secretory proteins from the two by the NF-(cid:1)B transcription factor. The present in vitro
strains might help explain the differences in terms of study showed that this immune activation could be
immunepotentiation. amplified by probiotic L. rhamnosus GR-1. Moreover,
TheroleofTLR4wasevaluatedbyblockingLPSbind- augmentation of NF-(cid:1)B was accompanied by an increase
ing tothereceptorusingpolymyxinB,whicheliminated in inflammatory TNF expression. The important recog-
the observed NF-kB potentiation. We initially saw that nition molecule TLR4 was found to be up-regulated by
expression of TLR4 at genetic and protein levels was L. rhamnosus GR-1 on both mRNA and protein level in
increasedduringco-stimulationcomparedtocontrols,or cells concomitantly challenged with E. coli. Moreover,
during individual stimulationwith E. colior lactobacilli. the blocking agonist binding to TLR4 completely inhib-
AlthoughTLR4hasLPSasanatural ligand,otherE.coli ited the augmentation of NF-(cid:1)B by L. rhamnosus GR-1.
components such as pili have been shown to be able to Due to the importance of TLR4 in the process of patho-
activateTLR4.However,inthisstudy,polymyxinBcom- gen clearance we suggest that this represents a pathway
pletely inhibited NF-(cid:1)B activation in E. coli stimulated in which probiotic immunomodulatory lactobacilli work
cells,therefore pilior othersurface structurescouldnot to increase immunity and prevent infections.
havecontributedtothiseffect[31].Weconsiderthatan
increasednumberofTLR4presentonthecellfacilitated Methods
activationbyligandsonE.coliandlactobacillialike. Cell culture
TLRs are important in UTI disease progression, as The T24 human bladder carcinoma cell line (ATCC
shown in C3H/HeJ mice with a mutation in theTlr4 HTB-4) was cultured in RPMI 1640 (Hyclone) supple-
gene. After an E. coli infection, these mutant mice have mented with 2.05 mM of L-glutamine and 10% fetal
problems removing the pathogens from their urinary bovine serum (FBS; Hyclone) at 37°C with 5% CO in a
2
tract[32].ArecentstudyscoringTLR4expressionlevels humidified environment.
inhealthycontrolsubjectsandUTIpatientsshowedthat
the latter have a lower TLR4 expression than healthy Bacterial strains and growth conditions
controls[9].ThisimportantfeatureofTLR4isconsistent L. rhamnosusGR-1 (urethral isolate) andGG (intestinal
with the effect that certain E. coli strains expressing isolate) were cultured on de Man Rogosa Sharp (MRS)
immunomodulatory compounds have on TLR signaling agar(Difco)anaerobically usinganaerobic packs(BD)at
and NF-(cid:1)B activation. The effect of lactobacilli on NF- 37°C for 24 h under static condition. For cell culture
(cid:1)B, TNF and TLR4 represents one possibility that challenge,lactobacilliweregrownfroma1%inoculumin
increasestheurothelialimmunecellresponses.Thisaug- MRSbrothfor24hfollowedbywashingandresuspend-
mentation might facilitate early detection and clearance ingintheoriginalvolumewithphosphatebufferedsaline
ofpathogens. (PBS;pH7.4).UropathogenicE.coliGR12wasgrownin
AsdefinedbyFAO/WHO,probioticmicrobesmustbe Luria-Bertani (LB)medium(Difco)at 37°Cand constant
alive when administered in order to confer health bene- shaking. Heat-killed bacteria were prepared by washing
fits[33].TheinvitroeffectsonNF-(cid:1)Baugmentationhas cultures in PBS and heating at 70°C for 1 h followed by
been reported to be dependent on lactobacilli viability, plating100μlontherespectivegrowthmedium(MRSor
sinceafterheat-killingtheyonlyhadamarginaleffecton LB)toconfirmlossofviability.Heat-killedL.rhamnosus
NF-(cid:1)B activation in co-stimulation experiments with GR-1 and E. coli were stored at -20°C until used for cell
E.coli.ThissupportsmodulationofNF-(cid:1)Basapotential challenge.
probioticmechanism.Theabilityofprobioticlactobacilli
to interfere with UPEC colonization in the vagina, and Transfection and luciferase reporter assay
therebythepathogens’ascensionintothebladder,could To detect activation of NF-(cid:1)B, a luciferase vector com-
thereforeinvolveimmunomodulatoryactivity,specifically posed of multiple (cid:1)B enhancer regions followed by the
viaNF-(cid:1)Bactivation. fireflyluciferasegenewasused(pNF(cid:1)B-Luc,Clontech).A
vector with constitutively active Renilla luciferase (pRL-
Conclusions CMV,Promega) waschosenasinternal control. Oneday
The main cause of UTI is ascending E. coli that colo- prior to transfection, approximately 0.5 × 105 cells per
nizes the vagina, urethra then bladder. To remove well were seeded in a 24-well format. Transfection was
unwanted pathogens, the urothelial cells of the mucosa performed for 6 h using 1.5 μl/well of Lipofectamine
Karlssonetal.BMCMicrobiology2012,12:15 Page8of10
http://www.biomedcentral.com/1471-2180/12/15
2000 (Invitrogen), 0.54 μg/well of pNF(cid:1)B-Luc and 0.06 centrifugedtoremovecelldebrisand100μlofthesuper-
μg/well of pRL-CMV. Lipofectamine 2000 and plasmids natantorstandardwasaddedtothewellsandincubated
weredilutedinserum-freeOpti-MEM(Invitrogen)during at room temperature for 2 h. After washing five times
preparation of DNA-liposome complexes. All plasmids with PBST, 100 μl detection antibody:HRP conjugate
were isolated by an endofree plasmid isolation kit (diluted1:250inPBSwith10%heat-inactivatedFBS)was
(Macherey-Nagel) according to the manufacturer’s added to the wells and incubated for 1 h at room tem-
instructions.Luciferasewasdetectedusingthedual-luci- perature. After extensive washing (seven times using
ferase reporter assay system (Promega) and a Turner PBST),100μlofH O /3,3’,5,5’-tetramethylbenzidinepre-
2 2
TD20/20 luminometer (Turner biosystems) set to 10s paredaccordingtothemanufacturer’sinstructions(TMB
measurementwithaninitial2sdelay.Transcriptionfactor substratereagentset,BDBiosciences)wasaddedtoeach
activation was expressed as relative NF-(cid:1)B activation, well and incubated at room temperature for 30 min in
definedastheratiobetweenfireflyluciferase andRenilla thedark.Thereactionwasstopped with2NH SO and
2 4
luciferase activity. Ratios were normalized against either absorbance read at 450 nm using a Multiskan MS plate
non-stimulated control cells or cells stimulated with reader (Labsystems). Difference between means was
E.coli.The difference betweenmeanswastestedstatisti- tested statistically by using the Student’st-test, with the
callybyusingStudent’st-test,withthelimitforstatistical limitforstatisticalsignificancesettop-values<0.05.
significancesettop-values<0.05.
Quantitative polymerase chain reaction
Epithelial cell line challenge TotalRNAwasextractedusingtheNucleospinRNAIIKit
T24 bladder cells transfected with luciferase vectors (Macherey-Nagel) with a DNase treatment step. cDNA
(pNF(cid:1)B-Luc and pRL-CMV) were challenged for 24 h was synthesized from 1 μg of extracted total RNA using
in a 24-well plate format with 2 × 107 cfu/ml of viable qScriptcDNASynthesisKit(QuantaBiosciences).Quanti-
or the equivalent number of heat-killed lactobacilli tative realtimePCRwasperformedusing PerfectaSYBR
(L. rhamnosus GR-1 or GG). For activation of NF-(cid:1)B, as Green Fastmix on a Stratagene MX3000 QPCR system
well as cytokine and chemokine release, epithelial cells (Agilent Technologies) according to the manufacturer’s
were stimulated with heat-killed E. coli (108 cfu/ml). instructions. Primers were designed to bind to different
Cell culture supernatants for ELISA were collected from exons within the genesthereby avoiding riskofgenomic
challenge experiments using non-transfected cells and DNA amplification. The primers had a Tm = 60°C with
stored at -20°C until use. For qPCR, cells were stimu- the following sequences: GAPDH: 5’ CCGTCTA-
lated in the same way although all experiments were GAAAAACCTGCCA 3’ and 5’ TGTGAGGAGGGGA-
done in 6-well plates (with proportional increase in GATTCAG 3’; TLR4: 5’ CTGAGCTTTAATCCCCTG
number of cells and bacteria) for increased amounts of AGGC3’and5’AGGTGGCTTAGGCTCTGATATGC3’.
RNA. Cell viability was determined by staining dead Allreactionswererunintriplicate.Resultswereanalyzed
cells using propidium iodide followed by flow cytometry usingMxProQPCRsoftware(Agilent Technologies)and
(Cytomics FC500, Beckman Coulter). To inhibit agonist statisticswereperformedonadjustedratiosusinganon-
activation of TLR4 in T24 cells, transfected cells were parametricMann-WhitneyUtest.Thelimitforstatistical
exposed to Polymyxin B (Invivogen), which effectively significancewassettop-values<0.05.
binds to LPS and thereby inhibits TLR4 activation, at a
concentration of 50 μg/ml for 1 h prior to the experi- Immunoblot
ment and subsequently challenged with bacteria, as pre- Cellswere grown andchallenged as previously described
viously described. inasix-wellformat,andthereafterlysedusingRIPAbuf-
fer.ImmunoblottingofcelllysateontoaPVDFmembrane
Enzyme-linked immunosorbent assays (Amersham Biosciences) was performed using vacuum.
TNF, IL-6 and CXCL8 levels were determined by BD Unbound PVDF sites were blocked with blocking buffer
ELISA sets (BD Biosciences) according to the manufac- (Tris-buffered saline, TBS, containing 0.05% Tween-20
turer’s instructions. A volume of 100 μl of capture anti- and1%BSA)for1h.Blottedmembranewasincubatedin
body (diluted 1:250 coating buffer) was added to each primaryantibodysolution(anti-TLR4,cloneHTA125;BD
well of a 96-well ELISA microplate (Nunc) and allowed Biosciencesoranti-b-actin,cloneAC-15;Sigma-Aldrich)
tobindovernight at4°C.Wellswerewashedthree times resuspendedinblockingbufferataconcentrationof1μg/
with PBST (PBS pH 7.0 with 0.05% Tween-20) and ml(anti-TLR4)or10,000timesdilution(anti-b-actin)for
blocked with PBS supplemented with 10% heat-inacti- 1 h at room temperature and thereafter washed 3 times
vated FBS (HyClone) for 1 h in room temperature after for5mininwashbuffer(TBSand0.05%Tween-20).For
whichthewellswerewashedthreetimeswithPBST.Tis- visualization, the membrane wasincubated withthe sec-
sue culture medium from challenged cells was briefly ondary antibody (anti-mouse IgG HRP-conjugated, GE
Karlssonetal.BMCMicrobiology2012,12:15 Page9of10
http://www.biomedcentral.com/1471-2180/12/15
Healthcare) at a 10,000 times dilution for 1 h in room References
temperature. The membrane was washed 4 times for 5 1. SamuelssonP,HangL,WulltB,IrjalaH,SvanborgC:Toll-likereceptor4
expressionandcytokineresponsesinthehumanurinarytractmucosa.
minusing washbufferbeforethe addition ofchemilumi-
InfectImmun2004,72:3179-3186.
nescent substrate (Supersignal west pico, Pierce). Lumi- 2. CollartMA,BaeuerleP,VassalliP:Regulationoftumornecrosisfactor
nescence was detected using a photographic film (GE alphatranscriptioninmacrophages:involvementoffourkappaB-like
motifsandofconstitutiveandinducibleformsofNF-kappaB.MolCell
Healthcare). Inorderto usetheloadingcontrol antibody
Biol1990,10:1498-1506.
(anti-b-actin), the membrane was stripped using a mild 3. KunschC,LangRK,RosenCA,ShannonMF:Synergistictranscriptional
stripping agent (200 mM glycine, 0.01% (v/v) Tween-20, activationoftheIL-8genebyNF-kappaBp65(RelA)andNF-IL-6.J
Immunol1994,153:153-164.
3.5mMSDS,pH2.2).
4. LibermannTA,BaltimoreD:Activationofinterleukin-6geneexpression
throughtheNF-kappaBtranscriptionfactor.MolCellBiol1990,
10:2327-2334.
Confocal microscopy
5. HoffmannA,LevchenkoA,ScottML,BaltimoreD:TheIkappaB-NF-kappaB
Cells were grown in a 6-well format on cover slips over-
signalingmodule:temporalcontrolandselectivegeneactivation.
night and challenged as described above. The cells were Science2002,298:1241-1245.
washed twice in PBS and fixed in 4% paraformaldehyde 6. FischerH,YamamotoM,AkiraS,BeutlerB,SvanborgC:Mechanismof
pathogen-specificTLR4activationinthemucosa:fimbriae,recognition
for 10 min followed by washing twice for 5 min in PBS.
receptorsandadaptorproteinselection.EurJImmunol2006,36:267-277.
Cells were permeabilized with PBS containing 0.25% 7. CirlC,WieserA,YadavM,DuerrS,SchubertS,FischerH,StappertD,
Triton X-100 (PBST) for 10 min and washed 3 times WantiaN,RodriguezN,WagnerH,etal:SubversionofToll-likereceptor
signalingbyauniquefamilyofbacterialToll/interleukin-1receptor
with PBS prior to blocking with 1% bovine serum albu-
domain-containingproteins.NatMed2008,14:399-406.
min in PBST (PBST-BSA) for 30 min. Primary antibody 8. HunstadDA,JusticeSS,HungCS,LauerSR,HultgrenSJ:Suppressionof
(anti-TLR4, clone HTA125, BD Biosciences) was added bladderepithelialcytokineresponsesbyuropathogenicEscherichiacoli.
InfectImmun2005,73:3999-4006.
to cells at a concentration of 0.5 μg/ml in PBST-BSA
9. YinX,HouT,LiuY,ChenJ,YaoZ,MaC,YangL,WeiL:AssociationofToll-
and incubated overnight at 4°C. Cells were washed 3 likereceptor4genepolymorphismandexpressionwithurinarytract
times in PBS and thereafter incubated for 1 h at room infectiontypesinadults.PLoSOne2010,5:e14223.
10. RavelJ,GajerP,AbdoZ,SchneiderGM,KoenigSS,McCulleSL,KarlebachS,
temperature with anti-mouse FITC antibody (BD Bios-
GorleR,RussellJ,TacketCO,etal:Vaginalmicrobiomeofreproductive-
ciences) diluted in PBST-BSA at a concentration of 0.5 agewomen.ProcNatlAcadSciUSA2011,108(Suppl1):4680-4687.
μg/ml. FITC-staining was followed by washing with PBS 11. DongQ,NelsonDE,TohE,DiaoL,GaoX,FortenberryJD,VanderPolB:
Themicrobialcommunitiesinmalefirstcatchurinearehighlysimilarto
and subsequent staining of actin using Alexa555 phalloi-
thoseinpairedurethralswabspecimens.PLoSOne2011,6:e19709.
din (Molecular probes) for 30 min at room temperature. 12. GuptaK,StapletonAE,HootonTM,RobertsPL,FennellCL,StammWE:
The cells were rinsed with PBS twice and incubated InverseassociationofH2O2-producinglactobacilliandvaginal
Escherichiacolicolonizationinwomenwithrecurrenturinarytract
with a 30 nM DAPI solution for 1 min before mounting
infections.JInfectDis1998,178:446-450.
onto glass slides. Fluorescence was observed through a 13. ColladoMC,IsolauriE,SalminenS,SanzY:Theimpactofprobioticongut
Fluoview 1000 scanning confocal laser microscope with health.CurrDrugMetab2009,10:68-78.
14. ReidG,BruceAW,FraserN,HeinemannC,OwenJ,HenningB:Oral
the FV10-ASW software (Olympus).
probioticscanresolveurogenitalinfections.FEMSImmunolMedMicrobiol
2001,30:49-52.
15. VanderhoofJA:Probioticsinallergymanagement.JPediatrGastroenterol
Acknowledgements Nutr2008,47(Suppl):S38-40.
ThisworkwassupportedbyfundingfromMagnusBergvallsStiftelse,The 16. ReidG,BruceAW:Probioticstopreventurinarytractinfections:the
KnowledgeFoundationandSparbanksstiftelsenNya.Thefundingagencies rationaleandevidence.WorldJUrol2006,24:28-32.
hadnoinfluenceonthestudydesign,datacollectionandanalysis,and 17. KimYG,OhtaT,TakahashiT,KushiroA,NomotoK,YokokuraT,OkadaN,
writingandsubmissionofthemanuscript. DanbaraH:ProbioticLactobacilluscaseiactivatesinnateimmunityvia
NF-kappaBandp38MAPkinasesignalingpathways.MicrobesInfect2006,
Authordetails 8:994-1005.
1SchoolofScienceandTechnology,LifeScience,ÖrebroUniversity,70182 18. YeganegiM,LeungCG,MartinsA,KimSO,ReidG,ChallisJR,BockingAD:
Örebro,Sweden.2DepartmentofMicrobiologyandImmunology,University LactobacillusrhamnosusGR-1-inducedIL-10productioninhuman
ofWesternOntario,London,ONN6A5C1,Canada.3TheLawsonHealth placentaltrophoblastcellsinvolvesactivationofJAK/STATandMAPK
ResearchInstitute,StJosephsHospital,London,ONN6A4V2,Canada. pathways.ReprodSci2010,17:1043-1051.
19. ChanRC,BruceAW,ReidG:Adherenceofcervical,vaginalanddistal
Authors’contributions urethralnormalmicrobialfloratohumanuroepithelialcellsandthe
MKparticipatedinthestudydesign,carriedoutmajorityoftheexperimental inhibitionofadherenceofgram-negativeuropathogensbycompetitive
workandwritingofthemanuscript.NSwasresponsiblefortheqPCR exclusion.JUrol1984,131:596-601.
analysis.GRparticipatedinthestudyconceptionandrevisingofthe 20. KimSO,SheikhHI,HaSD,MartinsA,ReidG:G-CSF-mediatedinhibitionof
manuscript.JJconceivedandparticipatedinthestudydesign,coordinated JNKisakeymechanismforLactobacillusrhamnosus-induced
thestudyandwritingofthemanuscript.Allauthorsreadandapprovedthe suppressionofTNFproductioninmacrophages.CellMicrobiol2006,
finalmanuscript. 8:1958-1971.
21. ReidG,BurtonJ:UseofLactobacillustopreventinfectionbypathogenic
Competinginterests bacteria.MicrobesInfect2002,4:319-324.
Theauthorsdeclarethattherearenocompetinginterests. 22. SongJ,AbrahamSN:Innateandadaptiveimmuneresponsesinthe
urinarytract.EurJClinInvest2008,38(Suppl2):21-28.
Received:13July2011 Accepted:22January2012 23. vanBaarlenP,TroostFJ,vanHemertS,vanderMeerC,deVosWM,de
Published:22January2012 GrootPJ,HooiveldGJ,BrummerRJ,KleerebezemM:DifferentialNF-kappaB
Karlssonetal.BMCMicrobiology2012,12:15 Page10of10
http://www.biomedcentral.com/1471-2180/12/15
pathwaysinductionbyLactobacillusplantarumintheduodenumof
healthyhumanscorrelatingwithimmunetolerance.ProcNatlAcadSci
USA2009,106:2371-2376.
24. CadieuxPA,BurtonJ,DevillardE,ReidG:Lactobacillusby-productsinhibit
thegrowthandvirulenceofuropathogenicEscherichiacoli.JPhysiol
Pharmacol2009,60(6):13-18.
25. PenaJA,VersalovicJ:LactobacillusrhamnosusGGdecreasesTNF-alpha
productioninlipopolysaccharide-activatedmurinemacrophagesbya
contact-independentmechanism.CellMicrobiol2003,5:277-285.
26. PerdigonG,AlvarezS,deRuizP,HolgadoA:Immunoadjuvantactivityof
oralLactobacilluscasei:influenceofdoseonthesecretoryimmune
responseandprotectivecapacityinintestinalinfections.JDairyRes1991,
58:485-496.
27. OgawaT,AsaiY,SakamotoH,YasudaK:Oralimmunoadjuvantactivityof
Lactobacilluscaseisubsp.Caseiindextran-fedlayerchickens.BrJNutr
2006,95:430-434.
28. BackhedF,SoderhallM,EkmanP,NormarkS,Richter-DahlforsA:Induction
ofinnateimmuneresponsesbyEscherichiacoliandpurified
lipopolysaccharidecorrelatewithorgan-andcell-specificexpressionof
Toll-likereceptorswithinthehumanurinarytract.CellMicrobiol2001,
3:153-158.
29. KarlssonM,LamS,ScherbakN,JassJ:Releasedsubstancesfrom
lactobacilliinfluenceimmuneresponsesinhumanepithelialcells.
Abstractsofthe3rdSwedish-Helleniclifesciencesresearchconference;March
25-27,2010;Athens,Greece2010,341-376,Invivo.
30. SanchezB,SchmitterJM,UrdaciMC:Identificationofnovelproteins
secretedbyLactobacillusrhamnosusGGgrownindeMann-Rogosa-
Sharpebroth.LettApplMicrobiol2009,48:618-622.
31. FrendeusB,WachtlerC,HedlundM,FischerH,SamuelssonP,SvenssonM,
SvanborgC:EscherichiacoliPfimbriaeutilizetheToll-likereceptor4
pathwayforcellactivation.MolMicrobiol2001,40:37-51.
32. ShahinRD,EngbergI,HagbergL,SvanborgEC:Neutrophilrecruitment
andbacterialclearancecorrelatedwithLPSresponsivenessinlocal
gram-negativeinfection.JImmunol1987,138:3475-3480.
33. FAO/WHO:GuidelinesfortheEvaluationofProbioticsinFood.[http://
www.who.int/foodsafety/fs_management/en/probiotic_guidelines.pdf].
doi:10.1186/1471-2180-12-15
Citethisarticleas:Karlssonetal.:LactobacillusrhamnosusGR-1
enhancesNF-kappaBactivationinEscherichiacoli-stimulatedurinary
bladdercellsthroughTLR4.BMCMicrobiology201212:15.
Submit your next manuscript to BioMed Central
and 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
