Bacterial glucuronidase as general marker for oncolytic virotherapy or other biological therapies

Description

Oncolytic viral tumor therapy is an emerging field in the fight against cancer with rising numbers of clinical trials and the first clinically approved product (Adenovirus for the treatment of Head and Neck Cancer in China) in this field. Yet, until recently no general (bio)marker or reporter gene was described that could be used to evaluate successful tumor colonization and/or transgene expression in other biological therapies. Methods Here, a bacterial glucuronidase (GusA) encoded by biological therapeutics (e.g. oncolytic viruses) was used as reporter system. Results Using fluorogenic probes that were specifically activated by glucuronidase we could show 1) preferential activation in tumors, 2) renal excretion of the activated fluorescent compounds and 3) reproducible detection of GusA in the serum of oncolytic vaccinia virus treated, tumor bearing mice in several tumor models. Time course studies revealed that reliable differentiation between tumor bearing and healthy mice can be done as early as 9 days post injection of the virus. Regarding the sensitivity of the newly developed assay system, we could show that a single infected tumor cell could be reliably detected in this assay. Conclusion GusA therefore has the potential to be used as a general marker in the preclinical and clinical evaluation of (novel) biological therapies as well as being useful for the detection of rare cells such as circulating tumor cells.

Sujets

• Documents
• Savoirs
• Beta-glucuronidase
• Oncolytic virus
• Cancer
• Reporter

Informations

• Publié par: biomed
• Publié le: 01 janvier 2011
• Nombre de lectures: 6
• Langue: English
• Poids de l'ouvrage: 3 Mo

Exrait

Bacterial

glucuronidase

as

general

marker

for

oncolyticvirotherapyorotherbiologicaltherapies

Hess
etal
.

Hess
etal
.
JournalofTranslationalMedicine
2011,
9
:172
http://www.translational-medicine.com/content/9/1/172(11October2011)

Hess
etal
.
JournalofTranslationalMedicine
2011,
9
:172
http://www.translational-medicine.com/content/9/1/172

RESEARCH

OpenAccess

Bacterialglucuronidaseasgeneralmarkerfor
oncolyticvirotherapyorotherbiologicaltherapies
MichaelHess
1
†
,JochenStritzker
1,2,3
†
,BarbaraHärtl
1,2
,JuliaBSturm
1
,IvayloGentschev
1,3
andAladarASzalay
1,3,4*

Abstract
Background:
Oncolyticviraltumortherapyisanemergingfieldinthefightagainstcancerwithrisingnumbersof
clinicaltrialsandthefirstclinicallyapprovedproduct(AdenovirusforthetreatmentofHeadandNeckCancerin
China)inthisfield.Yet,untilrecentlynogeneral(bio)markerorreportergenewasdescribedthatcouldbeusedto
evaluatesuccessfultumorcolonizationand/ortransgeneexpressioninotherbiologicaltherapies.
Methods:
Here,abacterialglucuronidase(GusA)encodedbybiologicaltherapeutics(e.g.oncolyticviruses)was
usedasreportersystem.
Results:
Usingfluorogenicprobesthatwerespecificallyactivatedbyglucuronidasewecouldshow1)preferential
activationintumors,2)renalexcretionoftheactivatedfluorescentcompoundsand3)reproducibledetectionof
GusAintheserumofoncolyticvacciniavirustreated,tumorbearingmiceinseveraltumormodels.Timecourse
studiesrevealedthatreliabledifferentiationbetweentumorbearingandhealthymicecanbedoneasearlyas9
dayspostinjectionofthevirus.Regardingthesensitivityofthenewlydevelopedassaysystem,wecouldshow
thatasingleinfectedtumorcellcouldbereliablydetectedinthisassay.
Conclusion:
GusAthereforehasthepotentialtobeusedasageneralmarkerinthepreclinicalandclinical
evaluationof(novel)biologicaltherapiesaswellasbeingusefulforthedetectionofrarecellssuchascirculating
tumorcells.
Keywords:
beta-glucuronidase,oncolyticvirus,cancer,reporter,fluorescentprobe

Background
DNAintegrationintothehostgenome.Moreover,vac-
Theregainedinterestinoncolyticvirusesoverthepastciniavirusdisplaysabroadhostcellrange,rapidspread
severalyearsledtoanenormousleapinthefieldwithandahighcapacity(upto25kbp)forgeneticpayload
moreandmoreoncolyticvirusestobedescribedandofforeignDNA[3].Ofnoteandimportanceregarding
yettocome.Notonlywerethosevirusesgeneticallythesafetyofvacciniavirus,isalsoitsbillion-foldusein
alteredtoattenuatetheirvirulence,toimprovetheirhumansduringtheeradicationprogramofsmallpox,as
safetyprofileandenhancetheirtumorspecificity,butwellasthefactthatvacciniavirusisnotahuman
theyalsowereequippedwithadditionalgenesfore.g.pathogen.Ontopofthat,recombinantvacciniavirus
cytotoxins,cytokines,prodrugconvertingenzymesandstrains(rVACVs)specificallycolonizesolidtumorsin
reportergenesthatimprovedtheoverallperformanceofmicewhilenotinfectingotherorgans[4-7].Therefore,
theseviruses[1,2].Amongthose,vacciniavirusisoneofitsuseinhumanpatientswaspursuedandfirsthuman
themostpromisingcandidatesandhasseveraladvan-trialshavealreadybeencarriedoutsuccessfully[8-11].
tages:SincethislargeDNAvirusencodese.g.itsownAreliablemonitoringofsuccessfultumorcolonization
DNApolymeraseitisabletoreplicateinthecytoplasminhumanswouldhaveanenormousimpactnotonlyon
ofinfectedhostcellstherebyminimizingtheriskofclinicaltrials,butalsotopredictpossibleoutcomesof
oncolyticvirustherapy.Inthisaspect,weandothers
*Correspondence:[e-mail]
useddifferentkindofreportergenesforoptical(e.g.
1
†
DCeopnarttribmuetnetdoefqBuioalclyhemistry,Biocenter,UniversityofWürzburg,Würzburg,
[5]),orradiological(e.g.[12-15])imagingmodalities.
Germany
Thisenabledvisualizationofvirusreplicationwithinlive
Fulllistofauthorinformationisavailableattheendofthearticle
©2011Hessetal;licenseeBioMedCentralLtd.ThisisanOpenAccessarticledistributedunderthetermsoftheCreativeCommons
AatntyribmuetidoinumLi,cpernosveid(hetdtpt:h//ecroeriagtiivneaclowmormkoisns.porrogp/leircleyncsitees/db.y/2.0),whichpermitsunrestricteduse,distribution,andreproductionin

Hess
etal
.
JournalofTranslationalMedicine
2011,
9
:172
http://www.translational-medicine.com/content/9/1/172

animalmodels.However,asopticalimaginghasstrong
limitationsinpenetrationdepthandradiologicalima-
gingistimeconsumingandrequirestheneedofspecia-
lizedpersonnelandexpensiveequipment,acheapand
simplemethodwithshortturn-around-timesisurgently
needed.Inparticular,ifthismethodcouldalsobeused
inotherbiologicaltherapyapproaches.
Beta-glucuronidasescatalyzethehydrolysisofß-D-
glucuronidesintothecorrespondingD-glucuronateand
alcohol.WhilethemammalianenzymeswithapH-opti-
mumunderacidicconditions(pH4to5)havestrongly
reducedcapacityatnormal(neutral)tissuepH,the
E.
coli
enzymeencodedby
gusA
worksoptimalinthe
rangeofpH6.8to7.7[16].
Itsfirstuseasa(fusion-)reportergenewasdescribedby
Jeffersonetal.[17,18]andwasextensivelyusedinplant
physiologystudies.Inmammals,bacterialglucuronidase
wasmainlyusedinprodrugstudies,duetotheverylow
abundanceofhumanglucuronidaseinhumanserum[19].
Severalstrategiesweresuccessfullyemployed:e.g.fusion
ofcancerspecificantibody-fragmentswithbeta-glucuroni-
dase[20]ortumorselectiveexpressionoftheenzyme
usingbacteria[21]oradenoviruses[22,23].Thereporter
genepropertiesoftheenzymewerenotstudiedasexten-
sivelyinanimals.However,twoindependentapproaches
werepublishedthatlookedatthepotentialofusingbeta-
glucuronidaseasatargetstructureforradiotracersinposi-
tron-emission-tomography[24,25].Inanotherstudy,a
membrane-anchoredformofamouse-glucuronidasewas
usedincombinationwiththefluoresceindi-beta-D-glu-
curonide(FDGlcU)whichwashydrolyzedtoafluorescent
reporterthatcouldbeusedtoassessthelocationandper-
sistenceofgeneexpressioninvivo[26].
Here,weshowthatbeta-glucuronidaseincombination
withfluorogenicsubstratescannotonlybeusedfor
localizationofenzymeexpression,butalsoasageneral
biomarkerforforeignproteinexpressioninserumsam-
ples.Consequently,thedescribedtest-systemcouldbe
appliedtoallkindsofbiologicaltherapieswhichdepend
onheterologousgeneexpression.
Materialsandmethods
Forevaluationofthedescribedglucuronidaseassayitwas
necessarytoconfirmtheheterologousgeneexpressionof
thedescribedvacciniavirusstrainbyWesternblotanalysis
aswellasimmuno-stainingstudiesincellcultureand
infectedtumorsections.Theassayitself(describedinthe
“
Fluorogenicprobesanddetectionoffluorescencepro-
ducts
”
section)wastestedwithpurifiedenzymeaswellas
withsamplesfromvacciniavirusinjectedanimals.
Cellculture
HumanA549lungcancercells(ATCCNo.CCL-185)
wereculturedinRPMI-1640mediumcontaining10%

Page2of11

fetalbovineserum(FBS)and1%antibiotic-antimycotic
solution(PAALaboratories,Cölbe,Germany)under
standardcellcultureconditions(37°C,5%CO
2
).
MTH52cisderivedfromamalignantsmall-cellcanine
carcinomaofthemammarygland[27]andculturedin
DMEMsupplementedwithantibiotic-antimycoticsolu-
tionand20%FBS.
Vacciniaviruses
TheattenuatedvacciniavirusstrainGLV-1h68waspur-
ifiedaspreviouslydescribed[4].Forgenerationofcon-
trolviruses,
lacZ
and
gusA
ofGLV-1h68werereplaced
bynonrelevantgeneconstructstocreatevirusesnega-
tiveforbeta-galactosidase(rVACV-LacZ
-
)andbeta-glu-
curonidase(rVACV-GusA
-
)respectively(AdditionalFile
1,FigureS1).
Infectionofcellcultures
Twodaysbeforeinfection,cellswereseededin6-well
platesforwesternblotanalysisor12-wellplatescon-
tainingsterilecoverslipsformicroscopystudies.90%
confluentcelllayerswereeithermocktreatedor
infectedwithGLV-1h68,rVACV-LacZ
-
orrVACV-
-GusAatamultiplicityofinfection(MOI)of0.1for1h
at37°Cand5%CO
2
inmediumcontaining2%FBS.
Afterwardstheinfectionmediumwasaspiratedand
replacedbystandardcellculturemedium.
Westernblotanalysis
Fordetectionofproteins,infectedcellswereharvested
andlysedinSDSsamplebufferat6,12,24and48
hourspost-infection(hpi).Lysateswereseparatedby
10%SDS-Polyacrylamidegelelectrophoresisandsubse-
quentlytransferredontoanitrocellulosemembrane
(WhatmanGmbH,Dassel,Germany).Afterblockingin
5%skimmilkinPBS,themembranewasincubatedwith
anti-beta-glucuronidaserabbitpolyclonalantibody
(G5420,Sigma-Aldrich,Schnelldorf,Germany),anti-
beta-galactosidaserabbitpolyclonalantibody(A-11132,
MolecularProbes,Leiden,Netherlands),anti-GFPrabbit
polyclonalantibody(sc-8334,SantaCruz,Heidelberg,
Germany)oranti-beta-actinmousemonoclonalanti-
body(ab6276,Abcam,Cambridge,UK).Thefirstanti-
bodiesweredetectedusinghorseradishperoxidase-
conjugatedanti-mouse(ab6728,Abcam,Cambridge,
UK)oranti-rabbit(ab6721,Abcam,Cambridge,UK)
secondaryantibodies,followedbyenhancedchemilumi-
nescencedetection.
X-Gal/X-GlcUstainingandmicroscopystudies
Fortheanalysisofexpressionandactivityofbeta-galac-
tosidaseandbeta-glucuronidaserespectively,A549cells
wereseededoncoverslipsandinfectedwith200PFU
GLV-1h68,rVACV-LacZ
-
orrVACV-GusA
-
perwell.

Hess
etal
.
JournalofTranslationalMedicine
2011,
9
:172
http://www.translational-medicine.com/content/9/1/172

Afterincubationfor2days,cellswerefixedusing4%
paraformaldehydeandwashedtwiceinPBS.Staining
solutionsconsistedof40
μ
lX-Gal(5-bromo-4-chloro-3-
indolyl-
b
-D-galactoside,Invitrogen,Karlsruhe,Germany)
andX-GlcU(5-bromo-4-chloro-3-indolyl-
b
-D-glucuro-
nide,Invitrogen,Karlsruhe,Germany)respectivelyin
-1dimethylformamide(40mgml),ferricyanide(12mM
K
3
Fe(CN)
6
),5.2mMMgCl
2
andferrocyanidesolution
(12mMK
4
Fe(CN)
6
).Coverslipswerestainedwitheither
X-GalorX-GlcUsolutionandincubatedfor24hat37°
CbeforemountinginMowiol.Imagesweretakenwith
aZeissAxiovert200Mmicroscope.
Histologyandimmunofluorescence
Forhistologicalanalysis,snap-frozentumorswerefixed
in4%paraformaldehyde/PBSovernightat4°C.Samples
wereembeddedin5%(w/v)low-meltagarose(Appli-
Chem,Darmstadt,Germany)und100
μ
msectionswere
cutusingaLeicaVT1000SVibratome(Leica,Heer-
brugg,Switzerland)asdescribedbefore[28].Afterper-
meabilizingin0.3%TritonX-100/PBS,sectionswere
incubatedwithHoechst33342,anti-beta-glucuronidase
rabbitpolyclonalantibody(G5420,Sigma-Aldrich,
Schnelldorf,Germany)andanti-beta-galactosidase
chickenpolyclonalantibody(ab9361,Abcam,Cam-
bridge,UK)beforestainingwithCy-5-conjugateddonky
anti-rabbitandCy-3conjugateddonkyanti-chickensec-
ondaryantibodies(JacksonImmunoResearch,West
Grove,PA).Mowiol-embeddedsectionswereexamined
usingaLeicaMZ16FAStereo-FluorescenceMicro-
scopeequippedwithaLeicaDC500DigitalCamera.
DigitalImageswereprocessedwithPhotoshop7.0
(AdobeSystems,SanJose,CA)andmergedtoyield
pseudocoloredpictures.
Animalstudies
A549andMTH52cxenografttumorsweredevelopedin
6-to8-week-oldnudemice(NCI:Hsd:AthymicNude
Foxn1
nu
,HarlanBorchem,Germany)byimplanting5×
10
6
cellssubcutaneouslyintherightabdominalflank.
Twotothreeweeksafterimplantation,micewere
anesthetizedwithisofluraneandinjectedwithaviral
doseof5×10
6
PFUGLV-1h68,rVACV-LacZ
-
or
rVACV-GusA
-
in100
μ
lPBSviatheretro-orbitalsinus
evni.Bloodandurinecollectionofmicewascarriedout
underanesthesiabyaheparinisedcapillarypipet(No.
554/20,Assistent,Sondheim,Germany)viatheretro-
orbitalsinusveinforbloodandabladdercatheter(No.
381312,BectonDickinson,Heidelberg,Germany)for
urinerespectively.
Allanimalexperimentswerecarriedoutinaccordance
withprotocolsapprovedbytheRegierungvonUnter-
franken(Würzburg,Germany)(protocolnumberAZ

Page3of11

55.2-2531.01-17/08)and/orthetheInstitutionalAnimal
CareandUseCommittee(IACUC)ofExploraBIO-
LABS,locatedinSanDiegoScienceCenter(SanDiego,
USA)(protocolnumber:EB08-003).
Fluorogenicprobesanddetectionoffluorescence
products
ThelyophilizedfluorogenicprobesFDGlcU,FDGand4-
Methylumbelliferyl-b-D-glucuronide(4-MUG)(Invitro-
gen,Karlsruhe,Germany)weredissolvedinDMSO
(36.5mM).Forin-vivostudies,5
μ
lofeachstockdilu-
tionwasmixedwith195
μ
lPBSandinjectedintraperi-
toneally.Wholebodyandurinefluorescenceanalysis
wasperformedusingaMaestroEXimagingsystem
(CRI,Woburn,MA).Forserumanalysis,thecollected
mouseserumwasdiluted1:15withPBSand80
μ
lof
eachsampleweremixedwithofeither2.5
μ
gFDGlcU
or1.5
μ
g4-MUGifnototherwiseindicated.Human
serumofhealthyindividuals(Zen-BioInc,ResearchTri-
angle,NC)wasobtainedfromwholebloodand10
μ
l
wereusedinthedescribedassay.Afterincubationfor1
hat37°C(ifnototherwiseindicated),fluorescencewas
readinLumox384-wellplates(Sarstedt,Nümbrecht,
Germany)usinganInfinite200ProMicroplateReader
(Tecan,Crailsheim,Germany)oraSpectraMaxM5
(MolecularDevices,Sunnyvale,USA)andfluorescence
intensitiesarelistedasrelativefluorescenceunits.
Results
Biologicaltherapies,likee.g.oncolyticvirotherapy,are
dependendontheexpressionofgenesthatarenotfunc-
tionallyexpressedintherespectivepatient.Werecently
describedtheusetherVACVstrainGLV-1h68incom-
binationwithaprodrugseco-analogofduocarmycinSA
whichisactivatedbythevirusencodedbeta-galactosi-
dase[29].Althoughverypromisingresultswere
observedincellculture,thesynergisticeffectintumor
bearingmicewerelesspronounced,whichwasalsotrue
forabacterialglucuronidaseactivatableversionofthe
prodrug(Hessetal,unpublished).Amongotherpoten-
tialreasonsforthisobservationwasanon-favorable
pharmacokineticoftheprodrugs.Therefore,wewanted
toanalyzetheactivationandpharmacokineticsofother
beta-galactosidaseorglucuronidasesubstratesthatcould
actuallybevisualizedbyopticalimaging.
Asfluorescentprobescanbedirectlydetectedinsmall
animals,thecompoundsFDG(substrateforthebeta-
galactosidaseLacZ)aswellasFDGlcU(aglucuronidase
substrate)wereinjectedintumorbearingmicethathad
previouslybeeninjectedwithoncolyticrVACV(GLV-
1h68)encodingbothenzymes.Uponcleavagethe
probeswereconvertedtofluoresceinwhichresultedin
achangeoftheirfluorescentpropertiesthatwasreadily
detectablewithasmallanimalfluorescenceimaging

Hess
etal
.
JournalofTranslationalMedicine
2011,
9
:172
http://www.translational-medicine.com/content/9/1/172

system(Figure1A).Animalsthatwerepreviously
injectedwithPBSorthecontrolrVACVstrainsnot
-expressingbeta-galatosicase(rVACV-LacZ)andglucur-
onidase(rVACV-GusA
-
)respectivelyservedascontrols.
Itbecameevidentthateachfluorogenicprobecould
indeedbeactivatedinthetumorandthatthisactivation

Page4of11

isdependentontheexpressionofLacZandGusA
respectively(Figure1Aand1C).Themaximumfluores-
ceinfluorescenceinthetumorwasobserved120min
afterintraperitonealinjection(Figure1B).Uponintrave-
nousFDGlcU-injectionthemaximumfluorescencewas
alreadyobserved20minpostinjection(AdditionalFile

Figure1
FluorogeniccompoundactivationinrVACV-colonizedtumors
.A)TwohourspostintraperitonealinjectionofFDGandFDGlcU
tumorspecificfluoresceindependentfluorescencewasobservedinthepresenceofLacZ-andGusA-expressingvacciniavirusGLV-1h68
respectively.InrVACV-LacZ-negativeandrVACV-GusA-negativecolonizedtumors,onlyGFPfluorescencewasobserved(nodifferencein
fluorescencebetweenpre-and120minpostfluorogeniccompoundinjection).B)FluorescenceimmunohistochemicalanalysisofGLV-1h68,
rVACV-GusA-negativeandrVACV-LacZ-negativecolonizedtumors.Blue-Hoechst33342stainedDNA,green-virusencodedGFP,yellow-beta-
galactosidase,red-beta-glucuronidase.C)TumorspecificfluorescenceovertimepostFDGlcU-injection(n=4;averageplusstandarddeviationis
shown).

Hess
etal
.
JournalofTranslationalMedicine
2011,
9
:172
http://www.translational-medicine.com/content/9/1/172

2,Fig.S2).About6hourspostinjection(hpi),theGFP-
dependentfluorescenceremainedwhilemostofthe
compoundspecificfluorescencewasgone(Figure1A
and1B).
Consequentlyitwasinvestigatedhowtheproduced
fluoresceinwasremovedfromthetumor.Possible
explanationscouldbefluoresceininstabilityinorcom-
poundexcretionfromthetumor.
Infavourof(renal)excretionwasthepresenceof
fluoresceinintheurineofinjectedmice.Moreover,
whenfluoresceinitselfwasinjecteddirectlyintothe
tumorthecompoundaccumulatedinthebladderand
wasthensecretedwiththeurine(datanotshown)while
itdisappearedfromthetumor.
Thepresenceoftheactivatedprobeintheurineof
GLV-1h68injectedtumorbearingmiceofferedthepos-
sibilitytoevaluatethisasapossiblebiomarkerforsuc-
cessfultumorcolonization.Totestthis,micewere
anesthesizedandurinewasisolatedviaabladdercathe-
therbeforeand90minutesafteri.p.injectionof
FDGlcU.Indeed,fluoresceinwaspresentinurineof
GLV-1h68treatedtumorbearinganimalsbutcouldnot
beobservedinFDGlcUinjectedcontrolmicethateither
hadnon-colonizedorrVACV-GusA
-
colonizedtumors
(Figure2A,B).Therefore,onecouldestablishatestfor
thepresenceofGLV-1h68intumorsoflivemicewitha
simpleurinetestaftersystemicinjectionofFDGlcU.
Nevertheless,wedidnotproceedtoestablishaurine
basedtestsystemastheinjectionofFDGlcUinhuman
patientsseemedunlikely.
Althoughfluoresceinobviouslywasrenallyexcreted,it
wasalsoinvestigatedwhetheractiveenzymesthatwere
specificallyproducedinthetumortissueleakedoutand
werethenpresentintheserumofGLV-1h68injected

Page5of11

tumorbearingmiceleadingtoadditional(non-tumor
specific)cleavageofFDGlcU.Toinvestigatethis,the
serumoftumorbearingmicewhichwerepreviously
injectedwiththeGusAencodingstrainGLV-1h68was
incubatedwithFDGlcUor4-MUG.Bothsubstancescan
behydrolysedbyglucuronidasetothefluorescentpro-
ductsfluoresceinand4-MUrespectively.Aslatterhas
excitationandemissionmaximaof365nmand455nm
respectively,itwasnotsuitableforoptical
invivo
ima-
gingandthereforeonlyusedin
invitro
studies.Inour
experiments,nofluorescencewasobservedwhenmice
wereinjectedwithPBSorGusA-negativecontrol
rVACVwhilethefluorescentcompoundsclearlywere
detectableafterco-incubationwithserumderivedfrom
GLV-1h68injectedmice(Figure3A).Theserumthere-
forecontainedactive(non-secreted,butshed)enzymes
thatwereproducedinthetumortissue.
Theoretically,thisdoespavethewaytoasimpleblood
testthatcouldbeusedtodistinguishtumorbearing
fromhealthypatientsand/ortodistinguishbetweensuc-
cessfulfromunsuccessfultumorcolonizationofGusA-
encodingoncolyticvirusstrainsincancerbearing
patients.Toconfirmthis,weretrospectivelytested
serumsamples(n=99)fromdifferenttumorxenograft
models(GI-101A,A549,DU-145,PANC-1,HT-29)that
werecollectedinourlaboverabout4yearsfrommice
injectedwithPBS(n=33)ortreatedfordifferentperi-
odsoftime(7to53days)withseveralGusA-positive(n
=53)or-negative(n=13)rVACV(Figure3B).Thetest
oftheseprobesconfirmedasignificant(p<0.001)dif-
ferencebetweentheGusAcontaininggroupandthose
thatdidnotresultinGusA-production.Asthetestwas
doneinretrospect,wecouldnotconfirm/excludesuc-
cessfultumorcolonizationinthosemicewhichhad

Figure2
Analysisofurinesamplesderivedfrommicebefore,30minand90minpostFDGlcUinjectionrespectively
.A)Five
μ
lurine
wereanalyzedintheMaestroImagingsystem.Unmixedfluoresceinspecificfluorescenceisshown.B)ThreeandsevendayspostrVACV
injectionurinewassampledbeforeand90minpostFDGlcUinjection.Averageplusstandarddeviationoffluoresceinspecificfluorescenceof
GusA-positive(n=6for3dpiandn=8for7dpi)andGusA-negative(n=4)rVACVcolonizedtumors.*indicatesp<0.05.

Hess
etal
.
JournalofTranslationalMedicine
2011,
9
:172
http://www.translational-medicine.com/content/9/1/172

Page6of11

Figure3
Glucuronidasespecificfluorigeniccompoundactivationinserumoftumorbearingmice
.A)Tumorbearingmiceweremock
injected(n=2)orinjectedwithGLV-1h68(n=6)andrVACV-GusA-negative(n=6)respectively.Sevendayslater5
μ
lserumwasco-incubated
for1hat37°Cwith4-MUGandFDGlcUrespectivelyandsubsequentlyspecificfluorescencewasdetermined.B)Retrospectiveserumanalysis.
Serumsamples(n=99)fromdifferenttumorxenograftmodels(GI-101A,A549,DU-145,PANC-1,HT-29)wereretrospectivelytested.Samples
werederivedfrommock(n=33)injectedmiceormicetreatedfordifferentperiodsoftime(7to53days)withseveralGusA-positive(n=53)
or-negative(n=13)rVACV.*indicatesp<0.03;***indicatesp<0.001.

Hess
etal
.
JournalofTranslationalMedicine
2011,
9
:172
http://www.translational-medicine.com/content/9/1/172

GusA-negativeresultsinthebloodtest.Thiscould
explainthenegativeresultsfor5oftheGusA-rVACV
injectedtumorbearingmice.
Next,itwasinvestigatedwhetheronecouldconfi-
dentlydecideviathebloodtestifatumorwascolo-
nizedornot.Tumorbearingmice(n=5)were
5systemicallyinjectedwithalowdose(1×10PFU)of
GLV-1h68.Thisdosewasknowntoresultincoloniza-
tionofonlysometumorswhileothertumorsdonot
getcolonizedatall.Serumwasthenisolated1,3,7,10
and14dayslater,beforemiceweresacrificedand
tumorcolonizationwastestedbyconventionalplaque
assayconfirmingtheresultsobservedbytheFDGlcU/
4-MUGbasedbloodtest(Figure4A).Theverysame2
outof5micewerefoundtohaveGLV-1h68intheir
tumorsaswellasbeingpositiveintheFDGlcU/4-
MUGbasedbloodtests.
Inanotherexperimentwetestedthesuitabilityofthe
bloodtesttodifferentiatebetweentumorbearingand
controltumorfreemice.Unexpectedly,sevendayspost
injectionofGLV-1h68innon-tumorbearingmicelow
butsignificantglucuronidaseactivitywasdetectedinthe
serumoftumorfreemice.Closerexaminationofthose
micerevealedGFPexpressioninthepawsof2mice
(datanotshown).Therefore,themiceweresacrificed
andaplaqueassayofseveralorganswasusedtofind
theoriginofglucuronidaseproduction(Table1).Apart
fromthetwoinfectedpawsviruswasreproduciblyiso-
latedinlowconcentrationfromovariesofnon-tumor
bearingmice.Interestingly,ovariesoftumorbearing
micewereessentiallyfreeofvirus.
Backgroundexpressionofglucuronidaseinhealthy
subjectswouldofcourseexcludeusingthistestsystem
fordetectionoftumorsifnodifferenceexistedthat
alloweddifferentiationbetweencancerpatientsand
healthyindividuals.Timecoursestudiesinmale(n=12
tumorbearingand12tumorfree)andfemalemice(n=
24tumorbearingand6tumorfree)overaperiodof14-
16daysinwhichbloodwastakeneveryotherday(in
onehalfofthemicebloodwastakenoneven,inthe
otheronunevendayspostvirusinjection),againshowed
lowlevelsofglucuronidasepresentintheserumof
tumorfreemice.Theseweresimilartothoseobserved
intumorbearingmiceuntil8dayspostvirusinjection
(Figure4BandAdditionalFile3,Fig.S3).Afterthat
however,significantchangesoccurred.Whiletheglu-
curonidaseactivityintheserumoftumorfreemice
decreased,thefluorescencesignalinthetumorbearing
animalprobesincreased.
Takentogether,9daysafterinjectionofthevirusit
waspossibletodecidewithconfidencewhetherA)an
existingtumorwassuccessfullycolonizedand/orB)a
tumorwaspresentinthe
gusA
encodingrVACV
injectedmouse.

Page7of11

Inanextstepthesensitivityandperformanceofthe
describedtestwasevaluatedin
invitro
studiesregarding
itsclinicaltranslatability.Wefoundapositivecorrela-
tionbetweenthefluorescencesignalintensitiesand
increasingA)glucuronidaseconcentration,B)substrate
(FDGlcUor4-MUG)concentrationandC)incubation
time(AdditionalFile4,Fig.S4).Thedataalsorevealed
thatverylowglucuronidaseconcentrationscouldbe
detectedusingthefluorogenicFDGlcUor4-MUGsub-
strateswhichshouldallowdetectionoflysedtumorcells
notonlyinmicebutalsoinhumans.Feasibilityinthe
presenceofhumanserumwasshownbyrunningthe
assayinthepresenceorabsenceof10
μ
lhumanserum
(AdditionalFile5,Fig.S5).Thedatarevealedthat
neitherthesensitivitynorthefluorescenceintensityof
theassaywaschangedinthepresenceofhumanserum.
Asthepresenceofglucuronidasereliesontheproduc-
tionbyinfectedcancercells,theminimalamountof
infectedcancercellswastestedthatwasneededtogen-
erateapositivefluorescentsignal.Forthis,A549cells
wereinfectedwithGLV-1h68orcontrol-rVACV.One
daylater,theamountofinfectedcellswasdetermined
bycountingandflowcytometryandthecellswere
dilutedandseededinhalf-logdilutionsin384-well
plateswithconcentrationsvaryingfromapprox.1.0to
1000infectedcells/wellandco-incubatedwith6.3
μ
g
FDGlcUand3.4
μ
g4-MUGrespectively.Toobtainhigh
sensitivity,theprobeswereincubatedat37°Covernight
andanalyzedthenextday(Figure5).Surprisingly,even
asingleinfectedcancercellcouldbereadilydetected
usingthistestwhenusing4-MUGasasubstrate.This
allowedustocalculatethepotentialsensitivityofthe
describedtestforthedetectionoftumorsinhuman
patientsusingtwodifferentapproaches:1)Assuming
thatmoreserum(e.g.50
μ
l)wouldbeusedwhentesting
thesystemonhumanpatientsamples,andanaverage
totalbloodvolumeofabout4.7liters,onlyapprox.10
5
infectedcancercellswouldbesufficienttogeneratea
detectablefluorescentsignal.2)Thefluorescentsignal
generatedfromasinglecancercellwassimilartothat
obtainedfrom0.2unitsglucuronidase.Increasingthe
sensitivitybyaddingmorefluorescentsubstrate0.05
unitsglucuronidasewereeasilydetectable,correspond-
ingtoaconcentrationof1unitglucuronidase/mlserum
(againusing50
μ
lserum/test)or4700units/average
bloodvolumeofahumanpatient.Therefore,aslowas
2.4×10
4
infectedcancercellswouldbesufficientfora
positivesignal.
Discussion
Biologicaltherapiesincludingstemcelltherapy,gene
therapy,immunotherapy,oncolyticvirotherapyetc.are
gainingmoreandmoreimpactine.g.regenerativemed-
icine,immunology,oncologyandtreatingvarious

Hess
etal
.
JournalofTranslationalMedicine
2011,
9
:172
http://www.translational-medicine.com/content/9/1/172

Page8of11

Figure4
Signalgenerationisbothdependentonpresenceoftumorsand
gusA
-encodingrVACV
.A)Confirmationofsuccessfultumor
colonizationafterlow-doseinjectionof1×10
5
pfuGLV-1h68.Intheleftpaneltimedependent4-MUfluorescenceisshownforindividualsera
ofrVACVinjectedtumorbearingmice.Theseraoftwomice(#482and#486)resultedingenerationofsignificantamountsof4-MU.Viraltiter
analysisoftumorsfromthesamemicerevealedthatonlythosemice(#482and#486)hadviruscolonizedtumorswhichalsoresultedin
generationof4-MU.B)Tumorbearing(black)andnon-tumorbearing(grey)controlmale(leftpanel)andfemale(rightpanel)micewere
injectedwith5×10
6
pfuGLV-1h68intotheretro-orbitalsinusvein.AnalysisofserarevealedconversionofthefluorigeniccompoundsFDGlcU
(solidlines)and4-MUG(brokenlines)inallmice.Significantdifferences(p<0.05)betweentumorbearingandcontrolmicewasobservedafter
9dpi.Resultsshowaverageplusstandarddeviationoffluoresceinand4-MUfluorescence.

Hess
etal
.
JournalofTranslationalMedicine
2011,
9
:172
http://www.translational-medicine.com/content/9/1/172

Page9of11

Table1Viraldistributionintumorbearingandnon-tumorbearingmice14dpi
pfu/gtissuefluorescence
mouse#tumorbloodovariesspleenkidneysliverlungbrainpaw4-MUfluorescein
1NANDNDNDNDNDNDND2.10E+0685991179
2NAND2500NDNDNDND100ND4418607
3NAND5300NDNDNDND20ND6775855
4NAND12000ND100NDNDND5.00E+05145872641
5NAND8800NDND100NDNDND4325483
6NANDNDNDND0NDNDND5272667
72.70E+07NDND40NDND20NDND40812168977
85.15E+07NDND100NDNDNDNDND4373261137
92.65E+07NDNDND100NDNDNDND4386643572
109.00E+06ND2020100ND80NDND38449135454
118.50E+06NDNDND20NDNDNDND4164528754
NA-notapplicable.ND-notdetectable.Detectionlimit20pfu/g.

diseases.Alloftheserequiretheexpressionofcertaintherapeuticeffect:Activeenzymecanonlygetinthe
genesthatusuallyarenotoronlyweaklyexpressedinserumwhen(tumor)cellsarelysed.Futurestudieswill
thetargetedtissue.Whileprogressintermsofbiologicalshowifrespondingandnon-respondingtumorscanbe
effectivenessismadeineachfield,nocommonreporterdifferentiatedusingthedescribedtestsystem.
systemexistswhichiscost-efficient,easytouseinclini-Thedetectionofcellsurfaceassociatedbeta-glucuro-
callaboratoriesandatthesametimeallowsshortturn-nidasecouldbehelpfulinstudiesrelyingontransfection
around-times.ofcellsorfollowingbacteriaorparasiteinfectionsin
InourstudiesweusedoncolyticrVACVstrainsasawhichblood-bornepathogensexpressamembrane-or
modeltoinvestigatethepotentialofa(bacterial)beta-glu-cell-wall-anchoredglucuronidase.Also,thepreviously
curonidaseincombinationwithfluorogenicprobestobementioned(non-membranepassing)prodrugtherapies
usedasageneralreportersysteminbiologicaltherapies.wouldbenefitfromdetectionofbeta-glucuronidasein
Thebacterialglucuronidasewaschosenasreportergene,theblood:Theenzymewouldonlybeobservedupon
sincethepHinblood(pHofabout7.4)isinthepH-opti-
mumrange(pH6.8-7.7)ofthebacterialenzyme[16]
whilemammalianbeta-glucuronidasesaremostactiveat
pH4to5andpresentalmostexclusivelyinlysosomes.
Moreover,thebacterialenzymedisplaysmuchhigherspe-
cificactivitycomparedtothehumanbeta-glucuronidase
[30].Ontheotherhand,bacterialenzymesusuallyhave
muchhigherimmunogenicitythantheirmammalian
counterpartsandthusmightbehinderedintheirefficacy
[31].Consequently,thegroupofS.Rofflerselected
mutantsofhumanbeta-glucuronidasewithenhanced
activityatneutralpH[30]andfusedthisproteintosingle
chainhumanizedantibodiesforenhancedantibody-direc-
tedenzymeprodrugtherapy[32].Therefore,itshouldalso
bepossibletousethisoptimizedhumanbeta-glucuroni-
daseinthedetectionsystemdescribedherein.
Furthermore,thepossibilitytodirectactivebeta-glu-
curonidaseintothecytoplasm(thisstudyandmany
others,e.g.[17]),attachittothecellsurface(e.g.
Figure5
MinimalamountofGLV-1h68infectedcancercells
[23,33]),orsecretefromproducingcells(e.g.[32,34])
necessaryforpositivedetection
.A549
ne
c
g
ellswereinfectedwith
offersanumberofdifferentpotentialapplications.
GLV-1h68orcontrol-rVACV(rVACV-GusA).Onedaylater,the
amountofinfectedcellswasdeterminedbyflowcytometryand
Cytoplasmassociatedbeta-glucuronidaseseemstobe
thecellswereseededinhalf-logdilutionsin384-wellplateswith
averyattractivemarkerforoncolyticviruses.Notonly
concentrationsvaryingfrom1.0to1000cells/wellandco-incubated
doespresenceofglucuronidaseintheserumshowsuc-
withFDGlcUand4-MUGrespectively.Datarepresentaverageplus
cessfultumorcolonization,butalsoindicatesa
standarddeviation(n=6).