[PDF] Tumor Temporal Proteome Profiling Reveals the Immunological

View - Download Tumor Temporal Proteome Profiling Reveals the Immunological

Previous PDF

Next PDF

TumorTemporalProteome Pro!ling

RevealstheImmunological Triple

OffensiveInducedbySynthetic

Anti-CancerSalmonella

ShuxinYang

1†

,WenjuanZhao

1†

,MuchunZhu

2†

,HuijuanHu 1 ,WeijieWang 1

ZhongshengZang

1 ,MeilingJin 1 ,JiachengBi 1 ,JiandongHuang 1 ,ChenliLiu 1

XuefeiLi

1 ,PengYin 2 andNanLi 1 1

ChineseAcademy ofSciences(CAS) KeyLaboratoryfor QuantitativeEngineeringBiology, ShenzhenInstituteof Synthetic

Biology,ShenzhenInstitute ofAdvanced Technology,ChineseAcademy ofSciences,Shenzhen, China, 2

Guangdong-Hong

Kong-MacaoJointLaboratory ofHuman-Machine Intelligence-SynergySystems,Shenzhen InstituteofAdvanced Technology,Chinese AcademyofSciences, Shenzhen,China Theengineer ed"obligate"anaerobicSalmonellatyphimuriumstrainYB1sho wsa prominentabilityto represstumorgrowth andmetastasis,which hasgreatpotential as anovelcancer immunotherapy.However,the antitumormechanismof YB1remains unelucidated.Toresolve theproteomedynamics inducedbythe engineeredbacteria,we appliedtumortemporal proteomepro!lingonmurine bladdertumors afterintravenous injectionof eitherYB1or PBSasa negativecontrol.Our datasuggeststhat duringthetwo weekstreatmentof YB1injections, thecuredtumors experiencedthreedistinct phasesof theimmuner esponse.Twodaysaft erinjection,theinnateimmu nerespons ewas activated,particularly thecomplementandblood coagulationpathways.In the meantime,thephagocytosiswasinitia ted.Theprofe ssionalphagocytessuchas macrophagesand neutrophilswererecruited, especiallythein !ltrationofiNOS and CD68 cellswasenhanced. Sevendaysafter injection,substantialamount ofTcells was observedatthe invasionmarginof thetumor.As aresult,the tumorshrunk signi!cantly. Overall,thetemporalprot eomepro !lingcansystem aticall yrevealtheYB1induced immuneresponsesin tumor,showinggreat promiseforelucidating themechanismof bacteria-mediatedcancerimmunotherapy. Keywords:quantitati veproteomics,cancerimmunothera py,engineeredSalmonella,bloodcoagulation, phagocytosis,antitumor Tcellresponse

INTRODUCTION

Globallycanceris animportantleading causeofdeath, accountingfornearly 10million deathsin

2020alone (WHO).Thevast majorityofcancers aresolidtumors, whichdevelopin avariety of

organs,suchas breast,lung, colorectum,liver,prostate andbladderetc. Todate,the conventional methodsofcancer treatmentsuchas surgery,radiationtherapy andchemotherapyare stillthe preferredchoice.However, surgeryis notanef fectivetherapyfor metastaticcancerand needstobe FrontiersinImmunology |www.frontiersin.org August2021| Volume12| Article7129361

Editedby:

Yan-fengGao,

SunYat-senUniversity, China

Reviewedby:

RoslynKemp,

UniversityofOtago, NewZealand

YichuanXiao,

ChineseAcademy ofSciences(CAS),

China *Correspondence: NanLi nan.li@siat.ac.cn

Theseauthorshave contributed

equallytothis workand share!rstauthorship

Specialtysection:

Thisarticlewas submittedto

CancerImmunity

andImmunotherapy, asectionof thejournal

Frontiersin Immunology

Received:21May2021

Accepted:03August2021

Published:19August2021

Citation:

YangS,ZhaoW,ZhuM, HuH,

WangW, ZangZ,JinM,BiJ,

HuangJ,LiuC,LiX,YinPand LiN

(2021)Tumor TemporalProteome

Pro!lingReveals theImmunological

TripleOffensive InducedbySynthetic

Anti-CancerSalmonella.

Front.Immunol.12:712936.

doi:10.3389/fimmu.2021.712936

ORIGINALRESEARCH

published:19August 2021 doi:10.3389/fimmu.2021.712936 usedincombination withothertraditional therapiessuchas radiationorchemotherapy.Thep rocedureand therapeutic effectsofradiotherapyandchemotherapy areimpededby the necroticandhyp oxic regionsintumors(1).Cancer immunotherapyhademergedasastandardtr eatment with greatpotentialin cancertherapy.It includesadoptiveT cell therapy,immunecheckpoint inhibitorsand cancervaccinesetc. Theearliestcancer immunotherapycouldbe tracedbackto the bacteria-mediatedcancertherapy200yearsago(2).So far, variouski ndsofbacterialinfectionorinject ionhave been reportedtorelieve cancersymptoms orwerespeci !callyused forcance rtreatment,su chasStreptococcuspyogenes(S. pyogenes),Coley'stoxins,Clostridiumhistolyticum andthe

BacillusCalmette-Guerin(BCG) vaccine(3-7).

Recentprogress inthe!eldsofimmunology and

biotechnologyhasgenerated newinterestin themodi!cation oftumor-targetingbacteria, returningthem totheforefront of cancerresearch( 8).Manynon-pathogenic obligateanaerobes andfacultativeanaerobes havebeenshown selectivelyproliferate intumorcells possessinghypoxiaand abnormalangiogenesis. AttenuatedSalmonellaisanoutstanding exampleofone such obligateanaerobe(9-12).Inour previousst udies,syn thetic biologytechniqueswer eappliedtoconstructanengineered (SalmonellaYB1).Theengineered SalmonellaYB1actedas an obligateanaerobe,targeting thehypoxicand necroticregionsin tumorsandsigni !cantlysuppressingthe growthand metastasis ofabroad rang eofcancers,incl udingbladdertumor( 13), neuroblastoma,liver cancerandbreast cancer(14-17). Systemicadminist rationofSalmonellaYB1can effectively stimulatetheimm unesystem,resul tingintheincreased productionoftumornecrosis factor- a(TNF-a),and interferon-g(IFN-g),aswell asactivationof bothinnateand adaptiveimmunecells(14-17).Theses timulated immune responsesmightcreatea hostileenvironmentfortumor progression( 18).However,the underlyingsystemictherapeutic mechanismofSalmonellaYB1remains tobeelucidated. Theproteomicapproach isapromising techniquethatcan facilitatethesystematicchar acterizationoft heproteome dynamicsinpharmacology andinterspecies interactions(19-

21).Systematicallyrevealingthe dynamicsofthet umor

proteomeafterYB1 treatmentwillcontribute tounderstanding itsmechanism. Thequantitativeprot eomicsof tumors undergoingbacteria ltreatmentc anespeciallycontributet o understandingofthe interactionprocessbetween bacteriaand tumors.Wepro !ledthe temporalproteomeof thebladdertumor xenograftsonmice afterSalmonellaYB1injection.We then appliedalabel-free quantitativeproteomicapproach within- solutiondigestion toidentifythe differentiallyexpressed proteins bymassspectrometry analysis.Thetumors withYB1injection experiencedthreedistinctphasesofimmune responses, includingtheactivation ofcomplementand bloodcoagulation pathways,iNOS andCD68 cellsmediatedphagocytosis, and theaccumulationof Tcellsat theinvasionmargin oftumors.In summary,weperformed thesystematictemporal analysistothe proteomeofthe tumorwithYB1 treatment.Itwill bebene!cial torevealthe therapeuticmechanismof YB1andits applicationin cancerimmunotherapy inthefuture.

MATERIALSANDMETHODS

SalmonellaYB1andTumor CellsCulture

S.typhimuriumstrainYB1was cultivatedin Luria-Bertani(LB) brothc ontaining25mg/mLchloramphenicol at37°C,with shakingat220rpmoverni ght.TheYB1 cultureswere then transferredtwiceandg rownuntilthelog arithmicp hase. OD600wasmeasured todetermine thebacterialcount. The

MB49mouseb ladderca ncercelllinewas maintainedin

Dulbecco'smod i!edEagle'smedi um-highglucose(DMEM-

HG)su pplementedwith10%FBS(fetalbovinese rum),1%

streptomycinand1%penicillin.Them ediumwasren ewed everyotherday. Cellswerecultivated at37°Cin ahumidi!ed atmosphereof 5%CO 2

AnimalsandTumor TissuesCollection

Allanimal experimentswereapprov edbythea nimalcare regulationsof theInstitutionalA nimalCareandUse

CommitteeoftheShenzhenInstit uteso fAdvan ced

Technology,ChineseAcademy ofSciences. Four-tosix-week- oldfemaleC 57BL/6mice(Vital RiverLaboratoryA nimal TechnologyCo.Ltd, CHN)were subcutaneouslyinjectedwith

MB49cells(1

10 6 )inthe "ankregion.Tumor volumewas calculatedaccording tothefollowing formula:tumorvolume = length (width)2/2.When theaveragevolumes oftheMB49 tumorsreachedapproximately 200mm 3 ,theC57BL/6 micewere randomlydividedinto twogroups.One groupwasinoculated via thetailvein ofthemice withYB1(1 10 7 )dissolvedin 125mL PBS,whereasthe controlgroupwas treatedwith 125mLPBS only.Themice werekilled atdifferent timepointsafter injection. Thewho letumortissueswerewashed twicewit hice-cold contaminants,quick-frozenin liquidN 2 andstored at-80°C forproteinextractions.

ProteinPreparationand Peptide

Extraction

Atleast1 mgsamplesof tumortissueswere cutoff andlysed ina bufferthatconsist edof5mM EDTA,150mMNaCl,20mM

HEPES(pH8.5),1%SDS ,andaR ochecompleteprotease

inhibitorcocktailtabletfor1 0minonice.Followinglysi s, tissuesdebriswas furtherlysedusing sonicationundersuitable conditions.Thelysates werecentrifuged at20,000g for15min at

4°C,andthe sup ernatantswereharvested.Th eprotein

concentrationw asmeasuredbyBCAAssay Kit(Thermo FisherScienti!c,P/N23 225).Appro ximately100mgpr otein solutionwasreduced andalkylatedfor 30minat 37°Cusing

1ml0.5M TCEP,and2 ml1M CAA.Afteralkylation, quantitative

precipitationofsoluble andhydrophobicproteins fromdilute solutionswasbased ona de!nedmethanol-chloroform-water mixturemethodas describedinWessel Detal. (22).Inbrief, an aliquot(0.200ml)ofmethanol wasadded to50mlofprotein

Yangetal. ProteomicStudy onImmunotherapyMechanism

FrontiersinImmunology |www.frontiersin.org August2021| Volume12| Article7129362 sample(approximately100 mgproteins)and thesampleswere vortexed.Then, chloroform(50ml)was addedandthe samples werevortexedagain. Forphaseseparation, 150mlofwater (HPLC grade)wasadded, andthesamples werevortexedvigorously and centrifugedat 9000gfor 2minat roomtemperature.The upper phasewascarefully removedanddiscarded. Afurther 150ml methanolwasadded slowlytothe restofthe lowerchloroform phaseand theinterphase withtheprecip itatedprotein.The samplesweremixed gentlyandcentrifuged againat9000 gfor

2min atroomtemperature topelletthe protein.Thesupernatant

wasremovedand theprotein pelletwasdried underastream of airfor5 min.Theprotein pelletswerere-disolved bytheaddition of20mlofUrea buffer (8MUrea, 100mMHEPES)andvortexed fully.Theprotein pelletswerereconstituted bythe additionof

180mlof20mM HEPESandvortexed fully.For digestion,2mg

massspectrometrygrade trypsin(Promega, P/NV5280)were addedfor digestionovernightat 37°C.Thepeptide digestions werequenchedby 10mlof10% formicacid(FA).

PeptidesDesalting

Theacidifying peptidesampleswere desaltedusinga 100mg desaltingcolumn(Thermo FisherScienti !c,P/N 60108-302).In short,thedesalting columnwasactivated with1ml acetonitrile (ACN))twice,and 1ml bufferB (80%ACN,0.5% FAinwater) twice.The bufferA wasloadedtwicetoequilibratethe desalting column.Subsequently,the peptidesolution wasloadedonto the column,washedwith 1mlbuf ferAthree timesandeluted with bufferBtoaclean tube.Theeluent wasdriedcompletely ina

SpeedVaccentrifugeat 45°Candstore at-80°C.

LC-MS/MSAnalysis

Allpept ideswerereconstituted in0.1%FA(vol/vol)an d separatedononreversed-p hasecol umns(tr appingcolumn: particlesize= 3mm,C18,length =20 mm(ThermoFisher Scienti!c,P/N164535); analyticalcolumn:particle size=2 mm, C18,length =150mm(ThermoFisher Scienti!c,P/N164534)) onanUltimat e

3000RSLCnano system(ThermoFisher

Scienti!c,SanJ ose,CA, USA)coupledtoOrb itrapQ-

Exactive

HF(ThermoFisher Scienti!c).Peptide separation

wasachievedusing a120min gradient(buffer A:0.1%FA in water,buffer B:0.1%FAin80%ACN) ata"owrate of300nl/ min,thenanalyzed byOrbitrapQ-Exactive

HFina data-

dependentmode.TheO rbitrapQ-Exactive

HFmass

spectrometerwasoperated inpositiveion modewiththe ion transfertube temperaturesetat 320°C.Thepositive ionspray voltagewas2.1 kV.Full-scanMS spectra(m/z350 -2000)was acquiredintheOrbitrapwit hare solutionof60,0 00.HCD fragmentationwas performedatnormalized collisionenergyof

28%.TheMS2 automaticgaincontrol (AGC)targetwas setto

5e4witha maximuminjection time(MIT)of 50msand dynamic

exclusionwasset to45 s.

ProteomicsDataProcessing and

BioinformaticAnalysis

TheMS/MS dataweresearched againstaSwiss-Prot database (Musculusrelease-20190412and Salmonellarelease-20190426 downloadedfromUniProt) withMaxQuant1.5.3.30. Datawere searchedwitha precursormasstolerance of20 ppmanda fragmentmasstolerance of0.5Da. Searcheswereperformed withenzymespeci !cityandonly trypticpeptideswere allowedto remaininthe!naldata sets,anduptot womis-cleavag es allowed.Cysteine carboxamidomethylationwasspeci !edasa staticmod i!cation;o xidationofmethionineresiduea nd acetylation,(protein-N) wereallowedas variablemodi!cations. Reversedecoyd atabaseswereincludedforallsearches to estimatefalsed iscoveryrates.Peptide andpr otein identi!cationswerealso quanti!edand !lteredforless than

1%false-discoveryrate (FDR).

Theintensityvalues fromMaxQuant werenormalizedand

furtherprocessed usingtheVSN method.Weremoved proteins withfewerthan twosamplesin eachgroupof samplesateach timepoint. Then,themissing valueswereimputed byusingthe QRILCmethod.The Limmapackage wasusedfor determining differentiallyexpressedproteinsbetweentumor miceandtumor miceinjectedwith Salmonella.Proteinswith anaverage fold change>1andp-value<0.05werecons idereddif ferent.For proteinsofsigni !cantdif ference,theirmolecularfunctionsand associatedbiological processeswereanalyzed usingtheDAVID thresholdwasset at5%.In orderto observetheprotein change panelsatseven timepoints,we usedtheCLUSTER package(23) tocreatea k-meansclusteranalysis.

ImmunohistochemicalAnalysis

Tumortissues werecollected fromMB49-bearingmice

intravenouslyinjectedwith bacteria(1x10 7

CFUpermice) or

PBSforhematoxylin andeosin(H&E) staining.Macrophages werelabeledwith F-4/80antibody(Servicebio, GB11027).M1 subsetmacrophages weremarkedwith iNOSandCD68 antibody (Servicebio,GB11119,GB11067). Neutrophilswerelabeled with Ly-6Gantibody(Servicebio, GB11229).Thecomplement activity wasrecognizedby C3antibody(Abcam, ab200999).Tand Bcells werelabeled respectivelywithC D3antibody(Servicebio, GB13014)andC D19antibody (Servicebio,GB11061). The collagenwasstainedwi thSiriusr edstaining.Themouse primaryantibodies wasdetectedusinga goatanti-mouse secondaryantibody(Servicebio, gb111739).

Quanti!cationofthe

ImmunohistochemicalStaining

Expressionlevelsof effectorproteins aswellas theabundanceof varioustypeof immunecellswere quanti!edusing imagesof tumorsections withimmunohistochemicalstaining.Nine regionsofi nterest(ROI) weremanuallyselectedfromthe wholescanimage ofthe eachtumorunder 40xmagni!cation (about250umx 480um), andthereare threetumor-bearing miceforeach experimentalcondition.First, weadopteda color deconvolutionmodel (24)toseparate theunstainedand stained regionsintoseparate channelswiththe deconvolutionmatrixset as[0.6500.704 0.286;0.2680.570 0.776;0.711 0.4230.561]. Then,imagebinarization withanappropriate thresholdsetting (C3w/70,CD3 w/10,CD11cw/30, CD68w/20, F4/80w/70, iNOSw/70,Ly6G w/25andCD19 w/10)wereapplied toextract

Yangetal. ProteomicStudyon ImmunotherapyMechanism

FrontiersinImmunology |www.frontiersin.org August2021| Volume12| Article712936 3 theareaof differen tmarkers.Ne xt,nucleiboundarieswere detectedbyusing alevels etsegmentationand detection techniqueaspreviously described( 25).Notethat weusedthe areaof cellnucleusinstead ofcytoplasmto representthearea of cellregions,for itis moredif!culttode !nethe regionsof cytoplasm,andwehypothesizedthat thenucl eus/cy toplasm area-ratioisconsistentacro ssdiff erentregions.Finally,the areaofsta inedsig nals(S)andallcellnuc leus(N)canbe obtained,anda ratioofS toN(S/N) wasde!nedtoevaluate theexpressionlevels ofeffector proteinsorthe abundanceof varioustype ofimmunecells inthetumor-section images.

TCellSeparation andELISpotAssay

Themouse spleenwasseparated andgrindedin RPMI1640

medium,thenthe Tcellswere isolatedbymagnetic negative selectionusingthe CD8

Tcell isolationkit(Miltenyi Biotec.).

Theactivityof CD8

Tcellswas testedusing theMouseIFN- g

precoatedELISPOTKit (DAKEWE,2210005) asthedescribed in theproductmanual. 1 10 5 freshlyspleen CD8

Tcells wereplatedin triplicates

into96-wellEl ispotplatesprecoat edwithanti-mouse-IFN -g antibodyandstimulated for18h at37°Cunder 5%CO 2 with either1 10 4

MB49cellsor 1

10 5

YB1re suspendedinPBS

(positivecontrol). Asunstimulatedcontrol,cells were incubatedfor18 hincell culturemedium (backgroundvalue). ThePMAstimulation wasusedas apositivecontrol aswell.After cellremoval,plates wereincubatedwith biotinylatedantibody andstreptavidin-HRPrespectively for1h. Spotdetectionwas performedbyFluoroSpot andELISpot Reader(MabtechIRIS

StatisticalAnalysis

Allvalues areexpressedasm eans±SD.Statisticallys igni !cant differencesamongindividualtreatmentsandthe corresponding controlgroupswerede terminedbytheK olmogorov-Smirnovtest (K-StestorKStest)oranalysisofvariance(ANOVA). Experiments wereindepe ndentlyrepeatedatleastthreetimes.All analyses werecarried outusingGraph PadPrism9. Ap-value<0.05 wasconsidered tobestatistical lysigni!cant.

RESULTS

TemporalProteome Pro!lingUnraveling

DiverseTumorResponses DuringYB1

Treatment

Tobuildthe tumormodel, C57BL/6micewere subcutaneously injectedwithMB49 cellsinthe "ankregion(detail inmethod). Inbrief,when theaveragesize oftheMB49 tumorsreached approximately200mm 3 ,theC57BL/6micewererandomly dividedintotwogro upsandinject edintrave nouslywith SalmonellastrainYB1( treatmentgroup) orPBS(control group).Thetumorgrowth wasinhibited byYB1treatment (SupplementaryFigure1 ),sameas ourpreviousstudy (13). Tomapthe proteomicatlasof theSalmonellaYB1treatment bladdertumors, wecollectedthe wholetumorsin fourreplicates fromeachgroup ateachtime point.There wereseventime pointsintotal, including6hours, 12hours,24 hours,2days,3 days,7days and14days postSalmonellaYB1orPBS injection. Weapplieda label-freequantitativeproteomic approachwithin- solutiondigestionto 56identi !edsamples bymassspectrometry analysis(Figure1A).To better understandthemole cular functionsoftumor proteinsduringSalmonellaYB1treatment, GOtermenrichment analysisfor YB1proteinsdisplaying !1± log2(fold-change)revealed atotalof 454enrichedGO terms(p!

0.05,right-s idedhypergeometrictest,Bonfe rronicorrected),

with397being upregulatedand 57downregulated( Figure1B,

SupplementaryTable1 ).

Weobserveda numberofinfection-related GOtermswere

signi!cantlyenriched, aswellas theintegrin-mediatedsignaling pathwayrelat edtobacterialinvasionat 6hours( 20),an apoptoticpathwayat 24hoursand responsetovirus at2days. These!ndingsdemonstratethat YB1could invadethetumor andcausecell apoptosis,consistent withourprevious study(14). Moreover,a stronginnateimmune responseoccurredat 24 hourslastto7da ys,includ ingacute -phase response,bloo d cogulationandcomplementactivation etc.( Figure1B, SupplementaryTable1 ).Overall,our resultsarein linewith previousobservationin Salmonellabasedtumor therapy,where Salmonellademonstratesan intrinsicanti-tumor effect,largely attributedto itsimmunestimulatory activity,aswell asactivation ofbothinnate andadaptiveimmune cells(18,26).

ADynamicProteomic Atlasof theYB1

TreatedTumors

Summingupallthe5 6samples,4 ,812proteinswerequanti !edata

1%peptideF DR(fal sediscoveryr ate)(Figure2A,Dataset1in

SupplementaryTable2).At otalof 4,516proteinswereidenti!edas high-qualityIDsb yselectingthosethathavebeen measuredwithat leastoneuniquepeptide (Figure2A,Dataset2inSupplementary Table3 ).Further !lteringforproteinsidenti!edin atleast2ofthe 4 replicatesatonetime point resultedina !nall istof2,739prote ins forb ioinformaticanalyses(Figure2A ,Dataset3inSupplementary Table4 ).Repl icatecorrelationofsampl esshowedgoodconsistency andahig hdegr eeofcorr elation(R=0.85-0.93)betweent heYB1 andPBSexp eriments( Figure2B,SupplementaryFigure2). Principalcomponentanalysis(P CA)showedthatreplicates clusteredclosely andtheYB1separationfromcontrolclusters PBS(Figure2C,SupplementaryFigure 3).Overal l,ourtumor proteomicd atasetisofhighqualityandreproducibility. Afterthese!lteringsteps,weperfor medrelativeprote in quanti!cationbasedonthel og2fold changeofprotei n intensityofYB1versusPBSsampl esandp-valuefromthe

Limmatest( Figure2D,SupplementaryFigure4).This

analysisshowedthat, 1097outof the2739quanti !edproteins weresigni!cantlyupor downre gulatedleas twiseat one timepoint(p-value"0.05,|log2 foldchange|!1) (SupplementaryTa ble5).Theser esultsrevealed thatthe majorityofdif ferentialexpressionof proteinsoccurredfrom24 hoursto7 days(947of 1097).Mostnotably, at24hours and2 days,therewere 453and412 differentiallyexpressed proteins respectively(Figure2E).

Yangetal. ProteomicStudy onImmunotherapyMechanism

FrontiersinImmunology |www.frontiersin.orgAugust2021| Volume12| Article7129364

KeyImmuneResponses Revealedby

K-MeansClustering Analysis

Tobetterexhibit biologicalprocessesinduced byYB1treatment, weappliedthe k-meansalgorithm toclassifythe 1097 differentiallyexpressedtumorproteins ,accordingtotheir dynamicco-expression patterns.Weobtainedsixprotein moduleswiththe unsupervisedk-meanclustering analysis (SupplementaryTable6).Thenwe analyzeGO Biological

Processtermsineac hofthemo dules(Figure3,

SupplementaryTable7 ).

ProteinsinModule 1and2 weregraduallydownregulated, droppedtoa troughat24hours and2 daysrespectively, includingp roteinsofRAD50,CCAR2,CDK5etc.These proteinsaremainlyi nvo lvedincell-cycle, includingDNA repairandregulation ofcellgrows. Ithasrecently beenshown thatcell-cycle arecorrelatedwithSalmonellaintracellular proliferation(20).Thesuppression ofcell-cyclemay indicatea host-drivenresponsetoSalmonella.Modu le2proteinsalso participateinthe post-translationalmodi!cationofproteins, suchas dephosphorylationandubiquitination. Itappearsthat cellproliferationand proteindegradationare thepredominantquotesdbs_dbs25.pdfusesText_31

[PDF] Beno?t STROH - IMJ-PRG

[PDF] Benoit VAILLOT - European University Institute

[PDF] Benoît Virole Sauvons les garçons!

[PDF] Benoît XV, le «pape de la paix

[PDF] BENOIT XVI (avril 2005-février 2013) Elu Pape à 78 ans, 60 ans - Gestion De Projet

[PDF] BENOÎT XVI ANGELUS Bressanone Dimanche 3 août 2008 Après l

[PDF] Benoît XVI est un saint - Riposte

[PDF] benoit XVI et le concile Vatican II

[PDF] Benoit XVI et l`Islam - Vignobles

[PDF] Benoît XVI invite à écouter la Parole de Dieu - Anciens Et Réunions

[PDF] Benoit XVI Pensees sur l`Eucharistie - Saints Des Derniers Jours

[PDF] Benoit XVI – Angelus du 4 novembre 2012 - Garderie Et Préscolaire

[PDF] Benoît XVI, Angelus 10 mai 2007 - Anciens Et Réunions

[PDF] Benoît – Mon lecteur MP3 - France

[PDF] BENOIT-LAPOINTE, Christian, Finance et Investissement, mi