The Role of Host Factors in Semliki Forest

VirusInfection

LauriIlmariAureliusPulkkinen

013865957

Progradu

Master’sPrograminGeneralMicrobiology

UniversityofHelsinki,DepartmentofBiosciences&InstituteofBiotechnology

SupervisedbyDr.GiuseppeBalistreri,Docent

©LauriPulkkinen,2017

1

Tiedekunta–Fakultet–FacultyFacultyofBiologicalandEnvironmentalSciences

Laitos–Institution–DepartmentDepartmentofBiosciences

Tekijä–Författare–AuthorLauriIlmariAureliusPulkkinenTyönnimi–Arbetetstitel–TitleTheRoleofHostFactorsinSemlikiForestVirusInfectionOppiaine–Läroämne–SubjectGeneralmicrobiologyTyönlaji–Arbetetsart–LevelProGradu

Aika–Datum–Monthandyear7/2017

Sivumäärä–Sidoantal–Numberofpages52

Tiivistelmä–Referat–AbstractHost factorsplaycrucialroles invirus infections.Virusesexploitvariouscellularprocessesand are counteracted by an arsenal of host antiviral defenses. Characterization of theseinteractions is crucial for understanding the viral life cycle and developing novel antiviraltreatments. Semliki Forest virus (SFV) is a positive-strand RNA alphavirus that has beenusedasamodelvirusformultipleclinicallysignificantdiseasessuchas lethalencephalitis.The aim of this thesis was to identify host factors that affect SFV infection to betterunderstandthebiologyofSFV,andtoprovidecandidatetargetsfortherapiesagainstmoreseriousalphavirusinfections.Here I have conducted follow up studies on a previously performed genome-wide siRNAscreen thathinted thatanumberofgeneshavenovel functions inSFV infection. Iusedanautomated high-throughput imaging-based approach to confirm the roles of these hostfactorsinSFVinfection.Forcomparison,Ialsousedasimilarstrategytotestifthesegenesaffect negative-strand RNA virus infections, using vesicular stomatitis virus (VSV).Additionally,IstudiedwhetherthehostfactorsaffectingSFVinfectionsperformtheirrolesin the entry and penetration, or post-penetration steps using a previously developedendocyticbypassassay.Iidentifiedtheγ-aminobutyricacid(GABA)transporter,SLC6A13,asapotentialreceptorforSFV.IalsodescribeothernovelgenesthathaverolesinSFVorVSVinfections.Inaddition,IshowthatTNP01,RPL18,ETF1,DMN2,andGNDPA1promote,andHDAC6counteractsSFVinfection in the entry andmembrane penetration steps. Furthermore, I report that in thelater stages of the infection DDX54 boosts and EIF2B3, EIF4G1, PHB2, EDF1, DDX47, andDHX57hinderSFV.Avainsanat–Nyckelord–KeywordsHost-virusinteractions,siRNA,Automatedhigh-throughputmicroscopy,RNAvirus,SLC6A13,GAT-2,SemlikiForestvirus,VesicularstomatitisvirusOhjaajataiohjaajat–Handledare–SupervisororsupervisorsGiuseppeBalistreriSäilytyspaikka–Förvaringställe–WheredepositedViikkiScienceLibrary/YMBOlibraryMuitatietoja–Övrigauppgifter–Additionalinformation

2

Tiedekunta–Fakultet–FacultyBio-jaympäristötieteellinentiedekunta

Laitos–Institution–DepartmentBiotieteidenlaitos

Tekijä–Författare–AuthorLauriIlmariAureliusPulkkinenTyönnimi–Arbetetstitel–TitleTheRoleofHostFactorsinSemlikiForestVirusInfectionOppiaine–Läroämne–SubjectYleinenmikrobiologiaTyönlaji–Arbetetsart–LevelProGradu

Aika–Datum–Monthandyear7/2017

Sivumäärä–Sidoantal–Numberofpages52

Tiivistelmä–Referat–AbstractIsäntäsolujenproteiinitovatmerkittävässäroolissavirusinfektioissa.Viruksethyödyntäväthuomattavaa määrää isäntiensä luontaisista ominaisuuksista, ja solut puolustautuvatviruksia vastaan monin tavoin. Virusten ja isäntäsolujen välisten vuorovaikutustenselvittäminen on välttämätöntä virusten biologian ymmärtämiseksi. Näidenvuorovaikutustentunteminenonmyöstärkeäävirusinfektioitahoidettaessa.SFV (Semliki forest virus) on positiivisjuosteinen RNA-virus, joka toimii malliviruksenamonille taudeille, kuten tappavalle virusaivokuumeelle. Tässä pro gradu -tutkielmassatavoitteenani oli löytää uusia geenejä, jotka toimivat SFV-infektioissa. Akateemisenmielenkiinnon tyydyttämisen lisäksi uusien SFV-infektioihin liittyvien proteiinientunnistaminen voi auttaa uusien hoitomuotojen kehittämiseksi vakavampia virustautejavastaan.Aikaisempi genominlaajuinen siRNA-kartoitus paljasti joukon geenejä, jotka saattavatvaikuttaa SFV-infektioihin. Tässä tutkielmassa selvitin näiden vaikutusta SFV-infektioihinkäyttäen siRNA-teknologiaa sekäautomatisoituakuvantamista ja kuva-analyysiä.Vertailunvuoksi selvitin myös, kuinka nämä geenit vaikuttavat negatiivisjuosteisten RNA-virusteninfektioihin käyttäen VSV-virusta (Vesicular stomatitis virus). Tämän lisäksi selvitin,toimivatkoSFV-infektioonvaikuttavatproteiinitinfektionalku-vailoppuvaiheissa.Havaitsin,ettäγ-aminovoihappoa(GABA)kuljettavaproteiiniSLC6A13,saattaatoimiaSFV-viruksen reseptorina. Löysin myös muita SFV- ja VSV-infektioihin vaikuttavia geenejä.Huomasinmyös,ettäTNP01-,RPL18-,ETF1-,DMN2- jaGNDPA1-proteiinejatarvitaanSFV-infektion alkuvaiheissa. Tämän lisäksi selvitin, että DDX54-proteiini edistää ja EIF2B3-,EIF4G1-,PHB2-,EDF1-,DDX47jaDHX57-proteiinitestävätSFV-infektiotavaikuttamallasenmyöhäisiinvaiheisiin.Avainsanat–Nyckelord–KeywordsVirus-isäntäsoluvuorovaikutus,siRNA,Automatisoitumikroskopia,RNA-virus,SLC6A13,GAT-2,Semlikiforestvirus,VesicularstomatitisvirusOhjaajataiohjaajat–Handledare–SupervisororsupervisorsGiuseppeBalistreriSäilytyspaikka–Förvaringställe–WheredepositedViikintiedekirjasto/YleisenmikrobiologianosastonkäsikirjastoMuitatietoja–Övrigauppgifter–Additionalinformation

3

DeclarationofauthorshipThe work presented in this thesis is my own and performed as a part of my

master’s thesis project, except for the following: I amplified and titered the

SemlikiForestvirusexpressingZsGreen,andvalidatedthehostfactorsinvolved

in Semliki Forest virus infection as a part of the course “Research Project in

Virology1” (528041).Dr.GiuseppeBalistreri andDr.KirsiHellströmgrew the

BHK-21 cells, and Dr. Giuseppe Balistreri amplified the recombinant vesicular

stomatitisvirusexpressingthegreenfluorescentprotein.

4

Contents1. Listofabbreviations...........................................................................................................6

2. Introduction

2.1. Virus-hostinteractions.....................................................................................................7

2.2. SemlikiForestvirusandvesicularstomatitisvirus.......................................................9

2.3. TheroleofhostfactorsinSFVinfection.....................................................................................11

3. Materialsandmethods

3.1. Celllinesandviruses......................................................................................................13

3.2. Virusproductionandtitration........................................................................................................13

3.3. VirusinfectionsofsiRNA-transfectedcells...............................................................................14

3.4. Endocyticbypassassay..................................................................................................15

3.5. Automatedhigh-throughputimageanalysis...............................................................16

4. Results

4.1. KnockingdownpreviouslyimplicatedhostfactorsaffectsSFVandVSV

infections.........................................................................................................................17

4.2. LowpH-inducedPMfusioncanbeusedtobypassthenormalentryandpenetration

stepsofSFVinfection.....................................................................................................25

4.3. SLC6A13isneededintheearlystagesofSFVinfection.............................................27

4.4. TNP01andHDAC6affecttheentryandpenetrationstepsofSFVinfection...........30

4.5. Componentsofthetranslationmachineryaffectpenetrationandpost-penetration

stepsofSFVinfection.....................................................................................................30

4.6. SFVinfectionisaffectedbyotherhostfactorsinpenetrationandpost-penetration

steps.................................................................................................................................33

5. Discussion

5.1. HostfactorsinSFVandVSVinfections........................................................................36

5

5.2. SLC6A13isacandidatereceptorforSFV.....................................................................37

5.3. HostfactorsthataffectSFVinfection...........................................................................38

5.4. Otherhostfactors...........................................................................................................42

5.5. Methodologicalconsiderations.....................................................................................44

5.6. Conclusions.....................................................................................................................45

6. Acknowledgements..........................................................................................................45

7. References............................................................................................................................46

8. Supplementaryinformation.........................................................................................52

6

1.ListofabbreviationsAP2 Adaptorcomplex2 kb kilobase

AP2M1 Adaptorrelatedproteincomplex2μ1 KIF11 Kinesinfamilymember11

ATP Adenosinetriphosphate KPNB1 karyopherinsubunitβ1

ATP6V1B2 ATPaseH+transportingV1subunitB2 KREMEN2 Kringlecontainingtransmembrane

protein2

ATP6V1G1 ATPaseH+transportingV1subunitG1 MEM MinimumEssentialMedium

BSA Bovineserumalbumin miRNA Micro-RNA

CLTC Clathrinheavychain mRNA MessengerRNA

CSPG5 Chondroitinsulfateproteoglycan5 NRAMP2 Naturalresistanceassociated

macrophageprotein2

DDX DEAD-boxhelicase nsP Non-structuralprotein

DHX DEAH-boxhelicase PBS Phosphate-bufferedsaline

DMEM Dulbecco’sModifiedEagle’sMedium PFA Para-formaldehyde

DMN2 Dynamin-2 PHB2 Prohibitin-2

EDF1 EndothelialDifferentiationRelated

Factor1

PM Plasmamembrane

EIF2B3 Eukaryotictranslationinitiationfactor

2Bsubunitγ

PPBI CyclophilinB(human)

EIF4G1 Eukaryotictranslationinitiationfactor

γ1

Ppbi CyclophilinB(mouse)

ESCRT Endosomalsortingcomplexes

requiredfortransport

PVR Poliovirusreceptor

ETF1 Eukaryotictranslationtermination

factor1

+RNA Positive-strandRNA

FBS Fetalbovineserum -RNA Negative-strandRNA

GABA γ-aminobutyricacid RPL18 RibosomalproteinL18

GAT-2 GABAtransporter2 RT Roomtemperature

GDP Guanosinediphosphate SFV SemlikiForestvirus

GFP Greenfluorescentprotein siRNA SmallinterferingRNA

GNPDA1 Glucosamine-6-phosphatedeaminase

1

SLC6 Solutecarrierfamily6

GTP Guanosinetriphosphate SLC6A13 Solutecarrierfamily6member13

HCV HepatitisCvirus TNP01 Transportin-1

HDAC Histonedeacetylase UPF1 Regulatorofnonsensetranscripts-1

HDAC6 Histonedeacetylase6 UTR Untranslatedregion

HIV-1 Humanimmunodeficiencyvirus1 vATPase VesicularATPase

IAV InfluenzaAvirus VSV Vesicularstomatitisvirus

7

2.Introduction

2.1Virus-hostinteractions

Asobligateparasites,virusesrequirediversecellularfactorsforthecompletion

oftheir lifecycles.Virusesusehost factorstobindandentercellsaswellasto

replicatetheirgenomesandproducenewvirions.Ontheotherhand,anumber

ofhostproteinsworktocounteractviralinfection.

Toinitiatesuccessful infection,virusesneedtobindtothesurfaceoftheirhost

cells. Inanimalcells, thisprocess ismediatedbyhost’sattachment factorsand

virusreceptors,whicharecommonlycellssurfaceglycoproteinsandglycolipids

(Smith & Helenius, 2004). Attachment factors bind the virus on the plasma

membrane (PM) of the host, often via non-specific electrostatic interactions

(Grove&Marsh,2011)(Figure1A).Thisinitialbindingallowstheboundvirion

torecruitthereceptormoleculesthatinitiatetheentryprocess(Grove&Marsh,

2011)(Figure1B).Virusentryisusuallymediatedbyendocytosis,aprocessby

whichthecellisabletointernalizeportionsofthePMandextracellularsolutes,

suchasnutrientsandhormones(Marsh&Helenius,2006).Multipleendocytosis

mechanisms exist and they canbe roughlydivided to two categoriesbasedon

the volume of the endocytic vesicle and the mechanism of vesicle formation

(Doherty &McMahon, 2009). In “micropinocytosis”, such as clathrin-mediated

endocytosis,thevesicleisformedbyinvaginationsofthePMandsmallvolumes

of the extracellular medium are internalized (Doherty & McMahon, 2009).

MacropinocyticvesiclesaretheresultofPMprotrusionsthat“grab”extracellular

mediumor even other cells (phagocytosis) and then fuse back to thePMwith

their cargo (Doherty&McMahon, 2009). Different viruses, such as hepatitis C

virus (HCV) (clathrin-mediated endocytosis) or Kaposi’s sarcoma-associated

herpesvirus (macropinocytosis) are able to exploit the full variety of these

processes(Meertensetal,2006;Raghuetal,2009;Merceretal,2010).

Recruiting the endocytic machinery gives viruses multiple advantages over

penetrating the PM directly. Internalizing the entire virion prevents the

accumulationofpotentiallyantigenicviralcomponentsonthehostsurface,gives

8

the virus a convenientway to infiltrate deep into the cell, and gives the virus

access to intracellular compartmentswhere chemical cuesdestabilize theviral

particle and initiate the disassembly process known as uncoating (Marsh &

Helenius,2006).ThesecuesareeitherchangesinthepHoftheendosomeorthe

action of endosomal enzymes (Grove & Marsh, 2011). The conditions of the

endosome cause conformational changes in the viral fusion proteins, which

allows the virus to penetrate the endosomal membrane into the cytoplasm

(Grove&Marsh,2011).Thisresultsinthedeliveryoftheviralnucleocapsidinto

thesiteofuncoatingand/orreplication(Grove&Marsh,2011)(Figure1D).

Thenext step in theviral life cycle is the transcriptionand translationof viral

messengerRNAs(mRNAs),orinthecaseofpositive-strand(+RNA)viruses,the

direct translation of the viral genome. RNA viruses use their own RNA-

dependentRNApolymerases for transcriptionbut theyusehostmachinery for

translation, which leads to a translational competition between host and viral

RNAs(Walsh&Mohr,2011)(Figure1E).Thisinteractionisfurthercomplicated

by host antiviral defenses that aim to shut down translation (Walsh & Mohr,

2011). Therefore it is understandable that viruses have evolved myriad

mechanismstousurptheproteinsynthesismachineryofthehostandtokeepit

operationaldespitehostantiviralresponse(Walsh&Mohr,2011).

Once the cell is under viral control, the final stage of the infectionbegins. The

viralgenomeisreplicatedandprogenyvirionsareassembledbyexploitinghost

resourcesandstructures,suchasthePM,theendoplasmicreticulum,ortheGolgi

apparatus(Kuismanenetal,1982;Gosertetal,2003;Spuuletal,2010)(Figure1

E).Oncethenewgenerationofvirions isready, theyexit thecellusingvarious

strategies. Virusesmay, for example, use the host’s ESCRT (endosomal sorting

complexes required for transport) system and bud from the PM (Votteler &

Sundquist,2014),useanexocytosis-likepathway (Johnson&Baines,2011),or

causethelysisofthehostcell(Tollefsonetal,1996)(Figure1F).

Asvirusinfectionishighlydependentonhostfunctions,elucidatingtheinterplay

between viral and cellular factors is crucial in understanding the biology of

9

viruses. Observing host-virus interactions also plays a key role in the

development of new treatments and therapies. In addition, by following the

different stages of virus infections, it is possible to assign novel functions to

poorlycharacterizedcellularfactors.

Figure1.The lifecycleofanendocytosis-utilizingenvelopedvirus.A:Thevirusinfectionbeginswithnon-specificbindingonhostattachment factorssuchasheparansulphate.B:Non-specificbinding leads toreceptorrecruitmentC:Receptorbindingisoftenmultivalentandtriggerstheendocytosisevent.D:Theendosomecarriesthevirusdeeperintothecellusing the tubularnetwork.Endosomalenzymesand/or theacidificationof theendosomebytheactionofhostvesicular ATPases (vATPases) leads to a conformational change in the viral spike proteins. This allows a membranefusioneventtooccur,whichreleasesthenucleocapsidintothecytosol.Theviralgenomeisuncoatedandtransportedtoits replicationsite.E:ViralRNAoutcompeteshostmRNAby limitinghost transcriptionorpreventinghost translation.Thisleadstotheproductionofviralproteinsandthereplicationoftheviralgenome.F:Theviralcomponentsassembleintonucleocapsids,whichexitfromthecellse.g.bybuddingfromthePM.

2.2SemlikiForestvirusandvesicularstomatitisvirus

Semliki Forest virus (SFV) is one of the best-studied members of the genus

Alphavirus. Like other alphaviruses, it is a +RNA virus that infects both

invertebrateandvertebratehosts(Griffin,2013).SFVvirionsareenvelopedand

haveadiameterof70nm,withicosahedrallysymmetricnucleocapsids(Mancini

etal,2000).SFVentershostcellsusingendocytosisandpenetratestheendocytic

vesicle after a low pH-induced conformational change in the SFV spike

glycoproteinsleadstothefusionofviralandendosomalmembranes(Heleniuset

al, 1980; White et al, 1980; Fuller et al, 1995). Once in the cytosol, the

nucleocapsid is immediately engaged by host ribosomes, which triggers the

10

uncoatingof thenucleocapsidand leads toviral translation (Singh&Helenius,

1992).The lowpHof theendosomes (

11

2013).However,VSVfusionoccurslaterthanSFVfusionanditismediatedbya

two-step fusion program (Le Blanc et al, 2005). First the VSV nucleocapsid is

releasedtothelumenofanintra-endosomalvesicleanditentersthecytoplasm

later, following a back-fusion of the intra-endosomal vesicle (Le Blanc et al,

2005). The transcription and replication of VSV follows the general scheme of

non-segmented -RNA viruses (Lyles et al, 2013). The genome is used as a

templateforviralmRNAsanda+RNA-antigenome,whichisusedastemplatefor

thenew-RNAgenomes.

2.3TheroleofhostfactorsinSFVinfection

Previously,agenome-widesiRNAscreenwasperformedtoidentifyhostfactors

thataffectSFVinfection(Balistrerietal,2014).Thescreenimplicatedanumber

ofgenes(“hits”) (Table1).Following thescreen, therolesofUPF1,ATP6V1B2,

andATP6V1G1inSFVinfectionwerecharacterized(Balistrerietal,2014).

Table1.GenesimplicatedinSFVinfectionbythegenome-widesiRNAscreen.

Genesymbol Function GeneID* Genesymbol Function GeneID*PHB2 Intracellular

signaling11331 DDX31 DEAD-box

helicase64794

EDF1 Transcriptionalcoactivation

8721 DDX41 DEAD-boxhelicase

51428

SLC6A13 GABA&taurinetransport

6540

DDX43 DEAD-boxhelicase

55510

EIF2B3 Translationinitiation

8891

DDX47 DEAD-boxhelicase

51202

ETF1 Translationtermination

2107 DDX54 DEAD-boxhelicase

79039

EIF4G1 Translationinitiation

1981 DHX37 DEAH-boxhelicase

57647

DDX18 DEAD-boxhelicase 8886

DHX57 DEAH-boxhelicase

90957

PVR Poliovirusreceptor 5817

KREMEN2 Transmembranereceptor

79412

CSPG5 Chondroitinsulfateproteoglycan5

10675

HDAC6

Deacetylationofvariousproteins

10013

GNPDA1 Glucosamine-6-phosphatedeaminase

10007

DNM2 Endocytosis 1785

RPL18 Ribosomalprotein 6141

AP2M1 VacuolarATPaseactivation

1173

UPF1 mRNAqualitycontrol,antiviral

activity

5976

CLTC Endocytosis 1213

ATP6V1B2 Vesicleacidification

526

KPNB1 Nuclearlocalization

3837

ATP6V1G1 Vesicleacidification

9550 TNPO1 Intracellularlocalization

3842

*O’Learyetal,2016

12

The firstaimof this thesiswastoconfirmtherolesof thehits inSFV infection

and investigate if their effectwas specific to SFV. Thiswas done using siRNA-

mediated knockdown and high-throughput imaging to determine if the hits

affectedSFVandVSVinfections.Thesecondaimwastoassignaroleintheearly

(entry andpenetration), or later (post-penetration) stages of SFV infection for

theconfirmedhostfactors.Thiswasperformedusingasimilarstrategyasabove

combinedwithanassaytobypasstheendocytosisstepofSFVinfection(Whiteet

al,1980)(Figure2).

Figure 2. The aim of this thesis. Previously, a genome-wide image-based siRNA screen identifiednumerouscandidates(or“hits”)forSFVinfectionaffectinghostgenes.Inthisthesis,Ihaveanalyzedtheirrole invirus infectionsusingsmall-scale image-basedsiRNAknockdownstudies. I confirmed if thesehitshadaroleinSFVinfectionandtestediftheyaffectedVSVinfectionforcomparison.ForthegenesaffectingSFV infection, I further studied if their role was related to either the entry and penetration, or post-penetration stages of the infection. The genome-wide screen image was kindly provided by GiuseppeBalistreri.

InthisthesisIdescribenovelgenesaffectingSFVinfection,aswellasnovelhost

factorswitharoleinVSVinfection.Ialsopinpointtheeffectsofgenesinvolvedin

SFV infectiontoentryandpenetration,orpost-penetrationsteps. Inaddition, I

show that SLC6A13, a membrane-spanning γ-aminobutyric acid (GABA) and

taurinetransporter(Kristensenetal,2011;Zhouetal,2012), isneededforthe

earlyeventsofSFVinfection,buthasnoroleinVSVinfection,andisthereforea

candidatereceptorforSFV.

13

3.Materialsandmethods3.1Celllinesandviruses

HeLa cells (ATCC) were cultured using Dulbecco’s Modified Eagle’s Medium

(DMEM) (Sigma-Aldrich, D7777) supplementedwith 10% fetal bovine serum

(FBS) (Sigma-Aldrich, F9665), GlutaMAX (Gibco, 35050-061), Non-essential

Amino Acids Solution (Sigma-Aldrich, M7145), and Antibiotics & Antimycotics

Solution (Sigma-Aldrich, A5955) (HeLa growthmedium). BHK-21 cells (ATCC)

were culture using Minimum Essential Medium (MEM) (Gibco, 61100-087)

supplementedwith 10% FBS and GlutaMAX. The cellswere grown at +37 °C

witha5%CO2atmosphere.

SFVexpressingZsGreenproteinfusedwiththevirusnsP3(SFV-ZsGreen)(Spuul

et al, 2010) and rVSV-GFP, a VSV strain expressing green fluorescent protein

(GFP)(Pelkmansetal,2005)havebeenpreviouslydescribed.Theviruseswere

originally produced at ETH Zurich (Switzerland) and were provided by Dr.

BalistreriwithpermissionfromprofessorAriHelenius.

3.2Virusproductionandtitration

VirusinoculawerepreparedinMEMsupplementedwith20mMHEPES(pH7.2),

GlutaMAX, and0.2%bovine serumalbumin (BSA)andconfluentBHK-21cells

were washed twice with phosphate-buffered saline (PBS) and infected with

eitherSFV-ZsGreenorrVSV-GFPusinganMOIof0.01byreplacingtheoldmedia

withthevirusinocula.Theinfectedcellswereincubatedat+37°Cwitha5%CO2

atmosphere.Themediawerecollectedafter22handcentrifugedat3900rpmfor

20 min (SFV-ZsGreen) or 10 min (rVSV-GFP) at +4 °C (Eppendorf Centrifuge

5810R)toeliminatecelldebris.Thesupernatantswerecollected,aliquotedand

storedat-80°C.

HeLacellswereseededontoblackclear-bottom96-wellplates(Corning,07-200-

568)atadensityof10000cells/wellandgrownfor16–20hat+37°Cwitha

5% CO2 atmosphere. The cells were washed with DMEM supplemented with

GlutaMAXandAntibiotics&AntimycoticsSolution(SFV-ZsGreen)orRPMI-1640

14

medium(ICN,1060122)containing20mMHEPESpH7.0andGlutaMAX(rVSV-

GFP) (100μl /well). The cellswhere then infectedwith1:2 (SFV-ZsGreen) or

1:10 (rVSV-GFP) serial dilutions in duplicate using 100 μl of virus inocula in

correspondingmediaTheinfectedcellsweregrownfor6h(SFV-ZsGreen)or7h

(rVSV-GFP)at+37°Cwitha5%CO2atmosphere.

After the incubation, themediawere aspirated and the cellswere fixed for20

minatroomtemperature(RT)using4%para-formaldehyde(PFA)inPBS(100

μl/well).Thefixedcellswerewashed3timeswithPBS(100μl/well).Thecells

werepermeabilizedandthenucleistainedbyincubatingthecellsfor10minat

RTin100μlofPBScontaining0.2%TritonX-100andHoechstDNAstain(1μg/

ml)perwell.Thecellswerethenwashed3timeswithPBS(100μl/well)andthe

plates covered with Black TopSeal-A plate seals (PerkinElmer, 6050173) and

storedat+4°C.Theplateswere imagedwithahigh-contentCellinsight Imager

microscope (Thermo Fisher) at the Light Microscopy Unit, Institute of

Biotechnology.16 imageswere takenperwell, usingboth the386nmand the

485nmfilterstovisualizethefluorescencesignaloftheHoechstandtheZsGreen

orGFP.TheimageswereanalyzedusingtheopensourceCellprofiler2software

(Carpenteretal,2006,www.cellprofiler.com)(seebelow).

3.3VirusinfectionsofsiRNA-transfectedcells

HeLa cells were reverse transfected using pooled siRNAs (Dharmacon

SMARTpoolsiRNAs)onablackclear-bottom96-wellplateusingaseparatewell

for each siRNA pool. The siRNA pools contained a mixture of 4 different

oligonucleotides against non-overlapping regions of each target gene in

equimolarconcentrations.ThesiRNApoolstargetedthegenesofinterestaswell

ascontrolgenes(KIF11,PpbiandPPBI)(tableS1).ThesiRNApoolagainstKIF11

wasusedasatransfectioncontrol.Sincetheproductofthisgeneisessentialfor

cell survival, monitoring cell death was used as a means to make sure the

transfectionwasefficient(>98%celldeathindicatedsuccessfultransfection).A

mixof4differentnon-specific (or “scrambled”)siRNAswasusedasanegative

controlinfourseparatewells.ThesiRNAsagainstPpbiandPPBIweresupplied

bythemanufactureraseasilyquantifiabletransfectioncontrols.

15

Foreachwell,10μlofasiRNAstock(100nM)wasmixedwith10μlofDMEM

containing0.1μl ofLipofectamine2000 (Invitrogen,11668-019)and theplate

was incubated at RT for 30–60 min. Then for each well, 2000 HeLa cells

suspended in 80 μl ofDMEMweremixedwith the siRNA–Lipofectamine 2000

mixyieldinga finalsiRNAconcentrationof10nMperwell.Theplatewasthen

incubatedat+37°Cwitha5%CO2atmospherefor6h,afterwhich,themedium

wasreplacedwithHeLagrowthmedium(200μl/well)andthecellsweregrown

at+37°Cwitha5%CO2atmosphere.

After72h theoldmediawereremovedand thecellswerewashedwitheither

DMEM supplemented with GlutaMAX and Antibiotics & Antimycotics Solution

(SFV-ZsGreen)orRPMI-1640medium(ICN,1060122)containing20mMHEPES

pH7.0andGlutaMAX (rVSV-GFP) (100μl /well).Thecellswere then infected

with100μl/wellofvirusinoculacontaining6*104pfuofeitherSFV-ZsGreenor

rVSV-GFPincorrespondingmedia.Thecellswereincubatedat+37°Cwitha5%

CO2atmospherefor5h(SFV-ZsGreen)or6.5h(rVSV-GFP)andtheplateswere

fixed, stained, imaged, and analyzed as above. Three biological repetitions per

viruswereperformed.

3.4Endocyticbypassassay

HeLacellswerereverse transfectedasabove.72hafter transfection, theplate

wasplacedoniceandcellswashedwithice-coldRPMI-1640mediumcontaining

20mMHEPES,pH7.0(100μl/well)andinfectedonicewith6*104pfuofSFV-

ZsGreen diluted in the samemedium (ice-cold) (100 μl /well). The platewas

incubatedonicefor1handthemediumwasremovedfromthewells.Toallow

virus fusion with the PM, the cells were treated with RPMI-1640 medium

containing 10 mM MES buffer, pH 5.5 for 90 s at +37 °C. The medium was

aspirated and replaced with HeLa growth medium supplemented with 20mM

NH4Cland20mMHEPES,pH7.2(200μl/well).Theplatewasincubatedat+37

°Cwith5%CO2atmospherefor4handtheplatewasfixed,stained,imaged,and

analyzedasabove.Controlsincludedwellsthatwerenotinfected,wellstreated

withpH7.0mediuminsteadofthepH5.5medium,andwellstreatedwithpH7.0

16

medium and HeLa growth medium instead of the pH 5.5 medium and HeLa

growth medium with NH4Cl, respectively. Three biological repetitions were

performed.

3.5Automatedhigh-throughputimageanalysis

The Cellprofiler 2 software (Carpenter et al, 2006, www.cellprofiler.com) was

usedtodeterminethepercentageof infectedcells ineachwell. Ineach image,

thenumberofcellswasdeterminedbydetectingthe386nmfluorescencesignal

ofthestainednuclei,designated“primaryobjects”(Figure3A&B).SFV-ZsGreen

producesZsGreen-labelednsP3,thusZsGreen-expressingcellsareinfectedwith

SFV. For the SFV experiments the perimeter of each detected nucleus was

digitally expanded to include a portion of the cytoplasm around the nucleus,

designated“secondaryobjects”(Figure3C,D,E&F).Thiswasdonetodetectthe

ZsGreensignalfromthecytoplasm,wheretheSFVproteinsynthesisoccurs.The

mean fluorescence signal on the 485nm channel (ZsGreen)wasmeasured for

each secondary object. For the rVSV-GFP-infected cells themean fluorescence

signalon the485nmchannel (GFP)wasmeasuredwithineachprimaryobject

(nucleus), as the GFP protein encoded by the virus was synthesized in the

cytoplasmand freelydiffused into thenucleusof thehost indicatingsuccessful

viraltranslation.

Usingathresholdvalueofmean485nmfluorescence,thecellswereclassifiedin

two categories: infected (above threshold) andnon-infected (below threshold)

(Figure4G).Non-infectedcellswerealwaysincludedtocalibratethethreshold

settings for each imagedplate.Once all images in eachwell of a 96well plate

wereanalyzed,thefinalresultswereexpressedastotalinfectedcellsperwell.To

calculatevirustiters, theestimatedamountofcellsperwell(16,000)(Rafferty,

1985) was multiplied with the percentage of infected cells to determine the

amountofinfectiousvirionsperwell.InthesiRNAexperiments,cellstransfected

with non-targeting (or “scrambled”) siRNAs were used as controls. Themean

numberofinfectedcellsinthescrambledcontrolswassetas1andtheinfection

percentages in the remaining transfected wells were normalized accordingly.

The mean infection percentage per gene was calculated from three biological

17

repetitions. A difference of 2 or more standard deviations from controls was

consideredtobesignificant.

Figure3.AutomateddetectionofinfectedcellsusingCellprofiler2.A&B:Fluorescencesignaldetectedusing395nm(blue)and485nm(green)filtersfromSFV-infectedcellspre-treatedwithscrambledcontrolsiRNA. C: Automatically detected nuclei (pseudocoloured 'primary objects'). E–F: Pseudocolouredsecondaryobjectsobtainedbydigitallyexpandingtheprimaryobjects(D)andsubtractingtheareaoftheprimaryobjects.G:Thesecondaryobjectswereusedtoautomaticallyclassifyinfected(red)ornon-infected(blue)cells,usingthemeanfluorescenceintensityfromthe485nmchannel(G).

4.Results

The aim of this thesiswas to study the effects of knocking down host factors

previously implicated in SFV infection, and compare those effects to VSV

infection. The genes affecting SFV infection were further characterized by

studying if theyaffectentryandpenetration,orpost-penetrationstagesofSFV

infection.

4.1KnockingdownpreviouslyimplicatedhostfactorsaffectsSFVandVSV

infections

To determine which of the previously implicated genes affect SFV and VSV

infections, I used siRNA-mediated knockdown, high-throughput imaging and

automated image analysis. I found that both SFV and VSV infection could be

reducedsignificantlybyknockingdownATP6V1B2andATP6VG1,twosubunits

ofavacuolarATPaseknowtoaffectSFVinfection(Balistrerietal,2014)aswell

18

as ribosomalproteinL18 (RPL18), apart of the60S ribosomal subunit (de la

Cruzetal,2015)(Figures4&5).BothSFVandVSVinfectionswereaffectedby

the knockdown of transportin-1 (TNP01), an intracellular transport molecule

(Twyffels et al, 2014), prohibitin 2 (PHB2), an intracellular signaling protein

(Bavellonietal,2015),andDEAH-boxhelicase(DHX)37,apoorlycharacterized

RNA-helicase (Gene ID: 57647) (Figures 4 &5). However, the knockdown of

these genes had the opposite effects on SFV and VSV infections, as TNP01

depletiondecreasedSFVinfectionandincreasedVSVinfectionsignificantly,and

in the case of PBH2 and DHX37 depletions SFV infection increased and VSV

infectiondecreasedsignificantly(Figures4&5).

Figure4.RelativeinfectionpercentagesincellsdepletedofhostfactorsthataffectbothSFVandVSVinfections. The relative infection percentages of siRNA-treated HeLa cells infected with SFV-ZsGreen(black)orVSV-GFP (white) are indicatedon they-axis.The siRNAsusedare indicatedon thex-axis.Theinfectionpercentageswerenormalizedbysettingthemeanofthescrambledcontrolsas1.Theerrorbarsrepresentstandarddeviationsof the treerepetitionsandanasterisksignifiesaresult that issignificantlydifferentfromthecontrol.

19

Figure 5. Representative images of cells depleted of host factors that affect both SFV and VSVinfections.A:HeLacellsinfectedwithSFV-ZsGreen.NucleiareshowninblueandnsP3-ZsGreen,indicatingSFV infection, is shown in green. The siRNAs used are indicated by the text in the images.B:HeLa cellsinfectedwithVSV-GFP.NucleiareshowninblueandGFP,indicatingVSVinfection,isshowningreen.ThesiRNAsusedareindicatedbythetextintheimages.Bydepletingcellsofsolutecarrierfamily6member13(SLC6A13),aGABAand

taurinetransporter(Kristensenetal,2011;Zhouetal,2012),GNPDA1,apoorly

characterized glucosamine-6-phosphate deaminase (Wolosker et al, 1998),

DEAD-boxhelicase(DDX)54,anRNA-helicasethathasatranscription-regulating

roleincells(Rajendranetal,2003;Kannoetal,2012),anddynamin-2(DMN2),

a protein involved in endocytosis (Kasai et al, 1999), SFV infection could be

significantly reduced (Figures 6 & 7). On the other hand, the knockdown of

eukaryotic translation initiation factor 2B subunit γ (EIF2B3) and eukaryotic

translation initiation factor γ 1 (EIF4G1) (Walsh & Mohr, 2011), eukaryotic

translation termination factor 1 (ETF1) (Taylor et al, 2012), UPF1, an RNA-

helicasethatpreventsSFVinfection(Balistrerietal,2014),DDX47andDHX57,

20

predicted RNA helicases (Gene IDs: 51202 and 90957, respectively), histone

deacetylase 6 (HDAC6), a multifunctional cellular deacetylase (Hubbert et al,

2002),andendothelialdifferentiation-related factor1(EDF1),a transcriptional

co-activator(Kabeetal,1999) increasedSFVinfectivitysignificantly(Figures6

&7).ThedepletionoftheseSFV-affectinggeneshadnosignificanteffectonVSV

infection(Figures6&7).

Figure 6. Relative infection percentages in cells depleted of host factors that affect only SFVinfection.TherelativeinfectionpercentagesofsiRNA-treatedHeLacellsinfectedwithSFV-ZsGreen(black)orVSV-GFP(white)areindicatedonthey-axis.ThesiRNAsusedareindicatedonthex-axis.Theinfectionpercentageswerenormalizedbysettingthemeanofthescrambledcontrolsas1.Theerrorbarsrepresentstandarddeviations of the tree repetitions and an asterisk signifies a result that is significantly differentfromthecontrol.

21

Figure 7. Representative images of cells depleted of host factors that affect only SFV infection. A:HeLacellsinfectedwithSFV-ZsGreen.NucleiareshowninblueandnsP3-ZsGreen,indicatingSFVinfection,isshowningreen.ThesiRNAsusedareindicatedbythetextintheimages.B:HeLacellsinfectedwithVSV-GFP.Nucleiareshown inblueandGFP, indicatingVSV infection, is shown ingreen.ThesiRNAsusedareindicatedbythetextintheimages.

VSV infectivity was significantly increased by the depletion of the poliovirus

receptor (PVR) (Mendelsohn et al, 1989), a Wnt/β-catenin-signaling receptor,

kringlecontainingtransmembraneprotein2(KREMEN2)(Maoetal,2002),and

adaptorrelatedproteincomplex2μ1(AP2M1),anadaptorofendocyticvesicles

and vATPases, (Heinaman, 1995) (Figures 8 & 9). Knocking down DDX18, a

poorly understood RNA-helicase (Dubaele & Chène, 2007) and karyopherin

22

subunitβ1(KPNB1),asubunitofanuclearimportprotein(Görlichetal,1995)

decreasedVSVinfectivitysignificantly(Figures8&9).Curiously,siRNAsagainst

bothmouseandhumancyclophilinB(PpibandPPBI,respectively)suppliedas

transfectioncontrolsbythemanufacturerincreasedVSVinfectivitysignificantly

(Figures 8 & 9). Cyclophilin B is a multifunctional signaling and anti-

inflammatoryprotein(Hoffmann&Schiene-Fischer,2014).Noneofthesegenes

hadasignificanteffectonSFVinfectivity(Figures8&9).

Figure 8. Relative infection percentages in cells depleted of host factors that affect only VSVinfection.TherelativeinfectionpercentagesofsiRNA-treatedHeLacellsinfectedwithSFV-ZsGreen(black)orVSV-GFP(white)areindicatedonthey-axis.ThesiRNAsusedareindicatedonthex-axis.Theinfectionpercentageswerenormalizedbysettingthemeanofthescrambledcontrolsas1.Theerrorbarsrepresentstandarddeviations of the tree repetitions and an asterisk signifies a result that is significantly differentfromthecontrol.

23

Figure 9. Representative images of cells depleted of host factors that affect only VSV infection. A:HeLacellsinfectedwithSFV-ZsGreen.NucleiareshowninblueandnsP3-ZsGreen,indicatingSFVinfection,isshowningreen.ThesiRNAsusedareindicatedbythetextintheimages.B:HeLacellsinfectedwithVSV-GFP.Nucleiareshown inblueandGFP, indicatingVSV infection, is shown ingreen.ThesiRNAsusedareindicatedbythetextintheimages.

Thedepletionofsomeofthegenesimplicatedbythegenome-widesiRNAscreen

didnothaveasignificanteffectonSFVorVSVinfection(Figures10&11).These

included chondroitin sulfate proteoglycan 5 (CSPG5), a cell surface protein

expressed in the brain (Watanabe et al, 1995), DDX31, an RNA-helicase

implicatedinrenalcancer(Fukawaetal,2012),DDX41,anRNAhelicasewitha

role in cellular immunity (Jiang et al, 2017), DDX43, a RNA and DNA helicase

(Tanuetal,2017),andclathrinheavychain(CLTC),awell-studiedendocytosis

molecule(Doherty&McMahon,2009).

24

Figure10.Relative infectionpercentages in cellsdepletedofhost factors thatdonotaffectSFVofVSV infections.Therelative infectionpercentagesofsiRNA-treatedHeLacells infectedwithSFV-ZsGreen(black)orVSV-GFP (white) are indicatedon they-axis.The siRNAsusedare indicatedon thex-axis.Theinfectionpercentageswerenormalizedbysettingthemeanofthescrambledcontrolsas1.Theerrorbarsrepresentstandarddeviationsof the treerepetitionsandanasterisksignifiesaresult that issignificantlydifferentfromthecontrol.

25

Figure 11. Representative images of cells depleted of host factors that do not affect SFV of VSVinfections.A:HeLacellsinfectedwithSFV-ZsGreen.NucleiareshowninblueandnsP3-ZsGreen,indicatingSFV infection, is shown in green. The siRNAs used are indicated by the text in the images.B:HeLa cellsinfectedwithVSV-GFP.NucleiareshowninblueandGFP,indicatingVSVinfection,isshowningreen.ThesiRNAsusedareindicatedbythetextintheimages

4.2LowpH-InducedPMfusioncanbeusedtobypassthenormalentryand

penetrationstepsofSFVinfection

To determine if depleted host factors play a part in the early stages (entry or

penetration) of SFV infection, these steps were circumvented by artificially

inducingviralenvelopefusionatthePM(Whiteetal,1980).Thiswasachieved

by allowing the viruses to bind cells on ice and treating the cells quicklywith

acidicmedium (pH5.5).This causes a conformational change in the SFV spike

26

proteins, which leads to membrane fusion, delivery of nucleocapsid into the

cytoplasm,anduncoatingofthenucleocapsid.Sincewithinthe90secondoflow-

pH treatment, someSFVvirions are internalizedby endocytosis, after the acid

treatment cells were incubated in NH4Cl-containing medium at pH 7.2 which

neutralized the acidic pH of endosomes preventing these viruses from fusing

(Figure12).ComparedtothenormalrouteofSFVinfection,theendocyticbypass

wasslightlylessefficient,restoringabout60%ofinfectedcells.Ifthecellswere

nottreatedwithacidicmedium,butincubatedinNH4Cl-containingmedium,the

infectionpercentagedroppedtothelevelofnon-infectedwells(Figure13).

Figure 12. Endocytic bypass assay.A:NormallySFVbinds to theplasmamembraneand is internalizedusingreceptor-mediatedendocytosis.TheendosomeprogressivelyacidifiesandapHlowerthan6.0causesaconformationalchangeinSFVspikeproteins,whichtriggersmembranefusionbetweenSFVenvelopeandthe endosomal membrane, freeing the nucleocapsid into the cytosol. The endosomal pathway can bebypassedbyallowingSFVvirionstobindtothehostPMinthecold,andsubsequentlytreatingtheinfectedcells with acidic medium for a short time. This leads to membrane fusion, and the release of the SFVnucleocapsid intothecytoplasm.Topreventendocytosedvirusesfromcompletingtheir infectionthroughtheendosomalroute,theacidicmediumisreplacedwithamediumcontainingNH4Cl.SomeoftheNH4+ionsare converted into NH3, a gas that can freely diffuse into the endosome and neutralize its acidification(Whiteetal,1980).B:Intheendocyticbypassassay,SFVvirionsareallowedtobindhostcellsonicefor60minutesandthevirus-containingmediumisremoved.Cellsaretreatedbriefly(90s)withacidicmedium,whichitisthenreplacedwithgrowthmedium(pH7.2)withorwithoutNH4Cl.Cellsareincubatedfor4hat+37°Candthenfixedusing4%PFA.

27

Figure13.HeLacellscanbeinfectedbyusingendocyticbypass.A:TherelativeinfectionpercentagesofHeLa cells infectedwith SFV-ZsGreen from two repetitions are indicated on the y-axis. The indicated pHvaluesrefertoinfectionwithnofusiononthePMbutincubationwithNH4Cl-containingmedium(pH7.0)and infection with PM fusion and incubation with NH4Cl-containing medium (pH 5.5). The infectionpercentageswerenormalizedbysettingthemeanof thenormal infectionsas1.Theerrorbarsrepresentstandard deviations of the two repetitions. B: Representative images of HeLa cells infected with SFV-ZsGreen.NucleiareshowninblueandnsP3-ZsGreen,indicatingSFVinfection,isshowningreen.Thetextintheimagesindicatesdifferentinfectionconditions(seeA).

4.3SLC6A13isneededintheearlystagesofSFVinfection

SLC6A13,alsoknownasGAT-2,isaGABAandtaurinetransporterthatcontains

12hydrophobicmembrane-spanningdomains,11loopregionsandcytoplasmic

NandCtermini(Kristensenetal,2011;Zhouetal,2012).Onthe3–4loopthere

arethreeN-glycosylatedresidues(Figure14)(Kristensenetal,2011).ThisPM

protein belongs to the solute-carrier 6 (SLC6) gene family, which contains

membrane proteins that transport neurotransmitters by utilizing Na+ and Cl-

ions(Kristensenetal,2011).AsitseemstoberequiredforSFVinfection,Itested

ifitisneededforentryandpenetrationstepsofinfection.ByinfectingHeLacells

using endocytotic bypass, I could revert SFV infectivity back to the level

observed in the scrambled controls (Figure 15). For further controls I used

siRNAsagainstATP6V1B2,ATP6V1G1,andUPF1.ATP6V1B2andATP6V1G1are

known to be critical for the penetration of SFV, and UPF1 works against SFV

28

infectioninthepost-penetrationstages(Balistrerietal,2014).Endocyticbypass

assay counteracted the effect of the siRNAs against ATP6V1B2 andATP6V1G1

(Figure15).Duetohighvariation,theSFVinfectivityonUPF1-depletedcellswas

not statistically different from controls in the endocytic bypass. However, the

meaninfectivitywassimilartothenormalinfectionassay(1.5-foldhigherthan

in controls) (Figure 15). Additionally, I used endocytic bypass to pinpoint

whether DDX54, an RNA helicase, affects the entry and penetration, or post-

penetration stages of SFV infection. The SFV infection inDDX54-depleted cells

didnotreverttothelevelofthecontrolswhenusingtheendocyticbypassassay

(Figure15).TheATP6V1B2,ATP6V1G1,andUPF1bypassresults,combinedwith

thefactthattheinfectivityinDDX54-depletedcellsremainedsignificantlylower

thanincontrolsshowthatendocyticbypassdoesnotreverttheeffectsofgene

knockdowninanunspecificmanner.Therefore,itcanbeusedtodistinguishthe

hostfactorsthatareneededfortheentryandpenetrationfromtheonesneeded

inthepost-penetrationstages.

Figure14.AschematicviewofSLC6A13.SLC6A13isamembraneproteinthatformsatransportchannel

forGABAandtaurineatthePM(adaptedfromKristensenetal,2011).

29

Figure 15. The roles of SLCA13 andDDX54 in SFV infection. A:The relative infectionpercentagesofsiRNA-treatedHeLacellsinfectedwithSFV-ZsGreennormally(black)orusingendocyticbypass(grey)areindicated on the y-axis. The siRNAs used are indicated on the x-axis. The infection percentages werenormalized by setting the mean of the scrambled controls as 1. The error bars represent standarddeviations of the tree repetitions and an asterisk signifies a result that is significantly different from thecontrol. B: HeLa cells infected with SFV-ZsGreen normally or using endocytic bypass (indicated on theright).NucleiareshowninblueandnsP3-ZsGreen,indicatingSFVinfection,isshowningreen.ThesiRNAsusedareindicatedbythetextintheimages.

30

4.4 TNP01 and HDAC6 affect the entry and penetration steps of SFVinfection.TNP01 is a cellular transportmolecule implicated innuclear transport,mitotic

spindleassembly,nuclearenvelopeassembly,ciliary import,andthe formation

ofPbodiesandstressgranules(Twyffelsetal,2014).HDAC6ismemberofthe

histone deacetylase (HDAC) family, but unlike other HDACs, it has also non-

histonictargetsandhasaubiquitin-bindingzinc-fingerdomain(Lietal,2013).

BothTNP01andHDAC6haverolesincellularstressresponses(Kawaguchietal,

2003;Twyffelsetal,2014)andseemtohaveroleinSFVinfection,eventhough

theeffectofHDAC6isminuscule. Intheendocyticbypassassay,SFVinfectivity

was not significantly different from the controls in both TNP01 and HDAC6-

depletedcells(Figure16).Thisindicatesthattheirrolesarerelatedtotheearly

stagesofSFVinfection.

4.5Componentsofthetranslationmachineryaffectpenetrationandpost-

penetrationstepsofSFVinfection

EIF2B3, EIF4G1, and RPL18 are components of the translational machinery

(Walsh&Mohr,2011;delaCruzetal,2015)(Figure17),andtheyappeartoplay

apartinSFVinfection.Astheircanonicalrolesarerelatedtohosttranslation,I

assumed that they would work similarly in SFV infection and affect the post-

penetrationstages.Itestedthisusingtheendocyticbypassassay.Thisapproach

revealed that EIF2B3 is not needed in SFV entry or penetration, as the

percentage of infected cells was three times that of controls in the endocytic

bypass assay (Figure 18). For the EIF4G1-depleted cells, the mean infection

percentageintheendocyticbypassassaywassimilartothatofnormalinfection

assay(1.5-foldhigher than incontrols),but thisdifferencewasnotstatistically

significant(Figure18).Surprisingly,bycircumventingendocytosis,Icouldfully

revert the infection-diminishing effect of the RPL18 knockdown (Figure 18),

indicating that this ribosomal subunit plays a part in the early stages of SFV

infection.

31

Figure 16. The roles of TNP01 and HDAC6 in SFV infection. A: The relative infectionpercentages ofsiRNA-treatedHeLacellsinfectedwithSFV-ZsGreennormally(black)orusingendocyticbypass(grey)areindicated on the y-axis. The siRNAs used are indicated on the x-axis. The infection percentages werenormalized by setting the mean of the scrambled controls as 1. The error bars represent standarddeviations of the tree repetitions and an asterisk signifies a result that is significantly different from thecontrol. B: HeLa cells infected with SFV-ZsGreen normally or using endocytic bypass (indicated on theright).NucleiareshowninblueandnsP3-ZsGreen,indicatingSFVinfection,isshowningreen.ThesiRNAsusedareindicatedbythetextintheimages.

32

Figure17.TherolesofeIF2B,eIF4,andRPL18ontheinitiationoftranslation.Duringtheinitiationofeukaryotic translation, a 43 S pre-initiation complex is formed by GTP-bound eIF2, the 40 S ribosomalsubunitandothertranslationfactors(Walsh&Mohr,2011).ThiscomplexbindseIF4andothertranslationfactors, formingthe48Spre-initiationcomplex(Walsh&Mohr,2011).FollowingGTPhydrolysisandthebinding of the 60 S subunit, GDP-bound eIF2 dissociates and translationmoves to the elongation phase(Walsh&Mohr, 2011).GDP-eIF2 is converted to activeGTP-eIF2by eIF2Band it can initiate translationagain(Walsh&Mohr,2011).RPL18isapartoftheribosomal largesubunit(delaCruzetal,2015).MosttranslationfactorsandmRNAareomittedforclarity.

33

Figure18.TherolesofdifferenttranslationalmachinerycomponentsinSFVinfection.A:Therelativeinfection percentages of siRNA-treated HeLa cells infected with SFV-ZsGreen normally (black) or usingendocytic bypass (grey) are indicated on the y-axis. The siRNAs used are indicated on the x-axis. Theinfectionpercentageswerenormalizedbysettingthemeanofthescrambledcontrolsas1.Theerrorbarsrepresentstandarddeviationsof the treerepetitionsandanasterisksignifiesaresult that issignificantlydifferent from the control.B: HeLa cells infected with SFV-ZsGreen normally or using endocytic bypass(indicatedontheright).NucleiareshowninblueandnsP3-ZsGreen, indicatingSFVinfection, isshowningreen.ThesiRNAsusedareindicatedbythetextintheimages.

4.6SFVinfectionisaffectedbyotherhostfactorsinpenetrationandpost-

penetrationsteps

SFV infectivity could be increased by using siRNAs against PHB2, EDF1, ETF1,

DDX47, DHX37, and DHX57. DDX47 and DHX57 are probable RNA helicases

(Gene IDs: 51202 and 90957, respectively), DDX54 is a RNA helicase that

functionsintranscriptionregulation(Rajendranetal,2003;Kannoetal,2012),

PHB2 and EDF1 are implicated in transcriptional processes (Kabe etal, 1999;

Bavelloni etal, 2015), andETF1 is known to interactwith nonsense-mediated

decay (Czaplinski et al, 1998), which is a part of the host antiviral defense

34

(Balistreri et al, 2014). Therefore my hypothesis was that all of these factors

wouldhavepost-penetrationaleffects.TheeffectsofPHB2andEDF1depletion

were similar in the endocytic bypass and normal infection (Figure 19) so I

concludedthattheireffectsarenotrelatedtoSFVentryorpenetration.InETF1,

DDX47andDHX37-depletedcellstheinfectivitywasnotstatisticallysignificantly

different from the controls, but for DDX47 and DHX37 it was still high and

comparable to the infectivity in thenormal infectionassay (2-foldand1.5-fold

higherthancontrols,respectively)(Figure19).Curiously,intheDHX57-depleted

cells,theeffectoftheendocyticbypasswastheoppositeofthenormalinfection,

astheinfectivitywassignificantlylowerthanincontrols(opposedtohigherthan

incontrolsinthenormalinfection).

ThedepletionofGNPDA1andDMN2decreasedSFVinfectivity.AsDMN2hasa

crucialroleinendocytosis(Kasaietal,1999),Iassumedthateffectofknockingit

downwouldbelimitedtotheearlystagesofSFVinfection.GNPDA1,ontheother

hand, isapoorly-characterizedglucosamine-6-phosphatedeaminase(Wolosker

et al, 1998) so it was difficult to hypothesize how it affects SFV infection. By

usingendocyticbypassassay,theeffectofthesiRNAtreatmentsagainstGNPDA1

and DMN2 were reversed, so it seems that they both have roles in the early

stagesofSFVinfection(Figure19).

35

Figure 19. The roles of PHB2, EDF1, ETF1, DDX47, DHX37, DHX57, GNPDA1, and DMN2 in SFVinfection. A: The relative infection percentages of siRNA-treated HeLa cells infected with SFV-ZsGreennormally(black)orusingendocyticbypass(grey)areindicatedonthey-axis.ThesiRNAsusedareindicatedonthex-axis.Theinfectionpercentageswerenormalizedbysettingthemeanofthescrambledcontrolsas1.Theerrorbarsrepresentstandarddeviationsofthetreerepetitionsandanasterisksignifiesaresultthatis significantly different from the control. B: HeLa cells infected with SFV-ZsGreen normally or usingendocytic bypass (indicated on the right). Nuclei are shown in blue and nsP3-ZsGreen, indicating SFVinfection,isshowningreen.ThesiRNAsusedareindicatedbythetextintheimages.

36

5.Discussion

5.1.HostfactorsinSFVandVSVinfections

Using siRNA knockdown, high-throughput imaging, and automated image

analysis,IstudiedtheeffectsofthedepletionofvarioushostfactorsonSFVand

VSVinfection(Table2).

Table2.TheeffectsofsiRNAtreatmentagainstvariousgenesonSFVandVSVinfection

Gene EffectonSFV

infection*

EffectonVSV

infection*

Gene EffectonSFV

infection*

EffectonVSV

infection*

PHB2 + - DDX41 0 0

EDF1 + 0 DDX43 0 0

SLC6A13 - DDX47 + 0

EIF2B3 + 0 DDX54 - 0

ETF1 + 0 DHX37 + -

EIF4G1 + 0 DHX57 + 0

DDX18 0 - KREMEN2 0 +

PVR 0

+ HDAC6

+ 0

CSPG5 0

0 DNM2 - 0

GNPDA1 -

0 AP2M1 0 +

RPL18 - - CLTC 0 0

UPF1 + 0 KPNB1 0 -

ATP6V1B2 - - TNPO1 - +

ATP6V1G1 - - Ppbi 0 +

DDX31 0 0 PPBI 0 +

*+=siRNAtreatmentincreasesinfection,-=siRNAtreatmentdecreasesinfection,0=siRNAtreatmenthasnoeffectoninfection

Byusingtheendocyticbypassassay,IpinpointedtherolesofSLC6A13,TNP01,

HDAC6, RPL18, ETF1, and GNDPA1 to the early events in SFV infection. I also

employedthesamemethodtoassignpost-penetrationrolesforDDX54,EIF2B3,

PHB2, and EDF1. My results also confirmed the previously reported roles of

ATP6V1B2andATP6V1G1 in the early stages of SFV infection (Balistrerietal,

2014). In the case of EIF4G1, DDX47, and DHX37 the mean infectivity in the

endocyticbypasswassimilartonormalinfection,eventhoughthedifferencewas

not statistically significant. In all cases this was caused by a single outlying

37

repetition(datanotshown),soIaminclinedtobelievethatthesefactorsaffect

the later stages of SFV infection. Similarly, in the endocytic bypass assay the

mean infectivity inUPF1-depletedcellswasnotsignificantlydifferent formthe

control because of a single outlier (data not shown). Still, themean infectivity

was similar to thenormal infection, andalmost two times thatof the controls.

Furthermore, UPF1 has been reported to prevent the translational and

transcriptional stagesof SFV infection (Balistrerietal, 2014). Curiously, in the

normal infection, DHX57 depletion increased SFV infection significantly but in

theendocyticbypass,virusinfectivitywassignificantlylowerthanincontrols.

5.2.SLC6A13isacandidatereceptorforSFV

SLC6A13 is structurally similar to the iron transporter NRAMP2 (natural

resistanceassociatedmacrophageprotein2)(Nevo&Nelson,2006;Kristensen

etal,2011).Theybothbelongto thesamefamilyofsolute-carryingmembrane

proteins that have 12 membrane-spanning domains and cytosolic N- and C-

termini (Nevo & Nelson, 2006; Kristensen et al, 2011). NRAMP2 is used as a

receptorinbothinsectandmammaliancellsbySindbisvirus,analphavirusthat

is closely related to SFV (Rose et al, 2012). Therefore SLC6A13 might be the

receptor for SFV.This is further supportedby the fact that SLC6A13depletion

does not affect VSV infectivity, as receptors are often virus-specific (Grove &

Marsh,2011).Additionally,SLC6A13mRNAistranscribedheavilyinthekidneys

(The Human Protein Atlas, http://www.proteinatlas.org/ENSG00000010379-

SLC6A13/tissue, Fagerberg et al, 2014). The currently used SFV strains have

been passaged multiple times in hamster kidney (BHK-21) cells (Atkins et al,

1999) and the reason thatBHK-21 cells supporthigh titers of SFVmaybe the

presence of the hamster analogue of SLC6A13. Passaging has affected SFV

bindingbyselectingforsingleaminoacidmutationsintheenvelopeproteinsof

the virus which allows SFV to bind to host heparan sulphate via simple

electrostaticinteractions(Smitetal,2002).Howeverbindingtohostreceptorsis

more complicated than binding on the surface, as it involvesmultiple specific

interactions between viral and host proteins (Marsh&Helenius, 2006). These

events need to be complex enough signal the host to initiate the endocytic

program(Marsh&Helenius,2006). It is thereforeconceivable thatadapting to

38

completelynewreceptors is anunlikelyeventeven in cell culture, andviruses

may retain specificity for suitable receptors on the surface of the cells from

whichtheyareproduced.

ToconfirmtheroleofSLC6A13inSFVinfection,furtherexperimentsareneeded.

WewilltestifSLC6A13-depletionblocksSFVentry(asopposedtopenetration)

using single-virus tracking in living cells and (Hoornweg et al, 2016).Wewill

also use fluorescent immunolabeling and confocalmicroscopy to compare the

rateofSFVendocytosisinSLC6A13-depletedandwildtypecells(Rizopouloset

al, 2015). Furthermore, will determine, if the transfection of a SLC6A13-

producing plasmidwill rescue SFV infection in conditional SLC6A13-knockout

cells. To further characterize the potential interaction between SLC6A13 and

SFV,astructuralapproachisneeded.Thismay,forexample,beintheformofX-

ray crystallography of the SFV spike proteins complexed with a SLC6A13

extracellular domain or domains (Peng et al, 2011), or a cryo-electron

microscopy single-particle reconstruction of SFV bound to SLC6A13 (He et al,

2002).

5.3.HostfactorsthataffectSFVinfection

ThefactthatTNP01RPL18,andETF1areneededforSFVentryandpenetration

stages issurprisingasTNP01 is involved in intracellular traffickingandRPL18

andETF1havecanonicalroles intranslation(Tayloretal,2012;Twyffelsetal,

2014; de la Cruz et al, 2015). TNP01 is used by adenoviruses (Hindley et al,

2007), human immunodeficiency virus 1 (HIV-1) (Arnold et al, 2006), and

humanpapillomavirus(Darshanetal,2004)forthenucleartransportofvarious

virus components. As the effect of TNP01 depletions are reversed by the

endocytic bypass assay, this cellular proteinmust be needed in the very early

stagesofSFVinfection.Therefore,aroleinthelaterstepsoftheviruslifecycle,

suchas interferencewithstressgranule formationorTNP01-mediatednuclear

import of viral replicase protein nsP2 (Twyffels et al, 2014; Fros & Pijlman,

2016),areunlikely tobe related to theobservedphenotypes. Itmightalsobe

thattheeffectofTNP01depletionisindirectanditaffectsthefunctionofsome

otherproteinsthathaverolesinSFVinfection.

39

In the case of VSV, the known role of TNP01 in stress granule formation, a

processthatisconsideredasamechanismofintrinsicantiviralimmunitymaybe

the reasonwhy TNP01 depletion increases VSV infection (Beckham& Parker,

2008;Twyffelsetal,2014).Ithaspreviouslybeenreportedthattheknockoutof

TIA-1, a stress granule-associated protein, increases the infectivity of VSV in

mouse embryonic fibroblasts (Li et al, 2002), which further supports this

hypothesis.

BasedontheinteractionofRPL18andotherviruses,onewouldexpectRPL18to

affect the late stages of SFV infection. This ribosomal proteinhas a role in the

translation or replication of Dengue virus (Cervantes-Salazar et al, 2015),

cauliflowermosaicvirus(Lehetal,2000)andHCV(Dharetal,2006).However,

inSFVinfection,RPL18functionsintheearlystagesofthevirallifecycle,which

indicates a mechanism different from the other known viruses. It has been

shownthatribosomesareneededintheuncoatingofSFVnucleocapsids(Singh

&Helenius,1992).Thus,RPL18mighthavearole inthisprocess,whichwould

beafunctionunlikeanydescribedforaribosomalproteinsofar.However,this

wouldalsomeanthatduringtheendocyticbypassassay,whenthenucleocapsids

aredelivereddirectlytothecytoplasmthroughthePM,SFVvirionsareuncoated

inadifferentwaythanduringthenormalentry.Thisunexpectedresultmayalso

be caused by an indirect interaction of RPL18 and some other cellular factor,

whichisperturbedbythedepletionofRPL18.AsIdidnotinvestigateatwhich

stageofVSVinfectionRPL18affects,myworkinghypothesis isthat itprobably

affectsthereplicationortranslation,ashasbeenreportedforotherviruses(Leh

etal,2000;Dharetal,2006;Cervantes-Salazaretal,2015).

ETF1isanotherproteinthatseemedtoberelatedtothepost-entrystagesofSFV

infection. ETF1 is a eukaryotic translation termination factor, with a role in

nonsense-mediated decay (Czaplinski etal, 1998), a pathway that is known to

counteract SFV infection during translation and replication (Balistreri et al,

2014). Thus, I initially assumed that ETF1 would have a role in the post-

penetrationstagesofSFVinfection.Myresultsruleoutthispossibility,because

the effect of ETF1 depletion was reversed by the endocytic bypass assay.

40

Therefore, itmight be that the effect I have observed is indirect, or ETF1 has

previouslyuncharacterizedfunctions.Additionally,thesefunctionsdonotseem

tobebroadlyantiviral,asETF1-depletionhadnoeffectonVSVinfection.

HDAC6isusedby influenzaAvirus(IAV)touncoat itsgenome(Banerjeeetal,

2014). It is also reported to inhibit oncolytic herpes simplex virus infection in

gliomacells,apparentlybyinterferingwithendocytictrafficking(Nakashimaet

al, 2015). As our initial result showed small but reproducible increase in SFV

infectionfollowedbyHDAC6knockdown,Iassumedthatitplaysaminorrolein

SFVentryandpenetrationprocessesasHDAC6hasarole in thepositioningof

endosomes (Li et al, 2013). This was confirmed by endocytic bypass.

SurprisinglyHDAC6depletiondidnotaffectVSV infection,eventhoughHDAC6

hasbeenreportedtoprotectmicefromVSV(Choietal,2016).Thismayindicate

that cell culturediffers significantly from invivoconditionsor thatHDAC6has

differentrolesinmouseandhumancells.ItisalsopossiblethatHDAC6functions

in the very late stages of VSV infection, such as egress, as our assay only

quantifiesviralreplicationandnotthelaterstagesofinfection.

DMN2isanotherproteinwithafunctioninendocytosis(Kasaietal,1999),and

therefore it was expected that its depletion would reduce SFV infectivity by

affecting theearlystepsof the infection.TheroleofDMN2 invirus infection is

well characterized, and it is needed for example by bovine ephemeral virus,

hepatitisEvirus,andPichindévirus(Velaetal,2008;Chengetal,2012;Hollaet

al,2015).SurprisinglyitwasnotrequiredbyVSV,eventhoughitisreportedto

becrucialforVSVentry(Curetonetal,2009;Johannsdottiretal,2009).

Incells,EIF2B3andEIF4G1function intranscription initiation(Walsh&Mohr,

2011).EIF2B3isacomponentofeIF2B,whichrecycleseIF2,akeyfactorinthe

formation of the 43 S pre-initiation complex and EIF4G1 has a function in

bringing the mRNA cap and poly-A tail together in the 48 S pre-initiation

complex(Walsh&Mohr,2011).Asapartofthehost’sbroadantiviralresponse

thataimstoshutdowntranslation,eIF2canbe inactivatedbyphosphorylation

(Walsh&Mohr,2011).SFVcancounteractthisresponse,andisknowntobeable

41

totranslateitsstructuralgeneseveninthepresenceofphosphorylatedeIF2.As

thedepletionofEIF2B3increasesSFVinfectioninthepost-penetrationstages,it

seems that SFV can translate even its early genes without a canonical set of

translationinitiationfactors.KnockingdownEIF4G1similarlyaffectslaterstages

of SFV infection and increases SFV infectivity. This shows that though cap-

dependent,thetranslationofSFVearlygenesdoesnotoccurinexactlythesame

way as the translation of hostmRNAs. Since SFV RNA is not as susceptible to

changesinthecompositionofthetranslationinitiationcomplexashostmRNAs

are,itgainsanadvantageinthetranslationalcompetition.Astheknockdownof

EIF4G1orEIF2B3doesnotaffectVSVinfection,theirfunctionismostlikelynot

broadlyantiviral.ThisindicatesthatSFVhasevolvedsomemechanismtobenefit

from the depletion of these host factorswhile VSV has not.Most likely this is

difference is mediated by the differences in SFV (sub)genomic RNA and VSV

mRNAs.

PHB2 knockdown increased SFV infectivity during the post-penetration steps

and it also increasedVSV infectivity.PHB2hasbeenreported tobeapro-viral

agent for HIV-1 (Emerson et al, 2010) and to associate with severe acute

respiratory syndrome coronavirus nsPs (Cornillez-Ty et al, 2009). Curiously,

PHB2-depletion has been previously shown to reduce VSV infection inmouse

cells (Kreit et al, 2015). The different result reported here might be due to

different roles of PHB2 in murine and human cells, or differences in VSV

infection between these two organisms. Additionally, Kreit et al (2015) used

small hairpin RNA technology, opposed to siRNAs used in this study. Even

though thesemethods are highly similar, thismay have a role in the differing

results. As PHB2 interacts with a myriad of host processes and proteins

(Bavellonietal,2015),itispossiblethattheeffectsofitsdepletionareindirect

and the actual effects are mediated by other host factors. PHB2 is also a

transcriptional regulator (Bavelloni et al, 2015) so its knockdown may just

reducehost translation, thusgiving theviralRNAsanedge in the translational

competition. This may also be the mechanism by which EDF1-depletion

promotesbothSFVandVSVinfectivity,asitisatranscriptionalactivator(Kabe

42

et al, 1999). This is further supported by the fact that the effect of EDF1

knockdownaffectsthelaterstagesofSFVinfection.

GNDPA1, DDX54, DDX47, DHX37, and DHX57 are all poorly characterized

cellular factors that have not been previously implicated in viral infections.

DDX47, DHX37, and DHX57 are predicted RNA helicases (Gene IDs: 51202,

57647, and 90957, respectively). DDX54 is an RNA-helicase that has been

reportedtobeatranscriptionalco-repressor(Rajendranetal,2003;Kannoetal,

2012). The knockdown of DDX53, and DDX47 increased SFV infection by

affecting the post-penetration stages, but had no effect on VSV infection. This

mayindicatethatbothofthesehelicasescantargetRNAstructuresfoundinSFV

but not in VSV. Or possibly SFV but not VSV requires these helicases for

translation or transcription. DHX37 depletion affected both SFV and VSV

infections.ItdecreasedSFVinfection,byaffectingthelatestages,andincreased

VSVinfection.Thismaybeduetoasimilarsituationasabove,butwithDHX37

depletion affecting host RNAs in a way, that allows VSV to succeed in the

translationalcompetitionagainstthehost.Itisdifficulttohypothesizethereason

why DHX57 depletion was beneficial to SFV in the normal infection, but

detrimentalintheendocyticbypass.ThismaybeduetoacomplexroleofDHX57

ormerely experimental anomalies. GNDPA1 is a highly conservedprotein that

plays a crucial role in metabolism (Wolosker et al, 1998). It is, therefore,

surprising that it would have a role in the early stages of SFV infection. This

effect would imply that GNDPA1 has previously uncharacterized functions, or

thatitsdepletionaffectsotherhostfactorsthat,inturn,affectSFVinfection.

5.4.Otherhostfactors

Apparently, VSV infection is counteracted by PVR, KREMEN2, AP2M1 and

cyclophilin B. On the other hand VSV seems to need ATP6V1B2, ATP6V1G1,

DDX18, and KPNB1 for successful infection. The requirement of the vATPase

subunits ATP6V1B2 and ATP6V1G1 by VSV is not surprising, as it penetrates

hostmembranesviapH-dependentfusion(Regan&Whittaker,2013).PVRand

KREMEN2arecellsurfacemolecules(Mendelsohnetal,1989;Maoetal,2002)

and AP2M1 is a part of AP2, and therefore has a role in clathrin-mediated

43

endocytosis (Heinaman,1995;Boucrotetal, 2010).Therefore it is conceivable

thatthesemoleculeswouldaffectVSVentryorpenetration.

DDX18isapoorly-studiedputativeRNAhelicaseforwhichhelicaseactivityhas

notbeendemonstratedyet(Dubaele&Chène,2007).IfDDX18doesaffectRNA,

the effect of its depletion on VSV but not SFV infection may be due to VSV

containing some RNAmotif that SFV lacks, or vice versa. KPNB1 is a protein

required for nuclear import (Görlich et al, 1995) and its depletion greatly

reducesVSVinfection.IthasbeenpreviouslyshownthattheMproteinofVSVis

transported into the nucleus, which allows it to reduce nucleocytoplasmic

transport(Petersenetal,2000;Glodowskietal,2002).Perhapsthetransportof

theMproteinismediatedbyKPNB1,whichreducesthetransportofhostmRNAs

into the nucleus, therefore giving VSV the upper hand in the translational

competition.

Cyclophilins affect the infections ofmultiple viruses (Frausto et al, 2013) and

cyclophilinAhasbeen reported tobe requiredbyVSV for successful infection

(Boseetal,2003).ThesiRNAsagainstPPBIincreasedinfectivitymorethanthe

siRNAsagainstPpbi.Mostlikelythiseffectisexplainedbysequencedissimilarity

between human andmouse cyclophilin B as the use of Ppbi-targeting siRNAs

wouldresultinonlypartialknockdownofPPBI(Petersenetal,2000;Glodowski

etal,2002).

CSPG6, a chondroitin sulfate proteoglycan (Watanabe et al, 1995), DDX31, an

RNA-helicasewithapossiblerole inrenalcarcinogenesis(Fukawaetal,2012),

DDX41aRNAhelicasewith functions in immuneprocesses (Jiangetal, 2017),

DDX43, an RNA and DNA helicase (Tanu et al, 2017), and CLTC, a well-

characterizedendocyticprotein(Doherty&McMahon,2009)didnothaveroles

inSFVofVSVinfections.Chondroitinsulfateproteoglycanshavebeenimplicated

inthebindingofvirusestohostcells(Banfieldetal,1995;Hsiaoetal,1999),but

SFVandVSVapparentlydonotutilize it,at leastexclusively.This isconsistent

withthereportthatSFVbindstoheparansulfateonthehostsurface(Smitetal,

2002). An siRNA screen showed a reduction in the infection of IAV (a double-

44

strandedDNAvirus) followedby the silencingofDDX31, but the investigators

didnotpursuethisresultanyfurther(Diotetal,2016)andDDX41seemstohave

aroleininnateimmunityinthedetectionofdouble-strandedDNAviruses(Jiang

etal,2017).AsSFVandVSVareRNAviruses,itseemsthatvirus-affectingroles

ofDDX31andDDX41might be exclusive toDNAviruses.DDX43hasnot been

previously implicated in viral infection so it is not surprising that it does not

affectSFVorVSVinfections.Clathrinhasbeenthoughttobecrucialfortheentry

ofbothSFVandVSV(Heleniusetal,1980;Sunetal,2005),butaccordingtomy

resultstheentryprocessmightbemorecomplicated,andCLTCisnotrequired

byeithervirus.

5.5.Methodologicalconsiderations

AssiRNA-basedscreeninghasbecomecheaperandsiRNAdesignhasimproved,

anumberoflarge-scalestudiesonhost-virusinteractionshavebeenpublishedin

recent years (Perreira et al, 2016). Even though siRNA-based screening is

definitelyapowerfultool, ithasitsdrawbacks.Traditionally,siRNAshavebeen

consideredveryspecificbutFranceschinietal(2014)showedthatsiRNAsused

inscreeningcanactasmicro-RNAs(miRNAs)andhavestrongoff-targeteffects.

This kind of indirect effects may explain the surprising results of siRNA

treatmentsagainstTNP01,RPL18,andETF1onSFVinfectionaswellasthoseof

siRNAsagainstPHB2andHDAC6onVSVinfection.Theapparentnonessentiality

of CLTC in both SFV and VSV infectionsmay also be explainedwith off-target

effects.Itmayalsobe,thatsomeofthesurprisingresultsareduetoincomplete

knockdown of the studied host factors. Due to possibility of off-targets or

incomplete knockdown, the results of siRNA screens need to be extensively

validated. The knockdown efficiency of the siRNAs should be tested using

immunoblotting.Preferably, therolesof thegenesof interestwouldbestudied

using knockout cells. As an ultimate test, a rescue assay needs performed, in

whichknockdownor(preferably)knockoutcellsaretransfectedwithaplasmid

thatproduces theprotein inquestion.This should lead to the reversionof the

effect on virus infection. Due to the lack of further validation, the genes

implicatedbymyexperimentsaretobeunderstoodascandidatefactors.

45

Theendocyticbypassassayusedinmyexperimentsisveryrobustandseparates

the function of the host factors to two rather broad categories. Therefore,

additional experiments are needed to perfectly understand the role of host

factors.Usingconfocalmicroscopy,itispossibletodetectifvirusesboundoncell

surface are endocytosed or not (Rizopoulos et al, 2015). To determine if the

depletion of the host factor affects SFV penetration or uncoating, specific

antibodies against the acidified spike proteins and SFV capsid proteins are

available (Liao & Kielian, 2006) Spike acidification is a sign of membrane

penetration,andthe localizationof thesignal fromanti-capsidantibodycanbe

used to detect the uncoating events (Singh&Helenius, 1992). Topinpoint the

effects of host factors to later stages of SFV infection, immunoblotting can be

used todetect theproductionofearlyand lateviralproteins. Similarly toSFV,

the endocytic part of VSV infection can be bypassed (Blumenthal etal, 1987).

Therefore,performinganendocyticbypassassaywouldbethenextstep inthe

elucidationoftherolesofthehostfactorsthataffectVSVinfection,followedwith

e.g.directmeasuringofVSVendocytosis(Rizopoulosetal,2015).

5.6.Conclusions

Themajor result of this study is that SLC6A13 is a candidate receptor for SFV

and if validated, it implies that alphaviruses in generaluse structurally similar

solutecarriersastheirreceptors.Thismayleadtothediscoveryofthereceptors,

of other,more clinically relevant alphaviruses,which in turnmay lead to new

therapeutic applications. Overall, it is clear that multiple previously

uncharacterized host factors affect both SFV and VSV infections and many of

thesefactorsarespecifictoSFVorVSV.Thisthesisalsoshowsthathostfactors

play important parts in both the early and late stages of SFV infection. Taken

together,theresultsofthisthesisfurtherconfirmthatvirus-hostinteractionsare

crucialinvirusinfections.Myfindingsserveasastartingpointtocontinuetoin-

depthstudiesintohowthesefactorsaffectSFVandVSVinfections.

6.AcknowledgementsI thank Dr. Giuseppe Balistreri for excellent supervision, Dr. Aušra Domanska,

ProfessorSarahButcher,Dr.JustinFlatt,andDr.CarlottaGlackinforthecritical

46

readingofmythesis,aswellasalltheothermembersoftheButchergroupfor

theirhelp and support. Iwould also like to acknowledge theLightMicroscopy

UnitoftheInstituteofBiotechnologyfortheuseoftheirequipmentandDr.Kirsi

Hellström,forthehelpprovidedintheamplificationofSFV-ZsGreen.

7.ReferencesArnoldM,NathA,HauberJ&KehlenbachRH(2006)MultipleImportinsFunctionas

NuclearTransportReceptorsfortheRevProteinofHumanImmunodeficiencyVirusType1.J.Biol.Chem.281:20883–20890

AtkinsGJ,SheahanBJ&LiljeströmP(1999)TheMolecularPathogenesisofSemlikiForestVirus:aModelVirusMadeUseful?J.Gen.Virol.80:2287–2297

BalistreriG,HorvathP,SchweingruberC,ZündD,McInerneyG,MeritsA,MühlemannO,AzzalinC&HeleniusA(2014)TheHostNonsense-MediatedmRNADecayPathwayRestrictsMammalianRNAVirusReplication.CellHostMicrobe16:403–411

BanerjeeI,MiyakeY,NobsSP,SchneiderC,HorvathP,KopfM,MatthiasP,HeleniusA&YamauchiY(2014)InfluenzaAVirusUsestheAggresomeProcessingMachineryforHostCellEntry.Science346:473–7

BanfieldB,LeducY,EsfordL,VisalliRJ,BrandtC&TufaroF(1995)EvidenceforanInteractionofHerpesSimplexViruswithChondroitinSulfateProteoglycansDuringInfection.Virology208:531–539

BavelloniA,PiazziM,RaffiniM,FaenzaI&BlalockWL(2015)Prohibitin2:AtaCommunicationsCrossroads.IUBMBLife67:239–254

BeckhamCJ&ParkerR(2008)P-bodies,StressGranulesandViralLifeCycles.CellHostMicrobe3:206–212

LeBlancI,LuyetP-P,PonsV,FergusonC,EmansN,PetiotA,MayranN,DemaurexN,FauréJ,SadoulR,PartonRG&GruenbergJ(2005)Endosome-to-CytosolTransportofViralNucleocapsids.Nat.CellBiol.7:653–664

BlumenthalR,Bali-PuriA,WalterA,CovellD&EidelmanO(1987)pH-DependentFusionofVesicularStomatitisViruswithVeroCells.MeasurementbyDequenchingofOctadecylRhodamineFluorescence.J.Biol.Chem.262:13614–13619

BoseS,MathurM,BatesP,JoshiN&BanerjeeAK(2003)RequirementforCyclophilinAfortheReplicationofVesicularStomatitisVirusNewJerseySerotype.J.Gen.Virol.84:1687–1699

BoucrotE,SaffarianS,ZhangR&KirchhausenT(2010)RolesofAP-2inClathrin-MediatedEndocytosis.PLoSOne5:e10597

CarpenterAE,JonesTR,LamprechtMR,ClarkeC,KangIH,FrimanO,GuertinDa,ChangJH,LindquistRa,MoffatJ,GollandP&SabatiniDM(2006)CellProfiler:ImageAnalysisSoftwareforIdentifyingandQuantifyingCellPhenotypes.GenomeBiol.7:R100

Cervantes-SalazarM,Angel-AmbrocioAH,Soto-AcostaR,Bautista-CarbajalP,Hurtado-MonzonAM,Alcaraz-EstradaSL,LudertJE&DelAngelRM(2015)DengueVirusNS1ProteinInteractsWiththeRibosomalProteinRPL18:ThisInteractionIsRequiredforViralTranslationandReplicationinHuh-7Cells.Virology484:113–126

ChengCY,ShihWL,HuangWR,ChiPI,WuMH&LiuHJ(2012)BovineEphemeralFeverVirusUsesaClathrin-MediatedandDynamin2-DependentEndocytosisPathwayThatRequiresRab5andRab7asWellasMicrotubules.J.Virol.86:13653–13661

ChoiSJ,LeeH,KimJ,ParkSY,KimT,LeeW,JangD,YoonJ,ChoiY,KimS,MaJ,KimC,YaoT,JungJU,LeeJ&LeeJ(2016)HDAC6RegulatesCellularViralRNASensingbyDeacetylationofRIG-I.35:429–442

47

CoordinatorsNR(2017)DatabaseResourcesoftheNationalCenterforBiotechnologyInformation.NucleicAcidsRes.45:D12–D17

Cornillez-TyCT,LiaoL,YatesJR,KuhnP&BuchmeierMJ(2009)SevereAcuteRespiratorySyndromeCoronavirusNonstructuralProtein2InteractswithaHostProteinComplexInvolvedinMitochondrialBiogenesisandIntracellularSignaling.J.Virol.83:10314–10318

CuretonDK,MassolRH,SaffarianS,KirchhausenTL&WhelanSPJ(2009)VesicularStomatitisVirusEntersCellsthroughVesiclesIncompletelyCoatedwithClathrinThatDependuponActinforInternalization.PLoSPathog.5:

CzaplinskiK,Ruiz-EchevarriaMJ,PaushkinSV.,HanX,WengY,PerlickHA,DietzHC,Ter-AvanesyanMD&PeltzSW(1998)TheSurveillanceComplexInteractswiththeTranslationReleaseFactorstoEnhanceTerminationandDegradeAberrantmRNAs.GenesDev.12:1665–1677

DarshanMS,LucchiJ,HardingE&MoroianuJ(2004)TheL2MinorCapsidProteinofHumanPapillomavirusType16InteractswithaNetworkofNuclearImportReceptorsTheL2MinorCapsidProteinofHumanPapillomavirusType16InteractswithaNetworkofNuclearImportReceptors.J.Virol.78:12179–12188

DharD,MapaK,PudiR,SrinivasanP,BodhinathanK&DasS(2006)HumanRibosomalProteinL18aInteractsWithHepatitisCVirusInternalRibosomeEntrySite.Arch.Virol.151:509–524

DiotC,FournierG,DosSantosM,MagnusJ,KomarovaA,vanderWerfS,MunierS&NaffakhN(2016)InfluenzaAVirusPolymeraseRecruitstheRNAHelicaseDDX19toPromotetheNuclearExportofViralmRNAs.Sci.Rep.6:33763

DohertyGJ&McMahonHT(2009)MechanismsofEndocytosis.Annu.Rev.Biochem.78:857–902

DubaeleS&ChèneP(2007)CellularStudiesofMrDb(DDX18).Oncol.Res.16:549–556EmersonV,HoltkotteD,PfeifferT,WangI-H,SchnolzerM,KempfT&BoschV(2010)

IdentificationoftheCellularProhibitin1/Prohibitin2HeterodimerasanInteractionPartneroftheC-TerminalCytoplasmicDomainoftheHIV-1Glycoprotein.J.Virol.84:1355–1365

FagerbergL,HallstromBM,OksvoldP,KampfC,DjureinovicD,OdebergJ,HabukaM,TahmasebpoorS,DanielssonA,EdlundK,AsplundA,SjostedtE,LundbergE,SzigyartoCA-K,SkogsM,TakanenJO,BerlingH,TegelH,MulderJ,NilssonP,etal(2014)AnalysisoftheHumanTissue-specificExpressionbyGenome-wideIntegrationofTranscriptomicsandAntibody-basedProteomics.Mol.Cell.Proteomics13:397–406

FranceschiniA,MeierR,CasanovaA,KreibichS,DagaN,AndritschkeD,DillingS,RamoP,EmmenlauerM,KaufmannA,Conde-AlvarezR,LowSH,PelkmansL,HeleniusA,HardtW-D,DehioC&vonMeringC(2014)SpecificInhibitionofDiversePathogensinHumanCellsbySyntheticMicroRNA-likeOligonucleotidesInferredfromRNAiScreens.Proc.Natl.Acad.Sci.111:4548–4553

FraustoSD,LeeE&TangH(2013)CyclophilinsasModulatorsofViralReplication.Viruses5:1684–1701

FrosJJ&PijlmanGP(2016)AlphavirusInfection:HostCellShut-OffandInhibitionofAntiviralResponses.Viruses8:166

FukawaT,OnoM,MatsuoT,UeharaH,MikiT,NakamuraY,KanayamaHO&KatagiriT(2012)DDX31Regulatesthep53-HDM2PathwayandrRNAGeneTranscriptionThroughItsInteractionwithNPM1inRenalCellCarcinomas.CancerRes.72:5867–5877

FullerSD,BerrimanJA,ButcherSJ&GowenBE(1995)LowpHInducesSwivelingoftheGlycoproteinHeterodimersintheSemlikiForestVirusSpikeComplex.Cell81:715–725

GeP,TsaoJ,ScheinS,GreenTJ,LuoM&ZhouZH(2010)Cryo-EMModeloftheBullet-ShapedVesicularStomatitisVirus.Science(80-.).327:689–693

48

GlodowskiDR,PetersenJM&DahlbergJE(2002)ComplexNuclearLocalizationSignalsintheMatrixProteinofVesicularStomatitisVirus.J.Biol.Chem.277:46864–46870

GörlichD,KostkaS,KraftR,DingwallC,LaskeyRA,HartmannE&PrehnS(1995)TwoDifferentSubunitsofImportinCooperatetoRecognizeNuclearLocalizationSignalsandBindThemtotheNuclearEnvelope.Curr.Biol.5:383–392

GosertR,EggerD,LohmannV,BartenschlagerR,BlumHE,BienzK&MoradpourD(2003)IdentificationoftheHepatitisCVirusRNAReplicationComplexinHuh-7CellsHarboringSubgenomicReplicons.J.Virol.77:5487–5492

GriffinDE(2013)Alphaviruses.InFieldsVirology,KnipeD&HowleyP(eds)pp651–686.Philadelphia:LippincottWilliams&Wilkins

GroveJ&MarshM(2011)TheCellBiologyofReceptor-mediatedVirusEntry.J.CellBiol.195:1071–1082

HeY,LinF,ChipmanPR,BatorCM,BakerTS,ShohamM,KuhnRJ,MedofME&RossmannMG(2002)StructureofDecay-acceleratingFactorBoundtoEchovirus7:aVirus-receptorComplex.Proc.Natl.Acad.Sci.U.S.A.99:10325–10329

HeinamanR(1995)ActivityandinVitroReassemblyoftheCoatedVesicle(H+)-ATPaseRequiresthe50-kDaSubunitoftheClathrinAssemblyComplexAP-2*.J.Biol.Chem.269:81–95

HeleniusA,KartenbeckJ,SimonsK&FriesE(1980)OntheEntryofSemlikiForestVirusIntoBHK-21Cells.J.CellBiol.84:404–420

HindleyCE,LawrenceFJ&MatthewsDA(2007)ARoleforTransportinintheNuclearImportofAdenovirusCoreProteinsandDNA.Traffic8:1313–1322

HoffmannH&Schiene-FischerC(2014)FunctionalAspectsofExtracellularCyclophilins.Biol.Chem.395:721–735

HollaP,AhmadI,AhmedZ&JameelS(2015)HepatitisEVirusEntersLiverCellsthroughaDynamin-2,ClathrinandMembraneCholesterol-dependentPathway.Traffic16:398–416

HoornwegTE,vanDuijl-RichterMKS,AyalaNuñezNV,AlbulescuIC,vanHemertMJ&SmitJM(2016)DynamicsofChikungunyaVirusCellEntryUnraveledbySingle-VirusTrackinginLivingCells.J.Virol.90:4745–56

HsiaoJC,ChungCS&ChangW(1999)VacciniaVirusEnvelopeD8LProteinBindstoCellSurfaceChondroitinSulfateandMediatestheAdsorptionofIntracellularMatureVirionstoCells.J.Virol.73:8750–8761

HubbertC,GuardiolaA,ShaoR,KawaguchiY,ItoA,NixonA,YoshidaM,WangX-F&YaoT-P(2002)HDAC6isaMicrotubule-associatedDeacetylase.Nature417:455–458

JiangY,ZhuY,LiuZJ&OuyangS(2017)TheEmergingrolesoftheDDX41ProteininImmunityandDiseases.ProteinCell8:83–89

JohannsdottirHK,ManciniR,KartenbeckJ,AmatoL&HeleniusA(2009)HostCellFactorsandFunctionsInvolvedinVesicularStomatitisVirusEntry.J.Virol.83:440–453

JohnsonDC&BainesJD(2011)Herpesvirusesremodelhostmembranesforvirusegress.Nat.Rev.Microbiol.9:382–394

KabeY,GotoM,ImaiT,WadaT,MorohashiK,ShirakawaM,HiroseS&HandaH(1999)TheRoleofHumanMBF1asaTranscriptionalCoactivatorTheRoleofHumanMBF1asaTranscriptionalCoactivator*.J.Biol.Chem.274:34196–34202

KannoY,SerikawaT,InajimaJ&InouyeY(2012)DP97,aDEADBoxDNA/RNAHelicase,IsaTargetGene-selectiveCo-regulatoroftheConstitutiveAndrostaneReceptor.Biochem.Biophys.Res.Commun.426:38–42

KasaiK,ShinHW,ShinotsukaC,MurakamiK&NakayamaK(1999)DynaminIIIsInvolvedinEndocytosisbutNotintheFormationofTransportVesiclesfromtheTrans-GolgiNetwork.JBiochem125:780–789

KawaguchiY,KovacsJJ,McLaurinA,VanceJM,ItoA&YaoTP(2003)TheDeacetylaseHDAC6RegulatesAggresomeFormationandCellViabilityinResponsetoMisfoldedProteinStress.Cell115:727–738

49

KreitM,VertommenD,GilletL&MichielsT(2015)TheInterferon-InducibleMouseApolipoproteinL9andProhibitinsCooperatetoRestrictTheiler’sVirusReplication.:1–22

KristensenAS,AndersenJ,JørgensenTN,SørensenL,EriksenJ,LolandCJ&StrømgaardK(2011)SLC6NeurotransmitterTransporters :Structure,Function,andRegulation.Pharma.Rev.63:585–640

KuhnRJ(2013)Togaviridae.InFieldsVirology,KnipeD&HowleyP(eds)pp629–650.Philadelphia:LippincottWilliams&Wilkins

KuismanenE,HedmanK,SarasteJ&PetterssonRF(1982)UukuniemiVirusMaturation:AccumulationofVirusParticlesandViralAntigensintheGolgiComplex.Mol.Cell.Biol.2:1444–58

delaCruzJ,KarbsteinK&WoolfordJL(2015)FunctionsofRibosomalProteinsinAssemblyofEukaryoticRibosomesInVivo.Annu.Rev.Biochem.84:93–129

LehV,YotP&KellerM(2000)TheCauliflowerMosaicVirusTranslationalTransactivatorInteractswiththe60SRibosomalSubunitProteinL18ofArabidopsisthaliana.Virology266:1–7

LeinonenR,AkhtarR,BirneyE,BowerL,Cerdeno-TarragaA,ChengY,ClelandI,FaruqueN,GoodgameN,GibsonR,HoadG,JangM,PaksereshtN,PlaisterS,RadhakrishnanR,ReddyK,SobhanyS,HoopenPTen,VaughanR,ZaluninV,etal(2011)TheEuropeanNucleotideArchive.NucleicAcidsRes.39:44–47

LiW,LiY,KedershaN,AndersonP,EmaraM,SwiderekKM,MorenoGT,BrintonMA&IrolJV(2002)CellProteinsTIA-1andTIARInteractwiththe3ЈStem-LoopoftheWestNileVirusComplementaryMinus-StrandRNAandFacilitateVirusReplication.76:11989–12000

LiY,ShinD&KwonSH(2013)HistoneDeacetylase6PlaysaRoleasaDistinctRegulatorofDiverseCellularProcesses.FEBSJ.280:775–793

LiaoM&KielianM(2006)Site-DirectedAntibodiesagainsttheStemRegionRevealLowpH-InducedConformationalChangesoftheSemlikiForestVirusFusionProtein.J.Virol.80:9599–9607

LylesD,KuzminI&RupprechtC(2013)Rhabdoviridae.InFieldsVirology,KnipeD&HowleyP(eds)pp885–922.Philadelphia:LippincottWilliams&Wilkins

ManciniEJ,ClarkeM,GowenBE,RuttenT&FullerSD(2000)Cryo-ElectronMicroscopyRevealstheFunctionalOrganizationofanEnvelopedVirus,SemlikiForestVirus.Mol.Cell5:255–266

MaoB,WuW,DavidsonG,MarholdJ,LiM,MechlerBM,DeliusH,HoppeD,StannekP,WalterC,GlinkaA&NiehrsC(2002)KremenProteinsAreDickkopfReceptorsthatRegulateWnt/β-cateninSignalling.Nature417:664–667

MarshM&HeleniusA(2006)VirusEntry:OpenSesame.Cell124:729–740MeadDG,RambergFB,BesselsenDG&MaréCJ(2000)TransmissionofVesicular

StomatitisVirusfromInfectedtoNoninfectedBlackFliesCo-feedingonNonviremicDeerMice.Science287:485–487

MeertensL,BertauxC&DragicT(2006)HepatitisCVirusEntryRequiresaCriticalPostinternalizationStepandDeliverytoEarlyEndosomesviaClathrin-CoatedVesicles.J.Virol.80:11571–11578

MendelsohnCL,WimmerE&RacanielloVR(1989)CellularReceptorforPoliovirus:MolecularCloning,NucleotideSequence,andExpressionofaNewMemberoftheImmunoglobulinSuperfamily.Cell56:855–865

MercerJ,SchelhaasM&HeleniusA(2010)VirusEntrybyEndocytosis.Annu.Rev.Biochem.79:803–833

NakashimaH,KaufmannJK,WangPY,NguyenT,SperanzaMC,KasaiK,OkemotoK,OtsukiA,NakanoI,FernandezS,GoinsWF,GrandiP,GloriosoJC,LawlerS,CripeTP&ChioccaEA(2015)HistoneDeacetylase6InhibitionEnhancesOncolyticViralReplicationinGlioma.J.Clin.Invest.125:4269–4280

NevoY&NelsonN(2006)TheNRAMPFamilyofMetal-ionTransporters.Biochim.

50

Biophys.Acta1763:609–620O’LearyNA,WrightMW,BristerJR,CiufoS,HaddadD,McVeighR,RajputB,RobbertseB,

Smith-WhiteB,Ako-AdjeiD,AstashynA,BadretdinA,BaoY,BlinkovaO,BroverV,ChetverninV,ChoiJ,CoxE,ErmolaevaO,FarrellCM,etal(2016)ReferenceSequence(RefSeq)DatabaseatNCBI:CurrentStatus,TaxonomicExpansion,andFunctionalAnnotation.NucleicAcidsRes.44:D733–D74