EMBO MolecularMedicine PeerReview ProcessFile- EMM-2013-02796 © EMBO 1 SynergisticCombination ofValproicAcid and Oncolytic ParvovirusH-1PV as a PotentialTherapy againstCervical and Pancreatic Carcinomas JunweiLi,Serena Bonifati,GeorgiHristov,TiinaMarttila,SéverineValmary-Degano,Sven Stanzel,MartinaSchnˆlzer,ChristianeMougin,MarcAprahamian,SvitlanaP.Grekova,Zahari Raykov,JeanRommelaereandAntonioMarchini Correspondingauthor:AntonioMarchini, GermanCancerResearchCenterDKFZ Review timeline: Submissiondate: 25 March 2013 EditorialDecision: 25 April2013 Revisionreceived: 23 July 2013 EditorialDecision: 05 August2013 Revisionreceived: 05 August2013 Accepted: 07 August2013 TransactionReport: (Note:With the exception ofthe correction oftypographicalor spelling errors thatcould be a source ofambiguity, letters and reports are not edited. The originalformatting oflettersand referee reportsmaynotbe reflected in this compilation.) Editor: CélineCarret 1stEditorialDecision 25 April2013 ThankyouforthesubmissionofyourmanuscripttoEMBO MolecularMedicine.Wehavenow heard back from thethreerefereeswhom weasked to evaluateyourmanuscript.Although the referees find the study to be of potentialinterest,they also raise a numberofconcernsthatneed to beaddressed in amajorrevision ofyourmanuscript. Asyouwillseefrom theenclosedreports,therefereesfindthestudyinterestingandcomprehensive. Whileref#1and#2arepositiveaboutit,ref#3ismorecritical.Assuch, we would like you to experimentally show the directrole ofVPA through NS1 acetylation assuggested by ref#1 and #3. In addition,ref#2 is concerned by the high titers of viruses and would be convinced of enhanced replication,spread and oncolysis if increased viraltitrescould be shown in isolated tumours.Finally, ref#3 has severalimportantcriticisms,however we would like you to particularly focus on testing anotherHDAC inhibitorin a key experimentand differentdosesofVPA.Ifyou have data on hand addressing the otherconcernsofref#3,we would strongly encourage you to include these in the manuscript. Shouldyoubeabletoaddresstheraisedconcernswithadditionalexperimentswhereappropriate, wewouldbewillingtoconsiderarevisedmanuscript. PleasenotethatitisEMBO MolecularMedicinepolicytoallow asingleroundofrevisioninorder to avoid the delayed publication of research findings. Consequently, acceptance or rejection of the EMBO MolecularMedicine PeerReview ProcessFile- EMM-2013-02796 © EMBO 2 manuscriptwilldependonthecompletenessofyourresponsesincluded in thenext version of the manuscript. EMBO MolecularMedicinehasa"scoopingprotection"policy,wherebysimilarfindingsthatare published by othersduring review orrevision arenotacriterion forrejection.Should you decideto submita revised version,Iwould like to ask you to getin touch afterthree months ifyou have not completed it,to update uson the status. Pleasealsocontactusassoonaspossibleifsimilarworkispublishedelsewhere.Ifotherworkis published we may notbe able to extend the revision period beyond three months. I look forward to receiving your revised manuscript. ***** Reviewer'scomments***** Referee#1: Thispaperreportsaremarkablycomprehensivestudyoftheuseofthehistone deacetylase inhibitor (HDAC) valproic acid (VPA) in combination therapy with the oncolytic parvovirus H-1PV,against cervicaland pancreatic cellsboth in culture and asxenograftsin immnuodeficientratsand mice. Theanalysisinculturedcellsisextensive,well-documented and reveals,ratherunexpectedly,that the major non-structuralprotein ofH-1PV isacetylated on lysineresidues,and thatthelevelof acetylation isincreased by the HDAC inhibitor,implying thatthe NS1 ofH-1PV may berecognized by theepigeneticmaintenancemechanismsoperating in itshostcell. Theauthorsgoontomapthetwolysineresiduesthatareacetylated,andusereversegeneticsand antibody pull-down experimentsto confirm thatthesearethemajorsitesofacetylation, and that, in turn, their de-acetylation isinhibited by VPA.Thisisan exciting,and unpredicted,new insightinto the biology of the parvoviruses that will surely lead to a greater understanding of the role of the multifunctionalNS1proteininviral replication, oxidative stress induction and resulting oncolytic activity.In thisrespect,the authorsshow,very convincingly,thatinhibiting the de-acetylation of these two residues with VPA leads to a synergistic elevation of viral DNA replication, progeny formation and oncolytic activity of H-1PV,both in vitro and in vivo. Indeed,the authors show thatcombined therapy with intermediate doses of H-1PV and low concentrationsofVPA can achieve complete remission in a numberofhuman xenografttumour models.Thatthemajorityofthesestudiesweredoneinnuderats,theratbeingthenaturalhost species ofH-1PV,goessomeway to alleviating many oftheconcernsassociated with using immunodeficient rodent xenograft models with oncolytic viruses that do not infect the host species. Thefindingthat"cured"immnunodeficientratsnotonlysurvivetheH1-PV infection,butremain healthy and tumour-free formore than a yearfollowing treatment,is very encouraging. In general,the manuscriptis veryclearly and tightly written and the authors'conclusionsare well- supported by the data presented.There are a smallnumberofminorpoints thatthe authors should considerthatwould benefitfrom clarification:- Page2,line11:towhatextentaretheK85andK257residuesconservedacrosstheparvoviruses? Page3,line18:Hristovetal.,2010? Page7,line2:"hallmarker"= "hallmark"? Page8,line15:althoughquantitationisnotprovidedforNS1,itappearsthatacetylationdoesnot affectthe levelofNS1 produced during infection - although itisknown to upregulate itsown,P4,EMBO MolecularMedicine PeerReview ProcessFile- EMM-2013-02796 © EMBO 3 promoteraswellasp38.Thismightbediscussed. Page9,line5:substitutingR forK atthesetwopositionsdoesnot"mimicthenon-acetylated state" - it conservesthebasicnatureoftheresiduewhilepresumably eliminating itsability to beacetylated. Page12,line17:"culminating"= "combining"? Page13,line11-12:thedirectroleforVPA through NS1 acetylation could betested using K85R;K257RdoublemutantH-1PV virus,forwhich oncolysisshould notbeenhanced by VPA if this is the sole effect of the HDAC inhibitor. Page18,line13:whatis"HMGS"? Page38,line6-7:thiscitation isincomplete. Page42,line16:"acetylatedin"= "acetylatedat"? Page42,line19:"trypsin-digested" Fig7:panelA - the key should be outside the plot - its current location makes this tiny figure even moredifficulttosee! PageS1,legendtoFigS2:sinceRT-qPCR istheoutputfortheseinfections,why use a recombinant H-1PV,and notjustmonitortheP4 transcriptsofthewildtypevirus? PageS2,legendtoFigsS4& S5:whatanimals- and whattumours?Presumably HeLa in Fig S5, butwhatin Fig S4? PageS3,legendtoFigS7:aren'tthetumoursinFigS5HeLa-derived,notAsPC-1 - ordo these PDAC cellsalsocontainHPV18sequences? PageS3,legendtoFigS8:what"analysis"wasperformed"after28days"? Referee#2: Severalaspectsofthisstudycallintoquestiontheclaim thattheHACiincreases viralreplication (by modifying directly NS1 and enhancing viralgene expression and replication). In Fig1A rather high MOI are used in severalcases implying thatthe virus is notable to spread in these cultures but rather is toxic if enough particles contact cells. In Fig4 a ten-fold change in MOIdoes notappreciably increase the titre produced,are these viral titres measured at plateau and so MOI does not matter? and then does VPA actually increase the plateau/maximaltitreachieved? Whenusing a rather resistant cell line (AsPC-1)in vivo theauthorsstillachievetumourregressions, buttheevidenceprovided forenhanced viralinfection oftumoursisnotconvincing (Fig 9c).Ifitis true that there is enhanced replication, spread and oncolysisthisshould bedemonstrated by increased viral titres in isolated tumours. Otherwise it is tempting to assume that the effects are largely due to some sort of enhanced bystander effect (possible due to innate immune mechanisms). Referee#3: ThemanuscriptbyLi,J.etal.describesaratherthoroughstudydemonstratingsynergybetween valproicacid and theratH-1PV oncolyticvirusin cervicaland pancreaticcarcinomacellmodels. Thestudiesaretechnicallyandmethodologicallyrobustandlogical conclusions are presented. EMBO MolecularMedicine PeerReview ProcessFile- EMM-2013-02796 © EMBO 4 Therefore,thequalityofthemanuscriptisverygood(besidessomelanguageissues)andwouldbe suitable forpublication.However,the findings are hardly noveland the proposed mechanism forthe synergistic effects is notconclusively verified. HDACinhibitorshavebeenexploredincombinationwithnumerousoncolyticmutantsandbeen demonstrated to almostuniversally increaseviruspotency through both virusand celllinedependent properties.Furthermore,only oneHDAC inhibitor (VPA) is tested and only 2 out of 14 Lys-residues in the NS1 protein are investigated and suggested as responsible for both viral gene expression, replication and synergistic cellkilling.Despite generation of stably expressing HeLa cells with the two lysine residues mutated to eliminate acetylation, it is not clear that this is the mechanism for the synergistic effects.How aboutacetylation ofNS1? Whatcellulareffects does VPA have - it is a very high dose(1mM)?How aboutotherHDAC inhibitors? BasedontheseconcernsIcannotrecommendpublicationofthemanuscriptincludingthecurrent datain Embo MolecularTherapy. Specificcomments: Thestatementintheintroduction"......tumourrelapseduetotheemergenceofvirus-resistantcancer cells." ismis-leading. To my knowledge, 'classical'resistance does not apply to most oncolytic viruses.Somecellsarenotsensitiveto virusbutemergenceofresistantcellsin thecourseofvirus- treatment has not been reported. Does this phrase refer specifically to H-1PV? Theuseofadoseof1mM VPA seemsextremelyhighandhardlywithinpharmacological applications.Please explain. In figure 1,why is notviralreplication measured in each cellline? Lysis could be caused by expression ofcytotoxic genes alone. Only in fig. 4 is replication measured and only in HeLa cells. In figure 3,the increase in NS1 acetylation by VPA is extremely faintand so is the VP induction (notsignificant?).Itis hard to believe thatthis is the cause of the increased cytotoxicity.Why were notacetylation induced? 1stRevision - authors'response 23 July 2013 ThankyouforsendingusyourcommentsandthoseoftheReviewersaboutourmanuscriptentitled “Synergisticcombination ofValproicacid and oncolyticparvovirusH-1PV asapotentialtherapy againstcervicaland pancreatic carcinomas”.We were pleased to receive positive feedback and gratefulforthesuggestionsgiven to improveourmanuscript. Wehavenowsuccessfully performed allthe additionalexperimentsrequested.Allnew resultssubstantiateourpreviousdataindicating that HDACIs enhance H-1PV oncolysisby multiplemechanismsincluding: 1) the induction of oxidativestressleading to increased DNA damage,apoptosisand celllysis; 2)theincreasein NS1 acetylation leading to a largervirusproduction in permissive tumor cells.Forthe sake ofclarity,we found itusefulto include an additionalfigure (new Fig.10)summarizing the results described in the presentstudy in theform ofatentativemodel. Wethankyouandthereviewersfortheconstructivecriticismsand we hope thatourmanuscript now fulfillsalltherequirementsforpublication in EMBO MolecularMedicine. Pleasefindbelow ourpoint-to point rebuttal letter to your and reviewers’ concerns: Yourcomments: ThankyouforthesubmissionofyourmanuscripttoEMBO MolecularMedicine.Wehavenow heard back from thethreerefereeswhom weasked to evaluate your manuscript. Although the referees find the study to be of potentialinterest,they also raise a number of concerns thatneed to beaddressed in amajorrevision ofyourmanuscript. Asyouwillseefrom theenclosedreports,therefereesfindthestudy interesting and comprehensive.EMBO MolecularMedicine PeerReview ProcessFile- EMM-2013-02796 © EMBO 5 Whileref#1and#2arepositiveaboutit,ref#3ismorecritical.Assuch,wewouldlikeyouto experimentally show 1.) the direct role of VPA through NS1 acetylation as suggested by ref#1 and #3. AUTHORS:Asmentionedabove,thepresentstudyrevealed two mechanismsthrough which Valproicacid(VPA)enhancesPV oncolyticactivities.However,theremaystillbeother mechanismsbywhichVPAcouldstimulateviruscytotoxicity.Inparticular,ithasbeenreported that HDACItreatment affects the expression of about 10% of all cellular genes. It is therefore conceivable thatsome ofthe genesaffected by VPA encode factorscontrolling virusreplication and/orcytotoxicity.On thebasisofourand othersdata,themechanismsby which VPA may boostH-1PV replication and cytotoxicityhavebeen extensivelydiscussed in thenew version of the manuscript on pages 13 and 14 (revised discussion section). 2) In addition,ref#2 is concerned by the high titters ofvirusesand would beconvinced of enhanced replication,spread and oncolysisifincreased viraltitrescould be shown in isolated tumours. AUTHORS:Weagreedwiththiscommentandperformednew experimentsusingbothHeLaand AsPC-1 xenograftratmodels,inwhichviralproductionwasmeasuredinisolatedtumours.Inboth animalmodels,weconfirmed thatVPA treatmentincreasesviralprotein levelsand infectiousvirus production (new Fig8D andE andnew Fig.9D andE). 3) Finally,ref#3hasseveral important criticisms, however we would like you to particularly focus on testing anotherHDAC inhibitorin a key experimentand differentdoses of VPA. AUTHORS:ThecombinationofH-1PV with a second HDAC inhibitor,namelysodium butyrate (NaB) has also been tested and isreported in therevised version ofthemanuscript.A seriesofnew experimentswasperformed in HeLa cellsshowing thatthe NaB treatmentactslike VPA in thatit increases: i) H-1PV mediated celllysisvia generation ofROS and DNA damage (new Supp.Fig3) ii) NS1intrinsiccytotoxicity(new SupportingInformation FigS4) iii) NS1acetylation(new Fig.3A andnew Fig5A) iv) NS1transcriptionalactivity(new SupportingInformation FigS5) VPA hasbeenalsotestedattheconcentrationof0.5mM in all seven cell lines analysed in this study,confirming previous results obtained using 1mM VPA (new SupportiveInformation Fig S1 and revised TableS1).We also explained thatboth 0.5 and 1mM doses wereselectedbasedonthe concentrationsthathave been established in patientswith epilepsy:0.5 mM (closeto thetypical therapeutic serum concentration of 0.6mM) and 1mM (close to the upper limit of antiepilectic range of0.9 mM).Thisinformation isprovided on page6 lines106-109 togetherwith references. 4) If you have data on hand addressing the other concerns of ref#3,we would strongly encourage you to include these in the manuscript. AUTHORS:Pleaserefertoourresponsetoreviewer3aboutoureffortstoaddresshis/her additionalconcerns. ***** Reviewer'scomments***** Referee#1: Thispaperreportsaremarkablycomprehensivestudyoftheuseofthehistonedeacetylaseinhibitor (HDAC) valproic acid (VPA) in combination therapy with the oncolytic parvovirus H-1PV,against cervicaland pancreaticcellsboth in cultureand asxenograftsin immnuodeficientratsand mice. Theanalysisinculturedcellsisextensive,well-documented and reveals,ratherunexpectedly,that the major non-structuralprotein ofH-1PV isacetylated on lysineresidues, and that the level of acetylation isincreased by the HDAC inhibitor,implying thatthe NS1 ofH-1PV may berecognized by theepigeneticmaintenancemechanismsoperating in itshostcell. EMBO MolecularMedicine PeerReview ProcessFile- EMM-2013-02796 © EMBO 6 Theauthorsgoontomapthetwolysineresiduesthatareacetylated,and use reverse geneticsand antibody pull-down experimentsto confirm thatthesearethemajorsitesofacetylation,and that,in turn, their de-acetylation isinhibited by VPA.Thisisan exciting,and unpredicted,new insightinto the biology oftheparvovirusesthatwillsurely lead to agreaterunderstanding oftheroleofthe multifunctionalNS1proteininviralreplication,oxidativestressinductionandresultingoncolytic activity.In thisrespect,the authorsshow,very convincingly,thatinhibiting the de-acetylation of these two residues with VPA leads to a synergistic elevation of viral DNA replication, progeny formation and oncolytic activity of H-1PV,both in vitro and in vivo. Indeed,the authors show thatcombined therapy with intermediatedosesofH-1PV and low concentrationsofVPA can achieve complete remission in a numberofhuman xenografttumour models.Thatthemajorityofthesestudiesweredoneinnuderats,theratbeingthenaturalhost species ofH-1PV,goessomeway to alleviating many of the concerns associated with using immunodeficient rodent xenograft models with oncolytic viruses that do not infect the host species. Thefindingthat"cured"immnunodeficientratsnotonlysurvivetheH1-PV infection,butremain healthy and tumour-free formore than a yearfollowing treatment,is very encouraging. In general,the manuscriptis very clearly and tightly written and the authors'conclusions are well- supported by the data presented. AUTHORS:Wethankthereviewerfor the kind comments.We were happy to know thathe/she found ourpaperinteresting and wellwritten. Thereareasmallnumberofminorpointsthattheauthorsshouldconsiderthatwouldbenefitfrom clarification: Page2,line11:towhatextentaretheK85andK257residuesconservedacrosstheparvoviruses? AUTHORS:Thisinformationhasbeenincludedinthediscussiononpage15lines348353. Page3,line18:Hristovetal.,2010? AUTHORS:Corrected.Thankyou Page7,line2:"hallmarker"= "hallmark"? AUTHORS:Corrected.Thankyou Page8,line15:althoughquantitationisnotprovidedforNS1,itappearsthatacetylationdoesnot affectthe levelofNS1 produced during infection - although itisknown to upregulate itsown,P4, promoteras wellas p38.This mightbe discussed. AUTHORS:InthisstudytheabilityofNS1toactivateitsownP4promoterinthepresenceor absenceofVPA hasbeen notinvestigated in details.Whilecellcultureexperimentsusing HeLa cells suggestthatVPA treatmentdoesnotresultin a significantchangein NS1 protein levelsduring H- 1PV infection,experimentsin animalsshow a clearinduction.Wefeelthatfurtherstudiesare required to clarify this issue.This pointis mentioned in the discussion on page 15 lines346 and 347. Page9,line5:substitutingR forK atthesetwopositionsdoesnot"mimicthenon-acetylated state" - it conserves the basic nature of the residue while presumably eliminating its ability to be acetylated. AUTHORS:Thankyouforthis comment. We have revised the text accordingly (line 202). Page12,line17:"culminating"= "combining"? AUTHORS:wehaverevisedtheentiresentencetomakeitclearer (Page 13 lines 304-305). Page13,line11-12:thedirectroleforVPA through NS1acetylationcouldbetestedusingEMBO MolecularMedicine PeerReview ProcessFile- EMM-2013-02796 © EMBO 7 K85R;K257RdoublemutantH-1PV virus,forwhich oncolysisshould notbeenhanced by VPA if this is the sole effect of the HDAC inhibitor. AUTHORS:wehaveaddressedthisquestionabove(ourresponsetotheeditor’scomments,point1). Page18,line13:whatis"HMGS"? AUTHORS:theabbreviationhasbeenexplained(line456). Page38,line6-7:thiscitation isincomplete. AUTHORS:Corrected.Thankyou. Page42,line16:"acetylatedin"= "acetylatedat"? AUTHORS:Corrected.Thankyou. Page42,line19:"trypsin-digested" AUTHORS:Corrected.Thankyou. Fig7:panelA - the key should be outside the plot - its current location makes this tiny figure even moredifficulttosee! AUTHORS:Wehaverevisedthe figure accordingly. PageS1,legendtoFigS2:sinceRT-qPCR istheoutputfortheseinfections,why usearecombinant H-1PV,and notjustmonitortheP4 transcriptsofthewildtypevirus? AUTHORS:Thepurposeofthisexperimentwasnottomonitorthe activity of the parvovirus P4 promoter.TherecPV-GFPviruscontainstheGFPgeneplacedunderthecontroloftheparvovirus P38promoterwhichistrans-activated byNS1.Thisvectorwasthusused to confirm thatin the presenceofVPA thereisan increase in NS1 transcriptionalactivity.Resultsare in agreementwith data obtained bydualluciferaseassayand with theVPA-dependentincreasein VP1 and VP2 protein levelsobserved upon H-1PV infection. PageS2,legendtoFigsS4& S5:whatanimals- and whattumours?Presumably HeLain Fig S5, butwhatin Fig S4? PageS3,legendtoFigS7:aren'tthetumoursinFigS5HeLa-derived,notAsPC-1 - ordo these PDAC cellsalsocontainHPV18sequences? PageS3,legendtoFigS8:what"analysis"wasperformed "after 28 days"? AUTHORS:Allthethreefigurelegendshavebeenrevisedaccordingtothereferee’ssuggestions. Thankyou. Referee#2: Severalaspectsofthisstudycallintoquestiontheclaim thattheHACiincreasesviralreplication (by modifyingdirectlyNS1andenhancingviralgeneexpressionandreplication). In Fig1A rather high MOI are used in severalcases implying thatthe virus is notable to spread in these cultures but rather is toxic if enough particles contact cells. AUTHORS:Weagreewiththereviewer.AmongthecelllinesusedonlyHeLacellssustainefficient viralmultiplication,while the othercelllinestested are low permissive forvirusreplication (our unpublished resultsand Dempeetal.IJC 2010).Thisinformation isprovided in thetexton page6 lines 104-106.EMBO MolecularMedicine PeerReview ProcessFile- EMM-2013-02796 © EMBO 8 In Fig4 a ten-fold change in MOIdoes notappreciably increase the titre produced,are these viral titres measured at plateau and so MOI does not matter? and then does VPA actually increase the plateau/maximaltitre achieved? AUTHORS:Wethankthereviewerforthiscommentwhichgaveusthepossibilitytoexaminein greaterdetailH-1PV production in HeLa cells.Wecarried outa timecourseexperimentin which virusproduction and release wasmonitored every day fora totaloffive days.Resultsfrom this experimentconfirmed thatthe presence ofVPA stimulatesboth virusproduction and release. Interestingly,while viraltitters in the absence of VPA reached a plateau after 72 hours, VPA stimulated virusproduction beyond thatlimit.Thisobservation hasbeen included in thenew version ofthemanuscripton page9 lines179-182 and in thenew Fig4B. Whenusingaratherresistantcellline(AsPC-1)in vivo theauthorsstillachievetumourregressions, buttheevidenceprovided forenhanced viralinfection oftumoursisnotconvincing (Fig 9c).Ifitis true that there is enhanced replication, spread and oncolysis this should be demonstrated by increased viral titres in isolated tumours. Otherwise it is tempting to assume thatthe effects are largely due to some sort of enhanced bystander effect (possible due to innate immune mechanisms). AUTHORS:PleaseseeouranswertotheEditor’scomments,point2. Referee#3: ThemanuscriptbyLi,J.etal.describesaratherthoroughstudydemonstratingsynergybetween valproicacid and theratH-1PV oncolyticvirusin cervicaland pancreaticcarcinomacellmodels. Thestudiesaretechnicallyandmethodologicallyrobustandlogicalconclusionsarepresented. Therefore,thequalityofthemanuscriptisverygood(besidessomelanguageissues)andwouldbe suitable forpublication.However,the findings are hardly noveland the proposed mechanism forthe synergistic effects is not conclusively verified. HDACinhibitorshavebeenexploredincombinationwithnumerousoncolyticmutantsandbeen demonstrated to almostuniversally increaseviruspotency through both virusand celllinedependent properties. AUTHORS:Weonlypartlyagreewith thereviewer’scomment.Whileoncolyticviruseshavebeen indeed tested in combination with various HDAC inhibitors as duly reported in our introduction and discussion,westillbelievethatthesynergisticoncosuppressiveeffectoftheH-1PV/HDACI combination againstcervicaland pancreatic carcinomaswasnotpredictable from the literature and representsan originaldiscovery.Besidesimproving substantiallyparvoviruscytotoxicity,this combination allowed forthe firsttime fulltumourremission to be achieved. To the best of our knowledge,no similarachievementhasbeen reported forotheroncolytic virusesin the animal modelsusedinthepresentstudy.Furthermore,weunravelledacompletelynovelmechanism of regulation ofthe cytotoxic NS1 protein bydiscovering thatNS1 isacetylated and thatVPA increases the acetylation status of the protein and thereby enhances parvovirus cytotoxicity. This finding not onlyimprovesourknowledgeofparvovirusbiology, buthasalso importantimplicationsregarding the clinical use of this virus in cancer therapy. Furthermore,onlyoneHDAC inhibitor(VPA)istestedandonly2outof14Lys-residues in the NS1proteinareinvestigatedandsuggestedasresponsibleforviralgeneexpression,replication and synergistic cellkilling. AUTHORS:ThereisamisunderstandingonthenumberofLys-residues investigated,which is by far higherthan 2.Ouranalysisactuallycovered 74.2% oftheentireNS1 sequenceincluding 36 outof 48 Lysresidues(see Fig.5B) Only 12 lysines were nottested by our MS analysis.This has been now explained in greaterdetailon page 9 line 199 (resultssection)and on page 15 lines360-361 (discussion section). Despitegenerationofstablyexpressing HeLa cellswith the two lysine residuesmutated to eliminate acetylation,itisnotclearthatthisisthe mechanism forthe synergistic effects.How about acetylation ofNS1? WhatcellulareffectsdoesVPAhave- it is a very high dose (1mM)? EMBO MolecularMedicine PeerReview ProcessFile- EMM-2013-02796 © EMBO 9 AUTHORS:Asmentionedabove,theVPAconcentrationof1mM usedinourcellculture experimentsisclose to the upperlimitofthe therapeutic concentration range (0.6-0.9 mM) established in patientswith epilepsy (see ouranswerto the editor’scomments,point 3). Nevertheless,toascertaintherelevanceofourresults,wehavefollowedtheEditor’ssuggestionand repeated LDH experiments using the VPA ata 0.5 mM concentration.Under these conditions also improved H-1PV oncolysiscould bedemonstrated in thepresenceofVPA.Itisalso noteworthythat the VPA concentration used in animal experimentation (100 mg/kg) correspondsto the human equivalentof16 mg/kg ascalculated according to the body surface area normalization method recommended by the Food and Drug Administration for conversion ofdrug doses between species (Reagan-Shaw etal,2008). This dose is within the clinical range of 15-30 mg/kg used forlong-term treatment of epileptic patients and far below the limit of 60 mg/kg considered safe and well tolerated in humans (Atmacaetal,2007). Thisimportantinformation hasbeen added in ourmanuscripton page19-20 lines461 -467 (materialsand methodssection). How aboutotherHDACinhibitors? AUTHORS:AssuggestedbytheEditorandthisReviewerafullseriesofexperimentshasbeen repeated using a second HDACinhibitor,namelyNaB.Thesedatahavenowbeenincludedinthe revised version ofthe manuscript(please see our answer to the Editor’s comments,point3). BasedontheseconcernsIcannotrecommendpublicationofthemanuscriptincludingthecurrent datain Embo MolecularTherapy. AUTHORS:Wehopethattheadditionaldataincludedinthisrevisedversionwillconvincethis reviewer thatour manuscriptis suitable for publication in EMM.In our opinion,we have addressed allhis/herconcerns. Specific comments: Thestatementintheintroduction"......tumourrelapseduetotheemergenceofvirus-resistantcancer cells." ismis-leading. To my knowledge, 'classical'resistance does not apply to most oncolytic viruses.Somecellsarenotsensitiveto virusbutemergenceofresistantcellsin thecourseofvirus- treatment has not been reported. Does this phrase refer specifically to H-1PV? AUTHORS:Emergenceofresistantcellsinthecourseofparvovirustreatmenthasbeenreported underin vitro conditions.Itsoccurrencein vivo hasnotindeed been described so far.However,this possibilitydeservesin ouropinion,to beconsidered dueto thehigh mutation rateofcancercells.As the parvovirus relies on host cell factors for its replication and cytotoxicity,inactivating mutations in some of the genes encoding these factors may result in acquisition of resistance. Theuseofadoseof1mM VPA seemsextremelyhighandhardlywithinpharmacological applications.Please explain. AUTHORS:Seeouranswerabove. In figure 1,why is notviralreplication measured in each cellline? Lysis could be caused by expression ofcytotoxic genesalone.Only in fig.4 isreplication measured and only in HeLa cells. AUTHORS:Itwasbeyondtheinitialscopeofthis study to investigate viral replication in all the cancercelllinesused.We stilltook thiscommentinto consideration and extended ouranalysisto two other cervical carcinoma cell lines (SiHa and CaSki). As in the case of HeLa cells, an increase in parvovirusreplication wasalso observed in thesecellsupon VPA treatment(NewSupportive Information Fig S6).Moreover,enhanced virus replication has been also demonstrated in vivo in tumours from VPA-treated animals (see our response to the Editor’s comments,point2). In figure 3,the increase in NS1 acetylation by VPA is extremely faintand so is the VP induction (notsignificant?).Itis hard to believe thatthis is the cause of the increased cytotoxicity.Why were notacetylation induced? AUTHORS:Asdocumented,inthenew versionofthemanuscriptwetestedthelevelsofNS1EMBO MolecularMedicine PeerReview ProcessFile- EMM-2013-02796 © EMBO 10 acetylation in thepresenceand absenceofVPA attwo differenttimepoints(16 and 32 hours) and showed thatVPA treatmentenhances NS1 acetylation in a time-dependentmanner(Newfigure3A). 2nd EditorialDecision 05 August2013 ThankyouforthesubmissionofyourrevisedmanuscripttoEMBO MolecularMedicine.Wehave now heard back from thetwo Reviewerswhom weasked to evaluateyourmanuscript. YouwillseethatwhileReviewer1issupportiveofyouwork,Reviewer3hasoneremaining issue that requires your action before we can accept your manuscript for publication. PleasefullyaddresstheReviewer3'sconcernasquicklyaspossibleandinanycasewithintwo weeks.Providedtheseissuesarefullyaddressed,thefinaldecisionwillbemadeattheEditorial level. I look forward to receiving your re-revised manuscriptas soon as possible and in any case within two weeks. ***** Reviewer'scomments***** Referee #1 (Comments on Novelty/ModelSystem): Theauthorshavesatisfactorilyaddressedallofthisreviewer'sconcerns. Referee#1(GeneralRemarks): In revising their manuscript,the authors have significantly improved an already comprehensive and ground-breaking study Referee#3(GeneralRemarks): Thecurrentversionofthemanuscriptismuchimprovedandinmyopinionitinterestingandshould bepublished.However,Ileaveitfortheeditorto decidewhetheritissuitableforpublication in EMBO MolecularMedicine. Thedataappeartobeofhighquality,reproducibleand severalkey findingswereverified in more than one model system and with two different HDAC inhibitors. While evidence supporting the claimsin the title and abstractare clear,evidence ofa directmechanism (through Lys-acetylation of the virus) for the increased cell killing when virus and drugs are combined are not completely convincing only strongly indicated. Mostofthisreviewer'squeriesandcommentsfromthepreviousversionhavebeenaddressedinthe revised manuscriptexceptthe following thatneed clarification to avoid misunderstanding: Line68-71:Thismakesnecessary to reinforcetheantineoplasticactivitiesofH-1PV to makeit moreeffective.Forthis,twoapproachesmightbeused:sensitizingthetumourcellsto parvovirus cytotoxicity and/orkilling virus-resistanttumour cells by other means. Comment:Theauthors'statementisstillmisleadingsincetheimpressionisthattheyaretalking abouttumor-tissue properties in the clinical setting not in cultured celllines.In the response to my commentsin the previousreview,they agreed thatviralresistance (using the same definition asfor drug-resistance) has notbeen demonstrated in vivo only in cultured celllines.Therefore,this should bespecified in the introduction.Resistance isa difficultterm to define and possible misunderstandingshouldbeavoided. EMBO MolecularMedicine PeerReview ProcessFile- EMM-2013-02796 © EMBO 11 2nd Revision - authors'response 05 August2013 Referee#1(CommentsonNovelty/ModelSystem): Theauthorshavesatisfactorilyaddressedallof this reviewer's concerns. Referee#1(GeneralRemarks): In revising their manuscript,the authors have significantly improved an already comprehensive and ground-breaking study AUTHORS:Thankyouverymuchforyourkindcomments. Referee#3(GeneralRemarks): Thecurrentversionofthemanuscriptismuchimprovedandinmyopinionitinterestingandshould bepublished.However,Ileaveitfortheeditorto decidewhetheritissuitableforpublication in EMBO MolecularMedicine. Thedataappeartobeofhighquality,reproducibleandseveralkeyfindingswereverifiedinmore than one model system and with two different HDAC inhibitors. While evidence supporting the claimsin the title and abstractare clear,evidence ofa directmechanism (throughLys-acetylation of the virus) for the increased cell killing when virus and drugs are combined are not completely convincing only strongly indicated. AUTHORS:Asmentionedabovewewerehappytoknow thatthereviewerfoundourmanuscript improved and interesting. Mostofthisreviewer'squeriesandcommentsfromthepreviousversionhavebeenaddressedinthe revised manuscriptexceptthe following thatneed clarification to avoid misunderstanding: Line68-71:Thismakesnecessary to reinforce the antineoplastic activities of H-1PV to makeit moreeffective.Forthis,twoapproachesmightbeused:sensitizingthetumourcellstoparvovirus cytotoxicity and/orkilling virus-resistanttumour cells by other means. Comment:Theauthors' statement is still misleading since the impression is that they are talking abouttumour-tissue properties in the clinical setting not in cultured cell lines. In the response to my commentsin the previousreview,they agreed thatviralresistance (using the same definition as for drug-resistance) has notbeen demonstrated in vivo only in cultured celllines.Therefore,this should bespecified in theintroduction.Resistanceisadifficultterm to defineand possible misunderstandingshouldbeavoided. AUTHORS:Toavoidanypossiblemisunderstandingwiththewordresistancewehavedecidedto substitute the old sentence (...In the framework ofcancer therapy and as also observed with other oncolyticviruses,thereisstilla riskoftumour relapse due to the presence ofparvovirus-resistant cancercells.Thismakesnecessary to reinforce the antineoplastic activitiesofH-1PV to makeit moreeffective.Forthis,twoapproachesmightbeused:sensitizingthetumor cellsto parvovirus cytotoxicity and/or killing virus-resistanttumor cellsby othermeans)with thisnew one: “Duetotheirgeneticheterogeneity,itislikelythatsomeofthecancercellswithinatumourwill havea differentsensitivityto H-1PV.Itisthereforeimportantto reinforce the antineoplastic activity ofthevirusin orderto improveitsclinicaloutcomein such a scenario. Thiscanbeachievedbydevelopingcombinationstrategiesbasedonvirusandotheranticancer agentsthatincreasecancercellkilling whileminimizingtoxicsideeffects.”(introduction,lines67- 71).