WO2016094869A1

WO2016094869A1 - Methods and compositions relating to high resolution structure of non-segmented negative strand virus l proteins

Info

Publication number: WO2016094869A1
Application number: PCT/US2015/065388
Authority: WO
Inventors: Bo Liang; Zongli LI; Simon JENNI; Nikolaus GRIGORIEFF; Stephen Harrison; Sean P.J. Whelan
Original assignee: Harvard University
Current assignee: Harvard University
Priority date: 2014-12-12
Filing date: 2015-12-11
Publication date: 2016-06-16
Anticipated expiration: 2017-06-12

Abstract

Described herein are methods and compositions relating to the high resolution 3D structure determination for non-segmented negative strand (NNS) virus multifunction L proteins. High resolution 3D structure as described permits the application of a range of rational drug design and screening approaches to the identification of inhibitors of any of the varying functions of the NNS virus L proteins and thereby facilitates the identification of antiviral agents.

Description

M ETH O D S A N D C O M P O SITIO N S REL A TIN G TO H IGH RESO L UTIO N STRUC TURE O F N O N -SEGM EN TED N EGA TIV E STRA N D V IRUS L P RO TEIN S

C RO SS-REFEREN C E TO REL A TED A P P L IC A TIO N This application claims benefitu nd er35U.S.C .§ 119(e)of U.S.P rovisionalA pplication N o.

62/091,261,filed D ecember12,2014,the contentof whichis incorporated herein by reference in its entirety. GO V ERN M EN T SUP P O RT

This invention was mad e with governmentsu pportu nd er R01-A I059371 award ed by the N ational Institu tes of H ealth(N IH ).The governmenthas certain rights in the invention. FIEL D O F TH E IN V EN TIO N The technology d escribed herein relates to the id entification of anti-viraltherapeu tics. B A C KGRO UN D The non-segmented negative-strand (N N S)RN A viru ses inclu d e some of the mostlethalhu man and animalpathogens,inclu d ing Ebolaviru s and rabies viru s.H u man respiratory syncytialviru s infects 95% of infants before the age of 2 and can resu lt in severe respiratory d isease requ iring hospitalization.A ccord ing to the C enter for d isease C ontrol,seasonalRSV infection accou nts for 172,000 hospitalizations and 1.5 million ou tpatientvisits amongchild ren you ngerthan 5 yrs old ,and 14,000 d eaths amongad u lts old erthan 65 years.C u rrenttherapeu tics forRSV infection are limited in efficacy and high in cost.Therapeu tics su ch as ribavirin have been u sed in the clinic,however evid ence thatitinhibits RS viralreplication has been lacking. Formany members of the N N S family viru ses there are no vaccines orefficaciou s antivirald ru gs. Fu rthermore,rapid d evelopmentof d ru gresistance to monotherapy has been observed forotherRN A viru ses,su ch as influ enza (van d er V ries etal.,2010;Zhu etal.,2012),su ggesting thatmu ltiple antivirals will be requ ired for long-term effective treatment of these d iseases.Therefore,the d evelopmentof new therapies is warranted ,especially ones thatcou ld targetmu ltiple members of this hu man-pathogen lad en viru s ord er. The mu ltifu nctional,large (L )polymerase proteins,carried within the virions (B altimore etal.,1970), are prod u ced by allN N S family members and have biochemicalproperties thatd istingu ishthem from mostother RN A polymerases of viru ses and polymerases of their hosts,which mightmake them su itable as antiviraltargets.In ad d ition to their RN A -d epend entRN A polymerase (Rd Rp)activity (Emerson and W agner,1973),N N S RN A viru s L proteins catalyze an u nu su alsequ ence of mRN A cappingreactions (H ercyketal.,1988 ),and the Rd Rpitself polyad enylates the viralmessage (H u ntet al.,1984). Relatively u ntargeted screens have id entified smallmolecu les thatappear to interactwith and /or inhibitcertain N N S viru s L proteins.B oehringerIngelheim has reported the id entification of asmall molecu le thatappears to targetL protein of RSV . A n inhibitorof the polymerase from the measles viru s id entified by mappingaresistance mu tation to the Rd Rpd omain of L ,has no activity againstthe polymerase of N N S family memberRSV .A n ad enosine analogwas reported thatinhibits Ebolaviru s in acellbased assay,bu tinterfered withreplication of many viru ses,raisingthe qu estion of its impact on the hostcell. N N S viru s L proteins have been d ifficu ltto stu d y becau se of theirsize,low expression and d earthof robu stbiochemicalassays of theirfu nctions. SUM M A RY O F TH E IN V EN TIO N The technology d escribed herein relates to the id entification of highresolu tion 3D stru ctu re forthe L proteins of N N S viru ses and the u se of su ch stru ctu res to id entify agents thatinhibitone ormore activities of the L protein. M olecu les or compou nd s thatinhibitL protein activity can provid e antiviraltherapies by d isru ptingviralreplication,whichd epend s u pon the activities of the L protein. W here the mu ltifu nctionalN N S viru s L protein is essentially u niqu e to the N N S family of viru ses, therapies thatselectively targetL protein activity have agood likelihood of inhibitingviralreplication withared u ced likelihood of u nwanted effects on the hostcellororganism.3D stru ctu res can permit the rationald esign and /orrefinementof cand id ate inhibitormolecu les.H owever,to d ate,the size and d ynamic stru ctu re of the N N S viru s L proteins has preclu d ed the application of stand ard approaches to d etermine their3D stru ctu re.The technology d escribed herein stems,in part,from the solu tion of an N N S viru s L protein at high resolu tion,which permits the application of compu ter-mod eling strategies to d esign orotherwise id entify cand id ate inhibitormolecu les. W hen u sed in combination withassays d escribed herein forthe variou s activities of the N N S viru s,the highresolu tion stru ctu re also permits the confirmation and refinementof inhibitory molecu les as antiviralagents. The N N S viru s L protein exemplified herein athigh resolu tion is thatof vesicu larstomatitis viru s (V SV ). The approaches d escribed herein to obtaining thathigh resolu tion stru ctu re are read ily applied to the L proteins of other N N S viru ses. H owever,notonly d oes the solu tion of the high resolu tion stru ctu re forthe V SV L protein d escribed herein d emonstrate the feasibility of generating su ch stru ctu res for any of the N N S viru ses by,e.g.,cryoelectron microscopy,bu t becau se the stru ctu res of N N S viru s L proteins are well-conserved between members of the N N S viru s family,the solu tion of the V SV protein’s stru ctu re also permits the u se of thatstru ctu re to pred ict,viacompu ter mod elingu singamino acid sequ ence information forotherN N S viru s L proteins,the highresolu tion stru ctu res of those L proteins as well. In one aspect,d escribed herein is amethod of id entifying an inhibitorof anon-segmented negative strand (N N S)viru s,the method comprising:(a)u singahighresolu tion 3D stru ctu re of an N N S viru s L protein to selectacand id ate inhibitorfrom avirtu allibrary of 3D stru ctu res of potentialcand id ates by performingstru ctu re based compu tationalmod eling;(b)obtainingthe cand id ate inhibitor;and (c) measu ringan activity of L protein when contacted with the cand id ate inhibitorobtained in step(b), wherein ad ecrease in the activity confirms thatthe cand id ate is an inhibitorof the N N S viru s. In another aspect,d escribed herein is a method of id entifying an inhibitor of a non-segmented negative strand (N N S) viru s polymerase,the method comprising:(a) u sing a high resolu tion 3D stru ctu re of an N N S viru s L protein to selecta cand id ate inhibitor from a virtu allibrary of 3D stru ctu res of potential cand id ates by performing stru ctu re based compu tational mod eling; (b) obtaining the cand id ate inhibitor;and (c)measu ring L protein polymerase activity when contacted withthe cand id ate inhibitorobtained in step(b),wherein ad ecrease in the activity of the polymerase confirms thatthe cand id ate is an inhibitorof the N N S viru s polymerase. In another aspect,d escribed herein is a method of id entifying an inhibitor of an N N S viru s,the method comprising:(a)screeningalibrary of cand id ate molecu les forinhibition of an N N S L protein in an in vitro biochemicalassay foran L protein activity;(b)u singahighresolu tion 3D stru ctu re of the L protein to d etermine the bind ing site of potentialcand id ates obtained in step(a)by compu ter mod eling;(c)id entifyingapotentialrefinementto the cand id ate inhibitor;(d )obtainingacand id ate inhibitormolecu le comprisingthe refinementid entified in step(c);and (e)measu ringthe activity of the L protein when contacted withthe cand id ate inhibitorobtained in step(d ),wherein ad ecrease in the activity of the L protein confirms thatthe cand id ate is an inhibitorof the N N S viru s. In anotheraspect,d escribed herein is amethod of id entifying an inhibitorof L protein of an N N S viru s,the method comprising:(a)u singahighresolu tion 3D stru ctu re of the L protein of vesicu lar stomatitis viru s (V SV )to create an atomic mod elof the L protein of asecond N N S viru s by homology mod elingthe amino acid sequ ence of the L protein of the second N N S viru s onto the highresolu tion 3D stru ctu re of the V SV L protein ;(b)u singthe atomic mod elof the second N N S viru s L protein to select a cand id ate inhibitor from a virtu al library of 3D stru ctu res of potential cand id ates by performing stru ctu re based compu tationalmod eling;(c) obtaining the cand id ate inhibitor;and (d ) measu ring an activity of L protein of the second N N S viru s when contacted with the cand id ate inhibitor obtained in step (b),wherein a d ecrease in the activity confirms thatthe cand id ate is an inhibitorof the second N N S viru s. In one embod imentof each of the aspects noted above,the high resolu tion 3D stru ctu re has a resolu tion of atleast3.8 Å . In anotherembod iment,the high resolu tion 3D stru ctu re is an atomic stru ctu re d etermined viacryo-electron microscopy and single particle analysis. In another embod imentof each of the aspects noted above,the N N S viru s L protein is vesicu lar stomatitis viru s L protein. In anotherembod iment,the V SV L protein high resolu tion stru ctu re has space grou pP 1,so as to form au nitcellof d imensions of a=112.0 Å ,b=143.0 Å ,c=106.0 Å and α,β,γ=90o. In one embod iment,the V SV L protein highresolu tion 3D stru ctu re has the coord inates as setou t,for example,in the stru ctu re d eposited withthe RC SB P rotein D ataB ankwithReference N u mber5A 22 (this stru ctu re is accessible on the W orld W id e W eb atrcsb.org/pd b/explore.d o?stru ctu reId =5A 22). In anotherembod imentof eachof the aspects noted above,the activity of the L protein measu red is selected from the grou p consisting of RN A d epend ent RN A polymerase activity, GD P polyribonu cleotid yltransferase (P RN Tase) activity, methyltransferase activity,and the ability to stru ctu rally rearrange L protein d omains d u ring L protein fu nctions.W hen the L protein activity measu red is RN A d epend entRN A polymerase activity,thatactivity can be fu rtherselected from the grou pconsistingof transcriptinitiation,elongation and termination. In anotherembod imentof certain aspects noted above,ratherthan selectingacand id ate inhibitorfrom avirtu allibrary of 3D stru ctu res of potentialcand id ates,the highresolu tion 3D stru ctu re of the N N S viru s L protein is u sed to d esign apotentialinhibitor.In one embod iment,the potentialinhibitoris d esigned to interactwith an L protein d omain selected from the grou p consisting of the RN A d epend entRN A polymerase (Rd Rp) d omain,the mRN A capping d omain,the methyltransferase d omain,the loopprimerd omain,the connectord omain,and the C -terminald omain. In anotherembod imentof eachof the aspects noted above,the id entified inhibitorbind s ad omain of the L protein ataregion selected from the grou pconsistingof the RN A d epend entRN A polymerase (Rd Rp)d omain,the mRN A cappingd omain,the methyltransferase d omain,the loopprimerd omain, the connectord omain,and the C -terminald omain of the L protein. In anotherembod imentof eachof the aspects noted above,the id entified inhibitorattenu ates the RN A d epend entRN A polymerase activity,the GD P polyribonu cleotid yltransferase (P RN Tase)activity,the methyltransferase activity,the loop priming activity,or the ability of the L protein to stru ctu rally rearrange its d omains d u ringL one ormore of its fu nctions. In anotherembod imentof eachof the aspects noted above,the method fu rthercomprises the stepof contactingthe confirmed cand id ate inhibitorwithacellinfected withan N N S viru s. In anotherembod imentof eachof the aspects noted above,the N N S viru s is selected from the viru ses of the Filovirid ae,P aramyxovirid ae,Rhabd ovirid ae and B ornavirid ae families. In one embod iment, the N N S viru s is a Filoviru s,selected from Ebola viru s,M arbu rg viru s and C u evaviru s. In one embod iment,the N N S viru s is a P aramyxoviru s selected from measles viru s,N ipah viru s,H end ra viru s,respiratory syncytialviru s and N ewcastle d isease viru s.In one embod iment,the N N S viru s is a Rhabd oviru s selected from L yssaviru s,V esicu loviru s,P erhabd oviru s,Sigmaviru s,Ephemeroviru s, Tibroviru s,Tu paviru s and Spriviviru s. In another aspect,d escribed herein is method of mapping the location of resistance to a known inhibitor of an N N S viru s L protein on the 3D stru ctu re of the L protein,the method comprising homology-based stru ctu ralmappingof the sequ ence comprisingthe resistance mu tation onto the high resolu tion 3D stru ctu re of an N N S viru s L protein. The high resolu tion 3D stru ctu re can be,for example,a3.8 Å orgreaterresolu tion atomic stru ctu re of the chosen L protein. In anotheraspect,d escribed herein is amethod of inhibiting an N N S viru s,the method comprising contactingan N N S viru s oracellinfected withan N N S viru s withan inhibitorid entified accord ingto the method of any one of the aspects d escribed herein above and /orin the Examples. The inhibitor can be ad ministered to an ind ivid u alin need of treatmentorprophylaxis forinfection withthe N N S viru s. In anotheraspect,d escribed herein is an N N S viru s inhibitorid entified u singthe method of any one of the aspects d escribed herein above and /orin the Examples. In anotheraspect,d escribed herein is amethod of d esigningacand id ate inhibitorof an N N S viru s,the method comprising:(a) id entifying an N N S viru s mu tantthatescapes the activity of a known inhibitorof the L protein of an thatviru s;(b) id entifyingthe site of the L protein resistance mu tation by sequ ence analysis; (c)u singhomology-based stru ctu ralmappingof the sequ ence comprisingthe resistance mu tation to mapthe mu tation site onto ahighresolu tion 3D stru ctu re of an N N S viru s L protein;(d )u sing compu termod eling with the high resolu tion 3D stru ctu re to selecting acand id ate bind ingmolecu le thattargets the site of the resistance mu tation mapped in step(c);and (e)obtaining the cand id ate inhibitor molecu le id entified in step (d ). In one embod iment,the method fu rther comprises the stepof contactingL protein of the N N S viru s withthe cand id ate inhibitorin an in vitro assay of L protein activity,wherein ad ecrease in L protein activity ind icates the cand id ate inhibitoris an N N S viru s inhibitor. In anotherembod iment,the method fu rthercomprises the stepof contacting the cand id ate inhibitor with the N N S viru s and assaying viralfu nction,wherein ad ecrease in viralfu nction ind icates the cand id ate inhibitoris an N N S viru s inhibitor. In anotheraspect,d escribed herein is amethod of d esigningacand id ate inhibitorof an N N S viru s,the method comprising:(a) id entifying an N N S viru s mu tantthatescapes the activity of a known inhibitorof the L protein of an thatviru s;(b) id entifyingthe site of the L protein resistance mu tation by sequ ence analysis; (c)u singhomology-based stru ctu ralmappingof the sequ ence comprisingthe resistance mu tation to mapthe mu tation site onto ahighresolu tion 3D stru ctu re of an N N S viru s L protein;(d )id entifyingapotentialrefinementto the known inhibitorthatpermits interaction withthe mu tantL protein;(e)obtainingthe inhibitormolecu le comprisingthe refinementid entified in step(d ); and (f) assaying mu tantviru s replication or infection in the presence and absence of the refined inhibitor molecu le,wherein a d ecrease in replication or infection ind icates the refined inhibitor is effective againstthe mu tantviru s. In anotheraspect,d escribed herein is the u se of an N N S viru s inhibitorid entified accord ing to the method s of any one of the aspects orembod iments d iscu ssed above and /ord escribed in the Examples, forthe treatmentof an ind ivid u alinfected with an N N S viru s. In su ch u se,the N N S viru s can be selected ,for example,from the grou p consisting of viru ses of the Filovirid ae,P aramyxovirid ae, Rhabd ovirid ae and B ornavirid ae families. In anotheraspect,d escribed herein is acompu ter-read able med iu m withhighresolu tion 3D stru ctu re coord inate information foran N N S L protein stored thereu pon. D EFIN ITIO N S: A s u sed herein,the term " non-segmented negative-strand (N N S)viru s" refers to aviru s thatbelongs to su perfamily M ononegavirales and inclu d es the families Rhabd ovirid ae (inclu d ing,bu tnotlimited to vesicu larstomatitis viru s (V SV )and rabies viru s),P aramyxovirid ae (inclu d ing,bu tnotlimited to measles and respiratory syncytialviru ses (RSV ),Filovirid ae (inclu d ing,bu tnotlimited to Ebolaand M arbu rgviru ses)and B ornavirid ae (inclu d ing,bu tnotlimited to B ornad isease viru s). A s u sed herein,the term“L protein of an N N S viru s”refers to the large mu ltifu nctionalproteins of non-segmented ,negative strand RN A viru ses thatcatalyzes RN A -d epend entRN A polymerization with viralribonu cleoprotein as template,a non-canonicalsequ ence of capping and methylation reactions,and polyad enylation of viralmessages. A s u sed herein,the term“L protein activity”refers to any of the biologicalactivities of an N N S viru s L protein involved in and /ornecessary forviralreplication.A ctivities can inclu d e any of,forexample, RN A d epend entRN A polymerase activity (incu d ing initiation,elongation and termination),GD P phosphoribonu cleotid yltransferase activity,methyltransferase activity,viralmRN A capping activity, and the ability of the L protein to rearrange one ormore if its d omains d u ring L protein fu nction necessary forviralreplication. A s u sed herein,the term“d omain,”when u sed in reference to an N N S viru s L protein orto aportion thereof,refers to aportion of an N N S viru s L protein thatperforms aparticu larfu nction,e.g.,an enzymatic fu nction inclu d ing RN A -d epend ent RN A polymerase [Rd Rp] , polyribonu cleotid yl transferase [P RN Tase] ,and methyltransferase fu nctions and /orthatd efines astru ctu ralfeatu re,e.g.,a connectord omain and /orC -terminald omain. The d omains of an N N S viru s L protein are shared across members of the family,su ch thatalignmentof,e.g.,the V SV L protein with thatof other family members permits the id entification of the correspond ingd omains in the otherfamily members. Sequ ence alignmentin conju nction withahighresolu tion 3D stru ctu re permitthe application of the 3D stru ctu re of one N N S viru s L protein to make pred ictions regard ing activities and inhibitors of otherN N S viru s L proteins. A s u sed herein,the term“RN A -d epend entRN A polymerase d omain”refers to thatportion of an L protein thatcatalyzes the protein’s RN A polymerase activity.In the 2109 amino acid V SV L protein, the d omain correspond s to amino acid resid u es 35-865 and encompasses arighthand ,“fingers-palm- thu mb”stru ctu re,withthe catalytic site on the palm in ad eepchannelbetween the fingers and thu mb su bd omains. A s u sed herein,the term“cappingd omain”refers to thatportion of the L protein thatcatalyzes viral mRN A cappingand encompasses the GD P polyribonu cleotid yltransferase (P RN Tase)catalytic site. In the 2109 amino acid V SV L protein,the d omain correspond s to amino acid resid u es 866-1334 and encompasses a GD P polyribonu cleotid yltransferase (P RN Tase) catalytic site of two conserved motifs— GxxT and H R, separated by ~ 70 resid u es, the largely α-helical, N -terminal half (resid u es 866–1100),whichabu ts the polymerase d omain and the C -terminalhalf (1100–1334). A s u sed herein,the term“methyltransferase d omain”refers to thatportion of the L protein responsible for methyltransferase activity. In the 2109 amino acid V SV L protein,the d omain correspond s to amino acid resid u es 1598 -1892,comprises the methyltransferase catalytic site,and contacts boththe connectorand cappingd omains. A s u sed herein,the term“C -terminald omain”refers to thatportion of the L protein occu rring C - terminalto the methyltransferase d omain. In the 2109 amino acid V SV L protein,the C -terminal d omain correspond s to amino acid resid u es 1893-2109. The C -terminal d omain is stru ctu rally an α- helical bu ndle, projecting a beak-like, β-hairpin su pported by a second inter-helical loop. It extend s back against the RdRp, au gmenting the β-hairpin that bears the catalytic A sp-A sn sequ ence at its tip, and terminates atthe three-way ju nction of the capping,connector,and methyltransferase d omains, where ithas one ormore contacts with each.The arm thu s contribu tes to closing the mu lti-d omain stru ctu re in the complex formed between the N N S viralL and P proteins. A s u sed herein,the term“connectord omain”refers to thatportion of the L protein occu rringbetween the capping and methyltransferase d omains. In the 2109 amino acid V SV L protein,the connector d omain correspond s to amino acid resid u es 1358 -1557 ,and is abu nd le of eighthelices. A s u sed herein,the term“high resolu tion 3D stru ctu re”refers to an atomic level protein or polypeptid e stru ctu re of atleast3.8 Å resolu tion,represented by asetof stru ctu re coord inates thatcan generate a3D representation of the protein stru ctu re when applied to appropriate software. In some embod iments,the highresolu tion 3D stru ctu re has aresolu tion of atleast3.8 Å ,e.g.,aresolu tion of at least3.7 Å ,atleast3.6Å ,atleast3.5Å ,atleast3.4Å ,atleast3.3Å ,atleast3.2Å ,atleast3.1Å ,atleast 3.0Å ,atleast2.9Å ,atleast2.8 Å ,atleast2.7 Å ,atleast2.6Å ,atleast2.5Å ,atleast2.4Å ,atleast2.3Å , atleast2.2Å ,atleast2.1Å ,atleast2.0Å ,oratleast1.9Å . A s u sed herein,the term“X Å atomic stru ctu re”refers to a three d imensionalstru ctu re or setof stru ctu re coord inates thatd escribes the 3D stru ctu re of aprotein ataresolu tion of atleastX Å . The term“X Å atomic stru ctu re of an N N S viru s L protein”refers to the stru ctu re orsetof coord inates that d escribes the stru ctu re of an entire N N S viru s L protein inclu d ingallof its d omains,catalytic regions and stru ctu ralelements atatleastX Å resolu tion.Thu s,a3.8 Å atomic stru ctu re of an N N S viru s L protein is athree d imensionalstru ctu re orsetof stru ctu re coord inates thatd escribes the 3D stru ctu re of an N N S L protein ataresolu tion of atleast3.8 Å .H ighresolu tion atomic stru ctu res su chas these can be obtained by method s inclu d ing bu tnotlimited to cryo-electron microscopy,single particle analysis and x-ray crystallography.A s an example,a3.8 Å atomic stru ctu re of V SV L protein,solved by cryo-electron microscopy,exhibits the resolu tion of atleast3.8 Å withaspace grou pP 1,so as to form a u nit cell of dimensions of a=112.0 Å , b=143.0 Å , c=106.0 Å and α, β, γ=90^o. A s u sed herein,the term“atomic mod el”refers to astru ctu ralmod elof aprotein atatomic resolu tion generated forexample,by tracingthe polypeptid e chain oramino acid sequ ence of the targetprotein onto the three d imensionalstru ctu re of targetprotein orthatof anotherrelated homologou s protein (e.g.,an N N S viru s L protein from anotherN N S viralspecies). A s u sed herein,the term“cryo-electron microscopy”or“cryo-EM”refers to amethod of imaging specimens in atransmission electron microscope u nd ercryogenic cond itions. C ryo-EM can permit high resolu tion stru ctu re d etermination of proteins for which crystalpreparation is d ifficu lt or impossible. A s u sed herein,the term“single-particle analysis”refers to a method of compu terized analysis of images from negative stainingand transmission electron microscopy by combiningseverald igitized images of similarparticles to bu ild u pathree-d imensionalreconstru ction of the particle. A s u sed herein,the term“virtu allibrary of 3D stru ctu res”refers to acollection of d atarepresenting the 3 d imensionalstru ctu res of potentialcand id ate molecu les,e.g.,smallorganic (or,forthatmatter, inorganic) molecu les,that,in concertwith appropriate software permitthe in silico d ocking of members of the library with an atomic stru ctu re mod elof,e.g.,a targetprotein. A non-limiting example of su ch a library or d atabase inclu d es the ZIN C d atabase; see, e.g., wiki.bkslab.org/ind ex.php/C ategory:ZIN C 15. A s u sed herein,the term“refinement”when u sed in relation to amolecu le orcompou nd refers to a mod ification of the chemicalstru ctu re agiven molecu le orcompou nd ,e.g.,amod ification of aknown orpred icted inhibitorof abiologicalfu nction.A“refinement”of agiven molecu le orcompou nd can, forexample,optimize the bind ingenergy of the molecu le comprisingthe refinementwithatargetsite and /orred u ce u nd esirable known orlikely bind ingto anotherfactor. A s u sed herein,the term“inhibitor”refers to amolecu le oragentthatattenu ates,inhibits,d ecreases, red u ces,orinterferes withthe stru ctu re,fu nction and /oractivity of an N N S viru s L protein in vitro,in situ,and /orin vivo.Inhibitors inclu d e bu tare notlimited to smallpeptid e orpeptid e-like molecu les, solu ble peptid es,and synthetic non-peptid ylorganic orinorganic compou nd s. A s u sed herein,the term“stru ctu re-based compu tationalmod eling”refers to the u se of the three- d imensionalstru ctu re or stru ctu re coord inates of a d ru g target,e.g.,a polypeptid e,inclu d ing,for example a large viralpolypeptid e,su ch as an N N S viru s L protein,in conju nction with compu ter mod eling to id entify and investigate cand id ate d ru gs and their bind ing geometries and affinities. Stru ctu re-based compu tationalmod eling u ses,forexample,3D stru ctu re coord inates obtained u sing method s inclu d ing bu t not limited to cryo-electron microscopy,X -ray crystallography or N M R spectroscopy. M ethod s and software packages thatpermitstru ctu re-based compu tationalmod eling are d escribed elsewhere herein and in the Examples. A s u sed herein,the term“homology mod eling”refers to constru ctingan atomic-resolu tion mod elof a " target" protein from its amino acid sequ ence and an experimentalthree-d imensionalstru ctu re of a related homologou s protein (the " template" ).The method relies on the id entification of one ormore known protein stru ctu res likely to resemble the stru ctu re of the qu ery sequ ence,and on the prod u ction of an alignmentthatmaps resid u es in the qu ery sequ ence to resid u es in the template sequ ence. M ethod s and software packages thatpermithomology mod elingare d escribed elsewhere herein and in the Examples. A s u sed herein,“RN A d epend entRN A polymerase (Rd Rp) activity”refers to the enzymatically catalyzed prod u ction of RN A from an RN A template. A ssays forRd Rpactivity of an N N S viru s L protein are d escribed herein and d emonstrated in the Examples. A s u sed herein,“GD P polyribonu cleotid yltransferase (P RN Tase)activity”refers to the transferof a nascentRN A onto aGD P acceptorto form aGpppN capstru ctu re,throu ghacovalentL protein–P protein RN A intermed iate.A ssays forP RN Tase activity are d escribed herein and d emonstrated in the Examples. A s u sed herein,“methyltransferase activity”refers to formation of the 7 mGpppN pmN pcapstru ctu re u pon methylation of the ribose O '2 and the gu anosine N 7. A ssays formethyltransferase activity are d escribed herein and d emonstrated in the Examples. The terms“attenu ate”,“d ecrease”,“red u ced”,“red u ction”,or“inhibit”and grammaticalvariations thereof are allu sed herein to mean a d ecrease by a statistically significant amou nt. In some embod iments,“attenu ate”“red u ce,”“red u ction" or“d ecrease" or“inhibit”typically mean ad ecrease by atleast10% as compared to areference level(e.g.,the absence of agiven treatment,cond ition or agent)and can inclu d e,forexample,ad ecrease by atleastabou t10%,atleastabou t20%,atleast abou t25%,atleastabou t30%,atleastabou t35%,atleastabou t40%,atleastabou t45%,atleast abou t50%,atleastabou t55%,atleastabou t60%,atleastabou t65%,atleastabou t70%,atleast abou t75%,atleastabou t80%,atleastabou t85%,atleastabou t90%,atleastabou t95%,atleast abou t98 %,atleastabou t99% ,or more.A s u sed herein,“red u ction”or“inhibition”d oes not encompass a complete inhibition or red u ction as compared to a reference level.“C omplete inhibition”is a100% inhibition as compared to areference level. The terms“increased " ,“increase" or“enhance" or“activate" and grammaticalvariations thereof are allu sed herein to generally mean an increase by astatically significantamou nt;forthe avoid ance of d ou bt,the terms“increased " ,“increase" or“enhance" or“activate" mean an increase of atleast10% as compared to areference level,forexample an increase of atleastabou t20%,oratleastabou t30%, oratleastabou t40%,oratleastabou t50%,oratleastabou t60%,oratleastabou t70%,oratleast abou t80%,oratleastabou t90% oru pto and inclu d inga100% increase orany increase between 10- 100% as compared to areference level,oratleastabou ta2-fold ,oratleastabou ta3-fold ,oratleast abou ta4-fold ,oratleastabou ta5-fold oratleastabou ta10-fold increase,atleastabou ta20-fold increase,atleastabou ta50-fold increase,atleastabou ta100-fold increase,atleastabou ta1000-fold increase ormore as compared to areference level. The term " location of resistance" refers to the stru ctu ralsite of the resid u es changed within amu tantL protein polypeptid e chain thatresu lts in the viru s being resistantto an inhibitor. The“location of resistance”can be mapped onto the 3D stru ctu re of the wild -type protein to elu cid ate the particu lar activity the mu tation changes and thereby permitthe d esign of an inhibitortargetingthatactivity. A s u sed in this specification and the append ed claims,the singu larforms " a," " an," and " the" inclu d e singu lar and plu ralreferences u nless the context clearly d ictates otherwise.Thu s for example, references to " the method " inclu d es one ormore method s,and /orsteps of the type d escribed herein and /orwhichwillbecome apparentto those persons skilled in the artu pon read ingthis d isclosu re and so forth. Forexample,the term " an N N S viru s L protein inhibitor" inclu d es reference to one ora plu rality (e.g.,two ormore)N N S viru s L protein inhibitor(s)and the term " N N S viru s L protein inhibitor " inclu d es reference to one or a plu rality (e.g.,two or more) of N N S viru s L protein inhibitor(s)and equ ivalents thereof known to those skilled in the art,and so forth.Itis fu rthernoted thatthe claims may be d rafted to exclu d e any optionalelement.A s su ch,this statementis intend ed to serve as anteced entbasis foru se of su ch exclu sive terminology as " solely," " only" and the like in connection withthe recitation of claim elements,oru se of a" negative" limitation. C ompu ter-read able storage med iaorcompu terread able med iacan be any available tangible med ia (e.g.,tangible storage med ia)thatcan be accessed by a compu ter,is typically of a non-transitory natu re,and can inclu d e bothvolatile and nonvolatile med ia,removable and non-removable med ia.B y way of example,and not limitation,compu ter-read able storage med ia can be implemented in connection with any method or technology for storage of information su ch as compu ter-read able instru ctions,program mod u les,stru ctu red d ata,or u nstru ctu red d ata.C ompu ter-read able storage med iacan inclu d e,bu tare notlimited to,RA M (rand om access memory),RO M (read only memory), EEP RO M (erasable programmable read only memory),flash memory orothermemory technology, C D -RO M (compactd isc read only memory),D V D (d igitalversatile d isk) or other opticald isk storage,magnetic cassettes,magnetic tape,magnetic d iskstorage orothermagnetic storage d evices, or other tangible and /or non-transitory med ia which can be u sed to store d esired information. C ompu ter-read able storage med iacan be accessed by one ormore localorremote compu tingd evices, e.g.,via access requ ests,qu eries or other d ata retrievalprotocols,for a variety of operations with respectto the information stored by the med iu m. Unless otherwise d efined herein,scientific and technicalterms u sed in connection with the present application shallhave the meanings thatare commonly u nd erstood by those of ord inary skillin the art.Fu rther,u nless otherwise requ ired by context,singu larterms shallinclu d e plu ralities and plu ral terms shallinclu d e the singu lar.In one respect,the presentinvention relates to the herein d escribed compositions,method s,and respective component(s)thereof,as essentialto the invention,yetopen to the inclu sion of u nspecified elements,essentialornot(" comprising). In some embod iments,other elements to be inclu d ed in the d escription of the composition,method or respective component thereof are limited to those thatd o notmaterially affectthe basic and novelcharacteristic(s)of the invention (" consisting essentially of" ). This applies equ ally to steps within ad escribed method as wellas compositions and components therein.In otherembod iments,the inventions,compositions, method s,and respective components thereof,d escribed herein are intend ed to be exclu sive of any elementnotd eemed an essentialelementto the component,composition ormethod (" consistingof" ).

B RIEF D ESC RIP TIO N O F TH E FIGURES FIGs. 1A-1E Electron C ryomicroscopic Reconstru ction of V SV -L atH igh Resolu tion.(FIG. 1A) Raw image of V SV -L particles in vitreou s ice record ed at1.8 mm d efocu s.Scale bar,10 nm.(FIG. 1B)P owerspectru m of the image shown in (A ),withplotof the rotationally averaged intensity versu s resolu tion.A rrow ind icates the spatialfrequ ency correspond ing to 3.8 Å resolu tion.(FIG. 1C) Representative class averages.Scale bar,10 nm.(FIG. 1D)Fou riershellcorrelation analysis:FSC , correlation between the half-setthree-d imensionalreconstru ctions (solid blu e line);C _ref,estimated correlation between the finalmapand aperfectreference mapcontainingno errors,calcu lated from FSC (Rosenthaland H end erson,2003);C C _work and C C _free,correlation between the finalmap and refined mod elforworkingand testsetof stru ctu re factors,respectively.(FIG. 1E)L eft:overview of V SV -L reconstru ction.In the view shown,the particle (241 kD a)is ~110 Å long and 80 Å wid e. Right:close-u pview of arepresentative region in the polymerase d omain (Rd Rp).The volu me shown in close-u p is from the protein interior,noton the Rd Rp su rface.D ensity is shown as gray mesh; polypeptid e-chain backbone of the refined mod el,as blackribbon;sid e-chain atoms (carbon,nitrogen, oxygen and su lfu r,as sticks.N ote the continu ou s backbone d ensity,a-helicalgrooves and resolu tion of bu lky sid e chains— featu res thatallowed bu ild ing and stereochemicalrefinementof the atomic mod el. FIGs. 2A-2E P reparation of Stru ctu re Factors for A tomic Stru ctu re Refinement.(FIG. 2A-2D) P reparation of Fou rier coefficients from the experimentalreconstru ction.(FIG. 2A) A mask is generated arou nd the mod el.(FIG. 2B)D ensity ou tsid e the maskis flattened ,and d ensity insid e the maskis pu ton absolu te scale.(FIG.2C)A mplitu d es and phases are calcu lated from the flattened map by Fou rier transformation (FFT).(FIG. 2D) Scaling of amplitu d es before refinement.(FIG. 2E) Estimation of figu res of merits from phase-angle d ifferences between the two half-setreconstru ctions. FIGs. 3A-3C Stru ctu re of V SV -L (FIG. 3A) D omain organization of V SV -L .The polymerase d omain (Rd Rp),cappingd omain (C ap),connectord omain (C D ),methyltransferase (M T)d omain,and C -terminald omain (C TD ) are ind icated .A mino-acid resid u e nu mbers ind icate fu nctionald omain bou nd aries.Flexible linkers 1 and 2 connectC ap to C D and C D to M T d omain,respectively. C onserved regions within L proteins of non-segmented negative-strand (N N S) RN A viru ses are labeled C R I–V I.A sterisks ind icate the position of active site resid u es.(FIG.3B)Ribbon d iagram of V SV -L polypeptid e chain;d omains ind icated as in (A ).(FIG. 3C)Su bstrate channels and internal cavities of V SV -L ,d epicted as white su rface enclosed by the stru ctu re in ribbon representation.In this orientation,the entrance to the template channellead ing to the active site faces d own;the channel ru ns between the Rd Rpand cappingd omains.N u cleotid es can access the Rd Rpactive site throu ghthe channelin the foregrou nd . FIG. 4 Second ary stru ctu re d iagram of V SV -L .Second ary stru ctu re elements along the V SV -L sequ ence are show as cylind ers and arrows for ^ helices and ^ strand s,respectively.D omains are ind icated as in FIG 3,withthe exception of the polymerase d omain (Rd Rp),whichis ind icated as in FIG 5,with the palm d omain,the fingers and the thu mb.D omain bou nd aries are ind icated by the correspond ingresid u e nu mbers. FIGs.5A-5C P olymerase Rd RpD omain (FIG. 5A)Stru ctu re of the Rd Rpd omain.Resid u es 35–865 are shown in ribbon representation in conventionalorientation (viewed from insid e the su rrou nd ing ‘‘cage’’and u psid e d own withrespectto the view in Figu re 2).The palm su bd omain,the fingers,and the thu mb are ind icated .The N -terminalregion is gray. (FIG. 5B)C lose-u pview of the active site. P alm,fingers and thu mb are ind icated as in (A ).The GD N active site motif is ata ^-hairpin in the palm d omain.A mod elforthe positions of the template RN A strand and two nu cleotid es is d erived from the reoviru s l3 initiation complex (P D B :1n1h) after su perposition on V SV -L Rd Rp.The priming loop (resid u es 1157–1173) intru d ing from the capping d omain (gray) positions the initial nu cleotid e of the transcript.(FIG. 5C)Similarity of the V SV -L Rd Rpd omain to those of otherviral polymerases.Stru ctu res of influ enzaviru s B polymerase (P D B :4wrt;P A resid u es 248–716 and P B 1 resid u es 1–616),reoviru s l3(P D B :1mu k;resid u es 2–890)and rotaviru s V P 1 (P D B :2r7 q;resid u es 2– 778 )are shown withthe same orientation and coloringscheme as the V SV -L Rd Rpin (A ). FIGs.6A-6D C appingD omain (FIG.6A)Stru ctu re of the cappingd omain.Resid u es 866–1334 are in ribbon representation.M otifs GxxT and H R are sites of gu anosine nu cleotid e bind ingand of covalent RN A attachment,respectively.Resid u es correspond ingto positions of inhibitor-resistance mu tations in hu man RSV polymerase (L iu zzietal.,2005)are shown as spheres.(FIG. 6B)C lose-u p of the active site.(FIG. 6C) C onfigu ration of the priming loop in V SV -L .O nly the Rd Rp and capping d omains are shown.The primingloop(resid u es 1157–1173)protru d es from the cappingd omain into the active site of the Rd Rpd omain.(FIG. 6D)P roposed d omain shifts to allow transcriptelongation and eventu altemplate release. FIG. 7 D ensity arou nd the P riming L oop,C ontou rs from a5 Å resolu tion d ensity mapou tline the flexible priming loop that projects from the capping d omain. H eavy black lines show the C α trace for the loopand ad jacentpolypeptid e chain. FIGs. 8A-8D D omain Reorganization (FIG. 8A)P rojection angle matching between class averages of negatively stained complexes of V SV -L and P protein (toprow)(Rahmehetal.,2010,2012)and projections calcu lated from the mod el(mid d le row).The bottom row shows the mod elin the same orientation with the ind ivid u ald omains colored as in FIG.3.N u mbers are correlation coefficients between mod eland negative-stain class averages.V SV -L :P (1–106) correspond s to the stru ctu re d etermined here.In the panelforV SV -L :P ,an arrow ind icates ad d itionald ensity observed in some class averages thatwas attribu ted to the bou nd P d imer.Scale bar,10 nm.(FIG. 8B) D ifference d ensity map(map_observed -map_{mod el})calcu lated to 5 Å resolu tion and shown togetherwiththe mod el. The map shows d ensity presentin the image reconstru ction thatcou ld notbe fitwith amolecu lar mod el.Strong d ensity— presu mably from linkers 1 and 2,which enter and leave this d ensity at d efined points,and withpotentialcontribu tion from P (tentative assignmentind icated by‘‘?’’)— lines the groove between the capping and connector d omains.P rojection angle matching of V SV -L fragments.ForV SV -L (1–1557 ),the negative-stain class averages su ggestaconformationally variable connection between the connector d omain (C D ) and the polymerase (Rd Rp) and capping d omain (C ap).The‘‘d ou ghnu t’’partof the image was selected and aligned resid u es 35–1334.(FIG. 8C, FIG. 8D)Fu ll-length V SV -L withou tP .C D ,M T,and C TD extend in variable orientation from the Rd Rp-C apd ou ghnu t. FIGs. 9A-9D C orrelation of Images of N egatively Stained V SV -L and Fragments with P rojected V iews of M od el-B ased D ensity,Foreach panel,an image from negative-stain electron microscopy (Rahmehetal.,2012)was correlated withprojections of ad ensity mapcalcu lated from the molecu lar mod elof V SV -L ,u singrou tines in SP ID ER (Shaikhetal.,2008 ).The images for(A )–(D )are those in the first,third ,fifthand sixthpanels,respectively,in the toprow of FIG.8 A .For(A )and (C ),which have a su bsid iary maximu m in the correlation plot,the projected views for both peaks have been shown.A ngu larcoord inates as d efined in SP ID ER. FIGs. 10A-10C M ethyltransferase D omain (FIG. 10A)Stru ctu re of the methyltransferase:resid u es 1598–1892 in ribbon representation.The consensu s fold of the S-ad enosylmethionine-d epend ent methyltransferase su bd omain is highlighted .The N -terminaland C -terminalregions are in gray.(FIG. 10B)C lose-u pof the active site.The SA M /SA H bind ing-site motif,GxGxG,is between b1 and aA . A n SA H molecu le’s location is d erived from asu perposition of its complex with d engu e viru s N S5 M T (P D B :1l9k).Resid u es thatparticipate in the methyltransferase activity are in stickd etail.(FIG. 10C) C omparison of V SV -L M T d omain with other viralA d oM et-d epend entmethyltransferases. Stru ctu res of d engu e viru s N S5 M T (P D B :1l9k;resid u es 7–267 )and vacciniaviru s V P 39 M T (P D B : 1av6;resid u es 3–297 )are shown in the same orientation and highlightscheme as in (A ). FIGs.11A-11B Second ary Stru ctu re D iagram of the M ethyltransferase D omain,(FIG.11A)D iagram of the consensu s fold forA d oM et-d epend entmethyltransferases; ^ helices, ^ strand s,and terminiare represented by circles,triangles and rectangles,respectively.The consensu s A d oM et-d epend ent methyltransferase fold is highlighted ,other regions in gray (as in FIG.10).The positions of SA M /SA H and the active site are ind icated . (FIG. 11B)D iagram of V SV -L M Tase,vacciniaviru s V P 39 M Tase,and flaviviru s N S5 M Tase,in the same highlight/colorscheme as (A ).Insertions or d eletions of α helices or β strand s are indicated as empty circles. FIG. 12 Sequ ence alignment.H omologou s L proteins inclu d e those of Rabies,Ebola,M easles and RSV .A lignmentof theirsequ ences shows the overallarrangementof variou s d omains,id entifies the active site resid u es of the protein forRd Rp,P RN Tase and methyltransferase activities and su ggests d omain bou nd aries forexpressingvariou s fragments of the proteins from V SV and related viru ses. FIGs.13A-13C P rod u ction and activity of M V polymerase complex.(FIG.13A)P u rification of M V polymerase complex.L and P were co-expressed in insectcells and pu rified as d escribed in M aterial and M ethod s and analyzed by SD S-P A GE.P osition of L ,P and co-pu rified H sp90 are ind icated . (FIG. 13B)D e novo RN A synthesis by M V L -P (lane 1)and M V L -P /H sp90 on anon-encapsid ated 19 nt long RN A template. Reactions were set u p in presence of [α32P ]-GTP , qu enched by the ad dition ED TA /formamid e and analyzed on a20% polyacrylamid e/7 M u reagel.D e novo initiation prod u ct sizes are ind icated on the left.(FIG. 13C)De novo RN A synthesis by M V L -P /H sp90 in presence of increase concentration of Geld anamycin. FIGs. 14A-14B Inhibition of M V L -P by A s136a.(FIG. 14A)Stru ctu re of the A s136ainhibitoris ind icated on top.Inhibitory effectof increasingA s136aconcentrations on RN A synthesis by M V L -P . Totalamou ntof prod u cts observed in (B ) were qu antified and the percentage of M V L -P RN A synthesis inhibition by the d ifferentA s136a concentrations were plotted .Error bars representthe stand ard d eviation from the mean of ind epend entexperiments.(FIG.14B)20% polyacrylamid e/7 M u reagelshowingthe inhibitory effectof increasingA s136aconcentrations on RN A synthesis by M V L -P . FIGs. 15A-15B Inhibition of M V L -P by ERD RP 0519.(FIG. 15A)Stru ctu re of the ERD RP 0519 inhibitor is ind icated on top.Inhibitory effectof increasing ERD RP 0519 concentrations on RN A synthesis by M V L -P .Totalamou ntof prod u cts observed in (B )were qu antified and the percentage of M V L -P RN A synthesis inhibition by the d ifferentinhibitorconcentrations were plotted .Errorbars represent the stand ard d eviation from the mean of ind epend ent experiments.(FIG. 15B) 20% polyacrylamid e/7 M u reagelshowing the inhibitory effectof increasing A s136aconcentrations on RN A synthesis by M V L -P . FIGs. 16A-16D Effectof ERD RP 0519 on M V L resistant-mu tants in complex with P .Effectof increasing ERD RP 0519 concentrations on RN A synthesis by M V (FIG. 16A) L _{T7 7 6A /T7 51I-P} ,(FIG. 16B)L _{T7 7 6A -P} and (FIG. 16C)L _{T7 51I-P} (C ).(FIG. 16D)Totalamou ntof prod u cts observed in (B )and (C ) were qu antified and the percentage of RN A synthesis inhibition by the d ifferent inhibitor concentrations were plotted .Errorbars representthe stand ard d eviation from the mean of ind epend ent experiments. FIGs. 17A-17C. Effect of inhibitor on recombinant measles containing su bstitu tions in L To d emonstrate thatthe two mu tations lead to complete resistance of the measles viru s to the non- nu cleosid e inhibitor,arecombinantmeasles viru s KS strain expressing eGFP was engineered .The kinetics of replication of the engineered variantM V -L _{T7 7 6A /T7 51I} were red u ced compared withthe wild type parentas evid enced by the d ecreased viralyield at24h postinocu lation of cells (FIG.17 A ) consistentwithafitness costto the resistance mu tations.The recombinantM V -L _{T7 7 6A /T7 51I}variantwas resistantto the non-nu cleosid e inhibitor on viralreplication as ju d ged by eGFP expression in the presence of 25 µM of the bioavailable inhibitor (FIG.17 B ,17 C ).This resu ltd emonstrates that su bstitu tions in measles L lead to complete resistance of the viru s to the non-nu cleosid e inhibitorin vitro and in cells bu tu ncoverafitness costto viralreplication associated withthose su bstitu tions.

FIG.18 M V L A mino acid sequ ence alignmentwithC D V and V SV .A mino acid sequ ence alignment of L from M V ,C D V and V SV has been realized u sing T-C offee and analyzed u sing Espript.The sequ ences are shown from M V L resid u e 266 to 1257.Resistantmu tations thathave been id entified forM V withA s136aand C D V withERD RP 0519 are noted withd ots and squ ares,respectively. FIGs.19A-19C L ocalization of resistantmu tation in L 3D stru ctu re.(FIG.19A)O verallstru ctu re of V SV L showingthe resistantmu tations obtained forM V and C D V in sid e chain d etail.(FIG. 19B) Rd Rpd omain of V SV L shown as the righthand orientation.The GD N resid u es responsible forRN A synthesis are shown in yellow and the resistantmu tants in sid e chain d etail.(FIG.19C)M od elof M V L ERD RP 0519 bind ing d omain bu ild u sing V SV L map(P D B )and coot.In yellow are shown the GD N resid u e,in sid e chain d etailthe resistantmu tants and in separate stickfigu re the ERD RP 0519 inhibitor. FIGs. 20A-20C P rod u ction and activity of RA B V polymerase complex.(FIG. 20A)P u rification of RA B V polymerase complex elements.L (left)and P (right)were ind ivid u ally expressed in insectcells and bacteria,respectively.B othproteins were pu rified as d escribed in M aterialand M ethod s analyzed by SD S-P A GE.(FIG.20B)D e novo RN A synthesis by RA B V L on aRA B V L e19 RN A template in absence or presence of either RA B V P or V SV P . Reactions were set u p in presence of [α32P ]-GTP , qu enched by the ad d ition ED TA /formamid e and analyzed on a20% polyacrylamid e/7 M u reagel.D e novo initiation prod u ctsizes are ind icated on the left.(FIG.20C).D e novo RN A synthesis by V SV L on aV SV L e19 RN A template in absence orpresence of eitherV SV P orRA B V P . FIG. 21 Id entification of the minimalP L -enhancing d omain.A llP proteins were ind ivid u ally expressed in bacteriaand pu rified as d escribed in Examples.(FIG. 21)D e novo RN A synthesis by RA B V L on aRA B V L e19 RN A template in absence orpresence of eitherRA B V P orV SV P ,P _1-91, P _21-91,P _11-91,P _1-40,P _1-50 orP _11-50. FIG. 22A-22C Template specificity forRN A synthesis by RA B V L -P .(FIG. 22A)D e novo RN A synthesis by RA B V L -P on RA B V L e19 and V SV L e19 RN A templates.Sequ ence of bothRN A s are ind icated on top,stars ind icating non-conserved nu cleotid e between the two templates.(FIG. 22B) C omparison of totalRN A synthesis by RA B V L -P on RA B V L e19 and V SV L e19.TotalRN A synthesis was qu antified ,normalized to levels of RN A synthesis on RA B V L e19 and graphed .Error bars representthe stand ard d eviation from the mean of three ind epend entexperiments.(FIG. 22C) Ind ivid u alreaction prod u cts of d ifferentsizes synthesized in (FIG. 22A)forRA B V L e19 (top)and V SV L e19 (bottom)were qu antified and normalized to the size of the 11 mer.Errorbars representthe stand ard d eviation from the mean of three ind epend entexperiments. FIG.23 Template specificity forRN A synthesis by V SV L -P .(FIG.23A)D e novo RN A synthesis by V SV L -P on V SV L e19 and RA B V L e19 RN A templates.(FIG. 23B)C omparison of totalRN A synthesis by V SV L -P on V SV L e19 and RA B V L e19.TotalRN A synthesis was qu antified , normalized to levels of RN A synthesis on V SV L e19 and graphed .Errorbars representthe stand ard d eviation from the mean of three ind epend entexperiments.(FIG. 23C)Ind ivid u alreaction prod u cts of d ifferentsizes synthesized in (FIG. 23A)for RA B V L e19 (top)and V SV L e19 (bottom)were qu antified and normalized to the size of the 11 mer.Errorbars representthe stand ard d eviation from the mean of three ind epend entexperiments. FIG. 24A-24B A ctivity of RA B V L on V SV L N -RN A .(FIG. 24A) M od elshowing d omain stru ctu re of P highlightingthe C TD whichwas exchanged .(FIG. 24B)A ctivities of L -P complexes on V SV N -RN A templates. FIG. 25 M od elfor Transcription of an RN P C omplex by V SV -L L eft:initiation complex,with d omains organized as in the stru ctu re d escribed here.A rrow shows d irection of capping d omain d isplacementrequ ired forthe transition to an elongation complex.Right:elongation requ ires both d isplacementof the capping d omain (with likely accompanying reorganization of the C D ,M T,and C TD ) and d isplacementof two to three N su bu nits from the template resid u es looped into the polymerase.The N su bu nits are shown linked as acontinu ou s chain,as su ggested by theirstru ctu re (Green etal.,2006). FIG. 26 C orrelation of M od eland D ensity,Related to ExperimentalP roced u res.Resid u e-by-resid u e correlation,with each panelcorrespond ing to a d omain.Resid u e nu mbers above the plots for the capping and connectord omains ind icate segments of poord ensity.The gapin the methyltrasferase plotcorrespond s to ashortloopomitted from the mod el.C alcu lated withC N S. FIG. 27 Start-stopmod elof sequ entialtranscription.A schematic of the V SV genome is d epicted 3'- 5'showingthe lead er(le)and trailer(tr)regions thatcontain essentialcis-actingsignals necessary to promote RN A synthesis.A teach gene ju nction ahighly conserved sequ ence 3'-A UA C UUUUUUU GA UUGUC N N UA G-5'provid es criticalsignals for the termination and polyad enylation of the u pstream mRN A and initiation and capping of the d ownstream mRN A .The viralN =nu cleocapsid , P =phospho,M =matrix,G=glyco-and L =large protein genes are shown alongwiththe mRN A thatare sequ entially transcribed from them. FIG. 28A-28D C ryoEM reconstru ction of V SV L stabilized by afragmentof P .(FIG.28 A )Su rface view of d ensity map.O valind icates approximate region of the d etailed map in (FIG.28 C ).The polymerase d omain is the large,rou ghly sphericallobe in the lowerhalf of the figu re;the cappingand methyltransferase d omains are atthe top.(FIG.28 B )Su rface view as in (FIG.28 A ),bu tsu perposed on rotaviru s V P 1 ribbon d iagram (pale magenta).The su perposition illu strates the similarity of the polymerase d omain and V P 1 (1086 resid u es).Ind ivid u alsecond ary stru ctu ralelements su chas alpha helices may have d ifferentlengths and somewhatd ifferentlocalpackings.The overalltopology of the chains are the same over mostof the polymerase d omain.The colored ovals atthe top are rou gh ou tlines of the three smallerd omains packed againstone su rface of the polymerase d omain.D ashed ovalshows the ou tline of ad omain to the rearof the othertwo.The channelthrou ghwhichnascent transcriptemerges (by analogy withV P 1)lead s toward this d omain.(FIG.28 C )D etailof d ensity for partof an a-helix,as ou tlined in (FIG.28 A ).(FIG.28 D )D etailof d ensity forthe polymerase catalytic site. FIG. 29A-29B (FIG. 29A) Isolation of an N -P -L complex.SD S-P A GE gelof apu rified complex comprisingV SV L togetherwithafragmentof P ,and an N -terminaltru ncation of N is shown.The N - protein was tru ncated by the d eletion of the N -terminalarm whichfavors the formation of the RN A - template associated form of N .The viralN and P proteins were co-expressed in E.coliand L from a recombinant bacu loviru s vector in insect cells. The complex pu rified by nickel affinity chromatography and size exclu sion.(FIG. 29B) Enzymatic properties of the N -P -L complex.RN A synthesis activity of L (1)L -P 1-106 (2)and L -P 1-106-N 22-423(3)on anaked 19 merRN A . FIG.30A-30C Stru ctu res of N in complex withP .(FIG.30A )Stru ctu re of three molecu les of N -RN A in complex withthe P c-rD P D B = 3H H Z (FIG 30B )Stru ctu re of N in complex withthe N -terminal d omain of P .P bind s in between the N -terminaland the C -terminalof N ,P D B = 3P M K^t451.(FIG. 30C )L ocation of the R146 resid u e thatis ahistid ine in the poIR1 mu tant.Shown on the stru ctu re of the N -RN A ,P D B =2GIC . FIGS. 31A-31B The P oligomerization d omain is d ispensable.(FIG.31A )Reconstitu tion of RN A synthesis on the N -RN A template.(FIG 31B ) P laqu e morphology of wild type recombinantviru s (P wt)orthatlackingthe P oligomerization d omain (P min0D ). FIGs. 32A-32C Inhibition by M of RN A synthesis on anaked RN A template by L and L -P .(FIG 32A )Stru ctu re of the C -terminald omain of V SV M ,note the N -terminald omain is notresolyed ⁽²⁴⁾. (FIG 32B )Inhibitory effectof increasingM concentrations on RN A synthesis by V SV L -P .Reactions were performed in presence of [³²P ] -GTP and analyzed on a20% polyacrylamid e/7 M u reagel.(FIG 32C )Inhibitory effectof increasing M concentrations on RN A synthesis by V SV L and L -P .Total amou ntof prod u cts was qu antified and the percentage of V SV L RN A synthesis inhibition by the d ifferentM concentrations was plotted .Errorbars representthe stand ard d eviation from the mean of ind epend entexperiments. FIG.33 N on nu cleosid e inhibitors of M easles viru s RN A -d epend entRN A polymerase, FIG.34 Effectof ERD RP 0519 on measles viru s L /P complex.The inhibitorinhibits withan IC 50 of 100nM FIG. 35 C ompou nd s A S-136aand ERD RP -00519 inhibitmeasles viru s Rd Rp activity.V alu es are expressed relative to vehicle-treated samples and representaverage of three experiments. FIG.36 Effectof inhibitors on M V L resistant-mu tants in complex withP . FIG. 37 M easles viru s L /P inhibitor mu tantsequ ence mod eled onto the V SV polymerase.The mu tation site id entified close to the active site. FIG.38 L ocalization of resistantmu tation in L 3D stru ctu re. FIG.39 RN A synthesis by M V L -P complex. FIG.40 Ribbon d iagram of protein stru ctu re.

D ETA IL ED D ESC RIP TIO N O F TH E IN V EN TIO N A s d iscu ssed in the Su mmary above,the technology d escribed herein relates to the id entification of high resolu tion 3D stru ctu re for the L proteins of N N S viru ses and the u se of su ch stru ctu res to id entify agents thatinhibitone ormore activities of the L protein.The highresolu tion 3D stru ctu res of N N S viru s L proteins have been elu sive,d u e in large partto the large size of the protein and the d ynamic natu re of the variou s protein d omains.The size and flexibility of the protein has preclu d ed , forexample,the u se of the stand ard X -ray crystallography approach,as these properties make the preparation of the necessary protein crystals d ifficu lt,if notimpossible to achieve. D espite these issu es,d escribed herein is the solu tion of an N N S viru s L protein athighresolu tion.The solu tion of the V SV L protein’s stru ctu re atthis resolu tion notonly d emonstrates the ability to apply the same or very similarmethod s to the solu tion of otherN N S viru s L proteins,bu t,u singthe V SV L protein high resolu tion 3D stru ctu re and homology mod eling based u pon the amino acid sequ ences of other L proteins,one can prepare u sefu l3D stru ctu res ormod els forothermembers of the family. Fu rther, the highresolu tion 3D stru ctu re of the V SV L protein permits the mappingof mu tations thatresu ltin resistance to known L protein inhibitors onto the highresolu tion stru ctu re to permitthe rationald esign of inhibitors withimproved activity oreven,forexample,broad eractivity againstmore than one N N S viru s. The following d escribes the variou s method s and consid erations necessary to u se the technology d isclosed herein. NNS viruses The non-segmented negative-strand (N N S)RN A viru ses inclu d e some of the mostlethalhu man and animalpathogens,inclu d ingEbolaviru s and rabies viru s.Theirmu ltifu nctional,large (L )polymerase proteins,carried within the virions (B altimore et al.,1970),have biochemical properties that d istingu ish them from mostotherRN A polymerases of viru ses orof theirhosts.In ad d ition to their Rd Rp activity (Emerson and W agner,1973),N N S RN A viru s L proteins catalyze an u nu su al sequ ence of mRN A capping reactions (H ercyk etal.,1988 ),and the Rd Rpitself polyad enylates the viralmessage (H u ntetal.,1984).A nu cleocapsid (N )protein sheathcoats the genomic RN A ,and the viralpolymerase u ses this N -RN A complex as template,ratherthan u ncoated RN A . Und erstand ing of RN A synthesis in N N S RN A viru ses comes principally from stu d ies of vesicu lar stomatitis viru s (V SV )— an enveloped ,bu llet-shaped ,rhabd oviru s,closely related to rabies viru s. V SV cau ses an acu te d isease of livestock.The V SV L protein d oes notbind the N -RN A template d irectly bu trequ ires a cofactor,the viralphosphoprotein (P ),as a brid ge (Green and L u o,2009). D elivery of the N -RN A -L -P complex into the cytoplasm when the viru s enters a cellinitiates infection.Transcription starts atthe 3'end of the genome and prod u ces atriphosphate 47 -nu cleotid e lead er RN A ,followed by sequ ential transcription of five capped and polyad enylated mRN A s (A braham and B anerjee,1976;B alland W hite,1976;W helan and W ertz,2002).A teach gene ju nction,L terminates synthesis of the u pstream gene,ad d ingapolyad enylate (polyA )tailby iterative transcription of aU7 tract(B arrand W ertz,2001;B arretal.,1997 ;Stillman and W hitt,1997 ).Itthen transcribes the d ownstream gene.A pproximately 30% attenu ation occu rs ateach su ccessive gene, withd issociation of the template (Iverson and Rose,1981).Replication also starts atthe 3'end of the genome,bu tencapsid ation by newly synthesized N accompanies synthesis of the nascentRN A strand , and in this mod e,L ignores allthe cis-actingsignals thatd ictate sequ entialtranscription of mRN A s (L aFerlaand P elu so,1989;P atton etal.,1984;P elu so and M oyer,1983). The variou s enzymatic activities of L are tightly linked .A GD P polyribonu cleotid yltransferase (P RN Tase)ad d s the capstru ctu re (O gino and B anerjee,2007 )when the nascentRN A chain length has reached 31 nu cleotid es,as shown by artificially stalling transcription atvariou s chain lengths (Tekes etal.,2011).The u nconventionalmechanism of cap ad d ition proceed s throu gh a covalent ad d u ctbetween a histid ine resid u e on L (H 1227 )and the monophosphate nascentRN A ,which is transferred onto aGTP -d erived GD P acceptor(L ietal.,2008 ;O gino and B anerjee,2007 ).Failu re to cap resu lts in the prematu re termination of transcription,which links cap ad d ition to Rd Rp processivity (L ietal.,2008 ;Stillman and W hitt,1999;W angetal.,2007 ).Su bsequ entmethylation is firstatthe 2'O position on the ribose of the firstnu cleotid e and then on the N 7 of the capping gu anylate— opposite to the typicalord erof mod ification (Rahmehetal.,2009).Failu re to methylate, eitherby exogenou sly manipu latingthe concentration of S-ad enosylhomocysteine (SA H )to compete with the S-ad enosyl methionine methyld onor or by mu tating criticalcatalytic resid u es in the methylase d omain,can resu ltin hyper-polyad enylation of mRN A transcripts,linkingthe methylation activity to the Rd Rp(Galloway and W ertz,2008 ;L ietal.,2009;Rose etal.,1977 ).B othcapad d ition and su bsequ entmethylation requ ire specific sequ ence elements within the first10 ntof the mRN A (W ang etal.,2007 ).The u ncapped lead erlacks those signals (L ietal.,2008 ;O gino and B anerjee, 2007 ). D u ring RN A synthesis,a few su bu nits of N d issociate from the template RN A for access to the catalytic site of Rd Rpand then reassociate as the process continu es (A lbertinietal.,2006;Green et al.,2006).P can help coord inate these events (Green and L u o,2009).There are nine nu cleotid es associated with each N su bu nitin the ribonu cleoprotein (RN P ).Fu llpolymerase processivity and correctrecognition of the cis-actingsignals in the viralgenome bothrequ ire N (M orin etal.,2012). Images from negative-stain electron microscopy (EM )of pu rified V SV -L show aring-like‘‘core’’ d ecorated by asetof three variably oriented globu larappend ages (Rahmehetal.,2010)Tru ncations of L map the Rd Rp to the ring-like d omain.Formation of a complex with P cau ses the globu lar append ages to reorganize into acompacttail,wrapped onto one sid e of the ring(Rahmehetal.,2010). C omplex formation withP enhances Rd Rpinitiation and processivity,bu tP itself has no enzymatic activity.A fragmentof P comprising resid u es 35–106 is su fficientto ind u ce the conformational rearrangement(Rahmeh etal.,2012).Segments in the center and atthe C -terminalend of the P polypeptid e chain med iate d imerization and N -RN A bind ing,respectively (D ing etal.,2006;Green and L u o,2009).Images from negative-stain EM of pu rified L -P complexes are amixtu re of single and d imeric L species,in whichthe two L molecu les have variable relative orientation (Rahmehetal., 2010).H oweverthese stu d ies d o notd efine whatform of complex prevails d u ringRN A sysnthesis. H erein,the stru ctu re of acomplex of L with P (35–106)by electron cryomicroscopy (cryo-EM )is d etermined .Into ad ensity mapat3.8 Å resolu tion,acomplete mod elof the 2109-resid u e polypeptid e chain is bu ilt.The resu ltshows thateven forafu lly asymmetric stru ctu re withmolecu larmass <250 kD a,the featu res in ad ensity mapfrom single-particle cryo-EM can be wellenou gh d efined fora complete d e novo chain trace.Five d omains have been d istingu ished :three— an RN A -d epend ent RN A polymerase (Rd Rp),an mRN A capping d omain,and a methyltransferase d omain— with assigned enzymatic activity and two— a connector between the capping and methyltransferase d omains and aC -terminald omain— thatappearto have largely organizationalroles.Itis believed that the conformation of the protein in the complex examined in the examples d escribed herein correspond s to an initiation state,read y to acceptthe 3'end of atemplate.Elongation beyond one or two nu cleotid es willrequ ire large-scale d omain reorganization.A s su ch,relative positions of the d omains and the ability to block movements then provid es another levelof activity to targetfor potentialantivirals. The N N S viru ses can be d ivid ed into two main grou ps based on theirgenomic RN A :non-segmented N S (N N S)RN A viru ses and segmented N S (SN S)RN A viru ses.The N N S RN A viru ses comprise fou rfamilies,the Rhabd ovirid ae (vesicu lastomatitis viru s (V SV )and rabies viru s),P aramyxovirid ae (measles and respiratory syncytialviru ses (RSV )),Filovirid ae (Ebolaand M arbu rg viru ses)and the B ornavirid ae (B ornad isease viru s).The SN S RN A viru ses comprise three families,the A renavirid ae (lymphocytic choriomeningitis viru s and M achu po viru s (M A C V )),B u nyavirid ae (RiftV alley fever viru s)and O rthomyxovirid ae (influ enzaA viru s). The N S RN A viru ses share common replicative machinery comprisingaprotein–RN A complex.Itis composed of the genomic RN A whichis wrapped overits entire lengthby the nu cleotid e protein (N ) to form the ribonu clé ocapsid e (N -RN A ),and the polymerase complex.O ne of the featu res of M ononegavirales is thatthis polymerase complex consists of two main elements:the protein L and its co-factor,the phosphoprotein (P )which allows interaction between L and N -RN A complex.Those N P -RN A templates are copied by the virally encod ed RN A d epend entRN A polymerase (Rd RP )in two synthetic mod es:mRN A transcription and genome replication.Forthe N N S RN A viru ses that single genome contains a tand em array of 5–10 genes thatare sequ entially copied by the viral polymerase,whereas the SN S RN A viru s polymerases copy eachtemplate into asingle mRN A . The enzymatic activities necessary forcopyingof the N P -RN A templates inclu d e an RN A -d epend ent RN A polymerase (Rd Rp),as wellas the enzymes requ ired for mRN A cap formation namely a GD P :polyribonu cleotid yltransferase (P RN Tase)and ad u alspecificity mRN A capmethyltransferase that are only u tilized d u ring mRN A transcription.A llof the necessary enzymatic activities resid e within a250 kD alarge (L )polymerase protein.B ioinformatics analyses have id entified six conserved d omains (C R Ito C R V I)in N N SV L proteins thatare connected by variable linkerregions (Figu re ; P ochetal.,1988 ,1990;Svend aetal.,1997 ).H owever,the precise roles foreachof these L d omains in Rd Rpfu nction are stilllargely u nclear.C R Ihas been implicated in L oligomerization (C eviketal., 2003,2004;Smallwood and M oyer,2004)and L–P interactions (H orikamietal.,1994;H olmes and M oyer,2002;C evik etal.,2003,2004;C hattopad hyay and Shaila,2004),C R III is involved in phosphod iesterbond formation forRN A polymerization (M alu retal.,2002b),and C R V Icontains methyltransferase activity (P och etal.,1990;Ferron etal.,2002).A conserved GX X TnH R motif in C R V of V SV L is thou ght to med iate u nu su al capping of the viral mRN A s throu gh transfer of 5′ - monophosphate-mRN A onto GD P . L protein as a target A n inherentchallenge of allpathogen-d irected d ru g d iscovery campaigns is a narrow ind ication spectru m of the therapeu tic cand id ate,limitinginhibitory activity to aspecific memberor,atbest,a single genu s within the viru s family.The L protein represents a rich targetfor d ru g d iscovery campaigns,d u e to its mu ltid omain organization and the concentration of severalessentialenzymatic activities in a single protein.The preced ence established by the d evelopmentof inhibitors of,for instance,H IV reverse transcriptase and H epatitis C viru s polymerase u nd erscores the valu e of high- resolu tion stru ctu ralinformation forthe id entification and optimization of hitstru ctu res,the molecu lar u nd erstand ing of the mechanism of inhibitory activity,and ,potentially,the proactive d esign of analogs withincreased resilience againstviralescape from inhibition (N ijhu is etal.,2009;A d ams et al.,2010;D as etal.,2011;H alfon and L ocarnini,2011;C anhou b,2012;L loyd etal.,2014).H owever, the d ru gd evelopmentfield is hampered by the cu rrentlackof high-resolu tion stru ctu ralinformation for any mononegavirales L protein.The high resolu tion N N S viru s L protein stru ctu re d escribed herein is amajorstepforward in this regard . A n envisioned d ru g application inclu d ing post-exposu re prophylactic u se affects the d ru g profile requ ested of ad esirable anti-N N S therapeu tic;asu ccessfu lcand id ate mu stbe safe and efficaciou s, amenable to cost-effective manu factu re,id eally be shelf-stable atambienttemperatu re,and mu stbe orally bioavailable.O f small-molecu le chemicalcompou nd s,large molecu le biologics,and peptid ic biopharmaceu tical as cand id ate d ru g classes,small-molecu les are most su itable to fu lfill these d ivergent d emand s (Ganellin et al.,2013).Two main classes of polymerase-targeted d ru gs are cu rrently in clinicalu se for,among others,antiretroviraltherapy,hu man cytomegaloviru s therapy, and H C V therapy,competitive nu cleotid e/nu cleosid e su bstrate analogs and non-nu cleosid e allosteric inhibitors (Su n etal.,2007 ,2008 ;A nd reietal.,2008 ;B rown,2009;Kru mm etal.,2011;M ercorelliet al.,2011). Assays for L protein RdRp assay The stru ctu ralinsights obtained from N S RN A viru s templates ind icate thatthe RN A mu stbe d isplaced from N P to facilitate its copying by polymerase.Those observations catalyzed the d evelopmentof an improved system to stu d y polymerase biology in vitro in whichpolymerase alone cou ld u se naked RN A as template.These in vitro mod els are powerfu ltools to valid ate and stu d y how selected inhibitors of L protein and /orN N S viru s polymerase fu nction.In agreementwithamod elof N d issociation,ithas been d emonstrated viaan in vitro system thatN N S L protein,in the absence of N P and P ,initiates RN A synthesis on anaked RN A correspond ingto the first19 nu cleotid es of the genomic RN A .The in vitro initiation assay u ses essentially only the enzyme and su bstrate components:L protein and naked RN A .RN A correspond ingto the first19 ntof the genome (L e19) can be chemically synthesized and u sed as template forRN A synthesis in vitro.Reaction prod u cts can be visu alized following30 min of incu bation,seperation on apolyacrylamid e/u reageland analyzed u sing au torad iography.The in vitro assay has been shown to be ad apted to stu d y the activity of L protein of otherviru ses of the N N S family,e.g.measles viru s L protein and rabies viru s L protein. D etailforthese assays are d escribed in the examples herein. In vitro capping assay to testfor capped prod u cts,in vitro transcription reactions are carried ou t in the presence of [α- 32P ] GTP ;RN A is pu rified and d igested with TA P ,and the prod u cts resolved by thin layer chromatography.P rod u ctthatcomigrated witha7 mGpmarker,confirms thatthe RN A is capped and methylated . In vitro methylation assay triphosphorylated RN A transcripts are capped in the presence of α-32P GTP (3,000 C i/mmol; P erkin- Elmer) and vaccinia viru s capping enzyme (Epicentre) in 15-μl reaction mixtu res. A fter 2 h at 37 °C , the reaction mixtu res are treated with 10 U of calf intestinalalkaline phosphatase (N EB )for 1 h, followed by pu rification. C apped RN A is qu antified by measu ring the incorporation of [α-32P ]GTP and extrapolation of molarqu antities from astand ard cu rve of moles versu s cou nts perminu te.For methylation at the G-N -7 position, 100 μM S-ad enosyl methionine is inclu d ed in the capping reaction mixtu re. For methylation at the ribose 2′ -O position or at both the G-N -7 and ribose 2′ -O positions, the capped RN A or the G-N -7 -methylated RN A is incu bated with 100 μM SA M and 7.5 U of vaccinia viru s 2′ -O M Tase (Epicentre). The prod u cts of this reaction are pu rified, digested with nu clease P 1, and resolved by TL C on P EI cellu lose F sheets. N u clease P 1 d igestion cleaves the 3′ -5′ phosphodiester bond s in single-strand ed RN A bu t d oes not cleave the 5′ -5′ bond of the cap stru ctu re, resu lting in the release of GpppA from u nmethylated RN A .Following incu bation of the RN A with pu rified L protein,the prod u cts of nu clease P 1 d igestion comigrated with a m7 GpppA m marker, d emonstrating that L possesses both G-N -7 and 2′ -O M Tase activities. Rational drug design In anotherembod iment,the technology d evised herein provid es amethod of u singan atomic stru ctu re of this invention in an inhibitorscreeningassay,the method comprisingselectingapotentialinhibitor by performingrationald ru gd esign withthe three-d imensionalstru ctu re d etermined forthe L protein. The selecting is performed u sing with compu ter mod eling.The rationally d esigned or improved cand id ate inhibitoris then contacted with an L protein and the ability of the potentialinhibitorto inhibiting L protein activity in vitro,infection or replication of an N N S viru s among others in measu red . N u merou s compu terprograms are available and su itable forrationald ru gd esign and the processes of compu ter mod eling,mod el bu ild ing,and compu tationally id entifying,selecting and evalu ating potential inhibitors of L protein.These inclu d e,for example, GRID (available form O xford University, UK), M C SS (available from M olecu lar Simu lations Inc., B u rlington, M ass.), A UTO D O C K (available from O xford M olecu larGrou p),FL EX X (available from Tripos,St.L ou is. M o.),D O C K (available from University of C alifornia,San Francisco),C A V EA T (available from University of C alifornia,B erkeley),H O O K (available from M olecu larSimu lations Inc.,B u rlington, M ass.),and 3D d atabase systems su ch as M A C C S-3D (available from M D L Information Systems, San L eand ro,C alif.),and UN ITY (available from Tripos,St.L ou is.M o.).P otentialagents can also be compu tationally d esigned“d e novo”u singsu chsoftW are packages as L UD I(available from B iosym Technologies,San D iego,C alif.),L EGEN D (available from M olecu larSimu lations Inc.,B u rlington, M ass.),and L EA P FRO G (Tripos A ssociates,St.L ou is,M o.).C ompou nd d eformation energy and electrostatic repu lsion, can be evalu ated u sing programs su ch as GA USSIA N 92, A M B ER, Q UA N TA /C H A RM M ,A N D IN SIGH T II/D ISC O V ER.These compu ter evalu ation and mod eling techniqu es can be performed on any su itable hard ware inclu d ingforexample,workstations available from Silicon Graphics,Su n M icrosystems,and the like.A ny su ch u se,as pertains to the present invention is to be consid ered an embod imentthereof.These techniqu es,method s,hard ware and software packages are representative and are notintend ed to be acomprehensive listing,and any su ch u se represents an embod imentof the invention.O thermod elingtechniqu es known in the artcan also be employed in accord ance withthis invention.See forexample,N .C .C ohen,M olecu larM od elingin D ru gD esign,A cad emic P ress (1996)(and references therein),and software id entified atinternetsites inclu d ing the C M B IC heminformatics Su ite atcheminf.cmbi.ru .nl/neW s/chemneW s_his.shtml(Feb. 27 , 2006), and the N IH M olecu lar M od eling H ome P age at W orld W id e W eb at ?.mu ni.cz/u sr/mejzlik/mirrors/molbio.info.nih.gov/mod eling/softW are list/(Feb.27 ,2006). The agent can be selected by performing rationald ru g d esign with the three-d imensionalhigh resolu tion stru ctu re (or stru ctu res) d etermined as herein,especially in conju nction with compu ter mod eling and method s d escribed above.The agentis then obtained from commercialsou rces oris synthesized from read ily available starting materials u sing stand ard synthetic techniqu es and method ologies known to those of ord inary skillin the art.The agent is then assayed ,in one embod iment,to d etermine its ability to inhibitstru ctu re,fu nction and /oractivity of an N N S viru s L protein orad omain thereof. The potentialinhibitory activity of achemicalcompou nd on an N N S L protein can be analyzed prior to its actu alsynthesis and testingby the u se of compu termod elingtechniqu es,as is known to those of ord inary skillin the art.O ne of ord inary skillin the artcan u se,in other embod iments of this invention,any one of severalmethod s to screen chemicalentities or fragments for their ability to associate with L protein or a d omain thereof.Selected fragments or chemicalentities can then be mod eled in a variety of orientations,or d ocked ,onto a specific targetd omain.D ocking can be accomplished u sing software su ch as Q u anta and Sybyl,followed by energy minimization and molecu lard ynamics withstand ard molecu larmechanics forcefield s,su chas C H A RM M and A M B ER. Specialized compu ter programs can also assistin the process of selecting fragments or chemical entities.These inclu d e,in one embod iment,the programs GRID ,M C SS,A UTO D O C K and D O C K. In anotherembod iment,the agentcan be d esigned as awhole or“d e novo”d esign u sing an empty bind ingsite.These method s can inclu d e the u se of programs su chas L UD I,L EGEN D and L eapFrog, eachof whichrepresents an embod imentof the technology d escribed herein. Embod iments of variou s aspects d escribed herein can be d efined in any of the following nu mbered paragraphs: 1.A method of id entifyingan inhibitorof anon-segmented negative strand (N N S)viru s,the method comprising:(a)u singahighresolu tion 3D stru ctu re of an N N S viru s L protein to selectacand id ate inhibitorfrom avirtu allibrary of 3D stru ctu res of potentialcand id ates by performingstru ctu re based compu tationalmod eling;(b) obtaining the cand id ate inhibitor;and (c) measu ring an activity of L protein when contacted with the cand id ate inhibitorobtained in step(b),wherein ad ecrease in the activity confirms thatthe cand id ate is an inhibitorof the N N S viru s. 2.A method of id entifyingan inhibitorof anon-segmented negative strand (N N S)viru s polymerase, the method comprising:(a)u singahighresolu tion 3D stru ctu re of an N N S viru s L protein to selecta cand id ate inhibitor from a virtu allibrary of 3D stru ctu res of potentialcand id ates by performing stru ctu re based compu tationalmod eling;(b)obtaining the cand id ate inhibitor;and (c)measu ring L protein polymerase activity when contacted withthe cand id ate inhibitorobtained in step(b),wherein ad ecrease in the activity of the polymerase confirms thatthe cand id ate is an inhibitorof the N N S viru s polymerase. 3.A method of id entifying an inhibitor of an N N S viru s the method comprising:(a) screening a library of cand id ate molecu les forinhibition of an N N S L protein in an in vitro biochemicalassay for an L protein activity;(b)u singahighresolu tion 3D stru ctu re of the L protein to d etermine the bind ing site of potentialcand id ates obtained in step (a) by compu ter mod eling;(c) id entifying a potential refinementto the cand id ate inhibitor;(d ) obtaining a cand id ate inhibitor molecu le comprising the refinementid entified in step(c);and (e)measu ringthe activity of the L protein when contacted with the cand id ate inhibitor obtained in step (d ),wherein a d ecrease in the activity of the L protein confirms thatthe cand id ate is an inhibitorof the N N S viru s. 4.A method of id entifying an inhibitorof L protein of an N N S viru s,the method comprising:(a) u singahighresolu tion 3D stru ctu re of the L protein of vesicu larstomatitis viru s (V SV )to create an atomic mod elof the L protein of asecond N N S viru s by homology mod elingthe amino acid sequ ence of the L protein of the second N N S viru s onto the highresolu tion 3D stru ctu re of the V SV L protein ;(b)u singthe atomic mod elof the second N N S viru s L protein to selectacand id ate inhibitorfrom a virtu allibrary of 3D stru ctu res of potentialcand id ates by performing stru ctu re based compu tational mod eling;

(c)obtaining the cand id ate inhibitor;and (d )measu ring an activity of L protein of the second N N S viru s when contacted with the cand id ate inhibitor obtained in step (b),wherein a d ecrease in the activity confirms thatthe cand id ate is an inhibitorof the second N N S viru s. 5.The method of any one of paragraphs 1-4 wherein the highresolu tion 3D stru ctu re has aresolu tion of atleast3.8 Å . 6.The method of paragraph 4 wherein the high resolu tion 3D stru ctu re is an atomic stru ctu re d etermined viacryo-electron microscopy and single particle analysis. 7.The method of any one of paragraphs 1-4 wherein the N N S viru s L protein is vesicu larstomatitis viru s L protein. 8.The method of paragraph7 wherein the V SV L protein highresolu tion stru ctu re has space grou p P 1, so as to form a u nit cell of d imensions of a=112.0 Å , b=143.0 Å , c=106.0 Å and α,β,γ=90^o. 9. The method of paragraph 7 wherein the V SV L protein high resolu tion 3D stru ctu re has the coord inates as setou t,forexample,in the stru ctu re d eposited withthe RC SB P rotein D ataB ankwith Reference N u mber 5A 22, and is accessible on the W orld W id e W eb at rcsb.org/pd b/explore.d o?stru ctu reId =5A 22. 10.The method of any of paragraphs 1,3 or 4 wherein the activity of the L protein measu red is selected from the grou p consisting of RN A d epend ent RN A polymerase activity, GD P polyribonu cleotid yltransferase (P RN Tase) activity,methyltransferase activity,and the ability to stru ctu rally rearrange L protein d omains d u ringL protein fu nctions. 11.The method of paragraph 2 or paragraph 10 wherein the RN A d epend entRN A polymerase activity measu red is selected from the grou p consisting of transcript initiation,elongation and termination. 12.The method of paragraph 1,2 or4 wherein,ratherthan selecting acand id ate inhibitorfrom a virtu allibrary of 3D stru ctu res of potentialcand id ates,the high resolu tion 3D stru ctu re of the N N S viru s L protein is u sed to d esign apotentialinhibitor.

13. The method of paragraph 12 wherein the potentialinhibitoris d esigned to interactwith an L protein d omain selected from the grou pconsistingof the RN A d epend entRN A polymerase (Rd Rp) d omain,the mRN A capping d omain,the methyltransferase d omain,the loop primer d omain,the connectord omain,and the C -terminald omain. 14.A method of mappingthe location of resistance to aknown inhibitorof an N N S viru s L protein on the 3D stru ctu re of the L protein,the method comprising:(a)homology-based stru ctu ralmappingof the sequ ence comprising the resistance mu tation onto the high resolu tion 3D stru ctu re of an N N S viru s L protein. 15. The method of paragraph 14 wherein the high resolu tion 3D stru ctu re is a 3.8 Å resolu tion stru ctu re. 16.A method of inhibitingan N N S viru s,the method comprisingcontactingan N N S viru s oracell infected with an N N S viru s with an inhibitor id entified accord ing to the method of any one of paragraphs 1-4. 17.A n N N S viru s inhibitorid entified u singthe method of any one of paragraphs 1-4. 18.The method of any one of paragraphs 1-4 wherein the id entified inhibitorbind s ad omain of the L protein ata region selected from the grou p consisting of the RN A d epend entRN A polymerase (Rd Rp)d omain,the mRN A cappingd omain,the methyltransferase d omain,the loopprimerd omain, the connectord omain,and the C -terminald omain of the L protein. 19.The method of any one of paragraphs 1-4 wherein the id entified inhibitorattenu ates the RN A d epend entRN A polymerase activity,the GD P polyribonu cleotid yltransferase (P RN Tase)activity,the methyltransferase activity,the loop priming activity,or the ability of the L protein to stru ctu rally rearrange its d omains d u ringL one ormore of its fu nctions. 20.The method of any one of paragraphs 1-4 which fu rther comprises the step of contacting the confirmed cand id ate inhibitorwithacellinfected withan N N S viru s. 21.The method of any of paragraphs 1-4 wherein the N N S viru s is selected from the viru ses of the Filovirid ae,P aramyxovirid ae,Rhabd ovirid ae and B ornavirid ae families. 22.The method of paragraph 21 wherein the N N S viru s is aFiloviru s,selected from Ebolaviru s, M arbu rgviru s and C u evaviru s. 23.The method of paragraph 21 wherein the N N S viru s is aP aramyxoviru s selected from measles viru s,N ipahviru s,H end raviru s,respiratory syncytialviru s and N ewcastle d isease viru s. 24.The method of paragraph21 wherein the N N S viru s is aRhabd oviru s selected from L yssaviru s, V esicu loviru s,P erhabd oviru s,Sigmaviru s,Ephemeroviru s,Tibroviru s,Tu paviru s and Spriviviru s. 25.A method of d esigning a cand id ate inhibitor of an N N S viru s,the method comprising:(a) id entifyingan N N S viru s mu tantthatescapes the activity of aknown inhibitorof the L protein of an thatviru s;(b) id entifyingthe site of the L protein resistance mu tation by sequ ence analysis;(c)u sing homology-based stru ctu ralmapping of the sequ ence comprising the resistance mu tation to mapthe mu tation site onto ahigh resolu tion 3D stru ctu re of an N N S viru s L protein;(d )u sing compu ter mod elingwiththe highresolu tion 3D stru ctu re to selectingacand id ate bind ingmolecu le thattargets the site of the resistance mu tation mapped in step (c);and (e) obtaining the cand id ate inhibitor molecu le id entified in step(d ). 26.The method of paragraph25fu rthercomprisingthe stepof contactingL protein of the N N S viru s withthe cand id ate inhibitorin an in vitro assay of L protein activity,wherein ad ecrease in L protein activity ind icates the cand id ate inhibitoris an N N S viru s inhibitor. 27.The method of paragraph25fu rthercomprisingthe stepof contactingthe cand id ate inhibitorwith the N N S viru s and assayingviralfu nction,wherein ad ecrease in viralfu nction ind icates the cand id ate inhibitoris an N N S viru s inhibitor. 28.A method of d esigning a cand id ate inhibitor of an N N S viru s,the method comprising:(a) id entifyingan N N S viru s mu tantthatescapes the activity of aknown inhibitorof the L protein of an thatviru s;(b) id entifyingthe site of the L protein resistance mu tation by sequ ence analysis;(c)u sing homology-based stru ctu ralmapping of the sequ ence comprising the resistance mu tation to mapthe mu tation site onto a high resolu tion 3D stru ctu re of an N N S viru s L protein;(d ) id entifying a potentialrefinementto the known inhibitor thatpermits interaction with the mu tantL protein;(e) obtaining the inhibitor molecu le comprising the refinementid entified in step (d );and (f)assaying mu tantviru s replication orinfection in the presence and absence of the refined inhibitormolecu le, wherein ad ecrease in replication orinfection ind icates the refined inhibitoris effective againstthe mu tantviru s. 29.Use of an N N S viru s inhibitorid entified accord ingto the method s of any one of paragraphs 1-4, 25and 28 forthe treatmentof an ind ivid u alinfected withan N N S viru s. 30.The u se of paragraph29 in whichthe N N S viru s is selected from the grou pconsistingof viru ses of the Filovirid ae,P aramyxovirid ae,Rhabd ovirid ae and B ornavirid ae families. 31.A compu ter-read able med iu m with high resolu tion 3D stru ctu re coord inate information for an N N S L protein stored thereu pon.

EX A M P L ES Example 1

Protein expression and purification

P rotein expression and pu rification were carried ou tby insectexpression of V SV -L ,u singbacu loviru s vector created with pFastB ac D u al(Invitrogen),placing L u nd er the controlof the polyhed rin promoter and green flu orescent protein (GFP ) u nd er controlof the P 10 promoter (to visu alize expression). Sf21 cells were infected ,incu bated at27 °C for 60-72 hou rs,and harvested as cell pellets by centrifu gation.Followinglysis by sonication and removalof celld ebris by centrifu gation, the L protein was pu rified by N i-nitrilotriacetic acid (N TA )chromatography followed by H i-TrapS and sizeexclu sion chromatography.V SV P ,resid u es 35-106 with an N -terminal6xH is-tag followed by atobacco-etchviru s (TEV )protease recognition motif,was expressed in RosettaB L 21 (D E3)E. colicells grown in L B med iu m containing 100 µg/mL ampicillin.P rotein expression with 0.8 mM IP TG atopticald ensity 0.8 ,and incu bated overnightat18 °C .The P (35-106) fragmentwas first pu rified by N i-N TA agarose chromatography followed by tag removalby incu bating with TEV protease overnightat4 °C .A second rou nd of N i-N TA chromatography separated cleaved P (35-106) from u ncleaved prod u ct.The cleaved proteins were d ialyzed against25 mM Tris pH 7.4,250 mM N aC l,1 mM D TT.P u rified V SV -L and P (35-106)were incu bated overnightat4 °C in amolarratio of 1:4,and the complex was isolated on aSu perd ex 200 gelfiltration colu mn in 25 mM H EP ES pH 7.4,250 mM N aC l,6 mM M gSO 4,0.5mM TC EP .

These method s can be applied to any N N S viru s L protein. Electron microscopy

The pu rified V SV -L :P complex was examined negatively-stained samples on aP hilips C M 10 electron microscope (EM ).Forcryo preparation,3.5µL of protein was applied at~0.35mg/mL to aQ u antifoil R1.2/1.3 C u grid (400 mesh)(Q u antifoil,Germany)thathad been glow d ischarged at40 mA for30 s. Grid s were plu nge-frozen withan FEIV itrobotM arkI,withthe followingsettings:65 % hu mid ity, offset-3,blottime 2 s,d rain time 1 s.Images were record ed withliqu id -nitrogen coolingon aTecnai F20 EM (FEI)withaC T3500 cryo-specimen hold er(Gatan);the microscope was operated at200 kV ; the d efocu s range was 0.9-2.3 µm.A semi-au tomated acqu isition program,UC SFImage4 to record movies with a K2 Su mmitd irectd etector (Gatan),operated in su perresolu tion mod e with d ose fractionation.The nominalmagnification was 29,000x,correspond ingto acalibrated magnification of 40,410x on the sensorplane of the camera.The beam intensity was setto 8 e/pixel/s.D u ring a6 s exposu re,30 frames of 200 ms eachwere collected foratotalelectron d ose of 31 e/Å 2.Frames were binned over2x2 pixels,yield ingapixel

size of 1.24 Å ,and aligned to eachotheru singthe program d osefgpu _d riftcorr. Image processing

A totalof 356,611 particles were picked by hand from 6x binned images from 1272 movies.Two- d imensionalclassification with mu ltivariate statisticalanalysis (M SA )was carried ou tin IM A GIC and K-means classification in TIGRIS (http://tigris.sou rceforge.net).Image d efocu s was d etermined with C TFFIN D 3 and class averages were calcu lated with fu llcontrast transfer fu nction (C TF) correction.Good class su ms were selected as references for particle alignmentand su bsequ entre- classification,iteratingtwice.A n initialmod el,calcu lated withEM A N 2 (e2initialmod el.py)u sed 292 class averages.Refinementand three-d imensionalclassification (3 classes)in FREA L IGN of the 292 C TF-corrected class averages resu lted in one class withabou t10 Å resolu tion,whichwas u sed as an initialreference forrefinementand classification of the fu llparticle stack. FREA L IGN was also u sed forrefinementand three-d imensionalclassification (3 classes)of the fu ll- d ose particle stack,initially u sing3x binned images (highresolu tion limit10 Å ).A fter11 cycles of refinementand classification,the compu tation switched to u nbinned particles.A fter160 cycles,the resolu tion had grad u ally extend ed to 7 Å ,atwhich pointthe bestsetof 155,443 particles were extracted .A fu rther100 cycles of refinementand classification (3 classes)extend ed the resolu tion to 6 Å .74,940 particles were then extracted with the bestscores from the two bestclasses.A final7 cycles of refinementof angles and shifts (u singa6 Å reference mod el)alternated between the fu ll- d ose (31 e/Å 2)stackand alow-d ose (12 e/Å 2)stack.The finalmap(3.8 Å resolu tion atFSC =0.143 criterion) was calcu lated from the low-d ose images.A similar approach can permit yet fu rther refinementin resolu tion. Model building

The polypeptid e chain of the Rd Rp,ring-like d omain was traced u singthe programs O and C oot.C oot was u sed to place standard poly-alanine α-helices into evid ent helical d ensity featu res, which were u su ally wellenou ghd efined to d etermine polarity,and connected the helices withpoly-alanine loops, following strong d ensity. The reoviru s and rotaviru s polymerases (λ3 and V P 1, respectively) were u sed as connectivity gu id es,having established correspond ence of helicalsegments overaspan of abou t600 totalresid u es.The methyltransferase d omain was bu iltfollowingasimilarstrategy,gu id ed by the consensu s fold of S-ad enosylmethionine (SA M )-d epend enttransferases.The cappingd omain, connector d omain,and C -terminald omain have no known homologs;the initialmod els therefore were bu iltd epend ingon d ensity,confirmingand ad ju stingconnectivity su bsequ ently by reference to the amino-acid sequ ence.Sid e-chain d ensity was strong enou gh in second ary stru ctu re elements of each d omain to establish the sequ ence register. Second ary stru ctu re pred iction (www.pred ictprotein.org) helped locate principal helices and strand s.The entire stru ctu re was checked and corrected with O ,u sing the lego-loop provision to rebu ild many of intersecond ary- stru ctu re loops and ad ju sting sid e-chain torsion angles to fitd ensity.Forthe following segments of the capping and connector d omains,the d ensity d id notallow confid entassignmentof backbone stereochemistry, and C α positions will have larger errors than in the rest of the mod el: 1159-1171; 1210-1226;1308 -1334;1512-1518 ;1534-1541. The stru ctu re coord inate information for the V SV L protein d etermined as d escribed herein is d eposited withthe RC SB P rotein D ataB ankwithReference N u mber5A 22,and is accessible on the W orld W id e W eb atrcsb.org/pd b/explore.d o?stru ctu reId =5A 22.See also L iangetal.,C ell162:314- 327 ,2015,whichprovid es stru ctu re figu res withd etails in color;the reference is incorporated herein by reference.The 2109 amino acid V SV L protein sequ ence is available atGenB ankA ccession N o. P 03523 (Ind ianastrain)and P 16379 (N ew Jersey strain). L protein sequ ences of otherN N S viru ses are read ily available.Examples inclu d e,bu tare notlimited to Rabies viru s L protein (A ccession N o. P 16289),Synochu s yellow net viru s (A ccession N o.P 31332),M easles viru s (A ccession N o. P 12576),Send ai viru s (A ccession N o.Q 06996),hu man P arainflu enza viru s 2 (A ccession N o. P 26676), hu man P arainflu enza viru s 3 (A ccession N o.P 12577 ),hu man P arainflu enza viru s 5 (A ccession N o.Q 03396),N ewcastle d isease viru s (A ccession N o.P 11205),hu man Respiratory syncytialviru s (A ccession N o.P 28887 ),Ebolaviru s Zaire (Q 05318 ),M arbu rgviru s (A ccession N o. A A R85466.1), C u evaviru s (A ccession N o. Y P _004928143), N ipah viru s (A ccession N o. A EZ01398.1), H end ra viru s (A ccession N o. N P _047113.3), L yssaviru s (A ccession N o. Y P _001285397.1),P erhabd oviru s (A ccession N o.Y P _007641367.1),Eel viru s (A ccession N o. A H D 46130.1), Sigmaviru s (A ccession N o. A C U65445.1), Ephemeroviru s (A ccession N o. Y P _009094434.1), Tibroviru s (A ccession N o. Y P _007641376.1), Tu paviru s (A ccession N o. Y P _238534.1)and Spriviviru s (A ccession N o.A B W 24037.1). The sequ ences forthe L proteins of otherstrains of the variou s N N S viru ses are also read ily available in,e.g.,GenB ank. Structure refinement

The d ensity mapwas fine sampled on agrid with0.72 Å spacingand transferred the d ensity (and mod el)into aP 1 cell(a=112 Å ,b=143 Å ,c=106 Å ,with90° angles)u singM A P RO T from the C C P 4 su ite.This mapwas fu rthersolventflatted (FIG.2)by calcu lating amask arou nd the mod elwith a probe rad iu s of 3.9 Å and setting grid points ou tsid e the masked region to a constant valu e correspond ingto 0.33e/Å 3.D ensity within the maskwas the setto an absolu te scale by d etermininga scale factorassu ming0.33 e/Å 3 and 0.43 e/Å 3,respectively forsolventand protein within the mask (d etermining the " d ry" protein volu me from its mass and partialspecific volu me;the " hyd ration" calcu lated this way is abou t0.3 w/w).M apand mask operations were carried ou twith M A P M A N . The amplitu d es (FP )and phases (P H IO )were calcu lated from the solvent-flattened map.A lthou gh amplitu d e stand ard d eviations (SIGFP )was notu sed in any of ou rcalcu lations,d u mmy valu es were su pplied (SIGFP = 0.1 FP )to satisfy inpu trequ irements of certain programs.4 % of the stru ctu re factors was flagged as a cross valid ation setfor calcu lating Rfree.Figu res of merit(FO M ) were estimated from the phase angle d ifference between the two half-setreconstru ctions and calcu lated H end rickson-L attman coefficients from P H IO and FO M .The stru ctu re was refined againstamplitu d es and phases,inclu d ing d ata to a minimu m B ragg spacing of 3.8 Å .P roced u res in P H EN IX were applied ,u sing a protocolwith severalrou nd s of ind ivid u ally restrained positionaland B -factor refinement,inclu d ing one rou nd of torsion-angle simu lated annealing and real-space refinement. A ppropriate weights were d etermined forthe experimentalterms in the targetfu nction by monitoring Rfree,mod elgeometry,and B -factorstatistics.Second ary stru ctu re restraints were applied throu ghou t the refinementand Ramachand ran restraints in the finalrou nd .The finalmod elwas analyzed with M olP robity.Refinementand mod elstatistics are shown in Table 1 below.

Example 2

High resolution structure determination

Images from V SV -L bou nd with P (35–106) were record ed from grid s prepared for cryo-EM as d escribed in ExperimentalP roced u res.A lld atawere taken on an FEIF20 microscope withaGatan K2 Su mmitd etector.Followingtwo-d imensionalclassification withIM A GIC (van H eeletal.,1996) and TIGRIS (http://tigris.sou rceforge.net) and calcu lation of an initialthree d imensionalreference d ensity withEM A N 2 (Tangetal.,2007 ),FREA L IGN was u sed forrefinementand three-d imensional classification.FIG.1 shows images,su mmarizes stages of the analysis,and illu strates the final3.8 A ° resolu tion d ensity map. Domain Organization of VSV-L

The polypeptid e chain from resid u e 35 to the carboxy terminu s,resid u e 2109 was traced ,leavingou t poorly ord ered linkersegments 1335–1357 and 1558–1597 and ashort,poorly ord ered loop1840– 1849.Some segments of ~10–15 resid u es between 1100 and 1334 and between 1358 and 1557 are in poord ensity,and the chain trace in those regions is approximate.B ecau se the beginningand end of the segmentand‘‘pu d d les’’of d ensity leave little u ncertainty abou tthe overallcou rse of the segment in qu estion,those resid u es are keptin the mod el. C onsid erable attention was paid to correctstereochemistry d u ringmod elbu ild ing,bu tto ad ju stboth the fitand the stereochemistry beyond the capacity of visu alinspection,one rou nd of refinementwas carried ou t,by calcu latingstru ctu re factors from the mapand u singbothamplitu d es and phases in the targetfu nction,with second ary-stru ctu re hyd rogen bond s imposed as restraints.A fter some minor ad ju stments and one more cycle,Rfree and Rworkwere 29.6 and 26.2,respectively (FIG 2;Table 1). The resid u es assigned to d omains and linkers as follows:Rd Rp,35–865;cappingd omain,866–1334; linker1 1335–1357 ;connectord omain,1358–1557 ;linker2,1558–1597 ;methyltransferase,1598– 1892;C -terminald omain,1893–2109 (FIG 2,FIG 3,FIG4 ;Table 1).The bou nd aries between the capping d omain and the connector d omain and between the latter and the methyltransferase are evid entfrom the d isord ered linkers thatintervene.The bou nd ary between the Rd Rpand the capping d omain correspond s to apreviou sly id entified tryptic cleavage site.A lthou gh the interface between the two d omains is relatively extensive and well packed ,fragment 1–860 can be expressed ind epend ently (Rahmeh et al., 2010).The bou nd ary between the connector d omain and the methyltransferase d omain also correspond s to two ind epend ently stable fragments:1–1593and 1594– 2109.B etween the methyltransferase d omain and the C -terminald omain is an extend ed bu trelatively polarinterface.N egative stain electron microscopy of afragmentthatinclu d es resid u es 1594–2109 shows thatstain can penetrate between the two d omains and thattheirconnection mightbe flexible (Rahmehetal.,2010).There are welld efined second ary-stru ctu re featu res in the d ensity foreachof the d omains,lead ingu s to believe thatpoord ensity forcertain segments reflects locald isord er,rather than failu re of three-d imensionalclassification algorithms to d etectoverallvariation in position or orientation of any particu lard omain withrespectto the others. RdRp

The Rd Rphas atits core aright-hand ,‘‘fingers-palm-thu mb’’stru ctu re (resid u es 360–865)common to avery large grou pof RN A and D N A polymerases (FIG 5).The catalytic site on the palm is in a d eepchannelbetween the fingers and thu mb su bd omains (extend ed from the palm as if in aloose hand grip).A ppend ed to the core on the N -terminalsid e is aglobu larregion (resid u es 1–359)that closes the channelon one end and reinforces the relatively slend er thu mb su bd omain.From the appearance of V SV -L in negative-stain electron microscopy,and in particu lar from the size and stainingof a‘‘d ou ghnu t-like’’part,the Rd Rpcan be similarin cage-like stru ctu re to the d sRN A viru s polymerases (Rahmeh etal.,2010).Those enzymes have their catalytic sites atthe center of an enclosed cavity,connected to the exterior by fou r channels,for template entrance,template exit, transcriptexit,and N TP access (L u etal.,2008 ;Tao etal.,2002).C omparison of the chain trace with their stru ctu res shows thatthis su ggestion was correct,with one mod ification.The d sRN A viru s Rd Rps have aC -terminal‘‘bracelet’’d omain thatencircles the exitpathforthe template and inclu d es asite forbind ingthe methylG capon the non-template,plu s-sense strand (L u etal.,2008 ;Tao etal., 2002).In V SV L ,the cappingd omain,whichhas no stru ctu ralsimilarity to the braceletd omain of the d sRN A viru s Rd Rps,occu pies the correspond ingspace.Thatis,resid u e 865,taken as the end of the Rd Rp,is atthe C terminu s of the thu mb.

The positions of secondary stru ctu ral elements were compared within V SV L , reoviru s λ3 (Tao et al., 2002),rotaviru s V P 1 (L u etal.,2008 ),and the heterotrimeric influ enzaviru s polymerase (P flu getal., 2014;Reichetal.,2014).The second ary stru ctu ralelements withcorrespond ences in the three other polymerases extend from abou tresid u e 107 in V SV -L to the end of the Rd Rpd omain (FIG 5).The analogou s parts of reoviru s l3 encompass resid u es 150–860 (approximately);those of rotaviru s V P 1, resid u es 135–750;those of hu man influ enzaviru s B polymerase,resid u es 415to the C terminu s (714) of the P A su bu nit,and resid u es 8–586 of the P B 1 su bu nit.The homology thu s extend s from the mid d le of P A into P B 1.The region in common between V SV -L and influ enzaviru s P B 1 correspond s to the fingers-palm-thu mb core stru ctu re,and the region shared withP A is alarge partof the Rd RpN - terminald omain. Capping Domain

Unlike the correspond inghost-cellprocess,the cappingreaction of N N S RN A viru ses proceed s from acovalentlinkage between the 5'end of the RN A and ahistid ine resid u e,withattackon thatlinkage by agu anosine nu cleotid e.The enzyme is thu s apolyribonu cleotid yltransferase (P RN Tase)rather than agu anylyltransferase.Two conserved motifs— GxxT and H R,separated by ~70 resid u es— mark the catalytic site (FIG 6A and 6B ).The formerparticipates in gu anosine nu cleotid e bind ing;the latter is the site of covalentRN A attachment.The d omain has no stru ctu ralhomologs thatcou ld be d etected with stand ard search method s.The largely a-helical,N -terminalhalf (resid u es 866–1100),which abu ts the polymerase d omain,is wellord ered .The C -terminalhalf (1100–1334)has severalpoorly d efined segments,inclu d ingthe loopthatbears the H R sequ ence.D espite u ncertain d efinition of sid e chains,the separation of the two conserved sites is 10 Å ,appropriate if aGTP bou nd atthe formeris to attack the histid ine-ligand ed 5'phosphate of the nascentRN A atthe latter(FIG 6B ).P ositions correspond ingto sites in hu man respiratory syncytialviru s (hRSV )of resistance mu tations to asmall- molecu le cappinginhibitor(L iu zzietal.,2005)impinge on the active site from three sid es (FIG 6A ); theirlocations,and the relatively poord efinition of the active site in the map,su ggestthatactivation of the d omain,perhaps by bind ing the 5'end of the nascentmessage,ind u ces a conformational rearrangement,similarto the d omain closu res seen in many enzymes when they bind theirsu bstrate. Two cand id ate Zn sites,one withcleard ensity where two C ys (resid u es 1120 and 1123)and two H is (1294 and 1296)ligand the likely Zn ion,and one withthree C ys (resid u es 1081,1299,and 1302)and aGlu (1108 ),contribu te stru ctu ralintegrity to the cappingd omain.The sites are close to eachother and wellou tsid e the catalytic center.In bothcases,the ligand ingresid u es are presentas aconserved setin mostN N S viru s L proteins and absentas asetin the others. A loop between resid u es 1157 (the threonine of the GxxT motif)and 1173 projects back into the cavity of the polymerase d omain (FIG 5B ,6C ,and 7 ).The poorly ord ered tipof this loopoccu pies the same position as the primingloopin the reoviru s polymerase (Tao etal.,2002).The loopin V SV -L polymerase d omain thatcorrespond s to the l3 primingloopis shorterthan its reoviru s homolog,and the capping-d omain loop projects over it.N either polymerase requ ires a polynu cleotid e primer to initiate, and the priming loop in the reoviru s polymerase su pports the initiating nu cleosid e triphosphate.A s elongation proceed s,the tip of the loop reced es to make room for the d sRN A replication prod u ctorforthe shortd ou blestrand ed region ju stu pstream of the newly ad d ed nu cleotid e d u ring transcription (Tao etal.,2002).This loop,which contacts the minor groove of the nascent prod u ct,can also enhance fid elity,by retard ing elongation of mismatches d etected by poorminor- groove geometry and allowing more time forA TP based pyrophosphorolysis of the mismatch.The position of the likely primingloopof V SV -L on the cappingd omain,ad jacentto the GxxT resid u es, su ggests cou plingof cappingto initiation of polymerization.

Connector Domain

The connector d omain is a bu nd le of eight helices (FIG 3B );it appears to have largely an organizationalrole in positioning or spacing the catalytic d omains.D isord ered linkers,23 and 40 resid u es long,respectively,lead into and ou tof the connectord omain.The end points of these linkers in well-d efined d ensity show thatthey mu stoccu py an extend ed groove between the capping and connector d omains; the groove also extend s into the interface between the capping and methyltransferase d omains

(FIG 8 ).Strong,low resolu tion d ensity featu res fillthis groove,bu tthey are notsharp enou gh to su ggestparticu lar linker conformations (FIG 8 B ).The location of P ind icated by negative-stain electron microscopy (FIG 8 A and 9;Table 2)lead s u s to su ggestthatthe groove also hold s some or allof the P fragmentpresentin the L -P complex imaged .B ecau se P (35–106) locks the smaller d omains of L into a fixed configu ration,itis plau sible thatitmightd o so by stabilizing fold ed stru ctu res for the two linker segments— glu ing them d own,so to speak,alongsid e the connector d omain (FIG.8 B ).

Methyltransferase Domain

The methyltransferase d omain has the stru ctu re characteristic of many otherd omains thatcatalyze transferof amethylgrou pfrom S-ad enosylmethionine (FIG.10 and FIG.11).Itmethylates boththe ribose O 20 and the gu anosine N 7 (Rahmeh etal.,2009).M ostof the d omain su perposes extremely wellon the flaviviru s methyltransferases,also d u alspecificity enzymes (Egloff etal.,2002;Ray etal., 2006;Zhou etal.,2007 ).Evid ence forfu nctionalfeed backfrom the V SV -L methyltransferase to the Rd Rp comes from the observation that ad d ition of S-ad enosyl homocysteine, which inhibits methylation,lead s to hyperpolyad enylation;mu tations thatpreventmethyltransfer have a similar effect(Galloway and W ertz,2008 ;L ietal.,2009;Rose etal.,1977 ).The methyltransferase d omain contacts both the connector and the capping d omains,bu tithas no d irectcontactwith the Rd Rp. M oreover,there is no obviou s‘‘tu nnel’’thatwou ld allow the 5'end of the transcriptto move from the catalytic site of the cappingd omain to the catalytic site of the methyltransferase d omain.Therefore, the L protein u nd ergoes asu bstantialconformation change followinginitiation of polymerization and thatthe inter-d omain commu nication shown in this stru ctu re is relevantto formation of the firstone or two phosphod iesterbond s,bu tnotto su bsequ entelongation and 5'-end mod ification. C-Terminal Domain

L ike the connectord omain,the C -terminald omain,whichterminates in a~25-resid u e longC -terminal ‘‘arm,’’ appears to have an essentially organizational role (FIG 3B ). It is largely an α-helical bu ndle, bu t a projecting, almost beak-like, β-hairpin su pported by a second interhelical loop, imparts a noticeable asymmetry.The C -terminalarm,afeatu re thatappears from sequ ence alignments to be conserved among N N S viralpolymerases,bu tvariable in length,extend s back againstthe Rd Rp, au gmentingthe b-hairpin thatbears the catalytic A sp-A sn sequ ence atits tip,and terminates atthe three-way ju nction of the capping,connector,and methyltransferase d omains,where ithas one or more contacts witheach.The arm thu s contribu tes to closingthe mu lti-d omain stru ctu re seen in the L - P complex,fu rtherstabilized by the phosphoprotein,P . Template Channel

Su perpositions of related positions in reoviru s λ3 (Tao et al., 2002) and rotaviru s V P 1 (L u et al., 2008 )have allowed u s to mod elabou nd template and atemplate-primer-N TP complex,becau se l3 was catalytically active in the crystals stu d ied u sed to d etermine the stru ctu re and V P 1 in its crystals incorporated template in asequ ence-specific register(Tao etal.,2002).The template entrance channel is atthe interface between the Rd Rpand the capping d omain (FIG 3C ).P olarand especially basic resid u es projectinto the groove from both sid es.A s in allpolymerases of this family,the template ru ns across a‘‘fingers loop’’ (resid u es 523–545 in V SV -L ) and twists sharply to present the templatingbase to the catalytic center.A hyd rophobic resid u e in the loop(P he541 in V SV -L )bears on the templatingbase to enforce correctbase pairingwiththe incomingnu cleosid e triphosphate.For initiation atthe 3'end of the viralRN A (eitherforreplication orfortranscription of lead erRN A ),the primingnu cleosid e triphosphate willrestagainstthe loopfrom the cappingd omain d escribed above (FIG 5B ).A ny fu rtherelongation,afterformingthe initialphosphod iesterbond ,willrequ ire this loop to move,and su bstantialelongation willalmostcertainly requ ire d isplacementof the entire capping d omain (FIG 6D ).To accommod ate transcriptionalelongation,the entire array of smallerd omains can reorganize. Domain Reorganization

The configu rations of V SV -L characterized herein by negative-stain electron microscopy illu strate the potentialforlarge-scale d omain reorganization (Rahmehetal.,2010).Images of L alone show acore ‘‘d ou ghnu t,’’which ad mits stain atits center,three globu lar append ages,in apparently variable positions and orientations withrespectto eachotherand to the core.A d d ition of P ,orof the peptid e, resid u es 35–106,u sed to stabilize the complex stu d ied here,locks the append ages in place (Rahmehet al.,2010,2012).M any of the projections of this locked stru ctu re resemble a FIG‘‘10.’’C lass averages from these images agree extremely wellwith projections of the stru ctu re d escribed (FIG 8 A ),as d o class averages from images of fou rd ifferentL fragments (FIG 8 C ).O ne of these (1–860) correspond s precisely to the Rd Rp.A nother,1–1121,inclu d es the Rd Rpand the largely helical,N - terminalhalf of the cappingd omain.The tryptic cleavage thatinitially generated thatfragmentis in a su rface loop. The last fragment previou sly imaged by negative staining comprises the methyltransferase and C -terminal d omains (FIG 8 C ) These comparisons show that previou s assignmentof the three globu larappend ages to the capping d omain,the methyltransferase d omain, and one u nassigned d omain (Rahmehetal.,2010)shou ld be mod ified .The cappingd omain is partof the d ou ghnu t,and the append ages correspond to the connector,methyltransferase,and C -terminal d omains,respectively (FIG 8 D ) The linkers between the capping d omain and the connector and between the connectorand the methyltransferase clearly allow the lattertwo to move away from the restof the molecu le;good d efinition in negative stain forthe third globu larappend age su ggests thatin the u nlocked stru ctu re,the C -terminalarm also pu lls away from the Rd Rp.M any of its interactions,as itinserts backagainstthe restof the molecu le in the stru ctu re d etermined herein,are ind eed withthe connectorand methyltransferase d omains.Images of negatively stained complexes of L with intact, d imeric P often show two,linked ,FIG-“10’’L molecu les,bu toccasionally the P d imer d oes not recru itasecond L and appears as asu rface featu re on the hookof the‘‘10’’(FIG 8 A ).C omparison of the L stru ctu re withthese projections is consistentwith ou rproposalthatthe interaction with P (35– 106)thatstabilizes the‘‘10’’conformation is withthe linkersegments ateitherend of the connector d omain (FIG 8 B ). Cryo-electron microscopy

H igh-resolu tion cryo-EM stru ctu re d etermination has u ntilrecently relied on eitherhighsymmetry or large size— forexample,icosahed ralviru ses,whichhave both,orribosomes,whichare large enou gh to prod u ce reasonable contrastforgettingstarted withiterative d etermination of particle orientations and centers (Grigorieff and H arrison,2011).The molecu larstru ctu re of an asymmetric protein of total mass <250 kD a.D ose fractionation (‘‘movies’’)has been d etermined herein,permitted by u se of a d irectelectron d etector,and refinementand maximu m-likelihood classification proced u res (L yu mkis etal.,2013),implemented in FREA L IGN ,facilitated aresolu tion ad equ ate to bu ild an atomic mod el. Sequential Transcription

A d e novo initiation eventwithA TP as initiatingnu cleotid e appears to startsynthesis of eachmRN A transcript.The stru ctu re is interpreted as thatof an early initiation state,representingan L -P complex read y forload ing onto the end of the template to synthesize lead erRN A .D u ring the transition to elongation the priming loop— contribu ted by the capping d omain— mu stshiftou tof the way to accommod ate the prod u ct.Inspection fu rther su ggests that after ad d ition of only a few more nu cleotid es,the cappingd omain as awhole mu stwithd raw from tightcontactwiththe Rd Rpto allow fu rtherelongation,as there d o notappearto be clearexitchannels fortranscriptand template.Upon termination of atranscript,the polymerase reinitiates on the nextgene,bu tthe efficiency of prod u cing the su cceed ing transcriptis only ~70% (Iverson and Rose,1981).The template entrance channelin V SV -L is atthe interface of the cappingd omain and the Rd Rp,and d issociation of the template will be straightforward (when initiation orearly elongation aborts)if thatinterface opens as su ggested . O therwise,the entire template wou ld have to thread throu gh the active site and emerge throu gh another channel.Transcription of the d ownstream gene probably requ ires reestablishing the inter- d omain contactseen in ou rstru ctu re,so thatthe primingloopcan reinsertinto the active site of the Rd Rpd omain forsu bsequ entd e novo initiation.Inhibition atany one of these rearrangements cou ld be effective to inhibit protein activity.That is,the id entification of these necessary d omain movements id entifies new targets forinhibition.Su chtargets have the benefitthatthey are likely to be u niqu e to the N N S viru ses oreven to aspecific viru s examined ,preventinglikelihood of sid e effects. Coupling of Capping, Polymerase, and Methyltransferase Activities

D isplacement of the capping d omain from the Rd Rp as elongation proceed s might have two consequ ences.First,the active site of the P RN Tase can reorganize (e.g.,by‘‘d omain closu re’’)into a betterord ered configu ration than the one seen in the presentstru ctu re.Second ,becau se the capping d omain faces boththe connectorand methyltransferase d omains,its d isplacementmightalso ind u ce rearrangementof the restof the capping machinery.A large-scale reorganization of this kind cou ld accou ntforsome of the observed fu nctionalcrosstalkbetween

the cappingand polymerase activities. A capis ad d ed only when the lengthof atranscripthas reached 31 nu cleotid es (Tekes etal.,2011).In vitro,very short(u pto 5 nt)transcripts can be capped in trans by L ,bu tthis process is inefficientand fails completely with longer transcripts (L ietal.,2008 ;O gino and B anerjee,2007 ).C onversely, mu tations in L thatd isru ptcapad d ition cau se prematu re termination (L ietal.,2008 ,2009).M u tations in the specific,cis-actingsignals atthe 5'end of the nascentstrand ,whichare absentfrom the lead er RN A ,also block capad d ition and resu ltin prematu re termination of thattranscript(L ietal.,2008 ; O gino and B anerjee,2007 ;Stillman and W hitt,1997 ;W angetal.,2007 ).The precision of the 31-nt requ irementsu ggests thatthe reorganized stru ctu re thatallows elongation is a well-d efined state, ratherthan aloosely ord ered one.

Reorganization of the cappingmachinery can also accou ntforwhy mRN A capmethylation requ ires no ad d itionalchain length (Tekes etal.,2011).In the configu ration represented by the stru ctu re d etermined herein,the catalytic sites forcappingand methylation are d istantfrom each other.If the smaller d omains move away from the polymerase core,the capped ,nascentRN A cou ld probably release from the cappingenzyme and gain immed iate access to the methylase d omain.M ethylation in trans can occu ru nd ersome circu mstances,bu tpreviou s workhas shown thattranscripts stalled ata chain lengthof 31 ntare fu lly methylated— presu mably in cis— by the stalled L (Tekes etal.,2011). Inhibition of mRN A capmethylation by highconcentrations of S-ad enosylhomocysteine can resu ltin hyper-polyad enylation,d emonstratingalinkage between the methylase and Rd Rpd omains (Galloway and W ertz, 2008 ; L i et al.,2009; Rose et al.,1977 ).The Rd Rp d omain of L carries ou t polyad enylation by iterative transcription of agene-end U tractelement.Some,bu tnotall,mu tations thatinhibitmethylation resu ltin the hyper-polyad enylation phenotype (Galloway and W ertz,2008 ), ind icating that the crosstalk mechanism is not a read ou t of cap mod ification bu t probably a consequ ence of interactions between d omains and between the protein and the nascenttranscript. The capmethylase of V SV participates in bothribose 2'O and gu anine-N 7 methylation reactions.The preferred su bstrate forallotherribose 2'O methyltransferases is 7 mGpppN and like otherproteins that recognize the mRN A cap stru ctu re— su ch as eIF4E–2'O methylases— those enzymes position the ribose in the active site by p-pstackinginteractions withthe 7 mG RN A .The ord erof capmethylation in V SV is reversed .M ethylation of 2'O preced es and facilitates su bsequ entmethylation of gu anine- N 7.The absence of aromatic resid u es thatcou ld participate in su ch interactions with a7 mGpppN RN A in the V SV methyltransferase is consistentwiththis altered reaction sequ ence.

The N Protein

The template forpolymerase is notnaked RN A ,bu tacomplex in whichthe template RN A is encased within the nu cleocapsid protein sheath.In thatcomplex the RN A bases are notaccessible to the Rd Rp of L ,and the N protein mu sttransiently d issociate from the RN A forthe Rd Rpto proceed (Green et al.,2006).The stru ctu re of L allows u s to estimate that20–25 ntof the template strand are thread ed throu ghthe polymerase d omain.A ccord ingly,becau se eachmolecu le of N covers 9 ntof RN A ,two or three molecu les of N mu stbe d isplaced from the template strand atany one time.A d jacentN su bu nits in the RN P interactstably,embracing each other throu gh N -and C -terminalextensions (Green etal.,2006).Thu s,looping ou tof template RN A need notentaild issociation of N from the RN P coil.Ind eed ,if the linked chain of N su bu nits is consid ered as the analogof acRN A strand ,then the d isplaced N is the cou nterpartof alooped -ou tplu s-sense strand d u ringtranscription by the related polymerases of d sRN A viru ses.This N -protein brid ge cou ld accou ntforthe precision of the 31-nt lengthof nascenttranscriptrequ ired forcapad d ition,perhaps by creatingad efined spacingbetween the Rd Rpand the popped -ou tcappingd omain.The N protein influ ences L activity,as recognition of the cis-acting signals in the genome requ ires itand as its presence influ ences incorporation by L of su bstitu ted nu cleotid e analogs (M orin and W helan,2014).Itcan also be necessary forcapping. The P Protein

P is an ad aptorthatengages boththe N -RN A template complex and the L protein.A small,globu lar d omain atthe C -terminalend of P (resid u es 195–265)interacts withthe N -RN A complex (Green and L u o,2009).This d omain cou ld in principle move from one su bu nitto the nextas polymerization proceed s.The stru ctu re d escribed herein contains only partof the N -terminalregion of P .A lthou ghit is poorly ord ered in the d ensity map,the P (35–106)can occu py some of the strong,low resolu tion d ensity featu res between the cappingand connectord omains,lockingin the linkersegments atboth end s of the latter.D epend ingon the chain polarity and on the flexibility of interveningsegments,the C -terminald omain of P cou ld lie nearthe openingthrou gh which RN A enters the active site of the Rd Rp and thu s,throu gh its interaction with N ,be partof the process thatfeed s template rapid ly throu ghthe polymerase channel. Homologies and Comparisons

H omologou s L proteins inclu d e those of rabies,Ebola,measles,and respiratory syncytialviru ses. A lignmentof theirsequ ences (FIG.2)shows the same overallarrangementof the variou s d omains, id entifies the active site resid u es of the protein forRd Rp,P RN Tase,and methyltransferase activities and su ggests d omain bou nd aries for expressing variou s fragments of the proteins from V SV and related viru ses.A llN N S RN A viru ses have a polymerase complex thatcomprises the enzymatic su bu nit,L ,and an equ ivalentof the V SV phosphoprotein,P .In some cases,ad d itionalviralproteins (V P 24 in the case of the filoviru ses,M 2-1 in the case of respiratory syncytialviru s)are necessary for fu llpolymerase processivity.The three-d imensionalinterconnections among d omains in the V SV polymerase su ggestthatbind ing of these accessory proteins to any of the smaller d omains might influ ence large-scale rearrangements and affectenzyme processivity. P arts of L are stru ctu rally similar to the heterotrimeric polymerase of influ enza viru s.The major d ifferences are the capping machineries,reflecting the d istinction between influ enza-viru s cap- snatchingand N N S RN A -viru s capsynthesis (Reichetal.,2014).A stru ctu ralfeatu re of the influ enza viru s polymerase thatis also absentfrom the V SV polymerase is the longP B 1 arm,whichis probably requ ired forastepof mRN A transcription (P flu getal.,2014;Reichetal.,2014). The V SV -L stru ctu re allows preliminary interpretation of mechanisms forinhibitors of its homologs. Formeasles viru s,resistance to anon-nu cleosid e analoginhibitorthatblocks gene expression maps to the polymerase d omain (Kru mm etal.,2014).The positions of those mu tations,mapped onto V SV -L , flankthe GD N Q motif atthe Rd Rpcatalytic site,su ggestingan allosteric mechanism,like thatof the non-nu cleosid e analog reverse transcriptase inhibitors for H IV .A compou nd active againstRSV targets the cappingd omain.The locations of positions in V SV -L thatcorrespond to sites of resistance mu tations in RSV -L are consistentwithou rproposalthatad omain closu re accompanies activation of the cappingactivity afterpolymerization has commenced (L iu zzietal.,2005). Example 3

Extension of in vitro L protein assay to Measles virus

RecombinantM easles viru s L -P complex was expressed as d escribed above forV SV L protein with some mod ifications.B riefly,7 xH is tagged L and P were co-expressed in Spod opterafu giperd a21 (Sf21) cells and affinity pu rified with N i-N TA agarose (Q iagen) followed by size exclu sion chromatography (Su perd ex 200 H R 10/30,GE healthcare)forthe W T protein,in abu ffercontaining 50 mM H epes pH 7.4,400 mM N aC l,1 mM D TT and 10% glycerol. Development of an in vitro RNA synthesis assay for MV

To d evelopan in vitro assay formeasles viru s RN A d epend entRN A polymerase the L and P proteins of the Khartou m Su d an (KS)strain of measles viru s were firstexpressed in insectcells from asingle recombinantbacu loviru s.The resu ltingL -P complex was pu rified by affinity forN i-N TA agarose and su bsequ entsize exclu sion chromatography (FIG 13A ).A portion of the L -P complex co-pu rified with an ad d itionalprotein whichid entified by mass-spectrometry as heatshockprotein (H sp)90 (FIG 13, lane 2).To d etermine whether the pu rified polymerase retains RN A d epend entRN A polymerase activity asimple in vitro assay u singas template achemically synthesized naked RN A correspond ing to the first19-ntof the measles viru s genome was d eveloped .The L -P ,and the L -P -H sp90 complexes exhibitRd RP activity in this assay (FIG 13B ),and inclu sion of the H sp90 inhibitorgeld anamycin had no effect on this activity (FIG 13C ).These resu lts d emonstrate that the pu rified polymerase complexes are equ ally active in vitro and thatH sp90 is notrequ ired forthe Rd RP activity of the complex. In vitro P olymerase assays formeasles viru s L protein were carried by u sing1.4 µM of template with 0.1 µM of M V L -P in areaction mixtu re containing 20 mM Tris-base pH 8 ,10 mM N aC l,6 mM M gC l2, 200 µM UTP , 1.5 mM A TP , 1.5 mM C TP and 165 nM of [α32P ]-GTP (3000 C i/mmol). Reactions were incu bated at 30°C for 3 hou rs,and stopped by ad d ition of ED TA /formamid e. Reactions prod u cts were resolved on a d enatu ring polyacrylamid e gel electrophoresis (20% polyacrylamid e,7 M u rea)in TB E bu ffer,and analyzed by au torad iography.The sizes of the prod u cts were d etermined by comparison witha19 ntmarkerRN A labeled by T4 polynu cleotid e kinase (N ew England B iolabs) u sing [γ32P ]-A TP (3000 C i/mmol). A ssays stu d ying the inhibitory effects of A s136a and ERD RP 0519 were carried ou tu sing the ind icated concentrations of the inhibitor.A ll rad ioisotopes were pu rchased from P erkinElmer. Modeling resistance mutations

Fu ll-length L proteins of M V (GenB ank:A D O 17332.1),C D V (GenB ank:A C Z56434.1)and V SV (UniP rotKB /Swiss-P rot:P 03523.2)have been aligned u sing T-C offee and shown u sing Espriptand location of mu tations were id entified and position on the V SV L stru ctu re (P D B :5A 22)u singP ymol. Forthe mod eling of ERD RP 0519 into the bind ing d omain of M V L ,ERD RP 0519 has been d rawn u singC hemd raw and mod eled u singP H EN IX ref).The polypeptid e chain of M V bind ingd omain has been traced with the M V L sequ ence (GenB ank:A D O 17332.1)and the V SV L map(P D B :5A 22) u singC oot. Effects of a non-nucleoside inhibitor and a more bioavailable derivative on RNA synthesis activity by MV L-P

Using cell-based lu ciferase replicon reporterassays forM V and canine d istemperviru s (C D V )the non-nu cleosid e inhibitor, 1-M ethyl-N -[4-(1-piperid inylsu lfonyl)phenyl] -3-(triflu oromethyl)-1H - pyrazole-5-carboxamid e (A s136a) and a more bioavailable d erivative (S)-1-methyl-N -(4-((2-(2- morpholinoethyl)piperid in-1-yl)su lfonyl)phenyl)-3-(triflu oromethyl)-1H -pyrazole-5-carboxamid e (ERD RP 0519),were shown to inhibitviralgene expression.Selection of viralmu tants resistantto those inhibitors id entified mu tations in the L gene have been d escribed { Kru mm,2014 #281;N d u ngu , 2012 #280 } .To d irectly d emonstrate thatthe non-nu cleosid e inhibits RN A synthesis theireffectwas tested on Rd RP activity in vitro. The inhibitors blockRN A synthesis initiation by L -P withan IC 50 of 70 nM forA s136aand 100 nM forERD RP 0519 (FIG.14)as d emonstrated by the u niform inhibition of allof the prod u cts of RN A synthesis.This resu ltshows thatthe non-nu cleosid e inhibitors inhibit the M V Rd RP . Effect of L gene mutations that lead to resistance to the non-nucleoside inhibitor

Resistance mu tations were previou sly shown to mapto the Rd Rpd omain of L { Kru mm,2014 #281} . A s a first step to u nd erstand which mu tations provid e resistance we expressed and pu rified polymerase complexes carrying either the mu tation L _{T7 7 6A} (L _{T7 7 6A} -P ) or L _{T7 51I} (L _{T7 51I}-P ).W hen compared to wild type L -P ,eachof the su bstitu tions led to asignificantd ifference in IC 50 of 1.3 µM forL _{T7 7 6A} -P and 110 µM forL _{T7 51I}-P (FIG.16B ,C ,D )-a10 and 1000 fold change.The effectof combined mu tations on M V L by generating an L protein thatcarries both T776A and T751I (L _{T7 7 6A /T7 51I}-P )was stu d ied .The combination of these two mu tations on L conferacomplete resistance to L _{T7 7 6A /T7 51I}-P as no inhibition is d etected even in the presence of 2.5 mM of inhibitor(FIG.16A ). These resu lts su ggestthatthe inhibitors bind d irectly M V L in an areaof L located close to its RN A synthesis active site. Recombinant measles containing substitutions in L exhibits complete resistance to non-nucleoside inhibitor

To d emonstrate thatthe two mu tations lead to complete resistance of the measles viru s to the non- nu cleosid e inhibitorarecombinantmeasles viru s KS strain expressing eGFP was engineered (FIG. 17 ).The kinetics of replication of the engineered variantM V -L _{T7 7 6A /T7 51I}were red u ced compared with the wild type parentas evid enced by the d ecreased viralyield at24hpostinocu lation of cells (FIG. 17 A ) consistentwith a fitness costto the resistance mu tations.The recombinantM V -L _{T7 7 6A /T7 51I} variantwas resistantto the non-nu cleosid e inhibitoron viralreplication as ju d ged by eGFP expression in the presence of 25 µM of the bioavailable inhibitor.This resu ltd emonstrates thatsu bstitu tions in measles L lead to complete resistance of the viru s to the non-nu cleosid e inhibitorin vitro and in cells bu tu ncoverafitness costto viralreplication associated withthose su bstitu tions. Example 5

Application of findings for VSV L protein to Rabies virus

Rabies viru s (RA B V )is anon-segmented negative-strand (N N S)RN A viru s from the Rhabod ovirid ae family thatgenerates lethalinfection in the brain of hu mans and mammals cau sing an estimated 55000 d eaths peryearwith4 ou tof 10 d eaths occu rringin child ren u nd erthe age of 15.A lthou ghan efficientvaccine-based treatmentexists,its pooraccess in ru ralregions of A siaand A fricasu ggests thatthe d evelopmentof antiviralcou ld be apossible alternative. N N S RN A viru ses share common replicative machinery thatconsists of aribonu cleoproteic (RN P ) complex composed by the genomic RN A completely coated by the viralnu cleoprotein (N )to form the N -RN A template fortranscription and replication by the polymerase complex.This latterconsists of the RN A -d epend entRN A polymerase (Rd RP )thatresid es within the 242-kD alarge (L )protein and the essential33-kD aco-factorphosphoprotein (P )thatbrid ges the interaction between N -RN A and the Rd RP .RA B V P is amu lti-fu nctionalprotein composed of three d omains,aC -terminald omain thatbind s the genomic N -RN A ,a centrald omain responsible for au to-oligomerization and a N - terminald omain withad u alfu nction of L -bind ingand N 0-bind ingd omain (withN 0 beingthe free- RN A monomeric solu ble form of N ).Extensive stu d ies of the vesicu lar stomatitis viru s (V SV ), anothermemberof the rhabod oviru s,have shown thatL also carries two more enzymatic activities,a GD P :polyribonu cleotid yltransferase thatcaps the newly synthesized mRN A ,and a d u alspecificity mRN A cap methyltransferase. M oreover, electron microscopy revealed that L u nd ergoes a conformationalchange viabind ingwithP and rend ers the Rd RP more active and processive. The N N S RN A viru ses minimalRN A synthesis machinery consists of L and anaked RN A has been d emonstrated .Usingpu rified recombinantV SV L and asynthetic RN A correspond ingto the first19 ntof the V SV genome,anew in vitro assay to stu d y RN A synthesis forV SV has been d eveloped . This assay became apowerfu lmethod formu ltiple stu d ies of the replicative machinery:investigate the RN A synthesis initiation mechanism and requ irements,stu d y the effects of viralor other or cellu larproteins on RN A synthesis and processivity,d iscovery and characterization of the mechanism of action of smallmolecu le inhibitors.W e sou ghtan RN A synthesis in vitro assay forRA B V for whichno RN A synthesis assay exists. M aterials and method s

Protein expression and purification

RecombinantL and P were expressed as d escribed in (Rahmehetal,2010).B riefly,6xH is tagged L was expressed in Spod opterafu giperd a21 (Sf21)cells and affinity pu rified with N i-N TA agarose (Q iagen)followed by ion exchange chromatography.10xH is tagged P was expressed in Escherichia coliB L 21 (D E3)cells orSf21 cells then affinity pu rified withN i-N TA agarose (Q iagen)followed by gelfiltration (Su perd ex 200 H R 10/30,GE healthcare). In vitro V SV L polymerase assay.RN A templates were chemically synthesized and pu rified (Integrated D N A Technologies).Stand ard polymerase assays were carried ou tu sing 0.2 µM of template with 0.2 µM of V SV L in areaction mixtu re containing 20 mM Tris-base,pH 8 ,50 mM N aC l,2 mM D TT and 0.5% (vol/vol)Triton X -100,6 mM M gC l2,200 µM UTP ,1.5 mM A TP ,1.5 mM C TP and 165 nM of [α32P ]-GTP (3000 C i/mmol). Reactions were incu bated at 30°C for 3 hou rs, and stopped by the ad d ition of ED TA /formamid e.Reactions prod u cts were resolved u singd enatu ring polyacrylamid e gelelectrophoresis (20% polyacrylamid e,7 M u rea)in TB E bu ffer,and analyzed by au torad iography.The sizes of the prod u cts were d etermined by comparison to a19 ntmarkerRN A labeled by T4 polynu cleotid e kinase (N ew England B iolabs) u sing [γ32P ]-A TP (3000 C i/mmol). A ssays stu d yingthe effects of N TP concentration were carried ou tu singthe ind icated concentrations. Formonitoringthe effectof varyingGTP concentration reactions were su pplemented with165 nM of [α32P ]-A TP (3000 C i/mmol). To d etermine the first nu cleotid e incorporated d u ring initiation, standard reactions were carried ou t with 165 nM [γ32P ]-GTP (3000 C i/mmol) or [γ32P ]-A TP (3000 C i/mmol).O n rand om RN A templates,A C RN A and A C G RN A ,200 µM of GTP ,1.5 mM of each, A TP and C TP and 165 nM of [α32P ]-UTP (3000 C i/mmol) were u sed . Internal initiation reactions were performed u sing 3’-A UGC UUC UGUUUGUUUGGUA -5’ (L e+1), 3’- GA UGC UUC UGUUUGUUUGGUA -5’ (L e+2) or 3’-GA A UGC UUC UGUUUGUU UGGUA -5’ (L e+3)as atemplate.The effectof P was examined by the ad d ition of 0.2 µM of P to the stand ard reaction.The effectof ribavirin triphosphate (RTP )was evalu ated on the C 3A template with165 nM of [α32P ]-UTP (3000 C i/mmol), 1.5 mM of each, A TP and C TP and 200 µM of either GTP or RTP . A ssays stu d ying the effects of RTP concentrations were carried ou t by ad d ing the ind icating concentrations of RTP in the stand ard polymerase assay.A llrad ioisotopes were pu rchased from P erkinElmer. Resu lts

Establishment and characterization of a minimal in vitro RNA synthesis assay for RABV

To establish aminimalin vitro RN A synthesis assay forRA B V ,the RA B V L and P proteins were expressed ind epend ently in insectcells and bacteria respectively,and pu rified them by N i-N TA affinity followed by asize exclu sion chromatography (FIG 20A ).Usingasystem d eveloped forV SV { M orin,2012 #96} ,anothermemberof the rhabd oviru s family,itis shown thatRA B V L can u se a non-encapsid ated RN A correspond ingto the first19 ntof the 3’end of RA B V genome (RA B V L e19) to synthesize RN A (FIG.20B ).This resu ltd emonstrates thatthe in vitro system previou sly d eveloped forthe vesicu loviru s can be ad apted forthe lyssaviru s.In the presence of RA B V P ,L RN A synthesis activity and processivity is enhanced (FIG.20B ).In ord erto observe if P of d ifferentrhabd oviru ses cou ld enhance the activity of any otherL proteins,the RA B V L has been incu bated witheitherRA B V orV SV P ,and V SV L witheitherV SV P orRA B V P .RA B V P can only enhance RA B V L activity, while V SV P can only enhance V SV L (FIG.20B ,C )This resu ltshows thatonly when L and P are comingfrom the same viru s L has an activity enhanced by P ,d emonstratingthatthe interaction L -P is specific of eachviru s species. Identification of the minimal P L-enhancing domain

A previou s stu d y have d emonstrated by co-immu noprecipitation thatthe region from aa1 to 19 of P is su fficientforL -bind ing bu tcou ld be stabilized by extend ing this region to aa52 { C henik,1998 #262} .Interestingly,the V SV L -bind ingd omain is located ju stu pstream the oligomerization d omain (aa92-131){ Gerard ,2009 #267 ;Rahmeh,2012 #165} .B ased on these d atasix P -d eletions mu tants were generated within the region 1-91 to id entify the minimalP region thatcou ld stillenhance both RN A synthesis and processivity of L ,P 1-91,P 21-91,P 11-91,P 1-40,P 1-50,and P 11-50.W hile P 21- 91 d oes notenhance L RN A synthesis activity,both P 1-91 and P 11-91 enhance L activity that d emonstrates thatthe region 11-21 of P is essentialforthe bind ingof L (FIG.21).In the same way P 1-40 d oes notmod ify the activity of L while P 1-50,enhances bothRN A synthesis and processivity showingthatthe region 40-50 is also criticalforL enhancement.B ased on these resu lts,P 11-50 was generated and observed thatthis d omain enhanced L polymerase activity.Thu s the minimalL - bind ing d omain of P from aa11 to 50 have been d efined herein.The resu lts confirms the previou s resu lts showingthatthe L -bind ingregion of L was encompassed within resid u es 1-52 { C henik,1998 #262} and show thatthe first10 aaare notessentialforthe enhancementof L . Template specificity of the RABV polymerase complex

The sequ ence properties of the template on the RN A synthesis activity of L were investigated .The activity of RA B V L on RA B V L e19 and asynthetic RN A correspond ingto the first19 ntof the 3’ end of V SV genome (V SV L e19)were compared .The RA B V L -P can initiate RN A synthesis on V SV L e19 (FIG 22A ).W e have previou sly d emonstrated thatthe firstthree ntof the template are essentialforV SV L to initiate RN A synthesis { M orin,2012 #96} .RA B V L e19 and V SV L e19 share the same firstthree nu cleotid es withthe firstd ifferentnu cleotid e located atposition 4,and 10 others located atposition 5,6,8 ,9,13,15,16,17 ,18 and 19 (FIG.22A ).A lthou gh RA B V L initiates on V SV L e19 itefficiency of initiation is lowerthan on RA B V L e19 (FIG.22B ),su ggestingarole of the ntlocated fu rtherthe firstthree positions in stabilizingthe initiation complex.L ookingatthe prod u cts size level,eachRN A template is observed to generate aspecific pattern (FIG.22C ).W hen RA B V L -P u se RA B V L e19 as atemplate the prod u cts-size thatare synthesized the mostare abou tfrom 9 to 13 and 19 ntlongwiththe 12 mersynthesized atalargerqu antity than the otherone.Interestingly very few RN A prod u cts are synthesized from 14 to 18 ntlong ind icating a good processivity of the polymerase arou nd this length.O n V SV L e19,RA B V L -P presentaprod u cts d istribu tion d ifferent than the one observe withRV L e19.W hile the 5 meris prod u ced atalargerlevel,L -P synthesized largerqu antity of the longsize prod u ct(>10 nt)withaslightred u ction arou nd the size of 15 nt.This resu ltsu ggests thatinteractions between the template and the entry tu nnelmightinflu ence the RN A synthesis by the polymerase. To compare if these resu lts wou ld be similar with another rhaboviru s L -P complex,a similar experimentwas cond u cted u singV SV L -P instead of RA B V L -P (FIG.22).A s observed withRA B V L -P ,V SV L -P can initiate on both V SV L e19 and RA B V L e19 templates,howeverinitiation with RA B V L e19 is less efficientthan withV SV L e19,witharou nd 50% less synthesis (FIG 23A ,23B ). Su rprisingly the profile of levelof synthesis size by size with V SV L -P on RA B V L e 19 is very similaras the one observed withRA B V L -P showingthe 12 mersynthesized atalargerqu antity and few RN A synthesis for the sizes from 15 to 18 ntlong (FIG.23C ).This su ggests thatthe RN A sequ ence itself can d rive the RN A synthesis property of bothpolymerases.H oweverwhen V SV L e19 is u sed as atemplate the 2 polymerases presentsome slightd ifferences. RABV L cannot copy the VSV N-RNA template

W e have d emonstrated herein thatRA B V L initiate RN A synthesis on asynthetic RN A correspond ing to the 3’end on the V SV genome (FIG.22).H oweverthe genomic RN A of the N N S RN A viru ses is completely enveloped by N to form the N -RN A meaningthatd u ringthe copy of the template by the polymerase complex asmallnu mberof N have to d issociate locally to letaccess the RN A into L active site.B ased on these two facts,the RA B V L can copy V SV N -RN A by u singachimerP protein thatcou ld bind both RA B V L and V SV N -RN A .D u ringtranscription,P bind s L viaits N -terminal end and N -RN A via its C -terminalend to brid ge the interaction between the polymerase and its template.A chimerP (cRA B V P )composed of the first91 aaof the N -terminalend of RA B V P and the oligomerization and C -terminald omains of V SV P ,resid u es 107 to 265was d esigned (FIG.24A ). W hile V SV L -P can u se V SV N -RN A as atemplate and synthesized the 47 ntlonglead erRN A ,the RA B V L /cRA B V P complex cannotcopy the template (FIG.29B ).M oreoverthe lack of any band s below the size of 47 ntind icates thatinitiation d oes notseem to occu ru singthe two complex RA B V L /cRA B V P and V SV -N -RN A .These resu lts su ggestthatd u ring RN A synthesis initiation of the polymerase complex on the N -RN A another interaction than L /P /N -RN A mightbe involved fora correctinitiation of RN A synthesis.P erhaps,ad irectbind ingL -N occu rs atearly stage of the RN A synthesis initiation as wellas d u ringelongation. In vitro assays RNA synthesis and initiation in vitro assay

Standard polymerase assays were carried ou t u sing 0.2 μM of template with 0.2 μM of V SV L in a reaction mixtu re containing 20 mM Tris-base,pH 8 ,50 mM N aC l,2 mM D TT and 0.5% (vol/vol) Triton X -100, 6 mM M gC l2, 200 μM UTP , 1.5 mM A TP , 1.5 mM C TP and 165 nM of [α32P ]-GTP (3000 C i/mmol).Reactions were incu bated at 30°C for 3 h,and stopped by the ad d ition of ED TA /formamid e. Reactions prod u cts were resolved u sing d enatu ring polyacrylamid e gel electrophoresis (20% polyacrylamid e,7 M u rea)in TB E bu ffer,and analysed by au torad iography. The sizes of the prod u cts were d etermined by comparison with a19-ntmarkerRN A labelled by T4 polynu cleotid e kinase (P N K) (N ew England B iolabs) u sing [γ32P ]-A TP (3000 C i/mmol). A ssays stu d ying the effects of N TP concentration were carried ou tu sing the ind icated concentrations.For monitoring the effectof varying GTP concentration,reactions were su pplemented with 165 nM of [α32P ]-A TP (3000 C i/mmol). To d etermine the first nu cleotid e incorporated d u ring initiation, standard reactions were carried ou t with 165 nM [γ32P ]-GTP (3000 C i/mmol) or [γ32P ]-A TP (3000 C i/mmol). O n rand om RN A templates, A C RN A and A C G RN A , 200 μM of GTP , 1.5 mM of each, A TP and C TP and 165 nM of [α32P ]-UTP (3000 C i/mmol) were u sed . Internal initiation reactions were performed u sing 3′ -A UGC UUC UGUUUGUUUGGUA -5′ (L e+1), 3′ - GA UGC UUC UGUUUGUUUGGUA -5′ (L e+2), or 3′ -GA A UGC UUC UGUUUGUUUGGUA -5′ (L e+3) as a template. The effect of P was examined by the ad d ition of 0.2 μM of P to the standard reaction. The standard polymerase assay was su pplemented with 250 μM GTP and 0.2 μM of P when N -RN A was u sed as atemplate.The effectof RTP was evalu ated on the C 3A template with165 nM of [α32P ]-UTP (3000 C i/mmol), 1.5 mM of each, A TP and C TP and 200 μM of either GTP or RTP . A ssays stu d ying the effects of RTP concentrations were carried ou t by ad d ing the ind icating concentrations of RTP in the stand ard polymerase assay.A llrad ioisotopes were pu rchased from P erkin-Elmer. in vitro capping assay

To testfor capped prod u cts,in vitro transcription reactions are carried ou t in the presence of [α- 32P ] GTP ;RN A was pu rified and d igested with TA P ,and the prod u cts resolved by thin layer chromatography.P rod u ctthatcomigrated with a 7 mGp marker,which confirmed thatthe RN A is capped and methylated . in vitro methylation assay

Triphosphorylated RN A transcripts are capped in the presence of 20 μC i of [α-32P ]GTP (3,000 C i/mmol; P erkin-Elmer) and 7.5 U of the vaccinia viru s capping enzyme (Epicentre) in 15-μl reaction mixtu res.A fter2 h at37 °C ,the reaction mixtu res were treated with 10 U of calf intestinalalkaline phosphatase (N EB )for1 h,followed by pu rification.C apped RN A was qu antified by measu ringthe incorporation of [α-32P ]GTP and extrapolation of molar qu antities from a standard cu rve of moles versu s cou nts per minu te. For methylation at the G-N -7 position, 100 μM SA M is inclu d ed in the capping reaction mixtu re. For methylation at the ribose 2′ -O position or at both the G-N -7 and ribose 2′ -O positions, the capped RN A or the G-N -7 -methylated RN A was incu bated with 100 μM SA M and 7.5 U of vaccinia viru s 2′ -O M Tase (Epicentre). The prod u cts of this reaction were pu rified, digested withnu clease P 1,and resolved by TL C on P EIcellu lose F sheets.N u clease P 1 d igestion cleaves the 3′ -5′ phosphodiester bond s in single-strand ed RN A bu t d oes not cleave the 5′ -5′ bond of the cap stru ctu re,resu lting in the release of GpppA from u nmethylated RN A .Following incu bation of the RN A with pu rified L protein,the prod u cts of nu clease P 1 d igestion comigrated with am7 GpppA m marker, d emonstrating that L possesses both G-N -7 and 2′ -O M Tase activities. A ll patents, patent applications, and pu blications id entified in this d ocu ment are expressly incorporated herein by reference for the pu rpose of d escribing and d isclosing,for example,the method ologies d escribed in su ch pu blications thatmightbe u sed in connection with the present invention.These pu blications are provid ed solely fortheird isclosu re priorto the filing d ate of the presentapplication.N othingin this regard shou ld be constru ed as an ad mission thatthe inventors are notentitled to anted ate su ch d isclosu re by virtu e of prior invention or for any other reason.A ll statements as to the d ate or representation as to the contents of these d ocu ments is based on the information available to the applicants and d oes notconstitu te any ad mission as to the correctness of the d ates orcontents of these d ocu ments. Itis u nd erstood thatthe foregoing d etailed d escription and the following examples are illu strative only and are notto be taken as limitations u pon the scope of the invention.V ariou s changes and mod ifications to the d isclosed embod iments,whichwillbe apparentto those of skillin the art,may be mad e withou td epartingfrom the spiritand scope of the presentinvention.Fu rther,allpatents,patent applications,and pu blications id entified are expressly incorporated herein by reference for the pu rpose of d escribingand d isclosing,forexample,the method ologies d escribed in su chpu blications thatmightbe u sed in connection with the presentinvention.These pu blications are provid ed solely fortheird isclosu re priorto the filingd ate of the presentapplication.N othingin this regard shou ld be constru ed as an ad mission thatthe inventors are notentitled to anted ate su chd isclosu re by virtu e of prior invention or for any other reason.A llstatements as to the d ate or representation as to the contents of these d ocu ments are based on the information available to the applicants and d o not constitu te any ad mission as to the correctness of the d ates orcontents of these d ocu ments. Itshou ld be u nd erstood thatthis invention is notlimited to the particu larmethod ology,protocols,and reagents,etc.,d escribed herein and as su chmay vary.The terminology u sed herein is forthe pu rpose of d escribing particu lar embod iments only,and is notintend ed to limitthe scope of the present invention,whichis d efined solely by the claims. REFEREN C ES

A braham,G.,and B anerjee,A .K.(1976).Sequ entialtranscription of the genes of vesicu larstomatitis viru s.P roc.N atl.A cad .Sci.USA 73,1504–1508.

A d ams,P .D .,A fonine,P .V .,B u nko´ czi,G.,C hen,V .B .,D avis,I.W .,Echols,N .,H ead d ,J.J.,H u ng, L .W ., apral,G.J.,Grosse-Ku nstleve,R.W .,etal.(2010).P H EN IX :acomprehensive P ython-based system formacromolecu larstru ctu re solu tion.A ctaC rystallogr.D B iol.C rystallogr.66,213–221. A lbertini, A .A ., W ernimont, A .K., M u ziol, T., Ravelli, R.B ., C lapier, C .R., Schoehn, G., W eissenhorn,W .,and Ru igrok,R.W .(2006).C rystalstru ctu re of the rabies viru s nu cleoprotein-RN A complex.Science 313,360–363.B all,L .A .,and W hite,C .N .(1976).O rd erof transcription of genes of vesicu larstomatitis viru s.P roc.N atl.A cad .Sci.USA 73,442–446.

B altimore,D .,H u ang,A .S.,and Stampfer,M .(1970).Ribonu cleic acid synthesis of vesicu lar stomatitis viru s,II.A n RN A polymerase in the virion.P roc.N atl.A cad .Sci.USA 66,572–576.

B arr,J.N .,and W ertz,G.W .(2001).P olymerase slippage atvesicu larstomatitis viru s gene ju nctions to generate poly(A )is regu lated by the u pstream 30-A UA C -50 tetranu cleotid e:implications forthe mechanism of transcription termination.J.V irol.75,6901–6913.

B arr,J.N .,W helan,S.P .,and W ertz,G.W .(1997 ).cis-A cting signals involved in termination of vesicu lar stomatitis viru s mRN A synthesis inclu d e the conserved A UA C and the U7 signalfor polyad enylation.J.V irol.71,8718–8725.

C hen,V .B .,A rend all,W .B .,3rd ,H ead d ,J.J.,Keed y,D .A .,Immormino,R.M .,Kapral,G.J.,M u rray, L .W .,Richard son,J.S.,and Richard son,D .C .(2010).

M olP robity:all-atom stru ctu re valid ation for macromolecu lar crystallography.A cta C rystallogr.D B iol.C rystallogr.66,12–21.

D ing,H .,Green,T.J.,L u ,S.,and L u o,M .(2006).C rystalstru ctu re of the oligomerization d omain of the phosphoprotein of vesicu larstomatitis viru s.J.V irol.80,2808–2814.

Egloff,M .P .,B enarroch,D .,Selisko,B .,Romette,J.L .,and C anard ,B .(2002).A n RN A cap (nu cleosid e-20-O -)-methyltransferase in the flaviviru s RN A polymerase N S5:crystalstru ctu re and fu nctionalcharacterization.EM B O J.21,2757–2768.

Emerson,S.U.,and W agner,R.R.(1973).L protein requ irement for in vitro RN A synthesis by vesicu larstomatitis viru s.J.V irol.12,1325–1335.

Emsley,P .,L ohkamp,B .,Scott,W .G.,and C owtan,K.(2010).Featu res and d evelopmentof C oot. A ctaC rystallogr.D B iol.C rystallogr.66,486–501.

Galloway,S.E.,and W ertz,G.W .(2008 ).S-ad enosylhomocysteine-ind u ced hyperpolyad enylation of vesicu lar stomatitis viru s mRN A requ ires the methyltransferase activity of L protein.J.V irol.82, 12280–12290. Green,T.J.,and L u o,M .(2009).Stru ctu re of the vesicu larstomatitis viru s nu cleocapsid in complex with the nu cleocapsid -bind ing d omain of the smallpolymerase cofactor,P .P roc.N atl.A cad .Sci. USA 106,11713–11718.

Green,T.J.,Zhang,X .,W ertz,G.W .,and L u o,M .(2006).Stru ctu re of the vesicu larstomatitis viru s nu cleoprotein-RN A complex.Science 313,357–360.

Grigorieff,N .,and H arrison,S.C .(2011).N ear-atomic resolu tion reconstru ctions of icosahed ral viru ses from electron cryo-microscopy.C u rr.O pin.Stru ct.B iol.21,265–273.

H ercyk,N .,H orikami,S.M .,and M oyer,S.A .(1988 ).The vesicu lar stomatitis viru s L protein possesses the mRN A methyltransferase activities.V irology 163,222–225.

H u nt,D .M .,Smith,E.F.,and B u ckley,D .W .(1984).A berrant polyad enylation by a vesicu lar stomatitis viru s mu tantis d u e to an altered L protein.J.V irol.52,515–521.

Iverson,L .E.,and Rose,J.K.(1981).L ocalized attenu ation and d iscontinu ou s synthesis d u ring vesicu larstomatitis viru s transcription.C ell23,477–484.

Jones,T.A .,Zou ,J.Y .,C owan,S.W .,and Kjeld gaard ,M .(1991).Improved method s for bu ild ing protein mod els in electron d ensity maps and the location of errors in these mod els.A ctaC rystallogr. A 47 ,110–119.

Kru mm,S.A .,Y an,D .,H ovingh,E.S.,Evers,T.J.,Enkirch,T.,Red d y,G.P .,Su n,A .,Saind ane,M .T., A rrend ale,R.F.,P ainter,G.,etal.(2014).A n orally available,small-molecu le polymerase inhibitor shows efficacy againstalethalmorbilliviru s infection in alarge animalmod el.Sci.Transl.M ed .6, 232ra252.

L aFerla,F.M .,and P elu so,R.W .(1989).The 1:1 N -N S protein complex of vesicu larstomatitis viru s is essentialforefficientgenome replication.J.V irol.63,3852–3857.

L i,J.,Rahmeh,A .,M orelli,M .,and W helan,S.P .(2008 ).A conserved motif in region v of the large polymerase proteins of nonsegmented negative-sense RN A viru ses that is essentialfor mRN A capping.J.V irol.82,775–784.

L i,J.,Rahmeh,A .,B ru sic,V .,and W helan,S.P .(2009).O pposingeffects of inhibitingcapad d ition and capmethylation on polyad enylation d u ringvesicu larstomatitis viru s mRN A synthesis.J.V irol. 83,1930–1940.

L iu zzi,M .,M ason,S.W .,C artier,M .,L awetz,C .,M cC ollu m,R.S.,D ansereau ,N .,B olger,G., L apeyre,N .,Gau d ette,Y .,L agace´ ,L .,et al.(2005).Inhibitors of respiratory syncytialviru s replication targetcotranscriptionalmRN A gu anylylation by viralRN A -d epend entRN A polymerase. J.V irol.79,13105–13115.

L u ,X .,M cD onald ,S.M .,Tortorici,M .A .,Tao,Y .J.,V asqu ez-D elC arpio,R.,N ibert,M .L .,P atton, J.T.,and H arrison,S.C .(2008 ).M echanism forcoord inated RN A packagingand genome replication by rotaviru s polymerase V P 1.Stru ctu re 16,1678–1688.

L yu mkis,D .,B rilot,A .F.,Theobald ,D .L .,and Grigorieff,N .(2013).L ikelihood based classification of cryo-EM images u singFREA L IGN .J.Stru ct.B iol.183,377–388. M ind ell,J.A .,and Grigorieff,N .(2003).A ccu rate d etermination of locald efocu s and specimen tiltin electron microscopy.J.Stru ct.B iol.142,334–347.

M orin,B .,and W helan,S.P .(2014).Sensitivity of the polymerase of vesicu larstomatitis viru s to 20 su bstitu tions in the template and nu cleotid e triphosphate d u ring initiation and elongation.J.B iol. C hem.289,9961–9969.

M orin,B .,Rahmeh,A .A .,and W helan,S.P .(2012).M echanism of RN A synthesis initiation by the vesicu larstomatitis viru s polymerase.EM B O J.31,1320–1329.

O gino,T.,and B anerjee,A .K.(2007 ).Unconventionalmechanism of M rna capping by the RN A - d epend entRN A polymerase of vesicu larstomatitis viru s.M ol.C ell25,85–97.

P atton,J.T.,D avis,N .L .,and W ertz,G.W .(1984).N protein alone satisfies the requ irementfor protein synthesis d u ringRN A replication of vesicu larstomatitisviru s.J.V irol.49,303–309.

P elu so,R.W .,and M oyer,S.A .(1983).Initiation and replication of vesicu larstomatitis viru s genome RN A in acell-free system.P roc.N atl.A cad .Sci.USA 80,3198–3202.

P flu g,A .,Gu illigay,D .,Reich,S.,and C u sack,S.(2014).Stru ctu re of influ enzaA polymerase bou nd to the viralRN A promoter.N atu re 516,355–360.

Rahmeh,A .A .,L i,J.,Kranzu sch,P .J.,and W helan,S.P .(2009).Ribose 20-O methylation of the vesicu larstomatitis viru s mRN A cappreced es and facilitates su bsequ entgu anine-N -7 methylation by the large polymerase protein.J.V irol.83,11043–11050.

Claims

C L A IM S 1.A method of id entifyingan inhibitorof anon-segmented negative strand (N N S)viru s,the method comprising:

(a)u singahighresolu tion 3D stru ctu re of an N N S viru s L protein to selectacand id ate inhibitorfrom avirtu allibrary of 3D stru ctu res of potentialcand id ates by performingstru ctu re based compu tational mod eling;

(b)obtainingthe cand id ate inhibitor;and

(c)measu ringan activity of L protein when contacted withthe cand id ate inhibitorobtained in step(b), wherein ad ecrease in the activity confirms thatthe cand id ate is an inhibitorof the N N S viru s.

2.A method of id entifyingan inhibitorof anon-segmented negative strand (N N S)viru s polymerase, the method comprising:

(b)obtainingthe cand id ate inhibitor;and

(c)measu ringL protein polymerase activity when contacted withthe cand id ate inhibitorobtained in step (b),wherein a d ecrease in the activity of the polymerase confirms thatthe cand id ate is an inhibitorof the N N S viru s polymerase.

3.A method of id entifyingan inhibitorof an N N S viru s the method comprising:

(a) screening a library of cand id ate molecu les for inhibition of an N N S L protein in an in vitro biochemicalassay foran L protein activity;

(b)u sing ahigh resolu tion 3D stru ctu re of the L protein to d etermine the bind ing site of potential cand id ates obtained in step(a)by compu termod eling;

(c)id entifyingapotentialrefinementto the cand id ate inhibitor;

(d )obtainingacand id ate inhibitormolecu le comprisingthe refinementid entified in step(c);and (e)measu ring the activity of the L protein when contacted with the cand id ate inhibitorobtained in step(d ),wherein ad ecrease in the activity of the L protein confirms thatthe cand id ate is an inhibitor of the N N S viru s.

4.A method of id entifyingan inhibitorof L protein of an N N S viru s,the method comprising:

(a)u singahighresolu tion 3D stru ctu re of the L protein of vesicu larstomatitis viru s (V SV )to create an atomic mod elof the L protein of a second N N S viru s by homology mod eling the amino acid sequ ence of the L protein of the second N N S viru s onto the highresolu tion 3D stru ctu re of the V SV L protein ; (b)u singthe atomic mod elof the second N N S viru s L protein to selectacand id ate inhibitorfrom a virtu allibrary of 3D stru ctu res of potentialcand id ates by performing stru ctu re based compu tational mod eling;

(c)obtainingthe cand id ate inhibitor;and

(d )measu ring an activity of L protein of the second N N S viru s when contacted with the cand id ate inhibitor obtained in step (b),wherein a d ecrease in the activity confirms thatthe cand id ate is an inhibitorof the second N N S viru s.

5.The method of any one of claims 1-4 wherein the highresolu tion 3D stru ctu re has aresolu tion of atleast3.8 Å .

6.The method of claim 4 wherein the highresolu tion 3D stru ctu re is an atomic stru ctu re d etermined viacryo-electron microscopy and single particle analysis.

7.The method of any one of claims 1-4 wherein the N N S viru s L protein is vesicu larstomatitis viru s L protein.

8.The method of claim 7 wherein the V SV L protein highresolu tion stru ctu re has space grou pP 1,so as to form a u nit cell of d imensions of a=112.0 Å , b=143.0 Å , c=106.0 Å and α,β,γ=90^o.

9. The method of claim 7 ,wherein the V SV L protein high resolu tion 3D stru ctu re has the coord inates as setou tin the stru ctu re d eposited with the RC SB P rotein D ataB ank with Reference N u mber5A 22.

10.The method of any of claims 1,3 or4 wherein the activity of the L protein measu red is selected from the grou p consisting of RN A d epend ent RN A polymerase activity, GD P polyribonu cleotid yltransferase (P RN Tase) activity,methyltransferase activity,and the ability to stru ctu rally rearrange L protein d omains d u ringL protein fu nctions.

11.The method of claim 2 or claim 10 wherein the RN A d epend entRN A polymerase activity measu red is selected from the grou pconsistingof transcriptinitiation,elongation and termination.

12.The method of claim 1,2 or4 wherein,ratherthan selectingacand id ate inhibitorfrom avirtu al library of 3D stru ctu res of potentialcand id ates,the highresolu tion 3D stru ctu re of the N N S viru s L protein is u sed to d esign apotentialinhibitor.

13.The method of claim 12 wherein the potentialinhibitoris d esigned to interactwithan L protein d omain selected from the grou pconsistingof the RN A d epend entRN A polymerase (Rd Rp)d omain, the mRN A capping d omain,the methyltransferase d omain,the loopprimerd omain,the connector d omain,and the C -terminald omain.

14.A method of mappingthe location of resistance to aknown inhibitorof an N N S viru s L protein on the 3D stru ctu re of the L protein,the method comprising;

(a)homology-based stru ctu ralmappingof the sequ ence comprisingthe resistance mu tation onto the 3.8 Å atomic stru ctu re of an N N S viru s L protein.

15.The method of claim 14 wherein the high high resolu tion 3D stru ctu re is a 3.8 Å resolu tion stru ctu re.

16.A method of inhibitingan N N S viru s,the method comprisingcontactingan N N S viru s oracell infected withan N N S viru s withan inhibitorid entified accord ingto the method of any one of claims 1-4.

17.A n N N S viru s inhibitorid entified u singthe method of any one of claims 1-4.

18.The method of any one of claims 1-4 wherein the id entified inhibitorbind s ad omain of the L protein ata region selected from the grou p consisting of the RN A d epend entRN A polymerase (Rd Rp)d omain,the mRN A cappingd omain,the methyltransferase d omain,the loopprimerd omain, the connectord omain,and the C -terminald omain of the L protein.

19.The method of any one of claims 1-4 wherein the id entified inhibitor attenu ates the RN A d epend entRN A polymerase activity,the GD P polyribonu cleotid yltransferase (P RN Tase)activity,the methyltransferase activity,the loop priming activity,or the ability of the L protein to stru ctu rally rearrange its d omains d u ringL one ormore of its fu nctions.

20.The method of any one of claims 1-4 whichfu rthercomprises the stepof contactingthe confirmed cand id ate inhibitorwithacellinfected withan N N S viru s.

21.The method of any of claims 1-4 wherein the N N S viru s is selected from the viru ses of the Filovirid ae,P aramyxovirid ae,Rhabd ovirid ae and B ornavirid ae families.

22.The method of claim 21 wherein the N N S viru s is aFiloviru s,selected from Ebolaviru s,M arbu rg viru s and C u evaviru s.

23.The method of claim 21 wherein the N N S viru s is aP aramyxoviru s selected from measles viru s, N ipahviru s,H end raviru s,respiratory syncytialviru s and N ewcastle d isease viru s.

24.The method of claim 21 wherein the N N S viru s is a Rhabd oviru s selected from L yssaviru s, V esicu loviru s,P erhabd oviru s,Sigmaviru s,Ephemeroviru s,Tibroviru s,Tu paviru s and Spriviviru s.

25.A method of d esigningacand id ate inhibitorof an N N S viru s,the method comprising: (a) id entifyingan N N S viru s mu tantthatescapes the activity of aknown inhibitorof the L protein of an thatviru s;

(b) id entifyingthe site of the L protein resistance mu tation by sequ ence analysis;

(c)u singhomology-based stru ctu ralmappingof the sequ ence comprisingthe resistance mu tation to mapthe mu tation site onto ahighresolu tion 3D stru ctu re of an N N S viru s L protein;

(d )u sing compu termod eling with the high resolu tion 3D stru ctu re to selecting acand id ate bind ing molecu le thattargets the site of the resistance mu tation mapped in step(c);

(e)obtainingthe cand id ate inhibitormolecu le id entified in step(d ).

26.The method of claim 25fu rthercomprisingthe stepof contactingL protein of the N N S viru s with the cand id ate inhibitor in an in vitro assay of L protein activity,wherein a d ecrease in L protein activity ind icates the cand id ate inhibitoris an N N S viru s inhibitor.

27.The method of claim 25 fu rthercomprisingthe stepof contactingthe cand id ate inhibitorwiththe N N S viru s and assaying viralfu nction,wherein ad ecrease in viralfu nction ind icates the cand id ate inhibitoris an N N S viru s inhibitor.

28.A method of d esigningacand id ate inhibitorof an N N S viru s,the method comprising:

(a) id entifyingan N N S viru s mu tantthatescapes the activity of aknown inhibitorof the L protein of an thatviru s;

(d )id entifyingapotentialrefinementto the known inhibitorthatpermits interaction withthe mu tantL protein;

(e)obtainingthe inhibitormolecu le comprisingthe refinementid entified in step(d );and

(f)assayingmu tantviru s replication orinfection in the presence and absence of the refined inhibitor molecu le,wherein a d ecrease in replication or infection ind icates the refined inhibitor is effective againstthe mu tantviru s.

29.Use of an N N S viru s inhibitorid entified accord ingto the method s of any one of claims 1-4,25 and 28 forthe treatmentof an ind ivid u alinfected withan N N S viru s.

30.The u se of claim 29 in whichthe N N S viru s is selected from the grou pconsistingof viru ses of the Filovirid ae,P aramyxovirid ae,Rhabd ovirid ae and B ornavirid ae families.

31.A compu ter-read able med iu m with high resolu tion 3D stru ctu re coord inate information for an N N S L protein stored thereu pon.