Interference among defective interfering particles of vesicular stomatitis virus.

0022-538X/82/010210-12$02.00/0

Interference

Among

Defective

Interfering

Particles

of

Vesicular Stomatitis

Virus

DONALDD.RAOt ANDALICE S. HUANG*

Departmentof Microbiology and MolecularGenetics,HarvardMedical School, and DivisionofInfectious

Diseases, Children'sHospitalMedicalCenter, Boston, Massachusetts 02115

Received 11 May1981/Accepted 3September1981

Three defectiveinterfering(DI)particles of vesicular stomatitis virus

(VSV),

all

derivedfrom thesameparentalstandard San Juan strain(Indianaserotype),were

used invarious combinations toinfect cells togetherwith theparentalvirus.The

replication of their RNA genomes in the presence of othercompetinggenomes

wasdescribed by the hierarchical sequence: DI 0.52particles>DI 0.45particles DI-T particles> standardVSV.TheadvantageofoneDIparticleoveranother was not due simply to multiplicity effects nor tothe irreversible occupation of

linmitedcellular sites. Interference, however, did correlate with a changein the

ratio of plus and minus RNAtemplatesthataccumulated

intracellularly

andwith the presence ofnewsequencesatthe 3' end of the DIgenomes. DI 0.52 particles containedsignificantlymorenucleotidesatthe 3' end thatwerecomplementaryto those at the 5' end of its RNA than did DI-Tor DI 0.45 particles. The first 45 nucleotides at the 3' ends of allofthe DI RNAswereidentical. VSV and its DI

particlescan be separated into three

classes,

depending

on their terminal RNA

sequences. These sequences suggesttwomechanisms,onebasedontheaffinityof

polymerase binding and the otheron theaffinity ofN-protein binding, that may

account for interference by DI particles against standard VSV and among DI

particlesthemselves.

When viruses grow to high titer, deletion mutants arise. They become enriched in the

population, especially ifthey have thecapacity

tointerferewith the growth of the standard

wild-type parent (12,14, 45). To study the molecular

basis of thisinterference, we have used

vesicu-larstomatitis virus (VSV) because of its

asym-metric bullet shape, which permits the

separa-tion of shorter deletion mutants from the parental wild type(15). Thisseparation has led tothecharacterization of individual size classes

of defective interfering (DI) particles and the

reconstruction of experiments with known amounts of standard and DIparticles.

Theprimarycompetition between DI particles andstandardVSV occurs intracellularly during the step of RNA replication or genome amplifi-cation (17,

29,

30). Very little is known about this mechanism, because RNA replication is initiated early during infection, and all of the proteinsthat are involved have not been identi-fied. Itisknown, however, that input

multiplic-itiesaffect this interference, that the small size

of theinterferingRNAalone cannotaccount for its ability to compete during replication, and that the derivation ofaDIparticle inrelation to the

tPresentaddress:Department of Microbiology,

Washing-tonUniversity, St. Louis, MO 63110.

strain ofstandardhelper virus determines in part the degree of competition and the subsequent ability of the DI RNAtobereplicated(14).

Continuous undiluted passages of the same VSV strain may result in the appearance of several different size classes of DI particles, withsome orone-sizeparticleseventually domi-nating the viral population (11, 21, 31, 37). These results show that interference occurs not only between DI particles and standard virus, but also among different DI particles. Byexamining DI particles,all derived from the sameparental standardvirus,itis suggested herethat, besides

RNA replication, there is another rate-limiting step during virusmaturation when interference

occurs. We found that the ability to interfere increased with the extent of complementary sequences foundat the termini of the RNA.

MATERIALS AND METHODS

Materials. The source of all materials has been

recently reported (35) except for thefollowing.

Cyti-dine 3',5'-bisphosphate and [,y-32P]ATP were

pur-chased from New England Nuclear Corp., Boston,

Mass. Actinomycin D was a kind gift from Merck

Sharp& Dohme, Rahway, N.J. Practical-grade

acryla-mide andelectrophoretic-grade

N,N'-methylene-bisa-crylamide were purchased from Eastman Kodak

Chemical Co., Rochester, N.Y. Aminobenzyloxy-methyl (ABM)-cellulose paper was purchased from

210

on November 10, 2019 by guest

http://jvi.asm.org/

Schleicher & Schuell Co., Keene, N.H.

Polynucleo-tide kinase andRNAligase fromT4-infected

Escheri-chia coliwerepurchased fromP-LBiochemicals,Inc.,

Milwaukee, Wisc. Calfintestine alkaline phosphatase

waspurchased from Boehringer Mannheim

Biochemi-cals, Indianapolis, Ind.RNasesTl,T2, A, andU2were

purchasedfrom Calbiochem-BoehringCorp., San

Die-go, Calif. RNase physarum Mwas akindgift from

Helen Donis-Keller.

Cells andviruses. BHK, CHO, andmouseL cells

were grown assuspensioncultures in Joklik modified

Eagle minimal essential medium supplemented with

5% bovine serum and nonessential amino acids.

Cloned andsucrosegradient-purified standardVSVof

theIndianaserotype(SanJuanstrain) and three size

classes of DIparticles derived from thissamestandard

strain were used. According to a recent agreement

(36), the threeDIparticlesaredesignated: VSIts+ SJ

DI-T/DI 0.33 (5') (15,19);VSIts+SJDI0.52 (5') (19,

34); and VSI ts+ SJDILT/DI0.45 (5') (5). DI 0.45was

kindly supplied by SueEmerson(University of

Virgin-ia,Charlottesville). All DI particles werepurified by

tworounds ofsucrosegradient centrifugation. Their

effectivemultiplicitywasdeterminedasdescribed by

Huangand Manders (17).

Infection ofceUsandextraction of cytoplasmicRNA.

Infection of4 x 106 cellsat34°C with the indicated

multiplicities for each experimentwasdone essentially

as described (9, 35). When RNAwas examined,the

cellsweretreated with actinomycinDat5 p.g/mlatthe

beginning of infection and0.1mCi of 32ppermlatthe

indicatedtimes. Cytoplasmicextractsweremade with

the nonionic detergent Nonidet P-40 (17).

Phenol-chloroform extractions of the RNA from the

cyto-plasmweredoneasdescribed previously(9,35).

Separation of RNA. For the separation of RNA

species ranging from 30 to 12,000 nucleotides, two

different gel systems (10%opolyacrylamide and1.5%

agarose)wereused as previously described(35). To

provide quantitative comparisons ofthe amounts of

RNA, either the autoradiogram was scanned orthe

bandswereexcised from the gels and theradioactivity

was counted. Conclusions basedontheincorporated

radioactivity(countsperminute)weresimilartothose

made whenmolaramountsof RNAwerecalculated.

Elution of RNA from 10%polyacrylamdegels. After

electrophoresis of RNAon apolyacrylamidegel, the

gel was covered with Saran Wrap and

autoradio-graphed. Using the autoradiogramas atemplate, the

region of thegel containing labeled RNAwasexcised

and slicedintopieces of 2mm3.Smallgel cubeswere

thentransferredtoasiliconized tubetogether with2

volumes ofelution buffer (0.4 M sodiumacetate,0.01

M Tris, pH 7.8, and 0.2% sodium dodecyl sulfate)

containing 50,ug ofyeastcarrier RNAperml.Elution

proceededovernightat22°C with intermittent mixing. The small gel pieces were rinsed with 1 additional

volumeof elution buffer. Thecombinedeluateswere

then filtered through a 0.45-p.m filter to retain any

remainingparticulatepolyacrylamide. The RNAwas

thenprecipitatedby2volumesofethanol.

Transfer ofRNAfrom agarosegels toDBMpaper

andhybridization to32P-labeledRNA.The procedure

of Alwine et al. (1) was used with the following

modifications. Potassium phosphate buffer, pH 6.5,

was used instead of sodium borate buffer, pH 8.0.

Unlabeled VSV-specificRNAextracted from virions

andfrom thecytoplasm was firstseparatedon a1.5%

agarose gel. After urea was diffused out of the gel, the

gel wasplacedin200ml of 50 mMNaOH and rocked

gently for 40 min at roomtemperature. The gel was

thentreated twice with 200 ml of 200 mMpotassium phosphate buffer,pH6.5, for 5 min at22°C.

Conver-sion of the ABM paper to diazobenzyloxymethyl

(DBM) paper, subsequent transferof RNA from the

gel to DBM paper, and hybridization conditions all

followedthe

procedures

describedbyAlwineetal.(1),

exceptthat3P-labeledsmallRNA from cells

coinfect-ed withDI-T andstandard VSV was thehybridization

probe ratherthan 32P-labeled DNA. After the

unhy-bridizedlabeledRNA wasthoroughlywashed off, the

DBMpaper was placed inSeal-N-Saveboilable bags

(SearsRoebuck and Co.) and exposedto X-rayfilm

with anintensifying screenat-70°C.

Extractionandpurificationof DIparticleRNA.

Puri-fied DI particles were diluted two- tothree-foldwith

Earle saline and then pelleted by centrifugation at

89,300 x g for90 min. The pelleted particles were

suspendedinETS buffer (0.01 M EDTA, 0.01 M Tris,

pH 7.4, and0.2%sodiumdodecyl sulfate)and

repeat-edly extracted with phenol-chloroform. Virion RNA

wasprecipitatedwithethanol.The RNAresuspended

in ETS bufferwas loaded onto a 15 to 30o (wt/wt)

sucrosegradientmadeinNETS buffer (0.15 MNaCIin

ETS buffer).Centrifugationwasfor 17 h at 89,300xg.

Thegradient fractions werecollected andmonitored

by UV absorbance. Full-length DI-particle genomic

RNA wasthenpooledandethanolprecipitated.

Radioactive

labeling

of theterminiofthe RNA of DI

partices.

Purified RNAs from DI particles were

la-beled at the 3'endbyligationof[32P]cytidine

3',5'bis-phosphate accordingtoEnglandandUhlenbeck(6) as

modified by Schubert et al. (40). Afterlabeling, the

RNA was again purified through sucrose gradients.

For labeling atthe 5' end, purifiedRNAs were first

suspendedin 10 mMTris,pH 8.4, boiledfor1.5min,

and thenrapidlychilled to4°C. Calf intestine alkaline

phosphatasewasaddedto aratio of16-3Uper Fg of

RNA. After 30 min at34°C,the RNAwasextracted

with phenol-chloroform repeatedly and thenethanol precipitated. The RNA was suspended in 20 Ill of

buffer consisting of 50 mM Tris, pH 8.4, 10 mM

MgCi2,and 5 mMdithiothreitol. [y-32P]ATP at 100,uCi

was added, and the whole sample was lyophilized.

Then 20,ulof steriledouble-distilledwatercontaining

10 Uofpolynucleotidekinasewasaddedfor 30 minat

370C. RNA was again extractedwith

phenol-chloro-formrepeatedlyand thenethanolprecipitated. solationof

self-complementary

terminal sequences.

End-labeledRNAsat1to4,ug from DIparticleswere

mixed with 50 ,ug ofyeastcarrier RNA in 25 p.l of

distilled water. Equal volumes of 2x hybridization

buffer(1x = 0.4 MNaCl, 0.01 MTris, pH 7.4,and

0.01 MEDTA)wereadded,and then the RNAsample

wassealed inacapillarytube andincubatedat70°Cfor

2 h.Sampleswerepipettedinto 200p.lof

hybridization

buffer containing, per milliliter, RNase T. at 50 U,

RNaseT2at5U,and RNase Aat100 ,ug. Digestion

proceededat37°Cfor 30min. Afterdigestion,

samples

wereextractedrepeatedly with

phenol-cjlloroform

in thepresenceof addedcarrierRNA(50to100

p.g)

and precipitatedwithethanol.Todenature

double-strand-VOL. 41,1982

on November 10, 2019 by guest

http://jvi.asm.org/

ed RNA, the RNAwassuspendedin 10

7.4,boiled for 1.5min, andrapidlychill4

RESULTS

Interference among different D Three different DIparticles,allisola

San Juan strain ofVSV, were use(

combinations together with standa

infect cells in orderto testwhether

the DIparticles was particularly mc

than the others during interferenc4 measuredbycomparingthe amount

lular RNAspecifictoeach DIparti4 As shown by the absence of stai

specific 40S RNA, each of the DI ;

0.45, DI 0.52, and DI-T) was capa pletely inhibiting the replication

VSV without affecting primary tran

the L,G, N,NS, andMmRNA's(Fi

f, and g). These resultsmeasuringRI ed well with inhibition of infecti identical combinations were used

forming abilityof standardVSV was three DI particles, singly or in c

a

b

c

d

e f

g

2.:r

um-L

FIG. 1. Interference amongDIparti

BHK cells in five cultures were infec

indicated virusesatmultiplicitiesof 20(

addedtotheculturesat1to6 hpostini

plasmic RNAwasthenextracted, and e

were separated on a1.5%agarose gel.

VSV plusDI 0.45 particles; (b)standard

0.45plus DI 0.52particles;(c) standard

0.45plus DI-T particles;(d) standard VSI

plusDI0.52 plusDI-Tparticles; (e) stand

DI0.52plusDI-T particles; (f) standard

0.52particles; (g) standardVSV plus Dl

mMTris, pH completely inhibited the growth ofVSV (D. D. ed onice. Rao and A. S. Huang, unpublished data).

Radio-active material between theoriginand40S RNA

probably represents contaminating DNA and

1I particles. some trapped RNA. Agarose gel patterns of

tedfromthe RNA obtained from cells infected with only d in various standard VSV have beenpublished (19, 34,35).

ird VSV to When cells weretriply infected with anytwo ranyone of of the DIparticlesandstandard VSV,40S RNA reeffective synthesis was still inhibited, but there were

e. This was varying degrees ofreplication of thegenomes of

tofintracel- DIparticles. DI0.45 and DI-Tparticles replicat-cle. ed about equally well in each other's presence

ndard virus- (Fig. 1, slot c), whereas DI 0.52particles hadan

particles

(DI advantage over the two smaller DI particles. Lble of com- When theindividual bandswereexcised and the

of standard radioactivitywascompared by Cerenkov

count-iscription of ing, DI 0.52 particles synthesized between 1.2

ig. 1, slotsa, and 2.5 times the molar amount of plus and

NA correlat- minus strands synthesized by the smaller DI

ivity. When particles. The advantage increased with time

and plaque- (Fig. 1, slots b and e). When all three DI assayed, all particles were added to cells in the presence of

,ombination, standardvirus, therewasanoverall reductionin

the totalamountof RNAreplication for each of the genomes, but DI 0.52stillappearedto

repli-cate its RNA somewhat better than the other

twoDI particles (Fig. 1, slot d).

- O These results indicate that all three DI parti-cles interfered effectively against the parental standard VSV. Among the DI particles, the -40£)S

largest

one

appeared

tohavea

replicative

advan-tageoverthetwosmallerones.Inonlyonecycle

- L of growth, interference against standard virus

was complete, whereas interference among DI

-DI 0.52 particles was onlypartial.

-DI0.4

Symmetry

of

replication.

During

standard

VSV genome replication, plus and minus RNA templates accumulate asymmetrically, with the bulk of the RNA being of the minus polarity(27,

- DI-T 44). It has been postulated thatdifferences

be-G tween plus and minus strands in their binding

affinities forpolymerases may account for this - N

asymmetry

and that if

binding

affinities of DI

templates

are

altered,

the differential

accumula-N S tion of the plus and minus strands may be +

changed

as well

(16,

33).

Because the

plus

and

M minusstrands of RNA specific to thereplication of VSVDIparticleshavebeen recently separat-icles ofvsv. ed on gels (19), the two templates were

mea-:ted with the sured directly.

each. 32pwas To test the above hypothesis, the doublet

fection. Cyto- bands of DI RNAs from coinfected cells were

,qual

Portions

quantitated, and the ratio of plus strands

(bot-VSV

plustDI

tom

band)

to minus strands

(top

band)

was

VSV plus DI determined

(Table 1).

DI 0.52

particles

synthe-V plus DI0.45 sized

virtually

equivalent

amounts of

plus

and

lardVSVplus minus

strands,

whereas thetwosmallerDI

parti-VSVplus DI cles made plus and minus RNA in the ratio of

I-Tparticles. 40:60.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:3.504.61.250.343.575.2]

[image:4.504.56.246.97.158.2]

MECHANISMS INTERFERENCE

TABLE 1. Ratio of intracellular plus and minus

RNAspecific to DI particles'

DIparticles

Plusstrand Minusstrand

DI0.52 49±3 51±2

DI 0.45 40±1 60±1

DI-T 40±1 60±1

aBHK

cells

were coinfected with standard VSV

and oneof the three DI particles, each at amultiplicity

of20, andexposed to32pfrom 0 to 6 h. Then the RNA

was extracted, separated on an agarose gel, and

ex-posedtoX-rayfilm. The bands representing

intracellu-lar RNAof the size of the genome of each DI particle

werequantitated by scanning thefilm.Theassumption

was made thatthefaster-migratingband is plus

strand-ed and theslower-migrating band is minus stranded

(19).

Thesedirect measurements confirm previous results for DI-T particles obtained by RNA hybridization studies (18). This is in contrastto

the ratio of20:80 for the intracellular plus and minusRNAstrands of standard VSV (27). These resultsindicate that as the interfering capability of the DIparticles increased, there appeared to be a moreequal accumulation of plus and minus RNAstrands.

Interference by DI 0.52 particles against DI-T

particles and standard VSV in mouse L, CHO,

and BHKcells. Because DI-Tand DI 0.45

parti-cles appeared to be nearly equivalent in their

interfering potential and because DI 0.45

parti-cles have been studied elsewhere (40, 41), we chose tofocuson DI 0.52and DI-Tparticles for

detailed comparative studies. Initial attempts

were to select a cell line which showed the greatestdegree ofinterference. Three different cell lineswereeachtriply infectedwith standard

VSVand bothDIparticles,all atequal

multiplic-ities. When equal portions of the RNA from

each celltypewere separated on1.5% agarose

gels, RNA specific to DI 0.52 particles was

alwaysthepredominant

species

(Fig. 2).RNAof

thesize ofstandard virus genomes,migratingat

40S, was undetectable, but some RNA

specific

toDI-Tparticleswasseen(Fig. 2).Theseresults

indicate that all three cell linessupported

inter-ference by DI 0.52 particles

against

standard

VSV and DI-T particles. In

addition,

as was

seeninFig. 1, interference among the DI

parti-cles was

quantitatively

different from

interfer-ence withstandard virusbyDI

particles.

Parenthetically,the totalamountof32p

incor-poratedinto the RNA from

equivalent

numbers

ofcellswashighestfor BHK cells. This reflects

thedifferent abilities of thesecelltypesto incor-porate

32p

into total

virus-specific

RNA,

even

thoughthe

production

ofinfectiousprogenywas

equivalentfor each of thesethreecell lines

(G.

J.

Freeman, Ph.D. thesis, Harvard University, Cambridge, Mass., A79). Because of theability ofsuspended BHKcellsto incorporate

32p

into VSV RNA, these cells were chosen for all subsequent experiments.

Relative accumulation of RNA species during interference between DI 0.52 and DI-Tparticles. Tomeasure the rate of inhibition of synthesis of RNA at hourly intervals after infection,

32p-labeled RNAfrom cells coinfected with standard VSV and each orboth of the DI particles was separated and quantitated. Figure 3 shows the accumulation of the intracellular RNAs from each of the three sets of infected cells. The accumulation of total plus andminus strands of DIRNAs was compared with that of the mRNA for the N protein from the same cells. When

cellswere coinfected with either one of the DI

particles and standard VSV, DI RNA replication occurred to about the same extent (Fig. 3a,b). When cells were triply infected, total RNA synthesis was reduced to approximately 70% of that in cells coinfected with only one type of DI

particle and standard VSV (Fig. 3c). In these

c

040S

-

4K-DI-0.52

-DI-T

-

G-

N-.4.

NS

M

FIG. 2. Competition between DI-T and DI 0.52

particlesindifferent cell lines.Mouse L(a),CHO(b),

and BHK (c) cells were each triply infected with

standardVSV, DI0.52particles, and DI-Tparticles,

eachat amultiplicityof 20.32Pwasadded from 1to6 h

after infection. Equal amounts of the cytoplasmic

RNAobtained from these infected cellswere

separat-ed on1.5%agarosegels.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:4.504.298.422.335.602.2]

214 RAO AND HUANG

re)

0 0

0

HOURS

POST

INFECTION

FIG. 3. Accumulationofintracellular RNAspeciesincells coinfected with standard VSV and DIparticles.

BHKcells werecoinfectedwithstandard VSV and either DI 0.52orDI-Tparticles.Another culture of cells was

triplyinfected with standard VSV and both DIparticles.Allmultiplicitieswereat20 each.32Pwasadded at the

time of infection together with actinomycin D. At hourly intervals, equal portions were removed, and

cytoplasmicRNA wasextracted andseparatedon1.5%agarosegels.Each of theVSV-specificRNA-containing

bands wasexcised,and theradioactivitywasquantitated byCerenkovcounting. (a)Standard VSV and DI 0.52

particles; (b)standard VSV and DI-Tparticles; (c)standardVSV and both DIparticles. Symbols: (@) plus-and

minus-strand RNA of DI 0.52particles; (0) plus-and minus-strand RNA of DI-Tparticles; (x)mRNA for the

nucleocapsid protein.

triply infected cells, DI-T particles replicated

less well than DI 0.52particlesand atonly 25%

theamountwhen DI 0.52particles were absent

(Fig. 3c).

Interference among the DI particles was de-tectedasearlyas2to3h afterinfection,withthe

difference in accumulation between the two

RNA genomes increasing as the infection

pro-gressed. Thisgradualaccumulation ofgenomes of DI-Tparticles, in the absence ofany detect-able synthesis ofstandard virusgenomes, again

suggests that the mechanism of interference

betweenthetwoDIparticleswasdifferentfrom thatbetweenDIparticles andstandardVSV.

Effectsofmultiplicity oninterference between the two DI partides. Previous studies showed thatinput multiplicitiesof standard virus and DI

particlescouldaffect the degree of interference, thereby giving an erroneous result concerning theabilitytointerfere byaparticularDIparticle (20,43). Toensurethat thepartial inhibition by DI0.52particles against DI-T particles wasnot

due simply to multiplicity, cells were triply infected with DI 0.52 particles together with standardVSV,eachataconstantmultiplicity of 20, and then with DI-T particlesover afivefold

range. As a measure ofinterference, the total

radioactivityinplus and minus RNA of thetwo

DIparticles wascompared.

In these triply infected cells, RNA of DI-T particles accumulated to the same reduced

amountdespite increased multiplicities of DI-T particles (Fig. 4). The lack of enrichment of RNA of DI-T particles at higher multiplicities indicated that the inhibitory effect of DI 0.52 particlesonDI-Tparticleswasmultiplicity

inde-pendent.

On the other hand, at the higher total multi-plicities, RNA of DI 0.52 particles was de-creased(Fig. 4). This reverse effect of increased

multiplicity of DI-Tparticles on the interfering

DI 0.52 particles relates to an expected

multi-plicity-dependent event where the competition

forproteins involved in RNA replicationis equal

between the two incoming DI genomes.

Effectofcycloheximide on the initial

competi-tion between DI-T andDI 0.52 particles.

Previ-ously, itwasfound that an advantage was recov-ered by DI-T particles over DI 0.52 particles if cellswereinfected with DI-T particles and stan-dard VSV for a period before superinfection with DI 0.52 particles (19). To determine

wheth-er this superinfection exclusion was due to the

ability ofinput DI-T particles to bind

on November 10, 2019 by guest

http://jvi.asm.org/

[image:5.504.97.412.71.295.2]

4

0--

0-20:100 20:50 2.20

multiplicity DI-0.52 DI-T

FIG. 4. Effects of relative multiplicities on the

competition between DI-T and DI 0.52 particles.

Threecultures of BHKcellswereinfected with

stan-dardVSV and DI 0.52 particles, eachat amultiplicity

of 20, and with multiplicities of DI-T particles at20,

50, and 100. 32P was present from 1 to 6 h

post-infection. Cytoplasmic RNA was extracted and

sepa-rated on1.5% agarose gels. Bands containing

subgeno-mic-sized DI RNA, both plus and minus strands, were

excised, and the radioactivity was quantitated by

Cerenkov counting. Symbols: (d) RNA of DI 0.52 particles; (0) RNA of DI-T particles.

of DIparticlestocompeteagainst standard VSV (13, 23). It therefore becomes important to

ex-aminethesequences attheRNAterminiofboth DI genomes inorder todetermine whether any

differences could account for the ability of the

largerDIparticletoinhibit the smaller DI

parti-cle.Hybridization ofthe RNAgenomesin

solu-tiongave9%oself-annealing for both DI0.52and DI-Tparticles (19). Thissuggeststhatthelarger

genome mayhave moreself-complementary se-quences than the smaller genome. Oligonucleo-tide analysis indicates thatboth arederivedfrom the Lgeneandcontainsequencesspecifictothe 5' end of the VSV RNA (10, 19).

Totestforsequencedifferences atthe 3' end ofthe twoDI RNAs,wetook advantage of the

presence of a small RNA, 46 nucleotides in length,which is synthesized in largeamountsin cellscoinfected with standard VSV andDI

parti-a

bcde

ibly to limited resources or sites during the

period before superinfection, cycloheximide

was added during the period of superinfection. This prevents genome amplification without in-hibiting attachment and uncoating (17, 24). Cells coinfected with standard VSV and DI-T

parti-cles were treated or not with cycloheximide for 0.5 or 1 h and then superinfected with DI 0.52

particles, followedby the removal of

cyclohexi-mide0.5 h later.

Figure 5a shows the usual predominance of

thelarger DI RNA when cells were triply infect-ed at the sametime.When DI-Tparticles had an

advantage of 0.5 or 1 h, there was a relative

decrease in theabilityof DI 0.52 toreplicate its

RNA (Fig. Sb,d). However, if cycloheximide

were present to prevent the amplification of

inputDI-Tgenomes, then the RNA of DI 0.52

was replicated to advantage over DI-T RNA

(Fig.

5c,e).

These results indicate that input DI-T parti-cles didnotirreversiblytieup limitedresources

of thecell, butthat superinfectionexclusion of

DI 0.52particlesrequired the initial increase in DI-Ttemplates. The end result ofpriorinfection by one DI particle was an increase in their

effectivemultiplicity.Therefore, prior infection, without RNAreplication,didnotirreversiblytie

up limitedresources oroccupy sites which con-trol RNAreplication.

Hybridization of small RNA to RNA from standardVSVand DIparticlesand from infected cells.Sequencedifferencesatthe3'end of RNA

genomesarepostulatedtoaccountfor theability

LJ

!

-DI-0.52

is

-DI-T -

= ::G2

u'-

N

NS

M

FIG. 5. Reversal by cycloheximide of the competi-tive advantage of DI particles due to prior infection.

BHK cells at 20 x 106 cells/ml were infected with

standardVSV and DI-Tparticlesat amultiplicityof 20

each andthen divided into fiveequal portions.

Addi-tion of cycloheximide at 100 Fig/ml, superinfection

with DI 0.52 particles at a multiplicity of 20, and

exposureto32pwere asfollows. RNAwasharvested

at6 hafter infection andseparatedon a1.5%agarose

gel. (a)DI0.52particlesat0h,32Pat1to6h;(b)DI

0.52particlesat0.5h, 32Pat1.5to6h;(c) cyclohexi-mideat0to1h,DI0.52particlesat0.5h,32pat1.5to 6 h;(d) DI 0.52particles at1 h, 32P at2to6h; (e)

cycloheximideat0to1.5h,DI0.52particlesat1h, 32p

at2to6h.

- som 4:,::*:*4-a

on November 10, 2019 by guest

http://jvi.asm.org/

[image:6.504.55.214.55.193.2] [image:6.504.272.439.281.538.2]

216

cles(34). Because small RNA is co0

tothe 3' end of DI RNAs(23), itca

determine the identity or diverg quencesatthe 3' endby cross-hyb

To do so, 32P-labeled small Rls pared from cells coinfected with D and standard VSV (34, 35). Sma hybridized against unlabeled virio pared from standard VSV or DI

wellasagainst total unlabeled intra<

from singly or coinfected cells. TI

RNA was first separated on agar

then transferred topaper before hl Small RNA failed to anneal to t strand RNA from standard VSV bound to the negative-strand RNA DI 0.52orDI-Tparticles(Fig.6c,e

coinfectedcellswasrecognizedby

theywere DI genomes ortheircoI plus strand (Fig. 6d,f).

a

be

0-40S

-L

-DI-0.52

9

DI-T

-G

-N

-NS

[image:7.504.69.240.292.562.2]

M

FIG. 6. Northern blot hybridization

toVSV-specificRNAspecies. Virion RI

,ug fromstandard VSVoritsDIparticle

lularRNAat3,Lgfromsinglyorcoinfec

separatedon a1.5%agarosegel, transfi

paper,andhybridizedto32P-labeledsm

(a) RNA fromstandard VSV;(b) RNA

VSV-infected cells; (c) RNA from DI(

(d) RNA fromcells coinfectedwith stan4

DI 0.52 particles; (e) RNA from DI-T

RNAfromcells coinfectedwithstandard

Tparticles.

mplementary Smallamounts ofhydridization occurred

be-an be used to tween small RNA and intracellular 40S RNA,

rence of se- presumably due to the complementary se-)ridization. quences onthelarge plusstrandwhich

comigrat-4A was pre- ed with virion 40S RNA (Fig. 6b). Annealingof

)I-Tparticles small RNA to the L mRNA from standard

VSV-,ll RNA was infected cells suggeststhat some L mRNAwas In RNA pre- not terminated, but contained run-through se-particles, as quences with 3' ends identical to those of the cellular RNA plus strand of 40S RNA (42). In Fig. 6f, the 'he unlabeled minor species whichbound small RNA was due ose gels and toacontaminatingDIparticlein ourDI-T prepa-iybridization. ration.

he negative- These hybridization results show that there

whereas it were sequences in common between the

ge-from either nomes of DI 0.52and DI-Tparticles whichwere

). RNA from not found on the negative-strand RNA of stan-small RNA if dard VSV. Thesesequenceswerepresumablyat

mplementary the 3' end. Therefore, if the two DI genomes

differed in their RNA sequenceatthe 3'

termi-nus,it would bedue to less than 20%divergence within the first 46 nucleotidesorinternaltothese

e nucleotides. Moreover, hybridization of small

RNA to both plus and minus strands of DI RNAs indicated sequence homology at the 3'

endsof theseRNAsandsuggested the presence

ofatleast 46nucleotidesthatwereterminal and

complementary at the 3' and 5' ends of both strands.

Isolation of self-hybridized duplex molecules

from the RNA of DI 0.52 and DI-T particles.

Because theprevious hybridizationresultswere indirect, adirectapproachwas taken to isolate

these terminal sequences by annealing. To do

so, RNAs from purified DI particles were la-beled at either the 5' or 3' end, rapidly

self-annealed, and digested with RNase. The

result-ing

duplex

molecules (stems) were electrophoresedon a10%polyacrylamide gelto

determine theirrelative sizes.

Discrete self-hybridizable pieces of 45, 48, and 55nucleotides wereformed from theRNA

of

DI-T,

DI

0.45,

and DI 0.52

particles,

respec-tively (Fig. 7). These lengths were determined by comparison with tRNA markers and the small RNA as well as from the sequencing results to follow. Because radioactive RNA, labeled in vitroateitherterminus before hybrid-ization and RNase digestion, appeared in the

of smallRNA isolated stems from all three DI particles, the

NAsat0.5to1 particles must be formed by self-annealing of :s and intracel- RNAsequences at the 3' and 5'terminiandnot

ted

cetlls

were

by loop-back or other

internal

complementary

erredtoDBM sequences. These results prove that terminal

from

_O.52

standard complementarity existed in the RNAs of all

particles;

threeDI

particles.

Differences in the stemsizes

dardVSV and correlated with their ability tocompete against r particles; (f) each other.

J VSVandDI- Sequencing ofend-labeled RNA from DI 0.52

and DI-Tparticles.Toensurethat thesequences

on November 10, 2019 by guest

http://jvi.asm.org/

MECHANISMS OF INTERFERENCE 217

a b c d e f incomplete when measured by RNAsynthesis,

tRN A - whereas inhibition of standard virus growth by

any one of the three DI particles resulted in complete inhibitionof standard genome

replica-XC 10- a* tion as well as infectious progeny. Moreover,

_m

inhibitionof one DIparticle by another could be

shown to be multiplicity independent, resulting VW

do

in the same partial degree of RNA replication,

abc de

abcd

e

FIG. 7. Terminal complementary stems obtained

from5'- or 3'-end-labeled DI RNAs. Purified RNAs

from DI 0.52, DI 0.45, and DI-T particles were

termi-nally labeled at either the 5' or 3' terminus. The

labeled RNAs were again purified on sucrose

gradi-ents,self-annealed, and digested with RNases.

RNase-resistant RNAs were then separated on a10%o

poly-acrylamide gel. tRNA and xylene cyanol (XC) dye

markers are indicated by arrows. (a) DI 0.52 RNA,

5'-endlabeled; (b) DI 0.45 RNA,5'-endlabeled;(c) DI-T

RNA, 5'-end labeled; (d) DI 0.52 RNA, 3'-endlabeled;

(e) DI 0.45 RNA,3'-endlabeled;(f)DI-T RNA,3'-end

labeled.

at the termini ofthe RNAs of the DI particles differed only in the extent of complementarity, sequence analysis was done on the 3'-end-la-beled stems obtained from DI 0.52 and

com-pared withDI-Tparticles.The 3' and 5'ends of

theRNA from DI-T and DI 0.45 particles have been sequenced (40).

Figure8shows thepatternof RNA fragments

from DI RNAslabeled at the 3' end obtained by

digestionwith alkali or several different RNases.

The overallpattern was identical for 45

nucleo-tides,indicatingthat theonly difference between

DI 0.52 and DI-Tparticleswas in theextentof

the self-complementarity at the termini. When

wholeRNAfromDI0.52 and DI-Tparticleswas

similarly sequenced, divergence at the 3' end

wasdetected from the first45 nucleotides. The

exact sequences in these first 45 nucleotides agree with previously published sequences for DI-T and DI 0.45particlesexceptfor nucleotide

positions35 and 36(40).Theseresults, however,

were notclearenough above the first45 nucleo-tides topermitfurthersequencedeterminations.

DISCUSSION

These studies relate sequences on the RNA

genomesof VSV and itsDIparticlestopossible mechanisms of interference. Interference among VSVDIparticles themselveswasquantitatively

differentfrom their inhibition of standard VSV

[image:8.504.77.223.81.214.2]

growth. Interference among DI particles was

FIG. 8. Sequence comparisons of 3'-end-labeled

complementaryRNAfrom DI 0.52 and DI-Tparticles.

RNA from the two DI particles were prepared as

described inMaterials andMethods. RNA stems

la-beledatthe 3' endwerethendenaturedby boilingand

quickcoolingbefore partialnucleasedigestion bythe

method ofDonis-Kelleretal.(4). Separationwas ona

20% polyacrylamide gel. (a-e) 3'-end-labeled RNA

from DI 0.52 pa~rticles; (a'-e') 3'-end-labeled RNA

from DI-Tparticles; (a, a')notreatment;(b, b')RNase

T1;(c, c')RNase U2; (d,d')RNase physarumM; (e,

e')alkalinedenatured.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:8.504.257.446.189.572.2]

218 RAO AND HUANG

irrespective of the initial concentration of the suppressed DIparticle.

When interference is multiplicity dependent,

the mechanism is likely to be an early event

involving input virions, whereas

multiplicity-independentinterference suggestsamechanism which occurs after the initial steps of viral multiplication. DIparticleshaveonlytwo

essen-tial functions: genome replication and

matura-tion intovirus particles. One is anearly event;

the other is a late event. The early event is multiplicitydependent,and the lateoneis multi-plicity independent.Therefore, itcanbe postu-lated that a competitive advantage might be gainedduringeachofthese functional steps.

How competitive advantages can be gained

during these steps may befoundinanevaluation ofthe RNA in VSV and in the DI

particles.

DI genomes contain additional complementary

se-quences aswellasthesingle-base

mutations

(8, 10, 40). Because single-base mutations are in

different coding regions for the L

protein,

it is

unlikely that they are involved in interference.

Complementary sequences atthe 3' end

of

the

DI RNAs are suspected to be responsible for

interferenceby DIparticles with the growthof

standard VSV (13, 23).

The ability ofDI particlestointerfere among

themselves correlated with the extent of

addi-tional complementary sequences atthe 3' end.

The greater thecomplementarity, thebetterthe

abilitytointerfere. Similardatawith DIparticles

obtainedfromdifferent strains ofVSV support this

conclusion

(32). This, however is the first time suchstudies have been carried out with DI particles all derived from thesame parent..

Oursequencing results and thosepublished by others on several DIparticles indicate that the first 45 to 47 nucleotides at the 3' terminus of the majority of DI RNAs are identical and comple-mentary to the 5' end (40). Normal asymmetry ofplus-andminus-strand synthesis during VSV RNA replication is thought to be due to the

differentialbinding affinity for polymerase at the

3' ends of standard RNA templates (16, 22, 33).

InFig. 9, these binding sequences are indicated by T and R, with R having the higher affinity for

polymerase.

With symmetrical terminal se-quences, the plus and minus DI RNAs would bothcontain R-like sequences at their 3' ends. Theresults should be equal production of both plus- and minus-strand templates.

Measurement of RNA templates in thisstudy,

however, indicates approximately equal

plus-and minus-strplus-and synthesis for only DI 0.52 particles. The other two DI particles synthesize

minus-strand

RNA in excess over plus-strand RNA(60:40), but not at the same ratio (80:20) as standard VSV (27, 44). These results suggest that the asymmetry of RNA synthesis may be

the result of other factors beside polymerase binding.

If the first 45 to 47 nucleotides at either

terminus ofthegenomeaccounted for

polymer-asebinding, how could the next 12 nucleotides

be related to maturation? By examining pub-lished standard VSV sequences (summarized in

Fig.9),noncodingregionsaccountfor 50

nucle-otidesatthe3' end and 59 nucleotides atthe 5'

end (25, 39). Transcriptional initiation of the N genebeginsatposition51 from the 3' end(7, 26,

38), leaving only a triplet AAA without an

assigned function. However,atthe 5'end,after

termination ofthe coding regionfor the L

pro-tein, there are 12 nucleotidesin addition tothe terminal 47 nucleotides. SinceDI 0.52particles contain sequences at the 3' end of their RNA which are complementary to some of these

additionalnucleotides,thisregionbetween

posi-tions 48 and59,just internalto the terminal47, atthe 5' end may determine interactionsduring

morphogenesis. Becausetheinitial stepof

mor-phogenesisisencapsidationof the RNA with N

protein, the interactioncould be viewed as the

determinant forN-protein binding.

Apartof the sequencewithinthe 12

nucleo-tides, 5' AUCAAA 3', is also found in the

noncoding regionof the N mRNAjust beforethe

first AUGforinitiating translation(5'

m7AmA-CAGUAAUCAAAAUG 3'). Such ahexamer is

notfound in any othernoncodingregion of

VSV-specificRNAs. It is prematuretodecide

wheth-erthis sequenceatthe 5'endsoritscomplement

at the 3' ends determines the interaction be-tween VSV RNA and Nprotein. Nevertheless, if the hexamer and thenucleotides surrounding it affect the affinity of the association between RNA andN protein, then the same arguments

forthe polymeraseconcerning bindingaffinities

andsymmetryof sequences would hold for the

interactions of RNA and the N protein.

Al-thoughthe RNA of DI 0.52 particles has

termi-nalcomplementarityonlyforaportion of the 12

nucleotides,including the hexamer, the

extend-ed symmetrical sequence in its RNA may be enough to account for its ability to compete during encapsidation as well as during replica-tion. Blumberg et al. (B. M. Blumberg, M. Leppert, and D. Kolakofsky, Cell, in press), reasoning from a differenttack, conclude simi-larly that thesequencesresponsible for initiating N-protein binding are internal to the termini. Ourassignment of theinternalposition of these sequences differsfrom theirs; however, regula-tory sequences and binding sequences need not be the same.

Sincethe general understanding of VSV RNA

replication and encapsidation is scant, other hypotheses for interference mechanisms could be entertained. It is possible that the ability of

on November 10, 2019 by guest

http://jvi.asm.org/

59

48

- 3'

WIEAA

]AEGE]ECGAI

L

IGAAACUAGGAAUE

3'

UUGUCAUUAGUUU...

5

AACAGUAAUCAAAAUG...

51

+

5'

LflUUU4

)CUUUGAUCCUEUACI

48 51

T

= TRANSCRIPTIONAL INITIATION SITE

R

=

REPLICATIONAL

INITIATION SITE

N, NS, [M, G,

AND

L

=

VSV

GENES

FIG. 9. Sequences at the termini of VSV RNA postulated to be responsible for RNA synthesis and

encapsidation. Sequencesaretakenfrom Schubertetal. (39), McGeoch and Dolan (25), McGeochetal. (26),and

Rose(38). T, Transcriptionalinitiation site;R,replicationalinitiation site;N,NS, M,G, and L, VSVgenes.

RNAtemplatestocircularize maycontribute in

some waytotheir abilitytobereplicated (28). If preferential replication dependsonthe stability

of circular template RNA, then the degree of

sequencecomplementarityatthetermini would correlate directly with such stability. On the otherhand, wehave made the assumptionthat all VSV particles, standard or defective, have

about the same particle-to-infectivity ratio. If the effective multiplicities for the DI particles

vary greater than threefold from our

assump-tions, thenany hypothesisbasedonthe degree

of sequence complementarity may have to be

revised.

Despitewhatevermechanisms turn out tobe

true,therearethree classes ofcontrolling

termi-nalsequencesontheRNAgenomesof VSVand

its DIparticles. Standardvirus and the DI 0.46

particle derived from the Orsay (HR) strain contain 18 nucleotides at the 3' and 5' termini

whicharecomplementaryexceptforafew base

mismatches(32).Other DIparticlescontain ter-minal RNA sequences which are completely

complementary for 45 to 48 nucleotides, as a secondgroup,and for 55to60nucleotides,asa thirdgroup (40). These three typesofterminal

sequences relate to a hierarchical interference

pattern where the extent of complementarity

results inagreaterabilitytocompeteduringthe

virus growth cycle; These results focus the

control of VSVRNAreplication and encapsida-tion onthese terminal sequences. Direct

com-petitive binding experiments with these se-quencesandtheirregulatoryproteinsareneeded

to understand how interference occurs during nucleicacidreplication and encapsidation.

ACKNOWLEDGMENTS

Thisworkwassupported by Public Health Service research

grantAI 16625 from the National Institutes of Health andby researchgrantMV-54 from the American CancerSociety.

We thankTrudy Lanman forexcellenttechnicalsupport andSuzanne Ress forthecomputer-aided preparationof this manuscript. Helen Donis-Keller, Peter Besmer,and Ian Ken-nedy kindlyprovidedtechnical adviceandmaterials.

LITERATURECITED

1. Alawne,J. C.,D.J. Kemp,andG. R. Stark. 1977. Method

fordetection ofspecificRNA inagarosegels bytransfer

to diazobenzyloxymethyl-paper and hybridization with

DNA probes. Proc. Natl. Acad. Sci. U.S.A.

74:5350-5354.

2. Colonno,R.J.,and A. K.Banedee.1976.AuniqueRNA

species involvedin initiation of vesicular stomatitisvirus

RNAtranscriptionin vitro. Cell 8:197-204.

3. Coonno, R. J., and A. K. Banerjee. 1978. Complete

5'

3'

on November 10, 2019 by guest

http://jvi.asm.org/

[image:10.504.52.458.76.343.2]

220 RAO AND HUANG

nucleotide sequence of the leader RNA synthesized in vitro by vesicular stomatitis virus. Cell 15:93-101. 4. Donis-Keller, H., A. M. Maxam, and W. Gilbert. 1977.

Mapping adenines, guanines and pyrimidines in RNA. NucleicAcidsRes. 4:2527-2538.

5. Emerson, S. U., P. M. Dierks, and J. T. Parsons. 1977. In vitro synthesis of a unique RNA species by a T particlesof vesicular stomatitis virus. J. Virol. 23:708-716. 6. England, T. E., and0. C.Uhlenbeck. 1978. 3-Terminal

labeling of RNA with T4 RNAligase. Nature(London) 275:560-561.

7. Franze-Fernandez, M. T., and A, K. Banerjee. 1978. In vitro RNA transcription by the New Jersey serotype of vesicular stomatitis virus I. Characterization of the mRNA species.J. Virol.26:179-187.

8. Freeman, G. J., D. D. Rao, and A. S. Huang. 1978. Genome organization of vesicularstomatitis virus: map-ping of ts G41 and the defective interfering T particle, p. 261-270. In B. W. J. Mahy and R. D. Barry (ed.), Negative strand viruses and the host cell. Academic Press,Inc., New York.

9. Freeman, G. J., D. D. Rao, and A. S. Huang. 1979. RNA synthesis of vesicularstomatitisvirus. VIII. Oligonucleo-tides of the structural genes and mRNA. Gene 5:141-157. 10. Hagen, F. S., and A. S. Huang. 1981. Comparison of

ribonucleotide sequences from the genome of vesicular stomatitis virus and two of its defective interfering parti-cles. J. Virol. 37:363-371.

11. Holland, J. J., L. P. Villarreal, and M. Breindl. 1976. Factors involved in the generation and replication of rhabdovirus defective T particles. J.Virol. 17:805-815. 12. Huang, A.S. 1973. Defective interfering viruses. Annu.

Rev. Microbiol. 27:101-117.

13. Huang, A. S. 1977. Viral pathogenesis and molecular biology. Bacteriol. Rev. 41:811-821.

14. Huang, A. S., and D. Baltimore. 1977. Defective interfer-ing animal viruses, p.73-116. In H. Fraenkel-Conratand R. R. Wagner (ed.), Comprehensive Virology, vol. 10. Plenum Publishing Corp., New York.

15. Huang, A. S., J. W., Greenawalt, and R. R. Wagner. 1966. Defective T particles of vesicular stomatitis virus. I. Preparation, morphology, and some biologic properties. Virology 30:161-172.

16. Huang, A. S., S. P.Little,M. B. A. Oldstone, and D. D. Rao. 1978.Defectiveinterferingparticles:theireffecton gene expression and replication of vesicular stomatitis virus, p. 399-408.InG. Todaro, J. Stevens, and C. F. Fox (ed.), ICN-UCLA Symposium on Persistent Viruses. Ac-ademicPress,Inc.,New York.

17. Huang, A. S., and E. K. Manders. 1972.Ribonucleic acid synthesis ofvesicularstomatitis virus. IV. Transcription by standard virus in thepresence ofdefectiveinterfering particles.J. Virol. 9:909-916.

18. Huang, A. S., and E. L. Palma. 1974.Vesicularstomatitis virus: defectiveness and disease, p. 87-99. In W. S. Robinson and C. F. Fox (ed.), Mechanisms of virus disease. Benjamin,Inc.,Menlo Park, Calif.

19. Huang, A.S., D. D. Rao, and G.Lanman.1980.Defective

interferingparticles ofvesicular stomatitis virus:

struc-ture-function relationships. Ann. N.Y. Acad. Sci.

354:238-250.

20. Huang, A.S., and R. R. Wagner, 1966. Defective T particlesofvesicularstomatitis virus.II.Biologicrole in homologousinterference.Virology 30:173-181. 21. Kang,C. Y., T. Glimp, J. P.Clewley,and D. H. L.Bishop.

1978. Studies on the generation ofvesicular stomatitis virus (Indiana serotype) defective interfering particles.

Virology84:142-152.

22. Kolakofsky, D., M. Leppert, and L. Kort. 1978. Agenetic

map of Sendai virus DI RNAs, p. 539-553. In B. W. J. Mahy and R. D. Barry (ed.), Negative strandvirusesand the hostcell.AcademicPress,Inc., New York. 23. Leppert, M., L. Kort, and D. Kolakofsky. 1977. Further

characterization of Sendai virus DI-RNAs: a model for their generation. Cell 12:539-552.

J. VIROL.

24. Marcus, P.I.,D. L.Engelhardt, J.M.Hunt,andM. J. Sekellick.1971. Interferon action: inhibitionofvesicular stomatitis virus RNAsynthesisinducedbyvirion-bound

polymerase.Science 174:593-598.

25. McGeoch, D. J., and A. Dolan. 1979. Sequenceof200 nucleotides at the 3' terminus of the genome RNA of vesicular stomatitis virus. Nucleic Acids Res. 6:3199-3211.

26. McGeoch, D.,A.Dolan,andC. R.Pringle.1980.

Compari-sonsof nucleotide sequencesin the genomes of theNew

Jerseyand Indiana serotypes of vesicular stomatitisvirus.

J.Virol. 33:69-77.

27. Morrison, T. G., M. Stampfer, H. F. Lodish, and D. Baltimore.1975. Invitro translation of vesicular stomatitis virus messenger RNA'sandtheexistence ofa40 S"plus"

strand, p. 293-306. In B. W. J. Mahy and R. D. Barry

(ed.), Negative strand viruses. Academic Press, Inc.,

New York.

28. Naeve, C. W., C. M. Kolakofsky, and D. F. Summers. 1980.Comparisonof vesicularstomatitisvirus intracellu-lar andvirionribonucleoproteins. J. Virol.33:856-865. 29. Palma, E. L., S. M. Perlman, and A. S. Huang. 1974

Ribonucleic acid synthesis ofvesicular stomatitis virus. VI.Correlation of defectiveparticleRNAsynthesiswith standard RNAreplication.J. Mol. Biol.85:127-136. 30. Perrault, J.,andJ.J.Holland.1972.Absence of

transcrip-taseactivity andtranscription-inhibiting abilityin defec-tive interfering particles of vesicular stomatitis virus.

Virology50:159-170.

31. Perrault, J.,andJ. J.Holland.1972.Variabilityof vesicu-lar stomatitis virus autointerference with different host cells and virus serotypes. Virology 50:148-159. 32. Perrault, J., and B. L. Semler. 1979. Internal genome

deletions in twodistinct classes of defective interfering particles of vesicular stomatitis virus. Proc.Natl. Acad. Sci.U.S.A. 76:6191-6195.

33. Perrault, J., B. L. Semler, R. W. Leavitt, and J. J. Holland. 1978. Inverted complementary terminal se-quencesin defectiveinterfering particleRNAs of vesicu-larstomatitisvirus andtheirpossiblerole in autointerfer-ence, p. 527-538. In B. W. J.Mahyand R. D.Barry(ed.), Negative strand viruses and the host cell. Academic Press,Inc., New York.

34. Rao, D. D., and A. S. Huang. 1979. Synthesis of a small RNAincells coinfectedby standardanddefective

inter-feringparticles ofvesicular stomatitisvirus. Proc. Natl.

Acad.Sci. U.S.A. 76:3742-3745.

35. Rao, D. D., and A. S. Huang. 1980. RNA synthesis of vesicular stomatitis virus. X. Transcription and replica-tion bydefective interfering particles. J. Virol. 36:756-765.

36. Reichmann,M.E., D. H. L.Bishop, F. Brown, J. Crick, J.

J. Holland, C.-Y. Kang, R. Lazzarini, S. Moyer, J. Per. rault, L. Prevec, C. R. Pringle, R. R. Wagner, J. S. Youngner, and A. S. Huang. 1980. Proposal for auniform nomenclature for defective interferingviruses of vesicular stomatitis virus. J. Virol.34:792-794.

37. Relchmann, M. E., D. R. Pringle, and E. A. C. Follett.

1971. Defective particles in BHK cells infected with temperature-sensitive mutants of vesicularstomatitis vi-rus. J. Virol. 8:154-160.

38. Rose, J. K. 1980. Complete intergenic andflanking gene sequences from the genome of vesicularstomatitis virus. Cell 19:415-421.

39. Schubert, M., J. D. Keene, R. C. Herman, and R. A.

Lazzarnln. 1980. Site of the vesicular stomatitis virus

genome specifying polyadenylation and theend of the L gene mRNA. J. Virol. 34:550-559.

40. Schubert, M., J. D. Keene, and R. A.Lazzarinl. 1979. A specificinternalRNA polymeraserecognition siteof VSV RNA is involved in the generation of DI particles. Cell 18:749-757.

41. Schubert, M., J. D. Keene, R. A. Lazzarla, and S. U. Emerson. 1978. The completesequence of aunique RNA species synthesized bya DI particle.Cell15:103-112.

on November 10, 2019 by guest

http://jvi.asm.org/

42. Schubert, M., and R. A. Lazzarinl. 1981. In vivo transcrip-tion of the 5'-terminal extracistronic region of vesicular stomatitis virus RNA. J. Virol. 38:256-262.

43. SekeUlick,M.J.,and P. .Marcus. 1980. Viral interference

by defective interfering particles of vesicular stomatitis virus measured in individual cells. Virology 104:247-252.

44. Soria, M., S. P. Little,andA.S. Huang.1974. Character-izationofvesicular stomatitis virus nucleocapsids I. Com-plementary 40 S RNA molecules in nucleocapsids.

Virolo-gy61:270-280.

45. vonMagnus, P.1954. Incompleteforms of influenza virus.

Adv.Virus Res. 2:59-79.

on November 10, 2019 by guest

http://jvi.asm.org/