Vesicular Stomatitis Virus mRNA and Inhibition of Translation of Cellular mRNA—Is There a P Function in Vesicular Stomatitis Virus?

(1)

Vol. 38, No. 2 JOURNAL OFVIROLOGY, May1981,p.504-517

0022-538X/81/050504-14$02.00/0

Vesicular

Stomatitis Virus mRNA and

Inhibition of

Translation

of Cellular mRNA-Is There a P Function in

Vesicular Stomatitis Virus?

HARVEYF. LODISH*AND MARY PORTER

Department of Biology, Massachusetts Institute of Technology,Cambridge, Massachusetts 02139 Received 15December1980/Accepted4February1981

Infectionof animalcellsbyvesicular stomatitis virus(VSV) results in inhibition

of translation of cellular mRNA. We showed previously that, in BHK cells

infected by the Glasgow isolate of VSV Indiana, this is duetocompetition during

the initiationstepofproteinsynthesis of viral and cellular mRNA fora constant,

limiting number of ribosomes. We show herethat infection of thesamecellswith

the San Juan isolate of VSV resulted in a morerapid shutoff of host protein

synthesis and that this was paralleled by a more rapid accumulation of viral

mRNA.Extendingourconclusion that shutoff is duetomRNAcompetition,we

show further that the average size ofpolysomes translating viral and cellular

nmRNA wasthreefold smaller in cellsinfectedby VSV San Juan than by VSV

Glasgow,which, inturn, wasabout one-half that of uninfectedcells.In allcases,

cellular andviral mRNA's which encoded the same-sized polypeptideswerefound

onthesame-sizedpolysomes,aresultindicatingthat theefficiency of translation

of bothtypesofmRNA's is about thesamein the infected cell.Also, therewasno

preferential sequestration of viralorcellular mRNA's in ribonucleoprotein

par-ticles.Additionalcorrelations between the levels of viral mRNA's and the

inhi-bition ofproteinsynthesiscamefromstudies of three otherwild-type VSV strains

andalso fromstudies with Vero and Lcells. Inparticular,therateof shutoff of

L-cellproteinsynthesisafter infectionbyanyVSV isolatewasslower than thatin

BHK cells, and this was correlated withaslowerrate of accumulation of viral

mRNA. VSVtemperature-sensitive mutantswhich synthesized, atthe

nonper-missive temperature,noVSV mRNAfailed toinhibitsynthesisofcellular

pro-teins. Stanners andco-workers (C. P. Stanners, A. M. Francoeur, and T. Lam,

Cell 11:273-281,1977)claimed that VSVmutantRlinhibitedsynthesis of Lcell

protein synthesis less rapidly than did its parent wild-type strain HR. They

concluded that this effectwasdueto amutation inanunspecified VSV protein,

"P."Wefound, in bothLand BHK cells, that Rl infection resulted inaslightly

slower inhibition ofcellular mRNA translation than did HRinfection and that

this was correlated with aslightly reducedaccumulation of VSV mRNA. The

level of VSV mRNA, rather thananyspecific VSV protein, appeared tobe the

keyfactorin

determining

therateofshutoff of host protein synthesis.

Infection of mammaliancellsbyvesicular

sto-matitis virus(VSV) results ininhibitionof syn-thesisofcellularprotein and RNA andeventual cell death. Neither replication of the genomic RNA norproduction ofinfectious virus is nec-essary for theinhibitionofcellularprotein

syn-thesis;however, sometranscriptionofthe

nega-tive-strandedviralgenome into mRNA is

essen-tial(3,4,6-8).Much,ifnot all, of theinhibition

ofcellular proteinsynthesis ismanifestedat the

translational level: after infection cellular

mRNAremainsintact and fullyfunctionalin in

vitro translationsystems, yet istranslatedin the

cellsat adecreasingrate (2).

Atleast three different mechanisms for this

translationalcontrol have beenproposed. Nuss

etal. suggested thatVSV mRNAinitiates

pro-tein synthesis severalfoldmore efficiently than

does cellular mRNA, and thus out-competes

mRNA for ribosomes (9). Stanners et al. (12) have isolated a mutant ofthe HR (Winnipeg)

wild-type isolateof theIndiana serotype ofVSV.

Infection ofL cells with this mutant does not

resultintherapid,precipitous inhibition of

pro-tein synthesisobserved with the HRwild-type

strain. Theyproposed thataspecific viralgene

product,termedP, isrequiredfor thisinhibition

andthat themutant,Rl,is P-.

504

on November 10, 2019 by guest

http://jvi.asm.org/

(2)

IS VSV?

More recently, we showed that in growing

BHKcells over 80% of most of thepredominant

species of cellular mRNA are bound to

poly-somes, and over 60% remainonpolysomes 4 h

after VSV infection (2). The average size of

polysomestranslating individual cellular mRNA

isreduced about two- to threefold after infection.

Forexample, in uninfectedcells, actin

(molecu-lar weight, 42,000) mRNA is found

predomi-nantly onpolysomes with12ribosomes;4hafter

infection it isfound onfivesomes, the same size

aspolysomes that aretranslating VSV N

(mo-lecularweight 52,000) and M(molecularweight

35,000) mRNA. We concluded that the

inhibi-tion ofcellular protein synthesis after VSV

in-fection isdue, in large measure, to competition

for ribosomes by alarge excess of viral mRNA.

Theefficiency of initiation of translation on

cel-lular and viral mRNA's is about the same in

infectedcells; cellular ribosomes are simply

dis-tributedamong more mRNA's (about threefold

more) than are present in growing cells.About

20 to 30% of each of the predominant cellular

andviral mRNA's are present in

ribonucleopro-tein (RNP) particles in infected cells and are

presumably inactive in protein synthesis. There

is no preferential sequestration of cellular or

viral mRNA's in RNPs after infection (2). All of the above-mentioned studies were done with different isolates ofVSV Indiana and in

different cultured cells.Aspart of asystematic

approach to the resolution of thesedifferences,

we have studied the effects of four commonly

used wild-type VSV strains and also the Rl mutant on three of the cell linesused in these

studies. Ourresults indicate that the rate and

extentofinhibition isafunction of both the cell

line and virus used. In allcasestested,however,

therewas agoodcorrelation between the

accu-mulation of viral mRNA and the extent of

inhi-bition of translation of cellular mRNA, thus

extendingourconclusionsonthe mechanism of

translational control to other cell line-VSV

strain combinations. Our studies also suggest

that Rlmaynotbeasimplemutantof the HR

wild-typestrain.

MATERIALS AND METHODS

Virus strains. The Glasgow isolate ofVSV was obtainedfrom RobinWeiss(ImperialCancer Research Fund,London),whooriginallyobtainedit fromCraig Pringle(InstituteofVirology,Glasgow) (3). The San Juan isolate of VSVwasobtained from Robert Laz-zarini(NationalInstitutes ofHealth),who obtained it from MarthaStampferandAliceHuangin 1972. The Mudd-Summers strain was also obtained from Dr. Lazzarini. According to him (personal communica-tion), itwasfirstusedatHarvard and thenpassedon

insuccessiontoDrs.Carver, Marcus,Summers,Mudd, andHolland beforereachinghim.StrainsHR andRl

aredescribed by Stanners et al. (12) and were obtained from C. Stanners (Ontario Cancer Institute). All vi-ruses used in these studies were grown on Vero cells.

Animalcelis and infections. Growth of suspen-sion BHKcells and monolayer Verocells and proce-dures for infection with VSV weredetailed in previous papers (2, 3). Mouse L cells were obtained from C. Stanners and were cultured as monolayers in minimal essential a medium (GIBCO no. 410-1900) with 7% calf serum and 3% fetal calf serum.Infection of these with VSV used a protocol identical to that for Vero cells. In all cases, 10PFU of VSV per cell was used.

Labeling ofinfected cells with[3'Smethionine

andanalysis of proteins by sodium dodecyl sul-fate (SDS)-gel electrophoresis. Procedure for la-belingandelectrophoresis have been fully described inpreviouspublications (2, 3).

Isolation of total cytoplasmic RNA. A previous publication detailed our procedures for isolation of polysomes and polysomal RNA from BHKcellsand forcell-free protein synthesis (2). The procedure de-veloped for extraction of RNA from BHKcellswas used withonlyminormodificationsfor monolayer L cells.About 5x 107cellsin three 150-mm-diameter plates were chilled on ice and washed several times withphosphate-bufferedsaline.Thecells were scraped off theplatesintophosphate-buffered saline and then treated as in theprotocol for suspension cells (2). Note that total cytoplasmic RNA, not selected for polya-denylatedcomponents,wasused.Thus, mRNA activ-itiesof RNAfragmentsarenormalized to totalcellular RNA, i.e., mostlyrRNA,whichisproportional to the number of cellsfractionated.

RESULTS

Virus andcell strains. Much of this study

is a comparison of the effects of five strains of

VSV onthree different host cells. Tracing the

historyof thecommonlaboratorystrainsofVSV Indiana is only slightly less complicated than

followingthegenealogyofOld Testament

patri-archs,but our bestattemptispresentedin

Ma-terials andMethods.Strain Rl is a

temperature-resistantrevertantoftemperature-sensitive (ts)

mutantT1023, which is itselfreported to be a ts

mutant of the HR strainwith alesionin

com-plementation group I (12). The BHK culture

used wasgrowninsuspension, whereas the Vero

andL-cellculturesweremaintained andinfected

as monolayers. Analysis by one-dimensional

SDS-gels of the [3S]methionine-labeled

pro-teinssynthesizedbythese (uninfected) cells

re-veals fewsimilarities except for actin,which is

inall cases the mostpredominant cellprotein.

Thus, to facilitate comparison ofdifferent cell

lines,we have focused onthe translation ofactin

mRNA. After infection of BHK cells by the

Glasgow isolateofVSV,inhibition ofactin

syn-thesis parallels, we have shown, that of most cellularproteins (2).

BHKcellsinfectedby differentstrains of VSV. Infection of BHK cells with four ofthe VOL. 38,1981

http://jvi.asm.org/

(3)

506 LODISH AND PORTER

five virus strains tested (Glasgow,

Mudd-Sum-mers,HR, andR1) resulted in similar kinetics of

inhibition of synthesis of actin andofmost

cel-lular proteins (Fig. 1 and 2). At 2.5 h after

infection, the rate of actin synthesiswas 60 to

75% that of mock-infected cells, andat4h itwas

25 to30%. Infectionby VSV SanJuan,by con-trast,resulted in amuchmore rapidinhibition

of synthesis of predominant cellular proteins

(Fig. 2-4); actinsynthesiswasreducedto45%of

the control value at 2.5 h after infection, 1 h

sooner than after infection by the other VSV

isolates.Inhibition ofsynthesisofother

predom-inant cellular proteins also occurred about 1 h

sooner after infection by VSV San Juan than

VSVGlasgow (Fig. 4).In allcases,by5h

post-infection actin synthesis wasless than 10% of

controllevels (datanotshown).Thisresult,

to-gether with microscope studies on these cells,

indicated thatover90%,andprobablyover98%,

wereinfected with VSV.

The inhibition of cellular protein synthesis

after infection ofBHKcellsby VSVGlasgow is

due, wehave shown (2), tothe competition of

cellular and viral mRNA for theconstant,

lim-iting numberof functional ribosomes. Were this

toapplytoinfection of BHK cellsby VSV San

Juan, thenonewouldexpect a morerapidrate

ofaccumulation of viral mRNA after infection

by VSV San Juan. Figure 3 andTable 1 show

that thiswasthecase.Inthis study,aconstant,

limitingamountofcytoplasmic RNAwas

trans-lated inawheatgermcell-freesystem.The

one-dimensional gelanalysis of Fig.3(lanes11to18)

makestheimportantpoint that all translatable

cellularmRNA's remained intact andfunctional

after infectionby both VSV Glasgow and VSV

San Juan. Permicrogram of cellular RNA,

prep-arations from infected and uninfected cells

di-rected the synthesis of equal amounts of all

predominantcellular polypeptides.Many of the

translation products of cellular mRNA

comi-gratedwith predominantprotein species

synthe-sizedby the growingcells.Anotableexample is

actin (polypeptide y), but conclusive

identifica-tion ofthesecellularspecies must awaitpeptide

mapping. The additional translation activity

present in RNA from infectedcellsdirected

syn-thesis of the five VSV structural proteins, plus

afew otherpolypeptides which may have been

incompleteordegraded products of these. The

amounts ofboth total translatable mRNA and viral mRNAincreased steadily from 2.5 h after

infection, and did so about 1 h earlier after

infection by VSV SanJuan than byVSV

Glas-gow (Table 1).Therefore,the differential

accu-mulation of viral mRNA paralleled the

differ-entialinhibition ofhostprotein synthesis by the

twostrains. The increased levelof VSVmRNA's

inSanJuan-infectedcells wasparalleledby an

c

- -~~~~d

I

-~~~N/NS

BHsactin C

2.5(aeh -)o 4.5h(lns713hferifcin (Lane 6, 7,l

andly3)

Mock

prtinfecnted;s(1

and8)vsv

Glasgow; (2and9)VSV SanJuan; (3 and10) VSV Mudd-Summers;(4and11)VSVHR; (5and12) VSV Rl. Samples ofthe cell lysates were analyzed by electrophoresis throughalinear10 to15% polyacryl-amidegel containing SDS. Shown is a

radioauto-grainofthe driedgel. Theuppercaselettersonthe

rightdenoteVSVproteins,andthelowercaseletters indicatepredominant species ofcellularpolypeptides. J. VIROL.

http://jvi.asm.org/

(4)

a.

40-20_

2 Hours

FIG. 2. Synthesis of actin inBHI different VSVstrains. Theradioau

were scannedwith a microdensitoa

ditions where the darkeningof the proportional to the amount ofrac

Theamountofradioactivityinthe(

relativetothatofmock-infectedcel function of time after infection. Alt inFig.1,therewas noreduction ofo

Ihafter infection ofanyofthestraii

(0) VSV SanJuan (SJ): (-O-) VS (A) VSV HR (HR); (0) VSVMud (--O--) VSVRI (Rl).

accelerated rate of synthesis oi comparedwithinfectionbyVSV

parelanes 4and7and lanes 5a] TheamountofviralmRNAs3 infection of BHKcellsbyVSVs

Rlwassimilartothat in VSV G

cells (Table 2). Togetherwitho

1and2) thatthekinetics of inhil protein synthesisarethesame,t withourconclusion that host shu a function of the amount of tr mRNA.

Is the increasedlevel of viral infected by VSV San Juan, r4

Glasgow, indeedthe causeofti hibition of translation of cellu]

answerthisquestion,westudied distribution of the predom mRNA's and of theviral mRNA tion.Figure5 shows that83% of3

ribosomeswere onpolysomesin and 75% were on polysomes af eithervirus. However,theaver

somes was smallerin the VSVSa cells than in those infected by

I thesewere,inturn,smaller than those in mock-Cells infected cells. At this time the rate ofprotein

synthesis,

as measured by incorporation of [

S]methionine,

relative to that ofuninfected

cellswas82% after VSV San Juan infectionand

87% after VSVGlasgowinfection.Thus,mostof

the polysomes in the infected cells must have

been functional. Theprofiles in Fig. 5a,

there-fore, indicate that thesamenumber of functional

ribosomesaredistributedamong moremRNA's

(i.e., viralplus cellular) after infection by VSV

Glasgow, andyet more after infection by VSV

\\ sHR San Juan.

\ MS

Considerably

moreinformationis obtainedby

\ GL

analyzing

the translationproducts directedby RI- RNA from individualpolysome and RNP

frac-tions onalong polyacrylamide gradient gel (2).

SJ By quantitating with a microdensitometer the amountof eachpolypeptide synthesizedby each i i

gradient

fraction,

it is

possible

todetermine the

subcellular localization of many predominant

cellularmRNA's encoding proteins of molecular

Kcellsifected by weights from 20,000 to 70,000 (Fig. 6).

tgramsof

Fig.

I With few exceptions (proteins

/i,

8, and D in

fhlmwas strictly

Fig.

3), in uninfectedcells over 85% of the

trans-dioactive

protein. latable mRNA encoding the 12 predominant actinpolypeptide, cellular

proteins

was bound to

polysomes

(Fig.

fIs, is plotted as a

6;

2).

In

general,

the size of 'hough not shown polysomes translatinganygivencellularmRNA actinsynthesis at wasproportionalto the size of theprotein

prod-n-sused.Symbols: uct.ActinmRNA (y;molecularweight, 43,000)

VGlasgow

(GL):

was localized predominantly in fractions

con-'d-Summers

(NS);

taining

12 ribosomes, and 0 (molecular weight,

31,000) mRNAwas localized inpolysome

frac-tionscontaining6to 8,whereasmRNA's

encod-fVSV proteins ing proteins 1, ic, X, and ,u (molecular weight,

rGlasgow(com- about20,000)wereenriched in fractions

contain-nd8ofFig. 3). ing5ribosomes.

ynthesized

after Three hours after infectionby VSV Glasgow,

;trains HR4 and actin mRNA waslocalized on hexasomes, 50%

'lasgow-infected the sizeofpolysomes translatingactin mRNA in

urfinding (Fig. growing cells.

Importantly,

viral and cellular

)ition

ofcellular mRNA's encoding proteins of about the same

,hisisconsistent size (N,52,000daltons; M,35,000daltons; actin,

itoff isprimarily 42,000 daltons: and0,31,000daltons)were

trans-anslatable viral latedonpolysomes containingthesamenumber

ofribosomes,inthiscasefourtosix(Fig.6). We

mRNA incells take thisasevidence that therateof

polypeptide

elative to VSV chaininitiation,relativetoelongation,ontypical

ie enhanced in- viral and cellular mRNA's is about the same.

lar mRNA? To The polysomal mRNA's encoding the smaller

Ithesubcellular proteins (I,K,

X,

and,u) were enriched in

struc-iinant cellular turescontainingtwo tothreeribosomes,avalue

k3h after infec- significantly smaller than the fivesomes which

IH-labeled

BHK translate these mRNA's in

growing

cells.

uninfectedcells, The sizeof

polysomes translating

all

predom-terinfection by inant cellular and viral mRNA's was much age size ofpoly- smaller after infection byVSV San Juan than

inJuan-infected by VSV Glasgow. Actin mRNA, to cite one

r VSV Glasgow: example,waslocalized

predominantly

todi-and

VOL. 38,1981

http://jvi.asm.org/

[image:4.500.48.244.61.291.2]

(5)

-G

-a

_R

N/NS

-

y

- E

I

M

~.

m

-'I

-K

-I"

1 2 3 4

5

6 8 9

10

11

12

13 1415

1E

Uninf.

VSV

Unninf

VSV

Glasgow

San

Glasgow

San

Juan

Cell

-

free Translation

508

a-d

actin e

Cells

http://jvi.asm.org/

(6)

IS THERE A P FUNCTION IN VSV? 509

100_

80 VSV Glasgow

60_

40-20 VSV SonJuan

b

100

60

40

-c 20 I

, e (actin)

800

60

-

40-

20-1

h

100

soo-80__

60

40-20

F

2 3 4 5

Hours

FIG. 4. Synthesis ofBHKproteinsafterinfection with VSVGlasgowor VSV San Juan. The autora-diogram of Fig. 3 (lanes 1-9) was scanned with a

microdensitometer. Theareas underthepeaks

cor-respondingtofour predominantcellular proteins,a,

b,c(actin),andh, weredetermined and normalized tothevalueof mock-infectedcells.

trisomes, only 17 to 25% the size ofpolysomes translatingactin mRNA ingrowingcells.Again,

theviral and cellular mRNA'sencoding proteins of thesamesize (N, M, actin,and 6)were trans-latedonpolysomes containingthesamenumber

ofribosomes, heretwo tothree.

Togetherwith the data ofFig. 3 and4, the study inFig.6indicatedthat the reduction after infection in the numberofribosomestranslating each moleculeofactin mRNA and otherspecies of cellularmRNAwas areflection, primarily,of

the increased amount of total translatable

mRNAspecies (predominantlyviral mRNA)in thecell.Virtually thesamenumber ofribosomes were distributed among a larger number of

mRNAmolecules.Thus, the twofold increase in

the totalamountoftranslatablemRNA 3 hafter

infection by VSVGlasgow observed in this

ex-perimentwasreflectedinatwofold reduction in

thenumber ofribosomesperactin mRNA. The

fivefoldincrease foundinthis experimentafter VSV San Juan infectionresulted in a

propor-tionately larger reduction inthe number of

ri-bosomes translating actin mRNA, and thus in the rate of actin synthesis. In neithercase did

there appear to be preferential translation of viral mRNA's or of cellular mRNA's in the

infectedcells,norwastherepreferential

seques-tration of particular cellular RNAs intoinactive

RNPs;about thesamefractions of viral mRNA's

and typicalcellular mRNA'swerein RNPs

(Ta-ble3).

Vero cells.Inhibition ofcellular protein

syn-thesis byall strains of VSV occurredata

some-what fasterrateinVero cells than in BHKcells

(Fig. 7 and 8). The rate of synthesis of total protein 3 h afterinfectionwasgreaterthan 80% of that obtained inuninfected cells (seefigure

legends). Depending on the virus strain, actin

synthesiswasreducedto15 to30% of the control value by 3h. In allcases,by 5 h actin synthesis wasless than 10% ofnornal, andallofthecells

hadrounded andpartly detached from the plate. Thus, wellover 90% of the cells were infected.

Note, however, that the relativeefficiencyofthe different VSV strainstoinhibit cellularprotein synthesis differed in the two types ofcells. In

BHKcells, VSVSan Juaninhibited more

rap-idly than did VSV Glasgow, whereas in Vero cells, both inhibitedatabout thesamerate(Fig.

8). However, in both Vero and BHK cells,

shutoffby VSV HR and Rl strainswasslightly

slower than by VSV Glasgow (compare Fig. 8 and 2). Because the differences among VSV

strainswaslesspronouncedin Verocells thanin

BHKcells,we didnotattempttocorrelatethe

different profiles of shutoffof Veroproteinsby

FIG. 3. Gelanalysis ofproteins synthesizedininfectedBHKcells andofcell-freetranslationproductsof RNAfrom infectedcells. A linear 10to 15%gradientpolyacrylamide gelwas used. (Lanes 1-9) Proteins

synthesizedduringa30-minlabeling periodwith[3S]methionine, beginningatthe indicated time.(1) Mock-infectedcells,2.5h; (2)3.5h; (3)4.5h; (4) VSVGlasgow-infected cells,2.5h; (5)3.5h;(6)4.5h; (7) VSV San Juan-infected cells,2.5h; (8)3.5h;(9)4.5 h.AsinFig. 1,the lower-caselettersrepresentpredominantspecies ofhostproteins. (Lanes10-18)Proteinssynthesizedinawheatgermcell-freesystemusing2 gofthe indicated

RNAsper25-,dreaction.A5-,ulsample ofthereactionwasanalyzed. Table1presentsthe totalamountof

acid-precipitable radioactivity incorporatedperreaction, as well astherelative amountofVSVproteins made. (10)RNAfrommock-infected cells,2.5h; (11)3.5h; (12)4.5h; (13)RNAfrom VSV-Glasgow-infected cells,2.5h;(14)3.5h;(15)4.5h; (16)RNAfromVSV SanJuan-infected cells,2.5h;(17)3.5h; (18)4.5 h. The

Greek lettersontherightdenotepredominanttranslationproducts ofcellularmRNA,and thecapitalletters

arethe VSV structuralproteins.

r

a

I

VOL. 38,1981

http://jvi.asm.org/

[image:6.500.67.223.64.358.2]

(7)

the various strains(whicharereproducible)with

the levels of VSV mRNA's.

MouseLcells.Therateof inhibition of

cel-lular proteinsynthesis after VSVinfectionwas

much slower in L cells (Fig. 9 and 10) and in

TABLE 1. Messenger activityin RNAfrom VSV-infected BHK celisa

Relative amt of mRNA activity VSV mRNA

Cell

5['S]methi-onine incor-

Normal-porated (105 °zed N/NS M cpm/,ug of

RNA)

Uninfected 4.18 1.00 VSVGlasgow

2.5h 4.98 1.19 1.00 1.00

3.5h 6.99 1.67 3.50 3.62

4.5 h 13.86 3.31 11.50 12.10

VSV San Juan

2.5 h 8.41 2.01 4.85 5.31

3.5 h 12.98 3.10 10.85 11.10

4.5 h 20.59 4.92 19.63 21.60

'Total cellular RNA was extracted from BHK cellsatthe indicatedtimes after infection. Wheatgermcell-freeprotein synthesis reactions(25Ll) containing 1,2, or4itgoftotal cytoplasmic RNAwereused,andduplicate5-,ulsampleswere

assayed foracid-precipitableradioactivity.For allRNAs,the amount of[3S]methionineincorporatedintoacid-precipitable radioactivity wasproportionaltotheamountof RNAadded; shown incolumn 2isthe totalamountof'S radioactivity incorporated in the entire 25-1lreaction. Incolumn3, the results are normalizedto themRNAactivityof RNAfrom uninfected cells. Each of these reactionswasanalyzed by SDS-gelelectrophoresis,andtheamountofN/NS and Mproteins wasdeterminedfrom the area under thepeakof the micro-densitometer scan of theradioautogram. These valuesare

normalizedtothe value characteristic of RNA from cells 2.5 h after infectionbyVSVGlasgow.

either Vero or BHK cells. At 3 h, synthesis of

actinin Lcellswas70to 100% ofthat of

unin-fectedcells, dependingontheVSV strain used.

In all ofourexperiments, shutoff of actin

syn-thesis afterinfectionof LcellsbyVSV Rlwas

slightlyless rapid than after infection by VSV

HR, althoughthis difference wasnotobserved

in thetwoother cell lines tested (Fig. 2 and8). Inallcases,the totalamountofproteinsynthesis

atanytime up to 5 h ofinfection was greater

than 90% that of uninfected cells (data not

shown).

Two typesof comparisonsindicated thatthe

shutoffofsynthesisofL-cellproteinwas

corre-latedwiththe levelof VSVmRNA. First, VSV

mRNA accumulated to a slightly lower level

after infection of L cells by VSV Rl than by

VSV HR (Table 2); this corresponded to a

slightlyslowerrate ofinhibition ofL-cellactin

synthesis by VSV Rl. Second, with all three

VSV strains tested (Glasgow, HR, and Rl),

much less translatable VSV mRNA

accumu-lated afterinfectionin L cells than inBHK cells (Table 2); this correlatedwellwith thereduced

shutoff observed in Lcells.Note that theamount oftranslatable mRNA (per milligram of cyto-plasmicRNA) recovered from uninfected Lcells

and BHK cellswasthesame and thatthe

pre-dominantspecies oftranslatable L-cell mRNA

also remainedfullyfunctional after VSV infec-tion(Fig. 11).

VSV ts mutants. As another approach to

investigating the role of VSV mRNA in

inhibi-tion of cellular proteinsynthesis,weusedtsVSV

mutants in each of the five complementation

groups (10, 13). All of the mutants used were

TABLE 2. Messenger activity in RNA from VSV-infected L cells and BHKceilsa

Total mRNA activity Relative amount of VSV mRNA Time after

Cells Virus infection [3S]methionine

(h) incorporated(105 Normalized N/NS M

cpm/igofRNA)

BHK None 2.60 1.00

VSVGlasgow 3 6.11 2.35 1.00 1.00

4 8.40 3.23 1.58 1.68

VSV HR 4 9.31 3.58 1.88 1.94

VSVRi 4 7.46 2.87 1.39 1.36

Lcels None 2.20 1.00

VSVGlasgow 3 3.26 1.48 0.46 0.41

4 4.36 1.98 0.86 0.79

VSVHR 3 4.03 1.83 0.68 0.64

4 5.02 2.28 1.03 0.97

VSVRi 3 3.32 1.51 0.53 0.44

4 4.49 2.04 0.99 0.96

aAnalysis of the total

cellular

RNAswas detailed in the legend to Table 1. Values for total mRNA activity

arenormalizedtothatobtained with RNA fromuninfected BHK or Lcells.In all cases, the amount of viralN/ NS andMmRNAactivity is nornalizedto that obtained from RNAisolated3 hafterinfectionof BHK cells withVSVGlasgow.

J. VIROL.

http://jvi.asm.org/

[image:7.500.56.249.165.313.2] [image:7.500.60.454.467.631.2]

(8)

IS THERE A P FUNCTION IN 511

VSV Son Juan

.3

c

2_

L1a,

VSV Glasgow

Uninfected

z

b

l

2 3 4 5 6 7 L 9 10 I 4

VSV SanJuan

3-VSV Glasgow a V

0

o-'2

Uninfected w VS G o

Polyribosome

i * Size

I I 1018 6 5 4 31 2 I1 5 10 15 20 25 30 35

Fraction

FIG. 5. Subcellular localizationofribosomes and translatable mRNA in growing BHK cells and in cells 3 h after infectionby VSVGlasgow and VSV San Juan.Asampleof6xplO BHKcellswasgrown for20hinthepresenceof1.5mCiofa'HJuridine (The RadiochemicalCentre)and harvestedby centrifuga-tion. One-thirdofthe cells were mock infected, and one-third each wereinfectedwithVSVGlasgow and VSV San Juan. Three hoursafter infection, a post-nuclearsupernatant fr-omthe cultureswasprepared

andcentrifuged througha 15 to30%(wt/vol)sucrose

gradient,asdetaiedpreviously(2).Shown inpanel b isthe 3Hradioactivityper50-ldportion offractions from thegradient ofthe extractfrom uninfectedor

infectedcells. The sizeofpolyribosomes was deter-mined by extrapolation oftheprofile ofan extract

fr-om rabbit reticulocytes analyzed in parallel. As

indicatedbythebrackets, 1.0mlfromthreeadjacent

fr-actions was pooled. RNA wats then isolated and

translated in a wheat germn cell-fr-ee system. Two differentamountsofRNAwereused,generallyabout 1.5 and3.0%ofthe total RNArecoveredfr-om each

derived fromtheGlasgow isolateofVSV. Three

of themutantsmade littleor no VSVmRNAat

39.50C,

the nonpermissive temperature:

tsG114(I),

tsG22(ll);

andtsG41(IV).Infectionof BHK cells (Fig. 12) or Vero cells (data not shown) at39.5°Cbythesemutantshad noeffect at all on the rate ofsynthesis ofpredominant

cellularproteins. Shutoffofcellularproteinsdid occur, by contrast, after infection by the two

RNA' mutantsused: tsG33(III) and tsL513(V)

(Fig.12).Thelowerrate of VSVprotein

synthe-sis intsL514(V)-infected

cells,

relativeto

infec-tion by the wild type, was a consequence of

increased cell death; at 3 h these celLs

synthe-sized

wild-type

levels of VSV proteins (14).

Again, there was an excellent correlation

be-tweensynthesis ofVSV mRNA andinhibition

oftranslation of

cellular

mRNA. Acorollary of

thisfinding is that the infecting particles

them-selves,atleastup to alevelof 10PFU/cell,had

noeffectonmRNAtranslation.

DISCUSSION

Dependingontheisolate of VSV Indiana and

onthe host cell used, therearemarked

differ-ences in the kinetics of inhibition of cellular

proteinsynthesis. In all casestested, this

inhi-bition was manifest at the translational level:

theamountoftranslatablemRNA's for the

pre-dominant cellular proteins which can be

ex-tracted from the cells remainsunchangedafter

infection, whereas therateofsynthesisof

cellu-larproteins decreasessteadily.

Weemphasize again (2) thatwehaveusedin

vitrosynthesisof discrete viral and cellular

pro-teinsas ameasureof theamountof

correspond-ingmRNA in theinfected cells.Especiallyinthe

caseof VSVRNAs, thistechniqueis

preferable

to the morecommonly used

hybridization,

gel

separation,or isotopeincorporation techniques

toquantitateVSVmRNA,since itmeasuresthe

biologically relevant

parameter-its ability

to

direct protein

synthesis.

Not all

VSV-specific

polyadenylated RNAs are functional mRNA's,

and thus physical measurements canseriously

overestimate the amount of VSV mRNA. A

fraction. Within this range, protein synthesis was

invariablyproportionaltotheamountofRNAadded (cf.reference 2). Plottedontheordinateinpanelais thetotal amountoftranslatable mRNA perpooled

fraction:this is theproductofcpm

of[3SJmethionine

incorporated/cpm of

[3HJRNA

added and total

amountof

[3H]RNA

in thepooledfractions. Since recoveries of[3H]RNAwereverysimilarinall frac-tionsandaveraged85%,thecorrectionfordifferential recovery of[3H]RNA in the different fractions is small anddoesnotaffecttheposition ofanyofthe datapoints shownbymorethan10%.

VOL. 38,1981

http://jvi.asm.org/

[image:8.500.44.235.71.462.2]

(9)

c 0 0

.-LL

z

E

120 100

80

60 40 20

80

40

80

40

120

80

40

mRNA Fraction

l I I I 1

18 12 9 6 4.5 3 2 I RNP 18 12 9 6 45 3 2 I RNP

Polysome Size

FIG. 6. Subcellularlocalizationof specificcellular and viral mRNA. RNAfromdifferent fractions ofthe polysomegradientof Fig.5 wastranslated inawheatgermcell-freesystem,and theproductswereresolved byelectrophoresis through a10to15%gradientpolyacrylamide gel(cf. Fig. 3).In all cases, anamountof recoveredRNA,equivalentto0.76% ofthe total RNA present in the initialpooledfraction,wasaddedto a 25-ulcell-freereaction. Threedifferent exposuresofthegel ofthetranslationproducts werescanned witha Joyce-Loeblmicrodensitometer,andtheareasofthepolypeptidebandsindicated(cf. Fig. 3)weredetermined. A1-h exposurewasusedforthe VSV G and Mprotein; a24-or48-h exposurewasusedforcalculationof cellular bands.Theareas(in arbitrary units)shownontheordinatearenormalizedtotheequivalentofa 24-h exposureofthefilm; thus,the valuesforthedifferentpolypeptidesareproportionaltotheirrelativeextents

of synthesisin the wheatgermextract. The average sizeofthegradientfractions fromwhich theRNAwas isolatedwastakenfromFig.4.

significant fraction of normal VSV 13 to 18S polyadenylated RNAs possess the 5' termini

pppAandpppGininfectedcelLs. Thesearenot

foundonpolysomes andarepresumably inactive asmessengers (11).Incellsinfectedatthe

non-permissivetemperaturebytsmutantsdefective inthe Mprotein, there isavastoverproduction

of VSVpolyadenylatedRNAs(1, 5).Gel electro-phoresisandhybridizationstudies indicatethat they are indistinguishable from nonnal VSV

mRNA (1, 5). However, they are

undermeth-ylated andare notfoundonpolyribosomes (5).

Previously,westudied the mechanism of this

inhibition in BHKcellsinfected by the Glasgow

isolate of VSV Indiana (2). At 4 h the total

amountoftranslatable mRNA (viral plus cellu-lar) was about threefold that of growing cells.

Most species of cellularand viral mRNA were

localizedonpolysomes.Infection did result in a

reduction(about threefold)inthe rate of initia-tion oftranslationofallpredominantspeciesof

cellular mRNA, but translation ofcellularand

J. VIROL.

http://jvi.asm.org/

[image:9.500.67.461.60.437.2]

(10)

[image:10.500.46.242.77.214.2]

TABLE 3. Fractionofviraland cellular mRNA's in RNP particlesa

Fraction of mRNA

mRNA Uninfected VSVGlasgow- VSV San Juan-cells infectedcells infected cells

a 0.19 0.14 0.20

Y 0.07 0.14 0.30

( 0.18 0.20 0.29

0.16 0.18 0.29

Kc 0.15 0.31 0.26

A 0.17 0.34 0.28

N/NS 0.16 0.39

M 0.13 0.37

aThe

data in Fig. 6 were used to calculate the fractionof mRNA activity for each of six cellular and two viral mRNA's which is localized in the RNP faction(gradient fractions32 to37;pools11and12).

viral mRNA'swhich encoded for thesame-size

proteins occurredon the same size of

polyribo-somes.We concludedthat the rates ofinitiation

ofprotein synthesisontypicalcellularandVSV

mRNA'sarethesame and that the large amount

ofviralmRNA synthesizedafterinfection com-peteswithcellularmRNAforaconstant number

of ribosomes. Our results are in disagreement

with the prediction of Nusset al. (9), who

con-cluded,onthe basis ofthe effects ofhypertonic

treatment ofcells, that VSVmRNA istranslated

inpreferencetohost mRNA. However, neither

themechanism of inhibition ofproteinsynthesis

by hypertonic treatment nor the relevance of

these results to mRNA translation in nornal

media isatall clear.

Thepresentresults extendourconclusionsto additional VSV strain-host cell combinations.

First, inhibition of cellular proteinsynthesis is

dependent on synthesis of VSV mRNA. After

infection at the nonpermissive temperatureby

VSV tsmutants defectiveinsynthesis of VSV

mRNA,therewas noinhibitionatall of

synthe-sisof cellularproteins (Fig. 12). Identical results

were obtained withmutantsdefectivein the N

(complementation group IV),NS (groupII), or

L(groupI)proteins. Mutants defectivein the M

orG protein, bycontrast,made closetonormal

levels of VSV mRNA and inhibited synthesis of

cellular proteins with normal kinetics (Fig. 12

and datanotshown).

Second, inBHKcells, threeof the four VSV

isolates tested [Mudd-Summers, HR

(Winni-peg),and

R]

inhibitedsynthesisofcellular

pro-teinsatabout thesame rate asdid theGlasgow

strainofVSVIndiana.Similarly,theamountof

VSV mRNA accumulated after infection by

these strains was approximately the same, as was the rate of VSV protein synthesis in the

infected cells (Fig. 1). Bycontrast, infectionof

BHK cells by the San Juan isolate of VSV

resulted ininhibition of synthesis of allcellular

proteins, and actin inparticular, about45 to60

min earlier than did infection by the other

vi--N/NS

-actin

-M

78 9 10 11 12 13 14

!

_C,2

_e5

6

;Z

A-

C

FIG. 7. Gel analysis ofproteins synthesized in Verocellsinfected bydifferentstrainsofVSV. Sam-plesofinfectedormock-infected Vero cellswere la-beledfor30minwith[3S]methioninebeginningat2 h(odd-numberedlanes)or3h(even-numberedlanes) after infection. Analysisofthe samples was as de-scribed in thelegendtoFig.1.Inparenthesesisgiven the total amount of

[OSJmethionine

incorporated intoprotein3hafterinfectionrelativetothe valueof

uninfectedcells. (1 and2) Uninfectedcells(1.00); (3

and 4) infected by VSV Glasgow (0.90); (5 and 6) infected by VSV San Juan(0.85); (7and8) infected byVSVMudd-Summers(0.81);(9and10)infectedby VSVHR (0.81);(11and12)infected byVSVR1(0.91). 38,

http://jvi.asm.org/

[image:10.500.250.442.124.547.2]

(11)

00

I1

ruses (Fig. 2 and 4). This inhibition was

corre-tXVero CelIlIs latedwithan increase in therateof

accumula-VtO Cellsdtionof VSVmRNA and inanaccelerationof the

rateofVSVproteinsynthesis: thesameamount

80L

ofVSV

mnRNA

was accumulated inthe cell45

min sooner after infection by VSV San Juan

thanbyVSVGlasgow. (Thehistoryof the

lab-oratorystrain of VSV SanJuan is highly rele-vant[A. Huang,personal communication]. The

60\ strain, as originally obtained in 1963 from the

American TypeCulture collection, formed

rela-tively smallplaques andyieldedastock ofvirus

of about 108 PFU/ml. To select for a variant

ICL| \ \ \\ _ which grew to ahigher titer, Dr.Huangpassaged

40\ R thevirusinCHOcells17timesinsuccessionat

RI

\\tlow multiplicities of infection. After this, the

HR virus was cloned repeatedly. The yield of

infec-tious virus in a

single-step

growth

curve was

20 MS increased about 10-foldby this enrichment

pro-GL cedure, and it is apparent thata more rapidly

Si replicatingvariant [at least in some cells]was

selected. The differences between this isolate

and the other VSV wild-type strains is most

1 2 3 pronounced in the BHKcellline.) The effects of

Ho u r s this increased amount of inRNAonthe

[image:11.500.59.254.60.327.2]

subcel-lular distribution ofcellular (and viral) mRNA

FIG. 8. Synthesis of actin in Vero cellsinfectedby are incomplete accord with our

conclusion

that different strains of VSV. The gel profile shown in viraland cellular mRNA'scompete on anequal Fig. 7 wasanalyzedasdetailed in Fig.2.Plotted is basis during the initiation step ofprotein syn-theamountof radioactivityin the actinpeptide, rel-

..

.

a n

ative to thatofmock-infected cells, at several times thesis for a

limiting

number ofribosomal sub-after infection. Although not shown, there is no unts. At 3 h after infection by either virus, at change in the rate ofactin synthesisIh afterinfection least 75% ofcellularribosomesremain in poly-by any VSVstrain. Symbols are as in Fig. 2. somes,andthetotalrateof totalprotein

synthe-L Cells

M

N/NS ctin

2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30

hr | 2hr 3hr | 4hr | 5hr

FIG. 9. Gelanalysis ofproteins synthesized in L cells by different strains ofVSV.Samples of mock-infected orVSV-infected cellswerelabeledfor 30 min with[3S]methioninebeginning at1h(lanes1-6),2h (7-12), 3 h (13-18),4h (19-24),or5h (25-30)after infection. Total cellular protein was analyzed byelectrophoresis througha10%polyacrylamide gel containing SDS; shown is a radioautogram of the dried gel. (Lanes 1, 7, 13, 19,21,25,27)Uninfected cells; (2, 8, 14, 20,26) infected by VSV Glasgow; (3, 9, 15) infected by VSV San Juan; (4, 10, 16, 22, 28)infected by VSV Mudd-Summers; (5, 11, 17, 23, 29) infected by VSV HR; (6, 12, 18, 24,30) infected by VSV RI.

J. VIROL.

http://jvi.asm.org/

[image:11.500.65.456.419.599.2]

(12)

IS THERE A P FUNCTION IN VSV? 515

L

G N/NS -actin

-M

[image:12.500.50.243.74.240.2]

2 3 4 5 6 7 8 9 10 11 12

FIG. 10. Cell-free translation products of RNA from VSV-infectedLcells.Proteins were synthesized inawheatgermcell-freesynthesisreaction, using 1 pgoftheindicated RNAs per25-,ilreaction. A

5-jtl

sampleof the reactionwasanalyzed on a 10 poly-acrylamide gel containing SDS. Table 2 presents the total amountofacid-precipitableradioactivity incor-porated perreaction,aswellasthe relative amounts of VSVproteins produced. (Lane 1) No RNA; (2) RNAfrommock-infectedLcells,3h;(3)RNAfrom VSVGlasgow-infected L cells, 3 h; (4) RNA from VSVHR-infectedLcells,3h; (5) RNAfrom VSVR1-infected Lcells, 3h; (6) RNAfrommock-infected L cells, 4 h; (7) RNA from VSV Glasgow-infected L cells,4h;(8) RNAfrom VSVHR-infectedLcells,4 h; (9) RNAfrom VSVRI-infected Lcells, 4h; (10) RNAfromadifferent preparation ofVSV Glasgow-infectedL cells,4h; (11)RNAfrom VSV Glasgow-infected BHKcells,4h; (12)RNAfrom VSVGlasgow-infected BHKcells,3h.

sis is atleast 80% that of

giowing

cells. Thus,

the vastmajority of ribosomesarefunctioning

in protein synthesis. However, the averagesize

of polysomes translating cellular and viral

mRNA's ismuchsmaller intheVSV San

Juan-infectedBHKcells.After infectionbyVSV

Glas-gow,mRNA's fortwotypicalcellularproteinsof

molecularweight of 30,000to40,000, 0 andactin,

are found to be pentasomes, the same size as

polysomeswhicharetranslatingVSV N(52,000)

and M (35,000) mRNA. This is 40 to 50% the

size ofpolysomeonwhich isfound actin mRNA

in growing cells, a result consistent with the twofold increase intotal amount of functional mRNA per cell. In the VSV San Juan-infected

cells, by contrast, actin, 0, N, and M mRNA's

arealllocalizedtodi-ortrisomes-one-sixth the

number of ribosomes per actinor0 mRNA found

ingrowing cells.Thiscorrespondstothe five-to

sixfoldincrease in total translatable mRNA per cell observed in thisexperiment.

Although ourresults on infection ofmouseL

cells by VSV are different in several respects

fromthoseusing BHK cells, theyareconsistent

with the notion that host shutoff is a

conse-quenceofmRNAcompetition for limiting

ribo-somes. First, for all VSV strains studied the kinetics of hostshutoffareslower in L cells than inBHK cells:about1hin the case ofinhibition

oftranslationofactin mRNA. Thisisparalleled

by a correspondingly slower rate of

accumula-tion ofVSV mRNA. Inparticular, the amount

ofVSV mRNA found3 h afterinfection of BHK

cells with VSVGlasgowis about the same as at 4h afterinfectionof L cells by thesamevirus

preparation (Table2).

Anadditional correlation emerges from a

com-parisonoftheinfectionby L cells by VSV strains

HR4 and Rl. AsStanners et al. (12) reported,

inhibition of cell protein synthesis isless after

Rlinfection than after HRinfection. However,

theconsistentdifferencewefind(Fig. 11) isnot

as marked as that previously described (10).

Nonetheless, thereducedlevel of VSVmRNA

which accumulates by 3 h in Rl-infected cells

relativetoHR4-infected cells correlates with the

reducedinhibitionofactinmRNAtranslation.

Stannersetal. (12)foundamuchmorerapid

inhibition ofL-cell protein synthesis after

infec-tion by the HR strain thanwe have observed.

Since we have used cell linesand virus grown

fromstocksprovidedby C. P.Stanners,wehave

L Cell s I0oo

80k

r-60

a)

a. H

401-

201-1 2 3 4 5

Hou r

FIG. 11. Synthesis of actin in L cells infected by four strains of VSV. The gelradioautogramof the gel in Fig.9 wasanalyzedasdetailedin thelegendto Fig.2.Symbolsarealsodescribed there.

38,

HF

http://jvi.asm.org/

[image:12.500.252.448.378.627.2]

(13)

234 5 67 8

b~~\/ V

'V

5

no explanation for this difference, except that

we are culturing thecells asmonolayers. Also, in allofourstudieswe takecare that the cells

aregrowinglogarithmicallyatthetimeof

infec-tion; thephysiologicalstateof the host cellcan

L affect the response to VSV infection (12). We

have confirmed the finding of Stanners et al.

that inhibition ofL-cellprotein synthesis is less

rapidafter infectionby strain Rl thanby HR. However, wequestion their assumption that Rl

is, in fact, a single-step mutant derived from

strain HR, and thus the conclusion that the

differences in kinetics of shutoff of L cells

be-tweenHR4 andRl aredue toamutation in a

single viralgenefunction, P. Recall that Rlwas

isolatedas atemperature-resistant revertantof

tsT1023(I), which issupposedtobea mutantof

strain HR. We havefoundconsistently that the

Rl Mprotein migrates 10% fasteronSDS-gels

thandoes the HRMprotein. This istruefor M

proteinssynthesized inBHK, L, and Vero cells

(Fig. 1, 7,and 9). It is also thecasefor HR and

-0 Rl M

proteins

synthesized

inawheat

germ

cell-free system (Fig. 10). It isdifficult to imagine

howasingle-step mutation could result insuch

alarge shift ingelmobility.Clarification of this

point willrequireisolation of P- mutations

di-rectly fromaknownwild-type strain.

- N / N S We do not know, in detail, why the kinetics of

inhibition of cellular proteins synthesis varies

-ac

ti|

n with the host cell and with the strain of VSV

used. Thereis,asnotedabove,agoodcorrelation

between the kinetics of host shutoff and the

accumulation of viralmRNA, but it isnotclear

why the rate of accumulation of viral mRNA

should besodependentonthe host cell line and

:M on the strain of VSV. Our preliminary studies

indicate that the inhibitionof actin synthesis 3

hafter infection ofBHKcells by10 or 50 PFU/

cellisnotsignificantly(10%)different,so

differ-entialadsorptionorpenetration ofthevirionsis

unlikelytobeinvolved. As noted in thetext, in all cases all of the cells are infected by VSV.

Whereas a VSV virion contains its own RNA

polymerase activity which will, in vitro, direct

synthesis of all five VSV mRNA's, it is not

known whether, in the infected cells, specific

[image:13.500.69.253.58.603.2]

host proteins (which might be present also in the virionin minute amounts) are also essential, nor is it known how the intracellular levels of

FIG. 12. Gelanalysis synthesisofproteins synthe-sizedinBHKcellsinfectedat40.0°Cbytsmutantsof VSV. As inFig. 1, cellswereinfectedby the indicated mutantsandlabeled with[3S]methionine4hafter infectionat40°C. Electrophoresisofthe labeled pro-tein utilized at 10 to 15%gradient polyacrylamide containing SDS.(Lanes1and2) Uninfected cells; (3) infection bywild-typeVSVGlasgow;(4) infection by tsG114 (I): (5)infection bytsG22(II); (6)infection by tsG33(III);(7)infection by tsG41(IV); (8)infection by tsL513(V).

516

http://jvi.asm.org/

(14)

A P

ribonucleoside triphosphatesorSAdoMetinthe

cells might influence therateof viral RNA

tran-scription.

Taken together, our results indicate that in-hibitionof cellularprotein synthesisisprimarily a consequence ofcompetition by viral and

cel-lular mRNAfor a constant limiting number of

ribosomes. Shutoffclearlyisdependenton

syn-thesis ofviral mRNA. However,ourdataonthe

subcellular distribution of mRNA's after

infec-tion andonthecorrelation between the levels of

viral mRNA and inhibition of cellular mRNA

translationareonly semiquantitative. It is

diffi-culttoeliminaterigorously other factors which

could reduce the rateoftranslation ofcellular

mRNA's after infection. Previously we listed

someof these: (i) reduction in therateof

poly-peptide chainelongation, (ii) death ofacertain

fraction of cells, and (iii) sequestration into

RNPs of a significant fraction (about 25%) of

cellspecies of cellular and viral mRNA's. These

mRNA's could be localized only to the dead,

inactive cells. Whatever the mechanismfor

in-crease inthelevel of RNPs afterinfection, it is

significant that there is no difference in the

extentofsequestration oftypicalviral and

cel-lular mRNAspecies.

ACKNOWLEDGMENT

Thisworkwassupported byPublicHealthService grant AI-08814-13from the National Institutes of Health.

LITERATURE CiTED

1. Clinton, G.M., S. P.Little,F. S. Hagen,and A. S. Huang.1978.The matrixproteinof vesicular stomatitis virusregulatestranscription. Cell15:1455-1462. 2. Lodish,H.F.,and M. Porter. 1980. Translational control

ofproteinsynthesisafter infectionbyvesicular stoma-titis virus. J. Virol. 36:719-733.

3. Lodish, H. F., and R. A.Weiss. 1979. Selective isolation of mutants of vesicular stomatitis virus defective in production of the viral glycoprotein. J. Virol. 30:177-189.

4. Marcus, P.I., M. J.Sekellick, L Johnson, and R. A. Lazzarini.1977. Cellkillingby viruses. V. Transcribing defective interfering particles of VSV function ascell killing particles. Virology 82:242-246.

5. Martinet, C., A. Combard, C. Printz-Ane, and P. Printz. 1979. Envelope proteinsreplicationof VSV: in vivo effects ofRNA' temperature-sensitive mutations on viral RNAsynthesis. J.Virol. 29:123-133. 6. Marvaldi,J., J. Lucas-Lenard, M.Sekellick,and P.

I.Marcus. 1977. Cellkilling by viruses. IV. Cell killing and protein synthesis inhibition by VSV require the same gene functions. Virology 79:267-280.

7. Marvaldi, J., M. J. Sekellick, P. I. Marcus, and J. Lucas-Lenard. 1978. Inhibition of mouse L-cell protein synthesis by ultraviolet-irradiated VSV requires tran-scription. Virology 84:127-133.

8. McAllister, P. E., R. R. Wagner. 1976. Differential inhibition of host protein synthesis inL-cells infected with RNA- temperature-sensitive mutants of vesicular stomatitis virus. J.Virol.18:550-558.

9. Nuss, D. L, H. Opperman, and G. Koch. 1975. Selective blockage of initiation of host protein synthesis in RNA-virus infected cells. Proc. Natl. Acad. Sci. U.S.A. 72: 1258-1262.

10. Pringle, C. R. 1977. Genetics of rhabdoviruses, p. 239-290.In H. Fraenkel-Conrat and R. R. Wagner (ed.), Comprehensive virology, vol. 9. Plenum Publishing Corp., New York.

11. Rose,J. K. 1975. Heterogeneous 5'-terminal structures occur on vesicular stomatitis virus mRNAs. J. Biol. Chem. 250:8098-8104.

12. Stanners,C. P., A. M. Francoeur, and T. Lam. 1977. Analysis of a VSV mutant with attenuated cytopatho-genicity: mutation in viral function, P, forinhibitionof protein synthesis.Cell11:273-281.

13. Wagner, R. R. 1975. Reproduction of rhabdoviruses, p. 1-94. In H. Fraenkel-Conrat and R. R. Wagner, (ed.), Comprehensive virology, vol. 4. Plenum Publishing Corp., New York.

14. Zilberstein,A., M. D. Snider,M.Porter, and H. F. Lodish. 1980. Mutants of vesicular stomatitisvirus blocked at different stages in maturation of the viral glycoprotein. Cell 21:417-427.

http://jvi.asm.org/