Establishment and maintenance of persistent infection by Sindbis virus in BHK cells.

(1)

Vol. 33,No. 1 JOURNAL OF VIROLOGY, Jan. 1980,p.463-474

0022-538X/80/01-0463/12$02.00/0

Establishment

and Maintenance

of Persistent Infection

by

Sindbis Virus in BHK Cells

BARBARAWEISS,RICHARD ROSENTHAL,AND SONDRA SCHLESINGER*

Department ofMicrobiologyandImmunology, Washington UniversitySchoolof Medicine,St.Louis, Missouri63110

We have establishedapersistentinfection of BHKcellswithapreparationof

Sindbis virusheavily enriched in defective interfering (DI) particles. Thesmall

fraction ofcells that survived the initial infection grew out to form a stable

population ofcells[BHK(Sin-1) cells],mostof whichsynthesized viral RNA and viral antigens. The presence of DI particles in this virus stockwas required to

establish thispersistentstate.BHK(Sin-1) cellsreleasedasmall-plaque,

temper-ature-sensitive virus (Sin-i virus) as well as DI particles containing DI RNAs

larger than thosepresent in the original stock used to establish the persistent

state. A cloned stock of Sin-i virus, free of detectableDIparticles, was able to

initiateapersistent infectionmorequickly and withgreatercellsurvival than the

original stock of Sindbis virus containing DI particles. About 2 weeks after the Sin-1 virus-infected cells were cultured, DI RNAs arose and soonbecame the

dominant viral RNAspecies produced by thesecells.

Manyhighly cytopathic RNAvirusesare

ca-pableoflong-term persistenceandreplicationin

cultured mammalian cells. In the most

exten-sively studied system

involving

vesicular

sto-matitis virus (VSV), three major factors were

implicated in the establishment and

mainte-nanceof such persistently infectedcultures: (i)

defective interfering (DI) particles, (ii)

muta-tions in the standard virus (particularly ts

mu-tations), and (iii) interferon production. Huang

and Baltimore (7)werethe first tosuggest that

DI particles might have an important role in

viral persistency, and Holland et al. (6) have

providedstrong evidencesupportingthis idea in

a model system

involving

VSV grown in BHK

cells. DIparticleshavealsobeen shown tobe a

major factorinpersistency involving Sendai

vi-rus (19), lymphocytic choriomeningitisvirus (9,

27), and reovirus (4, 23).Temperature-sensitive

mutants are frequently recovered from

persist-entlyinfected cultures and are oftenrequiredto

establishpersistencyeveninthe presence of DI

particles (16, 28-30). Recently Holland et al.

demonstrated that numerous mutations in the genome of VSV accumulate withtimein long-term carrier cultures (5). This selection for a highly mutated virus contrasts with the genetic

stability observedduringsuccessivelyticcycles

ofinfection with VSV and suggests that these mutations enhancethe survivalof infected cells. Ramseur and Friedman (17) have shown that interferon can facilitate the establishment of

persistency with VSV in L cells. In addition,

Sekellick and Marcus (21) have shown that

cer-tain ts mutants and DI particles of VSV are

particularly goodinducers ofinterferon. In

fur-ther support of arole for interferoninpersistent infection,Nishiyama et al. (13, 14) have isolated

fromapersistently infected L-cell culturea

mu-tant of VSV which is a better inducer ofand moresensitive to interferon thanwild-type VSV.

Althoughthemechanismsofviralpersistency

are becoming increasingly clear for

rhabdovi-ruses and other negative-strand virhabdovi-ruses, the

fac-tors mediating viral persistency with the

posi-tive-strand alphaviruses are less well

under-stood.Infectionofinvertebrate cellcultureswith

alphavirusesisnoncytopathic,andpersistent

in-fections are

readily

established. During the

maintenance ofthese infected

cultures,

temper-ature-sensitive mutants (22), antiviral factors

(18), and DI

particles

(2, 11) can be

demon-strated,

although

the relative

importance

of each of these factors in persistency has not been determined. Infection of vertebrate cells with

alphaviruses ishighly cytopathic, butit hasalso

been

possible

to establish

persistent

infections in thesecultures. Severalinvestigatorshave

sug-gested that DI

particles

mediate persistent

in-fections in vertebrate

cells,

but the carrier cul-tures were notwell characterized (1, 8, 20). In

addition, Inglot et al. (8) have presented

evi-denceimplicatinginterferon in theregulationof

thepersistentstatewithSindbisvirus in mouse

cells.

Wehave initiatedexperimentstoexamine the relative importance of DIparticles, mutations in thestandardvirus,andinterferonproductionin 463

on November 10, 2019 by guest

http://jvi.asm.org/

(2)

464 WEISS, ROSENTHAL, AND SCHLESINGER

the establishment andmaintenance of persist-ently infected cultures involving the alphavirus Sindbis virus, grown in BHK cells. Our studies show that DI particleswererequiredtoestablish a persistent infection with wild-type standard virus. The resulting infected cell cultures syn-thesized and released DI particlesas well as a mutated standardvirus. Both themixed popu-lation of virus released from infected cellsand the cloned mutant standard virusreestablished persistent infection with much greater facility than the original DI-enriched virus stock. DI particles were not involved in establishing the persistent infection initiated by the mutant vi-rus, but during subsequent long-term mainte-nance of these cultures DI particles invariably became the dominant viral species detected in these cells.

MATERIALS AND METHODS Cells. BHK-21cells andpersistentlyinfected

cul-turesderived from themweremaintainedat37'C in

Eagle minimal essential mediumsupplemented with 6% fetal calf serum. Primary chicken embryo fibro-blastsgrownin minimal essential mediumcontaining 3% fetal calf serum were used in plaque assays as

describedpreviously (26).

Viruses. The Sindbis virus stock, containing DI particles, that wasused to establishpersistency was

obtained by high-multiplicity passaging of wild-type Sindbis virus in BHK cells. The properties of this

particularpopulationof DIparticles havebeen

exten-sively described (10, 25, 26).The stocks ofwild-type Sindbis virus, Semliki Forest virus, and VSV (San Juan)werepreparedbypassageatlowmultiplicity of infection (MOI) onBHK cells. The tsmutants (ts6 andts24) werederived fromthe HR strainof Sindbis

virus(14).

Plaque purification of virus released from

BHK(Sin-1) cells. Cloningof thesmall-plaquevirus released by BHK(Sin-1) cells was carried out on

chickenembryo fibroblast monolayers. To obtain large plaquesof thismutant, assayswerecarried out in 0.5% agar containing DEAE-dextran (100 Lg/ml). Plaques

were picked after growthat 30°C for3 days. Three consecutiveplaque purificationswereperformedwith no intermediate amplification ofthe virus obtained

from asingle plaque.Thethrice-clonedvirus wasthen

passagedin chickenembryofibroblastsat300Cat an MOIof0.1.Allexperimentswith thecloned viruswere

donewith this virus stock.

Labeling andpurificationof viral RNA.

Lyti-callyinfected andpersistently infectedcellswere

la-beled with[3H]uridine (20 jtCi/ml, 28.5Ci/mmol) in

thepresence ofactinomycinD(1.0jig/ml)foraperiod

of8 or 16 h. [3H]uridine was added to the lytically

infectedcells1 hpostinfection.RNA wasisolated from

BHK cells essentially as described previously (25). Cell extracts were obtained by lysingcells in buffer

(0.05 M Tris, 0.1 M NaCl, 0.001 M EDTA, pH 7.5)

containing0.5%Triton and 0.5% recrystallized

naph-thalene disulfonate. All reagentsweresterilized when

possible and redistilled or recrystallized when heat

sterilizationwas notpossible.

Agarose gel electrophoresis. All RNA samples weredenatured withglyoxal before electrophoresis by

the methodofMcMasterandCarmichael(12).

Dena-turationwascarriedoutfor1h at50°C, usually 1 day

priorto electrophoresis. Electrophoresis was carried outusinghorizontal, 1.1% agaroseslab gels (5 mm in

thickness), prepared in 0.01 M sodium phosphate

buffer (pH7.0) andrunfor 3.5 h at approximately 3

V/cm (measured across the gel itself). Continuous

recirculation of buffer was essential to maintain a

constantpH during therun.

Immunofluorescenceassay.Cellsweregrown on

coverslips and, where indicated, infectedwithSindbis

virus. Cover slips were washed with phosphate-buffered saline (threetimesfor 5 min each), drained,

and dried overnight. Acetone fixationwas carriedout

at room temperature for 10 min. Dried cover slip

cultures werethen incubated with0.2 mlof a 1:100

dilution of rabbit antibodytoSindbisvirus.After30

min ofincubationinahumid chamberat room

tem-perature,thecellswerewashedthreetimes with

phos-phate-buffered saline and then incubatedwith 0.2 ml

ofa1:100dilution offluorescein-conjugatedgoat

anti-rabbitantibody (Gateway Immuno Sera Co., St. Louis, Mo.). Cells were mounted on slideswith Permount

and examined withaZeiss microscopeequippedwith anIVFlepi-fluorescence condenser.

Autoradiography. Approximately 6 x 105 cells

grown on coverslipswerepretreated withactinomycin

D (1 Ag/ml) for 3 h at 37°C and then labeled with [3H]uridine (10 jiCi/ml) for 3 h in the presence of

actinomycin D (1 ,ug/ml). Cover slips were then

washed withphosphate-buffered saline, fixed in two

changes of5%trichloroaceticacid for 10 min at 4°C,

anddehydrated with 70% ethanol followed by ether. Cover slips were counted in a scintillation counter

withatoluene-basedscintillant, rinsed with fresh tol-uene and ether, and then air dried. Mounted cover

slipswerecoatedwithNTB2(Kodak)liquidemulsion andexposedfor3to5daysat4°C.Cellswerestained withGiemsa stain afterdevelopment for better visu-alization.

Immunelysis assay.The complement-mediated

lysisof5'Cr-labeledcellstreated withviralantibody

wascarriedout

using

theproceduredescribedby Fan

andSefton(3).TheSindbis-infectedBHK cellcultures

werelabeledwith51Crat5hafter infectionat anMOI

of 100. Both the VSV and Sindbis antibodies were

used ata1:100dilution.

Superinfection assay. Carrier or normal BHK

cellswereinfected with virusat anMOI between 10

and100asindicated in the tables. Aftera 1-h

adsorp-tionperiod,monolayerswerewashed three timeswith warmphosphate-bufferedsaline and incubated for16 to 18 h ingrowth medium. At 1 h afterinfection, a

samplewas removedto determine the residual virus

presentafterthiswashing procedure.

RESULTS

Persistent infection of BHK cells with

Sindbis virus. Infections of BHK cells with

Sindbis virus are usually very cytopathic,

whereasinfections with virus stocksenriched in J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

(3)

SINDBIS VIRUS PERSISTENT INFECTION 465

DI particles showdecreased cytopathogenicity. Wewere unable to establisha persistent

infec-tion of BHK cellsusingour stocks ofstandard

Sindbisvirus; theobservation that DIparticles reducedcellkillingprovidedabasis fortheiruse

in establishingpersistentlyinfectedcultures.

The preparation of Sindbis virus usedto

es-tablishapersistent infectionwas alate-passage

virus stock with a high degree of interfering

activity (seeTable4).BHKcells,infectedwith this virus at an MOI of5, were maintained at

37°C. Cytopathogenicity developed slowly, but

after4daysmostof the cells had detachedfrom

the plate. Dead cells were removed from the

plate by severalchanges of medium.By 9days afterinfection,ninecolonies of cellswerevisible. By 22 days the colonies were large; cells were

thendispersed using3 mM EDTA andreplated

without dilution. From this time on cells were

subcultured ata 1:10 dilution when confluency

was reached, usually every 4 days. Cells were

dispersedwithEDTA untilpassage 5 (38 days

in culture). Thereafter trypsin was used in all

subculturing. These cells have beendesignated BHK(Sin-1).

BHK(Sin-1) cells could not be distinguished

from theoriginaluninfected BHK cellsbyeither

morphology orgrowthrate. Theyproducedlow

levels ofvirusatarateof lessthan 1 PFU/cell

perdayandwereresistant to infectionby

Sind-bis virus (Table 1). These properties are

com-moncharacteristicsassociated withpersistently

infected cultures. Theexperiments described be-lowwere undertaken to examine the extent of

Sindbis virus expression inthese cells

approxi-mately3 monthsafterthey had been in culture,

and to determinethe factors thatcontributedto

thestabilityofthevirus-infectedculture.

Absence of interferon in BHK(Sin-1)

cells. We tested acid-treated medium from

con-fluentBHK(Sin-1)cellsforexogenousinterferon

using a plaque reduction assay with VSV; we wereunabletodetect interferon.Youngneretal.

(29) and Ramseur and Friedman (17) reported similar results in L cells persistently infected

with VSV,but demonstrateda role for

endoge-nous interferon in maintaining the persistent

state.Inthese latterexamples,the infected cells

wereresistanttochallenge bybothhomologous andheterologousviruses. Asshown inTable 1,

infection of BHK(Sin-1) cells with VSV

pro-duced normalyieldsof virusand total cellkilling, providing further evidence that interferon was

not a major factor in the maintenance ofthis

persistentinfection.

Presence ofvirus-specific RNAin

BHK-(Sin-1) cells. The synthesis of virus-specific

RNA was detected inBHK(Sin-1) cells by the

incorporation of [3H]uridine into actinomycin D-treated cells (Table 2). Theincorporation of

[3H]uridine in BHK(Sin-1) cells was increased

overthat obtained in uninfectedcells,but itwas

onlyabout one-fourth the amount obtained ina

culture infected with standard virus. To

deter-minewhether the decreased rateof [3H]uridine incorporation was due to the synthesis ofviral RNAby onlya smallfraction ofthe BHK(Sin-1)cellsortothesynthesisoflessRNApercell, we exposed the [3H]uridine-labeled culturesto autoradiography. The micrographs shown in

Fig. 1 demonstrate that most, ifnotall, of the BHK(Sin-1) cellsweresynthesizing viral RNA.

The grains over the BHK(Sin-1) cells and the

cells infected with standard virus were

distrib-utedmainlyoverthecytoplasm,whereas in the uninfected cells the grains were localized over

the nucleus. The grain densities over the

BHK(Sin-1) cells were uniformly lower than

thoseovercells infected withSindbisvirus,

dem-onstrating that the former synthesizeless viral RNApercell than the latter.

Expression of virus-specific protein in

TABLE 1. Infection ofBHK(Sin-1)cells withhomologous and heterologous virus

Cell line Virus MOI Cytopatho- Virusyieldh(PFU/ml)

genicitya 1h 16 h

BHK Sindbis 135 + ND' 5.5x 108

BHK(Sin-1) Sindbis 135 0 1.3 x 105 1.2 x 105

BHK VSV 10 + ND 7.5x 108

BHK(Sin-1) VSV 10 + ND 6.0x 108

BHK(Sin-1) 0 ND 3.5x 105d

aCytopathogenicitywasscored at16hpostinfection.+,Severesigns of cell damage; 0, no difference from an

uninfected control cellpopulation.

bCells (106per35-mmdish) wereinfectedwithvirusfor 1 h at 37°C. Virus wasremoved, and the

cells

were

washed three times with phosphate-buffered saline containing 1% fetal calf serum. Fresh medium was then

added, andasamplewasremoved forplaquingattheindicated times.

'ND,Not done.

d Inthis case, all theplaquesweresmall and characteristic of the virus released fromthese cells.

VOL. 33,1980

on November 10, 2019 by guest

http://jvi.asm.org/

(4)

BHK(Sin-1) cells. To determine whether

BHK(Sin-1) cellssynthesized virus-specific

pro-teins as well as viralRNA,weanalyzed acetone-fixedBHK(Sin-1) cells for thepresence of viral antigens using indirect immunofluorescent

staining.These data (Fig. 2)showed thatabout

60% of thepopulation expressedviralantigens.

Lysis of infected cells by virus-specific

anti-body andcomplement, as measured by 5'Cr

re-lease,is a sensitive methodtodetectviral

anti-gens on cell surfaces (3). We have used this procedure to compare the expression of viral

antigensonthe surface ofBHK(Sin-1)and cells

infectedwith standard Sindbis virus. Our results

(Table 3) show that theamountof51Cr released

bythepersistently infectedcellswassimilarto,

althoughreproduciblylowerthan,thatreleased

by cells infectedfor 6 h with Sindbis virus.These

data provide additional evidence that most of

the cells in the

BHK(Sin-1) population

were

synthesizing viral

antigens.

In

addition,

someof

these antigens were

expressed

on the cell

sur-face.

Characterization of

virus-specific

RNAin

BHK(Sin-1) cells. The particles released

by

BHK(Sin-1) cellsinterfered with the

replication

of standard Sindbis

virus, suggesting

that DI

particleswerealso

being

released

by

these cells

(Table 4).Moredefinitive

proof

forthepresence

of DI particles was obtained

by

an

analysis

of the viral RNAs

synthesized

by these cells. A

sample of the

persistently

infected culture was

labeledwith

[3H]uridine

inthe presence of

acti-nomycin D for 16 h at 80 and 100

days

after

TABLE 2. Labeling of BHKandBHK(Sin-1)cells

with[IH]uridinea

Cell type Virus cpm of_{incorporatedb}[3H]uridine

BHK None 7,400±723C

BHK Sindbis 83,515±885

BHK(Sin-1) None 19,475+3,120

aBHKorBHK(Sin-1)

cells

(3x 105

cells

perdish)

weregrownonglasscoverslips in35-mmdishes.At 24 hafterplating, actinomycinD(1,tg/ml)wasaddedto

each plate, and,where indicated, cellswereinfected withSindbisvirus(MOI= 50). One hour later,

[3H]-uridine(10,uCi/ml)wasintroduced, and4hlater the

coverslipswereprocessedasdescribed in thetext.

bMean values ± standard error determined from

threeexperiments.cpm, Counts perminute.

'The addition ofhigher concentrations of actino-mycinDdidnotreduce thisbackground.

infection. The RNA was purified, denatured

with glyoxal, and subjected to high-resolution

gel electrophoresis. [3H]uridine-labeled RNA

wassimultaneously isolated from cellsinfected

with standardSindbis virus and from cells

in-fected with thelate-passage stock ofvirus

con-tainingDIparticles,which was used toinitiate

the persistent infection. As shown in Fig. 3,

BHK(Sin-1)cellssynthesizedalargenumber of

RNAspecies thatweredistinct from the

stan-dard viral 42S and 26S RNAs and from the

original DI RNAs. The largest major band of

RNA synthesized by BHK(Sin-1) cells (Fig. 3,

lanes4and5)had acalculatedmolecularweight

of1.50x 106,comparedtothat of 1.65 x106for

26S RNA.Thedifference between themobility

of 26S RNA and that of the largest DI RNA

bandwas moreeasily discernedwhen thelargest DI RNAbandwas not asoverexposed as it is in

lanes4and 5. Themolecularweights of the DI

RNAsdetected in thesecellsranged from1.5 x

106 to 0.8 x 106. The two major species ofDI RNA present in BHK cells infected with the

original DI particle-containing viral stock had

molecular weights of0.75 x

106

and 0.68 x 106

(lane 3 of Fig. 3), which were lower than the

smallest majorDIRNAobservedin

BHK(Sin-1) cells.

ThemultipleDIRNAs observed in

BHK(Sin-1) cells were also seen when BHK cells were

infected with the uncloned virus preparation

released from BHK(Sin-1) cells. This result

demonstrated that the different size classes of DIRNAs present inBHK(Sin-1) cells could be

packagedintoparticles.

Stability

of BHK(Sin-1) cells to

varia-tions intemperature.The virusreleasedfrom

BHK(Sin-1) cellswas a small-plaque,

tempera-ture-sensitive variant (Table 5). If the survival

ofthepersistently infectedcellsweredependent

on the temperature-sensitive phenotype of the

virus, growing the cellsat alower temperature

(30°C) might have led toincreased viral

repli-cation and cell death, whereas growth at the restrictivetemperature(40°C) might have led to curing of the cells or toselection of a

non-tem-perature-sensitiverevertant.We were unable to

affect the stability of the persistently infected

cells,

however, by altering the temperature.

BHK(Sin-1) cellsgrown at30°Cfor 25days did

not show any increase in cell death. The

cells

grown at40°C for 34daysshowed no evidence

ofcuring. They remained resistant to

superin-FIG. 1. Autoradiography ofcellspulse-labeledwith[3H]uridinefor3h. (A) Uninfected BHKcells. (B)

BHKcells infectedwith standard Sindbis virus(MOI= 100;pulse-labeledfrom3 to 6hpostinfection). (C)

BHK(Sin-1)cells. InallcasescellswereexposedtoactinomycinD(1

tLg/ml)

for3hbefore labeling.

J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:4.504.71.261.447.500.2]

(5)

&1 7.

. b.

''I

'U 4

467

A' *

t'

.

I

V.

--AO* io

-a'

V,", ,I&.

,,,

.0 I.. .wR.

AAAL.

on November 10, 2019 by guest

http://jvi.asm.org/

(6)

468

on November 10, 2019 by guest

http://jvi.asm.org/

(7)

fectionbySindbis virus and continuedtorelease

atemperature-sensitivevirus.

Curing of BHK(Sin-1) cells. The

autora-diographyandantibodystudies establishedthat

a large fraction ofBHK(Sin-1) cells were

syn-thesizing Sindbis virus-specific products.

Al-though persistent infections with RNA viruses

may appear, as in this case, to be completely

stable, it has been found that the association

between virus and cell may be lost when individ-ualcells are cloned from thepopulation (6).To

determine whether this would also be true for

BHK(Sin-1) cells, we examined 11 clones

de-rived from this culture. The cloning efficiency

was68%, providingevidence that we were

ana-lyzingarepresentativesampleof the cells.Nine

of the clones were considered to be cured of

virus based on plaque assays of the medium,

immunofluorescence, andcomplete

susceptibil-ity to infection by Sindbis virus. Two of the

cloneswere still

infected,

basedon these tests.

Thesamepercentageofcellsstained

positive

for viralantigens in these clones as in the original

persistentlyinfectedculture, indicatingthat the

original BHK(Sin-1) cells did not represent a stable mixture of viralantigen-positive and

an-TABLE 3. Immunelysisofpersistently infected cellsa

Celltype Virus Antisera Percent51Cr

released'

BHK None Sindbis 2.5±0.61

BHK None VSV 2.2±0.15

BHK Sindbis Sindbis 73 ±2.4

BHK Sindbis VSV 4.3±3.2

BHK(Sin-1) None Sindbis 53±9.9 BHK(Sin-1) None VSV 3.5±0.93

aThose

cells

infected with Sindbis virus received

100PFU/cell4.5hbeforelabelingwith"'Cr.Eachcell

culture containing 3 x 105 cellswas labeled for 1 h

with13 &Ciof"'Crperml in minimal essential medium

plus 1% fetal calfserum. The plateswerethen rinsed five times and incubated in 0.4 ml of this medium, containing a 1:80 dilution of antiserum and a 1:20

dilution of guinea pigcomplement,at37°C for 45 min. Thesupernatantfluidwasremoved, and the culture

wasrinsed twice. The rinses were combined with the first supernatant fluid. The remainingcellswere then lysed by freezing andthawingtwotimes with water.

bThe percentageof"'Crreleasedwascalculated as described by Fan and Sefton (3) and is based on a totalincorporation (supernatantplus lysate from

fro-zen and thawed cells) of approximately 40,000 cpm. The dataaremeanvalues+standard error determined fromfourexperiments.

tigen-negative cells.

Isolation and characterization of the

small-plaque variant produced

by

BHK-Sin-1) cells. We have initiated a more

de-tailedstudyof the virus isolated from BHK(Sin-1) cells.As a first step the virus waspurified free of DIparticles by three successiveplaque isola-tions. Itwasconsideredtobe free of DIparticles based on itsinability to interfere with the repli-cation of standard virus (Table 4). A further criterion for the absence of DI particles was that BHK cells infected with this virus synthesized only the standard viral RNA species(Fig. 4, lane 3).

Thepurified variant (Sin-1) was temperature sensitive in the ability to synthesize RNA in infected cells (Table 6). Even at 30°C, cells infected with this newly isolated virus synthe-sized less viral RNA than cells infected withthe standardwild-type virus.

Establishment of persistent infection

with the cloned virus from BHK(Sin-1)

cells.Our initialattempttoestablishpersistent

infection of BHK cells with Sindbis virus

re-quired the presence of DIparticlesin thevirus

TABLE 4. Homologousinterferenceasameasureof

thepresenceofDIparticles

Virus added Virus yield

(MOI) (PFU/ml)

StandardSindbisa (10) 2.3x 1010

Standard Sindbis (10)pluslate- 7.0x 106

passageSindbis(5)b

StandardSindbisC (10) 6.0x 109

StandardSindbis (10)plus uncloned 6.0x 107

Sin-1(5)d

Standard Sindbis(10)plus cloned 5.3x 109

Sin-1 (5)

aBHK-21

cells

(106

cells) wereinfected with

stan-dard virus alone or simultaneously coinfected with

standard virus and a late-passage stock of Sindbis

virus. Thevirusyieldwasdetermined 16h

postinfec-tion.

'Thispreparationistheoneusedto establish the

BHK(Sin-1)culture.

c BHK-21cellswereinfectedwithstandard virusin

thepresence ofactinomycinD(2

jg/ml)

for1hbefore

challenge with either the uncloned or cloned Sin-i

viruspreparation.Thevirusyieldwasdetermined16

hpostinfection.

dThe medium from BHK(Sin-1) cellswas

centri-fuged (3 h at 102,000 x g), and a sample from the

resuspended pellet was used to obtain an MOI of

standard virusof5.

FIG. 2. Indirectimmunofluorescence staining of BHK cells. (A) Uninfected cells. (B) Cells infected with

standardSindbisvirus. (C) Persistently infectedcells. Cells were prepared as described in the text.

VOL. 33,1980

on November 10, 2019 by guest

http://jvi.asm.org/

[image:7.504.246.438.351.475.2] [image:7.504.42.235.355.454.2]

(8)

1 2 3 4 5 1 2 3 4 5 6 7 8

28S- _ _ _

-_.

28S-18S

--^ - 26S

.b

FIG. 3. Viral RNA species synthesized by

BHK(Sin-1) cells. [3H]uridine-labeled intracellular

viral RNAs denatured inglyoxalwereseparated by

high-resolutionhorizontalagarosegel

electrophore-sis. Lane 1, BHK ribosomal 28S and 18S marker

RNAs;lane2,viral RNAssynthesized duringa

stan-dard Sindbis virus infection; lane 3, viral RNAs

synthesized in cells infected with the late-passage

virus usedtoestablishBHK(Sin-1) cell cultures. The

viral RNAspresentinBHK(Sin-1)cells labeledafter

80 days and 100 days in culture, respectively, are

shown in lanes 4 and 5.

TABLE 5. Temperature sensitivity of virus released by BHK(Sin-1) cells

Efficiency ofplaquingb

Tempa(°C) StandardSind- S v bis virus

Sm-1

virus

30 1.0 1.0

37 0.9 0.88

38 0.95 0.29

40 0.85 0.00059

aTemperature of incubation for plaque titration. Plaquingwascarriedout onchickenembryo fibroblast monolayers.

b Theefficiencyofplaquingisthe ratioof the

num-ber ofplaques observedatthetemperaturetested over

that obtained at300C.

preparationandconsiderable timetoallowthose

fewsurviving cellstogrowout. In contrast, we

wereabletoestablishpersistent infections quite

readily withSin-i whether or not DI particles

were present in thepreparation. When the

in-fections withSin-1 viruswere carriedout at37

or 38°C, the level of cell death was greatly

FIG. 4. Agarose gel electrophoretic analysis of

glyoxal-denatured viral RNAs synthesized by per-sistently infected BHK cells established with the clonedSin-1 virus. Cellswerelabeled with

[3H]uri-dine atvarious times after infection, and the RNA

wasprepared for electrophoresisasdescribed in the

text.rRNA'sfromBHKcells and standardSindbis

viral RNAsareshown in lanes1 and 2,respectively. Lanes3through8show the patternsoflabeled viral RNAsobtained with cells at1, 4, 9, 15, 26, and 18

daysafterinfectionwiththe clonedSin-1 virusatan

MOI of50 orgreater. Lanes 3, 4, and 8 arefrom differentcarrier cultures and represent cultures1, 4, and 18daysafter infection.Lanes5,6,and7arefrom asinglecultureat9,15, and26days after infection.

Theslight discrepancy in mobility of42S and 26S RNAseenin certain lanes arises because allofthe

sampleswerenotelectrophoresed simultaneouslyand representstheminormobilityvariation observedfrom oneelectrophoresisrun toanother.

TABLE 6. Temperature sensitivity of Sin-1 virus for RNAsynthesis in BHK cells

[3H]uridineincorporation'(cpm)

Temp

(OC) Standard Sindbis

Sin-i

virus virus

30 34,962 6,420

38 69,562 628

40 978 0

aThe data have beencorrectedfor the

incorpora-tion (counts per minute; cpm) incomparable

unin-fected cells treated with actinomycin D (2 ,ug/ml).

Adsorptionofeach virus (MOI=100) was at 30°C for 1 h. Cellswerethenshiftedtotheappropriate

tem-perature.Labelingwasfrom5 to 7hpostinfectionwith

20,uCiof[3H]uridineperml.

reduced.Asmeasuredbytrypanblueexclusion, about 15% of the cellsdied 24hafter infection

withSin-iplusDI

particles

and about30% after

infection with clonedSin-i. In each case,by6to - - 42S

-42 S

S

18S- _

-26S

J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:8.504.82.254.73.308.2] [image:8.504.272.468.75.266.2] [image:8.504.73.263.452.530.2] [image:8.504.274.464.481.550.2]

(9)

7 days afterinfection, a stable virus-producing cellpopulation was obtained. The ability to sur-vive after infection with cloned Sin-i was de-pendent on temperature; at 300C essentiallyall of the infected cells died. We considered the

possibility that the enhanced ability of the

cloned virus to establishpersistencywasdirectly related to itstemperature-sensitive phenotype. Ifthisweretrue, other RNA- temperature-sen-sitive mutants might also show this property.

We examined parallel cultures of BHK cells

infected either with ts6 or ts24, both

RNA-temperature-sensitivemutants (15),orwiththe

cloned Sin-i virus. Infections were initiated at

30°C for 6 h at MOIs of 0.5, 5.0, and 50.0.

Thereafter, cultures were maintained at 37°C.

All infections with ts6 resulted in total cell

kill-ing.Infectionswith ts24 alsovirtually

destroyed

these cultures. A few cells, however, survived

andslowlygrewtoform colonies.Repeated

cy-clesofgrowth and cell death ensued for several

months, untila morestable carrierstateevolved.

Thus, an RNA- phenotype per se is not

suffi-cient to confer on Sindbis virus the ability to

establishastablepersistentinfection.

About 6 days after infection with

Sin-i

at

37°Cwith orwithout DIparticles,thecultures

weretestedfor theirsensitivitytosuperinfection

by Sindbis virus and were found to be

com-pletely resistant. Theywere sensitive to

heter-ologousviruses

(Table

7),suggestingthat

inter-feronwas notinvolved in the establishment of

persistency. Immunofluorescent staining was

carriedout oncells infectedwithclonedSin-1or

with Sin-i plus DI particles. In both cases a

largepercentage ofthe cellswere viralantigen

positive.

Generation ofDIRNAin BHKcells

per-sistently

infected with

Sin-i.

Carrier cultures initiatedbyinfection withcloned

Sin-i

virus did not contain detectablelevels of DI RNA for at least 1 week after infection. As shown inFig. 4, lane 3, the RNApatternobserved in BHKcells infected with the cloned virus at anMOI of70 and labeled within the first24hafterinfection

TABLE 7. Infection of BHK cells carrying cloned

Sin-i withhomologousandheterologousvirus

Days Virus yield

postin- (PFU/rnl)

fection Virus added MOI

with

Sin-I 1h 18h

O Sindbis 100 NDa 2.1x 109

8 Sindbis 100 1.5x 104 3.0x 105

0 VSV 10 ND 3.3x 105

8 VSV 10 <102 3.6x 108

0 SemlikiForest 100 ND 2.4x 109

8 SemlikiForest 100 ND 2.0x 109 a ND, Not done.

was indistinguishable from the

early-passage

wild-typepatternshown inlane2.The

remain-ing lanes in this figure show the viral RNA pattern seen in three independent persistently infected cultures established with the cloned virusandlabeledatselectedtimesduringtheir growth. Lane 4shows that in one ofthese cul-tures at 4dayspostinfectionno DIRNAs were

detectable. Lane 5 shows another independent

carrier culture, in which the dominant species made at 9dayspostinfectionwere still 42Sand 26S RNA. Lanes 6 and 7 show the viral RNAs

detectable in this same culture at two

subse-quent times afterinfection, 15daysand 26days. At 15 days postinfection, the standard viral

RNAs still predominated; a number of new

bandswerevisible, both largerandsmallerthan

26S RNA. Thus, by 15 days, DI RNAs were present but were not markedly enriched for in

these cultures. The diffuse band between 42S

and 26S RNA may representa

large

setofDIs

with small size differences. Noneofthese RNA

bandspersisted forverylong (Fig.4,lane7),and

by26 days there was one dominant species of

DI RNA. Lane 8 represents another culture

persistently infected with Sin-i and illustrates

that the major species of DI RNA appearing

duringthefirstmonthafterinfection maydiffer

in size inindependentlyestablished carrier cul-tures.

Wehaveattemptedtoincreasethe sensitivity

of detection ofDIsby using the virus released

from early passages of Sin-l-infected cells to

infect BHK cells. This type ofinfectionwould

amplify any DIparticlespresent in the

popula-tion. Virus stockswereobtainedfrom three

in-dependently established virus-producing

cul-tures at7, 13, or 16days afterinfection. In each case the cells from which the virus was taken wereconsidered to be DInegativebasedonthe

analysis of virus-specific intracellularRNA

pat-terns.Byamplification,asshown inFig.5, it was

possible todetect minorspeciesof

[3H]uridine-labeled RNA migrating between 42S and 26S

RNA. These results suggested that DI RNAs

werebeing generatedat alow levelin these

cells,

even though at these times they could not be

detecteddirectlyin thepersistently infected cul-tures. The data clearly demonstrate, however, that DI RNAs did not represent a significant fraction of the viral RNA found in BHK cells

duringthe initialperiod establishing persistent

infection with

Sin-i

virus.

DISCUSSION

We initiated a study of

persistent

infection with Sindbis virus to

identify

the factors in-volved inestablishing thistype of infection and

inmaintaining this virusin cultured cells.

Sev-VOL. 33,1980

on November 10, 2019 by guest

http://jvi.asm.org/

[image:9.504.47.237.554.659.2]

(10)

2 3 4

42S

28S -.

_ _ 26S

18S _

FIG. 5. Viral RNA species synthesized in BHK

cells infected with virus releasedfrom DI-negative

persistently infected cultures. Lane 1, 18Sand28S rRNA's. Lanes2,3,and4,[3H]uridine-labeled

intra-cellular viral RNAssynthesized in actinomycin

D-treated cells after infection with virusfrom 7-, 13-,

and 16-day cultures, respectively. Each virus was

obtainedfrom an independently initiated

Sin-i-in-fected cell culture. The virus was concentrated by

pelletingandresuspendedtogiveanMOI of3.RNAs

were denatured with glyoxal and electrophoresed

througha 1.1%agarosegelasdescribedin thetext.

Lane 2 showsaparticularly high ratio of 42Sto26S

RNA. This anomalous resultwasnotobtainedwith

anyofourotherpreparations ofviral RNA.

eral conclusions can be drawn from ourinitial results.First,withourstrain ofstandardSindbis

virus, the presenceof DI particles in the stock

wasessential toestablish a persistent infection

in BHK cells. The cells that grew out after infection remained viable and havenever under-goneaperiodof crisisduringthemorethan200

days they have now been in culture. Second,

basedonthesynthesisof RNA in thecytoplasm ofactinomycinD-treatedcells,mostof thecells wereinfected. Third, interferon did notappear

tobeacritical factorinestablishingor maintain-ing this infection inBHKcells.Fourth,mostof thecellscloned from theBHK(Sin-1)cultureno

longer retainedanyofthe characteristics of

per-sistently infected cells, demonstratingthat the expression of viral genes was not a permanent

genetic propertyof the cells. The loss ofvirus

from these cells after cloning indicated that a

significantnumber ofuninfectedcells canarise

inthepopulation. Theappearanceofuninfected cells could be explained if the replication of

standardvirusweresuppressedtosuchanextent by DI RNA that the level of virion genomes

would be too low to be passed on to progeny

cells. In a mass culture of BHK(Sin-1) cells, uninfected cells would be rapidlyreinfected by virus and DI particles released by neighboring cells. This type of reinfection would not occur during the growth of a colony from asingle cell.

The standard virusobtainedfrom

BHK(Sin-1) cells was asmall-plaque, temperature-sensi-tivevariant. Our initial characterization of this virus showed that it had a temperature-sensitive defect in RNA synthesis. Animportant distinc-tion between Sin-i and the original standard

Sindbis viruswasthatinfection with the former

was much less cytopathic. BHK cells infected

with Sin-i at 37°C establisheda persistent

in-fection inthe absence ofDIparticles. Sincethis decrease in cytopathogenicity may be a crucial factor in the initial establishment ofpersistent

infection, itwould be important to knowwhat

alterations in the viral genome causedthe phe-notypeof decreasedcytopathogenicity. Wewere

not able to isolate either ts+ or large-plaque

revertants ofSin-i, anditisprobable that this variant hasmultiple mutations. Althoughwe are

presently examining the properties of this

mu-tant,it may be difficult toidentifytheparticular

changes associated with decreased

cytopatho-genicity.

Themajorviral RNAspeciesin BHK(Sin-i)

cellswereDIRNAs,butthese RNAswerelarger

andmore heterogeneous than those present in

the originalstock used to establish the carrier

culture.This observation suggeststhatDI RNAs

were being generated continually from virion

RNAand that thosepresent

originally

didnot

have a strong selective advantage over newly

generatedones.Themultiple speciesofDIRNA

synthesizedin

BHK(Sin-1)

cells may result from

thesynthesis of differentDI RNAsin different

cells. There are atleasttwo reasons to believe

thatthepopulation of

BHK(Sin-1)

cellsis

het-erogeneous: the

heterogeneity

of the DI RNA

pattern, and the

discrepancy

between the

per-centof cells scoredasinfectedby

autoradiogra-phyand the percent scored as infected by

im-munofluorescence. Onepossible explanation for

thisdiscrepancyis that thelevel of viralantigen

producedin acell reflected theextent towhich

viral mRNA synthesis could be suppressed in

thatcellbyaparticularDI RNA.

Onegoalofourpresentstudywasto investi-gate the role of DIparticlesofSindbis virus in

persistentinfections. When BHK cellswere

in-fected with the cloned stock of Sin-1 virus at

37'C, a large fraction of the cells survived the

infection andgrewout toformastable infected culture in whichmostof thecells contained viral antigens and interferon was not detectable. Dur-ing this period little or no DI RNA synthesis

was detected, demonstrating that DI particles

J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:10.504.103.224.71.259.2]

(11)

werenotrequired duringthe initialevents

lead-ingto the establishment ofpersistentinfection.

To establish apersistently infectedculture it

is essentialtoobtainabalance betweensufficient

viral replication to maintain the virus in the

culture and ceii survival. DI particlescan

pro-vide one means of protecting cells from the cytopathic effects of infection. They, or some

otheragent (suchasinterferon) which would be

capable of dampening an infection, would be

required toinitiate apersistent infection when

the standard virusisparticularly virulent. Thus,

DI particles were required in establishing

per-sistent infectionwithouroriginal standard

Sind-bis virus, andboth DI particles and interferon

have been important in establishing persistent infections withVSV (6, 17, 21, 29). Infection of

BHK cells by Sin-1virus at 37°C orby rabies

virus, asshownby Holland etal. (6), was

suffi-cientlynoncytopathicsothatapersistent

infec-tion could be established in the absence of DI particles or interferon. In boththese cases, DI particles arose in the culture soon after the

infected cellswere stabilized andhadbegunto

grow.

In discussions ofpersistent infections a

dis-tinction has been madebetween persistency at

thepopulation level,inwhichonlya small

frac-tion of the cells are infected, and a persistent

state in which most ofthe cells contain viral antigens (24; J. J.Holland, S.I. T. Kennedy, B.

L. Semler, C. L. Jones, L. Roux, and E. A. Grabau, Compr. Virol., inpress). In theformer

case, mostof thecells areresistant toinfection

due to thepresence ofinterferon or someother meansofpreventingviralreplication. Hollandet al. have proposed that thelatter case is one in

which DIparticles play animportant role. Our

finding that the majority ofBHK(Sin-1) cells were stably infected and did notundergo crisis

even at30°C suggeststhat suppression of viral replication byDIRNAisanimportant factor in

protecting these cells from environmental

changesthatcould enhance viralreplicationand lead tocelldeath. Our future studiesof

persist-ent infection by Sindbis virus in different cell lines will address the question of whether DI

particles arealways associated with stable

per-sistent infectionscausedbythisvirus.

ACKNOWLEDGMENTS

This research was supported by Public Health Service researchgrantAI-11377 from theNationalInstituteofAllergy and Infectious Diseases.R.R.wassupported by Public Health Servicetraining grantGM-07067from theNationalInstitute ofGeneral MedicalSciences.

LITERATURE CITED

1. Chambers, V. C. 1957. The prolonged persistence of westernequine encephalitisincultures ofstrain Lcells.

Virology 3:62-75.

2. Eaton, B. 1977.Evidence for thesynthesisof defective interferingparticlesby Aedesalbopictuscells persist-ently infectedwith Sindbis virus.Virology77:843-848. 3. Fan, D. P., and B. M. Sefton.1978.The entry into host cells of Sindbis virus, vesicular stomatitis virus and Sendai virus. Cell15:985-992.

4. Graham, A. F. 1977.Possible roleof defective virusin persistent infection, p.445-450.InD.Schlessinger(ed.), Microbiology-1977. American Societyfor Microbiol-ogy,Washington, D.C.

5. Holland, J. J.,E. A.Grabau,C.L. Jones, andB.L. Semler.1979.Evolution ofmultiple genome mutations during long term persistent infectionby vesicular sto-matitis virus.Cell16:495-504.

6. Holland,J.J.,L.P.Villarreal,R. M.Welsh,M. B. A.

Oldstone,D.Kohne,R.Lazzarini,and E. Scolnick.

1976. Long-term persistentvesicular stomatitis virus and rabies virus infectionof cells in vitro.J. Gen.Virol. 33:193-211.

7. Huang, A. S., and D. Baltimore.1970.Defective viral

particlesand viral disease processes. Nature(London)

226:325-327.

8. Inglot, A.,M.Albin,and T.Chudzio.1973.Persistent

infection of mouse cells with Sindbis virus: role of virulence ofstrains,auto-interferingparticlesand inter-feron.J.Gen. Virol. 20:105-110.

9. Jacobson, S., F. J. Dutsko, and C. J. Pfau. 1979. Determinantsof spontaneous recovery andpersistence

inMDCK cells infected withlymphocytic choriomen-ingitis virus. J. Gen. Virol.44:113-121.

10.Kennedy, S. I. T., C. J. Bruton, B.Weiss, and S.

Schlesinger. 1976. Defectiveinterfering passages of

Sindbisvirus: nature of the defective virion RNA. J. Virol.19:1034-1043.

11. Logan, K. B. 1979. Generation ofdefective interfering particlesofSemliki Forest virusin a clone of Aedes

albopictus (mosquito)cells. J. Virol.30:38-44.

12. McMaster,G.K.,andG. C. Carmichael.1977.Analysis

ofsingle- and double-stranded nucleic acids on

poly-acrylamideand agarosegelsbyusingglyoxaland

acri-dineorange.Proc.Natl. Acad. Sci.U.S.A. 74:4835-4838. 13. Nishiyama, Y. 1977. Studies of L cellspersistently in-fectedwithVSV: factors involved in theregulationof

persistentinfection. J.Gen. Virol.35:265-279.

14. Nishiyama,Y.,I.Yasuhiko,and K. Shimokata.1978.

Properties of the virus selectedduring persistent infec-tionof L cells withVSV. J. Gen. Virol.40:481-484. 15. Pfefferkorn, E. R., andB.W.Burge. 1967. Genetics

and biochemistry of arbovirus temperature-sensitive

mutants, p.403-426.InJ.S. Colterand W.Paranchych (ed.),Themolecularbiology ofviruses.AcademicPress Inc.,NewYork.

16. Preble,0. T.,and J. S.Youngner.1975.Temperature sensitive mutantvirusesand theetiologyof chronic and inapparent infections. J. Infect. Dis.131:467-471.

17. Ramseur,J.M.,and R. M. Friedman.1978.Prolonged

infection of L cells with vesicular stomatitis virus. De-fectiveinterferingforms andtemperature-sensitive mu-tantsasfactors in the infection. Virology85:253-261.

18. Riedel, B., and D. T. Brown. 1979. Novel antiviral

activity found in the media of Sindbis virus-persistently infectedmosquito (Aedes albopictus) cell cultures. J. Virol. 29:51-60.

19. Roux, L., and J. J. Holland. 1979. Role of defective interferingparticlesofSendai virus in persistent infec-tions.Virology93:91-103.

20. Schwobel,W.,and R.Ahl.1972.Persistence ofSindbis

virus in BHK-21 cell cultures. Arch. Gesamte Virus-forsch. 38:1-10.

21. Sekellick, M.,and P.I.Marcus. 1979. Persistent

infec-tion.II.Interferon-inducing temperaturesensitive mu-tantsasmediators ofcellsparing: possiblerole in per-VOL. 33,1980

on November 10, 2019 by guest

http://jvi.asm.org/

(12)

sistentinfectionbyvesicular stomatitis virus.Virology 95:36-47.

22. Shenk, T., K. A. Koshelnyk, and V. Stollar. 1974. Temperature-sensitive virus from Aedes albopictus cellschronically infected with Sindbis virus. J. Virol. 13:439-447.

23. Spandidos,D.A.,and A. F. Graham. 1976. Generation of defectivevirus after infection ofnewbornratswith reovirus. J. Virol. 20:234-247.

24. Walker,D. L. 1964. The viral carrier state in animalcell cultures.Prog.Med. Virol.6:111-148.

25. Weiss, B.,D.Goran,R.Cancedda,and S. Schlesin-ger. 1974. Defective interfering passages of Sindbis virus:natureof theintracellular defective viral RNA. J. Virol. 14:1189-1198.

26. Weiss, B.,and S.Schlesinger. 1973. Defective interfer-ing passages of Sindbis virus: chemical composition, biologicalactivity,and modeof interference. J. Virol.

12:862-871.

27. Welsh, R. M., and M. B. A.Oldstone. 1977. Inhibition of immunologicinjury of cultured cellsinfected with lymphocytic choriomeningitis virus: role of defective interfering virusinregulating viral antigenicexpression. J. Exp. Med.145:1449-1468.

28. Youngner, J. S., E.J.Dubovi,D.0.Quagliana,M.

Kelly, and0.T.Preble. 1976. Role of

temperature-sensitivemutantsinpersistent infections initiated with vesicularstomatitis virus. J. Virol. 19:90-101. 29. Youngner, J.S.,0.T.Preble, andE. V.Jones.1978.

Persistentinfection of Lcellswith vesicularstomatitis virus: evolutionof virus populations. J. Virol.28:6-13. 30. Youngner, J. S., and D.0.Quagliana.1975.

Temper-ature-sensitivemutantsisolated from hamster and

ca-ninecelllines persistentlyinfected with Newcastle dis-easevirus.J.Virol. 16:1332-1336.

J. VIROL.