Defective interfering influenza viruses and host cells: establishment and maintenance of persistent influenza virus infection in MDBK and HeLa cells.

JOURNAL OF VIROLOGY, Dec. 1980,p.847-859

0022-538X/80/12-0847/13$02.00/0 Vol.36,No.3

Defective

Interfering Influenza

Viruses and Host Cells:

Establishment

and

Maintenance

of Persistent Influenza Virus

Infection

in

MDBK and HeLa Cells

BARUN K. DE ANDDEBI P. NAYAK*

Department of Microbiology and Immunology, University of California-Los Angeles School ofMedicine, Los Angeles, California90024

WSN (HONI) influenza virus upon undiluted passages in different species of

cells, namely, bovine kidney (MDBK), chicken embryo (CEF), and HeLa

cells,

producedavarying amount ofdefective interfering (DI) viruswhich correlated

well with theability ofthe species of cell to produceinfectious virus. However,

thenatureof theinfluenzaDI viralRNA produced from a single clonal stock was

essentiallyidentical in allthreecelltypes, suggesting that these cells do not exert

agreatselectivepressurein the amplification of specific DI viral RNAs either at

early or late passages. DI viruses produced from one subtype (HON1) could

interfere with the replication of infectious viruses belonging to other subtypes

(HlNl, H3N2). DI viralRNAs could alsoreplicate with the helper function of

othersubtype viruses. The persistent infection of MDBK and HeLa cells could

be initiated by coinfecting cells with both temperature-sensitive mutants (ts-)

and DI influenza viruses. Persistently infected culturesat early passages (up to

passage 7) showed a cyclical pattern of cell lysis and virus production (crisis), whereas, at later passages (after passage 20), they produced little or no virus and

wereresistant toinfection byhomologous virus but not by heterologous virus.

Themajority of persistently infectedcells,however, contained the complete viral

genome since they expressed viral antigens and produced infectious centers.

Selectionof aslow-growing temperature-sensitive variant rather than the

pres-ence of DI virus or interferon appears tobe critical in maintaining persistent

influenzainfection in these cells.

Over the last fewyears we have studied the

properties ofvonMagnus influenza viruses (37)

produced in MDBK cells (25a). Like other

de-fective interfering (DI) viruses, these particles

appearnoninfectious, replicatein the presence

ofhelper infectious virus, and interferewiththe

replication ofhelpervirus. Furthermore, these

influenza DIparticlesalsopossess, inadditionto

viral RNA (vRNA)segments,smallerRNA

mol-ecules (DI RNA) of vRNA origin which are

absentininfectious viruspreparation (6, 17,25,

26).These DI RNA segmentsarise fromspecific

vRNAsegments(mostlyfrom Pgenes) by

inter-naldeletion(7, 8,25)andareresponsibleforthe

interferingproperty(18).

TheproductionofDIvirusesdependsonthe

natureof the host cell. With vesicular stomatitis virus (VSV), Holland et al. (14), and more

re-cently Kang et al. (19), have shown that the

sameclonal stockproduces different DI viruses

in different cell types, although the same DI

virus isproducedinagivencell type. Stark and

Kennedy (21, 36) also reported a strong host

influenceintheproductionofSemlik Forest DI

virus. The hostrange of influenza viruses varies

widely (2, 16, 32-34). WSN virus (HON1) has

been shown to possess adifferentialgrowth

po-tentialindifferenthostcells.Forexample, WSN

virus produceslargely infectious virus inMDBK

cells,predominantly incomplete virusinchicken

embryonic fibroblasts (CEF),andonlyabortive

infection with littleor novirus inHeLacells(5).

However, little information is available on the

role of different species of host cells on the

productionof influenza DI viruses.

Since DI viruses appearto reducethe

repli-cation ofinfectiousvirus andtherebyinhibit the

cellkilling effectofhomologouslytic virus, they

have been implicatedinestablishingand

main-taining persistentinfection both in vivo(15)and

in cultured cells for a number ofviruses, e.g.,

VSV(13), rabies virus(20),reovirus(1),measles virus(29),Japaneseencephalitisvirus(31), Sen-dai virus(30),lymphocytic choriomengitisvirus (28, 39), Sindbis virus (38), and Semliki Forest virus (23). In addition, temperature-sensitive

mutation andinterferon have been showntoaid

inestablishing thepersistently infected culture

in vitro (1, 11,41). Althoughthe mechanism of viral persistency has been studied extensively

847

on November 10, 2019 by guest

http://jvi.asm.org/

848 DE AND NAYAK

for a number of the viruses referenced above, verylittle information isavailable forinfluenza viruspersistent infections either innature orin cellculture.

Inthis paper we haveinvestigated the role of host cells, e.g.bovine (MDBK), chicken (CEF), and human (HeLa) cells, on both the amount and the nature of the influenza DI virus pro-duced in these cells. We have also analyzed the role of DI influenza virus in establishing and maintaining persistent infection of MDBK and HeLacells.

MATERLALS AND METHODS Cells and viruses. MDBK (bovine kidney), MDCK (caninekidney), and HeLa cellsweregrown

andmaintained inEagle minimal medium containing

10% heat-inactivated fetal calf serum (26). Chicken fibroblast(CEF) cultureswereprepared from 11-day-old embryonated eggs and grown in Eagle minimal mediumcontaining 10% heat-inactivated fetal calf se-rum and 10%tryptose phosphate broth. A/WSN/33 (HON1),A/USSR/90/77 (H1N1),A/PC/1/73 (H3N2), andA/Vic/3/75 (H3N2) influenza virus strains were

usedintheseexperiments. Cloneswereisolatedafter six successiveplaque purifications,in MDBKfor WSN

virusandinMDCKcells forotherviruses(26).Clonal stocksweregrownin MDBKcellsforWSN virus and

in MDCK cells for othervirusesasreported earlier

(17). Bothwild-type (ts+)and agroup II

temperature-sensitivemutant(ts52)ofWSN viruswereused(26). Production ofinfluenza DI virus in different cells.Samplesofclonal stocks of WSN virusprepared

in MDBK cells were passed serially, undiluted, in MDBK, HeLa,orCEFcellstoprepareDI virus (26). Clonal stocks ofts52initiallygrown in MDBKcells containednodetectableamountofDIvirus,as deter-mined by infectious center reduction assay and by RNAanalysisusing polyacrylamide gel electrophore-sis(17).

Long-term passage of DI virus. These

experi-ments were initiated with DI virus produced at the

undilutedpassage 2 (p-2).Subsequently,ateach pas-sage,cellswereinfectedwith DIvirusproducedatthe precedingpassageaswellasadditionalinfectious stock

virus. Briefly,MDBK, HeLa, and CEF cultures (150

cm2) wereinfected with 5 ml of DIvirusproducedat

theprecedingpassage, plus0.5mlofinfectious virus

(1PFU/cell).Viruswasharvestedat 14hpostinfection by subjectingthe cellsand culture supernatant to two

cycles of rapidfreezingandthawingandwasused as thesourceof DIvirus inoculum forthe next passage. Thisprocedureofinfectionwith DI andadded infec-tious viruses was continued forup to30 passages in

respectivecell lines. Themultiplicityof infection of DI virus in theselong-termpassages variedfrom2 to 20 defective interferingunits (DIU)per cell atdifferent passages.

Isolation and analysis of 32P-labeled viral RNA.MDBK, MDCK, CEF,andHeLacells(150-cm2 flasks)wereinfected with a mixture ofinfectiousvirus (1 PFU/cell) and DI virus (2 DIU/cell). The labeled viral RNA was isolated (27) and analyzed by slab gel

electrophoresis asdescribed before (17).

Electropho-resis wascarriedouteitherat4°Cor at room temper-aturefor21husing2.2%polyacrylamide-0.8%agarose

gelscontaining6Murea,0.036 MTris-hydrochloride (pH 8.1), 0.03 M NaH2PO4, and0.001 M EDTA as

described byFloydetal.(10).

Oligonucleotide mapping.32P-labeled infectious

orDI viral RNA (approximately 5 x 106to20x 106

cpm) waselectrophoresed inasingle lane ofaslabgel

asdescribed earlier(8). The positions of viral and DI

RNA segments werelocated fromanautoradiograph obtained after30min to 1hofexposure of Du Pont Cronexfilm. RNAwaseluted from thecorresponding gelsegmentsby theprocedure of Maxam and Gilbert

(22). Labeled RNA segments were digested with RNaseT,andanalyzedby two-dimensionalgel elec-trophoresisusingamodifiedprocedure of DeWachter andFiers(9). Thespecificconditions of electrophore-sis used foroligonucleotide mapping have been de-scribedpreviously (8).

Establishment of persistent infections in MDBK and HeLa cells. MDBK and HeLa cellswere

infected with either the infectious virus (1PFU/cell) aloneor amixtureof infectious (1 PFU/cell) andDI virus (p-2for MDBK and p-2forHeLa,2DIU/cell). At 14hpostinfection, infected cultures werewashed

thoroughly withphosphate-buffered saline (PBS) free

of Ca2" and Mg2" (PBS-) and overlaid with fresh

mediumwhich wasreplacedeveryday. Surviving cells eventually formed confluent monolayers of persist-ently infected (Pi) cells. These cellsweresubcultured twice a week at a ratio of 1:4. These Pi cells were

termedMP1,MP2, and MP3orHP1and HP2 when theywere ofMDBK orofHeLa cell origin,

respec-tively. Occasionally these Pi cells underwent crisis with extensivecytopathic effect andvirusproduction, butthey recovered after frequent changes of growth medium.

Challenge with infectious homologous and heterologous viruses. Picell cultures (MP1, MP2, MP3, HP1, and HP2) wereinfected with infectious virus(2to 10PFU/cell)atdifferentpassages. At14 h

postinfection the flaskswerechecked for cytopathic effect, and the supernatantswereanalyzed for PFU, DIU, andhemagglutinationunits(HAU)permilliliter

asdescribed earlier (17). For heterologous virus chal-lenge, Pi cellswere infected withNewcastle disease

virus (2to 10PFU/cell) andanalyzedforvirus pro-duction.

Infectiouscenterassay.Picellsweretrypsinized

and added at different concentrations to fresh MDBK

monolayers.After1 h at37°C, the agaroverlay me-dium was added andkepteither at34°Cor at39°C;

after4to5days, plaqueswerecounted (17).DIUper milliliterweredeterminedbyusingtheinfectious cen-terreductionassay as described previously (17).

Indirect fluorescent-antibody staining.

Unin-fected cells (MDBK and HeLa), acutely infected

MDBK and HeLacells, and a number of Pi cellsof different passages were grown onmicroscopic slides for 24 h and fixed with acetone in cold (-20°C) as

described by Roux and Holland (30). Subsequently theywerecovered with asolutionofanti-WSNrabbit serum(1:20dilutioninPBScontaining 500 hemagglu-tination inhibition units per ml) for 30min at room J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

DEFECTIVE INTERFERING INFLUENZA VIRUS 849

temperature, washed thoroughly, andfinally stained with fluorescein-labeled goat anti-rabbit 7S globulin (1:25in PBS-; GIBCO) for 30 min at room tempera-ture. Theselabeled cells were further washed twice with PBS- and examined under a microscope equipped withaUVlight source.

To detect the presence of viral antigen on the cell surface, live cells were used.Cells were trypsinized and incubated for 2 to 3 h at 37°C with the original conditioned growth medium. These cells were then washed with PBS- and incubated with anti-WSN an-tibody (1:20inPBS)at0°C for 30min.Subsequently they were washed in PBS- and stained with fluores-cein-conjugated goat anti-rabbit 7S globulin for 30min

at0°C. The suspended cells were washed twicefurther

withPBS- and examined for immunofluorescence.

RESULTS

Production of DI viruses in MDBK, CEF,

and HeLa cells. To determine the effect of

differentspecies of host cellsontheproduction

ofDIvirus,infectiousts52 WSN clones,sixtimes

plaque purified, were used to prepare clonal

stocks. Samples of clonal stock ofts52 (clone4;

2.8 x 107PFU/ml, 2,048HAU/ml) were

inocu-lated into MDBKandCEF (0.01PFU/cell)and

HeLa cells (1 PFU/cell). At low multiplicity

(0.01 PFU/cell), Helacellsproducedverylittle

virus. Totalvirus washarvested from these

in-fected cultures at 14 h postinfection and was

calledp-0. p-1viruswasproducedinrespective

cells by usingp-0 virus at 20 PFU/cell.

Subse-quently,5ml of undiluted viruspreparationwas

passedserially for the production ofDIvirus in

respective cells (26). Infectious virus (PFU per

milliliter), DI virus (DIU per milliliter), and

total virusproduction (HAUpermilliliter)were

assayedatdifferent passages by theprocedure

describedpreviously(17, 26).

DI virus was notdetectedat p-0 inMDBK,

CEF, or HeLa cells. At p-1, MDBK cells

pro-duced4 x 105DIU/ml,whereasboth CEFand

HeLa cells produced 1 x 105 DIU/ml.

Subse-quently,inseriallyundiluted passages(p-3to p-5), more DI virus wasproducedinall threetypes

of cells. However, both CEF and HeLa cells

produced only a moderate amount of DI virus

(CEF, 2 x 105 DIU/ml atp-4; HeLa, 2.5 x 105

DIU/mlatp-3) whencomparedtothe maximum

productionin MDBK cells(2.4x 106 DIU/mlat

p-5). Thetotalvirusparticleyield,asdetermined

by HAU,wasreduced inallcellsupon undiluted

passages. As the production of DI virus

in-creased,theoverallyield of infectious virus

de-creased in all threespeciesofcells. Atp-4,HeLa cells contained a detectable hemagglutination

titer(16HAU/ml),althoughnoDIorinfectious viruscould be detected. Theproductionof

hem-agglutinating antigen at p-4 may represent an

abortivereplicationofWSNvirus in HeLacells

(e.g.,intracytoplasmic inclusions[4]).

Inexperiments involving long-tern passage of

DI virus,cellswerecontinuously coinfected with DI virus andadditional infectious virus at each passage. Theresults (Fig. 1)show that the

over-all patterns of the production ofDI virus and

infectious virus, as well as the hemagglutinin

titer, were essentially similar inall three types

ofcells. Therewas agradual decrease of

infec-tiousvirusyield, accompanied by an increase in

DIvirus production, irrespective of the nature

ofthe cell. Again, the greatest amount of both

infectious and DI virus was produced in MDBK

cellsascomparedtothatproducedinCEFand

HeLacells.Finally, the DIU/PFU ratio showed

asteadyincrease in allcells(from0.125inp-3 to

19inp-10 for MDBK; from 0.005 in p-3 to 5.5 in

p-10for CEF; from0.007 in p-3 to 11 in p-10 for

HeLa cells). After passage 10, the same DIU/

PFUratiowasmaintainedup top-30 by

contin-ued coinfection with DI and infectious viruses

(data not shown). Thus, it appears that after

continued coinfection with infectious and DI

viruses over many passages, an equilibrium in

the production of infectious and DI viruses is

established.

Nature ofDI virus produced in MDBK,

HeLa, and CEF cells. The natureofRNA of

DIvirusreleased bythesecellsatdifferent

pas-sages was analyzed bypolyacrylamide gel

elec-trophoresisandoligonucleotidemapping. Figure

2showsthat the [32P]RNA ofpurified virus at

p-O contained only eightvRNA segments

with-out any detectable amount of DI virus in all

three cell species. However, the virus at the

undilutedp-3,atcoinfected p-9, andatp-30 from

all cell types contained a number of DI RNA segments in addition to the eight standard

vRNA's,althoughpresent invarying ratios.

Fur-thermore, we have rather consistently found

here(Fig.2and4), andpreviously (7,8, 18,26),

that the molar ratio ofM generelativetoother

viralgenesincludingthe NS gene isincreasedin DI particles. This would suggest either that someof the DIparticlescontainonlythe M gene or thatDIparticles arepolyploid in respect to the M gene. Anotheralternative, although

un-likely,isthepresenceofaDIRNAcomigrating

with the M gene. At presentwe have no idea which, if any, of thesepossibilitiesistrue.

The DI RNAsegmentsproducedatp-3, p-9,

and p-30 in all cell types were essentially the same in nature. At least six distinct DI RNA segments were found. At p-3, DIa migrating

faster than the NS genewas mostconspicuous

inall celltypes(Fig. 2A).Atp-30,DIadecreased

and a smaller DI RNA (DIb) became more

VOL. 36,1980

on November 10, 2019 by guest

http://jvi.asm.org/

850 DE AND NAYAK

I

E

LL U.

8

E

I-102

10

II

-1

al

10 X

-2 10

FIG. 1. Analysis oftotal infectious virus (PFUper milliliter, 0), defective interfering virus (DIUper milliliter,A)andhemagglutinin (HA Uper milliliter, 0) produced bythesameclonal isolate(no. 4)inMDBK, CEF, and HeLa cells during continued coinfection using infectious virus and DI virus frompreceding passagesasdescribed in thetext.ThecorrespondingDIU/PFU(A)ratioateach passage isalsogivenunder

thesameconditionsforthe threecell types.

prominent in all three types ofcells (Fig. 2B).

At p-9, anintermediate resultwasobtained(data

not shown). Thus it appears that after many

passages theearlyDI virus isgradually replaced byaDI viruswith asmallerDI RNA segment.

Further sequence relationship among the DI

RNAs wasdeterminedbyoligonucleotide

map-ping (8). DIa isolated from MDBK cells

con-tained spotscharacteristic ofthe P3 (polymer-ase) gene andtherefore originates from the P3

gene(Fig.3AandB).Similarly,DIafrom HeLa

cells and CEF cells containedspotsidentical to

those present in DIa of MDBK cells and is,

therefore, also of P3 origin (data not shown).

Both DIb RNAs isolated from p-30 ofMDBK

and HeLa cells produced essentially identical

oligonucleotide maps and are subsets of DIa

(Fig. 3C). These results suggest that both DIa

and DIb areinterrelatedand originate from the same P3 gene. These results also confirm our previous observation that the majority of influ-enzaDIRNAsoriginate from one of the polym-erasegenes (8).

DI virus-mediated interference against

different subtypes of influenza A viruses.

Interference mediatedbyDIviruses appears to

behighly specificandeffectiveagainst

homolo-gous or closely related viruses but not against

heterologous viruses. With VSV,heat-resistant

(HR)DIparticlesof the Indianaserotype

inter-fere poorly withNewJerseyMserotype.

How-ever, DI NewJerseyMvirusinterferesequally

with M and Ogden virus (24). Since a large

number ofserotypesof influenza virusare

avail-able, we tested the interfering ability ofWSN

DI (HON1) against USSR (HlNl), Port

Chal-mers(H3N2), and Victoria (H3N2) viruses. We

determined theinterfering titer of WSNDIvirus

against these virusesby usingadirect infectious

center reduction assay as described previously

(17). Theresults (Table 1) show thatessentially

thesametiter ofinterfering activityofWSNDI

virus (DIUper milliliter) wasobtained against

different subtype A viruses. Additionally, in a

separateexperimentwefound that the yield of

different subtypeviruses was reduced by

coin-fecting with WSN DI virus (data notshown).

Theseresults show thatWSNDI virusisequally

effective againstall influenzaAviruses.

Prelim-inarydatasuggest thatWSNDIis noteffective

againstinfluenza B viruses.

Tofurther determine whether WSN DI can

replicate by using thehelperfunctionof other influenza Aviruses,MDCKcellswerecoinfected with both DIandinfectiousviruses and labeled

with

32pO4

analyzedinpolyacrylamide gels. Figure4shows

that inspiteofrepeated cloning allthree viruses containedvisibleDI RNAsegments. These re-sults

confirn

on November 10, 2019 by guest

http://jvi.asm.org/

DEFECTIVE INTERFERING INFLUENZA VIRUS

A OCO 0CO OC°

CL l a. a- a-Q

NS 2S

DHaA up

B 0

I--.

.ow.

_o~--f #

Nw

--

Dib-FIG. 2. Analysis of[3'P]RNAofstandard virusatp-Oand DI virusesproduced atp-3(A) andp-30(B) duringcontinuedcoinfectioninMDBK, CEF,and HeLa cellsusingthesameclonal stockvirus. Cellswere

infected with a mixtureofstandard virus(1 PFU/cell) and DI virus (2 DIUlcell). 32P-labeled RNA was

isolatedfrom purifiedvirus andanalyzed by polyacrylamide gel electrophoresis for21 hat4°C using180 V

(17).Arrows indicate thepositions ofsomeofthe DI RNA segments. DIa and DIbappear to be themost prominent DI RNAs in allthree typesofcellsatp-3and p-30, respectively. M(PO), M(P3), M(P30), C(PO), C(P3), C(P30), H(PO), H(P3), H(P30)representMDBK, CEF,and HeLa cells atp-0,p-3,andp-30, respectively.

all, wild-type viruses contain DI viruses.

Whether these visible DI RNA segments are

newly formed during plaque isolation and

pro-duction ofclonal stockor arepre-existing

con-taminants is being currently investigated. Our

gel analyses showed that WSN DI RNA can

replicateefficientlyin the presence ofHlNland

H3N2helperviruses andcaninterfere withthe

replicationof vRNA segments.AnalysesofRNA

by polyacrylamide gel electrophoresis at two

temperatures (room temperature and 4°C)

clearlyshowed that themajority ofvRNA

seg-ments,asexpected,arepredominantlyofhelper

virus origin (Fig. 4). A reduction of specific

vRNA segments ofchallenge virus as reported

previously (5, 17, 25) was also observed (e.g.,

band2from top ismissinginUSSR xDI,lane

6,Fig. 4).When thehelpervirus also contained

DI RNA segments, both WSN DI RNA and

helper virus DI RNA could replicate in

coin-fected cells. Similarly,DIvirusmade from Port

Chalmers(H3N2)virus couldalso interfere with

WSN, USSR,and Victoria subtypes, and Port

Chalmers DI RNA could replicate with the

helper function of these viruses (data not

shown). These results suggest that WSN DI

virus can interfere with as wellas replicate in

the presence of other subtypes ofinfluenza A viruses.

Establishment ofPi cultures. Since

per-sistentinfectionof cultured cellswith manylytic

viruses has been established, itwas of interest to determine whether influenza virus can also

establishpersistent infectionandtoanalyzethe

conditions thatfavortheestablishment of

per-sistent infection.

No Pi cultures of MDBK (Pi MDBK) and

HeLacells(Pi HeLa)could be established when

cells were infected with either wild-type (ts+)

virus (0.1 to 1 PFU/cell) or temperature-sensi-tivemutantalone(0.1to1PFU ofts52percell).

Picultures couldnot beestablishedeven when cellswerecoinfected with ts+ and DI virus. On

the other hand, Pi MDBK (MP1, MP2, and

MP3)and Pi HeLa cell(HP1andHP2) cultures could be routinely obtained by infecting cells

VOL. 36,19860 851

0 0.0

CO CY) CY)

13- a_

a_-m

C)

2

on November 10, 2019 by guest

http://jvi.asm.org/

852 DE AND NAYAK

with both an infectious temperature-sensitive

(ts52) virus and DI virus simultaneously. The source of DIvirus, whether obtained from ts+ or ts-mutants, did not affect the establishment of Pi cultures.

At earlypassages (<p-7),cells showed a

typi-calcyclicalpattern of cytopathiceffect and virus

production (crisis) usually immediately after

subculture. Pi cultures subsequentlyrecovered

upondaily washing with fresh medium and

be-*

tel.

*w

-~~

17

. .

A

%.*4-H11

!-

7

%ft\1 0-

Ell

came confluent again. These Pi cultures

ap-peared more rounded and refractile than the

uninfected cells and wereeasily dispersed with

trypsin.

Presenceand expression of viralgenome

in Pi cultures. Pi cultures, once established,

can grow and multiply as well as produce a

relativelysmallamount of virus. There are two

likely explanations fortheseresults: (i) asmall

percentage ofthecells in cultures are infected

*;s

J8--J

4-

t-f

1

1

E

7-a

x

P3

X

DDI

a

(MDBK

)

a

x

*-H11

*

Fi

x

Dlb(MDBK)

FIG. 3. Oligonucleotidemaps of P3, DIa,and DIbRNAs. Individual[32P]RNA segments were isolated,

digestedwithT,ribonuclease,andanalyzed by two-dimensional gel electrophoresis (8). Detailed oligonucleo-tideanalyses ofindividual genes have been reportedpreviously (17, 25a). Only the spots common among P3,

DIa,and DIb areidentified. x's (bottomandtop)represent xylene cyanol and bromophenol blue,respectively.

J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

DEFECTIVE INTERFERING INFLUENZA VIRUS

andare producing virus at a given time or (ii)

the majority, if not all, of the cells are infected

andcontain viralgenome, but either some orall

of them areproducingsmall amounts of virus. A number ofexperiments describedbelowsupport

the latter explanation. First, the Pi cultures

(MP1, MP2, MP3, and HP1, HP2) were resistant

tosuperinfectionbyhomologousbut not

heter-ologous viruses. Both early- (<p-16) and

late-passage (>p-30) Pi MDBK and Pi HeLa cells

showed little cytopathic effect (<15%) after

su-perinfection with WSN virus. Control MDBK

celLs, wheninfected with either ts52 or ts+ WSN

virus, produced 2,048 HAU/ml and 300 x 106

PFU/ml, whereas Pi MDBK cells (MP2 and

MP3)produced only32HAU/ml and3 x 103to

5 x 103PFU/ml uponsuperinfection (Table 2).

Similarresults were obtained with HP1 and HP2

cells as compared tonormal HeLa cells. How-ever, whencells were challengedwith a

heter-ologous virus (Newcastle disease virus), both

normal cells and Pi cellswere equally sensitive.

Over 80% of cells showed cytopathic effect

within24hafterinfection, and high HAU(4,096

to8,194HAU/mlinMDBK;512to1,024HAU/

mlin HeLacells) and infectiousvirus titers (50

x 106to 70x 106PFU/ml for MDBK and 3 x

105 to 5 x 105 PFU/ml for HeLa cells) were

obtainedinbothnormal andPicultures.

Second, the expressionof viralantigensin Pi

cellswasexaminedbyimmunofluorescence and

electron microscopy. Accordingly, MP2, MP3,

HP1, and HP2 cultureswere examined forthe

TABLE 1. Titerof WSN DI virus measured against challengeinfectiousvirusesof different Asubtypesa

Challengevirus DIU/ml

WSN(HONi) ... ... 16.5x106

PortChalmers(H3N2) .... 7.5 x106

USSR(HlNl) ... .... 10.5x106

Victoria (H3N2) ... 7.8x 106

'Titerof WSN DI viruswasdeterminedby infec-tiouscenterreduction assay against different A sub-type viruses. WSN DI viruswas prepared by high-multiplicity passages as discussed previously (26). MDCK cellswereinfected with different dilutions of WSN DI virus andsuperinfected with4PFU/cellof challenge viruses of different subtypes. Cells were trypsinized and platedonmonolayers of MDCK cells for infectious center assay. Reduction of infectious

centersof eachchallengevirus atdifferentdilutionsof

WSN DI virus was determined and the number of DIUofWSNDI viruspermilliliterwascalculatedin

comparison with each challenge virus as described previously (17). Clonal stocks of infectious WSN

(HON1),PortChalmers(H3N2),Victoria(H3N2),and

USSR(HlNl) viruspreparationscontained30x 106 PFU/ml and 1,024 HAU/ml, 15 x 106PFU/ml and

512 HAU/ml, 17x 106PFU/mland1,024 HAU/ml, and10x106PFU/mland512HAU/ml,respectively.

Z xCCzz

co) z x cn) co D

cp Q , CL)D

D)

-5- n L

FIG. 4. Replication of WSN DI RNA with the helperfunctionof other influenzaAsubtypes: USSR (HINI),Port Chalmers(H3N2), and Victoria (H3N2). WSN DIwaspreparedfromasix-times-cloned virus by three consecutive passages. Subsequently, cells

wereeitherinfected with standard virusorcoinfected with both standard virus and WSN (p-3) DI virus. 32P-labeledviral RNAwasisolatedandanalyzed by polyacrylamide gel electrophoresis as described in Fig.2.Arrowsshowanumberofprominent DI RNA segments.In spite of repeated cloning, standard Port Chalmers and Victoria viruses also contain DI RNA segments.

expression of viral antigens by indirect

immu-nofluorescence using antisera against WSN

vi-rus.Approximately70 to80% of thecells showed

apositive reaction. Both acetone-fixed (Fig. 5)

andlive cells(Fig. 6) demonstrated thepresence

of viral antigen in the cytoplasm and cellular

membrane ofinfectedcells.Tofurtherconfirm

the expression of viral genes, thin sections of

bothacutelyinfected and Picellswere studied

by transmission electron microscopy. Results

confirmed that althoughno viruswas released

in the supernatant,buddingvirusparticleswere

present on the membrane of Pi cells (data not shown). Finally,infectious centerassay further

demonstratedthepresence of the viral genome

in Picells. PiMDBK and Pi HeLacells,when

platedonmonolayersofuninfectedMDBKcells,

showed a 70 to 80% plating efficiency as

com-paredtothe 100% plating efficiencyofacutely

infectedMDBK and HeLacells (Table3).This

slightlylowerplatingefficiencyof Pi cellscanbe due toeither DI virus present insome Pi cells

853

VOL. 36,1980

on November 10, 2019 by guest

http://jvi.asm.org/

854 DE AND NAYAK

TABLE 2. Virusproductionafter superinfection of persistentlyinfected cellsbyhomologousand

heterologous virusesa

Yield Cells Superinfecting

(passageno.) virus(MOjb) PFU/ml HAU (X106)

HUml'

MDBK

Control WSN(2) 300 2,048

MP3(p-16) WSN(2) 0.005 32

MP3 (p-30) WSN(2) 0.0035 32

MP3 (p-44) WSN (2) 0.003 32

HeLa

Control WSN (10) 4.0 1,024

HP2(p-14) WSN(10) 0.003 16 HP2 (p-29) WSN(10) 0.002 16 HP2 (p-41) WSN (10) 0.0015 16 MDBK

Control NDV (2) 70 8,192

MP3 (p-16) NDV (2) 50 4,096

MP3 (p-30) NDV(2) 60 4,096

HeLa

Control NDV(10) 0.5 1,024

HP2(p-14) NDV(10) 0.4 1,024

HP2(p-29) NDV (10) 0.3 512

aNormalMDBKandHeLa cellsand anumber of persist-ently infected cells at different passages were superinfected withWSN virus or Newcastle disease virus (NDV) for 14 h as described in thetext.Thesupernatantswerethenharvested andanalyzed for HAU as well as PFU (for WSN virus we used MDBKmonolayersat34°C,and for Newcastledisease virus weused CEFat37°C).

bMOI,Multiplicity of infection.

preventing plaque formation (17) or somecells

notproducinganyinfectious virusatall. These

above experiments suggest that the virus

ge-nomeispresent andexpressed (at leastpartly)

in the majority of Pi cells, although complete

infectiousparticlesarenotproduced.

Nature ofvirus expressed by the Pi

cul-tures.Asdiscussedabove,littleornodetectable

infectious viruswasreleasedby the Picultures

(Table 4),althoughPicellsproduced infectious

centers onmonolayersofMDBKcells(Table 3).

Todetermine thenatureof the virusreleasedby

the establishedPicells,anumber ofexperiments

were done. In one, persistently infected cells

were trypsinized and plated on monolayers of

MDBKcells for infectious centerassay at

340C

(permissive) and

390C

tempera-ture). Results show that little or no wild-type

(ts+)viruswasreleased from Picultures (Table

3). This was not surprising since Pi cultures could beestablishedonlywith ts-mutants, and wedid notdetectanywild-typerevertants in Pi

cultures. In thesecond type ofexperiment, the

virus released byPi cellswas amplifiedby

co-cultivating with monolayers of MDBK cells.

Again,virus was releasedonly at34°C andnot

at

390C.

340C

form

plaques at 390C. Viruses isolated from Pi

cul-turesformed smallerplaqueseven at340C than

did the original clone4 ofts52virus.

Addition-ally, the clones isolated from Pi cultures

pro-ducedaloweryieldat340C.

Role ofts- mutants in establishing and maintaining persistent cultures. Our data

suggestthat both DI virus andts- mutants are

needed inestablishing Pi cultures with influenza

virus.Two typesof experimentswereperformed

tofurther evaluate the role ofts mutantsinboth

establishingandmaintainingthepersistent

cul-tures. First, MDBK and HeLa cells were

in-fected with thets- mutantalone and incubated

eitheratpermissive (340C) andsemipermissive

(3700)or atrestrictive(390C)temperatures. We

found thatover99% of thecellsweredestroyed

in 3 dayseither at 34 or37°C,whereasonly40

to 50% of cells were killed at the restrictive

(390C) temperature, using the ts- virus alone.

However, eventually all cells underwent lysis,

andno Picultureswereobtainedateither

per-missive orrestrictive temperatures. Inthe

sec-ondtype ofexperiment, Pi cultures (MP2, MP3,

HP2,HP3)that wereestablishedby coinfection withts- mutantand DIvirus and maintainedat

370C were shifted to either 34 or 390C. After

shiftdown, thesePiculturesdemonstrated

nei-ther crisisnor anyincreasedamounts of

infec-tiousvirus, as comparedtoPicultures keptat

370C.Norwasthereanyobservedchange when

thePicultureswereincubatedat390C.

Absence of detectable interferon in Pi

cultures. Culturesupernatants fromPiMDBK

and PiHeLa cells were used for preparing

ex-ogenous interferon andwereassayed for

inter-feronbyplaquereductionassay inhomologous

cells with VSV (3). There was no detectable

amount of interferon ineitherPiMDBKorPi

HeLacells. Additionally, both of these Pi

cul-tures, although resistant to homologous

infec-tion, were sensitive to heterologous infection,

namely Newcastle disease virus (Table 2) and

VSV (datanotshown). Thesetwoexperiments

suggest that neither exogenous nor endogenous interferon was amajorfactor in themaintenance

ofpersistent influenza virus infection ofHeLa

andMDBK cells.

Role ofDIvirus in themaintenance ofPi

cultures. Since DI particles have been

impli-cated inpersistentinfection,anumber of

exper-imentsweredone todetermine whether DI

vi-ruses werepresent inpersistently infected cul-tures. Accordingly, the releasedvirus was

am-plified and labeled by plating Pi cells on the

monolayerof MDBKcellsandincubating in the

presence of32PO4 medium.

[32P]vRNA

an-alyzedinpolyacrylamide agarose gels (26). The

J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

FIG. 5. Presence of viral antigen inpersistently infectedMDBK (MP3) andHeLa (HP2) afteracetone fixation. Cultures (24 h old) grown on coverslipswerefixed in acetone and examined for viral antigens by indirectimmunofluorescenceasdescribed in the text. (A)UninfectedMDBK; (B) acutelyinfectedMDBK; (C)

persistentlyinfected MDBK, MP3; (D) uninfectedHeLa; (E) acutelyinfectedHeLa; (F)persistentlyinfected

HeLa, HP2.

855

on November 10, 2019 by guest

http://jvi.asm.org/

I

I

FIG. 6. Thepresence of viral antigen on the membranes of persistently infected cells. Live cells were examined for the presence ofviral antigen by indirect immunofluorescence as described in the text. (A) UninfectedMDBK; (B) acutelyinfected MDBK; (C) persistently infected MDBK, MP3; (D) uninfected HeLa;

(E)acutelyinfectedHeLa; (F) persistently infected HeLa,HP2. 856

on November 10, 2019 by guest

http://jvi.asm.org/

DEFECTIVE INTERFERING INFLUENZA VIRUS 857

TABLE 3. Infectious center assay of Picellsa

No. of infectivecenters per Cells(passage no.) 100cells

34°C 390C

MDBKxts52b 100 1

MP1 (p-16) 70 0

MP2 (p-17) 75 0

MP3 (p-17) 71 0

MP1(p-30) 55 0

MP2(p-31) 50 0

MP3 (p-35) 60 0

HeLaxts52b 70 0

HP1 (p-14) 90 0

HP2 (p-16) 80 0

HP1 (p-41) 78 0

HP2 (p-43) 70 0

aResults were calculated from duplicate assays us-ing 400, 200, and 100 trypsinized cells. Pi cells of different passages were trypsinized, plated on fresh MDBKmonolayers(5-cmplates)indifferent numbers (400, 200,and100cellsperplate), and incubated for1

hat37°C after the addition of 5 ml of agar overlay. Theseplates were then incubated at34and390C,and plaques were counted on days 3and4.

bControl MDBK and HeLacellswereinfected with infectious ts52 virus (4 PFU/cell) for 1 h at 370C, trypsinized, and assayed by thesameprocedure.

results showthatonlystandard vRNA segments

couldbedetected,withno appreciable amount

of specific DI RNA segment in the released

virus. Thiswassurprising becauseDIvirus was

usedinestablishing the persistent infection. The

lackof DI RNAcannotbe duetotheinabilityof

DI virustoreplicateineither of thesecells, since

DI viruses can be produced and amplified in

both of these cells.

Tofurtheranalyze thepresenceofDIvirusin

Picultures, culture supernatant obtained after

amplification was analyzed for interference by

deterniining the reduction of virusyieldand the

reduction of infectious center after coinfection

with infectious virus (17). Although the virus

released from Pi cultures had a lower PFU/

HAU ratio, it did not possess any interfering

activity againsthomologous virus eitherin the

yield reduction (Table 4) or in the infectious

center reduction assay (data notshown). This

would suggest that the virus released by Pi

cultures was defective but noninterfering. We arein theprocess of further characterization of

thisvirus.

DISCUSSION

Effects of host on DI virus production.

The production and amplification of DI virus

depend partly onthe parental infectious virus

andpartlyonthe hostcells(14, 19).WSN virus

replicates most efficiently in MDBK cells and

least efficiently in HeLa cells (5). Experiments

reported here show thatboth infectious and DI

viruses are produced most in MDBK cellsand least in HeLa cells. CEFcellsproduce an inter-mediate amount of both infectious and DI vi-ruses.Such a result would suggest that the steps required in the replication for influenza DI virus

areessentiallythe same as those needed for the

replication of infectiousvirus.

The nature of DI RNA(s) generated in all

three species of cells fromasingle clonal stock

appearsidentical bypolyacrylamide gel

electro-phoresis and oligonucleotidespotanalysis. The

results, although apparently different, are not

-totally inconsistent with those observed for VSV

(14, 19). The host factor(s) influencing the

na-ture of VSV DI virus does not appear to be

absolutely selective because (i) someofthe DI

viruses produced in different species of cells

appear identical and (ii) any predominant DI virus, irrespective of its source of origin, can

replicatein agiven cell. Our results suggest that

the nature of influenza DI virus is primarily

determined by those thatarealreadypresentin

the clonal stock and that MDBK, CEF, and

HeLacells donot exert agreatselectivepressure

intheamplificationofinfluenzaDIvirus.

DI virus-mediated interference among

different subtypes of influenza viruses.DI

virus prepared from WSN virus (HON1) can

effectively interferewith thereplicationofother

influenza A viruses, namely HiN1 and H3N2

viruses. Equally, WSN DI RNAs can replicate

TABLE 4. Analysis of virus particles produced by Pi cells

Exptno. HAU/ml PFU/ml PFU/HAU ratio 1. MP3(p-20) supernatanta <2 0

2. HP2(p-21)supernatanta <2 0

3. MP3(p-20)trypsinized 256 20x 10' 0.78x 103 cellsb

4. HP2(p-21)trypsinized 512 70x104 1.3x 103

cells'

5. MP3supernatant(expt 3) 512 15x106 3.0x 104

xts52c

6. HP2supernatant (expt 4) 512 30x106 6x104 xts52c

7. ts52aloned 1,024 90 x106 8.8x 104

aSupernatantsof MP3 (p-20) andHP2(p-21) were

ana-lyzedfor HAUandPFU at340C.

bMP3and HP2 cellsweretrypsinized, and5x 106cells

wereaddedtothemonolayerof fresh MDBK cellsand incu-batedat 34°C for 48 h. Supernatants wereharvested and analyzed for HAU and PFUat34°C.

cMonolayersof MDBK cells (150cm2) were coinfected with 5 ml of the supernatant (experiments 3 and 4) and infectious ts52virus(1 PFU/cell). The virusyieldwas har-vestedat 48hand34°CandassayedforHAU and PFU.

dMDBK cellswereinfectedwith ts52 virusalone(1PFU/ cell)for 48 h at34°C.

VOL. 36,1980

on November 10, 2019 by guest

http://jvi.asm.org/

858 DE AND NAYAK

in the presenceof infectious virus of other

sub-types.These resultssuggestthat thegene

prod-uctsrequired for thereplication of DIRNAs,as

wellasthose for the interference ofvRNA's,can

functionequallywell amongthe viralsubtypes.

These results are not surprising since genetic

reassortmentexperiments amongdifferent

sub-types suggest that the genes among different

influenza A subtypes can be exchanged freely

and thatgeneproducts ofonesubtypeare

func-tional with othersubtypes. Our results extend

the idea offunctional genetic exchange among

the viralsubtypestotheDIvirus-mediated

in-terference.

Establishment and maintenance of

per-sistent influenza infection. The results

re-ported here show that influenza viruses under

certain conditionscanestablishpersistent

infec-tion ofMDBK and HeLacellsin culture. The

presence of both DI virus andts- mutants ap-pears to be critical in establishing the initial

condition favorable for persistent infection. As

yet we have been unable toobtain Picultures

byusing eitherts+ virusalone, ts- virusalone,

or ts+ virus with DIvirus. A similar initial

re-quirement of bothDIvirus andts- mutanthas

been reportedfor VSVand BHK-21 cells (13),

whereas DI virus in thepresenceofts+virus has beenshown toestablishpersistentinfection ofa

numberof other cell-virussystems (23, 30,38).

Whether different ratios of DIvirustots+virus

orother conditions suchasinterferontreatment

could help in establishing persistent influenza

infection incellculturesisbeingcurrently

inves-tigated.

Persistent influenza infection in MDBK and

HeLacellsappears not tobemaintainedat the

population level (i.e., a small fraction of cells

being infectedat agiventime)since themajority

ofcellscontainandexpressviralantigen(s).Nor

doesinterferonappear toplayasignificant role

in the maintenance of persistent infection of

influenzavirus in thesecells. Our resultssuggest

thatthe selection ofvirus variantswithgreatly

reduced growthcapacityat

370C

host cells is probably required for maintaining

persistent influenza infection. Repeated crises

observed in the early passages would assist in

the selection of a viruspopulationwith agrowth behaviorcompatibletocellulargrowth. The vi-ruspopulation obtained from stable Pi cultures

producedsmallplaquesand areducedyield at

340C

Whether these virus clones isolated from Pi cultureswouldbelesscytolyticand establish Pi

infectionmoreeasily, andprobably without the

helpof DIviruses,isbeingstudiedcurrently.

AlthoughDI virusappears to be critical in the

initialphaseofestablishingpersistent infection,

we wereunabletodetect either the DI virusor the DIRNAinreleased virus from Pi cultures.

During the initial phase ofinfection, DI virus is

needed to suppress the cytopathic effect of

in-fectious virus and to help the survival of host

cells during crisis. In addition, DI virusesmay

accelerate the selection ofanappropriate variant

of infectious virus (12, 35). However, once the

variant isselected andastable cell-virus

associ-ation isestablished, DI virusmay notbe needed

and thus may be eliminated from Pi cultures.

The presence of viral antigens and defective

budding particles on the cellmembrane would

block thereceptors forsuperinfecting

homolo-gous infectious virus but not for heterologous

virus.

In nature aswell as in cell culture there ap-pears tobe acontinuous evolution of influenza

viruses (40).DIviruses, which appear to occur

commonly in influenza virus replication (17),

mayfurther aid inthe selection ofvariants and

thushelpin theevolution of the virus and the

creation ofdiversityamong theviruspopulation.

ACKNOWLEDGMENTS

A/USSR/90/77, A/PC/1/73, A/Vic/3/75 wereobtained from Alan Kendall, Center for Disease Control, Atlanta, Ga. We thankCharles Samuel, University ofCalifornia, Santa Barbara forperforming the interferon assays of supernatants ofpersistentlyinfected HeLa cells. Wealso thank Zane Price forelectronmicroscopyand forphotographs.

These studiesweresupportedin partbyaPublicHealth Service grant from the National Institute of Allergy and Infectious Diseases (AI-12749) andby the National Science Foundation(PCM 7823220).

LITERATURE CITED

1. Ahmed,R., and A. F.Graham.1977.Persistent infec-tions in Lcells with temperature-sensitive mutants of reovirus. J.Virol. 23:250-262.

2.Almond, J. W. 1977.Asingle gene determines the host range ofinfluenza virus. Nature(London) 270:617-618. 3. Baron, S.1969.Interferonproductionassay and charac-terization, p.399-410.In K. Habeland N. P. Salzman (ed.),Fundamental techniques invirology. Academic Press,Inc., New York.

4. Caliguiri,L.A.,and K. V. Holmes.1979. Host-depend-entrestriction of influenza virus maturation. Virology 92:15-30.

5.Choppin,P.W., andM.W. Pons.1970.The RNAs of infective and incomplete influenza virions grown in MDBKand HeLa cells.Virology 42:603-610. 6. Crumpton,W. M., N. J.Dimmock,P. D.Minor, and

R.J.Avery. 1978.TheRNAsofdefectiveinterfering influenza virus.Virology90:370-373.

7.Davis,A.R.,A.L.Hiti,and D. P.Nayak.1980. Influ-enzadefective interfering viral RNA is formed by inter-nal deletion of genomic RNA. Proc. Natl. Acad. Sci. U.S.A.77:215-219.

8. Davis,A.R., and D. P. Nayak. 1979. Sequence relation-ships among defectiveinterferinginfluenzaviralRNAs. Proc. Natl. Acad. Sci. U.S.A.76:3092-3096.

9. DeWachter, R., and W. Fiers. 1972. Preparative two dimensionalpolyacrylamidegelelectrophoresis of 'P-labeledRNA.Anal. Biochem.49:184-197.

10. Floyd,R.W.,M. P. Stone, and W. K. Joklik. 1974. J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

DEFECTIVE INTERFERING INFLUENZA VIRUS 859

Separation of single stranded ribonucleic acids by acryl-amide-agarose urea gelelectrophoresis. Anal. Biochem. 59:599-609.

11. Friedman,R.M.,and J. M. Ramseur. 1979. Mecha-nisms ofpersistent infections by cytopathic viruses in tissue culture. Arch. Virol.60:84-103.

12. Holland, J. J., E. A. Grabau, C. L. Jones, and B.L. Semler.1979.Evolution ofmultiple genome mutations during long-term persistent infection by vesicular sto-matitis virus.Cell 16:495-504.

13. Holland, J. J., and L. P. Villarreal. 1974. Persistent noncytocidal vesicular stomatitis virus infections me-diated by defective T particles that suppress virion transcriptase. Proc. Natl. Acad. Sci. U.S.A. 71:2956-2960.

14.Holland, J. J., L. P. Villarreal, and M. Breindl. 1976. Factorsinvolved in the generation and replication of rhabdovirus defective T particles. J. Virol. 17:805-815. 15. Huang,A.S., andD.Baltimore. 1970.Defective viral particles and viral disease processes. Nature (London) 226:325-327.

16. Israel, A., M. Semmel, and J.Huppert. 1975. Host-range mutants offowl plague virus (FPV): comparison of the genome and virus proteins. Virology 68:504-509. 17. Janda, J. M., A. R.Davis, D. P. Nayak, and B. K. De. 1979.Diversity and generation of defective interfering influenza virus particles. Virology95:48-58.

18.Janda, J. M., and D. P. Nayak.1979.Defective influenza virus ribonucleoproteins cause interference. J.Virol. 32: 697-702.

19. Kang,C. Y., T.Glimp,J. P. Clewley, and D. H. L. Bishop. 1978. Studiesonthegeneration of vesicular stomatitisvirus(Indiana serotype) defective interfering particles.Virology84:142-152.

20. Kawai,A.,S.Matsumoto, and K. Tanabe.1975. Char-acterization of rabies viruses recovered from persist-ently infected BHKcells. Virology67:520-533. 21. Kennedy,S.L.T.1976.Sequencerelationships between

the genomeand theintracellular RNA species of stand-ard and defectiveinterferingSemliki Forest virus.J. Mol. Biol.108:491-511.

22. Maxam, A.M., andW.Gilbert.1977.A newmethodfor sequencing DNA. Proc. Natl. Acad. Sci. U.S.A. 74:560-564.

23. Meinkoth, J., and S. L. T. Kennedy. 1980. Semliki Forestviruspersistencein mouseL929cells. Virology 100:141-155.

24. Metzel,P.S., W. M. Schnitzlein, and M. E. Reich-mann. 1978.Characterization of distinct vesicular sto-matitisvirus, NewJersey serotype, isolates with respect tonucleic acidhomologies, interferenceby DI particles andprotein structure, p.515-526.In B.W.J.Mahy and R. D. Barry (ed.), Negative strand viruses and host cells.AcademicPress, Inc., New York.

25. Nakajima,K.,M.Ueda,and A.Sugiura. 1979.Origin of small RNA in von Magnusparticles of influenza virus.J.Virol. 29:1142-1148.

25a.Nayak, D. P.1980.Defectiveinterferinginfluenza viruses.

Annu. Rev.Microbiol. 34:619-644.

26. Nayak, D. P., K. Tobita, J. M. Janda, A. R. Davis, and B. K. De. 1978. Homologous interference mediated bydefective interfering virus derived from a tempera-turesensitivemutant of influenza virus. J. Virol. 28: 375-386.

27. Palese, P., and J. L. Schulman. 1976. Differences in RNA patterns of influenza A viruses. J. Virol. 17:876-884.

28. Popescu, M., and F.Lehmann-Grube. 1977. Defective interfering particles in mice infected with lymphocytic choriomeningitis virus. Virology77:78-83.

29. Rima, B. K., B.W.Davidson,and S. J. Martin. 1977. The role of defective interferingparticles in persistent infection of Verocellsbymeaslesvirus.J.Gen. Virol. 35:89-97.

30. Roux, L., and J. J. Holland. 1979. Role of defective interferingparticles ofSendai virus in persistent infec-tions.Virology 93:91-103.

31. Schmaljohn, C., and C. D. Blair.1977.Persistent infec-tion of cultured mammalian cells by Japanese enceph-alitis virus. J.Virol.24:580-589.

32. Scholtissek, C.,L.Koennecke, and R. Rott. 1978. Host range recombinants of fowlplague (influenza A) virus. Virology 91:79-85.

33. Scholtissek,C.,and B. R.Murphy. 1978.Hostrange mutantsofaninfluenzaAvirus.Arch. Virol. 58:323-333.

34. Schulman, J.L.,andP.Palese.1977.Virulence factors of influenza A viruses: WSN virus neuraminidase re-quired for plaque productioninMDBKcells. J. Virol. 24:170-176.

35. Semler, B. L., and J. J. Holland.1979.Persistent vesic-ular stomatitis virus infection mediates base substitu-tionsinviral termini. J. Virol.32:420-428.

36. Stark,C.,and S.L.T.Kennedy.1978.Thegeneration and propagation of defective interfering particles of Semliki Forestvirus indifferentcelltypes.Virology89: 285-299.

37. von Magnus, P. 1954. Incomplete forms of influenza virus. Adv.VirusRes. 2:59-78.

38. Weiss, B., R. Rosenthal, and S. Schlesinger. 1980. Establishment andmaintenanceofpersistent infection by Sindbis virusinBHKcells.J.Virol. 33:463474. 39. Welsh, R. M.,P.A.Burner, J. J.Holland,M. B. A.

Oldstone, H. A. Tompson, and L. P. Villarreal. 1975. Acomparison of biochemical andbiological prop-ertiesof standard and defective lymphocytic chorio-meningitis virus. Bull. W.H.O.52:403-408.

40. Young, J.F., U.Desselberger,and P. Palese. 1979. Evolution of human influenzaA virusesinnature: se-quentialmutation inthe genomes ofnewHlNl isolates. Cell 18:73-83.

41.Youngner, J.S.,andD.0.Quagliana. 1975. Temper-ature sensitive mutantsisolatedfromhamster and ca-ninecelllinespersistently infectedwith Newcastle dis-easevirus. J.Virol.16:1332-1336.

VOL. 36,1980

on November 10, 2019 by guest

http://jvi.asm.org/