Viral and cellular factors governing hamster cell infection by murine and gibbon ape leukemia viruses.

(1)

Viral and Cellular

Factors

Governing

Hamster

Cell

Infection by

Murine

and Gibbon

Ape

Leukemia Viruses

CAROLYN A. WILSON ANDMARIBETH V. EIDEN*

Laboratory of Cell Biology, NationalInstitute of Mental Health, Building36, Room2D10,

9000Rockville Pike, Bethesda, Maryland 20892

Received 17 May1991/Accepted 1 August 1991

Hamster cells are resistant to infection by most retroviruses, including Moloney murine leukemia virus

(MoMLV) and gibbonapeleukemiaviruses(GaLVs). Wehave constructedMoMLV-GaLVhybrid virionsto identify viraland cellular determinants responsible for the inability of GaLVand MoMLVto infect hamster cells. Thesubstitution of MoMLVcorecomponentsforGaLVcorecomponentscircumvents theresistanceof

hamster cells toinfection by GaLV, demonstrating that hamster cells havereceptorsfor GaLV butarenot

efficiently infectedbythis primateretrovirus because ofapostpenetration block.Incontrast,hamstercells are apparently resistanttoMoMLVinfection because although they bearareceptorforMoMLV, thereceptoris

nonfunctional. Treatment of CHO Kl orBHK21hamstercells with theglycosylation inhibitor tunicamycin

allows the cells to be infected by MoMLV. The construction of MoMLV-GaLV hybrid virions that can efficiently infect resistant cells has allowed the identification of viral and cellular factors responsible for

restricting infectionof hamster cells by MoMLV and GaLV.

Both host cell and viral factors regulate the efficient replication of retroviruses in cells, eitherbyinfluencing viral

entry orby affecting viral replication at a postpenetration

level(32).

The entryof retroviruses into cells dependson the

inter-actionbetween theenvelope glycoprotein andaspecific cell

surfacereceptor.Thepresenceofafunctional viralreceptor

is an absolute requirement for viral infection. Host range restrictions due tothe absence of the viral receptorcanbe overcome by introducing and expressing the appropriate

receptor in the resistant cells (1) orby replacing the viral

envelope oftherestricted virus withonederivedfromavirus

which can infect the cell of interest (26, 33, 34). Cellular resistance toa virus may also be imposed by saturation of the cellular receptors with viral envelope glycoproteins synthesized within a preinfected cell, which then block infection ofsusceptible cellsbythehomologous virusorany virusthatutilizes thesame receptor. This iscalled

superin-fection interference (25).

Postpenetration blocks can also result in substantial

de-creasesin theefficiency ofinfection. Bothcellularand viral

factors may modulate susceptibility to infection by murine retroviruses at the postpenetration level. The Fv-1 gene locus in murinecellsgovernssusceptibility tointegration of retroviral DNA into a murine host cell genome (11-13). SusceptibilitytoFv-1 restriction has beenphysically mapped

to twoadjacent amino acids in the p30core protein (5, 20,

28). Other factors thatregulate viral replicationin the later stagesofinfection arethoseaffectingefficientexpressionof the integrated viralDNA. Negative regulators inhost cells havebeen proposed toaccount for therefractory status of stem cells in vivo and of undifferentiated embryonic stem

cells in vitro with respecttoefficient infection by Moloney

murine leukemia virus (MoMLV) (27, 29, 30). The viral determinants responsibleforrestricting expressionin these cells reside in the long terminal repeat and primer-binding

sitesof MoMLV (6, 7, 14, 21).

*Correspondingauthor.

We prepared hybrid gibbon ape leukemia virus (GaLV)-MoMLV virions, containing various proportions ofprotein

and genomic components contributed either by GaLV or

MoMLV, to determine what viral component(s) restricts efficient infection of hamster cells by GaLV and murine

ecotropic and amphotropic viruses. The 4070A virus is a

murine amphotropic virus with a broad host range that

includesmanytypes ofmouse, cat, dog, guinea pig, rabbit,

rat,and human cells. Hamster cells, however, are resistant

to4070Avirus(8, 23). GaLV replicates wellinawidevariety ofcells; however, hamster and mouse cells are resistant to

productive infectionby GaLVs (31, 33).

Two types of hybrid virions were used to assess the

relative contributions of viral genome, core proteins, and

envelope proteins tothe inability ofMoMLV, 4070A, and GaLV to infect hamster cells. One type ofhybrid virion, historicallyreferredtoasapseudovirion,containsagenome

from one virus and structural proteins from another virus. Suchahybrid virionorpseudovirionis namedaccordingto theviruscontributingthe structuralproteins. Asecondtype ofhybrid virion, typified by virions produced by the pack-aging cell lines TAM or PA317 (4, 17) and PG13 (18), contains both genome and corecomponents from MoMLV andenvelope proteinsfromasecond virus. Thus,PA317or

TAM hybrid virions contain an envelope from the murine 4070A virus and PG13 virions contain GaLV envelope proteins.

In additionto resolvingthe contribution of differentviral component(s) to theblock ofreplicationof these viruses in

hamstercells, we also soughtto determine whether a com-monmechanism of resistancetoMoMLVorGaLV infection

(suchastheabsence ofexpressionof theappropriatecellular

receptor) is sharedby five different hamster cell lines. We

examined cell lines derived from three different species of

hamster to determine whether different mechanisms that

accountfor the resistanceofhamster cells tovirus infection by MoMLV, 4070A, and GaLVs exist between hamster species orwithin cell lines established from the same

spe-cies.

5975

0022-538X/91/115975-08$02.00/0

on November 10, 2019 by guest

http://jvi.asm.org/

(2)

MATERIALS AND METHODS

Cell lines and viruses. The cells used in this studyinclude

the following: NIH 3T3 murine fibroblasts (ATCC CRL 1658);rat2embryo fibroblasts (ATCC CRL1764); CHOKl

Chinese hamsterovary cells

(ATCC

CCL61); HaK

Syrian

hamsterkidney cells (ATCC CCL 15);DonChinesehamster

lung

cells(ATCC CCL 16); BHK 21 Syrian hamster kidney

cells (provided by Noel Bouck, Northwestern University,

Chicago,

Ill.), AHL Armenian hamster lung fibroblasts (ATCCCCL 195), EJ human bladder carcinoma cells

(pro-videdby S. Aaronson, National Cancer Institute, Bethesda, Md.), HOS human osteosarcoma cells (ATCC CRL 1543), minklung fibroblasts (ATCC CCL 64), PA317 (ATCC CRL

9078), T2/BAG and TAM/BAG (provided by C. Cepko, Harvard Medical School, Boston, Ma. [24]) and PG13 (ATCC CRL 10,683). All cell lines (except CHO K1)were

maintained in Dulbecco's modified essential medium

(DMEM; Whittaker Bioproducts, Inc., Walkersville, Md.), supplemented with 5% fetal bovine serum(FBS), 100 U of

penicillin

perml, 100 ,ugofstreptomycinperml,and 40 mM

glutamine.

CHO Kl cells were propagated in similarly supplemented alpha MEM (Whittaker Bioproducts, Inc.). Mink lung fibroblasts infected with the 4070A virus were

provided

by Janet Hartley (National Institute of Allergyand

Infectious

Diseases,-

Bethesda, Md.). GaLV SEATO-in-fected batlung fibroblasts(CCL 88) and biologically cloned

MoMLV was obtained from Electronucleonics,

Incorpo-rated (Silver Spring, Md.).

Wild-type virus preparation and RT assays.

Wild-type

virus obtained from the culture medium oftheinfected cells

wasconcentratedbycentrifugationat143,000x g for 75 min

at

4°C.

The viral pellet was resuspended in phosphate-buffered saline andstored in

20-,lI

aliquotsinliquid nitrogen. Serial dilutionsoftheconcentrated viruswereusedtoinfect

rat 2fibroblasts. Thirty hours postinfection, a reverse

tran-scriptase

(RT)assay was

performed

onthe cellular

superna-tantsandthedilution of virus which generated >10,000cpm

ofRT

activity

per 20

[lI

ofsupernatant was used in

subse-quent

experiments.

Hamster or rat 2 cells were seeded in 24-well dishes at

20,000to30,000cells per well and theninfected with virusin

medium

containing

3

pug

ofPolybrene per ml. Twenty-four hourspostinfection, thecellswererinsedandfed withfresh

medium. Forty-eight hours postinfection, the cells were

seeded at a 1:10 dilution in 24-well dishes.

Eight

days

following infection,

the supernatant was

assayed

for RT

activity.

Five microliters ofasolution

containing

50 mMTris (pH 8.3), 120 mMNaCl, and 1% TritonX-100 wasadded to 20 ,u of

supernatant

and incubatedatroomtemperaturefor

10to 15min

prior

totheaddition of25

Ru

ofRTcocktail (50

mM Tris [pH 8.3], 10 mM

dithiothreitol,

1.2 mM

MnCl2,

10

pug

of

oligo(dT)12_18

perml, 20 ,ugofpoly(A)per ml, 10 nM

[3H]dTTP).

Samples

wereincubatedanadditional2hat37°C

before the reactions were stopped by the addition of

5-pdl

quantities

of 0.25 M EDTA.

Samples

were filtered onto

DEAE paper

by

using

a Titertek cell harvester and were

counted in a

liquid

scintillation counter.

Virionproductionandinfection.

T2/BAG

andTAM/BAG

cell lines were used as the source ofT2 and

PAM

virions

employed

in these studies (22). Both the Escherichia coli lacZ

,3-galactosidase

and TnS neo resistance genes are

present in the defective BAG genome transduced

by

T2/ BAGand TAM/BAG

virions,

and viral titers cantherefore bedetermined

by

counting neomycin-resistant

cellclonesor

by counting

bluefoci.

Transcription

oftheneogeneinitiates

from an internal simian virus 40 promoter, whereas tran-scription of the lacZ gene initiates from the retroviral pro-moter contained within the U3 of the long terminal repeat.

Helper-free GaLV virions were prepared by infecting the PG13 GaLV packaging cell line (18) with TAM virions

containingthe BAGgenome. PG13 virions contain MoMLV genomeandcorecomponentsand GaLVenvelope proteins.

Minkcellsproducing 4070A pseudovirionswereprepared by superinfecting 4070A virus-producing mink lung

fibro-blasts with PG13 virionscontainingthe BAGgenome. Mink cell clones expressing pseudovirions were selected in me-diumcontaining800pugof G418 per ml. GaLVpseudovirions

werepreparedin ananalogousmanner, thatis, by infecting mink fibroblasts

producing

GaLV SEATO

wild-type

virus

withTAM/BAG virionsandselecting forminkcells

express-ingtheBAG genome with mediumcontaining800p.gof G418 per ml.

Determination of virion titersby

0-galactosidase

andG418 assays.Targetcells wereseeded at50,000cells per well ina

12-welldish. Thenextday, targetcellswereexposed to2.0 mlof virusinmediumadjustedto3

pug

ofPolybreneper ml. Allvirussuspensionswereobtained fromthe filtered super-natantof confluentvirus-producing cells. Twenty-fourhours

postinfection,

cellswere

trypsinized

andseededat1:2,

1:10,

1:100,and 1:1,000 dilutions into 10-cm-diameterdishes. The

cells werefed with mediumcontaining G418. All cells were fed with DMEM

containing

5% FBS and the

indicated amount

(in micrograms

per

milliliter)

of active

G418asfollows:allhamster cellsandrat2

fibroblasts, 450;

NIH3T3cells, 400; EJandminkcells, 800; HOS

cells,

250.

Cells were selected for 10 to 15 days after which

G418-resistant cell clones were stained witha methanol solution

containing

0.75%

methylene

blue and0.25% carbol fuchsin

(Sigma,

St.

Louis,

Mo.)and

quantitated.

Relativetiterswere

determined by

dividing

the actual titer obtained with a

particular

cellline(e.g.,the numberof G418-resistantclones permilliliterofvirussupernatant) bythetiter obtained with

rat 2 fibroblasts. Serial dilutions of infected cells

prior

to

selection with G418 allowed differences in the apparent viral titer to be assessed over a

100,000-fold

range. To

determine virus titerson thebasisof

P-galactosidase

activ-ity,

cells

exposed

tovirionsweregrownto

confluence,

fixed

in a2%

glutaraldehyde-formaldehyde solution,

and stained

with

5-bromo-4-chloro-3-indolyl-p-D-galactopyranoside

(X-Gal)

(27). Relative titers were determined

by

dividing

the number ofblue

foci,

obtained with a

particular

target cell

exposed

to 1.0 ml of virus supernatant,

by

the number of bluefociobtainedon rat2cellsinfected with1.0mlofvirus supernatant.

Infection of target cellspretreatedwithtunicamycin.Target

cellswereseededat

50,000

cellsper wellinDMEM

contain-ing

5% FBS with or without0.15 p,g of

tunicamycin

(Cal-biochem

Corporation,

La

Jolla,

Calif.)

per ml in a 12-well tissue culture dish.

Tunicamycin

solutions were

prepared

andstoredasdescribed

by

Heifetzetal.

(9).

Approximately

18 h later, cellswere rinsed three times with fresh medium without

tunicamycin

andinfected with 2 ml of filtered virus

suspension.

Thenext

day,

cellswere

trypsinized

andseeded

at 1:2, 1:5, and 1:10 dilutions in a 10-cm-diameter tissue

culture dishand selected withG418 asdescribed above.

Construction of hamster and human cells

expressing

the murine MoMLV receptor cDNA. The

pLNS-JET

retroviral

expression plasmid

wasconstructed

by

introducing

a2.28-kb EcoRI-to-BamHI restriction

fragment

from the

pJET

plas-mid (J.

Cunningham,

Harvard Medical

School, Boston,

Mass.)

into the

pLNSXe plasmid (19).

The

pJET-derived

on November 10, 2019 by guest

http://jvi.asm.org/

(3)

[image:3.612.54.295.99.210.2]

TABLE 1. Susceptibility of hamster cells to infection by GaLV, GaLVpseudovirions, and PG13 virions

Susceptibility of cells to infection byvirusa

Cells GaLV PG13

Wild-type Pseudo- virionc virion'b virionc

CHOKl - 0.001 0.4

Don - 0.1 1.0

BHK 21 - <0.0001 0.002

HaK - 0.0003 0.015

AHL - 0.0001 0.01

Rat 2 + 1.0 1.0

aWild-type GaLV virions contain GaLV-derived genome and core and envelope proteins. GaLV pseudovirions contain GaLV-derived core and envelope proteins and a MoMLV-based genome. PG13 virions contain

MoMLV-derivedgenome and core components and GaLV envelope proteins. bSusceptibility based on RT assay. Symbols: -, <twofold over back-ground RT activity; +, >200-fold over backback-ground RT activity.

cSusceptibility based on resistant colony titers, normalized to

G418-resistant colony titers on rat 2 cells.

fragment contains the coding region for the ecotropic re-ceptor. CHOKl, BHK 21, and AHL weretransfected with 20

p.g

of pLNS-JET (2). Transfected cells were selected with 450 ,ug of G418perml, andresistant colonies were pooled. PA317 cells were similarly transfected and selected with 400 ,ug ofG418 per ml. Supernatant from the stably

trans-fected PA317 cells was used to infect human EJ bladder

carcinomacells.

RESULTS

Roleof GaLV core componentsin restricting GaLV infec-tion of hamster cells.Weexamined fivedifferent hamstercell lines for susceptibility toinfectionby GaLV. These include BHK 21 fibroblast cells and HaK epithelial cells (derived from Syrian hamsterneonatalkidney),twoChinese hamster cell lines, CHO Kl and Don (derived from Chinese

ham-ster ovary and lung, respectively), and AHL fibroblast

cells (derived from Armenian hamster lung). All of these cells were resistant to infection by wild-type GaLV and

demonstrated anincreasedsusceptibility toinfection bythe PG13virions comparedwith the GaLVpseudovirions (Table 1).

GaLV pseudovirions contain GaLV viral proteins and a

MoMLV-based defective genome, whereas PG13 virions

containMoMLVcorecomponentsinadditionto aMoMLV genome and GaLVenvelope. GaLVpseudovirions did not

efficiently

infect CHO Kl, HaK, BHK 21,and AHL cells.

This resultindicates that

replacing

theGaLV genome witha

MoMLV-based genome alone is notsufficient to overcome

GaLV restriction in these cells. Substitution of both a

MoMLV core and MoMLV genome facilitates

infection;

therefore, core components are responsible for

restricting

GaLVinfection of CHOKl, HaK,BHK21,and AHL cells. The resistance ofthese hamstercellstoGaLV is thus dueto apostpenetrationblocktoinfectionandnot totheabsence of

afunctionalGaLV receptor.

Different hamster celllines demonstrate variableefficiencies of infection by PG13 virions.

Susceptibility

to infection

by

PG13 virions was quite variable among the different

ham-stercells.The HaK and BHK21cell lineswereestablished

from Syrian hamsters, which are members of the genus

TABLE 2. Susceptibility of hamster cells to infection by

MoMLV, 4070A,4070A pseudovirions, andNP2andPAM virions

Susceptibility of cells to infection by virus

Cells Wild-typeviriona 'T2 4070A 'PAM virionb psuo virionb

MoMLV 4070A virionb

CHO Kl - - <0.0001 <0.0001 <0.0001

Don - - <0.0001 0.01 0.01

BHK 21 - - <0.0001 <0.0001 <0.0001

HaK - - 0.01 0.0003 0.015

AHL - - <0.0001 <0.0001 <0.0001

Rat 2 + + 1.0 1.0 1.0

a Susceptibility based on RT assay. Symbols: +, >200-fold over back-ground RT activity; -, <twofold over backback-ground RT activity.

bSusceptibility based on G418-resistant colony titers, normalized to G418-resistantcolony titers on rat 2 cells.

Mesocricetus. The Chinese and Armenian hamster cell lines used in these studies were established from two

differ-ent species (Cricetulus griseus and Cricetulus

migra-torius, respectively) within the same genus. Chinesehamster cells were the most efficiently infected, while BHK 21 cells were the least efficiently infected (a 1,000-fold lower effi-ciency of infection of BHK 21 cells compared with Don cells, [Table 1]). Don cells were unique in their ability to be much moreefficiently infected by GaLVpseudovirions than other hamster strains were albeit they were 10 times less efficiently infected by GaLV pseudovirions than by PG13 virions.

Hamster cells lack appropriate receptors for MoMLV and 4070A. Wild-type MoMLV and 4070A viruses failed to in-fect any of the five hamster cell lines employed (Table 2). No G418-resistant clones or ,-galactosidase-positive blue foci (data not shown) were observed with CHO Kl, BHK 21, and AHL cells following exposure to T2 virions, 4070A pseudovirions, or TAM virions. Substitution of the GaLV envelope is sufficient to overcome the resistance of these cells toinfection by T2 or TAMvirions, demonstratingthat MoMLV-derived genome and core components function efficiently in hamster cells (Table 1). The viral determinants that restrict the MoMLV and 4070A virions from infecting

these hamster cells therefore reside in the envelope. Hamster cells exhibit variable susceptibility to infection by P2 and PAM virions. The only hamster cell line that could be infected with T2 virions was the Syrian hamster cell line, HaK, albeit at a reduced efficiency relative to

that of rat 2 fibroblasts (15). The ability to infect HaK with T2 but not wild-type MoMLV may simply reflect differ-ences in the sensitivity between the RT assay and the

G418-resistant colony-forming assay.

Cells from the same species of hamster vary in their susceptibility to infectionby 4070Apseudovirionsand TAM virions (Table 2). Don cells are infected 100 times more

efficiently than CHO Kl, though both are derived from the same species, C. griseus. The observed differences in the apparent titer of TAM virions on BHK 21 compared with that on HaK demonstrates thatvariation in susceptibility to

infection can exist not onlyincells from thesame

species

of hamster (Mesocricetus auratus) but even from the same

tissue (Table 3). The mechanism for the observed variation in theefficiency of PAM infection ofdifferent hamster cells is unknown.

Effect ofinhibition ofglycosylation oninfection of hamster

on November 10, 2019 by guest

http://jvi.asm.org/

[image:3.612.314.557.101.206.2]

(4)

[image:4.612.65.566.103.190.2]

TABLE 3. Summaryof infectionefficiencies of hamstercellsobtained withPG13,T2,andTAMvirionsa

Infectionefficiencyof cells

Typeofvirion Chinese hamster Syrian hamster Armenian hamster

Type of

virlon

(Cricetulusgriseus) (Mesocricetusauratus) (Cricetulusmigratorius) Rat 2

CHOKl Don BHK21 HaK AHL

PG13virions +++ ++++ + ++ ++ ++++

T2virions - - - ++ - ++++

PAM virions - ++ - +++ - ++++

a This tablesummarizes datapresented in Tables1and2.Oneplussignisequivalenttoanincrease intiterof1orderofmagnitude.

cells by P2 and IAM virions. The resistance to MoMLV observed with the hamster cells couldbe aconsequence of thelackofexpression ofappropriate MoMLV receptors. It wastherefore necessary todetermine whetheramRNAfor

the MoMLV receptor was expressed in hamster cells. All

cells examined expressed mRNAs ofthe appropriate sizes that hybridized to the murine MoMLV receptor cDNA(data

not shown), despite the fact that only HaK cells demon-strated even limited susceptibility to infection by MoMLV

enveloped virions.

Todetermine whetherhamster cellresistancetoMoMLV infection was glycosylation dependent, three hamster cell

lines resistanttoinfectionby

P2

were analyzed for

suscep-tibilityto infection following treatmentwith an inhibitor of N-linked glycosylation, tunicamycin. The human cell line

EJ, whichisresistant toinfectionbyP2 but susceptibleto

infection by

PAM,

wasalsoanalyzed. While treatmentdid notalter theresistance of EJ or AHLcells toP2 infection,

resistance to infection was reversed in both BHK 21 and

CHOKlcellsby exposingthese cellstotunicamycin priorto

infection (Fig. 1). EJ cellsretained their ability to be infected by

PAM

virions (data not shown), indicating that the inabil-ity to overcome EJ cell resistance to

P2

infection following treatment with tunicamycin was not attributable to

tuni-camycin-induced cytopathicity.

Resistancetoinfection by

PAM

virions was alsoanalyzed

after tunicamycin treatment. In contrast to the results

ob-tained with

P2

virions, AHL cells were rendered sensitive

to

PAM

infection after tunicamycin treatment (Fig. 2), whereas CHO

Ki

and BHK 21 remained resistant to

PAM

infection (data not shown). Failure to detect any G418-resistant AHL cells following tunicamycin treatment and exposure to

P2

virions is probably not attributable to drug

toxicity, because AHL cells exposed to

PAM

survived the treatment. Instead, our results suggest that pretreatment of targetcells with tunicamycin renders AHL cells selectively

permissive to infection by

PAM,

but not

P2,

virions. Likewise, tunicamycin renders BHK 21 and CHO Kl

selec-tivelypermissive to

P2,

but not

PAM

virions.

The apparent titer of the

P2

virions on rat 2 cells was

500-fold less than the titer determined on untreated rat 2

cells,presumably reflecting some toxicity of the tunicamycin

treatment. Nonetheless, equivalent

P2

titers were obtained with tunicamycin-treated rat 2 cells (data not shown) and tunicamycin-treated, MoMLV-resistant BHK 21 and CHO Klcells (approximately

102

G418-resistantCFU/ml [Fig. 1]). These findings demonstrate that inhibition of glycosylation renders functional the existing MoMLV receptors on CHO

Kland BHK 21 cells and receptors for 4070A on AHL cells and thus show the treatment is not acting nonspecifically to

allow virus entry.

Expression ofatransfected murine MoMLV receptor cDNA in hamster cells conferssensitivity toMoMLV infection. To resolve more precisely how

glycosylation

renders BHK 21 and CHO Kl cells susceptible to MoMLV infection, the murine MoMLV receptor cDNA was transfected and

ex-pressed in these cells. The MoMLV receptor cDNA was

transfectedinto CHOKl, BHK21,and AHLcells. Suscep-tibilitytoinfectionwasdeterminedbyexposingthese cellsto

T2virionsandcounting X-Gal-positiveblue cell foci.

Intro-duction and

expression

ofa functional MoMLV receptor cDNAin CHO Kl, BHK 21, and AHLcells overcame the resistance to infection by T2 virions (Fig. 3). Armenian hamster lung fibroblasts are unlike BHK 21 and CHO Kl

cells but are similar tohuman EJ cells, in that these cells express areceptorhomologthatcannotfunctionally

substi-tute for this viral receptor even after pretreatment with

tunicamycin (1).

DISCUSSION

Hamstercells express GaLV receptors butareresistantto

GaLV at apostpenetration stageof infection. The blockto

GaLV infection of hamster cells appears to reside in the

GaLVcore, since substitution ofMoMLVcorecomponents circumvents the restriction of hamster cellstoGaLV infec-tion. Thisrestriction forGaLV in hamster cells is therefore similartoFv-J restriction in murine cells.

Therelativeefficiencywith which MoMLV-GaLVhybrid virions infect different hamster cell lines varies (Table 3). BHK 21cellsare not asefficiently infectedby PG13virions

asthetwoChinesehamster celllines, CHOKlandDon. It isunlikelythat therelativelyinefficient infectionofBHK 21 cells with PG13 reflects a restriction affecting appropriate functioning of the MoMLV genome or core components, because BHK 21 cells expressing the transfected MoMLV receptor areefficiently infectedby T2 virions. The observed variations insusceptibilitytoinfection of BHK 21 compared with thatof Chinese hamster cells to PG13 infection could be dueto subtlealterationsin theviral receptors present in the

twolines,resulting in inefficient binding of the virus tothe receptorsexpressedonBHK 21cells, orsimply adecreased GaLV receptordensityon thesurface of these cells.

Hamstercells areresistant to infection by MoMLV even

though they express a mRNA that is highly related to the MoMLV receptor cDNAisolated from NIH 3T3 cells. There

areseveralmechanismsthatcould account for theresistance

ofhamster cellstoinfectionbyMoMLV.First, the MoMLV receptorhomolog in these cells may be dysfunctional, sim-ilartothe MoMLV receptorexpressed in human EJ cells (1).

Second,theamountof MoMLV receptoractually expressed may be insufficientto allow efficient virus binding or entry

on November 10, 2019 by guest

http://jvi.asm.org/

(5)

I~~~~~~(

i

LI li~~~~~~~~~~~

45~~~~~4 5~~~~~~~~~~~~~~,

.4

elk%

.%9-. e4 *vX9

FIG. 1. Analysis of the effect of tunicamycintreatmentontarget cell susceptibility to T2 infection. FourT2-resistant cell lines, a

humancell line(EJ), aSyrian hamster fibroblast cell line (BHK 21), a Chinese hamster ovary cell line (CHO Kl), and an Armenian hamsterfibroblast cellline (AHL)wereassayed for infection by T2

virions followingtreatmentwithaninhibitor of glycosylation,

tuni-camycin. Shown herearecells whichweretrypsinized and seededat

a1:5 dilutionina10-cm-diameter tissueculturedishandselectedwith G418afterexposureto2.0mlof filtered virus. The culture disheson

the left(containing the untreated cells) and the culture dishesonthe right (containing tunicamycin-treated cells) depict the presence or

absence ofG418-resistant clones afterexposuretoT2 virions. Clones

were stained with amethanol solution containing 0.75% methylene blue and0.25% carbol fuchsintooptimizevisualization.

into cells. Third,resistancetoMoMLV infectionmaybe due toaviralinterference-like mechanismwherebyhamster cells express afunctional MoMLV receptor that iscompetitively blocked by endogenous ecotropic gp70s. Fourth, the

FIG. 2. Analysis of the effectoftunicamycin treatment on the hamster cells CHOKlandAHLsusceptibilityto TAMinfection. Depicted above are untreated cells (plates on the left) and cells treatedovernight with tunicamycin(plates on the right) exposed to PAM virions and seededat a1:5dilution priorto selection inculture mediumcontaining G418.

MoMLV receptorexpressed by these cells may be rendered nonfunctional as aresult ofposttranslational modifications

of theprotein.

Our finding that CHO Kl and BHK 21 cells were made

susceptible to MoMLV infection following treatment with

tunicamycin,aninhibitor of N-linkedglycosylation, suggests

thattherefractory status of thesecells cannot be attributed

to aprimary receptor protein product thatis either grossly

dysfunctionalorinsufficientlyexpressed.Tunicamycin treat-mentcouldreverseresistance ofCHO Kl and BHK 21 cells

toMoMLV infectiondirectlybyremovingoligosaccharides thatareN-glycosidically attachedtothe receptorproteinand block virusbindingorindirectly by preventing bindingofa

competing glycoprotein to the MoMLV receptor. Neither

viral glycoprotein interference nor a hamster cell-specific posttranslational inactivation of an otherwise conserved (i.e., murinelike) MoMLV receptor can explain the

tuni-camycin-sensitive block to MoMLV infection of hamster cells, since transfection and expression of the murine

MoMLV receptor protein facilitated MoMLV entry into

BHK 21 and CHO Kl hamster cells in the absence of

inhibitorsofglycosylation. Thus, glycosylationinactivation of the endogenous hamster MoMLV receptor is the most

likely mechanism for the blocktoMoMLV infection ofthe

hamster cells and must reflect differences in the primary

structure of thehamsterandmousereceptorproteins.

Com-parisonof the nucleotide sequence ofthe MoMLVreceptor mRNAhomolog expressed inCHO Kl or BHK 21 cells to

the murine MoMLV receptorcDNAmayresolve nucleotide

differences in the

receptor-coding

region that account for

alterations in glycosylation. If so, then

glycosylation

sites

thatarepresent inthe hamster MoMLV receptor cDNAand

HO}hl ₍_fie0 _K1_SANI

on November 10, 2019 by guest

http://jvi.asm.org/

[image:5.612.54.296.74.539.2] [image:5.612.315.557.78.326.2]

(6)

5980 WILSON AND EIDEN

(p

i:

AHL

/.

p..

'Er

4

r

r.

._

/) / /

~~~A

401

.4: +

.I

Cw

AHL + MoMLV receptor

....*

..s

t.S t ; t + .) i 8 .

>; *g v.*

t>8k,tS

+

'Sa

$.$w

> e tw

;Sa \ \ ,. * \ ;

..

,.s-

, ., $ *

f

.

#S {i QI '- 8

'^tO

.. t

CHO Ki

Ott,' * 0'

0

U

* -fit

7

1

I

t R. .X}

t

1

.*

IF <.~

0 A

s Ct.

p.

CHO Ki

+

MoMLV

receptor

or:

*

p

S

BHK

21 _BHK

₂₁ +

MoMLV

receptor

FIG. 3. Susceptibilityofresistant hamster cell lines to infection withT2recombinant virions before and aftertransfectionofMoMLV receptor cDNA. Aplasmid expressingthe murine MoMLV receptor cDNA was introduced into hamster cells viaCaPO4-mediated gene transfer(2). A Nikon invertedmicroscopewasused to obtainphase-contrast micrographsofcontrolAHL, CHOKi,andBHK21 cellsand their transfected counterpartsthat stablyexpressed MoMLV receptor cDNA after exposure toP2virions containingthe

p-galactosidase

gene.Threedaysafter exposuretovirus, cells were assayed for

P-galactosidase

activity by using thehistochemicalproceduredescribedin thetext. Magnification, x40.

J.VlIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

[image:6.612.70.553.93.641.2]

(7)

notfound inthe murine MoMLVreceptorcDNAarelikely to representvirus-binding regions.

No commonmechanism of cellularresistancetoinfection

byT2orTAMwasdiscerned for hamster cells derived from

the same genus, species, or even from the same hamster

tissue (Table 3). The AHL and CHO Ki cell lines, both establishedfrom the hamstergenus Cricetulusdiffer in that

tunicamycin pretreatment renders CHO Kl but not AHL

cells permissiveforT2infection and AHL butnotCHOKl

cells permissive toTAMinfection. TAM virionscaninfect

some but not all cells derived from C. griseus (Chinese hamster). The Chinese hamster ovary cell line CHO Ki is resistant to TAM infection, whereas Chinese hamster lung

fibroblasts (Don cells) arenot. Both BHK 21 and HaKcell lineswereestablishedfromaSyrian golden hamsterkidney,

yet these two celllines demonstrate marked differences in their susceptibility toTAMandT2virions aswellasPG13 virions. One factor governing increased susceptibility to infection among hamster cells may be cellular phenotype. BHK 21 cells are fibroblasts, whereas the more readily

infectible HaK cells are epithelioid cells. A

phenotype-dependent induction of functional viral receptor upon in

vitro culture of Syrian hamster kidney cells may explain these results. An analogous observation has been reported for infection of cultured fetal kidney cells by poliovirus. Expression of poliovirus receptor mRNAoccurs in kidney

cells bothin vivo and in vitro (16); however, primatekidney cells donotbindpoliovirus in vivo butcanbe inducedtodo

so after they are trypsinized and grown as a monolayer in

vitro(10). Ithas been proposed thatregulation ofpoliovirus bindingcouldbe affected by posttranslationalmodifications

of itsreceptor inducedby culturing kidneytissue in vitro. The general approachemployed heremay proveusefulin

theelucidation of the viral and cellulardeterminants

respon-sible for the refractory status of many human cells to

infection by recombinantvirions and in designing recombi-nanthybrid virions that cansuccessfully mediate transduc-tionandexpression offoreigngeneticmaterialinthesecells.

It has been previously reported that some human

hemato-poietic cell lines are resistant to infection by

TAM/PA317-derived virions due to postpenetration blocks to infection

(3). Attempts to design hybrid virions similar to the ones described in thisreport thatcan efficiently infectthese cells

may provefruitful.

ACKNOWLEDGMENTS

Weespecially thankLarry Mahanfor help inthepreparation of

this manuscript. Weare alsoindebtedtoJonMarshand A. Dusty

Miller for helpful discussions; Karen Farrell for technical

assis-tance;andMikeBrownstein and LeeGreen forsupportand encour-agement.

REFERENCES

1. Albritton, L. M., L.Tseng, D.Scadden,and J.M.Cunningham.

1989. A putative murine ecotropic retrovirion receptor gene

encodes a multiple membrane-spanning protein and confers susceptibilitytovirusinfection. Cell57:659-666.

2. Chen, C., andH. Okayama. 1987. High-efficiency

transforma-tion of mammalian cells by plasmid DNA. Mol. Cell. Biol. 7:2745-2752.

3. Collins, S.1988. Different mechanisms accountforthe relative

resistance ofKG-1 andHL-60celllines toretrovirusinfection.

J. Virol.62:4346-4348.

4. Cone, R. D., and R. C. Mulligan. 1984. High-efficiency gene

transferintomammaliancells:generation of helper-free recom-binantretroviruswithbroadmammalian hostrange.Proc. Natl.

Acad. Sci. USA81:6349-6353.

5. DesGroseillers,L.,andP.Jolicoeur. 1983. Physical

mapping

of the Fv-J tropism host range determinant of BALB/c murine leukemia viruses. J. Virol. 48:685-696.

6. Feuer,G., M.Taketo, R. C. Hanecak, and H. Fan. 1989.Two blocks inMoloneymurine leukemiavirusexpression in undif-ferentiated F9 embryonal carcinoma cells as determined

by

transientexpressionassays. J.Virol. 63:2317-2324.

7. Gorman,C.M.,P. W.J.Rigby,and D. P. Lane.1985.

Negative

regulation of enhancers in undifferentiated embryonic stem cells.Cell 42:519-526.

8. Hartley, J. W., and W. P. Rowe. 1976. Naturally

occurring

murineleukemia viruses in wild mice:characterization ofa new

"amphotropic" class. J. Virol. 19:19-25.

9. Heifetz, A., R. Keenan, and A. Elbein. 1979. Mechanism of actionoftunicamycinontheUDP-GlcNAc:dolichyl-phosphate

GlcNAc-1-phosphatetransferase. Biochemistry18:2186-2192. 10. Holland, J. J. 1961.Receptoraffinitiesasmajordeterminants of

enterovirustissuetropisminhumans. Virology 15:312-326. 11. Jolicoeur,P.1979.The Fv-1geneofthemouseand its control of

murineleukemiavirusreplication.Curr.Top.Microbiol. Immu-nol.86:68-122.

12. Kai, K., H. Ikeda, Y. Yuasa, S. Suzuli, and T. Okada. 1976. Mouse strain resistance to N-, B-, and NB-tropic murine leukemiaviruses. J. Virol. 20:133-140.

13. Lilly, F.,and T. Pincus. 1973. Geneticcontrolofmurinevirus

leukemogenesis.Adv. CancerRes. 17:231-277.

14. Loh, T.

P.,

L. L. Sievert, and R. W. Scott. 1987. Proviral sequences thatrestrict retroviralexpressionin mouse

embryo-nalcarcinoma cells. Nature(London) 308:470-472.

15. Mann,R.,R. C.Mulligan,and D.Baltimore.1983.Construction ofaretroviruspackagingmutantand itsusetoproduce

helper-freedefective retrovirus. Cell 33:153-159.

16. Mendelsohn, C. L., E. Wimmer, and V. R. Racaniello. 1989. Cellular receptorfor poliovirus: molecular

cloning,

nucleotide sequence,andexpressionofanewmember ofthe

immunoglob-ulinsuperfamily. Cell 56:855-865.

17. Miller, A. D.,and C. Buttimore. 1986.

Redesign

of retrovirus

packaging cell lines to avoid recombination

leading

to

helper

virusproduction. Mol.Cell. Biol. 6:2895-2902.

18. Miller, A. D., J. V. Garcia, N. von Suhr, C. M. Lynch, C. Wilson,andM. V. Eiden. 1991. Construction and

properties

of retrovirus packagingcells basedongibbonapeleukemia virus. J. Virol. 65:2220-2224.

19. Miller, A. D., and G. J. Rosman. 1989.

Improved

retroviral

vectorsforgenetransferandexpression.

BioTechniques

7:980-990.

20. Ou,C.-Y.,L. R. Boone,C.-K. Koh,R. W.Tennant,and W. K. Yang. 1983. Nucleotide sequences ofthegag-pol

regions

that determinetheFv-1host rangeproperty of BALB/c

N-tropic

and

B-tropic murine leukemiaviruses. J. Virol. 48:779-784. 21. Peterson, R., G. Kempler, and E. Barkis. 1991. A stem

cell-specific silencer in the

primer-binding

site ofaretrovirus. Mol. Cell. Biol. 11:1214-1221.

22. Price,J.,D.Turner,andC.Cepko.1987.

Lineage

analysis

in the vertebrate nervous system by retrovirus-mediated gene trans-fer. Proc.Natl. Acad. Sci. USA84:156-160.

23. Rasheed, S., M. B. Gardner, and E. Chan. 1976.

Amphotropic

host range ofnaturallyoccurringwildmouseleukemiaviruses. J. Virol. 19:13-18.

24. Rein, A., A. Schultz, J. P. Bader, and R. H. Bassin.

1982.

Inhibitors ofglycosylation reverse retroviral interference.

Vi-rology 119:185-192.

25. Rubin, H. 1960. A virus in chick embryos which induced resistance to in vitro infection

by

Rous sarcoma virus. Proc.

Natl. Acad. Sci. USA46:1105-1119.

26. Rubin, H. 1965. Genetic control of cellular

susceptibility

to

pseudotypesof Rous sarcoma virus. Virology 26:270-276. 27. Sanes, J., L. Rubenstein, and J.-F. Nicolas. 1986. Use of a

recombinant retrovirusto

study

post-implantation

celllineagein

mouseembryos. EMBO J. 5:3133-3142.

28. Schindler, J., R. Hynes, and N.

Hopkins.

1977. Evidence for recombination between N- and

B-tropic

murine leukemia

vi-ruses: analysis of three virion

proteins by

sodium

dodecyl

on November 10, 2019 by guest

http://jvi.asm.org/

(8)

sulfate-polyacrylamide gel electrophoresis.J.Virol. 23:700-707. 29. Speck, N. A., and D. Baltimore. 1987. Six distinct nuclear factors interact with the 75-base-pair repeat ofthe Moloney murineleukemia virus enhancer.Mol. Cell. Biol. 7:1101-1110. 30. Teich, N.M., R. A.Weiss,G. R.Martin,and D.R.Lowy.1977. Virus infection of murine teratocarcinomastemcell lines. Cell 12:973-982.

31. Teich,N. M., R. A. Weiss,S.Z. Salahuddin, R. E.Gallagher, D. H. Gillespie, and R. C. Gallo. 1975. Infective transmission and characterization of a C-type virus released by cultured

humanmyeloid leukaemia cells.Nature(London) 256:551-555.

32. Weiss, R. A. 1985. Experimental biology andassayof

retrovi-ruses, p. 209-260.InR. A.Weiss,N. M.Teich, H. E. Varmus,

and J. Coffin (ed.), RNA tumorviruses, vol. 1. Cold Spring HarborLaboratory, Cold Spring Harbor,N.Y.

33. Weiss, R. A., and A. L. Wong. 1977.Phenotypic mixingbetween avianand mammalian RNAtumorviruses.I.Envelope

pseudo-typesof Roussarcomavirus. Virology76:826.

34. Zavada,J. 1972.Pseudotypesof vesicularstomatitisviruswith the coat of murine leukaemia and of avian myeloblastosis viruses. J.Gen. Virol. 15:183-191.