Down regulation of poliovirus receptor RNA in HeLa cells resistant to poliovirus infection.

0022-538X/91/041829-07$02.00/0

Copyright C) 1991, AmericanSociety for Microbiology

Down

Regulation of Poliovirus

Receptor

RNA

in

HeLa

Cells

Resistant

to

Poliovirus

Infection

GERARDO KAPLAN AND VINCENTR. RACANIELLO*

Departmentof Microbiology, College of Physicians and Surgeons, Columbia University,

701 West168th Street, New York, New York 10032

Received 10 August 1990/Accepted 21 December 1990

A lineofHeLa cells (SOFIA) was previously isolated that is resistanttopoliovirus infection and does not expressfunctional virus binding sitesatthe cellsurface. The expression of the poliovirusreceptor(PVR)gene

inSOFIAcells wasexaminedtodetermine the molecular basis for the failure of these cellstoexpressPVRs. Southern blot analysis ofgenomic DNA revealed that the PVR gene in SOFIA cells did notcontain gross

alterations. However,PVRtranscriptswerenotdetected in Northern (RNA) blotanalysis of SOFIAcell RNA. Invitronuclearrun-onanalysis showed that transcriptionof PVR-specific RNAwasreduced inSOFIA cells.

Treatmentof SOFIAcells with 5-azacytidine restored susceptibilityto poliovirus infection, which correlated

with theappearanceofPVRsatthe cellsurface,asdetected withanti-PVR monoclonal antibody D171. PVR RNAwas detected in clones derivedfrom5-azacytidine-treated SOFIA cells. SOFIA cellswereconvertedto

poliovirus sensitivity ata rateof 5to 7%, suggesting that down regulationof PVR expressioninvolved few

cellular targets. Resistance of SOFIA cells to poliovirus infection therefore appears to result from down

regulation ofPVR RNA, leadingtolackofPVR expressionatthe cellsurface. Methylationmayplayarole in regulating the expression of thePVR gene, whichis notessentialforsurvivalofHeLa cells.

Poliovirus infection of primate cell cultures usually results

in rapid cell death. Cell killing depends on both viral and

cellular determinants, few of which have been identified.

Oneapproachto elucidating cellular requirementsfor virus-mediated cell killing has been to isolate cell lines that are

partially or totally resistant to picornavirus infection.

Al-though a number ofsuch cell lines have been isolated (14, and references therein), in most cases, the mechanism of

resistanceto viralinfectionis notknown.

To identify host requirements for poliovirus replication,

we recently isolated HeLa cell lines that are resistant to

infectionwith this virus (14). These cell lineswereobtained

by cotransfecting HeLa cellswithpoliovirusRNAandanin vitro-synthesized subgenomic replicon lacking sequences

encoding the viral capsid proteins (15). Some cells survived

the viral infection and could be propagated, despite

contin-uousproductionof virusduringpassage. Over 6months,the

cells underwentperiods of crisis accompanied byextensive celldeath, followed by periodsofrecoveryandproliferation.

After6 months, cell lines were isolatedthatgrewstably in

culture and no longer produced poliovirus but supported

viral infection with different levels of efficiency. One line, SOFIA, was isolated that was resistant to poliovirus infec-tion duetorestrictionsatthe cell surface and within thecell

(14). The nature of the intracellular block to poliovirus infectioninSOFIA cellsisnotyetknown, whereasthe basis of the cell surface restriction is thetopic of this study.

SOFIAcells donotadsorbpoliovirus and donotexpress

the epitope recognized by monoclonal antibody D171, di-rectedagainstthecellularreceptorforpoliovirus(14). Here

we show that poliovirus receptor (PVR) mRNA cannotbe detected in SOFIA cells and that transcription ofthe PVR

gene appears to be down regulated. Treatment of SOFIA

cells with 5-azacytidine (5-azaC), a potent demethylating agent, resulted in partial restoration of susceptibility to

* Correspondingauthor.

poliovirus infection, accompanied by increased levels of PVR mRNA and cell surface reactivity with D171. These results demonstratethat SOFIA cells donotbindpoliovirus becausestablePVR mRNA isnotpresent.Treatment witha

demethylating agent, whichrestores susceptibilityto

infec-tion, mayactivate transcription of the PVRgene or agene

required forPVRexpression.

MATERIALS ANDMETHODS

Cells and viruses. HeLa S3 and SOFIA cells (14) were

grown in either suspension cultures or monolayers. For

poliovirus titration, 2 x 106 HeLa cells grown in spinner

cultureswereplated in6-cmplasticcellculture dishes24 h beforeuse(16).ForpurificationofgenomicDNAorcellular

RNA,2 x 107cellswereplatedin15-cmplasticcell culture dishes 48 h beforeuseandgrowninDulbecco modifiedEagle

medium containing 10% horse serum (16). Stocks of

polio-virusP1/Mahoneywerepreparedin HeLacellsasdescribed

previously (16).

For infectious centers assay, monolayers were infected with poliovirus at a multiplicity of infection of 10 and

processed asdescribedpreviously (14).

Antisera. Monoclonal antibody D171 directedagainstthe HeLa cell PVR (21) was purified and used as described

previously (14). Sheep anti-mouse immunoglobulin 1251I-la-beled antibodies(2,000 Ci/mmol)wereobtained from Amer-sham Co.

Genomic DNA extraction and Southernhybridization

anal-ysis. Genomic DNA from HeLa and SOFIA cells was

isolated frommonolayersas describedpreviously (2).DNA was digested tocompletion with Sacl as recommended by

the manufacturer (New England BioLabs), fractionated on

0.8% agarose gels, and transferred to nitrocellulose paper

(Schleicher& Schuell)in10x SSC(1x SSC is 0.15 M NaCl

plus 0.015 M sodium citrate) as described previously (18).

Membrane-bound DNA was hybridized to a 32P-labeled XhoI-EcoRI cDNAfragmentofpSVL-H20A (20) containing

1829

on November 10, 2019 by guest

http://jvi.asm.org/

nearly all the coding sequence of the PVR, prepared by

oligonucleotide-primed DNA synthesis (Pharmacia) to a

specific activity of 109 cpm/,ug of DNA. The filter was

washed three times(15min each)with 2x SSC-0.1% sodium

dodecyl sulfate (SDS) atroomtemperature and three times at65°C. Autoradiographywasfor 48h at -70°CwithKodak XAR film (Eastman Kodak Co.) and a single intensifying screen.

Isolation of RNA and Northern(RNA) hybridization

anal-ysis. Total RNA from HeLa and SOFIA cellswasprepared

by theguanidine thiocyanate-CsCl technique (18). Poly(A)+

RNA was selected by chromatography on prepacked

oli-go(dT)-cellulose columns (Collaborative Research)

accord-ing to the manufacturer. RNA was fractionated in 1%

agarose-formaldehyde gels, transferredtonitrocellulose pa-per,andhybridizedwitha32P-labeled double-strandedDNA

prepared asdescribed above. DNA probes were either the

cDNAfragmentdescribed above for Southernanalysisor a

1.0-kb PstI-XbaI cDNA fragment encoding human y-actin

(obtainedfrom L. Kedes). Prehybridizationwasdonefor 4 h at42°Cinasolutioncontaining50%formamide, 5x SSCPE

(1x SSCPE is 150 mM NaCl,15 mM sodiumcitrate, 13 mM

KH2PO4, 1 mM EDTA, pH 7.2), 5x Denhardt's solution

(50x Denhardt'ssolution is 1% FicollDL, 1% bovineserum

albumin,1%polyvinylpyrrolidone),and 500 ,ugof denatured

herring sperm DNA per ml. Hybridization reactions were

done for 22 h at 42°C in 50% formamide-5x SSCPE-1x

Denhardt's solution-100 ,ug of denatured herring sperm

DNAperml-10% dextransulfate-150ngof 32P-labeled DNA

probe.

In vitro nuclear transcription assay. This assay was a

modification ofpublished procedures (3, 7, 19, 24). Nuclei

were isolated from 7 x 108 HeLa or SOFIA cells by suspending phosphate-bufferedsaline-washed cellsin 4 mlof

reticulocyte standard buffer (RSB) containing 1% Nonidet

P-40(23), incubatingthem for 10minonice,andcentrifuging them for 1 min at 15,000 x g. The nuclear pellet was

resuspendedin 1 ml of RSB-1% NonidetP-40, pelleted,and washed three times with reaction mix consisting of20 mM Tris(pH 7.4),140 mMKCl,10 mMMgCl2, 1 mM

dithiothre-itol,and20%glycerol. Nucleartranscription wasdone with

5 x 108 nuclei in 1ml ofreaction mix containing GTP, CTP,

and ATP (1 mM each), and 1 mCi of [a-32P]UTP (3,000 Ci/mmol; Dupont, NEN Research Products)at 30°C for 20

min. The reactionwas terminated with 1 ml of 2x HSB(20 mM Tris [pH 7.4], 1 MNaCl, 100 mM MgCl2,4mMCaCl2)

and20,ugof DNase Ipermlfor 30minatroomtemperature; 2.5 mlof2x TES(20mMTris[pH 7.4],20 mMEDTA,0.4%

SDS)and 300 ,gof tRNA were addedpriorto two

phenol-chloroformextractions at65°C. RNAwasprecipitatedfrom

the aqueous phase with 5% trichloroacetic acid and 60 mM

Na2P2O7 and filtered through 25-mm-diameter filters

(HAWP; 0.45-,um pore size; Millipore). After extensive

washing with 5%trichloroacetic acid and 60 mM Na2P2O7,

filterswereplacedinglassvialscontaining500

RI

of 1x TES and incubated for 10 min at 65°C to elute RNA from the filter. This elutionstep was repeateduntil less than0.1% of

the counts remained associated with the filter. RNA was

precipitatedwith 2volumesofethanol, resuspended in 50

[lI

of H20, and boiled for 3 min and placed on ice before addition to thehybridizationbuffer.

Filters were spotted with 5 ,ug ofa 1.0-kb XhoI-EcoRI cDNAfragmentfromplasmid pSVL-H20A (20)thatcontains

almost allthe coding sequence of the PVR, a 1.0-kb PstI-XbaI cDNA fragment encoding human-y-actin, and EcoRI-linearizedpUC19. Labeled RNA (5 x 107cpm)from either

HeLa or SOFIA cells was hybridized with a separate filter

containingthese three DNAs. Prehybridizationand hybrid-izationweredoneat42°Cfor 4 h and 3days, respectively,in

1mlof thesame solutions used for Northernhybridization. Filterswerewashed with 2x SSCat65°Cthree times(10min each) andonce at roomtemperature. High background was eliminatedby washing filters with 30 ml of 2x SSC

contain-ing100 units of boiled RNase T1 and 10 ,ug of boiled RNase Afor 1 hat37°C;750,ul of 20% SDS and 150 ,ul ofproteinase

K(20mg/ml)were added and incubatedfor 30 minat37°C. Finally,filterswere washed three times with 2x SSC-0.2% SDS at 65°C andexposed for 3to 10days as described for Southernanalysis.

Treatmentofcells with5-azaCand analysis ofPVR induc-tion. To optimize the concentrations of 5-azaC suitable for

ourexperiments, wegrew cells foratleast twogenerations

in the presence of 5, 10, or 20 ,uM 5-azaC as described

previously (8, 9); the latter concentration was found to be

cytotoxicandwasnotused further.Monolayersgrownin the presence of 5 and 10 FLM of 5-azaC were infected with

poliovirusat amultiplicityofinfection of10,and the kinetics

ofvirusproduction were determined. Therewere no

differ-ences in

poliovirus

yield in the presence ofS or 10 ,uM of 5-azaC(datanotshown),and therefore the latter

concentra-tion was used for subsequent experiments. For poliovirus infection and infectious centers assay, 6-cm dishes were

seeded with 106 HeLa or SOFIA cells containing regular growthmedia with 5-azaC. Cellsweregrowninthe presence of 5-azaC for 60 hto80% confluence and thenwereinfected withpoliovirustype 1 Mahoneyat amultiplicity of infection

of 10. After 60 minofadsorption, monolayerswerewashed

extensively and either trypsinized for infectious centers

assayorincubated for 48 hat37°Cforanalysisof virusyield;

for the latter,

aliquots

of medium were taken at different

times and titratedon HeLacells(16).

For detection ofPVR at the cell surface, 96-well plates

were seeded with 104 HeLa or SOFIA cells per well and grown for 60 h in growth media with or without 10 ,uM 5-azaC. Cells were stained with 10-fold dilutions of

mono-clonal

antibody

D171 and then with 125I-labeled sheep

anti-mouse immunoglobulin (1 ,uCi per well), as described

previously (14). After extensivewashing, plates were

auto-radiographed.

RESULTS

Analysis of PVR gene in SOFIA cells. SOFIA cells are

variants of HeLa cells thatareresistant topoliovirus

infec-tion. We previously showed that poliovirus cannot infect

SOFIA cells because cell receptors for the virus areabsent

(14). As afirst steptoward determiningthe molecular basis

forthe lack of functional PVR in SOFIAcells,weexamined

thestateof thePVRgeneby Southern hybridization

analy-sis. Therewere nodifferences in the patternsofSacI DNA

fragmentsobserved in HeLacells, SOFIAcells,andcloned SOFIA cell lines(Fig. 1). Similar resultswereobtained when other restriction enzymes were used (data not shown), suggesting thatloss ofPVRexpression in SOFIA cellswas notduetogrossalterations in thePVRgene.

Down regulation of PVR RNA in SOFIA cells. In HeLa

cells, a 3.3-kb PVR RNA is detected by Northern blot

analysis

of total cellular RNA, and in addition a 5.6-kb mRNA can be detected when poly(A)+ RNA is analyzed (20). Northernhybridization analysis of total cellularRNA, usingaPVRcDNAprobe,revealed thatSOFIA cells donot

express stable levels of the 3.3-kb PVR RNA (Fig. 2).

on November 10, 2019 by guest

http://jvi.asm.org/

Sofia

clones

1C3 1C6

1E7 1G52F3 2G5 2G72G9 3E5 S H

'

,,*t ;

**g,i;;

3--2.1

*__|E;x a--964

,S,,....,.r... ^ St. |w

>. ...

;r S .* 8 z s

* # *l,

:E, . e

., ,,l iF 8 *

.+

8

.4..

i

--2.3

.? -2.2

z. "

It

0

a

O'

Y..

3k 1..os i *

#.

#: *

*i X..zr

FIG. 1. Southernblot analysis ofPVRgeneinSOFIAcells. DNA samples (10,ug) of HeLa cells (H), uncloned SOFIA cells (S), and

severalsingle-cellclonesderivedfromSOFIAcells (1C3, 1C6, 1E7, 1G5, 2F3, 2G5, 2G7, 2G9, and 3E5)weredigestedwith Sacl, fractionated ina0.8%agarosegel,transferredtonitrocellulose,andhybridized witha32P-labeledcDNAprobe representing almost all the codingsequence

ofthe PVRgene. Positions ofmarkerDNAs (HindIIIfragmentsof bacteriophageX DNA)areindicated, with sizes in kilobase pairs.

Northernanalysisofpoly(A)+ RNAfromoneSOFIAclone,

2G9, confirmed the absence of stable 3.3-kb mRNA and in

additiondemonstrated absenceof the5.6-kb mRNAaswell

(Fig. 3).

To determine whether the absence of PVR mRNA in SOFIA cellswasduetotranscriptional down regulationorto

posttranscriptional events such as degradation, we

deter-mined the transcriptionrate ofthe PVRgene in HeLa and SOFIA cells by in vitro nuclear run-on analysis. Similar

Sofia clones

c

.x

-

:

--288

tf~8

s:

lk

*

[image:3.612.164.462.78.318.2]

--18S

FIG. 2. Northern blot analysis oftotal cell RNA from SOFIA cells. Samples (10 ,ug)of total cellular RNA from HeLa cells (H), uncloned SOFIA cells (S), and several single-cell clones derived fromSOFIA cells(1C3, 1C6, 1E7, 2G5, 2D5,2D11, 2E6, 2F3, 2G9, 2H2,and3E5)werefractionated ina1%agarose-formaldehyde gel, transferred to nitrocellulose, and hybridized with a 32P-labeled

cDNAproberepresentingalmost allthecodingsequenceof the PVR

gene. Positions of 28S and 18S rRNA markersareindicated.

amounts of32P-labeled RNA probes isolated from in vitro

nuclear run-on transcription assays were hybridized to ni-trocellulose filters containing PVR cDNA, human -y-actin

cDNA as an internal standard, and pUC18 as negative

control. Transcriptionof the PVRgeneisverylow in HeLa cellscomparedwith-y-actin(Fig. 4).InSOFIA 1E7and 2G9

clones, transcriptionof the PVRgenewasbarely detectable.

Scanningof thefilters indicated thattranscriptionof the PVR

gene wasthree-tofivefoldlower in 1E7 and2G9cells thanin

HeLa cells. The transcription rateof -y-actin wassimilarin

all three cell lines.

5-azaC-treated SOFIA cellsregain susceptibility to

poliovi-rus infection. Many silent genes can be activated by treat-ment of cells with the demethylating agent 5-azaC (4). Therefore, itwasof interesttodetermine whether suscepti-bility of SOFIA cells to poliovirus infection could be

re-storedby treatment of SOFIA cells with 5-azaC.

Theeffect of 5-azaCtreatmenton poliovirus yieldintwo

single-cell clones derived from SOFIA cells(1E7 and 2G9)

andin HeLa cellswasdetermined(Table1).Poliovirustiters increased 105-fold in 5-azaC-treated SOFIA cells after24h of infection, whereas very little increase in viral titerwas

observed in untreated cellscompared with untreatedHeLa cells. Virus yield inHeLa cellswasnot affectedby5-azaC treatment. Therewas nofurtherincrease inpoliovirus yield

at72 hpostinfectioninSOFIAcells, andmostof theSOFIA cells didnotshowcytopathiceffect. Incontrast, HeLacells were completely destroyed at24 h postinfection. Mouse L

cells, which do not express PVR, remained resistant to

poliovirus infection after treatment with 5-azaC (data not

shown).

An infectious centers assay was used to determine the fraction of SOFIA cells thatregained susceptibilityto

polio-virus infection after 5-azaC treatment. The 1E7 and 2G9

on November 10, 2019 by guest

http://jvi.asm.org/

[image:3.612.63.301.499.655.2]

1E7

2G9

HeLa

PVR

*

|

w

y-a~~~~~~~ctin

pUC18

FIG. 4. Invitro nucleartranscription. Nuclei from7x 10 HeLa cellsorSOFIAcell clones1E7and 2G9wereisolated. RNAchains wereallowed toelongate inthe presence of[32P]UTP and usedto

probe nitrocellulose filters containingPVR cDNA, y-actincDNA, andpUC18. Afterwashing and treatment withRNases, the filters wereexposedtoX-rayfilm.

_~

--18S

FIG. 3. Northern blot analysis ofpoly(A)+ RNA from clone 2G9 ofSOFIA cells. Samples (10 ,ug) ofoligo-dT-selected RNAfrom HeLacells (H), clone 2G9 cells (2G9), and clone 2G9 cells treated with 10 ,ug of 5-azaCperml for 48 h(2G9/5-AC)werefractionated in a 1% agarose-formaldehyde gel and blotted onto nitrocellulose

paper. (a) The filter was hybridized toa 32P-labeled PVR cDNA probe. (b) The filter was stripped of probe and rehybridized toa

32P-labeled -y-actin cDNA probe and autoradiographed for 24 h. Positionsof28S and18S rRNA markersareindicated.

clones ofSOFIA cellsand HeLa cellsweretreatedwith 10

,uM of 5-azaC for 60 h, infected with poliovirus at a

multi-plicity of infection of 10, dispersed with trypsin, and seeded

onto monolayersofHeLacells. Only 1 in 103 1E7 and 2G9

cellsappearedtoproduce infectious virus, whileeach HeLa

cellproduced infectious progeny(Table 2). However, since previous results indicate that SOFIA cells do notproduce infectious poliovirus (14), it is likely that virus detected in

the infectious centers assay represented particles that

ad-hered nonspecifically to the cells. Treatment with 5-azaC increased the number of SOFIA cell infectious centers by 100-fold or more; 7.6% of 1E7 cells and 5.5% of 2G9 cells

becamesusceptibletopoliovirus infection after 5-azaC treat-ment.

Induction of PVR after treatment of SOFIA cells with 5-azaC. To determine whether treatment of SOFIA cells

with 5-azaC induced cell surface expression of PVR, we

assayedfor the presenceofthe receptor using antireceptor monoclonal antibodyD171 (21). Cellsweregrownin96-well

plates in the presence or absence of 10 ,uM 5-azaC and

incubated first with D171 and then with 125I-labeled

anti-mouse antibody. Treatment of SOFIA clones with 5-azaC

resulted in the induction of PVR comparedwith untreated cells (Fig. 5). Thus, 5-azaCappearsto restore sensitivityof

SOFIA cellstopoliovirusinfectionby restoringcell surface

expression of PVR.

It was not possible to detect PVR RNA by Northern

hybridization analysis of total RNA (data not shown) or

poly(A)+ (Fig. 3, compare lanes 2G9 and 2G9/5-azaC) in

5-azaC-treatedSOFIAcells. Thisresultwasnotunexpected,

since only 5 to 7% of 5-azaC-treated SOFIA cells became

susceptibletopoliovirusinfection(Table 2), andthelevels of

PVRtranscripts inthese cells arelower than in HeLa cells

(seebelow and Fig. 5). Tocircumvent thisproblem, SOFIA

cells were treated with 5-azaC and cloned, and

poliovirus-susceptible and -resistant clones were identified. Northern hybridization analysisdemonstrated thatpoliovirus-resistant

clones did not contain stable PVR RNA, while different levels of the 3.3-kb PVR RNA were observedinthe

polio-virus-susceptible celllines(Fig. 6). Although theamount of

PVR RNA detectedwas verylow,exceptinclone 13 which

expressed levels comparable to those in HeLacells, allthe clonesdisplayed completecytopathiceffect 24 h after infec-tionwithpoliovirus, withnodifferences inthe kineticsof cell

killingorvirus production (datanot shown). DISCUSSION

SOFIA cells were derived from HeLa cells that were

persistentlyinfectedwithpoliovirus andunderwent several

crisisperiodsbefore stabilizationand lossofsusceptibilityto

[image:4.612.112.243.74.308.2]

poliovirus infection (14). SOFIA cells do not express cell

TABLE 1. Effectof5-azaCtreatment on susceptibilityof cells to poliovirus

Viraltiter

(PFU/ml)a

Hour SOFIA

postinfection 1E7 2G9HeLa

_ ~~+ _+ -+

0 <2 x 102 <2 x 102 <2 x 102 <2 x 102 <2 x 102 <2 x 102

24 2 x 103 2.5 x 107 2.5 x 103 1.8x 107 3.6 x 108 4 x 108

72 1.5 x 103 2.3 x 107 2 x 103 2x 107 3.2 x 108 3.8 x 108

a

-,

without

5-azaC;

+, with 5-azaC.

CG

10,

99\

--28S

a

--18S

on November 10, 2019 by guest

http://jvi.asm.org/

[image:4.612.314.551.74.159.2] [image:4.612.64.556.625.716.2]

TABLE 2. Infectious center assayof 5-azaC-treated cells

No.of cells/infectiouscenter

5-azaCa SOFIA

HeLa

1E7 2G9

- 1,300 3,600 1

+ 13 18 1

a+,monolayerswere treatedwith10

F.M

5-azaC for60 h and theninfected

with poliovirus; -, no 5-azaC treatment.

receptors for poliovirus. Here we demonstrated that stable PVR mRNAis absent from SOFIA cells and that treatment

ofSOFIA cells with 5-azaC restores PVR expression and

sensitivitytopoliovirus infection. These results indicate that PVRexpression is notessential for viability of HeLa cells.

Although the PVR genein SOFIA cells does not appear to be grossly altered, it waspossible that undetected mutations were present which result in loss of PVR function. The

abilitytoinduce functional PVR with 5-azaC demonstrated the absence of such mutations in SOFIA cells. The doses of

5-azaCused inthisstudy are onlyslightly mutagenic (17, 22), and therefore it is unlikely that restoration of poliovirus

sensitivitytoSOFIAcells resultsfrom 5-azaC-induced

mu-tations. The high frequency at which SOFIA cells were converted topoliovirussusceptibility also makes this

possi-bility unlikely.

10 20 40 80 160 320

}

-

5-aza C

HeLa

}

Our data suggest that5-azaC restores transcription of PVR RNAin SOFIA cells. Stable PVR mRNAs were not detected in SOFIA cells, and nuclear run-on experiments indicated a reduced level of transduction of the PVR gene. The rate of PVR transcription in SOFIA cells was reduced three- to

fivefold compared with HeLa cells. Although this difference is small, it results in barely detectable transcription in SOFIA cells. This difference in transcription rate could accountfor thefailure to detect PVR mRNA in these cells, ifthis mRNA was veryunstable. Onespecies of PVR mRNA

contains a consensus mRNA destabilizing sequence in the

3'-noncoding region (20). Ourresults showed that the level

of PVR mRNA observed in HeLa cells is apparently not

required for susceptibility to poliovirus infection, since several poliovirus-susceptible 5-azaC-treated SOFIA cell clones expressed low but detectable levels of PVR mRNA

(Fig.6).

Themechanismby which5-azaC restorestranscriptionof the PVR geneis not known. Thecytidineanalog 5-azaC isa

potent demethylating agent that causes permanent

alter-ations ofgeneexpressionviaincorporation atmultiplesites in DNA and by inhibition by hemimethylases (fora recent

review, see reference 4). Activation of genes by 5-azaC

treatmenthas beencorrelated with changes in DNA

meth-ylation (5, 6, 11, 12), but 5-azaC also induces phenotypic changesinorganismswithunmethylated genomes (1, 25)and genetic changessuchasmitoticrecombination,chromosome

decondensation, and chromatid exchanges (10, 26).

There-10

20 40 80 160

320

{

1C6

+

5-azaC {

-5-azaC

{

+5-aza

C

...

J,,,

ii-,

FIG. 5. Inductionof PVRatthe cellsurface of SOFIA cells treatedwith 5-azaC. HeLa cells and SOFIA cell clones 1C6, 1E7,and 2G9

weregrownin96-well plates for 48 h in thepresence(+ 5-azaC)orabsence (- 5-azaC) of10,uM5-azaC. Growth mediumwasremoved,and monolayersweretreated with dilutions ofanti-PVR monoclonalantibodyD171(1/10, 1/20, 1/40, 1/80, 1/160,and1/320), extensively washed, and stainedwith125I-labeled sheep anti-mouse immunoglobulin (1 ,uCiperwell). Plateswere exposedto X-rayfilm in the presenceofan

intensifyingscreenfor 5days. Concentrations of monoclonal antibody D171areexpressedasthereciprocalofthecorrespondingdilution.

1E7

d,RI

:R

....

}

2G9

on November 10, 2019 by guest

http://jvi.asm.org/

[image:5.612.98.521.371.672.2]

pvresistant pv

susceptiL

4

5

6

2

13

14

4^

FIG. 6. Induction of PVR-specific RNA in,' with 5-azaC. Northern blot analysis of total R

clones of SOFIA cells treated with 5-azaC described in the legendto Fig. 2. Three clon poliovirus infection (4, 5, and 6), whereas fou susceptible (2, 13, 14, and 15). The position of 3 indicated.

fore, we cannot rule out the possibility

other thandemethylationareinvolvedintl

onSOFIA cells. Whatever the mechanisn induces the expression of PVR in SOFI frequency of conversion to poliovirus set tentwithasmall number oftargetsrequire thisgene.

If hypomethylation is involved in re,

expression in SOFIA cells, the target of

cis-actingsequencesof the PVRgeneitsel

regulatorysequences of trans-actingfacto

expression. Examples of both types of

activationhave beenreported previously

(

ence4). Structural analysis of the PVRgi progressto address these possibilities.

Whileexpressionof PVR in SOFIAcellD atthe level oftranscription, it is not yet geneisregulatedinanywayinvivo. The to dateindicates that PVR RNA andprot4 mosthumanorgans(9a, 20). However,iti

PVRexpression is regulated ina cell-typ

For example, although monkey kidney poliovirus replication, primary cultures (

susceptible to poliovirus infection after culture (13). It has been reported thai

susceptibilitytopoliovirus correlateswith

PVR-specific mRNA and PVR-specific el

surface of tubularepithelial cells after 8to of theprimary culture (lOa). The

develor

bilityinmonkey kidneymayhavemechan

the restoration of poliovirus susceptibilit

with 5-azaC. Further studies on the mec

actiononSOFIA cellsmight be relevanttc

activityin primate tissues.

ACKNOWLEDGMENTS This work was supported by Public He

GM38155fromtheNational Institute of Gener

REFERENCES

1. Antequera F., M. Tamame, J. R. Villanu

1984. DNAmethylation in the fungi. J. Bi 8036.

2. Bollon, A.,andS. Silverstein. 1982. Techn atedgene transfer foreukaryotic cells,p.

)le (ed), Techniques in somatic cell genetics. Plenum Publishing Corp., New York.

1 15 3. CairoG., L. Bardella, L. Schiaffonati, P. Arosio, S. Levi, and A. Bernelli-Zazzara. 1985. Multiple mechanisms of iron-induced ferritinsynthesis in HeLa cells. Biochem. Biophys. Res.

Com-mun.133:314-321.

4. Cedar, H. 1988. DNA methylation and gene activity. Cell

-3.3 kb 53:3-4.

5. Charache,S.,G.Dover,K.Smith,C.Conover,J. R.Talbot,M. Moyer,andS.Boyer.1983. Treatment ofsickle cellanemia with

5-azacytidine results in increasedfetal hemoglobin production

and is associated with non-random hypomethylation ofDNA Ib: around theac-&--globingenecomplex. Proc. Natl. Acad. Sci. SOFIAcells treated USA 80:4842-4846.

'NAfrom single-cell 6. Chen, E., R. W. Karr, J. P. Frost, T. A. Gonwa, and G. D. was performed as Ginder. 1986. Gamma interferon and 5-azacytidine cause

tran-es were resistant to scriptional elevation of class I major histocompatibility complex

rother clones were gene expression in K-562 leukemia cells in the absence of .3-kbPVR mRNAis differentiation. Mol. Cell. Biol. 6:1698-1705.

7. Compere,S.J.,S.McKnight,and R.D.Palmiter.1981. Andro-gens regulate ovomucoid and ovalbumin expression indepen-dentlyofestrogen.J. Biol. Chem. 256:6341-6347.

8. Constantinides, P. G., P. A. Jones, and W. Gevers. 1977. that mechanisms Functional striated muscle cells from non-myoblast precursors he effect of 5-azaC following 5-azacytidine treatment. Nature (London) 267:364-nbywhich 5-azaC 366.

[A cells, the high 9. Constantinides, P. G., S. M. Taylor, and P. A. Jones. 1978.

nsitivity is consis- Phenotypic conversion of culturedmouseembryo cells by aza d for activation of pyrimidine nucleosides. Dev. Biol. 66:57-71.

9a.Freistadt, M.,and V. R. Racaniello. Unpublisheddata. storation of PVR 10.

Hori,

T. A. 1983. Induction of chromosome decondensation,

5-azaC

mho sister-chromatid exchange and endoreduplications by

5-azacy-Ifor

5altmight

be tidine, an inhibitor of DNA methylation. Mutat. Res.

121:47-Lfor, alternatively, 52.

orsthat affect PVR 10a.Horie, H., S. Koike, T. Kurata,I. Ise, A. Okitu, M. Kohara, K. 5-azaC-mediated Tago,and A. Nomoto.1990. Abstr. Fifth RinshokenInt.Conf.:

(reviewed

inrefer- Genetic Determinants ofPicornavirusDisease, p. 62.

ene iscurrently in 11. Jeggo, P. A., and R. Holliday. 1986. Azacytidine-induced reac-tivation ofaDNArepairgenein Chinese hamsterovarycells. smayberegulated Mol. Cell. Biol. 6:2944-2949.

clear whetherthe 12. Jones, P. A. 1985. Altering gene expression with 5-azacytidine. evidenceobtained Cell 40:485-486.

evidence obtained 13. Kaplan, A. S. 1955. Comparison of susceptible and resistant

ein is

expressed

in cells to infection with poliomyelitis virus. Ann. N.Y. Acad. Sci.

sstillpossiblethat 61:830-839.

e specific manner. 14. Kaplan, G., A. Levy, and V. R. Racaniello. 1989. Isolation and is not a site of characterization ofHeLacell lines blockedatdifferentstepsin

of kidney become thepolioviruslifecycle. J. Virol. 63:43-51.

several hours of 15. Kaplan, G., and V. R. Racaniello. 1988. Construction and t development of characterization of poliovirus subgenomic replicons. J. Virol.

the appearanceof 62:1687-1696.

pitopes at the cell

_16ihtofe

16. La Monica, N., C. Meriam, and V. R. Racaniello. 1986. Mapping

at the cell of sequences requiredformouse neurovirulence ofpoliovirus

16hofincubation type 2 Lansing. J. Virol. 57:515-525.

pment ofsuscepti- 17. Landorph, J. R., and P. A. Jones. 1982. Mutagenicity of

listic

parallelswith 5-azacytidine and related nucleosides inCsH/1OT1/2clone 8 and

ty in SOFIA cells V79 cells. Cancer Res. 42:817-823.

,hanism of 5-azaC 18. Maniatis, T., E. Fritsch, and J. Sambrook. 1982. Molecular

induction of PVR cloning:alaboratorymanual. ColdSpringHarborLaboratory,

ColdSpring Harbor,N.Y.

19. Mattia, E., J.denBlaauwen,G.Ashwell,andJ.vanRenswoude. 1989. Multiple post-transcriptional regulatory mechanisms in ferritingeneexpression.Proc. Natl. Acad. Sci. USA 86:1801-ealth Service grant 1805.

alMedical Sciences. 20. Mendelsohn, C., E. Wimmer, and V. R. Racaniello. 1989. Cellularreceptor forpoliovirus: molecularcloning, nucleotide sequence andexpressionofanewmemberofthe

immunoglob-ulinsuperfamily. Cell 56:855-865.

ieva, and T. Santos. 21. Nobis, P.,R.Zibirre,G.Meyer,J.Kuhne,G.Warnecke,andG.

iol. Chem. 259:8033- Koch. 1985. Production ofamonoclonal antibody againstan

epitopeonHeLacells that is the functional poliovirusbinding

iquesof DNA medi- site. J.Gen. Virol. 6:2563-2569.

415-429. InJ. Shay 22. Olsson, L., andJ. Forchammer. 1984. Induction ofthe

on November 10, 2019 by guest

http://jvi.asm.org/

[image:6.612.62.300.67.183.2]

static phenotype inamousetumormodel by 5-azacytidine, and characterization ofanantigenassociated with metastatic activ-ity. Proc. Natl. Acad. Sci. USA 81:3389-3393.

23. Racaniello, V. R.,andC. Meriam.1986. Poliovirus temperature-sensitive mutantcontainingasingle nucleotide deletion in the

5'-noncoding region of the viral RNA. Virology 15:498-507. 24. Rao, K., J. Harford, T. Rouault, A. McClelland,F. Ruddle,aind

R.Klausner.1986.Transcriptionalregulation by iron of thegene

for thetransferrinreceptor. Mol. Cell. Biol.6:236-240. 25. Tamame, M., F. Antequera, and E. Santos. 1988. Development

ofcharacterization and chromosomal mapping of 5-azacytidine-sensitivefluF locus ofAspergillus nidulans. Mol. Cell. Biol. 8:3043-3050.

26. Zimmerman, F. K., and J. Scheel. 1984. Genetic effects of 5-azacytidine in Saccharomyces cerevisiae. Mutat. Res. 139:21-24.

on November 10, 2019 by guest

http://jvi.asm.org/