Rearrangement of integrated viral DNA sequences in mouse cells transformed by simian virus 40.

0022-538X/81/060872-08$02.00/0

Rearrangement

of

Integrated

Viral DNA

Sequences

in

Mouse

Cells

Transformed by Simian Virus

40

MARY ANN BENDER ANDWILLIAM W. BROCKMAN*

DepartmentofMicrobiology, University of MichiganMedicalSchool,AnnArbor, Michigan48109

Received 15 December1980/Accepted 26 February 1981

The organization ofviral DNA sequences inseveral celllines derived from a

primary colony of simian virus 40(SV40)-transformed mousecellswasanalyzed

to examine the origin of the various distinctive patterns of SV40 sequence

arrangement present in transformed cells. This analysis revealed a complex

arrangement of viralsequencesintheuncloned transformed cellsbutsimplified arrangements incloned derivatives of theprimary transformant. The cell lines

studied hadcertainSV40sequencearrangementsincommon,but thecloned lines

had lost some parental arrangements and acquired new arrangements. These

results indicate that the arrangement of viral sequences in some

SV40-trans-formed cellsis not fixed but that alterations occurafter integration, creating a

heterogeneous population oftransformants. In the process, expression of viral

genes maybe altered. Possible causesforandimplications ofthisgenetic

insta-bilityare discussed.

Cellular transformation inducedbysimian

vi-rus 40 (SV40) results from expression ofa

por-tion of the viral genome which is integrated

within the host cell DNA (19). Recently, the

patterns of integration of SV40 sequences in

several transformed cell lines have been

ana-lyzed (4,5, 14;G.Ketner and T. J.Kelly,J.Mol.

Biol., in press). From the results ofthese studies,

it may beconcludedthat therearemanysites in

cellular DNA into which SV40 may integrate

and that there are multiple sites in the viral

genome at which recombination with host

se-quences mayoccur. An assumption underlying

such conclusions is that rearrangementof

inte-grated viral DNA does not occur after the

pri-maryintegration event.

To gaina betterunderstanding of the

mech-anismresultinginstableintegrationpatterns in

transformed cells, we analyzed the SV40

se-quences present in cells grown from a primary

colony of SV40-transformed BALB/c 3T3

mousecells as wellastheviral sequences in each

ofseveral lines generated by consecutive cloning

of the primary transformant. This analysis

re-vealed that the pattern of integration ofSV40

sequences in the primary transformant is not

stable. Alterations in the arrangement of viral

sequencesresult in the appearance of new

inte-gration patternsinclones derived from the

pri-mary transformant. These results suggest that

the arrangement ofSV40 sequences present in

serially cloned,stabletransformants may not be

atruereflection of theinitial integration of SV40

into cell DNA but rather may represent the

cumulative result of eventsoccurring after the

primaryrecombination event.

(A portion of this work was carried out by

M.A.B.inpartialfulfilmentofrequirementsfor

the Ph.D. degree from the Departmentof

Mi-crobiology, University of Michigan Medical

School.)

MATERIALS AND METHODS Virus and cells. Thepropagation ofBALB/c 3T3

mousecells and their transformationby infection with tsA58 virionswereperformedat33°Cas wepreviously described (6, 12).Aprimary focus of transformed cells

wasisolatedinliquid medium and recloned thrice in successionat33'Cby endpointdilution. In thiscloning procedure,trypsinized cellswereplatedathigh dilu-tion in2-cm2wells and allowedtogrowintocolonies. Individual cloneswerepropagated from wells contain-ingasinglecolony.

Viraland recombinant DNAs.SV40 DNA was prepared from BSC-1 cells infected with wild-type

SV40 strain776asdescribedpreviously(9) and

puri-fiedbyelectrophoresisin 1.4% agarosegels (7). To ensure thatSV40 DNAused forhybridization

was not contaminated with mammalian cell DNA, SV40 DNA cloned in aplasmid vectorwas used in somehybridization experiments. Recombinant DNA

consisting ofSV40 strain 776 DNA cleaved at the

BamHIsite and insertedinto plasmid pBR322DNA

at the BamHI site was generously supplied by D. Nathans. Transformation ofEscherichiacoli HB101 (3) byrecombinant molecules wasperformed

essen-tiallyasdescribedbyWensinketal.(20).Recombinant

plasmidsweregrown and purifiedasdescribed else-where (8)with the exceptionthat theG-100 column

fractionation was omitted. Propagation of recombi-872

on November 10, 2019 by guest

http://jvi.asm.org/

nant DNA molecules was done in accordance with National Institutes of Health guidelines.

SV40and recombinant DNAs were labeled with 32P by nick translationessentially as described by Man-iatis etal. (16). The reaction mixture contained 50 mM Tris-hydrochloride (pH 7.8), 5 mM MgCl2, 10 mM 2-mercaptoethanol, 17.5

i.M

each dGTP and TTP, 2.5 ,LMeach dATP anddCTP, 4.4 mCi each of

[nP]dATP

and[32P]dCTP (specific activity, 400 to 600 Ci/mmol; Amersham Corp.), perml 10 ,ug of SV40 or

recombi-nant DNAperml, 175 U ofDNApolymerase I (Be-thesda ResearchLaboratories) per ml, and 0.012,ugof DNase I (Worthington Biochemicals Corp.) per ml. Reactions were carried outat 14°C and stopped by addition of EDTA to 30 mM, followed by phenol and chloroform extractions, when the incorporation of3P intoacid-precipitable material ceased.DNAwas sep-arated from soluble radioactive material by chroma-tographyon aSephadexG-75 column (1.2 by 10 cm)

runin10mM Tris(pH 7.4),0.1mMEDTA, and 100 mM NaCl. The final specific activity of the labeled

DNA was 4x 107to 20 x107cpm/,Lg.

Cell DNA. High-molecular-weight cell DNA was prepared byamodificationof the procedure of Gross-Bellardetal. (11). Confluent cells were harvested by scraping or trypsinization, washed in phosphate-buffered saline, and lysed by incubation in 10 mM Tris-hydrochloride (pH 7.9), 1 mMEDTA, 0.5% so-diumdodecyl sulfate (SDS), and0.1mg of proteinase K perml(Boehringer Mannheim Corp.)12to 16 h at 37°C. The lysate was extracted twice with an equal volume of phenol saturated with 0.1 M Tris-hydro-chloride (pH 8.6) followed by two extractions with chloroform-isoamyl alcohol (24:1). After exhaustive dialysis against 10mMTris-hydrochloride (pH 7.4), 1 mMEDTA, and 10 mM NaCl, RNase A (Sigma Chem-icalCo.; DNase heat inactivated) was added to 40,ig/ ml, and theresulting solutionwasincubated for 2 h at 37°C. This was followed by addition of 0.1 mg of proteinase K per ml and incubation for1 h at 37°C. Aftertwophenol andtwochloroform-isoamyl alcohol extractions, the aqueous component was dialyzed against10mM Tris(pH 7.4) and1mMEDTA.DNA was concentrated either by spooling on a glass rod from 33%isopropyl alcohol-0.3 M NaCl or, in studies wheredetection of free viral DNAwasattempted, by ethanolprecipitation, followed by resuspension in 10 mM Tris (pH 7.4) and 1 mM EDTA. Final DNA concentrationsweredeterminedby absorbance at 260 nm or by the fluorometric method of Labarca and Paigen (15).

Restriction enzyme digestion. Digestion of transformed cellDNAbyendo R *BglIIandTaqIwas

performed in

200-pl

reaction mixtures containing 30

,ugof DNA,10 mMTris-hydrochloride (pH 7.6),6mM

MgCl2, 6 mM2-mercaptoethanol and, forBglII and

TaqI digestions, 60and100 mM NaCl, respectively. Digestionswerefor3h at37°C forBglII andat65°C forTaqI.In acontrolreaction,30pgofSV40form I

and II DNA and30,ugof normalBALB/c3T3DNA weredigested under the above conditions.

Electrophoresis and transfer. Thirty

micro-grams ofdigestedcellDNA wasloadedinto aslot(6 by3by 10mm) in ahorizontal slab(20 by24by 1.0

cm) of0.7 or 1% agarose (Seakem or Sigma). One

adjacent slot contained a digested mixture of SV40 DNA and BALB/c 3T3 DNA (3 x 105 and 30 ug,

respectively) towhich, in somecases, form I and II

SV40 DNAwasadded. Another slot contained2,tg of Hind-III-digested phage ADNA.Electrophoresiswas

performedat 1V/cm (1% gels) or 2V/cm (0.7% gels) for 16hin thepreviously described buffer(7). After electrophoresis, the distribution of cell DNA

frag-ments andposition of theDNAfragments were doc-umented byphotography of the gel after staining with ethidiumbromide.

DNAwastransfered from thegelto anitrocellulose sheet (Schleicher & Schuell Co.) essentially as de-scribed by Southern (18). The DNA was denatured in thegelby immersion of thegelin0.5MNaOH-1.5M

NaCl for2h. TheNaOHwasneutralized by immersing thegel in1M Tris(pH 5)-3 M NaCl for2h. Transfer of denatured DNAtonitrocellulosewasmediatedby flow of 20x SSC (SSC is 0.15 M NaClplus 0.015 M

sodium citrate) through the gel over a 48-h period. After transfer, the filter was washed in 2x SSC and dried for2hat80°C under vacuum.

Hybridization.Thenitrocellulose sheets

contain-ing DNA weresoakedin 2xSSC and 1Ox Denhardt solution(Denhardt solution is 0.02% Ficoll, 0.02% pol-yvinyl pyrrolidine, and 0.02% bovineserumalbumin), 0.1%SDS, and50ugof heat-denatured salmon sperm DNA per mlfor4h at65°C. Hybridizationwascarried

out at68°C for65 to 72h inaplastic bag containing

0.05 M Na2HPO4-NaH2PO4 (pH 6.5), 2x SSC, lx

Denhardt solution, 0.1% SDS, 50

jg

of denatured salmon sperm DNA per ml, 10% (wt/vol) dextran

sulfate,and30 to150ngof denatured32P-labeledSV40

DNA (viralorrecombinant; specific activity,4 x 107

to 20x 107cpm/,g).Afterhybridization,the nitrocel-lulose was washed in 2x SSC, 0.05 M Na2HPO4-NaH2PO4 (pH 6.5), and 0.5% SDS for 4 hat 65°C. Afterdrying, thefilterwassubjected to autoradiogra-phy withanintensifyingscreenfor24 hto 3 weeks.

RESULTS

Derivation and biological properties of

celllines.All cell lines discussed in thisreport

werederivedfromasingle colonyoftransformed

cells. Oneday after infectionwith tsA58 virions

at avirus-to-cellratioof 2, 105BALB/c3T3cells

wereplatedina 10-cmculturedish. Four weeks

later,a

single

isolated

colony

ofovergrowthwas

selected for further

study.

With this

transfor-mation

procedure,

the

frequency

of

transforma-tionwas about 105 PFU per

transforming

unit.

The unclonedprogeny of the

original

transfor-mation colony has been

designated

line B1-0.

Thislinewas subjectedto cloning by

endpoint

dilution(seeMaterials and

Methods),

giving

rise

tolinesBl-la, -b, -c, and -d (see

pedigree, Fig.

1).One ofthese,Bl-ld,was

subjected

tofurther

cloning to generate lines B1-2a,

-b,

-c, and -d.

Finally,B1-3wascloned from B1-2a.

LineB1-0 and the majorityof clonesderived

from it

expressed

SV40T

antigens

and

displayed

a

transfoIrmed

phenotypeasassessed

by

on November 10, 2019 by guest

http://jvi.asm.org/

cJX

_{m -~} o .0 a

m X m

Kb

2.0-7.5

52

-

3.6-2.8

-1.9

-i.T

-a

b

FIG. 1. SV40-specific sequences in DNA from lineBl-0 and its derivatives; autoradiogram of 32P-labeled SV40 DNAhybridizedtocellDNAfragmentsgenerated by endoR. TaqI cleavage, separated by electropho-resis in1.0%oagarosegels,transferredtonitrocellulose sheets, and hybridizedto32P-labeledSV40 DNAas

described in Materials and Methods. The cell linesfromwhich DNAwasisolatedareindicatedatthe topof eachslot,asis therelationship of the cell linestoeachother.Slot M containsamarkerof linear SV40 DNA generated by TaqI digestion ofcircular SV40 DNA in the presence of BALBIc 3T3 DNA. The size (in kilobases)ofseveralfragmentsasdeterminedfromthemobility of A DNAfragmentsin the samegel is shown

attheleft. (a)SV40-specific fragmentsinBl-0 and its clonal derivatives; (b)SV40-specific fragments in BI-ld and its clonal derivatives.

tion density, ability to grow in soft agar, and

abilitytoovergrownormalmonolayers (datanot

presented). The two exceptions were B1-2a and

itssubclone B1-3,whose growth characteristics

resembled those of normal BALB/c 3T3 cells

and which did not contain SV40 T antigens

detectable by immunoprecipitation. Detailed

characterization of thesetwotransformation

re-vertantswill bedescribed in a subsequent com-munication.

SV40sequence arrangement in cell DNA.

To investigatethe stability of SV40 sequences

intransformed cellsand to attempt a

determi-nation of thebasis for the spontaneous evolution

of the transformation revertants, the

high-mo-lecular-weight

DNAs from each of the cell lines

were analyzed byrestriction digestion followed

byhybridizationto32P-labeledSV40DNA.

Fig-ure 1 shows results of hybridization of SV40

DNA to cell DNAsdigested by TaqI. This

re-striction endonuclease cleaves once within the

SV40 genome. The pattern of integration of

SV40 sequences in the original transformant,

B1-0, was complex (Fig. 1). Therewere over a

dozen distinct SV40-specific fragments, and

these varied considerably in radioautographic

intensity. Onefragmenthadthe

mobility

of

lin-earSV40DNA(5.2 kilobases[kb])and could be

derivedfrom either free viral DNAor

integrated

viral DNA present inatandem array. The

large

numberof distinct

SV40-specific

fragments

pres-entmayreflectmultiplerearrangementsofviral

I -9

la cj .0 0

"Y

IN

"I

I

m

E

ai

a

ro

on November 10, 2019 by guest

http://jvi.asm.org/

[image:3.495.117.404.71.407.2]

sequences, multiple sites of integration of viral

DNA within cellDNA, or both. The variation in

intensity of these fragmentsmay be due to

dif-ferent amounts ofSV40DNA in different

frag-ments, reiteration of someSV40-containing

frag-ments, heterogeneity of the transformed-cell

population, or all three. Additional analyses

(datanotpresented) revealed that the intensity

of certain fragments relative to others varied

with the number of passages cells had undergone

atthe time of DNA extraction.

To further investigate the nature of SV40

integration in B1-0 cells, the arrangement of

viralsequences inlinesBl-la,-b, -c,and-d was

analyzed by TaqI digestion and hybridization.

Theselines,allof which wereclonedfromB1-0,

showedasimplificationinpattern of SV40DNA

integration (Fig. 1). Each clone retained some

SV40-specific fragments havingthe same

elec-trophoretic mobilityasfragmentsin theparental

line (e.g., the 7.5-, 2.8-, 1.8-, and 1.5-kb, aswell

as the5.2-kb,fragments)but each hadlostother

fragments (e.g., the 3.6-kb fragment in Bl-ld

and Bl-lc and the 12.0- and1.7-kb fragments in

all Bl-1lines).Inaddition,eachclonecontained

adistinctmajor bandnotdetectedinthe parent

(e.g., the 8.0- and 3.5-, 3.5- and 1.6-, 3.2-, and

1.9-kb fragments in Bl-ld, -c, -b, and -a,

respec-tively). Moreover, the integration pattern in

each clone was distinctfrom that found in the

other three.

It appears, then, that line B1-0 is not a

ho-mogeneous population. Heterogeneity of this

line could result from the presence of several

distinct primary transformants. This does not

appear tobealikely explanation,since B1-0 was

selected as awell-isolated colony from cells

in-fected at arelatively lowvirus-to-cell ratio. On

the other hand, the observed heterogeneity

could result from rearrangement of viral

se-quences afterintegration. Indeed, thepresence

of certaincommonbandsinall four B1-1 clones

andtheparental transformantindicates that the

four clones have in fact evolved from a

single

primary transformant and are not merely

de-rivedfromamixedpopulation of several primary

transformants. This conclusion is further

sub-stantiatedbytheresults of

analysis

ofthe DNA

inlines derivedfrom one of thesecondary cell

lines, Bl-ld. The pattern of TaqI-generated

DNAfragmentscontainingSV40sequenceshad

been further altered in linesB1-2a, -b, -c (Fig.

1). Each line retained at least one fragment

whosemobility corresponded tothat of a

frag-ment in theBl-ld and B1-0 progenitors. Each

had lost at leastoneparental fragment,andtwo

lines contained fragments not detected in the

Bl-ld parent. The transformationrevertant

B1-2a and its subclone B1-3 had a

single

major

SV40-specific fragment. Analysis of SV40

se-quencesinthe clonesderived from asecondary

line, then, revealedalterations qualitatively

sim-ilar to those observed in the initial cloning. All

seven clones and subclones derived from B1-0

showed related but distinct patterns of SV40

sequencearrangement. Thus, the diverse

inte-gration patterns in the various cell lines

ap-peared to have a common origin.

To confirm the results obtained by TaqI

digestionofthe various cellDNAs and to

deter-mine the number ofSV40 integration sites in

each cellline, the cellDNAs were analyzed by

BglII digestion.Thisenzymedoesnotcleavethe

SV40genome. Cell DNAfragments containing

SV40 sequences are therefore generated when

BglIIcutscell sequencesadjacent tointegrated

viral DNA; the number of SV40-specific bands

in agivencellline DNA indicates the number of

sitesat whichSV40 was integrated within that

cell line. Figure2shows theresult of

hybridiza-tion of SV40 DNA to cell DNAs digested by

BglII.Nofreeviral genomes weredetected,since

noSV40-specific fragmentswere seenmigrating

with the mobility of closed circular (I), open

circular (II), or linear (L) SV40 DNA. As

ex-pected,the number ofSV40-specific fragments

present in the variouscelllines wassmallerthan

wasthecasewithTaqI cleavage.Inthemajority

oflines, however, SV40 DNA wasintegratedat

morethanonesite.Again,aswithTaqI analysis,

anevolutioninthearrangement ofSV40-specific

fragments was observed. The various cell lines

had distinctive but related SV40 sequence

ar-rangements. In most cases, sisterclonesshared

some common parental fragments, lacked

oth-ers, and containedone or more newfragments.

The evolution of new arrangements of SV40

sequencesobservedin boththe TaqIandBglII

analyses indicated that at least some of the

observed arrangements of

integrated

SV40

se-quences wereunstable.

Assessment of free viral DNA. The

ob-served

reorganization

of

integrated

viral

se-quences inSV40-transformedcells

might

be

me-diated by free SV40 genomes.

TaqI

digestion

(seeFig. 1)of DNA from all cell lines except the

revertants

generated

SV40-specific fragments

having the

electrophoretic

mobility

of

unit-length linear SV40 DNA. The

linear-length

SV40-specific fragments observed here

might

haveresulted from TaqI

digestion

of either free

viral genomes or

tandemly

integrated

SV40

DNA. That the latter situation is the case is

indicated by

analysis

of the

SV40-specific

cell

DNAfragments

generated by

BglIl.

After

diges-tion ofcell DNAs with

BglII,

no

SV40-specific

bands

comigrating

withSV40form

I,

form

II,

or

linear DNA was observed

(Fig. 2);

unit-length

on November 10, 2019 by guest

http://jvi.asm.org/

Kb

27.s-

15-L

ID

I

1

I I I

E

>:

m

m

mn

mn

di

m

n

I

23-2..O

FIG. 2. Autoradiogram of 3P-labeledSV40 DNAhybridizedtocell DNAfragments generated byendo

R-BglIIandseparated byelectrophoresis in0.7%agarose.Slot M contains markersofclosed circular(I),linear

(L), and open circular(II)SV40 DNAs which have been addedtoBglII-digestedBALBIc 3T3 DNA. The size (in kilobases) and positionofA DNAHindIIIfragmentsinthesamegelareshownattheleft.On shorter exposure,six distinctSV40-specific bands larger than6.7kbaredetectable in line Bl-0.

SV40fragments, then, weregenerated by TaqI

cleavage oftandem arrays ofSV40 sequences.

ThecellDNAsused inthese analyses, however,

were prepared byspooling from isopropanol, a

procedure which might exclude the extraction of

low-molecular-weightviral DNA. To circumvent

this possibility, DNA extracted from several

lineswas concentrated by ethanol precipitation

rather thanbyspooling and subjected to

diges-tionbyBglII. Nofree viral DNA was detected

in any ofthe lines tested (Fig. 3). This procedure

should becapable of detecting an average of one

free viral genome per cell. We cannot rule out

the presenceof lesser amounts of free viral DNA.

DISCUSSION

The arrangement ofintegrated viral sequences

in a primary isolate of SV40-transformed cells

andin thecellscloned from it has been

analyzed.

This analysis has revealed differences in the

arrangementof viral sequences when individual

clonesare

compared

with each other and with

their progenitors. Thesedifferences are

attrib-utedto rearrangements of the

integrated

viral

sequencessubsequenttotheprimary

integration

event, and not to different sequence

arrange-mentsin a mixed populationof

primary

trans-formants,forseveralreasons.First,the

primary

transformantwasselectedas awell-isolated

col-ony derived from cells infected at a

relatively

low virus-to-cell ratio. Ifno rearrangement

oc-curred afterintegration,thento accountfor the

various patterns ofintegration observed in the

variouscell lines, the primary

colony

of

trans-formation would havetobe amixture ofmore

than eight different transformants.

Second,

on November 10, 2019 by guest

http://jvi.asm.org/

[image:5.495.139.383.72.423.2]

REARRANGEMENT OF SV40 SEQUENCES 877

a

O

',

co

"I

m lm m m

7-I

FIG. 3. SV40-specificDNAinBl-Oandit8 serially

clonedderivativelines. The analysiswasperformed asinFig.2 exceptthatcellDNAswerenotspooled fromisopropanol (see Materials and Methods) before digestionwithendoR.BglII. Thespoolingstepwas

omittedtoenhance thepossibilityof recovering free

viralgenomes. SlotMcontains closedcircular (I),

linear(L),andopencircular(I) SV40DNA.

clone Bl-ldwasderived from theprimary

trans-formantby endpoint dilution cloning, yetgave

rise to four subclones with different viral

se-quence arrangements. Finally, the presence in

various clones ofcommonSV40-containing

frag-mentssuggests that these clones share a com-mon ancestry. The evolution of new

arrange-mentsofSV40DNA in these various cell lines

indicates,then, thatatleastsomearrangements

ofintegrated SV40sequences areunstable.

Ap-parently, during the establishment of at least

thetwoprogenitor lines,B1-0andBl-ld,certain

viral sequences have been rearranged in some

cells. The minorspecies ofSV40-specific frag-mentsinthe digestsofparental DNA (see Fig.

1)presumably reflect such arrangements.

Rear-rangementsbecome more obvious when the viral

sequencesin progeny clones are analyzed.

In the series of cells analyzed here, certain

arrangements of SV40 sequences appear to be

more stable than others. Forexample,the

BglII-generated fragment (see Fig. 2) migrating with

asize of 6.3 kbappears in all four Bl-1 lines and

inB1-2aandB1-3.Retention of certain

arrange-ments of viralsequences might afford the host

cell aselective growthadvantage. Itis not

read-ily apparent, however, that this is the case for

the example just cited. Lines B1-2 and B1-3

contain SV40 sequences integrated at a single

site in the host DNA (i.e., within the 6.3-kb

BglII fragment). T antigens, however, are not

expressed in these linesbecausetheSV40 early

coding region is not intact (M. A. Bender and

W. W.Brockman, unpublisheddata).As aresult,

B1-2 and B1-3 havegrowthproperties similarto

thoseofnormalBALB/c3T3cells.Ontheother

hand, whether SV40 sequences are retained or

lost may be determined by the nature of the

cellular sequences to which the SV40 DNA is

linked orby the absence or presence of tandem

duplicationsof viralsequences (see below).

ComparisonofSV40-specificrestrictiondigest

patterns of theprogeny clonesdescribedinthis

communication relative with those ofthe

pri-mary transfornantreveals two types of

altera-tions: (i)loss of certainSV40-specific fragments

and(ii) acquisitionof new

SV40-containing

frag-ments. Loss of SV40-specific fragments could

result from loss ofentirechromosomesorfrom

deletion of smaller segments ofDNA. The

ap-pearance of new SV40-containing fragments

may result from point mutation at restriction

sites or from deletions within either SV40 or

flankingcellsequences. On theother

hand,

new

fragments might arise from transposition of

SV40sequenceseither aloneoraccompanied

by

flanking

cell sequences. The presence of

tan-demly integrated SV40 sequences in line B1-0

and its progeny clones suggests that

complete

copies of viral DNA might be released by

ho-mologous recombination between

duplicated

re-gions of viral DNA

(13).

Free SV40 genomes

generatedinthis fashion

might

then

integrate

at

othersiteswithin the cell DNA. In thecaseof

the related

virus,

polyoma,

free viral genomes

(1, 2) and their apparent excision from tandem

arrays of integrated viral sequences

(1)

have

beendetected in

semipermissive

ratcells. In the

present and earlier studies (17),

however,

free

viral DNA isnot

physically

detectable in

SV40-transformed

nonpermissive

mouse cells.

Daya-Grosjean and Monier

(10),

in contrast,

using

a

biologicalassay have

reported

the

detection,

in

oneSV40-transformedmouse

line,

of low levels

on November 10, 2019 by guest

http://jvi.asm.org/

[image:6.495.83.204.77.422.2]

(<0.005 molecule/cell) of infectious material

having the buoyant densityof form ISV40 DNA.

Although we are unable to detect free SV40

genomes in the cells we have examined, we

cannotexcludethe possibility that the observed

rearrangements of integrated SV40 sequences

aremediatedby excisionorintegration (or both)

ofasmallnumber of free viral genomes.

Alter-natively, transposition of SV40 sequencesmight

be facilitatedby integration adjacentto

reiter-ated sequences ormovable elements in the cell

DNA. In such cases, rearrangement of SV40

sequences might result from (i) recombination

facilitated by homology between flanking cell

DNA andsequences repeated elsewhere inthe

cell genome or (ii) transposition of a movable

element together with covalently linked viral

sequences. Ifsuch mechanisms wereoperative,

it might be expected that the arrangement of

integrated SV40 in some transformants would

be stable whereas those in otherswould be

la-bile, depending on the proximity of the viral

sequences to reiterated sequences or movable

elements.Indeed,asnotedabove,in the present

studysomeSV40sequence arrangements appear

to be more stable than others. Theprevalence

of labile associations between SV40 and cell

DNA in transformed cellsaswellasthe

relation-ship oflability ofSV40 integration patterns to

the presence oftandemly integratedviral DNA

andtothe nature ofadjacent cell sequences is

presently under investigation.

In previous studiesofthe integration of SV40

DNA in rodent cellDNA (4, 5, 14; Ketner and

Kelly, in press), transformants, which after

re-cloning atleast once were found to have

rela-tivelysimplepatternsofSV40 integration, were

analyzed in some detail. Conclusions regarding

thelack of site specificity for SV40 integration

were based on the assumption that

rearrange-mentof the integrated viral DNA does not occur

after the primary integration event. More

re-cently,twolaboratories have found thatcellular

DNA sequences flanking integrated SV40

se-quencesare notcolinear in thenontransformed

parentalcell line(4; P.Mounts and T. J.Kelly,

personal communication), indicating that

rear-rangementofcellsequenceshas occurred

coin-cident with or after the initial recombination

betweenSV40and cell DNAs. This finding,

to-getherwiththe data presented in this

commu-nication, suggests that the arrangement of SV40

sequences present in serially cloned, stable

transformants may represent the cumulative

re-sult ofeventsoccurring after the primary

inte-grationevent.

ACKNOWLEDGMENTS

WethankJoanChristensen forexcellent technical

assist-ance and Phoebe Mounts and T. J. Kelly for theiraid in mastering theintricacies of DNA transfer andhybridization.

M.A.B.istherecipient ofanF.G.Novypredoctoral fellow-ship.This work wassupported byPublic Health Servicegrant CA-19816 from theNational Institutes of Health.

LITERATURE CITED

1. Basilico, C., S. Gattoni, D. Zouzias, and G. Della Valle. 1979. Lossofintegrated viral DNA sequencesin polyoma-transformedcells is associated with anactive viral A function. Cell 17:645-659.

2. Birg, F.,R.Dulbecco,M.Fried,and R. Kamen. 1979. State and organization of polyoma virus DNA se-quences intransformedratcell lines. J. Virol. 29:633-648.

3. Bolivar, F., R. Rodriguez, M. Betlach, and H. W. Boyer.1977.Constructionand characterization ofnew vehicles. I.Ampicillin-resistant derivatives of the plas-midpMB9.Gene2:75-93.

4. Botchan, M., S.Stringer, T. Mitchison, andJ. Sam-brook. 1980. Integration and excisionofSV40 DNA from thechromosome of atransformed cell. Cell20: 143-152.

5. Botchan,M., W. Topp, andJ.Sambrook. 1976.The arrangement ofSV40 sequencesinthe DNA of trans-formedcells. Cell9:269-287.

6. Brockman,W. W.1978.Transformation ofBALB/c-3T3

cells by tsA mutantsofsimianvirus 40: temperature sensitivity of the transformedphenotype and retrans-formationbywild-type virus. J. Virol. 25:860-870. 7. Brockman,W.W.,andD.Nathans.1974.The isolation

of simian virus 40 variants with specifically altered genomes.Proc.Natl. Acad. Sci. U.S.A. 71:942-946. 8. Dagaiczyk, A.,and S. L. C. Woo. 1980.Mappingand

cloning ofeukaryoticgenes, p.1-38.In W. T. Schrader andB. W.O'Malley (ed.), Laboratorymethods manual for hormone action andmolecularendocrinology, 4th ed. Texas MedicalCenter,Houston.

9. Danna, K.J.,and D. Nathans.1971.Specific cleavage of simian virus40DNAbyrestrictionendonuclease of Hemophilusinfluenzae. Proc. Natl.Acad. Sci.U.S.A. 68:2913-2917.

10. Daya-Grosjean, L.,and R. Monier. 1978.Presence of freeviral DNA in simianvirus40-transformed nonpro-ducer cells.J.Virol. 27:307-312.

11. Gross-Bellard, M.,P.Oudet,and P. Chambon.1973. Isolation ofhigh molecularweight DNAfrom mam-malian cells. Eur.J.Biochem. 36:32-38.

12. Kelley,S.,M. A.Bender,and W. W. Brockman. 1980. TransformationofBalb/c-3T3 cells bytsA mutants of simian virus 40: effect of transformation technique on the transformedphenotype. J. Virol. 33:550-552. 13. Kelly,T.J.,A. M.Lewis, A. S. Levine, andS.Siegel.

1974.Structure oftwoadenovirussimian virus 40 hy-bridswhich contain the entireSV40 genome. J. Mol. Biol.89:113-126.

14. Ketner,G., and T. J. Kelly. 1976. Integrated simian virus 40sequencesin transformedcell DNA:analysis usingrestrictionendonucleases. Proc. Natl. Acad. Sci. U.S.A.73:1102-1106.

15. Labarca,C.,and K.Paigen.1980.Asimple, rapid,and sensitive DNAassayprocedure. Anal. Biochem. 102: 344-352.

16.Maniatis, T., A.Jeffrey, andD. G.Kleid.1975. Nu-cleotidesequence of therightward operatorofphage A. Proc. Natl. Acad. Sci.U.S.A.72:1184-1188. 17. Sambrook, J., H.Westphal,P. R.Srinivasan, andR.

Dulbecco. 1968. Theintegratedstateof viral DNA in SV40-transformedcells. Proc. Natl. Acad. Sci. U.S.A. 60:1288-1295.

18. Southern,E. 1975. Detection ofspecific sequences among DNA fragments separated by gel electrophoresis. J. Mol. Biol. 98:503-515.

on November 10, 2019 by guest

http://jvi.asm.org/

19.Tooze,J. 1980.Transformation by simian virus40 and 20. Wensink,P.C.,D.J.Finnegan,J. E.Donelson,and polyomavirus, p. 205-296. In The molecularbiology of D. S. Hogness. 1974. A system for mapping DNA tumorviruses, vol. 2, 2nd ed. ColdSpring Harbor Lab- sequencesin thechromosomes ofDrosophila melano-oratory, ColdSpring Harbor,N.Y. gaster.Cell3:315-325.

on November 10, 2019 by guest

http://jvi.asm.org/