Identification of adenovirus genes that require template replication for expression.

0022-538X/83/060737-12$02.00/0

Copyright © 1983, AmericanSocietyforMicrobiology

Identification of Adenovirus Genes

That

Require

Template

Replication for

Expression

LYLE D.CROSSLANDtANDHESCHEL J. RASKAS*

Department of Pathology, Washington University School of Medicine, St. Louis, Missouri 63110

Received 15 November 1982/Accepted 1 March 1983

Therelationshipbetween adenovirus type 2 DNA replication and expression of

intermediate stage viral genes was investigated. The 1.03-kilobase mRNA from

early region lb (Elb)and the mRNAscoding for proteins IX and IVa2 were first

detectedbetween 6and8 hpostinfection. Inhibition of viral DNAreplication with

hydroxyurea prevented expressionof the IX andIVa2 mRNAs, but not of the Elb

mRNA. Pulse-labelingexperiments demonstrated that the block ofIX and IVa2

expression in hydroxyurea-treated cells was at the level oftranscription. By a

series of superinfectionexperiments, it was determined that the viral and cellular

factors present during the late stage of adenovirus infection are insufficient to

activate IX geneexpression. The viral DNAtemplate must firstreplicatebefore

IXtranscription canbegin.

Lytic infection by humanadenoviruses is

di-vided into four temporal stages of gene

expres-sion: pre-early, early, intermediate, and late (5,

19, 34).Theassignmentofgenes to aparticular

stagehas been basedonthetime ofappearance

of the gene product. mRNAs that appear

be-tween 6 and 8 h after infection have been

as-signedtotheintermediate class(34, 38).

Includ-edin this classarethe0.52-kilobase (kb)mRNA

from region Ela, the 1.03-kbmRNAfrom Elb,

and themRNAsforprotein IVa2 and proteinIX

(Fig. 1). Previous work has indicated that the

appearance at intermediate times ofthe0.52-kb

and 1.03-kb mRNAs is due to increased

cyto-plasmic stability ofmRNA species producedin

the nucleus at early times (41). The IX gene,

however, is regulated atthetranscription level,

andits promoterisnotactive untilintermediate

times (42). The mode ofregulation of the IVa2

gene has notbeen determined.

Thepromoterfor themajor late transcription

unit (Fig. 1) is active at early times, although

premature termination of the transcripts pre-vents formationof the mRNAs for late regions L2 through L5 (2, 29, 35). The onset of viral

DNA replication has traditionally marked the

beginning ofthe late stage of adenovirus

infec-tion (8). Thomas and Mathews (39)

demonstrat-ed that the gene products of late regions L2

throughL5 areexpressedonlyfrom a replicated

DNAtemplate;eventhough themajor late

pro-moterisactiveonparental genomes,replication

tPresent address: The Biological Laboratories, Harvard University,Cambridge,MA 02138.

and some coincident physical change in the

template are required for expression of the

downstream late regions. The relationship of

intermediate gene expression to DNA

replica-tion is lesscertain. Although intermediate gene

expression begins between6and8 hafter

infec-tion (about thesametime thatviral DNA

repli-cation begins), evidence of IX and IVa2 gene

expressionhas been foundincells infected inthe

presenceofcytosine arabinoside, aninhibitorof

DNAreplication (24, 30). Itiscertain, however,

that theexpression ofIXandIVa2mRNAis at

itsmaximum levelatlate timesafter viral DNA

replication has begun (24, 38, 42). Because all

fiveearly transcription units (28) and the major

latetranscription unit (2, 29, 35)areactiveby3 h

postinfection, while the IX transcription unit is

inactive atthis time (42), the IXpromoter may

be the last viral promoter to be activated. The

exact times of activation of the other known

viral promoters, IVa2 and the alternative E2

promoter at map position 72, are notknown.

The goal of this work was to determine the

relationshipbetween viral DNAreplication and

intermediate gene expression, focusing on the

geneforproteinIX. Thefirstsetofexperiments

was designed to determine how hydroxyurea

(HU) blockade ofDNA synthesis affects

inter-mediate gene expression. HU was chosen

be-cause at10 mM itblocks detectable DNA

syn-thesis,does notbreak downduringthe courseof

the experiment (39), and does not significantly

affect RNAsynthesis. Thesecondsetof experi-mentsexamined thebasis ofthedependence of

IXgeneexpressionupon viral DNAreplication.

737

on November 10, 2019 by guest

http://jvi.asm.org/

738 CROSSLAND AND RASKAS

E1 B EIA

1.02 kbi-' A _M,l

.88 kbi 'A .4

1.03kkbbIV

.52kb_,' "

*.49kbl

IZ2

iQ

Ad2

dl313

Hpa E Hind m G

C2 cl

---4C1

HindM C

Sma F

FIG. 1. Map positions of Ad2 left-end mRNAs (6, 9, 21) and DNA fragments cloned foruse as probes. Spliced-out regions of mRNAsareindicatedby dashed lines, and theapproximatesize in kbof each mRNA is given. ML is the major late promoter (13, 44). The hashed box indicates theregiondeleted indl313 (18). All DNA fragments were inserted into pBR322, with the exception of Sma F whichwasinserted into M13 phage, andwere

used asrecombinant DNAs. The M13-Sma F recombinant used released the left strand of Sma F in phage particles.

MATERIALSAND METHODS

Cells and virus. Cells were grown as monolayer

cultures in Dulbeccomodified Eaglemedium (DME)

containing10%calforhorseserum(KC Biologicals)

for KB or 293 cells, respectively. Cells were split 2 days before theexperiments,tobeapproximately 70%

confluent when infected. Virus stockswereprepared

asdescribed (37). tsl25 virus wasobtained from Jim Williams, d1313 was obtained from TomShenk, and Ad2+ ND5wasobtained from Andrew Lewis.

Infectionconditions. Cells for infectionweregrown in 150-mm dishes. Themediumwasaspirated off,and the monolayers were washed with serum-free DME. Virus was added in 5 ml of DME per plate, and absorptionwascontinued for1 hwithgentleagitation every 15 min.Theinoculumwasremoved, and 30 ml of DMEcontaining10%serumand, insomecases, 10 mMHUwasaddedtoeachplate.

Thesuperinfection protocolwasidenticaltothatfor

infection, except thata45-min absorptionwasused. Greatcarewasneededtoavoiddetachingthe 293 cells whenadding thesuperinfection inoculum and overlay medium.

Recombinant DNAprobes. The left strandof the Ad2

SmaF fragment (map units 11.5-18.5) cloned in the

single-strandedbacteriophageM13 wasobtained from MarkBoguski. The HpaE (map units 0-4.4), HindIII C (7.9-17.0), EcoRI C (89.7-100), HindIll H (72.8-79.9), andEcoRIFl (72.8-75.9)fragments of

adenovi-rustype2(Ad2) cloned in pBR322wereobtained from Beth Ladin, Elizabeth Slattery, Michael Tigges, D.

Huang, and Richard Rosenthal, respectively. The

HindIIl Cl (11.5-17.0) and C2 (7.9-11.1) fragments

were obtained by digesting the original HindIII C clone with SmaI, which cuts at 11.1 and 11.5. The linear DNAwasthendigested withasecond enzyme

that cutstheplasmid sequenceneartheinsert.EcoRI

digestionwasusedtoexcise the11.5-17.0region,and

BamI digestion was used to excise 7.9-11.1. The

DNAs werethen recircularized with T4 DNAligase and usedtotransform Escherichiacoli HB101.

Pulse-labeling RNA. Topulse-label cells, the

medi-umwasaspirated off and replaced with 8 ml per plate of DME with serum,containing300,uCi of13H]uridine

permland, whereappropriate,10mMHU.After 5or 10minof incubation, the plateswereplacedonice, the labeling solution was removed, and the cells were scraped into coldphosphate-buffered saline.

RNA isolation. Cytoplasmic, polyadenylic acid

[poly(A)]-containing RNAwas isolatedby Nonidet

P-40leakage in isotonic buffer, extraction of the post-nuclear supernatant with phenol-chloroform-isoamyl

alcohol, andoligodeoxythymidylate-cellulose chroma-tography, allasdescribedpreviously (25).

Totalcell RNAwasisolatedbyanadaption of the procedure of Sauerbrier and Brautigam (32). Harvested cells were pelleted from coldphosphate-buffered

sa-line and taken up in1 mloflysisbuffer (1% sodium dodecyl sulfate, 1 mM EDTA, 0.1 M NaCl, 0.1% diethyloxydiformate, and 10 mM Tris, pH 7.5). The mixturewasvortexedvigorously,heatedat85°C for 30

sand cooledonice.An8-mlvolumeof8.5molal CsCl inlysis buffercontaining 0.1% sodium dodecyl sulfate

was added, and the solution was pipetted several times. Thelysed cell mixturewaslayeredover2.5ml of 9.5 molal CsCl inanSW41 tube andcentrifugedat

26,000 rpm for 22 h at 8°C in an SW41 rotor. The

supernatantwas aspirated, and the RNA pellet was

suspended inwaterandprecipitated bytheaddition of sodiumacetate to0.2Mand 2.5volumes of ethanol.

RNA blots. RNA samples for Northern analysis were incubatedwith 1 Mglyoxal in 10 mMNa2PO4 (pH6.8)at50°C for1 h(26) and thenelectrophoresed

in 1.1 or1.4%agarosegels.Inall cases,RNAfroman

equivalentnumberof cellswasloadedoneachlaneof

agivengel.The RNAsweretransferredto nitrocellu-lose as described by Thomas (40) and hybridized to

nick-translatedDNAprobes.

RNAgradients. Pulse-labeledRNAforsucrose gra-dientfractionationwastakenupin 250,ulofasolution containing80%formamide,2 mMEDTA,and40mM

PIPES [piperazine-N,N'-bis(2-ethanesulfonic acid)]

(pH7.4). The mixturewasheatedat68°C for7.5min

i

on November 10, 2019 by guest

http://jvi.asm.org/

and layered on a preformed gradient of 5 to 20%

sucrose in the same buffer solution. The gradients werecentrifuged in an SW41 rotor at 30,000 rpm for 48 hat 4°C, and350-,ul fractions were collected from the bottom of the tube. The RNA in each fraction was precipitated with ethanol, taken up in buffer, and partiallyhydrolyzed with alkali as described by Nev-ins (27). Selected fractions were then hybridized to DNA bound tonitrocellulose filters as described previ-ously (25).

DNAblots. To parepare DNA forSouthernanalysis, nuclei were isolated as described (16) andlysedby the addition of DNA extraction buffer (0.1 M NaCl, 10 mMEDTA,0.5% sodium dodecyl sulfate, 50 mM Tris, pH 8.5). The mixture was digested with 1 mg of pronase B per mlovernight at 37°C, extracted with phenol-chloroform (1:1) three times, and precipitated withethanol. The DNA was taken up in anappropriate buffer, digested with the restriction enzyme KpnI, and electrophoresedon a1% agarosegel. The DNA bands were transferred to nitrocellulose by the method of Southern (36) and hybridizedto nick-translated Ad2 DNA.

RESULTS

Time course ofintermediate gene expression.

Asa firststeptoward studying their regulation,

the time of appearance of the intermediate

mRNAs was examined. To determine the

steady-state concentration of the mRNAsinthe

cytoplasmasinfectionprogressed,

poly(A)-con-tainingcytoplasmicRNAwasisolated from2 x

107 KB cells infected for4, 6, 8, or 10h. The

RNA was subjected to electrophoresis in an

agarose gel and then transferred to

nitrocellu-lose. The Ad2 HindlIl C fragment (7.9 to 17.0

mapunits), cloned in pBR322,was

nick-translat-ed and hybridized to the RNA blot. The three

intermediate mRNAs that hybridize to this

probe, the 1.03-kb Elb mRNA and the IXand

IVa2 mRNAs,werefirstdetected by this

analy-sis between6and 8hpostinfection(Fig. 2, lane

3). As expected, the 2.28-kb early mRNA from

Elb was detected at 4 h postinfection (Fig. 2,

lane 1).

Intermediategene expression in theabsenceof DNA replication. When DNA synthesis was

blockedby theaddition of 10mM HUafter virus

absorption, the normally large quantity of IX

mRNApresentin 2 x

107

postinfec-tionwasundetectable(Fig. 2, lanes 8 and 9).In

contrast, the 2.28-kb and 1.03-kb Elb mRNAs werebothpresent in HU-treated cells,although

the1.03-kbspecieswasreduced inquantity.The

HindIlI C2probe (7.9-11.1)used inthis

experi-mentdoesnothybridize IVa2 mRNA. When the

infectionwascontinued for15h and theHindIII

Cprobe (7.9-17.0)wasusedin thehybridization

analysis,neither the IX nor theIVa2 mRNA was

detected inHU-blockedcells(Fig.2, lanes 5and

6). Similarexperiments(not shown) suggest that

the 0.52-kb Ela mRNA has the same time

course and resistance to HU blockage as the

1.03-kb Elb mRNA. Although the time course

of appearance of the four intermediate mRNAs was the same and thelevels ofthe0.52-kb Ela

and 1.03-kbElb mRNAs were reduced by HU,

only the IX and IVa2 mRNAs were dependent

uponviralDNAreplication for expressionin the

cytoplasm.

Activation of IX and IVa2expression is inde-pendent of template number. At 12 h

postinfec-tion, cells infected in thepresence ofHU

con-tained fewer templates for viral transcription

than control cells infected without drugs. To

ensure that this difference in template number was notresponsible forourinabilitytodetectIX

mRNAinHU-treated cells, infectionconditions

weredetermined that would place

approximate-ly thesamenumber oftemplates in the nuclei of

thecontrol andHU-treated cells. Fourplates of

KBcellswereinfectedat amultiplicityof

infec-tion (MOI) of 10, and at various times after

infection nuclear DNA was isolated. A fifth

plate was infected at an MOI of 500 in the

presence of HU, and at 10 h postinfection its

nuclear DNA was isolated. Equal amounts of

these DNA samples were bound to

nitrocellu-lose filters andhybridized tosaturatingamounts

3

L 3 5 6

S

FIG. 2. Time course of intermediate mRNA

expression. Cytoplasmic, poly(A)-containing RNA was isolated from cells infected for 4, 6, 8, or 10 h

(lanes1through4), for 12 h(lanes8 and9),orfor 15 h (lanes5and6). Lanes5and9arefrom cells infected in

the presence of10 mM HU. Lane 7is a 14C-labeled

rRNA marker. The mRNAswere electrophoresedin

an agarose gel, transferredto nitrocellulose, and hy-bridizedto32P-labeledHindIII C(lanes1through 7)or

HindIII C2(lanes 8 and 9)DNA. Theadditional band

comigrating with the 18SrRNAmarker inlanes3, 4,

and 5 is occasionally seen with several Ad2 DNA

probes.Presumably, it results fromRNAaggregation

due to rRNA contamination of poly(A)-containing

RNAandsubsequent overloading of thisregion ofthe

gel. VOL 46,1983

-3, 61 -r

i

;-I .k,

a

06.

-OW.

a

on November 10, 2019 by guest

http://jvi.asm.org/

- 21

o 18

z

0

0 .

HOURS POST INFECTION

FIG. 3. Quantitation of Ad2 DNA in infected cells. The amount of Ad2 DNA per cell was determined by binding nuclear DNA from 3.3 x104infected KB cells tonitrocellulose andhybridizingto saturatingamounts of3H-labeled Ad2 DNA. Under identical conditions, 3.9x 10-3jigoffilter-bound Ad2 DNA (the equivalent of 9.4 x 10'Ad2 DNA molecules) hybridized 24,400 cpm.Itwasconcluded that each 8.6 cpm hybridized to the experimental filters represented one Ad2 DNA molecule per cell. Symbols: 0,DNAfrom cells infect-edatanMOIof 10;0,DNAfrom cellsinfected at an MOI of 500 in the presence of HU.

of3H-labeled Ad2 DNA (Fig. 3). Theopencircle

inthe center ofFig. 3 indicates the number of

adenovirus DNA moleculespercellat10 hafter

aninfection at 500 MOI in the presence of HU. Theamountof intranuclear Ad2DNApresent at

12 h after a10 MOI infection without drugs is

approximatelyequaltotheamountpresentafter

infection at an MOI of500 in the presence of

HU.

Basedontheresults described above,

experi-ments were performed to determine the

influ-enceoftemplatenumberonIXandIVa2

expres-sion. KB cells were infected at an MOI of10

withnodrugor at anMOI of500in thepresence

of HU.At12 hpostinfection, RNAwasisolated

for blot hybridization analysis. When ablot of

cytoplasmic,poly(A)-containingRNAfrom2 x

107cellswas hybridized to the HindlIl Cprobe

(7.9-17.0), the IX and IVa2 mRNAs were only

detected indrug-free control cells, whereas the

Elb 2.28-kb and1.03-kbmRNAsweredetected

in control andHU-treated cells (Fig. 4, lanes 1

and2) (the 1.03-kb bandwastoofaint to

repro-duce in the

1photograph).

fragment C2 (7.9-11.1) (Fig. 4, lanes 4 and5) a

similar resultwasobserved,except that the level

ofthe 1.03-kb mRNAwas relatively greater in

total cell RNA. Hybridization ofthe total cell

RNA blots to the EcoRI C (89.7-100) (Fig. 4,

lanes 7 and 8) probe demonstrated that the

predominant earlymRNAsfromregionE4were

present in equal or greater amounts in the HU-treated cells.

Newlysynthesized IX and IVa2 mRNA can only be detected when viral DNA replication occurs.

Regulation ofmRNAlevels in the cytoplasmcan

occur atseveral levels (11).Todetermine

wheth-erthe absence of IX and IVa2 mRNA in

HU-treatedcellswasbecause of lack of transcription

of these genes, newly synthesized molecules

were quantitated after a briefpulse label.

Be-causethe entiresequenceofIX mRNAisshared

with the Elb transcript, no DNA probe can

specifically hybridize onlyIX mRNA. To

quan-titate IXtranscription it was therefore necessary

to size-fractionate IX mRNA and the Elb

mRNAs before hybridization. KB cells were

again infected at10 MOI without drugs or 500

MOI in thepresence of HU. At12 h

postinfec-tion, when the numberof viral DNAtemplates

in thetwo setsof cellswasequal,thecellswere

pulse-labeled with [3H]uridine for10min. Total

cell RNA was isolated and fractionated by su-crosegradient centrifugation.Tolocatethe frac-tions containing mRNAs of interest, samples of

fractions 9 through 25 were electrophoresed in

anagarosegel, transferredtonitrocellulose, and

hybridized to 32P-labeled HindIII CDNA

(7.9-17.0). Transcription activity was assayed by

hybridization of gradient fractionsto

nitrocellu-lose-bound HindlIl C2DNA(7.9-11.1)to assay

Elb and IX, or SmaF left-strand DNA

(11.5-1 2 3 4 5 6 7 8

L;')5kb *

03D S

FIG. 4. Effect of HU on early and intermediate

mRNA accumulation. KB cells were infected at an

MOI of500 inthe presenceof HU(lanes 1, 4, and 7)or at an MOI of 10 without drugs (lanes 2, 5, and 8).

Twelve hours postinfection, cytoplasmic, poly(A)-containing RNA (lanes 1 and 2) or total cell RNA (lanes 4-8)wasisolated,electrophoresedinanagarose gel, and transferredtonitrocellulose. TheRNAblots

werehybridizedto32P-labeled HindlllC(lanes1and 2), HindIII C2(lanes3through5),orEcoRI E(lanes7

and 8) DNA. Lanes 3 and 6 are 14C-labeled rRNA markers.

VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

18.2) to assay IVa2 (Fig. 5). In both RNA

gradients,asignificant peak of [3H]RNA, about

22Sin size, hybridizedto the HindIlI C2 probe

(7.9-11.1). This peak represents either the Elb precursor (2.36 kb) or the 2.28-kb message, or both. The absence of apeak corresponding to

the 1.03-kb mRNA in either gradient suggests

that 10min is insufficient time for labeling and

splicing ofasignificantamountof theprecursor

to occur.Transcriptionfrom the IX gene,

repre-sentedby the 9S peak (see Fig. 5), and the IVa2

gene, represented by the 17S peak, was not

detectable after a 10-min pulse unless DNA

replicationwasallowed. Similarly, aftera5-min

[3H]uridine pulse, transcription from the IX

gene was not detected unless DNA replication

was allowed (Fig. 6). Other pulse-labeling

ex-periments (not shown) havedemonstrated that

IX mRNAcontinuestobe transcribed if HU is

added after viral DNA replication has begun.

Weconclude that, when viral DNAreplication is

blocked with HU,transcriptionofIXandIVa2 is

reduced atleast90% and isverylikely absent.

Development of asuperinfection protocol.

Su-perinfection experiments were used to

deter-mine how IX transcription is related to viral

DNAreplication. Activation ofthe IX promoter

might be dependent upon trans-acting, viralor

cellular, regulatory factors that are produced

onlyasthevirusbeginsDNAreplication. If this

is thecase, anunreplicated viralgenome

enter-ingapreviously infected, late-stage cell should

begin transcribing the IX gene. Alternatively,

the IX promoter might require activation by a cis-acting change in the template itself during

replication. In this case, a superinfecting viral

genomewould first haveto bereplicatedbefore

it could act as a template forIX transcription,

even if it had entered a late-stage cell. To

distinguish between these possibilities, we

per-formed a series ofsuperinfection experiments

utilizingadenovirus deletion and insertion

mu-tants.

Inall of thesuperinfection experiments, d1313

wasused as the preinfecting virus. d1313 is an

AdSdeletion mutantlackingpartofEl,

includ-ing the IXpromoter(18). The AdS-transformed

human cell line 293 contains integrated AdS

DNAfromregion El (1, 17); d1313 canlytically

infect 293 cells because themissing El functions

areprovided bythe hostcell.However, because

theviralIXpromoterisdeleted and the

integrat-ed IX gene cannot beactivatedbyinfection, no IX mRNA or protein is produced during the

infection(37).

It wasessentialthatd1313-infectedcells be in thelatestageof infectionand thus beproducing

any factors that might be necessary for IX

transcription before superinfection. To

deter-minewhen the late stage begins,d1313-infected

A

--2.28kb

am -1,61blib2)

0 - 1 03kb

* -0.49kb1(E)

8

'-4

x U I

2

28s 18s

I 4s

20 FRACTION NUMBER

30

B

s -2.28kb

-103 kb

T0

x

C-I.

18s

10

4S

FRACTION NUMBER

FIG. 5. Newly synthesized IX and IVa2 mRNA. Cellsinfectedat anMOIof 10 withoutdrugs(A)or at anMOIof 500 in the presence of HU (B)were

pulse-labeled for 10 min with [3H]uridine. Total cell RNA was isolated and fractionated on sucrose gradients, andfractionswerehybridizedtonitrocellulose-bound HindIII C2 DNA (0)or SmaFleft-strand DNA(0).

Thelocations ofthe28S, 18S, and 4S RNA markers in the gradient are indicated. The insets show agarose gels of fractions9through 25, transferredto nitrocellu-lose andhybridizedto32P-labeledHindIIl CDNA.

293 cells were pulse-labeledwith

[3H]thymidine

at various times after infection. DNA, isolated from cellsfromeach timepoint, washybridized

tonitrocellulose-boundAdSDNA(Fig.7). Viral

on November 10, 2019 by guest

http://jvi.asm.org/

A

F

_B

I

03kb-20 21 22 23 24 25 26

1.61kb- so

i

03kb-3

49kb--'A,,R

..t... ..rs

il_

FIG. 6. Newly synthesized IX mRNA. Cells were infected at 500 MOI in the presence of HU (upper panels)

or at10MOI in the absence ofdrugs(lower panels) and were pulse-labeled for 5 min with[3H]uridine.Total cell RNA wasfractionated on sucrose gradients, and fractions were hybridized to nitrocellulose-boundHindlllC2 DNA(A). RNAfromfractions 20 through 26 was electrophoresed in an agarose gel, transferred to nitrocellulose, andhybridized to32P-labeled HindIlI C DNA (B).

DNAreplicationin cells infected withdl313and

the Ela deletion mutant d1312 began at 12 h

postinfection, 6 h later than in wild-type

Ad5-infected cells.

We next demonstrated that prior infection

with d1313 can provide functions needed for a

superinfection toproceed. For this experiment

we utilizedd1313and, as asuperinfectingvirus,

the temperature-sensitive AdS mutant ts125.

ts125 codes for a temperature-sensitive

DNA-binding protein and cannotreplicate viralDNA

at40°C(12,43). Mock-infectedand

d1313-infect-ed293 cells wereshifted from37 to40°Cat13 h

postinfection and superinfected with ts125.

Southern blot analysis (36) ofthe viral DNA in

these cells revealed that mock-infected,

ts125-superinfected cells contained only about as

muchts125 DNAafter 12 h at40°Cas asimilar

preparation blocked with HU (Fig. 8, lanes b

and e). Little viral DNA replication had

oc-curred, and the bands observed were

presum-ably due mostly to parental ts125 DNA.

dl313-infected, tsl25-superinfected cells containDNA

from both viruses.Toobserve thereplicationof

only the superinfecting virus by Southern blot

hybridization, amarker specific for ts125 DNA

wasrequired. Because of the deletion in d1313,

the HindIII G fragment present in Ad5 (and

tsl25) is missing. Prior infection with d1313

allowed significant productionof ts125 DNA in

12h at40°C,asevidencedby thedensity of band

G in lane a versus lane b ofFig. 8. A similar

resultwasobtained whents125wasallowed24 h

toreplicate(Fig. 8,lanes candd).These results

prove that the superinfectingts125 DNAenters

the nucleus, where an early protein coded by

d1313 cancomplement its missingfunction.

Only replicated templates can express IX

mRNA.Thesuperinfection protocolwasutilized to assay IX expression from unreplicated and

replicated templates in a late environment. In

addition toIX, the superinfecting virus should

code foran early RNA (one that is expressed

prior to viral DNA replication) that can be

J. VIROL.

.1 .4

-4.

ell"---k

)6".,

a,4,

cl;

E 'i... r, I t-I

-L,

on November 10, 2019 by guest

http://jvi.asm.org/

ment.Through analysis of the Ad2-SV40 hybrid mRNAs, one can independently confirm that the superinfecting template is transcriptionally ac-tive for anearly promoter. Ad2+ND5 replication was monitored in d1313-infected,

Ad2+ND5-su-perinfected cells by [3H]thymidine

pulse-label-ing and hybridization of cellular DNA to SV40

DNA. Replication of the superinfecting viral

DNA was just detectable at 5 to 7 h

post-superinfection (datanotshown).It wasexpected

that asuperinfectionallowed toproceed for 5 to

7 h before the addition of HU would produce

Ad2+ND5-coded mRNAs whose expression

was dependent upon template replication.

To perform the superinfection experiment,

fourplates of 293 cellswereinfected withd1313,

5

3 6 9 12 15 18

HOURS POST INFECTION

FIG. 7. Viral DNAreplicationin deletion

mutant-infected cells. Infected cells were pulse-labeled with [3H]thymidine for 1 h atthe times indicated. Their

DNA was isolated and hybridized to

nitrocellulose-bound AdS DNA, and the quantity ofhybridizable [3H]DNAwasdetermined. WT, Wild-type virus.

detectedagainstabackground of d1313 mRNAs. Ad2+ND5 is a nondefective derivative of Ad2 carrying a 1.3-kb insertion of simian virus 40

(SV40)DNA in early region 3 (E3) (15, 20, 23).

Early in Ad2+ND5 infection, mRNAs of 1.5 and 2.3 kbweredetectedby RNA blot hybridization withSV40 DNAas aprobe (Fig. 9, lane a). The 2.3-kb mRNA hybridized to HindIl H DNA (72.8-79.9) (Fig. 9, lane e) but notto EcoRI Ft DNA(72.8-75.9) (lane d). Thissuggeststhatthe 2.3-kb mRNA originates at theE3 promoter at

76mapunits. Thestructureof the 1.5-kb mRNA is unknown. Late in AD2+ND5 infection, mRNAsof3.1, 3.5, 4.0, and 6.8 kb hybridizedto

SV40 DNA (Fig. 9, lane b). These species also hybridized to the EcoRI Ft DNA (72.8-75.9) (lane c), suggesting that they result from fusion of late region L4 mRNA bodies with SV40 sequences.

The superinfection protocol consists of infec-tion of 293 cells withd1313, followed by superin-fection with AD2+ND5 in thepresence of HU. Thisprocedure allows observation of the ability

of the unreplicated, superinfecting template to

transcribe the IXgene while in a late

environ-__I A

fe

J.

;,

3

H.Ciz!R S

m2 C4 Xr b c de f ' ' c

i.

Lm a

- - -

-37

-A

-B

-D

--Fp

G

H

me 4 go -I

FIG. 8. d1313 preinfectioncan complement

defec-tive early function ofsuperinfecting ts125. Line 293 cells were preinfectedwithd1313 (aandc) or

mock-preinfected (bandd).After 13 hat37°C,the cellswere

shiftedto40°C, superinfectedwithts125,and incubat-ed for 12 h(aandb)or24 h(candd)more.For(e),

cellswereinfected with ts125at40°Cin thepresence

of HU and incubated for 12 h. For (f), cells were

infected with ts125 at 33°C and incubated for 24 h. DNA from each sample was digested with KpnI, electrophoresedina1%agarosegel,and transferredto nitrocellulose. KpnI-digested d1313, Ad2, and Ad5 DNAswereincludedasmarkers. The DNA blotwas

hybridizedto32P-labeled Ad5 DNA.

0

x

U

I

(V)

15r

dl 313 10j

x

220

I 15

en,

dl 312

10 46,

r,

L

on November 10, 2019 by guest

http://jvi.asm.org/

65 70 5 79 86

___> 23kb 3kb

abcd e~~~~~~~O

E->oS;RI

6.84b

43Okb235kb

-

1.5kb-expression detected (Fig. 10, lanes c and d). The

0.66-kbmRNAhybridizingtheHindIII Cprobe

is transcribed from d1313 and is a fusion product of Ela and Elb sequences(37).

Hybridization with the SV40 probe detected low levels of the 2.3- and 1.5-kb early hybrid mRNAs from the superinfecting virus, even

when DNA replication was blockedimmediately aftersuperinfection (Fig. 10, lane a). A low level of the early hybrid mRNAs is to be expected after superinfection into late-stage cells, as E3 promoteractivity declineslatein infection(28). The 3.1-, 3.5-, 4.0-, and 6.8-kb late hybrid mRNAs were detected only when more than 3 h wasallowed for the superinfecting virus to enter andreplicate (Fig. 10, lanes c andd).

The results presented above demonstrated that IX mRNA was expressed only in cells in which the superinfecting viral DNA had

replicat-FIG. 9. Ad2+ND5 hybrid mRNAs. The hatched boxdepicts the SV40 insert in Ad2+ND5. Numbers above the lineareAd2 mapunits;those below the line areSV40mapunits.Proposedstructuresof thehybrid

mRNAs(except 1.5kb)aregiven, alongwiththe map locations of the Ad2 DNAfragmentsclonedforuse as

probes. The unbroken lines represent Ad2 sequences, the dashed lines indicate spliced-out sequences, and the open boxes represent SV40 sequences. The four

largermRNAbodies areproposedtobejoinedtothe Ad2 late tripartite leader (4, 10, 22). Line 293 cells wereinfected withAd2+ND5 foi 4 h (a,d, e)or15h (b, c). Their cytoplasmic, poly(A)-containing RNA was isolated, electrophoresed in anagarosegel, and transferred to nitrocellulose. The RNA blots were

hybridizedto32P-labeledSV40DNA(a,b),EcoRI Fl

DNA(c,d),orHindlll H DNA (e).

and four were mock-infected. At 13 h

postin-fection, the cells were superinfected with

AD2+ND5. One plate was changed to

HU-containing mediumat0h,anotherat3 h,anda

thirdat6hpost-superinfection;all were

harvest-ed at 9 h. Cytoplasmic,poly(A)-containingRNA wasisolatedfromeachplate, electrophoresed in anagarosegel,andtransferredtonitrocellulose.

Duplicate RNA blots were hybridized to

nick-translated HindIII C (7.9-17.0) DNA or SV40

DNA (Fig. 10 and 11). The0.49-kb IX mRNA

was notdetectedby theHindlIl Cprobe when

DNAreplicationwasblockedimmediately after

superinfection (Fig. 10, lane a) or when it was

blocked 3 h post-superinfection (lane b). Only

when more than 3 hwasallowed forthe

superin-fecting virus to enter and replicate was IX

I_.

..~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.

.7

ricurs

h o le, -2

*U

:...4

"A o b t-- d

.s ,kE

....

_ 4i-k b

SF _ 3 kb 3 i kb

-x13-rE - 5.Kb

-R:_,!. ' De.v-NE PR

FIG. 10. Expression ofIXmRNArequires

replica-tion. Line 293 cellswereinfected withd1313(athrough d)and13hlaterwere superinfected withAd2+ND5. The cellswerechangedtomediumcontaining10 mM HU at thetimes indicated (dashed lines indicate the presence ofHU) andwereharvested 9hafter

superin-fection. Cytoplasmic, poly(A)-containing RNA was

isolated, electrophoresed inanagarosegel,and

trans-ferredtonitrocellulose. TheRNAblotswere

hybrid-izedto32P-labeledHindIll C DNAorSV40 DNAas

indicated. Autoradiogram of the hybridized blot. M

denotes14C-labeledrRNAmarkers.

J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

A

nJ

0-d

fll

B D D

a1

b 1\

ARV

post-superinfection, the poly(A)-containing RNA was isolated from

107

and 100-MOI, no-drug superinfections (Fig. 12,

lanes a and b), but no IX mRNA from the 10-MOI and 100-10-MOI plus HU superinfections (lanes a and c).Despitea10-fold greater input of superinfecting virus, expression of IX mRNA could not be detected in the absence of DNA replication.

DISCUSSION

This paperdescribesexperiments designedto test the effect oftemplate replication upon the

expression of Ad2 intermediategenes. All four

intermediate mRNAsassayed appearedat6 to 8

hpostinfection, a timecourse ofexpression in

agreementwithprevious work(30, 38).

We have compared the steady-state levels of

the IX andIVa2mRNAs in cellsreplicatingviral

DNA and cells blocked in DNAsynthesis.IX or

IVa2 mRNA was detectable by RNA blot

hy-bridizationonlywhen viral DNAreplicationwas

allowed. Densitometer scans of the

autoradio-graphs from these experiments were used to

quantitate the reductionin IX and IVa2 mRNA

levels. Inthree experiments in which both sets

of cells were infected at the same MOI, HU

treatment reduced the steady-state level of IX

IM a b c d e

c

d

FIG. 11. Densitometer scansof lanes athrough d ofFig.10,hybridizedtoHindIlI C DNA (A) andSV40

DNA(B).

ed. Duetoreplication,those cells also contained

a larger number ofpotential templates for IX

transcription. To demonstrate that the

replica-tion-dependent regulation of IX transcription is

independent oftemplatenumber, the amount of

superinfecting virus was varied 10-fold.

d1313-infected 293 cells were superinfected with

Ad2'ND5at10 or 100MOI. Half of each set of

cells wasincubatedin the presence of HU, and half was incubated without drugs. Nine hours

I

-0.66kb

-0.49kb(2)

FIG. 12. Expression ofIX mRNA is independent of MOI.Line 293 cellswere infected withd1313 (MOI,

20; lanes athrough d)ormock-infected (e) and 13 h

later were superinfected with Ad2+ND5 as follows: lane a, MOIof10plus HU; laneb, MOI of 10 without HU;lanec, MOIof100plus HU; lanes d and e, MOI of100without HU. Nine hours aftersuperinfection, cytoplasmic, poly(A)-containing RNA was isolated, electrophoresed inanagarosegel, and transferred to

nitrocellulose. Each lane represents the material from

107cells. TheRNAblotwashybridizedto32P-labeled HindIIICDNA.

on November 10, 2019 by guest

http://jvi.asm.org/

mRNA an average of 800-fold and that of IVa2 mRNA an average of 350-fold. Underinfection

conditions that resulted in equal numbers of

viral templates in both sets of cells, HU

treat-ment reduced the steady-state level of IX

mRNA500-fold and that ofIVa2 mRNA50-fold.

Thesefiguresrepresent minimum values, as the

films wereprobably exposed beyond thelinear

response range in the no-drug lanes inan effort

to revealfaint bands intheHU-treated lanes.

We havealsocomparedthe rateof

transcrip-tion ofIX andIVa2 mRNAs in cells replicating

viralDNAand cellsblocked inDNA synthesis.

When equal numbers of viral templates were

presentin both sets of cells, newly synthesized IX and IVa2 mRNA could only be detected

when viral DNA replication was allowed.

Be-cause of the high backgrounds inherent in this

pulse-labeling experiment,we can saywith

cer-tainty only that the rate of transcription of IX

and IVa2 mRNAs is reduced 10-fold in the

absenceof viralDNAreplication. The

accumu-lated evidence suggests, however, thatno IX or

IVa2 mRNAis produced when viral DNA

repli-cation is blocked.

Other investigators have observed some IX

protein (30)orIVa2 mRNA (24) in cellsinfected

inthe presence of the DNA synthesis inhibitor

cytosine arabinoside, although the quantity of

these products present was not directly

com-pared tothat present in control cells. Cytosine

arabinoside has been reported to break down

during prolonged incubation (39). The IX and

IVa2 geneactivity observed by these

investiga-torscould haveresulted fromasmallamountof

viral DNAreplication due to cytosine

arabino-side breakdownduringthecourseof the

experi-ment. Alternatively, the observed activity could

have resulted frominefficient expressionofthe

IX and IVa2 genes in cells in which no viral

DNA replication had taken place. Our results

have demonstrated, however, that this

expres-sion isinsignificantly low.

Bacteriophage T4lategenes, like the Ad2 IX

gene, areexpressed only after viral DNA repli-cation hasbegun. Investigations of T4late-gene

regulation suggest that bothphage-encoded

sol-uble geneproducts and aphysical changein the

template fortranscriptionarerequiredforphage

late-gene expression (31). The superinfection

experiments presented inthis paper demonstrate

thattheAd2 IX promoter cannot beactivatedon

aninfecting, parental template. Diffusible,

trans-acting factors produced late in infection with

d1313are notsufficienttoactivatethe IX gene on

aparentaltemplate.The possible objectionthat

d1313 does not produce the needed regulatory

factorsbecause of mutation is dismissed by the

observation thatd1312, whichwasderived from

the same parent virus and thus shoulddifferonly

in region El (18), produces IX mRNA and

protein wheninfecting 293 cells. BesidesIX,the

onlyEl mRNAmissing from d1313-infected293 cellsisthe0.52-kb mRNA from Ela (37), which is also absent fromd1312-infected 293cells.

Activation of theIX gene promoteris

proba-bly dependent upon a change in the physical

structureof the template during replication. An

alteration in the distribution or composition of

nucleosomes (33), exposing the IXpromoter to

the action of RNA polymerases, could be

re-sponsible for this change. Another possibility is

that soluble regulatoryfactors necessaryfor IX

promoter activity cannot bind to the parental

DNAbecause of the nucleosomestructure.

Dur-ingorafter replication, accessibilitytothe

bind-ing site of these regulatory factorson the

tem-plate is afforded by the perturbation of the

deoxyribonucleoproteinduringreplication. This

model of IX regulation resembles the stable,

active and inactivetranscription complexes

ob-servedby Bogenhagen etal. in theXenopus5S

RNA gene (7). A third possibility is that the

replicatedgenomes enter aspecific nuclear

com-partment,andonly in thatcompartment arethe

factorsnecessaryforIXtranscription available.

Further work will be needed to determine

which, ifany, of these models is correct.

Although IX and lateregions L2 through L5

are activated by the same event, viral DNA

replication,IXprotein is observed earlier during

infection than proteins encoded by L2 through

L5(14, 30).Thistime differencemayresult from

morerapidtranscription, processing, transport,

and translation ofthe small, unspliced IX

mes-sage(3).Also, the different mechanisms of

regu-lation-promoter activation for the IX gene,

versusantitermination and poly(A) site selection

for L2 through L5-may be responsible for

some of the time difference. The observation

that fiber protein from L5 isnot seen until2 h

after the appearance of hexon protein protein

fromL3(30) indicates thatlate-geneactivation is

complex andmayinvolvefactorsor eventsother

than DNAreplication.

Several characteristics of the IX and IVa2 genesdifferentiatethemfromthe 0.52-and 1.03-kb intermediate genesfrom El:(i) IXandIVa2

aretheonlyAd2transcriptionunits toproducea

single transcript and polypeptide, whereas the

other intermediates are membersof families of

mRNAspecies originating at the same

promot-er;(ii)IX andIVa2,andpossiblythe E2

promot-eratmapposition72,arethelast Ad2promoters

to be activated during lytic infection, whereas

the other intermediates derive frompromoters

active at early times; (iii) IX and IVa2 are

regulated by promoter activation, whereas the

other intermediates are regulated by mRNA

stability.

J. VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/

ACKNOWLEDGMENTS

We thank thecolleagues who kindly provided recombinant DNAs for thisstudy. We alsothank EllenJohnson and Scott Hauser for technicalassistance,andBetsy Klein for typing the manuscript.

This work was supported by research grants from the American CancerSociety (MV-31H) and the National Cancer Institute (CA-16007). Cell culture mediawereprepared in a Cancer Centerfacility funded bythe NationalCancer Institute (CA-16217). This studywasalsosupported byBrown

William-son Tobacco Corp., Philip Morris, Inc., R. J. Reynolds TobaccoCo.,and the United StatesTobacco Co.

LITERATURE CITED

1. Aieio,L.R.,R.Guifoyle,K.Huebnpr,andR. Weinmann. 1979. Adenovirus 5 DNA sequencespresent and RNA

sequencestranscribed in transformed human embryonic kidney cells(HEK-Ad5or293). Virology 94:460-469. 2.Akusjarvi, G., and H. Persson. 1981. Controls of RNA

splicing and termination in the major late adenovirus transcription unit. Nature (London) 292:420-426. 3. Alestrom, P., G.Akusjarvi,M.Perricaudet,M. B.

Math-ews,D.F.Kiessig,and U. Pettersson. 1980. Thegenefor polypeptide IX ofadenovirus type 2 and its unspliced

messengerRNA.Cell 19:671-681.

4. Berget,S.M.,C.Moore,andP. A.Sharp. 1977. Spliced

segments atthe 5'terminusof adenovirus 2late mRNA. Proc. Natl.Acad. Sci. U.S.A.74:3171-3175.

5. Berk, A. J., F. Lee,T. Harrison, J.Williams, and P. Sharp. 1979. Pre-early adenovirus 5gene product

regu-lates synthesis of early viral messenger RNAs. Cell 17:935-944.

6. Berk, A. J., and P. A. Sharp. 1978. Structure of the adenovirus 2early mRNAs. Cell 14:695-711.

7.Bogenhagen,D.F.,W. M.Wormington, andD.D.Brown. 1982.Stabletranscription complexes of Xenopus 5S RNA

genes: a means to maintain the differential state. Cell 28:413-421.

8. Carter,T.H.,and H. S.Ginsberg.1976. Viral transcrip-tion in KB cells infectedby temperature-sensitive "early"

mutantsof adenovirustype5.J.Virol. 18:156-166.

9. Chow, L. T., T. R. Broker, and J. B. Lewis. 1979. Complex splicingpatterns ofRNAs from theearly regions of adenovirus 2. J. Mol.Biol. 134:265-303.

10. Chow, L. T.,R. E. Gelinas, T. R. Broker, and R. J. Roberts. 1977.Anamazingsequencearrangementatthe 5' ends of adenovirus 2messengerRNA.Cell 12:1-8. 11. DarneUl,J.E. 1982.Varietyin the levelofgenecontrol in

eukaryoticcells. Nature(London)297:365-371. 12. Ensinger,M.J.,and H. S.Ginsberg.1972.Selectionand

preliminary characterizationoftemperature-sensitive

mu-tantsoftype5adenovirus.J. Virol.10:328-339. 13. Evans, R., J. Weber, E. Ziff, andJ. E. Darnell. 1979.

Premature termination during adenovirus transcription. Nature(London)278:367-370.

14. Everitt, E.,and L.Philipson.1974.Structuralproteinsof adenovirus. XI. Purification of three low molecular weight virion proteins of adenovirus type 2 and their synthesis during productive infection. Virology 62:253-269.

15. Flint,S.J.1977. Thetopographyandtranscriptionof the adenovirusgenome.Cell10:153-166.

16. Goldenberg, C. J., and H.. J. Raskas. 1979. Splicing patternsof nuclearprecursorstothemRNAfor adenovi-rus2DNAbinding protein.Cell16:131-138.

17. Graham,F. L.,J. Smiley, W.C. Russell,andR. Nairn. 1977. Characteristics ofahumancell linetransformedby DNA from human adenovirus type 5. J. Gen. Virol. 36:59-72.

18.Jones, N., and T. Shenk. 1979. Isolationofadenovirus type5 hostrangedeletionmutantsdefective for transfor-mation ofratembryocells. Cell17:683-689.

19. Jones, N.,and T.Shenk. 1979. Anadenovirustype5early

gene function regulates expression of otherearly viral genes.Proc.Natl. Acad. Sci.U.S.A.76:3665-3669. 20. Kelly, T. J., andA.M.Lewis.1973. UseQf nondefective

adenovirus-simian virus 40hybridsformappingthesimian virus 40 genome.J.Virol. 12:643-652.

21. Kitchingman, G. R., S.-P. Lai,and H. Westphal. 1977. LoopstructuresinhybridsofearlyRNAand the separat-ed strands of adenovirus DNA. Proc. NatI. Acad. Sci. U.S.A. 74:4392-4395.

22. Klessig, D. F. 1977. Two adenovirus mRNAs have a common5' terminal leader sequence encodedatleast 10 kb upstreamfrom their maincoding regions.Cell 12:9-21. 23. Lebowitz, P.,T.J.Kelly,D.Nathans,T. N.H.Lee,and A. M. Lewis. 1974. Acolinear maprelating the simian virus 40(SV40) DNA segmentsof sixadenovirus-SV40 hybrids to theDNA fragments produced byrestriction endonucleasecleavage of SV40. Proc. NatI. Acad. Sci. U.S.A. 71:441-445.

24. Lewis,J. B.,H.Esche, J.E.Smart,B. W.Stillman,M. L. Harter, and M. B. Mathews. 1980. Organization and expressionof the left third of the genome of adenovirus. ColdSpringHarborSymp. Quant. Biol. 44:493-508. 25. McGrogan, M.,andH.J.Raskas.1977.Species

identifica-tion and genomemapping ofcytoplasmicadenovirus type 2RNAssynthesizedlateininfection. J. Virol. 23:240-249. 26. McMaster, G. K.,andG.C. Carmichael. 1977.Analysisof

single- and double-stranded nucleic acidson polyacryl-amide and agarose gels by using glyoxal and acridine orange.Proc. Natl. Acad.Sci. U.S.A.74:4835-4838. 27. Nevins,J.R.1980.Definition andmappingof adenovirus 2

nucleartranscription. MethodsEnzymol.65:768-785. 28. Nevins,J. R., H. S.Ginsberg, J.-M. Blanchard, M.C.

Wilson,andJ.E.Darnell.1979.Regulationof theprimary expressionof theearlyadenovirustranscriptionunits. J. Virol. 32:727-733.

29. Nevins, J. R., and M. C. Wilson. 1981. Regulation of adenovirus 2 geneexpressionatthelevel of transcription-al termination and RNA processing. Nature (London) 290:113-118.

30. Persson, H., U. Pettersson, and M. B. Mathews. 1978. Synthesis ofa structural adenovirus polypeptidein the absence of viral DNAreplication.Virology90:67-79. 31. Rabussay,D.P.,and E. P.Geiduschek. 1977.Regulation

of gene action in thedevelopmentoflytic bacteriophages, p. 1-196. In H. Fraenke-Conrat and R. Wagner (ed.), Comprehensive virology, vol. 8. Plenum Publishing Corp.,New York.

32. Sauerbrier,and A.R.Brautigam. 1970.Asimplemethod forisolatingRNAfrom bacteria. Biochim.Biophys.Acta

199:36-40.

33. Sergeant, A., M. A. Tigges, and H. J. Raskas. 1979. Nucleosome-like structural subunits of intranuclear pa-rentaladenovirus type 2 DNA. J. Virol. 29:888-898. 34. Sharp, P.A., J. Manley, A. Fire,and M. Gefter. 1980.

Regulation of adenovirusmRNA synthesis. Ann.N. Y. Acad.Sci. 354:1-15.

35. Shaw,A. R., and E.B. Ziff. 1980. Transcriptsfrom the adenovirus-2 major late promoter yield a single early familyof 3' coterminal mRNAs and five late families.Cell 22:905-916.

36. Southern, E. M. 1975. Detection of specific sequences amongDNAfragments separatedby gel electrophoresis. J.Mol.Biol. 98:503-517.

37. Spector, D. J., D. N. Halbert, and H. J. Raskas. 1980. Regulation of integrated adenovirus sequences during adenovirus infection of transformed cells. J. Virol. 36:860-871.

38. Spector, D. J., M. McGrogan, and H. J. Raskas. 1978. Regulation of the appearance ofcytoplasmic RNAsfrom region 1 of the adenovirus 2 genome. J. Mol. Biol. 126:395-414.

39. Thomas,G. P.,and M.B. Mathews. 1980. DNA replica-tion and theearlyto late transition inadenovirus infec-tion.Cell22:523-533.

40. Thomas,P.S. 1980. Hybridizationof denaturedRNA and

on November 10, 2019 by guest

http://jvi.asm.org/

748 AND RASKAS

small DNAfragments transferredtonitrocellulose. Proc. Natl. Acad.Sci. U.S.A. 77:5201-5205.

41. van derViet, P. C., A. J. Levine, M. J. Ensinger, and H.S. Ginsberg. 1975. Thermolabile DNA binding proteins from cells infected withatemperature-sensitivemutantof adenovirus defective in viral DNA synthesis. J. Virol. 15:348-354.

42. Wilson, M. C., and J. E. Darnel. 1981. Control of

messengerRNAconcentration by differential cytoplasmic half-life. AdenovirusmessengerRNAs fromtranscription

units IA and 1B. J. Mol. Biol. 148:231-251.

43. Wilson,M. C., N. W. Fraser, and J. E.DarneDl. 1979. Mapping of RNA initiation sites by high doses ofuv irradiation: evidence for three independent promoters

within theleft 11% of the Ad2genome.Virology 94:175-184.

44. Ziff, E. B., and R. M. Evans. 1978. Coincidence of the

promoter and capped 5' terminus of RNA from the adenovirus 2 major latetranscription unit. Cell 15:1463-1475.

J.VIROL.

on November 10, 2019 by guest

http://jvi.asm.org/