0022-538X/78/0028-0227$02.00/0
Copyright X 1978 American SocietyforMicrobiology PritedinU.S.A.
Transcription Map for Adenovirus Type 12 DNA
J. R.
SMILEYt
AND S.MAK*Departmentof Biology, McMaster University, Hamilton,Ontario,CanadaL8S 4K1 Received forpublication 23 March 1978
The regionsof the adenovirus type 12 genomewhich encode 1- and r-strand-specific cytoplasmicRNAwere mapped by thefollowingprocedure. Radioactive, intact, separatedcomplementary strands oftheviralgenome werehybridized to saturatingamounts of unlabeled late cytoplasmic RNA. The segmentsof each DNA strand complementary to the RNA were then purified by
Si
nuclease digestion of the hybrids. Thearrangement of the coding regions of each strand wasdeducedfromthe pattern ofhybridization of these probes to unlabeled viral DNA fragments producedby digestion with EcoRI, BamHI, and HindIII. The resultingmapissimilar,
ifnot identical, to that of adenovirus type 2.The subset of the late cytoplasmic RNAsequenceswhich are expressed at early times were located on the map by hybridizing labeled, early cytoplasmic RNA to both unlabeled DNA fragments and unlabeled complementary strands of specificfragments.
Early cytoplasmic RNA hybridized to the r-strand toEcoRI-C andBamHI-Bandtothe
1-strand
of BamHI-E. HybridizationtoBamHI-Cwasalso observed. The relativeratesof accumulation of cytoplasmic RNA complementary to individual restriction fragments was measuredat bothearly and late times. Early during infection, most of the viral RNA appearing inthe cytoplasm was derived from the molecular ends of thegenome.Later (24to26hpostinfection)the
majority
of thenewly labeled cytoplasmic RNAwastranscribed from DNAsequencesmappingbetween25and60mapunitsonthegenome.
Theavailability of bacterial restriction endo-nucleases which cleavea
given
DNA molecule to generate a defined set offragments has al-lowedconsiderable progress to be made in theanalysis
of thetranscription
program of thegroupC human adenoviruses.
The DNA sequences encoding the mRNA's
synthesized during
theearly
andlatephases
ofproductive infection,
aswellasthoseexpressed
intransformed celllines, have been mappedon theadenovirus type 2 (Ad2) (2,4,6, 28, 31, 39) andAd5 (5,
7)
genomes. Thepositions
and po-larities of the sitescomplementarytoearly
and late mRNAaresimilar ifnotidenticalinthetwogenomes,a
finding
notexpected
in view oftheir extensive nucleotide sequencehomology
(10)
and similar
biological
characteristics.We areinterested in theanalysis of the
tran-scription
of the genome ofAdl2,
amember ofthe highly oncogenic group A human adenovi-ruses. These viruses share very fewnucleotide sequences with the group C adenoviruses (17) anddifferwidelyinanumberof
biological
prop-ertiesfromboth Ad2 andAd5. Wehaveprevi-ously shown that early in
productive
infectiont Present address: DepartmentofTherapeutic Radiology, RadiologyLaboratories, YaleUniversitySchool ofMedicine, NewHaven, CT06510.
withAdl2, RNA complementaryto about 30% of theviralgenome,derived from both comple-mentary DNAstrands, is detectable in the cy-toplasm of infectedcells(33).The fractionof the genome
expressed
ascytoplasmic RNA increases toabout 80%atlate times. The majorityof the RNAsequenceswhichappearexclusivelyatlate times are derived from the r-strand of the ge-nome. Further experiments showedthat
the early RNAsequences are asubset of those pres-entlate in infection. Wereporthere experiments which map the physical locations and strand derivation of the DNA sequences expressed asearly
and late cytoplasmic RNA, using a newmethod. In addition, the relativeratesof accu-mulation of cytoplasmic RNA derived from dif-ferentregions of thegenome weremonitoredat different times after infection.
MATERIALS
AND METHODSVirusandcells.PurifiedAdl2 strain1131(24)and human KB cells grown in suspension were used throughout this study. Proceduresfor infection and viruspurificationhavebeen described(12,21).Viral DNAwasextracted frompurifiedvirionsasdescribed by Green and Pina(13).
Separationof the intactcomplementaryAdl2
DNAstrands. Thecomplementarystrands ofAdl2 DNAwereseparated bycentrifugingdenatured viral 227
on November 10, 2019 by guest
http://jvi.asm.org/
SMILEY AND MAK
DNAtoequilibriuminneutral cesiumchloridedensity gradients in the presence ofpolyuridylic acid-poly-guanidylic acid [poly(U,G)], as previously described
(18, 33).
Material pooled from the gradients was further
purified byexhaustive self-annealingat37°C in 0.14 Msodium phosphate buffer, pH 6.8, containing50% formamide,followedbysedimentation in5to20%(wt/ wt) high-saltneutralsucrosegradients (35)at38,000 rpmfor 90mininaBeckmanSW40rotor.Intactsingle strands (sedimenting at 70S) were pooled,
exhaus-tively dialyzed against0.01 M Tris(pH7.3) containing 2%formamide, andused forhybridization.
RNAextraction. RNAwasextractedfrominfected
cells aspreviouslydescribed(33). EarlyRNA(labeled
5.5 to 7.5 h postinfection) wasprepared by
concen-tratinginfected cellsto3 x 106/ml by centrifugation andadding[3H]uridine (New England Nuclear,20Ci/
mmol)toafinalconcentration of 10 MCi/ml.
Restriction endonucleases and electrophore-sis. Endonuclease EcoRIwaspurifiedbythemethod
of Pettersson and Philipson (26). Endonucleases BamHI and HindIIIwerepurchased from Bethesda
Research Laboratories. Digestionbufferswere0.1 M
Tris (pH7.9)-0.01 M MgCl2forEcoRI, 0.02 M Tris (pH 7.4)-0.01MMgCl2-0.06MNaClforHindIII and 0.02 M Tris (pH 7.4)-0.01 MgCl2-0.002 M,B-mercap-toethanol for BamHI.Fragments wereseparated by
electrophoresis through0.7%(EcoRI)or1%(HindIII and BamHI) agarose gels for 11 h at 2 V/cm. The DNA bands were visualized under short-wave UV
illumination after staining the gels with 0.5 Mg of
ethidium bromideperml.
Separation of the complementary strands of DNA fragments by electrophoresis. DNA frag-mentsinagarosegelslicesweredenatured insitu and
subjectedtore-electrophoresisin1.0%agarosegelsas
describedbyHorowitz(16).
Restriction endonuclease cleavage map of Adl2 (1131) DNA. The cleavage patterns of Adl2 (1131) DNA by endonucleases EcoRI, BamHI, and HindIIIwereexamined. TheEcoRI and BamHI
frag-mentsof strain 1131DNAwerefoundtobeidentical to those produced from Adl2 (Huie) DNA (Smiley andMak, unpublished data;C.Mulder,personal
com-munication). Since identical sets offragmentswere
produced by digestionof the DNAofbothstrainsof Adl2,we have assumed that the EcoRI and BamHI
cleavage maps of these strains are identical. The HindIII cleavagemapofstrain 1131 DNAwas
deter-mined (Smileyand Mak, manuscriptinpreparation) and is shown with those forEcoRIand BamHI in Fig. 2.
Purification of the segments of each DNA strandcomplementarytolatecytoplasmic RNA. 3H-labeled (leftward transcription)-andr(rightward
transcription)-strands (5 x 105 cpm/,ug, 0.5 Mg/ml)
wereseparately hybridizedtounlabeled late cytoplas-mic RNA (2 mg/ml) for 24 h at 67°C in 0.3 M NaCl-0.01 M Tris(pH7.3),conditionspreviously de-termined to besaturating. The material wastreated
withS1 nucleaseasdescribed previously (33)and then
made 0.2 M sodium phosphate (pH 6.8)-1% sodium dodecylsulfate(SDS),andyeastRNA (1mg/ml)was
added as carrier. The mixture was extracted with phenol saturated with 0.2 M sodium phosphate (pH 6.8), dialyzed against 0.01 M Tris (pH 8.1), and ethanol precipitated.
The pelletsweretaken up in2ml of0.3NNaOH, boiled for 15min,neutralized, bufferedatpH 7.3 with 0.1 MTris, and dialyzed against 0.1x SSC (lx SSC = 0.15 MNaCl plus 0.015 M sodiumacetate) after the addition of1mg of yeast RNAper mlascarrier.
Hybridization of labeled nucleic acidstoDNA fragments generated by restriction endonuclea-ses. Unlabeled DNAfragments separated by agarose gelelectrophoresisweredenatured in situ and trans-ferredtonitrocellulosestrips by the method of South-em(34). Each stripwasprepared fromagel containing 2 ,ug of Adl2 DNA. The labeled nucleic acid was hybridizedtothestrip in2xSSCcontaining0.1% SDS and 500 yg of yeast RNA per mlat65°C for 20 h. The unhybridized nucleic acidwasremovedbyoneoftwo procedures. In [3H]RNA-DNA hybridization, the stripswerewashed and treated withpancreaticRNase as described by Southern (34). In [3H]DNA-DNA hybridization, the unhybridized [3H]DNA was re-movedbywashing for 1h in 500ml of2xSSC plus 0.1% SDSat 65°C followed by1 h in500ml of4 x 103 MTrisplus0.1xSSC (pH 9.4) at room temper-ature (21). Thehybridized radioactivitywasdetected and quantitated by fluorography. Strips were oven
dried,immersed in 20%(wt/vol) Omnifluor (New Eng-land Nuclear) in toluene, air dried, and exposed at -70°C to Kodak RP Royal X-omat (RP R14) film previously exposedto ahypersensitizing flash of white light (19). Preexposure wasdone tobring the back-ground density of the film to 0.15 D above that of unexposed film. Theresulting negatives were scanned withaJoyce-Loeblmicrodensitometer, and the rela-tiveamountofradioactivity hybridized to each band was estimated from the peak areas. In some cases, therewassignificantbackground radioactivityonthe
fluorographs.Estimationof thespecifically hybridized radioactivitywasmoredifficult,and thepeaks were extrapolatedtothe baseline of the microdensitometric trace. It is clear that this procedure results in an overestimation of thespecifically hybridized radioac-tivity where there isasignificant background.
Hybridization of labeled RNA to immobilized 1-andr-strands. To prepare complementary strand-specific RNA, samples containing5 x 106cpm of 3H-labeledlatecytoplasmic RNA (about 150Mgof RNA) werehybridizedto1-orr-strandAdl2 DNA immobi-lized onnitrocellulose filters.Hybridizationwasfor 20 h in 2x SSC containing 0.1% SDS at 65°C. After hybridization, filters were washed with 1 liter of 2x SSC containing 0.1% SDS at 65°C for 20 min with stirring, followed byafurtherwash with 2xSSCunder suction at room temperature. The hybridized RNA waseluted by boiling for5minin0.lxSSC, and yeast RNA(500 ,ug/ml)wasadded ascarrier.After dialysis against 0.01 M Tris (pH 7.3), any contaminating
['4C]DNA was degraded with 100 Mg ofpancreatic DNase per ml in the same buffer containing 0.01 M
MgCl2. The RNA was then extracted by the hot phenol-SDS method, ethanol precipitated, and di-alyzed against0.1xSSC.
on November 10, 2019 by guest
http://jvi.asm.org/
ADENOVIRUS TYPE 12 RNA 229
RESULTS
Mapping the DNA sequences expressed
aslatecytoplasmic RNA. Most of the hybrid-ization mapping studies of adenovirus published
todate have utilized asprobes labeled
comple-mentarystrand-specificsequencesof individual
DNA fragments generated by restriction endo-nucleases. The complementary strands of indi-vidual fragments have been prepared by hybrid-izationtoexcess,intact complementary strands
(39) or by electrophoresis of denatured DNA fragments (7, 31). The complement excess
hy-bridization method (39) required rather large
amountsofintactcomplementary strands,
quan-tities thataredifficulttoobtain with Adl2 DNA because of the relatively low yield of virus and the small density difference between the
com-plements in poly(U,G)-cesium chloride density gradients (18, 33). Some of the Adl2 DNA
frag-mentscouldnotbe separated directlybyeither electrophoresis orbybinding topoly(U,G) fol-lowed bycesium chloride density centrifugation (23;Smiley, unpublished data). We have, there-fore,adoptedamappingstrategywhich obviates
therequirement for strand separating the indi-vidual DNA fragments. Radioactively labeled intact1-andr-strandswerehybridized to
satu-ratingamountsofunlabeled cytoplasmic KB cell RNA extracted at 24 h after infection (late RNA). Late cytoplasmicRNAiscomplementary to65%ofthe r-strand and 15% of the 1-strand of Adl2 DNA (33). In this preparative hybrid-izationtheadded RNA protected 63.8% of the
r-strand and 16.2% of the 1-strand from S1
nu-clease, demonstrating that the hybridization
re-actionwassaturated.Thehybridized portions of
eachstrandwerepurified byS1nuclease
diges-tion ofthe unhybridized nucleic acid, and the RNA was removed by alkaline hydrolysis as
described inMaterialsand Methods. The DNA sequences recovered by this procedure
corre-spondtothesegmentsof1-andr-strands which
are complementary to late cytoplasmic viral RNA andshouldbe inequimolaramounts. We
cannot, however, exclude the possibility that sequencesfromasmallfraction of the genome
whichencodesextremelyrareRNAspeciesare
underrepresented in the DNA selected in this
manner.
These [3H]DNA preparations were used as
probes to determine the regions of the Adl2
genome complementary to these sequences.
Eachprobewas hybridizedtoexcess (2
,ug)
un-labeled DNA fragments generated by various
restriction endonucleases. The pattern of hy-bridization of these selected sequences to the
fragmentsetsgeneratedbydifferent restriction
endonucleasesprovidesinformationonthe
dis-tribution of thecoding regions of the 1- and r-strands throughout the genome. In addition, within the limits discussed below, the relative amountofradioactivity hybridizedtoeach frag-mentgivesanestimate of thedegree of comple-mentarity of each fragmentto the probe used. This inturnprovidesa measureof the fraction of the DNA fragment transcribed into stable cytoplasmic RNA in the infected cell (see be-low).
To determine how accurately the relative amountofradioactivity hybridizedtoeach frag-mentreflects thedegree of
complementarity
of the probe to that fragment, total 3H-labeled Adl2 DNA, degraded toabout 350nucleotides inchain length byboiling in alkali,was hybrid-izedtoEcoRIfragments AtoE, using the South-erntechnique (34). Theamountofradioactivity bound toeachfragmentwasquantitatedas de-scribed in Materials and Methods. The results are shown in Table 1. It canbe seen that the relative amount of 3H-labeled Adl2 DNA hy-bridizedtothe unlabeledfragments
isapproxi-mately equaltothe
expected
values of comple-mentarity between the labeled probe and the unlabeled DNAfragments.
Severalpoints
con-cerning these data and theirinterpretation
are noteworthy. First,although
thereisclose agree-mentbetweenexperiments, thedegree
of com-plementarity of theprobe
tofragments
AandCis
overestimated,
whereas that to D and E isunderestimated. It is
possible
that this result reflects therelativeguanine-plus-cytosine
con-tentsof the
fragments,
since the DNAprobe
was labeled with[3H]thymidine,
or some other source ofsystematic
error. In some cases de-scribedbelow,
thedegree
of complementarity between ther-specific probe
and individualDNA
fragments
is alsoslightly overestimated,
perhaps for similar reasons.
Second,
since the resultsaredisplayed
asthe fraction ofthetotal hybridized radioactivitywhich hasannealedtoTABLE 1. Hybridizationof3H-labeledAdl2 DNAto Adl2 DNAfragments generated byEcoRIa
Relativeamtof3H-Adl2 hy-Adl2 DNA
Fragmentisize
bridizedfraction of the
fragment gnm
genome ExptI ExptH
A 0.36 0.39 0.39
B 0.27 0.26 0.30
C 0.165 0.21 0.20
D 0.115 0.096 0.095
E 0.07 0.042 0.047
aFragmented
3H-labeled Adl2
DNAwashybridizedtonitrocellulosestripsbearingelectrophoretically sep-aratedDNAfragments,andthehybridized radioactiv-itywasquantitatedby fluorography.
on November 10, 2019 by guest
http://jvi.asm.org/
AND
each fragment, the determinations within a given
experiment
arenotindependent.
System-atic or spurious errors in the measurement of thefraction
hybridized
to onefragment
neces-sarily introduceerrorintothe other determina-tions. One of the factorscontributing
tospurious erroris thebackgroundradioactivity
on someof thefluorographs (e.g.,Fig.
1).To maptheviralDNAsequences complemen-tary tolateRNA,3H-labeled 1- andr-strandviral DNA sequencesselectedby hybridizationtolate cytoplasmic RNAwere
hybridized
toDNAfrag-ments generated by EcoRI, BamHI, and
HindIII, usingtheSouthern technique. The re-sultsareshown inFig. 1and Table2.Since the r-strand probe is derivedfrom0.65 of the total r-strand (33), the fraction of the
probe
hybrid-izedtoeach unlabeledfragmentmustbe multi-pliedby0.65 togive the fraction of the total r-strand transcribed within this
particular
DNA fragment. Only0.15of thetotal1-strand
is com-plementary to latecytoplasmic
RNA (33). Therefore, the fraction of the total1-strand
tran-scribed from each DNAfragment
can be simi-larly estimated.Data from Table 2 wereused toconstruct a map of the sites
expressed
aslatecytoplasmic
RNA (Fig. 2). The map shown wasassembled from the data asdescribed below byassuming
that
cytoplasmic
RNA isasymmetrically
tran-scribed and
by reducing
the number ofseparate,contiguously
transcribed regions to themini-Eco RI
E D C BA
m r
O
0
mum demanded bythe data. Other, more com-plex arrangements for the coding regions are possible.
Consider first the transcription from the 1-strand. A minimum oftwo separate regions of transcribed DNA were detected by hybridiza-tiontothe EcoRI
fragments;
onewithin EcoRI-A, the other within EcoRI-C. the leftmost re-gion, withinEcoRI-C,
mayextend a small dis-tancebeyond the EcoRI-C-D junction, sincein oneexperiment the probehybridizedto asmall degree with EcoRI-D. This suggestion is sup-ported by the weakhybridizationtoHindIII-C. The data obtained with thefragmentsgenerated by BamHI demonstrate that EcoRI-A contains at least two separate regions of transcribed DNA,onewithin BamHI-C and the other within BamHI-E.(Although
BamHI-C and -Dare notseparable by gel
electrophoresis,
the failure ofthe
1-strand-specific
probe to hybridize witheither EcoRI-B orHindIII-A, which both con-tain BamHI-D, allows the assignment to BamHI-C.) The fragments produced by HindIII allowa more precise localization of these three regions. The probehybridizedtoonly
HindIII-C,
-D, and -E(however, fragments
J, K, andL were run outof the gels used in the analysis).Qualitative
consideration of the data shows thatthe leftmost region is located to the right of HindIII-F
(6.3
mapunits)
andtothe left ofany of BamHI-D, -H, or -I (25 map units). The region detected within BamHI-C was notde-MIGRATION
FIG. 1. Hybridization of labeled1-andr-strand-specific probes complementary to late cytoplasmic RNA to unlabeled DNAfragments. Nitrocellulose strips bearing 2 pg of unlabeled DNA fragmentsproduced by EcoRI, BamHI, and HindIII were incubated with 0.02 pg (104 cpm) of r-strand probe or 0.005
pg
(2.5 x 103cpm)of I-strand probe in2xSSC plus 0.1% SDS at 65°C for 20 h. The hybridized radioactivity was detected byfluorographyto
flash-hypersensitized
X-ray film.Microdensitometer scans of the resulting negatives are shown.on November 10, 2019 by guest
http://jvi.asm.org/
[image:4.500.116.405.408.598.2]ADENOVIRUS TYPE RNA 231
TABLE 2. Distribution of the 1- and r-strand sequences complementary to late cytoplasmic RNA among specific DNAfragmentsa
ProbeFragmentFragment size(frac-
Fraction
ofgenomerepre-Probe FragmentFent ntSztion) fa-Fractionofprobe hybridized sented inprobementbwithin
frag-r EcoRI-A r EcoRI-B r EcoRI-C r EcoRI-D r EcoRI-E r EcoRI-F r BamHI-A r BamHI-B r BamHI-C r BamHI-D r BamHI-E r BamHI-F r BamHI-G
r BamHI-H + I
r HindIII-A r HindIII-B r HindHI-C r HindIII-D r HindIII-E r HindIII-F r HindIII-G r HinduII-H r HindIII-I r HindIII-J-L 1 EcoRI-A 1 EcoRI-C 1 EcoRI-D BamHI-A BamHI-C BamHI-E 0.36 0.27 0.165 0.115 0.07 0.02 0.25 0.16
0.131
0.13} 0.115 0.075 0.063 0.065 0.23 0.155 0.13 0.115 0.09 0.063 0.060 0.049 0.029 0.071 0.36 0.165 0.115 0.25 0.13 0.115 0.35 0.36 0.17 0.06 0.06 ND 0.23 0.24 0.11 0.04 0.04 ND 0.13 0.16 0.32C 0.04 0.12 0.10 0.12 0.08 0.11 0.08c 0.13c 0.03 0.08 0.07 0.08 0.29 0.19 0.06 0.02 0.01 0.13 0.12 0.10 0.07 ND 0.56(0.49)d 0.36 (0.51) 0.08(<0.01) 0.39(0.53) 0.31 (0.25) 0.30 (0.22) 0.18 0.12 0.04 0.01 0.006 0.08 0.07 0.06 0.04 ND 0.08(0.07) 0.05(0.08) 0.01(0.00) 0.06(0.08) 0.05(0.04) 0.05(0.03)1 HindIII-C 0.13 0.09(0.07) 0.01(0.01)
1 HindIII-D 0.115 0.67(0.57) 0.07(0.06)
1 HindIII-E 0.09 0.24(0.36) 0.03(0.04)
a3H-labeled1-and r-strandprobes complementary to late cytoplasmic RNA were hybridized tonitrocellulosestrips bearing
electrophoretically separatedDNAfragments,and thehybridizedradioactivity was quantitated by fluorography (Fig. 1). DNA fragments whichhybridizenondetectable amounts of radioactiveprobe are not presented in this table. ND, Not done.
'Calculatedastheproductof the effective sequencecomplexityoftheprobe used and the fraction of the total hybridized radioactivity annealedtoeachfragment.The effectivecomplexityofthe r-strandprobewas 0.65whenhybridizedtoEcoRI and BamHIfragmentsand0.61when annealedtotheHindIII fragmentsA toI. Effectivecomplexitiesof0.15 and0.106 were
assumed for the 1-strandprobe(see text).
cFragments BamHI-C and -D comigrate. Together they contain sequences homologous to 32% of the r-strand probe (=21%
of thetotalr-strand).The rationale forapportioningthese sequences between thetwofragments is giveninthetext.
dThe two numbersrepresent results from two separateexperiments.
tectable withinHindIII-A to -I and, therefore, resides withinoneor more of theHindIII
frag-ments excluded from the analysis (J-L). This
regioncanbeplacedbetween EcoRI-E(62map units) and HindIII-B (68mapunits).The
right-most 1-specific region maps to the right of 91
mapunits (entirelywithinHindIII-E).The rel-ativesizes of these threeregionswereestimated
fromthe fraction of theprobewhichhybridized
toeach DNAfragment (Table 2). Fromleftto
right these regionsare 6to8,4to 5,and 3to5 mapunitsinlength.If it is assumed thatthe
1-specific segment ofHindIII-C is contiguous to
that of
HindIII-D,
then theleftmost regioncanbe
placed
from about 11.5to18.5mapunits.The r-strand probe hybridized, tovaried de-grees, with all of the DNA fragments assayed (Fig. 1). Given the above-mentioned constraints onthepositions of thesegments ofthe
1-strand
complementarytocytoplasmic RNA,thisresult demonstrates thatatleastthree separateregions of the r-strandaretranscribed intocytoplasmic RNA.One of theseregionsliestothe leftof the leftmost
1-specific
block. The size of this seg-ment,estimatedfrom theextentofhybridization of the r-probe to EcoRI-C,BamHI-A,
andl
on November 10, 2019 by guest
http://jvi.asm.org/
232 SMILEY AND MAK
Eco RI F
C
I
DI
B IEUj
ABam HI
A I G
I(D.H.I)
I
FI
C I B I EHin dll (G.J,K,L)
F1
D I CI
A 1 1 B jHjlj Er---~~~~~-
_E_D
I a
0 0.25 0.5 0.75 1.0
FRACTIONAL LENGTH
FIG. 2. Map of sites of theAdl2genome expressed asearly and late cytoplasmic RNA. The map for late RNAwasdeducedfrom the data shown in Fig.1and
Table2. The mapfor early RNA was deduced from data shown in Table 3 and Fig. 5, together with
information fromOrtinetal. (24).Symbols: (_) late RNA; (o) early RNA.
HindIII-F and
-D,
is8to 11mapunits.Thenext region, whichbegins
totheright
ofthis1-strand
block (i.e., some
point
beyond
position
18.5) should encompass all or part ofEcoRI-D,
-B,
and -E,
BamHI-G, -D, -H, -I,
and-F,
andHindIII-C,
-A, and -G.(Although
the relativeorder of
HindIII-G, -J, -K,
and-Lisnotknown,
HindIII-G is
entirely
contained withinEcoRI-E [Smiley,unpublished
data]).
Therelativeextentof
hybridization
ofther-specific probe
tothesefragmentssuggeststhatmost orallof
EcoRI-B,
BamHI-G, -D, -H, -I,
and-F,
and HindIII-A and-Gare
complementary
totheprobe. However,
itappears that
only
aportion
of EcoRI-D and HindIII-C is complementary tothe probe. Be-cause ofthe uncertainties inherent in interpret-ingthequantitativedata,aspointedoutabove, it is difficulttoprecisely locatethe endsofthis block. Theavailable data suggest that it starts atorclosetoposition25and ends with EcoRI-E or -F andis,therefore, about 35mapunitsin length. The third r-strand region maps toEcoRI-A,
BamHI-C,
-B,and-E,andHindIII-B,
-H, and -I. Thissegment beginsto therightof
the
1-strand
region, which maps toBamHI-C,
and, basedonthe low levels of
hybridization
toHindIII-E,
appears to end aroundposition
92.Several difficulties were encountered in the interpretation of the data obtained with the r-strand probe. First, because BamHI-C and -D were not separable electrophoretically, we ob-tainedanestimate of the fraction of the r-strand transcribed from within both theseregions. The data obtained with thefragments generated by EcoRIandHindIII indicate thatvirtually all of the r-strand within EcoRI-B and HindIII-A is transcribed ascytoplasmic RNA. The BamHI-Dfragmentmapsentirelywithinthisregion, and consequently wehave assumed total transcrip-tion of the r-strand of this fragment andhave assignedexcesshybridizationtoBamHI-C. Sec-ond, the methodcalculating the fraction of each strand transcribed within each region assumes that all ofthe DNA sequences
complementary
to the probe used are present in the analysis. FragmentsHindIII-Jto-L andEcoRI-Fwere in fact excluded because of their small size. The exclusion of EcoRI-F (2% ofAdl2) hasa negli-gibleeffect, but theexclusion ofHindIII-Jto -L (approximately 8% of Adl2) complicates the analysis. However, asdiscussedabove,aregion comprising 4 to 5 map units located entirely withinfragments HindIII-J to-Lis transcribed from the1-strand andcanbeexcluded from the analysis of r-strand-specific RNA. Effectively, only about3 to4% of Adl2 is then excludedby not examining thesefragments. Assuming that all of this DNAisrepresented as r-strand-spe-cific cytoplasmic RNA, the effective complexity of the r-strand DNA probe becomes 0.61. The effects of this alterationincomplexityaresmall. The values given in Table 2 were calculated using 0.61 as the complexity of the r-strand probe.
Analysis of the RNA sequences actively synthesized at late times. The results out-lined above serve to map the positions and strand specificities of the DNA sequences rep-resented in late cytoplasmic RNA but give no indication of whichsegmentsofthegenome are actively transcribedatlate times.Todetermine whether both strandsaretranscribed, Adl2-in-fectedcellswerelabeled from 21 to 24 h postin-fection with
[3H]uridine,
and the cytoplasmic RNA waspurified.It wasfoundthat this
[3H]RNA
could hybrid-izetoboth 1- and r-strands ofAdl2
DNA (data not shown). To determine which segments of each strand areactivelytranscribed at 24 h after infection,the3H-labeledviral RNAwasselected byhybridizationto1- and r-strandDNA,eluted, andrehybridizedtounlabeled EcoRIfragments
(Fig. 3).Itis clear thatmostRNA
species
pres-ent in thecytoplasm atlatetimes are actively synthesizedandtransportedtothecytoplasmat J. VIROL.on November 10, 2019 by guest
http://jvi.asm.org/
[image:6.500.60.253.55.310.2]w
u z
rstrand RNA
0 (I)
I strand RNA
MIGRATION
FIG. 3. Hybridization oflabeled late complemen-tary strand-specific RNA to unlabeled EcoRI frag-ments. CytoplasmicRNA labeled with[3H]uridine from21to24 hpostinfection wasstrandselectedas
described in thetextandhybridizedtonitrocellulose strips bearing 1 pg of denatured, separated Adl2
EcoRIfragments. Hybridized radioactivity was de-tectedasinFig.1.
late times. This is in contrast tothe result
re-ported thatatlate times after infection with Ad2 the RNA is transcribed exclusively from the
r-strand of thegenome (26).
It is possible that the labeling times chosen
are not comparable to the 16- to 18-h period
usedby PetterssonandPhilipson (26)with Ad2
and that analysis of RNA synthesized later
would reveal only r-strand transcription. It is
worthnoting, however, that the maximal rates
of viralDNA and RNAsynthesisoccurat 24 h
postinfection with Adl2 (21), whereas this oc-curs at 18 hpostinfection with Ad2 (14). Since
the strand-specific RNA used in the previous experimentwasisolatedbyhybridizing
near-sat-urating amounts of RNA to each strand, the
results of the experiment displayed in Fig. 3 cannotbe usedtoestimate the relative amounts
of radioactivity incorporated in RNA derived
fromdifferentregionsof thegenome.
To obtain estimates of RNA transcription
ratesfrom different regionsof the genome, cy-toplasmic RNAlabeled 24to26 hpostinfection with[3H]uridinewasdegradedin alkalitoabout
500 nucleotides in chainlength (1), and 2,tgof
this RNA (specific activity, 3.4 x 104 cpm/,ug)
was hybridized directly to excess (2 ,tg) frag-ments generated by EcoRI and BamHI. The
relative amount ofradioactivity hybridizing to
eachbandwas taken as a measure of the relative ratesofaccumulation of cytoplasmic RNA de-rived from each fragment. The data obtained (Table 3) indicate tht RNA species derived from different regions of the genome accumulate at widely differing rates. During the labeling period chosen,RNA derived from about 25 to 50 map units onthe genome (EcoRI-B and -D) accu-mulatesat a greater rate than any other class. Mapping of the early regions. At early times,viral RNA forms an extremelysmall frac-tion of the total cytoplasmic RNA. Conse-quently, themethod used to map the late RNA wasdifficultto apply to early RNA, since a large amount of viral RNA is required to saturate substantial quantities of radioactively labeled viralDNA.
Someinformationabout the DNA sequences encoding earlycytoplasmic RNA was obtained by hybridizing cytoplasmic RNA labeled with [3H]uridine from 5.5 to 7.5 h postinfection to
TABLE 3. Distributionof labeled late cytoplasmic RNA sequences among DNAfragmentsa
Fraction of total Specific accumula-Fragment RNA hybridized tion rateb
EcoRI-A 0.07 0.19
EcoRI-B 0.60 2.22
EcoRI-C 0.07 0.42
EcoRI-D 0.20 1.74
EcoRI-E 0.06 0.86
BamHI-A 0.06 0.24
BamHI-B 0.03 0.19
BamHI-C
.05d
0.42CBamHI-D * 3.93C
BamHI-E 0.02 0.17
BamHI-F 0.07 0.87
BamHI-G 0.16 2.55
BamHI-H+I 0.12 1.80
aLate
cytoplasmic
RNAlabeled24to 26 hpostin-fection with[3H]uridine degradedtoabout500 nucleo-tidesin chainlengthwashybridized toexcess (2,&g)
unlabeled DNAfragments, and the hybridized radio-activitywasquantitatedbyfluorography.
bDefined as the fraction of the total hybridized radioactivity annealing to each fragment divided by thefractionallength of that fragment.
cTherelative distribution ofhybridized
radioactiv-ity between BamHI-C and-D wasestimatedby meas-uring hybridizationtofragmentsproducedby simul-taneouscleavagewithEcoRI andBamHI.The ratio of hybridized radioactivity between BamHI-C/ EcoRIa (within BamHI-C) and BamHI-Dwas0.07.
BamHI-C/EcoRIa represents 0.71 of BamHI-C. As-suming that accumulation within BamHI-C is
uni-form,thisyields aspecific accumulationrateof0.42 forBamHI-C.
d BamHI-C and-D are notseparable
electrophoret-ically.This numberrepresentsthesumof the radio-activityhybridizingtoeachfragment.
E D C BA
II
on November 10, 2019 by guest
http://jvi.asm.org/
[image:7.500.55.249.72.288.2] [image:7.500.258.451.316.471.2]234 SMILEY AND MAK
DNA fragments produced by EcoRI, BamHI, andHindIII. Relativeamountsof[3H]RNA hy-bridizedtoeachfragmentareshown in Table 4. Hybridization was detected with
EcoRI-A,
-B, and-C, BamHI-A, -B,-C, and -E and HindIII-B, -C, -E, -F, and -H. FragmentsEcoRI-F and HindIII-J, -K, and -Lwerenotanalyzed.
These data demonstrate that early RNA is derived fromatleastthree separateregions
of the viral genome.PositivehybridizationtoHindIII-E butnotto HindIII-Iindicated thatone
region
is contained entirely within HindIII-E.Similarly,
positive hybridization to HindIII-F but not -D places anotherregionnearthe left end of the genome. Results from hybridization with EcoRI and BamHI fragments areconsistent with thiscon-clusion.Atleastone
region
is located in BamHI-B and -C. The [3H]RNAhybridized
to theBamHI-C/D
doubletwasassigned
toBamHI-Cby the
following experiment.
Adl2 DNA wascleaved
simultaneously
withEcoRI andBamHI,and the
fragments
wereseparated
by
electro-phoresis (Fig. 4A). BamHI-D isnot cleaved
by
EcoRI,but BamHI-C is cleaved into three
frag-ments
(BamHI-C/EcoRIa
[fragment
VI inFig.
4A, lane2]andtwosmallerones). Resultsshown inTable4show that
early [3H]RNA hybridized
onlytoBam-C/EcoRIfa
(designated
asBamHI/
EcoRI VI in Table 4) but not to BamHI-D (designated asBamHI/EcoRI III). The lack of hybridizationtoHindIII-Aconfirmsthis conclu-sion.
The observed hybridization of early RNA to EcoRI-B (6%ofthetotalhybridized radioactiv-ity; Table 4) does not agree with the data ob-tainedwith BamHI, where no hybridization was
seenwithfragmentsD,H, I, and F and only very small amounts (<1% of the total) were found withG.Similarly,nohybridizationwasobserved
to HindIII-A and -G (<1%), and only small amounts of hybridization were observed to HindIII-C (1%). The rather high levels of hy-bridization ofearly RNA to EcoRI-B can per-haps be explainedby postulating the existence
ofminorityrearrangement genomes bearing se-quencesderived from the normal EcoRI-A
and/
[image:8.500.258.449.95.307.2]or-Cfragments onfragments of about the size ofEcoRI-B. Such genomes have in fact been
TABLE 4. DistributionoflabeledearlyRNA sequences amongDNAfragmentsa
Fractionof
Maxi-total radio- mumsize Specific
Fragment activityhy- ofregion
accumu-bridizedan- encoding lation
nealed to the early rateb fragment RNA
EcoRI-A 0.47
EcoRI-B 0.06
EcoRI-C 0.47
BamHI-A 0.60 0.067 8.96
BamHI-B 0.11 0.15 0.73
BamHI-C+D 0.09 0.05 1.80
BamHI-E 0.21 0.05 4.20
BamHI-G <0.01
HindIII-B 0.05
HindIII-C 0.01
HindIII-E 0.40
HindIII-F 0.50
HindIII-H 0.13
BamHI+EcoRI-I+IIf 0.63 BamHI+EcoRI-IV 0.28 BamHI+EcoRI-VI 0.08
'CytoplasmicRNA labeled 5.5 to7.5 h postinfection was
hybridizedto2yg of unlabeled Adl2 DNAfragments. The
analyses withEcoRIandBamHIfragmentsweredone with the same RNApreparation,whereasanotherlotof RNAwas usedfor the experimentswiththeHindIII fragments.
Frag-mentsEcoRI-A-E, BamHI-A-I,HindIII-A-I,and BamHI +
EcoRI-I-VIIwere assayed.Onlythosefragmentswhich
hy-bridized detectableradioactivityaretabulated. The limit of detection of the assay is about 1% of the totalhybridized
radioactivity.
b Defined as the fraction of the totalhybridized radioactivity
which annealed to agivenfragment divided bythe maximum size of the regionencodingtheearlyRNA.
'The identityoffragmentsproducedby the double diges-tionare:I,EcoRI-C; II,BamHI-B; III,BamHI-D; IV,
BamHI-E;and VI,Bam-HI-C/EcoRIa.
observed within Adl2 (Huie) populations (I. Mak, H.Ezoe, and S. Mak,manuscript in prep-aration). From data in Table 4, it can be seen thatthedifferent earlyRNA species accumulate atwidely differing rates, with RNA encoded by themolecular ends of the genome accumulating
mostrapidly.These RNA species together com-prise from 75 to 90% of the total viral RNA detected.
Strand assignment of the early regions. To determine directly the strand derivation of
FIG. 4. Electrophoreticstrandseparation of someAdl2DNAfragments.Adl2DNA wascleaved simulta-neouslywith BamHI and EcoRI, and the resulting fragments were separated in 1% agarose gels. Fragments EcoRI-C, BamHI-B, BamHI-E, andBamHI-C/EcoRIa were isolated in gel slices, denatured in situ, and again subjectedtoelectrophoresis.(A)Electrophoreticseparation of fragments produced byEcoRI(1),EcoRI plusBamHI (2, 3), and BamHI (4) in 1% agarose gels,allowing theidentificationof theorigin of some of the fragmentsproduced by doubledigestion. (B)Electrophoresisofdenatured DNAfragments in 1% agarose gels. 1,EcoRI-CDNA;2,BamHI-BDNA; 3,BamHI-E DNA; 4,BamHI-C/EcoRIaDNA.Electrophoresis was for 8hat 2V/cm.
J. VIROL.
on November 10, 2019 by guest
http://jvi.asm.org/
B
1234
235
AI
on November 10, 2019 by guest
http://jvi.asm.org/
236 SMILEY AND MAK
the regions coding for the early RNA, we at-tempted to resolve the strands of the DNA fragments(i.e.,EcoRI-C, BamHI-B and -E, and the BamHI-C/EcoRIa) by gelelectrophoresis. Themethod used is that used by Horowitz (16). Figure 4B shows that fragments EcoRI-C and BamHI-B and -Ewere resolved into doublets, whereasdenatured
BamHI-C/EcoRIa
migrated as asingleband. To determine the strand deri-vationof the fast- and slow-migrating bands of EcoRI-C, BamHI-B, and BamHI-E, the rele-vantportion of each gel wasexcised and com-positesweremade. TheDNAfromthe compos-ite gels was transferred to nitrocellulose strips by the Southern method (34) without further denaturation. These stripswerethenhybridizedto purified,
3H-labeled,
intact l- and r-strandsand
fluorographed (Fig.
5).Thelabeledcomple-mentary strands
hybridized
primarily
to one band of thedoublet, allowing
strandassignments
to each band to be made. The
fast-migrating
band of
EcoRI-C
is derived from the intact1-strand, whereas the fast bands of BamHI-B and -E are part of the intact r-strand. The strand derivations of
early
RNAs were thendirectly
determined
byhybridizing
labeled RNAtosuchstrips (Fig. 5). The results demonstrate that early
cytoplasmic
RNA is derived from the r-strand ofEcoRI-C
and BamHI-B and the 1-strand ofBamHI-E. The strand derivation of early RNA with BamHI-Cwas notdirectly
de-termined.A
.. ;.I
.-EP.i
r8
r'
41 M iR.-I~~~~~~~~,s
a...
.. ffY"T.
U
I
I
RI
..
A.
A.
41
Pl
Bam
F
I
kI
It#
i
i i -kiI
[image:10.500.115.397.254.575.2]Zi
FIG. 5. Hybridization of H-labeledintact1-andr-strand DNA and early cytoplasmic RNAto complemen-tarystrandof EcoRI-C,BamHI-B,andBamHI-E DNA.Each DNA fragmentwasdenatured and subjected
toelectrophoresisinagarosegels.The gelswereexaminedunderUV illumination after staining with0.5 pug ofethidiumbromideper mLTheregion of eachgel containing the double bandswasexcised, andcomposite
gels werecreated byend-to-end alignment of the excisedsegments. The DNA in the composite wasthen
transferredtoanitrocellulose membrane andhybridizedto10i cpmof 3H-labeled r-strandAdl2 DNA (A)or
1-strandAdl2DNA(B) andto5x106cpmof 3H-labeledearly cytoplasmic RNA (C). Hybridized radioactivity wasdetectedbyfluorography.Aschematicrepresentation of the positionsof each of the DNA bands in the
composite gelsis shown.
.^
_ _
'SF
_
_ _ :.
t.2
I
I
f.
on November 10, 2019 by guest
http://jvi.asm.org/
12 237 DISCUSSION
We havedevelopeda newmethod of mapping
viral genes expressed late during infection. It
involves theisolation of radioactively labeled l-and r-strl-and DNAsequencescomplementaryto
latecytoplasmic RNA and, subsequently, rehy-bridization of these sequences to DNA
frag-ments generated by restriction endonucleases. The method has several advantages over the
standardones. First, the requirement for strand
separating the individual DNA fragments is by-passed. Second, the transcribed segments of each strand can be hybridized to any set of ordered DNAfragmentstorefine themapwith a minimum of additional effort. Drawbacks of
the method include the lack ofinformation on
the abundance of different RNA species, the requirement for rather large quantities of viral RNA toprepare the DNA probes, and the
re-duced precision of the method ascompared to
saturationhybridization. The ends of blocks of transcribed DNAcanbe locatedonlytowithin
several map units by this procedure, and it is likely that small gaps withinwhatwe have
as-sumedareregions of contiguous
complementar-itytocytoplasmic RNA wouldgounnoticed.
Using this new method and the DNA
frag-mentsgeneratedby three restriction
endonucle-ases, we have mapped the sites of the Adl2
(strain 1131)genomewhichareexpressedaslate
cytoplasmic RNA. The map is similar to the Adl2(Huie)mapobtained by saturation
hybrid-ization of labeled, complementary strand-spe-cificsequencesgeneratedby EcoRI and BamHI (29). The closeagreementofthetwo transcrip-tionmaps indicates thatourmethod is suitable
for mapping late mRNA. The Adl2 transcrip-tional map for late cytoplasmic RNA is very
similar to those previously obtained with Ad2 (27, 31) and Ad5, with the exception that a
region of Adl2 DNA extending from approxi-mately 17 to 25 map units is apparently not
transcribed into stable cytoplasmic RNA. The corresponding region of Ad2 and Ad5 DNA is transcribed asstable RNA (27, 31).
Well before the onset ofviral DNAsynthesis,
we have detected cytoplasmic RNA derived
fromaminimum ofthreeseparateregions of the
Adl2 genome. Ortin et al. (23) have detected
fourseparateblocks of DNAsequencesencoding
early Adl2 cytoplasmic RNA, mapping within the r-strand of EcoRI-CandBamHI-B, and the 1-strand of BamHI-C and BamHI-E. We were
not able to separate the regions mapped in BamHI-B and -C. The data ofOrtin etal. (23) and those obtained in this studyallow a quite
precise mapping of the four major early regions.
The leftmost block, transcribed from the r-strand, maps between 0 and 6.3 map units (withinHindIII-F). Ortinet al. have estimated thesize of this region as about 7.5%ofAdl2.If itis assumed that the earlyRNA derivedfrom BamHI-C is transcribed entirely from the 1-strand (23), this segment mustlie betweenthe EcoRI-E-F junction (62 map units) and the leftmost end of the HindIII-B (68 map units). The maximum size ofthis region is 4 to 5 map units. The total lack of hybridization of early cytoplasmic RNA to HindIII-I allows placing the right terminus of the early region within BamHI-B to the left of 88.1 map units (the HindIII-H-I boundary).Themaximumpossible size ofthisblockisthen 15mapunits,which is identicaltothe estimates obtainedby saturation hybridization (23). The rightmostearly region, transcribed from the
1-strand,
mapsentirelyto the right of91 mapunitsand isprobably
iden-ticaltotherightmost late1-specific
region. The sites of these early regions are also similar to those of Ad2 (27, 31). Itwas ofinteresttonote that the extreme left r-strand early region of Adl2 (at most 6.3 map units) is considerably smaller than the analogous region of Ad2 (11 mapunits) (27, 31).At both early and late times postinfection, cytoplasmic RNAs complementary to different segments ofthe genome accumulate at
widely
differentrates.Figure6showsthe
specific
rates of accumulation in thecytoplasm
of RNA-de-rived different regions of the genome atearly
and late times. The data obtained with
early
RNA have been normalized to the maximum
possible
sizeof eachearly coding
region,
whereasthelate data have been normalizedto
fragment
size. Ifnotall DNAsequencesare a
template
forthe
synthesis
of latecytoplasmic
RNA(for
ex-ample,
theregion
from18to25mapunits),
thisprocedure willunderestimate the
specific
accu-mulation rate of RNA derived fromsequencesflanking
the nontranscribedsegments.Thema-jor RNAspecies
synthesized
andtransported
tothe
cytoplasm during
the interval between 5.5and 7.5 h postinfection are derived from the molecular ends of the genome
(HindHII-E
and -F).Smalleramounts aredetected from BamHI-B and -C. By 24 to 26 h after infection the patternhasdrastically
changed,with RNA de-rived from between 25 to about 60 map units being byfar themostrapidly
accumulating
class. This RNA constitutes the major classsynthe-sizedexclusivelyatlate times. The
correspond-ingregionof Ad2 encodes themajorvirion pro-teins(20).Ifsimilar
regions
of theAd2 andAdl2 genomes encodeanalogous
proteins,
thisresultisnot
surprising.
on November 10, 2019 by guest
http://jvi.asm.org/
238 SMILEY AND MAK
class in the nucleus and cytoplasm. At 32 h postinfectionRNAmappingbetween 66 and 80 map units comprises over half of the total viral cytoplasmicRNA. Our data indicate thatduring the interval for 24 to 26 h postinfection with
Adl2,
RNA derived from 25 to 60 map units accumulates most rapidly in the cytoplasm. Since the labeling time chosen corresponds to the time at which the maximal rate of viral RNA synthesisoccurs(14),this RNAwillpresumably eventually constitute the most abundant cyto-plasmic class. This may indicate a difference in the fine control of Ad2 and Adl2 RNA accu-mulation.The distribution ofcontiguous regions encod-ing early and late cytoplasmic RNA is virtually identical in Ad2, Ad5, Ad7 (37), and Adl2. The gross homology of genome organization of the group A, B, and C adenoviruses may indicate that genes with similar functions share common I _ t v map
positions.
Thissupposition
isstrengthened
by the finding that the gene(s) responsible for
0 0.25 0.5 0.75 1.0 the maintenance of in vitro transformation in
bothgroupsof viruses maptothe left (guanine FRACTIONAL LENGTH pluscytosine rich) end of thegenome (7, 9, 11; Correlationof the rates of accumulation of Mak et al., manuscript inpreparation). The ob-tic RNAcomplementary to DNAfragments served
simflarity
further impliesthat during theD position. The data obtained with the evolutionary divergence of the group A, B, and
'ragments, shown in Tables 3 and 4, are Cadenoviruses theoveralllayout of the genome
l.Specificaccumulationratesfor early RNA
ere calculated by dividing thefraction of the has beenpreserved,whereastheDNAsequences !y RNApresent in each class by the maxi- of the idividual genes have diverged to the sible fractional size of the corresponding pointwhere very little sequence homology can
ling region. (Upper) Cytoplasmic RNA la- be detected (17). beled5.5 to 7.5 hpostinfection. (Lower) Cytoplasmic
RNAlabeled24 to 26hpostinfection.
These results clearly show that both the "early" and "late" cytoplasmic RNAs can be subdivided accordingtotheirrates of accumu-lation.Byexamining onlytwoperiods postinfec-tion we are able to distinguish three classesof RNA onthisbasis: 1, RNA whichaccumulates rapidly early (derived from HindIII-E and -F); 2, those early RNA species which accumulate moreslowly (withinBamHI-B and-C);3, RNA not detectableearly which accumulates rapidly late (derived from EcoRI-B, -D, and -E). It is possiblethat an examination of moretime inter-valswill reveal furthercomplexity. It istempting tospeculatethat these RNA classesdiffer in the timesof their maximalrate ofaccumulation.
Flint andSharp (8) have quantitated the lev-els of Ad2 nuclear andcytoplasmic RNA derived from different segments of the viral genome at varioustimespostinfection. At both 18 and 32 h postinfection, RNA derived from theregion of thegenomeextendingfromapproximately65 to 90 map units constitutes the most abundant
ACKNOWLEDGMENTS
This work was supported by grants from the National
Cancer Institute of Canada and National Research Council of Canada. J.R.S.isaResearch Student of th-e National Cancer Institute of Canada.
LITERATURE CITED
1. Bachenheimer,S., and J. E. Darnell. 1975. Adenovirus 2 mRNA is transcribed as partofa high molecular
weightprecursorRNA. Proc. Natl. Acad. Sci. U.S.A.
72:445-449.
2. Chinnadurai, G., H. M. Rho, R. B. Horton, and M. Green. 1976. mRNAfrom thetransforming segment of theadenovirus 2genomeinproductivelyinfected and transformedcells.J. Virol.20:255-263.
3. Craig, E. A., and H. J. Raskas. 1974. Two classes of earlycytoplasmicviral RNAsynthesizedearly in
pro-ductiveinfection with adenovirus 2. J. Virol. 14:26-32.
4. Craig, E. A., S. Zimmer, and H. J. Raskas. 1975. Analysis of early adenovirus 2 RNA using Eco R R1 viral DNAfragments. J. Virol. 15:1202-1213. 5. Flint, S. J., S. M. Berget, and P. A. Sharp. 1976.
Adenovirustranscription. III. Mapping of viral RNA
sequencesincellsproductivelyinfectedbyadenovirus type5.Virology 72:443-455.
6. Flint, S. J., P. H. Gallimore, and P. A.Sharp. 1975.
Comparison of viral RNAsequencesin adenovirus
2-transformed cellsandlytically-infected cells.J. Mol.
Biol.96:47-68.
7.Flint, S. J., J. Sambrook, J. F. Williams, andP. A.
10
8
6 4
2 0
3
2 z 0
4
-i
U D
0. CL)
0
FIG. 6. cytoplasm with maj
BamHIf,
displayed
specieswi total earl mum pos early cod
on November 10, 2019 by guest
http://jvi.asm.org/
[image:12.500.59.252.59.318.2]Sharp. 1976. Viral nucleic acid sequences in trans-formed cells. IV. A study of the sequences of adenovirus 5DNA and RNA in four lines of adenovirus 5-trans-formed rodent cells using specific fragments of the viral genome.Virology 72:456-470.
8. Flint,S. J., and P. A. Sharp. 1976. Adenovirus transcrip-tion. V. Quantitation of viral RNA sequences in ade-novirus 2-infected andtransformed cells. J. Mol. Biol. 106:749-771.
9. Gallimore, P. H., P. A. Sharp, and J. Sambrook. 1974. Viral DNA in transformed cells. HI. A study of the sequencesof adenovirus 2 DNA in ninelines of trans-formed rat cells using specific fragments of the viral genome.J. Mol. Biol. 89:49-72.
10.Garon,C. F., K. W. Berry, J.C. Hierholzer, and J. A. Rose. 1973. Mapping of base sequence heterologies between genomes fromdifferent adenovirus serotypes. Virology 54:414-426.
11.Graham, F. L., P. J. Abrahams, C. Mulder, H. L. Heijneker, S.0. Warnaar, F. A. J. de Vries, W. Fiers,andA.J.vander Eb.1974.Studiesonin vitro transformationbyDNA and DNAfragmentsof human adenoviruses and simian virus 40. ColdSpring Harbor Symp. Quant.Biol. 39:637-650.
12. Green, M.,and M. Pina.1963.Biochemical studies on adenovirus multiplication. IV. Isolation, purification and chemical analysis of adenovirus. Virology 20:199-207.
13.Green,M., and M. Pina. 1964. Biochemical studies of adenovirusmultiplication.VI.Properties of highly pur-ified tumorigenic human adenoviruses and their DNAs. Proc.Natl. Acad.Sci. U.S.A. 51:1251-1259.
14.Green, M.,M. Pina, K.Funjinaga, S.Mak, and D. Thomas. 1968.Transcriptionof viral genes in adeno-virus-infected and transformed cells. Perspect. Virol. 6:15-37.
15.Hayward, G. S.1972.Gelelectrophoretic separationof thecomplementarystrands ofbacteriophageDNA. Vi-rology 49:342-344.
16. Horowitz,M. S. 1976.Bidirectionalreplication of ade-novirus type2DNA. J.Virol. 18:307-315.
17.Lacy,S., and M. Green. 1964.Biochemical studieson
adenovirus multiplication. VII. Homology between DNAs oftumorigenicandnontumorigenic human ade-noviruses. Proc.Natl. Acad. Sci. U.S.A.52:1053-1059.
18. Landgraf-Leurs, L,and M. Green. 1971. Adenovirus DNA. III.Separationof thecomplementarystrands of adenovirus2, 7,and 12DNAmolecules. J.Mol. Biol. 60:185-202.
19. Laskey,R.A.,and A.D.Mills.1975.Quantitativefilm detection of3Hand"4Cinpolyacrylamide gels by fluo-rography. Eur. J. Biochem.56:335-341.
20. Lewis, J. B., J. C. Atkins, C.W. Anderson, P. R. Baum, and R. F. Gesteland. 1975.Mappingoflate adenovirus genesbycell-free translation of RNA
se-quencesselectedby hybridizationtospecificDNA
frag-ments.Proc. Natl. Acad. Sci.U.S.A.72:1344-1348. 21. Mak,S.1969.Transcriptionandreplicationof viral
de-oxyribonucleic acid in cells coinfected with adenovi-rusestypes2and12.J. Virol.4:651-656.
22. Mulder, C.,P. A.Sharp,H.Delius,and U.Pettersson.
1974. Specific fragmentation of DNA of adenovirus
serotypes 3, 5, 7 and 12, and adeno-simian virus 40
hybrid virus Ad2'ND1 by restriction endonuclease R EcoRI. J. Virol. 14:68-77.
23. Ortin, J.,K.H.Scheidtmann,R.Greenberg,M. West-phal, and W.Doerfler. 1976. Transcription of the genomeof adenovirus type12.III.Maps of stable RNA from productively infected humancells and abortively infected and transformed hamster cells. J. Virol. 20:355-372.
24. Pereira, M. S., and F.0.MacCallum. 1964. Infection with adenovirus type 12. Lanceti:198-199.
25. Pettersson, U., C.Mulder,H.Deluis,and P. A.Sharp. 1973. Cleavage of adenovirus type 2 DNA into six uniquefragments by endonuclease RI. Proc. Natl. Acad. Sci.U.S.A. 70:200-204.
26. Pettersson, U., and L. Philipson. 1974. Synthesisof complementary RNA sequencesduring productive ad-enovirus infection. Proc. Natl. Acad. Sci. U.S.A. 71:4887-4891.
27. Pettersson,U.,C. Tibbetts,andL.Philipson. 1976. Hybridizationmapsofearly and late messenger RNA sequences of theadenovirus type 2genome. J. Mol. Biol. 101:479-501.
28. Philipson, L., U. Pettersson, U. Lindberg, C. Tib-betts, B.Vennstrom, and T. Persson. 1974. RNA synthesis and processing in adenovirus-infected cells. ColdSpring HarborSymp.Quant. Biol. 39:447-456. 29. Scheidtmann,K.-H.,and W.Doerfler.1977.
Transcrip-tion of the genome of adenovirus type12.VI.Mapsof stable late RNA from productively infected human cells. J. Virol. 22:585-590.
30. Sharp,P.A.,and S. J.Flint.1976.Adenovirus transcrip-tion. I. Curr.Top.Microbiol. Immunol. 74:137-166. 31. Sharp,P.A.,P. H.Gallimore,and S. J.Flint.1974.
Mapping of adenovirus2RNA sequences inlytically infectedcells and transformed cell lines. Cold Spring HaborSymp. Quant.Biol. 39:457-474.
32. Sharp, P. A., B. Sugden, and J. Sambrook. 1973.
Detection oftworestriction endonuclease activities in Hemophilus parainfluenzae using analytical agarose ethidium bromide electrophoresis. Biochemistry 12:3055-3063.
33. Smiley, J. R., and S. Mak. 1976. Adenovirus type 12-specificRNA sequences during productive infection of KBcells.J. Virol.19:36-42.
34. Southern, E. M. 1975. Detection of specific sequences amongDNAfragmentsseparated by gel electrophore-sis.J. Mol. Biol.98:503-517.
35. Thomas, D. C., and M. Green. 1969. Biochemical studies
onadenovirusmultiplication.XV. Transcription of the adenovirus type 2 genome during productive infection. Virology39:205-210.
36. Thuring,R. W.J., J. P. M. Sanders, and P. Borst. 1975. Afreeze-squeeze method for recovering long DNA from agarosegels. Anal. Biochem. 66:213-220. 37. Tibbetts,C. 1977. Physical organization of subgroup B
humanadenovirus genomes. J. Virol. 24:564-579. 38. Tibbetts, C., K.Johansson,and L. Philipson. 1973.
Hydroxyapatite chromatography and formamide dena-turation ofadenovirus DNA. J. Virol. 12:218-225. 39. Tibbetts, C.,and U. Pettersson. 1974. Complementary
strand-specific sequences from unique fragments of
ad-enovirus 2 DNA for hybridization mapping experi-ments. J.Mol. Biol. 88:767-784.