0022-538X/82/120772-10$02.00/0
Accurate Transcription of Simian Virus 40 Chromatin in
aHeLa Cell Extract
JOHN N.BRADY, MICHAEL RADONOVICH,ANDNORMAN P. SALZMAN*
Laboratory of Biology of Viruses, National Institute of Allergy and InfectiousDiseases,Bethesda, Maryland
20205
Received 17 May1982/Accepted 23 August 1982
During simian virus 40 viral maturation,a series ofmodifications occurwhich
alter the composition of viral nucleoprotein complexes. As a consequence, the
chromatin that is extracted from extracellular simian virus 40 virions exhibits
properties thataresimilar tothose of transcriptionally active eucaryotic
chroma-tin.Theinfluence of this chromatin structureonspecificRNAinitiation by RNA polymerase IIwasexaminedby using the in vitro HeLa cellextractof Manleyet
al. (Proc. Natl. Acad. Sci. U.S.A. 77:3855-3859, 1980). The 5' ends of RNA
transcripts were positioned by the "run-off' assay, in which transcripts extend from the initiation sitetotermination sitescreated by restrictioncleavage and by
S1 nuclease analysis, using DNA probes labeled at their5' termini. Two major
earlyRNAtranscripts, which originated atmappositions 5,240 ± 10and5,145 ±
10, andtwomajor late RNA transcripts, which originatedatmappositions 325 ±
10 and 185 ± 10, were identified. Transcripts were initiated with comparable
relativeefficienciesatthesame5' site wheneither purifiedDNAorchromatinwas usedasthe template. Our resultssuggestthat extracellular simian virus40virion
chromatin modifications do notregulate simian virus 40promoter selection but
function to increase the accessibility of RNA promoter sequences to RNA
polymerase IIand allowefficient elongation of theRNAchain afterthe initiation
event.
The DNA of eucaryotic cells is associated with octomers of four histone proteins which condense the DNAinto the basicrepeatunit of chromatin, the nucleosome (29). The
nucleo-some structure effectively blocks the access of
enzymes to DNA and therefore inhibits RNA
chain initiationandelongation (7, 9, 46). Recent
studies have shown that changes in chromatin
structure occur asgenes areactivated for
tran-scription; thissuggeststhat chromatinstructure
isonedeterminant ofregulation oftranscription
ineucaryoticcells(6,8, 10, 12,16,25,32, 33,37,
44, 47,48). Analyses of severalgeneticlocihave
demonstrated that during periods of
transcrip-tionalactivity, chromatin DNAbecomes
hyper-sensitive to DNase I, apparently due to modifi-cations in the chromatin structure, including
histone acetylation, decrease in H1 stoichiom-etry,and interaction of nonhistoneproteinswith chromatin.
Two in vitro RNA transcription systems
which accurately transcribe purified DNA
tem-plateshave beendeveloped recently. Weiletal.
(43) useda combination ofpurified RNA
poly-merase II and a crude HeLa cell cytoplasmic
extract toprovide specifictranscriptionof DNA
templates, whereasManley etal. (26)promoted
specifictranscription byadding DNAtemplates
to aconcentrated HeLawhole-cellextract.
Us-ingavariety ofDNAtemplates, several investi-gators have demonstrated that both of these soluble systems recognize specificDNA
regula-tory sequences(e.g., theGoldberg-Hogness
se-quence TATA) and that RNA chain initiation
starts atsitesidenticaltothose thatoccurin vivo
(2, 18, 20, 21, 27, 38, 41, 42).
We have previously reported that the 110S
chromatin complex isolated from extracellular
simian virus 40 (SV40) virions is an efficient template forRNA transcription byEscherichia coli RNApolymerase(4, 5). We hadnot charac-terized the suitability of this complex as a
tem-plate forthehomologous eucaryoticRNA
poly-merase II or determined if the chromatin
structurealtered theabilityof RNApolymerase
II torecognize promoter sequences andinitiate
specific RNA synthesis. In this paper, we
de-scribe the transcription of extracellular virion chromatin by RNA polymerase II in the HeLa
extractofManleyet al. (26).
MATERIALSANDMETHODS
Viruses and cells. SV40 strain 776 was originally
obtained from K. Takemoto, National Institute of
772
on November 10, 2019 by guest
http://jvi.asm.org/
Allergy and Infectious Diseases. Theviruswasgrown
inBSC-1 cells.
Infection andpurification of SV40 virions.Confluent
monolayers of BSC-1 cells maintained in Eagle mini-malessential medium supplemented with 0.03% gluta-mine, 10% fetal calfserum, streptomycin, penicillin, ampicillin, andmycostatinwereinfected withSV40at amultiplicity of infection of 10PFU/cell. After2 h of
adsorption at 37°C, Eagle minimal essential medium
supplemented with 2% fetal calf serum was added. Whencelllysiswascomplete(approximately9days), theinfected cells and supernatantwerecollected,and
the viruswaspurifiedaspreviously described (5). Preparation of chromatin complexes and conditions for in vitro RNAsynthesis. Preparation of SV40
chro-matincomplexeswascarriedoutbyethylene glycol-bis(O-aminoethyl ether)-N,N'-tetraacetic acid
(EGTA)-dithiothreitol dissociation ofSV40 virionsas
described previously (4, 5), except that the NaCl
concentration was decreased to 100 mM. HeLa cell extractswereprepared bythemethodofManleyetal.
(26), using cells thatwereharvestedatacell concen-tration of5 x 105to6x 105 cellspermlinstead of 8 x 105cellspermlasoriginally described. The concentra-tion of template DNArequired for specific transcrip-tionwasdecreased whenthe cells wereharvestedat the lower cell density. Each standard transcription reaction mixture (60 ,ul) contained 10 mMHEPES
(N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid)
(pH 7.9), 8 mM Tris (pH 8.6), 50 mM KCl, 16 mM NaCl, 6.25 mM MgC92, 0.8 mM EGTA, 0.05 mM EDTA, 1.5 mMdithiothreitol, 0.16 mM phenylmethyl-sulfonyl fluoride, 0.008% Triton X-100, 8.5%glycerol, 0.4 mM ATP, 0.4 mM GTP, 0.4 mM CTP, 0.04 mM UTP,20,uCi of [c_-32P]UTP (500 Ci/mmol), and 0.5 to 2.5 ,ug of DNAas chromatin orpurified DNA. Tran-scriptionassaymixtureswereincubated for 60 minat 30°C. DNase I(RNase-free; Worthington Diagnostics) was then addedto afinal concentration of15 ,ug/ml, and the reaction mixturewas incubated for 5 min at 30°C. Sodium dodecyl sulfate (SDS)wasthen addedto afinalconcentrationof 0.2%,and thiswasfollowed by the addition of 25 ,ug of proteinase K per ml and incubation at 30°C for 15 min. The RNA was then purifiedasdescribed by Lavialleetal. (24).
Analysis of RNA transcripts by glyoxalgel
electropho-resis.SV40 DNA fragments whichwereusedas size markers were generated by restriction endonuclease cleavage ofsuperhelical SV40DNA with (i) BamHI,
(ii) HinFI, or (iii) BamHI and HpaII. Anequimolar mixture of the three digests was labeled at the 5' termini with phage T4 polynucleotide kinase and
[y-32P]ATPas described by Maxam and Gilbert (28).
Purified RNAsamples and DNA markerswere dena-turedbyglyoxalasdescribed by McMaster and Carmi-chael (30). Denatured samples wereelectrophoresed
on a 1.5% agarosegel (10 mM phosphate buffer, pH
7.0)at150V(50 mA) for 2.5 h. The gelwasfixed with
1% streptomycin for 30 min at room temperature, dried, and exposedtoX-rayfilm.
S1nucleaseanalysisof RNAtranscripts. SV40DNA
fragmentswerelabeledattheir 5'terminiasdescribed
above. RNAtranscripts and the labeled DNA probe weresuspended in 20 ,ul ofasolution containing0.4 M
NaCl, 0.04 M PIPES
[piperazine-N,N'-bis(2-ethane-sulfonic acid)] (pH 6.4), 1.25 mM EDTA, and 80% formamide and heatdenatured by incubationat85°C
for5min. Thesampleswerethentransferred
immedi-atelyto a50°Cwaterbath and incubated foratleast 3
h. Hybridization reactions were stopped by adding9
volumes of a solution containing 0.25 MNaCI,0.03 M
sodium acetate (pH 4.6), and 1 mM ZnSO4
(pre-equilibratedto4°C). SI nucleasewas addedtoafinal
concentration of 5,000 U/ml,and thereactionsample
was incubated for 45 min at 30°C. Protected DNA fragments were thenpurified and analyzed byglyoxal
gelelectrophoresis.
Analysis of the superhelical structure of chromatin
DNA.A 2.5-,ugsample of SV40 chromatinorpurified
SV40 F, DNAwas incubated with the soluble HeLa
cellextract asdescribedabovefortranscription
reac-tions.After0, 0.5, 5, 15, 30,and60 minofincubation,
portionswereremoved, andthereactionwasstopped
by adding an equal volume of20 mM EDTA-0.4%
SDS. Sampleswerethen treated withproteinaseK(50
,ug/ml) for 1 h at 37°C. The DNA was purified and
analyzed by gelelectrophoresisona1.5%agarosegel
at50Vfor16to20 h witha40 mMTris-acetate(pH
7.2)-20 mM sodium acetate-1 mM sodium EDTA
buffer. The gelwasstained withethidium bromide and
photographed underaUVlightsource. Todistinguish
betweennicked circularandcovalently closedcircular SV40 DNA molecules, DNA samples were
electro-phoresedin the presenceof6 ,ugof ethidium bromide
perml.
Isolation of RNAby using Southern blots.The sepa-rated strandsoftwo SV40 DNA fragmentsgenerated
by cleavagewithBamHIand HpaII restriction
endo-nucleases were electrophoresed and transferred to
nitrocellulosepaperbythemethod of Southern(36), as
previously described (3). Hybridization of
[c-32P]UTP-labeled RNAs was carried out in a buffer
containing 50% formamide, 2 mM EDTA, 0.75 M
NaCl, 0.5% SDS, and 200 p.goftRNA permlin 100
mMTris(pH7.4)for16hat37°C. Hybridization strips
werewashed with10successiverinse solutions (5 ml
each)containing0.2%SDS and2 mM EDTAin 10 mM
Tris (pH 8.0). Nitrocellulose strips were dried and
exposedtoX-rayfilm. Individual bandswereexcised,
and the RNA waselutedby incubation at 85°C for5 min inhybridization washbuffer.
RESULTS
Stability of SV40 chromatin in HeLa cell ex-tract. HeLacell extracts prepared by the
meth-od of Manley et al. (26) contain topoisomerase. This enzyme removes superhelical turns from covalently closed DNA molecules by introduc-ingasingle-strand nick inthe DNA andresealing
thenickafter the torsional tensionof the double
helix is relieved, allowing the release of one superhelical turn each time that the cycle is
repeated. By virtue of the torsional constraints that nucleosomes exert on the DNA duplex,
chromatin structure prevents the loss of super-helical turns (17, 29). Therefore, an analysis of DNAstructureafterincubationwith the extract
provides anassay for the stability ofthe SV40
chromatin complex. Control superhelical SV40 DNA
(F1)
and SV40 chromatin were added tocomplete transcription assayreaction mixtures.
Sampleswereremovedatdifferenttimes,
on November 10, 2019 by guest
http://jvi.asm.org/
774
teinized, and analyzed by agarose gel
electro-phoresis. The action of topoisomerase on F1
SV40DNAwas apparent after 0.5 min of
incu-bation since most of the superhelical DNA had
been converted either to intermediates with a
lower superhelix density or to relaxed
covalent-ly closed DNA molecules and nicked circular DNA(SV40DNA
Fll)
molecules (Fig. 1,lanes Aand B). The latter two forms of SV40 DNA comigrated in this gel system. Further incuba-tion of purified DNA resulted in complete con-version ofF1 DNA to relaxed covalentlyclosed
DNA molecules and nicked circular molecules (Fig. 1, lanes C throughF). In contrast, incuba-tion of chromatin DNA under identical condi-tions revealed that the superhelical DNA was
stable in the presence of the enzyme (Fig. 1,
lanes G through L). After 0.5 and 5 min of
incubation, no significant change in superhelical
DNA structure was detected in the chromatin
samples. After10, 30,and 60 min ofincubation, there was a partial loss of the FI superhelical DNA band, with a concomitant increase in the amountofFII orrelaxed covalentlyclosed circu-lar DNA. To differentiate between F11 and
re-laxed covalently closed circular DNA
mole-cules, duplicate DNA samples were
electrophoresedin an agarosegel containing0.6 mg of ethidium bromide per ml. This analysis demonstratedthat inadditiontoDNAmolecules
with an unaltered superhelix density, the only DNA species present in the chromatin samples was DNA F1l(data not shown).
The stability of SV40 chromatin during incu-bation with HeLa cell extracts was also ana-lyzed by velocity sedimentation and isopycnic
banding in CsCl. [3H]thymidine-labeled SV40 virions were dissociated and incubated with the soluble HeLa cell extract as described above. After incubation, the samplewascompared with control chromatin on isokinetic 10 to 30%
su-crosegradients and CsCl density gradients. The
results of these experiments demonstrate that chromatin structure was preserved since the sedimentation coefficient (110S) and buoyant density (1.40 g/cm3)were notaltered after incu-bation with the extract (data not shown). The fact that no free DNAwas detected in eitherof
these analyses also demonstrates that the
F1l
circular DNA molecules noted above were in a chromatin structure before deproteinization of the DNA forgel analysis.
Transcription of SV40 chromatin in soluble HeLa cell extracts. SV40 virions were
dissociat-1*
II
FIG. 1. Superhelicalstructureof chromatinDNAafter incubation withHeLacellextract.A2.5-,ugsampleof
SV40 chromatinorSV40F,DNA wasincubatedinthe standardtranscriptionmixture,whichincluded allfour
nucleotide triphosphates. After 0, 0.5, 5, 15, 30, and 60 min ofincubation, portions were removed, and the
reactionwasstopped by addinganequalvolume of 20mmEDTA-0.4%SDS.DNAsampleswerethentreated
withproteinaseK(50,ug/ml)for1hat37°C,purified,andanalyzed by gel electrophoresison a1.5% agarosegel
at50Vfor16 to 20 hwitha40 mMTris-acetate (pH 7.2)-20mM sodiumacetate-1 mM EDTA(sodiumsalt)
buffer. Thegelwasstained withethidium bromide andphotographedunderaUVlightsource. LanesAthrough
F,zerotime and0.5-, 5-, 15-, 30-, and 60-minpurifiedDNAincubationsamples,respectively;lanesGthrough L,
zerotime,and0.5-, 5-,15-, 30-,and60-minchromatinDNAincubationsamples,respectively.IandIIindicate the migration positions of SV40 superhelical (F1) and nicked circular (F,,) DNAs, respectively. Relaxed
covalentlyclosed circularDNAmigratedatthesamepositionasF1, DNA.
J. VIROL.
on November 10, 2019 by guest
http://jvi.asm.org/
[image:3.491.85.407.372.584.2]EcoRI* Bam HI
l I
I
AvaI1
Ta Ava 11 Hpa11Hind 11 HaeII* Eco RI
0 I I
1
1
lll1
I Il I
2.0 2.5 3.0 3.5 4.0 4.5 5.0 3.45
EARLY --I
.5 1.0 1.5
1.6
I 1.45
[image:4.491.97.397.76.170.2]LATE
FIG. 2. TranscriptionalmapofSV40 DNA.ThediagramrepresentsanSV40 DNAtemplatewhich hasbeen
digested with endonuclease EcoRI (Fig. 3). The orientations andlengthsof RNAtranscriptsfromearlyandlate
SV40 promoters are shown below the map. The restriction endonucleasesites used fortranscription "run-of'
assays(single asterisks) (Fig. 3 and 4) andS1nucleaseanalysis (doubleasterisks)(Fig. 6)areindicated abovethe
map.The mapcoordinates andRNAtranscriptlengthsareindicated(inkilobases).
ed with EGTA and dithiothreitol as described
previously (4, 5), and the chromatin DNA was
restricted with endonuclease EcoRI (map posi-tion 1,783) (Fig. 2). The restricted chromatin
template was then added to the transcription reaction mixture at different concentrations
ranging from 10to50,ug/ml.Afterincubation at
30°C for 1 h,the RNAtranscriptswerepurified and analyzed by glyoxal gel electrophoresis
(Fig. 3). The majorRNA bandsdetectedin this
analysis were transcripts of approximately 3,450, 1,600, and 1,460 nucleotides. Assuming that transcription terminates at the
endonucle-ase restriction site, the 3,450-nucleotide
tran-script is consistentwithan RNAmolecule
origi-natingneartheearlySV40promoter and
extend-ingto the EcoRl restriction site. Similarly, the 1,600- and 1,450-nucleotide transcripts would
originatefrom late SV40 promoters. The radio-active band migrating at 1,800 nucleotides was not aSV40 transcript, but apparently was gener-ated by radiolabeling of the endogenous RNA present in the extract (26). The lengths of all RNAtranscriptshavebeencorrectedby5.6%to compensate for the different migration rates of RNAand DNA inglyoxal gels.
Thesurprisingaspectof this experimentwas
thataccurateinitiation oftranscription onSV40 chromatin occurred over awiderange of DNA concentrations. This is in directcontrast to the
strict DNA concentrationrequirementfor accu-rateinitiation oftranscriptionon nakedDNA,as
noted by Manley et al. (26) in the original description of the soluble transcription system.
We observed asimilar DNA concentration
de-pendence in our analysis of transcription of purified DNA templates. Specific transcription from SV40 early and late promoters occurred
only over averynarrowrange (20 to30,ug/ml)
when the template was purified DNA. Increas-ingpurified DNAconcentrations to 40 to 50,ug/
mlresultedina decrease inspecifictranscription
and aconcomitant increase innonspecific tran-scription(datanotshown).Transcription of both
DNAand chromatinatconcentrations lessthan
10 ,g/ml reduced transcription below the level of detection.
We havepreviously shown that extracellular SV40 virion chromatin is an efficient template
A B C D E F
5,243
3,036
2,190
- 1,087
768
uI 533
FIG. 3. Transcriptionof SV40 chromatin in HeLa
cell extracts. SV40 virions were dissociated by
treat-mentwith 5 mM EGTA-3 mM dithiothreitol-100 mM
NaCl in 50 mM Tris (pH 8.7) for 30 min at 32°C. After
dissociation, MgCl2was added to a concentration of 6 mM, and the chromatin template was restricted by
incubationat32°C for 1.5 h in the presence of 5 to 10 U
ofEcoRI per ,ug of DNA. [a-32]UTP-labeled RNAs
were synthesized in standard reaction mixtures con-taining 10 ,ug (lane A), 20 Fg (lane B), 30 ,ug (lane C),
40 ,ug (lane D), or 50 ,ug (lane E) of the
EcoRI-restricted chromatin template perml. Purified RNA
transcripts were purified and denatured with glyoxal
asdescribed by McMaster and Carmichael (30).
Dena-turedsamples wereelectrophoresed in a 1.5% agarose
gel in 10 mMphosphate buffer (pH 7.0) at 150 V (50
mA)for 2.5 h. Gels were fixed with 1%streptomycin
(30min,25'C),dried, and exposedtoX-ray film. Lane
F contained a 32P-end-labeled DNA marker. The
lengths of all RNA transcripts have been corrected by
5.6%tocompensatefor the differentmigrationratesof
RNAand DNAmolecules in glyoxalgels.
on November 10, 2019 by guest
http://jvi.asm.org/
[image:4.491.271.422.297.483.2]776 BRADY,
for E. coli RNA polymerase (4, results were observed in a compan
transcriptional efficiencies of chro
SV40DNAin theextractof Manleye
quantitative analysis ofdensitometr [a-32P]UTP-labeled transcript auto
andhybridization data indicated thati, DNA concentrations, the chromatii
was approximately 75% as efficien
DNA(data notshown).
Termination of transcription occur!
tion endonuclease sites. To determir termination of transcriptionoccurrec
tion endonuclease sites, the chromat was restricted with either (i) EcoRI, and BamHI, or (iii) EcoRI and Hi restriction, the templates were ad( transcription reaction mixtureatacoi
of 30 ,xg/ml and incubated at 30°C f
The results of this experiment oriente tion oftranscription and demonstrate scription terminates at therestrictioi 4). Transcription of EcoRI-restricted (Fig. 4, laneB) produced dominantI
A B C D F
FIG. 4. Polarity and termination of ii
chromatin template transcripts. SV40 dissociated as described in thelegend t
restricted with either EcoRI (lane B),
BamHI (lane C), orEcoRI and HaeII
32P]UTP-labeled RNAsweresynthesizec
reaction mixtures, purified, andanalyze
gel electrophoresis. Lanes A and Ecor
markers.Thelengthsof all RNA transcril corrected by 5.6% to compensate for
migrationratesofRNA and DNA molecu
gels.
5). Similar of 3,450, 1,600, and 1,460 nucleotides.
Tran-ison of the scriptionof EcoRI-BamHI-restricted chromatin matin and (Fig. 4, lane C) produced RNA transcripts of tal. (26).A 2,700, 1,600, and 1,460 nucleotides. The specific
ic scans of decrease in the size of the 3,450-nucleotide
radiograms transcript to approximately 2,700 nucleotides atoptimum suggests not only that termination occurred at n template therestriction site but also that this RNA tran-t as SV40 scriptoriginatedat the earlySV40promoterand proceeded counterclockwise toward the EcoRI
satrestric- site. Transcription of the
EcoRI-HaeII-restrict-ne whether ed chromatin generated RNA transcripts of i atrestric- 3,450, 650,and515nucleotides (Fig. 4, laneD). :in complex Thespecificdecrease inlengthof the1,600- and (ii) EcoRI 1,460-nucleotide transcripts suggests that these
aeII. After transcripts originated at late SV40 promoter
ded to the sites andproceeded clockwise toward the EcoRI
ncentration site. An RNA transcript of approximately 850 for 60 min. nucleotides, which did not correspond to any of dthe direc- thetranscripts discussed above, was
reproduci-d that tran- bly observed in the EcoRI-HaeII transcription n site (Fig. samples. Although the origin of this transcript
chromatin was notstudied in detail, a preliminary analysis
RNA bands suggested that it is transcribed from the early SV40 strand and originates near map position
1,680.
Orientation ofRNA transcripts by hybridiza-tion,elution, andglyoxalgelelectrophoresis.
Re-striction of SV40DNAwithHpaII andBamHI,
5f243
followed by strand separation on agarose gels, allows for the separation of the main coding 3036 sequencesforearly
andlate SV40 mRNAs(3).
Hybridization of RNA transcripts synthesized
27
it90 from EcoRI-restricted SV40 chromatin to the 2 early/late Southern blot resulted in the pattern 1,829 shown inFig.5, lane A. RNAhybridizationwaspredominantly to the EA and LB DNA
frag-ments, which correspond to the major coding
0(J87 sequences of early and late mRNAs,
respective-ly. When the RNA that hybridized to the four _7 8. DNA fragments was quantitated, 70 to 80% of
768
the RNA hybridized to the EA and LB DNAfragments. To determine which RNA species
533 hybridizedtothespecificDNAfragments,RNA was eluted from each fragment, purified, and
analyzedby glyoxal gelelectrophoresis.
Togeth-er with datapresented above, the distinct
hy-235 bridizationof the3,450-nucleotideRNA toDNA
fragment EA(Fig. 5, laneC)and the
hybridiza-n vitro SV40 tion of the 1,600- and 1,460-nucleotide RNAs to virions were the LB DNAfragment (Fig. 5, laneE) demon-.o Fig. 3 and stratedthat these RNAscorrespondtoinitiation
EcoRI and of RNAsynthesis at early and lateSV40
promo-(lane D). [a- ters, respectively. The heterogeneous smearof
din standard RNA in Fig. 5, lane C, apparently was due to
d byglyoxal degradation of the 3,450-nucleotide RNA
spe-ts
havebeen cies. Thelight hybridization
of the3,450-nucleo-the
different
tide RNAtoDNAfragment
EB and thehybrid-les inglyoxal ization of the 1,600- and1,460-nucleotide RNAs
to DNA fragment LA were due tooverlaps of
J.VIROL.
on November 10, 2019 by guest
http://jvi.asm.org/
[image:5.491.75.224.329.570.2]ACCURATE SV40 777
AB C D F
EA
--LA
LB
EB
FIG. 5. Hybridization, elution, and
electrophoresis of RNAs synthesized froi
matin. RNAswere synthesized froman
matin reaction mixture and purified. [c
beledRNAswerehybridizedtothesepa
ofaBamHI-HpaIIdigest ofSV40(3)in
buffercontaining 50% formamide, 2 mM
MNaCl, 0.5% SDS, and 200 ,ug of tRNA
mMTris (pH 7.4) for 16 hat37°C. Hybric
werewashed with 10 successive rinsesc
each) containing 0.5% SDS and 2 mM ED
Tris (pH 8.0). The hybridization stripsw
exposedtoX-rayfilm. Radioactive band
ized andcutout,andthe radioactiveRN
by incubationat85°C for 5min in hybri(
buffer. RNA transcripts were then puril
lyzed by glyoxal gel electrophoresis. La
late hybridization (EA, early strand, A fi
latestrand, Afragment; LB, late strand
EB, early strand, B fragment); lanes (
[a-32P]UTP-labeled RNA transcripts h
EA,LA, LB, and EB DNA fragments,
lanes BandG, DNA markers.
RNAsequences atthe3' end ofthe
and the 5'end of thelate RNA. Si nudease analysis of RNA trai
determine thepositionof theinitiatio
RNA transcripts moreprecisely, we
Si nuclease mapping of unlabeled scripts with several DNA fragmentv their5' termini (Fig. 6). Hybridizati
matin RNAtoAvaIl fragmentD lab 5'termini(nucleotides 5,123and 561
the protection ofthree DNAfragm4 230, and 375(lower bandofdoublet)
(Fig. 6a, lanes D and E). The 43( DNAfragmentwasnotareproducib ase fragment and appeared to be tl
minor contamination in the probe.
tentative location of RNA initiatioi above), the 115-nucleotide bandwa:
with anearlyRNAtranscript thato
Gs map position 5,240 ± 10, whereas the 230- and 375-nucleotide bands correspondedtolate RNA
5,243 transcripts thatoriginated atmappositions 325
± 10 and 185 ± 10, respectively. These map 3,036 positionswere confirmed
by performing
a simi-lar S1 nuclease analysis with DNA fragments2,190 labeledatthe
TaqI
site(Fig. 6a,lanes Gthrough_
m.w.1829 L) (mapposition 4,741),theHpaIIsite(Fig. 6b,
lanes A through D) (map position 349), orthe Hindll site (Fig. 6b, lanes E through H) (map
#NO 1,087 position473). Briefly, the TaqI probe protected
two bands of 495 and 410 nucleotides (Fig. 6a,
~EM 768 lanesJ and K) the HpaII probe protectedaband _ 533 of 170 nucleotides (Fig. 6b, lanes C and D), and the HindIIprobe protected bands of 285 and 145 nucleotides (Fig. 6b, lanes G and H). Since the
structure of linearized chromatin might allow
ai 235 recognition of promoter sequences which are
notaccessibleinchromatinthat contains a cova-glyoxal gel lently closed circular DNA, each analysis
in-mSV40 chro- cluded an RNA samplesynthesized from uncut EcoRI chro- SV40chromatin. Our resultsdemonstrated that
X-32P]UTP-1a- specific initiation oftranscription occurred on
rated strands both intact and restricted SV40 chromatin and
hybridization thattheseinitiation sitescorrespondedto
initia-lEDTA,
0.75 tion sites used in vivo (2, 20, 27, 39). It is ofpermlin100 interest that whereas the levels of RNA
synthe-)lutions (5ml sis fromthe late
SV40
promoter werecompara-)TA in 10 mM ble in intact and restricted SV40 chromatin ere dried and templates, RNA synthesis from the early SV40
Is were local- promoter appeared tobe enhanced in the intact
[A was eluted chromatin template. The significance of this
dlization
wash observation ispresently being investigated.fied and
ana-me A, Early/ DISCUSSION
B
fragment;
Our results demonstrate that extracellular Cthrough
F SV40virion chromatin isaccurately
transcribed ybridized to in the in vitrotranscription systemofManleyetrespectively; al. (26). The lengths of RNA transcripts (Fig. 3 and 4), the hybridization patterns (Fig. 5), and
the S1 nuclease mapping data (Fig. 6) demon-early RNA strate that two major early RNA transcripts, which originate at map positions 5,240± 10 and nscripts. To 5,145 ± 10, and two major late transcripts,
nsite of the which originate at map positions 325 ± 10 and e performed 185 ± 10, are synthesized from the chromatin RNA tran- template. These map positions in turn
corre-s labeled at spond to RNA initiation sites which have been ion ofchro- mappedpreviously by either primer extension or yeledatboth Si nuclease analysis (2, 20, 27, 39). In contrast ) resulted in totranscription of purified DNA templates, ac-ents of115, curate transcription of SV40 chromatin is less nucleotides dependent upon the addition of specific DNA
)-nucleotide concentrations to the transcription assay
mix-leSi nucle- ture.
he result of Recent analyses of transcriptionally active From our and inactive eucaryotic genes suggest that one
n sites (see level oftranscriptional regulation isby
modifica-s consistent tion of chromatin structure. The nucleosome
riginated at structureof transcriptionally inactive eucaryotic
on November 10, 2019 by guest
http://jvi.asm.org/
[image:6.491.47.240.78.271.2].i
i
;~~~~~~~~~~~~~~~~~~~~~~~~~~1i-:e
*-e- b_ _~~~oww
J K L
_- 1087
a- 768
as.
.o..).O. Om _m
a
_- t_~~~~440
; en.
... ...No":
533
510 235
am 110
[image:7.491.103.394.93.498.2]83
FIG. 6. Sinucleaseanalysis of in vitro RNA transcripts. UnlabeledRNAtranscriptsweresynthesized from
EcoRI-restrictedorintact chromatin in standardreaction mixtures,exceptthat[a-32P]UTPwasomitted. Purified
RNAtranscripts andalabeledDNAprobewerelyophilized, suspended in20 ,ul of 0.4MNaCl-0.04 M PIPES
(pH6.4)-1.25mM EDTA-80% formamide,and heatdenatured byincubationat85°Cfor 5 min. Hybridization
samples were transferredimmediatelyto a50°Cwaterbath andincubated foratleast 3 h.Hybridization reactions were stopped by adding 9 volumes of 0.25 M NaCI-1 mM ZnSO4-0.03 M sodium acetate (pH 4.6)
pre-equilibratedto4°C. Sinucleasewasaddedto afinal concentration of 5,000U/ml, and the reaction mixturewas
incubated for 45 min at 30°C. Protected DNA fragments were then purified and analyzed by glyoxal
electrophoresis in a 2.0% agarose gel (10 mM phosphate buffer, pH 7.0) at 150 V (50 mA) for2 h. (a) Si
hybridizationwithAvallfragmentD(mappositions 5,123 and 561)orTaqI(mapposition 4,741)probe.LaneA,
AvaIl fragmentDprobe; lane C,AvaIl fragmentDprobehybridization, Si treated; lane D, AvallfragmentD
intact chromatinRNAhybridization,SItreated; lane E,AvaIl fragmentD-EcoRIchromatinRNAhybridization,
Sitreated; lane G, TaqIprobe;laneI, TaqIprobehybridization, SI treated;laneJ,TaqI-intactchromatinRNA
hybridization, S1treated;laneK, TaqI-EcoRIchromatinRNAhybridization,Si treated.LanesB, F,H, andL
containedDNAmarkers.(b)SihybridizationwithHpaII(mapposition 349)orHindll(mapposition473)probe.
LaneA, HpaII probe; lane C, HpaII-intact chromatin RNAhybridization, Si treated; lane D, HpaII-EcoRI
chromatin RNAhybridization, Si treated; lane E,HindIl probe, Si treated;laneG, Hindll-intact chromatin
RNAhybridization, Sitreated;laneH,HindII-EcoRIchromatin RNAhybridization,Si treated.Lanes BandF
containedDNAmarkers.
J. VIROL.
*.i.Mwdw
on November 10, 2019 by guest
http://jvi.asm.org/
genes consists of an octomer of four histones (H2A, H2B, H3, and H4) which are associated with 145 base pairs of DNA. Histone H1 inter-acts with this basic chromatin structure to con-dense the DNA into 25- to 30-nm higher-order chromatin fibers. (For a review of chromatin structure, see reference 29.) By virtue of the nucleosome structure, chromatin DNA is quite resistant to enzymatic activity and is an
ineffi-cient templateforeither the initiationor
elonga-tion reactions of RNA synthesis in vitro (7, 9, 46).Analyses oftranscriptionally active
eucary-otic genes, such as the genes for globin (44), ovalbumin (16), integrated viral genomes (6, 8, 12, 14, 15, 33), and the Drosophila GP-70 heat shock gene (47, 48), have demonstrated that thesegeneticloci reside within chromatinhaving an altered structure, as indicated by DNase I
hypersensitivity.Wuetal.(47,48) demonstrated thisclearly in theircomparisonof the transcrip-tionallyactive and inactive chromatinstructures
of the major heat shock protein gene of
Dro-sophila. When the GP-70 gene was
transcrip-tionally activated by heat shock, disruption in nucleosomal organization was evidenced by a
loss orsmearing of discrete DNAfragments in micrococcal nuclease digests. When the cells recovered from the heat shock and the GP-70 gene returned to the inactive state (47, 48),
micrococcal nuclease DNA ladderpatternswere
restored. Further evidence thattranscriptionally active chromatinis modifiedcomesfromrecent
studies by Weisbrod and Weintraub (45), who demonstrated that "active" nucleosomes may be separated from "inactive" nucleosomes by affinity chromatography with HMG proteins
bound to a solid matrix (45). Although these
experiments demonstrated a cleardifference in
chromatin structure, the basis of nuclease hy-persensitivity and differential binding of HMG proteins to transcriptionally active chromatin remains to be established. The contribution of chromatin modifications which include histone acetylation, loss of histoneH1, and nonhistone protein interaction with chromatin all require furtheranalysistoclarifytheir role in chromatin
transcriptional activation. Chromatin
modifica-tionsmayaffectpromoter selection or function primarily to "open" chromatin toallowentry of thepolymerase at the promoter site and subse-quent elongation of the RNA chain through nucleosomes.
We have recently reported that extracellular SV40 viralchromatin, which is similar to
tran-scriptionallyactiveeucaryoticgenes, has a "re-laxed" chromatin structure which makes the viral DNA highly accessible toenzymatic activi-ties(4, 5). The conversion ofSV40 chromatin to
therelaxedconformation occurs asintracellular
75S chromatin sequentially matures into a 200S
previrion, a 240S nuclearvirion, and finally an extracellular virion (4, 5, 11, 23). These studies demonstrate that the maturation process is
ac-companied by significant changes in chromatin composition, including increased acetylation of histones H3 and H4, the lossof histone H1, and thespecific association of nonhistone viral pro-tein with the chromatin complex. After in vitro dissociation ofnuclear and extracellular SV40 virions, we have shown by comparative enzy-matic analysisthatextracellular virion chroma-tin is nucleasehypersensitive and 16- to 20-fold more efficient as a transcriptional template (4). To determine what effect chromatin
modifica-tions might have on promoter selection, we compared thetranscription of extracellular viri-on chromatin and purified DNA in three
inde-pendentsystemswhich included(i)E. coli RNA
polymerase, (ii)purified polymerase II,and (iii)
HeLa cell extracts (Manley extract). Although
the utilization of specific promoters differed
qualitativelyin eachsystem,wefoundno
signifi-cant difference in chromatin and purifiedDNA
transcription patterns, which suggests that extracellular SV40 virion chromatin structure
regulates neither strand selection nor promoter
recognition bythe RNApolymerase.Our results suggest that SV40 virion chromatin modifica-tions, rather thanregulatingpromoterselection, functionprimarilytoincreasetheaccessibilityof RNApromoter sequencestopolymeraseIIand allow efficientelongationof theRNAchain after theinitiation event.
In contrast, intracellular SV40 chromatin,
which has quite different structural properties
thanextracellularSV40 virionchromatin2 appar-ently has the potentialtoaffect SV40 promoter
selection. Jakobovitsetal. have shown that the initiation specificity of E. coli polymerase is altered when transcription is carried out on
intracellular SV40 chromatin rather than on pu-rified DNA (22). Initiation oftranscription on the viralchromatin occurredpreferentially from the late strandin theregion close to the in vivo latepromoter. Thisobservation isstriking since
no E.coli promoters, which initiatesynthesis of RNA onthe late strand, havebeen mapped in thisregion ofthe genome (24). Precisely within the later promoterregion, approximately 20%of
the intracellularchromatincomplexescontain a
nuclease-sensitivegapped stretch of DNA which is not contained in a typical nucleosome
struc-ture. These results may suggest that in vivo transcription initiates withinthisregion on such gapped molecules.Theobservationthatactively transcribed eucaryotic chromatin also displays nucleosome-deficient stretches ofDNA makes
thismodelanattractive mechanism for promoter selection forsomegenes in eucaryotic cells.
TheSV40genome is divided into two sets of
VOL.44, 1982
on November 10, 2019 by guest
http://jvi.asm.org/
780 BRADY, SALZMAN
geneswhichare temporally regulatedduringthe
infection cycle. Before DNA replication, stable SV40 RNAtranscriptsare synthesized
predomi-nantly from the early gene which codes for the
T-antigens. After expression of the early gene
products and subsequentDNAreplication,
tran-scription occurs predominantly from the late
SV40genes which code for viral structural
pro-teins. Transcriptional analyses basedon in vivo and in vitro studies demonstrate that regulation
oftheSV40earlygeneis under the control of the
Goldberg-Hogness-like sequence TATTTA which is located 25 to 30 nucleotides upstream
fromthe majorRNA cap site and a72-base
pair
tandem DNA repeat sequence locatedmorethan 100nucleotidesupstreamfrom the cap site(1,
2,
18-20, 27). Expression ofSV40early mRNA is
repressed by its own gene product,
large
T-antigen (31, 35, 40). The regulatory sequences
whichcontrol expression ofthe SV40late genes
and the mechanism by which accumulation of late RNA transcripts is suppressed early in the
infection cycle are not clear at present. The
observation that both late and early SV40
pro-moters were efficiently expressed in the invitro
transcription system of Manley et al. suggests
that promoterstrengthcouldnot accountfor the 20- to 100-fold-lower amount of late mRNA
found during the initial stages of infection
(34).
Sinceour transcriptional analyses of extracellu-larvirionchromatindemonstrate thatchromatin modifications do not regulate promoter
selec-tion, it also appears
unlikely
that chromatinstructure can accuratelyaccountfor the
unequal
expression ofearly and late SV40mRNAs
dur-ingtheearlystages ofSV40infection. It appears
more plausible, as
previously
hypothesized
by
Birkenmeieretal. (3),thatinitiation of lateSV40
RNA synthesis occurs in vivo
during
theearly
stages of infection but stable
transcripts
are notproducedduetoattenuation oflate RNA
synthe-sis by cellular proteins or
by specific
degrada-tion of late RNA after
synthesis.
LITERATURECITED
1. Benoist, C.,and P.Chambon.1980. Deletionscoveringthe putativepromoterregionofearlymRNAsof simianvirus 40 do not abolish expression. Proc. Natl. Acad. Sci. U.S.A. 77:3865-3869.
2. Benoist, C., and P. Chambon. 1981. In vivo sequence requirements oftheearlySV40 promoterregion. Nature (London)290:304-310.
3. Birkenmeier, E. H., E. May, and N. P. Salzman. 1977. Characterizationof simian virus 40 tsA58transcriptional intermediatesatrestrictivetemperatures:relationship
be-tween DNA replication and transcription. J. Virol. 22:702-710.
4. Brady,J.,M.Radonovich,C.Lavialle,and N. P.Salzman. 1981. Simian virus 40 maturation: chromatin modifica-tions increase theaccessibilityof viral DNAtonuclease and RNApolymerase.J.Virol. 39:603-611.
5. Brady, J. N., C. Lavialle, and N. P. Salzman. 1980. Efficient transcription ofa compact nucleoprotein
com-plex isolated from purified simian virus 40 virions. J. Virol. 35:371-381.
6. Breindl, M., and R. Jaenisch. 1979. Conformation of Moloney murine leukemiaproviralsequencesin chroma-tinfrom leukemic and non-leukemic cells. Nature
(Lon-don) 277:320-322.
7. Bustin,M. 1978.Binding ofE. coli RNApolymeraseto chromatin subunits. Nucleic Acids Res. 59:925-932. 8. Chae, C., T. Wong,andR.Gadski. 1978. Transcription
processing, and structure of chromatin in SV40-trans-formed cells. Biochem.Biophys.Res.Commun.
83:1518-1524.
9. Cremisi,C.,A.Chestier,C.Dauget,andM.Yaniv. 1977.
Transcription of SV40 nucleoprotein complexesin vitro. Biochem. Biophys. Res.Commun. 78:74-82.
10. Defer,N.,M.Crepin, C. Terrioux, J. Krun,and F.Gros. 1979.Comparisonof non-histoneproteinsassociated with nucleosomes withproteinsreleasedduringlimited DNase
digestions.Nucleic AcidsRes.6:953-966.
11. Fernandez-Munoz, R., M. Coca-Prados,and M.-T. Hsu. 1979.Intracellular formsof simian virus 40nucleoprotein complexes.I. Methods of isolation andcharacterization in CV-1 cells. J. Virol.29:612-623.
12. Flint, S. J., and H. M. Weintraub. 1977. An altered subunitconfigurationassociated withactively transcribed
DNAofintegratedadenovirus genes. Cell 12:783-794. 13. Frolova, E., and E. Zalmanzon. 1978. Transcription of
viralsequencesin cells transformedby adenovirus type5. Virology89:347-359.
14. Frolova, E.,E.Zalmanzon,E.Lukanidin,andG.Georgiv. 1978.Studies oftranscriptionof viral genome in adenovi-rus5transformed cells.Nucleic Acids Res. 5:1-11. 15. Garber, E. A., and M. M. Seidman, and A. J.Levine.
1978. The detection and characterization of multiple
forms of SV40 nucleoprotein complexes. Virology
90:305-316.
16. Garel, A., and R. Axel. 1976. Selective digestion of
transcriptionally active ovalbumin genes from oviduct nuclei. Proc.Natl. Acad.Sci. U.S.A. 73:3966-3970. 17. Germond, J.,B.Hirt,P.Oudet,M.Gross-BelHard,and P.
Chambon. 1975. Folding of the DNA double helix in chromatin-like structures from simian virus 40. Proc. Natl. Acad.Sci. U.S.A. 72:1843-1847.
18. Ghosh, P., P. Lebowitz, R. Frisque, and Y. Gluzman. 1981. Identification ofapromoter componentinvolved in
positioningthe5'terminiof simian virus 40earlymRNAs. Proc.NatI.Acad. Sci. U.S.A. 78:100-104.
19. Gruss, P.,R.Dhar,andG.Khoury. 1981. Simian virus 40 tandem repeated sequences as an element of the early
promoter. Proc.Natl. Acad.Sci. U.S.A. 78:943-947. 20. Handa, H., R. Kaufman, J. Manley,M. Gefter, andP.
Sharp. 1981. Transcriptionof simian virus40DNA ina
HeLawhole cell extract.J.Biol.Chem. 256:478-482. 21. Hu, S.,andJ. Manley. 1981. DNAsequencerequiredfor
initiation of transcription in vitro from the major late promoter of adenovirus2. Proc.Natl.Acad. Sci. U.S.A. 78:820-824.
22. Jakobovits, E.,S.Saragosti,M.Yaniv,and Y.Aloni.1980. EscherichiacoliRNApolymerasein vitromimics simian virus40 in vivotranscription whenthetemplateisviral nucleoprotein. Proc. Natl. Acad. Sci. U.S.A. 77:6556-6560.
23. LaBella, F., andC. Vesco. 1980. Late modifications of simian virus40chromatinduringthelyticcycleoccurin
animmature form of virion.J.Virol.33:1138-1150. 24. Lavialie,C.,Y.Reuveni,M.Thoren,and N. P.Salzman.
1982. Molecularinteractionbetweensimian virus40 DNA and Escherichia coli RNA polymerase. J. Biol. Chem. 257:1549-1557.
25. Levy,B., W.Connor, andG. Dixon. 1979. Asubset of
trout testis nucleosomes enriched in transcribed DNA sequencescontainshighmobilitygroupproteinsasmajor structuralcomponents.J. Biol.Chem. 254:608-620. 26. Manley, J.,A. Fire,A. Cano,P.Sharp,andM.Gefter.
1980.DNAdependenttranscriptionofadenovirusgenes
on November 10, 2019 by guest
http://jvi.asm.org/
ACCURATE TRANSCRIPTION OF SV40 CHROMATIN
in a soluble whole cell extract. Proc. Natl. Acad. Sci. U.S.A. 77:3855-3859.
27. Mathis, D., and P. Chambon. 1981. TheSV40 earlyregion TATA box isrequired for accurate in vitro initiation of transcription. Nature (London)290:310-315.
28. Maxam, A. M., and W.Gilbert. 1977.A newmethodfor sequencing DNA. Proc. Natl. Acad. Sci. U.S.A. 74:560-664.
29. McGhee,J., andG.Felsenfeld.1980.Nucleosome struc-ture.Annu.Rev.Biochem.49:1115-1156.
30. McMaster, G., and G. Carmchael. 1977. Analysis of
single-and double-stranded nucleic acids on
polyacryl-amide andagarose gels by using glyoxal and acridine orange. Proc.Natl.Acad.Sci.U.S.A. 74:4835-4838. 31. Myers,R., D. Rio, A. Robbins, and R.TUan.1981. SV40
geneexpression ismodulated by thecooperative binding
of Tantigen to DNA. Cell 25:373-384.
32. Nelson, D., M. Perry, and R.ChaLldey.1979. Acorrelation
between nucleosome spacer region susceptibility to DNAse I and histone acetylation. Nucleic Acids Res. 6:561-574.
33. Panet, A., and H. Cedar. 1977. Selectivedegration of
integratedmurine leukemiaproviralDNAby
deoxyribo-nucleases.Cell11:933-940.
34. Parker, B. A., and G. R. Stark. 1979.Regulation ofsimian virus 40 transcription: sensitive analysis of the RNA
speciespresentearly ininfections byvirusorviral DNA. J.Virol. 31:360-369.
35. Rio, D., A. Robbins, R. Myers, and R. Than. 1980.
Regulation of simian virus 40 early transcription bya purifiedTantigen.Proc. Nati. Acad. Sci. U.S.A. 77:5706-5710.
36. Southern,E.1975.Deletion ofspecificsequences among DNAfragmentsseparated by gelelectrophoresis.J. Mol.
Biol.98:503-517.
37. Stadler, J., T.Seebeck, and R. Braun. 1978.Degradation oftheribosomalgenesbyDNase I inPhysarum polyce-phalum.Eur. J.Biochem.90:391-395.
38. Talkington,C.,Y.Nishioka,and P. Leder.1980. In vitro
transcriptionof normal, mutant and truncated mouse
a-globingenes. Proc. Natl. Acad.Sci. U.S.A.77:7132-7136. 39. Thompson, J. A., M. F. Radonovich, and N. P.Salzman. 1979. Characterization of the 5'-terminal structure of simianvirus 40 early mRNA's. J. Virol. 31:437-446.
40. Tijan, R. 1978. The binding site onSV40DNAfor a T
antigen-related protein.Cell13:165-180.
41. Wasylyk, B., R.Derbyshire, A. Guy, D. Molko, A. Roget, R. Teoule, and P.Chambon. 1980. Specific in vitro tran-scription of conalbumin gene is drastically decreasedby
single-point mutation in T-A-T-A box homology se-quence. Proc.Natl. Acad. Sci. U.S.A. 77:7024-7028.
42. Wasylylk,B., C.Kedinger, J. Corden, B. Brison, and P.
Chambon. 1980.Specific in vitro initiation of transcription on conalbumin and ovalbumin genes and comparisonwith adenovirus-2 early and late genes. Nature (London) 285:367-373.
43. Weil,P., D.Luse, J. Segall, and R. Roeder. 1979. Selective and accurateinitiation of transcription at the Ad2 major late promoter in a soluble system dependent on purified RNA polymerase II and DNA. Cell 18:469-484. 44. Weintraub, H., and M. Groudine. 1976. Chromosomal
subunits in active genes have an altered conformation. Globin genes aredigested bydeoxyribonuclease I in red blood cell nuclei but not in fibroblast nuclei. Science 193:848-856.
45. Weisbrod, S., and H. Weintraub. 1981.Isolation of active-ly transcribed nucleosomes using immobilized HMG 14 and 17 andanalysis of a-6 chromatin. Cell 23:391-400. 46. WilHamson,P., and G. Felsenfeld. 1978.Transcriptionof
histone-coveredT7DNA byEscherichia coli RNA poly-merase.Biochemistry17:5695-5704.
47. Wu, C., P.Bingham,K.Livak, R.Holmgren,and S.Elgin.
1979. The chromatin structure ofspecific genes. I. Evi-dence forhigher order domains of defined DNA sequence. Cell16:797-806.
48. Wu, C., Y. Wong, and S. Elgin. 1979. The chromatin structure ofspecific genes. II. Disruption of chromatin structureduringgeneactivity.Cell 16:807-814.
1982
on November 10, 2019 by guest
http://jvi.asm.org/