JOURNAI OF VIROLOGY, June 1992, p. 3476-3484 0022-538X/92/063476-09$02.00/0
Copyright ©D 1992,American Society for Microbiology
In
Vitro Transactivation
of Baculovirus
Early
Genes by Nuclear
Extracts
from
Autographa californica
Nuclear Polyhedrosis
Virus-Infected
Spodoptera
frugiperda
Cellst
BARBARA GLOCKER, RICHARD R. HOOPES,JR., AND GEORGE F. ROHRMANN*
Department of Agricultural Chemistry, Oregon State University, Corvallis, Oregon 97331-6502
Received 16December 1991/Accepted 16February 1992
Nuclear extracts, prepared from Autographa californica nuclear polyhedrosis virus-infected Spodoptera
frugiperda cells duringatimecourseofinfection,wereanalyzed for activation of earlygenetranscriptionand
forlate gene transcription. The templates used in the in vitro transcription assays contained promoters for
baculovirus genesthat have been classified as immediate early,delayed early,and late. Thepromoters were
derived from thebaculovirus 39K, p26, gp64,and DNA polymerasegenes.In addition,the adenovirus major
late promoter was included in these studies. We found that transcription from promoters classified as
immediate earlyordelayed earlywasaccuratelyinitiated by usingextractsfromuninfected cells.Furthermore, transcription from all earlypromoters testedwasfound tobe transactivated by nuclearextracts prepared at
4and8 hpostinfection. However, baculovirus enhancer-dependent transcriptionalactivationwasnotobserved intestswith templates containing the hr5 enhancersequence.Transcription from baculovirus late promoters
wasalsonotobserved. A decline in transcriptionby nuclearextractsprepared from cells late in infectionwas
associated withthe presence ofDNase activity. TheBaculoviridae are a family of insect viruses
charac-terized byacomplexreplication cycle that culminates in the
occlusion ofvirions in a crystalline protein matrix (3). The
Autographa califonica multicapsid nuclear polyhedrosis virus(AcMNPV) is themostwell-characterizedbaculovirus and has a double-stranded, circular, supercoiled DNA ge-nome of approximately 128 kbp. Progression through the
AcMNPVinfection cycle is governed bya cascade ofearly,
late, andverylategenetranscription (7). Earlytranscription
begins before the initiation of replication of the viralgenome
andis inhibitedbyo-amanitin, consistent with it being RNA polymerase II dependent. Investigations with reportergene
constructs transfected into insect cellssuggest that the level oftranscription from earlygene promoters ismodulated by viral transactivating factors (4, 5, 10, 23) and enhancer
sequences(12, 18, 22). Afterinitiationof viral DNA replica-tion, lategeneexpression isinitiated and thetranscriptionof
some host nuclear genes is repressed (19). Very late in infection, two genes involved in occlusion body formation (polyhedrin and plO) arehyperexpressed (for a review, see
reference 3). The transcriptionoflategenes isdependenton
thepresenceofana-amanitin-resistantRNApolymerase(8, 9, 16) having a unique subunit composition that is differcnt
from those of the three host RNApolymerases (26). Althoughoneof themostintriguingaspects of baculovirus replication concerns the control of the viral transcription cascade, fewofthe hostandviral factorsparticipatinginthe regulation ofbaculovirusgene expression have been
identi-fied.Thedevelopmentof in vitrosystemsthatreflectin vivo
gene transcription is crucial for the identification and analy-sisof these factors. We have recently reported that
baculo-virus genes classified as immediate early are accurately
initiated by nuclear extracts from uninfected insect
(Spodoptera frugiperda [Sf9]) and human (Namalwa) cells
* Correspondingauthor.
tTechnical report 98t)7 from theOregon State University
Agri-cultural Experiment Station.
(15). In these studies, we concluded that host cell RNA
polymerase II and its associated factors were the only
components required for basal transcription from
baculovi-rus immediate-early promoters. Wc have now cxtended
these initial studies to cxamine in vitrotranscription, using nuclear extracts prepared from both uninfected and AcM NPV-infected Sf9 cells. Uninfected-cell nuclear extracts
were tested for the ability to support the transcription of
genes categorized as immediate early or delayed early,
whereas nuclearextractsfrom infected cellswere examined
forthe transactivation ofearlygenetranscription and for latc
promoter-dependent transcription. We also tested for
bacu-lovirus hrS enhancer sequence-dependent transcription in vitro. Finally,wecomparedthe relativelevels of in vivo and
in vitro transcription from a specific promoter at various
times afterinfection.
MATERIALS AND METHODS
Construction oftemplates. The templates constructed for this study are described below and are shown in Fig. 1. These templates were derived from the genomcs of AcM NPV, Orgyiapseludotsutgata multicapsidnuclcar polyhcdro-sisvirus (OpMNPV), and adenovirus type 2. The sequence
ofthe promoter and mRNA initiationregions of thesegenes
is shown inFig. 5.
(i) AcMNPV 39K gene promoter (p39K). Plasmid p39K containsa1-kbPstI-SstIfragmentfromthe AcMNPVPstI K fragment (10) cloned into thecorresponding restriction sitcs ofpBlucscript (pKS-; Stratagcnc Cloning Systcms).
(ii) 39K gene promoter plus enhancer hr5 (pHR39K). A plasmid, pBG8, containing hi-5wasconstructcd bydigcstion
of the AcMNPV HinidIlI 0 fragmcnt (17) with MluI. The
484-bp fragment containing lit-5was isolated, the ends wcrc
filled with Klcnowpolymerase,and the fragmcntwascloncd
into the StinaI site of pBluescribe (pBS; Strataigene). To characterize the lr-5 insert, it wals sequcnecd and found to
have a single nucleotide difference (nuclecotide
Intl
-525, T3476
Vol. 66, No. 6
on November 9, 2019 by guest
http://jvi.asm.org/
A) p39k
pHR39k
Ad
hr5 i Kpn.I HixtII
B) p64CAT-319
EcoRI
C) pAML
D) pOpp26
E) pAcHRp26
Hlnd!!!
F) pDNAP
Ec I
200bp
FIG. 1. Schematic representation of the p
vitro transcription. The arrowsindicate the ru duced fromthe RNA start sitewhen thetemplal
indicatedrestrictionenzyme.Thenumbersindi
sizes in nucleotides (nt). The derivation of tk
scribed inMaterials andMethods.
replacedwithC)from the hr5 sequencerep (17). ThekIpnl-HindIII fragmentofplasn ing thehr5 enhancerregionwascloned inl ing restriction sites ofplasmid p39Ktocr
(iii) AcMNPVp26gene promoterwith hi plasmid containing theAcAfNPV p26pror constructed by cloning a 1.5-kb HindIII AcMNPV HindIII-Q (17) into pKS-. Pla
contains the 1.3-kbHindIII-SalI fragmen
smaller portion of the p26 open readin,
constructedforuse inprimerextensiona] (iv) AcMNPV DNA gene polymerase pr An 800-bp AcMNPV EcoRI V fragmet promoter region of the DNA polymera.
subcloned inpKS-.
(v) OpMNPV gp64 gene promoter (pl plasmidcontains 319 bp of the5' untrans gp64 cloned into pBS- upstream of a
acetyltransferase (CAT)geneandis descri
(vi) OpMNPV p26 gene promoter (pOpp
contains thep26genewith 363bpofthe 5' cloned intopBS-andwasprovided by C:
(vii) Adenovirus major late gene promol
midpMLH1, provided byD.Hawley (14),
PstI and SstI, and the fragment containii
majorlatepromoter (260 bp5' and 536bp3
site)wascloned intopBluescript(pKS-)
-Cells, virus, and nuclear extract prep
(ATCC CRL 1711), obtained from GIBI search Laboratories, were grown in seru
dium (GIBCO-Bethesda Research Lab(
214nt mentedwith penicillin G (50 U/ml)-streptomycin
(50
pg/ml)
- (Whittaker Bioproducts)-amphotericin B (Fungizone, 375
ssi
ng/ml;
Flow Laboratories) in shaking flasks (135 rpm) at 214 nt 27°C. Cells were grown to a density of 2.0 x 106 cells/ml and X > infected withAcMNPV(E-2 strain, a gift from L. Volkman) at amultiplicityofinfection of 10. The time point defined as zerohours postinfection (h p.i.) represents the timeimme-316nt _ diately before the addition of virus to the cells. During the first hour ofinfection, the cells were not shaken. Nuclear EwRI extractswere prepared asdescribed by Hoopes and Rohr-5%4nt mann (15).
In vitrotranscription reactions. Invitro transcription reac-tionsemploying uninfected- or infected-cell nuclearextracts were carried out as previously described (15), with minor 143nt changes. Determination of the protein concentration (Bio-r -EZ Rad Protein Assay kit) ensured thatequivalent amounts of
Safl
HindIII
nuclear extract were included in the reactions. Reaction378nt mixtures (final volume, 20 ,ul) contained nuclear extract (3
Tr5 mg/ml), DNA template (30 ,ug/ml), and MgCI2 (6 mM).
-s,gn
W Transcription was initiated after 25 min of preincubation(30°C) by the addition of nucleotides (600 ,uM
[each]
ATP,218 nt CTP, and GTP; 25 p.M UTP [Pharmacia]; 5
,uCi
of[a-32P]UTP [DuPont, NEN Research Products]), followed
xioI by incubation for 25 to 30 min at 30°C. The reaction was
stopped bytheaddition of 30p.lofH20-50p.1 of stop buffer. The RNAwas extractedoncewithwater-saturated
phenol-ilasmids used for in chloroform and
precipitated
with ethanol. The labeled RNA noff transcriptspro- products weresubjected
togel
electrophoresis
on 7 Mteistruncatedbythe urea-5%
acrylamide gels
for 35 min on a Bio-Radminigel
catepredictedrunoff apparatus at 175 V. Relative levels of runoff transcription tese plasmids is de- products were determined by laser densitometry (GS300; Hoefer Scientific Instruments) of autoradiographs. Results from thesemeasurementswereconfirmed by excising the gel bands anddirectly countingthe [32P]UTP incorporated.
Isolation of in vivo RNA. During the time course of an )ortedbyLiuetal. infection,cell samples (7.5ml)wereremoved from shaking aid pBG8contain- flasks,pelleted, quick frozen in liquid nitrogen, and stored at tothecorrespond- -80°C. RNA was isolated by a slight modification of the
*eate
pHR39K.
procedure ofSambrooket al. (21). Three hundred microli-r5(pAcHRp26). A ters of RNA extraction buffer (100 mM NaCl, 10 mM moterandhr5was Tris-HCI [pH 8.1], 1 mM EDTA, 5 mM MgCI2, 0.5%[-SstI fragment
of NonidetP-40, 1 mMdithiothreitol; 4°C)was added directly smidpAcHRp26A
tofrozen cellpellets,and thesuspensionwasvortexed for 30 t(Fig. 1E)
with a s toresuspendthe cellsandlyse
the membranes.Cellswere g frame and was centrifuged for 2 minat4°Cin a microcentrifuge(12,000 xnalysis.
g).Tothe supernatant, 17.5 p.1 ofamixture containing 0.5M romoter (pDNAP). EDTA,8.5 ,ul of20% sodiumdodecyl sulfate, and proteinase nt containing the K(100 ,ug/,I final concentration) was added and incubated se gene(24)
was for 30 minat37°C.Thesampleswereextractedwithanequal volume of hot phenol (65°C), followed by extraction with64CAT-319). This phenol-chloroform and then chloroform-isoamyl alcohol. lated sequence of Finally, the RNAwas precipitatedwith ethanol and
resus-chloramphenicol
pended in 30to60 p.1 of TE(10mMTris-HCI,1mMEDTA,bed elsewhere
(4).
pH 8.0).,26).
Thisplasmid
Primer extensionanalysisand DNAsequencing.Forprimerflanking region (1)
extensionanalysis
of invitrotranscription
startsites,
RNA 'hristian Gross. was prepared as described above, except that no labeled ter (pAML). Plas- UTPwasincluded in the reactionmixture,the UTP concen-wasdigested
with trationwasraisedto200 ,uM,and thepelletwasdissolved in ng the adenovirus TE(10mMTris-HCl, 1 mMEDTA, pH8.0). Primer exten-Vof the RNAstart sionanalysis
wasdonebyusingforward andreverseprimers andcalledpAML.
(5'-GTAAAACGACGGCCAGT-3' and 5'-TCACACAGGA iaration. Sf9 cells AACAGCTATGAC-3', respectively) complementary to se-CO-Bethesda Re- quencesof the vector,aspreviously
described(15).tm-free
Sf900 me- Toconfirm that the in vitrotranscriptionstartsitescorre-)ratories) supple-
sponded
to the in vivo startsites,
RNA isolated fromon November 9, 2019 by guest
http://jvi.asm.org/
[image:2.612.57.298.73.339.2]3478 GLOCKER ET AL.
AcMNPV-infected Sf9 cells was purified as described above and analyzed by primer extension. The primers used for start site determination were 39K, 5'-AACCTTTGAAACAAC CCG-3' (nt -15 to -32 upstream of the ATG [10]); DNA polymerase, 5'-GCATATTCTGCAAAGCGC-3' (nt +50 to +33 downstream of the ATG [24]); and AcMNPV p26, 5'-TGTTGGATCAATTGCATA-3' (nt +39 to +22 down-streamof the ATG[17]). Primer extension analysis of in vivo RNA was performed as described above, except that after precipitation with ethanol, the pellets were resuspended in 4 ,l of 0.1 N NaOH-1 mM EDTA and incubated for 25 min at 30°C to hydrolyze the RNA. To these samples, 4
p,l
of 98% formamide-2% xylene cyanol-bromophenol blue was added and the samples were analyzed on DNA sequencing gels. DNA sequencing was performed by using a Sequenase sequencing kit (United States Biochemical Corp.), according to the manufacturer's instructions.Tests for the presence of nucleases. To test for RNase activity in the nuclear extracts, radioactively labeled RNA wasmade in a standard transcription reaction mixture using an uninfected-cell nuclear extract with SstI-digested pHR39K (Fig. 1A) as template DNA. After the normal incubation time, the reaction mixture was divided into four
aliquots and treated as follows. One aliquot was stopped immediately by the addition of stop buffer. To the second aliquot, RNase A was added to a final concentration of 1
,ug/,uland incubated for a further 10 min at 30°C. To the third aliquot, uninfected extract was added at a concentration of
30%oftotal reaction volume, and the reaction was incubated for a further 30 min at 30°C. This reaction tested for the presence of RNase activity in the uninfected-cell nuclear extract. The fourth aliquot was used to test for RNase activity in late infected-cell nuclear extracts. This was done by mixing it with a nuclear extract prepared at 34 h p.i. at a finalconcentration of30% of the total reaction volume and
incubating for 30 min at 30°C. All samples were then
processed like standard in vitro transcription reactions and analyzed by polyacrylamide gel electrophoresis and
autora-diography.
To test for DNase activity, mock transcription reactions were prepared in which SstI-digested pHR39K template DNA (30 ,ug/ml) was incubated with nuclear extracts (3
mg/ml) produced at various times postinfection. Two reac-tions for each extract sample were prepared, and EDTA was added at a final concentration of 10 mM to one of the samples. After incubation for 60 min at 30°C, the DNA templates were purified by extraction with phenol and
pre-cipitationwith ethanol. Samples were loaded onto an agar-osegel in sample buffer containing RNase A (100
p,g/ml)
to remove RNApresent in nuclearextracts. Afterelectropho-resis, the DNA was stained with ethidium bromide and
photographed on a UVtransilluminator. RESULTS
Rationale fortemplates used for in vitro transcription. The purpose of this investigation was to determine whether the transcriptional events observed during the baculovirus infec-tioncycle were mirrored by the in vitro transcription system. The transcriptional processes investigated and the gene promoters included in these studies are listed as follows. (i) Gene class: the genes selected are examples of promoters termed in the literature as immediate early, delayed early, early, and late. (ii) Promoter structure: the AcMNPV DNA
polymerase gene was investigated because unlike several other baculovirus early promoters it lacks a conventional
TATA promoter element (see Fig. 5). (iii) Transactivation: theAcMNPV39KandOpMNPVgp64 geneswere employed
because they both have been shown to be transactivated in vivo (4, 11). (iv) Enhancement by hr sequences: the AcM NPV 39K gene was used as a template because it is en-hanced by hr sequences (12). The AcMNPV p26 gene was investigated because it isnext to anenhancer sequence(hr5)
in its native location (17). The OpMNPV p26 gene(1) has a sequencesimilar to that of theAcMNPVp26 gene but lacks an adjacent enhancer sequence and would allowexamination of the influence of enhancer proximity for comparison tothe AcMNPV p26 gene. (v) A non-baculovirus promoter: the adenovirus major late promoter wasemployed because it is commonly used for investigations of transcription in higher eukaryotesand would allow comparison of events occurring in baculovirus-infected cell extracts to better-characterized systems.
In vitro transcription by nuclear extracts from uninfected
Sf9 cells. Previously, it was demonstrated that plasmids containing promoters from the OpMNPV gp64, AcAMNPV
IE-1 (a gene that encodes a transactivating factor) (13), and the adenovirus major late genes are transcribed by nuclear extracts from uninfected Sf9cells (15). In these assays they behaved as immediate-early genes. In order to characterize the transcription of early promoters in vitro, the ability of nuclear extracts from uninfected Sf9 cells to transcribe a variety of other baculovirus genes was compared with in vitro transcription of the gp64 and adenovirus major late promoter constructs previously investigated. The constructs used in these investigations included 5' regulatory sequences of genes reported to be (i) immediate early (the gp64 pro-moter p64CAT-319) (15); (ii) early (the DNA polymerase promoter pDNAP) (24); (iii) delayed early (the 39K promot-ers p39K and pHR39K) (11); the AcMNPV p26 promoter (16) pAcHRp26 and theOpMNPV p26 promoter pOpp26 (1); and (iv) a commonly used RNA polymerase II promoter (the adenovirus major late promoter pAML). For constructs see Fig. 1. All these templates gave runoff transcripts of the expected sizes when used in the in vitro transcription system with uninfected nuclear extracts (Fig. 2A through F, lanes 1). Primer extension analyses confirmed that the transcripts initiated at the same position both in vitro and in vivo (see below). Whereas the p64CAT-319, pHR39K, and pAML promoter templates produced strong signals, runoff tran-scripts from pDNAP, pAcHRp26, and pOpp26 were de-tected at a lower level (Fig. 2A through F, lanes 1). These experiments indicate that nuclear extracts from uninfected Sf9 cells accurately initiate and elongate RNA from the promoters of these genes. This suggests that basal levels of in vitro transcription from these promoters are dependent solely on host enzymes and transcription factors.
Transactivation of transcription by nuclear extracts from AcMNPV-infected Sf9 cells. In vitro transcription of tem-plates was examined by using Sf9 nuclear extracts prepared at 0 (uninfected), 4, 8, 12, 16, 24, and 34 h p.i. (Fig. 2A through F). Although the signal produced with uninfected-cell nuclear extracts was relatively weak, there was a marked increase in the intensity of the runoff transcript from most templates employing nuclear extracts prepared at 4 and 8 h p.i.(Fig. 2Athrough F, compare lanes 2 and 3 with lanes 1). The ratios of the intensity of the runoff signal produced by 8-h p.i. extracts to that produced by uninfected-cell nuclearextracts for each template are as follows: pHR39K, 8;pDNAP, 4.7; pOpp26, 3.3; p64CAT-319, 2.7; pAML, 1.3; andpAcHRp26, 1.2. This activation decreased when nuclear extracts prepared after 12 and 16 h p.i. were employed (Fig. J. VIROL.
on November 9, 2019 by guest
http://jvi.asm.org/
B
603
-- 316nt 310- -U" _gp
234 -194
-118
-0
4- 21 43
0 4 8 12 16 24 34
hourspostinfection
1 2 3 4 5 6 7
603
-..raw.
D
-- 544nt 310
-234 -194
-0 4 8 12 16 24 34
hourspostinfection
1 2 3 4 5
603
-310 -234 -194
-- 143nt
118- 118
-6
21
Iii
3
62
0 4 8 12 16 24 34
hourspostinfection
'uilii
0 hours post4 8 infection12 16F
1 2 3 4 5
9
q
W
'.
p
.F
--- 378nt
1 2 3 4 5
603-
I9
310
-234
-194
-- 218nt
118
-t
0 4 8 12 16 0 4 8 12 16
hourspost infection hourspostinfection
FIG. 2. In vitrotranscriptionoftemplates,usingAcMNPV-infectedSf9cell nuclearextractspreparedatdifferenttimespostinfection.The transcription productsshown in thepanelswereobtainedbyusingthefollowingtemplateDNAs:SstI-digestedpHR39K(A),
EcoRI-digested
p64CAT-319(B),SstI-digested pAML(C),HindIll-digestedpOpp26(D), SstI-digestedpAcHRp26(E),and
XhoI-digested
pDNAP
(F).
The numbersontheleft indicate thepositionsof selected radiolabeledHaeIII-digested +X174DNAfragments. Thepositions
and sizes of the runofftranscription productsaremarked witharrows. Below eachautoradiogramisagraphicrepresentationofadensitometry
scanof therunofftranscript signal, indicatingrelative levels oftranscription. O.D. is the relativeabsorptionoftherunoff
transcript.
3479
603
-310 -234 -194
-118
-- 214nt
6t
I
U2
ill
I
C
E
603
-310 -234 -194
-118
-64
¶111.1
on November 9, 2019 by guest
http://jvi.asm.org/
[image:4.612.61.554.59.678.2]3480 GLOCKER ET AL.
infected uninfected
214nt
1 2 3 4
FIG. 3. Lack of in vitro enhancement by the hr5 enhancer
sequence. Lanes 1 and 3 (+) contain Sstl-digested pHR39Kwith
intact hr5astemplate; lanes 2 and4(-) contain thesametemplate aslanes 1and3exceptitwasalsodigestedwithEcoRI todestroy
hr5.Transcriptsinlanes 3 and 4wereproducedfromuninfected-cell
nuclearextracts,andtranscriptsinlanes1 and 2wereproduced by using nuclearextracts preparedat8 hp.i. The position and sizeof
theexpectedrunoffproductis marked withan arrow.(Note:in these
reactions, the DNA concentrations used [45 ,ug/ml] account for
levels of activation thataredifferentfrom those shown inFig. 2A).
Foradescription of the graph,seethelegendforFig.2.
2Athrough F,lanes 4 and5). In vitrotranscriptionwasalso examined forIE-1 by usingatemplatedescribedpreviously
(15).We found thatIE-1showedanactivationprofilesimilar
to those of the other templates (Fig. 2), being activated maximallyat8 hp.i. andshowingabout 10-fold stimulation (datanot shown).
No transcription was observed from late promoters. The
pHR39Kandp64CAT-319 templatescontain both earlyand latepromoter elements thatare utilized in vivo (2, 10) (see Fig. 7). There was no evidence for specific transcription
fromlatepromotersbyextractsfrom late-infected cells(Fig. 2A and B).
Presence of the baculovirus enhancer sequence hr5 on
templatesdid notincrease levels of in vitrotranscription.The AcMNPV genome has several homologous repeated (hr) sequences (6) that contain palindromes centered around EcoRI sites. Fusion of hrsequencestoa39K-CATconstruct
wasshowntoresultinenhanced levelsofCAT activitywhen theconstructwastransfected into cellsalongwithaplasmid
containing IE-1 (12). Nuclear extracts from Sf9 cells were
tested for enhancer sequence-dependent activation of in vitrotranscription (Fig. 3) by usingseveralconstructs. Inthe first tests, a plasmid containing the hr5 enhancer element
upstream of the AcMNPV 39K gene promoter was used
(Fig. 1A, pHR39K). Digestion of pHR39K with SstI, which leaves the hr5 region intact and located in cis to the 39K promoter region, resulted in a runoff transcript of 214 nt,
usingnuclearextractspreparedfrom uninfected cells(Fig. 3, lane3). Doubledigestionofthesame plasmid withSstI and
EcoRI, which destroyed the enhancer element but left the promoter intact, failed to reduce the level of transcription (Fig. 3,lane 4). Similar resultswereobtained when nuclear
extracts produced from cells at 8 h p.i. were tested with
these templates (Fig. 3, lanes 1 and 2). A repeat of this experimentwithtwodifferenttemplates, p39K (without hr5; Fig. 1A) and pHR39K (containing hr5; Fig. 1A), gave the samenegative result (data notshown). Furthertests witha
plasmid comparable to theCATconstruct
reported
togive
maximal levels of enhancement by Guarino and Summers(12) (this plasmid, calledp39CAT-Q-, includes theHindlIl Q fragment
containing
thehr5 enhancer sequence inserted upstream of the 39K promoter region) also showed no in vitrotranscriptional activation byhr5(data
notshown).
Toinvestigateenhancementbyagenethatnormallycontainsan
adjacent hr sequence, transcription of the AcMNPV p26
gene(which is locatedimmediatelydownstream ofhr5)was examined. However, this gene could not be tested for hr5-mediatedenhancement becausewefound that the RNA start site was located 17 nt upstream of the published location. As a result, the promoter overlapped the hr se-quence (see below and Fig. 4C) and EcoRI digestion to eliminate thehr sequence destroysthe AcMNPVp26 gene promoterregion. Therefore, the OpMNPVp26gene, which ishighly homologous totheAcMNPVp26 gene (1)but does not contain an adjacent enhancer sequence in its native state, wasused. The hr5 sequence wascloned upstream of the promoterregion,usingtheSall site of theOpMNPV p26
geneindicatedinFig. 1D. No evidence of enhancementwas obtained by testing pOpp26 with and without the hr5 en-hancer sequence(resultsnotshown). Therefore, this system doesnotappeartodemonstrateenhancer-dependent activa-tion of geneexpression.
Primer extension analysis of in vitro and in vivo RNA. Primer extension analysis was used to confirm that the in vitro RNAs produced by usingnuclearextracts from unin-fected cells initiatedatthestartsites usedin vivo. To avoid detection of viral mRNAs in the nuclear extracts from infectedcells,weusedprimers complementarytosites in the vector. To obtaina primer extension product forthe AcM NPVp26geneofsuitablelength,anadditionaltemplatewas constructed(pAcHRp26A,seeMaterials andMethods).The
primerextensionproductswereresolvedon
sequencing gels
next to asequencingladderof the gene,generated by
using
thesameprimer.Results from theprimerextensionanalysis
ofthe 39K gene, theOpMNPVp26gene,and the AcMNPV
p26 gene areshown in Fig. 4, whereas the positionsof the mRNA start sitesin the correspondingDNA sequences are shown inFig. 5.
The invitro transcription start sites of the 39K gene and the OpMNPV p26gene
(Fig.
4AandB)
areidenticalto the startsitespublished
for invivo RNA(1, 10).
However, the AcMNPVp26geneprimer
extensionproduct from in vitro RNAmappedto atranscriptionstart site 17nt upstream of theonepreviouslysuggested (17, 20) (Fig. 4C).This sitewas confirmed by primer extension analysis of in vivo mRNA(Fig. 6A). Invitro transcriptionofpDNAP showed arunoff
transcriptofabout 218nt(Fig.2F).Our attemptstomapthe in vitro transcription start sites by primer extension were unsuccessful
(probably
becausetheRNAconcentration pro-ducedbyuninfected-cellextracts was solow).However,we did map the in vivoDNApolymerasemRNA startsite and found amajorstart site(Fig. 5 and Fig. 6B) located 122bpupstreamof theDNApolymerase ATG,whichcorresponds
to one of the major start sites previously reported (24).
Initiation from this site would produce a runoff transcript
similar in size to that observed by in vitro transcription
assays. Therefore thein vitro and in vivoDNApolymerase
RNAappearto initiateat thesamesite.
Comparison of in vitro transcription by nuclear extracts from a time course of infection with the mRNA levels present in infected cells. Todetermine whether the increased levels of in vitro transcription seen with 4- and 8-h p.i. extracts were a reflection of invivoRNA levels, RNA was isolated
J. VIROL.
on November 9, 2019 by guest
http://jvi.asm.org/
[image:5.612.118.258.84.241.2]A
B
C
WTG C A
4 a
T\ _
Am
C W
A G
A d
_ _
_*
;-A_w
P TGCA P
A A
G
C
A
*>-
T-Ar
A i
-Is
P
TGCA
T
ZA
A T T
C
-<-AW".*
0 T [image:6.612.120.482.80.357.2]T T
FIG. 4. Primer extension analysistodetermine startsitesof RNA transcribed in vitro. Primer extensionproductswere synthesized by usingRNAproduced byuninfected-cell nuclearextracts.32P-labeledprimerextensionproducts and35S-labeleddideoxysequencingladders
generated byusingthesameprimerwere subjectedtoelectrophoresison a6% sequencing gel. RNAstartsitesweredetermined for genes
p39K(A), pOpp26 (B),andpAcHRp26A (C). Theprimerextensionproduct iscontained in lane P. Thearrows indicate thepositionsonthe sequence of the mRNAstart sites.
from Sf9 cells duringa time course of AcMNPV infection. The relative levels oftranscription from the AcMNPV 39K gene were analyzed by primer extension. In in vitro tran-scription assays, the highestlevels of 39K RNA (and RNA from all other genes tested) were produced by nuclear extracts prepared at 4 and 8 h p.i. In vitro RNA levels decreased with 12- and 16-hp.i. extracts, andnodetectable
transcriptswere present withextracts prepared at 24 h p.i.
(Fig. 2A). In contrast, in vivo early 39K RNA transcripts
werefirst detectedatlow levelsat8 hp.i. and reachedhigh
levelsat12 and 18 hp.i.Thesetranscriptswerestill apparent at 27 hp.i., butearlytranscriptswere not detectableat67 h
p.i. (Fig. 7). With the in vivo time course, late transcripts
werefirstseen at12 hp.i.; this latesignalwasvery strongat 18hp.i. butwasreduced at 27and 67 hp.i. (Fig. 7).
Characterization of nuclease activity in nuclear extracts prepared late in infection. Nuclear extracts prepared from insect cells after 8 hp.i. showedadecline intranscriptional activity.Reduced levels of runoff
transcripts
wereproduced
byusingnuclearextracts
prepared
at12or16 hp.i.,
whereas those prepared at 24 or 34 hp.i.
showed nospecific
tran--40 -30 -20
scription signals. This loss of transcriptional activity oc-curredregardless of thetemplateused(Fig. 2AtoC,lanes6 and 7). Polyacrylamide gel protein profiles from uninfected and late-infected nuclear extracts incubated for 1 hat30°C
showednoevidence ofproteindegradationwhencompared
with theoriginal unincubatedsamples(datanotshown).The decline intranscriptionalactivitywasthereforeprobablynot duetogeneralproteolysis. Todetermine whether the reduc-tion in transcriptionwas due to nucleases degrading either the mRNAorthetemplateDNA, the nuclearextractswere tested for RNase and DNase activity. When transcription
products from uninfected-cell nuclearextracts(Fig. 8A,lane 1)weretreated withRNase, allbandswereeliminated
(Fig.
8A, lane 2),
indicating
that all of the labeled productsconsisted of RNA. When the same transcription products
were incubated with nuclearextracts from uninfected cells orfrom 24-hp.i.cells,nochangein the
intensity
of theRNAsignalwasobserved(Fig. 8A, lanes 3 and
4).
Therefore,
we found no evidence ofhigh levels of RNase activity in late nuclearextracts.Nuclear extracts prepared during atime course of
infec--10 +1
AML CTG AAGGGGGGCL5ATAAAAGGGG GTGGGGGCGC GTTCGTCCTC ACTCTCT... / 60 nt/ ...ATG
OPP6 TGA TAAGCATGGG IAIATAAGGG CCTACAGTGT TCTGGTAAAT
C ATGC
.../ 37 nt/...ATG Ac39K GAG CGTATAAAAG AAIA8ATAAG AGCTAATTTA GGCCATTTCA CAGIAAT .../126 nt/ ...ATGOpp26 ATT TGTCACATCG CTATTTAAAA GGAAACGCGG TCGCAAAAGC I.AAACG...1/ 20 nt/ ...ATG
Acp26 GGG TAGAATTCTA CGCGTAAAAC ATGATTGATA ATTAAATAAT TLATTTG.../ 29 nt/...ATG
AcDNApol GTT TCAAAATTAC AACGACTTTG TGAATCTCTC GCAAAGCGTG =ATAC.../115 nt/...ATG
FIG. 5. Sequence of promoterregions. The mRNAstartsites and thepresumedTATAboxesareunderlined.
on November 9, 2019 by guest
http://jvi.asm.org/
[image:6.612.107.500.660.721.2]3482 GLOCKER ET AL.
hrp.i.
8 18 67
4 12 27 TG CA
FIG. 6. Primer extension analysis to determine start sites of RNAproducedinvivo. Toproducetheprimerextensionproducts, primerscomplementarytothefollowinggenes wereused:AcMINPV p26 (A) and DNA polymerase (seeMaterials andMethods) (B).The
sameprimerwereusedtogeneratethe 35S-labeleddideoxy
sequenc-ing ladders. Lanes: 1 and 2, primer extension product from 12-and
18-h p.i. RNA, respectively; T, G, C, and A, sequencing ladders
usedassizestandards.Thesequencecomplementarytoaportionof
thesequencing ladder is indicated.Thearrowsindicate thepositions onthesequenceof the mRNAstartsites.
tion were also tested for DNase activity by incubating the
extracts with template DNA under standard transcription preincubation conditions in the presence or absence of
EDTA. With extracts prepared at 0, 4, and 8 h p.i., no
difference in templaterecoverywasapparent(Fig. 8B, lanes 1, 2, and 3). Extracts from 12 and 16 hp.i. caused aslight
degradation of the DNAtemplatein reactionmixtures from which EDTA had been omitted (Fig. 8B, lanes 4 and 5). However,whentemplatewasincubated inextractsprepared
at 24 or 34 h p.i., the DNA incubated without EDTAwas
almostcompletely degraded (Fig. 8B,lanes 6 and7). There-fore,itwasconcluded that late-infected cell nuclearextracts containedDNase activity.Nevertheless,itwaspossiblethat nuclear extracts prepared from cells late in infection
con-tainedstrongDNAbinding proteinsthatmighthavereduced the recovery of DNA template during phenol extraction. Therefore radiolabeled pBS- DNA, along with nonlabeled template, was added to the in vitro transcription reaction mixtures as described above. After a 60-min incubation at
30°C, DNAwasprecipitated withperchloric acid and
cen-trifuged, and the radioactivityreleasedinto the supernatant
was counted. In samples incubated with nuclear extracts
prepared at 34 h p.i., all the radioactivity was released,
indicating the presence of high levels of DNase activity
(resultnotshown).
DISCUSSION
The data presented in this report were derived from in
vitrotranscription ofavariety of baculovirus earlygenesand
demonstrate that basal levels of transcription require only host RNA polymerase II and its associated transcription
factors.These datasuggestthat there is onlyonecategoryof early genes andthat previous distinctions between baculo-virus immediate-early and delayed-early genes are due to
differences in levels ofpromoteractivity early ininfection.
T
.A
C
-
A
G
T
A
E
->
E
->
E
E
FIG. 7. Primer extension analysis of 39K RNA over a time
course of infection. Primerextension lanes are designated bythe
hour postinfection, indicating the time at which mRNAwas
pre-pared. A sequencing ladder generated fromaplasmid containing the 39Kgene and the same primer as used in the primer extension
analysis is indicated. E marks the start sites of the early 39K transcripts and L marks the latetranscriptionstart site.
Infected-cell nuclear extracts prepared at 4 and 8 h p.i. activated transcription of most baculovirus genes tested from3- to 10-fold. Transactivation has beeninvestigated in vivobycotransfection ofAcMNPV39KorOpMNPV gp64
CATconstructswith theAcMNPVtransactivatinggeneIE-1 into Sf9 cells(4,11). Although gp64showedsimilar levelsof transactivation in vitro and in vivo, the 39K gene was
transactivatedtomuchhigherlevels when measured in the in vivosystem. Directcomparison of these activatedlevels is limited by fundamental differences in the two methods of analysis, and identical results arenotnecessarily expected. The activity measured by in vivo CAT assays may be
amplified by multiplerounds of translation ofasinglemRNA
and by recycling of the CAT enzyme. In contrast, invitro transcription by infected-cell nuclear extracts involves a
complexmixturelikelytocontainnotonly IE-1 and IE-N (5) butavarietyof otheras-yetuncharacterized viral
transacti-vatorsandattenuators and isassayed solelyon thebasis of
thelevels of RNA transcribed.
The baculovirus promoters tested showed various levels of activation with infected-cell extracts. This activation of transcriptioncould be duetoageneral, nonspecific
improve-ment in the extracts, possibly as a result of more efficient
extraction of transcription factors. Such extracts should
cause a higherlevel ofbasal transcription fromall
promot-ers. Alternatively, the activation could be due to the pres-ence of specific transactivators in the infected-cell nuclear
extracts, in which case only targeted promoters would be expected to be activated. Our data show that different promoterswere activated to different levels. The 39K and gp64 promoters were activated to the highest extent,
A
B
12 TGCA
A T T C
A *
T T
1 2 T G C A
-itit
-_
_. I0
_ A.
*.
a
B-a ._4
aw... 4t
I.*
-.-^:
IGJ
C
G
T
G
C.* IT
C
J. VIROL.
on November 9, 2019 by guest
http://jvi.asm.org/
[image:7.612.69.308.67.265.2] [image:7.612.345.550.79.338.2]B
1 2 3 4
1 2 3 4 5 6 7
M C - + - + - + - + - + - + - +
214 nt
61
2
1 2 3 4
6108
5090\-4072 3054 2036
FIG. 8. Assays for nuclease activity in nuclear extracts. (A) Assay for RNase activity. A standard transcription reaction using an
uninfected-cell nuclearextractandSstI-digested pHR39K template DNAwasperformed. After the normal incubation time, the reactionwas
divided in fourpartsand treatedasfollows. Lane1, stopped immediately; lane 2, RNase Awasaddedtoafinal concentration of 1 ,ug/,u, and the reaction wasincubated forafurther 10minat30°C; lane 3,to testfor RNase activity in the uninfected-cell nuclearextract,thisextract
wasaddedataconcentration of 30% of total reaction volume, and the reactionwasincubated forafurther 30minat30°C; and lane 4,to test
for RNase activityinlate-infected cell nuclearextracts, nuclearextractpreparedat34hp.i.wasaddedat aconcentration of30% of total
reaction volume and incubated for 30min at30°C. The position and size of the expected runoff productaremarked withan arrow.Fora
descriptionof the graph, see thelegendforFig. 2. (B) Assay for DNaseactivity. Extractswere tested for DNaseactivity by incubating SstI-digested pHR39K template with nuclearextracts in the presence (+) orabsence (-) of EDTA. The following indicate the stage of infectionatwhich the nuclearextractswereprepared: lane 1, uninfected; lane 2, 4 h p.i.; lane 3, 8 h p.i.; lane 4, 12 h p.i.; lane 5, 16 h p.i.;
lane6, 24 h p.i.; lane 7, 34 h p.i. Lane C, nuclear extractwasnotincluded in thetranscription reaction; lane M, 1-kb ladder (Bethesda Research Laboratories) usedassize standard.
whereas the adenovirus major late and AcMNPV p26 pro-
-moters were not significantly activated. These data argue
against nonspecific activation and strongly suggest the in-volvement ofspecific transactivation bycomponentsof the infected-cell nuclearextracts.
Constructs containingthe baculovirus enhancerhr5were
examined for elevated levels of transcription, using both uninfected- and infected-cell nuclearextracts. No evidence for enhancementwas seen.Conditionsorfactor(s) essential
for enhancement could have been either lacking or
inacti-vatedintheextracts.Similarproblems couldaccountfor the lack oftranscription from the latepromoters present onthe
39K andgp64 constructs.
The decline in activity of nuclear extracts prepared late after the initiation of infectionsuggested thatan inhibitor of
transcriptionwaspresentorthat nucleasesorproteasesmay
be affecting thesystem. Wefoundthat, although therewas no evidence ofhigh levels of RNase orprotease activityin the nuclear extracts, levels of DNase activity capable of completely hydrolyzing templateDNAweredetected in the late-infected nuclear extracts. Wilson and Miller (25) have shown that during thecourseofbaculovirusinfection, host cell DNA is notdegraded. The DNaseactivityweobserved
may be the result of viral infection, possibly caused by a
turnoff of the synthesis of a host DNase inhibitor or by
destabilization of cell compartmentalization (e.g., DNase-containing lysosomes). The nuclear extraction preparation protocol may cause the release of such nucleases into the
extract.
The 39K promoter was used to compare the
transcrip-tionalactivityofnuclearextractsprepared atdifferent times
after theonsetof infection with the levels of 39Ktranscripts presentin the infected cells. The results indicated that the nuclearextractssupported transcription of the39Kgene at high levels well before 39KmRNA accumulated to
measur-able levels in infected cells. The marked contrast between these resultsmayreflect the difference in thetwomethods of analysis. Invitrotranscriptionassaysmeasurethe abilityof
enzymesand factors present in a nuclearextract toinitiate transcription from an abundant exogenous DNA template. In contrast, primer extension analysis of in vivo RNA
measures the accumulation of specific RNA transcribed
from an endogenous template. This accumulation is
gov-ernedbythe quantity oftemplate, the rateoftranscription from the template (influenced by such factors as promoter strength and level of transactivation), the length of time RNA is allowedtoaccumulate, and therateof RNA degra-dation.Comparisonof the in vitro and in vivo datasuggested that although enzymes and transcription factors necessary
forhigh levels oftranscription exist at4 and 8 h p.i., they
werenotmaximally exploitedatthis time in vivo. Thismay
be caused bya lack oftemplate DNA in cells infectedat a
relativelylowmultiplicityof infectionorthat levels ofearly
gene mRNA may not accumulate until after DNA replica-tion,when the number ofcopies ofearlygenesin a cell is amplified, thereby permitting increased levels of transcrip-tion.
Thestudy described in thisreportwasdesignedto deter-mine whether this in vitro system mirrored transcriptional
events occurring in cells during baculovirus infection. The datawehavepresentedindicate that theuseof infected-cell
nuclear extracts may prove useful for the identification of
A
on November 9, 2019 by guest
http://jvi.asm.org/
[image:8.574.47.516.71.288.2]3484 GLOCKER ET AL.
factors involved in transactivation ofearly genes. However, the presence of higher levels of DNase activity in nuclear extracts from late in infection complicates the use of this approach for the study of late gene regulation. In addition,
factors involved in enhancement of earlygene expression, through interaction with known baculovirus enhancer se-quences, were not evident in these extracts. We are cur-rently investigating the use ofdifferent extraction protocols to develop in vitro systems suitable for the investigation of all transcription events occurring during the baculovirus infection cycle.
ACKNOWLEDGMENTS
The authors thank C. Rasmussen, C.Gross, andG.Blissard for suggestions and criticisms ofthemanuscript.
This workwassupported bygrantAl 21973 from theNIH. REFERENCES
1. Bicknell, J. N., D. J. Leisy, G. F. Rohrmann, and G. S. Beaudreau. 1987. Comparisonof thep26regionoftwo baculo-viruses.Virology 161:589-592.
2. Blissard, G. W., and G. F. Rohrmann. 1989. Location, se-quence, transcriptional mapping, and temporal expression of thegp64envelope glycoproteingeneof theOrgyia pseudotsug-atamulticapsid nuclearpolyhedrosisvirus. Virology 170:537-555.
3. Blissard, G. W., and G. F. Rohrmann. 1990. Baculovirus diversityandmolecularbiology. Annu. Rev. Entomol. 35:127-155.
4. Blissard, G. W., and G. F. Rohrmann. 1991. Baculovirus gp64 geneexpression: analysisof sequencesmodulating immediate-early transcription andtransactivationbyIEL.J. Virol. 65:5820-5827.
5. Carson, D. D., M. D. Summers, and L. A. Guarino. 1991. Transient expression of the Autographa californica nuclear polyhedrosis virus immediate-earlygene, IE-N, is regulated by three viral elements. J. Virol. 65:945-951.
6. Cochran, M. A., and P. Faulkner. 1983. Location ofhomologous DNA sequencesinterspersed atfive regionsin the baculovirus AcMNPV genome. J. Virol. 45:961-970.
7. Friesen, P. D., and L. K. Miller. 1986. The regulation of baculovirusgene expression. Curr. Top. Microbiol. Immunol. 131:31-49.
8. Fuchs, Y. L., M. S. Woods, and R. F. Weaver. 1983. Viral transcription during Autographacalifomnicanuclear polyhedro-sis virusinfection:anovelRNApolymeraseinduced ininfected Spodopterafrugiperdacells.J. Virol.48:641-646.
9. Grula, M. A., P. L. Buller, and R. F. Weaver. 1981. Alpha amanitin-resistantviral RNA synthesis in nuclei isolated from nuclear polyhedrosis virus-infected Heliothis zea larvae and Spodopterafrgiperdacells. J. Virol.38:916-921.
10. Guarino, L. A., and M. W. Smith. 1990. Nucleotide sequence andcharacterization of the39K gene region of the Autographa califomnica nuclearpolyhedrosisvirus. Virology 179:1-8. 11. Guarino, L. A., and M. D. Summers. 1986. Functional mapping
ofatrans-activating generequired for expression ofa baculo-virusdelayed-early gene. J. Virol.57:563-571.
12. Guarino, L. A., and M. D.Summers. 1986.Interspersed homol-ogous DNA ofAutographa califomnica nuclear polyhedrosis virus enhancesdelayed-earlygeneexpression.J.Virol. 60:215-223.
13. Guarino, L. A., and M. D. Summers. 1987. Nucleotide sequence and temporal expression ofa baculovirus regulatory gene. J. Virol. 61:2091-2099.
14. Hawley, D. K., and R. G. Roeder. 1987. Functional steps in transcription initiation and reinitiation from the major late promoterinaHeLanuclearextract.J. Biol. Chem. 262:3452-3461.
15. Hoopes, R. R., Jr., and G. F. Rohrmann. 1991. In vitro transcription of baculovirus immediate early genes: accurate
mRNA initiation by nuclear extracts from both insect and human cells. Proc. Natl. Acad. Sci.USA88:4513-4517. 16. Huh, N. E., and R. F. Weaver. 1990. Identifying the RNA
polymerases that synthesize specific transcripts of the Au-tographa californica nuclearpolyhedrosis virus.J. Gen. Virol. 71:195-201.
17. Liu, A., J. C.Qin, C. Rankin, S. E. Hardin, and R. F. Weaver. 1986. Nucleotide sequence of a portion of the Autographa califomica nuclear polyhedrosis virus genome containing the EcoRIsite-rich region(hr5)and an openreading framejust5' of the plOgene.J. Gen.Virol. 67:2565-2570.
18. Nissen, M. S., and P. D. Friesen.1989.Molecularanalysisof the transcriptionalregulatory region ofanearly baculovirusgene. J. Virol. 63:493-503.
19. Ooi, B., and L. K. Miller. 1988.Regulationof host RNAlevels duringbaculovirus infection. Virology 166:515-523.
20. Rankin, C., B. F. Ladin, and R. F. Weaver. 1986. Physical mapping of temporally regulated, overlapping transcriptsin the region of the 10K protein gene in Autographa californica nuclearpolyhedrosis virus. J. Virol.57:18-27.
21. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory,Cold Spring Harbor,N.Y.
22. Theilmann, D. A., and S. Stewart. 1992. Tandemly repeated sequenceatthe 3' end of the IE-2 gene of the baculovirusOrgyia pseudotsugata multicapsid nuclear polyhedrosis virus is an
enhancer element.Virology187:97-106.
23. Theilmann, D. A., and S. Stewart. 1991. Identification and characterization of the IE-1 gene of Orgyia pseudotsugata multicapsidnuclearpolyhedrosis virus. Virology180:492-508. 24. Tomalski, M. D., J. Wu, and L. K. Miller. 1988. The location,
sequence, transcription, and regulationof abaculovirusDNA polymerasegene.Virology 167:591-600.
25. Wilson, M. E., and L. K. Miller. 1986.Changes in the nucleo-protein complexes of a baculovirus DNA during infection. Virology151:315-328.
26. Yang, C. L., D. A.Stetler,and R. F. Weaver. 1991. Structural comparisonof theAutographacalifornica nuclear polyhedrosis virus-induced RNA polymerase and the three nuclear RNA polymerasesfrom thehost, Spodopterafrugiperda. Virus Res. 20:251-264.
J. VIROL.