Proc. Nadl. Acad. Sci. USAVol. 88, pp. 4513-4517, May 1991Biochemistry

In vitro transcription of baculovirus immediate early genes:Accurate mRNA initiation by nuclear extracts from bothinsect and human cells

(Spodopterafrugiperda cells/Namalwa cells/Autographa californica nuclear polyhedrosis virus IEI gene/Orgyia pseudotsugata gp64 gene)

RICHARD R. HOOPES, JR., AND GEORGE F. ROHRMANNDepartment of Agricultural Chemistry, Oregon State University, Corvallis, OR 97331

Communicated by Max D. Summers, February 22, 1991 (received for review December 19, 1990)

ABSTRACT The production and characterization of nu-clear extracts from uninfected Spodoptera frugiperda cells,capable of accurately initiating transcription of baculovirusimmediate early genes in vitro, are described. Optimal in vitrotranscription was dependent on the presence of a TATA boxpromoter element and was abolished by a-amanitin. Nuclearextracts from the S. frugiperda cells primed with plasmid DNAcontaining the adenovirus major late promoter produced run-off transcripts of the size predicted for initiation from theadenovirus promoter. In addition, nuclear extracts preparedfrom a human cell line accurately initiated transcription fromthe promoter of the baculovirus immediate early gene encodinggp64. Primer extension analysis showed that transcripts de-rived from the gp64 gene promoter using both the S.frugiperdaand human cell nuclear extracts initiated at the same nucleotideas transcripts produced in vivo.

Nuclear polyhedrosis viruses (NPVs), a subgroup of theBaculoviridae, are a diverse family of viruses pathogenic forinvertebrates, particularly insects of the orders Lepidoptera,Hymenoptera, and Diptera (1). NPV virions have largedouble-stranded DNA genomes of 88-165 kilobases (kb) andare occluded in polyhedron-shaped protein crystals com-prised of a 29-kDa protein termed polyhedrin (2). Occlusionprotects the virions from exposure to the environment andallows them to remain infectious to susceptible insects in-definitely. Polyhedrin is expressed at high levels late in thevirus infection, and this hyperexpression appears to befacilitated by an a-amanitin-resistant RNA polymerase de-rived either from the host or synthesized de novo by the virus(3). Because of the high levels of polyhedrin expression, alarge number of laboratories are using engineered baculovi-ruses for the hyperexpression of foreign genes under thecontrol of the polyhedrin gene promoter (4, 5). Although theregulation of baculovirus gene expression is of considerableinterest, progress to define the components ofthis system hasbeen limited by the lack of an in vitro transcription system.In this report we describe the production and characteriza-tion of an in vitro transcription system derived from nuclearextracts of uninfected Spodopterafrugiperda cells that faith-fully initiates the transcription of baculovirus immediateearly genes. This system should prove useful for the isolationand identification of host and viral transcription factorsinvolved in the regulation of baculovirus gene expression.

MATERIALS AND METHODSTemplates. A plasmid (p64CAT-166) containing the gp64

promoter (the promoter of the immediate early gene encodinga 64-kDa glycoprotein, gp64) (6) was used as the transcriptiontemplate and was provided by Gary Blissard. p64CAT-166

contains 166 base pairs (bp) of the 5' flanking sequence and21 bp of the gp64 gene open reading frame fused to thebacterial chloramphenicol acetyltransferase (CAT) gene andcloned into a pBluescribe plasmid [pBS(-)] (Fig. 1B). Aderivative of p64CAT-166, containing a change in the DNAsequence of the gp64 gene TATA box, was designed toinvestigate the role of the TATA sequence in gp64 geneexpression. The mutation was produced by site-directedmutagenesis (see below) using a 35-base oligonucleotidecomplementary to nucleotides -58 to -90 upstream of thegp64 gene ATG, which altered the native sequence (GGG-TATATAA) to GTCTAGATAA (the changes are italic).These changes, which introduced an Xba I site to facilitateplasmid screening, were verified by DNA sequencing.The adenovirus major late (ML) promoter was assayed

from the plasmid pBRD, a gift of Diane Hawley (Universityof Oregon). This plasmid contains adenovirus sequencesfrom -260 to +10, followed by a guanosine-less cassettederived from pMLCAAT (7). When digested with BamHI,this template produces a run-off transcript of 389 nucleotidesin a human cell transcription system. A plasmid containingthe AcMNPV IEJ gene (ref. 8; see. Fig. 1) was the gift ofLinda Guarino (Texas A&M University).To prepare templates for transcription, plasmid DNA was

digested with the desired restriction enzyme; extracted withphenol/chloroform, 1:1 (vol/vol); precipitated with ethanol;and resuspended in 10 mM Tris, pH 7.5/1 mM EDTA (TEbuffer) at a concentration of about 400 ,ug/ml.Maintenance and Growth of Sf9 Cells. S. frugiperda cells

(Sf9 cells; ATCC CRL 1711) were obtained from GIBCO/BRL and grown in serum-free Sf900 medium (GIBCO/BRL)in 250- or 500-ml sterile disposable Erlenmeyer flasks (Corn-ing) on an orbital shaker (VWR Scientific model 2001) at 135rpm. Cells were grown in a VWR Scientific model 2020incubator at 27°C without CO2 and were maintained atdensities of 0.2-3.6 x 106 cells per ml.Nuclear Extract Preparation. Nuclear extracts were pre-

pared with few modifications as described by Dignam et al.(9) except that all buffers contained Tris HCl (10) rather thanHepes. Sf9 cells were harvested at a density of3-5 x 106 cellsper ml by pelleting at 1500 rpm for 10 min in a Beckman GPcentrifuge (swinging bucket rotor) at room temperature. Allsubsequent steps were carried out at 4°C. Cells were resus-pended in 4 packed-cell volumes of buffer A (10 mMTris-HCI, pH 7.9/1.5 mM MgCI2/10 mM KCI/0.5 mM di-thiothreitol) and were left on ice for 10 min. Cells then werepelleted in a clinical centrifuge (1500 rpm for 10 min),resuspended in 2 packed-cell volumes of buffer A, and lysedwith 10 strokes of a Kontes Dounce homogenizer with a B

Abbreviations: NPV, nuclear polyhedrosis virus; CAT, chloram-phenicol acetyltransferase; MNPV, multicapsid NPV; AcMNPV,Autographa californica MNPV; gp64 promoter, immediate earlypromoter of the gene expressing glycoprotein gp64.

4513

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

July

10,

202

1

4514 Biochemistry: Hoopes and Rohrmann

A In vitro TranscriptionM 1 2 3 4 5

1078-

872-

603-

310-271-118-

.

B gp64cat-1 66316 nt

%XXXXXX x'''' ''' CATE TATA CAGT

8

gp64 5' Loader166nt

6

-.o-857

Wq -3394-316

57 nt

:7,I/EkV

339 ntC IEI

i T CTATA CAGT .-

FIG. 1. (A) In vitro transcription of baculovirus immediate earlygenes. Lanes: 1, no template; 2, HindIII-digested p64CAT-166; 3,EFcoRI-digested p64CAT-166; 4, EcoRI-digested p64CAT-166 with a

mutated TATA (TAGA) box; 5, same as in lane 3 except that 1 ,&gpf a-amanitin per ml was added to the transcription reactions; 6,Hinfl-digested immediate early transacting factor gene (JEJ) of theAutographa californica multicapsid NPV (AcMNPV). All reactionswere carried out using standard reaction conditions. The numbers on

the left indicate the position of radio-labeled Hae III-digested OX174markers (M). The numbers on the right indicate the location andestimated size of run-off transcripts. (B and C) Diagram of p64CAT-166 and IEL. The arrows with numbers in nucleotides (nt) indicate thetranscripts produced from the RNA start site. The locations of theTATA box and CAGT consensus sequences are indicated.

pestle. Nuclei were pelleted (1500 rpm for 10 min) andresuspended in 1 nuclear volume of buffer C [20 mMTris-HCl, pH 7.9/25% (vol/vol) glycerol/420 mM NaCl/1.5mM MgCl2/0.2 mM EDTA/0.5 mM dithiothreitol]. The nu-

clei were then subjected to 15 strokes of the homogenizer (Bpestle), stirred slowly in a beaker with a stir bar for 30 min at4°C, and centrifuged in a Sorval SS-34 centrifuge (25,000 xg average for 30 min). (At this point the supernatant fractionof the extract can be frozen in liquid nitrogen and stored at-80°C.) The supernate was dialyzed (3500-Da cutoff) against40 volumes of buffer D (20 mM Tris HCI, pH 7.9/100 mMKCI/20% glycerol/0.2 mM EDTA/10 mM 2-mercapto-ethanol) for 3 hr. No precipitate was evident after this dialysisstep. Aliquots (50 ,ul) of the resulting dialysate were frozen inliquid nitrogen and stored at -80°C. From 500 ml of Sf9 cellsat 3-5 x 106 cells per ml, the yield was typically 6 ml ofnuclear extract with a protein concentration of 12-14 mg/mlas determined by the Bradford protein assay (11).

In Vitro Transcription Reactions. Standard conditions foroptimal in vitro transcription were determined as described inResults. Reactions were carried out in 25-,ul volumes, with 20,ug of digested p64CAT-166 plasmid per ml as template. Finalconcentrations of the reaction components were 20 mMHepes (pH 8.4 at 25°C), 6 mM MgCl2, nuclear extract at 20%

of the final reaction volume (about 3 mg/ml), and the fol-lowing components contributed by the extract storage buffer:20 mM KCI, 4 mM Tris HCl, 4% glycerol, 0.04 mM EDTA,and 2 mM 2-mercaptoethanol. Reaction mixtures of nuclearextracts and templates were preincubated for 20 min at 30'C.Transcription was then initiated by the addition of nucleo-tides, and reaction mixtures were incubated for 40 min at300C. The final concentrations of components of the nucle-otide mix used were: 600 AuM each of ATP, CTP, and GTP;25 A&M UTP (Pharmacia); 10 mM creatine phosphate (Boeh-ringer Mannheim); and 5 ,uCi (185 kBq) of [a-32P]UTP (800Ci/mmol; DuPont/NEN) per reaction. Reactions werestopped by the addition of 25 A.l of stop buffer (0.5% SDS/10mM EDTA/100 mM NaOAc, pH 5.2/1 mg of tRNA per ml)and were extracted once with an equal volume of 1:1 phenol/chloroform. The phenol phase was then back-extracted with55 A1d of stop buffer, and the combined aqueous phases wereprecipitated with ethanol. After 25 min at -700C, RNA waspelleted by microcentrifugation for 15 min at 4°C, air-dried atroom temperature for 25 min, and suspended in 15 ,ul of 98%formamide/2% each xylene cyanol and bromophenol blue.Samples (5-10 ,ul) were electrophoresed at 550 V for 1.5 hrthrough a 7 M urea/5% polyacrylamide gel in TBE buffer(0.09 M Tris/0.09 M borate/2 mM EDTA) in a Bio-RadProtean II apparatus, which was cooled with circulating tapwater. Hae III restriction fragments of 4X174 DNA, radio-actively labeled by using T4 DNA polymerase (12), were usedas size standards on gels. Gels were dried by vacuum, andKodak XAR film was exposed for about 12 hr at -80°C witha DuPont intensifying screen. Autoradiograms were scannedby using a Zeineh soft-laser densitometer (Biomed Instru--ments, Fullerton, CA), and areas under peaks were mea-sured; because the autoradiogram bands may not have beenin a linear range for densitometry, our data are not consideredquantitative.The Namalwa cell nuclear extracts were derived from a

human Burkitt lymphoma cell line (26) and were the gift ofBarbara Hoopes (University of Oregon).Primer Extension Analysis. The in vitro transcription start

sites were confirmed by primer extension analysis of the invitro RNA prepared as described above, except that labeledUTP was not included in the reactions. In vitro transcribedRNA, purified from a 25-,ul transcription reaction, was hy-bridized to a 5'-end-labeled 22-mer oligonucleotide comple-mentary to the 5' end (nucleotides 15-37) of the CAT geneopen reading frame on p64CAT-166. End-labeling and primerextensions were carried out as described (6, 13).DNA Sequencing and Site-Directed Mutagenesis. The dide-

oxy chain-termination method of Sanger et al. (14), withmodifications using Sequenase (United States Biochemical)(15), was used to sequence double-stranded plasmid DNA.Site-directed mutagenesis was accomplished by following themethods of Kunkel et al. (16) and using oligonucleotidessynthesized on an Applied Biosystems oligonucleotide syn-thesizer.

RESULTSIn Vitro Transcription from the Baculovirus gp64 Promoter.

The baculovirus gp64 gene encodes a glycoprotein thatmodifies the infected insect cell plasma membrane and be-comes a major component ofthe budded virus envelope whenthe virus buds through the modified membrane. Investiga-tions of the regulation of the gp64 gene of Orgyia pseudo-tsugata MNPV (OpMNPV) indicated that it has both an earlypromoter and a series of four late promoters (6). Subsequentinvestigations of sequences that regulate expression of theimmediate early gp64 promoter identified regions of the 5'flanking sequence that modulated the expression of a gp64-CAT gene fusion transfected into uninfected Sf9 cells (G.Blissard and G.F.R., unpublished). One construct, p64CAT-

Proc. Natl. Acad Sci. USA 88 (1991)

Dow

nloa

ded

by g

uest

on

July

10,

202

1

Proc. Natl. Acad. Sci. USA 88 (1991) 4515

166 (Fig. 1B), contains 166 nucleotides of the gp64 gene 5'flanking sequence and gave maximal CAT expression relativeto a variety of other constructs. These results suggested thatit would be a useful template for developing an in vitrotranscription system for immediate early genes.The rationale we used to produce an in vitro system for

baculovirus transcription was to start by applying the sim-plest and most straightforward methods previously used tomake transcriptionally active eukaryotic cell extracts. Wetherefore applied the procedures described by Dignam et al.(9) for the preparation of mammalian cell nuclear extracts toour uninfected Sf9 cells. To assay for in vitro transcriptionactivity from the nuclear extracts, templates produced bydigestion of p64CAT-166 with HindIII or EcoRI were used.HindIII linearized the plasmid, while EcoRI produced a DNAfragment of 439 nucleotides containing the gp64 promoter(Fig. 1B). Correct initiation from the gp64 promoter wouldproduce run-off transcripts of 857 or 316 nucleotides from theHindIII or the EcoRI-digested plasmids, respectively (Fig.1B).

Several observations indicated that the Sf9 nuclear ex-tracts correctly transcribed the immediate early gp64 pro-moter. In an initial experiment, we observed that transcrip-tion reactions containing no template produced no specifictranscripts (Fig. 1A, lane 1). However, when HindIII- orEcoRI-digested templates were added, major run-off prod-ucts (about 857 and 316 nucleotides, respectively) werepresent, and their lengths were consistent with the distancefrom the in vivo transcription initiation site to the restrictionsite (Fig. 1A, lanes 2 and 3). Furthermore, transcription of agp64 mutant promoter in which the sequence of the TATAbox was changed to TAGA resulted in a reduction by at leasta factor of 15 in the level of transcription (Fig. lA, lane 4) asdetermined by densitometric scanning. When transcription ofthe wild-type gp64 gene was examined in the presence of 1 Agof a-amanitin per ml, transcription was abolished (Fig. lA,lane 5). As further support for these results, another bacu-lovirus gene transcribed from an immediate early gene pro-moter (the IEJ gene of AcMNPV) was examined. The IEJplasmid was digested with Hinfl to yield a DNA fragmentcontaining the IEJ start site 339 nucleotides upstream of therestriction site (8). This template produced a runoff transcriptof length consistent with the expected 339-nucleotide product(Fig. 1A, lane 6). These results suggested that this systemproduced transcripts from the correct start site for baculov-irus immediate early gene promoters and that optimal levelsof transcription were dependent on the presence of a TATAbox. Furthermore, the sensitivity of transcription to a-aman-itin suggested that the mRNA was synthesized by RNApolymerase II. Altogether, these results indicated that nu-clear extracts from Sf9 cells were capable of producing RNApolymerase II-mediated in vitro transcription.

Optimization of Extract and Template Concentrations. Weexamined the effects of varying extract and template concen-tration on the production of specific transcripts by the Sf9system. To determine the optimal extract concentration, twosets of experiments were done with either 20 or 60 mM KCI(see the legend to Fig. 2). A broad range of nuclear extractconcentrations (1.4-5.6 mg/ml) was found to produce signif-icant levels of run-off transcripts, with 2.8 mg/ml being theoptimal concentration (Fig. 2A). Dilutions of less than 1 mg ofextract per ml produced transcripts of a size consistent withend-to-end transcription of the DNA fragments but no specifictranscripts (data not shown). Such transcripts may be causedby the dilution of specific transcription factors, resulting in thepreferential interaction ofRNA polymerase II with the ends ofthe DNA fragment. High concentrations of extract alsostrongly inhibited specific transcription (Fig. 2A).We found that a range of template concentrations (10-50

,ug/ml) produced specific transcription, with 20,g/ml being

A20 mM KC1 60 mM KCI

316Ii

1.4 2.1 2.8 2.8 5.6 8.4Extract. mg/mi

B

I:Ip q. -.

U11 5 10 20 30 50Template, big/ml

FIG. 2. Optimization of nuclear extract and template concentra-tion. (A) Nuclear extract concentration. (B) Template concentration.Both sets of experiments were done by using standard transcriptionconditions except for the determination of extract concentration.Because the extract is stored in a 100mM KCI buffer, KCI was addedto the reaction mixtures containing 1.4 and 2.1 mg of extract per mlto make them equivalent to the reaction containing 2.8 mg of extractper ml (2.8 mg/ml is 20%6 of the original extract concentration andtherefore contains 20 mM KCl). Likewise, similar adjustments weremade to the reaction mixtures containing 2.8 and 5.6 mg of extractper ml to make them equivalent to the reaction mixture containing 8.4mg of extract per ml (600% extract concentration). The numbers (316nucleotides) indicate the location and estimated size of the run-offtranscript. A is relative absorption.

the optimal concentration (Fig. 2B). At the higher templateconcentrations (50 ug/ml), specific transcription was re-duced, and larger nonspecific transcripts were generated(Fig. 2B).

In an early attempt to reduce the level of nonspecificbackground transcripts, aDNA fragment containing the gp64promoter was gel-purified and used as the template in astandard transcription reaction. Under those conditions,specific transcripts were not observed. However, whenDNAlacking baculovirus immediate early gene promoters (e.g.,pBS) was added to reactions containing the gel-purified gp64template specific run-off transcripts were produced (data notshown). Related observations were reported by Dignam et al.(9), who found that the addition of nonspecific DNA reducedthe concentration requirements for template DNA. Becauseof these observations, the templates used in our subsequentstudies were not gel-purified and included DNA fragmentsderived from the original plasmid in addition to the specificpromoter-containing fragment.Further Optimization of Reaction Conditions. To further

characterize the in vitro transcription conditions, the influ-ence of Mg2+ concentration, KCI concentration, tempera-ture, pH, and preincubation and incubation times were

examined. Specific transcription from the early gp64 pro-

moter was observed for Mg2+ concentrations from 2 to 10mM, with the optimum being 6 mM MgCl2 (Fig. 3A). Theoptimal Mg2+ concentration for another baculovirus imme-diate early gene promoter (AcMNPV IE-I) (8) was also foundto be 6 mM (data not shown). Concentrations of <2 mM or

>15 mM MgCl2 resulted in greatly reduced levels of tran-scription.The temperature dependence of in vitro transcription with

the Sf9 extracts was determined by analyzing the gp64 run-offproducts of reactions carried out at temperatures between16°C and 35°C. While temperatures from 25°C to 30°C sup-

ported high levels of transcription, the optimal temperaturefor the production ofrun-offtranscripts was determined to be30°C (Fig. 3B). In this analysis, reactions were both prein-cubated and incubated at the temperature tested. Thus, anydifferences in the temperature dependence of specific tran-

-_316

Biochemistry: Hoopes and Rohrmann

Dow

nloa

ded

by g

uest

on

July

10,

202

1

4516 Biochemistry: Hoopes and Rohrmann

A gp64 AML

St Na Sf Na

L d|0 0.5 1 2 4 6 10 20

Mg2. mM

--316 1

5 20 45 60 90

PreincLubation ir'-

-316

16 21 25 30 35TermoeratlUre. C

.:t-31 96

s;1~~~~~~~~~7-Olll,0 1 5 10 15 30 45 65

Time course. mrimn

- 389

316-.- 1M

FIG. 3. Optimal conditions for transcription of the immediateearly gp64 promoter. (A) Mg2+ concentration. (B) Temperature. (C)Preincubation time. (D) Incubation time. Preincubation was carriedout for 20 min under standard reaction conditions at 300C withoutdNTPs and was followed by 40-min incubations with dNTPs. Allreactions took place under standard transcription conditions exceptfor the reaction condition being tested. The size of the majortranscript (316 nucleotides) is indicated on all panels.

scription steps, such as initiation or elongation, would nothave been distinguished.The effect of varying pH and KCl concentration in the

reactions was also tested and found to have little effect overa broad range. While all extracts were prepared with Tris'HClat pH 7.9 (40C), we varied the pH of the in vitro transcriptionreactions by adding Hepes buffers (20 mM final concentra-tion) of pH 7.4, 7.9, 8.4, and 8.9. No significant differencewas seen in the production of transcripts in this range of pH(data not shown). Varying KCl concentrations in the tran-scription reactions from 20 mM to 60 mM had only a slighteffect on transcription, although results indicated that 20mMKCI may be optimal (Fig. 2A).

Transcription reactions were carried out with a preincuba-tion step to allow for the binding oftranscription factors to thetemplate before elongation by RNA polymerase II. There-fore, the effect of varying preincubation time in the absenceof the nucleotide substrates (which allow elongation) was

investigated. Very little transcription from the gp64 genepromoter was seen when the preincubation step was omitted,even though the reactions were incubated for 40 min after theaddition of nucleotides (Fig. 3C). A significant increase in theproduction of transcripts was seen after only 5 min ofpreincubation, while 20 min of preincubation was sufficientfor near maximal transcription (Fig. 3C).The time course of transcription after a 20-min preincuba-

tion period was measured under standard assay conditions,and the results indicated that maximal production of tran-scripts occurred after 30 min of incubation (Fig. 3D).The gp64 and the Adenovirus Major Late Promoters Are

Transcribed by Both Sf9 and Human Cell Nuclear Extracts. Totest whether transcription from the gp64 promoter wasunique to the Sf9 system or was based on expression fromuniversal promoter elements, we investigated the ability ofnuclear extracts from Namalwa cells, a human Burkitt lym-phoma cell line (26), to produce run-off transcripts. Similar-size run-off transcripts from the EcoRI-digested p64CAT-166construct were produced with extracts from both Sf9 andNamalwa cells (Fig. 4A). In addition, run-off transcripts ofsimilar size were produced from the adenovirus major latepromoter by nuclear extracts from both cell types (Fig. 4).These results indicate that the transcription factors of both

B AML389 nt

TATA CATT

Em

FIG. 4. Comparison of transcription by extracts from Sf9 andhuman cell nuclei. (A) Comparison of the transcription of theimmediate early gp64 promoter and adenovirus major late promoterby Sf9 (Sf) and human cell (Na) nuclear extracts. (B) Diagram of theplasmid containing the adenovirus major late (AML) promoter. Thelengths of the run-off transcripts are indicated by the arrows.

insect and human cells are capable of recognizing and initi-ating transcription from baculovirus and human virus pro-moters interchangeably.Primer Extension Analysis of in Vitro-Generated gp64 Gene

Transcripts Produced by Nuclear Extracts from Both Insectand Human Cell Lines. To determine if the run-offtranscriptsproduced in these investigations were initiating at positionsidentical to the in vivo transcripts, primer extension analysiswas performed on RNA purified from the in vitro transcrip-tion reactions (Fig. 5). The insect and human cell-derivedextracts produced primer extension products of identicallength, and these mapped to the adenosine of the gp64 geneCAGT motif (Fig. 5A). This is identical to the mRNA startsite of gp64 RNA produced in vivo (6). The CAGT motif is asequence conserved at the start sites of several baculovirusearly genes (6) and is similar to a consensus sequence formRNA initiation for a variety ofgenes from other eukaryotes(17, 18). In control reactions where no gp64 template wasadded to the transcription reactions, no primer extensionproducts were evident (Fig. SA).The uninfected Sf9 extracts did not support baculovirus

late gene expression, as we found no evidence of primerextension products mapping to the late promoter sequences(ATAAG) in the gp64 gene 5' flanking region.

DISCUSSIONThe baculovirus life cycle is characterized by a uniquecascade of differential gene expression that terminates in thehyperexpression of two late genes that encode polyhedrinand p10. In this report, we have described an in vitrotranscription system for baculovirus immediate early genesand have used it to begin to characterize the contributions ofthe host cells to the initial events of baculovirus infection. Inaddition, we have demonstrated that nuclear extracts fromboth uninfected lepidopteran and human cells are capable ofaccurately initiating the transcription of baculovirus imme-

A B-316

C

Proc. Natl. Acad. Sci. USA 88 (1991)

U

%:.. .r-.. i. L

Dow

nloa

ded

by g

uest

on

July

10,

202

1

Proc. Natl. Acad. Sci. USA 88 (1991) 4517

A Sf NaG A T C + --+

AnT

CAGT

T

G

A

B

TATA C

0a

FIG. 5. Determination by primer extension analysis of the startsite of in vitro transcribed gp64 RNA. (A) GATC indicates thesequencing ladder produced from the plasmid p64CAT-166; the sameCAT primer used for the primer extension analysis (see B) was usedfor the DNA sequencing. Sf and NA indicate primer extensionanalysis of RNA made from nuclear extracts of Sf9 and Namalwacells, respectively. The lanes designated + and - indicate analysisof RNA purified from transcription reactions with or without tem-plate added. The sequence complementary to a portion of thesequencing ladder in the region of the primer extension products isindicated on the left. (B) Diagram of strategy for primer extensionanalysis of the gp64 early gene transcripts. The dashed arrowindicates the primer extension product. The cross-hatched area is the5' flanking and coding sequences of the gp64 gene.

diate early genes in a manner similar to infected insect cells.Furthermore, we have shown that the adenovirus major latepromoter is transcribed by insect cell nuclear extracts.

Several observations from these investigations are consis-tant with RNA polymerase II-mediated transcription by Sf9nuclear extracts. The optimal conditions for transcription ofbaculovirus immediate early genes are similar to the condi-tions described for RNA polymerase TI-mediated transcrip-tion using human cell extracts (9, 19). Furthermore, thebaculovirus immediate early gene promoters resemble typi-cal RNA polymerase II promoters (20, 21), and mutation ofthe gp64 gene TATA box to TAGA caused a large reductionin specific run-off transcription. Finally, transcription fromthe gp64 promoter was abolished by a-amanitin at a concen-tration (1 ,ug/ml) consistent with RNA polymerase II-mediated transcription (22).

Transcription from the adenovirus major late promoter bySf9 nuclear extracts indicates that the lepidopteran transcrip-tion system recognizes universal consensus sequences com-mon to invertebrates and vertebrates, which signal both thepositioning of the transcriptional apparatus and the site ofmRNA initiation (20). Transcription of the adenovirus majorlate promoter has been reported in HeLa cell extracts (9) andin a variety of heterologous systems including extracts from

Bombyx mori silk glands (23), Drosophila cells (24), and yeast(25) cells.The development of a system for in vitro transcription of

baculovirus immediate early genes based on a run-off assayshould assist in the identification, fractionation, and purifi-cation of host and viral regulatory factors involved in bacu-lovirus transcription. In addition, the run-off transcriptionassay will provide a sensitive method for analysis of theactivity of reconstituted transcriptional complexes. Suchanalyses will be essential for defining the role of specificfactors in baculovirus transcription.

We thank Linda Guarino for the IEJ plasmid and Barbara Hoopesand Diane Hawley for the Namalwa cell nuclear extracts, theadenovirus major late plasmid, and advice throughout this project.The excellent technical assistance of Rebecca Russell is gratefullyacknowledged. We thank Charlotte Rasmussen for her assistancewith the transcription of the IEJ promoter and comments on thismanuscript, Christian Gross for making the site-directed promotermutation, and Gary Blissard for his comments on this manuscript andfor the gift of the p64CAT plasmid used in this project. We also thankStefan Weiss for his assistance with the culture of Sf9 cells. Thisproject was supported by a grant from the National Institutes ofHealth (Al 21973). This is Technical Report no. 9468 from the OregonState University Agricultural Experiment Station.

1. Blissard, G. W. & Rohrmann, G. F. (1990) Annu. Rev. Ento-mol. 35, 127-155.

2. Rohrmann, G. F. (1986) J. Gen. Virol. 67, 1499-1513.3. Grula, M. A., Buller, P. L. & Weaver, R. F. (1981) J. Virol. 38,

916-921.4. Pennock, G. D., Shoemaker, C. & Miller, L. K. (1984) Mol.

Cell. Biol. 4, 399-406.5. Smith, G. E., Summers, M. D. & Fraser, M. J. (1983) Mol.

Cell. Biol. 3, 2156-2165.6. Blissard, G. W. & Rohrmann, G. F. (1989) Virology 170, 537-

555.7. Sawadago, M. & Roeder, R. G. (1985) Proc. Natl. Acad. Sci.

USA 82, 4394-4398.8. Guarino, L. A. & Summers, M. D. (1987) J. Virol. 61, 2091-

2099.9. Dignam, J. D., Lebovitz, R. M. & Roeder, R. G. (1983) Nu-

cleic Acids Res. 11, 1475-1489.10. Wiest, D. K. & Hawley, D. K. (1990) Mol. Cell. Biol. 10,

5782-5795.11. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254.12. Maniatis, T., Fritsch, E. F. & Sambrook, J. (1982) Molecular

Cloning:A Laboratory Manual (Cold Spring Harbor Lab., ColdSpring Harbor, NY), p. 117.

13. Blissard, G., Quant-Russell, R., Rohrmann, G. F. &Beaudreau, G. S. (1989) Virology 168, 354-362.

14. Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Natl.Acad. Sci. USA 74, 5463-5467.

15. La Baer, J. (1989) U.S. Biochem. Corp. Comments 15, 19-20.16. Kunkel, T. A., Roberts, J. D. & Zakour, R. A. (1987) Methods

Enzymol. 154, 367-403.17. Smale, S. T. & Baltimore, D. (1989) Cell 57, 103-113.18. Smale, S. T., Schmidt, M. C., Berk, A. J. & Baltimore, D.

(1990) Proc. Natl. Acad. Sci. USA 87, 4509-4513.19. Dignam, J. D., Martin, P. L., Shastry, B. S. & Roeder, R. G.

(1983) Methods Enzymol. 101, 582-598.20. Bucher, P. (1990) J. Mol. Biol. 212, 563-578.21. Penotti, F. E. (1990) J. Mol. Biol. 213, 37-52.22. Roeder, R. G. (1976) in RNA Polymerase, eds. Losick, R. &

Chamberlain, M. (Cold Spring Harbor Lab., Cold Spring Har-bor, NY), pp. 285-329.

23. Hirose, S. & Suzuki, Y. (1988) Proc. Natl. Acad. Sci. USA 85,718-722.

24. Heiermann, R. & Pongs, 0. (1985) Nucleic Acids Res. 13,2709-2730.

25. Lue, N. F., Flanagan, P. M., Sugimoto, K. & Kornberg, R. D.(1989) Science 246, 661-664.

26. Matsuo, T., Heller, M., Petti, L., O'Shiro, E. & Kieff, E. (1984)Science 226, 1322-1325.

Biochemistry: Hoopes and Rohrmann

TRANSCRI"..w - - - - - - - -

P.

, CATAGT

I-.t cc0

I LU

Dow

nloa

ded

by g

uest

on

July

10,

202

1