- Email: [email protected]
DNA binding activity of the baculovirus late expression factor PP31
Virus Research 90 (2002) 187 /195 www.elsevier.com/locate/virusres
DNA binding activity of the baculovirus late expression factor PP31 Linda A. Guarino a,b,*, Toni-Ann Mistretta a, Wen Dong b a
Department of Biochemistry and Biophysics, Texas A&M University, College Station, TX 77843-2128, USA b Department of Entomology, Texas A&M University, College Station, TX 77843-2128, USA Received 12 April 2002; received in revised form 18 July 2002; accepted 18 July 2002
Abstract PP31 is a baculovirus protein that is essential for viral late gene expression. To study the role of PP31 in late transcription in vitro, it was purified from infected insect cells. A combination of heparin affinity, cation exchange chromatography, and gel filtration was used to purify native non-tagged protein. Nearly 5 mg of PP31 was obtained from 95 mg of nuclear extract confirming that PP31 is an abundant viral protein. DNA binding assays revealed that PP31 binds to single-stranded and double-stranded DNA with equal affinities. Addition of PP31 to in vitro transcription assays with purified baculovirus RNA polymerase resulted in a strong inhibition of transcription. This indicates that the viral RNA polymerase was not able to displace PP31, and suggests that other late expression factors may function to help RNA polymerase bind to PP31-coated templates. # 2002 Elsevier Science B.V. All rights reserved. Keywords: Function; Purification; Transcription; Baculovirus; Gene regulation
1. Introduction Nineteen viral genes are required for expression of baculovirus late genes in a transient expression assay (Rapp et al., 1998). These 19 genes are known as late expression factors (LEFs). In baculoviruses, late gene expression requires early gene expression and viral DNA replication, and so these 19 genes include factors necessary for early events as well as those directly involved in late
* Corresponding author. Tel.: /1-979-845-7556; fax: /1979-845-9274 E-mail address: [email protected] (L.A. Guarino).
gene expression. Ten of the LEFs are required for early gene expression and DNA replication expression. These are ie-1 , ie-2 , dnapol , p143, p35, lef-1, lef-2, lef-3, lef-7, and lef-11 (Kool et al., 1994; Lin and Blissard, 2002; Lu and Miller, 1995). It is currently unknown whether these factors function solely during the early phase or whether they also have roles in late transcription. The remaining 9 factors are believed to be directly involved in late gene transcription. LEF4, LEF-8, LEF-9, and P47 make up the viralencoded RNA polymerase (Guarino et al., 1998b). LEF-5 has a ribbon domain common to the transcription elongation factor TFII-S, and has been predicted to function in elongation (Har-
0168-1702/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 8 - 1 7 0 2 ( 0 2 ) 0 0 1 5 2 - 1
188
L.A. Guarino et al. / Virus Research 90 (2002) 187 /195
wood et al., 1998). LEF-6 and LEF-12 have no obvious sequence homologies with non-baculovirus proteins and little is known concerning their potential functions in late expression. PP31 is a phosphoprotein that binds DNA non-specifically (Guarino et al., 1992). PP31 also interacts with the virogenic stroma, which is an altered form of the nuclear matrix formed at the site of viral DNA replication and packaging (Fraser, 1986; Okano et al., 1999; Wilson and Price, 1988). pp31 is highly expressed and is regulated by tandem early and late promoters (Guarino and Summers, 1986). Together, these observations suggest that PP31 has a structural role in viral life cycle, and may serve to organize the viral replication factories and late transcription apparatus. To analyze the functions of late transcription factors, we have attempted to purify all of the essential LEFs and analyze the effect of adding these factors to in vitro transcription assays containing purified baculovirus RNA polymerase. To this end, we developed a simple scheme for purification of PP31 from Spodoptera frugiperda cells infected with Autographa californica nuclear polyhedrosis virus (AcNPV). Here we report that PP31 bound single and double-stranded DNA with equal affinities. Furthermore, we found that addition of PP31 to in vitro transcription assays with purified baculovirus RNA polymerase resulted in an inhibition of transcription activity.
Superose sp. column. Bound proteins were eluted with a linear salt gradient from 100 to 550 mM KCl. The peak fraction from Superose sp. chromatography was filtered through Superdex 200. Molecular standards were filtered through the same column in the same buffer to establish a standard curve. Purified protein was dialyzed against 50 mM Hepes (pH 7.2), 100 mM KCl, 1 mM DTT, 20% glycerol frozen in liquid nitrogen droplets and stored at /80 8C. 2.2. DNA /protein interactions An oligonucleotide corresponding to the 5?-end of the LEF-12 (nucleotides 33361 /33402 on the genomic map) gene was 5?-end labeled with (g-32P)ATP (Sambrook and Russell, 2001). To make double-stranded probe, the probe was mixed with an equimolar amount of the unlabeled complementary strand. The single-stranded probe was mixed with an equimolar amount of the same unlabeled oligo. In all reactions, the DNA was held constant at 120 nM for dsDNA or 240 nM for ssDNA (a nucleotide concentration of 5 mM in both cases). The protein concentration was varied in twofold increments from 0.01 to 2 mM. All reactions were normalized with BSA diluted in enzyme storage buffer so that the buffer concentrations and total protein amounts were the same in every tube.
2. Materials and methods
2.3. In vitro transcription assays
2.1. Purification of PP31
The transcription template pPolh/CFS was linearized with EcoRI before use in DNA binding assays. The standard assay mix contained 50 mM Tris (pH 7.9), 100 mM KCl, 0.1 mM EDTA, 5 mM DTT, 5% glycerol in a final reaction volume of 20 ml. The DNA concentration was held constant at 0.1 mg/ml (15 mM nucleotide), and the concentration of PP31 was varied from 0.06 to 4 mM in twofold increments. All reactions were normalized with BSA diluted in enzyme storage buffer so that the buffer concentrations and total protein amounts were the same in every tube. In vitro transcription assays were performed as previously described (Xu et al., 1995) using the same
Nuclear extract was prepared from 2.5 l of Sf9 cells infected with AcNPV, as previously described (Guarino et al., 1998b). Protein was loaded onto a 5-ml heparin column in 50 mM Tris (pH 7.9), 250 mM KCl, 2% taurine, 1 mM DTT. Bound proteins were eluted with a linear salt gradient from 250 to 625 mM KCl. Two milliliters fractions were collected and analyzed on 12% acrylamide-SDS gels stained with Coomassie Blue. The peak fraction of PP31 from heparin was dialyzed against 50 mM Hepes (pH 7.5), 100 mM KCl, 2% taurine, 1 mM DTT and loaded onto a
L.A. Guarino et al. / Virus Research 90 (2002) 187 /195
189
ratio of PP31 to template as in the corresponding DNA binding assays. Reaction products were separated on a 6 M urea, 6% polyacylamide gel, and reaction products were visualized and quantitated by phosphorimager analysis.
3. Results 3.1. Purification of PP31 Nuclear extracts prepared from Spodoptera frugiperda cell cultures infected with AcNPV contain an abundant protein an apparent molecular weight of 35 000. This protein elutes over a broad range from heparin affinity and cation exchange columns, and was a major contaminant of fractions containing RNA polymerase and the very late transcription factor called VLF-1 (Guarino et al., 1998b; McLachlin and Miller, 1994; Mistretta and Guarino, in preparation). The abundant 35 000 mol wt. protein also binds to Mono S and elutes over a broad range, compromising purification of baculovirus proteins by cation exchange chromatography (Mistretta and Guarino, in preparation). Therefore, we decided to identify the protein so that we could better design purification strategies to remove this contaminant. The protein sequence was obtained from a Mono S fraction during purification of VLF-1. This protein preparation was similar to that shown in Fig. 1a (lane 4) and Fig. 1. Purification of PP31. (A) SDS-PAGE analysis. Nuclear extract prepared from 2.5 l of infected cells (hep load; lane 3) was loaded onto a 5 ml heparin column and bound proteins were eluted with a linear salt gradient from 250 to 625 mM KCl. The peak fraction of PP31 from heparin (S load; lane 4) was dialyzed and loaded onto a Superose SP column. Bound proteins were eluted with a linear salt gradient from 100 to 550 mM KCl. The sole peak of UV absorbance corresponded to the peak of PP31, and is shown in lane 5 (S peak). Lane 2 contains crude nuclear extract prepared from uninfected cells. Each lane contains 2.5 mg of protein. The positions of protein molecular markers (lane 1) are shown in kilodaltons on the left. The arrow on the right indicates the position of PP31. Samples were separated on an 11% SDS-polyacrylamide gel and stained with Coomassie blue. (B) Immunoblot analysis. Aliquots of the same fractions were probed with rabbit antibody prepared against PP31 (Guarino et al., 1992).
Fig. 1
consisted of two bands, a major band and less abundant faster migrating bands that copurified with it. The sequence of the N-terminal 12 amino acids of major band was VNVPEQQSPEEA, and sequencing of the faster migrating bands yielded the sequence VNVPE. Both sequences correspond to the published N-terminal sequence of the viral protein PP31, minus the initiating methionine (Guarino and Smith, 1990). We have previously shown that PP31 is a phosphoprotein that exists in different phosphorylation states (Broussard et al.,
190
L.A. Guarino et al. / Virus Research 90 (2002) 187 /195
1996), suggesting that the differently migrating bands reflect hypo and hyper-phosphorylated versions of the protein. PP31 is a highly basic protein (pI /10.7) and so should bind to cation exchange columns with high affinity. The fact that it eluted over a broad range from both heparin and Mono S suggests that it either aggregates or associates with other proteins, thus weakening its affinity for the negativelycharged column matrices. To decrease the likelihood of these interactions and to improve separation, we included 2% taurine in all of the column buffers. Under these conditions, PP31 eluted from heparin in a sharp peak at 470 mM KCl. This is a very high salt concentration for heparin affinity chromatography, and thus the protein was nearly pure at this stage, as judged by Coomassie staining of SDS-PAGE gels (Fig. 1a). Taurine also helped to sharpen the elution profiles of RNApol and VLF-1, and addition of taurine to in vitro transcription assays at concentrations up to 2% had no effect on RNA polymerase activity (data not shown). Further purification of PP31 was performed on a Superose SP column. Again the viral protein eluted from the cation exchange column at very high salt (440 mM). This chromatography step removed all visible contaminants from the heparin column. Immunoblot analysis confirmed that all three protein bands correspond to PP31 (Fig. 1b). This rabbit antibody was raised against PP31 expressed in bacteria (Guarino et al., 1992), and so it is unlikely that the serum would react so strongly against non-related proteins. Minor immunoreactive bands were detected at approximately 64 kDa. We believe that these bands represent unresolved dimers of PP31 because their intensities are proportional to the amount of monomer PP31 in each lane. The total amount of PP31 purified from 2.5 l of infected cells was 4.9 mg from 95 mg of nuclear extract. Assuming a recovery of 100%, these figures indicate that PP31 accounts for 5% of total nuclear protein. We have previously hypothesized that PP31, which binds to both DNA and the virogenic stroma, serves as a scaffold for the assembly of viral replication factories in the nucleus (Guarino et al., 1992). The high intracel-
lular concentration of PP31 reported here is consistent with the idea that PP31 functions stoichiometrically and not catalytically. 3.2. Gel filtration of PP31 An aliquot of the Superose SP peak was filtered through Superdex 200 in the presence of 2% taurine (Fig. 2). A single peak of UV absorbance was detected at 10.17 ml. Extrapolating from a set of standards proteins filtered through the same column, we calculated an apparent molecular mass of 496 000 for PP31. This indicates that PP31 forms a higher order complex at physiological salt. Assuming that the conformation of the complex is globular, this would correspond to 16 monomers per oligomer. It is unlikely that the migration of PP31 was affected by contaminants in the protein prep because Coomassie staining is equivalent to protein concentration, and the absence of stained bands suggests that there are no contaminants of a high enough molar concentration to bind all of the PP31. 3.3. DNA binding activity of PP31 Previously we demonstrated that PP31 binds to DNA in a sequence non-specific manner by southwestern blot analysis (Guarino et al., 1992). The availability of purified protein allowed us to extend these analyses, and measure the relative affinity of PP31 for double-stranded and singlestranded DNA by gel electrophoretic mobility shift analysis (Fig. 3). The probe used in these experiments was a 5?-end labeled 44-mer, either used singly (panel B) or hybridized to its complement to make dsDNA (panel A). The DNA sequence of this probe corresponds to the viral DNA sequence surrounding the translation start site for lef-12 . Additional experiments using a non-viral fragment confirmed our previous experiments showing that PP31 binds in a sequencenonspecific manner (data not shown). The concentration of DNA in the binding reactions was held constant at 120 nM (a nucleotide concentration of 5 mM), and the protein concentration was varied in twofold increments from 0.01 to 2 mM. Bound and free probe were
L.A. Guarino et al. / Virus Research 90 (2002) 187 /195
191
Fig. 2. Gel filtration of PP31. (A) The peak fraction from Superose sp. chromatography was filtered through Superdex 200. The positions of elution of blue dextran 2000, thyroglobulin (T; 669 kDa), ferritin (F; 443 kDa), and catalase (C; 232 kDa) are indicated. (B) The fractions across the peak of UV absorbance were electrophoresed on an 11% SDS-polyacrylamide gel and visualized by staining with Coomassie blue. Lane 2 contains an aliquot of the Superose sp. peak (S) that was loaded onto the column. The position of PP31 is indicated on the right and the migration of protein molecular weight markers are indicated on the left.
then separated on native gels. Addition of protein caused a dramatic shift in the migration of the probe, so that the bound probe migrated near the top of the wells, and at the highest concentration
barely entered the gel. A plot of the fraction of free probe remaining as a function of input PP31 concentration is presented in Panel C. The ds and ss probes shifted to a slower migrating form at
192
L.A. Guarino et al. / Virus Research 90 (2002) 187 /195
Fig. 3. Binding of PP31 to ds (A) and ss (B) DNA. An oligonucleotide corresponding to the 5?-end of the LEF-12 gene was 5?-end labeled with (g-32P)ATP (Sambrook and Russell, 2001) (Lanes 2 /9), Lane 1, no PP31. DNA was held constant at 120 nM (a nucleotide concentration of 5 mM), and the protein concentration was varied in twofold increments from 0.01 to 2 mM. The positions of free and bound probe are indicated.
approximately the same input concentrations of PP31. This suggests that PP31 binds to ssDNA and dsDNA with equal affinities. To directly compare the binding of PP31 to ds and ssDNA, a competitive gel shift assay was performed. Equimolar amounts of both forms of DNA were mixed, and incubated with different concentrations of PP31 (Fig. 4). After separation of bound and free probes, the fraction of unbound probe was calculated and plotted as a function of PP31 concentration. The curves for singlestranded and double-stranded DNA were essentially superimposable (Fig. 4b), indicating that PP31 binds ds and ssDNA with equal affinities.
3.4. Effect of PP31 on in vitro transcription We next wanted to determine the effect of PP31 on transcription efficiency. For these experiments, a nucleotide-free transcription template linked to the very late polyhedrin promoter, called pPolh/ CFS, was used (Xu et al., 1995). To monitor the formation of DNA /protein complexes with pPolh/CFS, it was first linearized with EcoRI, and then incubated with concentrations of purified PP31 ranging from 0.06 to 4 mM. DNA /protein complexes were resolved on 0.5% agarose (Fig. 4a). Low concentrations of PP31 caused only a modest shift in the migration of the probe.
L.A. Guarino et al. / Virus Research 90 (2002) 187 /195
193
In vitro transcription reactions were performed using the same ratios of protein to nucleotide on an uncut DNA template. PP31 and pPolh/CFS were pre-incubated for 15 min on ice to allow for binding prior to addition of transcription mix. Analysis of the in vitro products revealed that transcription was inhibited by PP31 in a concentration dependent manner. Saturating amounts of PP31 completely inhibited transcription activity (Fig. 4b, c).
4. Discussion
Fig. 4. Competitive binding to ds and ssDNA. ssDNA and dsDNA oligonucleotides were mixed in equimolar proportions and titrated as described in the legend to Fig. 3. Panel C shows a plot of the fraction of unbound probe remaining at different concentrations of input PP31.
Increasing amounts caused a larger shift in a stepwise fashion, indicating that PP31 binds to DNA in a non-cooperative manner. Between 2 and 4 mM protein, there was only a slight decrease in the mobility of the shifted complexes. This suggests that the DNA template was saturated or nearly saturated. At this concentration, the ratio of protein to nucleotide was 7.5. Addition of higher amounts of protein resulted in the formation of visible precipitates and no visible DNA band on the gel, possibly due to aggregation of DNA-bound proteins (data not shown).
Here we report a method for the purification of native, untagged PP31 from baculovirus-infected cells. Previously, we tried to express PP31 in bacteria, but were unable to purify active protein due to the accumulation of PP31 in insoluble inclusion bodies (Guarino et al., 1992). The results reported here reveal that PP31 is highly expressed during normal infection of insect cells, so that use of an overexpression system or a purification tag is unnecessary. The availability of soluble protein allowed us to extend upon our previous studies on the DNA binding activity of PP31 (Guarino et al., 1992). We had previously shown that PP31 binds to DNA in a sequence non-specific manner, using southwestern analysis. This is a qualitative approach and measurements of binding affinity are not possible. Therefore, we used gel shift assays to directly compare the interactions of PP31 with double-stranded and single-stranded DNAs. These studies revealed that PP31 bound both forms with equal affinity. Addition of PP31 to in vitro transcription assays inhibited the activity of the viral RNA polymerase. This result is somewhat surprising considering that PP31 is required for transient expression of reporter genes under the control of late and very late promoters (Todd et al., 1995). The most likely explanation is that one or more of the other viral LEFs is needed for efficient transcription of DNA that is coated with PP31. The purification of PP31 reported here provides us with an assay to characterize the activities of additional LEFs on
194
L.A. Guarino et al. / Virus Research 90 (2002) 187 /195
Fig. 5. Transcription. (A) DNA binding. The transcription template pPolh/CFS was linearized with EcoRI before use. The standard assay mix contained 50 mM Tris (pH 7.9), 100 mM KCl, 0.1 mM EDTA, 5 mM DTT, 5% glycerol in a final reaction volume of 20 ml. The DNA concentration was held constant at 0.1 mg/ml (15 mM nucleotide), and the concentration of PP31 was varied from 0.06 to 4 mM in twofold increments. (B) Transcription assays. In vitro transcription assays were performed as previously described (Guarino et al., 1998b). Reaction products were separated on a 6 M urea, 6% polyacylamide gel. (C) Reaction products were quantitated by phosphorimager analysis. The amount of product obtained relative to that in the absence of pp31 is plotted versus input pp31.
a template that may more closely resemble the in vivo state.
PP31 is localized to the virogenic stroma (Guarino et al., 1992). This is the site of DNA replication and packaging (Fraser, 1986; Okano et al., 1999; Wilson and Price, 1988). The values for PP31 obtained here allow us to estimate the relative ratios of PP31 and viral DNA in the stroma. We obtained 4.9 mg of protein from 2.5 /106 cells, indicating that there are at least 38 million molecules of PP31 per cell. The actual numbers are probably higher, because only the peak fractions from each column were included in this calculation. This would be more than enough protein to coat 200 /500 copies of viral DNA at a ratio of one molecule of PP31 per 7.5 nt of DNA. Electrophoretic mobility shift assays suggest that this ratio of protein to DNA was sufficient to saturate DNA and to completely inhibit transcription activity. Previously, we reported that transcription activity in crude nuclear extracts of baculovirus-infected cells was not linear with respect to amount of protein added (Guarino et al., 1998b). Therefore, we were not able to calculate the specific activity of the crude extract. Heparin chromatography, however, removed the inhibitory compound, and the activity of the heparin peak was proportional to amount of protein added to assays. We speculated that the inhibition was due to the presence of proteins that bound DNA or inactivated the RNA polymerase. We now believe that the inhibition of RNA polymerase activity in crude extracts is primarily due to the presence of PP31. We were able to remove the inhibitor by passing extracts through heparin at high salt, and under these conditions, PP31 was the major protein bound (data not shown). Furthermore, addition of the PP31 fraction to the flow through containing RNA polymerase inhibits activity. We have previously shown that the viral RNA polymerase is composed of only four subunits, which together have the ability to recognize viral promoter, transcribed the linked gene, cap the 5?end of the RNA, terminate transcription at T-rich sequences, and polyadenylate the terminated transcript (Guarino et al., 1998a,b; Jin et al., 1998; Jin and Guarino, 2000). This accounts for all of the essential features of a transcription complex, yet there are still 4 additional factors that are required
L.A. Guarino et al. / Virus Research 90 (2002) 187 /195
for late transcription, and we need to identify essential roles for these proteins. LEF-5, for example, has homology with the transcription elongation factor TFIIS (Harwood et al., 1998). One of the activities attributed to TFIIS is the ability to help RNA polymerase transcribe through regions of the DNA template containing tightly bound proteins (Reines et al., 1993). Therefore, it will be of interest to determine whether LEF-5, or one of the other LEFs, increases transcription efficiency on PP31-coated DNA templates.
Acknowledgements This research was supported by the National Science Foundation under Grant No. 0110925.
References Broussard, D.R., Guarino, L.A., Jarvis, D.L., 1996. Dynamic phosphorylation of Autographa californica nuclear polyhedrosis virus pp31. J. Virol. 70, 6767 /6774. Fraser, M., 1986. Ultrastructural observations of virion maturation in Autographa californica nuclear polyhedrosis virus infected Spodoptera frugiperda cells. J. Ultrastruct. Mol. Struct. Res. 95, 189 /195. Guarino, L.A., Smith, M.W., 1990. Nucleotide sequence and characterization of the 39 K gene region of Autographa californica nuclear polyhedrosis virus. Virology 179, 1 /8. Guarino, L.A., Summers, M.D., 1986. Functional mapping of a trans -activating gene required for expression of a baculovirus delayed-early gene. J. Virol. 57, 563 /571. Guarino, L.A., Jin, J., Dong, W., 1998a. Guanylyltransferase activity of the LEF-4 subunit of baculovirus RNA polymerase. J. Virol. 72, 10003 /10010. Guarino, L.A., Xu, B., Jin, J., Dong, W., 1998b. A virusencoded RNA polymerase purified from baculovirus-infected cells. J. Virol. 72, 7985 /7991. Guarino, L.A., Dong, W., Xu, B., Broussard, D.R., Davis, R.W., Jarvis, D.L., 1992. Baculovirus phosphoprotein pp31 is associated with virogenic stroma. J. Virol. 66, 7113 /7120. Harwood, S.H., Li, L., Ho, P.S., Preston, A.K., Rohrmann, G.F., 1998. AcMNPV late expression factor-5 interacts with
195
itself and contains a zinc ribbon domain that is required for maximal late transcription activity and is homologous to elongation factor TFIIS. Virology 250, 118 /134. Jin, J., Guarino, L.A., 2000. 3?-end formation of baculovirus late RNAs. J. Virol. 74, 8930 /8937. Jin, J., Dong, W., Guarino, L.A., 1998. The LEF-4 subunit of baculovirus RNA polymerase has RNA 5?-triphosphatase and ATPase activities. J. Virol. 72, 10011 /10019. Kool, M., Voeten, J.T., Goldbach, R.W., Vlak, J.M., 1994. Functional mapping of regions of the Autographa californica nuclear polyhedrosis viral genome required for DNA replication. Virology 198, 680 /689. Lin, G., Blissard, G.W., 2002. Analysis of an Autographa californica nucleopolyhedrovirus lef-11 knockout: LEF-11 is essential for viral DNA replication. J. Virol. 76, 2770 / 2779. Lu, A., Miller, L.K., 1995. The roles of eighteen baculovirus late expression factor genes in transcription and DNA replication. J. Virol. 69, 975 /982. McLachlin, J.R., Miller, L.K., 1994. Identification and characterization of vlf-1, a baculovirus gene involved in very late gene expression. J. Virol. 68, 7746 /7756. Mistretta, T.A., Guarino, L.A., Functional Analysis of Transcription Activity of Autographa californica nucleopolyhedrosis virus Very Late Factor 1, in preparation. Okano, K., Mikhailov, V.S., Maeda, S., 1999. Colocalization of baculovirus IE-1 and two DNA-binding proteins, DBP and LEF-3, to viral replication factories. J. Virol. 73, 110 / 119. Rapp, J.C., Wilson, J.A., Miller, L.K., 1998. Nineteen baculovirus open reading frames, including LEF-12, support late gene expression. J. Virol. 72, 10197 /10206. Reines, D., Mote, J., Jr., 1993. Elongation factor SII-dependent transcription by RNA polymerase II through a sequencespecific DNA-binding protein. Proc. Natl. Acad. Sci. USA 90, 1917 /1921. Sambrook, J., Russell, D.W., 2001. Molecular Cloning. A Laboratory Manual, Third ed.. Cold Spring Harbor Laboratory Press, Cold Spring Harbor. Todd, J.W., Passarelli, A.L., Miller, L.K., 1995. Eighteen baculovirus genes, including lef-11, p35, 39K, and p47, support late gene expression. J. Virol. 69, 968 /974. Wilson, M.E., Price, K.H., 1988. Association of Autographa californica nuclear polyhedrosis virus (AcMNPV) with the nuclear matrix. Virology 167, 233 /241. Xu, B., Yoo, S., Guarino, L.A., 1995. Differential transcription of baculovirus late and very late promoters: fractionation of nuclear extracts by phosphocellulose chromatography. J. Virol. 69, 2912 /2917.
DNA binding activity of the baculovirus late expression factor PP31 Linda A. Guarino a,b,*, Toni-Ann Mistretta a, Wen Dong b a
Department of Biochemistry and Biophysics, Texas A&M University, College Station, TX 77843-2128, USA b Department of Entomology, Texas A&M University, College Station, TX 77843-2128, USA Received 12 April 2002; received in revised form 18 July 2002; accepted 18 July 2002
Abstract PP31 is a baculovirus protein that is essential for viral late gene expression. To study the role of PP31 in late transcription in vitro, it was purified from infected insect cells. A combination of heparin affinity, cation exchange chromatography, and gel filtration was used to purify native non-tagged protein. Nearly 5 mg of PP31 was obtained from 95 mg of nuclear extract confirming that PP31 is an abundant viral protein. DNA binding assays revealed that PP31 binds to single-stranded and double-stranded DNA with equal affinities. Addition of PP31 to in vitro transcription assays with purified baculovirus RNA polymerase resulted in a strong inhibition of transcription. This indicates that the viral RNA polymerase was not able to displace PP31, and suggests that other late expression factors may function to help RNA polymerase bind to PP31-coated templates. # 2002 Elsevier Science B.V. All rights reserved. Keywords: Function; Purification; Transcription; Baculovirus; Gene regulation
1. Introduction Nineteen viral genes are required for expression of baculovirus late genes in a transient expression assay (Rapp et al., 1998). These 19 genes are known as late expression factors (LEFs). In baculoviruses, late gene expression requires early gene expression and viral DNA replication, and so these 19 genes include factors necessary for early events as well as those directly involved in late
* Corresponding author. Tel.: /1-979-845-7556; fax: /1979-845-9274 E-mail address: [email protected] (L.A. Guarino).
gene expression. Ten of the LEFs are required for early gene expression and DNA replication expression. These are ie-1 , ie-2 , dnapol , p143, p35, lef-1, lef-2, lef-3, lef-7, and lef-11 (Kool et al., 1994; Lin and Blissard, 2002; Lu and Miller, 1995). It is currently unknown whether these factors function solely during the early phase or whether they also have roles in late transcription. The remaining 9 factors are believed to be directly involved in late gene transcription. LEF4, LEF-8, LEF-9, and P47 make up the viralencoded RNA polymerase (Guarino et al., 1998b). LEF-5 has a ribbon domain common to the transcription elongation factor TFII-S, and has been predicted to function in elongation (Har-
0168-1702/02/$ - see front matter # 2002 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 8 - 1 7 0 2 ( 0 2 ) 0 0 1 5 2 - 1
188
L.A. Guarino et al. / Virus Research 90 (2002) 187 /195
wood et al., 1998). LEF-6 and LEF-12 have no obvious sequence homologies with non-baculovirus proteins and little is known concerning their potential functions in late expression. PP31 is a phosphoprotein that binds DNA non-specifically (Guarino et al., 1992). PP31 also interacts with the virogenic stroma, which is an altered form of the nuclear matrix formed at the site of viral DNA replication and packaging (Fraser, 1986; Okano et al., 1999; Wilson and Price, 1988). pp31 is highly expressed and is regulated by tandem early and late promoters (Guarino and Summers, 1986). Together, these observations suggest that PP31 has a structural role in viral life cycle, and may serve to organize the viral replication factories and late transcription apparatus. To analyze the functions of late transcription factors, we have attempted to purify all of the essential LEFs and analyze the effect of adding these factors to in vitro transcription assays containing purified baculovirus RNA polymerase. To this end, we developed a simple scheme for purification of PP31 from Spodoptera frugiperda cells infected with Autographa californica nuclear polyhedrosis virus (AcNPV). Here we report that PP31 bound single and double-stranded DNA with equal affinities. Furthermore, we found that addition of PP31 to in vitro transcription assays with purified baculovirus RNA polymerase resulted in an inhibition of transcription activity.
Superose sp. column. Bound proteins were eluted with a linear salt gradient from 100 to 550 mM KCl. The peak fraction from Superose sp. chromatography was filtered through Superdex 200. Molecular standards were filtered through the same column in the same buffer to establish a standard curve. Purified protein was dialyzed against 50 mM Hepes (pH 7.2), 100 mM KCl, 1 mM DTT, 20% glycerol frozen in liquid nitrogen droplets and stored at /80 8C. 2.2. DNA /protein interactions An oligonucleotide corresponding to the 5?-end of the LEF-12 (nucleotides 33361 /33402 on the genomic map) gene was 5?-end labeled with (g-32P)ATP (Sambrook and Russell, 2001). To make double-stranded probe, the probe was mixed with an equimolar amount of the unlabeled complementary strand. The single-stranded probe was mixed with an equimolar amount of the same unlabeled oligo. In all reactions, the DNA was held constant at 120 nM for dsDNA or 240 nM for ssDNA (a nucleotide concentration of 5 mM in both cases). The protein concentration was varied in twofold increments from 0.01 to 2 mM. All reactions were normalized with BSA diluted in enzyme storage buffer so that the buffer concentrations and total protein amounts were the same in every tube.
2. Materials and methods
2.3. In vitro transcription assays
2.1. Purification of PP31
The transcription template pPolh/CFS was linearized with EcoRI before use in DNA binding assays. The standard assay mix contained 50 mM Tris (pH 7.9), 100 mM KCl, 0.1 mM EDTA, 5 mM DTT, 5% glycerol in a final reaction volume of 20 ml. The DNA concentration was held constant at 0.1 mg/ml (15 mM nucleotide), and the concentration of PP31 was varied from 0.06 to 4 mM in twofold increments. All reactions were normalized with BSA diluted in enzyme storage buffer so that the buffer concentrations and total protein amounts were the same in every tube. In vitro transcription assays were performed as previously described (Xu et al., 1995) using the same
Nuclear extract was prepared from 2.5 l of Sf9 cells infected with AcNPV, as previously described (Guarino et al., 1998b). Protein was loaded onto a 5-ml heparin column in 50 mM Tris (pH 7.9), 250 mM KCl, 2% taurine, 1 mM DTT. Bound proteins were eluted with a linear salt gradient from 250 to 625 mM KCl. Two milliliters fractions were collected and analyzed on 12% acrylamide-SDS gels stained with Coomassie Blue. The peak fraction of PP31 from heparin was dialyzed against 50 mM Hepes (pH 7.5), 100 mM KCl, 2% taurine, 1 mM DTT and loaded onto a
L.A. Guarino et al. / Virus Research 90 (2002) 187 /195
189
ratio of PP31 to template as in the corresponding DNA binding assays. Reaction products were separated on a 6 M urea, 6% polyacylamide gel, and reaction products were visualized and quantitated by phosphorimager analysis.
3. Results 3.1. Purification of PP31 Nuclear extracts prepared from Spodoptera frugiperda cell cultures infected with AcNPV contain an abundant protein an apparent molecular weight of 35 000. This protein elutes over a broad range from heparin affinity and cation exchange columns, and was a major contaminant of fractions containing RNA polymerase and the very late transcription factor called VLF-1 (Guarino et al., 1998b; McLachlin and Miller, 1994; Mistretta and Guarino, in preparation). The abundant 35 000 mol wt. protein also binds to Mono S and elutes over a broad range, compromising purification of baculovirus proteins by cation exchange chromatography (Mistretta and Guarino, in preparation). Therefore, we decided to identify the protein so that we could better design purification strategies to remove this contaminant. The protein sequence was obtained from a Mono S fraction during purification of VLF-1. This protein preparation was similar to that shown in Fig. 1a (lane 4) and Fig. 1. Purification of PP31. (A) SDS-PAGE analysis. Nuclear extract prepared from 2.5 l of infected cells (hep load; lane 3) was loaded onto a 5 ml heparin column and bound proteins were eluted with a linear salt gradient from 250 to 625 mM KCl. The peak fraction of PP31 from heparin (S load; lane 4) was dialyzed and loaded onto a Superose SP column. Bound proteins were eluted with a linear salt gradient from 100 to 550 mM KCl. The sole peak of UV absorbance corresponded to the peak of PP31, and is shown in lane 5 (S peak). Lane 2 contains crude nuclear extract prepared from uninfected cells. Each lane contains 2.5 mg of protein. The positions of protein molecular markers (lane 1) are shown in kilodaltons on the left. The arrow on the right indicates the position of PP31. Samples were separated on an 11% SDS-polyacrylamide gel and stained with Coomassie blue. (B) Immunoblot analysis. Aliquots of the same fractions were probed with rabbit antibody prepared against PP31 (Guarino et al., 1992).
Fig. 1
consisted of two bands, a major band and less abundant faster migrating bands that copurified with it. The sequence of the N-terminal 12 amino acids of major band was VNVPEQQSPEEA, and sequencing of the faster migrating bands yielded the sequence VNVPE. Both sequences correspond to the published N-terminal sequence of the viral protein PP31, minus the initiating methionine (Guarino and Smith, 1990). We have previously shown that PP31 is a phosphoprotein that exists in different phosphorylation states (Broussard et al.,
190
L.A. Guarino et al. / Virus Research 90 (2002) 187 /195
1996), suggesting that the differently migrating bands reflect hypo and hyper-phosphorylated versions of the protein. PP31 is a highly basic protein (pI /10.7) and so should bind to cation exchange columns with high affinity. The fact that it eluted over a broad range from both heparin and Mono S suggests that it either aggregates or associates with other proteins, thus weakening its affinity for the negativelycharged column matrices. To decrease the likelihood of these interactions and to improve separation, we included 2% taurine in all of the column buffers. Under these conditions, PP31 eluted from heparin in a sharp peak at 470 mM KCl. This is a very high salt concentration for heparin affinity chromatography, and thus the protein was nearly pure at this stage, as judged by Coomassie staining of SDS-PAGE gels (Fig. 1a). Taurine also helped to sharpen the elution profiles of RNApol and VLF-1, and addition of taurine to in vitro transcription assays at concentrations up to 2% had no effect on RNA polymerase activity (data not shown). Further purification of PP31 was performed on a Superose SP column. Again the viral protein eluted from the cation exchange column at very high salt (440 mM). This chromatography step removed all visible contaminants from the heparin column. Immunoblot analysis confirmed that all three protein bands correspond to PP31 (Fig. 1b). This rabbit antibody was raised against PP31 expressed in bacteria (Guarino et al., 1992), and so it is unlikely that the serum would react so strongly against non-related proteins. Minor immunoreactive bands were detected at approximately 64 kDa. We believe that these bands represent unresolved dimers of PP31 because their intensities are proportional to the amount of monomer PP31 in each lane. The total amount of PP31 purified from 2.5 l of infected cells was 4.9 mg from 95 mg of nuclear extract. Assuming a recovery of 100%, these figures indicate that PP31 accounts for 5% of total nuclear protein. We have previously hypothesized that PP31, which binds to both DNA and the virogenic stroma, serves as a scaffold for the assembly of viral replication factories in the nucleus (Guarino et al., 1992). The high intracel-
lular concentration of PP31 reported here is consistent with the idea that PP31 functions stoichiometrically and not catalytically. 3.2. Gel filtration of PP31 An aliquot of the Superose SP peak was filtered through Superdex 200 in the presence of 2% taurine (Fig. 2). A single peak of UV absorbance was detected at 10.17 ml. Extrapolating from a set of standards proteins filtered through the same column, we calculated an apparent molecular mass of 496 000 for PP31. This indicates that PP31 forms a higher order complex at physiological salt. Assuming that the conformation of the complex is globular, this would correspond to 16 monomers per oligomer. It is unlikely that the migration of PP31 was affected by contaminants in the protein prep because Coomassie staining is equivalent to protein concentration, and the absence of stained bands suggests that there are no contaminants of a high enough molar concentration to bind all of the PP31. 3.3. DNA binding activity of PP31 Previously we demonstrated that PP31 binds to DNA in a sequence non-specific manner by southwestern blot analysis (Guarino et al., 1992). The availability of purified protein allowed us to extend these analyses, and measure the relative affinity of PP31 for double-stranded and singlestranded DNA by gel electrophoretic mobility shift analysis (Fig. 3). The probe used in these experiments was a 5?-end labeled 44-mer, either used singly (panel B) or hybridized to its complement to make dsDNA (panel A). The DNA sequence of this probe corresponds to the viral DNA sequence surrounding the translation start site for lef-12 . Additional experiments using a non-viral fragment confirmed our previous experiments showing that PP31 binds in a sequencenonspecific manner (data not shown). The concentration of DNA in the binding reactions was held constant at 120 nM (a nucleotide concentration of 5 mM), and the protein concentration was varied in twofold increments from 0.01 to 2 mM. Bound and free probe were
L.A. Guarino et al. / Virus Research 90 (2002) 187 /195
191
Fig. 2. Gel filtration of PP31. (A) The peak fraction from Superose sp. chromatography was filtered through Superdex 200. The positions of elution of blue dextran 2000, thyroglobulin (T; 669 kDa), ferritin (F; 443 kDa), and catalase (C; 232 kDa) are indicated. (B) The fractions across the peak of UV absorbance were electrophoresed on an 11% SDS-polyacrylamide gel and visualized by staining with Coomassie blue. Lane 2 contains an aliquot of the Superose sp. peak (S) that was loaded onto the column. The position of PP31 is indicated on the right and the migration of protein molecular weight markers are indicated on the left.
then separated on native gels. Addition of protein caused a dramatic shift in the migration of the probe, so that the bound probe migrated near the top of the wells, and at the highest concentration
barely entered the gel. A plot of the fraction of free probe remaining as a function of input PP31 concentration is presented in Panel C. The ds and ss probes shifted to a slower migrating form at
192
L.A. Guarino et al. / Virus Research 90 (2002) 187 /195
Fig. 3. Binding of PP31 to ds (A) and ss (B) DNA. An oligonucleotide corresponding to the 5?-end of the LEF-12 gene was 5?-end labeled with (g-32P)ATP (Sambrook and Russell, 2001) (Lanes 2 /9), Lane 1, no PP31. DNA was held constant at 120 nM (a nucleotide concentration of 5 mM), and the protein concentration was varied in twofold increments from 0.01 to 2 mM. The positions of free and bound probe are indicated.
approximately the same input concentrations of PP31. This suggests that PP31 binds to ssDNA and dsDNA with equal affinities. To directly compare the binding of PP31 to ds and ssDNA, a competitive gel shift assay was performed. Equimolar amounts of both forms of DNA were mixed, and incubated with different concentrations of PP31 (Fig. 4). After separation of bound and free probes, the fraction of unbound probe was calculated and plotted as a function of PP31 concentration. The curves for singlestranded and double-stranded DNA were essentially superimposable (Fig. 4b), indicating that PP31 binds ds and ssDNA with equal affinities.
3.4. Effect of PP31 on in vitro transcription We next wanted to determine the effect of PP31 on transcription efficiency. For these experiments, a nucleotide-free transcription template linked to the very late polyhedrin promoter, called pPolh/ CFS, was used (Xu et al., 1995). To monitor the formation of DNA /protein complexes with pPolh/CFS, it was first linearized with EcoRI, and then incubated with concentrations of purified PP31 ranging from 0.06 to 4 mM. DNA /protein complexes were resolved on 0.5% agarose (Fig. 4a). Low concentrations of PP31 caused only a modest shift in the migration of the probe.
L.A. Guarino et al. / Virus Research 90 (2002) 187 /195
193
In vitro transcription reactions were performed using the same ratios of protein to nucleotide on an uncut DNA template. PP31 and pPolh/CFS were pre-incubated for 15 min on ice to allow for binding prior to addition of transcription mix. Analysis of the in vitro products revealed that transcription was inhibited by PP31 in a concentration dependent manner. Saturating amounts of PP31 completely inhibited transcription activity (Fig. 4b, c).
4. Discussion
Fig. 4. Competitive binding to ds and ssDNA. ssDNA and dsDNA oligonucleotides were mixed in equimolar proportions and titrated as described in the legend to Fig. 3. Panel C shows a plot of the fraction of unbound probe remaining at different concentrations of input PP31.
Increasing amounts caused a larger shift in a stepwise fashion, indicating that PP31 binds to DNA in a non-cooperative manner. Between 2 and 4 mM protein, there was only a slight decrease in the mobility of the shifted complexes. This suggests that the DNA template was saturated or nearly saturated. At this concentration, the ratio of protein to nucleotide was 7.5. Addition of higher amounts of protein resulted in the formation of visible precipitates and no visible DNA band on the gel, possibly due to aggregation of DNA-bound proteins (data not shown).
Here we report a method for the purification of native, untagged PP31 from baculovirus-infected cells. Previously, we tried to express PP31 in bacteria, but were unable to purify active protein due to the accumulation of PP31 in insoluble inclusion bodies (Guarino et al., 1992). The results reported here reveal that PP31 is highly expressed during normal infection of insect cells, so that use of an overexpression system or a purification tag is unnecessary. The availability of soluble protein allowed us to extend upon our previous studies on the DNA binding activity of PP31 (Guarino et al., 1992). We had previously shown that PP31 binds to DNA in a sequence non-specific manner, using southwestern analysis. This is a qualitative approach and measurements of binding affinity are not possible. Therefore, we used gel shift assays to directly compare the interactions of PP31 with double-stranded and single-stranded DNAs. These studies revealed that PP31 bound both forms with equal affinity. Addition of PP31 to in vitro transcription assays inhibited the activity of the viral RNA polymerase. This result is somewhat surprising considering that PP31 is required for transient expression of reporter genes under the control of late and very late promoters (Todd et al., 1995). The most likely explanation is that one or more of the other viral LEFs is needed for efficient transcription of DNA that is coated with PP31. The purification of PP31 reported here provides us with an assay to characterize the activities of additional LEFs on
194
L.A. Guarino et al. / Virus Research 90 (2002) 187 /195
Fig. 5. Transcription. (A) DNA binding. The transcription template pPolh/CFS was linearized with EcoRI before use. The standard assay mix contained 50 mM Tris (pH 7.9), 100 mM KCl, 0.1 mM EDTA, 5 mM DTT, 5% glycerol in a final reaction volume of 20 ml. The DNA concentration was held constant at 0.1 mg/ml (15 mM nucleotide), and the concentration of PP31 was varied from 0.06 to 4 mM in twofold increments. (B) Transcription assays. In vitro transcription assays were performed as previously described (Guarino et al., 1998b). Reaction products were separated on a 6 M urea, 6% polyacylamide gel. (C) Reaction products were quantitated by phosphorimager analysis. The amount of product obtained relative to that in the absence of pp31 is plotted versus input pp31.
a template that may more closely resemble the in vivo state.
PP31 is localized to the virogenic stroma (Guarino et al., 1992). This is the site of DNA replication and packaging (Fraser, 1986; Okano et al., 1999; Wilson and Price, 1988). The values for PP31 obtained here allow us to estimate the relative ratios of PP31 and viral DNA in the stroma. We obtained 4.9 mg of protein from 2.5 /106 cells, indicating that there are at least 38 million molecules of PP31 per cell. The actual numbers are probably higher, because only the peak fractions from each column were included in this calculation. This would be more than enough protein to coat 200 /500 copies of viral DNA at a ratio of one molecule of PP31 per 7.5 nt of DNA. Electrophoretic mobility shift assays suggest that this ratio of protein to DNA was sufficient to saturate DNA and to completely inhibit transcription activity. Previously, we reported that transcription activity in crude nuclear extracts of baculovirus-infected cells was not linear with respect to amount of protein added (Guarino et al., 1998b). Therefore, we were not able to calculate the specific activity of the crude extract. Heparin chromatography, however, removed the inhibitory compound, and the activity of the heparin peak was proportional to amount of protein added to assays. We speculated that the inhibition was due to the presence of proteins that bound DNA or inactivated the RNA polymerase. We now believe that the inhibition of RNA polymerase activity in crude extracts is primarily due to the presence of PP31. We were able to remove the inhibitor by passing extracts through heparin at high salt, and under these conditions, PP31 was the major protein bound (data not shown). Furthermore, addition of the PP31 fraction to the flow through containing RNA polymerase inhibits activity. We have previously shown that the viral RNA polymerase is composed of only four subunits, which together have the ability to recognize viral promoter, transcribed the linked gene, cap the 5?end of the RNA, terminate transcription at T-rich sequences, and polyadenylate the terminated transcript (Guarino et al., 1998a,b; Jin et al., 1998; Jin and Guarino, 2000). This accounts for all of the essential features of a transcription complex, yet there are still 4 additional factors that are required
L.A. Guarino et al. / Virus Research 90 (2002) 187 /195
for late transcription, and we need to identify essential roles for these proteins. LEF-5, for example, has homology with the transcription elongation factor TFIIS (Harwood et al., 1998). One of the activities attributed to TFIIS is the ability to help RNA polymerase transcribe through regions of the DNA template containing tightly bound proteins (Reines et al., 1993). Therefore, it will be of interest to determine whether LEF-5, or one of the other LEFs, increases transcription efficiency on PP31-coated DNA templates.
Acknowledgements This research was supported by the National Science Foundation under Grant No. 0110925.
References Broussard, D.R., Guarino, L.A., Jarvis, D.L., 1996. Dynamic phosphorylation of Autographa californica nuclear polyhedrosis virus pp31. J. Virol. 70, 6767 /6774. Fraser, M., 1986. Ultrastructural observations of virion maturation in Autographa californica nuclear polyhedrosis virus infected Spodoptera frugiperda cells. J. Ultrastruct. Mol. Struct. Res. 95, 189 /195. Guarino, L.A., Smith, M.W., 1990. Nucleotide sequence and characterization of the 39 K gene region of Autographa californica nuclear polyhedrosis virus. Virology 179, 1 /8. Guarino, L.A., Summers, M.D., 1986. Functional mapping of a trans -activating gene required for expression of a baculovirus delayed-early gene. J. Virol. 57, 563 /571. Guarino, L.A., Jin, J., Dong, W., 1998a. Guanylyltransferase activity of the LEF-4 subunit of baculovirus RNA polymerase. J. Virol. 72, 10003 /10010. Guarino, L.A., Xu, B., Jin, J., Dong, W., 1998b. A virusencoded RNA polymerase purified from baculovirus-infected cells. J. Virol. 72, 7985 /7991. Guarino, L.A., Dong, W., Xu, B., Broussard, D.R., Davis, R.W., Jarvis, D.L., 1992. Baculovirus phosphoprotein pp31 is associated with virogenic stroma. J. Virol. 66, 7113 /7120. Harwood, S.H., Li, L., Ho, P.S., Preston, A.K., Rohrmann, G.F., 1998. AcMNPV late expression factor-5 interacts with
195
itself and contains a zinc ribbon domain that is required for maximal late transcription activity and is homologous to elongation factor TFIIS. Virology 250, 118 /134. Jin, J., Guarino, L.A., 2000. 3?-end formation of baculovirus late RNAs. J. Virol. 74, 8930 /8937. Jin, J., Dong, W., Guarino, L.A., 1998. The LEF-4 subunit of baculovirus RNA polymerase has RNA 5?-triphosphatase and ATPase activities. J. Virol. 72, 10011 /10019. Kool, M., Voeten, J.T., Goldbach, R.W., Vlak, J.M., 1994. Functional mapping of regions of the Autographa californica nuclear polyhedrosis viral genome required for DNA replication. Virology 198, 680 /689. Lin, G., Blissard, G.W., 2002. Analysis of an Autographa californica nucleopolyhedrovirus lef-11 knockout: LEF-11 is essential for viral DNA replication. J. Virol. 76, 2770 / 2779. Lu, A., Miller, L.K., 1995. The roles of eighteen baculovirus late expression factor genes in transcription and DNA replication. J. Virol. 69, 975 /982. McLachlin, J.R., Miller, L.K., 1994. Identification and characterization of vlf-1, a baculovirus gene involved in very late gene expression. J. Virol. 68, 7746 /7756. Mistretta, T.A., Guarino, L.A., Functional Analysis of Transcription Activity of Autographa californica nucleopolyhedrosis virus Very Late Factor 1, in preparation. Okano, K., Mikhailov, V.S., Maeda, S., 1999. Colocalization of baculovirus IE-1 and two DNA-binding proteins, DBP and LEF-3, to viral replication factories. J. Virol. 73, 110 / 119. Rapp, J.C., Wilson, J.A., Miller, L.K., 1998. Nineteen baculovirus open reading frames, including LEF-12, support late gene expression. J. Virol. 72, 10197 /10206. Reines, D., Mote, J., Jr., 1993. Elongation factor SII-dependent transcription by RNA polymerase II through a sequencespecific DNA-binding protein. Proc. Natl. Acad. Sci. USA 90, 1917 /1921. Sambrook, J., Russell, D.W., 2001. Molecular Cloning. A Laboratory Manual, Third ed.. Cold Spring Harbor Laboratory Press, Cold Spring Harbor. Todd, J.W., Passarelli, A.L., Miller, L.K., 1995. Eighteen baculovirus genes, including lef-11, p35, 39K, and p47, support late gene expression. J. Virol. 69, 968 /974. Wilson, M.E., Price, K.H., 1988. Association of Autographa californica nuclear polyhedrosis virus (AcMNPV) with the nuclear matrix. Virology 167, 233 /241. Xu, B., Yoo, S., Guarino, L.A., 1995. Differential transcription of baculovirus late and very late promoters: fractionation of nuclear extracts by phosphocellulose chromatography. J. Virol. 69, 2912 /2917.