|Home | About | Journals | Submit | Contact Us | Français|
The Bombyx mori nucleopolyhedrovirus (BmNPV) genome contains five related members of the bro gene family, all of which are actively expressed in infected BmN cells. Although their functions are unknown, their amino acid sequences contain a motif found in all known viral and prokaryotic single-stranded DNA binding proteins. To determine if they bind to nucleic acids, we fractionated the nuclei of BmNPV-infected BmN cells using a histone extraction protocol. We detected BRO-A, BRO-C, and BRO-D in the histone H1 fraction using anti-BRO antibodies. Micrococcal nuclease treatment released these BRO proteins from the chromatin fraction, suggesting their involvement in nucleosome structures. Chromatographic fractionation showed that BRO-A and/or BRO-C interacted with core histones. Expression of partial sequences of BRO-A proved that the N-terminal 80 amino acid residues were required for DNA binding activity. We also demonstrated that BmNPV BRO proteins underwent phosphorylation and ubiquitination followed by proteasome degradation, which may explain their distribution in the cytoplasm as well as the nucleus. We propose that BRO-A and BRO-C may function as DNA binding proteins that influence host DNA replication and/or transcription.
Bombyx mori nucleopolyhedrovirus (BmNPV) belongs to the Baculoviridae, a large family of viruses with double-stranded (ds) DNA genomes that are pathogenic mainly for lepidopteran insects. The BmNPV genome is about 128 kb in length and is predicted to contain 136 open reading frames (ORFs) (9). Among these ORFs, five genes (bro-a, bro-b, bro-c, bro-d, and bro-e) were found to belong to a unique baculovirus multigene family, since they demonstrated high homology to each other (13). Multiple members of this gene family have been reported in the genomes of the Orgyia pseudotsugata NPV, Lymantria dispar NPV (LdNPV), and Xestia c-nigrum granulovirus (1, 11, 16). However, the well-characterized Autographa californica MNPV (AcMNPV) genome contains only a single member (ORF2), which is related to BmNPV bro-d (2, 13) with 80% amino acid sequence identity. This is much lower than the average identity of predicted proteins from these two viruses, which is over 93% (9). In addition, NPV pathogenic for Spodoptera exigua lacks a bro homolog (12).
Most bro genes share a related core sequence and demonstrate differing degrees of similarity in other regions (16). Although BmNPV BRO proteins show high homology within the family and with other baculovirus BROs, they have no strong similarity with any known proteins. Thus, it has been difficult to predict their function during the viral infection cycle.
Recently, we reported that all bro genes of BmNPV are actively transcribed as delayed-early genes and that BRO proteins are produced at high levels between 8 and 14 h postinfection (p.i.) (13). We also reported that one BmNPV bro gene (bro-d) is essential for viral infection and that bro-a and bro-c may functionally complement each other (13). Since our immunohistochemical analysis using confocal microscopy showed nuclear localization of BRO proteins, we investigated whether they were able to bind to DNA. In this report, we describe that BRO-A and BRO-C are novel DNA binding proteins that show a stronger affinity for single-stranded (ss) DNA than for dsDNA.
Nuclei were isolated from 2 × 107 BmN cells as described by Durandel et al. (7). The purified nuclei were subjected to histone extraction with 20 mM Tris-HCl (pH 8.0) containing 75 mM NaCl–25 mM EDTA (step a), 350 mM NaCl (step b), or 600 mM NaCl (step c) and then 0.2 M H2SO4 (step d) (6). Extractions were performed twice. Micrococcal nuclease (MN) (Worthington Biochemical Corp.) treatment was introduced between steps b and c for 15 min at room temperature (RT). Aliquots of the collected fractions were precipitated with 20% trichloroacetic acid (final concentration) in the presence of bovine serum albumin (10 μg) and subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) followed by Western blot hybridization.
BmNPV-infected BmN cells (107) were collected at 14 h p.i. and extracted by sonication with 1.5 ml of extraction buffer containing 20 mM Tris-HCl (pH 7.5), 2 M NaCl, 2 mM EDTA, and 0.5% Nonidet P-40 (NP-40). Cell debris was removed by centrifugation at 15,000 × g for 30 min, and the supernatant (cell extract) was used for column chromatography. Three columns with 0.75 ml of ssDNA- or dsDNA-cellulose or poly(U)-agarose (Sigma Aldrich) were equilibrated with elution buffer I (20 mM Tris-HCl [pH 7.5], 2 mM EDTA, 0.1% NP-40, and 10% glycerol, containing 0.2 M NaCl). The cell extract was diluted to 0.2 M NaCl with elution buffer I and loaded onto the columns. Each column was washed with 5 column volumes of elution buffer I containing 0.2 M NaCl, and then elution buffer I (5 column volumes each) containing NaCl at final concentrations of 0.3, 0.5, 0.7, 0.9, 1.2, 1.5, and 2.0 M was applied. Proteins from each fraction were precipitated by trichloroacetic acid in the presence of bovine serum albumin (20 μg) and analyzed by SDS-PAGE and Western blotting. For histone-agarose column chromatography, the nuclear fraction extracted with 600 mM NaCl (see above) was treated with MN (360 U) for 30 min at RT, diluted to 0.05 M NaCl with elution buffer II (elution buffer I without NP-40), and applied to a 0.25-ml histone-agarose column (Sigma Aldrich) equilibrated with elution buffer II. The column was treated as described above, and fractions were collected at NaCl concentrations of 0.1, 0.15, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 1.0, and 2.0 M.
BmNPV-infected BmN cells (4 × 106) were incubated in the presence of 100 μCi of [32P]H3PO4 (ICN Radiochemicals)/ml for 4 h at 8, 14, 20, 26, 36, and 48 h p.i. The cells were collected by centrifugation and extracted with 0.25 ml of buffer containing 40 mM Tris-HCl (pH 7.5), 500 mM NaCl, 2 mM EDTA, 0.5% SDS, and 50 mM NaF for 1 h on ice with vortexing. After centrifugation at 7,000 × g for 30 min, the supernatants were used for immunoprecipitation. To prepare the immunoaffinity matrix, 2 μl of BRO-A antiserum was diluted with 1 ml of TBS-T (20 mM Tris-HCl [pH 7.5], 135 mM NaCl, 0.1% Tween-20) and mixed with 25 μl of a 50% slurry of protein A-Sepharose CL-4B (Sigma). After 1 h of incubation at RT with gentle mixing, the Protein A complex was collected by brief centrifugation. Half the volume of the cell extracts was diluted with 500 μl of buffer A (40 mM Tris-HCl [pH 7.5], 500 mM NaCl, 2 mM EDTA, 0.5% NP-40, 50 mM NaF) and mixed with 15 μl of the protein A–anti-BRO complex. This was incubated for 1 h at RT with gentle mixing, and the immunoprecipitates were collected by a brief centrifugation. The immunoprecipitates were then washed twice with buffer A containing 0.1% SDS for 30 min at RT. After brief centrifugation, the immunoprecipitates were suspended in SDS sample loading buffer and analyzed by SDS–8% PAGE. The gels were dried under vacuum, and the labeled bands were exposed to X-ray film (Kodak X-OMAT AR).
BmN cells infected with BmNPV were incubated in the presence of proteasome inhibitor MG-132 (Calbiochem) at appropriate concentrations for 6 h prior to harvest and harvested at 16 h p.i. The cell homogenates were analyzed by SDS-PAGE followed by Western blot hybridization.
BmNPV bro-a fragments were amplified by PCR using combinations of primers. Sense and antisense primers were designed with EcoRI and SalI sites, respectively (Table (Table1).1). The amplified fragment was digested with EcoRI and SalI and then inserted into pET28a(+) (Novagen). The resulting plasmids were transformed into Escherichia coli BL21 (DE3) LysE. Recombinant proteins (Novagen) were expressed following the manufacturer's manual. The cells from a 1.5-ml Luria broth culture were collected by centrifugation and extracted with 0.5 ml of buffer (50 mM Tris-HCl [pH 7.5], 0.2 M NaCl, 2 mM EDTA, 0.5% NP-40) for 1 h at 4°C with periodic sonication. After centrifugation at 12,000 × g for 20 min, the supernatants (cell extracts) were used for ssDNA-cellulose batch chromatography.
Because of our observation that BRO proteins were localized in the nuclei of infected cells, we used a motif search analysis (DNASIS) to determine if the sequences of these proteins contained motifs associated with DNA binding. We found that BmNPV BRO proteins contained a motif found in single-stranded binding (SSB) proteins from prokaryotic and eukaryotic organisms (22). BmNPV DBP and LEF-3, known baculovirus SSB proteins, also contain this motif (Fig. (Fig.1)1) (20). The motif is composed of a pattern of basic and aromatic amino acid residues with the consensus sequence K/RX2–5K/RX4–12F/YX2–14F/YX6–13F/YX1–19K/RX3–26F/Y/WX6–12K/R. As shown in Fig. Fig.1,1, the N termini of BRO proteins from AcMNPV and BmNPV present perfect matches to the above consensus. This motif was also found in most BRO proteins from various baculoviruses (data not shown, but see Fig. Fig.1).1). This finding suggested the possibility that these proteins could be interacting with nucleic acids.
To determine if the BRO proteins were interacting with nucleic acids, we fractionated the nuclei of BmN cells from 14 h p.i. by extraction with different concentrations of NaCl and performed Western blot analyses using anti-BRO antibodies to detect the presence of BRO proteins. We recently reported that anti-BRO antibodies were able to recognize all three groups of BmNPV BRO proteins (28 [BRO-B and -E], 38 [BRO-A and -C], and 41 [BRO-D] kDa) (13). As shown in Fig. Fig.2A,2A, most of 37-kDa polypeptides, which likely correspond to BRO-A and/or BRO-C (BRO-A/C), could be extracted with 0.6 M NaCl (Fig. (Fig.2A,2A, lanes 5 and 6). This is the concentration necessary to extract histone H1 from chromatin (6). Other histone proteins are extracted with H2SO4. Soluble nuclear proteins and proteins with moderate affinity to genome DNA (i.e., high-mobility-group and low-mobility-group proteins) would be extracted during the first two extraction steps (75 mM NaCl and 350 mM NaCl, respectively) (6). There were also immunoreactive bands of 37 and 75 kDa detected during these extractions (Fig. (Fig.2A,2A, lanes 1 to 4). These could be unbinding BRO proteins, since BRO proteins are produced in excess at 14 h p.i. (13). Extraction analysis of nuclei isolated from BmN cells at 4 h p.i. also showed that BRO-A/C was detected predominantly in the 0.6 M NaCl fraction (Fig. (Fig.2B),2B), suggesting that BRO-A/ C was already bound tightly to nuclear structures even at this very early stage of infection. Although we detected BRO-A/C, we were not able to distinguish individual BRO proteins because of their similar molecular masses (35.4 kDa for BRO-A and 35.9 kDa for BRO-C).
The results described above suggested that BRO-A/C interact either with proteins that bind to DNA with high affinity or with nucleic acids directly. To investigate this, we treated the nuclei of BmNPV-infected cells with MN. Nuclei were isolated from BmN cells at 4, 8, and 14 h p.i. and incubated with or without MN. After centrifugation, the supernatants were subjected to SDS-PAGE and Western blotting. As shown in Fig. Fig.3A,3A, significant amounts of 37-kDa BRO species (BRO-A/C) were detected in supernatants even after MN treatment. At 4 h p.i., the 38-kDa polypeptide corresponding to BRO-D was also clearly detected after MN treatment (Fig. (Fig.3A;3A; upper band of lane 2), suggesting that BRO-D is also involved in nuclear structure. The polypeptide with a molecular mass of 75 kDa was considered to be a homo- or heterodimer of BRO-A/C (Fig. (Fig.3A,3A, lanes 4 and 6). Yeast two-hybrid screening showed the interactions between BRO-A and itself or between BRO-C and BRO-C, supporting this idea (W.-K. Kang, unpublished data).
To further examine the interaction of BRO-A/C with nucleic acids, MN treatment was introduced between the 350 and 600 mM NaCl extraction steps. Abundant amounts of BRO-A and/or -C remained associated with chromatin after the extraction with 350 mM NaCl, since MN treatment was able to release these proteins from the chromatin (Fig. (Fig.3B,3B, lanes 5 to 8). These results strongly indicated that BRO-A/C was associated with nucleosomes in BmN cells. To further investigate this, we determined whether BRO-A/C interacted with core histones by performing histone-agarose chromatography. A protein fraction was prepared by extracting nuclei with 600 mM NaCl and then treating them with MN to digest DNA in the sample, thereby excluding it from interaction via DNA binding. The complete digestion of genomic DNA was confirmed by agarose gel electrophoresis in the presence of ethidium bromide (data not shown). The extract was diluted to 0.05 M NaCl and then loaded onto a histone-agarose column, eluted with increasing concentrations of NaCl, and then analyzed by Western blotting (Fig. (Fig.3C).3C). Polypeptides corresponding to BRO-A/C were detected in all fractions except the fraction of 2.0 M NaCl (Fig. (Fig.3C).3C). BRO-A/C was eluted from the histone-agarose column as a single asymmetric peak with a maximum at 0.15 to 0.2 M NaCl (Fig. (Fig.3C,3C, lanes 2 and 3), although traces of the antigen were still detected in the fractions up to 1.0 M NaCl (lane 9). Thus, this elution profile suggested that BRO-A/C was able to interact with core histones. Taking the data together, we concluded that BRO proteins are involved in nucleosome organization.
We examined the ability of BRO-A/C to bind nucleic acids by using column chromatography on ss- or dsDNA-cellulose or poly(U)-agarose. Polypeptides corresponding to BRO-A/C showed binding affinities for ss- and dsDNA as well as poly(U) used as a model RNA resin (Fig. (Fig.4).4). The chromatographic profiles for dsDNA (Fig. (Fig.4B)4B) and poly(U) (Fig. (Fig.4C)4C) were similar, with most of the protein eluting at 0.5 M (lanes 2). However, the profile of ssDNA-cellulose elution was different. There are two peaks for elution, one at 0.5 M and the other at 1.2 M NaCl, and this peculiarity of ssDNA chromatography was reproduced in several experiments. The reason for this is unknown, although it could suggest two methods of binding, e.g., a dimer or monomer and a higher-order conformation. However, it is clear that the retention ability of BRO-A/C for ssDNA was stronger than that for dsDNA or poly(U).
To determine the region responsible for nucleic acid binding properties, a series of BRO-A fragments were expressed in E. coli as His-tagged polypeptides and used in ssDNA-cellulose binding experiments. Polypeptides corresponding to the N-terminal half (1 to 159 amino acids [aa]) were eluted in fractions up to 0.7 M NaCl (Fig. (Fig.5A),5A), whereas polypeptides of the C-terminal half (160 to 317 aa) showed no affinity for DNA (Fig. (Fig.5B).5B). This data suggested that DNA binding activity might lie on the N terminus of BRO-A. To further define the precise region for DNA binding activity, we expressed two BRO-A fragments containing 1 to 78 (Fig. (Fig.5C)5C) and 79 to 159 (Fig. (Fig.5D)5D) aa. Whereas the 79- to 159-aa region was unable to bind to DNA, the most N-terminal polypeptide containing 78 amino acid residues demonstrated strong DNA binding activity. As described above, the N-terminal region of BRO-A showed an ssDNA binding motif (Fig. (Fig.1).1). This also strongly supports the idea that the N-terminal region containing 78 amino acid residues of BRO-A was responsible for DNA binding activity.
Posttranslational modification(s) of BmNPV BRO proteins may account for the immunoreactive bands (e.g., 40 to 60 kDa) that do not conform to the size of the predicted proteins. Such increases in molecular mass may be due to ubiquitination of BRO proteins followed by proteasome degradation. To investigate this, a specific proteasome inhibitor (MG-132) was tested. We postulated that the inhibition of the proteolytic activity of the proteasomes would increase the amount of BRO proteins as well as that of their ubiquitinated forms. As expected, our result showed that the amounts of BRO-A/C and higher-mass forms (40 to 60 kDa) were increased by the introduction of proteasome inhibitor (Fig. (Fig.6A),6A), suggesting that ubiquitination plays a role in the degradation of these proteins. The immunoreactive bands at around 25 and 27 kDa shown in Fig. Fig.6A6A might include smaller BRO proteins (BRO-B and BRO-E) and proteolytic fragments of BRO-A/C. Next, we investigated which BRO proteins are predominantly ubiquitinated by using four BmNPV recombinants, in which each bro gene was deleted by replacement with the β-galactosidase cassette (13). Figure Figure6B6B indicated that BRO-A was the main target of ubiquitination, since extra polypeptides were not present in the cells infected with the bro-a deletion mutant (Fig. (Fig.6B,6B, lanes 2 and 7).
We also found that BmNPV BRO proteins served as substrates for protein kinase(s) in infected cells. Immunoprecipitation of in vivo-labeled proteins with 32P using BRO-A antiserum revealed the incorporation of phosphates into polypeptides corresponding to BRO-A/C (Fig. (Fig.6C).6C). These proteins were phosphorylated at least by 8 h p.i. The level of phosphorylation reached a maximum between 14 and 20 h p.i. and persisted through 48 h p.i.
Our investigations have demonstrated that BmNPV BRO proteins, especially BRO-A, BRO-C, and BRO-D, have nucleic acid binding activities and are involved in nucleoprotein complexes in the nuclei of infected cells. BmNPV BRO-A, -C, and -D proteins were reported to be localized in the nucleus; however, BRO-B and -E showed only cytoplasmic distributions (13). This also supports the notion that BRO-A, -C, and -D might be nuclear proteins. We have been concentrating on BRO-A and BRO-C due to their apparent abundance in infected cells. Although it was difficult to separate BRO-A from BRO-C by SDS-PAGE in most experiments because of their similar molecular weights, anti-BRO-A antibodies were able to recognize the BRO-C protein in the experiments using bro-a deletion mutants and vice versa. The failure to obtain double-deletion mutants of bro-a and bro-c (13) suggests not only that this group plays an important role in infection but also that they may carry out the same function(s). Thus, we postulated that our data are true for both BRO-A and BRO-C. The extraction of nuclei by following a histone extraction protocol and MN treatment analyses indicated that BRO-A/C are involved in nucleosome organization by binding to nucleic acids directly. Similar concentrations (500 to 600 mM) of NaCl are required for eluting BRO-A/C from either the chromatin of infected cells or DNA columns, suggesting that BRO-A/C interact with DNA in a sequence-independent manner.
Wilson and Miller reported that viral DNA acquired a chromatinlike structure in the nuclei of Sf-21 cells infected with AcMNPV (23). They also showed that this nucleosome structure contained two major virus-induced proteins with molecular masses of 15 and 39 kDa. BmNPV BRO-A/C revealed similar molecular weights, nuclear localization, and binding affinities for nuclear structures, suggesting that they are the counterparts of the 39-kDa nucleoprotein in AcMNPV. Polyclonal antibodies against BmNPV BRO-A recognize a single polypeptide of 35 kDa that corresponds relatively well to the predicted molecular mass of AcMNPV BRO protein (37.8 kDa) (E. Zemskov, unpublished data). This polypeptide was specific for infected cells, appeared no later than 4 h p.i., and persisted through at least 26 h p.i. The expression pattern and molecular mass of AcMNPV BRO suggested that BRO could be the 39-kDa protein in the nucleosome structure of infected Sf-21 cells as described by Wilson and Miller (23). Although it remains unclear whether these two are the same, our data support the idea that baculoviral proteins are involved in nucleoprotein complexes in the nuclei of infected cells.
DNA-cellulose chromatography experiments using overexpressed BRO-A fragments in E. coli indicate that the DNA binding ability lies in the N-terminal region of BRO-A containing 80 amino acid residues. Further alignment by computer confirmed the presence of an ssDNA binding motif in this region. This motif was originally found in SSB proteins from prokaryotic and eukaryotic organisms (22). We also found that most BRO proteins contain this motif. The N-terminal localization of the motif is common to all BmNPV BRO proteins, AcMNPV BRO protein, and some BRO proteins from LdNPV; however, several BRO proteins of LdNPV and X. c-nigrum granulovirus contain the motif in a central or C-terminal region. Baculoviral LEF-3 and DBP, which have also been described as SSB proteins, contain this motif (20). Interestingly, LEF-3 and DBP have no homology with any known SSB proteins. Thus, this consensus seems important for baculoviral SSB proteins.
Due to the limited number of viral proteins, one protein could have several functions in infected cells. This has already been demonstrated for LEF-3 of AcMNPV. It functions as an SSB protein in DNA replication and also participates in the translocation of virus-encoded DNA helicase from the cytoplasm to the nucleus (8, 24). Among the products of the five BmNPV bro genes, at least three BRO proteins, BRO-A, BRO-C, and BRO-D, are nucleic acid binding proteins. Before the onset of viral DNA replication, these proteins are already associated with nuclear structures, most likely with chromatin. BRO-A/C especially showed very strong affinity for ssDNA. Based on these data, we propose a number of possible functions for these proteins. They could block cellular replication and/or transcription and switch host machinery to viral DNA or RNA synthesis by binding to host chromosomal DNA. In addition, RNA binding activity of BRO-A/C demonstrated by poly(U)-agarose chromatography suggests that they could participate in the nuclear export of mRNA. The presence of such proteins is known in eukaryotic cells and some viruses (3, 4, 10, 21). BmNPV BRO proteins seem to be abundant in the early stage of infection. Therefore, their ubiquitination and involvement in proteasome-directed cleavage could protect other viral proteins from degradation and increase the efficacy of the infection. Phosphorylation of BRO proteins may also regulate DNA and RNA binding activity as shown in many DNA binding proteins as well as LEF-3 and DBP of baculovirus (5, 14, 15; Zemskov, unpublished). A switch of functions might be modulated by factors such as the ratio of host DNA to viral DNA, interaction with specific proteins, and posttranslational modifications (phosphorylation and ubiquitination). Although a distinct feature of the bro gene family is its extensive repetition in baculovirus genomes (up to 17 copies in the LdNPV genome ), the necessity for this amplification is unclear. It could be involved in binding to a variety of different DNA-protein conformations that may be present in most cells or specific for different cell types.
We thank Keiju Okano and George F. Rohrmann for critical reading of the manuscript and M. Kurihara for providing insect cells.
This work was supported by a CREST award from Japan Science and Technology Corporation (S.M.). The work was also supported by grants from the COE (Center of Excellence) program and Biodesign Research program of the Science and Technology Agency, the President Special Research Grant of RIKEN (W.K.), and an STA fellowship (E.A.Z.).