|Home | About | Journals | Submit | Contact Us | Français|
The Autographa californica multiple nucleopolyhedrovirus orf92 (p33), ac92, is one of 31 genes carried in all sequenced baculovirus genomes, thus suggesting an essential function. Ac92 has homology to the family of flavin adenine dinucleotide-linked sulfhydryl oxidases and is related to the ERV/ALR family of sulfhydryl oxidases. The role of ac92 during virus replication is unknown. Ac92 was associated with the envelope of both budded and occlusion-derived virus (ODV). To investigate the role of Ac92 during virus replication, an ac92-knockout bacmid was generated through homologous recombination in Escherichia coli. Titration and plaque assays showed no virus spread in ac92-knockout bacmid DNA-transfected insect cells. Deletion of ac92 did not affect viral DNA replication. However, ac92-knockout bacmid DNA-transfected cells lacked multiply enveloped occlusion-derived nucleocapsids; instead, singly enveloped nucleocapsids were detected. To gain insight into the requirement for sulfhydryl oxidation during virus replication, a virus was constructed in which the Ac92 C155XXC158 amino acids, important for sulfhydryl oxidase activity, were mutated to A155XXA158. The mutant virus exhibited a phenotype similar to that of the knockout virus, suggesting that the C-X-X-C motif was essential for sulfhydryl oxidase activity and responsible for the altered ODV phenotype.
Baculoviruses infect insects and are characterized by a circular double-stranded DNA genome ranging from 80 to 180 kbp (10), which is packaged within a rod-shaped capsid and surrounded by an envelope. Two types of virions are produced during the baculovirus replication cycle: the budded virus (BV) and the occlusion-derived virus (ODV). Although the nucleocapsids are similar in the two, the envelopes differ in composition, reflecting their different functions during infection (4). The virions have distinct tissue tropism: ODVs infect midgut epithelial cells up to 10,000-fold more efficiently than do BVs, whereas BVs infect cultured cells 1,000-fold more efficiently than do ODVs (24).
Alphabaculovirus and Betabaculovirus are two genera of Baculoviridae, based primarily on occlusion body morphology and phylogeny (11). Alphabaculoviruses have been divided into single nucleopolyhedrovirus (SNPV) and multiple nucleopolyhedrovirus (MNPV) occlusions. The M or S refers to whether the ODVs contain multiply or singly enveloped nucleocapsids, respectively (33). The MNPVs have been isolated only from Lepidoptera (23). It has been proposed that the M phenotype has an advantage over the S phenotype. Although the establishment of primary infections may be the same, the presence of multiple virions in MNPVs accelerated the onset of systemic infections (32, 33).
The Autographa californica MNPV (AcMNPV) open reading frame 92 (orf92) (ac92, p33) is one of 31 core genes present in all sequenced baculovirus genomes (24). Previous studies showed that Ac92 and its homologs, Culex nigripalpus NPV orf14 and Helicoverpa armigera SNPV orf80, are associated with the ODV (3, 8, 18). Predicted protein sequence and functional analyses indicate that Ac92 is a flavin adenine dinucleotide (FAD)-linked sulfhydryl oxidase with a conserved C-X-X-C sulfhydryl oxidase motif (15).
Coexpression of Ac92 and human P53 in insect cells leads to the interaction of the two proteins. Ac92 also enhanced P53-mediated apoptosis approximately 2-fold (20). Information on the role of Ac92 during virus replication and the significance of its interaction with P53 is lacking; thus, studying ac92 during baculovirus replication promises to lead to insightful information.
In this study, we generated an ac92-knockout virus, Ac92KO-PG, to investigate the role of Ac92 during AcMNPV replication. Ac92KO-PG was compared to viruses carrying ac92 with respect to virus production kinetics, plaque formation phenotype and production, and viral DNA and protein synthesis. Since we found that Ac92 localized to both BV and ODV, the morphology of ODVs in Ac92KO-PG was also examined by electron microscopy. In addition, to characterize ac92, we examined its temporal expression, identified the transcriptional initiation site, and determined protein synthesis during virus replication. In order to determine if sulfhydryl oxidase activity is responsible for the phenotypes of the knockout virus, we constructed an Ac92 mutant virus with disrupted cysteines within the motif important for sulfhydryl oxidase activity. Characterization of this virus is important to determine if this enzymatic activity is responsible for the observed phenotype and if this activity is essential for virus replication.
Bacmid bMON14272 (Invitrogen), herein referred to as AcBAC, containing an AcMNPV genome, was propagated in Escherichia coli BJ5183 as previously described (2). DH10B cells with helper plasmid pMON7124 were purchased from Invitrogen. The Sf9 insect cell line, the clonal isolate 9 from IPLB-Sf21-AE cells, derived from the fall armyworm Spodoptera frugiperda (31), was purchased from ATCC and cultured at 27°C in TC-100 medium (Invitrogen) supplemented with 10% fetal bovine serum, penicillin G (60 μg/ml), streptomycin sulfate (200 μg/ml), and amphotericin B (0.5 μg/ml).
Sf9 cells were infected with AcMNPV at a multiplicity of infection (MOI) of 5. Cells were treated with the protein synthesis inhibitor cycloheximide (100 μg/ml) or DNA synthesis inhibitor aphidicolin (5 μg/ml) as previously described (9). Total cellular RNA was isolated using TRIzol reagent (Sigma) according to the protocol of the manufacturer. RNA was quantified by optical density measurements at 260 nm.
RT-PCR was performed using the Access RT-PCR System (Promega) according to the handbook provided. Two primers, Ac9251 (5′-GAGCTCATGATACCGCTGACGCCGCT-3′) and Ac9233 (5′-TTGTATTCTCCCCATGTCATGCGT-3′), were used for PCR amplification to detect transcription of ac92.
5′ rapid amplification of cDNA 5′ ends (RACE) was performed using a 5′ RACE system kit, version 2.0, according to the handbook provided (Invitrogen). ac92-specific reverse primer Ac92SP1 (5′-CCATATCGTCGATGATGAG-3′) and Ac92SP2 (5′-CCCATATGGTAGTAAACG-3′) were used for PCR amplification. PCR products were gel purified with a gel extraction kit (Qiagen) and cloned into pCRII (Invitrogen) prior to derivation of the nucleotide sequence.
A monolayer of Sf9 cells (1.0 × 106) was infected at an MOI of 5 with Ac92HARep-PG, an ac92-knockout bacmid expressing ac92 at a different locus (see below). The protein synthesis inhibitor cycloheximide and the DNA synthesis inhibitor aphidicolin were added as described above. At different time points, cells were collected and centrifuged at 1,000 × g for 5 min at room temperature. The pelleted cells were washed twice with phosphate-buffered saline (PBS), pH 6.2 (19), and resuspended in PBS with an equal volume of 2× protein loading buffer (PLB; 0.25 M Tris-Cl, pH 6.8, 4% SDS, 20% glycerol, 10% 2-mercaptoethanol, and 0.02% bromophenol blue). Proteins were analyzed by SDS-12% PAGE (sodium dodecyl sulfate-polyacrylamide gel electrophoresis) followed by immunoblotting or stored at −20°C until further use.
Protein samples were mixed with equal volumes of 2× PLB and incubated at 100°C for 5 min. Samples were resolved by SDS-12% PAGE, transferred onto a polyvinylidene fluoride (PVDF) membrane (Millipore), and probed with one of the following primary antibodies: (i) mouse monoclonal antihemagglutinin (anti-HA) antibody (Covance), (ii) rabbit polyclonal VP39 antiserum (13), (iii) mouse monoclonal anti-GP64 antibody (eBioscience), or (iv) mouse monoclonal anti-IE-1 antibody; this probing was followed by incubation with the horseradish peroxidase-conjugated secondary antibodies (Sigma). Blots were developed using the SuperSignal West Pico chemiluminescent substrate (Pierce) and exposed to X-ray films.
The ac92-knockout bacmid was generated through homologous recombination in E. coli as previously described (38) in which ac92 in AcBAC was replaced with the chloramphenicol resistance gene (Cm). A 1,178-bp PCR fragment was amplified using pCMR (34) as the template and primers Ac92D1 (5′-AGCCTCTAAATCGACGCACCTCACCAAACTGTTGTCTTCGCAAGCAACTTACTTATACCATTTTGCGTGTCCCT TTCGTCTTCGAATAAA-3′) and Ac92D2 (5′-TGTTGTATTATTGCAAATTTAACAATTTTTTGTATTCTCCCCATGTCATGCGTTCGTAATGAGCGGG CGGTAAACCAGCAATAGACATAA-3′). This fragment contains the 1,038-bp Cm cassette and 70 bp of ac92 flanking regions at each end. Purified PCR fragments (1 μg) were electroporated into electrocompetent BJ5183 cells, and colonies resistant to chloramphenicol and kanamycin were selected for further confirmation by PCR analysis. The resulting ac92-knockout bacmid was named Ac92KO. To facilitate further study, the Ac92KO bacmid was isolated and electroporated into DH10B cells which contained the helper plasmid pMON7124. Colonies that grew on Luria-Bertani (LB) agar plates containing 20 μg/ml chloramphenicol, 50 μg/ml kanamycin, and 10 μg/ml tetracycline were selected, and the Ac92KO bacmid was isolated and confirmed by PCR again.
Four primers were used to confirm the replacement of ac92 by Cm in the ac92 locus of AcBAC. Primers Cm5 (5′-CTTCGAATAAATACCTGTGA-3′) and Cm3 (5′-AACCAGCAATAGACATAAGC-3′) were used to detect the correct insertion of the Cm cassette. Primers Ac9251 and Ac9231, which bind outside the flanking sequences used for recombination, were used to confirm the deleted region and the insertion of the Cm. Primer pairs Ac9251/Cm3 and Cm5/Ac9231 were used to examine the recombination junctions at the flanking regions upstream and downstream of the insertion site, respectively.
AcBAC and Ac92KO bacmids were isolated from BJ5183 cells and electroporated into DH10B cells containing the helper plasmid pMON7124. To facilitate the examination of virus-infected cells and to determine if deletion of ac92 had any effect on occlusion morphogenesis, polyhedrin and enhanced green fluorescent protein (egfp) were inserted into the polyhedrin locus by site-specific transposition as previously described (38). To this end, several donor plasmids were constructed as follows. Two primers, Ac9252 (5′-GAGCTCAATGCACATATGTCTTATAC-3′ with SacI restriction site underlined) and Ac9234 (5′-TCTAGAAGGTGTAGGTAACGGTACAA-3′ with XbaI restriction site underlined) were designed to amplify a 1,281-bp fragment containing the native ac92 promoter and ac92 open reading frame (ORF) (Ac92POA) using AcBAC DNA as a template. This PCR product was ligated into pCRII for sequencing, and the resulting plasmid, pCRII-Ac92POA, was digested with SacI and XbaI. Next, the Ac92POA fragment was inserted into the pFB1-PH-GFP plasmid (38) to generate a donor plasmid named pFB-PG-Ac92POA. Primers OppAT5 (5′-TCTAGACCGTGGTGGCGTGCGGCAA-3′ with XbaI underlined) and OppAT3 (5′-CTGCAGCGAAAATCGGGCTAGATT-3′ with PstI underlined) were designed to amplify a 473-bp fragment encoding a polyadenylation signal derived from the Orgyia pseudosugata MNPV inhibitor of apoptosis gene (OpIAP-pA-Tail) using plasmid pHSGFP (7) as a template. The PCR product was ligated into the pCRII vector for nucleotide sequencing, and the resulting plasmid, pCRII-OpIAP-pA-Tail, was digested with XbaI and PstI to obtain the OpIAP-pA-Tail fragment. The OpIAP-pA-Tail fragment was inserted into pFB1-PH-GFP to generate the plasmid pFB-PG-pA. Primers Ac9252 and Ac9232 (5′-TCTAGATTAGGCGTAATCTGGGACGTCGTATGGGTATTGCAAATTTAACAATTTTTTGTATTCTCCCC-3′ with XbaI restriction site underlined) were used to amplify an 1,109-bp fragment that contained the ac92 native promoter, ORF, and an HA tag prior to the stop codon (Ac92POHA) by using AcBAC DNA as template. The PCR product was ligated into the pCRII vector for sequencing, and the resulting plasmid, pCRII-Ac92POHA, was digested with SacI and XbaI to insert the Ac92POHA fragment into pFB-PG-pA and generate a donor plasmid, pFB-PG-Ac92POHA. Electrocompetent DH10B cells containing helper plasmid pMON7124 and AcBAC were transformed with pFB1-PH-GFP to generate the control bacmid AcWT-PG. Electrocompetent DH10B cells containing helper plasmid pMON7124 and Ac92KO were transformed with donor plasmid pFB1-PH-GFP, pFB-PG-Ac92POA, or pFB-PG-Ac92POHA to generate the Ac92 knockout virus Ac92KO-PG and two repair viruses, Ac92Rep-PG, with ac92, and Ac92HARep-PG, with an HA-tagged ac92. Bacmid DNA was isolated, and successful transposition was confirmed by PCR. The correct recombinant bacmids were electroporated into E. coli DH10B cells and screened for tetracycline sensitivity to ensure that the isolated bacmids were free of helper plasmids. Bacmid DNA was extracted and purified with the Qiagen large-construct kit and quantified by optical density measurement.
Sf9 cells (1.0 × 106 cells/35-mm-diameter dish) were transfected with 1 μg of a bacmid construct using Grace's unsupplemented medium (Invitrogen) and Lipofectin (7). Cell monolayers were incubated for 5 h after transfection and then washed twice with TC-100 medium and replenished with 2 ml of fresh TC-100 medium supplemented with 10% fetal bovine serum. Plaque assays were performed as previously described (38). The supernatants of the transfected cells were collected at various time points to determine titers by 50% tissue culture infective dose (TCID50) endpoint dilution assays (17) on Sf9 cells.
To detect viral DNA replication in virus-transfected cells, QPCR was performed as previously described (30). The gp64-knockout virus, AcGP64KO, was used as a noninfectious control since the virus is unable to infect from cell to cell (16). Sf9 cells (1.0 × 106 cells/35-mm-diameter dish) were transfected in triplicate with 1 μg of bacmid DNA from Ac92KO-PG or AcGP64KO, and cells were collected at different time points. Total DNA was prepared with the Wizard genomic DNA purification kit (Promega) according to the protocol of the manufacturer. Purified DNA was quantified by optical density measurement. Prior to PCR, 1 μg of total DNA from each time point was digested with 20 units of DpnI restriction enzyme (NEB) overnight in a 25-μl total reaction volume. QPCR was performed with 1 μl of the digested DNA and the Absolute QPCR SYBR green fluorescein mix (Thermo Scientific) according to the protocol of the manufacturer by using the same primers to amplify a 100-bp region of the AcMNPV gp41 as described previously (30). Standard DNA samples were used from purified AcMNPV BV DNA and serially diluted to 100, 10, 1, 0.1, 0.01, and 0.001 ng. Genomic equivalents of DNA samples were determined by extrapolation from standard curves. One copy of AcMNPV genomic DNA was estimated to be 1.36 × 10−4 pg (5). A melting-curve analysis of each amplified sample was carried out to check the specificity of each reaction. The results were analyzed using GraphPad Prism version 5.01 (GraphPad Software, Inc.).
To analyze the effect of deleting ac92 on viral protein synthesis, Sf9 cells were transfected with A92KO-PG or AcWT-PG bacmid DNA. At different time points, cells were collected and samples were prepared for immunoblotting as described above. IE-1 antibodies and VP39 antisera were used as primary antibodies, and immunoblotting was performed as described above.
Sf9 cells were infected with Ac92HARep-PG at an MOI of 0.01. At 96 h postinfection (p.i.), the pellet of infected cells and the supernatant were separated by centrifugation. BV was concentrated and purified as described previously (17). To collect polyhedra (4), 600 fourth-instar Spodoptera exigua larvae were infected by contamination of artificial diet with Ac92HARep-PG polyhedra, which were prepared from Ac92HARep-PG-infected Sf9 cells. ODVs were purified as previously described (37).
BV and ODV were fractionated into envelope and nucleocapsid fractions as previously described (4).
Sf9 cells (1.0 × 106 cells/35-mm-diameter dish) were transfected with 1 μg of bacmid DNA. Transfected cells were collected at different time points, fixed, dehydrated, embedded, sectioned, and stained as described previously (14). Samples were viewed with a Philips CM 100 transmission electron microscope at an accelerating voltage of 100 kV.
Cysteines 155 and 158 were altered to alanines by a previously described method (6) and primers Ac9255FT (5′-GCAGCCATGGCACGCGATCATTATATGAACGTCA-3′), Ac9256FS (5′-CGCGATCATTATATGAACGTC-3′), Ac9237RT (5′-TGCCATGGCTGCCTGTAGTATGAAAAATACGTTG-3′), and Ac9238RS (5′-CTGTAGTATGAAAAATACGTTG-3′) in PCRs using pCRII-Ac92POHA as a template. The PCR was started at 94°C for 3 min, followed by 25 cycles of 94°C for 20 s, 60°C for 20 s, and 68°C for 4 min, with a final 7-min extension step at 68°C. The PCR product was digested with DpnI (NEB) for 2 h and denatured at 99°C for 3 min. Hybridization was performed using two cycles of 65°C for 5 min and 30°C for 15 min. An aliquot of 20 μl from the reaction mixture was transformed into E. coli XL1-Blue cells. The mutations in the final product were confirmed by DNA sequencing. The correct plasmid pCRII-Ac92POHA-C155C158toA155A158 was digested with SacI and XbaI, and the mutated DNA fragment was inserted into pFB-PG-pA to generate the donor plasmid pFB-PG-Ac92POHA-C155C158toA155A158. DH10B cells containing helper plasmid pMON7124 and the Ac92KO bacmid were transformed with the donor plasmid pFB-PG-Ac92POHA-C155C158toA155A158 to generate the recombinant virus Ac92HAM1-PG. Bacmid DNA was isolated, and successful transposition was confirmed by PCR.
The ac92 gene was PCR amplified from plasmid pCRII-Ac92POHA using primers Ac9257 (5′-GCTAGCATGATACCGCTGACGCCGCTTT-3′ with NheI restriction site underlined) and HindIII-HA-3 (5′-AAGCTTTTAGGCGTAATCTGGGACGT-3′ with HindIII restriction site underlined), and the nucleotide sequence was confirmed. The resulting plasmid, pCRII-Ac92HA, was digested with NheI and HindIII, and the Ac92HA fragment was inserted into the pET28a vector (Novagen) to generate the expression plasmid pET28a-Ac92HA. The primers Ac9257 and HindIII-HA-3 were also used to amplify the ac92 mutated gene using pCRII-Ac92POHA-C155C158toA155A158 as a template. The PCR product was ligated into the pCRII vector for sequencing to generate pCRII-Ac92HA-C155C158toA155A158. This plasmid was subsequently digested with NheI and HindIII, and the fragment was inserted into pET28a to generate the expression plasmid pET28a-Ac92HAM1.
For expression, BL21(DE3)/pLysS competent cells transformed with plasmid pET28a-Ac92HA or pET28a-Ac92HAM1 were grown at 30°C in LB broth containing 50 μg/ml kanamycin and 20 μg/ml chloramphenicol to an A600 of 0.6, induced by adding IPTG (isopropyl-β-d-thiogalactopyranoside) to a final concentration of 1 mM, and incubated overnight at 18°C. For protein purification, 2-liter cultures were harvested by centrifugation at 3,000 × g for 15 min at 4°C. The pellet was resuspended in 50 ml cold extraction buffer (300 mM NaCl, 50 mM NaPO4, 20 mM imidazole, pH 7.5) containing a protease inhibitor cocktail tablet (Roche), and cells were disrupted by 15 10-s sonications at maximal intensity, with cooling of the sample between sonications. The lysate was clarified by centrifugation at 15,000 × g for 20 min at 4°C. The clarified lysate was mixed with 500 μl Talon metal affinity resin (Clontech) that had been equilibrated with extraction buffer, incubated for 3 h at 4°C with gentle agitation, transferred to a column, washed three times with extraction buffer, and eluted sequentially with 4 column volumes of extraction buffer containing 50 mM, 100 mM, 200 mM, or 300 mM imidazole.
Sulfhydryl oxidase activity was measured using the method of Long et al. (15). Briefly, Ac92HA or Ac92HAM1 protein was used as an enzyme and dithiothreitol (DTT) was used as a substrate. After specific time periods, aliquots were mixed with Ellman's reagent [5,5′-dithio-bis(2-nitrobenzoic acid) (DTNB); Sigma] and thiol content was measured by absorbance at 421 nm and by calculating an extinction coefficient of 13.6 mM−1 cm−1.
Activity of Ac92 as an FAD-linked sulfhydryl oxidase has been previously determined (15); however, the role of this activity during virus replication or morphogenesis is unknown. In order to determine the effects of this gene during viral replication, we constructed an ac92-null virus, Ac92KO-PG, and a virus with mutations within a conserved motif necessary for sulfhydryl oxidase activity, Ac92HAM1-PG. Both viruses were characterized and compared to three other viruses, AcWT-PG, carrying unaltered ac92 at its native locus; Ac92Rep-PG, carrying ac92 at the polyhedrin locus; and Ac92HARep-PG, carrying an epitope-tagged ac92 at the polyhedrin locus. This approach allowed us to contrast the effects of Ac92 and its known enzymatic activity during virus replication and determine if ac92 has a function independent of sulfhydryl oxidation.
We determined the timing of ac92 transcription in an effort to determine when the gene was important during virus replication. RT-PCR was performed using total RNA extracted from AcMNPV-infected cells at different time points. ac92 transcripts were detected from 9 h p.i. to 72 h p.i. (Fig. (Fig.1A,1A, RT+). No signals were obtained when avian myeloblastosis virus (AMV) reverse transcriptase was not included (Fig. (Fig.1A,1A, RT−). Addition of the protein synthesis inhibitor cycloheximide and DNA synthesis inhibitor aphidicolin to infected cells abolished transcription, indicting that ac92 is a late gene, requiring DNA and prior protein synthesis (Fig. (Fig.1B1B).
The ac92 transcriptional initiation site was determined by 5′ RACE analysis using total RNAs isolated from AcMNPV-infected cells at 24 h p.i. Five clones derived from the 5′ RACE products were sequenced, and transcripts were initiated at the first A of the typical baculovirus late promoter motif TAAG, located 184 nucleotides (nt) upstream of the ac92 translation initiation codon ATG (Fig. (Fig.1C).1C). Together, RT-PCR and the 5′ RACE analyses indicate that ac92 is a late gene.
Synthesis of Ac92 from Ac92HARep-PG was analyzed by immunoblotting lysates from Ac92HARep-PG-infected cells by using HA antibody. An immunoreactive band of approximately 33 kDa, the expected mass, was first detected at 24 h p.i. and persisted at 96 h p.i. (Fig. (Fig.1D).1D). The protein synthesis inhibitor cycloheximide and the DNA synthesis inhibitor aphidicolin were individually added to infected cells; no proteins were detected in the presence of cycloheximide as expected, and Ac92 was not detected in aphidicolin-treated cells (Fig. (Fig.1E).1E). The requirement for DNA synthesis indicates that Ac92 is produced at late times.
Ac92KO was constructed by deleting most of ac92 but retaining 150 nt from the 5′ end and 61 nt from the 3′ end of the ac92 coding region to ensure expression of the neighboring genes, ac93 and ac91 (Fig. (Fig.2A).2A). The remaining 567-nt coding sequence (nt 78762 to nt 79328 of AcMNPV ) was replaced with the Cm cassette (Fig. (Fig.2A2A).
PCR was performed to confirm that ac92 had been deleted from the ac92 locus of AcBAC and that the Cm cassette was inserted (Fig. (Fig.2B).2B). Primers Ac9251 and Ac9231 were used to confirm the deletion of 567 bp within the ac92 coding region and its replacement with the 1,038-bp Cm. Primer pairs Ac9251/Cm3 and Cm5/Ac9231 (Fig. (Fig.2A)2A) were used to examine the recombination junctions upstream and downstream of ac92, respectively. Primers Cm5/Cm3 were used to confirm the insertion of the Cm cassette. The sizes of the PCR-amplified products are as expected (Fig. (Fig.2B2B).
To determine if deletion of ac92 had any effects on occlusion morphogenesis and to facilitate examination of virus-infected cells, an ac92-knockout mutant, Ac92KO-PG, containing polyhedrin and egfp, was constructed by transposition of polyhedrin and egfp into the polyhedrin locus of Ac92KO (Fig. (Fig.2D,2D, row i). Two repair bacmids were made to rescue and confirm the phenotype resulting from the deletion of ac92, Ac92Rep-PG (Fig. (Fig.2D,2D, row ii) and Ac92HARep-PG, with an HA tag at the C terminus (Fig. (Fig.2D,2D, row iii). AcWT-PG, AcBAC with polyhedrin and egfp, was used as a positive control (Fig. (Fig.2C).2C). All constructs were confirmed by PCR, expression of egfp, and the formation of occlusions (data not shown).
To determine the effect of deleting ac92 on viral replication, cells were transfected with DNA from Ac92KO-PG, Ac92Rep-PG, Ac92HARep-PG, or AcWT-PG. Transfected cells were monitored by fluorescence microscopy, and no obvious differences in the numbers of fluorescent cells and fluorescence intensities were observed at 24 h posttransfection (p.t.) (Fig. (Fig.3A,3A, 24 h p.t.). eGFP fluorescence was observed in most AcWT-PG-, Ac92Rep-PG-, or Ac92HARep-PG-transfected cells by 72 h p.t. (Fig. (Fig.3A,3A, 72 h p.t.). In contrast, the number of cells transfected with Ac92KO-PG DNA did not increase in eGFP fluorescence significantly from 24 to 72 h p.t. (Fig. (Fig.3A,3A, 72 h p.t.). Inspection of cells by light microscopy showed no differences in occlusion body formation in any of the viruses up to 48 h p.t. (data not shown). However, many cells transfected with AcWT-PG, Ac92Rep-PG, or Ac92HARep-PG DNA contained occlusion bodies by 72 h p.t., while Ac92KO-PG DNA-transfected cells showed no increase in the number of cells that contained occlusion bodies (Fig. (Fig.3A,3A, 72 h p.t.). By 96 h p.t., most cells transfected with AcWT-PG, Ac92Rep-PG, or Ac92HARep-PG DNA showed eGFP fluorescence and contained occlusion bodies (data not shown); Ac92KO-PG DNA-transfected cells had no obvious differences, except for several small plaques (data not shown).
One-step virus growth curve analyses were performed to assess the effect of deleting ac92 on virus replication. Cells were transfected with bacmid DNA, and at selected time points the BV titers were determined by TCID50 endpoint dilution assays. Cells transfected with AcWT-PG, Ac92Rep-PG, or Ac92HARep-PG revealed a steady increase in virus production (Fig. (Fig.3B).3B). However, Ac92KO-PG-infected cells had single-cell infections as indicated by single eGFP-positive cells. Accounting for single eGFP-positive cells in titers yielded a very low titer starting at 24 h p.t., where the production of BV was 10 million times less than that in ac92-carrying viruses at 120 h p.t. (Fig. (Fig.3B,3B, panel ii). If single-cell infections were excluded, no detectable infectious BV was produced by Ac92KO-PG (Fig. (Fig.3B,3B, panel i). The supernatants from Ac92KO-PG DNA-transfected cells were collected and used for subsequent infections. At 96 h p.i., only 10 to 20 individual cells per 106 cells were eGFP positive. Supernatant from infected cells was collected and used to infect naive cells, and by 96 h p.i., no additional eGFP-positive cells were detected (data not shown). Therefore, deletion of ac92 is essential for budded virus production.
Plaque assays were performed to better assess virus spreading to neighboring cells. By 4 days p.t., large plaques were formed in AcWT-PG-, Ac92Rep-PG-, and Ac92HARep-PG DNA-transfected cells (Fig. (Fig.3C),3C), while plaques from Ac92KO-PG DNA-transfected cells consisted of single cells (Fig. (Fig.3C),3C), indicating a defect in virus production.
To assess the role of ac92 in viral DNA replication, we performed quantitative real-time PCR. AcGP64KO, lacking the envelope fusion protein, was used as a noninfectious control since the virus was reported to be unable to infect from cell to cell (16). Cells transfected with Ac92KO-PG or AcGP64KO bacmid DNA were harvested at 0, 24, 48, and 72 h p.t. Total intracellular DNA was extracted, and virus-specific DNA was quantified by real-time PCR (Fig. (Fig.4).4). We assumed that the efficiencies of viral and cellular DNA purification from viral DNA-transfected cells were equivalent. A standard calibration curve was generated, and a highly reproducible linear curve was obtained when the log concentration of purified AcMNPV genomic DNA was plotted against the cycle threshold (data not shown). Melting-curve analysis revealed that PCR products were specific (data not shown). There were no significant differences when comparing Ac92KO-PG and AcGP64KO at 0 and 24 h p.t.; however, viral DNA copy numbers were significantly higher from 0 to 24 h p.t. for each virus, clearly demonstrating that DNA replication was occurring in both viruses (Fig. (Fig.4).4). However, AcGP64KO produced more viral DNA at 48 h p.t. (P < 0.001) and 72 h p.t. (P < 0.05) than did Ac92KO-PG. Although it is possible that Ac92KO-PG has slight defects in DNA synthesis, we have observed that AcGP64KO has a slight leaky spreading defect that would lead to additional rounds of infections and thus more DNA copies (W. Wu and A. L. Passarelli, unpublished results) and therefore favor the later scenario.
To evaluate the effect of deleting ac92 on specific viral protein synthesis, proteins from AcWT-PG- or Ac92KO-PG DNA-transfected cells were harvested at designated time points and the early regulatory gene product IE-1 or the late major capsid gene product VP39 was detected with anti-IE-1 antibodies or anti-VP39 antiserum, respectively. IE-1 and VP39 were detected from 12 h p.t. or 18 h p.t., respectively, from AcWT-PG or Ac92KO-PG DNA-transfected cells (Fig. (Fig.5).5). The results indicate that deletion of ac92 did not affect IE-1 or VP39 protein production and, possibly, does not affect genes in these temporal phases.
BV or ODV was prepared from the supernatant of Ac92HARep-PG-infected cells or whole larvae, respectively. Virions were purified and biochemically fractionated into nucleocapsid and envelope fractions (Fig. (Fig.6,6, lanes NC and E), and HA-tagged Ac92 was detected by immunoblotting using anti-HA antibody. Ac92 was detected in both BV and ODV and was associated with the envelope fraction of BV. In the ODV, Ac92 appeared on both nucleocapsid and envelope fractions; however, proteins in the envelope fraction are more prominent than those in the nucleocapsid fraction. At this time, we cannot rule out contamination of the ODV nucleocapsid fraction with proteins from envelopes. To confirm fractionation efficiency, the nucleocapsid protein VP39 and the BV envelope protein GP64 were also analyzed by immunoblotting, and both VP39 and GP64 were detected in the expected fractions. These results indicate that Ac92 is associated with both BV and ODV and localized to the envelope fractions.
To determine whether deletion of ac92 affected virion morphogenesis, thin sections from bacmid DNA-transfected cells were visualized by electron microscopy. At 72 h p.t., cells transfected with AcWT-PG DNA showed typical characteristics of baculovirus infection, including the presence of the virogenic stroma (VS), an electron-dense net-shaped baculovirus-induced structure, containing numerous rod-shaped nucleocapsids (Fig. (Fig.7A,7A, panel i), and virus-induced intranuclear microvesicles and bundles of enveloped nucleocapsids aligning within the ring zone (Fig. (Fig.7A,7A, panel ii) and the mature multiply enveloped nucleocapsid within the ring zone (Fig. (Fig.7A,7A, panel ii, arrows, and Fig. Fig.7A,7A, panel iii) or embedded in polyhedra (Fig. (Fig.7A,7A, panel iv). Cells transfected with Ac92HARep-PG showed characteristics similar to those of AcWT-PG DNA-transfected cells (data not shown).
In Ac92KO-PG DNA-transfected cells, the VS and rod-shaped nucleocapsids were also observed in the nucleus (Fig. (Fig.7B,7B, panel i). Intranuclear microvesicles appeared within the ring zone (Fig. (Fig.7B,7B, panels ii and iv). However, ODVs with multiply enveloped nucleocapsids were not detected in the ring zone; instead, ODVs with singly enveloped nucleocapsids accumulated along with the microvesicles (Fig. (Fig.7B,7B, panels ii and iv). The shape and the size of polyhedra formed after transfection of cells with Ac92KO-PG DNA were similar to those of polyhedra formed by AcWT-PG or Ac92HARep-PG DNA-transfected cells; however, some did not contain ODVs (Fig. (Fig.7B,7B, panel vi), at least in the cross sections viewed, and some contained ODVs with singly enveloped nucleocapsids (Fig. (Fig.7B,7B, panel v). Together, these observations indicate that deletion of ac92 did not affect the formation of the VS, nucleocapsids, or polyhedra. However, it affected the formation of ODVs with multiply enveloped nucleocapsids. To our knowledge, ac92 is the only gene affecting this step in virion assembly.
ac92 encodes an FAD-linked sulfhydryl oxidase, a member of the ERV1/ALR family of sulfhydryl oxidases (15). The presence of a C-X-X-C active-site motif, where X is a variable amino acid, is the hallmark of proteins that are involved in forming or breaking disulfide bonds (29). To assess whether the C-X-X-C motif in Ac92 is essential for sulfhydryl oxidase activity and to determine the effects of this activity on virus morphogenesis, Ac92HAM1-PG, in which the predicted cysteine amino acids 155 and 158 of Ac92 were mutated to alanines and an HA tag was inserted at the C terminus of ac92, was constructed. The ac92 in Ac92HAM1-PG was driven by the ac92 native promoter, and the virus carried polyhedrin and egfp at the polyhedrin locus (Fig. (Fig.2D,2D, row iv).
Cells were transfected with Ac92HARep-PG or Ac92HAM1- PG DNA and harvested at 24 and 48 h p.t., and proteins were detected by immunoblotting using anti-HA antibody. A specific immunoreactive band of approximately 33 kDa, corresponding to epitope-tagged Ac92, was detected in transfected cells (Fig. (Fig.8).8). The signal was stronger in Ac92HARep-PG DNA-transfected cells at 48 h p.t., probably due to virus spread to and protein production from adjacent cells, while there was no virus spread from Ac92HAM1-PG-transfected cells (data not shown and Fig. Fig.33).
Replication of Ac92HAM1-PG in insect cells was similar to that of Ac92KO-PG (Fig. (Fig.3),3), including infection of single cells. In addition, plaques from Ac92HAM1-PG DNA-transfected cells consisted of single cells, reminiscent of those formed by Ac92KO-PG (Fig. 3B and C). Electron microscopy analysis of Ac92HAM1-PG DNA-transfected cells also showed results similar to those for Ac92KO-PG DNA-transfected cells, where singly enveloped nucleocapsids accumulated with microvesicles in the ring zone (Fig. (Fig.7B,7B, panel vii). Thus, altering the C-X-X-C motif affected both BV production and ODV envelopment of multiple nucleocapsids, suggesting that sulfhydryl oxidase activity affected these processes.
Ac92 (Ac92HA) and Ac92HA cysteine mutant protein (Ac92HAM1) were expressed in E. coli and affinity purified (data not shown). The purified proteins were assayed for sulfhydryl oxidase activity by using DTT as a substrate. Reactions were initiated by adding purified proteins to the substrate mixture, and aliquots were withdrawn at different times thereafter. The thiol content was measured in collected aliquots by DTNB reactivity (15). Ac92HA was able to oxidize DTT in a time-dependent manner and with similar oxidation efficiencies from 0 to 120 min, while Ac92HAM1 was not able to oxidize the substrate (Fig. (Fig.9).9). Thus, mutation of the C-X-X-C motif in Ac92 hinders sulfhydryl oxidase activity.
The AcMNPV ac92 is one of 31 baculovirus core genes, genes present in all sequenced baculovirus genomes (24). In this report, we investigated its enzymatic activity, transcription, and protein synthesis profiles and successfully generated recombinant viruses having an ac92 gene missing or altered to study its function and effects on AcMNPV replication.
We mapped the ac92 transcriptional start site to a late promoter element at nt 79622 (1). It is the furthermost of three such elements upstream of the gene. We confirmed that it is a late gene by its requirement for both DNA replication and protein synthesis. In addition, we found that Ac92 was not produced until 24 h p.i. Overall, these data are consistent with ac92 belonging to the late gene phase.
Deletion of ac92 led to a defect in infectious BV production. However, viral DNA replication was not affected, suggesting that a step after or not requiring viral DNA replication blocked BV production. The production of specific early and late gene protein products was also examined by immunoblotting. Production of neither IE-1 nor VP39 appeared to be affected by ac92.
Previous studies showed that Ac92 was associated with the ODV (3, 8, 18). Our results showed that Ac92 was a membrane component of BV and was associated with both the nucleocapsid and envelope of ODV. However, analysis of the sequence of Ac92 does not predict a transmembrane domain or signal sequence. Ac92 may be membrane associated.
During virus infection, an intranuclear viral replication center, the VS, is formed. Viral DNA replication and late gene transcription take place in the electron-dense VS (39). During the late phase of infection, viral DNA is condensed and packaged into capsids to form nucleocapsids within the VS. These nucleocapsids egress from the cell nucleus to bud through a modified plasma membrane to form BV. Later in infection, nucleocapsids remain in the nucleus and become bundled together, and it is thought that they acquire envelopes from intranuclear microvesicles to form ODVs. The resulting ODVs are embedded within large proteinaceous paracrystalline occlusion bodies (36). Our data showed that the deletion of ac92 did not affect the formation of VS, nucleocapsid assembly, the formation of intranuclear microvesicles, or polyhedra (Fig. (Fig.7).7). Interestingly, when cells were transfected with DNA from Ac92KO-PG, multiply enveloped nucleocapsids of the ODV were not observed; instead, we observed what appeared to be singly enveloped nucleocapsids in the ring zone associated with intranuclear microvesicles. Consistent with this result, Ac92 localizes to the ring zone of infected cells (D. A. Theilmann, personal communication). It would be interesting to know if these singly enveloped nucleocapsids are orally infectious to insect larvae and how their infectivity compares to that of multiply enveloped nucleocapsids.
Alignments between Ac92 homologs predicted in the genomes of all sequenced M and S phenotype baculoviruses did not show any significant differences that could provide insight into a relationship between the gene and the phenotype (results not shown). Although lack of ac92 affects virion assembly resembling S-like phenotype virions, since ac92 is carried in both S and M phenotype baculoviruses, we do not think that it is a direct determinant of the M virion phenotype. There is no information available concerning factors affecting the M virus phenotype.
A previous study determined that Ac92 had a 90% probability of being an FAD-binding sulfhydryl oxidase related to the ERV/ALR family (24). This relationship was based on the prediction of helical domains surrounding a C-X-X-C motif with potential oxidoredox activity (15). Ac92 was expressed and purified, and its biophysical and enzymatic properties were characterized to confirm that it indeed encoded an FAD-containing sulfhydryl oxidase (15). In the present study, mutation of the C-X-X-C sulfhydryl oxidase motif in ac92 abolished the ability of Ac92 to oxidize DTT, confirming that the cysteine motif is important for sulfhydryl oxidase activity in Ac92 and is essential for function. Our data showed that Ac92HAM1-PG had characteristics similar to those of the knockout virus. Although we cannot rule out that this mutation resulted in a misfolded protein, it suggests that the sulfhydryl oxidase activity might play a major role in virus formation and infectivity.
In vaccinia virus, three proteins, E10R, A2.5L, and G4L, comprise a complete pathway for the formation of disulfide bonds in intracellular virion membrane proteins (26-28, 35). The upstream component, E10R, belongs to the ERV1/ALR family of FAD-containing sulfhydryl oxidases, participates in disulfide bond formation, is associated with virion membranes, and has a role in virion morphogenesis (26, 27). The second component, A2.5L, is a small α-helical protein with a C-X-X-X-C motif that forms a stable disulfide-linked heterodimer with E10R and a transient disulfide-linked complex with the third component, G4L (28). G4L is a thioredoxin-like protein that directly oxidizes thiols of L1R, the virion membrane protein (28); F9L, a virion structurally related protein (28); and A28L, another virion membrane-associated protein (25). These proteins involved in a disulfide bond formation pathway are conserved throughout the poxvirus family, suggesting that the pathway is an ancestral mechanism in this group of viruses (28). In African swine fever virus, the protein pB119L, with a C-X-X-C motif similar to that of E10R and belonging to the ERV/ALR family of sulfhydryl oxidases, has been described as a late nonstructural protein required for virion maturation (12). Biochemical experiments showed that pB119L is an FAD-containing sulfhydryl oxidase that interacts with a viral protein, A151R. A151R was found to interact with the viral structural protein pE248R, which contains disulfide bridges and belongs to a class of myristoylated proteins related to vaccinia virus L1R (21, 22). Vaccinia virus replicates in the cytoplasm of infected cells in contrast to baculovirus nuclear replication. It is interesting that both viruses require sulfhydryl oxidase activity for virus morphogenesis in distinct nonoptimal redox cellular compartments.
Ac92 has redox functions necessary for virus morphogenesis. Ac92 shares functional properties and has localization similar to those of other viral sulfhydryl oxidases. Substrates of Ac92 will help to elucidate the mechanisms and requirements for virion assembly in baculoviruses.
We thank Kai Yang (Sun Yat-sen University) for his generous gifts of pFB1-PH-GFP plasmid and the VP39 polyclonal antiserum and Linda Guarino (Texas A&M) for IE-1 antibodies. We also thank George R. Rohrmann (Oregon State University) for discussions and David A. Theilmann for communicating results prior to publication.
This research was supported by the United States Department of Agriculture 2008-35302-18849 award.
This is contribution number 10-388-J from the Kansas Agricultural Experiment Station.
Published ahead of print on 22 September 2010.