Abstract
Nucleopolyhedroviruses (NPVs), family Baculoviridae, are insect-specific viruses with the potential to control insect pests in agriculture and forestry. NPVs are occluded in polyhedral occlusion bodies. Polyhedra protect virions from inactivation in the environment as well as assisting virions in horizontal transmission in the insect population. The process of virion occlusion in the polyhedra is undefined and the genes that regulate the virion occlusion process have not been well investigated yet. An Autographa californica multiple nucleopolyhedrovirus (AcMNPV) mutant (AcDef) that has a 2136 bp DNA deletion, including p26, p10 and p74 genes, has been isolated. No virions were detected in the polyhedra of AcDef. Restoration of all the missing sequences into AcDef led to proper virion occlusion. Individual gene deletion of either p10 or p26 could not abolish virion occlusion in the polyhedra of AcMNPV, but p10 deletion reduced virion occlusion efficiency more than threefold compared with the wild-type AcMNPV. Previous studies by other research groups on deletion of AcMNPV gene p74 suggested that p74 is a per os infectivity factor, and deletion of the p74 gene did not eliminate virion occlusion. Collectively, the three genes (p26, p10 and p74) may act in concert to regulate the virion occlusion process. Therefore, p26, p10 and p74 are all required for proper virion occlusion in the polyhedra of AcMNPV.
-
↵†Present address: Entomology Department, 420 Bio Science Building, University of Georgia, Athens, GA 30602, USA.
-
↵‡Present address: Entomology and Nematology Department, IFAS, University of Florida, Gainesville, FL 32611, USA.
-
A supplementary figure and two tables are available with the online version of this paper.
INTRODUCTION
The polyhedrin gene of nucleopolyhedroviruses (NPVs), in the family Baculoviridae, is highly active in the late phase of viral infection in permissive insect cells; it synthesizes large quantities of polyhedrin protein that form polyhedra or occlusion bodies (OBs) to occlude the newly assembled virus particles or occlusion-derived virus (ODV). The polyhedra protect ODV from inactivation by natural hazards, such as UV irradiation, and allow it to infect and kill insects. The polyhedra are also responsible for the horizontal transmission of baculoviruses in insect populations in the environment (Federici, 1997). Natural infection of baculoviruses starts when insects feed on vegetation contaminated with virion-containing polyhedra from a previous infection cycle. The polyhedra are ingested and dissolved in the alkaline environment of the insect midgut, releasing the virions which subsequently enter the midgut cells and replicate to produce budded virus (BV). BVs then traverse the basal lamina to the haemocoel and begin infecting tracheal and fat body cells (Granados & Lawler, 1981). It has also been reported that an Autographa californica multiple nucleopolyhedrovirus (AcMNPV) infects the tracheoblasts of Trichoplusia ni larvae to pass through basal laminae and spread infection in the host (Engelhard et al., 1994). Multiplication of viruses within host tissues eventually kills the insects, and polyhedra from killed insects are released on the vegetation to initiate another cycle of infection (Vlak & Rohrmann, 1985). Polyhedra without virions are unable to kill insects, as such, polyhedra must contain as many virions as possible to maximize their effectiveness as biological insecticides. The mechanism and genetic basis of the occlusion process are unknown and the frequency of polyhedra without occluded virions in nature has never been systematically investigated.
Polyhedra without virions are often observed in infected cells in cell culture (Goodwin & Adams, 1980). This might be due to a lack of selection pressure in a cell culture system compared with nature. For example, a mutant forest insect baculovirus Lymantria dispar NPV that produces polyhedra without occluded virions was isolated from tissue culture, but the cause of the non-virion occlusion phenomenon has not been identified (Slavicek et al., 1998). Virion-free polyhedra also occurred in vivo or in insects. When a Pseudoplusia includens larva had been infected by Thysanoplusia orichalcea NPV (ThorNPV), ThorNPV polyhedra occluded virions in one infected tracheal tissue cell but not in the neighbouring one (Cheng & Carner, 2000). The shape and polyhedral structure of the two viral populations were indistinguishable, which suggests that the non-virion occlusion may not have resulted from mutation of the polyhedrin. Therefore, it was suggested that the virion-negative polyhedra were caused by mutations in genes other than the polyhedrin gene in the viral genome (Cheng & Carner, 2000). In addition, early reports suggested that mutations in the polyhedrin or FP25K genes also impaired virion occlusion in polyhedra (Carstens et al., 1986; Fraser & Hink, 1982).
Although the virion occlusion mechanism is not clear, there are indications that the baculovirus virion envelope has a strong affinity to polyhedrin (Blissard & Rohrmann, 1990). In Spodoptera frugiperda multiple NPV (SfMNPV), deletion of open reading frame (ORF) 29 (or Sf29) reduced virion occlusion efficiency by sixfold compared with wild-type (wt) SfMNPV in viral DNA isolation assays; however, Sf29 is not an ODV envelope protein (Simon et al., 2008b). In this paper, we describe an AcMNPV mutant whose polyhedra contain no virions. We also provide evidence from gene restoration and gene knockout studies that AcMNPV p26, p10 and p74 gene products are required for proper virion occlusion.
METHODS
Cells, insects and viruses.
Sf21 cells were used throughout this study and maintained at 27 °C in Grace's insect media supplemented with 10 % fetal bovine serum. Insects used for recombinant virus OB production (T. ni and S. frugiperda) were reared on artificial diets purchased from Southland Products. Wt AcMNPV (C6) and a commercial recombinant virus AcUW1-lacZ, containing a copy of the lacZ gene at the p10 locus (Pharmingen), were propagated in Sf21 cells for BV production. Polyhedra were produced by intrahaemocoelic injection of insect larvae with BVs from cell culture (Wang et al., 2008).
Screening for defective AcMNPV (AcDef) incapable of infecting larvae per os.
In order to express the Bacillus thuringiensis (Bt) CryIA(c) gene at the p10 locus using the AcUW1-lacZ vector (Pharmingen) as a parental virus, white plaques not showing lacZ activity were selected and the virus was amplified in Sf21 cells. At day 5 post-infection (p.i.), cells containing OBs were pelleted and used to infect third instar T. ni larvae (Cheng et al., 2005). Infected larvae were harvested for OB purification by sucrose gradient centrifugation (Cheng et al., 1990). The OB concentrations were estimated using a haemocytometer. To isolate AcDef, the OBs purified from larvae were fed to third instar T. ni larvae at a dose of 30 000 polyhedra per larva by using the diet plug method (Cheng et al., 2001); mortality was recorded.
Identification of deleted gene sequences.
Since the flanking genes of the p10 locus (p26 and p74) were targeted for homologous recombination, there was a chance that these genes might have been deleted. Primers specific to the p26 and p74 genes were designed to determine whether both genes were intact and free of mutations (all primer sequences are given in Supplementary Table S1, available in JGV Online). The primer pair Acp26F and Acp26R was used for p26. The forward primer Acp26F is located in the hr5 region upstream of the putative p26 promoter (Fig. 1⇓). The primer pair Acp74F and Acp74R was used for p74. AcMNPV polyhedrin primers polh-F and polh-R5 were used as a control to confirm the presence of AcDef DNA as a template in the PCR amplification. The negative control contained no template DNA and the positive control contained AcUW1-lacZ DNA. PCR products were analysed by agarose gel electrophoresis.
Schematic of wild-type AcMNPV and the recombinant viruses that were used to investigate genes regulating virion occlusion in polyhedra. Mutated genes are indicated in parentheses for each viral construct. Hatched boxes in AcΔp10-GFP denote the remaining sequences of the p10 gene after insertion of the gfp expression cassette. Primers (Supplementary Table S1) used for PCR to detect gene integrity in AcDef are shown.
To pinpoint the missing sequences in AcDef, the primers flanking p26 and p74 of AcMNPV (Acp26F and Acp74F; Fig. 1⇑) were used for PCR amplification using template and controls as above. PCR products were visualized by agarose gel electrophoresis then purified and cloned into pGEM-T Easy vector (Promega) and sequenced. The sequence was compared with the AcMNPV genome sequence using Lasergene (dnastar) to identify missing sequences.
Restoration of missing sequences.
To restore the missing DNA sequences in AcDef, a transfer vector was constructed that contained 3.4 kb insert DNA, including p10 and the flanking regions from nt 117038 to 120464 of AcMNPV (Ayres et al., 1994). This 3.4 kb insert was amplified using primers P5 and P6 and wt AcMNPV genomic DNA template in a standard PCR; this was cloned into pGEM-T Easy (Promega) to construct plasmid pT3.4K. A 0.9 kb DNA fragment containing the AcMNPV polh promoter and the green fluorescent protein (GFP) gene excised from pBlueGFP (Cheng et al., 2001) was inserted between the BglII and HindIII sequences of p10 in pT3.4K to construct the transfer vector pT3.4KGFP for GFP expression (Fig. 1⇑). The transfer vector was confirmed by DNA sequencing.
Sf21 cells were co-transfected with pT3.4KGFP and AcDef genomic DNA. A plaque assay was performed to isolate recombinant virus AcΔp10-GFP. Next, Sf21 cells were co-transfected with plasmid pT3.4K and AcΔp10-GFP genomic DNA to isolate GFP-negative and polyhedra-positive plaques that contained all the sequences missing from AcDef. The recombinant virus was designated AcBack.
Inactivation of p26.
Both p10 and p74 had previously been mutated in AcMNPV and were reported not to impair virion occlusion (Faulkner et al., 1997; Kuzio et al., 1989; van Oers et al., 1993). Deletion of the AcMNPV p26 gene (ORF136) had no effect on viral infectivity in cell culture and larvae or on BV and OB production (Simon et al., 2008a). Effects of p26 deletion on virion occlusion in polyhedra have not been studied.
A transfer vector with a GFP expression cassette flanked by the up- and downstream sequences of the p26 gene was constructed to delete the p26 gene ORF. Two primer pairs were used to amplify two 1 kb homologous recombination DNA fragments up- and downstream of p26. These were amplified by PCR with primers P26up-F and P26up-R (containing HindIII and BamHI site sequences, respectively) and primers p26dn-F and p26dn-R (contain XhoI and EcoRI site sequences, respectively), respectively. The up- and downstream fragments and the GFP expression cassette from pBlueGFP (Cheng et al., 2001) were assembled and cloned between the HindIII and EcoRI sites of pUC19 to generate a transfer vector pAcΔp26-GFP. pAcΔp26-GFP and wt AcMNPV DNA were used to co-transfect Sf21 cells to generate recombinant virus AcΔp26-GFP, which contained a p26 gene deletion with the GFP gene insertion, by plaque purification (O'Reilly et al., 1992).
OB DNA analyses.
DNA was extracted in triplicate from equal numbers of OBs (5×105) of wt AcMNPV, AcUW1-lacZ, AcDef, AcΔp10-GFP, AcBack and AcΔp26-GFP using proteinase K treatment and phenol extraction (O'Reilly et al., 1992). The amounts of DNA in the virions of the polyhedra were compared. DNA was dissolved in 20 μl TE buffer (pH 8.0). DNA solution (1 μl) from each of the samples was analysed by agarose gel electrophoresis with λ phage DNA (100–1 ng) standards as reference. DNA yields were quantitatively estimated using a NanoDrop spectrophotometer.
To detect trace amounts of DNA in the AcDef polyhedra, real-time quantitative PCR (qPCR) was used. A serial dilution of wt AcMNPV DNA (from 5 to 5×10−4 ng μl−1) was used as a template in the qPCR to construct a standard curve using polh gene sequence primers polh-F and polh-R1 to generate a 100 bp amplicon (Wang et al., 2008). qPCR was performed in triplicate. Differences in DNA yields among different viral constructs were statistically analysed by using anova by Tukey's comparisons (using Minitab).
SDS-PAGE and immunoblotting.
Equal numbers (5×105 OBs in 10 μl) of OBs of different viral constructs were mixed with an equal volume of alkaline solution (0.1 M Na2CO3, pH 10.5) and incubated at 22 °C until the solution became clear. A solution of each OB sample was divided equally and separated using two identical 12 % acrylamide SDS-PAGE gels. One gel was stained with Coomassie blue and the other was used to transfer the proteins onto a nitrocellulose membrane for Western blot analysis with a monoclonal antibody (kindly provided by Dr Loy Volkman, University of California, Berkeley) against AcMNPV VP39 (a major capsid protein). Bonded anti-VP39 antibodies were detected with a horseradish peroxidase (HRP)-conjugated anti-mouse IgG antibody and chromogenic substrates (Bio-Rad). The SDS-PAGE and Western blots were performed in triplicate. Virion occlusion efficiencies of AcDef and its derivatives (AcΔp10-GFP and AcBack) were analysed further using 1.5×106 OBs of each recombinant virus. The SDS-PAGE and immunoblotting analyses were similar to that described above except an alkaline phosphatase (AP)-conjugated anti-mouse IgG antibody and BCIP/NBT were used. Virion contents in polyhedra were compared by densitometry. The integrated density values (IDV) of both VP39 from Western blot and polyhedrin from Coomassie blue stains were measured using AlphaImager 2200 to identify differences in virion contents in the polyhedra of different viral constructs. These were statistically analysed by using anova by Tukey's comparisons.
Electron microscopic analysis of OBs.
BVs of each virus (wt AcMNPV, AcUW1-lacZ, AcDef, AcΔp10-GFP, AcBack and AcΔp26-GFP) were used to infect Sf21 cells at an m.o.i. of 5 p.f.u. per cell. At day 3 p.i., cells were harvested and processed for transmission electron microscope (TEM) examination of virion occlusion (Granados & Lawler, 1981).
RESULTS
Isolating AcDef
In a study aimed at constructing a Bt toxin expressing baculovirus with AcUW1-lacZ as a parental virus, 19 putative recombinant plaques were picked and used to infect Sf21 cells. At day 5 p.i., infected cells were pelleted to infect T. ni larvae per os. All larvae were killed by the viral infections and OBs were observed in tissues such as the fat bodies of larvae. When OBs from these viral plaques were used to infect third instar larvae at a single high dose (30 000 polyhedra per larva), the OBs of three viral plaques (AcDef plaques 1, 3 and 16) could not cause mortality to T. ni larvae, whereas the OBs of the other viral plaques killed all the larvae tested. However, AcDef plaques 1, 3 and 16 replicated normally in the Sf21 cell line (data not shown).
Identifying the gene sequences deleted from AcDef
Three genes, including p26, p10 and p74, were not intact in AcDef. The p26 gene primer pair amplified the expected 1 kb amplicon when AcUW1-lacZ DNA was used as a template, whereas no amplicon was produced when AcDef was used as a template (Fig. 2a⇓). Similar results were obtained with the p74 gene in AcDef (Fig. 2a⇓). This suggests that neither p26 nor p74 were intact in the AcDef genome. When primers Acp26F and Acp74F were used to amplify AcDef, a 1 kb PCR product was obtained, whereas a 3 kb fragment was produced when wt AcMNPV DNA was used as a template (Fig. 2b⇓). Therefore, an ∼2 kb DNA sequence was missing from the AcDef genome compared with wt AcMNPV.
PCR analysis of gene deletion in AcDef. (a) PCR amplification of p26 and p74 genes using gene-specific primers (see Fig. 1) with AcUW1-lacZ (AcU) and AcDef DNA as templates. The AcMNPV polh primer pair was used to confirm the presence of DNA in the PCR. (b) PCR detection of sequence deletion in AcDef. Acp26F and Acp74F primers were used to amplify AcDef and wt AcMNPV DNA. A mock amplification used water in place of a DNA template as a no-template control (NTC).
To determine the exact sequences deleted from the AcDef genome, the 1 kb PCR fragment from AcDef was cloned into pGEM-T vector. Sequencing data analysis showed that there was a 2136 bp DNA deletion in the AcDef genome, which corresponds to nt 117 989–120 124 in the wt AcMNPV genome sequence (Ayres et al., 1994). In addition to the lack of the p10 gene in AcUW1-lacZ, the p26 ORF and part of the upstream sequences were missing from AcDef. Furthermore, the 3′ end of the p74 gene (988 bp) was also lost (Fig. 1⇑). The hr5 and p35 genes were not affected by this deletion and remained intact (Fig. 1⇑). The p10 gene is flanked by the p26 and p74 genes in AcMNPV; in AcUW1-lacZ, the p10 ORF is substituted with lacZ, therefore, AcDef lacks not only the p10 gene but also lacZ, which has been deleted from the parental AcUW1-lacZ (Fig. 1⇑). The impact of p26, p10 and p74 gene deletions in AcMNPV on virion occlusion was investigated further.
Virion occlusion did not occur in AcDef
Three independent experiments were performed to examine the significance of the deletion of p26, p10 and p74 on virion occlusion of AcMNPV; it was found that AcDef is defective in virion occlusion.
Firstly, AcDef, wt AcMNPV and AcUW1-lacZ OBs were used for DNA isolation. Using ethidium bromide staining, DNA was not detected from AcDef OBs but was detected from both wt AcMNPV and AcUW1-lacZ OBs; the DNA concentration was higher in wt AcMNPV (Fig. 3a⇓). As control DNA was detected using 5 ng ethidium bromide ml−1 (Fig. 3a⇓), this suggests that the DNA isolated from AcDef OB was less than 5 ng μl−1. Spectrophotometry showed a statistically significant sixfold reduction of OB DNA in AcUW1-lacZ compared with wt AcMNPV (Fig. 3b⇓); no DNA was detected in AcDef using this method. This suggests that the p10 gene deletion in AcUW1-lacZ had an impact on virion occlusion of AcMNPV. qPCR detected 2.2 ng AcDef DNA μl−1 from AcDef OB DNA isolation (see Supplementary Fig. S1 and Supplementary Table S2, available in JGV Online). Approximately 44 ng DNA was obtained from 5×105 AcDef polyhedra, demonstrating an approximately 68-fold reduction in DNA yields from polyhedra compared with those from wt AcMNPV (Fig. 3b⇓). This detected DNA might have been purified from non-occluded virions attached to the surface of polyhedra during the polyhedron purification (Fig. 5c⇓). Therefore, the DNA isolation assay suggested that AcDef had few or no virions occluded in the polyhedra.
Polyhedra DNA content analysis of different viral constructs (Fig. 1) to understand the genes regulating virion occlusion in polyhedra. (a) DNA extracted from the same number of polyhedra (5×105 OB) of viral constructs with the indicated gene deletions to compare the virion occlusion efficiency. DNA in equal volumes (1 μl) of solution of a each viral DNA was separated. Bacteriophage λ DNA of known concentration was used as a reference. (b) Spectrophotometry of viral DNA yields from polyhedra of the viral constructs. The AcDef DNA yield was obtained by qPCR (see Supplementary Fig. S1). Error bars represent sem (n=3). Bars with different letters are statistically different with 95 % confidence.
Secondly, Western blot analysis (using an anti-VP39 antibody) detected VP39 or virions from OBs of wt AcMNPV and AcUW1-lacZ but not in the polyhedra of AcDef, when 5×105 OBs were used (Fig. 4a⇓). However, VP39 antibodies detected a faint VP39 or virion signal in AcDef OBs when 3× more OBs (1.5×106) were used (Fig. 4b⇓). AcUW1-lacZ and AcDef showed significant 3-fold and 12-fold reductions in virion occlusion efficiency, respectively, compared with that of wt AcMNPV (Fig. 4c⇓). This supported the notion that AcDef contained few or no virions or viral DNA in the polyhedra and that the p10 gene deletion reduced virion contents in OBs.
SDS-PAGE and Western blot analysis of virion occlusion efficiency in polyhedra. (a) Equal amounts of polyhedra (2.5×105 OBs) from each of the viral constructs were processed for SDS-PAGE (upper panel). A duplicate of the gel was used for Western blot detection of a major capsid protein VP39 with anti-VP39 monoclonal antibodies, using the HRP detection system (lower panel). (b) Three times more OBs of AcDef and its derivatives than in (a) (7.5×105 OBs) were processed for SDS-PAGE (upper panel) and Western blot detection (lower panel) as before, but using the AP detection system. (c) Quantitative analysis of virion occlusion efficiency. Densitometry was used to measure the integrated density values (IDV) of VP39 signals from the Western blot and polyhedrin from the SDS-PAGE. IDVs of VP39 relative to those of polyhedrin in experiments for (a) were used to compare virion contents in polyhedra of the different viral constructs. AcDef and AcΔp10-GFP values were calculated from the experiments in (b). Error bars represent sem (n=3). Bars with different letters are statistically different with 95 % confidence.
Thirdly, TEM examination for the presence of virions in the OBs showed that wt AcMNPV replicating in the nuclei of Sf21 cells assembled polyhedra with occluded virions (Fig. 5a⇓). However, no virions were found in the polyhedra of AcDef (Fig. 5b⇓); instead, these were found outside the OBs of AcDef (Fig. 5c⇓).
TEM analysis of virion occlusion in polyhedra. (a) wt AcMNPV with proper virion occlusion in Sf21 cells. (b) AcDef in Sf21 without virions in polyhedra. (c) AcDef in Sf21 showing virions surrounding empty polyhedra. (d) Polyhedra containing virions after p26 and p74 genes were restored in AcΔp10-GFP. Bars, 1 μm.
Restoring missing sequences restored virion occlusion
In order to confirm that the missing AcDef sequences were responsible for the empty polyhedra phenotype of AcDef, the missing AcDef sequences were restored, which resulted in virion occlusion in the polyhedra. AcΔp10-GFP was constructed, with AcDef as the parent, to restore the p26 and p74 loci. This contained a GFP expression cassette at the p10 locus, which allowed the recombinants to be screened. AcΔp10-GFP was later used as a parent to restore the complete missing sequences in AcDef to produce AcBack.
An OB viral DNA isolation assay indicated that AcΔp10-GFP and AcBack contained virions in the OBs; AcΔp10-GFP OBs contained noticeably less DNA or virions than OBs from wt AcMNPV and AcBack (Fig. 3a⇑). Spectrophotometry showed that the DNA yield from AcΔp10-GFP OBs was significantly lower (about sixfold) than that from wt AcMNPV OB (Fig. 3b⇑) and was similar to that of AcUW1-lacZ, which also has a p10 gene deletion (Fig. 3a, b⇑). DNA and virion contents in the OBs were significantly reduced in both of these. When gfp was replaced with p10 in AcΔp10-GFP, to generate AcBack (Fig. 1⇑), the DNA content of OBs increased to the levels in wt AcMNPV (Fig. 3a, b⇑). This suggests that the p10 gene deletion reduces DNA content or virion occlusion of AcMNPV OBs.
These data were confirmed by Western blot analysis with the anti-VP39 antibody. This detected virions in the OBs of AcBack, but did not detect VP39 from 2.5×105 AcΔp10-GFP OBs (Fig. 4a⇑). When 3× more OBs (1.5×106) were used in the SDS-PAGE and Western blot analyses, virions were detected in the OBs of AcΔp10-GFP (Fig. 4b⇑). Quantitative virion occlusion efficiency analysis by SDS-PAGE and Western blot showed that deletion of the AcMNPV p10 gene significantly reduced virion occlusion efficiency in polyhedra by fourfold. Restoration of all the missing sequences in AcDef (AcBack) resulted in normal virion occlusion, with no significant differences in virion contents compared to wt AcMNPV (Fig. 4c⇑). This also suggested that no other region of AcDef had mutations which led to the defective virion occlusion in polyhedra and that the p26, p10 and p74 genes might be responsible for proper virion occlusion.
Deletion of p26 alone could not abolish virion occlusion
Since AcMNPV strains with deletions in either p10 or p74 still produced polyhedra with virions occluded (Faulkner et al., 1997; Kuzio et al., 1989; van Oers et al., 1993), it seemed that deletion of the p26 gene should produce polyhedra with a clear reduction in the number of occluded virions. However, the p26 gene knockout experiment suggested that deletion of the p26 gene alone had no negative effect on virion occlusion in the polyhedra of AcMNPV.
We observed that virions were occluded in the polyhedra of AcΔp26-GFP (p26 replaced with the GFP expression cassette) in both a viral DNA extraction assay and a Western blot assay with an anti-VP39 antibody (Figs 3⇑ and 4⇑). In the DNA extraction assay, no significant differences in DNA yields were observed between AcΔp26-GFP, AcBack and wt AcMNPV OBs, even though AcΔp26-GFP showed lower DNA content in OBs than wt AcMNPV and AcBack. Western blot analysis supported these results, indicating that the single p26 gene deletion could not impair virion occlusion of AcMNPV. SDS-PAGE and Western blot assays also showed that AcΔp26-GFP had a similar virion occlusion efficiency to wt AcMNPV and AcBack (Fig. 4⇑). TEM examination confirmed that AcΔp26-GFP had virions occluded in the polyhedra (data not shown).
DISCUSSION
NPV polyhedra that contain no virions have been reported in the literature, but genes that play roles in the virion occlusion of baculoviruses have not been identified in AcMNPV. AcMNPV contains about 150 genes (Ayres et al., 1994), but reverse genetics by knockout of specific genes may be tedious as there are no clues as to which gene should be mutated in order to find those that regulate the virion occlusion process. We were fortunate to isolate an AcMNPV mutant (AcDef) that is defective in virion occlusion and has a large piece of DNA deleted. This allowed us to use forward genetics to identify the genes deleted in the AcDef genome and eventually to restore the missing genes to confirm that these were responsible for virion occlusion in polyhedra. We were able to successfully produce polyhedra with virions occluded. Understanding the genes that control virion occlusion may enable the development of baculoviruses with more virions occluded; this would allow more efficient insect control, allowing baculoviruses to compete with chemical insecticides for pest control in agriculture. This may also lead to an understanding of the evolution of virion occlusion in baculoviruses. By using three independent approaches, we found that the concerted action of three AcMNPV genes, p26, p10 and p74, is required for proper virion occlusion in the polyhedra of AcMNPV. The three genes were found in a cluster in AcMNPV (Ayres et al., 1994) and several other NPVs, including Helicoverpa armigera single nucleocapsid NPV (Chen et al., 2001), Spodoptera exigua multiple NPV (IJkel et al., 1999), Bombyx mori NPV (Gomi et al., 1999) and Choristoneura fumiferana defective NPV (Lauzon et al., 2005).
Baculoviruses replicating in the nuclei of insect cells synthesize large amounts of polyhedrin and P10. Newly assembled virions are occluded in the polyhedra as polyhedrin crystallizes to form polyhedra. If the virions are not occluded at this stage, virions cannot enter the polyhedra later. It is therefore hypothesized that virion envelope proteins may have an affinity for polyhedrin (Blissard & Rohrmann, 1990). Of the three genes that have been completely or partially deleted in the AcDef genome, p74 is a virion envelope protein that is involved in per os infectivity in larvae (Faulkner et al., 1997). We did not perform the p74 gene knockout for virion occlusion studies, but previous work has shown that deletion of p74 abolishes the infectivity of OBs per os and that virions are occluded in the OBs (Faulkner et al., 1997). It is likely that the p74 gene of AcMNPV has an additional function in virion occlusion, as individual gene deletion of either p26 or p10 cannot abolish virion occlusion, but AcDef with the three genes (p26, p10 and p74) completely or partially deleted contains few or no virions in OBs (Figs 3⇑, 4⇑ and 5⇑). The function of the p10 gene is not well-defined but reports have suggested that P10 protein is associated with polyhedra (Quant-Russell et al., 1987; Carpentier et al., 2008). Deletion of p10 results in OBs incapable of being released from nuclei (Williams et al., 1989). In this study, deletion of p10 still produced OBs with virion occlusion, but the efficiency of virion occlusion dropped substantially (Figs 3⇑ and 4⇑). This suggests that another function of the AcMNPV p10 gene is its involvement in virion occlusion. The p10 gene locus of AcMNPV was explored to insert insecticidal genes for improved recombinant viral insecticide development (McCutchen et al., 1991). In fact, the p10 deletion mutant AcUW1-lacZ increased LD50 to third instar T. ni by ninefold compared with wt AcMNPV (Cheng et al., 2001, 2005). Based on the results obtained in this investigation, the p10 gene should not be interrupted for viral insecticide development as its deletion leads to reduced virion occlusion efficiency, which may reduce the per os infectivity of the recombinant viruses to insects.
The p26 gene of AcMNPV was first characterized at the transcriptional level (Liu et al., 1986). Proteomic analyses of AcMNPV ODV proteins did not locate P26 in the ODV (Braunagel et al., 2003). Furthermore, deletion of the AcMNPV p26 gene did not reveal any biological function (Simon et al., 2008a). We first hypothesized that the p26 gene is the only gene required for virion occlusion in polyhedra of AcMNPV, as deletion of either p10 or p74 could not produce OBs without virions (Faulkner et al., 1997; van Oers et al., 1993). However, when the AcMNPV p26 gene is knocked out, virion occlusion efficiency in OBs is similar to that of wt AcMNPV (Figs 3⇑ and 4⇑). Our studies also confirm that p26 is dispensable and not required for AcMNPV DNA replication (Simon et al., 2008a).
Since AcDef OBs contain no virions, and deletion of the p10 gene significantly reduced the virion occlusion efficiency but could not completely abolish virion occlusion in polyhedra (Figs 3⇑ and 4⇑), it is plausible that p10 and p26, as well as p74, work together to regulate the virion occlusion process. Further double gene deletion of the p26, p10 and p74 genes is required to fully address the non-occlusion phenomenon in AcDef. Nevertheless, we conclude that all of the three genes p26, p10 and p74 are essential for optimal virion occlusion in AcMNPV.
Acknowledgments
We thank Dr Basil Arif for his help during the early stages of this project. Andy Brownwright and Blair Wormer are credited for helping with TEM work. C. M. S. K. is supported by an NEIST-DBT biotechnology overseas fellowship from the Government of India. This research is partially supported by an Ohio Plant Biotechnology Consortium grant (401120) awarded to X. W. C. and a start-up fund to X. W. C. from the College of Arts and Science, Miami University.