Introduction

Members of the Baculoviridae family can infect larvae within the orders Lepidoptera, Hymenoptera, and Diptera; the family consists of four genera, namely, the alpha-, beta-, gamma-, and delta-baculoviruses [1]. Alphabaculoviruses are Lepidoptera-specific nucleopolyhedroviruses (NPVs) that can be divided into two groups known as group I and group II as based on phylogenetic studies [2, 3]. The Autographa californica multiple nucleopolyhedrovirus (AcMNPV) genome is approximately 134 kbp and contains 154 predicted open reading frames (orfs) [4]. Comparative analysis of the 40 or so baculovirus genomes completely sequenced to date has revealed 30 conserved core genes. These genes are central to DNA replication, RNA transcription, nucleocapsid assembly, and budding off of the virus [5]. There are some unique genes that only exist in the group I NPV, including gp64, ie2, ptp1, odv-e26, ac5, ac30, ac72, ac73, ac114, ac124, and ac132 [5]. These genes have been found to be involved in various processes such as cellular attachment (gp64) [6], viral DNA replication and gene expression (ie2) [7], larvae locomotion (ptp1) [8], and transcription regulation (ac16) [9]. However, thus far, several unique genes (i.e., ac72, ac114, ac124, and ac132) remain uncharacterized.

AcMNPV orf114 (ac114) is predicted to encode a protein of 424 amino acids with a molecular mass of 49 kDa. Mass spectrometry has shown that the Ac114 protein is associated with AcMNPV BV and occlusion-derived virus (ODV) particles [10, 11]. Evolutionary rate analysis has shown that within the ac114 gene the average ω and K a values were around two times lower than that of the gp64 gene, suggesting that ac114 has evolved under relatively relaxed purifying selection [12]. Phylogenetic analysis has also revealed that Ac114 shares homology with a hypothetical protein from the human malaria parasite Plasmodium falciparum (3D7 strain) [13]. HHpred analysis predicted that Ac114 contains the general transcription factor II-I repeat domain that is involved in signal-dependent transcription regulation in various biological contexts [14]. However, information on the function of ac114 or any of the other homologs in the baculovirus life cycle is currently limited.

In this study, an ac114 knockout bacmid was generated using an AcMNPV bacmid. The effects of the ac114 knockout on the production of budded viruses (BVs) and occlusion bodies (OBs), temporal transcription, cellular localization, and the infectivity of OBs were examined.

Materials and methods

Virus, cell lines, and insects

The AcMNPV bacmid bMON14272 (Invitrogen Life Technologies, Carlsbad, CA, USA) was derived from AcMNPV, which could propagate in Escherichia coli DH10B cells. Sf9 cells were cultured at 27 °C in Grace’s medium (Invitrogen) supplemented with 10 % fetal bovine serum (Gibco, Grand Island, NY, USA). Spodoptera exigua larvae were reared on an artificial diet [15] at 28 °C.

Construction of AcMNPV bacmids containing enhanced green fluorescence protein (gfp) and polyhedrin

The ac114 knockout bacmid was constructed using the λ-recombination system in E. coli as previously described [16]. First, the chloramphenicol resistance gene (Cm) was amplified from the pKD3 plasmid by PCR with primers ac114-KO-F and ac114-KO-R (Table 1), which contained a 40 bp fragment homologous to one terminus of hr4B and the 3′ terminus of ac115, respectively, in accordance with the 97,886–97,925 and 99,121–99,160 nt positions within the AcMNPV genome (NC 001623.1) [4], to preserve the functions of hr4b and ac115. The gel-purified Cm cassette PCR fragment was electroporated into E. coli BW25113 harboring the AcMNPV bacmid bMON14272 and pKD46 (encoding the λ-Red recombinase). After incubation at 30 °C in 0.8 ml of SOC medium [17] for 2 h, the electroporated cells were spread onto agar plates with medium containing 50 μg/ml of kanamycin and 25 μg/ml of chloramphenicol. The plates were incubated at 37 °C for 2–3 days and colonies were selected for PCR analysis. The replacement of ac114 by the Cm cassette was confirmed using the primers ac114-F/ac114-R and Cm-F/Cm-R (Table 1). The recombination junctions between the Cm cassette and the flanking sequences of ac114 were examined using the ac114-F/Cm-R and Cm-F/ac114-R primers (Table 1). A 1,195 bp fragment of the ac114 coding region in the AcMNPV bacmid was replaced with the Cm cassette to produce AcMNPVac114KOph− (Fig. 1a).

Table 1 List of primers
Full size table
Fig. 1
figure 1

Generation of the ac114 knockout and wild-type AcMNPV bacmids. a Construction strategy for AcMNPVac114KOph−. The partial coding region of ac114 in AcMNPV bMON14272 was replaced with the chloramphenicol resistance gene (Cm) by homologous recombination in E. coli. b PCR confirmation of ac114 knockout bacmid. The primer pairs are indicated above each line. Sizes markers are indicated on the right-hand side of the panel. c Schematic diagram of the recombinant bacmids AcBac and AcBac-ac114KO, showing the polyhedrin (polh) and enhanced green fluorescent protein (gfp) genes inserted into the polyhedrin locus of AcMNPV bMON14272 or AcMNPVac114KOph− by Tn7-mediated transposition. d Verification of the ac114 knockout by RT-PCR. Total RNA was extracted from mock-infected Sf9 cells (Mock), or Sf9 cells infected with either vAcBac (wt) or vAcBac-ac114KO (KO). RT-PCR was conducted using the primers ac114-Orf-F and ac114-Orf-R flanking the ac114 knockout region, and ac114-IN-F and ac114-IN-R within the knockout region. DNA size markers are shown on the right-hand side

Full size image

The AcMNPV bacmid bMON14272 has an occlusion-minus phenotype [17]. To obtain bacmids containing the polyhedrin and gfp genes, the donor plasmid Dual-gfp-ph was constructed. In brief, the gfp PCR product, amplified from pEGFP-N1 (Clontech, Heidelberg, Germany) with the primers gfp-F and gfp-R (Table 1), was digested with XbaI and HindIII and ligated into pFastBacDual (Invitrogen), downstream of the polyhedrin promoter, to generate Dual-gfp. Subsequently, the polyhedrin PCR product, amplified from AcMNPV (C6) genomic DNA with the primers polyhedrin-F and polyhedrin-R (Table 1), was digested with SphI and NotI, and cloned into Dual-gfp under the control of a p10 promoter, to generate Dual-gfp-ph. Thereafter, the donor plasmid Dual-gfp-ph was transformed into electrocompetent DH10B cells containing the helper plasmid pMON7124 and bacmid AcMNPVac114KOph− (or bMON14272) to generate the bacmid AcBac-ac114KO (or AcBac) (Fig. 1c). After incubation at 37 °C for 4 h in 0.8 ml of SOC medium, DH10B cells were plated onto LB agar medium containing 50 μg/ml of kanamycin, 7 μg/ml of gentamicin, 10 μg/ml of tetracycline, 100 μg/ml of X-Gal, and 40 μg/ml of IPTG. The plates were incubated at 37 °C for 48 h, white colonies were selected, and the presence of the correct construct confirmed by PCR using universal M13-F/M13-R primers (Table 1).

Transfection of cells and virus growth curves

Sf9 cells (1.0 × 106 cells/35-mm-diameter plate) were transfected with 2.0 μg of the recombinant bacmids AcBac or AcBac-ac114KO using 8 μl of Cellfectin reagent (Invitrogen) according to the manufacturer’s instructions. After incubation for 4 h, the transfection supernatants were discarded and the cells were replenished with 2 ml of fresh Grace’s medium (Invitrogen) supplemented with 10 % fetal bovine serum, 100 μg/ml of penicillin, and 30 μg/ml of streptomycin. Green fluorescent protein (GFP) expression and OB formation were observed by fluorescence microscopy. Supernatants containing BVs (designated vAcBac or vAcBac-ac114KO) were collected at 72 h post-transfection (p.t.) and the BV titers were determined by end-point dilution assays in Sf9 cells.

To determine the kinetics of BV production, Sf9 cells (1.0 × 106 cells/35-mm-diameter plate) were infected with vAcBac or vAcBac-ac114KO at a multiplicity of infection (MOI) of 5. At designated time points (i.e., 0, 12, 24, 48, 72, and 96 h post-infection (p.i.)), the supernatants were harvested from the infected cells and centrifuged at 8,000×g for 5 min to remove the cell debris. The BV titers were determined by end-point dilution assays on Sf9 cells in triplicate using 96-well microtiter plates. The BV titers of the recombinant viruses at each time point were compared using a Student’s t test after logarithm transformation.

Transmission electron microscopy

Sf9 cells (1.0 × 106 cells/35-mm-diameter plate) were infected with each BV (vAcBac, vAcBac-ac114KO) at a MOI of 5. At 96 h p.i., the supernatants were discarded, the cells were then pelleted, fixed in 2.5 % glutaraldehyde at 4 °C overnight, washed three times with 0.1 M PBS (pH 7.2), followed by fixation in 1 % osmium tetroxide for 2–3 h at room temperature. After dehydration in graded ethanol (30–100 %), cells were soaked in acetone and embedded in spur resin. After staining with uranyl acetate and lead citrate, ultrathin sections were examined using a Hitachi H-800 transmission electron microscope (Hitachi Co., Ltd., Tokyo, Japan).

To determine whether a lack of ac114 had any effect on the number of ODVs embedded in one OB, OBs isolated from vAcBac- or vAcBac-ac114KO-infected S. exigua larvae were alkali-treated and the ODVs with an intact envelop were counted using electron microscopy as previously described [18]. ODV counts per OB of the two viruses were transformed into square roots to normalize the data and compared by a Student’s t test.

Total RNA extraction and reverse transcription (RT)-PCR

Sf9 cells (1.0 × 106 cells/35-mm-diameter plate) were infected with AcMNPV at a MOI of 5. Total intracellular RNAs were isolated from mock- and viruses-infected cells at 0, 6, 9, 18, 24, 48, and 72 h p.i. using TRIZOL (Invitrogen). The extracted RNA samples were treated with RNase-free DNase I (TaKaRa Biotechnology Co., Ltd., Dalian, China) to remove genomic DNA contamination. The First-strand cDNA was synthesized using reverse transcriptase (Invitrogen) and an oligo(dT) primer (TaKaRa) with 2 μg total RNA as template. The cDNA mixtures were amplified using the ac114 primers ac114-Orf-F/ac114-Orf-R (Table 1). Another reaction that included no RT prior to PCR amplification was used as a control to detect any possible viral DNA contamination. The AcMNPV lef3 gene was amplified with the primers aclef3-F/aclef3-R (Table 1) and used as a control. PCR products were analyzed on 0.8 % agarose gels.

The 5′ ends of the ac114 transcripts were determined using total RNA harvested at 24 h p.i. According to the manufacturer’s protocol from the 5′ full RACE kit (TaKaRa), the first-strand cDNA was synthesized with the random primer (TaKaRa). After the cDNAs were generated, the RACE-PCRs were conducted with the ac114-specific primer GSP1/5′ RACE outer primer (TaKaRa) followed by the GSP2/5′ RACE inner primer (TaKaRa). The PCR products were gel purified and cloned into pMD-18T (TaKaRa) for sequencing.

To verify the knockout of ac114 from AcMNPV, RNAs were isolated from vAcBac- or vAcBac-ac114KO-infected Sf9 cells as described above. RT-PCR was carried out using primers ac114-Orf-F/ac114-Orf-R flanking the ac114 knockout region and ac114-IN-F/ac114-IN-R (Table 1) spanning the ac114 knockout region.

Subcellular localization of Ac114 during viral infection

To monitor the localization of Ac114 in Sf9 cells, a recombinant bacmid containing the Ac114-GFP chimera was constructed. The ac114 orf (without the TAA stop codon) was amplified from the AcMNPV bacmid with primers ac114-Orf-F and ac114-Orf-R (Table 1). The ac114 PCR product was digested with EcoRI and XbaI and cloned into Dual-gfp-ph in frame with the gfp gene to generate Dual-ac114-gfp. Dual-ac114-gfp was transformed into DH10B cells, which contain the helper plasmid pMON7124 and bacmid AcMNPVac114KOph− to generate AcBac-ac114Rep (Fig. 7a). Sf9 cells were transfected with the bacmid AcBac-ac114Rep, and the BVs (designated vAcBac-ac114Rep) harvested from the supernatant of the transfected cells were used to infect Sf9 cells at a MOI of 5. The vAcBac-ac114KO construct was used as a control. At 96 h p.i., cells were stained with Hoechst 33258 and visualized with fluorescence microscopy (Zeiss, Oberkochen, Germany).

Bioassays

The median lethal dose (LD50) of the AcMNPV variants and median survival time (ST50) of the third-instar S. exigua larvae inoculated with the viruses were determined by a droplet-feeding assay as described previously [18]. The LD50 assay was conducted by incubating the groups of 48 larvae with different doses of the OB suspensions for 10 min. The larvae that ingested the OBs were reared at 27 °C and examined daily until all larvae had died or pupated. The ST50 was determined by feeding larvae with OBs at a concentration of 3 × 107 OB/ml. Mortality was recorded every 8 h until the remaining survivors pupated. These experiments were repeated in triplicate for each virus.

The LD50 values were determined using probit analysis (SPSS Inc., 2003) and compared with a relative median potency method. The ST50 was calculated using the Kaplan–Meier estimator and subsequently compared using the log-rank test (SPSS Inc., 2003) [19].

Results

Generation of the ac114 knockout and positive control AcMNPV bacmids

First, an ac114-null bacmid (AcMNPVac114KOph−) was generated, and the displacement of ac114 with the Cm cassette was examined by PCR (Fig. 1b). To test whether the ac114 knockout would affect OB morphogenesis and to facilitate examination of viral infection, the polyhedrin and gfp genes were introduced into the polyhedrin locus in AcBac-ac114KO or AcBac (Fig. 1c). RT-PCR was performed using the primer pairs flanking or within the ac114 knockout region; the results confirmed that ac114 had been successfully deleted from the original bacmid (Fig. 1d).

Analysis of AcBac and AcBac-ac114KO replication in transfected Sf9 cells

To determine whether the ac114 knockout had any effect on virus replication, Sf9 cells were transfected with AcBac or AcBac-ac114KO. The transfected cells were monitored by the fluorescence output from GFP expression. No difference was observed between the two viruses at 24 h p.t., indicating that the constructs had equal transfection efficiencies (Fig. 2a). For both bacmids, there were gradual increases in the number of infected cells at 72 h p.t., indicating that infectious BVs were produced from the cells that had been initially transfected (Fig. 2a). At 96 h p.t., fluorescence was observed in almost all AcBac or AcBac-ac114KO transfected cells (Fig. 2a). Light microscopy analysis at 96 h p.t. showed that OBs with normal appearance had formed in the Sf9 cells transfected with both bacmids (Fig. 2b). These results indicate that the ac114 knockout had no discernible effect on the production of infectious BVs, or the formation of OBs.

Fig. 2
figure 2

Analysis of viral replication in Sf9 cells. a Fluorescence microscopy of Sf9 cells transfected with AcBac or AcBac-ac114KO from 24 h p.t. to 96 h p.t. b Light microscopy of OBs in the AcBac or AcBac-ac114KO transfected Sf9 cells at 96 h p.t

Full size image

To quantify the effect of the ac114 knockout on virus replication and to determine the replication kinetics of the two viruses, virus growth curves were performed. Sf9 cells infected with the two different viruses showed a steady increase in BV production throughout the infection process (Fig. 3). At each time point, the titer of vAcBac-ac114KO was nearly equivalent to vAcBac (all p > 0.05). These results indicated that the ac114 knockout had no apparent effect on BV proliferation in the cultured cells.

Fig. 3
figure 3

Viral growth curve analysis. Sf9 cells were infected by vAcBac or vAcBac-ac114KO at a MOI of 5. The cell supernatants were harvested at various time points and the titers of the BVs were determined by TCID50 end-point assays. Each data point represents the average titer of three independent TCID50 assays. The bars represent the standard deviations

Full size image

Electron microscopy analysis

To examine whether the ac114 knockout would affect virus morphogenesis, electron microscopy observations were performed on the thin sections generated from the vAcBac-ac114KO-infected Sf9 cells. Rod-shaped nucleocapsids associating with the electron-dense edges of the virogenic stroma were observed in the nuclei of the vAcBac-ac114KO-infected cells (Fig. 4a). The bundles of nucleocapsids were aligning with de novo envelopes (Fig. 4b). OBs with normal shapes and sizes were observed; it was also noted that the ODVs had been occluded into the polyhedra within the ring zone (Fig. 4c). ODVs in OBs isolated from vAcBac-ac114KO-infected S. exigua larvae were observed by dissolving the polyhedron envelope on nickel grids (Fig. 4d). The numbers of virions per OB for vAcBac and vAcBac-ac114KO are 29.1 ± 11.5 (mean ± SD, n = 36) and 23.3 ± 10.5 (mean ± SD, n = 36), respectively, which are significantly different (t = 2.289, df = 70, p = 0.025) (Fig. 5).

Fig. 4
figure 4

Electron microscopic analysis of ultrathin sections derived from vAcBac-ac114KO-infected Sf9 cells (a–c) and alkali-treated OBs extracted from infected S. exigua larvae (d)

Full size image
Fig. 5
figure 5

The comparison of numbers of ODVs per OB between vAcBac and vAcBac-ac114KO. Data are shown as means and S.E. (n = 36 for each virus; asterisk represents that difference is significant, Student’s t test, α = 0.05)

Full size image

Transcription analysis

The orf114 of AcMNPV is 1,275 bp in length and encodes a 424 aa protein with a predicted molecular mass of 48 kDa. Computer analysis showed that a late transcription initiation motif (GTAAG) was found 13 nt upstream of the start codon ATG, suggesting that ac114 may be a late gene. RT-PCR confirmed that a 1,275 bp ac114 transcript was detected at 9 h p.i., and was still present at 72 h p.i. (Fig. 6a). 5′ RACE revealed that the transcription initiation site was located 16 nt upstream of the ATG start codon of ac114 at an A residue (Fig. 6b).

Fig. 6
figure 6

Transcription analysis of ac114. a Time course analysis of the transcription profile of ac114. Total RNA was extracted from AcMNPV-infected Sf9 cells at designated time points. PCR products from different genes are indicated on the left. The sizes of the different PCR products are indicated on the right. b 5′ RACE analysis of the ac114 transcript. The transcription initiation site is indicated by an arrowhead. The baculovirus late transcription motif is underlined

Full size image

Subcellular localization of Ac114 in Sf9 cells

To investigate the subcellular localization of Ac114, Sf9 cells infected with vAcBac-ac114Rep (or vAcBac-ac114KO as a control) were observed for GFP expression using fluorescence microscopy. At 96 h p.i., punctate patches of AC114-GFP fluorescence were mostly observed in the cytoplasm of vAcBac-ac114Rep-infected cells, whereas GFP fluorescence was evenly distributed in both the cytoplasm and nucleus of the vAcBac-ac114KO-infected cells (Fig. 7b).

Fig. 7
figure 7

Intracellular localization of the ac114-GFP fusion protein in Sf9 cells. a Schematic diagram of the recombinant bacmid AcBac-ac114Rep. The ac114 orf in frame with the gfp gene driven by the polyhedrin promoter was inserted into the polyhedrin locus of AcMNPVac114KOph− to generate AcBac-ac114Rep. b Sf9 cells infected with vAcBac-ac114Rep and vAcBac-ac114KO were observed under fluorescence microscopy. The nuclear DNA was stained with Hoechst 33258 (blue). Cells infected with vAcBac-ac114KO were used as a control

Full size image

Effect of the AcMNPV ac114 knockout on infectivity to S. exigua larvae

To determine whether the lack of ac114 had any effect on the infectivity and killing speed of AcMNPV for third-instar S. exigua larvae, bioassays were performed using a droplet-feeding method. In three independent experiments, no significant differences in the ST50 values were detected between vAcBac and vAcBac-ac114KO (Table 2). However, the LD50 of vAcBac-ac114KO was significantly higher than vAcBac (Table 3). These results indicate that the loss of ac114 has an effect on viral infectivity, but does not affect the viral killing speed in S. exigua larvae.

Table 2 Time–mortality responses of third-instar S. exigua larvae inoculated with vAcBac or vAcBac-ac114KO
Full size table
Table 3 Dose–mortality responses of third-instar S. exigua larvae inoculated with vAcBac or vAcBac-ac114KO
Full size table

Discussion

Homology searches have identified 17 unique genes that are conserved in group I alphabaculovirus, indicating that the function of unique genes may be host dependent [5]. Although ac114 is one of these unique genes, it is still not known whether ac114 or its homologs play any roles in the viral infection process. In this study, we demonstrated that ac114 was not critical for the production of BVs and OBs in vitro. However, interestingly, the ac114 knockout led to the reduction of viral infectivity in S. exigua larvae.

Thus far, few of the unique genes in the group I NPV appear to be directly or indirectly involved in the virus infection process. Mutation of ie2 in AcMNPV reduced the oral infectivity of OBs in S. frugiperda and Trichoplusia ni larvae, owing to the absence of virions in the OB [20]. Bm21 of Bombyx mori NPV, a homolog of AcMNPV ac30, was found not to be essential for virus replication in vitro, but its knockout delayed the killing speed in the larvae [21]. In the present study, we found that the LD50 of the ac114 knockout virus was significantly higher than the control virus. Although Ac114 was identified as a component of BV particle [10], our analysis of the virus growth curve detected no significant difference between the ac114 knockout virus and the control virus, both at MOI = 5 and 2 (Fig. 3 and unshown data), showing that ac114 knockout did not affect BV production and the increased LD50 in the ac114 knockout virus did not result from the impairment of BV production. Electron microscopic observations found that the numbers of ODVs embedded in one OB between the two viruses were statistically significantly different, indicating that the reduced infectivity of the ac114 knockout virus, observed in our study based on administration of OBs, might be due to the reduction of virions embedded in the OBs. Ac114 has also been detected as a structural protein of ODV [11], which might interact with one or more of the other ODV proteins and thus affect the efficiency of embedding of ODVs into OBs. It is worth noting that the difference of numbers of ODVs/OB between two viruses was not as dramatic as those of LD50 values (Table 3), we cannot exclude the possibility that ac114 may affect the virus infectivity by other unknown mechanisms. Zhang et al. observed that when AcMNPV orf150 was deleted, from 4- to 18-fold more OBs of mutant virus than wild-type virus were needed to achieve 50 % mortality. They concluded that the difference in virulence was due to the reduction of the number of foci generated by each OB in the primary infection [22].

Transcription analysis of ac114 by RT-PCR showed that the ac114-specific transcript could be detected from 9 to 72 h p.i., indicating that ac114 was a late transcriptional gene, a result which is consistent with the findings of Jiang et al. [23]. However, a microarray analysis of gene expression in AcMNPV-infected Sf9 cells detected the ac114 transcript by 2 h p.i., after which time it increased steadily from 12 to 24 h p.i., but then declined [24]. The discrepancy between these studies may result from differences in the AcMNPV genotypes (C6 vs. E2), or in the sensitivity of the detection methods used in the experiments.

The localization assay showed that Ac114 was localized primarily in the cytoplasm of Sf9 cells, a result that is in agreement with the NetNES analysis that detected the leucine-rich nuclear export signal at the C terminus of Ac114 [25]. Our own data verify the localization of Bm94, a homolog of Ac114, which is primarily distributed in the cytoplasm of BmN cells, as displayed by immunofluorescence analysis [26]. It is noteworthy that multiple phosphorylation sites in Ac114 were predicted by Netphos [27]. Phosphorylation modification has been shown to control the nuclear transport of a number of proteins in eukaryotic cells [28]. Whether Ac114 can be phosphorylated and whether its phosphorylation state might regulate its subcellular distribution in Sf9 cells is not known at present.

In summary, this study has shown that ac114 was not pivotal for the amplification of AcMNPV in vitro, but that it made an effect on the viral infectivity in vivo. To shed light on the potential roles for ac114 in group I, NPV infection will require further experiments. Such experiments are needed to elucidate whether Ac114 might interact with other proteins and thus participate in the infection process.