Immediate-Early Protein ME53 Forms Foci and Colocalizes with GP64 and the Major Capsid Protein VP39 at the Cell Membranes of Autographa californica Multiple Nucleopolyhedrovirus-Infected Cells

Abstract

me53 is an immediate-early/late gene found in all lepidopteran baculoviruses sequenced to date. Deletion of me53 results in a greater-than-1,000-fold reduction in budded-virus production in tissue culture (J. de Jong, B. M. Arif, D. A. Theilmann, and P. J. Krell, J. Virol. 83:7440-7448, 2009). We investigated the localization of ME53 using an ME53 construct fused to green fluorescent protein (GFP). ME53:GFP adopted a primarily cytoplasmic distribution at early times postinfection and a primarily nuclear distribution at late times postinfection. Additionally, at late times ME53:GFP formed distinct foci at the cell periphery. These foci colocalized with the major envelope fusion protein GP64 and frequently with VP39 capsid protein, suggesting that these cell membrane regions may represent viral budding sites. Deletion of vp39 did not influence the distribution of ME53:GFP; however, deletion of gp64 abolished ME53:GFP foci at the cell periphery, implying an association between ME53 and GP64. Despite the association of ME53 and GP64, ME53 fractionated with the nucleocapsid only after budded-virus fractionation. Together these findings suggest that ME53 may be providing a scaffold that bridges the viral envelope and nucleocapsid.

INTRODUCTION

Autographa californica multiple nucleopolyhedrovirus (AcMNPV), a member of the Baculoviridae, is the type species for the lepidopteran baculovirus genus Alphabaculovirus. A hallmark of lepidopteran baculoviruses is the production of two morphologically distinct virion phenotypes, occlusion-derived virus (ODV) and budded virus (BV). ODVs are responsible for horizontal insect-to-insect transmission of virus, whereas BVs establish a systemic infection within a susceptible host. Budded virus can enter a wide variety of cell types, including mammalian cells (22, 25). Entry of BVs into insect cells involves clathrin-mediated endocytosis, and release of budded-virus nucleocapsids from endosomes is pH-dependent (3, 39). However, evidence for entry into mammalian cells supports both clathrin-dependent and clathrin-independent pathways (22, 25). Once free of the endosome, nucleocapsids associate with F-actin, which facilitates transport to the nucleus (4, 5, 23). After nuclear entry, viral early transcription, DNA replication, and late transcription occur in a cascade fashion. At late times, de novo nucleocapsids are transported from the nucleus to the plasma membrane, where they acquire a lipid envelope and associated viral proteins and bud from the cell, forming BVs. During the very late phase of infection, progeny nucleocapsids are retained in the nucleus, become enveloped, and are embedded into occlusion bodies (OBs).

me53 is a conserved gene found in all lepidopteran alpha- and betabaculoviruses sequenced to date. me53 is transcribed from a dual early/late promoter and was originally identified as a major early transcript (19, 20). In contrast to a report that deletion of me53 resulted in complete abrogation of BV production and DNA replication (45), we have shown that deletion of me53 did not abrogate BV production, although there was a greater-than-1,000-fold reduction in budded-virus production in both Sf-21 and Tn-5 cells, while DNA replication and OB production remained unaffected. This discrepancy likely arose because our studies analyzed viral DNA replication during primary infection over the first 24 h posttransfection (hpt), while that of Xi et al. (45) analyzed cells for a longer period, allowing for secondary infection of the wild-type bacmid but not the Δme53 bacmid. These results suggest a role for me53 in budded-virus production (8).

Although several baculovirus genes have been implicated in either nuclear or cellular egress, the mechanism of budded-virus egress is poorly understood. Deletion of either exon 0 or ac66 results in a greater-than-100-fold reduction in BV production and nucleocapsid accumulation within the nuclei of transfected cells, implicating these two genes in nuclear egress (10, 17). Additionally, exon 0 interacts with β-tubulin, and drug-induced microtubule depolymerization leads to a reduced BV titer, linking microtubules to nucleocapsid transport to the plasma membrane (11). To date the only virus protein identified in egress of group I alphabaculoviruses at the plasma membrane is the major envelope glycoprotein GP64, which exists as trimers that span the membranes of infected cells (32). BVs that lack GP64 are noninfectious, and deletion of gp64 results in a greater-than-50-fold decrease in BV production (26, 31).

Here we demonstrate that ME53 localizes to distinct foci at the plasma membrane at late times postinfection and that these areas correspond to accumulation of GP64, the major envelope fusion protein of BVs. Additionally, the major capsid protein VP39 can be found associated with the ME53/GP64 foci, suggesting that these membrane foci may represent viral budding sites. Furthermore, we demonstrate that deletion of GP64, but not VP39, abolishes ME53 focus formation at the plasma membrane.

MATERIALS AND METHODS

Cells.

Spodoptera frugiperda-derived IPLB-Sf21-AE (Sf-21) and Trichoplusia ni-derived BTI-Tn-5b1 (Tn-5) cells were maintained in Grace's insect medium (Invitrogen) supplemented with 3.33 g/liter lactalbumin hydrolysate, 3.33 g/liter yeastolate, 10% fetal bovine serum, 25 units/ml penicillin, and 25 ng/ml streptomycin.

Generation of ME53:GFP repair virus.

The me53 promoter region and entire open reading frame (ORF) was amplified from AcMNPV bacmid (bMON14272; Invitrogen) with primers me53promFSac (TTACTGAGCTCGTATGTCGGCGTTGTACATG) (underlined characters in primer sequences here and elsewhere represent restriction enzyme sites) and me53REV (TAAGATATCGTTATTTACAATATTAGAATTCTTA) and cloned into the SacI and EcoRV sites of pBluescript (Stratagene), generating pBlueme53. Enhanced GFP (EGFP) was PCR amplified from pFACTGFP with primers GFPFEV (TAAGATATCGTGAGCAAGGGCGAGGA) and GFPRHin (TAAAAGCTTATTCTTGTACAGCTCGTCCATGG) and cloned into the EcoRV and HindIII sites of pFastBachta (Invitrogen), generating pFastBaceGFP. EGFP and the simian virus 40 (SV40) poly(A) signal were amplified from pFastBaceGFP with primers GFPFEV and SV40RXho (TAACTCGAGTCAAGCAGTGATCAGATCC) and cloned into pBlueme53 using EcoRV and XhoI, generating pBlueme53:GFP:40. The me53 promoter and ORF, with a C-terminally fused GFP and SV40 poly(A) signal, was then subcloned from pBlueme53:GFP:40 into pFACT using SacI and XhoI, generating pFACTme53:GFP. The Tn7 cassette from pFACTme53:GFP was transferred to the AcΔ53 bacmid (8) as described in the Bac-to-Bac expression manual (Invitrogen), creating AcΔme53Repme53:gfp (Fig. 1).

Fig. 1.

(A) Schematic representation of the me53 locus and the polh locus of AcΔme53Repme53:gfp. (B) Mean viral titer from three transfections at 120 h for AcΔme53Repme53:gfp and AcGFP. (C) Anti-GFP Western blot analysis of Sf-21 cells infected with AcΔme53Repme53:gfp at the indicated times postinfection. Ac48 denotes infection with wild-type AcMNPV (lacking GFP) at 48 hpi. The major band migrates at about 80 kDa, as expected for an intact ME53:GFP.

The growth kinetics of AcΔme53Repme53:gfp were confirmed by endpoint dilution after transfection with AcΔme53Repme53:gfp bacmid DNA as previously described (8).

Western blotting.

The presence of ME53:GFP fusion protein was confirmed by Western analysis. Briefly 1 × 10⁶ Sf-21 cells were seeded onto 35-mm dishes, allowed to attach for 1 h, and infected with AcΔme53Repme53:gfp at a multiplicity of infection (MOI) of 10. Total protein was harvested at 6, 12, 18, 24, and 36 h postinfection (hpi) by removal of medium and addition of SDS-PAGE loading buffer supplemented with E-64 protease inhibitor. Samples were heated for 10 min at 95°C, and equal volumes were separated by 10% SDS-PAGE and transferred to polyvinylidene difluoride (PVDF) membranes. Membranes were blocked with Tris-buffered saline-Tween 20 (50 mM Tris, 150 mM NaCl, 0.1% Tween 20) (TBST) supplemented with 2.5% skim milk for 1 h and incubated with primary anti-GFP antibody (Roche; 1:1,000) diluted in TBST overnight. Membranes were washed 3 times for 5 min each in TBST and incubated with secondary antibody (rabbit anti-mouse-horseradish peroxidase [HRP]; Sigma) diluted 1:20,000 in TBST with 2% skim milk for 1 h. Membranes were washed 3 times for 5 min in TBST. Detection was with SuperSignal West Pico chemiluminescent substrate (Pierce).

Budded-virus fractionation.

Fractionation was accomplished as described in reference 10. Western analysis was carried as described above except with antihemagglutinin (anti-HA) (Sigma), anti-GP64 (15), or anti-VP39 (15) primary antibodies.

Colocalization of ME53:Venus, VP39:Cerulean, and GP64.

Venus (an enhanced yellow fluorescent protein [EYFP] variant) was PCR amplified from pC32V (21) with primers VenusFEV (TAAGATATCGTGAGCAAGGGCGAGGA) and VenusRHin (TAAAAGCTTATTCTTGTACAGCTCGTCCATGG) and cloned into the EcoRV and HindIII sites of pFastBachta (Invitrogen), generating pFastBacVenus. Venus and the SV40 poly(A) signal were amplified from pFastBacVenus with primers VenusFEV and SV40RXho and cloned into pBlueme53 using EcoRV and XhoI, generating pBlueme53:Venus:40. The me53 promoter and ORF, with a C-terminally fused Venus and SV40 poly(A) signal, were then subcloned from pBlueme53:Venus:40 into pFACT using SacI and XhoI, generating pFACTme53:Venus. The Tn7 cassette from pFACTme53:Venus was transferred to the AcΔ53 bacmid (8) as described in the Bac-to-Bac expression manual (Invitrogen), creating AcΔme53Repme53:ven.

The vp39 promoter and ORF were amplified from AcMNPV bacmid (bMON14272; Invitrogen) with primers vp39promFSac (TTACTGAGCTCGCCATCGTGGAATCAAATAG) and vp39REV (TAAGATATCGACGGCTATTCCTCCACCT) and cloned into the SacI and EcoRV sites of pBluescript (Stratagene), generating pBluevp39. Cerulean (a cyan fluorescent protein [CFP] variant) was PCR amplified from pC32V with primers cerFEV (TAAGATATCGTGAGCAAGGGCGAGGA) and cerRHin (TAAAAGCTTATTCTTGTACAGCTCGTCCATGG) and cloned into the EcoRV and HindIII sites of pFastBachta (Invitrogen), generating pFastBacCerulean. Cerulean and the SV40 poly(A) signal were amplified from pFastBacCerulean with primers cerFEV and SV40RXho and cloned into pBluevp39 using EcoRV and XhoI, generating pBluevp39:cer:40. The vp39 promoter and ORF, with a C-terminally fused Cerulean and SV40 poly(A) signal, were then subcloned from pBluevp39:GFP:40 into pFACT using SacI and XhoI, generating pFACTvp39:cer. The Tn7 cassette from pFACTvp39:cer was transferred to the AcΔ53 bacmid (8) as described in the Bac-to-Bac expression manual (Invitrogen), creating AcΔme53-vp39:cer.

For generation of a mixed viral stock, 2 × 10⁶ Sf-21 cells were seeded into a 60-mm dish and cotransfected with 2 μg of AcΔme53Repme53:ven and 2 μg of AcΔme53-vp39:cer using 15 μl of Cellfectin (Invitrogen) as per the manufacturer's protocol. At 7 days posttransfection, supernatant was harvested, spun at 1,000 × g for 5 min, and transferred to a new tube.

Generation of vp39/me53 and gp64/me53 double-knockout bacmids.

The λ Red recombination system was used to delete either vp39 or gp64 from bacmid AcΔme53 (8) harbored in Escherichia coli DH10 cells as described previously (8). The zeocin resistance gene, under the control of a bacterial promoter, with 5′ 50-nucleotide (nt) arms homologous to either the 5′ or 3′ flanking region of vp39 (nt 75780 to 75829 and nt 76262 to 76311 of AcMNPV; NC_001623) or gp64 (nt 108151 to 108200 to nt 109718 to 109767) was PCR amplified from pIZ-V5His (Invitrogen) using primers TATCAAGCGAATTTTAATTCCAAGCGCAACCAATTACCAAGACGTGTTTACATATGAATATCCTCCTTAG and CAAACGATTGGGTTGACTTCTCATAATGTCACTGCTTCTTATCGGGTTGTTGTAGGCTGGAGCTGCT for deletion of vp39 and TACTAGTAAATCAGTCACACCAAGGCTTCAATAAGGAACACACAAGCAAGCATATGAATATCCTCCTTAG and AATAATGATACAATTTTTATTATTACATTTAATATTGTCTATTACGGTTTTGTAGGCTGGAGCTGCT for deletion of gp64. Ampli- cons were verified by agarose gel electrophoresis and purified using the QIAquick PCR purification kit (Qiagen) as per the manufacturer's protocol. Electrocompetent E. coli DH10 cells harboring AcΔme53 bacmid and pKD46 (λ Red recombinase genes) were transformed with 250 ng of the purified amplicon using the GenePulser Xcell (Bio-Rad). Cells were supplemented with 1 ml of fresh LB medium and incubated for 4 h at 37°C with gentle shaking. Cells were pelleted for 1 min at 10,000 × g, resuspended in 200 μl of LB broth, and plated on LB plates supplemented with kanamycin (30 μg/ml), chloramphenicol (34 μg/ml), and zeocin (20 μg/ml). Colonies were subjected to a PCR screen to confirm deletion of either vp39 or gp64 and the expected genomic insertion of the zeocin resistance gene.

The vp39 promoter and ORF were amplified from AcMNPV bacmid (bMON14272; Invitrogen) with primers vp39promFSac and vp39RHin (TAAAAGCTTTTAGACGGCTATTCCTCCACCT) and cloned into the SacI and HindIII sites of pFastBachta (Invitrogen), generating pFastBacvp39. The vp39 promoter and ORF and SV40 poly(A) signal were then amplified by PCR from pFastBacvp39 with vp39promFsac and SV40XhoR and cloned into pFACT or pFACTGFP using SacI and XhoI, generating pFACTvp39:40 and pFACTGFPvp39:40. The Tn7 cassette from pFACTvp39:40 or pFACTGFPvp39:40 was transferred to the AcΔ53Δvp39 bacmid as described in the Bac-to-Bac expression manual (Invitrogen), creating AcΔme53Δvp39-Repvp39 and AcGFPΔme53Δvp39-Repvp39.

Construction of AcΔme53Δgp64-Repgp64 and AcGFPΔme53Δgp64-Repgp64 was the same as described above except using the gp64 promoter and ORF (gp64promoF [TAAGAGCTCTCGCAAAGCGGATCAAATACT] and gp64RHin [TAAAAGCTTTTAATATTGTCTATTACGGTTTCTAAT]). Construction of AcΔme53Δvp39-Repme53:gfp and AcΔme53Δgp64-Repme53:gfp was accomplished by transfer of the Tn7 cassette from pFACT53:GFP:40 into either AcΔme53Δvp39 or AcΔme53Δgp64. Construction of AcGFPΔme53Δvp39 and AcGFPΔme53Δgp64 and of AcΔme53Δvp39pFACT and AcΔme53Δgp64pFACT was accomplished by transfer of the Tn7 cassette from pFACTGFP or pFACT, respectively, into either AcΔme53Δvp39 or AcΔme53Δgp64.

Imaging conditions.

Unless otherwise stated, imaging conditions were the following.

Sf-21 cells (9 × 10⁵) were seeded onto 35-mm dishes, allowed to attach for 1 h, and infected with AcΔme53Repme53:gfp at an MOI of 10. Cells were rinsed three times in phosphate-buffered saline (PBS) and overlaid with 3 ml of fresh medium or PBS before imaging.

For ME53:GFP, images were acquired with a Leica SP5 confocal laser scanning microscope (CLSM), using a 63× lens for slides or a 63× dipping lens for live cells, excitation with the 488-nm laser with collection between 500 and 523 nm, 1,024 × 1,024 pixels, 200 Hz, and line average of 3 and frame average of 2 for fixed cells or line average of 3 and frame average of 1 for live cells.

For ME53:GFP and GP64 colocalization, at 30 hpi cells were fixed with 4% paraformaldehyde in PBS for 10 min, washed 3 times in 2 ml PBS, and blocked for 30 min with blocking buffer (2% bovine serum albumin [BSA] in PBS). Monolayers were incubated with AcV1 (39) diluted 1:50 in 1 ml of PBS-2% BSA for 1 h, rinsed 3 times with 2 ml of blocking buffer, and incubated for 30 min with goat anti-mouse Alexafluor 594 (Invitrogen) diluted 1:100 in blocking buffer. Monolayers were rinsed 3 times with 2 ml PBS and overlaid with 3 ml of PBS. Imaging was accomplished with the Leica SP5 CLSM using a 63× dipping lens, simultaneous excitation with the 488-nm and 594-nm lasers, and acquisition between 500 and 523 nm for ME53:GFP and 607 and 671 for Alexafluor 594.

For colocalization of ME53:Ven, VP39:Cer, and GP64, images were acquired sequentially with the Leica SP5 CLSM using a 63× dipping lens. For VP39:Cer excitation was with the 458-nm laser, with acquisition between 470 and 500 nm, 1,024 × 1,024 pixels, 200 Hz, a line average of 3 and a frame average of 1, and for ME53:Ven excitation was with the 514-nm laser, with acquisition between 525 and 574 nm, 1,024 × 1,024 pixels, 200 Hz, a line average of 3, and a frame average of 1. Localization of GP64 was carried out as described above. Images were overlaid using ImageJ.

For overlay and Pearson's correlation analysis, a total of 30 regions of interest (ROIs) from 5 separate background-corrected Sf-21 or BTI-Tn5b1 imaged cells were selected and analyzed with the ImageJ intensity correlation analysis plug-in. ROIs were selected based on areas demonstrating the highest intensities in the green channel (ME53:GFP).

Viral spread of AcΔme53Δvp39 and AcΔme53Δgp64 and repairs thereof.

To test viral spread following transfection, 9 × 10⁵ Sf-21 cells were seeded onto 35-mm dishes and transfected with 2 μg of either AcGFPΔme53Δvp39, AcGFPΔme53Δgp64, AcGFPΔme53Δvp39-Repvp39, AcGFPΔme53Δgp64-Repgp64, AcΔme53Δvp39-Repme53:gfp, AcΔme53Δgp64-Repme53:gfp, or a combination of AcΔme53Δvp39-Repme53:gfp and AcGFPΔme53Δvp39-Repvp39 or of AcΔme53Δgp64-Repme53:gfp and AcGFPΔme53Δgp64-Repgp64 using 10 μl of Cellfectin (Invitrogen) as per the manufacturer's protocol. Monolayers were imaged using a Leica DM 4500B epifluorescence microscope at 24 and 96 h posttransfection (hpt).

Extracellular progeny virions for the transfections described above were quantified by quantitative PCR (qPCR) (14). At 96 hpt, supernatant was harvested and extracellular DNA was isolated using the Roche viral DNA minikit as per the manufacturer's instruction for cells grown in monolayer. Extracellular DNA was quantified against a standard curve generated by making eight 5-fold serial dilutions of an AcMNPV viral stock of determined titer and isolating the DNA as described above. Four microliters of isolated DNA was subjected to qPCR using primers GTGTTCAAAACGCACTGCAG and CGGTGTTATTGATGGCGTTG, based on the odve56 gene, and the Roche SYBR green master mix, using the following conditions on the Corbett 6000RG cycler: 95°C for 15 min and then 35 cycles of 95°C for 20 s, 55°C for 30 s, and 72°C for 30 s, followed by acquisition and melt curve analysis. Extracellular DNA was quantified against the standard curve using the Corbett software and expressed in PFU equivalents. The titers presented represent the averages for 3 technical replicates from each of two independent transfections.

Cellular localization of ME53:GFP from AcΔme53Δvp39-Rep53:gfp and AcΔme53Δgp64-Rep53:gfp.

For localization of ME53:GFP in AcΔme53Δvp39-Rep53:gfp and AcΔme53Δgp64-Rep53:gfp, Sf-21 cells (9 × 10⁵) were seeded onto 35-mm dishes, allowed to attach for 1 h, and transfected with 7 μg of either AcΔme53Δvp39-Rep53:gfp or AcΔme53Δgp64-Rep53:gfp bacmid DNA with 15 μl of Cellfectin (Invitrogen). At 48 h pt, cells were rinsed three times in PBS and overlaid with 3 ml of fresh medium before imaging. Imaging conditions were the same as those used for the live-cell imaging of ME53:GFP described above.

Localization of ME53:GFP and GP64 in uninfected cells.

The ME53:GFP ORF was amplified from pFACT53:GFP:40 using primers ME53FnotI (TAAGCGGCCGCATGAACCGTTTTTTTCGAGAGAAT) and SV40RXba (TAATCTAGATCAAGCAGTGATCAGATCC) and cloned into the NotI and XbaI sites of pIZ (Invitrogen), to create pIZME53:GFP. The gp64 ORF and SV40 poly(A) were amplified from pFACTgp64:40 using primers gp64FXba (TAATCTAGAATGGTAAGCGCTATTGTTTTATAT) and SV40RApa (TAAGGGCCCTCAAGCAGTGATCAGATCC) and cloned into the XbaI and ApaI sites of pDIE (36), creating pDIEGP64. Sf-21 monolayers were seeded onto 35-mm dishes at 9 × 10⁵ cells and transfected with 2 μg of either pIZME53:GFP, pDIEGP64, pFACTGFP, or a combination of pIZME53:GFP and pDIEGP64 or of pFACTGFP and pDIEGP64. Cells were fixed with 4% paraformaldehyde in PBS at 48 hpt and probed with AcV1 antibody as described above.

Localization of ME:53 in cells treated with colchicine and cyclochalasin D.

Sf-21 cells (1 × 10⁶) were seeded onto 35-mm dishes, allowed to attach for 1 h, and infected with AcΔ53Rep53:gfp at an MOI of 10. At 1 hpi cells were treated with Grace's insect medium containing either 10 μM colchicine, 2.5 μg/ml cytochalasin D, or carrier alone. Cells were imaged as described above at 30 hpi.

RESULTS

Localization of ME53:GFP in infected Sf-21 and Tn-5 cells.

A recombinant Δme53 repair bacmid (AcΔme53Repme53:gfp) was constructed from AcΔme53 (8), whereby me53 with a C-terminally fused EGFP was introduced under the control of the native me53 promoter (Fig. 1A). The ME53:GFP fusion bacmid behaved similarly to the AcGFP bacmid in terms of budded-virus production as measured by endpoint dilution at 5 days posttransfection (Fig. 1B). Additionally, the full-size fusion protein (approximately 80 kDa) was detected by Western blot analysis from 6 to 36 hpi using anti-GFP antibodies. A GFP-only fragment (27 kDa) was not detected in the virus-infected lanes, indicating that the GFP was not being cleaved from the fusion protein. This confirmed that ME53:GFP was synthesized and that any GFP fluorescence observed was due to the ME53:GFP fusion (Fig. 1C).

The localization of ME53:GFP in infected Sf-21 or Tn-5 cells was determined by confocal fluorescence microscopy. Generally, fixed (see Fig. S1 in the supplemental material) and live (Fig. 2A) samples displayed similar localization patterns. However, fixation led to a higher accumulation of ME53:GFP at the cell periphery at the earlier time points than in the live cell samples. Each time point in Fig. 2A is depicted by a magnified view of a representative live cell in either the Sf-21 or Tn-5 cell line infected with AcΔme53Repme53:gfp or AcGFP as a free-GFP control. A monolayer overview of each time point in both cell lines, demonstrating the overall nature of the images in Fig. 2A, can be seen in Fig. S2 in the supplemental material.

Fig. 2.

(A) Live-cell images of Sf-21 or Tn-5 cells infected with AcΔme53Repme53:gfp or AcGFP as a control at 6, 12, 18, 24, and 36 hpi. Representative cells for each time point are shown. Cells were imaged with a Leica SP5 CLSM with excitation at 488 nm and collection from 500 to 523 nm. (B) Montage of 16 z-stack slices representing a total of 6.72 μm from an AcΔme53Repme53:gfp-infected Sf-21 cell at 30 hpi. (C) 3D maximum projection of the same 16 slices. ME53 localizes to discrete foci at or near the plasma membrane at late times in infection. These foci appear randomly distributed throughout the cell surface. The scale bar represents 10 μm.

At 6 hpi of Sf-21 cells, ME53:GFP localized mainly to the cytoplasm, displaying a slight presence at the plasma membrane in certain cells, with little fluorescence in the nucleus. A similar pattern was seen at 12 hpi, with some evidence of more ME53:GFP in the nucleus and continued accumulation at the cell periphery. By 18 hpi, ME53:GFP began to accumulate to higher levels in the nuclei of some cells, as well as concentrating at the cell periphery. In some of the cells, ME53:GFP was found in discrete foci localized at the cell periphery. This trend continued through 24 to 36 hpi, with an increased accumulation of ME53:GFP within the nucleus and in discrete foci at the cell membrane. Concurrently, there was a decrease in the relative amounts of cytoplasmic ME53:GFP.

A similar localization pattern was seen in live confocal images from infected Tn-5 cells (Trichoplusia ni) (Fig. 2A). Early in the infection (6 to 12 h), ME53:GFP was primarily in the cytoplasm, with a slight concentration of ME53:GFP at the cell periphery and little fluorescence originating from the nucleus. However, at the late and very late stages of infection (18 to 36 hpi), there was a strong nuclear localization as well as accumulation of discrete foci of ME53:GFP at the cell periphery near or at the plasma membrane. As seen in the Sf-21 cells, there was a concurrent decrease in the amount of ME53:GFP in the cytoplasm of Tn-5 cells at late times postinfection.

AcGFP (parental AcMNPV bacmid expressing nonfused EGFP under the control of the Orgyia pseudotsugata multiple nucleopolyhedrovirus [OpMNPV] IE2 promoter) was used as a control to demonstrate the differences in the distributions of ME53:GFP and GFP alone. At all time points, GFP was distributed almost evenly throughout both the cytoplasm and nucleus in the Sf-21 and Tn-5 cell lines (Fig. 2A).

To determine the overall distribution of ME53:GFP near the cell surface, a z-stack series of infected cells was acquired at 30 hpi. ME53:GFP was found in discrete foci distributed throughout the cell periphery (Fig. 2B). This was especially evident when the z-stack was viewed as a three-dimensional (3D) maximum projection, where patches of ME53 could be seen over the cell periphery near or at the plasma membrane (Fig. 2C).

ME53:GFP colocalizes with GP64.

The ME53:GFP localization at the peripheral foci was similar to that described for GP64 (2, 33). With this observation, we needed to ascertain whether ME53:GFP colocalizes with GP64 and investigate any functional implications. GP64 is trafficked to the endoplasmic reticulum (ER) via a signal peptide and is transported to and displayed on the outer surfaces of infected cells. De novo nucleocapsids transit from the nucleus to the cytoplasm and the cell membrane, where they acquire GP64 and plasma membrane-derived envelope as they bud from an infected cell.

Colocalization of ME53 and GP64 was investigated in both qualitative (by dye overlay) and quantitative (by Pearson's correlation coefficient Rr and overlap coefficient R) manners. Sf-21 and Tn-5 cells were infected with AcΔme53Repme53:gfp at an MOI of 10. At 30 hpi, cells were fixed, but not permeabilized, to ensure detection of only surface-exposed GP64. Fixed cells were incubated with AcV1 antibody and detected with secondary antibody conjugated to Alexafluor 594. AcV1 is a neutralizing antibody specific for a conformational epitope on GP64 (39). Much like ME53:GFP (Fig. 3A), GP64 (Fig. 3B) accumulated in distinct foci at the periphery of infected Sf-21 cells at 30 hpi. Dye overlay of the green (ME53:GFP) and red (GP64) channels demonstrated that the ME53:GFP found in foci at the plasma membrane colocalized with GP64 clusters, as seen by the yellow in the overlay images (Fig. 3C and E). This was especially evident in the 3D maximum projection (Fig. 3D). Colocalization was also seen in Tn-5 cells (Fig. 3G) and was detectable by fluorescence microscopy from 18 hpi onwards in both the Tn-5 and Sf-21 cell lines. All ME53:GFP foci observed at the plasma membrane demonstrated overlap with corresponding foci of GP64.

Fig. 3.

Colocalization of ME53:GFP and GP64 at 30 hpi. Sf-21 or Tn-5 cells were infected with AcΔme53Repme53:gfp at an MOI of 10, fixed with 4% paraformaldehyde, and probed with AcV1 antibody and goat anti-mouse Alexafluor594 secondary antibody. (A to C) Representative infected Sf-21 cells showing ME53:GFP localization by GFP fluorescence (A), GP64 localization by AcV1 antibody and Alexafluor 594 secondary antibody (B), and dye overlay of ME53:GFP and GP64 localization (C). (D) Maximum-projection dye overlay of 16 z-stack slices representing 6.72 μm. (E to H) Colocalization analysis of ME53:GFP and GP64 in Sf-21 (E and F) or Tn-5 (G and H) cells at 30 hpi. Shown are dye overlays of representative cells, with regions of interest (ROIs) indicated and numbered (E and G) and the corresponding magnified ROIs (F and H). The table shows the mean Pearson's coefficient (Rr) and overlap coefficient (R) for 30 ROIs each from 5 Sf-21 or 5 Tn-5 cells.

Quantitative analysis of colocalization was employed to validate the dye overlay data. To this end, a total of 30 regions of interest (ROIs) from each of 5 distinct Sf-21 and 5 distinct Tn-5 cells were selected for quantitative analysis. Representative Sf-21 and Tn-5 cells and their respective ROIs are shown in Fig. 3E to H. As can be seen in the magnified ROIs (Fig. 3F and H), there was strong overlap of green (ME53:GFP) and red (GP64) pixels as demonstrated by the high intensity of yellow seen in the images. This strong overlap was supported by the overlap coefficient (R), which measures how often pixels overlap from each channel and had mean values of 0.931 for Sf-21 cells and 0.949 for Tn-5 cells (maximum R value = 1). The strong overlap was also supported by the numerical data for Pearson's coefficient (Rr), which measures how well pixel intensities from each channel correlate (0.735 for Sf-21 cells and 0.706 for Tn-5 cells; maximum Rr value = 1). Taken together, these results indicated that almost all of the red and green pixels overlapped in the ROIs and that the relative intensities of the respective green and red pixels varied synchronously.

ME53:GFP, GP64, and VP39 associate at the plasma membrane.

Given the membrane location of GP64 and ME53:GFP colocalization and the role of GP64 in budded-virus production and infectivity, we hypothesized that the GP64/ME53:GFP foci represent viral budding sites. To test this hypothesis, Sf-21 cells were coinfected with AcΔme53Repme53:ven (an AcΔme53 repair virus that expresses ME53:VEN [ME53 fused to Venus, a fast-folding bright variant of YFP]) and AcΔme53vp39:cer (an AcΔme53 virus expressing VP39:CER [VP39 capsid protein fused to Cerulean, a bright variant of CFP]) and imaged at 30 hpi. Although most of the VP39:CER was nuclear, some VP39:CER (Fig. 4B) formed foci with ME53:VEN at the peripheries of cells coinfected with AcΔme53Repme53:ven and AcΔme53vp39:cer (Fig. 4A and B). Furthermore, these foci colocalized, as shown when images were overlaid (Fig. 4C). The insets in Fig. 4C represent magnified images of the boxed areas, and overlay is depicted as cyan pixels indicating Venus and Cerulean colocalization. However, this was not seen in every cell, as some cells were found with ME53:VEN-only foci lacking VP39:CER (Fig. 4D to F). AcV1 antibody was used to localize GP64 (red), together with ME53:VEN and VP39:CER in cells coinfected with AcΔme53Repme53:ven and AcΔme53-vp39:cer. Some foci showed evidence of all three proteins colocalized at the plasma membrane (white pixels), as can be seen in the representative images in Fig. 4G to J and insets.

Fig. 4.

Sf-21 cells coinfected with AcΔ53Rep53:Venus (ME53:VEN) and AcΔ53-VP39:Cerulean (VP39:CER), fixed with 4% paraformaldehyde in PBS, and imaged at 30 hpi using the Leica CLSM-SP5. (A and D) Localization of ME53:VEN (green); (B and E) localization of VP39:CER (blue). Merged images of panels A and B and of D and E are shown in panels C and F, respectively. (G) ME53:VEN; (H) VP39:CER; (I) localization of surface-exposed GP64 (red) as detected by AcV1 antibody and Alexafluor594-conjugated secondary. (J) Merged images. Insets in panels C and J are magnified images of the boxed areas.

Role of GP64 and VP39 in ME53:GFP localization.

To determine the potential role(s) of GP64 and VP39 in the localization of ME53:GFP, AcΔme53Δvp39 and AcΔme53Δgp64, dual me53 and vp39, and dual me53 and gp64 knockout bacmids were generated (see Fig S3 and S4 in the supplemental material). Successful deletion of vp39 or gp64 was confirmed by PCR (see Fig. S3 and S4 in the supplemental material). In transfected monolayers, infection with either AcΔme53Δvp39 or AcGFPΔme53Δgp64 was limited to initially transfected cells (Fig. 5A and B, first columns). ME53:GFP repair bacmids AcΔme53Δvp39-repme53:gfp and AcΔme53Δgp64-repme53:gfp also failed to spread beyond the initially transfected cells (Fig. 5A and B, third columns). Single repair bacmids AcGFPΔme53Δvp39-Repvp39 and AcGFPΔme53Δgp64-Repgp64 (with reintroduction of vp39 and gp64, respectively) also demonstrated very limited spread (Fig. 5A and B, second columns). Sf-21 cells cotransfected with both an me53:gfp repair bacmid and a vp39 repair bacmid (AcΔme53Δvp39-Repvp39) and cells cotransfected with both an me53:gfp repair bacmid (AcΔme53Δgp64-Repme53:gfp) and a gp64 repair bacmid (AcΔme53Δgp64-Repgp64) demonstrated viral spread comparable to that of monolayers transfected with AcGFP (Fig. 5A and B, fourth columns).

Fig. 5.

Viral spread assay. (A) Sf-21 monolayers transfected with bacmids AcGFPΔme53Δvp39 (ΔME53, ΔVP39, +GFP), AcGFPΔme53Δvp39Repvp39 (ΔME53, +rvp39, +GFP), AcΔme53Δvp39-Repme53:gfp (ΔVP39, +rME53:GFP), or a combination of both AcΔme53Δvp39-Repme53:gfp and AcΔme53Δvp39-Repvp39 (+rME53:GFP, +rvp39) at 24 and 96 hpt. (B) Sf-21 monolayers transfected with AcGFPΔme53Δgp64 (ΔME53ΔGP64, +GFP), AcGFPΔme53Δgp64-Repgp64 (ΔME53, +rGP64, +GFP), AcΔme53Δgp64-Repme53:gfp (ΔGP64, +rME53:GFP), or both AcΔme53Δgp64-Repme53:gfp and AcΔme53Δgp64-Repgp64 (+rME53:GFP, +rGP64) at 24 and 96 hpt. (C) Extracellular DNA expressed as PFU equivalents from AcGFPΔme53Δvp39 and Δvp39 “repair” transfected monolayers at 96 hpt. (D) Extracellular DNA expressed as PFU equivalents of AcGFPΔme53Δgp64 and Δgp64 “repair” transfected monolayers at 96 hpt. Values below 10³ PFU equivalents (*) are below the dynamic range of the standard curve.

To further confirm repair, the viral titer was measured at 96 hpt by absolute quantification of extracellular viral DNA with qPCR. The titer from monolayers transfected with AcΔme53Δvp39-Repvp39, showed a lower, at least a 1,000-fold, reduction in extracellular viral DNA compared to that for the AcΔme53Δvp39-Repvp39/AcΔme53Δvp39-Repme53:GFP cotransfected cells (Fig. 5C).

For the AcΔme53Δgp64 and GP64 “repair” bacmids (Fig. 5D), both AcΔme53Δgp64 and AcΔme53Δgp64Repme53:gfp demonstrated a greater-than-1,000-fold-lower level of PFU equivalents than monolayers cotransfected with both AcΔme53Δgp64Repgp64 and AcΔme53Δgp64Repme53:gfp. An approximately 500-fold smaller amount of PFU equivalents was detected from monolayers transfected with AcΔme53Δgp64Repgp64 than from cotrans- fected monolayers. Monolayers cotransfected with both AcΔ53Δ64Repgp64 and AcΔ53Δ64Repme53:gfp produced an amount of PFU equivalents similar to that of monolayers transfected with AcGFP.

In cells transfected with a bacmid lacking vp39, AcΔme53Δvp39-Repme53:gfp (Fig. 6B), localization of ME53:GFP was similar to that in cells transfected with AcΔme53Repme53:gfp (Fig. 6A) or cells cotransfected with both AcΔme53Δvp39-Repme53:gfp and AcΔme53Δvp39-Repvp39 (Fig. 6C), with ME53:GFP localizing to both the nucleus and foci at the plasma membrane. The cells shown are representative of the ME53:GFP pattern seen in the majority of transfected cells. This suggests that ME53:GFP localization is independent of vp39.

Fig. 6.

Localization of ME53:GFP in Sf-21 cells transfected with a dual me53/vp39 deletion bacmid, a dual me53/gp64 deletion bacmid, and gp64 and vp39 repair bacmids at 48 hpt. (A) AcΔme53Repme53:gfp (rME53:GFP); (B) AcΔme53Δvp39Repme53:gfp (ΔVP39+rME53:GFP); (C) both AcΔme53Δvp39Repme53GFP and AcΔme53Δvp39Repvp39 (rVP39+rME53:GFP); (D) AcΔme53Δgp64Repme53GFP (ΔGP64+rME53:GFP); (E) both AcΔme53Δgp64Repme53GFP and AcΔme53Δgp64Repgp64 (rGP64+rME53:GFP). Lower panels, transient expression and localization in uninfected cells of plasmids expressing GP64 only (F), GFP only (G), or both GP64 and GFP (H) (a merged image is in panel I) or expressing ME53:GFP only (J) or GP64 (K) in cells cotransfected with both ME53:GFP- and GP64-expressing plasmids or ME53:GFP in cells cotransfected with both ME53:GFP- and GP64-expressing plasmids (L) (a merged image is in panel M). GP64 was localized with AcV1 and Alexafluor 594 secondary antibody and GFP or ME53:GFP by GFP fluorescence. The merged images show no evidence of colocalization of both GP64 and ME53:GFP (or GFP only).

In contrast, cells transfected with a bacmid lacking gp64, AcΔme53Δgp64-Repme53:gfp, demonstrated an altered localization for ME53:GFP compared to cells transfected with AcΔme53Repme53:gfp. In the absence of gp64, ME53:GFP still localized to the nucleus, but it failed to form foci at the plasma membrane (Fig. 6D). In cells cotransfected with both AcΔme53Δgp64-Repme53:gfp and AcΔme53Δgp64-Repgp64, thus restoring both ME53 and GP64, ME53:GFP was found in foci at the plasma membrane, confirming that reintroduction of a copy of gp64 was sufficient to “repair” the peripheral localization of ME53:GFP (Fig. 6E). The cells shown are representative of the ME53:GFP pattern seen in the majority of transfected cells.

Since deletion of gp64 altered the localization of ME53:GFP, the localization of ME53:GFP in uninfected cells when expressed alone or with GP64 at 48 hpt was investigated. In uninfected Sf-21 cells GFP adopted only a diffuse cytoplasmic and nuclear distribution (Fig. 6G, GFP). In contrast to GFP, ME53:GFP maintained a predominantly cytoplasmic localization (Fig. 6J, ME53:GFP alone). This is similar to what is seen at early time points in the infection cycle (Fig. 2, 6 hpi); however, in the uninfected cells there is no accumulation near the cell periphery. Cotransfection with an ME53:GFP- or GFP-expressing plasmid and a plasmid expressing GP64 did not alter the GFP localization of the GFP control (Fig. 6H, GP64+GFP, and I, GP64+GFP merge) or of the ME53:GFP (Fig. 6L, GP64+ME53:GFP, and M, GP64+ME53:GFP merge). GP64, when expressed alone (Fig. 6F, GP64), was found in discrete foci at the cell membrane, similar to the localization pattern seen in infected cells (Fig. 3). In cells cotransfected with GFP- or ME53:GFP-expressing plasmids, the GP64 similarly localized to foci at the cell periphery, with no evidence of colocalization with ME53 in these foci (Fig. 6I, K, and M). This suggested that envelope localization of ME53:GFP but not GP64 was dependent on infection and that the localization of GP64 was independent of ME53.

ME53 is in the nucleocapsids of budded virions.

To determine whether ME53 associates with either the envelope or nucleocapsid portion of the budded virion, budded virus was purified from Sf-21 cells infected with AcΔme53Repme53:HA (an ME53 repair virus with a C-terminal HA epitope tag) (8) and fractionated into envelope and nucleocapsid fractions. To ensure complete fractionation, samples were probed using anti-VP39 (nucleocapsid protein) and anti-GP64 (envelope pro- tein) antibodies (15). While both VP39 and GP64 were detected as part of the whole virion, VP39 was detected only in the nucleocapsid fraction (capsid), as expected, while GP64 was detected exclusively in the envelope (env) fraction, as expected (Fig. 7, VP39 and GP64), demonstrating that fractionation was effective. An anti-HA-positive band at approximately 53 kDa, as expected for ME53:HA, was detected by Western analysis in the whole virus (BV) as well as the nucleocapsid fraction but was not detected in the envelope fraction (Fig. 7, me53:HA), demonstrating that ME53 is associated with the nucleocapsid.

Fig. 7.

Fractionation of budded virus from Sf-21 cells infected with AcΔme53Repme53:HA. Budded virions were purified with sucrose cushions and sucrose gradients and left intact (BV) or fractionated into envelope (Env) or capsid fractions. Samples were separated by SDS-PAGE (12%) and probed with anti-HA, anti-VP39, or anti-GP64 antibodies as indicated to the left of the panels.

Role of cytoskeleton in ME53/GP64 focus formation.

Since microtubules have been implicated in baculovirus egress (11, 40), cells were treated with colchicine (10 μM) (11, 40), a microtubule-depolymerizing agent, to determine whether microtubules play a role in localization of GP64 and ME53:GFP. Treatment with colchicine did not disrupt focus formation in infected Sf-21 cells, and both ME53 and GP64 were present in foci at the plasma membrane at 30 hpi (Fig. 8D to F). Although colchicine-treated cells were morphologically less spherical than nontreated ones, there were no discernible differences between the untreated controls (Fig. 8A to C) and colchicine-treated cells in terms of localization of ME53:GFP and GP64.

Fig. 8.

(A to I) Effect of cytoskeleton depolymerization on cellular distribution of ME53:GFP and GP64. Sf-21 cells were infected with AcΔ53Rep53:gfp at an MOI of 10, fixed at 30 hpi, probed with AcV1 and Alexafluor 594 secondary antibody to detect GP64 (red) and green fluorescence to detect ME53:GFP (green), and imaged using confocal microscopy. (A to C) Infected cells treated with carrier only (dimethyl sulfoxide [DMSO]). (D to F) Infected cells treated with 10 μM colchicine from 1 hpi onward. (G to I) Infected cells treated with 2.5 μg/ml cytochalasin D from 1 hpi onward. (J to L) Localization of actin by phalloidin-TRITC (red) relative to ME53:GFP (green) at 30 hpi. Insets represent magnified images of boxed areas.

Similarly, to investigate the potential role of actin microfilaments in focus formation, infected cells were treated with cytochalasin D (2.5 μg/ml) (4) from 1 h postinfection onward. Although ME53:GFP/GP64 foci still formed, some of these foci were larger than those seen in the untreated controls (Fig. 8G to I, circled). Additionally, fewer of the GP64 foci correlated with ME53 foci, which was especially evident on the left-hand side (boxed) of the representative cell shown in Fig. 8I.

Considering the above observation, we localized actin, using phalloidin-tetramethyl rhodamine isocyanate (TRITC), in cells infected with AcΔme53Repme53:gfp at 30 hpi. As can be seen in Fig. 8J to L, actin filaments (red) in infected cells lay in a band just beneath where ME53:GFP localizes, and actin projections were often associated with ME53:GFP at the tips or along their projections (Fig. 8J to L, insets).

DISCUSSION

me53 of AcMNPV is highly expressed at early and late times postinfection from a dual early/late promoter (19, 20), suggesting that ME53 may have a dual function, one early and one late in infection. Our data support the idea of a dual function, as ME53:GFP demonstrated both a nuclear localization and localization to discrete foci at the periphery of the cell. The nuclear localization is consistent with previous work which suggested that ME53 may act as a transcription factor due to the presence of a highly conserved zinc finger motif (8, 20). Furthermore, deletion of me53 results in an over-1,000-fold reduction in budded-virus production (8). However, the localization of ME53 in infected cells, potential interaction partners, and the role of ME53 in budded-virus production have not yet been investigated. In the present study we used an ME53:GFP repair virus to demonstrate the localization pattern of ME53:GFP throughout the infection cycle in two different insect cell lines, Sf-21 and Tn-5. The replication kinetics of AcΔme53Repme53:gfp, as measured by endpoint dilution, was similar to that of the parental bacmid AcGFP demonstrating its wild-type functionality. Additionally, Western blot analysis of infected cells produced an anti-GFP-positive band at approximately 80 kDa, as would be expected for an intact ME53:GFP fusion protein (53 kDa for ME53 plus 27 kDa for GFP). This indicates that the fusion protein was being properly translated and remained intact and that the addition of GFP to the C terminus of ME53 did not alter ME53's functionality in terms of budded-virus production.

The pattern of ME53:GFP localization was consistent between the two cell lines, with ME53:GFP found predominantly in the cytoplasm at early times postinfection, while a primarily nuclear localization was seen at late times postinfection. Additionally, at late times, ME53:GFP accumulated in distinct foci near the periphery of the cell, giving a beaded-necklace appearance in two dimensions. Three-dimensional projections of a z-stack revealed that these ME53 foci were found randomly distributed throughout the cell surface, at or near the cell membrane. This localization pattern was different from that in control cells expressing GFP alone, which demonstrated an overall diffuse pattern of GFP within the cell throughout the infection cycle, with a slightly higher accumulation within the nucleus.

The accumulation of ME53:GFP in distinct foci at the plasma membrane was reminiscent of that reported for GP64 (2, 33) and led us to investigate whether ME53:GFP colocalizes with GP64. One of the primary roles of GP64 is to allow the nucleocapsid to escape the endosome (upon acidification) after entry into a cell. GP64 is essential for virus replication, as virions lacking GP64 are noninfectious. GP64 has also been implicated in efficient virus budding, as deletion of GP64 results in a drastic reduction in budded-virus production (31). GP64 is trafficked to the ER via a signal peptide and is transported to, and displayed on, the outer surfaces of infected cells (32). De novo nucleocapsids move from the nucleus to the cytoplasm and then to the cell membrane, where they acquire GP64 and a plasma membrane-derived envelope as they bud from an infected cell. We report here the colocalization of ME53:GFP and GP64, which occurs at very distinct areas of the plasma membrane and develops over time during the infection. The strong visual dye overlay of GP64 and ME53:GFP is supported by the data seen for the overlap coefficient (R), which had mean values of 0.931 for Sf-21 cells and 0.949 for Tn-5 cells. Additionally, numerical data for Pearson's coefficient (Rr) indicate that in areas of overlap the relative intensities of the respective green and red pixels vary synchronously. Taken together, the data are indicative of a functional colocalization and may represent an association between ME53:GFP and GP64 in AcMNPV-infected cells.

VP39 fluorescently tagged with Cerulean was also found at the periphery of the cell and associated with some, but not all, ME53/GP64 foci. This is consistent with the hypothesis that the ME53/GP64 foci represent viral budding sites and is in keeping with transmission electron microscopy (TEM) observations by Volkman and Goldsmith (39), who noted accumulation of GP64 at viral budding sites. Foci with only ME53/GP64 but not VP39 might be areas which precede the arrival of nucleocapsids at the membrane.

Current views of membrane fission associated with viral egress suggest a requirement for both “push” and “pull” forces (6, 13, 44). The pull force is due to accumulation of viral glycoproteins spanning the cellular membrane, which causes localized distortions and curvature of the membrane, thereby reducing the energy requirements for membrane fission (9, 44). The push force originates on the cytoplasmic side of the membrane, possibly through the accumulation of viral proteins on the inner side of the membrane and/or preassembled nucleocapsids, in conjunction with cellular factors (44). Considering the localization of ME53:GFP, ME53 might represent such an inner viral protein, assembling below patches of the membrane-spanning GP64 triplexes. Additionally, although microtubules have been implicated in baculovirus egress (11, 40) and may provide additional push, actin may also be involved. This is supported by the fact that actin was identified through mass spectrophotometry analysis of budded virions from Epstein-Barr virus (16), human cytomegalovirus (37), HIV (7), and, more recently, AcMNPV (42). Furthermore, actin-based motility has been implicated in AcMNPV nuclear bypass transport in gut epithelial cells of infecting ODV nucleocapsids to the cell periphery after initial entry, and these nucleocapsids accumulate at the tips of actin-rich cell surface spikes (30). Here we demonstrate a potential relationship between ME53-containing foci and cell surface actin-rich spikes. Previous studies (43) have implicated early-expressed GP64 in the “repackaging” of incoming ODV nucleocapsids as BV in midgut epithelial cells. It is interesting to note that this strategy requires early gene expression, including production of GP64 (43, 46). Our previous study (8) indicated that deletion of the me53 early promoter resulted in a moderate decrease in budded-virus production in tissue culture; however, it did not evaluate the effects of early me53 deletion in vivo. It is possible that in addition to GP64, ME53 may also be required for efficient budding of “repackaged” ODV nucleocapsids from initially infected epithelial cells.

To investigate the potential interactions between VP39, ME53, and GP64, a set of me53/vp39 and me53/gp64 double-knockout bacmids were generated. From our observations with a vp39-null bacmid, both VP39 and nucleocapsid assembly do not appear to influence the localization of ME53:GFP to the nucleus or formation of the ME53/GP64 foci at the cell periphery. This indicates that the ME53:GFP found at the periphery of the cell is not localized there as a consequence of association with nucleocapsids (42), and nucleocapsid-mediated budding events are not required for formation of ME53/GP64 foci.

Deletion of gp64 eliminates the accumulation of ME53:GFP at the periphery of the cell, and reintroduction of a wild-type copy of gp64 restores this normal localization of ME53:GFP. However, in a null me53 background and when expressed alone in uninfected cells, GP64 still accumulates at distinct areas at the plasma membrane. Cumulatively, this suggests that it is the localization of GP64 that determines the budding sites and influences the accumulation of ME53:GFP at foci at or near the cell surface and not vice versa. Considering that both proteins are present at early times in the infection but ME53:GFP/GP64 foci do not form until later times postinfection and that cotransient expression of only GP64 and ME53:GFP in an uninfected cell does not result in accumulation of GP64/ME53:GFP foci, another viral protein(s), or a virally induced cellular signal other than the arrival of a nucleocapsid, must cause ME53 to be recruited to areas of GP64 accumulation. Similarly in uninfected cells, ME53:GFP fails to localize to the nucleus, again suggesting the involvement of other factors, such as viral proteins or virally induced posttranslational modifications, in virus-infected cells for nuclear translocation of ME53.

In this study we demonstrated that despite the strong colocalization of ME53:GFP with GP64 at the membrane, ME53 is associated exclusively with the nucleocapsid fraction of BV. This is in agreement with recent mass spectrometry (MS) studies on the budded virion from AcMNPV, which also detected ME53 in the nucleocapsid fraction (42). The localization of ME53:GFP with GP64 and the influence of the presence of GP64 on the localization of ME53:GFP in an infected cell are suggestive of an interaction between these proteins during infection. However, the fractionation data imply an association between ME53 and capsid and/or capsid-associated proteins. ME53 may therefore be providing a link between the nucleocapsid and the envelope, similar to the role played by the matrix proteins of other enveloped viruses (29).

Since evidence presented here suggests that ME53/GP64 foci may represent viral budding sites and these foci can form in the absence of nucleocapsids, it is likely that ME53 and GP64 associate, with the help of other viral proteins, prior to the arrival of nucleocapsids, thereby creating localized “bud zones” at the plasma membrane. This would be similar to what is seen for other budding enveloped viruses such as orthomyxoviruses (influenza virus) or herpesviruses (herpes simplex virus type 1 [HSV-1]). In the case of influenza virus, budding results from the coordinated association of the M1 matrix protein with the cytoplasmic tails of glycoproteins HA, NA, and M2 and the association of M1 with ribonucleoprotein structures (RNPs) (1, 29). Similarly to ME53:GFP and GP64, M1 colocalizes with HA at the plasma membrane (1), with M1 on the cytoplasmic side of lipid raft microdomains and clusters of HA and NA on the extracellular face of the membrane, creating viral assembly and bud sites (34). Similarly, we also have preliminary immunogold EM evidence to suggest that ME53 accumulates on the cytoplasmic side of the membranes of infected Sf-21 cells (see Fig. S4 in the supplemental material).

A more complex situation for viral egress is seen with HSV-1, where multiple, possibly redundant, interactions between the outer tegument and the cytoplasmic tail domains of the 11 identified envelope glycoproteins, and tegument and capsid, drive viral assembly (18). For example tegument protein pUL48 (VP16) interacts with glycoprotein H as well as inner tegument protein pUL36 and capsid-associated proteins pUL35 and pUL38 (24, 38), thereby helping to bridge the envelope to the capsid. Additionally, deletion of pul48 abolishes secondary envelopment and release of virions (12, 27, 41). Although it is not known whether these interactions would occur in the absence of an intermediate virion, the interaction between the envelope glycoproteins and the capsid is central to driving viral egress. It is interesting to note that pUL48 is also involved in transcription of viral genes after viral entry (47). Although pUL48 tends to fractionate with the viral envelope, other tegument proteins such as pUL37 and pUL46 fractionate with the capsid (28, 35).

The data presented here and in previous studies (8) support the hypothesis that ME53 is a structural component of a multiprotein scaffold that prepares budding sites and helps bridge the viral envelope glycoproteins, such as GP64, with budding nucleocapsids. Specifically, (i) deletion of me53 results in a greater-than-1,000-fold reduction in budded-virus production, (ii) ME53:GFP localizes in a GP64-dependent manner to discrete areas at or near the plasma membrane and these foci colocalize with GP64 and often with VP39, (iii) deletion of gp64 abolishes ME53:GFP accumulation at the cell periphery, and (iv) despite strong colocalization of ME53:GFP and GP64, ME53 remains capsid associated after virion fractionation.

Here we implicate ME53 in concert with GP64 in egress at the plasma membrane; however, their colocalization at the plasma membrane likely relies on other viral proteins to mediate interaction or transport to the membrane. It would be interesting to identify other ME53- or GP64-interacting partners, perhaps through yeast two-hybrid analysis or tandem affinity purification (TAP) coupled with mass spectrometry analysis.