Autographa californica Multiple Nucleopolyhedrovirus Ac34 Protein Retains Cellular Actin-Related Protein 2/3 Complex in the Nucleus by Subversion of CRM1-Dependent Nuclear Export

Actin, nucleation-promoting factors (NPFs), and the actin-related protein 2/3 complex (Arp2/3) are key elements of the cellular actin polymerization machinery. With nuclear actin polymerization implicated in ever-expanding biological processes and the discovery of the nuclear import mechanisms of actin and NPFs, determining Arp2/3 nucleo-cytoplasmic shuttling mechanism is important for understanding the function of nuclear actin. A unique feature of alphabaculovirus infection of insect cells is the robust nuclear accumulation of Arp2/3, which induces actin polymerization in the nucleus to assist in virus replication. We found that Ac34, a viral late gene product encoded by the alphabaculovirus Autographa californica multiple nucleopolyhedrovirus (AcMNPV), is involved in Arp2/3 nuclear accumulation during virus infection. Further assays revealed that the subcellular distribution of Arp2/3 under steady-state conditions is controlled by chromosomal maintenance 1 (CRM1)-dependent nuclear export. Upon AcMNPV infection, Ac34 inhibits CRM1 pathway and leads to Arp2/3 retention in the nucleus.

Introduction cytoplasm (Reviewed in [33]). After its nuclear export, RanGTP is hydrolyzed to RanGDP, and the complex releases the cargo protein to the cytosol.
In this research, we found that Arp2/3 subcellular distribution is controlled by CRM1-dependent nuclear export under steady-state conditions. AcMNPV infection induced Arp2/3 nuclear retention by inhibiting the CRM1 pathway with a viral late gene product, Ac34. To our knowledge, this is the first study describing the nuclear retention mechanism of Arp2/3 under steady-state and virus-infection conditions. We also provide the first example of a virus specifically blocking the CRM1 nuclear export pathway to promote its replication.

Results
An AcMNPV late gene product induces P40 nuclear accumulation Previously, we and other groups have revealed the nuclear accumulation mechanism of P78/83 and G-actin [29][30][31], two key elements of the actin polymerization machinery, during AcMNPV infection. To investigate how AcMNPV accumulates Arp2/3, the central regulator of actin polymerization, in the nucleus, we cloned the cDNA sequences of Arp2/3 subunits from Sf9 cells, a commercially available Spodoptera frugiperda cell line commonly used for baculovirus infection (GenBank Accession: KJ187399.1, JQ364941.1, KJ187400.1, GU356595.1, KJ187401.1, KJ187402.1) [34]. Here, P40 was selected to represent Arp2/3 because P40 appeared to be the most abundant protein detected by either Western blot or fluorescence microscopy (Arp2 and P20 were less abundant than P40; Arp2 could only be detected by Western blot; other subunits were barely detected by Western blot or fluorescence microscopy when transiently expressed in Sf9 cells). We prepared plasmid-based expression constructs encoding P40 tagged with a V 5 epitope (P40-V 5 ) at its C-terminus or P40 fused to enhanced green fluorescent protein at its N-terminus (EGFP-P40) to monitor the Arp2/3 dynamics during AcMNPV infection.
Cytoplasmic localization was noted by immunofluorescence for P40-V 5 for mock infected cells (Fig 1A, left panel), but some nuclear localization was observed for cells infected with AcMNPV carrying an EGFP marker (vAc egfp , diagramed in S1A Fig). As evidenced by cell fraction and Western blot, P40-V 5 was present in only the cytoplasmic fraction of mock infected cells, while some P40-V 5 was found in the nuclear fraction of vAc egfp infected cells (Fig 1A, right panel). The nuclear and cytoplasmic control proteins, histone and tubulin respectively, were identified in the nuclear and cytoplasmic fractions, respectively, validating the effectiveness of the fractionation (Fig 1A, right panel). Similarly, by fluorescence microscopy, EGFP-P40 localized to the nucleus only in cells infected with AcMNPV expressing polyhedrin (vAc polh , diagramed in S1A Fig) (Fig 1B). This phenotype is in accordance with the observation described by Goley et al., in which yellow fluorescent protein-tagged P21 (P21-YFP) was observed to accumulate in the nucleus during AcMNPV infection [23].
We next investigated which class of viral genes needed to be expressed for P40 nuclear accumulation. Aphidicoline (APH), an inhibitor of DNA synthesis, was used to shut off AcMNPV . Subcellular distribution of P40 in plasmid-transfected cells. P40-V 5 was transiently expressed in Sf9 cells by plasmid transfection and then the cells were either mock-infected or infected with vAc egfp (MOI = 2). At 24 hpi, the cells were subjected to either immunofluorescence microscopy (left panel) or cell fractionation (right panel). For immunofluorescence microscopy, the cells were fixed, probed with anti-V 5 antibody to detect P40-V 5 , and visualized with Alexa Fluor 568-conjugated secondary antibody (red). Hoechst 33258 was used to stain the nuclear DNA (blue). For the cell fractionation assay, the cells were harvested and separated into nuclear (N) and cytoplasmic (C) fractions. The fractions were separated by SDS-PAGE and probed with anti-V 5 , anti-tubulin, and antihistone-H 3 antibodies to detect P40, tubulin, and histone-H 3 , respectively. (B). EGFP-P40 subcellular late gene expression [35]. The dynamic localization of P40 in AcMNPV-infected cells was monitored in the presence or absence of APH. Early during infection (0-12 hpi), P40 resided predominantly in the cytoplasm irrespective of APH treatment (Fig 1D). During the late phase of infection (After 12 hpi), AcMNPV infection resulted in detectable P40 accumulation in the nucleus in the absence of APH, suggesting that viral late gene products may play an important role in P40 nuclear accumulation. When the expression of viral late genes was shut off by APH, P40 failed to accumulate in the nucleus, as demonstrated by both immunofluorescence microscopy and cell fractionation assays ( Fig 1D). Together, these data indicated that viral late gene products are responsible for P40 nuclear accumulation.

Ac34 is involved in Arp2/3 nuclear accumulation induced by AcMNPV
To identify the viral protein responsible for the nuclear accumulation of P40, AcMNPV ORFs were individually cloned into a pIZ-V 5 transient expression vector (Invitrogen). Each individual viral ORF was co-expressed with EGFP-P40 and the subcellular distribution of P40 was determined using fluorescence microscopy. Among the 118 viral ORFs screened (S1 Table), only Ac34, a viral late gene product, appeared to be sufficient to induce P40 nuclear accumulation. Ac34 tagged with mCherry (mC-Ac34) was shown to accumulate EGFP-P40 or P40-V 5 in the nucleus when co-expressed in Sf9 cells (Fig 2A and 2B). As a control, we co-expressed P40 with non-fused mCherry (Fig 2A and 2B), resulting in a predominantly cytoplasmic localization of P40. Similar nuclear relocation induced by Ac34 also occurred for P20 (S2A Fig), indicating that Ac34 is sufficient to accumulate Arp2/3 in the nucleus.
Arp2/3 nuclear accumulation is dependent on the presence of Ac34 in the nucleus Previously, we revealed that virus-encoded NPF P78/83, another key element of the nuclear actin polymerization machinery during AcMNPV infection, is relocated to the nucleus by binding to and co-transportation with C42, which harbors a nuclear localization sequence (NLS) [31]. Based on this scenario and the fact that Ac34 is present in the nucleus (Fig 2A and  2B), we were prompted to explore whether P40 nuclear accumulation is also correlated to the presence of Ac34 in the nucleus. distribution. EGFP-P40 was expressed in mock-infected Sf9 cells or cells infected with vAc polh (MOI = 2), a bacmid derived from the AcMNPV genome with an inserted polyhedrin open reading frame (ORF). At 48 hpt, the cells were analyzed by fluorescence microscopy. (C). EGFP-P40 associates with Arp2 and P20. EGFP (approx. 27 kDa) or EGFP-P40 (approx. 69 kDa) was co-expressed with Arp2-Ha (approx. 46 kDa) or P20-Ha (approx. 21 kDa) in Sf9 cells, respectively. At 48 hpt, the cells were harvested and subjected to a co-IP assay using anti-Ha. The immunoprecipitated proteins and the WCL were separated by SDS-PAGE and probed with anti-Ha and anti-EGFP. (D). AcMNPV late gene products induce P40 nuclear accumulation in the late infection phase. Sf9 cells transfected with the P40-V 5 encoding plasmid were either mock infected or infected with vAc egfp (MOI = 2) in the presence or absence of APH. The cells were fixed at 6, 12 and 24 hpi or harvested for fractionation and further treated as described in (A). Scale bar: 20 μm; Scale bar in the zoomed panel: 10 μm. A series of mCherry-fusedC-terminal and N-terminal Ac34 truncations (S1C Fig) was prepared to identify the sequence responsible for Ac34 nuclear localization. Fluorescence microscopy demonstrated that the removal of amino acids (aa) 195-215 of Ac34 (mC-Ac34  ) resulted in the cytoplasmic localization of Ac34 (Fig 3), which is in sharp contrast to the full-length Ac34 (mC-Ac34) and all the tested N-terminal Ac34 truncations, which exhibited a predominantly nuclear localization pattern (S3 Fig). This phenotype indicated that the aa 195-215 region plays a major role in determining the presence of Ac34 in the nucleus, although sequence analysis did not show any classic NLS pattern (tandem repeats of lysine and arginine) within this region. Notably, when the C-terminal truncation of Ac34 was extended to aa 75 or further (mC-Ac34 1-75 and mC-Ac34 1-55 ), a diffuse cellular distribution of Ac34 was observed (Fig 3), which could be attributed to free nucleo-cytoplasmic shuttling of the resulting low-molecular-mass polypeptides.
Interestingly, among all the tested Ac34 truncations, only full-length Ac34 could accumulate EGFP-P40 in the nucleus, and the removal of aa 195-215 of Ac34 resulted in a lack of EGFP-P40 nuclear accumulation (Fig 3), thus supporting our hypothesis that P40 nuclear accumulation is dependent on the presence of Ac34 in the nucleus. Similar nuclear accumulation also occurred for P20 (S2A Fig), indicating that the aa 195-215 region is essential for Ac34 to accumulate Arp2/3 in the nucleus.
To verify the role of aa 195-215 of Ac34 in P40 nuclear accumulation during AcMNPV infection, Ac34  was used to rescue vAc34KO egfp , generating vAc34KO ac34Δ195-215 (diagramed in S1B Fig). When P40 was co-expressed in bacmid-transfected cells, only vAc34K-O ac34 could induce P40 nuclear accumulation at 48 hpt, in contrast to the cytoplasmic distribution pattern of P40 in vAc34KO egfp and vAc34KO ac34Δ195-215 -transfected cells ( Fig 4A). Similar nuclear accumulation also occurred for P20 (S2B Fig), further confirming that Ac34 is responsible for the Arp2/3 nuclear accumulation induced by AcMNPV, and aa 195-215 are required for the accumulation.
Nuclear actin polymerization requires the nuclear localization of Arp2/3. To explore whether Ac34 is involved in AcMNPV-induced nuclear actin polymerization, Sf9 cells were transfected with vAc34KO egfp , vAc34KO ac34 , or vAc34KO ac34Δ195-215 and stained with phalloidin at 48 hpt to visualize F-actin. Among all the transfected bacmids, only vAc34KO ac34 induced typical nuclear actin polymerization, with F-actin accumulating in the nuclear region ( Fig 4B). The cells transfected with the other bacmids showed no significant F-actin accumulation in the nucleus ( Fig 4B). This phenotype can easily be attributed to the absence of Arp2/3 in the nucleus due to either ac34 knockout (vAc34KO egfp ) or the loss of its nuclear localization determinant (vAc34KO ac34Δ195-215 ).
Arp2/3 cytoplasmic distribution is controlled by CRM1-dependent nuclear export CRM1 is a highly versatile transport receptor that mediates the nuclear export of a large number of proteins. Inhibition of CRM1 results in nuclear retention of NES-bearing protein. Bioinformatics assay (LocNES, http://prodata.swmed.edu/LocNES/) [37] predicted that the P40 Cterminus (aa 360-374), a leucine-rich sequence, is a putative NES. We then explored whether the cytoplasmic distribution of P40 is CRM1-dependent.
P40-V 5 was transiently expressed in Sf9 cells. Immunofluorescence microscopy showed that P40 exhibited significant nuclear accumulation after adding leptomycin B (LMB), a specific CRM1 inhibitor (Fig 5A) [38][39][40]. Removing aa 360-374 of P40 resulted in P40 (P40 Δ360-374 -V 5 ) accumulation in the nucleus (Fig 5A), implying that the P40 C-terminus functions as a NES to determine the cytoplasmic distribution of P40. To further confirm P40 nuclear accumulation is CRM1-dependent, cellular CRM1 was knocked-down using double-stranded RNA   AcMNPV infection inhibits cellular CRM1-dependent nuclear export Given the evidence that the cytoplasmic distribution of Arp2/3 is controlled by CRM1-dependent nuclear export, and AcMNPV infection induces Arp2/3 nuclear accumulation, one of the possible explanations is that AcMNPV infection inhibits cellular CRM1-dependent nuclear export and subsequently leads to Arp2/3 retention in the nucleus.
To evaluate the influence of AcMNPV infection on the CRM1 pathway, a classic NES peptide (LQNKLEELDL) [41] was fused to mCherry (mCherry-NES) and EGFP (EGFP-NES) to construct probes for CRM1-dependent nuclear export.
When mCherry-NES was transiently expressed in Sf9 cells, a predominantly cytoplasmic distribution pattern was observed (Fig 6). Adding LMB to the culture medium resulted in the accumulation of the majority of mCherry-NES in the nucleus (Fig 6), indicating that the nuclear export of mCherry-NES is CRM1-dependent. Similar nuclear retention upon CRM1 knockdown was also observed in EGFP-NES expressing cells (S4 Fig).
When virus stock solution (vAc egfp ) was added to the culture medium, mCherry-NES accumulated in the nucleus of infected cells and remained in the cytoplasm of uninfected cells (Fig  6). This differential distribution indicated that AcMNPV causes dysfunctional cellular CRM1-dependent nuclear export. To identify which class of viral genes was responsible for the dysfunction, APH was added to the culture medium after virus infection. All the cells showed cytoplasmic distribution of mCherry-NES (Fig 6), suggesting that AcMNPV late gene products were responsible for the virus-induced dysfunction in the CRM1 pathway.

Ac34 inhibits cellular CRM1-dependent nuclear export during AcMNPV infection
Because we demonstrated that Ac34 is responsible for virus-induced Arp2/3 nuclear accumulation, and AcMNPV inhibits CRM1-dependent nuclear export, which can lead to Arp2/3 retention in the nucleus, it is highly possible that Ac34 is involved in the dysfunction of the CRM1 pathway induced by AcMNPV.
To test this hypothesis, EGFP-NES was co-expressed with mCherry or mC-Ac34 in Sf9 cells. Fluorescence microscopy showed that EGFP-NES resided in the cytoplasm in the presence of mCherry, whereas it accumulated in the nucleus in the presence of mC-Ac34 or LMB (Fig 7A). This phenotype indicated that Ac34 is sufficient to inhibit CRM1-dependent nuclear export. Removing the Ac34 C-terminus (aa 195-215), which is essential for Ac34 nuclear localization and Arp2/3 nuclear accumulation, also abolished EGFP-NES nuclear retention (Fig 7A).
Taken together, this evidence demonstrated that Ac34 induces Arp2/3 nuclear retention by inhibiting CRM1-dependent nuclear export during AcMNPV infection.

Discussion
The nuclear import mechanisms of key elements of actin polymerization machinery, including actin and N-WASP, have been previously identified [7-10]. However, nucleo-cytoplasmic shuttling mechanism of Arp2/3, the central regulator of actin polymerization, has not been elucidated yet. In this study, a unique virus-infection system was employed to reveal how Arp2/3 . Ac34 is sufficient to inhibit CRM1-dependent nuclear export. EGFP-NES was co-expressed with mCherry, mC-Ac34, or mC-Ac34  in Sf9 cells. In cells co-expressing EGFP-NES and mCherry (row 1 and 2), ethanol (1 μl) or LMB (0.1 μg/ml) was added to the culture medium at 44 hpt. At 48 hpt, all the cells were fixed and subjected to fluorescence microscopy assays. (B). Ac34 is involved in AcMNPV-induced CRM1 pathway dysfunction. A mCherry-NES encoding plasmid was co-transfected is retained in the nucleus, which could shed light on the nucleo-cytoplasmic shuttling mechanism of Arp2/3 under different physiological or pathophysiological conditions. Viral manipulation of cellular the nucleo-cytoplasmic transport of proteins has been extensively documented in recent years (reviewed in [42]), in particular in cardioviruses and enteroviruses. Cardioviruses use their leader proteins to induce the hyper-phosphorylation of nucleoporins and disrupt the RanGTP gradient [43,44], thus inducing an efflux of the nuclear proteins required for viral replication and leading to interferon suppression. Enterovirus infection results in cellular protein retention in the cytoplasm via the degradation of nucleoporins mediated by the virus-encoded proteases 2A and 3C [45][46][47]. Other viruses, such as herpes simplex virus [48], human papillomavirus [49,50], severe acute respiratory syndrome coronavirus [51], Ebola virus [52], and measles virus [53], employ a variety of methods to interfere with the nucleo-cytoplasmic shuttling of cellular proteins, therefore facilitating viral replication and escape from the host anti-viral immune response.
Unlike most viruses, which primarily induce impaired protein nuclear import or enhance protein nuclear export, our results demonstrated that AcMNPV infection results in impaired protein nuclear export. As a nucleopolyhedrovirus, most AcMNPV replication processes, including viral genome replication, gene transcription, and nucleocapsid assembly, all occur in the nucleus. These processes require a variety of proteins, including, but not limited to, virusencoded transcription factors, transcriptases, and capsid proteins, as well as some cellular proteins (e.g., actin, Arp2/3), to accumulate in the nucleus. AcMNPV contains 156 predicted ORFs at least 50 aa in length. Aside from a limited number of exceptions, the nuclear import mechanisms of most viral and cellular proteins during AcMNPV infection remain unknown. Currently, at least 7 exportins have been identified in eukaryotic cells [8,32,[54][55][56][57][58]. Unlike other exportins that only transport highly specialized cargoes (Reviewer in [59]), CRM1 mediates the nuclear export of many NES-bearing proteins, and its dysfunction leads to the nuclear accumulation of these proteins. Based on bioinformatics prediction (NetNES, http://www.cbs. dtu.dk/services/NetNES/) [60], 98 AcMNPV proteins contain putative residues that could serve as a NES (S2 Table). Such a high percentage of viral proteins bearing putative NESs implies that CRM1-dependent nuclear export may determine the subcellular distribution of many viral proteins, and the inhibition of CRM1-dependent nuclear export by Ac34 could possibly play a key role in the AcMNPV-induced nuclear accumulation of proteins. Whether Ac34 also influences other exportins or these exportins also contribute to the virus-induced protein nuclear accumulation remain to be explored.
Ac34 homologues are presented in all sequenced alphabaculoviruses but absent in betabaculoviruses [61]. Alphabaculoviruses and betabaculovirusesbehave in significantly different ways. In respect to cytopathology, alphabaculoviruses assemble their nucleocapsid in the nucleus, whereas betabaculovirusesinduce nuclear membrane rupture, and nucleocapsid assembly occurs in a combination of the cytoplasm and the nucleoplasm [62]. This cytopathologic difference suggests that unlike alphabaculoviruses, betabaculoviruses do not need to accumulate the cytoplasmic actin polymerization machinery to the nucleus. As a consequence, betabaculoviruses do not need a viral protein or mechanism to induce nuclear accumulation of Arp2/3 (although only P40 and P20 were proved to be retained in the nucleus of AcMNPV-infected cells in this study, both Arp2/ 3 components behave in a similar way upon virus infection), which is supported by the evidence that Ac34 homologues are absent in the genomes of betabaculoviruses [61]. Nuclear G-actin is required for the transcriptional activity of RNA polymerases [63][64][65] and the epigenetic activation of chromatin (Reviewed in [5,66]). Among the three key actin polymerization elements that are accumulated in the nucleus during AcMNPV infection, only Gactin is recruited to the nucleus by early viral gene products [29,30]. This early nuclear accumulation of G-actin could increase the nuclear G-actin pool and promote the transcription of viral early genes that are transcribed by host RNA polymerase II [61]. Late in infection, P78/83 and Arp2/3 accumulate in the nucleus and induce nuclear actin polymerization that converts G-actin to F-actin. The resulting nuclear G-actin pool depletion could lead to the loss of the transcriptional activity of host RNA polymerases and the epigenetic reprogramming of host chromatin towards transcriptional inhibition, which could contribute to the host gene transcription shutoff that occurs in the late phase of baculovirus infection [67,68]. Consistent with this, cytochalasin D, a chemical that specifically prevents actin polymerization, behaves as an antagonist of the virus-induced shutdown of host gene expression [69]. In this respect, nuclear actin polymerization induced by baculovirus infection may also participate in the regulation of host/virus gene expression by the modulation of the nuclear G-actin pool, in addition to its role in assisting viral nucleocapsid assembly and transport, which has long been recognized.
In summary, Ac34 subversion of the CRM1-dependent nuclear export during AcMNPV infection suggests that alphabaculoviruses may employ an efficient way by encoding a single protein to accumulate multiple viral and host proteins in the nucleus to assist in virus replication. As a key element of actin polymerization machinery, Arp2/3 is present in both the cytoplasm and the nucleus. Our finding that Arp2/3 nuclear-cytoplasmic shuttling is CRM1-dependent sheds light on how cells manage to control actin polymerization machinery in different cellular compartments to exert different functions.

Cell culture, transfection, and infection
Sf9 cells from S. frugiperda were cultured in Grace's medium (Invitrogen) with 5% fetal bovine serum (Invitrogen) and 0.1% Antibiotic-Antimycotic (Invitrogen) at 27°C. Sf9 cells were transfected with the indicated plasmids or bacmids using the Cellfectin II reagent (Invitrogen) following the standard procedures. For infection, the Sf9 cells were incubated with virus stock solution for 1 h at a multiplicity of infection (MOI) of 2. (MOI = 2). The cells were then rinsed twice and then incubated in fresh medium or medium with APH (5 μg/ml) (Sigma). The cells were fixed for further immunofluorescence detection at 6, 12, and 24 hpi. To block CRM1-dependent nuclear export, LMB (0.1 μg/ml) (Beyotime) was added to the culture medium and the cells were incubated for 4 hours before the fluorescence assays.

Construction of AcMNPV ORF transient expression library
One hundred fifty-four ORFs of AcMNPV were cloned by polymerase chain reaction (PCR) and inserted into pIZ-V 5 (Invitrogen). All the viral ORFs began with ATG and ended without the stop codon to create an in-frame fusion with the V 5 epitope. All the constructs were sequenced, and 118 viral ORFs were tested for their impact on the change in P40 subcellular distribution (S1 Table).

Construction of plasmids and bacmids
All the plasmids used in this research for transient expression were prepared by standard molecular cloning protocols. The indicated genes, gene truncations, and genes with epitope tags were generated by PCR or site-directed mutagenesis (Transgene) and inserted into pIZ-V 5 /Ha vectors (Invitrogen).
To prepare recombinant bacmids, the Bac-to-Bac system was employed according to Invitrogen's protocol. In brief, Ac34 expression cassettes controlled by the native ac34 promoter were cloned into pFbdg, a pFastbac-Dual vector (Invitrogen) bearing an EGFP expression cassette controlled by the p10 promoter [31]. The resulting shuttle vectors were then used to transform DH10B E. coli cells harboring the vAc34KO bacmid provided by Cai et al. to generate the transposed bacmid constructs [36].
Maps of the plasmids and bacmids prepared in this research are diagramed in S1 Fig.

Cell fractionation
Cells were rinsed with ice-cold PBS and lysed with homogenization buffer (10 mM HEPES pH = 7.9, 10 mM KCl, 1.5 mM MgCl 2, 0.1 mM EGTA, 0.5 mM DTT, 2 mM PMSF, 1 μg/ml Proteinase Inhibitors (Roche)). The cell membranes were disrupted by passing through a 25G needle 5 times, and the lysates were then spun at 1000×g for 10 min at 4°C. The supernatant containing the crude cytoplasmic fraction was collected in 1.5 ml tubes and spun at 20,817×g for 30 min at 4°C, and the supernatant was collected as the purified cytoplasmic fraction. The nuclear pellet was rinsed in 1 ml homogenization buffer and centrifuged at 1000×g for 10 min at 4°C. The pellet was re-suspended in 100 μl extraction buffer (10 mM HEPES pH = 7.9, 0.4 M NaCl, 1.5 mM MgCl 2 , 0.1 mM EGTA, 0.5 mM DTT, 2 mM PMSF, 1 μg/ml Proteinase Inhibitors) under gentle shaking for 30 min at 4°C. The suspension was centrifuged at 20,817×g for 30 min at 4°C and the supernatant was collected as the nuclear fraction. The protein concentrations of all samples were determined using Bradford assays (Bio-Rad) and the samples were subjected to Western blot assays. Anti-histone H3 (Sigma) and anti-tubulin (Sigma) diluted to 1:1000 were used to verify the quality of the cytoplasmic and nuclear fractions, respectively. After HRP-conjugated secondary antibody (1:10,000 dilution, Jackson Laboratory) incubation, the blots were developed using an enhanced chemiluminescence kit (Pierce).

Immunoprecipitation
Sf9 cells were rinsed with ice-cold PBS and lysed with RIPA buffer (50 mM Tris, pH = 7.5, 1 mM EGTA, 1 mM EDTA, 1% Triton X-100, 150 mM NaCl, 2 mM DTT, 100 μM PMSF, 1 μg/ ml Proteinase Inhibitors). The cell lysates were centrifuged at 20,817×g at 4°C for 10 min and the supernatants (WCL) were collected. The protein concentrations of the WCL were determined by Bradford assays and 1500 μg was mixed with 2 μg anti-Ha (Sigma) and Protein G Agarose (Millipore) and incubated at 4°C overnight according to the manufacturer's protocol. The immunoprecipitated samples were centrifuged and washed three times and subjected to Western blot assays using anti-Ha (1:1000 dilution) and anti-EGFP (1:1000 dilution, Invitrogen).
For F-actin staining, the cells were transfected with different recombinant bacmids, fixed, and permeabilized as described above and then stained with 0.7 U/ml Alexa Fluor 568-phalloidin (Invitrogen) and Hoechst 33258 for 10 min. The cells were then washed three times with PBS and examined by confocal microscopy using a PerkinElmer UltraVIEW VoX microscope.

Fluorescence quantification and statistical assays
The fluorescence quantification data were obtained using Volocity 6.3 software (PerkinElmer) and Student's T-test was performed to compare the differences between the tested samples.
Supporting Information S1 Fig. Diagrams of plasmids and bacmids. (A). bMON14272-based bacmids. The coding sequences of EGFP or Polyhedrin were inserted into the downstream region of the p10 promoter (P p10 ) or the polyhedrin promoter (Ppolh). The resulting plasmids were transposed into bMON14272 (Invitrogen), a bacmid harboring the wild-type AcMNPV genome, at the polyhedrin (polh) locus using the Bac-to-Bac method. (B). vAc34KO-based bacmids. The coding sequences of EGFP, Ac34, and Ac34  were inserted into the downstream region of the p10 promoter (P p10 ) or the ac34 promoter (P ac34 ). The resulting plasmids were transposed into vAc34KO at the polyhedrin (polh) locus using the Bac-to-Bac method. (C). N-terminal or Cterminal Ac34 truncations were fused to mCherry. The expression of the fusion proteins was controlled by an OpIE2 promoter (P opIE2 ). (TIF) S2 Fig. Ac34 mediates P20 nuclear relocation. (A). Ac34 is sufficient to relocate P20 to the nucleus. P20-Ha was co-expressed with mCherry, mC-Ac34, or mC-Ac34  in Sf9 cells. At 44 hpt, LMB (0.1 μg/ml) was added to the culture medium and the cells were incubated for 4 hours. At 48 hpt, all the cells were fixed and subjected to immunofluorescence microscopy assays using anti-Ha. Scale bar: 20 μm. (B). The impact of CRM1 knockdown on P20 subcellular distribution. Sf9 cells were transfected with ds-crm1 1-1000 or ds-control. At 24 hpt, plasmids encoding P20-Ha were transfected to dsRNA-bearing cells. At 72 hpt, cells were fixed and subjected to immunofluorescence microscopy assay using anti-Ha. Scale bar: 5 μm. (C). Ac34 is involved in P20 nuclear relocation induced by AcMNPV. Plasmids encoding P20-Ha were cotransfected with vAc egfp , vAc34KO egfp , or vAc34KO ac34Δ195-215 . At 48 hpt, the cells were fixed and subjected to immunofluorescence microscopy assays using anti-Ha. The arrow pointed to the bacmid-free cells that showed different spatial pattern of P20-Ha in comparison with the adjacent cells bearing vAc34KO ac34 . Scale bar: 20 μm. Densitometry assays were performed simultaneously. The bars represent the means and standard errors of the means for three independent experiments. Each experiment involves the quantification of 30 transfected cells. ÃÃÃ ,