A new mechanism for nuclear import by actin-based propulsion used by a baculovirus nucleocapsid

The main mechanism of nuclear transport is by passage through the nuclear pore complex (NPC). Molecules from 5 to 39 nm (Mohr et al., 2009; Pante and Kann, 2002) cross the NPC by bearing one or more nuclear localization sequences (NLSs). NLSs are recognized by nuclear transport receptors (NTRs) that translocate the cargo through the NPC (Cautain et al., 2015; Wente and Rout, 2010). The small GTPase Ran regulates the binding and dissociation of NTRs and NLS-containing cargos. Numerous studies have provided evidence for deviations to this conventional nuclear import mechanism (reviewed by Wagstaff and Jans, 2009). For example, nuclear import of cyclin-B1–Cdc2 occurs independently of Ran but requires importin-β (Takizawa et al., 1999) and some nuclear import pathways are importin-β independent, for which the cargo acts as an NTR (Kumeta et al., 2012).

Viruses and subviral particles use both conventional and unconventional pathways to enter the nucleus (Fay and Pante, 2015a,,b). The cytoskeleton enhances this process by accelerating viral transport through the cytoplasm toward the nucleus (Wagstaff and Jans, 2009). Although most viruses use microtubule-mediated transport (Dodding and Way, 2011), baculoviruses use actin-based motility. Autographa californica nucleopolyhedrovirus (AcMNPV), the most-studied baculovirus, releases its nucleocapsid into the cytoplasm during cell entry (Rohrmann, 2013). The nucleocapsid is subsequently propelled towards the nucleus by inducing actin polymerization at one of its ends through VP78/83 (a single protein encoded by the baculovirus VP78/83 gene), a viral Wiskott–Aldrich syndrome protein (WASP)-like protein (Lanier and Volkman, 1998; Machesky et al., 2001; Mueller et al., 2014; Ohkawa et al., 2010).VP78/83 is a nucleation-promoting factor that activates actin polymerization by the host Arp2/3 complex (Goley et al., 2006). The rod-shaped nucleocapsid (30×250–300 nm) then enters the nucleus by crossing the NPC lengthwise (Au and Pante, 2012; Au et al., 2013). Whether baculoviruses use a conventional or an unconventional nuclear import pathway remains to be determined. In any case, proteins of the AcMNPV nucleocapsid do not contain any obvious functional NLS that could mediate the nuclear import of the nucleocapsid.

Here, we show that nuclear import of AcMNPV nucleocapsids occurs independently of the Ran-GTPase cycle, but depends on Arp2/3-mediated F-actin nucleation. We further demonstrate the involvement of actin polymerization in the nuclear import of AcMNPV by reconstituting nuclear import of nucleocapsids in purified nuclei supplemented with G-actin and Arp2/3 under permissive conditions for actin polymerization. Our data suggest that baculovirus-induced actin polymerization is involved in the mechanism of nuclear import of AcMNPV nucleocapsids.

In order to study nuclear import of AcMNPV nucleocapsids, we used the well-established nuclear import assay with digitonin-permeabilized cells (Adam et al., 1990; Cassany and Gerace, 2009). In this system, nuclear import is assayed by incubating the digitonin-permeabilized cells with a fluorescently labeled cargo and either exogenous cytosol or purified recombinant nuclear transport factors and an energy-regenerating system. Control experiments with a 70-kDa dextran labeled with Texas Red, which does not diffuse through the NPC, demonstrated that digitonin did not disrupt the nuclear envelope (data not shown). Control experiments with a Cy3-conjugated NLS-tagged bovine serum albumin (Cy3–NLS-BSA) showed that this cargo was efficiently imported into the nucleus in the presence of exogenous cytosol (rabbit reticulocyte lysate; RRL) and an energy-regenerating system, whereas in the absence of RRL and the energy-regenerating system, nuclear import was inhibited as expected (Fig. 1A). Strikingly, when the import assay was performed with purified AcMNPV nucleocapsids, nuclear import of the nucleocapsid occurred under both permissive (RRL with the energy-regenerating system) and non-permissive (without RRL and the energy-regenerating system) nuclear import conditions with comparable nuclear import efficiency (Fig. 1A–C; Movies 1,2).

Some viruses enter the nucleus through nuclear envelope disruptions, instead of the NPC (Cohen et al., 2006; Cohen and Pante, 2005). Thus, to determine whether AcMNPV nucleocapsids enter the nucleus through the NPC or through the nuclear envelope, we blocked transport through the NPCs with wheat germ agglutinin (WGA), which blocks the NPC by binding to nucleoporins (Nups), and found that nucleocapsids remained in the cytoplasm (Fig. 1D). Thus, the mode of nuclear entry of AcMNPV nucleocapsids observed in Fig. 1A was not the result of nuclear envelope disruptions by the virus. This is in agreement with Ohkawa et al. (2010) who found that nuclear import of nucleocapsids was inhibited in cells microinjected with WGA before AcMNPV infection and in cells that have their NPCs blocked with a dominant-negative form of importin-β, which does not release endogenous import cargo from the NPC, and therefore clogs the NPC and inhibits nuclear transport (Gorlich et al., 1996; Kutay et al., 1997).

In contrast to the results with WGA, nuclear import of nucleocapsids was not inhibited by importazole (Fig. 2), which disrupts the interaction of importin-β and Ran-GTP (Soderholm et al., 2011). This result indicates that the nucleocapsid either does not use importin-β or uses importin-β but the binding of Ran-GTP to importin-β is not required for its nuclear import. Consistent with the latter, nuclear import of the nucleocapsid was not inhibited in the presence of GTPγS (Fig. 2), a non-hydrolyzable GTP analog that inhibits Ran-dependent nuclear import (Melchior et al., 1993). Another virus that uses importin-β but has its nuclear import inhibited by GTPγS is hepatitis B virus (Rabe et al., 2003). Collectively, these results indicate that nuclear import of AcMNPV nucleocapsids occurs through the NPC and is independent of Ran.

It is possible that the nucleocapsid has importin-β like properties. However, comparison of all baculovirus proteins from currently sequenced viral genomes did not identify any protein with homology to importin-β (data not shown).

Given that the baculovirus nucleocapsid hijacks the host Arp2/3 complex during infection to promote actin polymerization at one of its ends, creating a ‘comet tail’ that propels the nucleocapsid towards the cell nucleus (Mueller et al., 2014; Ohkawa et al., 2010) and digitonin-permeabilized cells contain G-actin, F-actin and Arp2/3 (Fig. S1A–C), we hypothesized that nuclear import of the nucleocapsid could also be driven by the propulsive force of actin polymerization involving Arp2/3. To test our hypothesis, the import assay was performed with semi-permeabilized cells treated with CK666, an Arp2/3 inhibitor (Nolen et al., 2009). We found that nuclear import of nucleocapsids was significantly reduced in cells incubated with CK666 (Fig. 3). These results indicate that F-actin nucleation promoted by Arp2/3 is involved in the mechanism of nuclear import of the nucleocapsid. Given that the Arp2/3 activity for nucleating G-actin depends on nucleation-promoting factors (Firat-Karalar and Welch, 2011) and the only nucleation-promoting factor present in our experiment is the viral protein VP78/83 on nucleocapsids, we conclude that baculovirus-induced actin polymerization is involved in the mechanism of nuclear import of AcMNPV. In agreement with this conclusion, nuclear import of the nucleocapsid was inhibited when the digitonin-permeabilized cells were treated with cytochalasin D, which disrupts F-actin and inhibits actin polymerization (Fig. S1D).

To further test the involvement of actin polymerization in the nuclear import of nucleocapsids, we assayed for nuclear import of nucleocapsids in purified nuclei using an actin polymerization assay. In this assay, F-actin nucleation occurs in the presence of both an actin nucleation-promoting factor (such as VP78/83) and Arp2/3. Therefore, we incubated purified nuclei with nucleocapsids and performed the actin polymerization assay in the presence and absence of Arp2/3. As shown in Fig. 4B, nucleocapsids were found in the nucleus when the actin polymerization assay was performed in the presence of Arp2/3 under permissive conditions for actin polymerization (presence of G-actin and actin polymerization buffer containing ATP). In contrast, no nucleocapsids were found in the nucleus when the assay was performed in the absence of actin polymerization buffer (Fig. 4B). When Arp2/3 was omitted in the assay very few nucleocapsids were found in the nucleus (Fig. 4B). Moreover, experiments with VP39–3×mCherry nucleocapsids incubated with pyrene–G-actin under conditions that allow for actin polymerization, in which pyrene-conjugated actin becomes fluorescent upon polymerization, and visualized without fixation yielded actin comet tails attached to VP39–3×mCherry nucleocapsids (Fig. 4D).

The involvement of actin polymerization in the nuclear import of AcMNPV is supported by live-cell images of High Five cells transiently expressing EGFP–actin and infected with VP39–3×mCherry virus in which nucleocapsids physically depressing the nuclear envelope are depicted (Ohkawa et al., 2010). Although in infected cells it is not possible to separate the process of nuclear import from the cytoplasmic trafficking of the nucleocapsids, HeLa cells transduced with AcMNPV in the presence of CK666 had a significant decrease in nuclear accumulation of nucleocapsids (Fig. S2A,B). However, when examined by electron microscopy, although some nucleocapsids were found attached to the nuclear envelope (Fig. S2C), they were never found in the nucleus.

Based on our results, we propose that AcMNPV uses the propulsive force of actin polymerization to drive not only the migration of the nucleocapsid through the cytoplasm, but also the nuclear import of the nucleocapsid. In this proposed mechanism the actin comet tail formed at one end of the nucleocapsid pushes the nucleocapsid through the NPC for nuclear import. In agreement with this model, the nucleocapsid crosses the NPC lengthwise with the end containing the actin-assembly-inducing VP78/83 protein away from the NPC (Au and Pante, 2012). Our model implies that the phenylalanine-glycine-rich nucleoporin (FG-Nup) permeability barrier at the NPC central channel can be breached by the propulsive force of actin polymerization.

HeLa cells (American Type Culture Collection) were cultured in complete DMEM medium supplemented with 10% fetal bovine serum (FBS). Spodoptera frugiperda clone 9 (Sf9) cells were grown in Sf-900 II serum-free medium (Invitrogen) supplemented with 2% FBS. Antibodies used were mouse anti-VP39 (1:500, provided by Robert Kotin, National Heart Lung and Blood Institute, NIH), rabbit anti-Nup153 (1:100, Abcam-84872), rabbit anti-Arp2 (1:300, Abcam, ab47654), and mouse anti-fibrillarin (1:1000, Abcam, 38F3) antibodies. Phalloidin–FITC was from Sigma-Aldrich. Secondary antibodies were from Invitrogen. The Arp2/3 inhibitor CK666, its inactive form CK689, and GTPγS were from EMD Millipore. The importin-β inhibitor importazole was from Sigma-Aldrich.

Recombinant AcMNPV was propagated in E. coli (DH10B) and amplified at a multiplicity of infection (MOI) of 1 in Sf9 cells. Virus was maintained by infecting Sf9 cells grown in Sf-900-II serum-free medium at MOI=1 and was harvested 5 days post infection by centrifugation at 1000 g for 10 min at 4°C. The viral titre was determined by TCID₅₀ end-point dilution assay using Sf9 cells (O'Reilly et al., 1992). AcMNPV was purified through a continuous 15–60% (w/v) sucrose gradient centrifugation.

AcMNPV VP39–3×mCherry virus expressing both wild-type VP39 and VP39 fused to three copies of mCherry was generated as described previously (Biswas et al., 2016). The VP39–mCherry fusion construct was derived from the WOBCAT vector (Ohkawa et al., 2010).

AcMNPV nucleocapsids and VP39–3×mCherry nucleocapsids were obtained by de-enveloping the virus by treatment with 1% NP40 for 1 h at 30°C as described previously (Au et al., 2010; Au and Pante, 2012).

BSA covalently attached to the NLS of SV40T antigen (CGGGPKKKRKVED) at a ratio of 5:1 (NLS:BSA) was custom made (Sigma-Genosys). NLS-BSA was labelled with Cy3 (Amersham-Biosciences) according to the manufacture's protocol.

Adherent HeLa cells were permeabilized with 20 µg/ml digitonin (Sigma-Aldrich) in transport buffer (20 mM HEPES, pH 7.4, 110 mM potassium acetate, 1 mM EGTA, 5 mM sodium acetate, 2 mM magnesium acetate, and 2 mM dithiothreitol) for 4 min. Permeabilized cells were washed with transport buffer and incubated with transport buffer containing 70-kDa dextran conjugated to Texas Red (Invitrogen), Cy3-labeled cNLS-BSA, or purified nucleocapsids for 60 min at 37°C in the presence or absence of 20% rabbit reticulocyte lysate (RRL; Promega), a energy regenerating system (0.4 mM ATP, 0.45 mM GTP, 4.5 mM phosphocreatine and 18 U/ml phosphocreatine kinase; Sigma-Aldrich), complete-Mini EDTA-free Protease Inhibitor Cocktail (Roche) at 10 μg/ml, and 1.6 mg/ml BSA. Next, cells were washed with transport buffer three times and prepared for immunofluorescence microscopy.

For inhibitor treatment, permeabilized cells were pre-treated with 0.5 mg/ml WGA (0.5 mg/ml, 30 min), importazole (40 µM, 1 h), or CK666 or CK689 (1 mM, 1 h). For some treatments, the drug was also present in the import mixture (importazole at 40 µM, and CK666 or CK680 at 1 mM).

Nuclei from HeLa cells were isolated by osmotic swelling under isotonic conditions. Briefly, pelleted cells were resuspended in Earle's balance salt solution (1.8 mM CaCl₂, 5.3 mM KCl, 0.8 mM MgSO₄, 117 mM NaCl, 26 mM NaHCO₃, 1 mM NaH₂PO₄-H₂O, 5.6 mM glucose) and centrifuged for 5 min at 600 g, followed by resuspension in 10 volumes of isotonic buffer (30 mM Tris-HCl pH 8.3, 100 mM NaCl, 5 mM MgCl₂) containing 0.5% (v/v) of NP40, incubation on ice for 4 min with gentle mixing, and centrifugation for 5 min at 600 g. The supernatant was discarded and the resulting pellet containing isolated nuclei was resuspended in isotonic buffer. Isolated nuclei were adhered to glass coverslips by centrifugation for 10 min at 2000 g through a 30% sucrose cushion onto a coverslip.

Coverslips containing isolated nuclei were incubated with AcMNPV nucleocapsids, 40 µM importazole and skeletal muscle G-actin from an actin polymerization kit (Cytoskeleton Inc). Next, actin polymerization buffer (final concentration on the coverslip: 50 mM KCl, 2 mM MgCl₂, 5 mM guanidine carbonate and 1 mM ATP) was added and the actin polymerization assay was performed according to the manufacturer's protocol. After 60 min, cells were washed with PBS and prepared for immunofluorescence microscopy. For experiments in the presence of Arp2/3 complex, 50 nM of this complex was added to the coverslip prior to adding the actin polymerization buffer.

mCherry nucleocapsids were mixed with pyrene-conjugated skeletal muscle G-actin from an actin polymerization kit (Cytoskeleton Inc) in a well of an eight-well glass bottom μ-slide (Ibidi). Next, actin polymerization buffer was added and samples were incubated at 37°C in the dark for 30 min. After this incubation time, samples were visualized immediately by confocal microscopy.

HeLa cells or purified nuclei on glass coverslips assayed as indicated above were fixed with 3% paraformaldehyde (PFA) in PBS for 10 min, permeabilized with 0.2% Triton X-100 in PBS for 5 min, blocked with PBS containing 1% BSA and 10% goat serum for 30 min at 37°C, and labelled with a primary antibody against VP39 or Nup153 for 1 h at 37°C. Samples were washed three times for 10 min each with PBS, followed by incubation with fluorescently labeled secondary antibodies (1:1000 dilution) for 45 min at 37°C. Coverslips were washed three times for 10 min each with PBS, and mounted with Prolong Gold antifade reagent containing DAPI (Invitrogen).

All fluorescence microscopy images were acquired using a Fluoview FV1000 confocal laser-scanning microscope (Olympus). Quantification was performed using Prism (GraphPad Prism Software, Inc) and analyzed using an unpaired Student's t-test.

We are grateful to Dr Robert Kotin (NIH) for the VP39 antibody and to Drs Matthew Welch and Taro Ohkawa (University of California, Berkeley) for the VP39-3xmCherry construct.