J Biol Chem, Vol. 275, Issue 6, 4298-4304, February 11, 2000

Nuclear Import of Adenovirus DNA in Vitro Involves the Nuclear Protein Import Pathway and hsc70^*

Andrew C. S. Saphire, Tinglu Guan, Eric C. Schirmer, Glen R. Nemerow, and Larry Gerace

From the Departments of Cell and Molecular Biology, Scripps Research Institute, La Jolla, California 92037

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Adenovirus, a respiratory virus with a double-stranded DNA genome, replicates in the nuclei of mammalian cells. We have developed a cytosol-dependent in vitro assay utilizing adenovirus nucleocapsids to examine the requirements for adenovirus docking to the nuclear pore complex and for DNA import into the nucleus. Our assay reveals that adenovirus DNA import is blocked by a competitive excess of classical protein nuclear localization sequences and other inhibitors of nuclear protein import and indicates that this process is dependent on hsc70. Previous work revealed that the hexon (coat) protein of adenovirus is the only major protein on the surface of the adenovirus nucleocapsid that docks at the nuclear pore complex. This, together with our finding that in vitro nuclear import of hexon is inhibited by an excess of classical nuclear localization sequences, suggests a role for the hexon protein in adenovirus DNA import. However, recombinant transport factors that are sufficient for hexon import in permeabilized cells do not support DNA import, indicating that there are other as yet unidentified factors required for this process.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Proteins are imported into the nucleus through aqueous channels that span the nuclear envelope called nuclear pore complexes (NPCs).¹ Whereas ions and molecules less than ~20-40 kDa can diffuse passively through the NPC, larger proteins are transported by saturable pathways that are energy- and signal-dependent. The classical signals that specify nuclear protein import (nuclear localization sequences (NLSs)) are short stretches of amino acids rich in basic residues, although other classes of NLSs have been described recently (1, 2). The transport of NLS-bearing proteins into the nucleus requires the participation of soluble cytosolic factors (3). The advent of a cytosol-dependent in vitro assay for nuclear import (3) has led to the identification of a number of these factors (4), namely importin- (also called the NLS receptor and karyopherin-), importin- (also called p97 and karyopherin-), Ran (TC4), and NTF2 (also called p10 and pp15). An initial step in the import of proteins containing classical NLSs occurs in the cytoplasm, where the substrate binds to the receptor importin-. Subsequently, this complex docks at the cytoplasmic face of the NPC together with importin-. The complex is then translocated through a central gated channel of the NPC into the nuclear interior. Movement of the import complex through the NPC requires Ran and NTF2 (5-8), but the precise role of these components is unclear. An additional protein, hsp/hsc70, has been shown to be required for import of some substrates (e.g. the large T-antigen protein (9)), but not of others (e.g. the glucocorticoid receptor (10)).

Most viruses that replicate in the nucleus must use the NPC to introduce their genome into the nucleoplasm. Although it is known that the gated channel of the NPC can expand up to ~25 nm in diameter to allow signal-mediated transport (11), this is still considerably smaller than the diameters of many viruses known to replicate in the nucleus. For example, the diameters of SV40, adenovirus, and herpesvirus are 45-50, 60-90, and 120-200 nm, respectively. Enveloped viruses such as human immunodeficiency virus and influenza manage to circumvent the physical barrier of the NPC by releasing the genome from their large capsids during entry into the cytoplasm. The genome is subsequently transported through the NPC as a flexible protein-nucleic acid complex that appears to utilize a classical NLS-mediated pathway (12, 13). Most non-enveloped viruses must undergo restructuring for the import of their genomes into the nucleus. Both the DNA and associated capsid proteins of SV40 are transported into the nucleus after infection of cells, but the structural state of the nucleocapsid during import has not yet been characterized (14). However, the DNA viruses such as adenovirus, baculovirus, and hepadnavirus all appear to only partially uncoat at various stages of entry into the host cell (15, 16). In the case of adenovirus, the early events of infection such as cell attachment and internalization have been partially defined (17-20), but little is known about the transport of adenovirus DNA into the nucleus. As adenovirus is a strong candidate for gene therapy applications (21), it is important to obtain a detailed understanding of the interactions between this virus and the host cell.

Adenovirus is a non-enveloped double-stranded DNA virus containing 11-15 different virus-encoded polypeptides and a linear genome of ~36 kilobase pairs (22). The total mass of the nucleocapsid is estimated to be 150 MDa (22). The major capsid protein is the hexon coat protein. The capsid contains 240 hexon trimers, which form an icosahedron with 12 vertices and 20 facets. At the vertices are the fiber proteins, which contain a primary cell-surface attachment site (17). Anchoring each fiber to the capsid is the penton protein, which contains a secondary binding site whose recognition is a precursor to internalization (18, 19). Stabilizing the capsid are several other minor proteins that either associate with hexons alone (proteins VI, VIII, and IX) or strengthen the association between hexon and penton proteins (proteins IIIa and IX). The DNA is condensed with proteins V, VII, and µ and the covalently attached terminal protein, which together form the viral core.

The adenovirus capsid is a very stable structure that protects the viral genome against harsh environmental conditions. The virus enters the cell by receptor-mediated internalization in coated pits, where the combination of the energy of binding and the acidic environment of the early endosome is thought to cause a series of cooperative structural changes in the capsid (20). These structural changes appear to facilitate fusion with the endosomal membrane and penetration into the cytoplasm (23) as well as potentiate the loss of certain capsid proteins. By the time it is delivered to the cytoplasm, the virus loses the fiber penton, peripentonal hexon proteins, and the capsid-stabilizing proteins (proteins IIIa, VIII, and IX) (20). The nucleocapsid resulting from these rearrangements docks with the NPC, where it appears to undergo further disassembly, allowing its DNA to enter the nuclear interior (20). Because cellular nucleoprotein particles (e.g. messenger ribonucleoproteins) can be transported through the NPC by the machinery that mediates protein transport (4), adenovirus similarly might exploit the cellular nuclear protein import machinery for the import of its DNA into the nuclei of infected cells.

Here we use a digitonin-permeabilized cell assay to analyze the interaction of nucleocapsid derived from adenovirus type 2 with the NPC and the transport of viral DNA into the nucleus. Our results indicate that adenovirus DNA import utilizes components of the classical protein import machinery and hsp/hsc70. Furthermore, our findings, in conjunction with previous work, indicate that the major capsid protein (hexon) has an important role in viral DNA import. Since recombinant factors for classical NLS-mediated import are insufficient to support adenovirus DNA import, additional components appear to be involved. We discuss the possibility that these are involved in nucleocapsid uncoating at the NPC.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Nuclear Import Assays-- The bovine serum albumin (BSA)-NLS peptide conjugate used as a nuclear transport substrate and HeLa cytosol were prepared as described previously (3). The nuclear protein import assay was carried out as described (3), with the use of virus in place of the BSA-NLS conjugate at 0.1-10 viruses/pore (assuming 2000 NPCs/nucleus). For competition assays, transport substrate was prepared as described above, omitting the conjugation with fluorescein isothiocyanate, and used at up to 50 µM. The anti-hsp70 antibody SPA810 (Stressgen Biotech Corp., Victoria, British Columbia, Canada) was used at 50-100 µg/ml. Import was reconstituted with recombinant import factors using 200 nM importin-, 200 nM importin-, 1 µM NTF2, and 1 µM Ran with 1% BSA in transport buffer.

Immunofluorescence Detection of Nucleocapsids-- Import assays were performed as described above using pH 6.4 buffer-treated nucleocapsids. The latter were introduced into the import assay at a ratio of five viruses/NPC, or ~10,000 plaque-forming units/cell, assuming 2000 NPCs/cell. Estimates of virus number were taken from our observation that 1 µg of virus equals 10⁹ plaque-forming units. Nucleocapsids treated with pH 6.4 buffer cannot be quantitated via plaque assay since they no longer possess the fiber and penton proteins necessary for attachment and endocytosis. After incubating the permeabilized cells for 30 min with nucleocapsids and washing, cells were fixed for 6 min in 4% formaldehyde in PBS and permeabilized in 1% Triton X-100 in PBS. The nucleocapsid was detected by either monoclonal anti-hexon antibody 2HX (American Type Culture Collection) or polyclonal anti-hexon antibody AB1056 (Chemicon International, Inc., Temecula, CA) diluted in PBS containing 2% BSA. Primary antibodies were incubated on slides for 1 h at room temperature, washed, and incubated with either anti-goat or anti-rabbit rhodamine-labeled secondary antibody in PBS containing 2% BSA for 1 h at room temperature.

Preparation of Adenovirus-- A549 cells (American Type Culture Collection) were grown to 80% confluency in T-150 flasks in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 2 mM glutamine, 100 units/ml penicillin, and 0.1 mg/ml streptomycin at 37 °C and 5% CO₂. Cells were subsequently infected with adenovirus type 2 (American Type Culture Collection) for 16 h. The medium was replaced with a double volume of fresh medium. Between 48 and 60 h post-infection, when cells became rounded and detached but not lysed, the cells were harvested by centrifugation and resuspended in 1 ml of complete medium. To extract virus, cells were freeze-thawed three times by alternately placing tubes in dry ice/ethanol and a 37 °C incubator. Lysates were then centrifuged at 14,000 × g to remove cell debris. The resultant high titer stock was either stored at 80 °C or immediately utilized to purify virus by cesium chloride gradient ultracentrifugation as described previously (21). The virus particle was modified by treatment with pH 6.4 buffer as described by van Oostrum and Burnett (24).

Isolation of Viral DNA for Hybridization Probe-- 60 µg of adenovirus in Tris-buffered saline was pelleted in a Beckman Airfuge. The pellets were carefully washed with Tris-buffered saline and repelleted. To disrupt the capsid coat, a solution of Tris/EDTA (pH 8.0), 0.1% SDS, and 60 µg/ml proteinase K (Calbiochem) was added to each tube and allowed to incubate overnight at 50 °C. The next morning, the samples were boiled for 10 min to inactivate the proteinase K. Viral DNA was precipitated with 0.33 volume of 2 M ammonium acetate (pH 4.0) and 2.5 volumes of ethanol. Pellets were washed and allowed to air-dry.

Purification of Hexon Protein-- Hexon protein was isolated using the protocol of Waris and Halonen (15). HeLa cells were grown and infected as described above. After ~48 h, cells were harvested and washed twice with PBS. Cells were disrupted by freeze-thawing three times in the presence of 1,1,2-trichlorotrifluoroethane (Freon 113, Aldrich). Cell debris was removed by centrifugation at 1000 × g for 30 min. The supernatant was harvested and centrifuged at 100,000 × g for 30 min. The resulting supernatant was passed over a Mono Q column (Amersham Pharmacia Biotech) and eluted with a linear salt gradient of 0.0-0.5 M NaCl in bis-Tris buffer (14). The peak hexon fractions were combined and size-selected over a Superose 12 column in bis-Tris buffer containing 0.15 M NaCl. All samples were monitored by SDS-polyacrylamide gel electrophoresis.

Detection of Viral DNA Using in Situ Hybridization-- Nonradioactive probes were generated using the GENIUS SYSTEM® (Roche Molecular Biochemicals). 1 µg of DNA consisting of either whole viral DNA or the hexon coding region alone was used as a template to generate digoxigenin-incorporated DNA probes ~1 kilobase pair in length via random priming. The hybridization solution consisted of 6× SSC, 45% formamide, 5× Denhardt's solution, 10% dextran sulfate, and 100 µg/ml salmon sperm DNA (33). Upon completion of an import assay, slides were fixed with 4% formaldehyde in PBS for 6 min, washed, and then permeabilized with 1% Triton X-100 in PBS and washed. The slides were immediately dehydrated in cold 100% isopropyl alcohol. In this state, they could be stored overnight at 20 °C if necessary. Slides were then rehydrated stepwise with 70, 50, and 30% EtOH in 6× SSC to a final hydrated solution of 2× SSC. Cells were washed and then immersed in 0.1 N NaOH for 90 s at room temperature. Slides were again washed in 2× SSC for 30 s and then dehydrated stepwise using the solutions above. Slides were air-dried for 5 min. Prehybridization was not routinely performed, as the background staining was low. Probe was added to the hybridization fluid at a final concentration of 10 ng/10 µl, incubated for 10 min at 95 °C, and then plunged into slush-ice to cool. Enough probe solution was added to cover the slide, and then a coverslip was fixed in place. The entire slide was subsequently heated to 95 °C for 10 min and then incubated from 6 h to overnight at 50 °C. After incubation, slides were washed with room temperature 6× SSC, followed by PBS. A 1:10 dilution of fluorescein isothiocyanate-labeled anti-digoxigenin Fab fragments in PBS containing 1 mg/ml BSA was added and incubated at room temperature for 1 h. Slides were washed for 5 min in PBS and examined by fluorescence microscopy. For localization of adenovirus DNA with respect to the nuclear lamina, slides were prepared as described above and incubated with anti-lamin A/C antibody (guinea pig) for 1 h at room temperature. Slides were then washed with PBS, incubated with Texas Red-conjugated donkey anti-guinea pig IgG (Jackson ImmunoResearch Laboratories, Inc.), mounted in Slofade® (Molecular Probes, Inc., Eugene, OR), and examined with a laser scanning confocal microscope (MRC-1024, Bio-Rad).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Association of Adenovirus Nucleocapsid with the Nuclear Envelope in Permeabilized Cells-- We used a permeabilized HeLa cell assay that was previously developed to study nuclear protein import (3) to analyze the nuclear import of adenovirus DNA. The permeabilized cells permit the virus to pass directly into the cytosolic environment without prior exposure to the interior of the endosome. To simulate the exposure of the virus to the acidic environment of the endosome that occurs in vivo during infection, we pretreated the virus used for our in vitro assay with mildly acidic (pH 6.4) buffer (24). Consistent with previous studies (25), this treatment resulted in a near-quantitative removal of the penton base and fiber proteins, as detected by SDS-polyacrylamide gel electrophoresis (data not shown), whereas only a small fraction of hexon protein was lost (24).

The treated nucleocapsids were introduced into the permeabilized cell assay at ratios ranging from 0.1 to 10 nucleocapsids/NPC. Samples were incubated for 30 min at 30 °C in the presence or absence of cytosol, and the location of nucleocapsids was analyzed by immunofluorescence staining with anti-hexon antibodies. In the presence of cytosol, we observed strong accumulation of hexon at the nuclear periphery (Fig. 1A). At low concentrations of nucleocapsid, the staining resembled the punctate pattern obtained by immunofluorescence staining with anti-NPC antibodies (data not shown). At higher concentrations, the staining also localized to the nuclear rim, although some signal was apparent both within the nucleus and in the cytosol as well (Fig. 1A). In the absence of added cytosol, either weak or no perinuclear staining was observed (data not shown). The weak perinuclear staining seen in some experiments may be due to the influence of residual cell cytosol that persists after permeabilization. To examine whether acid pretreatment of the virus was necessary for hexon accumulation at the nuclear periphery, we compared the staining pattern obtained by carrying out the assay with pH 6.4 buffer-treated virus and untreated virus. We detected no significant difference between the two (data not shown). To see if hexon accumulation at the nuclear rim was dependent on basic amino acid-type NLSs, we performed the assay in the presence of a 50-fold excess of BSA-NLS conjugate competitor. In this instance, the amount of anti-hexon fluorescence at the nuclear envelope was greatly diminished (Fig. 1B). Nuclear rim staining was also diminished significantly by the presence of wheat germ agglutinin (WGA) (Fig. 1C), a lectin that binds to O-linked glycosylated proteins of the NPC and that is known to block the active import of nuclear proteins (26, 27). This suggests that an interaction with an O-linked glycoprotein of the NPC is required for nucleocapsid accumulation at the nuclear rim.

View larger version (78K): [in this window] [in a new window]   Fig. 1.   Localization of hexon protein at the nuclear envelope after incubation of adenovirus nucleocapsids with permeabilized cells. Permeabilized HeLa cells were incubated for 30 min with cytosol and adenovirus nucleocapsids under various conditions, washed, and analyzed by immunofluorescence staining with the anti-hexon antibody. Although a strong nuclear rim staining can be seen with cytosol alone (A), the addition of WGA (B) or competing BSA-NLS conjugates (C) strongly diminished hexon accumulation at the nuclear envelope.

The nuclear rim anti-hexon immunofluorescence staining is likely to be derived from intact nucleocapsids. To verify that nucleocapsids were indeed docking with the NPC under the conditions of our assay, samples were processed for thin-section electron microscopy (28) after a 30-min incubation (Fig. 2). Numerous examples were seen where nucleocapsids appeared to be docked with the NPC (Fig. 2). This is consistent with a recent analysis of adenovirus infection in vivo that showed that the nucleocapsid remains intact in the cytoplasm after release from endosomes and apparently does not disassemble until after prolonged contact with the NPC (29). Although other experiments demonstrate that purified hexon trimers were efficiently imported into the nucleus (see below), the majority of anti-hexon immunofluorescence staining observed here was excluded from the nuclear interior, suggesting that hexons are associated with structures that are too large to be imported. Considered together, our immunofluorescence and electron microscopic data indicate that the permeabilized cell assay allows the binding of adenovirus nucleocapsids to the NPC in a manner similar to that observed in vivo after normal viral infection.

View larger version (112K): [in this window] [in a new window]   Fig. 2.   Visualization of nucleocapsid binding to the nuclear envelope in digitonin-permeabilized cells by electron microscopy. Permeabilized cells were incubated with nucleocapsids and cytosol for 30 min, washed and processed for thin-section electron microscopy. Shown is a gallery of views of the nuclear envelope in cross-section. NPCs are denoted by arrows.

Import of Adenovirus DNA in Vitro-- After establishing that nucleocapsids bound to the nuclear envelope under our assay conditions, we next sought to determine if these conditions also supported the import of adenovirus DNA into the nucleus. To detect viral DNA, samples were fixed after various periods of incubation, and slides were processed for in situ hybridization with anti-DNA probes representing either part or all of the adenovirus genome. After 30 min of incubation, a strong accumulation of viral DNA at the nuclear envelope was seen, but little DNA was observed in the nucleus (data not shown). We could not directly compare hexon and DNA localization in the same samples due to the destruction of the hexon protein epitope by the conditions required for in situ hybridization. However, comparison of parallel assays revealed that the majority of viral DNA as well as hexon protein accumulated at the NPC at this time point. In contrast, when we performed the incubation for 45 or 60 min, we found that a significant amount of viral DNA appeared in the nucleus (Fig. 3, A and E). As observed with protein import, DNA import also was blocked by the addition of WGA (Fig. 3B) and by excess BSA-NLS conjugate (Fig. 3C). Furthermore, GTPS, an inhibitor of nuclear protein import that interferes with the activity of Ran (5, 6) and possibly of another GTPase (30), also blocked the import of viral DNA (Fig. 3D). To confirm that the nuclear signal of viral DNA observed by conventional microscopy (see above) corresponds to intranuclear DNA and not simply DNA adsorbed to the nuclear surface, in a separate experiment, slides were processed to detect both viral DNA and nuclear lamins, which line the inner surface of the nuclear envelope. Samples were examined by laser confocal microscopy taking 0.2-µm Z-sections through central regions of the nucleus; a single section is shown in Fig. 3E. Viral DNA was clearly localized in a number of intranuclear zones. In this particular experiment, dense foci of viral DNA could be seen at or near the cytosolic face of the nuclear envelope, presumably representing DNA that had not yet been released from nucleocapsids. The abundance of these foci varied considerably between different experiments. Taken together, these data suggest that the import of adenovirus DNA into the nuclei of permeabilized cells utilizes at least some of the components involved in the nuclear import of NLS-bearing proteins.

View larger version (56K): [in this window] [in a new window]   Fig. 3.   Nuclear localization of adenovirus DNA during nucleocapsid incubation with permeabilized cells is blocked by inhibitors of NLS-mediated nuclear import. Permeabilized cells and nucleocapsids were incubated for 1 h with cytosol alone (A) or with cytosol supplemented with WGA (B), competing BSA-NLS conjugates (C), or GTPS (D). Adenovirus DNA was detected by immunofluorescence microscopy after in situ hybridization with a digoxigenin-labeled DNA probe, and samples were viewed with conventional optics. In a separate experiment, adenovirus DNA was visualized together with lamins using a laser scanning confocal microscope (E).

Import of Purified Hexon Protein into the Nuclei of Digitonin-permeabilized Cells-- The apparent involvement of some components of the nuclear protein import machinery in the transport of adenovirus DNA into the nucleus suggested that a protein of the nucleocapsid could play a role in this process. The fiber, penton, and hexon proteins are all known to enter the nucleus during the assembly of new virions after viral DNA replication. However, because the hexon is the only major protein associated with the nucleocapsid after release from the endosome during infection and is exposed on the nucleocapsid surface, we considered it to be the most likely agent responsible for targeting to the NPC. Despite its known karyophilic properties, no NLS has yet been identified on the hexon. To investigate if the hexon has an intrinsic sequence that can target it to the nucleus in the absence of other viral components, we analyzed the nuclear import of purified fluorescently labeled hexon trimers (320 kDa) in permeabilized cells. The labeled hexon strongly localized to the nuclear interior within 10 min (Fig. 4A). Nuclear import of the purified hexon protein as well as of a control BSA-NLS conjugate was blocked by WGA (Fig. 4B). The nuclear accumulation of both proteins was strongly diminished when the assay was performed at 0 °C or when cytosol was first depleted of ATP by treatment with hexokinase/glucose, conditions that block NLS-mediated import (data not shown). Furthermore, when cytosol was preincubated with GTPS, both hexon and control BSA-SV40 NLS failed to accumulate in the nucleus (Fig. 4C). Under these conditions, the hexon protein appeared to become bound to cytoplasmic components, resulting in a low-moderate level of cytoplasmic staining.

View larger version (87K): [in this window] [in a new window]   Fig. 4.   Purified adenovirus hexon protein is imported into the nuclei of permeabilized cells by an NLS-mediated pathway. Fluorescently labeled hexon trimers were added to permeabilized cells and incubated for 30 min in the presence of cytosol alone (A) or of cytosol supplemented with WGA (B), GTPS (C), or competing BSA-NLS conjugates (D). Samples were visualized by fluorescence (Hexon) and phase-contrast (Phase) microscopy.

To investigate whether the hexon enters the nucleus by a basic amino acid-type NLS pathway, we introduced competing nonfluorescent BSA-NLS conjugates into our import reactions at a range of concentrations known to successfully compete for import of fluorescent BSA-NLS conjugates. The presence of excess competitor strongly reduced the levels of import observed (Fig. 4D). A BSA-peptide conjugate containing the sequence of the SV40 T-antigen NLS synthesized in reverse order, which is known to lack the capacity to direct import (3), competed poorly for import (data not shown), confirming that the competition was specific. These data show that the hexon protein is imported into the nucleus by an NPC-mediated process utilizing a basic amino acid-type NLS. Thus, the nuclear import characteristics of the hexon protein are consistent with a role for this protein in mediating the binding of nucleocapsid to the NPC.

Involvement of hsp/hsc70 and Other Factors in Hexon and DNA Import-- Because hsp/hsc70 has been implicated both in the early events of infection following adenovirus penetration of the endosome (31) and in nuclear import of some proteins (9), we examined its role in hexon and viral DNA import. Cytosol and permeabilized cells were preincubated with an anti-hsp70 antibody known to inhibit SV40 NLS-mediated import, and this system was tested for its capacity to support the import of purified hexon. The import of hexon coat protein into the nucleus was not detectably affected by the presence of anti-hsp70 antibodies (Fig. 5B), which do reduce the import of BSA conjugated with peptides containing the SV40 T-antigen NLS or the c-Myc NLS (data not shown). The specificity of the block was demonstrated by preincubation of recombinant hsc70 with the anti-hsp70 antibody, which alleviated the block in import of SV40 and the c-Myc NLS when added to cytosol during preincubation (data not shown). In contrast to the absence of an effect of the anti-hsp70 antibody on the import of the purified hexon protein, the antibody significantly diminished the import of viral DNA as measured by in situ hybridization (Fig. 5, C and D).

View larger version (68K): [in this window] [in a new window]   Fig. 5.   Antibody against hsc70 blocks the import of adenovirus DNA, but not of purified hexon protein. Nuclear import of fluorescent hexon was undiminished by a 1-h preincubation at 0 °C with an anti-hsc70 antibody (B) as compared with a control sample lacking the antibody (A). Antibodies against hsc70 strongly inhibited the accumulation of viral DNA in the nucleus (D) as compared with a control sample lacking the antibody (C).

To more extensively analyze mechanisms for the import of adenovirus DNA, we analyzed the ability of purified cytosolic factors to support this process. Hexon or a control SV40 NLS conjugate was added to permeabilized cells in the presence of transport buffer alone or in the presence of the four cytosolic factors that can reconstitute basic amino acid-type NLS-mediated import (importin-, importin-, NTF2, and Ran). In the absence of these exogenous import factors, there was only a small amount of accumulation of signal in the nuclei of permeabilized HeLa cells (Fig. 6A). This was probably due to residual levels of factors not washed out after permeabilization. However, with the added recombinant cytosolic transport factors, strong accumulation of signal in the nucleus was seen both for the fluorescein isothiocyanate-labeled hexon protein (Fig. 6B) and the BSA-NLS conjugate (data not shown). Thus, as observed previously (32), the four recombinant factors strongly stimulate NLS-dependent nuclear protein import in permeabilized cells.

View larger version (78K): [in this window] [in a new window]   Fig. 6.   Purified import factors are sufficient to support the nuclear import of purified hexon in permeabilized cells. A mixture of recombinant importin-, importin-, Ran, and NTF2 supported the import of BSA conjugated upon a 30-min incubation with purified hexon (B), but transport buffer without added import factors did not (A).

To assess the role of these factors in the import of the viral DNA, we incubated whole nucleocapsids and permeabilized cells either with the four purified transport factors analyzed above or with cytosol and compared the levels of DNA import. No significant DNA import was seen with buffer alone (Fig. 7A) or with the four transport factors (Fig. 7C). In contrast, whole cytosol produced a significant intranuclear signal in the same assay (Fig. 7B). Since hsc70 is required for the import of adenovirus DNA but not of purified hexon, we added it to the purified factors to see if it would restore viral DNA import in permeabilized cells. The addition of hsc70 at levels comparable to those reported to support nuclear import in vivo (9) gave no significant import of viral DNA into the nucleus (Fig. 7D). The activity of the hsc70 used in this experiment was verified in ATPase assays and by its ability to support the release of clathrin from coated vesicles (data not shown). Thus, hsc70, in conjunction with the factors that support NLS-mediated import, is not sufficient to restore the import of viral DNA. This indicates that other cytosolic components are required for adenovirus DNA import.

View larger version (90K): [in this window] [in a new window]   Fig. 7.   Purified protein import factors and hsc70 are not sufficient to support the import of viral DNA. Nucleocapsids were incubated with permeabilized cells for 1 h in the presence of buffer alone (A); cytosol (B); buffer supplemented with a mixture of importin-, importin-, Ran, and NTF2 (C); or buffer supplemented with a mixture of importin-, importin-, Ran, NTF2, and hsc70 (D). Buffer alone (A) and buffer supplemented with the import factors importin-, importin-, Ran, and NTF2 (C) did not support the import of viral DNA. The addition of hsc70 to the factor mixture did not significantly increase the level of viral import observed (D).

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Here we describe an assay for the import of adenovirus DNA into the nuclei of permeabilized HeLa cells. In this assay, a large amount of nucleocapsid becomes associated with the NPC within 30 min, as documented by both immunofluorescence staining and thin-section electron microscopy. A significant amount of viral DNA is subsequently imported into the nucleus by 45-60 min, as monitored by in situ hybridization. We have used this assay to define the requirements for adenovirus DNA import and to begin a mechanistic analysis of this process. As discussed in detail below, our experiments indicate that the docking of adenovirus to the NPC is a prerequisite for DNA import and that docking (and possibly subsequent steps) involves the hexon protein and components of the machinery for nuclear import of proteins containing basic amino acid-type NLSs. Although we could reconstitute adenovirus DNA import in vitro with cytosol and demonstrated a requirement for hsp/hsc70, the four classical nuclear protein import factors together with hsc70 are insufficient to reconstitute viral DNA import in permeabilized cells.

Our results showing that WGA blocks the docking of adenovirus at the NPC and the internalization of viral DNA are consistent with the effects of WGA in vivo (29). Furthermore, our electron microscopic data showing that nucleocapsids dock directly at the NPC prior to the appearance of adenovirus DNA in the nucleus closely resemble electron microscopic observations made for in vivo adenovirus infection (34-36). Thus, our assay reflects some of the basic features of adenovirus DNA import that have been described previously with in vivo studies and is likely to be physiologically relevant.

Nucleocapsid docking at the NPC appears to be a necessary precondition for DNA import since blocking the association of nucleocapsid with the nuclear envelope in vitro either with WGA or with an excess of a protein conjugate containing a classical NLS also inhibits the import of viral DNA into the nucleus. Further support for nucleocapsid docking at the NPC as a requirement for DNA import is the finding that the kinetics of DNA import in vitro (this study) and in vivo (29) are significantly slower than the kinetics of nucleocapsid binding at the NPC. This conclusion is also supported by immunofluorescence localization experiments involving cells infected with adenovirus that suggest that the DNA-associated protein VII (and presumably the associated DNA) is released from the capsid only after the nucleocapsid docks at the nuclear envelope (29).

A variety of data suggest that the hexon coat protein is the principal targeting determinant for adenovirus binding to the nuclear envelope. A priori, the fact that the hexon has karyophilic properties, as shown in our in vitro studies, and is the major protein on the surface of the viral capsid both in our in vitro assay and in vivo (22) implicates it as the principal agent responsible for docking to the NPC. Further evidence for this is the fact that excess basic amino acid-type NLSs competitively inhibit the import of purified hexon protein as well as the docking of the nucleocapsid to the NPC. It is unclear whether the role of the hexon protein is simply to dock the nucleocapsid to the NPC or to engage more distal components of the NLS-mediated import machinery for the import of the adenovirus DNA-protein complex into the nucleus. Other candidate proteins for this second function are proteins VII and V, which are associated with viral DNA in the nucleus and may be co-imported with the latter (29).

Although it is known that hexon trimers efficiently enter the nucleus during viral assembly (36), an NLS on the hexon has not yet been defined. However, scrutiny of the primary structure of the hexon protein reveals several sequences that have basic amino acid-type NLS character. One is at position 279 in the amino acid sequence, and others occur at positions 567, 780, and 807. Although there are fewer basic amino acids in each of these sequences than in the prototypical NLS of the SV40 T-antigen, the amino acid sequence starting at position 279 shares at least a superficial similarity with a number of sequences known to have NLS function (for a comprehensive list, see Ref. 37). However, it must be pointed out that the hexon may not possess a linear, continuous stretch of basic amino acids that serves as an NLS. Instead, some of the sites mentioned above may combine in the folded three-dimensional structure to form an active NLS. Finally, although less likely, the hexon may altogether lack an intrinsic NLS, but could rely on specifically binding to a host cytosolic protein that does possess an NLS.

Previous observations have suggested a role for hsp/hsc70 in the infective pathway of adenovirus. Immunolocalization studies in infected cells have shown that hexon protein and hsp/hsc70 associate after release from the early endosome (31), raising the possibility that hsp/hsc70 may act on the nucleocapsid before docking with the NPC. One possible function of this binding could be to induce conformational changes in the nucleocapsid that expose NLSs. Consistent with this model, previous studies have shown that the hexon protein undergoes conformational changes during uncoating (16). Although purified hexon is imported into the nucleus and must therefore have an easily accessible NLS, this signal may be concealed in the nucleocapsid until a conformational change involving hsp/hsc70 or some other component exposes it. It is tantalizing to speculate that the hexon in the intact adenovirus nucleocapsid has a concealed NLS to protect the latter from the host's immune system since most basic amino acid-type NLSs are charged, hydrophilic, exposed, and thus highly antigenic.

A second possible function for hsc70 could be to disassemble the nucleocapsid at the NPC. Because capsid disassembly must occur before viral DNA can be imported (20), it may be that this step is blocked by the antibodies to hsp70. Docking with components of the NPC may catalyze the conversion of nucleocapsid to a disassembled state, and hsc70 could stabilize the disassembled state, consistent with its role as a chaperone protein (38). Alternatively, hsp/hsc70 could actively unfold the capsid. Finally, hsc70 may act directly on components of the NPC itself, and it is their activity that is essential for the translocation of viral DNA into the nucleus. In yeast, hsp/hsc70 has been suggested to play a role in the binding of the NLS both to its receptor and on the NPC itself during protein import (39). Our data do not allow us to precisely identify the site of hsp/hsc70 activity in the import pathway of adenovirus. It may be that hsp/hsc70 plays a role in several steps in the infective pathway of adenovirus discussed above.

Permeabilized cell nuclear import assays similar to the one we describe for adenovirus have been employed in the study of two lipid enveloped viruses that replicate in the nuclei of infected cells: influenza virus (12) and woodchuck hepadnavirus (40). A simple ribonucleoprotein complex of influenza virus, consisting of nucleoprotein and the RNA genome, is sufficient for the import of the influenza genome in vitro and for the production of infectious virus in vivo (12). Similarly, a complex of hepadnavirus DNA and viral polymerase is sufficient to transport the hepadnavirus genome into the nucleus in vitro (40). A combination of the cytosolic factors that can reconstitute import of proteins with basic amino acid-type NLSs in permeabilized cells is also sufficient to reconstitute import of the influenza RNA in vitro (12).

Our work indicates that adenovirus requires a more complex set of interactions with the host cell than these other viruses. In particular, when the adenovirus nucleocapsid is introduced into a permeabilized cell nuclear transport assay, a combination of the cytosolic factors that support nuclear import of substrates with basic amino acid-type NLSs together with hsc70 is insufficient to reconstitute adenovirus DNA import, although DNA import is supported by complete cytosol. The unknown cytosolic factors required for adenovirus DNA import could be involved in nucleocapsid uncoating at the NPC and/or in translocation through the pore. It is expected that the in vitro assay reported here will permit detailed biochemical analysis of these components, which are expected to be key to adenovirus infectivity.

	ACKNOWLEDGEMENTS

We thank Patricia Mathius, Dr. Dan Von Seggern, and Rod Endo for supplies of virus, help, and stimulating discussions. We thank Dr. Sandra Schmid for the donation of hsc70. In addition, we thank Dr. Frauke Melchior and Dr. Bryce Paschal for helpful discussions.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grants GM41955 (to L. G.) and HL54352 and EI38948 (to G. R. N.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Depts. of Cell and Molecular Biology, Scripps Research Inst., 10550 N. Torrey Pines Rd., IMM 10, La Jolla, CA 92037. Tel.: 619-784-8514; Fax: 619-784-9132; E-mail: lgerace@scripps.edu.

	ABBREVIATIONS

The abbreviations used are: NPCs, nuclear pore complexes; NLS, nuclear localization sequence; BSA, bovine serum albumin; PBS, phosphate-buffered saline; bis-Tris, 2-[bis(2-hydroxyethyl)amino]-2-(hydroxymethyl)propane-1,3-diol; WGA, wheat germ agglutinin; GTPS, guanosine 5'-O-(3-thiotriphosphate).

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES