Translocation and accumulation of exogeneous hepatitis B virus preS surface proteins in the cell nucleus

Most viral proteins play multiple roles in the viral life cycle, thereby maximizing the functionality of each individual protein. The large hepatitis B surface protein (LHBs) of the hepatitis B virus (HBV) is a typical case of this (for review, see Ganem, 1996). The LHBs is transcribed from the env gene using the first of three overlapping reading frames and is composed of the preS1, preS2 and S domains (Fig. 1). Two membrane topological isomers have been described for this envelope protein, which differ in their display, either to the external or internal side of the viral envelope, of the N-terminal preS domain (Bruss et al., 1994; Ostapchuck et al., 1994; Prange and Streeck, 1995). LHBs with preS domains exposed to the external side has been implicated in binding to the cellular virus receptor. In particular, the preS1 region has been reported to serve as the principal binding site for hepatocytes (Neurath et al., 1986; Pontisso et al., 1989; Theilmann and Goeser, 1991) and the preS2 region has been independently reported to contain an auxiliary binding site using polymerized human serum albumin as an intermediate receptor (Machida et al., 1983; Sobotta et al., 2000). However, preS domains displayed to the inner side of the virus membrane are known to be involved in viral morphogenesis, establishing a physical interaction of the viral envelope with preformed cytosolic nucleocapsids (Bruss, 1997; Poisson et al., 1997).

Independent of the membrane topology, other crucial roles have also been assigned to the LHBs that involve various steps in the viral life cycle, including the regulation of viral replication (Summers et al., 1990; Summers et al., 1991) and the transactivation of a variety of promoter elements (Kekule et al., 1990; Hildt et al., 1996; Kim et al., 1997). Although it is true that such functions are performed by the LHBs, on closer inspection, one can easily see that these multiple roles are mostly, if not solely, performed by the preS region of the envelope protein. This observation naturally leads to the question whether preS protein fragments alone might also have some biological roles in the absence of the S region. Actually, in vitro studies with protease-treated HBV particles showed specific cleavage of preS fragments from LHBs, which indicate the presence of protease-sensitive sites within the junction of the preS and S proteins (Gerlich et al., 1993; Lu et al., 1996). So, while the proteolytic cleavage and generation of free preS fragments are quite evident, their biological meanings remain obscure.

With respect to the rapid degradation of exogenous proteins in the cytosol, a putative role for free preS proteins in the viral infection process can be assumed to be largely restricted to the early steps of infection. There are many examples that viral structural proteins have been shown to be associated with various subcellular structures, including actin filaments or the Golgi complex and even the nucleus. Depending on their intracellular localization, further studies have shown that these proteins are involved either in exerting some essential functions in the internalization of the virus (Freed, 1998) or in the transport of viral particles along the cytoskeletal fibers (Kasamatsu et al., 1983; Elliott and O’Hare, 1997), besides performing their primary functions as structural components. Accordingly, we have investigated the putative role of free preS proteins in the viral life cycle. To achieve this, fluorescently labeled preS(1-174) proteins of the LHBs were exogenously introduced into the detergent-permeabilized cells and the intracellular distribution and the interaction of preS proteins with cellular components were analyzed by confocal laser-scanning microscopy. Unexpectedly, we found that the preS proteins were specifically translocated into the nucleus in an energy-and cytosolic factor-independent manner. We will discuss the mechanism of this nuclear accumulation and its biological meaning in the viral life cycle.

The expression vector for the recombinant production of HBV-preS proteins was constructed by subcloning the coding sequence for the whole preS region of HBV (subtype adr) in fusion to the maltose-binding protein (MBP), using the pMAL-c2 vector (New England Biolabs, Beverly, MA). The insert was prepared by PCR amplification from the plasmid pHBV (Choi et al., 1986) using forward (5′-GGAATTCATGGGAGGTTTGTCTTCCAAA-3′) and backward primers (5′-TGCACTGCAGTTAGTTCGGCGGTGCAGGGTC-3′).

As these primers contain EcoRI and PstI recognition sites on their respective 5′ ends, the amplified 522 bp PCR fragment was subcloned into the EcoRI and PstI sites of pMAL-c2. The resulting expression vector, pMAL-preS, was then introduced into Escherichia coli (strain BL21) cells. For the expression of MBP-preS fusion protein, overnight culture of transformed cells was used to inoculate fresh media and this culture was induced at its log growth phase with 1 mM isopropyl-thio-β-D-galactopyranoside (IPTG). After incubation for further 2 hours, cells were harvested and the pellet was resuspended in ice-cold TBS (10 mM Tris-HCl, pH 7.4, 150 mM NaCl) containing 5 mM EDTA and 2 mM phenylmethylsulfonylfluoride (PMSF). All subsequent procedures were performed at 4°C. Cells were destroyed by sonication and the debris removed by centrifugation at 10,000 rpm for 20 minutes. The supernatant was then treated with 50% ammonium sulfate and the precipitate was resuspended in distilled water and mixed with the same volume of 0.1 M sodium citrate (pH 4.6). Denatured proteins were then removed by centrifugation and, to concentrate the protein solution, the supernatant was precipitated once again with 50% ammonium sulfate. The final pellet was resuspended and dialyzed against column buffer (20 mM Tris-HCl, pH 8.5, 25 mM NaCl, 10% glycerol, 1 mM EDTA). The solution was passed through a QAE-Sepharose (Amersham Pharmacia Biotech, Uppsala, Sweden) column and bound proteins were eluted in 25-300 mM NaCl gradient. The successful purification was confirmed by SDS-PAGE (see Fig. 2A). To confirm that the preS region was intact, western blot analysis was performed with preS epitope-specific antibodies (see Fig. 2B): mAb F35.25, which is specific for the preS1(21-47) region (Petit et al., 1991) and mAb H8, which specifically recognizes a conformational epitope covering the preS2(120-145) region (Chung and Kim, 1987). MBP was identified using the anti-MBP mAb (Fig. 2B), HAM-19 (Park et al., 1998) and these were detected by peroxidase-conjugated anti-mouse Ig antibodies (Sigma, St Louis, MO) and the corresponding substrate.

For fluorescein-isothiocyanate (FITC) labeling, MBP or MBP-preS protein solutions were dialyzed against conjugation buffer (0.1 M sodium carbonate, pH 9.0) and concentrated to 2 mg/ml. Each 10 μl of freshly prepared FITC solution (10 mg/ml in DMSO) was added to 1 ml of each protein solution and incubated in the dark for 2 hours at room temperature. The reaction was stopped by addition of 1/10 volume of 0.1 M glycine (pH 8.0). The reaction mixture was fractionated using a 1.5×8 cm Sephadex G50 column (Amersham Pharmacia Biotech) gel filtration chromatography.

Peptides were synthesized by the solid phase method (Merrifield, 1986) using Fmoc-chemistry. Fmoc-Wang-amino acids resins (Novabiochem, San Diego, CA) were used as support. For the peptide chain elongation, dicyclohexylcarbodiimide (DCC) and N-hydroxybenzotriazole (HOBt) were used as coupling agent. The side chains of the amino acids were protected with the following base (piperidine)-stable protecting groups: Asp (OtBu), Asn (Trt), Trp (Boc), Arg (Pmc), Ser (tBu) and Tyr (tBu). The final protected peptide-resins were cleaved and deprotected with TFA-based reagents (82.5% TFA, 5% phenol, 3% H₂O, 5% thioanisole, 2.5% 1,2-ethandithiol and 2% triisopropylsialine) for 2 hours, then precipitated with diethylether and dried in vacuum. The crude peptides were purified by a preparative reverse-phase (RP)-HPLC on a Waters 15 μm Deltapak C₁₈ column (300 Å, 1.9×30 cm). Purity of the HPLC-isolated peptides was checked by an analytical RP-HPLC on an Ultrasphere C₁₈ column (5 μm, 0.46×25 cm, Beckman, San Ramon, CA).

Cell lines were obtained from the ATCC (Rockville, MD). The human hepatoma cell line HepG2 and the human epitheloid carcinoma cell line HeLa were cultured in Dulbecco’s modified Eagle’s media (DMEM; GibcoBRL, Grand Island, NY) supplemented with 10% fetal bovine serum. Cells were cultured at 37°C in a 5% CO₂ atmosphere in a humidified incubator.

The day before analysis, HeLa or HepG2 cells were detached from stationary cultures and plated on 18×18 mm glass coverslips in six-well plates (NUNC, Roskilde, Denmark) to 50% confluency. The next day, adherent cells were washed with ice-cold PBS and then fixed with 2% paraformaldehyde in PBS for 20 minutes at 4°C. All subsequent procedures were performed at 4°C. The fixation reaction was blocked by incubation of the cells in 0.1 M glycine in PBS for 5 minutes. Afterwards, the plasma membrane was permeabilized with 0.1% Triton X-100 in PBS for 5 minutes and the detergent was removed by extensive washing with PBS. Permeabilized cells were then incubated with FITC-labeled MBP-preS or MBP to a concentration of 40 μg/ml for 1 hour. To determine nuclear integrity, permeabilized cells were incubated with anti-DNA antibodies (Roche Molecular Biochemicals, Mannheim, Germany) in 0.2% (w/v) bovine serum albumin (BSA) in PBS for 15 minutes at room temperature, rinsed and then incubated with FITC-labeled secondary antibodies. As there were no signals detectable in the nucleus, it was concluded that the nuclear membrane had been remained intact during the staining procedure (data not shown). For the competition assay, permeabilized cells were first incubated with competitor peptides, the preS(120-174) or the HIV gp41(584-618) peptide (RILAVERYLKDQQLLGIWGCSGKLICTT-AVPWNAS) at a concentration of 4 mg/ml in PBS for 30 minutes. Without washing, cells were incubated directly with FITC-labeled MBP-preS or MBP at a concentration of 40 μg/ml for 60 minutes. Finally, cells were washed with ice-cold PBS and analyzed by confocal laser-scanning microscopy.

The selective permeabilization of the plasma membrane and subsequent nuclear transport assays were performed by the method as described by Adam et al. (Adam et al., 1990). In brief, one day before the transport assay, HepG2 or HeLa cells were plated on 18×18 mm glass coverslips in six-well plates. On the day of analysis, the coverslips with the attached cells were washed with ice-cold import buffer (20 mM Hepes, pH 7.3, 110 mM potassium acetate, 5 mM sodium acetate, 2 mM magnesium acetate, 0.5 mM EGTA) and then immersed in ice-cold import buffer containing 40 μg/ml digitonin for the plasma membrane-specific permeabilization. After five minutes, the digitonin-containing buffer was removed by aspiration and replaced with ice-cold import buffer. Coverslips were drained and blotted to remove excess buffer and then placed in the inverted orientation over 40 μl of a complete import mixture. The complete import mixture contained the following components: 50% cytosol, 20 mM Hepes, pH 7.3, 110 mM potassium acetate, 5 mM sodium acetate, 2 mM magnesium acetate, 2 mM DTT, 0.5 mM EGTA, 1 mM magnesium ATP, 5 mM creatine phosphate (Roche Molecular Biochemicals), 20 U/ml creatine phosphokinase (Roche Molecular Biochemicals), substrates for the nuclear transport assay, and 1 μg/ml each of aprotinin, leupeptin, and pepstatin. As a source of cytosol fraction, rabbit reticulocyte lysate (Promega, Madison, WI) was used. The entire chamber was incubated at 37°C for 30 minutes. At the end of the assay, each coverslip was washed with import buffer and fixed by immersion in 4% (w/v) paraformaldehyde dissolved in import buffer prior to image analysis. As a positive control in the nuclear transport assay, BSA-conjugated with a classical nuclear localization signal (NLS) sequence was used (Adam et al., 1990). In brief, the synthetic peptide, CGGGPKKRKVED, which contains a well-known nuclear transport signal sequence of SV40 T-antigen (PKKKRKVE) was synthesized and crosslinked with FITC-labeled BSA via N-terminal cysteine residues. The successfully crosslinked conjugate was fractionated using 1.5×8 cm Sephadex G50 column (Amersham Pharmacia Biotech) gel filtration chromatography.

Cells treated with FITC-labeled proteins were analyzed by confocal laser-scanning microscopy. The confocal microscope system consisted of a Leica TCS 4D connected to a Leica DAS upright microscope (Leica Lasertech GmbH, Heidelberg, Germany). Fluorescent intensity of each cellular fraction (nucleus and cytosol) was quantitated using NIH Image 1.60 public domain software after the substraction of the fluorescence intensity of the background. Relative fluorescence intensity was determined using the fluorescence intensity of the FITC-MBP staining as standard.

Far-UV circular dichroism spectra of preS(120-174) peptides were acquired using a Jasco J720 spectropolarimeter. The peptide concentration was 100 μg/ml in solutions of PBS, pH 7.4, or 50% (v/v) NMR grade trifluoroethanol (TFE; Sigma-Aldrich) or 30 mM sodium dodecylsulfate (SDS) dissolved in 10 mM sodium phosphate buffer, pH 7.4, respectively. All samples were maintained at 25°C during the analysis. The spectra were measured at 0.2 nm intervals from 240 to 195 nm at 25°C using a 1 mm path-length cell. The average of four scans was calculated and used for the graphic analysis. The mean residue ellipticity (θ), is given in deg cm² dmol^-1.

The intracellular distribution of exogenously introduced preS proteins was analyzed for HepG2, human hepatoma cells and HeLa, an HBV non-permissive epithelial-like cell line (Qiao et al., 1994), for which recombinant preS proteins as expressed in C-terminal fusion to the maltose-binding protein (MBP), termed MBP-preS(1-174), was used (Fig. 2). In all of the experiments, MBP without co-expressed preS proteins was used as negative control. For the introduction of free preS proteins into the cytoplasm, these cells were first fixed with 2% paraformaldehyde and then the plasma membranes were permeabilized with 0.1% Triton X-100 and finally incubated with FITC-labeled preS proteins. Interestingly, incubation with FITC-labeled MBP-preS(1-174) fusion proteins resulted in a specific accumulation of these proteins in the nucleus of HepG2 and HeLa cells, as shown in Fig. 3A and Fig. 3B respectively. On the contrary, FITC-labeled MBP proteins were distributed only to a background level throughout the cytoplasm (Fig. 3C,D) but not in the nucleus. To test the accessibility of exogeneous molecules to nuclear components of Triton X-100 permeabilized cells, these cells were incubated with anti-DNA antibodies and their intracellular distribution was examined. As no signals were detectable in the nucleus (data not shown), it was concluded that the nuclear envelope was neither solubilized nor disrupted when using this concentration of Triton X-100. Therefore the possibility of nuclear accumulation of preS proteins, induced by nonspecific diffusion caused by nuclear envelope damage or passive leakage, could be excluded.

Another important observation is that there were no differences in the intracellular distribution of MBP-preS(1-174) between HepG2 and HeLa cells. HepG2 cells are known to be HBV permissive and previous studies have shown that they are able to bind HBV particles via the viral preS1 domain (Neurath et al., 1986; Pontisso et al., 1989; Theilmann and Goeser, 1991; Qiao et al., 1994; Lee et al., 1996) as well as to produce functional viral particles after transfection of the viral genome. However, the HeLa cell line is an HBV non-permissive epithelial-like cell line, generally used as a negative control in HBV infection (Choi et al., 1996). So it seems that the penetration and accumulation effects of the preS(1-174) protein on the nucleus are independent of HBV susceptibility.

The nuclear envelope is composed of two concentric membranes, the inner and outer nuclear membrane, and is perforated by nuclear pores. To accumulate within an intact nucleus, the preS proteins must translocate itself through this nuclear envelope. In live cells, molecular traffic between the cytoplasm and nucleus occurs through nuclear pores. These nuclear pores contain central hydrophilic channels, but most proteins are too large to cross these channels, and even when they are small enough to diffuse through the nuclear pore, specific mechanisms involving nuclear localization signals and energy supply are required to cross the pore complex (Mattaj and Engelmeier, 1998). In the case of fixed and permeabilized cells, as used in the present experiments (Fig. 3), cytoplasmic factors are thoroughly washed out and no energy source is available, which is not compatible with the conditions in living cells. To examine whether the nuclear accumulation effect of preS protein is also reproducible in intact cells, in vitro nuclear transport assays (Adam et al., 1990) (using digitonin-permeabilized HepG2 cells) were performed. To compare the present finding with the characteristics of conventional nuclear transport mechanisms, SV40 T-antigen nuclear localization signal (NLS)-conjugated BSA, a representative nuclear transport substrate, was used as positive control.

As shown in Fig. 4A, the SV40 T-antigen NLS conjugates were able to enter the nucleus at 37ºC when cytosol and energy source were supplied. But, when temperature was lowered to 4ºC (Fig. 4B), or when the cytosolic factors were depleted (Fig. 4C), nuclear accumulation of these substrates was inhibited.

However, preS proteins were still able to accumulate within the nucleus even under energy-and cytosolic factor-depleted situations (Fig. 4E,F). Nevertheless, when the temperature was shifted to 37ºC, the nuclear accumulation was more enhanced (Fig. 4D), perhaps because of the increase of the molecular motions induced by the increase of temperature. The same results were obtained for HeLa cells (data not shown). From these results, it has been concluded that the nuclear accumulation effect of preS proteins is independent of cytosolic factors in intact cells, and that this process would not follow conventional cytosolic factor-and energy-dependent nuclear translocation mechanisms.

preS-mediated nuclear accumulation has been not reported before. To explore the biological meaning and the mechanism behind this finding, it is necessary to define the effective domain more precisely. A clue for the determination of the active domain was found by analysis of the physical characteristics of the preS protein. After FITC-conjugation, the labeled preS proteins were stored in PBS at 4ºC. During the long-term storage, MBP-preS proteins were slowly degraded, which was detected by a mobility change in an SDS-PAGE (Fig. 5A). Interestingly, it was then observed that these truncated MBP-preS proteins had lost their ability to accumulate within the nucleus (Fig. 5B). Judging from their molecular weight and their reactivity with preS region-specific antibodies, the degradation of preS proteins was located around the C-terminal region. This C-terminal truncation was about the size of 10 kDa, which would cover the entire preS2 region (preS(120-174); 55 amino acids, corresponding to 6 kDa) and a part of the preS1 region (about 4 kDa). When these degraded proteins were analyzed for their reactivity to the preS region-specific antibodies, the reactivity to the mAb H8, which specifically detects the preS(120-145) region, was lost, but not the reactivity to F35.25, an anti-preS(21-47)-specific mAb. These results clearly show that the C-terminal region covering the preS2 domain (data not shown) was deleted by degradation. Accordingly, we conclude that the effective domain for nuclear accumulation has to be present within this region.

For explaining the nuclear translocation and accumulation of MBP-preS proteins, some kinds of interactions between nuclear components and preS proteins have been assumed. As shown above, the nuclear envelope of the detergent (Triton X-100 or digitonin)-permeabilized cells remained intact. Therefore, the preS protein itself must actively cross the nuclear envelope. This might be achieved either by using pre-formed channels on the nuclear envelope, i.e. the nuclear pore complex, or by direct pore formation in the nuclear envelope. In the former case, preS fragments might either interact with conventional transport mediators (importin alpha, beta) or otherwise directly bind to the nuclear pore complex. However, the experimental results of Triton X-100 or digitonin-permeabilized cells (Figs 3, 4) exclude the possibility of a conventional NLS-mediated transport. If there might be direct interactions between nuclear pore components and the preS protein, the number of binding sites should be restricted, and the binding should also be saturable and competitive. To examine this possibility, excess amounts of preS peptides were added to inhibit the accumulation of FITC-labeled MBP-preS protein in a competitive manner. As it was evident that the nuclear accumulation effect is dependent on the entire preS2 region (Fig. 5B), synthetic preS(120-174) peptides were used as competitors. Surprisingly, pre-incubation with excessive amounts of preS2 peptides (about 1000 molar ratio in excess) did not inhibit but rather enhanced (about 1.5-fold) the accumulation of the MBP-preS protein within the nucleus (Fig. 6B). Furthermore MBP alone, which itself has no nuclear accumulation effect, was now translocated into the nucleus, though not to the extent of MBP-preS proteins (Fig. 6E). To exclude the possibility of a nonspecific perturbation of the nuclear membrane caused by high concentration of synthetic peptides, the HIV gp41 (584-618) peptide was used in the same concentration in an independent experiment as negative control. In this case, no nuclear accumulation was observed, as has it been the case with preS peptides (Fig. 6C,F). The nuclear translocation mechanism of preS proteins could therefore be explained by the second possibility, which suggest a direct pore-forming activity of preS proteins in the nuclear envelope by the modulation of the nuclear envelope itself, even transiently, which results in the nuclear translocation and accumulation of these proteins. This mechanism also could explain the energy-and cytosolic factor-independency of the accumulation of preS proteins within the nucleus.

Based on the present results, it is possible that the nuclear translocation of the preS protein is caused by a membrane perturbation effect. If there were any interaction between the preS protein and the lipid bilayer of the nuclear envelope during the nuclear translocation, some changes might occur in the conformation of the preS protein or in the topological arrangement of the lipid bilayer. To study the conformational behaviors of the preS2 peptide upon interaction with the nuclear envelope, its secondary structure was analyzed by CD (circular dichroism) under different membrane-mimicking conditions. Micelles of sodium dodecyl sulfate served as an experimental model of the heterogeneous amphipatic environment of a membrane lipid-aqueous interface, whereby TFE (trifluoroethanol) solution was used to mimic the homogeneous hydrophilic face of the membrane (Du et al., 1998). According to the CD analysis, the whole preS2 peptide in the aqueous solution was unstructured, without any typical patterns of a α-helix or β-sheet structure in PBS (Fig. 7). But in an environment that mimics the lipid bilayer (30 mM SDS or 50% TFE), preS2 peptides exhibited a typical α-helical pattern, with the minimum at about 208 nm and the shoulder at 222 nm. The calculated α-helical contents were 19.8% in 30 mM SDS and 20.2% in 50% TFE. Although these values indicate relatively low α-helical contents, it is evident that the hydrophobic environments induce conformational changes of preS2 peptides, which would provide a further clue about the membrane-penetrating properties of preS2 peptide on the lipid bilayer.

The present study reports a novel characteristic of the HBV preS envelope protein as a membrane-penetrating polypeptide. The membrane-permeabilization effect of preS proteins was shown by the nuclear translocation of exogenously added recombinant (Figs 3, 4) or synthetic (Fig.6) preS peptides and further analysis of the active domain determined the preS2 region as the translocation motif (Fig. 5). The mechanism by which preS proteins can translocate themselves or fused proteins through the nuclear membrane is not known. However, mediation over classical NLSs can be largely ruled out as no cluster of basic amino acids or other defined NLS motifs were found within the preS region (Fig. 1). It seems rather that the membrane-penetration effect depends on the conformational configuration of the preS2 peptide, which might depend on the presence of α-helical regions (as has been shown in this study for the preS region (Fig. 7)) as well as has been reported for other membrane-penetrating peptides (Derossi et al., 1996). Another remarkable fact in this process is the absolute independence of preS proteins on energy sources (Fig. 4), similar in its property to the HSV VP22 or the third helix of the antennapedia homeodomain, which have been previously been described to migrate through cells in an energy-independent manner.

In general, the translocation of proteins across biological membranes is highly selective and restricted and several cellular factors are involved. However, there is also a growing family of proteins and peptides that act via a protein translocation mechanism, known as “nonselective membrane translocation”, in which proteins cross biological membranes (plasma membranes as well as nuclear envelopes) by themselves without any mediators or energy supply. The antennapedia homeodomain (AntpHD; Derossi et al., 1996), the HIV-Tat transcription factor (Fawell et al., 1994; Vives et al., 1997), the herpes virus VP22 protein (Elliot and O’Hare, 1997) and lactoferrin (He and Furmanski, 1995) are some examples of these. These proteins are produced by neighboring cells and can diffuse through the plasma membrane and the nuclear envelope of the target cells. This phenomenon is known to be temperature independent and to be not saturable by increasing concentration of the respective proteins. Furthermore, the reverse sequence of these proteins also showed translocation activities, indicating that these effects are not dependent on the specific binding of chiral protein factors (Derossi et al., 1996) but rather that they are induced by a nonselective membrane permeabilization.

The nuclear translocation properties of preS are identical to the characteristics of such membrane-permeable proteins in terms of their energy independency (Figs 3, 4), mediator-independent membrane translocation (Figs 3, 4), non-saturable translocation effect (Fig. 5) and translocation of preS peptide-conjugates (Fig. 3). Therefore, preS might also be regarded as a new member of energy-independent membrane-penetrating proteins and peptides. In this regard, it is expected that the membrane-penetrating pathways of other peptides would be similar to the mechanism of preS-mediated nuclear translocation. Considering the mechanism of other membrane-penetrating proteins, the most promising possibility seems to be a preS protein-lipid interaction, which might cause membrane pertubation that would lead to increased permissibility of biological membranes. Such a hypothesis has been tested out, for example, in the case of the AntpHD, where the cellular entry of various mutants was analyzed (Berlose et al., 1996). The results from this study proposed that the interactions between positively charged amino acid residues, negatively charged phospholipids and α-helical conformations in a membrane environment made these peptides prone to interactions with membrane lipids, which resulted in transmembrane conformations. The already known membrane-penetrating proteins, however, do not share sequence homology with each other or with the preS protein. Nevertheless, the amino acid residues found within preS consist mostly of those that have the tendency to form an α-helical conformation; indeed, CD analysis (Fig. 7) showed that preS peptides in membrane-mimetic environments adopt α-helical conformations, which could act as initiator for protein-lipid interactions. The importance of the preS α-helicity was recently confirmed in another study (Oess and Hildt, 2000), where the energy-independent membrane-penetration effect of an preS2(161-172) fragment was reported. Oess and Hildt made a similar observation to that described in the present study, which compromised the plasma membrane translocating effect of the preS2(161-172) region. However, neither the biological meaning of this effect nor its role in the viral life cycle could be clarified, which raised questions about the role of the membrane translocating ability of viral structural proteins in general.

However, for the HSV-1 VP22 protein, the HIV-Tat protein and other membrane-permeable proteins, their property of energy-independent membrane translocation has been already discussed and described in context of some possible biological functions. These include activities such as those undertaken by microtubule-associated proteins during infection (Elliott and O’Hare, 1998) and by gene regulatory molecules (Rappaport et al., 1999). Accordingly, it is possible that the unusual property as exerted by free preS proteins might also have a role in vivo. The biological roles for preS fragments can be proposed in two aspects. The first hypothetical role is based on the nuclear membrane-permeabilization effect of the preS proteins, while the second relies more on the observation of nuclear accumulation of the translocated preS/preS2 peptides. Regarding the first possibility, a functional role for preS peptides as mediators for transporting the viral genome into the nucleus can be proposed. For viruses of the hepadna family, including HBV, the delivery of the virus genome into the host cell nucleus is a prerequisite for further progress in the virus life cycle (Summers and Mason, 1982). However, the exact mechanism for uncoating and nuclear delivery of HBV is poorly characterized. Based on cases of other viruses, some hypotheses have been suggested. One is that the viral genome is imported by direct attachment of viral particles to the nuclear membrane and that it is subsequently released through the channel of the nuclear complex, as it is the case in adenoviruses (Greber et al., 1996). Recent reports on the direct binding of HBV core particles to the nuclear core complex support such view (Kann et al., 1999). Alternatively, it is also assumed that the virus capsid is disassembled in the cytosol and that the genome is translocated into the nucleus with the help of some viral proteins, such as the covalently associated HBV DNA polymerase (Foster et al., 1991; Kann et al., 1997) or core proteins (Yeh et al., 1990). Although, so far, none of these hypotheses has been definitively proved, relying on the results from the present study, a further scheme can be suggested where the preS protein acts as a putative mediator in targeting the core particle or the viral DNA into the nucleus.

This idea is not so far-fetched when regarding the reports that state the nuclear membrane is impermeable into both directions to core particles (Guidotti et al., 1994), which makes it necessary to disassemble the viral capsid in the cytosol for delivering the virus genome into the nucleus. In this case, the preS proteins would either directly carry the HBV-disassembled complex into the nucleus or facilitate the nuclear translocation of the HBV genome by permeabilization of the nuclear membrane (as has been described in this study; Fig. 4). Naturally, such hypothetical views need more supporting data, and further studies with HBV core particles or viral genome and free preS peptides are necessary to prove the validity of this hypothesis.

Regarding the second possibility on the in vivo function of preS-mediated nuclear translocation, another interesting aspect of free preS fragments must be considered. A large body of evidences has confirmed that the truncated form of MHBs (Kekule et al., 1990) and preS/S2 fragments (Hildt et al., 1995), as well as preS1 proteins (Kim et al., 1997), are novel transcriptional transactivators. The mode of transcativating promoters of cellular genes, including c-myc and c-fos, by these viral transactivators has been extensively studied (Lauer et al., 1994; Hildt and Hofschneider, 1998). It was shown that they may act in the cytoplasm by the activation of the protein kinase C (PKC). However, the PKC signal triggers the c-raf-1/MAP2-kinase signal transduction pathway, which finally results in the activation of transcription factors such as AP-1 and NF-κB (Hildt and Hofschneider, 1998). While such studies were exclusively based on the assumption of a cytosolic orientation of preS/S2 domains, the current finding opens a new possibility for preS proteins in exerting transactivator functions by acting directly in the nucleus.

Regarding this observation, it is assumable that truncated preS fragments would freely enter the nucleus and might interact with nuclear components (including other transactivation factors), resulting in the accumulation and perhaps direct transactivation of cellular genes.

While the novel functions for the HBV preS region, as suggested in the present study, might fill in the missing links in the HBV infection pathway, it is unlikely that such preS-mediated nuclear translocation or a possible nuclear transactivator function would be the sole mechanism for the HBV replication after the HBV infection. Le Seyec et al., for example, showed that the entire preS2 region was dispensable for a successful HBV infection (Le Seyec et al., 1998). The preS(3-77) region, however, was shown in the same study to be essential for infection, which is possibly due to the presence of a binding region for the cellular HBV receptors. As other reports have also observed that the preS2 region is not necessarily a major component in the HBV entry (Fernholz et al., 1993; Sureau et al., 1994) and infection pathway, the preS2-mediated membrane translocation might be not necessarily represent a major step for HBV infection or replication. Rather, it could describe just one of many redundant pathways that HBV can follow in the infection of the host cells.

In conclusion, the unveiling of the exact biological role of free preS proteins, either in serving as a mediator for nuclear translocation or as a nuclear transactivator, will greatly broaden our current understanding about the early events in HBV infection.

The authors thank Dr M.K. Lee for synthesis of the preS(1-174) peptide. This study was supported in part by a grant (BG630M) of the Ministry of Health and Welfare and a grant (HS2470, NL1010) from the Ministry of Science and Technology, Korea.