Carboxypeptidase D (gp180), a Golgi-Resident Protein, Functions in the Attachment and Entry of Avian Hepatitis B Viruses

gp180-expressing plasmids.

The gp180-coding sequence (16), kindly provided by Kazuyuki Kuroki, was subcloned via NcoI into a minimal pUC plasmid containing the cytomegalovirus promoter and the bovine growth hormone poly(A) site to create plasmid pC-gp180. pC-gp180^Tm was derived from a similar construct, pC-myc-gp180, containing additionally a Myc epitope tag inserted into the first KpnI site in the coding region. Mutant pC-gp180^Tm terminates after the predicted transmembrane region at Ile-1334 (...CVCSI.).

Fluorescent conjugates.

Recombinant DHBV pre-S was conjugated to fluorescein by using 5(6)-carboxyfluorescein N-hydroxysuccinimide ester (Boehringer Mannheim). Ten microliters of a freshly prepared stock solution of the dye (10 mg/ml in dimethyl sulfoxide) was added to 0.5 mg of pre-S in 1 ml of buffer C (200 mM Na-borate, pH 8.5). Following incubation for 1 h at room temperature, labeled pre-S was separated from the unreacted dye and transferred to phosphate-buffered saline (PBS) by using a PD-10 column (Pharmacia). DHBV particles were enriched from high-titer duck serum by one-step centrifugation (13). The peak fraction (500 μl), containing approximately 10¹⁴ particles/ml, diluted with 500 μl of 2× buffer C was labeled as described above.

Biochemical binding assay.

All of the His₆-tagged pre-S polypeptides were prepared and handled as reported elsewhere (30). Coupling of pre-S to CH-activated Sepharose (Pharmacia) was carried out according to the manufacturer’s instructions (2 h, room temperature; 100 mM NaP_i–100 mM NaCl [pH 6.5]–4 mg of DHBV or HHBV pre-S per mg of dry resin). Liver extracts were prepared by Dounce homogenization in buffer H (150 mM NaCl, 10 mM Tris-HCl [pH 7.4], 2 g of aprotinin/liter, 2 g of leupeptin/liter, 1 mM phenylmethylsulfonyl fluoride, 1% Triton X-100) (5 ml/g of liver) and removal of the nuclei and debris by centrifugation. The pre-S–Sepharose (500 pmol of pre-S) was incubated for 4 h (4°C) in the absence or presence of uncoupled ligand with 200 μl of extract and 600 μl of buffer H at 4°C. The Sepharose was washed four times with buffer W (buffer H without protease inhibitors). Specifically bound proteins were eluted for 1 h with excess free pre-S (360 nmol in 50 μl; 37°C), loaded on a sodium dodecyl sulfate (SDS) gel, and detected by silver staining; alternatively, the washed Sepharose was resuspended in 50 μl of Laemmli buffer and boiled for 10 min, and gp180 was detected by Western blotting (ECL system; Amersham). The binding of gp180 to 500 pmol of pre-S coupled to Sepharose was determined to be inhibited by 50% with approximately 5 pmol of free pre-S; 250 pmol of the free polypeptides was therefore used for competition experiments (see Fig. 2B).

Infection inhibition.

PDHs were prepared and cultured as previously described (23). Soluble gp180 (sgp180) was purified from culture supernatants of high five insect cells infected with a recombinant baculovirus encoding the extracellular domain of gp180 (amino acids 1 to 1308). Construction of the recombinant virus and purification of sgp180 have been described elsewhere (6, 31). sgp180 and DHBV-containing duck serum (100 DNA-containing particles/cell) were mixed and immediately applied in 500 μl of maintenance medium to 1 well of a 12-well culture dish containing about 8 × 10⁵ cells. After 14 h of incubation, the cells were washed and further cultivated. After 6 days, cells were harvested and intracellular viral DNA was prepared by using the QIAamp blood kit (Qiagen) and subjected to DNA dot blot analysis as previously described (23). The radioactive signals were quantified with a Molecular Dynamics PhosphorImager.

Internalization assay.

HuH7 cells were cultivated in six-well culture dishes or (for confocal microscopy) on coverglass chamber slides (Nunc) and transfected with gp180-expressing plasmids by using a standard Ca phosphate protocol or Superfect (Qiagen). At 1 to 2 days after transfection, the cells were incubated with labeled substrates (50 to 100 μl/ml of culture medium) for 4 h at 37°C, washed with PBS, and fixed with 3% paraformaldehyde for 20 min. The fixed cells were permeabilized with 0.25% Triton X-100 in PBS and immunostained either with a gp180 polyclonal antibody or, in the case of gp180^Tm, with a Myc monoclonal antibody (9E10) and then with a tetramethylrhodamine-conjugated secondary antibody. Fluorescence was analyzed with a Leica DM IRB inverted fluorescence microscope (20×/0.40 objective) equipped with an automatic photo camera and the Leica TCS NT confocal laser scanning microscope (63×/1.2 objective).

gp180 localization.

PDHs cultured on coverglass chamber slides were fixed and reacted with a gp180 antiserum, followed by a secondary fluorescein-conjugated antibody. HuH7 cells transfected with pC-myc-gp180 were fixed at 1 day posttransfection and costained with a polyclonal antiserum against TGN46 and monoclonal antibody 9E10. The secondary antibodies were conjugated with fluorescein and tetramethylrhodamine, respectively. Immunofluorescence was analyzed by confocal microscopy as described above.

gp180 is the dominant pre-S binding protein in duck hepatocytes.

DHBV infection of cultured PDHs can be efficiently inhibited by addition of recombinant DHBV pre-S polypeptide from Escherichia coli (30). Hence, recombinant DHBV pre-S is able to bind to the viral receptor on hepatocytes and to prevent virus binding and subsequent infection. We have used this recombinant protein to assess whether cellular pre-S binding proteins, other than the previously reported gp180, are involved in virus attachment and might be targeted by pre-S during infection inhibition. For this purpose, we immobilized DHBV pre-S covalently to Sepharose (see Materials and Methods) and then incubated this affinity matrix with Triton X-100 extracts of either duck liver or PDHs or with solubilized liver membranes. Bound proteins were eluted with an excess of the free ligand. Although SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and silver staining revealed that many proteins were eluted from the pre-S matrix (Fig. 1A, left lane), gp180 was the only protein consistently present in the eluate, regardless of the source of the input, and whose binding to the affinity matrix could be inhibited by the addition of free competing pre-S (Fig. 1A, right lane) or viral particles (not shown). The identity of gp180 was verified by Western blotting (Fig. 1B). The rate constants observed for the binding to the pre-S–Sepharose confirmed the specificity of this interaction: by measuring the disappearance of gp180 from the supernatant (detected by radioimmunoblotting) and its reappearance in the presence of free ligand, we estimated the on rate of this reaction to be approximately 10⁴ M⁻¹ s⁻¹; an off reaction was not detectable after 24 h at 4°C and therefore was assumed to be slower than 3 × 10⁻⁶ s⁻¹ (data not shown). These preliminary kinetic data are supported by a detailed BIAcore analysis with recombinant gp180 and pre-S polypeptide (31).

The domains within the DHBV L protein determining functional binding to a cellular receptor and physical binding to gp180 overlap extensively.

With a set of terminally and internally deleted pre-S polypeptides, infection inhibition studies mapped the essential region for efficient receptor binding to amino acids 30 to 115 of the viral L protein (30). To confirm that the cellular receptor interacting with this region was gp180, we used the identical deletion mutants (Fig. 2A) to map the domain required for gp180 binding. Mirroring the infection competition mentioned above, we performed a binding competition assay, where an excess of the free pre-S polypeptides was present during binding; uncompeted, matrix-bound gp180 was then detected by Western blotting (Fig. 2B).

With this procedure, mutants Δ22-30, Δ128-139, and 30-115 (Fig. 2B, lanes 2, 9, and 10) efficiently inhibited gp180 binding to the DHBV pre-S matrix, indicating that the gp180 binding domain maps between amino acids 30 and 115. Deletions between amino acids 30 and 115 all affected binding efficiency; some residual gp180 competition competence was observed with pre-S mutants Δ52-61, Δ62-73, and Δ74-84 (lanes 3, 4, and 5), defining an inner domain (amino acids 85 to 115) where deletions completely abolished gp180 interaction (lanes 6, 7, and 8).

These gp180 binding data strikingly correlate with complementary receptor interaction data obtained with the same mutants and synthetic peptides (30). A correlation between functional receptor interaction and gp180 binding was also seen with L protein from grey heron HBV (HHBV), another avian hepadnavirus; an HHBV pre-S polypeptide bound duck gp180 (1, 11) and blocked DHBV infection (30). Since all pre-S–gp180 binding data are in good accordance with corresponding cellular receptor binding data from infection competition experiments (30), we conclude that gp180 is the receptor for DHBV on the liver cell surface.

sgp180 efficiently inhibits DHBV infection.

To further corroborate that gp180 is the cellular DHBV receptor, we examined whether a recombinant gp180 which lacked the transmembrane region and the cytoplasmic domain (sgp180) could block DHBV infection of PDHs, as was shown to be the case for other virus-receptor systems, such as human immunodeficiency virus (HIV) and CD4 (34). sgp180, obtained from insect cells infected with recombinant baculovirus (31), was used in an infection inhibition assay similar to the one described for the recombinant pre-S polypeptides (30). With increasing amounts of sgp180 present during DHBV infection of PDH cultures, inhibition of infection occurred in a concentration-dependent manner (Fig. 3). An almost complete loss of infection (>97% inhibition) was observed at sgp180 concentrations of 7.5 μg/ml; 50% inhibition was achieved at about 0.13 μg/ml, a concentration which corresponds to seven molecules of sgp180 per DHBV particle, assuming that, in the serum used for infection, DHBV virions were accompanied by a 1,000-fold excess of noninfectious but receptor binding-competent SVPs.

gp180 is a Golgi-resident protein.

gp180 was formerly reported to be cell surface protein (15). Readdressing the question of the subcellular localization of gp180, we used confocal microscopy to perform immunofluorescence analysis of endogenous gp180 in cultured PDHs. The cells were fixed at 2 days postplating, immunostained with an anti-gp180 antiserum, and analyzed. By this method, gp180 was detected exclusively in tubular, perinuclear structures intracellularly (Fig. 4A) and was not found to reside at the plasma membrane to a detectable degree.

To examine the identity of the gp180-containing intracellular compartment, we transfected pC-myc-gp180, a plasmid expressing a Myc epitope-tagged gp180 under control of the cytomegalovirus promoter (see Materials and Methods), into HuH7 cells, a human liver-derived cell line. At 1 day posttransfection, the cells were coimunostained with monoclonal antibody 9E10 for gp180 and with an antiserum against TGN46, a Golgi marker protein (21). Analysis by confocal microscopy revealed an accurate colocalization of expressed gp180 and endogenous TGN46 (Fig. 4B), suggesting that gp180 localizes to the Golgi apparatus rather than to the cell surface.

gp180 transfection mediates uptake of DHBV particles in nonpermissive cells.

If gp180 is the cellular DHBV receptor, one would assume that transiently expressed gp180 could function as a receptor for DHBV, mediating virus uptake in nonpermissive cells. To examine this, HuH7 cells (nonpermissive for DHBV infection) were transfected with pC-gp180, incubated with fluorescein-labeled DHBV particles or recombinant DHBV pre-S polypeptide, washed, fixed, counterimmunostained for gp180, and examined for the presence and the localization of fluorescent viral proteins and gp180 (Fig. 5). Expression of gp180 was also verified by Western blotting (not shown).

In these experiments, the fluorescently labeled viral substrates not only bound to cells that expressed the receptor but also were selectively internalized. In Fig. 5A this is shown for viral particles by conventional fluorescence microscopy: the nuclei of total cells are stained with DAPI (4′,6-diamidino-2-phenylindole) and seen as blue fluorescence. From these, only the transfected cells (gp180, stained red) took up the labeled particles (green fluorescence). Confocal microscopy revealed that the internalized viral particles colocalized with gp180 in a vesicular, mostly perinuclear compartment (Fig. 5B). Another fraction of gp180 was localized to tubular structures and was not found to be associated with viral particles. Fluorescently labeled pre-S polypeptide was internalized with a similar efficiency (Fig. 5C).

In addition, we used a construct (pC-gp180^Tm) which expresses a C-terminally truncated form of gp180 (gp180^Tm) lacking the cytosolic domain. In this case, the ligand was found to be arrested on the cell surface along with the exposed mutant receptor, apparently being anchored in the plasma membrane but unable to undergo endocytosis (Fig. 5D). Taken together, these findings demonstrate that gp180 binds the virus at the cell surface and is capable of cointernalizing bound particles. However, gp180-expressing HuH7 cells were not infected, as judged from the results of immunofluorescence analysis, while intracellular core antigen was not detected at day 5 postincubation with DHBV-containing duck serum (not shown).

Virus-gp180 interaction is conserved and restricted to hosts of avian hepadnaviruses.

In the DHBV infection competition assay (30) and the gp180 binding competition assay mentioned above, the avian pre-S proteins from DHBV and HHBV were found to be functionally interchangeable, suggesting that initial gp180 binding may be a common step in the uptake mechanisms of these two avian hepadnaviruses. To substantiate this hypothesis, we investigated whether the gp180 homolog from grey heron liver extract was a potent binding protein for the HHBV L protein. Using an immobilized HHBV pre-S polypeptide and conditions similar to those used in the previous experiment with duck liver extract (Fig. 1A), we found an analogous 180-kDa binding protein (Fig. 6A). In addition, as already observed with duck liver extracts (Fig. 1A), many nonspecifically bound proteins, whose binding was unaffected by an excess of free HHBV pre-S, were eluted from the matrix. The heron gp180 was found to bind DHBV pre-S with an efficiency similar to that of duck gp180 (not shown). In contrast, when we analyzed chicken liver extracts, which contain similar levels of a serologically cross-reacting gp180 homolog, we could not detect, by silver staining, specific binding of this or any other protein to the DHBV pre-S matrix (not shown). A very weak binding of the chicken homolog was detected only by the more sensitive Western blotting technique (Fig. 6B, lower panel). In the absence of a chicken hepadnavirus, these observations suggest that an efficient pre-S–gp180 interaction is a common feature of hepadnavirus infection of avian hosts.