Cloning and expression of a novel hepatitis B virus-binding protein from HepG2 cells.

A direct involvement of the hepatitis B virus (HBV) preS1-(21-47) sequence in virus attachment to cell membrane receptor(s) and the presence on the plasma membranes of HepG2 cells of protein(s) with receptor activity for HBV have been suggested by many previous experiments. In this study, by using a tetravalent derivative of the preS1-(21-47) sequence, we have isolated by affinity chromatography from detergent-solubilized HepG2 plasma membranes a 44-kDa protein (HBV-binding protein; HBV-BP), which was found to closely correspond to the human squamous cell carcinoma antigen 1 (SCCA1), a member of the ovalbumin family of serine protease inhibitors. Comparison of SCCA1 sequence with the sequence of the corresponding HBV-BP cDNA, cloned by polymerase chain reaction starting from RNA poly(A)(+) fractions extracted from HepG2 cells, indicated the presence of only four nucleotide substitutions in the coding region, leading to three amino acid changes. Intact recombinant HBV-BP lacked inhibitory activity for serine proteases such as alpha-chymotrypsin and trypsin but inhibited with high potency cysteine proteases such as papain and cathepsin L. Direct binding experiments confirmed the interaction of recombinant HBV-BP with the HBV preS1 domain. HepG2 cells overexpressing HBV-BP after transfection of corresponding cDNA showed a virus binding capacity increased by 2 orders of magnitude compared with untransfected cells, while Chinese hamster ovary cells, which normally do not bind to HBV, acquired susceptibility to HBV binding after transfection. Native HBV particle entry was enhanced in transfected cells. Both recombinant HBV-BP and antibodies to recombinant HBV-BP blocked virus binding and internalization in transfected cells as well as in primary human hepatocytes in a dose-dependent manner. Our findings suggest that this protein plays a major role in HBV infection.

The hepatitis B virus (HBV), 1 a member of the hepadnavi-ruses family, is able to infect only humans or higher primates and has a strong organ tropism for hepatocytes (1). The mechanism of hepatocyte infection induced by this noncytolytic virus is not yet clear.
The infectious unit of HBV is an enveloped virion of 42-nm diameter that contains an icosahedral nucleocapsid that encompasses a circular, partly double-stranded DNA molecule with a single strand region of variable length, a DNA-linked protein with functions critical for packaging and DNA replication, including priming, RNA-and DNA-dependent DNA polymerase, and RNAH activities. The HBV envelope that determines the targeting to the host cells and the early entry steps is composed of three proteins anchored in a lipid bilayer, occurring in mature virions in both glycosylated and nonglycosylated forms. The three proteins are called small (S), middle (M), and large (L) surface HBV antigens (HBsAg). They are the translational products of three overlapping open reading frames that start from different initiation codons localized at the 5Ј end of the preS1, preS2, and S regions of the env gene (2). The role of the preS1 region of the L protein appears to be important in cell attachment and consequently in viral infectivity, since preS1 synthetic peptides and corresponding antibodies inhibit virus binding to HepG2 cells (3,4).
A cellular receptor is required for HBV binding and penetration in liver cells. Studies in rat hepatoma cells transfected with the HBV genome (5) and transgenic mice with the genome integrated (6) indicate that HBV is able to replicate in rodent cells once it bypasses the attachment and entry steps of infection, so that the absence of specific cellular receptor(s) constitutes the barrier of HBV infection and replication in nonhuman hepatocytes. A number of putative human cellular receptors for HBV have been proposed, including the receptor for immunoglobulin A (IgA), which shows a partial sequence similarity between the Fc region of the IgA ␣ chain and the preS1 domain of the virus (7,8), and the receptor for interleukin-6, due to the presence of a recognition site for the preS1 domain on interleukin-6 (9, 10). In addition, HBV particles bind isolated asialoglycoprotein receptors through the HBV preS1 domain (11), whereas the preS2 domain seems involved in the interaction with the transferrin receptor (12). It is also known that the preS region of the duck HBV, another member of the hepad-* This research was performed under a contract of the Italian Ministero dell'Università della Ricerca Scientifica e Tecnologica assigned to TECNOGEN s.c.p.a. within the "Programma Nazionale di Ricerca sui Sistemi Neurobiologici-Tecnologie della Trasduzione del Segnale." The costs of publication of this article were defrayed in part by the payment of page charges. This 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. Tel.: 39-0823-612-214; Fax: 39-0823-612-230; E-mail: fassina@tecnogen.it. 1 The abbreviations used are: HBV, hepatitis B virus; HBV-BP, HBVbinding protein; rHBV-BP, recombinant HBV-BP; HBs, surface HBV; HBsAg, surface HBV antigen(s); PM, plasma membrane; RIA, radioimmunoassay; BSA, bovine serum albumin; PBS, phosphate-buffered saline; MES, 2-(N-morpholino)ethanesulfonic acid; CHO, Chinese hamster ovary; FITC, fluorescein isothiocyanate; FACS, fluorescenceactivated cell sorting; MFI, mean log fluorescence intensity; PCR, polymerase chain reaction; SCCA, squamous cell carcinoma antigen; EMEM, minimum essential medium with Earles' salts; HBeAg, hepatitis B e antigen. naviruses family, binds to a 170-kilodalton cellular protein having significant similarities to human and animal carboxypeptidases H, M, and N (13). Other proteins have been identified that bind to different domains of the HBV envelope proteins with a proposed role as mediators in HBV attachment to the hepatocyte. Apolipoprotein H (14) binds to the S domain, whereas polymerized human serum albumin (15) and a human soluble serum factor (16) bind to the preS2 domain of HBsAg. Annexin V, previously known as endonexin II (17,18), binds with low affinity to HBsAg, but interaction involves HBsAg lipids and not HBV proteins (19). More recently, an 80-kDa protein that binds to the preS1 domain of HBV has been isolated but not identified. In addition, this protein has been detected in rat hepatocytes, which are not susceptible to HBV infection (20). These results are conflicting, and no conclusive and convincing biological data have been obtained to fully demonstrate the functional role of these proteins in HBV binding and internalization. None of these proteins has been obtained in a recombinant form to show inhibition of HBV binding to and internalization on primary human hepatocytes.
HBV receptor identification is hampered mainly by the virus-restricted host range and tissue tropism of the virus and the lack of a reproducible in vitro infection systems. Primary human hepatocytes are reported to be susceptible to HBV infection but are difficult to study because they are short lived and require primary explanted liver, and reproducibility of infection requires treatment of the cells with chemical agents that may induce unnatural mechanisms of viral entry (21). However, human hepatoma HepG2 cells, grown under conditions designed to maintain the hepatocyte-specific differentiated function, are thought to express the HBV receptor at maximum levels (22)(23)(24). HepG2 cells have characteristics similar to normal liver parenchymal cells (22,23) and can support HBV replication following transfection by cloned HBV DNA (24). HBV particles bind to HepG2 cells in a saturable manner but are refractive to virus replication, probably due to the inability of the incoming virus to reach the nucleus in order to gain access to the cellular transcriptional machinery (25). The binding of viral particles to HepG2 may be prevented by HBV preS1-(21-47) peptide or anti-preS1-  antibodies (3,4,8), suggesting that this sequence could be directly responsible for HBV attachment to the cellular surface (26).
To identify the protein responsible for HBV attachment to HepG2 cells, in this study we have developed a tetrameric form of the preS1-  peptide in order to increase its functional affinity. This peptide has been used to determine optimal conditions for the detergent solubilization of HepG2 plasma membranes (PM) and as a ligand to fractionate solubilized HepG2 PM extracts by affinity chromatography. By this procedure, a 44-kDa protein has been purified, which has been then cloned and expressed in Escherichia coli to characterize its functional activity. Its role as a cellular receptor for HBV has been studied in transfected CHO and HepG2 cells with the corresponding cDNA. We have analyzed the effect of overexpression of this protein in mammalian cells on virus particle binding and the inhibitory effects of the recombinant protein and corresponding polyclonal antibodies on HBV binding and internalization in transfected cells and primary human hepatocytes.

EXPERIMENTAL PROCEDURES
Peptide Synthesis-Peptide synthesis was performed following the Fmoc (N-(9-fluorenyl)methoxycarbonyl)/dicyclohexylcarbodiimide/Nhydroxybenzotriazole methodology on a fully automated 431A peptide synthesizer (Applied Biosystems). The 4-preS1-  peptide was prepared in a tetrameric form starting from a polylysine core on a 4-(hydroxymethyl)phenoxyacetic resin. After completion of the synthesis cycles, peptide resins were dried overnight under vacuum; peptides were cleaved from resins using trifluoroacetic acid/phenol/thioanisole/ H 2 O/triisopropylsilane (83:6:5:4:2; 5 ml/200 mg of resin), incubating at room temperature for 3.0 h. After synthesis and deblocking from the resins, peptide was purified by extensive dialysis against acetic acid using a membrane with a cut-off of 12,000 Da. The peptides were characterized by analytical HPLC, amino acid analysis, and time-offlight matrix-assisted laser desorption-ionization mass spectrometry. Synthetic myristyl-preS1 (adw2 subtype) was prepared as previously described (27,28).
Isolation of HepG2 Plasma Membranes-HepG2 PM were prepared by the method of Hubbard (29) with few modifications. The protein concentration of PM preparations was determined using the Bio-Rad protein assay, and bovine serum albumin was used as standard protein. PM were characterized by their enzymatic activity by measuring 5Јnucleotidase activity, an enzyme marker for the plasma membranes (Sigma Kit 265-UV, following the manufacturer's instructions), and glucose-6-phosphatase activity (30), an enzyme marker of endoplasmic reticulum.
100 l of HepG2 PM preparations (2 mg/ml) were incubated with detergents such as Tween 20, Triton X-100, or SDS at concentrations ranging from 0.1 to 1% (v/v) for 1 h at 4 or 25°C. After incubation, the samples were diluted with PBS (150 mM NaCl, 50 mM NaH 2 PO 4 , pH 7.5) and centrifuged for 30 min at 18,000 rpm, and the soluble fraction was recovered. To test the effect of acidic conditions, a 100-l sample of Triton-solubilized HepG2 PM fraction (1.8 mg/ml) was incubated with 900 l of acetic acid 0.1 M or glycine 0.1 M for 1 h at 25°C. Samples were then neutralized with NaOH and tested for functionality by RIAs with 125 I-4-preS1-(21-47) (see below). RIAs-HBV synthetic antigens preS1-(21-47) and 4-preS1-  were iodinated with Bolton and Hunter reagent, N-succinimidyl 3-(4hydroxy, 5-[ 125 I]iodophenyl)propionate (Amersham Pharmacia Biotech). 5-20 g of polypeptides were mixed with 1 mCi of reagent in 0.1 M sodium borate, pH 8.5, and incubated under gentle agitation overnight at room temperature. The reaction was stopped by the addition of glycine 0.2 M, and nonincorporated reagent was removed by ultrafiltration on Microcon R300 (Amicon); the 125 I incorporation was checked by trichloroacetic acid precipitation. The specific activity of the labeled polypeptides was 8 ϫ 10 6 cpm/g for 4-preS1-(21-47) peptide and 1.5 ϫ 10 6 cpm/g for preS1-  peptide. Polystyrene microtiter plates were coated with HepG2 or control PM in PBS overnight at room temperature. The plates were washed with PBS containing 0.05% Tween 20 (PBS-T) and then blocked with a 10% solution of bovine serum albumin (BSA) in PBS for 6 h at room temperature. Plates were washed again with PBS-T, labeled polypeptides were added, and binding reactions were performed overnight at room temperature. The unbound polypeptides were removed by washing with PBS-T, and the radioactivity associated with the wells was measured (␥-counter Isomedic 10/600 ICN).
Purification of HBV-BP from Detergent-solubilized HepG2 Plasma Membranes-Affinity column with immobilized 4-preS1-(21-47) peptide was prepared by incubating 10 mg of peptide ligand dissolved in 10 ml of 0.1 M sodium bicarbonate buffer, pH 8.5, with 1.0 g of CH-Sepharose 4B (Amersham Pharmacia Biotech) under gentle agitation for 24 h at room temperature. The extent of ligand incorporation was monitored by reverse phase HPLC and was close to 95%. HepG2 PM have been solubilized with 0.1% Triton X-100 for 1 h at 4°C, and soluble fraction was recovered by centrifugation at 20,000 ϫ g for 30 min at 4°C. The soluble fraction was diluted in loading buffer (50 mM sodium phosphate, pH 7.0) to reach a final concentration of 500 g/ml before loading on the 4-preS1-(21-47)-Sepharose affinity column, equilibrated at a flow rate of 1 ml/min with loading buffer. After elution of unretained material, the eluent was changed to 0.1 M glycine, pH 2.2, to elute bound proteins. The extent of purification was followed by RIAs and SDS-polyacrylamide gel electrophoresis. The bound fractions were dialyzed against 50 mM sodium phosphate, pH 8.0, concentrated on Microcon R300 (Amicon), and then digested with trypsin (1:500 (w/w) enzyme/protein ratio) for 1 h at room temperature. The extent of trypsin degradation was followed by reverse phase HPLC analysis of the reaction mixture using a linear acetonitrile gradient (0.1% trifluoroacetic acid) ranging from 5 to 60% in 50 min on a RP-8 column (Aquapore) and monitored at 225 nm. Peaks eluted were collected, lyophilized, and submitted to automated sequencing (model 477A connected to a PTHamino acid analyzer model 120A; Applied Biosystems).
Cloning and Sequencing of HBV-BP-Total RNA from HepG2 was extracted by cell lysis with guanidinium isothiocyanate, and mRNA fraction was purified using the oligotex kit (Qiagen). First strand of cDNA was synthesized (Superscript II kit; Life Technologies, Inc.) and used to isolate HBV-BP cDNA by nested polymerase chain reactions (PCR) (all of the oligonucleotide primers were purchased from Life Technologies). PCR-1 was performed with the first primer pair (primer-1 (sense) (5Ј-CACAGGAGTTCCAGATCACATCGAG) and primer-2 (antisense) (5Ј-CTGGAAGAAAAAGTACATTTATATGTGGGC), complementary to Ϫ31/Ϫ7 and ϩ1355/ϩ1384 regions of squamous cell carcinoma antigen 1 (SCCA1) mRNA, respectively) and Taq polymerase (PerkinElmer Life Sciences). Purified PCR-1 products were used as template for a more internal pair of primers in PCR-2 (primer-3 (sense) (5Ј-GTTCACCATGAATTCACTCAGTGAAGCC); primer-4 (antisense) (5Ј-GCAATCAGTTTACCAGAACATCTGCAG) complementary to Ϫ7/ϩ21 and ϩ1212/ϩ1238 regions of SCCA1 mRNA, respectively) performed with Pwo DNA polymerase (Roche Molecular Biochemicals) to obtain blunt end amplified products that were used directly in ligation with pUC18 vector digested with SmaI restriction endonuclease. XL1-blue-MRFЈ (Stratagene) CaCl 2 -competent cells have been transformed with ligated DNA and plated on LB plates containing ampicillin, 5-bromo-4-chloro-3-indolyl-␤-D-galactopyranoside, and isopropyl-␤-Dthiogalactopyranoside for selection of recombinant white colonies. The sequence of the insert of the constructed plasmid pUC18/HBV-BP was determined using the AmpliTaq FS Prism ready reaction cycle sequencing kit (PerkinElmer Life Sciences) electrophoresed and assembled on a 373 DNA sequencer (PerkinElmer Life Sciences).
Expression and Purification of Recombinant HBV-BP (rHBV-BP)-The HBV-BP cDNA was cloned in the expression vector pET21a (Novagen). The vector was digested with NdeI, which generates a 3Ј protruding extremity, digested with T4 DNA polymerase (Roche Molecular Biochemicals) to obtain blunt ends, and then digested with BamHI endonucleases. To obtain the HBV-BP cDNA with compatible ends for cloning and to place the ATG start codon of HBV-BP cDNA in the right context with respect to the ribosome binding site carried by the vector, the pUC18/HBV-BP plasmid was used as template for a PCR performed with the following primer pair: primer (sense), 5Ј-CATGAATTCACT-CAGTGAAGCCAAC-3Ј (complementary to Ϫ1/ϩ24 HBV-BP cDNA) and the universal sequencing primer M13 as antisense primer to amplify also the vector polylinker sequence that carries a recognition sequence for BamHI endonuclease located immediately at the 3Ј side of the HBV-BP cDNA insert. The reaction was performed with Pwo DNA polymerase to obtain blunt end products. The amplified product was BamHI-digested and ligated with the pET21a vector. The ligation product was then used to transform the E. coli XL1-blue-MRFЈ strain (Stratagene). The sequence of the insert of the constructed plasmid pET21a/ HBV-BP was determined as described above. The constructed expression vector pET21a/HBV-BP was then used to transform the E. coli strain BL21(DE3) (Novagen) suitable for expression.
A single transformed BL21 (DE3) bacterial colony was inoculated in 25 ml of LB broth containing ampicillin (100 g/ml) and grown at 37°C for 20 h to obtain a saturated culture. This culture was added to 1 liter of LB containing ampicillin and incubated at 37°C. At A 600 ϭ 0.6, the expression of foreign protein was induced with 1 mM isopropyl-␤-Dthiogalactopyranoside, and the culture was incubated at 30°C. The maximal level of recombinant protein production was reached after 3 h from the induction.
The bacteria were recovered by centrifugation and washed three times with PBS buffer. The recovered pellet was resuspended in PBS (4 ml/300 mg, wet weight) and sonicated with the Ultrasonic W-380, with the immersion probe microtip 419A, f 4.8 mm (Heat Systems Ultrasonics), with the following parameters: boost, 1Ј; duty cycle, 40%; output power, 4. The sonication was stopped at 95-98% of lysis of the bacteria, as checked with an optical microscope. The crude bacterial extract was centrifuged at 15,000 ϫ g, at 4°C, for 30 min, to separate the soluble from the insoluble fraction. The soluble fraction was dialyzed against 20 mM MES, pH 6 (buffer A), and the protein concentration was determined with the Bio-Rad Protein Assay kit. The dialyzed soluble fraction was diluted in buffer A to reach a final concentration of 500 g/ml. A glass column (25 ϫ 0.5-cm inner diameter) (Omnifit) was filled with anionic exchange Source 30S resin (Amersham Pharmacia Biotech); the prepared column was equilibrated at a flow rate of 3.0 ml/min with buffer A. The diluted soluble fraction was loaded on the column, and following the elution of unretained material, a gradient from 0 to 100% of buffer B (20 mM MES, pH 6.0, 0.5 M NaCl) in 20 min was applied to elute bound proteins. The retained fraction was recovered and dialyzed against PBS, pH 7.5. The purified fraction of rHBV-BP was stored at Ϫ80°C.
Production of Rabbit Anti-HBV-BP Polyclonal Antibodies-Polyclonal antibodies were produced by immunizing rabbits with 350 g of rHBV-BP in complete Freund's adjuvant and boosted every 3 weeks with the same dose of protein in incomplete Freund's adjuvant. Serum was collected, and the IgG fraction was purified by protein A-Sepharose chromatography. The polyclonal antibody anti-HBV-BP was then puri-fied by affinity chromatography on rHBV-BP-Sepharose columns. Preimmune serum, after protein A purification, was used in control experiments.
Binding of Myristyl-PreS1 to rHBV-BP-Polystirene microtiter plates (Nunc) were coated with rHBV-BP at 10 g/ml (100 l/well) in 0.1 M NaHCO 3 , pH 8.8, and incubated overnight at 4°C. The plates were washed three times with PBS buffer, and the wells were saturated with 200 l of 3% dried milk in PBS and incubated for 1 h at 37°C to block uncoated plastic surface. Plates were then washed three times with 0.5% dried milk in PBS buffer (PBS-M); myristyl-preS1 was added at different concentrations (0.2-20 g/ml in PBS-M, 100 l/well) and incubated for 1 h at 37°C. The plates were washed six times, and an anti-preS1 monoclonal antibody (T0606; Sorin Biomedica) was added at 1 g/ml in PBS-M, 100 l/well. After 1 h of incubation at 37°C and subsequent washing, wells were filled with 100 l of horseradish peroxidase-labeled goat anti-mouse (Sigma) solution diluted 1000-fold with PBS-M. The plates were left to stand for 1 h at 37°C, washed six times, and then filled with 120 l of o-phenylenediamine dihydrochloride (Sigma) chromogenic substrate solution freshly prepared according to protocols from the manufacturer. The color was allowed to develop for 20 min, and the A 450 was determined with an enzyme-linked immunosorbant assay plate reader.
Kinetic Analysis-Photometric substrates, N ␣ -benzoyl-L-arginine and N-benzoyl-L-tyrosine, were diluted in PBS buffer (150 mM NaCl, 50 mM NaH 2 PO 4 , pH 7.5) and were used with the enzymes ␣-chymotrypsin and trypsin, respectively. The fluorogenic substrate, used with the enzymes cathepsin L and papain, was diluted in the enzyme reaction buffer containing 50 mM sodium acetate (pH 5.5), 4 mM dithiothreitol, 1 mM EDTA. The fluorogenic substrate was a bisamide derivative of rhodamine 110, bis-(CBZ-L-phenylalanyl-L-arginine amide), dihydrochloride (Rh110). The amount of the monoamide liberated was determined fluorometrically with excitation and emission wavelengths of 499 and 521 nm, respectively, using a Shimadzu spectrofluorimetric detector, model RF-551. Protease inhibition by rHBV-BP was determined by use of a continuous assay procedure. Proteases were mixed with different concentrations of rHBV-BP, and the samples were incubated for 30 min at 25°C in the appropriate reaction buffer, as described. The reaction was started by the addition of protease-specific substrate. The amount of carboxylic acid liberated from the hydrolysis of the photometric substrates was continuously recorded at a fixed wavelength (253 nm for N ␣ -benzoyl-L-arginine and 256 nm for N-benzoyl-L-tyrosine) using a Beckman spectrophotometer, model DU 640.
Mammalian Cell Transfection and Selection of Stable Cellular Lines-The HBV-BP cDNA was cloned in the pcDNA3 (Invitrogen) mammalian expression vector that allows the selection of G-418 stable clones for the presence of the neomycin gene. The complete insert of pUC18/HBV-BP plasmid was recovered by digestion with KpnI and XbaI endonucleases and ligated in the pcDNA3 vector restricted in the same manner. E. coli XL1-blue-MRFЈ (Stratagene) CaCl 2 -competent cells were transformed with the product of ligation reaction, and the DNA of the resulting recombinant plasmid (pcDNA3/HBV-BP) was prepared (Qiagen Maxi-prep kit). HepG2 and CHO transfections were performed with pcDNA3/HBV-BP DNA using the calcium phosphate precipitation technique. Two stable clones, HepG2/HBV-BP and CHO/ HBV-BP, were selected for the maximal amount of the HBV-BP expression as established by Western blot and RIAs.
Immunofluorescence Analysis-Cells were fixed with 3.7% formaldehyde in PBS, washed with 0.1 M glycine in PBS, blocked with 5% normal goat serum in PBS, and incubated with polyclonal anti-HBV-BP antibodies at 5 g/ml in PBS containing 1% BSA for 30 min at room temperature. After washing with PBS, cells were incubated with the supernatant of hybridoma producing the monoclonal antibody W6/32 (ATCC number HB-95) diluted 1:50 in PBS plus 1% BSA. After 30 min at room temperature, cells were washed again with PBS, and immunoreactivity was visualized by incubation with a mixture of secondary antibodies composed of goat anti-rabbit, conjugated with FITC and goat anti-mouse conjugated with tetramethylrhodamine isothiocyanate (Sigma), both diluted 1:100 in PBS plus 1% BSA for 30 min at room temperature. Images were acquired for the different fluorescent emissions and overlapped.
HBV Particle Binding to HepG2/HBV-BP Cell Surface-Infectious viral particles were obtained under informed consent from serum of a chronically infected patient with HBsAg-and HBeAg-positive serology, in which HBV DNA levels ranged from 3.4 to 1.8 ng/ml (Digene Hybrid Capture HBV DNA Assay) in several occasions. Enriched viral particle preparations were obtained by discontinuous sucrose gradient ultracentrifugation, and preS1 protein content was assessed using a previously standardized enzyme-linked immunosorbent assay (7). All viral particle preparations showed a strong preS1 reactivity, with OD values at 492 Ͼ3. Samples containing 10 6 HepG2 nontransfected cells or HepG2/ HBV-BP cells were incubated with enriched HBV particle preparations (5 ϫ 10 6 genome equivalents/sample) for 90 min at 4°C, washed with phosphate-buffered saline containing 10% fetal calf serum (PBS plus 10% FCS), and incubated with rabbit anti-HBs antiserum (Behringwerke, Marburg, Germany) for 20 min. After an additional wash with PBS plus 10% FCS, FITC-conjugated goat anti-rabbit antiserum (Dako) was added at a 1:10 concentration for 20 min at 4°C. Similar conditions were used in inhibition experiments where HepG2/HBV-BP cells were simultaneously incubated with HBV particles in the presence of 18 pg of rHBV-BP. Cells were successively washed before the analysis. As control for the fluorescence-activated cell sorter (FACS) procedure cells were directly incubated with anti-HBs antiserum, washed, and then incubated with the FITC-conjugated anti-rabbit antibody. Cells were then scored using a FACScan analyzer (Becton Dickinson), and data were processed using CELLQuest software program (Becton Dickinson). Mean log fluorescence intensity (MFI) values were obtained by subtracting the MFI of the control from the MFI of the samples. To evaluate whether the differences between the different samples were statistically significant, the Kolmogorov-Smirnov test for analysis of histograms was used.
HBV Particle Adsorption and Entry in Different Cell Lines-The following cell lines were used to study the interaction with HBV particles: HepG2, HepG2/HBV-BP, CHO, CHO/HBV-BP, primary human hepatocytes, and RAT1. Stable cell lines grown at semiconfluence in 75-cm 2 flasks were incubated with EMEM plus 10% FBS containing 23 ϫ 10 7 viral genome equivalents. After a 30-min incubation at 37°C, medium was collected, and cells were washed three times with PBS containing 0.05% Tween 20 and three times with PBS alone before cellular DNA extraction and HBV DNA determination. In inhibition experiments, cells were incubated with serial dilutions of anti-HBV-BP antiserum or with preimmune rabbit serum, as control, for 30 min at 37°C and extensively washed with EMEM plus 10% FBS before incubation with the HBV preparation. HBV DNA was evaluated in cellular DNA extracts by molecular hybridization and by PCR analysis (31). Viral DNA was detected by a solution hybridization assay (Digene Hybrid Capture HBV DNA Assay) with a linear range of detection between 10 6 and 10 8 genome equivalents/ml. Molecular amplification of HBV DNA was carried out using primers derived from the core gene of HBV (primer sequence: C1, 5Ј-TTG CCT TCT GAC TTC TTT CC-3Ј (core positions 1955-1974), and C2, 5Ј-TCT GCG AGG CGA GGG AGT TCT-3Ј (core positions 2401-2381)) with a roughly linear detection range from 10 3 to 10 7 genomes/ml (data not shown). In similar experiments, cells were incubated with EMEM plus 10% FBS containing 100 l of hepatitis C virus (HCV)-positive serum with a known amount of RNA molecules (6 ϫ 10 6 genome equivalents/ml). After extensive washes, cellular RNA was extracted by RNAZOL (Biogenesis), and HCV RNA was measured using a branched DNA assay (Quantiplex 2.0; Bayer Corp.) as previously described (33). All experiments were repeated at least four times. To evaluate the relative amount of HBV particle binding and penetration into cultured cells, the following experiments were carried out in parallel. HepG2/HBV-BP cells grown to semiconfluence in 25-ml flasks were incubated with EMEM plus 10% FBS containing 22 ϫ 10 7 genome equivalents for 2 h at 4°C or at 37°C. Cells were then washed twice with PBS before treatment with 0.5 mg/ml trypsin, 0.5 mM EDTA. Trypsin was removed, and cells were collected after washing with PBS plus 10% FBS, centrifuged at 1000 rpm for 3 min, and transferred to Eppendorf tubes. To extract DNA from cytoplasm and nuclear cellular fractions, the cell pellet was resuspended in lysis buffer (0.14 M NaCl, 1.5 mM MgCl 2 , 10 mM Tris (pH 8), 0.5% Triton X-100) and homogenized with a loose fitting pestle, and after 5 min of settling, the supernatant was stratified over an equal volume of 24% (w/v) sucrose in lysis buffer. After centrifugation at 10,000 ϫ g for 20 min at 4°C, the cytoplasmic fraction (upper fraction) and the nuclear fraction (pellet) were recovered, the pellet was washed with cold PBS, and both cellular fractions were proteinase K-digested before nucleic acid extraction. In inhibition experiments, increasing concentrations of rHBV-BP ranging from 0.18 to 18 g were mixed with 100 l of HBV particle suspensions (11 ϫ 10 7 genome equivalents) and incubated at 37°C for 30 min before the addition to cells grown in 24-well plates. After a 2-h incubation at 37°C, medium was removed, and plates were washed before trypsinization and total cellular DNA extraction. For inhibition experiments of HBV entry in primary human hepatocytes, supplemented WE medium containing 12 ϫ 10 7 genome equivalents preincubated for 30 min at 37°C with increasing concentrations of rHBV-BP (0.18, 18, and 180 g) or anti-HBV-BP antibodies (0.1-1.0 g/ml) or PBS was added to individual wells of 24-well plates where 10 5 human hepatocytes were seeded. Cells were then washed twice with PBS plus 10% FBS before trypsin treatment and nucleic acid extraction.
Isolation of Primary Human Hepatocytes-Primary human hepatocytes were obtained from the explanted liver of a patient suffering from hypercystinuria, an inherited metabolism disorder, at the time of his double renal and liver transplant. Liver histology documented normal features. Approval for experimental use of the liver was obtained from the Medical Ethical Committee. Hepatocytes were isolated by the collagenase perfusion technique (34), and cell viability was Ͼ90%, as determined by the trypan blue exclusion method. Cells were seeded on culture plates precoated with human liver biomatrix (35), in Williams' E medium (Life Technologies) supplemented with 10% fetal calf serum containing 2 mM L-glutamine, 50 mM dexamethasone, 20 milliunits ml Ϫ1 insulin, 100 units ml Ϫ1 penicillin, and 100 g ml Ϫ1 streptomycin.

Isolation of the HBV-BP from HepG2 Plasma Membranes-
Molecules containing multiple repeats of the same ligand have enhanced the functional affinity of these ligands in several systems. The repeating ligands have also improved their function in affinity chromatography applications (36). In an attempt to obtain an efficient ligand to capture the HBV-BP from HepG2 PM, we synthesized a multimeric form of the preS1-(21-47) peptide, called 4-preS1-(21-47) containing four identical preS1-(21-47) chains linked to a central polylysine core (Fig. 1A). The evaluation of 4-preS1-(21-47) binding activity to HepG2 PM was carried out by RIAs. The tetrabranched peptide showed a selective binding to HepG2 PM preparation (Fig. 1B), with negligible binding to membranes from an unrelated cell line (Rat-1). Its functional affinity for HepG2 PM was higher than the monomeric peptide preS1-(21-47) (Fig. 1C). 4-preS1-(21-47)/HepG2 PM interaction was specific, since unlabeled 4-preS1-  or detergent-solubilized PM reduced binding in a dose-dependent manner (Fig. 1D). The RIA was used to evaluate the effect of various detergents, such as Tween 20, Triton X-100, and SDS, at different concentrations, temperatures, and lengths of treatment, on HepG2 PM solubilization. Treatment with 0.1% Triton X-100 for 1 h at 4°C led to the highest recovery of protein content (87.5%) and of binding activity (50%). The HBV-BP protein was isolated by affinity chromatography loading the Triton-solubilized HepG2 PM extract on an affinity column prepared by linking covalently 4-preS1-(21-47) to activated Sepharose. An acidic buffer was used to elute the bound proteins. One major component with an apparent molecular weight of 44 kDa unit of mass was detected in the acid-eluted fraction ( Fig. 2A). Fractions from the affinity purification step were also assayed by RIA, and almost all of the binding activity to 125 I-4-preS1-  was recovered in the bound fraction, while only negligible activity was found in the column flow-through fraction (Fig. 2B).
These results suggested that the affinity-purified 44-kDa protein is responsible for the binding activity of Triton-solubilized HepG2 PM.
The affinity-purified protein was then treated with trypsin in order to generate peptide fragments that were purified by reverse phase HPLC to be subjected to microsequencing. Two fragments were purified in adequate amounts to provide useful sequence data: peptide 1 (NTYK) and peptide 2 (TYLFLQEYL).
Homology search analyses on the Swiss-Prot data bank (BLITZ system, EBI network service) indicated that human SCCA1, a member of the ovalbumin family of serine protease inhibitors (Ov-serpins) (37) contained both sequences. In the SCCA1 protein, the two sequences matching the tryptic fragments were preceded by a lysine, as expected from a trypsin digestion. To confirm the occurrence of SCCA1 or related proteins in hepatic cells, we carried on nested PCR from the first strand of cDNA synthesized by reverse transcriptase reaction performed on the HepG2 mRNA fraction. PCR experiments performed with sense primer 1 and antisense primer 2 that recognized, respectively, the regions Ϫ31/Ϫ7 and ϩ1355/ϩ1384 of SCCA1 mRNA sequence produced on agarose gel only a smear without any discrete DNA band. The purified products of the PCR-1 reaction were then utilized to perform a second PCR made with a more internal primer pair that recognized the regions Ϫ7/ϩ21 (sense primer 3) and ϩ1212/ϩ1238 (antisense primer 4); the agarose gel analysis of PCR-2 showed a single DNA fragment of the expected size (not shown). In order to obtain blunt end products, the PCR-2 was carried out with the Pwo DNA polymerase. The amplified DNA fragment was purified and cloned in the pUC18 vector restricted with SmaI endonuclease. The sequence of cloned cDNA (Fig. 3) showed an open reading frame of 1173 bases encoding for a protein of 390 amino acids with a predicted molecular mass of 44 kDa, containing the sequences of the two tryptic fragments previously determined by protein microsequencing. The nucleotide sequence of HBV-BP cDNA was strictly correlated to the SCCA1 nucleotide sequence found by Suminami (37), with only four nucleotide substitutions in the coding region leading to three amino acid substitutions, Gly 351 to Ala, Ala 357 to Thr, and Ser 389 to Pro. One of these substitutions, Ala 357 to Thr, has been reported previously for the SCCA1 sequence. To confirm that the observed nucleotide differences were not due to incorrect incorporation during amplification cycles, PCR experiments on human genomic DNA with proofreading DNA polymerase Pwo were carried out. Since the nucleotide substitutions determining the three amino acid changes were located in the last exon of the SCCA1 gene, two primers amplifying a DNA fragment encompassing the nucleotide substitutions were used. The obtained amplification products were cloned using the primer sense corresponding to the ϩ770/ϩ798 region and The proteins were separated on polyacrylamide gel (12%) and stained by the Silver Stain method. B, HepG2 PM affinity purification monitored by RIA. Samples corresponding to HepG2 PM, HepG2 PM solubilized with Triton X-100 (loaded), unretained and bound fraction from the affinity purification step, were coated on microtiter plates at the same concentration (10 g/ml); 125 I-4-preS1-(21-47) was added (160,000 cpm/well), and the recovered cpm were determined counting directly the wells in a ␥-counter. the primer antisense corresponding to the ϩ1187/ϩ1212 region of coding sequence. Twelve independent recombinant clones were sequenced, showing that the three sequences corresponding to SCCA1, SCCA2, and HBV-BP were equally represented. Analysis of the deduced amino acid sequence (www.expasy.ch) indicated that the sequence encompassing residues 27-45 could constitute a transmembrane domain.
Expression of Recombinant HBV-BP-To confirm that the cloned cDNA coded for the same protein isolated from HepG2 PM by affinity chromatography using the synthetic 4-preS1-(21-47) HBV antigen and to preliminarily characterize the functional properties of HBV-BP, a recombinant form of HBV-BP was expressed in E. coli by the use of the pET expression vector system and purified by anionic exchange chromatography (Fig. 4A). The purification step was very effective, since rHBV-BP was recovered with a high degree of purity (90 -95%) as showed by SDS-polyacrylamide gel electrophoresis analysis (Fig. 4B). The purity and identity of the protein were also confirmed by reverse phase HPLC and matrix-assisted laser desorption-ionization mass spectrometry analysis (not shown). By following this protocol, up to 20 mg of purified rHBV-BP per liter of bacterial culture were obtained. The purified protein was used then to produce antibodies in rabbits, which were purified first on protein A and then on rHBV-BP affinity columns. As shown in Fig. 4C, the antibody stained equally well both the 44-kDa protein obtained from the 4-preS1-(21-47) affinity purification of HBV-BP from solubi-lized HepG2 PM and the recombinant protein from E. coli, thus confirming the occurrence of HBV-BP in HepG2 detergent solubilized PM. To further investigate the role of HBV-BP in HBV recognition, binding experiments were performed with a synthetic preparation of myristyl-preS1 (adw2 subtype) (27,28). This protein was prepared by stepwise solid phase chemical synthesis and found to interact specifically with HepG2 PM. The interaction between myristyl-preS1 and rHBV-BP was evaluated on microtiter plates by coating rHBV-BP to the plastic surface and detecting the binding of myristyl-preS1 by using a monoclonal antibody specific for preS1. As shown in Fig. 4D, myristyl-preS1 bound to rHBV-BP in a saturable and dose-dependent manner, with EC 50 ϭ 0.5 g/ml. Thus, in addition to preS1-  binding, rHBV-BP did recognize the entire preS1 domain, the proposed virus attachment domain.
HBV-BP Inhibition of Cysteine Proteases-SCCA1 and SCCA2 genes are tandemly arranged on chromosome 18 and encode for proteins that are 92% similar at the amino acid level (38). Protein data base comparison showed that SCCA1 and SCCA2 belong to the superfamily of serine proteases inhibitors  Fig. 2A) and of the recombinant HBV-BP obtained from E. coli. D, binding of myristyl-preS1 to recombinant HBV-BP. Recombinant HBV-BP was coated on microtiter plates at 10 g/ml (100 l/well) in 0.1 M NaHCO 3 , pH 8.8, overnight at 4°C. Synthetic myristyl-preS1 was added at different concentrations (0.2-20 g/ml) and incubated for 1 h at 37°C. Complex formation was detected by adding an anti-preS1 monoclonal antibody (T0606; Sorin Biomedica). Four replicate assays were carried out, and means Ϯ S.D. are shown.
(serpins), which in most cases are tightly binding inhibitors of serine proteinases (38). Despite the high sequence homology, significant differences do exist between the reactive site loop responsible for protease inhibition, especially those residues flanking the putative scissile bonds, suggesting that these serpins may inhibit different types of proteinases. Previous works suggested that SCCA1 inhibits the ability of chymotrypsin to degrade gelatin and ovalbumin (39), but other investigations indicated activity also for cathepsin L (40). Given the high homology at the amino acid level between HBV-BP and SCCA1, enzyme inhibition assays were performed with rHBV-BP in order to evaluate the protein's ability to act as an enzyme inhibitor. As shown in Fig. 5, rHBV-BP did inhibit with high potency the cysteine proteases papain and cathepsin L. On the contrary, no inhibitory effect was seen with ␣-chymotrypsin and trypsin (not shown). These data confirmed recent reinvestigations of SCCA1 activity, where it was found that unlike SCCA2, SCCA1 lacked inhibitory activity against any of the more common types of serine proteinases but was a potent cross-class inhibitor of the archetypal lysosomal cysteine proteinases cathepsins K, L, and S (41).
HBV Binding to HBV-BP-transfected Cells-In order to establish whether HBV-BP was effectively involved in HBV interaction, HepG2 and nonpermissive hamster CHO cells were transfected with HBV-BP cDNA, first to verify that HBV-BP was localized on the cell surface, an essential requirement for virus binding, and second to verify that HBV-BP transfection conferred HBV binding capability. Cells were transfected with the expression vector pcDNA3/HBV-BP, and two stable cell lines were selected for their high levels of HBV-BP expression, denoted HepG2/HBV-BP and CHO/HBV-BP, respectively.
The cellular localization of HBV-BP was evaluated first by flow cytometry experiments, detecting HBV-BP by using anti-HBV-BP polyclonal antibodies. Western blot analysis performed on total protein extracts obtained from HepG2/HBV-BP cell lines indicated that the affinity-purified antibody was highly specific, since only the 44-kDa HBV-BP protein was stained (not shown). As shown in Fig. 6A, HepG2/HBV-BP cells displayed a significantly higher proportion of HBV-BP on the cell surface than HepG2. Transfected HepG2 showed an enhanced ability to bind 125 I-4-preS1-(21-47) compared with nontransfected cells, with a 2-fold increase in the amount of bound ligand (Fig. 6B). Additional evidences of HBV-BP cell surface localization were obtained by co-localization experiments performed on HepG2/HBV-BP cells, by using as surface markers antigens of class I (HLA A, B, and C) of major histocompatibility complex. These proteins were detected by using the monoclonal antibody W6/32 (ATCC number HB-95) on the same sample probed with the polyclonal antibody anti-HBV-BP. As shown in Fig. 7, the two fluorescent signals were overlapping and perfectly corresponding to the cell surface. These results confirmed that in HepG2/HBV-BP cells the protein encoded by HBV-BP cDNA is localized on the cell surface. Experiments carried out on permeabilized cells indicated that the protein is localized also in the cytosol (data not shown), confirming previous findings (42).
The role of HBV-BP in viral particle binding was examined with transfected cells using hepatitis B viral particles obtained from the serum of an HBsAg-and HBeAg-positive patient with ongoing HBV replication. The binding of HBV particles to untransfected HepG2 cells was clearly measurable by flow cytometry experiments, detecting the attachment of viral particles to the cells by using anti-HBs polyclonal antibodies (Fig.  8A, top). As expected, transfected HepG2 cells showed an enhanced capacity to bind viral particles, as indicated by the much higher MFI compared with untransfected HepG2 (Fig.  8A, bottom). This result clearly indicated that overexpression of HBV-BP on the cell surface was accompanied by an enhancement of cellular ability to bind viral particles. The HBV-BP role in HBV binding was further confirmed by the effect of soluble rHBV-BP on virus/cell interaction. As shown in Fig. 8A (bottom), the presence of the rHBV-BP protein with HBV particles resulted in a consistent reduction of virus attachment to HepG2/HBV-BP cells. The observed shift in the fluorescence signal (MFI) from 417 to 160, a value even lower than that obtained with untransfected cells, suggested that rHBV-BP in solution binds to viral particles inhibiting the attachment of HBV to the cells' surface.
A quantitative evaluation of the effect of HBV-BP overexpression on HBV binding, determining the amount of viral DNA found in different types of cells exposed to the same number of HBV particles, indicated that in HepG2/HBV-BPtransfected cells, the total amount of viral DNA found after incubation with HBV was about 100 times higher than that observed in nontransfected HepG2 cells under the same conditions ( Fig. 8B and Table I). Even HBV-BP-transfected CHO cells acquired a remarkable virus binding capability compared with untransfected cells, where virus binding was negligible; however, the total amount of viral DNA was considerably lower than that observed with transfected HepG2 cells. Unrelated cell lines used in control experiments, such as Rat-1, did not significantly bind HBV under the same conditions, while primary human hepatocytes, obtained from the explanted liver of a patient suffering from hypercystinuria, bound HBV much better than nontransfected HepG2 cells. Viral specificity of HBV-BP was determined with HCV. Interaction of HCV was assessed in HepG2 and HepG2/HBV-BP cell lines, and HCV RNA was measured in cellular RNA extracts. Apparently, HBV-BP is an HBV-specific receptor and does not enhance HCV binding activity to transfected cells (Table I).
Further confirmation that the increase of viral DNA found in HBV-BP-transfected cells was directly correlated to the enhanced amount of HBV-BP protein on the cell surface was provided by competition experiments with polyclonal antibodies raised against HBV-BP. As shown in Fig. 8C, the addition of anti-HBV-BP antibodies suppressed in a dose-dependent manner viral attachment to HepG2 and CHO transfected cells. At a concentration of 1 g/ml in the culture medium, anti-HBV-BP antibodies reduced the amount of viral DNA found in the cells to roughly 2-5% of the original value. Similar results were obtained with primary human hepatocytes, where, at a concentration of 1.0 g/ml, anti-HBV-BP antibodies completely blocked HBV binding. Preimmune rabbit serum, after affinity purification on protein A, was ineffective in reducing HBV binding to transfected HepG2 and CHO cells and to primary human hepatocytes (Fig. 8C).
These data clearly indicate that when HBV-BP is masked by antibodies, cells lose their ability to bind viral particles.
HBV-BP Is Responsible for HBV Entry in Transfected Cells and Primary Human Hepatocytes-The role of HBV-BP in virus entry into hepatic cells was examined by measuring the ability of soluble rHBV-BP to inhibit viral particle penetration into HepG2/HBV-BP cells and primary human hepatocytes. Internalized viral DNA was detected according to a well established procedure (4,12,17,43,44) by incubating the cells first with the virus suspension for 2 h at 37°C and subsequently by treating the cells with trypsin. To prove that trypsinization removes bound HBV from the cell surface, control cells were incubated in parallel with virus suspension at 4°C, not allowing HBV penetration. Viral particles that were not internalized during incubation were therefore removed as shown in Fig. 9A, and analysis of cellular extracted DNA was representative only of the internal nucleic acid content. By Southern blot hybridization, HBV DNA was recovered in the cytoplasmic fraction of cells incubated with the virus at 37°C, but not in the nuclear fraction, and the amount of recovered HBV DNA was higher in the cytoplasmic fraction of HepG2/HBV-BP than in nontransfected cells (Fig. 9B), suggesting an increased ability of HepG2/ HBV-BP cells to internalize HBV. The electrophoretic pattern of the trypsin wash was similar to that of the HBV inoculum, with a major viral DNA form migrating at 3.6 kilobase pairs, corresponding to relaxed circular viral DNA molecules, known to exist in HBV-infected human liver and in virions isolated from plasma of infected patients (45). In the cytoplasmic fractions, predominant double-stranded linear molecules, migrating at 3.2 kilobase pairs, were detected in both cell types, as previously observed for internalized nonreplicating forms of HBV DNA (46). The addition of increasing concentrations of soluble rHBV-BP up to a concentration of 1.8 g/ml (40 nM) to HBV particles before incubation with HepG2/HBV-BP cells decreased virus penetration to almost undetectable values (Fig.  10A). These data were further confirmed by experiments with primary human hepatocytes. As shown in Fig. 10B, the amount of viral DNA found in human hepatocytes treated with HBV particles was reduced by the addition of recombinant HBV-BP in a dose-dependent manner. Inhibition of virus entry in primary human hepatocytes was efficiently achieved also by anti-HBV-BP antibodies at a concentration of 1 g/ml (Fig. 10C).
These data indicate that HBV-BP is responsible for HBV attachment and entry in primary human hepatocytes. DISCUSSION Data obtained in this study indicate that HBV-BP is important for HBV binding to HepG2 cells and primary human hepatocytes. This protein has been isolated by using a derivative of the preS1-(21-47) peptide, the proposed virus attachment site, and interacts specifically with the entire HBV preS1 domain. FACS and immunofluorescence analysis suggest that HBV-BP is present on the surface of transfected HepG2 cells, an essential requirement to support the proposed role as an HBV-binding protein. Further support for this contention was provided by the remarkable enhancement of virus binding capability by transfected cells, even in the case of CHO cells, where HBV binding is normally negligible. Viral particle attachment to transfected cells is specific, since the addition of soluble rHBV-BP or anti-HBV-BP polyclonal antibodies inhibit virus binding in a dose-dependent manner. Cells overexpressing HBV-BP on their surface internalize more viral DNA than untransfected cells, and virus entry is blocked by the addition of soluble rHBV-BP. Primary human hepatocytes, the natural target for HBV infection, are refractive to HBV penetration in the presence of rHBV-BP or the corresponding antibodies.
HBV-BP results strictly correlated to human SCCA1 (37), with only three deduced amino acid differences: Gly 351 to Ala, Ala 357 to Thr, and Ser 389 to Pro. Chymotrypsin (39), cathepsin L, and papain (47) react with SCCA1, but SCCA1 function is substantially unknown. Like other human serpins such as plasminogen activator inhibitor type 2, placental thrombin inhibitor, and FIG. 8. Recombinant HBV-BP inhibits HBV binding to transfected cell lines and primary human hepatocytes. A, adsorption of enriched HBV particle preparation (5 ϫ 10 6 genome equivalents/sample) to HepG2 and to HepG2/HBV-BP cells (10 6 ) as detected by FACS by incubation with rabbit anti-HBs serum (Behringwerke) followed by treatment with FITC-conjugated goat anti-rabbit antiserum (Dako). In inhibition experiments, HepG2/HBV-BP cells were simultaneously incubated with HBV particles in the presence of 18 pg of recombinant HBV-BP. Cells were successively washed before the analysis. As control for the FACS procedure, cells were directly incubated with anti-HBs antiserum, washed, and then incubated with the FITC-conjugated antirabbit antibody. B, quantitative determination of HBV DNA in cellular extracts by a solution hybridization assay (Digene Hybrid Capture HBV DNA assay). C, inhibition of HBV (23 ϫ 10 7 viral genome equivalents) attachment to HepG2/HBV-BP cells (black), CHO/HBV-BP cells (white), and primary human hepatocytes (gray) by serial dilutions of affinity-purified anti-HBV-BP polyclonal antibodies. The dashed bar corresponds to the effect of control antibodies derived from rabbit preimmune sera on HBV attachment to HepG2/HBV-BP cells. No inhibition was observed also for CHO/HBV-BP cells and primary human hepatocytes. b For HCV binding experiments, cells were incubated with EMEM plus 10% FBS containing 100 l of HCV-positive serum with a known amount of RNA molecules (5.8 ϫ 10 6 genome equivalents l/ml). After extensive washes, cellular RNA was extracted by RNAZOL (Biogenesis), and HCV RNA was measured using a bDNA assay (Quantiplex 2.0, Bayer Diagnostic Division), as previously described (33). All experiments were repeated at least four times. Meq, megaequivalents. ND, not determined. elastase inhibitor, SCCA1 belongs to the ovalbumin family of serine protease inhibitors (Ov-serpin) (48), a subfamily of the large serpin superfamily (49). Serpins have a highly ordered tertiary structure defined by the crystal structure of the prototype molecule ␣ 1 -antitrypsin, consisting of nine ␣-helices and three ␤-sheets arranged in a stressed configuration with the reactive center, which has the unusual feature of being the most variable region, located in an exposed loop. When serpins bind to their protease target, the inhibition occurs by a profound conformational change initiated by reaction of the protease active site with the reactive center of the serpin (50). HBV-BP does inhibit cysteine proteinases such as cathepsin L and papain, and studies are in progress in order to further characterize this activity. This property could be very important for the proposed biological role of this protein as an HBV-binding protein, since it could be speculated that receptors involved in HBV binding might lose their ability to inhibit proteases, thus creating an area on the cell surface where proteases can act, thus facilitating cell infection. Very often, the entry of enveloped viruses in the cells requires a proteolytic digestion of envelope protein close to a fusion domain; the cleaved form of the envelope protein allows the fusion of virus with PM or, after endocytosis, with the endosomal membrane (51,52).
Like SCCA1, HBV-BP displays a pair of serine residues in the reactive site P1-P1Ј peptide bond localized at positions 354 and 355. Bovine ␣ 1 -antichymotrypsin is the only other serpin known to contain a Ser-Ser at the reactive site (53). HBV-BP represents the first Ov-serpin differing from all the others for the presence of a proline residue in position 389, which is one of the three amino acids differing from the SCCA1 sequence. The other two amino acid differences between SCCA1 and HBV-BP, Gly 351 to Ala and Ala 357 to Thr, are both localized within the reactive center, and the last is particularly interesting, since the Thr 357 residue could represent a new putative site of glycosylation.
Post-translational modification of HBV-BP could be fundamental for its function. In addition to glycosylation sites, sequence analysis of HBV-BP shows the presence of many putative phosphorylation sites. Phosphorylation of cellular receptors for viral particles is often essential for their functionality in the initial events of the infective process, as shown for the receptors of Epstein-Barr virus (54), mouse mammary tumor virus (55), and polio virus (56,57).
The genes coding for many serpins of the ovalbumin family are localized closely on chromosome 18. The SCCA1 and SCCA2 genes are closely linked, tandemly arrayed, and flanked by two closely related members of the Ov-serpin family, plasminogen activator inhibitor type 2 and maspin. All four serpins are located within 300 kilobase pairs of 18q21.3 (38). Apparently, HBV-BP is not simply an allelic variant of SCCA1. PCR experiments performed on human genomic DNA using primers designed to amplify the last exon of the SCCA gene, where all of the three nucleotide substitutions responsible for the three amino acid changes in the HBV-BP are located, indicate that the HBV-BP sequence is amplified as well as the SCCA1 and SCCA2 exons. This suggests that a specific gene coding for HBV-BP is present in the human genome or that a sequence coding for the last exon of the SCCA gene is present and used for alternative splicing processes. Studies are in progress to identify the genomic sequences coding for HBV-BP.
Previous works did not report expression of SCCA1 and of the high homologous SCCA2 gene in cells derived from human liver. SCCA1 was isolated as a component of T-4 antigen complex from a human metastatic squamous cell carcinoma (SCC) (58) and has been used as circulating tumor marker for the management and diagnosis of SCC of the cervix, head and neck, lung, and esophagus (59). SCCA1 is detected in normal squamous epithelium and produced at high levels by the corresponding carcinoma cells. The lack of tissue specificity may resemble the situation occurring with carboxypeptidase D (gp180), the proposed cellular receptor for duck HBV. In this case, the protein is found not only in duck liver but also in other tissues, which are not susceptible to duck HBV infection (60). As in our case, gp180 expression increases duck HBV attachment to and entry in transfected heterologous cells, but productive virus infection does not occur, suggesting that gp180 may function as a primary virus attachment site and that a species-specific co-receptor is needed to complete viral replication. Although our data show that HBV-BP is the primary factor responsible for HBV binding to and entry in primary human hepatocytes, the possibility cannot be excluded that additional, still unknown cellular components might be involved for the active replication cycle of the virus.