Properties of Monoclonal Antibodies Directed against Hepatitis B Virus Polymerase Protein

Abstract

Hepadnavirus polymerases are multifunctional enzymes that play critical roles during the viral life cycle but have been difficult to study due to a lack of a well-defined panel of monoclonal antibodies (MAbs). We have used recombinant human hepatitis B virus (HBV) polymerase (Pol) expressed in and purified from baculovirus-infected insect cells to generate a panel of six MAbs directed against HBV Pol protein. Such MAbs were subsequently characterized with respect to their isotypes and functions in analytical and preparative assays. Using these MAbs as probes together with various deletion mutants of Pol expressed in insect cells, we mapped the B-cell epitopes of Pol recognized by these MAbs to amino acids (aa) 8 to 20 and 20 to 30 in the terminal protein (TP) region of Pol, to aa 225 to 250 in the spacer region, and to aa 800 to 832 in the RNase H domain. Confocal microscopy and immunocytochemical studies using various Pol-specific MAbs revealed that the protein itself appears to be exclusively localized to the cytoplasm. Finally, MAbs specific for the TP domain, but not MAbs specific for the spacer or RNase H regions of Pol, appeared to inhibit Pol function in the in vitro priming assay, suggesting that antibody-mediated interference with TP may now be assessed in the context of HBV replication.

Hepadnaviruses are a group of small, enveloped DNA viruses that cause acute and chronic hepatitis and strongly predispose to the development of hepatocellular carcinoma (11). The prototype member of this virus family is the human hepatitis B virus (HBV). Despite containing a small (3 to 3.3 kb) encapsidated DNA genome, hepadnaviruses are classified as viral retroelements, because the central step in their replication cycle is the reverse transcription of an RNA intermediate (called a pregenome) (57) by virtue of a protein-primed reaction (3, 31, 63). Reverse transcription occurs within the nucleocapsid (core particle) composed of the nucleocapsid protein, the reverse transcriptase (RT)-polymerase (Pol), and the pregenome which is used as an RNA template. Pol is composed of four domains (44). From the amino terminus, the domains are (i) the terminal protein (TP), which becomes covalently linked to negative-strand DNA through the protein-primed initiation of reverse transcription, (ii) the spacer, which is tolerant of mutations, (iii) the RT, which contains the YMDD consensus motif for RT, and (iv) the RNase H.

The mechanism of genome replication for hepadnaviruses has been determined in detail. The initial step appears to be the recognition of the pregenomic RNA by Pol. This recognition occurs best in cis, appears to be cotranslational (2, 18, 19, 22, 25, 43), and is mediated by an RNA sequence (designated ɛ) that is present at both ends of the terminally redundant pregenomic RNA. However, only the 5′ copy of ɛ appears to function in packaging, and the ɛ sequence in itself is sufficient to induce the packaging of foreign RNA by HBV Pol (19, 22). The packaging of Pol is dependent upon an RNA molecule possessing a 5′ copy of ɛ (4). Thus, neither Pol nor pregenomic RNA can be packaged in the absence of the other. The second critical event in genome replication involves a priming reaction in which a nucleotide becomes covalently attached to Pol (3, 5, 38, 58, 63). The addition of the first four nucleotides is templated by a sequence in a bulge in the 5′ copy of ɛ (42, 59, 62). The primed Pol complex is then translocated to a complementary sequence present in the 3′ copy of a genetic element termed direct repeat (DR) 1, where the synthesis of minus-strand DNA resumes (8, 33, 39, 47–49, 59, 62, 67). The synthesis of minus-strand DNA terminates at the 5′ end of pregenomic RNA (47, 67). The RNA template is degraded by the RNase H activity of Pol. Only a short terminal oligoribonucleotide remains, which is then translocated to a homologous site, DR 2, on minus-strand DNA where it serves as the primer for plus-strand DNA (32, 35, 50, 55). Once plus-strand DNA synthesis has reached the 5′ end of minus-strand DNA, a final translocation to the 3′ end of minus-strand DNA occurs, resulting in a noncovalently closed, partially double-stranded, circular DNA molecule.

Hepadnavirus Pol proteins play a central role in the viral life cycle. Recently it was demonstrated that the formation of the Pol-pregenomic RNA ribonucleoprotein complex in the avian hepadnavirus duck hepatitis B virus depends on host cellular factors including the heat shock protein 90 (Hsp-90) and p23, a chaperone partner of Hsp-90 (21). This chaperone complex also appeared to be incorporated into viral nucleocapsids. These findings lend support to the concept that interactions of molecular chaperones with Pol play a critical role in the maintenance of the enzyme in a conformational state that renders it competent for its various functions.

Several systems which permit the direct analysis of Pol function in the absence of viral replication and other viral proteins have been described (20, 29, 30, 51, 58, 63). The Pol system utilizing purified HBV Pol from baculovirus-infected insect cells has been employed to dissect protein-protein and protein-RNA interactions involving Pol (30).

The baculovirus system has enabled us to obtain large amounts of purified Pol protein which we have used in the present study to raise monoclonal antibodies (MAbs) against HBV Pol by using the entire protein as the antigen. Such reagents have been difficult to generate because antigen preparations of sufficient purity did not exist, and Pol appears to be poorly immunogenic at the B-cell level. We have characterized these MAbs in detail and have used them as probes for the mapping of B-cell epitopes of Pol, as well as for studies addressing the intracellular localization of the protein. In addition, TP-specific MAbs appeared to inhibit the in vitro priming reaction.

MATERIALS AND METHODS

Production and purification of recombinant Pol.

Recombinant HBV Pol protein carrying a FLAG epitope at the N terminus was produced in baculovirus-infected insect cells as previously described (29, 30). The immunoaffinity purification of Pol with the M2 MAb (International Biotechnologies Inc., New Haven, Conn.) has been described previously (29, 30). To obtain large quantities of gel-purified Pol for immunizations, Pol was produced in the High Five Trichoplusia ni cell line (Invitrogen, Carlsbad, Calif.). High Five cells were infected with the recombinant baculovirus feline panleukopenia virus (FPL)-Pol (29), and 48 h postinfection the cells were scraped into a TNM buffer (100 mM Tris-HCl, pH 7.5; 30 mM NaCl; 10 mM MgCl₂) and sonicated. The cell lysate was clarified, and the insoluble pellet was solubilized by sonication in TNM buffer containing 6 M urea. Pol was separated on sodium dodecyl sulfate (SDS)–8% polyacrylamide preparative gels (26), localized by staining with Coomassie brilliant blue (0.25%) in H₂O, and excised from the gel. The gel fragments were homogenized, and Pol was eluted by shaking in 0.1% SDS. Pol was concentrated in a Centricon 30 microconcentrator (Millipore Co., Bedford, Mass.).

Establishment of MAbs against Pol.

BALB/c mice were immunized intraperitoneally with purified Pol protein, and serum from immunized animals was periodically analyzed for reactivity against Pol by Western blotting. After a final intravenous boost with antigen 3 days prior to fusion, spleen cells were fused with the Sp2/O-Ag14 myeloma cell line (American Type Culture Collection, Rockville, Md.) as described previously (61). Hybridomas were selected and maintained as described previously (16, 61). The screening procedure was as follows. Preparations of purified Pol were separated by SDS–8% polyacrylamide gel electrophoresis (PAGE) and transferred to an Immobilon-P membrane (Millipore Co.). Undiluted supernatants from hybridoma colonies were applied as the primary antibody with a Miniblotter model 45 (Immunetics, Cambridge, Mass.), which allowed the testing of 45 supernatants on one 13- by 13-cm membrane. Antibodies that bound to Pol were visualized after incubation with a horseradish peroxidase-conjugated sheep anti-mouse antiserum (NA 931; Amersham Life Sciences Inc., Arlington Heights, Ill.) and subsequent chemiluminescence detection with the ECL system (Amersham Life Sciences Inc.). Hybridomas that were immunoreactive with recombinant Pol were cloned by limiting dilution. The MAb isotype was determined with the IsoStrip mouse MAb isotyping kit (Boehringer Mannheim, Indianapolis, Ind.). A protein G column (Pharmacia, Piscataway, N.J.) was used for the affinity purification of MAbs from ascites fluid.

EIA and immunoprecipitation.

Recombinant Pol (200 ng/well) was coated onto enzyme immunoassay (EIA) plates (Corning Costar Co., Cambridge, Mass.) for 12 to 16 h at room temperature and incubated for 1 h at room temperature with various MAbs (final concentration, 1 μg/ml), followed by incubation for 1 h at room temperature with a 1:5,000 dilution of a horseradish peroxidase-conjugated sheep anti-mouse antiserum (NA 931; Amersham Life Sciences Inc.). Bound antibodies were visualized with the OPD (o-phenylenediamine-2–HCl) reagent (Abbott Diagnostics, North Chicago, Ill.). For the immunoprecipitation of recombinant Pol with MAbs, Sf9 insect cells infected with FPL-Pol were labeled 42 to 46 h postinfection with 200 μCi of [³⁵S]methionine (NEN, Boston, Mass.) per ml. Cells were washed twice in phosphate-buffered saline (PBS) and lysed in an extraction buffer (EB) (50 mM Tris-HCl, pH 9.0; 100 mM NaCl; 1% Nonidet P-40), supplemented with protease inhibitors. Clarified lysates were incubated for 4 h at 4°C with MAbs against Pol prebound to protein G affinity beads (Life Technologies, Gaithersburg, Md.), followed by three washes with buffer WB (EB plus 0.5% sodium deoxycholate and 0.1% SDS). Proteins bound to pelleted beads were eluted with SDS-gel sample buffer containing 2% SDS and 2% β-mercaptoethanol and separated by SDS–12% PAGE as described previously (27).

Cells, transfections, and infections.

The human hepatocellular carcinoma (HCC) cell line HuH-7 (41) was grown in modified Eagle minimal essential medium (Cellgro Mediatech, Washington, D.C.), supplemented with 10% fetal calf serum, 1% nonessential amino acid solution (Life Technologies), and 1% penicillin-streptomycin stock solution (Cellgro Mediatech). Transfections were performed by using a modified calcium phosphate precipitation protocol (7) routinely with 20 μg of DNA plus 1 μg of reporter plasmid pTKGH (52) per 100-mm-diameter plate seeded with 7 × 10⁶ cells. HuH-7 cells were transfected either with the construct pMT-HBVpol (the kind gift of Heinz Schaller) (44) in which Pol expression is driven by the ubiquitously active human metallothionein II_A promoter (15), or with the construct pCH3142 (22) (the kind gift of Michael Nassal). pCH3142 bears a 1.1 HBV genome-length HBV DNA sequence under the control of the cytomegalovirus immediate-early promoter but carries a 42-nucleotide (nt) deletion from nt 1818 to 1859 (numbering according to reference 10). Transcription from this construct yields pregenomic RNA species carrying a short deletion in the lower stem of the 5′ ɛ signal that renders these transcripts noncompetent for encapsidation. As a consequence, pregenomic RNA, the core protein, and Pol are not assembled into nucleocapsids that support viral DNA synthesis, and Pol protein is expected to be present intracellularly in a nonencapsidated state. For some immunofluorescence experiments, HuH-7 HCC cells were infected with a recombinant vaccinia virus that allowed for the inducible expression of Pol (30a). The FPL-Pol insert was cloned into the pVOTE-2 vector, and a recombinant vaccinia virus was generated as described previously (64).

Protein analysis.

The reactivity patterns of the MAbs against Pol and Pol degradation products were determined by Western blot analysis with purified Pol. Pol was separated on a preparative minigel (SDS–12% PAGE) and transferred to a Sequiblot polyvinylidene fluoride membrane (Bio-Rad Laboratories, Hercules, Calif.). Antibodies were incubated with the membrane in individual lanes with a PR-150 Mini decaprobe (Hoefer Scientific Instruments, San Francisco, Calif.) at a final concentration of 5 μg/ml, followed by rabbit anti-mouse immunoglobulin G (IgG) (final concentration) and ¹²⁵I-labeled protein A (NEN).

Immunofluorescence and immunocytochemistry studies.

HuH-7 cells were grown on sterile glass slides and either transfected by the standard calcium phosphate precipitation protocol (7) or infected with recombinant vaccinia virus. Cells were washed once with PBS and fixed with HistoChoice (Amresco, Solon, Ohio) tissue fixative for 30 min at room temperature. After one wash with PBS, cells were permeabilized with 0.05% Saponin in PBS for 10 min at room temperature. After blocking for 1 h at room temperature in PBS–1% bovine serum albumin, MAbs directed against Pol (final concentration, 1 μg/ml) were added, and the solution was incubated 12 to 16 h at 4°C. After being washed three times with PBS, slides were incubated for 30 min with a 1:250 dilution of a biotinylated horse anti-mouse antiserum (Vector Laboratories, Burlingame, Calif.). For immunofluorescence, cells were equilibrated in 0.1 M NaHCO₃–1.5 M NaCl, pH 8.2, for 5 min, and the final incubation was performed with an avidin-fluorescein isothiocyanate conjugate (Vector Laboratories) at a 1:500 dilution. Cover slides were mounted in Vectashield (Vector Laboratories) and examined with a Nikon Labophot photomicroscope equipped with the epifluorescence attachment EF-D (Nikon, Garden City, N.Y.). Confocal microscopy was performed with a Leica TCS4D confocal scanner (Leitz, Wetzlar, Germany). For immunocytochemistry, cells were incubated with the Vectastain Elite ABC reagent and stained by using a 3,3′-diaminobenzidine substrate kit (both from Vector Laboratories) according to the instructions of the manufacturer.

Epitope mapping.

A set of deletion mutants of Pol produced in and purified from baculovirus-infected insect cells was used to test the reactivity of MAbs against Pol by Western blotting. The constructs represented a series of amino- and carboxy-terminal deletion mutants of the TP and RT domains and permitted the mapping of epitopes to within 10 to 32 amino acids. The details of the construction of these vectors are described elsewhere (28).

Pol assays.

Pol reactions were performed with Pol polypeptides immunoprecipitated by MAbs against Pol and still bound to the affinity beads. The beads were suspended in TNM (100 mM Tris HCl, pH 7.5; 30 mM NaCl; 10 mM MgCl₂) containing 100 μM concentrations of unlabeled deoxyribonucleoside triphosphates (dATP, dGTP, and dCTP) and 5 μCi of [α-³²P]TTP (3,000 Ci/mmol; NEN). Assays were routinely performed at 30°C for 30 min. Densitometry of gels was performed using the NIH Image 1.60 software (42a).

RESULTS

Generation of MAbs against Pol protein.

Initially, animals were immunized with Pol purified by the M2 MAb affinity column. When analyzed by Coomassie blue-stained SDS-PAGE, this material derived from insect cells contained several additional bands that copurified with Pol (29). Such proteins could not be removed from Pol without denaturing the protein. Mice immunized with this material showed a predominant immune response against a protein with an apparent molecular weight (MW of 70,000), but no reactivity against Pol was detectable. Pol appeared to be less immunogenic than one or several of the contaminating bands. Therefore, mice were repeatedly immunized with gel-purified Pol protein over the time course of 1 year. Finally, the serum of one animal that had been immunized intraperitoneally seven times exhibited a strong reactivity against Pol at a serum dilution of 1:5,000. This animal was used for the cell fusion with Sp2/O-Ag14 myeloma cells. The screening of hybridoma supernatants was performed by Western blotting. Six hybridomas producing Pol-specific MAbs were obtained from this fusion. Table 1 summarizes the characteristics of these MAbs. All MAbs functioned well in a Western blot format, and all except the IgM MAb 10B9 recognized Pol in an EIA format. MAbs 2C8, 8D5, and 9F9 were able to detect endogenously synthesized Pol by indirect immunofluorescence. These three antibodies also functioned well in immunoprecipitation studies as shown below.

TABLE 1.

Characteristics of MAbs

Characteristic or test	MAb
	1B4	2C8	7C3	8D5	9F9	10B9
Isotype	IgG1 κ	IgG1 κ	IgG2a κ	IgG1 κ	IgG1 κ	IgM κ
Immunoblot	+	+	+	+	+	+
EIA	+	+	+	+	+	−
Indirect immunofluorescence	−	+	−	+	+	−
Immunoprecipitation	(+)a	+	(+)a	+	+	−
Epitope	aa 20–30	aa 8–20	aa 20–30	aa 225–250	aa 800–832	aa 20–30
Pol domain	TP	TP	TP	Spacer	RNase H	TP

^a(+), weak reactivity. 

Epitope mapping.

Figure 1 shows the results from epitope mapping studies performed by using all six MAbs as primary antibodies in immunoblots against deletion mutants of Pol. MAbs 1B4, 7C3, and 10B9 all recognized an epitope within amino acid (aa) positions 20 to 30 of Pol in the TP region. MAb 2C8 recognized an epitope between aa positions 8 and 20 within the TP region. An epitope between aa positions 225 and 250 in the spacer region of Pol was recognized by MAb 8D5. The MAb 9F9 recognized an epitope between aa positions 800 and 832 within the RNase H domain of Pol. Thus, murine B-cell epitopes of HBV Pol appeared to be positioned at the very N and C termini of the protein and within the Pol spacer region.

FIG. 1.

Epitope mapping of Pol MAbs. MAbs 1B4, 2C8, 7C3, 8D5, 9F9, and 10B9 were used as primary antibodies in immunoblots against various deletion mutants of Pol expressed in and purified from baculovirus-infected insect cells (see Materials and Methods). MAbs 1B4, 7C3, and 10B9 recognize an epitope within aa 20 to 30 of Pol in the TP region of HBV subtype ayw. MAb 2C8 recognizes an epitope between aa 8 and 20 within the TP region. An epitope between aa 250 and 275 in the spacer region of Pol is recognized by MAb 8D5. The MAb 9F9 recognizes an epitope between aa 800 and 832 within the RNase H domain of Pol.

Detection of endogenously synthesized Pol by indirect immunofluorescence.

HuH-7 HCC cells were transfected with the construct pMT-HBVpol in which Pol expression is driven by the human metallothionein II_A promoter (Fig. 2A to F). The expression level of Pol from this construct was estimated to be at least 10 times higher than the level obtained from the endogenous Pol promoter. Alternatively, Pol was also inducibly expressed in HuH-7 cells after infection with a recombinant vaccinia virus coding for Pol (VV_Pol) (Fig. 2G to K). The expression levels of Pol reached with VV_Pol were at least 10 times higher than those with pMT-HBVpol. Cells were fixed 2 days after transfection or 4 h after vaccinia virus infection. The results shown in Fig. 2 illustrate intracellular staining patterns obtained with the MAb 2C8, specific for an epitope in the TP domain, and 8D5, specific for an epitope in the spacer region. Staining with the MAb 9F9 yielded similar results (data not shown). As demonstrated in Fig. 2A and B, a fine granular, cytoplasmic staining pattern (compare Nomarski images in panels D and E) was observed with both MAbs when Pol was expressed from the construct pMT-HBVpol. Control transfections with the HBV L protein expression construct pApLHBs (13) and incubation with Pol MAb 8D5 (Fig. 2C and F) or 2C8 (data not shown) did not result in specific signals. All VV_Pol-infected HuH-7 cells exhibited a very strong cytoplasmic staining pattern (Fig. 2G). When such cells were incubated with the HBV L protein-specific MAb 18/7 (17), no specific signals were visible (Fig. 2H). Nuclear staining was not detectable in all cases. Similar results were obtained when murine BALB/3T3 fibroblasts were infected with VV_Pol (data not shown). These data illustrated that MAbs 2C8, 8D5, and 9F9 were capable of detecting endogenously synthesized Pol protein in transfected or vaccinia virus-infected HCC cells. Pol expressed in HuH-7 HCC cells appeared to be exclusively localized in the cytoplasm.

FIG. 2.

Confocal microscopy studies of endogenously synthesized Pol in HuH-7 HCC cells with Pol MAbs. Immunofluorescence (A, B, C, G, and H) and Nomarski images (D, E, F, J, and K) are illustrated. MAbs used for the staining are indicated on the lower right of each immunofluorescence image. (A to F) HuH-7 HCC cells transfected with the construct pMT-HBVpol in which Pol expression is driven by the human metallothionein II_A promoter. (G to K) Expression of Pol in HuH-7 cells after infection with recombinant VV_Pol. Cells were fixed 2 days after transfection or 4 h after vaccinia virus infection. The intracellular staining patterns obtained with the MAbs 2C8 (specific for an epitope in the TP region) and 8D5 (specific for an epitope in the spacer region) are illustrated. MAb 9F9 yielded similar results (data not shown). A fine granular, exclusively cytoplasmic staining pattern (compare panel A with D, B with E, and G with J) is observed. Control transfections with the HBV L protein expression construct pApLHBs (13) and staining with Pol MAb 2C8 (data not shown) 8D5 do not result in specific signals (C and F). VV_Pol-infected cells are negative when stained with the HBV L protein-specific MAb 18/7 (17) (H and K). Similar results were obtained when murine BALB/3T3 fibroblasts were infected with VV_Pol (data not shown).

Immunocytochemistry was used to detect HBV Pol intracellularly in the presence of the other HBV proteins. When cells transfected by HBV constructs (of more than one genome length) that allow for viral DNA synthesis were analyzed with MAb 2C8, 8D5, or 9F9, no signals corresponding to HBV Pol could be detected, while viral core and envelope proteins were readily detectable (data not shown). Similar results were obtained when the cell line HepG2-2.2.15 (53), stably expressing HBV proteins and replicating the virus, was analyzed (data not shown). When liver tissue sections from mice transgenic for HBV were subjected to immunohistochemistry with the Pol-specific MAbs, no signal was detected (data not shown; analysis kindly performed by Luca Guidotti, Scripps Research Institute). Some possible reasons for the inability to detect Pol in these experimental settings included (i) low intracellular levels of Pol and/or (ii) the inaccessibility of Pol due to encapsidation of the protein into core particles. To test the latter hypothesis, we used the construct pCH3142 (22), coding for mutant pregenomic RNA species with a deletion in the 5′ ɛ signal rendering these transcripts noncompetent for encapsidation. After the transfection of cells with pCH3142, pregenomic RNA, core protein, and Pol were not expected to be assembled into nucleocapsids supporting viral DNA synthesis, and Pol protein was likely present intracellularly in a nonencapsidated state. HuH-7 HCC cells were transfected with either pMT-HBVpol or pCH3142, and Pol was detected with MAb 2C8 (Fig. 3). As a control, HBV core protein was detected with a polyclonal rabbit anti-HBcAg antiserum (DAKO, Carpinteria, Calif.). A faint cytoplasmic signal corresponding to HBV Pol was detectable in cells transfected with pCH3142 (Fig. 3B), which was at least 10 times weaker than the signal obtained after the transfection of pMT-HBVpol (Fig. 3A). The HBV core protein was readily detectable in the nuclei and cytoplasms of cells transfected with pCH3142 (Fig. 3D), while it was not observed in cells transfected with pMT-HBVpol (Fig. 3C). In addition, HBV envelope proteins were detectable after the transfection of pCH3142 (data not shown). These data suggested that HBV Pol was predominantly localized in the cytoplasms of transfected cells in the presence of other HBV proteins.

FIG. 3.

Detection of HBV Pol in the presence of other HBV proteins. The immunocytochemistry of cells transfected with pMT-HBVpol (A and C) or pCH3142 (B and D), a construct that allows for the intracellular expression of Pol in a nonencapsidated state in the presence of core and envelope proteins (22), is shown. The staining was performed with MAb 2C8 (A and B) or a polyclonal rabbit anti-HBcAg antiserum (DAKO) (C and D). A faint cytoplasmic signal corresponding to HBV Pol is detectable in cells transfected with pCH3142 (B). In addition, HBcAg is detectable in the cytoplasms and nuclei of pCH3142-transfected cells (D). pMT-HBVpol-transfected cells exhibit the previously detected, strong cytoplasmic staining pattern for Pol (A) (see Fig. 2), while no HBcAg is detectable (C).

Western blots and immunoprecipitations.

All Pol MAbs except 10B9 were analyzed for their staining patterns of Pol and Pol degradation products on Western blots. The M2 MAb served as a positive control. This MAb is known to bind to the N terminus of Pol. As demonstrated in Fig. 4, lane 6, M2 recognized full-length Pol as well as degradation products containing the N terminus with apparent MWs down to ca. 30,000. MAbs 1B4, 2C8, and 7C3 exhibited the same staining pattern as M2 (Fig. 4, lanes 1 to 3). These data are consistent with the observation that these Pol MAbs recognize epitopes at the N terminus of Pol. MAb 8D5 (Fig. 4, lane 4) stained all Pol-associated bands except the smallest one, which is consistent with its epitope being located in the spacer region. Finally, MAb 9F9 (Fig. 4, lane 5) identified only full-length Pol and a very minor degradation product with an apparent MW of 68,000. This pattern is consistent with the observation that MAb 9F9 recognizes an epitope at the C terminus of Pol.

FIG. 4.

Detection of recombinant Pol by MAbs on Western blots. Staining with the FLAG epitope-specific M2 MAb (lane 6) serves as a positive control. M2 recognizes full-length Pol (position indicated on the left) as well as C-terminal degradation products with apparent MWs down to ca. 30,000. MAbs 1B4 (lane 1), 2C8 (lane 2), and 7C3 (lane 3) exhibit the same staining pattern as M2, consistent with the observation that these Pol MAbs recognize epitopes at the N terminus of Pol. MAb 8D5 (lane 4) stains all Pol-associated bands except the smallest one, which is consistent with its epitope being located in the spacer region. MAb 9F9 (lane 5) stains only full-length Pol and a minor degradation product with an apparent MW of 68,000. Positions of molecular mass markers are indicated on the right.

Pol MAbs were analyzed for their potential to immunoprecipitate metabolically labeled Pol from cellular extracts of Sf9 insect cells. As demonstrated in Fig. 5, lanes 3, 5, and 6, MAbs 2C8, 8D5, and 9F9 were able to immunoprecipitate Pol well. The M2 MAb (Fig. 5, lane 10) served as a positive control. MAbs 1B4 and 7C3 yielded only very small amounts of Pol, whereas the IgM MAb 10B9 did not immunoprecipitate Pol (Fig. 5, lanes 2, 4, and 7). Several additional signals with lower apparent MWs than that of Pol were visible after immunoprecipitations with MAbs 2C8, 8D5, and 9F9. It is currently unclear which proteins correspond to the observed signals, but it is unlikely that they represent the Pol degradation products observed by Western blotting, since 2C8 and 9F9 would not be expected to recognize the same degradation products of Pol. Of note, the MAb 9F9 coimmunoprecipitated a protein that was visible as a band with an apparent MW of 27,000. This band did not appear to correspond to a degradation product of Pol since it was not recognized by Western blotting with the same antibody.

FIG. 5.

Immunoprecipitation of metabolically labeled Pol from cellular extracts of Sf9 insect cells. MAbs 2C8 (lane 3), 8D5 (lane 5), and 9F9 (lane 6) immunoprecipitate Pol well. The M2 MAb (lane 10) serves as a positive control. MAbs 1B4 (lane 2) and 7C3 (lane 4) yield only very small amounts of Pol, whereas the IgM MAb 10B9 (lane 7) does not immunoprecipitate Pol at all. Controls with protein G alone (lane 8) and an irrelevant antibody (lane 9) are negative. MAb 9F9 coimmunoprecipitates a protein that is visible as a band with an apparent MW of 27,000. The position of full-length Pol protein is indicated on the right. Lane 1, molecular mass markers.

Inhibition of in vitro priming with Pol MAbs.

All Pol MAbs that were capable of immunoprecipitating Pol were tested for their potential to inhibit the in vitro priming activity of the enzyme. For this purpose, Pol reactions were performed with Pol proteins immunoprecipitated by Pol MAbs. The Coomassie blue-stained protein gel (Fig. 6, top panel) illustrates the amounts of Pol precipitated with the various MAbs, and the lower panel in Fig. 6 shows the result from the priming reaction of the same samples with Pol bound to the protein G beads by the respective antibodies. Signals present on the gels were quantified by densitometry. Immunoprecipitation and in vitro priming with the M2 MAb (Fig. 6, lane 8) served as a positive control. No immunoprecipitation of Pol was observed with the negative-control antibodies C7-57, specific for the bacterial glutathione S-transferase protein (40) (Fig. 6, lane 6), and 12CA5 directed against an influenza virus hemagglutinin peptide sequence (Fig. 6, lane 7), and consequently, no in vitro priming was detectable. When Pol was immunoprecipitated with the MAb 9F9, no inhibition of priming activity was observed (Fig. 6, lane 5). In contrast, the MAb 2C8 immunoprecipitated equal amounts of Pol protein (Fig. 6, lane 4), but the extent of in vitro priming was inhibited by 86%. A 14% inhibition of priming was observed in the case of the MAb 8D5 (Fig. 6, bottom panel, lane 3). MAbs 1B4 (Fig. 6, top panel, lane 1) and 7C3 (top panel, lane 2) immunoprecipitated equal amounts of Pol when compared with MAb 8D5, but the in vitro priming obtained with MAbs 1B4 and 7C3 was strongly reduced (1B4, 98% inhibition; 7C3, 81% inhibition) when compared with MAb 8D5 (Fig. 6, bottom panel; compare lanes 1 and 2 with lane 3). These observations suggested that TP-specific MAbs were capable of inhibiting in vitro priming by Pol. However, the possibility existed that these MAbs recognized and immunoprecipitated an inactive fraction of the Pol expressed in insect cells. To examine this possibility, Pol was immunoprecipitated with M2 to ensure that the immunoprecipitated protein represented the active fraction, and then Pol still bound to the M2 beads was exposed to purified MAb 2C8 or 9F9 or to the buffer without antibodies. The beads were washed to remove excess antibody, and priming reactions were conducted with the bound Pol. The buffer control and 9F9-exposed Pol exhibited similar priming activities, while the priming reaction for 2C8-exposed Pol was reduced by more than 50% (data not shown). These observations confirmed that the MAbs directed to the TP domain were capable of inhibiting Pol in vitro priming activity.

FIG. 6.

Inhibition of in vitro priming with TP-specific MAbs. (Top panel) Immunoprecipitation (IP) of Pol protein by various MAbs and subsequent detection by Coomassie blue-stained SDS-PAGE. (Bottom panel) In vitro priming assay of the same samples (Pol protein radiolabeled by nascent HBV minus-strand DNA; see Materials and Methods). The Coomassie blue-stained protein gel shows the amounts of Pol precipitated with the various MAbs. Immunoprecipitation and in vitro priming with the M2 MAb (lane 8) serve as a positive control. No immunoprecipitation of Pol is observed with the negative-control antibodies C7-57 (specific for the bacterial glutathione S-transferase protein) (lane 6) and 12CA5 (directed against an influenza virus hemagglutinin peptide sequence) (lane 7), and consequently, no in vitro priming is detectable. The MAb 2C8 (bottom panel, lane 4) inhibits priming by 86% (analysis by densitometry) when compared with M2 (bottom panel, lane 8). In contrast, the MAb 9F9 (bottom panel, lane 5) does not inhibit priming. All three MAbs immunoprecipitate equal amounts of Pol (top panel, lanes 4, 5, and 8). MAbs 1B4, 7C3, and 8D5 immunoprecipitate less Pol than M2 (top panel, lanes 1, 2, 3, and 8). MAbs 1B4 and 7C3 strongly reduce priming (1B4, 98% inhibition; 7C3, 81% inhibition) when compared with MAbs 8D5 and M2 (bottom panel, lanes 1, 2, 3, and 8). The positions of mouse Ig heavy chains (HC) and light chains (LC) are indicated on the right. Positions of molecular mass markers are indicated on the left.

DISCUSSION

This report describes the generation and characterization of MAbs against the full-length HBV Pol protein and their value for study of this protein which plays a central role in the viral life cycle. A panel of six MAbs against Pol was generated from a mouse immunized seven times over the time course of 1 year. Several attempts to obtain Pol MAbs from animals that had been immunized fewer times over shorter time periods were unsuccessful. In addition, the purity of the Pol antigen used for immunizations turned out to be a critical factor. Standard purified Pol preparations from baculovirus-infected insect cells (29) still contained ample amounts of several additional proteins of which at least one contaminant was found to be very immunogenic. Eventually, only gel-purified material was able to elicit an immune response that was sufficient for the generation of Pol-specific MAbs. The recombinant Pol used in this study was poorly immunogenic with respect to eliciting a humoral immune response in mice. Interestingly, a fusion performed with spleen cells from an animal that had been immunized four times over half a year yielded only MAbs of the isotype IgM (data not shown), suggesting that the affinity maturation and isotype switching in mice during the humoral immune response to recombinant Pol occur slowly. In contrast, Rehermann et al. (45) found Pol to be quite immunogenic at the cytotoxic T-lymphocyte level. These authors also noted a rapid degradation of Pol from its C terminus, which also was detectable with the various MAbs used in this study. Of note, one other murine MAb produced against a recombinant Pol polypeptide derived from the TP region has been described (14).

Epitope mapping studies presented here demonstrated that the newly established MAbs recognized four different epitopes on Pol. Two of these epitopes are positioned adjacently to each other at the N terminus: one is in the spacer region and the fourth is located at the C terminus of Pol. It has been previously demonstrated that certain patients infected with HBV exhibit humoral immune responses against Pol (6, 9, 24, 56, 65, 66, 69). Most of these studies have identified antigenic regions of the Pol protein at the N and C termini, whereas the immune responses to central regions of the Pol protein were represented to a lesser extent. Our study shows that murine B-cell epitopes appear to be located in Pol regions that have previously been demonstrated to elicit humoral immune responses in HBV-infected individuals.

Indirect immunofluorescence and confocal microscopy as well as immunocytochemistry studies using the newly established MAbs demonstrated that full-length Pol appeared to be exclusively localized in the cytoplasms of HuH-7 HCC cells, even in the presence of other HBV proteins. In no case was nuclear staining detectable, neither by MAb 2C8, which binds to the terminal protein region, nor by MAb 8D5, reactive with the spacer region, or MAb 9F9, directed against the RNase H domain. Similar observations were made after the infection of HuH-7 HCC cells or BALB/3T3 fibroblasts with VV_Pol. Vaccinia virus-infected cells overexpressed Pol, and strong staining throughout the cytoplasm was observed. However, the presence of Pol in the nucleus of transfected or infected cells at very low levels that were not detectable by the MAbs cannot be excluded. Our observations suggest that the full-length HBV Pol protein alone does not contain a nuclear localization signal that is efficiently recognized by the cell types used in this study. This finding is relevant because Pol protein is covalently bound to the minus DNA strand of the virion-encapsidated form of the genome. Therefore, it has been suggested that Pol may play a role in the intracellular amplification of viral covalently closed circular DNA by facilitating the entry of viral genomes (68) from mature core particles that are located in the cytoplasm into the nucleus. Kann and coworkers (23) have addressed this question by analyzing the intracellular trafficking of viral components and complexes in digitonin-permeabilized HuH-7 HCC cells, whose cytosol was substituted by rabbit reticulocyte lysate and an ATP-generating system. They found that a woodchuck hepatitis B virus-derived Pol-DNA complex was efficiently transported into the nuclei of HuH-7 cells by an ATP-dependent mechanism, whereas deproteinized viral DNA remained completely outside the nucleus, suggesting that the viral Pol is sufficient for mediating the transport of the viral genome into the nucleus. However, a possible contribution of core protein subunits associated with Pol (3) to the transport of viral genomes into the nucleus could not be excluded.

An immunohistochemical study performed with liver specimens from patients chronically infected with HBV demonstrated that polyclonal antisera raised against portions of the TP regions of Pol stained hepatocytes predominantly in the nucleus (37). These observations suggested that either the entire Pol protein or the portion encoding the TP is translocated to the nucleus during the course of natural infection, although the mechanism of transport remained unclear. Our observation that the MAb 2C8, which is specific for an epitope within the TP region of Pol, did not exhibit a nuclear staining pattern is not necessarily contradictory to these findings, because we expressed Pol in transfected cells, either in the presence or absence of other viral proteins. Our observation that Pol alone remains in the cytoplasms of transfected cells even when strongly overexpressed makes it possible that the putative signal that mediates the entry of Pol or Pol subdomains into the nuclei of naturally infected cells consists of multiple components and may be active only during certain stages of the viral life cycle. Thus, the intracellular localization of Pol may be determined by viral and/or cellular factors that are associated with it. It will be of interest to reassess the intracellular localization of Pol and Pol subdomains in the infected liver by using the MAbs described here.

The MAbs developed in our study were used to investigate whether the inhibition of in vitro priming by Pol could be achieved with these reagents. We used the various MAbs to immunoprecipitate Pol and then performed priming-reverse transcription reactions. These experiments revealed that the TP-specific MAbs 1B4, 2C8, and 7C3 were able to inhibit priming and reverse transcription, whereas the spacer-specific MAb 8D5 and the RNase H-specific MAb 9F9 were not. These data suggest that the TP-specific MAbs were able to interfere with the TP function in this assay and point to the importance of the conservation of structural features within the TP region of Pol for the proper function of the enzyme. While the experimental system used for these studies is quite different with respect to in vivo viral replication, a further assessment of the potential of the MAbs described here to interfere with Pol functions within the cell will be of interest.

Considerable efforts are being made to develop compounds that block HBV replication. One present focus is to find chemical compounds that selectively interfere with essential steps of the viral life cycle and replication without significantly affecting host cell metabolism. HBV Pol is a candidate target protein, because it is essential for viral multiplication. Inhibitors that target Pol fall within two broad categories, nucleoside analogs and nonnucleoside derivatives. However, prolonged chemotherapy may result in the emergence of resistant viruses. Such resistance phenomena may be due to specific changes in the gene encoding Pol (1, 34, 60). Therefore, alternative strategies against HBV based on gene therapy approaches are being actively studied (70). One such strategy involves the expression of virus-specific recombinant antibodies targeted to intracellular compartments of infected cells in order to interfere in a specific manner with the corresponding viral antigen (46). This approach has been demonstrated to be effective against human immunodeficiency virus type 1 in experimentally infected cells. For example, intracellularly expressed engineered antibodies against human immunodeficiency virus type 1 RT were able to confer protection against viral infection and replication (36, 54). An important requirement for this approach is the availability of recombinant forms of a high-affinity antibody specific for the target antigen (12) and the characterization of the antibody with respect to its possible neutralization of essential viral functions. In this context, the development of HBV Pol-specific MAbs and the study of their impact on Pol function in the in vitro priming assay represent an important first step towards the further exploration of the intracellular antibody strategy against HBV. In the infected cell, Pol is encapsidated together with pregenomic RNA and possibly other host cell-derived proteins into nucleocapsids. In addition to the possible interference of Pol-specific MAbs with the priming of minus-strand DNA, it is likely that recombinant antibody fragments bound to Pol in the cell will reduce the efficiency with which the enzyme is packaged into nucleocapsids. If so, MAbs that bind to Pol regions other than TP may also interfere with Pol function intracellularly.