Fcγ Receptor-like Activity of Hepatitis C Virus Core Protein*

We have previously demonstrated that viral particles with the properties of nonenveloped hepatitis C virus (HCV) nucleocapsids occur in the serum of HCV-infected individuals (1). We show here that nucleocapsids purified directly from serum or isolated from HCV virions have FcγR-like activity and bind “nonimmune” IgG via its Fcγ domain. HCV core proteins produced in Escherichia coli and in the baculovirus expression system also bound “nonimmune” IgG and their Fcγ fragments. Folded conformation was required for IgG binding because the FcγR-like site of the core protein was inactive in denaturing conditions. Studies with synthetic core peptides showed that the region spanning amino acids 3–75 was essential for formation of the IgG-binding site. The interaction between the HCV core and human IgG is more efficient in acidic (pH 6.0) than in neutral conditions. The core protein-binding site on the IgG molecule differs from those for C1q, FcγRII (CD32), and FcγRIII (CD16) but overlaps with that for soluble protein A from Staphylococcus aureus (SpA), which is located in the CH2-CH3 interface of IgG. These characteristics of the core-IgG interaction are very similar to those of the neonatal FcRn. Surface plasmon resonance studies suggested that the binding of an anti-core antibody to HCV core protein might be “bipolar” through its paratope to the corresponding epitope and by its Fcγ region to the FcγR-like motif on this protein. These features of HCV nucleocapsids and HCV core protein may confer an advantage for HCV in terms of survival by interfering with host defense mechanisms mediated by the Fcγ part of IgG.

HCV 1 is a member of the Flavivirus family and has a positive-strand RNA genome of ϳ9.5 kb. The viral genome encodes a large precursor protein that is cleaved by both cellular and viral proteases into structural (core, E1, and E2) and nonstructural proteins (P7, NS2, NS3, NS3a, NS4b, NS5a, and NS5b). The HCV core protein is encoded by the 5Ј-terminal region of the open reading frame and is composed of a basic RNA-binding domain I (aa 1-122) and highly hydrophobic domains II (aa 123-174) and III (aa 174 -192) (2,3). The core protein has several functional motifs, including putative nuclear localization signals (4), a DNA-binding motif (SPRG), and several cAMP-dependent protein kinase and protein kinase C recognition sites (5,6), and it binds to lipid droplets via domain II (2,3). The HCV core protein has many effects on host cell functions: modulation of gene expression (7), apoptosis (8,9), lipid metabolism (10), and transforming activity (11). It may also interfere with host defense mechanisms by modulating Fasand tumor necrosis factor ␣-mediated signaling (12) or by suppressing the antiviral cytotoxic T lymphocyte response through interaction with the C1q complement receptor (13,14). These properties of the core protein suggest that, together with host cell factors, it may contribute to the pathogenesis of HCV infection.
Several human viruses have evolved mechanisms for decreasing the efficacy of the host immune response and interfering with viral clearance (15)(16)(17)(18). Human cytomegalovirus, herpes simplex virus (HSV), varicella zoster virus, and Epstein-Bar virus encode proteins that bind the Fc region of IgG (19 -22). These Fc␥R-like structures are expressed on viral particles (19) and on the cell surface or in the cytoplasm of infected cells (20,23,24). Virus-encoded proteins with the functional properties of Fc␥Rs may enable the virus to evade immune surveillance by avoiding the effector consequences of antibody binding, such as antibody-dependent cell cytotoxicity, cytokine release, and/or activation of the classical complement pathway (15)(16)(17). Moreover, some of the viral structures with IgG binding properties (such as HSV-1 and pseudorabies virus glycoproteins), when expressed on the plasma membrane of infected cells, might internalize bound antibody molecules after bipolar bridging, thereby promoting viral replication and the spread of infection in an immunized host (25)(26)(27).
Fc␥R-expressing cells, activated via various signaling cascades, lyse IgG-opsonized pathogens or kill IgG-coated cells, endocytose immune complexes, promote antigen presentation and cytokine release and induce the production of pro-inflammatory molecules (28 -31). The Fc␥R-mediated internalization of viral particles and/or viral proteins in antigen-presenting cells may optimize antigen presentation, resulting in an amplification of the immune response to the virus (32). Human Fc␥R (Fc␥RI/CD64, Fc␥RII/CD32, and Fc␥RIII/CD16) differ in structural characteristics, function, and cellular distribution (30,33). A new human receptor for IgG was recently identified, * This work was supported by a grant from Agence Nationale de Recherches sur le SIDA (ANRS) (MK/CP/2002/50). 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.
ʈ neonatal Fc␥R (FcRn), that is structurally related to the major histocompatibility complex class I family (34). FcRn mediates the transcytosis of IgG from serum to bile and protects the internalized IgG from catabolism (35). It is present on mouse (35) and human hepatocytes (36), intestinal epithelial (37), endothelial cells (35), and kidney cells (38). It is thought to be involved in immunosurveillance at epithelial surfaces and liver defense mechanisms against pathogens in biliary luminal fluids (35). It has been suggested that FcRn mediates the entry of neutralizing antibodies into HBV-infected hepatocytes, thereby inhibiting the encapsidation and secretion of HBV virions (36). FcRn is also functionally expressed on macrophages and dendritic cells, where it is thought to modulate antigen presentation (39).
We show here that HCV core protein binds "nonimmune" IgG via the Fc␥ region, in addition to reacting with specific anticore antibodies. Native HCV nucleocapsids purified directly from the plasma of HCV carriers or isolated from putative virions strongly bound nonimmune IgG and its Fc␥ fragments. SPR and ELISA studies of the core-IgG interaction revealed that this interaction resembled that between FcRn and IgG. The Fc␥R-like function of the core protein, expressed on HCV nucleocapsids and/or on HCV-infected cells, would help HCV to circumvent immune defense mechanisms and may therefore be of considerable importance in the immunopathology of HCV infection.

Serum and Plasma Samples
Plasma and serum samples were obtained from volunteer blood donors with normal alanine transaminase (ALT) levels who tested positive for anti-HCV antibodies by MONOLISA anti-HCV PLUS (Bio-Rad) and RIBA (ORTHO Diagnostics) and for HCV RNA by reverse transcription-PCR. The plasma samples were stored frozen at Ϫ80°C.

Isolation of HCV Nucleocapsids from the Plasma of HCV Carriers and from HCV Virions
HCV nucleocapsids were purified from the plasma of HCV carriers containing high levels of HCV RNA (reverse transcription-PCR titer 10 5 -10 7 ) of genotype 1a or 1b, as previously described (1). HCV nucleocapsids were also isolated by detergent treatment from putative HCV virions as previously described (1). Solid phase ELISA was used to test the fractions of the gradient for the presence of HCV core antigen.

Production of HCV Core Protein and "Nucleocapsid-like" Particles in Insect Cells
HCV core protein and nucleocapsid-like particles were purified from insect cells (Spodoptera frugiperda, Sf9) infected with a recombinant baculovirus bearing genes for structural HCV proteins, as previously described (1). Recombinant baculovirus was kindly provided by J. Liang (National Institutes of Health, Bethesda, MD). The insect cells were infected as described by Baumert et al. (40).

Recombinant Core Proteins and Synthetic Core Peptides
Recombinant HCV core protein NC 360 aa 1-120 was from Bio-Rad. This protein was produced in Escherichia coli and purified to 94% purity. Recombinant core protein aa 2-169 carrying a His 6 tag fused to the C terminus was produced in E. coli and purified by nickel-nitrilotriacetic acid-agarose chromatography and reverse phase high pressure liquid chromatography on a VYDAC C8 column. 2 HCV-core protein aa 2-122 was obtained by inserting the corresponding DNA fragment into the piVex 2.4a expression vector. A plasmid encoding the entire core protein (kindly provided by G. Inschauspé) was digested and used as a template for PCR amplification with primers 5Ј-AATAGACCGTGCGGCCGCAGCACGATTCCCAAA-3Ј and 5Ј-GCCGCACGTAAGGGATCCTAAAAGCTTACCCAA-3Ј. The amplicon was inserted between the NotI and BamHI sites of piVex 2.4a, and the resulting plasmid was used to transform E. coli BL21. HCV core protein was purified in native conditions on a nickel-nitrilotriacetic acid-agar-ose column (Qiagen). Synthetic peptides covering amino acids 1-169 of the core protein were kindly provided by A. Kolobov (St. Petersburg) and J.-F. Delagneau (Bio-Rad).

Monoclonal and Polyclonal Anti-core Antibodies
The anti-core monoclonal antibody (mAb) VT was obtained from Valbiotech (Paris, France). mAb ACAP-27 was kindly provided by J.-F. Delagneau (Bio-Rad). Polyclonal anti-HCV antibodies (HCIG) were a globulin fraction prepared from the serum of HCV carrier positive for anti-core, anti-NS3, and anti-NS4 antibodies in the Abbott HCV enzyme immunoassay. These antibodies were kindly provided by Ali Fattum (Nabi Biopharmaceuticals, Rockville, MD).
Fab and Fc fragments from anti-core mAb ACAP-27 were prepared and purified using an Immunopure® Fab preparation kit from Pierce. The Fab and Fc fragments were found to be 99% pure by SDS-PAGE. They were concentrated and dialyzed using Ultrafree concentrators from Millipore and stored in 20 mM NaPi, pH 7.0, 5 mM EDTA.
Purified polyclonal human IgG, IgA, IgM, Fab, and Fc␥ fragments of human IgG and Fc5 fragments isolated from human myeloma IgM, either unlabeled or labeled with horseradish peroxidase (HRPO), were obtained from Rockland Immunochemicals (Gilbertsville, PA).
A recombinant monoclonal human IgG (huIgG) that binds human Fc␥RII and Fc␥RIII through its Fc domain was kindly provided by Dr. C. de Romeuf (Laboratoire Français du Fractionnement et de Biotechnologie, Lille, France). Mouse anti-human Fc␥RII mAb IV.3 and mAb AT 10 and anti-human Fc␥RIII mAb 3G8 were purified from cell culture supernatants by affinity chromatography on protein A-Sepharose. Complement C1q protein from human serum was obtained from ICN Biochemicals (Aurora, OH), and soluble protein A from Staphylococcus aureus was purchased from Rockland Immunochemicals.

ELISA for the Detection of HCV Core Antigen
All of the ELISA procedures were carried out using mAbs VT and ACAP-27, as previously described (1).

Immunoglobulin Binding Assay
The binding of Igs of various isotypes and their Fab or Fc fragments to recombinant core protein was analyzed by solid phase ELISA, using recombinant core proteins aa 1-120, aa 2-169, and aa 2-122. Synthetic HBV pre-S1 protein aa 21-47 and HCV E2 protein were used as controls. The plates were coated with the proteins at a concentration of 1 g/ml in PBS. The wells were then saturated with 3% BSA, 0.05% Tween 20 in PBS and incubated with 100 l of a purified preparation of human Igs of various isotypes (IgG, IgM, and IgA) or their fragments (Fab, Fc␥, and Fc5) conjugated with HRPO. Bound Ig (or their fragments) were detected using tetramethyl benzidine as the enzyme substrate by determining the absorbance at 450 nm with a Titertek Multiscan ELISA reader.
The Ig binding properties of HCV nucleocapsids isolated directly from serum or liberated from putative HCV virions and of HCV core protein/nucleocapsid-like particles produced in baculovirus-infected insect cells were analyzed by solid phase ELISA. Fractions of CsCl gradient were either directly added to plates for coating or bound by intermediates of antibodies. For direct coating, the wells of ELISA plates were incubated with 100 l of the fraction overnight at 4°C, washed, and blocked with 3% BSA, 0.05% Tween 20 in PBS. For indirect ELISA, the plates were incubated first with anti-core mAbs or with nonimmune Ig and then with fractions of the gradient. The plates were washed, and 100 l of HRPO-labeled anti-core mAb ACAP-27 or HRPO-labeled "nonimmune" Igs of various isotypes or their fragments were added to the wells, which were then incubated for 2 h at 37°C. The plates were developed and read as described above.

Mapping of the Fc␥-binding Site on HCV Core Protein
Synthetic core peptides corresponding to fragments of HCV core were used to coat ELISA plates at a concentration of 1 g/ml. The plates were saturated with 3% BSA, 0.05% Tween 20 in PBS. HRPO-labeled Igs of various isotypes (IgG, IgM, and IgA) or their HRPO-labeled Fab and Fc fragments were incubated with the peptide-coated plates. The reaction was developed and read as described above.

Mapping of the Core Protein-binding Site on the IgG Molecule
Inhibition ELISA-To define the region of the IgG molecule binding HCV core protein, inhibition ELISA was carried out using soluble SpA and human C1q as inhibitors of IgG or Fc␥ fragment binding to the recombinant core protein NC-360 aa 1-120. All of the potential inhibitors were first incubated with HRPO-labeled IgG or Fc␥ at a concen-tration of 0 -50 g/ml, for 2 h at room temperature. The mixture was then added to the wells of ELISA plates coated with recombinant core protein (1 g/ml in PBS). ELISA was then carried out as described above.
Competitive Immunofluorescence Assay-To compare the site of HCV core protein interaction on the IgG molecule with binding sites for human Fc␥ receptors, competitive immunofluorescence assays were carried out with cells expressing these Fc␥Rs.
Recombinant core protein aa 2-122 was incubated at a concentration of 50 g/ml with 25, 50, or 10 g/ml purified monoclonal human IgG (huIgG) for 30 min on ice. The mixtures were then incubated on ice with 5 ϫ 10 5 indicator cells (K562, IIA1.6-huFc␥RIIB1, or Jurkat-huFc␥RIIIA/␥, respectively) in ice-cold phosphate-buffered saline containing 0.5% bovine serum albumin (PBS-BSA) for 30 min. The cells were washed once with PBS-BSA and incubated with FITC-conjugated mouse F(ab)Ј 2 anti-human IgG (HϩL) (Jackson ImmunoResearch Laboratories, West Grove, PA) for 30 min on ice. The cells were washed several times, and the binding of monoclonal huIgG to indicator cells was analyzed by flow cytometry with a FACScan (Becton Dickinson, Mountain View, CA) using Cell Quest Pro software. In control experiments, Fc␥RIIA, Fc␥RIIB ϩ , or Fc␥RIIIA ϩ expressing cells were incubated with 1 g/ml anti-Fc␥RIIA (mAb IV-3), 1 g/ml anti-Fc␥RIIA/IIB (mAb AT10), or 5 g/ml anti-Fc␥RIIIA (mAb 3G8) antibody for 30 min on ice to block IgG-binding sites. Then 25, 50, or 10 g/ml of monoclonal huIgG and cells were added and incubated on ice for 30 min. The binding of huIgG alone and in the presence of competitors was assessed as described above.

SPR Analysis
Real time binding experiments were performed on a BIAcore 2000 and BIAcore 1000 Upgrade biosensor system (BIAcore, Uppsala, Sweden). All of the experiments were performed at 25°C at a flow rate of 20 l/min. The eluent consisted of 10 mM HEPES at either pH 7.4 or 6.0, 150 mM NaCl, 50 M EDTA, and 0.005% surfactant P20. The dispenser buffer consisted of 10 mM HEPES, pH 7.4, 150 mM NaCl, 3 mM EDTA, and 0.005% surfactant P20. Recombinant core protein aa 2-169 was immobilized via its C-terminal His 6 tag on the surface of an nitrilotriacetic acid sensor chip previously activated with 20 l of 500 M NiCl 2 . The levels of immobilization, expressed in resonance units (RU), are indicated in the legends to the figures. VT or ACAP-27 monoclonal anti-core antibodies or nonimmune human Igs (at concentrations of 10 -100 g/ml) were used. Purified Igs of various isotypes (IgG, IgM, and IgA) and their Fab, Fc␥, or Fc5 fragments were injected in dispenser buffer, pH 6.0. Changes in surface concentration resulting from interaction of the antibody with surface-fixed antigen were detected as an optical phenomenon affecting the SPR signal, expressed in RU where 1 RU corresponds to an immobilized protein concentration of 1 pg/mm 2 . At the end of each cycle, immobilized core protein and bound proteins were removed by injecting 10 l of a regeneration solution consisting of 0.35 M EDTA, 0.05% SDS.

Reverse Transcription-PCR for Determination of HCV RNA
HCV RNA was determined by nested PCR, based on amplification of the cDNA from the core region of the viral genome as previously described (1).

Western Blot
The binding of IgG and Fc␥ to HCV core protein in denaturing conditions was investigated using HCV core protein produced in insect cells infected with recombinant baculovirus. Infected Sf9 cells were collected 48 h after infection, washed, and solubilized in 2% SDS, 5% 2-mercaptoethanol in Tris/HCl, pH 6.8, for 2 min at 100°C. The samples were then subjected to electrophoresis in 12% polyacrylamide gels and electroblotted onto nitrocellulose membranes. The membrane strips were incubated overnight at 4°C with 5% skimmed milk and 0.1% Tween 20 in PBS, washed, and incubated for 1 h at 37°C with anti-core mAbs ACAP-27 or VT and diluted in 1% skimmed milk powder in PBS followed by HRPO-labeled anti-mouse IgG (H ϩ L) (Fab)Ј 2 fragments (Amersham Biosciences). For investigation of the binding properties of IgG and Fc␥, the blots were incubated with HRPO-labeled IgG or Fc␥ fragments diluted in 1% skimmed milk powder in PBS. The blots were rinsed and developed with the ECL detection system (Amersham Biosciences).

Native HCV Nucleocapsids Bind Nonimmune IgG and Fc␥
Fragments-Studies of HCV nucleocapsids naturally occurring in the serum of HCV-infected individuals (1) have provided several lines of evidence that these viral particles, in addition to reacting with specific anti-core antibodies, may bind nonimmune globulins. If plasma samples from HCV carriers were subjected to isopycnic centrifugation in CsCl gradients, frac-FIG. 1. IgG binding properties of native HCV nucleocapsids. A, IgG binding activity of HCV nucleocapsids purified from the plasma of HCV carriers by isopycnic centrifugation in CsCl density gradients. HCV core antigen was detected at a density of 1.28 -1.30 g/ml, by ELISA, with mAb ACAP-27 used as the capture antibody and HRPOlabeled mAb ACAP-27 used as the tracer. IgG binding was detected using nonimmune IgG for coating ELISA plates and mAb ACAP-27 as the tracer. B, binding of nonimmune IgG and Fc␥ by HCV nucleocapsids isolated from HCV virions. An aliquot of a fraction of the gradient (shown in A), banding at a density of 1.10 g/ml and corresponding to the HCV RNA peak, was treated with Tween 80 and subjected to CsCl density gradient centrifugation. Fractions of the gradient were used directly to coat the plates, and the presence of HCV nucleocapsids was demonstrated by reactivity with HRPO-labeled mAb ACAP-27. Immunoglobulin binding properties of HCV cores were investigated using HRPO-labeled human IgG, Fc␥, and Fc5 fragments. The results are presented as the P/N ratio (absorption at 450 nm).
tions of the gradient banding at a density of 1.28 -1.30 g/ml and testing positive for HCV core antigen by ELISA also bound nonimmune IgG (Fig. 1A). HCV nucleocapsids were therefore isolated from putative ␤-lipoprotein-associated virions by detergent treatment and assayed for Ig binding capacity. For this, the fraction of the gradient banding at a density of 1.10 g/ml and corresponding to the HCV RNA peak (shown in Fig. 1A) was treated with 0.5% Tween 80 and subjected to centrifugation in the same conditions. The fractions of the gradient were used directly to coat the ELISA plate and were tested for reactivity with anti-core mAbs and with nonimmune Igs. HCV core antigen was detected at the density of 1.28 -1.30 g/ml (Fig.  1B), using anti-core mAbs. Fractions positive for HCV core antigen also showed IgG and Fc␥ binding activity. No binding of Fc5, IgM (Fig. 1B) or IgA or Fab fragments (data not shown) was detected for these fractions. No binding of IgG or Fc␥ fragments was detected in control gradients if normal human serum was subjected to the same centrifugation procedure.
Recombinant Core Proteins Bind Nonimmune IgG and Their Fc␥ Fragments-We investigated the capacity of HCV core protein to bind nonimmune Igs of various isotypes and their fragments, using recombinant proteins of various lengths and synthetic core peptides. Recombinant core protein NC-360 (aa 1-120) bound anti-core mAb and nonimmune human IgG; it also weakly bound IgM but did not bind IgA ( Fig. 2A). Control proteins (HCV E2 protein and HBV pre-S1 protein) did not bind to IgG or Fc␥. HCV core protein interacted with nonimmune IgG via its Fc region, as demonstrated by the binding of Fc␥ fragments and the lack of binding of Fab fragments. Binding to IgG and Fc␥ was dose-dependent and saturable, whereas binding to IgM was not (data not shown). Because native nucleocapsids isolated from serum did not bind Fc5 (Fig. 1B), the weak binding of IgM by recombinant core protein was considered nonspecific, related to the polyspecificity of polyclonal IgM and was not investigated further. His-tagged recombinant core proteins aa 2-169 and aa 2-122 also bound IgG and Fc␥ (Fig. 2B).
Further experiments were carried out to determine whether core protein (and/or nucleocapsid-like particles) produced in insect cells also bound Igs. An extract prepared from baculovirus-infected cells was subjected to CsCl density centrifugation. HCV core antigen was detected at a density of 1.25-1.16 g/ml, by sandwich ELISA (Fig. 3A). The core antigen-positive fractions also bound IgG and Fc␥ but not Fab fragments. Analysis of the core antigen-positive fractions by electron microscopy showed the presence of nucleocapsid-like particles associated with membrane fragments similar to those previously described (1).
We then investigated whether HCV core protein bound IgG and its Fc␥ fragment in denaturing conditions. HCV core protein extracted from Sf9 cells infected with recombinant baculovirus, for which IgG binding capacity was demonstrated by ELISA (shown in Fig. 3A), was subjected to SDS-PAGE and electroblotted onto nitrocellulose membranes. HCV core protein was detected on the blots with mAb VT, which recognizes a linear, denaturation-resistant epitope located in the stretch of amino acids from positions 24 to 37 (Fig. 3B, lane 1). The denaturated core protein did not bind IgG (Fig. 3B, lane 2) or Fc␥ (not shown). These results confirm that the folded conformation of HCV core protein is required for its interaction with the Fc region of IgG and that the IgG binding activity of HCV core protein is distinct from its reactivity with mAbs, recognizing sequential core epitopes.
Mapping of the Fc␥-binding Site on HCV Core Protein-We used a panel of synthetic core peptides to search for the amino acid sequence delineating the Fc␥ receptor-like site on HCV core protein. Efficient binding of IgG and Fc␥ fragments required the presence of a relatively long fragment of the core protein, spanning amino acids 3-75. Shorter core peptides encompassing aa sequences 10 -53, 11-45, 16 -40, and 39 -75 displayed only limited Fc␥ binding capacity (Fig. 4). A similar pattern of reactivity was obtained with nonimmune IgG. Thus, the Fc␥R-like activity of HCV core protein requires a folded conformation and the N-terminal aa sequence spanning amino acids 3-75 is essential for Fc␥R-like activity.
Mapping of the Core Protein-binding Site on the IgG Molecule-We investigated whether the HCV core protein-binding site on the IgG molecule was similar to the binding sites of "classical" human Fc␥Rs by carrying out competitive immunofluorescence assays with cells expressing Fc␥RIIA, Fc␥RIIB, or Fc␥RIIIA. The binding of a monoclonal huIgG to the three types of Fc␥R was assessed in the presence and absence of recombinant core protein aa 2-122. Prior incubation of huIgG with 50, 100, or 200 g/ml of the core protein did not modify its binding to Fc␥RIIA, Fc␥RIIB, or Fc␥RIIIA (Fig. 5A). In control experiments, mAbs directed against the IgG-binding sites of Fc␥RII and Fc␥RIII inhibited the huIgG binding (Fig. 5B). Thus, the binding site of the core protein on the Fc region of the IgG molecule is different from the binding sites for Fc␥RII and Fc␥RIII, which are located in the lower hinge region and the C␥2 domains (43)(44)(45).
To define further the region of nonimmune IgG interacting with HCV core protein, we analyzed the ability of soluble SpA from S. aureus and C1q purified from human serum to inhibit the binding of IgG and Fc␥ to HCV core protein. SpA inhibited (by 40%) the binding of Fc␥ to HCV core protein, whereas C1q had no effect on Fc␥ binding (Fig. 6). Similar results were obtained if human IgG was used instead of Fc␥ fragment. Thus, the region of nonimmune IgG interacting with the HCV core protein differs from the binding site for C1q and the classical Fc␥Rs(IIA/IIB/IIIA) but overlaps with the SpA binding site, which is located in the CH2-CH3 interface.
Analysis of the Interaction of HCV Core Protein with Nonimmune IgG by SPR-For analysis of the molecular interactions of the core protein with nonimmune IgG by SPR, recombinant core protein aa 2-169 was immobilized on the nitrilotriacetic acid sensor chip via its C-terminal His residues. We observed binding of nonimmune IgG to the HCV core protein (Fig. 7A) as well as their Fc␥ fragments, whereas the Fab fragments did not bind to the core protein (data not shown), consistent with the results obtained by ELISA. IgG binding to HCV core protein was more effective and stable at pH 6.0 (1600 RU) (Fig. 7A) than at pH 7.4 (50 RU) (Fig. 7B). The kinetic constants for the interaction of IgG with the immobilized recombinant core protein were measured by surface plasmon resonance with a range of IgG concentration of 13, 33, 66, 660, and 1,320 nM. An apparent dissociation constant (K d app ) of 84 nM was calculated from the dissociation rate constant (k off ) of 5.67 ϫ 10 Ϫ4 s Ϫ1 and the association rate constant (k on ) of 6.78 ϫ 10 3 M Ϫ1 s Ϫ1 .
When anti-core mAb ACAP-27 (1000 and 4200 RU; Fig. 7C) and mAb VT (not shown) recognizing continuous epitopes in aa sequences 40 -53 and 24 -37, respectively (1), and nonimmune IgG were sequentially injected, 150 RU of IgG were bound to the immobilized core protein. In reciprocal experiments, when injected in reverse order, 500 RU (bottom curve) and 1500 RU (top curve) of nonimmune IgG and 3400 RU of anti-core mAb ACAP-27 (bottom and top curves) were bound to the core protein (Fig. 7D). Similar profiles were obtained with polyclonal anti-HCV antibodies from sera from HCV patients (HCIG) and nonimmune IgG.
We investigated further the interaction of Fab fragments prepared from mAb ACAP-27 (Fab ACAP-27 ) and IgG or Fc␥ fragments with the recombinant core protein. Fab ACAP-27 fragments bound to immobilized core protein with a slightly lower affinity (4.87 nM) than full-length ACAP-27 mAb (0.14 nM). Both Fab ACAP-27 (800 RU) and IgG (600 RU) were found to bind (Fig.  7E, bottom curve) when Fab ACAP-27 was used at a concentration of 1 g/ml. At a Fab ACAP-27 concentration of 10 g/ml (3400 RU), no binding of nonimmune IgG to HCV core protein was detected (Fig. 7E, top curve). IgG (2400 RU) and Fab ACAP-27 (1700 RU) FIG. 3. IgG binding properties of HCV core protein produced in insect cells. A, binding of IgG and Fc␥ to HCV C protein produced in insect cells infected with recombinant baculovirus. Cell extract from infected Sf9 cells was subjected to CsCl density gradient centrifugation. HCV core antigen was detected in the fractions by sandwich ELISA with one of the mAbs (VT or ACAP-27) as the capture antibody and HRPO-labeled mAb ACAP-27 as the tracer. IgG binding properties of HCV core protein/particles were investigated using human IgG, Fab, and Fc␥ fragments or specific anti-core mAbs as capture molecules and HRPO-labeled IgG or Fc␥ or Fab fragments as tracers in the various combinations shown on the graph. The results are presented as the P/N ratio (absorption at 450 nm). B, Western blot analysis of IgG binding to HCV core protein. Extract from Sf9 cells infected with recombinant baculovirus was subjected to SDS-PAGE; the proteins were transferred onto nitrocellulose strips and probed with the anti-core mAb VT followed by HRPO-labeled anti-mouse IgG (H ϩ L) F(ab)Ј 2 fragments (lane 1). We investigated IgG and Fc␥ binding by incubating the blots with HRPO-labeled IgG (lane 2). Binding was detected with the enhanced chemiluminescence detection system. The sizes of the molecular mass markers are indicated in kDa on the right. Identical results were obtained if the HRPO-labeled human Fc␥ fragment was used as a probe instead of HRPO-labeled IgG. These experiments showed two possibilities of the binding of anti-core antibodies to the core protein: via their paratopes to the corresponding epitopes and via the Fc␥ region of the IgG to the Fc␥R-like site formed by the N-terminal part of the core protein. DISCUSSION We have previously shown that nonenveloped HCV nucleocapsids occur naturally in the serum of HCV-infected individ-uals (1). In these studies, several observations suggested that native HCV core particles, in addition to reacting with specific anti-core antibodies, bind nonimmune Igs. The binding of human IgG, through its Fc region, to an unknown HCV component present in the serum of HCV-infected individuals has been reported (46). The results reported herein show that HCV nucleocapsids isolated directly from the serum of HCV carriers or liberated from putative HCV virions by detergent treatment bind IgG and their Fc␥ fragments. The binding of IgG and Fc␥ was also demonstrated for recombinant core proteins of various lengths produced in E. coli and in insect cells infected with recombinant baculovirus. Studies with synthetic core peptides showed that a relatively long amino acid sequence spanning residues 3-75 was crucial for the optimal activity of this site. The lack of binding of IgG and their Fc␥ fragments to HCV core protein in denaturing conditions confirmed that a folded conformation is required for IgG binding.
The capacity of HCV nucleocapsids to bind nonimmune IgG and Fc␥ fragments (evaluated by ELISA) was as strong as their capacity to bind specific anti-core antibodies and stronger than IgG binding by recombinant core proteins and their fragments. This was probably due to the native conformation and the polymeric nature of the core protein in these viral particles. The detection of nonenveloped HCV nucleocapsids associated with IgG and IgM as constitutive components of cryoglobulins (47) and deposits in the glomeruli of the kidneys of HCVinfected patients (48) in the absence of HCV envelope proteins is consistent with this notion and suggests its pathological relevance.
Although the role of the Fc␥R-like function of viral proteins remains unclear, such a function appears to be common in the herpes virus family. Like HCV, these viruses establish persistent infection, targeting various host cell reservoirs, and replicate in the context of an immunized host. Therefore, the Fc␥ FIG. 6. Mapping of the HCV core-binding site on the IgG molecule by inhibition ELISA. Inhibition ELISA was carried out using soluble SpA and human C1q. HRPO-labeled Fc␥ fragment was incubated with potential inhibitors and then allowed to react with recombinant NC-360 aa 1-120 core protein, which had been used to coat the ELISA plate. The results are expressed as the percentages of inhibition with respect to control samples (HRPO-labeled Fc␥ preincubated with PBS). Similar results were obtained if HRPO-labeled IgG was used instead of HRPO-labeled Fc␥. binding properties of the core protein and HCV nucleocapsids may also confer an advantage for the virus in terms of survival.
It seems probable that because of its Fc␥R function, HCV core protein can bind anti-core antibodies by "bipolar bridging," as previously described for HSV-1 (19) and pseudorabies virus IgG-binding proteins (27). In this model, the Fab part of antibody molecule (paratope) binds to its antigenic target (epitope), whereas the Fc␥ part of the antibody binds to the Fc␥R-like site on the viral protein. The binding profiles obtained with im-mune and nonimmune IgG and HCV core protein in our SPR studies and ELISA were very similar to those described for HSV-1 (19). Moreover, if Fab fragments were prepared from anti-core mAb ACAP-27, we observed simultaneous binding of both Fab ACAP-27 and nonimmune IgG (or Fab ACAP-27 and Fc␥ ACAP-27 fragments) to immobilized core protein. This raises the possibility that, like the Fc-binding proteins of HSV and pseudorabies virus, HCV core protein may scavenge the Fc␥ domains of anti-core antibodies after the binding of their para- FIG. 7. SPR analysis of the interaction of HCV core protein with IgG. For all experiments, 2500 -3500 RU of recombinant C protein aa 2-169 was immobilized on nitrilotriacetic acid sensor chips via C-terminal His residues. A and B, binding of nonimmune human IgG (50 g/ml) to HCV core protein immobilized on the sensor-chip at pH 6.0 (A) and pH 7.4 (B). C, binding of anti-core mAb ACAP-27 (1 and 5 g/ml; bottom and top curves, respectively) and IgG (50 g/ml) when sequentially injected. D, binding of nonimmune IgG (50 g/ml and 1 mg/ml; bottom and top curves, respectively) and mAb ACAP-27 (50 g/ml) when injected in reverse order. E, binding of Fab fragments of mAb ACAP27 (1 and 10 g/ml; bottom and top curves, respectively) and nonimmune IgG (50 g/ml) to immobilized core protein. F, binding of Fab fragments of mAb ACAP-27 (50 g/ml) after injection of human IgG (50 g/ml). topes to their antigenic target, thereby interfering with the effector functions mediated by the Fc␥ domain of the bound antibody. Folded conformation and the polymeric nature of the core protein in the HCV nucleocapsid would increase the possibility of bipolar binding of anti-core antibodies in vivo.
Strikingly, the HCV core-IgG interaction displays two major features previously reported for a novel human Fc␥R-FcRn. First, the interaction of the core protein with IgG is markedly inhibited by SpA. The SpA-binding site is located in the CH2-CH3 interface region of the IgG (49). This region of the IgG molecule also interacts with herpes simplex virus IgG-binding proteins gE-gI (50) and binds to FcRn, which competes with SpA for binding to the IgG molecule (49,51). Second, our SPR studies demonstrated that the HCV core protein interacted much more efficiently with IgG and Fc␥ fragments at pH 6.0 than at neutral pH. Thus, the recombinant core protein-IgG interaction shows a pH dependence similar to that observed for the FcRn-IgG interaction, which is more effective at low pH (6.0 -6.5) than in neutral pH (34,35,39,52). FcRn was recently identified as a candidate receptor mediating endocytosis of IgG in rat (35) and human hepatocytes (36). FcRn is expressed on the canalicular membranes of hepatocytes, where it is thought to be involved in delivering immune complexes from the canalicular space to Kü pffer cells and bile (35) and in the binding of IgG at the hepatocyte surface (36). IgG binding to FcRn is followed by endocytosis of the complex in the acidic endosome and its transport through cellular conduits, resulting in the final release of IgG into the extracellular fluids (35). It was recently suggested that FcRn mediates the cellular uptake of neutralizing anti-HBV antibodies in hepatocytes, which neutralize the virus within the cell, thereby interfering with the secretion of HBsAg and infectious HBV (36). FcRn is also expressed on cells of the immune system such as monocytes, tissue macrophages, and dendritic cells (39), where it may affect Ag presentation (39) and deliver antigens to intestinal cells, inducing either tolerance or immune activation (34). It is tempting to suggest that the HCV core protein and/or circulating HCV nucleocapsids may mimic FcRn functions, thereby interfering with some of the mechanisms mediated by interactions between the Fc region of the IgG antibody molecule and this particular receptor.