SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Two determinants of infectivity have been identified in the hepatitis B virus (HBV) envelope proteins: a pre-S1 receptor-binding site and an uncharacterized determinant in the antigenic loop (AGL), which is structurally related to the antigenic a-determinant. Infection would proceed through virus attachment to cell surface heparan sulfate (HS) proteoglycans (HSPGs) before pre-S1 engages a specific receptor for uptake. Using heparin binding and in vitro infection assays with hepatitis D virus as a surrogate for HBV, we established that HS binding is mediated by the AGL. Electrostatic interaction was shown to depend upon AGL residues R122 and K141, because their substitution with alanine modified the virus net-charge and prevented binding to heparin, attachment to hepatocytes, and infection. In addition to R122 and K141, the HS binding determinant was mapped to cysteines and prolines, which also define the conformational a-determinant. The importance of AGL conformation was further demonstrated by the concomitant loss of a-determinant and heparin binding upon treatment of viral particles with membrane-impermeable reducing agent. Furthermore, envelope proteins extracted from the viral membrane with a nonionic detergent were shown to conserve the a-determinant but to lose heparin affinity/avidity. Conclusion: Our findings support a model in which attachment of HBV to HSPGs is mediated by the AGL HS binding site, including only two positively charged residues (R122 and K141) positioned precisely in a three-dimensional AGL structure that is stabilized by disulfide bonds. HBV envelope proteins would individually bind to HS with low affinity, but upon their clustering in the viral membrane, they would reach sufficient avidity for a stable interaction between virus and cell surface HSPGs. Our data provide new insight into the HBV entry pathway, including the opportunity to design antivirals directed to the AGL-HS interaction. (HEPATOLOGY 2013)

Worldwide, hepatitis B virus (HBV) affects more than 350 million chronically infected individuals whose infection may progress to severe liver disease, including cirrhosis and hepatocellular carcinoma.1 HBV has a very narrow host range and a prominent hepatotropism that, for the most part, is determined at viral entry. Although much information has been gathered in recent years in many aspects of the HBV replication cycle, the HBV entry pathway has not been fully explored due to cellular receptors that await identification.2

Three types of membrane-associated glycoproteins are present in the envelope of HBV virions, each bearing the HBV surface antigen (HBsAg): (1) the small envelope protein (S-HBsAg); (2) the middle protein (M-HBsAg), which differs from S-HBsAg by an N-terminal pre-S2 ectodomain; and (3) the large protein (L-HBsAg), which includes a further N-terminal pre-S1 extension. Surprisingly, the envelope proteins are produced in amounts far exceeding the need for virion assembly, leading to the production of empty subviral particles (SVPs) that outnumber virions by more than 10,000:1. In addition, the HBV envelope proteins have the capacity to interact with hepatitis D virus (HDV) ribonucleoprotein in case of HBV/HDV coinfection,3, 4 leading to the assembly of HDV ribonucleoprotein-containing virions coated with HBV envelope proteins.4, 5 HBV and HDV envelopes may differ from each other with regard to the relative amount of L-HBsAg, estimated at up to 25% of the total surface proteins for HBV6 and only 5% for HDV.7 The envelope proteins fulfill identical functions at the surface of HDV and HBV, including binding to cell surface receptors for promoting viral entry.8-10 The HDV model can thus be used as a surrogate for HBV to study the HBV envelope protein functions at viral entry; in fact, it offers several interesting technical advantages.11

Infectivity of HBV or HDV depends upon at least two elements of the envelope proteins: (1) a receptor binding site located within the N-terminal pre-S1 domain of L-HBsAg,7 which is responsible for tissue specificity (pre-S1–specific synthetic peptides are potent inhibitors of viral entry, with therapeutic potential12), and (2) an infectivity determinant in the surface-exposed antigenic loop (AGL), a polypeptide present in the S domain of all envelope proteins.13 The AGL function has not been established, but its activity was shown to depend upon cysteine residues involved in structuring the AGL-associated a-determinant.13 The a-determinant was the first discovered HBV marker; it is a conserved immunodominant determinant bearing most of the HBV neutralizing epitopes, with which no essential function had yet been associated. In a previous study, we mapped the AGL infectivity determinant to a set of conserved residues—cysteines and noncysteines—that are predicted to cluster together in a disulfide-bridges network that underlies the a-determinant.9 Interestingly, two conserved, positively charged residues (R122 and K141) were identified as crucial to infectivity. Positively charged amino acid side chains, if properly exposed in a polypeptide, can eventually mediate attachment to negatively charged molecules such as cell surface glycosaminoglycans (GAGs). In fact, HBV virions bind immobilized heparin, a structural homologue of heparan sulfate (HS),14 and HBV entry initiates with virus attachment to HS proteoglycans (HSPGs).15, 16 This attachment was described as dependent upon L-HBsAg,16 but the direct contribution of pre-S1 in GAG binding was not demonstrated. In a more recent study, binding of HBV SVPs to immobilized heparin was shown to be independent of L-HBsAg,17 leaving open the possibility for a role of the AGL in mediating attachment to GAGs.

Cell surface HSPGs fulfill many biological functions by interacting with a variety of proteins, and they are used by many viruses for attachment to their target cells.18 In this study, we hypothesized that R122 and K141 in the AGL of HBV envelope proteins are crucial to mediating the initial attachment to cell surface HSPGs. We created viral particles bearing mutants of the HBV envelope proteins, and evaluated their affinity for heparin, their cell binding capacity and in vitro infectivity. Our results are in agreement with an HS binding site centered at R122 and K141 within the conformational, surface-exposed AGL of the HBV envelope proteins.

Materials and Methods

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Production and characterization of HDV and HBV particles were performed as described.13, 19, 20 HDV infection assays were conducted as reported.19 Cell binding assays are described in detail in the Supporting Information.

Trypsin Digestion of HBV Particles.

Preparations (200 μL) of 10 × HBV particles (see Supporting Information for description of the purification procedure) were mock treated or treated with 1,800 μL of 0.05% trypsin (Gibco) for 1 hour at 37°C. After treatment, 1 mL of each sample was subjected directly to heparin affinity chromatography, and 1 mL was subjected to sedimentation by centrifugation for 2 hours at 50,000 rpm in an SW55 rotor (Beckman) on a 1-mL, 30% sucrose cushion in 10 mM Tris-HCl (pH: 7.4), 150 mM NaCl, 1 mM EDTA. After centrifugation, the particle-containing pellet was resuspended in 20 mM Tris-HCl (pH 7.4), 140 mM NaCl and subjected to western blot analysis for the detection of envelope proteins, and to ETI-MAK-4 HBsAg enzyme-linked immunosorbent assay (ELISA) for detection of the a-determinant prior to heparin affinity chromatography.

Heparin-Affinity Chromatography.

Prior to heparin chromatography, samples were normalized for HBV envelope protein concentration via western blotting and ELISA. Chromatography was performed using new 1-mL bed HiTrap Heparin sepharose HP columns (GE Healthcare) that were run in parallel using a peristaltic pump equipped with a multichannel pump head. Each column was washed with 10 mL of binding buffer (20 mM Tris-HCl [pH 7.4], 150 mM NaCl). One milliliter of sample (10-fold concentration of cell culture–derived HDV or HBV particles in binding buffer, trypsin-digested serum-derived particles) was loaded on the column, and the column was washed with 5 × 1 mL of binding buffer. The flow rate was controlled with the peristaltic pump at 0.5 mL/min as described.14 We used conditions in which the sample was applied twice to the column (except for the experiment reported in Fig. 3, in which the sample was applied only once) before collection of the flow-through, washes, and eluants. One-milliliter flow-through was collected, followed by 5 × 1-mL washes in binding buffer. Elution was performed with 6 × 1 mL of 20 mM Tris-HCl (pH 7.4), 500 mM NaCl. After elution, 1 ml of 20 mM Tris-HCl (pH 7.4), 2 M NaCl was applied. All fractions, including flow-through and 2 M NaCl elution were analyzed via ELISA for detection of HBsAg, northern blot analysis for detection of HDV RNA, or Southern blot analysis for detection of HBV DNA.19 Typically, HBsAg or RNA signals in the 2 M NaCl fraction amounted for less than 5% of the total eluate.

Results

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

As reported, there are technical advantages in using the HDV model as a surrogate for HBV for in vitro study of the HBV entry mechanism.9 Because infection of human hepatocytes with HBV was shown to depend on an initial attachment of virions to cell surface–associated HSPGs,15, 16 we sought to determine whether this holds true for HDV. HDV infection assays were thus conducted in HepaRG cells in the presence of HS or dextran sulfate (DS) as soluble forms of GAGs. The results show that both HS and DS were able to block infection in a dose-dependent manner (Supporting Fig. 1), in agreement with HDV entry depending on an initial attachment to cell surface HSPGs.

Removal of the Pre-S Domains at the Surface of HBV Particles by Trypsin Digestion Does Not Abrogate Heparin-Binding.

According to Schulze et al.,16 HBV affinity for heparin would require the integrity of the pre-S domain of L-HBsAg. This conclusion was based on the fact that trypsin digestion of HBV virions, aimed at removing the surface-exposed pre-S polypeptides, resulted in a loss of heparin binding activity. However, in that particular experiment, the integrity of the viral particles—and that of the AGL structure, in particular—after exposure to trypsin were not documented. In the present study, a preparation of purified, serum-derived, HBV particles was mock-treated or treated with 0.05% trypsin at 37°C for 1 hour. After treatment, we verified that the pre-S1 and pre-S2 polypeptides were removed (Fig. 1A, left panel) and the S-HBsAg proteins were unaffected (Fig. 1A, middle panel). A small amount of S-HBsAg cleavage product was observed at an apparent molecular weight of 15 kDa (Fig. 1A). Because the R247 antibody used for immunoblotting recognizes a linear epitope encompassing the Gln54-to-Ser64 sequence in S-HBsAg, cleavage is likely to have occurred at R122 in the AGL and not at K141. This is suggested by two observations: (1) the size (15 kDa) is compatible with S-HBsAg polypeptide 1-122 and (2) cleavage at carboxyl side of lysine-141 would not be efficient for the presence of proline-142. We ascertained that the a-determinant of treated particles was conserved using ELISA based on a conformation-dependent anti-HBsAg monoclonal antibody (A1.2) (Fig. 1A, right panel). When the two preparations were compared for their capacity to bind immobilized heparin, we observed no difference between mock- and trypsin-digested particles, and this was true for both virions (HBV DNA signals) and SVPs (HBsAg ELISA). We concluded that the pre-S domain does not directly contribute to HBV binding to heparin. Note that virions always appear to bind more efficiently to heparin (DNA signals) than SVPs (HBsAg ELISA) (Fig. 1B). This likely reflects a better heparin avidity of virions over SVPs, a consequence of their respective size.

thumbnail image

Figure 1. Removal of pre-S polypeptides at the surface of HBV particles by trypsin digestion does not prevent binding to immobilized heparin. Preparations of purified HBV particles in 10 mM Tris-HCl (pH: 7.4), 150 mM NaCl, 1 mM EDTA were mock- or trypsin-digested for 1 hour at 37°C. (A) Digestion of the pre-S1 and pre-S2 polypeptides was evidenced by immunoblotting using an anti–pre-S2 antibody. The integrity of the S-HBsAg proteins was monitored using an anti-S (R247).13 The integrity of the a-determinant was evidenced by ELISA using a conformation-dependent anti-HBsAg monoclonal antibody (A1.2). (B) Mock- and trypsin-digested particles were subjected to heparin affinity chromatography as described in Materials and Methods. Each fraction, including flow-through (fl), washes, and eluants, was tested for HBsAg, and for HBV DNA via Southern blot analysis. The results are expressed as percentages of the total recovered material.

Download figure to PowerPoint

Binding of SVPs or HDV Virions to Immobilized Heparin or Target HepaRG Cells Is Not Dependent upon the Large HBV Envelope Protein.

In a recent study, Chai et al.17 presented data supporting that SVP binding to heparin is independent of L-HBsAg. Here, we produced HDV virions coated with S-HBsAg only (S-HDV), or S-, M-, and L-HBsAg (SML-HDV) and demonstrated that both could bind to immobilized heparin as measured via northern blot analysis for the detection of HDV RNA (Fig. 2A). Present also in the preparations were S- and SML-SVPs, the heparin affinity of which was monitored using an HBsAg ELISA. We concluded that both HBV SVPs and HDV virions could bind heparin in an L-HBsAg–independent manner.

thumbnail image

Figure 2. Binding of SML-HDV and S-HDV particles to immobilized heparin or HepaRG cells. (A) Preparations of particles (10-fold concentrate) were subjected to heparin affinity chromatography as described in Materials and Methods. Each fraction, including flow-through (fl), washes, and eluants, was tested for HBsAg and for HDV RNA. The results are expressed as percentages of the total recovered material. (B) HepaRG cells were cultured in 6-well plates (1 × 106 cells/well) as described.20 Cells were exposed to inoculum containing approximately 108 genome equivalents, in the presence of 4% polyethylene glycol for 16 hours. For the attachment assay, cells were washed three times with PBS postexposure and lysed in RNA extraction buffer. RNA extracted from 70 μL of inoculum (ino) or 6.6 × 104 cells (cell) were analyzed for genomic HDV RNA. Ribosomal RNA (rRNA) signals from nonspecific hybridization served as loading controls. For infection assay, cells were harvested at day 8 postinoculation, and HDV RNA was analyzed as described above. Quantification of HDV RNA signals by phosphorimager is indicated as percentages of the SML-HDV value. Error bars indicate the SD of three independent infection assays.

Download figure to PowerPoint

We also exposed HepaRG cells to S- and SML-HDV virions for 16 hours at 37°C. Inoculums were then removed, and cells were washed with phosphate-buffered saline (PBS). Note that the amount of cell-associated HDV RNA at the end of the incubation period reflects the number of internalized virions during exposure in addition to surface-bound virions. Both S- and SML-HDV particles could attach HepaRG cells as evidenced by equivalent levels of cell-associated HDV RNA after inoculum removal (Fig. 2B). When cells were maintained in culture after inoculation, the amount of cell-associated HDV RNA at day 9 postinoculation was used as a marker of infection. As expected, high levels of HDV RNA were detected only in cells inoculated with SML-HDV. These results demonstrate that both S- and SML-HDV can bind to cell surface HSPG, in agreement with an HS binding site located in the AGL.

Alanine Substitution for Basic R122 and K141 Residues of the AGL Abrogates Binding to Immobilized Heparin and HepaRG Cells.

In a previous study, we mapped the AGL infectivity determinant to a set of conserved residues, including two basic residues (R122 and K141). To address the contribution of R122 and K141 in electrostatic interaction with HS, we produced SML-SVPs bearing a substitution of alanine for R122 or K141 or both. Particles were analyzed via immunoblotting and northern blotting and normalized for envelope proteins and HDV RNA (Supporting Fig. 2). When compared with wild-type (WT) particles in the heparin binding assay, both R122A and K141A particles had a reduced binding capacity (Fig. 3A), whereas R122A/K141A particles demonstrated a near to complete loss of heparin affinity, appearing only in the flow-through and column wash fractions. The results demonstrate that R122 and K141 are crucial for heparin binding. To further test the role of positive charges in heparin binding, we prepared particles bearing the D144A or G145R substitutions that have been reported in the literature as arising under immune pressure selection.21 As reported,9 not only were D144A and G145R permissive to infectivity, they were slightly more infectious than WT. In the heparin binding assay, D144A and G145R particles demonstrated an enhanced heparin binding capacity. These results suggest that the overall charge of the AGL polypeptide is crucial to heparin affinity, because removing positive charges (R122A or K141A) reduced binding, whereas adding a positive charge (G145R), or removing a negative charge (D144A), led to a gain of binding activity.

thumbnail image

Figure 3. Analysis of HDV particles bearing substitutions in the HBV envelope proteins for their capacity to bind immobilized heparin or HepaRG cells. (A) Preparations of particles (10-fold concentrate) were subjected to heparin affinity chromatography as described in Materials and Methods. Each fraction, including flow-through (fl), washes, and eluants, was tested for envelope proteins using a pre-S2–specific ELISA.9 The results are expressed as percentages of the total recovered material. (B) HepaRG cells (1 × 106 cells/well) were exposed to 1 mL of inoculum containing 108 genome equivalents for 16 hours at 37°C. Inoculums consisted in HDV particles bearing the indicated substitutions in the AGL. After exposure, cells were washed with PBS to remove unbound virus and were lysed for RNA extraction. Cellular messenger RNAs (mRNAs) and genomic HDV RNA were captured together, with a mixture of biotinylated oligo (dT) and HDV RNA-specific oligonucleotides. HDV RNA from 70 μL inoculum (ino) and captured RNAs from 0.5 × 106 cells (cell) were analyzed for genomic HDV RNA and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) mRNA. Quantification was achieved using a phosphorimager. Histograms show the amounts of cell-associated HDV RNA at day 1 postinoculation (attachment assay), or at day 8 postinoculation (infection assay). Histogram values are expressed as a percentage of WT SML-HDV normalized to GAPDH mRNA. Error bars indicate the SD of three independent infection assays.

Download figure to PowerPoint

To determine whether virion attachment to susceptible cells also depends on AGL-positive charges, we exposed HepaRG cells to SML-HDV bearing individually the R122A, K141A, R122A/K141A, D144A, or G145R substitutions. D144A- and G145R-HDV demonstrated an increased cell binding capacity (2.8- and 3-fold increase, relative to the WT, respectively), whereas the R122A and K141A single or double mutants were highly deficient (<5% relative to the WT) (Fig. 3B). In the infection assay, D144A- and G145R-HDV demonstrated a gain of infectivity (1.5- and 1.4-fold relative to WT, respectively).9 Therefore, increasing the positive net charge of the AGL (D144A or G145R) led to a gain of cell binding capacity and, to a lesser extent, increased infectivity. These results are evidence that positive charges of R122 and K141 in the HBV envelope proteins' AGL are involved in HSPG binding.

Surface Exposed R122 and K141 Residues of the HBV Envelope Proteins Are Crucial to HBV Virions Binding to Heparin.

To investigate whether R122 and K141 mediate heparin binding for HBV virions, we produced HBV bearing G145R or R122A/K141A envelope proteins and measured their relative affinity to immobilized heparin. The behavior of virions and SVPs in the heparin binding assay was monitored via Southern blot analysis and HBsAg ELISA for the detection of HBV DNA or envelope proteins, respectively. Note that DNA extraction was performed from virions immunoprecipitated with anti–pre-S1 antibodies to exclude nonenveloped NCs. Under the indicated affinity chromatography conditions (Fig. 4), 79% of WT virions and 43% of WT SVPs could bind immobilized heparin. As expected, substitution of alanine for R122 and K141 was inhibitory to HBV (virions and SVPs) affinity for heparin, whereas G145R substitution has the opposite effect. Note that, virions always demonstrated a significantly higher heparin binding capacity than did SVPs. This is likely related to avidity, itself a consequence of particle size. Overall, these results are evidence for the AGL acting as the major determinant of HBV binding to HSPGs through the contribution of positively charged residues R122 and K141 and eventually arginine at position 145 in the case of the G145R variant.

thumbnail image

Figure 4. Heparin binding of HBV particles bearing R122A-K141A or G145R substitutions in the envelope proteins. Preparations of HBV particles (100-fold concentration) were subjected to heparin affinity chromatography as described in Materials and Methods. Each fraction, including flow-through (fl), washes, and eluants, was tested for envelope proteins using a pre-S2–specific ELISA,9 and for HBV DNA via Southern blot analysis. The results are expressed as percentages of the total recovered material.

Download figure to PowerPoint

Using a native agarose-gel electrophoresis approach, we demonstrated that R122 and K141 were indeed solvent-exposed and were major contributors to the global charge of HBV particles (Supporting Results and Supporting Fig. 2).

The AGL Heparin-Binding Determinant Maps to a Set of Conserved Residues That Are Essential to Infectivity and the a-Determinant.

To provide a complete account of the AGL residues involved in heparin binding, we used a panel of previously generated SML-SVPs9 bearing alanine substitution for each of the noncysteine residues in the AGL sequence (positions 101-172 of the S domain) or serine substitution for cysteines. Expressed at the surface of HDV virions, the mutant envelope proteins served to map the AGL infectivity determinant to amino acid residues that are also essential to the a-determinant.9

We used the panel of normalized SVPs in the heparin binding assay (Fig. 5A). Cysteine-to-serine substitutions that were shown to be detrimental to infectivity and the a-determinant were inhibitory to heparin binding, whereas most of the alanine substitutions were tolerant, except for P108A, I110A, P120A, T123A, P142A, I150A, and W156A, which, in addition to R122A and K141A, reduced binding to <25% of that of the WT. These results suggest that in addition to positive charges, the AGL conformation, which is dependent upon cysteine disulfide bonds and proline residues, plays an essential role in defining the HSPG binding site. When the histogram for heparin affinity (Fig. 5A) is compared with that of the infectivity determinant (Fig. 5B) that has been established,9 most of the infectivity-deficient substitutions appeared to also impair heparin binding. However, there are a few exceptions to this observation: P105A and V106A, though detrimental to infectivity, had no impact on heparin binding. More interestingly, the N146A glycosylation-defective mutation that impairs infectivity increased heparin affinity. The loss of infectivity for N146A was previously interpreted as resulting from the amino acid side chain substitution, not from the lack of N-linked carbohydrates, since nonglycosylated N146T retained infectivity.22 Based on the fact that the heparin binding and infectivity determinants closely match, one can conclude that the main function of the AGL infectivity determinant resides in providing attachment to HSPGs. Furthermore, this genetic analysis confirms that, in addition to negative charges, the three-dimensional structure of the AGL is crucial to heparin binding.

thumbnail image

Figure 5. Heparin binding of SML-SVPs bearing single amino acid substitutions in the HBV envelope proteins. Preparations of SML-SVP mutants normalized for envelope proteins with a pre-S2–specific ELISA were subjected to heparin affinity chromatography as described in Materials and Methods. A total of 400 μL of each preparation was applied to a 400-μL heparin column. Flow-through was discarded, and the column was washed five times with binding buffer. Elution was performed with 400 μL of 20 mM Tris-HCl (pH 7.4), 500 mM NaCl, and eluants were tested for envelope proteins using a pre-S2–specific ELISA. (A) Effect of a substitution on heparin binding activity. (B) Effect on infectivity as reported.9 Values are given as percentages relative to that of the WT. Error bars indicate the SD of three independent assays. Residues essential for heparin binding (A) or infectivity (B) are indicated in black letters. Gray bars indicate amino acids that have not been tested. Serine was substituted to cysteine and alanine to noncysteine.

Download figure to PowerPoint

The Effect of a Membrane-Impermeable Reducing Agent and a Nonionic Detergent on Heparin-Binding.

To further explore the importance of AGL conformation for heparin binding, we conducted a biochemical analysis. We had shown previously that treatment of HDV particles with tris(2-carboxyethyl)phosphine (TCEP), a membrane-impermeable reducing agent, led to an alteration of the AGL cysteine disulfide network and to a loss of both AGL conformation and infectivity, without altering the integrity of the viral particles.13 Here, HDV virions were mock-treated or treated with 1:2 dilutions of 0.5 mM TCEP before they were subjected to heparin binding assay (Fig. 6). Clearly, a 0.03-mM TCEP concentration reduced heparin binding more than three-fold, confirming that in addition to basic residues, the three-dimensional structure of the AGL is essential to heparin binding activity.

thumbnail image

Figure 6. Effects of TCEP and NP-40 on the capacity of HBV envelope proteins to bind immobilized heparin. A total of 10 μL of a 100-fold concentration of SML-HDV particles was mock-treated or treated with 1:2 dilutions of 0.5 mM TCEP or 0.1% NP-40 for 1 hour at 37°C. Preparations were diluted 100-fold in 20 mM Tris-HCl (pH 7.4), 150 mM NaCl and subjected to (1) a pre-S2–specific ELISA (anti–pre-S2); (2) a conformational a-determinant–specific ELISA using the A1.2 anti-HBsAg monoclonal antibody37; (3) Northern blot analysis for the detection HDV RNA as a marker of HDV particle integrity; and (4) heparin affinity chromatography as described in Materials and Methods. Values are given as the percentage relative to that of the untreated particles.

Download figure to PowerPoint

To question the importance of HBV surface proteins clustering in the viral membrane for heparin binding, we treated HDV particles with the nonionic detergent Nonidet P-40 (NP-40) to extract the envelope proteins from the viral membrane while preserving the AGL conformation as described.13 Particle disassembly was evidenced by NP-40 exerting a dose dependent loss of HDV RNA, a marker of HDV virion's integrity. Using conformation-sensitive antibodies, we confirmed that NP-40 treatment, did not affect the a-determinant, but it prevented binding of HBV envelope proteins to heparin in a dose-dependent manner, suggesting that virus binding to heparin is achieved through avidity conferred by the clustering of envelope proteins in the viral membrane.

Discussion

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

HSPGs have been shown to promote HBV attachment to hepatocytes at the initial phase of viral entry.15, 16 In the present study, we show that the same requirement applies to HDV entry, further substantiating the relevance of HDV as a surrogate for HBV to study the role of the envelope proteins in infectivity. We have demonstrated that HSPG binding is dependent upon the AGL determinant we characterized previously as conformational, with structural bases shared by the a-determinant.9 We have shown that attachment of HBV and HDV particles to immobilized heparin, or to cell surface HSPGs, is dependent upon basic residues (R122 and K141), consistent with a binding energy contributed, in part, by electrostatic interactions. HBV variants bearing a more positively charged AGL such as D144A and G145R mutants, demonstrated a three-fold increase in cell attachment, but only a 1.4-fold increase in infection (Fig. 3B). This finding is in agreement with hepatocytes bearing a large number of cell surface HSPGs and a limited number of pre-S1–specific receptors.15, 16 Note, however, that the electrical charge of the AGL of all HBV genotypes (amino acids 100-154) is neutral in including two strictly conserved plus charges at positions 122 and 141 and two minus charges at 100 and 144.23

The G145R substitution—and to a lesser extent, D144A substitution—is found in HBV isolates referred to as “escape mutants” as the result of selection pressure exerted by the immune response. G145R-HBV is able to replicate in individuals in the presence of anti-HBsAg antibodies, and to escape detection by immunoassays for HBsAg. Although partially deficient for particle secretion,24 G145R-HBV can be horizontally transmitted, thereby representing a threat for the population of vaccinated individuals. One can speculate that G145R preferentially emerges under immune pressure because its enhanced infectivity9 (i.e., a greater potential for transmission) compensates for its defect in particle secretion.

Regarding the role of L-HBsAg in binding to heparin, we came to conclusions conflicting with the study of Schulze et al.16 In our hands, when trypsin digestion of pre-S polypeptides from the surface of HBV virions was achieved in conditions that preserved S-HBsAg,25 the AGL, and the a-determinant, heparin binding was preserved (Fig. 1). In contrast, Schulze et al.16 observed that trypsin-digested HBV virions would lose heparin binding capacity. Because the integrity of both the AGL and the a-determinant was not monitored by Schulze et al., the possibility exists that the lack of heparin binding was due not to the removal the pre-S domain, but to an alteration of the AGL structure upon trypsin treatment. In addition, the authors used, as a substrate for trypsin digestion, a preparation of HBV particles that had been purified on a heparin affinity column. It cannot be excluded, then, that heparin binding has induced a conformational change in the HBV envelope proteins, a phenomenon that has been described as part of the entry mechanism for other viruses, particularly the adeno-associated virus serotype 2.26 If HBV binding to heparin induces a nonreversible conformational change of the AGL, heparin-purified virions could display a heparin binding–deficient AGL and, possibly, a protruding pre-S. Loaded a second time on a column, virions would then bind through the positively charged pre-S domain of L-HBsAg as a result of a cation exchange effect and not necessarily a heparin affinity effect. After trypsin digestion of pre-S from the surface of heparin-purified virions, binding would be lost. L-HBsAg accounts for as much as 25% of virion's envelope proteins, and pre-S is clearly positively charged, (an average of +5 to +9, according to genotypes). Hence, pre-S might contribute to virions attachment to immobilized heparin, but it may just be through an electrostatic and not a heparin affinity contribution. In general, there is more to a heparin binding site than a set of basic residues18: three-dimensional structure is important, and binding also involves noncharged residues engaged in hydrogen-bonding, van der Waals forces, and hydrophobic interactions.

The finding that the electrostatic contribution of the AGL is based on only two nonsequential basic residues is somewhat unusual. HS binding sites often consist in linear amino acid sequences, including several positively charged residues27; however, as indicated, they also include noncharged residues, and they may be defined by the association of sequentially distant amino acids in a specific spatial organization upon protein folding and quaternary structure acquisition. Our genetic and biochemical analyses converge in suggesting that the role of the AGL disulfide bonds, at the base of the a-determinant, is to approximate R122 and K141 in a precise folding of the AGL polypeptide, and to cluster a critical number of R122/K141 pairs through multimerization. HBV envelope proteins are known to form dimers that are cross-linked by disulfide bonds,28 and dimers could assemble as tetramers in the asymmetric unit of the SVP structure29, 30 and perhaps as hexamers in the virion's envelope.31 One can speculate that R122/K141 pairs are localized at the interface of the oligomeric structures resulting in the clustering of 8 or 12 surface-exposed positive charges per asymmetric unit.

Overall, the AGL HS binding determinant consists in (1) basic and polar residues and (2) structure-defining cysteines and prolines, distributed in three discrete subdomains: amino acids 118-124, 137-142, and 146-150. Two distal hydrophobic residues, I110 and W156, also contribute to HS binding. Understanding how these elements come together to form the active HS binding site and how a conformationally fixed arrangement of R122 and K141 can line up with specific sulfate groups of the HSPG will require cocrystallization on the interacting partners.

Liver HS chains consist of, on average, 40-60 disaccharides corresponding to a 20- to 30-nm extended length close to the diameter of viral particles. If AGL clustering on the virus surface is necessary for binding to HS chains, the latter must also achieve some degree of clustering on the cell surface to anchor particles to the basal membrane. Syndecan-1, which is abundant at the surface of human hepatocytes, offers this possibility because its ectodomain bears three HS chains. Furthermore, syndecan-1 itself can multimerize at the basal membrane to form clusters that are generally involved in lipoprotein endocytosis.32 Liver-specific HSPGs (including syndecan-1) can undergo endosomal cycling and, hence, might be able to internalize HBV prior to pre-S1 activity.

Interestingly, an AGL-like polypeptide, including the structure-defining cysteine and proline residues and positive charges at positions 122 and 141, is present in the envelope proteins of all Orthohepadnaviruses, suggesting a conserved HS binding function among all genus members.33, 34 In contrast, the envelope proteins of the duck hepatitis B virus (DHBV), an Avihepadnavirus, do not bear an ALG-like polypeptide, implying that DHBV would not use HS for initial attachment to duck hepatocytes, in agreement with the observation that heparin or DS demonstrated no inhibitory effect on DHBV infection.15, 35

Positive charges, conformation and dimerization appear crucial to the AGL HS binding activity, but not sufficient for viral particle attachment. This was demonstrated when HBV envelope proteins were extracted from the particle with a nonionic detergent in conditions that preserved the a-determinant. The detergent-isolated surface proteins no longer attached to immobilized heparin, implying that binding was achieved by avidity acquisition upon protein clustering in the viral envelope. Differences in avidity levels may, at least in part, explain the previously reported selectivity of virions, over SVPs, for binding to immobilized heparin.14 Considering that the HBV virion diameter is approximately twice that of SVP, it should bear four times the number of heparin binding sites, assuming an equivalent surface protein density.29, 31 HBV virions indeed demonstrated a moderate advantage over SVPs in heparin binding, which was clearly independent of L-HBsAg (Figs. 1 and 4). Note that binding of HBV virions to an immobilized heparin column could involve the participation of pre-S if the column were to be used under ion exchange rather than heparin affinity chromatography conditions.

Although SVPs were initially reported to bear only trace amounts of L-HBsAg,36 recent studies estimated an average of 5%-6% of the total number of envelope proteins, a proportion that was shown sufficient to confer infectivity to HDV.7, 17 It is thus reasonable to assume that most SVPs would be able to compete with virions at viral entry, and because SVPs can outnumber virions by at least 4 logs, the interference at viral entry could have important consequences for transmission, propagation in the liver, and pathogenesis. The filamentous form of SVPs would appear as better entry competitor than spherical SVPs because of their larger size. However, their competitive effect may be limited because they outnumber virions only by 1 log compared with 4 to 5 logs for SVPs. The competitive effect of SVPs at viral entry might account for the long incubation period that characterizes the course of an HBV infection and explain the difficulties encountered in propagating HBV in tissue culture.

Finally, since cell surface proteoglycans are not restricted to human hepatocytes, tissue and species specificities of HBV infection are unlikely to be determined at the initial stage of viral entry. In fact, HBV can attach to nonliver cells and to nondifferentiated HepaRG cells.16 Interestingly, even pre-S1 binding, though liver-specific, is not restricted to human hepatocytes.12 Thus, species specificity may lie at an intermediate level between AGL and pre-S1 activities. Whether HSPGs in the space of Disse participate to selectivity is unknown.

In the current study, the characterization of AGL function at viral entry provides a foundation for understanding the true significance of HS binding in HBV infection, in particular to what degree this process participates to species specificity and tissue tropism. Progress in this direction will require, in addition to the structural data, the identification of oligosaccharide motifs involved in AGL recognition. This will help design new classes of antivirals such as HS-like molecules directed to the AGL or compounds targeting liver-specific HSPGs.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

We thank Stephan Urban for providing the pre-S1 lipopeptides.

References

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information
  • 1
    Ganem D, Prince AM. Hepatitis B virus infection—natural history and clinical consequences. N Engl J Med 2004; 350: 11181129.
  • 2
    Glebe D, Urban S. Viral and cellular determinants involved in hepadnaviral entry. World J Gastroenterol 2007; 13: 2238.
  • 3
    Bonino F, Heermann KH, Rizzetto M, Gerlich WH. Hepatitis delta virus: protein composition of delta antigen and its hepatitis B virus-derived envelope. J Virol 1986; 58: 945950.
  • 4
    Wang CJ, Chen PJ, Wu JC, Patel D, Chen DS. Small-form hepatitis B surface antigen is sufficient to help in the assembly of hepatitis delta virus-like particles. J Virol 1991; 65: 66306636.
  • 5
    Sureau C, Lanford R. Analysis of hepatitis B virus envelope proteins in assembly and infectivity of human hepatitis delta virus. Prog Clin Biol Res 1993; 382: 4551.
  • 6
    Heermann KH, Goldmann U, Schwartz W, Seyffarth T, Baumgarten H, Gerlich WH. Large surface proteins of hepatitis B virus containing the pre-s sequence. J Virol 1984; 52: 396402.
  • 7
    Le Duff Y, Blanchet M, Sureau C. The pre-S1 and antigenic loop infectivity determinants of the hepatitis B virus envelope proteins are functionally independent. J Virol 2009; 83: 1244312451.
  • 8
    Engelke M, Mills K, Seitz S, Simon P, Gripon P, Schnolzer M, et al. Characterization of a hepatitis B and hepatitis delta virus receptor binding site. HEPATOLOGY 2006; 43: 750760.
  • 9
    Salisse J, Sureau C. A function essential to viral entry underlies the hepatitis B virus “a” determinant. J Virol 2009; 83: 93219328.
  • 10
    Sureau C, Jacob JR, Eichberg JW, Lanford RE. Tissue culture system for infection with human hepatitis delta virus. J Virol 1991; 65: 34433450.
  • 11
    Sureau C. The use of hepatocytes to investigate HDV infection: the HDV/HepaRG model. Methods Mol Biol 2010; 640: 463473.
  • 12
    Petersen J, Dandri M, Mier W, Lutgehetmann M, Volz T, von Weizsacker F, et al. Prevention of hepatitis B virus infection in vivo by entry inhibitors derived from the large envelope protein. Nat Biotechnol 2008; 26: 335341.
  • 13
    Abou-Jaoude G, Sureau C. Entry of hepatitis delta virus requires the conserved cysteine residues of the hepatitis B virus envelope protein antigenic loop and is blocked by inhibitors of thiol-disulfide exchange. J Virol 2007; 81: 1305713066.
  • 14
    Zahn A, Allain JP. Hepatitis C virus and hepatitis B virus bind to heparin: purification of largely IgG-free virions from infected plasma by heparin chromatography. J Gen Virol 2005; 86: 677685.
  • 15
    Leistner CM, Gruen-Bernhard S, Glebe D. Role of glycosaminoglycans for binding and infection of hepatitis B virus. Cell Microbiol 2008; 10: 122133.
  • 16
    Schulze A, Gripon P, Urban S. Hepatitis B virus infection initiates with a large surface protein-dependent binding to heparan sulfate proteoglycans. HEPATOLOGY 2007; 46: 17591768.
  • 17
    Chai N, Chang HE, Nicolas E, Han Z, Jarnik M, Taylor J. Properties of subviral particles of hepatitis B virus. J Virol 2008; 82: 78127817.
  • 18
    Sarrazin S, Lamanna WC, Esko JD. Heparan sulfate proteoglycans. Cold Spring Harb Perspect Biol 2011; 3.
  • 19
    Blanchet M, Sureau C. Infectivity determinants of the hepatitis B virus pre-S domain are confined to the N-terminal 75 amino acid residues. J Virol 2007; 81: 58415849.
  • 20
    Gripon P, Rumin S, Urban S, Le Seyec J, Glaise D, Cannie I, et al. Infection of a human hepatoma cell line by hepatitis B virus. Proc Natl Acad Sci U S A 2002; 99: 1565515660.
  • 21
    Carman WF. The clinical significance of surface antigen variants of hepatitis B virus. J Viral Hepat 1997; 4( Suppl. 1): 1120.
  • 22
    Sureau C, Fournier-Wirth C, Maurel P. Role of N glycosylation of hepatitis B virus envelope proteins in morphogenesis and infectivity of hepatitis delta virus. J Virol 2003; 77: 55195523.
  • 23
    Norder H, Courouce AM, Coursaget P, Echevarria JM, Lee SD, Mushahwar IK, et al. Genetic diversity of hepatitis B virus strains derived worldwide: genotypes, subgenotypes, and HBsAg subtypes. Intervirology 2004; 47: 289309.
  • 24
    Kalinina T, Riu A, Fischer L, Will H, Sterneck M. A dominant hepatitis B virus population defective in virus secretion because of several S-gene mutations from a patient with fulminant hepatitis. HEPATOLOGY 2001; 34: 385394.
  • 25
    Stibbe W, Gerlich WH. Characterization of pre-s gene products in hepatitis B surface antigen. Dev Biol Stand 1983; 54: 3343.
  • 26
    Levy HC, Bowman VD, Govindasamy L, McKenna R, Nash K, Warrington K, et al. Heparin binding induces conformational changes in adeno-associated virus serotype 2. J Struct Biol 2009; 165: 146156.
  • 27
    Esko JD, Linhardt RJ. Proteins that bind sulfated glycosaminoglycans. In: Varki A, Cummings RD, Esko JD, Freeze HH, Stanley P, Bertozzi CR, et al, editors. Essentials of Glycobiology. 2nd ed. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 2009.
  • 28
    Wounderlich G, Bruss V. Characterization of early hepatitis B virus surface protein oligomers. Arch Virol 1996; 141: 11911205.
  • 29
    Mulder AM, Carragher B, Towne V, Meng Y, Wang Y, Dieter L, et al. Toolbox for non-intrusive structural and functional analysis of recombinant VLP based vaccines: a case study with hepatitis B vaccine. PLoS One 2012; 7: e33235.
  • 30
    Short JM, Chen S, Roseman AM, Butler PJ, Crowther RA. Structure of hepatitis B surface antigen from subviral tubes determined by electron cryomicroscopy. J Mol Biol 2009; 390: 135141.
  • 31
    Seitz S, Urban S, Antoni C, Bottcher B. Cryo-electron microscopy of hepatitis B virions reveals variability in envelope capsid interactions. EMBO J 2007; 26: 41604167.
  • 32
    Foley EM, Esko JD. Hepatic heparan sulfate proteoglycans and endocytic clearance of triglyceride-rich lipoproteins. Prog Mol Biol Transl Sci 2010; 93: 213233.
  • 33
    Galibert F, Chen TN, Mandart E. Nucleotide sequence of a cloned woodchuck hepatitis virus genome: comparison with the hepatitis B virus sequence. J Virol 1982; 41: 5165.
  • 34
    Cote PJ Jr, Gerin JL. Nonoverlapping antigenic sites of woodchuck hepatitis virus surface antigen and their cross-reactivity with ground squirrel hepatitis virus and hepatitis B virus surface antigens. J Virol 1983; 47: 1523.
  • 35
    Offensperger WB, Offensperger S, Walter E, Blum HE, Gerok W. Sulfated polyanions do not inhibit duck hepatitis B virus infection. Antimicrob Agents Chemother 1991; 35: 24312433.
  • 36
    Heermann KH, Gerlich WH. Surface proteins of hepatitis B viruses. In: Maclachlan A, ed. Molecular Biology of the Hepatitis B Virus. Boca Raton, FL: CRC Press; 1991: 109144.
  • 37
    Shearer MH, Sureau C, Dunbar B, Kennedy RC. Structural characterization of viral neutralizing monoclonal antibodies to hepatitis B surface antigen. Mol Immunol 1998; 35: 11491160.

Supporting Information

  1. Top of page
  2. Abstract
  3. Materials and Methods
  4. Results
  5. Discussion
  6. Acknowledgements
  7. References
  8. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
HEP_26125_sm_SuppInfo.docx167KSupporting Information
HEP_26125_sm_SuppFigS1.tif1616KSupporting Figure S1
HEP_26125_sm_SuppFigS2.tif674KSupporting Figure S2
HEP_26125_sm_SuppFigS3.tif1460KSupporting Figure S3

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.