Gamma-adaptin, a novel ubiquitin-interacting adaptor, and Nedd4 ubiquitin ligase control hepatitis B virus maturation

Hepatitis B virus (HBV) budding from infected cells is a tightly regulated process that requires both core and envelope structures. Here we report that HBV uses cellular gamma2-adaptin and Nedd4, possibly in conjunction with ubiquitin, to coordinate its assembly and release. In search of interaction partners of the viral L envelope protein, we previously discovered gamma2-adaptin, a putative endosomal sorting and trafficking adaptor of the adaptor protein complex family. We now demonstrate that the viral core interacts with the same gamma2-adaptor and that disruption of the HBV/gamma2-adaptin interactions inhibits virus production. Mutational analyses revealed a hitherto unknown ubiquitin-binding activity of gamma2-adaptin, specified by a ubiquitin-interacting motif, which contributes to its interaction with core. For core, the lysine residue at position 96, a potential target for ubiquitination, was identified to be essential for both gamma2-adaptin-recognition and virus production. The participation of the cellular ubiquitin system in HBV assembly was further suggested by our finding that core interacts with the endosomal ubiquitin ligase Nedd4, partly via its late domain-like PPAY sequence. Overexpression of a catalytically inactive Nedd4 mutant diminished HBV egress, indicating that protein ubiquitination is functionally involved in virus production. Additional evidence for a link of HBV assembly to the endosomal machinery was provided by immunolabeling studies that demonstrated colocalization of core and L with gamma2-adaptin in compartments positive for the late endosomal marker CD63. Together, these data indicate that an enveloped DNA virus exploits a new ubiquitin receptor together with endosomal pathway functions for egress from hepatocytes.     

Functions of the internal pre-S domain of the large surface protein in hepatitis B virus particle morphogenesis.

The large hepatitis B virus (HBV) surface protein (L) forms two isomers which display their N-terminal pre-S domain at the internal and external side of the viral envelope, respectively. The external pre-S domain has been implicated in binding to a virus receptor. To investigate functions of the internal pre-S domain, a secretion signal sequence was fused to the N terminus of L (sigL), causing exclusive expression of external pre-S domains. A fusion construct with a nonfunctional signal (s25L), which corresponds in its primary sequence to sigL cleaved by signal peptidase, was used as a control. SigL was N glycosylated in transfected COS cells at both potential sites in pre-S in contrast to s25L or wild-type L, confirming the expected transmembrane topologies of sigL and s25L. Phenotypic characterization revealed the following points. (i) SigL lost the inhibitory effect of L or s25L on secretion of subviral hepatitis B surface antigen particles, suggesting that the retention signal mapped to the N terminus of L is recognized in the cytosol and not in the lumen of the endoplasmic reticulum. (ii) SigL was secreted into the culture medium even in the absence of the major HBV surface protein (S), while release of an L mutant lacking the retention signal was still dependent on S coexpression. (iii) s25L but not sigL could complement an L-negative HBV genome defective for virion secretion in cotransfections. This suggests that the cytosolic pre-S domain, like a matrix protein, is involved in the interaction of the viral envelope with preformed cytosolic nucleocapsids during virion assembly.     

An 80-kilodalton protein that binds to the pre-S1 domain of hepatitis B virus

It has been suggested that hepatitis B virus (HBV) binds to a receptor on the plasma membrane of human hepatocytes via the pre-S1 domain of the large envelope protein as an initial step in HBV infection. However, the nature of the receptor remains controversial. In an attempt to identify a cell surface receptor for HBV, purified recombinant fusion protein of the pre-S1 domain of HBV with glutathione S-transferase (GST), expressed in Escherichia coli, was used as a ligand. The surface of human hepatocytes or HepG2 cells was biotinylated, and the cell lysate (precleared lysate) which did not bind to GST and glutathione-Sepharose beads was used as a source of receptor molecules. The precleared lysate of the biotinylated cells was incubated with the GST-pre-S1 fusion protein, and the bound proteins were visualized by Western blotting and enhanced chemiluminescence. An approximately 80-kDa protein (p80) was shown to bind specifically to the pre-S1 domain of the fusion protein. The receptor binding assay using serially or internally deleted segments of pre-S1 showed that amino acid residues 12 to 20 and 82 to 90 are essential for the binding of pre-S1 to p80. p80 also bound specifically to the pre-S1 of native HBV particles. Analysis of the tissue and species specificity of p80 expression in several available human primary cultures and cell lines of different tissue origin showed that p80 expression is not restricted to human hepatocytes. Taken together the results suggest that p80 may be a component of the viral entry machinery.     

gamma2-Adaptin, a ubiquitin-interacting adaptor, is a substrate to coupled ubiquitination by the ubiquitin ligase Nedd4 and functions in the endosomal pathway

gamma2-Adaptin is a putative member of the clathrin adaptor protein family with unknown physiological function. We previously reported that gamma2-adaptin acts as a ubiquitin receptor by virtue of its ubiquitin-interacting motif. Here we demonstrate that this motif mediates a specific physical interaction with the ubiquitin ligase Nedd4 and promotes ubiquitination of gamma2-adaptin. By mapping regions of Nedd4 involved in binding to gamma2-adaptin, we identified its C2 domain to be essential, whereas the WW and HECT domains are dispensable. Consistent with this, we uncovered that the C2 domain of Nedd4 is ubiquitinated itself and as such is recruited by the ubiquitin-interacting motif of gamma2-adaptin for subsequent ubiquitin conjugation. Unlike known coupled ubiquitination reactions, this novel type of interaction leads to mono- and multi/polyubiquitinated gamma2-adaptin. In addition, we show that gamma2-adaptin functions in the endosomal/multivesicular body (MVB) pathway. Depletion of gamma2-adaptin impairs the degradation of internalized epidermal growth factor and results in defective MVB morphology characterized by significantly enlarged vesicles. These defects cannot be rescued by gamma1-adaptin, a closely related homolog of gamma2-adaptin, which is unable to bind ubiquitin. Together, these results indicate that gamma2-adaptin may operate within the MVB sorting system in a manner different from that of classic adaptins.     

Hepatitis B virus maturation is sensitive to functional inhibition of ESCRT-III, Vps4, and gamma 2-adaptin

Hepatitis B virus (HBV) is an enveloped DNA virus that presumably buds at intracellular membranes of infected cells. HBV budding involves two endocytic host proteins, the ubiquitin-interacting adaptor gamma 2-adaptin and the Nedd4 ubiquitin ligase. Here, we demonstrate that HBV release also requires the cellular machinery that generates internal vesicles of multivesicular bodies (MVBs). In order to perturb the MVB machinery in HBV-replicating liver cells, we used ectopic expression of dominant-negative mutants of different MVB components, like the ESCRT-III complex-forming CHMP proteins and the Vps4 ATPases. Upon coexpression of mutated CHMP3, CHMP4B, or CHMP4C forms, as well as of ATPase-defective Vps4A or Vps4B mutants, HBV assembly and egress were potently blocked. Each of the MVB inhibitors arrested virus particle maturation by entrapping the viral core and large and small envelope proteins in detergent-insoluble membrane structures that closely resembled aberrant endosomal class E compartments. In contrast, HBV subvirus particle release was not affected by MVB inhibitors, hinting at different export routes used by viral and subviral particles. To further define the role gamma 2-adaptin plays in HBV formation, we examined the effects of its overexpression in virus-replicating cells. Intriguingly, excess gamma 2-adaptin blocked HBV production in a manner similar to the actions of CHMP and Vps4 mutants. Moreover, overexpressed gamma 2-adaptin perturbed the endosomal morphology and diminished the budding of a retroviral Gag protein, implying that it may act as a principal inhibitor of the MVB sorting pathway. Together, these results demonstrate that HBV exploits the MVB machinery with the aid of gamma 2-adaptin.     

Immunoadhesins containing pre-S domains of hepatitis B virus large envelope protein are secreted and inhibit virus infection

Hepatitis B virus (HBV) replication produces three envelope proteins (L, M, and S) that have a common C terminus. L, the largest, contains a domain, pre-S1, not present on M. Similarly M contains a domain, pre-S2, not present on S. The pre-S1 region has important functions in the HBV life cycle. Thus, as an approach to studying these roles, the pre-S1 and/or pre-S2 sequences of HBV (serotype adw2, genotype A) were expressed as N-terminal fusions to the Fc domain of a rabbit immunoglobulin G chain. Such proteins, known as immunoadhesins (IA), were highly expressed following transfection of cultured cells and, when the pre-S1 region was present, >80% were secreted. The IA were myristoylated at a glycine penultimate to the N terminus, although mutation studies showed that this modification was not needed for secretion. As few as 30 amino acids from the N terminus of pre-S1 were both necessary and sufficient to drive secretion of IA. Even expression of pre-S1 plus pre-S2, in the absence of an immunoglobulin chain, led to efficient secretion. Overall, these studies demonstrate an unexpected ability of the N terminus of pre-S1 to promote protein secretion. In addition, some of these secreted IA, at nanomolar concentrations, inhibited infection of primary human hepatocytes either by hepatitis delta virus (HDV), a subviral agent that uses HBV envelope proteins, or HBV. These IA have potential to be part of antiviral therapies against chronic HDV and HBV, and may help understand the attachment and entry mechanisms used by these important human pathogens.     

Gamma-BAR, a novel AP-1-interacting protein involved in post-Golgi trafficking

A novel peripheral membrane protein (2c18) that interacts directly with the gamma 'ear' domain of the adaptor protein complex 1 (AP-1) in vitro and in vivo is described. Ultrastructural analysis demonstrates a colocalization of 2c18 and gamma1-adaptin at the trans-Golgi network (TGN) and on vesicular profiles. Overexpression of 2c18 increases the fraction of membrane-bound gamma1-adaptin and inhibits its release from membranes in response to brefeldin A. Knockdown of 2c18 reduces the steady-state levels of gamma1-adaptin on membranes. Overexpression or downregulation of 2c18 leads to an increased secretion of the lysosomal hydrolase cathepsin D, which is sorted by the mannose-6-phosphate receptor at the TGN, which itself involves AP-1 function for trafficking between the TGN and endosomes. This suggests that the direct interaction of 2c18 and gamma1-adaptin is crucial for membrane association and thus the function of the AP-1 complex in living cells. We propose to name this protein gamma-BAR.     

Protease-induced infectivity of hepatitis B virus for a human hepatoblastoma cell line.

The human hepatoblastoma cell line HepG2 produces and secretes hepatitis B virus (HBV) after transfection of cloned HBV DNA. Intact virions do not infect these cells, although they attach to the surface of the HepG2 cell through binding sites in the pre-S1 domain. Entry of enveloped virions into the cell often requires proteolytic cleavage of a viral surface protein that is involved in fusion between the cell membrane and the viral envelope. Recently, we observed pre-S-independent, nonspecific binding between hepatitis B surface (HBs) particles and HepG2 cells after treatment of HBs antigen particles with V8 protease, which cleaves next to a putative fusion sequence. Chymotrypsin removed this fusion sequence and did not induce binding. In this study, we postulate that lack of a suitable fusion-activating protease was the reason why the HepG2 cells were not susceptible to HBV. To test this hypothesis, virions were partially purified from the plasma of HBV carriers and treated with either staphylococcal V8 or porcine chymotrypsin protease. Protease-digested virus lost reactivity with pre-S2-specific antibody but remained morphologically intact as determined by electron microscopy. After separation from the proteases, virions were incubated with HepG2 cells at pH 5.5. Cultures inoculated with either intact or chymotrypsin-digested virus did not contain detectable levels of intracellular HBV DNA at any time following infection. However, in cultures inoculated with V8-digested virions, HBV-specific products, including covalently closed circular DNA, viral RNA, and viral pre-S2 antigen, could be detected in a time-dependent manner following infection. Immunofluorescence analysis revealed that 10 to 30% of the infected HepG2 cells produced HBV antigen. Persistent secretion of virus by the infected HepG2 cells lasted at least 14 days and was maintained during several reseeding steps. The results show that V8-digested HBV can productively infect tissue cultures of HepG2 cells. It is suggested that proteolysis-dependent exposure of a fusion domain within the envelope protein of HBV is necessary during natural infection.     

Quantitative analysis of the interaction between the envelope protein domains and the core protein of human hepatitis B virus

Interaction between preformed nucleocapsids and viral envelope proteins is critical for the assembly of virus particles in infected cells. The pre-S1 and pre-S2 and cytosolic regions of the human hepatitis B virus envelope protein had been implicated in the interaction with the core protein of nucleocapsids. The binding affinities of specific subdomains of the envelope protein to the core protein were quantitatively measured by both ELISA and BIAcore assay. While a marginal binding was detected with the pre-S1 or pre-S2, the core protein showed high affinities to pre-S with apparent dissociation constants (K(D)(app)) of 7.3+/-0.9 and 8.2+/-0.4microM by ELISA and BIAcore assay, respectively. The circular dichroism analysis suggested that conformational change occurs in pre-S through interaction with core protein. These results substantiate the importance of specific envelope domains in virion assembly, and demonstrate that the interaction between viral proteins can be quantitatively measured in vitro.     

Infection Process of the Hepatitis B Virus Depends on the Presence of a Defined Sequence in the Pre-S1 Domain

During the life cycle of hepatitis B virus (HBV), the large envelope protein (L) plays a pivotal role. Indeed, this polypeptide is essential for viral assembly and probably for the infection process. By performing mutagenesis experiments, we have previously excluded a putative involvement of the pre-S2 domain of the L protein in viral infectivity. In the present study, we have evaluated the role of the pre-S1 region in HBV infection. For this purpose, 21 mutants of the L protein were created. The entire pre-S1 domain was covered by contiguous deletions of 5 amino acids. First, after transfection into HepG2 cells, the efficient expression of both glycosylated and unglycosylated L mutant proteins was verified. The secretion rate of envelope proteins was modified positively or negatively by deletions, indicating that the pre-S1 domain contains several regulating sequences able to influence the surface protein secretion. The ability of mutant proteins to support the production of virions was then studied. Only the four C-terminal deletions, covering the 17 amino acids suspected to interact with the cytoplasmic nucleocapsids, inhibited virion release. Finally, the presence of the modified pre-S1 domain at the external side of all secreted virions was confirmed, and their infectivity was assayed on normal human hepatocytes in primary culture. Only a short sequence including amino acids 78 to 87 tolerates internal deletions without affecting viral infectivity. These results confirm the involvement of the L protein in the infection step and demonstrate that the sequence between amino acids 3 and 77 is involved in this process.     

Role of the large hepatitis B virus envelope protein in infectivity of the hepatitis delta virion.

The hepatitis delta virus (HDV) is coated with large (L), middle (M), and small (S) envelope proteins encoded by coinfecting hepatitis B virus (HBV). To study the role of the HBV envelope proteins in the assembly and infectivity of HDV, we produced three types of recombinant particles in Huh7 cells by transfection with HBV DNA and HDV cDNA: (i) particles with an envelope containing the S HBV envelope protein only, (ii) particles with an envelope containing S and M proteins, and (iii) particles with an envelope containing S, M, and L proteins. Although the resulting S-, SM-, and SML-HDV particles contained both hepatitis delta antigen and HDV RNA, only particles coated with all three envelope proteins (SML) showed evidence of infectivity in an in vitro culture system susceptible to HDV infection. We concluded that the L HBV envelope protein, and more specifically the pre-S1 domain, is important for infectivity of HDV particles and that the M protein, which has been reported to bear a site for binding to polymerized albumin in the pre-S2 domain, is not sufficient for infectivity. Our data also show that the helper HBV is not required for initiation of HDV infection. The mechanism by which the L protein may affect HDV infectivity is discussed herein.     

Mapping of the hepatitis B virus pre-S1 domain involved in receptor recognition

Hepatitis B virus (HBV) and woolly monkey hepatitis B virus (WMHBV) are primate hepadnaviruses that display restricted tissue and host tropisms. Hepatitis D virus (HDV) particles pseudotyped with HBV and WMHBV envelopes (HBV-HDV and WM-HDV) preferentially infect human and spider monkey hepatocytes, respectively, thereby confirming host range bias in vitro. The analysis of chimeric HBV and WMHBV large (L) envelope proteins suggests that the pre-S1 domain may comprise two regions that affect infectivity: one within the amino-terminal 40 amino acids of pre-S1 and one downstream of this region. In the present study, we further characterized the role of the amino terminus of pre-S1 in infectivity by examining the ability of synthetic peptides to competitively block HDV infection of primary human and spider monkey hepatocytes. A synthetic peptide representing the first 45 residues of the pre-S1 domain of the HBV L protein blocked infectivity of HBV-HDV and WM-HDV, with a requirement for myristylation of the amino terminal residue. Competition studies with truncated peptides suggested that pre-S1 residues 5 to 20 represent the minimal domain for inhibition of HDV infection and, thus, presumably represent the residues involved in virus-host receptor interaction. Recombinant pre-S1 proteins expressed in insect cells blocked infection with HBV-HDV and WM-HDV at a concentration of 1 nanomolar. The ability of short pre-S1 peptides to efficiently inhibit HDV infection suggests that they represent suitable ligands for identification of the HBV receptor and that a pre-S1 mimetic may represent a rational therapy for the treatment of HBV infection.     

Analysis of host range phenotypes of primate hepadnaviruses by in vitro infections of hepatitis D virus pseudotypes

Hepatitis B virus (HBV) and woolly monkey hepatitis B virus (WMHBV) have natural host ranges that are limited to closely related species. The barrier for infection of primates seems to be at the adsorption and/or entry steps of the viral replication cycle, since a human hepatoma cell line is permissive for HBV and WMHBV replication following transfection of cloned DNA. We hypothesized that the HBV and WMHBV envelope proteins contain the principal viral determinants of host range. As previously shown by using the hepatitis D virus (HDV) system, recombinant HBV-HDV particles were infectious in chimpanzee as well as human hepatocytes. We extended the HDV system to include HDV particles pseudotyped with the WMHBV envelope. In agreement with the natural host ranges of HBV and WMHBV, in vitro infections demonstrated that HBV-HDV and WM-HDV particles preferentially infected human and spider monkey cells, respectively. Previous studies have implicated the pre-S1 region of the large (L) envelope protein in receptor binding and host range; therefore, recombinant HDV particles were pseudotyped with the hepadnaviral envelopes containing chimeric L proteins with the first 40 amino acids from the pre-S1 domain exchanged between HBV and WMHBV. Surprisingly, addition of the human amino terminus to the WMHBV L protein increased infectivity on spider monkey hepatocytes but did not increase infectivity for human hepatocytes. Based upon these data, we discuss the possibility that the L protein may be comprised of two domains that affect infectivity and that sequences downstream of residue 40 may influence host range and receptor binding or entry.     

Fibronectin of human liver sinusoids binds hepatitis B virus: identification by an anti-idiotypic antibody bearing the internal image of the pre-S2 domain.

Anti-idiotypic antibodies (anti-Ids) have been successfully used to characterize and isolate receptors of several cell ligands. To prepare an immunological probe for identification of cellular components interacting with the hepatitis B virus (HBV), polyclonal antisera against a panel of five HBV-specific monoclonal antibodies (MAbs) were produced in syngeneic BALB/c mice. MAbs to HBV used for immunization (Ab1) recognized biologically important and potentially neutralizing epitopes, located in the pre-S1, pre-S2, or S region-encoded domains of HBV proteins. All the anti-Ids (Ab2) were specific to idiotopes of the homologous Ab1 and inhibited their interaction with the corresponding viral epitopes, suggesting that they recognized unique determinants on the paratope of each immunizing Ab1. Therefore, all five generated polyclonal anti-Ids were of the Ab2 beta type and could represent internal images of viral epitopes. Ab2 raised against the pre-S2 region-specific MAb F124 bound to the extracellular matrix fibronectin of human liver sinusoids. Immunohistochemical studies demonstrated the attachment of viral and recombinant (S, M) hepatitis B surface antigen particles with the pre-S2 region-encoded epitopes to the fibronectin of human liver sinusoids. In contrast, recombinant (S, L*) hepatitis B surface antigen particles, in which the epitope recognized by F124 MAb was not expressed, did not show any binding capacity. These findings suggest that human liver fibronectin may bind HBV in vivo by the pre-S2 region-encoded epitopes in a species-restricted manner. Furthermore, binding of the circulating virus to liver sinusoids could facilitate its subsequent uptake by hepatocytes.     

Infectivity determinants of the hepatitis B virus pre-S domain are confined to the N-terminal 75 amino acid residues

The N-terminal pre-S domain of the large hepatitis B virus (HBV) envelope protein plays a pivotal role at the initial step of the viral entry pathway. In the present study, the entire pre-S domain was mapped for infectivity determinants, following a reverse-genetics approach and using in vitro infection assays with hepatitis delta virus (HDV) or HBV particles. The results demonstrate that lesions created within the N-terminal 75 amino acids of the pre-S region abrogate infectivity, whereas mutations between amino acids 76 and 113, overlapping the matrix domain, had no effect. In contrast to the results of a recent study (L. Stoeckl, A. Funk, A. Kopitzki, B. Brandenburg, S. Oess, H. Will, H. Sirma, and E. Hildt, Proc. Natl. Acad. Sci. 103:6730-6734, 2006), the deletion of a cell membrane translocation motif (TLM) located between amino acids 148 and 161 at the C terminus of pre-S2 did not interfere with the infectivity of the resulting HDV or HBV mutants. Furthermore, a series of large deletions overlapping the pre-S2 domain were compatible with infectivity, although the efficiency of infection was reduced when the deletions extended to the pre-S1 domain. Overall, the results demonstrate that the activity of the pre-S domain at viral entry solely depends on the integrity of its first 75 amino acids and thus excludes any function of the matrix domain or TLM.     

Gamma-adaptin interacts directly with Rabaptin-5 through its ear domain

In yeast two-hybrid screening using gamma1-adaptin, a subunit of the AP-1 adaptor complex of clathrin-coated vesicles derived from the trans-Golgi network (TGN), as bait, we found that it could interact with Rabaptin-5, an effector of Rab5 and Rab4 that regulates membrane docking with endosomes. Further two-hybrid analysis revealed that the interaction occurs between the ear domain of gamma1-adaptin and the COOH-terminal coiled-coil region of Rabaptin-5. Pull down assay with a fusion protein between glutathione S-transferase and the ear domain of gamma1-adaptin and coimmunoprecipitation analysis revealed that the interaction occurs in vitro and in vivo. Immunocytochemical analysis showed that gamma1-adaptin and Rabaptin-5 colocalize to a significant extent on perinuclear structures, probably on recycling endosomes, and are redistributed into the cytoplasm upon treatment with brefeldin A. These results suggest that the gamma1-adaptin-Rabaptin-5 interaction may play a role in membrane trafficking between the TGN and endosomes.     

ARH is a modular adaptor protein that interacts with the LDL receptor,clathrin, and AP-2

Mutations in the phosphotyrosine binding domain protein ARH cause autosomal recessive hypercholesterolemia, a disorder caused by defective internalization of low density lipoprotein receptors (LDLR) in the liver. To examine the function of ARH, we used pull-down experiments to test for interactions between ARH, the LDLR, and proteins involved in clathrin-mediated endocytosis. The phosphotyrosine binding domain of ARH interacted with the internalization sequence (NPVY) in the cytoplasmic tail of LDLR in a sequence-specific manner. Mutations in the NPVY sequence that were previously shown to decrease LDLR internalization abolished in vitro binding to ARH. Recombinant ARH bound purified bovine clathrin with high affinity (K(D), approximately 44 nm). The interaction between ARH and clathrin was mapped to a canonical clathrin box sequence (LLDLE) in ARH and to the N-terminal domain of the clathrin heavy chain. A highly conserved 20-amino acid sequence in the C-terminal region of ARH bound the beta(2)-adaptin subunit of AP-2. Mutation of a glutamic acid residue in the appendage domain of beta(2)-adaptin that is required for interaction with the adapter protein beta-arrestin markedly reduced binding to ARH. These data are consistent with the hypothesis that ARH functions as an adaptor protein that couples LDLR to the endocytic machinery.     

Deletion or alteration of hydrophobic amino acids at the first and the third transmembrane domains of hepatitis B surface antigen enhances its production in Escherichia coli

To investigate the failure of high-level production of hepatitis B viral (HBV) surface antigen (HBsAg), including three authentic forms, large (L), middle (M) and major/small (S) HBsAg, in Escherichia coli, we employed the high-expression vector pGEX containing the glutathione S-transferase-encoding gene (GST) to study HBsAg production. Different fragments of HBV DNA containing the entire pre-S1/pre-S2/S region (for L protein), or partial pre-S1, pre-S2, pre-S1/pre-S2 and pre-S2/S region (for M protein), were fused downstream from the GST gene, in order to obtain five plasmids which encode GST-HBsAg fusion proteins. SDS-PAGE analyses revealed that cells containing plasmids with a full-length S region (pGLS and pGMS) produced undetectable GST-HBsAg fusion proteins, in contrast to those cells harboring plasmids without the S region (pGS1, pGS2 and pGS1S2), which synthesized fusion proteins in 3–10% of the total cellular protein. Using an immunoblot method to screen HBsAg production in cells which harbored plasmids derived from exonuclease BAL 31-digested pGLS, we obtained eight positive clones. Nucleotide sequence analyses of plasmids from the positive clones revealed that termination, deletion or frameshift occurred at the regions encoding either the first or the third transmembrane domain of the major HBsAg. Correlation between the production level of GST-HBsAg fusion proteins and their constituent and arrangement of amino acids (aa) at the last 20 as among 15 clones suggested that the fusion protein ended with a longer stretch of or a higher ratio of hydrophobic as had a lower production in E. coli.