Sera of patients with acute (AH) and chronic active hepatitis (CAH) were tested for anti-hepatitis B virus (HBV) x-protein (HBx) by immunoblotting, using recombinant MS2- and βgal-HBx fusion proteins as substrate. Antibodies against HBx were detected in 5 out of 17 patients with AH at an early stage of infection, and in 13 out of 35 patients with CAH. Positive sera from AH patients showed a relatively weak anti-HBx reactivity when compared to sera from CAH patients. In follow up studies we tested serial serum samples from patients positive for anti-HBx. Patients with AH were observed for 3 to 6 weeks and CAH patients for up to 51 months. In general anti-HBx reactivities appeared to be stable although significant differences in apparent antibody levels were noted when sera from individual patients were compared. Our data further support an early expression of HBx-antigen in HBV-infected individuals. There was no correlation between HBe-antigen and anti-HBx in CAH.

All human hepatitis B viruses characterized so far express three envelope proteins, pre-S1, pre-S2, and HBs, which are believed to function as binding proteins for the cellular receptor, as targets for immune-mediated virus elimination, and in virion morphogenesis and secretion. Here we report the characterization of infectious HBV variant genomes that are unable to express a pre-S2 protein and which were derived from serum of a highly viremic chronic carrier. Direct sequencing of the amplified pre-S region and sequencing of 50 cloned amplified pre-S DNA fragments revealed that in all molecules, in addition to numerous nucleotide changes, there were deletions of the pre-S2 translation initiation codon and three codons 54 nucleotides downstream thereof. No pre-S2 protein and altered pre-S1 proteins were found in the serum of the patient. Cloned infectious HBV DNA genomes having the pre-S region substituted by the variant pre-S region were replication competent in cultured hepatoblastoma cells. Morphologically normal virions were efficiently secreted and were infectious for primary human hepatocyte cultures. These data demonstrate that HBV devoid of pre-S2 protein can occur in vivo as a dominant or exclusive virus population and that expression of the pre-S2 protein is not essential for HBV replication, virion morphogenesis, secretion, or in vitro infectivity.

The hepatitis B virus genome encodes a transcriptional transactivator protein designated HBxAg. We have investigated whether this antigen is a target structure for human T-lymphocytes. Using recombinant HBxAg protein, we found HBxAg-specific stimulation of peripheral blood mononuclear cells in patients with acute hepatitis B virus infection (6 of 6) and chronic hepatitis B virus infection (6 of 17) but not in healthy individuals. With HBxAg-specific synthetic polypeptides, several T-cell epitopes were identified. Most were located in the carboxyterminal half of the HBxAg protein. Five T-cell clones specific for a T-cell epitope located at the carboxyterminal region of HBxAg were established and found to belong to the CD2/CD4-positive, CD8-negative subtype. These data establish for the first time HBxAg as an antigen in the cellular immune response.

Serological responses to hepatitis B virus-X determinants have been noted in human sera, but conflicting findings concerning the correlation of anti-HBx antibodies with different stages of hepatitis B virus infection or pathological sequelae have been reported. Using an adenovirus-based eukaryotic vector, the 17 kD X protein was efficiently expressed in 293 cells. Cellular extracts containing the eukaryotic X protein have been used to screen for anti-HBx antibodies by immunoblot analysis in a large panel of sera from patients affected by hepatitis B virus chronic hepatitis, hepatocellular carcinoma and acute viral hepatitis. Sera from 32 of 171 (19%) chronic hepatitis B virus patients were positive for anti-HBx antibodies. Only one of thirty-two (3%) HBsAg-negative, anti-HBs/anti-HBc-positive chronic hepatitis serum was anti-HBx positive. Very few sera from primary hepatocellular carcinoma patients showed positivity for anti-HBx (8 of 149 or 5%). Anti-HBx were also detected in 8 of 48 (17%) acute viral hepatitis patients. In the four cases that were followed up weekly, anti-HBx antibodies appeared 3 to 4 wk after the onset of the clinical signs. To compare the X protein expressed in eukaryotic and prokaryotic cells as a substrate for anti-HBx antibody detection, 171 sera were screened with HBx fusion proteins expressed in Escherichia coli. The prokaryotic cell extract test seems to be more sensitive. During the chronic phase of hepatitis B virus infection, the presence of anti-HBx antibodies detected with the eukaryotic cell extract correlates with the presence of well-established markers of ongoing viral replication: serum hepatitis B virus-DNA (p less than 0.001) and intrahepatic HBcAg expression (p less than 0.001).

Mutant HBV genomes unable to produce preS2 protein were identified in the serum of a highly viremic chronic carrier. By direct sequencing of the amplified preS region and by sequencing of 50 cloned amplified preS DNA fragments all molecules were found to have deleted the preS2 translation initiation codon and three amino acids 54 nucleotides downstream thereof. In addition, numerous amino acid changes, predominantly located between both deletions, were revealed. HBV-DNA genomes containing the mutated preS sequences were shown to be replication competent and to secrete efficiently virions that were morphologically and by the type of DNA encapsidated not distinguishable from wild-type virus. These data demonstrate that HBV with mutated preS sequences and unable to express preS2 protein can occur as a dominant or exclusive virus population in a highly viremic chronic carrier. The data also show that expression of the preS2 protein is not essential for HBV replication, virion morphogenesis and secretion.

The immune response to the X protein of human hepatitis B virus (HBV) was studied by epitope mapping by using a set of MS2-HBx fusion proteins and synthetic peptides. Antibodies in sera of patients with acute and chronic HBV infection showed a multispecific immune response. Each serum contained antibodies to a different set of epitopes, which taken together cover most of the HBx sequence. Some of the epitopes were detectable only by immunoblotting with fusion proteins; others were detectable only by an enzyme-linked immunosorbent assay (ELISA) with synthetic peptides. The carboxy-terminal half of the HBx protein was preferentially recognized by antibodies from patients with chronic hepatitis and contained a short immunodominant antigenic region with at least two major nonoverlapping epitopes. Anti-HBx antibody titers as revealed by peptide ELISAs were highest and most frequent in patients with chronic hepatitis and usually low in acutely infected patients and asymptomatic carriers. The data demonstrate a remarkable qualitative and quantitative heterogeneity of the humoral HBx immune response which can be monitored by HBx-specific peptide ELISAs. Such tests may become useful diagnostic tools.

In hepatitis B virus carriers who are anti-HBe positive despite ongoing viral replication (HBcAg in liver and HBV-DNA in serum) the natural course of hepatitis is severe and the response to interferon is low. We investigated whether a new hepatitis B virus (HBV) strain could be involved. A translational termination codon at the carboxyterminal end of the pre-C region responsible for the lack of HBeAg secretion was found in 18 of 19 HBV clones isolated from seven pedigreed patients with this clinical syndrome. The same findings were confirmed by direct sequencing. One of these patients underwent a liver transplant and HBV infection of the new liver resulted in high titered viremia and intrahepatic expression of HBcAg, without detectable HBeAg in serum. Another patient was superinfected by hepatitis delta virus (HDV) and developed high titres of total and IgM anti-HD. In spite of this, chronic hepatitis remained unchanged during 7 years of follow-up. These data strongly suggest that a viable precore minus mutant of hepatitis B virus is responsible for the lack of HBeAg in the serum of these patients. The HBV variant may explain the peculiar geographic distribution of anti-HBe positive hepatitis. The variations in the virus genome sequence may cause the more severe form of liver disease and modify the pathogenicity in the case of HDV superinfection.

To examine whether there are HBV variants not yet described which cannot express HBe due to a mutation in the pre-C region, and, if they exist, whether they cause a particular course of infection and disease, we analyzed the HBV genomes of a HBs/anti-HBe positive chronic carrier who had several episodes of acute reexacerbations of chronic hepatitis with at least two viremic phases. Direct sequencing of the precore/core and the pre-S regions of the HBV sequences of both viremic phases amplified by the polymerase chain reaction revealed that they were very similar to each other but substantially divergent from published HBV genomes. Both virus populations contained a mutation in the first nucleotide of the pre-C translation initiation codon (AUG/CUG) which prevents HBeAg expression. These data demonstrate the existence of a new HBV variant which can enter a latent phase from which it can be reactivated with acute reexacerbation of liver inflammation.

In human patients infected with hepatitis B virus (HBV) seroconversion from HBe to anti-HBe often signals virus elimination. Occasionally, a second viremic phase is observed which may be due to superinfection with a variant or reactivation of a latent virus. To differentiate between these two possibilities we investigated the nucleotide sequences of virus populations from sera of a chronically infected patient who had two distinct viremic phases, one e-antigen positive and one antiHBe antibody positive. By direct sequencing of amplified HBV C- and pre-S gene sequences we found that the viruses in the two populations differed by only two point mutations, one of which prevents expression of precore derived e-antigen. In the nonviremic phase viral DNA and antigen were found in the liver but nucleocapsid protein expression appeared drastically downregulated. These data demonstrate that HBV can enter a latent phase with low expression of core protein and selection for HBV variants which cannot synthesize e-antigen. This suggests that e-antigen expression can be a critical target for virus elimination and that loss of its expression can prolong chronicity and provoke latency of HBV infection.

Human sera were examined by immunoblotting for antibodies against the polymerase (reverse transcriptase) believed to be encoded in the P open reading frame (ORF) of human hepatitis B virus (HBV). Sera from patients with self-limited and chronic hepatitis reacted specifically with fusion proteins containing different domains of the P ORF. The results indicate that this ORF is expressed, and that the corresponding proteins contain at least two immunogenic domains. In contrast to human immunodeficiency virus, induction of antibodies against reverse transcriptase appears to be less common for HBV, and may depend on long persistence of infection.