Coexistence of Hepatitis B Surface Antigen (HBs Ag
http://www.100md.com
病菌学杂志 2006年第6期
Departments of Virology
Hepato-Gastro-Enterology, AP-HP Pitie-Salpetriere Hospital, Paris, France
ABSTRACT
In chronic hepatitis B (CHB), the persistence of hepatitis B surface antigen (HBs Ag) is sometimes associated with antibodies (Ab) to HBs (anti-HBs). To assess the hypothesis of the selection of HBs Ag immune escape variants in CHB patients, the variability of the HBV S gene was determined for patients persistently carrying both HBs Ag and anti-HBs antibodies and patients solely positive for HBs Ag. We selected 14 patients who presented both markers (group I) in several consecutive samples and 12 patients positive for HBs Ag only (group II). The HBs Ag-encoding gene was amplified and cloned, and at least 15 clones per patient were sequenced and analyzed. The number of residue changes within the S protein was 2.7 times more frequent for group I than for group II patients and occurred mostly in the "a" determinant of the major hydrophilic region (MHR), with 9.52 versus 2.43 changes per 100 residues (P = 0.009), respectively. Ten patients (71%) from group I, but only three (25%) from group II, presented at least two residue changes in the MHR. The most frequent changes in group I patients were located at positions s145, s129, s126, s144, and s123, as described for immune escape variants. In CHB patients, the coexistence of HBs Ag and anti-HBs Ab is associated with an increase of "a" determinant variability, suggesting a selection of HBV immune escape mutants during chronic carriage. The consequences of this selection process with regard to vaccine efficacy, diagnosis, and clinical evolution remain partially unknown.
INTRODUCTION
More than 350 million people, or 5% of the world's population, are chronic carriers of hepatitis B virus (HBV), and this infection represents a worldwide public health problem (21). Many patients chronically infected with HBV, as defined by the persistence over >6 months of hepatitis B surface antigen (HBs Ag), will develop life-threatening diseases such as liver cirrhosis and hepatocellular carcinoma. It has been estimated that up to 30% of them will die from the consequences of their infection (20).
In the natural course of HBV infection, virus clearance likely results from the intimate coordination of both humoral and cellular immune systems. Overall, the antibody-mediated immune response to HBV proteins aims at the clearance of circulating HBV particles, whereas the cellular effectors contribute to eliminating infected hepatocytes (30). Biologically, virus clearance is classically characterized by the emergence of anti-HBs antibody (Ab) in the serological profile.
During chronic hepatitis B infection, two clinically important phases can be defined. While the immune tolerance phase is usually characterized by little liver damage and the presence of hepatitis B e antigen (HBe Ag) in the serum, the second phase can be described as more aggressive for the liver, with selection of HBe Ag-negative variants and detection of anti-HBe antibody. Ultimately, chronic carriers can be classified as inactive carriers, with little viral replication and anti-HBe antibody and normal liver biochemical markers, or as chronic hepatitis patients, with abnormal liver enzyme levels and higher viral loads (23).
Notably, several reports have described the persistence of HBs Ag associated with anti-HBs antibodies in 10 to 25% of chronic hepatitis B (CHB) patients (19, 27). The mechanism underlying the presence of both HBs Ag and anti-HBs antibodies remains unknown, but one possibility could be the selection of immune escape mutants.
The envelope gene of HBV has three open reading frames (ORF), PreS1, PreS2, and S, which code for three proteins, the small, middle, and large HBs Ag, translated from distinct mRNAs. Common to all three proteins, the "a" dominant epitope is located at codon positions 124 to 147 within the major hydrophilic region (MHR) of the S gene. This determinant is one of the main targets of anti-HBs antibodies during the course of the initial immune response in acute hepatitis B.
HBs Ag immune variants with mutations in single or multiple sites of the "a" determinant may emerge as a result of selective pressure on the S protein. Indeed, these escape mutants mainly arise in vaccinated patients or in patients with orthotopic liver transplantation treated with monoclonal or polyclonal antibody to HBs Ag (5, 7, 11). Some of the common immune escape HBs Ag residue changes include G145R, D144A, P142S, Q129H I/T126N/A, and M133L (36). Site-directed mutagenesis experiments in which each amino acid was replaced with all other possible residues have since confirmed that the amino acids at positions 141 to 145 are crucial for binding of anti-HBs antibodies induced by the recombinant HBV vaccine (33). Furthermore, escape mutation in the "a" determinant may also arise in the natural course of HBV infection, resulting in active viral replication and liver disease despite seroconversion to anti-HBs in patients with chronic hepatitis B (1, 13, 25, 28, 38, 39). Similarly, several reports have identified HBs Ag mutants in serum samples that tested positive for both HBs Ag and anti-HBs (3, 19, 26). However, the limitations of these former studies were the limited number of clones or patients considered for analysis and the fact that the prevalence of naturally occurring mutations in the "a" determinant remains unknown for CHB patients. As a practical consequence, the detection of anti-HBs associated with S variants in chronic carriers may lead to a misdiagnosis of former hepatitis B virus infection if the diagnostic test fails to detect the mutated HBs Ag. The mechanism underlying the emergence of anti-HBs in chronic hepatitis B patients remains unclear, but one reason might be the selection of HBs Ag immune escape variants. To assess this hypothesis, the variability of the HBV S gene in CHB carriers was determined for patients carrying both HBs Ag and anti-HBs antibodies and patients solely positive for HBs Ag.
MATERIALS AND METHODS
Patients and sera. Eight hundred sixty-six patients are regularly followed in La Pitie-Salpetriere Hospital for CHB, defined by HBs Ag carriage beyond a 6-month period. Although the vast majority of them (91.1%) have not developed detectable anti-HBs antibody during follow-up, 77 patients (8.9%) concomitantly carry both HBs Ag and anti-HBs antibodies. For this study, we tested 26 sera from CHB patients selected on the basis of their anti-HBs antibody status. Selection criteria for patients with both HBs Ag and anti-HBs were mainly focused on anti-HBs titers at least three times above the analytical threshold of the technique (10 mIU/ml) on at least three consecutive visits.
Detection of HBV serological markers. Measurements of HBs Ag, anti-HBs antibodies, HBe Ag, and anti-HBe antibodies were done using standard, commercially available microparticle enzyme immunoassays (HBs Ag V2.0, anti-HBs V2.0, HBe Ag 2.0, and anti-HBe 2.0) (AxSYM assay; Abbott, Rungis, France). HBV DNA viral load quantification was performed using the commercially available Hybrid Capture II Digene (Abbott, Rungis, France) and HBV Monitor Cobas (Roche Diagnostics, Meylan, France) assays.
Extraction and amplification of full-length HBV genome. HBV DNA was extracted from 200 μl of each patient's serum, using a QIAmp DNA Blood mini kit (QIAGEN, Les Ulis, FRANCE), and collected in 60 μl of water. HBV DNA amplification was done according to the method described by Günther et al., with slight modifications (12). Briefly, PCR was performed in a 50-μl reaction mixture containing 10 μl of HBV DNA template, 2 mM MgSO4, a 0.2 mM concentration of each deoxynucleoside triphosphate, 60 mM Tris-SO4 (pH 8.9), 18 mM (NH4)2SO4, 0.2 μM (each) primers P1 (5'-CCG GAA AGC TTG AGC TCT TCT TTT TCA CCT CTG CCT AAT CA-3') and P2 (5'-CCG GAA AGC TTG AGC TCT TCA AAA AGT TGC ATG GTG CTG G-3'), and 1 μl of Taq Platinum High Fidelity DNA polymerase (Invitrogen, Cergy-Pontoise, France) on a thermal cycler (PTC-200 DNA engine; MJ Research). After a hot start and denaturation at 94°C for 1.5 min, 45 cycles were run, with denaturation at 94°C for 30 s, annealing at 56°C for 30 s, and elongation at 68°C for 3.5 min.
In case of unsuccessful amplification using the full-length approach, a method was developed to amplify the S protein-encoding gene. A PCR using the specific primers VTS1 (5'-TTC TTG GAA CAA GAG CTA C-3') and VT1022 (5'-GCA AAG CCC AAA AGA CCC ACA AT-3') under the same conditions as those described above, but with a 1-min elongation time, was performed to amplify the full S gene.
Cloning of full-length HBV genome. Following PCR amplification and electrophoresis, the PCR products were recovered from an agarose gel, purified using a QIAquick gel extraction kit (QIAGEN, Les Ulis, France), and cloned using a TOPO XL PCR cloning kit (Invitrogen, Cergy-Pontoise, France). Plasmids were prepared with Qiaprep 8 Turbo miniprep kits (QIAGEN, Les Ulis, France).
Sequencing of S gene. The nucleotide sequence of the S gene was determined with a BigDye Terminator v3.1 cycle sequencing ready reaction kit and run on an ABI 3100 DNA sequencer (Applied Biosystems, Les Ulis, France). The primers used for sequencing the S ORF are summarized in Table 1. Sequence analysis was performed with Seqscape software (Applied Biosystems, Les Ulis, France), Clustal software (34a), and Mutation Master software (http://tandem.bu.edu/tools.html). Briefly, genomic sequences obtained for the HBV S gene were compared with all HBV reference sequences used on the NCBI website to genotype HBV (www.ncbi.nlm.nih.gov/projects/genotyping/), as follows (accession numbers are given): subtype A, X02763, X51970, and AF090842; subtype B, D00329, AF100309, and AB033554; subtype C, X04615, M12906, and AB014381; subtype D, X65259 and X85254; subtype E, X75657 and AB03243; subtype F, X69798, AB036910, and AF223965; subtype G, AF160501, AB064310, and AF405706; and subtype H, AY090454, AY090457, and AY090460.
Statistical analysis. Comparisons were performed using the Mann-Whitney or t test for quantitative data. Statistical significance was determined at a P level of <0.05. Data were analyzed using NCSS 2001 (Kaysville, Utah).
RESULTS
Patient characteristics. Among the 864 chronically HBV-infected patients monitored at La Pitie-Salpetriere Hospital, 77 (8.9%) have been identified as carrying both HBs Ag and anti-HBs antibodies in consecutive samples. In order to analyze the mechanisms underlying the presence of both HBs Ag and anti-HBs in chronically infected patients, we selected two groups of patients on the basis of their serological status.
Fourteen CHB patients who presented with both HBs Ag and anti-HBs at consecutive visits made up group I, while 12 CHB patients solely positive for HBs Ag represented group II. As classically described, all patients were also positive for other HBV infection markers, such as anti-HBc and HBe Ag or anti-HBe. Importantly, as detailed in Table 2, the anti-HBs levels of patients from group I were rather high, with a median titer of 81 IU/liter (range, 43 to 512).
The two groups were comparable with regard to their HBe Ag status (9 HBe Ag-positive patients in each group), anti-HBe antibody status (6 patients in both groups), and viral load above 6 log copies/ml (10 and 11 patients for groups I and II, respectively). The persistence of an unchanged serological status was also taken into consideration for patient selection, as demonstrated by the median HBs Ag status in follow-ups of 35 and 92 months for patients from groups I and II, respectively. The HBV genotype distributions were heterogeneous in both groups, and two coinfections by different genotypes were observed in group I. Nine patients from group I and eight patients from group II had received treatment for their infections with either interferon or a nucleoside (or nucleotide) analogue. More importantly, two patients from group I and four patients from group II had developed lamivudine resistance, as evidenced by the presence of lamivudine resistance mutations in the polymerase gene.
Sequencing of HBV S region and statistical comparison. A cloning-sequencing approach was used to study the entire S-encoding gene sequences for patients from each group. Two pairs of primers were designed to allow for amplification of the entire S gene and, due to overlapping ORFs, the N terminus of the polymerase-encoding gene. First, a comparison of all clone sequences from each patient was performed; then, all clone sequences derived from a unique patient were compared to a consensus sequence of the same genotype to identify uncommon residues. In order to standardize our comparison, we have arbitrarily chosen to select as reference sequences those used in the NCBI genotyping tool (www.ncbi.nlm.nih.gov/projects/genotyping/) (31).
Full-length sequences generated from all of the clones were analyzed, and the percentages of residue substitutions were compared between the two groups (Fig. 1). The distributions of changes were heterogeneous along the S ORF, and changes were more prevalent for group I than for group II. For simplicity of presentation, the full-length S protein (226 amino acids [aa]) has been hereafter divided into three regions: the N-terminal region (aa 1 to 99), the MHR (aa 100 to 169), which contains the "a" determinant (aa 124 to 147), and the C-terminal region (aa 170 to 226) (34, 35). The most striking accumulation of residue changes was observed in the region encompassing the "a" determinant. Within this region, the numbers of residue changes for group I were two to four times higher than those for group II.
The overall analysis of the S protein sequences for each patient, independent of the serological status, confirmed the heterogeneous distribution of residue changes. Indeed, the highest frequency of residue changes was observed within the "a" determinant (6.25 mutated residues per 100 amino acids). The MHR (3.74 mutated residues per 100 amino acids) and the C-terminal part of the protein (3.58 mutated residues per 100 amino acids) showed less variability, but slightly more than that of the N-terminal region (2.64 mutated residues per 100 amino acids) (Table 3). For statistical analysis, the N-terminal region was taken as the background variability level since it is likely less influenced by immune pressure or antiviral treatment.
Furthermore, a comparison of variability along the S protein indicated a marked difference between patients in groups I and II (Fig. 1; Table 3). When considering the full-length S coding region, the number of mutated residues for group I patients was markedly increased compared to that for group II patients (4.55 versus 1.66 substitutions per 100 amino acids; P = 0.009) (Table 3). Importantly, the amino acid exchange distribution along the S protein indicated that residue changes occurred mostly within the MHR, at 5.71 versus 1.43 mutated residues per 100 amino acids (P = 0.004) for groups I and II, respectively. This difference was even more striking for the "a" determinant (9.52 versus 2.43 mutated residues per 100 amino acids; P = 0.009) (Table 3). These results indicate that the presence of anti-HBs concomitantly with HBs Ag is associated with an accumulation of mutated residues, mostly, but not only, within the antigenic loops. For the S protein C-terminal region, the residue change frequencies between groups I and II were comparable (3.88 versus 3.22 substitutions per 100 amino acids; P was not significant). It should be pointed out that this region overlaps the polymerase ORF and was likely affected by resistance mutations induced by antiviral treatments.
The accumulation of several residue changes for group I patients within the "a" determinant was also much more frequent than that for group II patients (Fig. 2). Indeed, 10 patients (71%) from group I and only 3 patients (25%) from group II carried at least two residue changes in the antigenic loops. The accumulation of several residue changes is likely to change the immunogenicity of the protein.
The percentages of MHR variants infecting individual patients were also different between the two groups (Table 4). Patients from group I were more likely infected by a homogeneous population of clones carrying one or more mutations. Inversely, patients from group II often had a minor population of mutated HBV, with many clones showing no particular mutation. One should point out that for each group I patient, the presence of mutated clones was always associated with the detection of at least one wild-type HBV clone. Interestingly, it was noticed that the same patient could be infected by a mixture of HBV strains with several associations of residue changes in the "a" determinant of the S protein.
The most frequent changes observed in anti-HBs patients were located at positions s145, s126, and s129, but changes were also seen throughout loops 2 to 4 of the MHR. Furthermore, only patients from group I had the G145R, Q129N, I/T126 S/N, D144A, and T123N amino acid changes, which are known to alter immunogenicity.
Analysis of polymerase reading frame. For group I patients, the numbers of residue changes between the S and the polymerase reading frames were compared for the overlapping region encompassing the antigenic loops. The frequency of residue changes was less important for the polymerase than for the S reading frame, at 1.92 versus 3.74 substitutions per 100 amino acids (P = 0.003), respectively, and this difference was even accentuated in the "a" determinant (6.25 versus 3.53 substitutions per 100 amino acids; P = 0.001) overlapping region. This difference in residue substitution between the HBs Ag and polymerase reading frames was not significant for group II patients.
The absence of a difference in the C-terminal portion of the S protein with regard to the number of residue changes between patients in groups I and II is likely explained by the influence of resistance mutations detected in the polymerase reading frame. Indeed, two patients from group I and four patients from group II had been treated with lamivudine and had developed lamivudine-resistant HBV strains.
DISCUSSION
In contrast to what is usually described as a favorable outcome for CHB, the acquisition of anti-HBs antibody is not systematically associated with the loss of HBs Ag. Indeed, the coexistence of HBs Ag and anti-HBs in chronic hepatitis B virus carriers has been described for 10 to 25% of patients by several authors (19, 27). In accordance with these studies, the presence of both HBs Ag and antibody was documented in almost 9% of 864 chronic hepatitis B virus carriers followed in our hospital. The mechanism underlying the presence of both HBs Ag and anti-HBs despite viral replication is unknown, but one reason might be the selection of HBs Ag immune variants. Indeed, an accumulation of residue changes in the MHR of the S protein, the main target of anti-HBs, could lead to escape from recognition by the host immune system. Several studies have described point mutations resulting in amino acid changes in the S protein antigenic loops in vaccines and hepatitis B immune globulin recipients (5, 10, 14, 36). However, HBV escape mutants may also arise naturally in chronic hepatitis B virus carriers due to the sole pressure of the host immune system (13, 19, 22, 25, 26, 28, 39). This hypothesis is reinforced by the results of our study based on a clonal analysis of HBV strains isolated from CHB patients with (n = 14) or without (n = 12) anti-HBs. It should be stressed that one important criterion for selection of our 14 patients in group I was the detection of high anti-HBs titers in several consecutive samples despite detectable viral replication.
In comparing HBs Ag sequences from both groups, it was striking to notice an accumulation of residue changes in strains from Ag/anti-HBs patients (group I) compared to HBs Ag-only carriers (group II). Moreover, the distribution of these amino acid changes was not homogeneous along the protein but, rather, gathered within the MHR, including the highly conformational and cysteine-rich "a" determinant. This antigenic loop region of HBs Ag is described as the main target of the humoral response, and any change of its primary sequence may alter its antigenic structure and would render any anti-HBs humoral defense against this region less effective (29).
Two previous studies using a similar approach, but on a more limited number of patients and clones, have characterized the most prevalent residues affected by such changes (19, 39). Thus, positions 145, 144, 129, 126, and 130 were the most likely to be modified. The description of these residue changes, particularly the G145R change, is remarkable because these substitutions also correspond to common mutations described for HBV vaccine escape variants or for patients who have been therapeutically administered monoclonal anti-HBs (14, 36). One should also point out that these very residues are part of MHR loops 3 and 4 and that other residues within these same loops were also affected by additional changes. In our study, the most frequent residue changes in the MHR were observed principally in patients from group I with both markers (71% versus 22% of patients carried more than one mutation), and they mainly included sG145R, sI/T126N/S, sQ129N, and sD144A. Our results are comparable to those observed by Yamamoto et al. and Kohno et al. but included a larger number of patients and clones per patient. It is important that residue changes observed in Ag/anti-HBs patients are not only at these characteristic positions but also throughout loops 2 to 4 of the antigenic loops, but it is difficult to predict the structural and biochemical effects of these amino acid substitutions. Nevertheless, it is likely that residue changes located at position s146, representing an N-glycosylation site, or s147, the location of a disulfide bridge, could alter the immunological properties of the S protein (19, 38). Altogether, several residue changes within the MHR that are specific to patients carrying both HBs Ag and anti-HBs antibodies need to be further characterized immunologically. Although it seems quite convincing that these changes may alter antibody recognition of the S protein, very little is known about the consequences of such modifications on T-cell epitope recognition. As recently described by Bauer et al. for vaccines, accumulations of changes in CHB patients' T-helper-cell epitopes may also alter immune response efficacies and may facilitate virus persistence (4).
Overall, the accumulation of residue changes in and around the "a" determinant may generate HBs variants with alterations of antigenicity and possibly T-cell epitope structure, leading to a subsequent failure to neutralize HBV (5). Surprisingly, our results are quite similar to those observed by Weinberger et al., who described a high genetic variability of the "a" determinant region for HBs Ag-negative chronic virus carriers (38). In the former study, patients were selected on the basis of anti-HBc reactivity, and a comparison was made according to the detection or lack of detection of HBs Ag. Due to different experimental approaches, i.e., direct sequencing in Weinberger's study and clonal analysis in our work, a comparison of the results is difficult. However, our findings underscore an increased protein variability of the MHR in Ag/anti-HBs-positive patients. The amino acid exchanges found by Weinberger et al. in HBs Ag-negative patients were also found in some clones from Ag/anti-HBs-positive patients from our study. In both cases, a very appealing hypothesis is the selection of immune escape mutants under strong host immune pressure.
The analysis of group I patients revealed the presence of a viral mixture consisting of the wild type and antigenic loop variants, suggesting that the acquisition of anti-HBs is associated with an emergence of immune escape variants persisting for several years in association with wild-type strains. Our study suggests that the presence of HBs Ag variants would arise naturally in chronic carriers and is likely a viral strategy to evade the immune system. In light of this hypothesis, it is possible that the detectable HBs Ag (as measured by the diagnostic assay) is a mixture of the wild type and variants, whereas the detectable anti-HBs antibodies are targeted against the wild-type viruses. In order to eliminate bias due to the serological assay used for anti-HBs determination and to confirm the specificity of the circulating anti-HBs antibodies for wild-type HBs Ag, all patients' anti-HBs titers were controlled using a second anti-HBs assay (Monolisa anti-HBs Plus; Bio-Rad, France), with very comparable results (data not shown).
The selection of immune escape mutants during the natural history of the disease could have several deleterious consequences. First, the accumulation of mutated residues within the MHR may lead to an absence of HBs Ag recognition by commercially available diagnostic tests. This issue has been largely reported by several authors (9, 37). Indeed, many assays are unable to detect G145R mutants, and some of them have dramatically decreased sensitivities for the detection of other mutants. The reasons for such failure depend on the type of antibody (monoclonal or polyclonal) and the targeted epitope used in the assays, with some tests being largely affected by mutations within loops 2 to 4 (6, 9, 10, 16, 18).
Second, the accumulation of HBV carrying MHR mutations in CHB patients with relatively high viral loads raises the problem of transmission of such variants (8, 15). As previously mentioned, MHR variants may not be fully neutralized by vaccine-induced antibodies, and we are still lacking data on their possible transmission to vaccinated subjects. In light of our data, chronic carriers with both Ag and anti-HBs could be considered potential reservoirs of immune escape variants. Further epidemiological studies are needed to scientifically assess the potential threat of such chronic carriers in areas of high endemicity where infant vaccination has been implemented (15).
Third, when screening a patient for the presence of vaccine-induced antibodies, it may also be useful to propose at least the detection of HBs Ag since both markers may be present simultaneously. Testing for anti-HBc antibodies may also be a wise strategy.
We also noted that amino acids located at positions 122 and 160, which determine the d/y and w/r subtypes of HBs Ag, were completely conserved (2). Indeed, amino acid substitutions located at these positions result in changes of the HBs Ag subtype (32). Thus, we may observe over time in the same patient a co-occurrence of HBs Ag particles of a certain subtype and anti-HBs with specificity for another subtypic determinant not borne by the circulating particles. Indeed, for patient 3 (P3) belonging to group I, no amino acid substitutions were evidenced in any of the tested clones, despite the presence of anti-HBs. As recently demonstrated, it is possible that in this particular case the detected antibodies targeted a previously present serotype; however, this hypothesis could only be ruled out by a longitudinal study over time on serial samples from the same patient (24). This mechanism could be another way for HBV variants to persist in spite of immune pressure (19, 28).
Although it was not possible to thoroughly assess the natural history of the disease in this study, the coexistence of HBs Ag and anti-HBs does not seem to affect disease evolution. However, a larger study without selection bias linked to ethnicity, duration of infection, age at infection, and genotype would be needed to specifically address this issue. In light of the recently described selection of a secretion-incompetent variant by Kalinina et al., one may wonder if MHR variants are associated with certain forms of disease (17).
The selection process leading to the emergence of HBs variants could therefore be compared to the well-described selection of HBe-negative variants over the course of chronic hepatitis B virus carriage.
REFERENCES
Asahina, Y., N. Enomoto, Y. Ogura, M. Kurosaki, I. Sakuma, N. Izumi, F. Marumo, and C. Sato. 1996. Sequential changes in full-length genomes of hepatitis B virus accompanying acute exacerbation of chronic hepatitis B. J. Hepatol. 25:787-794.
Ashton-Rickardt, P. G., and K. Murray. 1989. Mutations that change the immunological subtype of hepatitis B virus surface antigen and distinguish between antigenic and immunogenic determination. J. Med. Virol. 29:204-214.
Bahn, A., P. Gerner, U. Martine, F. Bortolotti, and S. Wirth. 1997. Detection of different viral strains of hepatitis B virus in chronically infected children after seroconversion from HBsAg to anti-HBs indicating viral persistence. J. Hepatol. 27:973-978.
Bauer, T., K. Weinberger, and W. Jilg. 2002. Variants of two major T cell epitopes within the hepatitis B surface antigen are not recognized by specific T helper cells of vaccinated individuals. Hepatology 35:455-465.
Carman, W. F. 1997. The clinical significance of surface antigen variants of hepatitis B virus. J. Viral Hepat. 4:11-20.
Carman, W. F., J. Korula, L. Wallace, R. MacPhee, L. Mimms, and R. Decker. 1995. Fulminant reactivation of hepatitis B due to envelope protein mutant that escaped detection by monoclonal HBsAg ELISA. Lancet 345:1406-1407.
Carman, W. F., A. R. Zanetti, P. Karayiannis, J. Waters, G. Manzillo, E. Tanzi, A. J. Zuckerman, and H. C. Thomas. 1990. Vaccine-induced escape mutant of hepatitis B virus. Lancet 336:325-329.
Chakravarty, R., M. Neogi, S. Roychowdhury, and C. K. Panda. 2002. Presence of hepatitis B surface antigen mutant G145R DNA in the peripheral blood leukocytes of the family members of an asymptomatic carrier and evidence of its horizontal transmission. Virus Res. 90:133-141.
Coleman, P. F., Y. C. Chen, and I. K. Mushahwar. 1999. Immunoassay detection of hepatitis B surface antigen mutants. J. Med. Virol. 59:19-24.
Cooreman, M. P., G. Leroux-Roels, and W. P. Paulij. 2001. Vaccine- and hepatitis B immune globulin-induced escape mutations of hepatitis B virus surface antigen. J. Biomed. Sci. 8:237-247.
Ghany, M. G., B. Ayola, F. G. Villamil, R. G. Gish, S. Rojter, J. M. Vierling, and A. S. Lok. 1998. Hepatitis B virus S mutants in liver transplant recipients who were reinfected despite hepatitis B immune globulin prophylaxis. Hepatology 27:213-222.
Günther, S., B. C. Li, S. Miska, D. H. Kruger, H. Meisel, and H. Will. 1995. A novel method for efficient amplification of whole hepatitis B virus genomes permits rapid functional analysis and reveals deletion mutants in immunosuppressed patients. J. Virol. 69:5437-5444.
Hou, J., Z. Wang, J. Cheng, Y. Lin, G. K. Lau, J. Sun, F. Zhou, J. Waters, P. Karayiannis, and K. Luo. 2001. Prevalence of naturally occurring surface gene variants of hepatitis B virus in nonimmunized surface antigen-negative Chinese carriers. Hepatology 34:1027-1034.
Hsu, H. Y., M. H. Chang, Y. H. Ni, and H. L. Chen. 2004. Survey of hepatitis B surface variant infection in children 15 years after a nationwide vaccination programme in Taiwan. Gut 53:1499-1503.
Huang, X., D. Lu, G. Ji, Y. Sun, L. Ma, Z. Chen, L. Zhang, J. Huang, L. Yu, and S. Levicnik-Stezinar. 2004. Hepatitis B virus (HBV) vaccine-induced escape mutants of HBV S gene among children from Qidong area, China. Virus Res. 99:63-68.
Jeantet, D., I. Chemin, B. Mandrand, A. Tran, F. Zoulim, P. Merle, C. Trepo, and A. Kay. 2004. Cloning and expression of surface antigens from occult chronic hepatitis B virus infections and their recognition by commercial detection assays. J. Med. Virol. 73:508-515.
Kalinina, T., A. Riu, L. Fischer, T. Santantonio, H. Will, and M. Sterneck. 2003. Selection of a secretion-incompetent mutant in the serum of a patient with severe hepatitis B. Gastroenterology 125:1077-1084.
Kfoury Baz, E. M., J. Zheng, K. Mazuruk, A. Van Le, and D. L. Peterson. 2001. Characterization of a novel hepatitis B virus mutant: demonstration of mutation-induced hepatitis B virus surface antigen group specific "a" determinant conformation change and its application in diagnostic assays. Transfus. Med. 11:355-362.
Kohno, H., T. Inoue, F. Tsuda, H. Okamoto, and Y. Akahane. 1996. Mutations in the envelope gene of hepatitis B virus variants co-occurring with antibody to surface antigen in sera from patients with chronic hepatitis B. J. Gen. Virol. 77:1825-1831.
Lai, C. L., V. Ratziu, M. F. Yuen, and T. Poynard. 2003. Viral hepatitis B. Lancet 362:2089-2094.
Lavanchy, D. 2004. Hepatitis B virus epidemiology, disease burden, treatment, and current and emerging prevention and control measures. J. Viral Hepat. 11:97-107.
Liu, C. J., J. H. Kao, W. Y. Shau, P. J. Chen, M. Y. Lai, and D. S. Chen. 2002. Naturally occurring hepatitis B surface gene variants in chronic hepatitis B virus infection: correlation with viral serotypes and clinical stages of liver disease. J. Med. Virol. 68:50-59.
Lok, A. S. 2000. Hepatitis B infection: pathogenesis and management. J. Hepatol. 32:89-97.
Margeridon, S., A. Lachaux, C. Trepo, F. Zoulim, and A. Kay. 2005. A quasi-monoclonal anti-HBs response can lead to immune escape of ‘wild-type’ hepatitis B virus. J. Gen. Virol. 86:1687-1693.
Mathet, V. L., M. Feld, L. Espinola, D. O. Sanchez, V. Ruiz, O. Mando, G. Carballal, J. F. Quarleri, F. D'Mello, C. R. Howard, J. R. Oubina, F. Clement, P. Dewint, G. Leroux-Roels, S. Louisirirotchanakul, C. Kanoksinsombat, A. Theamboonlert, P. Puthavatana, C. Wasi, Y. Poovorawan, N. Khan, M. Guarnieri, S. H. Ahn, J. Li, Y. Zhou, G. Bang, K. H. Kim, J. R. Wands, S. Tong, H. Singh, R. Aggarwal, R. L. Singh, S. R. Naik, S. Naik, C. Osiowy, W. N. Chen, C. J. Oon, and M. C. Moh. 2003. Hepatitis B virus S gene mutants in a patient with chronic active hepatitis with circulating Anti-HBs antibodies. J. Med. Virol. 69:18-26.
Mesenas, S. J., W. C. Chow, Y. Zhao, G. K. Lim, C. J. Oon, and H. S. Ng. 2002. Wild-type and ‘a’ epitope variants in chronic hepatitis B virus carriers positive for hepatitis B surface antigen and antibody. J. Gastroenterol. Hepatol. 17:148-152.
Mimms, L. 1995. Hepatitis B virus escape mutants: pushing the envelope of chronic hepatitis B virus infection. Hepatology 21:884-887.
Ogura, Y., M. Kurosaki, Y. Asahina, N. Enomoto, F. Marumo, and C. Sato. 1999. Prevalence and significance of naturally occurring mutations in the surface and polymerase genes of hepatitis B virus. J. Infect. Dis. 180:1444-1451.
Oon, C. J., K. L. Tan, T. Harrison, and A. Zuckerman. 1996. Natural history of hepatitis B surface antigen mutants in children. Lancet 348:1524.
Rehermann, B., and M. Nascimbeni. 2005. Immunology of hepatitis B virus and hepatitis C virus infection. Nat. Rev. Immunol. 5:215-229.
Rozanov, M., U. Plikat, C. Chappey, A. Kochergin, and T. Tatusova. 2004. A web-based genotyping resource for viral sequences. Nucleic Acids Res. 32:W654-W659.
Schaefer, S. 2005. Hepatitis B virus: significance of genotypes. J. Viral Hepat. 12:111-124.
Steward, M. W., C. D. Partidos, F. D'Mello, and C. R. Howard. 1993. Specificity of antibodies reactive with hepatitis B surface antigen following immunization with synthetic peptides. Vaccine 11:1405-1414.
Stirk, H. J., J. M. Thornton, and C. R. Howard. 1992. A topological model for hepatitis B surface antigen. Intervirology 33:148-158.
Thompson, J. D., T. J. Gibson, F. Plewniak, F. Jeanmougin, and D. G. Higgins. 1997. The CLUSTAL_X Windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25:4876-4882.
Tiollais, P., C. Pourcel, and A. Dejean. 1985. The hepatitis B virus. Nature 317:489-495.
Torresi, J., L. Earnest-Silveira, G. Civitico, T. E. Walters, S. R. Lewin, J. Fyfe, S. A. Locarnini, M. Manns, C. Trautwein, T. C. Bock, G. Deliyannis, K. Edgtton, H. Zhuang, T. Sozzi, and D. C. Jackson. 2002. The virological and clinical significance of mutations in the overlapping envelope and polymerase genes of hepatitis B virus. J. Clin. Virol. 25:97-106.
Weber, B. 2005. Genetic variability of the S gene of hepatitis B virus: clinical and diagnostic impact. J. Clin. Virol. 32:102-112.
Weinberger, K. M., T. Bauer, S. Bohm, and W. Jilg. 2000. High genetic variability of the group-specific a-determinant of hepatitis B virus surface antigen (HBsAg) and the corresponding fragment of the viral polymerase in chronic virus carriers lacking detectable HBsAg in serum. J. Gen. Virol. 81:1165-1174.
Yamamoto, K., M. Horikita, F. Tsuda, K. Itoh, Y. Akahane, S. Yotsumoto, H. Okamoto, Y. Miyakawa, and M. Mayumi. 1994. Naturally occurring escape mutants of hepatitis B virus with various mutations in the S gene in carriers seropositive for antibody to hepatitis B surface antigen. J. Virol. 68:2671-2676.(Olivier Lada, Yves Benham)
Hepato-Gastro-Enterology, AP-HP Pitie-Salpetriere Hospital, Paris, France
ABSTRACT
In chronic hepatitis B (CHB), the persistence of hepatitis B surface antigen (HBs Ag) is sometimes associated with antibodies (Ab) to HBs (anti-HBs). To assess the hypothesis of the selection of HBs Ag immune escape variants in CHB patients, the variability of the HBV S gene was determined for patients persistently carrying both HBs Ag and anti-HBs antibodies and patients solely positive for HBs Ag. We selected 14 patients who presented both markers (group I) in several consecutive samples and 12 patients positive for HBs Ag only (group II). The HBs Ag-encoding gene was amplified and cloned, and at least 15 clones per patient were sequenced and analyzed. The number of residue changes within the S protein was 2.7 times more frequent for group I than for group II patients and occurred mostly in the "a" determinant of the major hydrophilic region (MHR), with 9.52 versus 2.43 changes per 100 residues (P = 0.009), respectively. Ten patients (71%) from group I, but only three (25%) from group II, presented at least two residue changes in the MHR. The most frequent changes in group I patients were located at positions s145, s129, s126, s144, and s123, as described for immune escape variants. In CHB patients, the coexistence of HBs Ag and anti-HBs Ab is associated with an increase of "a" determinant variability, suggesting a selection of HBV immune escape mutants during chronic carriage. The consequences of this selection process with regard to vaccine efficacy, diagnosis, and clinical evolution remain partially unknown.
INTRODUCTION
More than 350 million people, or 5% of the world's population, are chronic carriers of hepatitis B virus (HBV), and this infection represents a worldwide public health problem (21). Many patients chronically infected with HBV, as defined by the persistence over >6 months of hepatitis B surface antigen (HBs Ag), will develop life-threatening diseases such as liver cirrhosis and hepatocellular carcinoma. It has been estimated that up to 30% of them will die from the consequences of their infection (20).
In the natural course of HBV infection, virus clearance likely results from the intimate coordination of both humoral and cellular immune systems. Overall, the antibody-mediated immune response to HBV proteins aims at the clearance of circulating HBV particles, whereas the cellular effectors contribute to eliminating infected hepatocytes (30). Biologically, virus clearance is classically characterized by the emergence of anti-HBs antibody (Ab) in the serological profile.
During chronic hepatitis B infection, two clinically important phases can be defined. While the immune tolerance phase is usually characterized by little liver damage and the presence of hepatitis B e antigen (HBe Ag) in the serum, the second phase can be described as more aggressive for the liver, with selection of HBe Ag-negative variants and detection of anti-HBe antibody. Ultimately, chronic carriers can be classified as inactive carriers, with little viral replication and anti-HBe antibody and normal liver biochemical markers, or as chronic hepatitis patients, with abnormal liver enzyme levels and higher viral loads (23).
Notably, several reports have described the persistence of HBs Ag associated with anti-HBs antibodies in 10 to 25% of chronic hepatitis B (CHB) patients (19, 27). The mechanism underlying the presence of both HBs Ag and anti-HBs antibodies remains unknown, but one possibility could be the selection of immune escape mutants.
The envelope gene of HBV has three open reading frames (ORF), PreS1, PreS2, and S, which code for three proteins, the small, middle, and large HBs Ag, translated from distinct mRNAs. Common to all three proteins, the "a" dominant epitope is located at codon positions 124 to 147 within the major hydrophilic region (MHR) of the S gene. This determinant is one of the main targets of anti-HBs antibodies during the course of the initial immune response in acute hepatitis B.
HBs Ag immune variants with mutations in single or multiple sites of the "a" determinant may emerge as a result of selective pressure on the S protein. Indeed, these escape mutants mainly arise in vaccinated patients or in patients with orthotopic liver transplantation treated with monoclonal or polyclonal antibody to HBs Ag (5, 7, 11). Some of the common immune escape HBs Ag residue changes include G145R, D144A, P142S, Q129H I/T126N/A, and M133L (36). Site-directed mutagenesis experiments in which each amino acid was replaced with all other possible residues have since confirmed that the amino acids at positions 141 to 145 are crucial for binding of anti-HBs antibodies induced by the recombinant HBV vaccine (33). Furthermore, escape mutation in the "a" determinant may also arise in the natural course of HBV infection, resulting in active viral replication and liver disease despite seroconversion to anti-HBs in patients with chronic hepatitis B (1, 13, 25, 28, 38, 39). Similarly, several reports have identified HBs Ag mutants in serum samples that tested positive for both HBs Ag and anti-HBs (3, 19, 26). However, the limitations of these former studies were the limited number of clones or patients considered for analysis and the fact that the prevalence of naturally occurring mutations in the "a" determinant remains unknown for CHB patients. As a practical consequence, the detection of anti-HBs associated with S variants in chronic carriers may lead to a misdiagnosis of former hepatitis B virus infection if the diagnostic test fails to detect the mutated HBs Ag. The mechanism underlying the emergence of anti-HBs in chronic hepatitis B patients remains unclear, but one reason might be the selection of HBs Ag immune escape variants. To assess this hypothesis, the variability of the HBV S gene in CHB carriers was determined for patients carrying both HBs Ag and anti-HBs antibodies and patients solely positive for HBs Ag.
MATERIALS AND METHODS
Patients and sera. Eight hundred sixty-six patients are regularly followed in La Pitie-Salpetriere Hospital for CHB, defined by HBs Ag carriage beyond a 6-month period. Although the vast majority of them (91.1%) have not developed detectable anti-HBs antibody during follow-up, 77 patients (8.9%) concomitantly carry both HBs Ag and anti-HBs antibodies. For this study, we tested 26 sera from CHB patients selected on the basis of their anti-HBs antibody status. Selection criteria for patients with both HBs Ag and anti-HBs were mainly focused on anti-HBs titers at least three times above the analytical threshold of the technique (10 mIU/ml) on at least three consecutive visits.
Detection of HBV serological markers. Measurements of HBs Ag, anti-HBs antibodies, HBe Ag, and anti-HBe antibodies were done using standard, commercially available microparticle enzyme immunoassays (HBs Ag V2.0, anti-HBs V2.0, HBe Ag 2.0, and anti-HBe 2.0) (AxSYM assay; Abbott, Rungis, France). HBV DNA viral load quantification was performed using the commercially available Hybrid Capture II Digene (Abbott, Rungis, France) and HBV Monitor Cobas (Roche Diagnostics, Meylan, France) assays.
Extraction and amplification of full-length HBV genome. HBV DNA was extracted from 200 μl of each patient's serum, using a QIAmp DNA Blood mini kit (QIAGEN, Les Ulis, FRANCE), and collected in 60 μl of water. HBV DNA amplification was done according to the method described by Günther et al., with slight modifications (12). Briefly, PCR was performed in a 50-μl reaction mixture containing 10 μl of HBV DNA template, 2 mM MgSO4, a 0.2 mM concentration of each deoxynucleoside triphosphate, 60 mM Tris-SO4 (pH 8.9), 18 mM (NH4)2SO4, 0.2 μM (each) primers P1 (5'-CCG GAA AGC TTG AGC TCT TCT TTT TCA CCT CTG CCT AAT CA-3') and P2 (5'-CCG GAA AGC TTG AGC TCT TCA AAA AGT TGC ATG GTG CTG G-3'), and 1 μl of Taq Platinum High Fidelity DNA polymerase (Invitrogen, Cergy-Pontoise, France) on a thermal cycler (PTC-200 DNA engine; MJ Research). After a hot start and denaturation at 94°C for 1.5 min, 45 cycles were run, with denaturation at 94°C for 30 s, annealing at 56°C for 30 s, and elongation at 68°C for 3.5 min.
In case of unsuccessful amplification using the full-length approach, a method was developed to amplify the S protein-encoding gene. A PCR using the specific primers VTS1 (5'-TTC TTG GAA CAA GAG CTA C-3') and VT1022 (5'-GCA AAG CCC AAA AGA CCC ACA AT-3') under the same conditions as those described above, but with a 1-min elongation time, was performed to amplify the full S gene.
Cloning of full-length HBV genome. Following PCR amplification and electrophoresis, the PCR products were recovered from an agarose gel, purified using a QIAquick gel extraction kit (QIAGEN, Les Ulis, France), and cloned using a TOPO XL PCR cloning kit (Invitrogen, Cergy-Pontoise, France). Plasmids were prepared with Qiaprep 8 Turbo miniprep kits (QIAGEN, Les Ulis, France).
Sequencing of S gene. The nucleotide sequence of the S gene was determined with a BigDye Terminator v3.1 cycle sequencing ready reaction kit and run on an ABI 3100 DNA sequencer (Applied Biosystems, Les Ulis, France). The primers used for sequencing the S ORF are summarized in Table 1. Sequence analysis was performed with Seqscape software (Applied Biosystems, Les Ulis, France), Clustal software (34a), and Mutation Master software (http://tandem.bu.edu/tools.html). Briefly, genomic sequences obtained for the HBV S gene were compared with all HBV reference sequences used on the NCBI website to genotype HBV (www.ncbi.nlm.nih.gov/projects/genotyping/), as follows (accession numbers are given): subtype A, X02763, X51970, and AF090842; subtype B, D00329, AF100309, and AB033554; subtype C, X04615, M12906, and AB014381; subtype D, X65259 and X85254; subtype E, X75657 and AB03243; subtype F, X69798, AB036910, and AF223965; subtype G, AF160501, AB064310, and AF405706; and subtype H, AY090454, AY090457, and AY090460.
Statistical analysis. Comparisons were performed using the Mann-Whitney or t test for quantitative data. Statistical significance was determined at a P level of <0.05. Data were analyzed using NCSS 2001 (Kaysville, Utah).
RESULTS
Patient characteristics. Among the 864 chronically HBV-infected patients monitored at La Pitie-Salpetriere Hospital, 77 (8.9%) have been identified as carrying both HBs Ag and anti-HBs antibodies in consecutive samples. In order to analyze the mechanisms underlying the presence of both HBs Ag and anti-HBs in chronically infected patients, we selected two groups of patients on the basis of their serological status.
Fourteen CHB patients who presented with both HBs Ag and anti-HBs at consecutive visits made up group I, while 12 CHB patients solely positive for HBs Ag represented group II. As classically described, all patients were also positive for other HBV infection markers, such as anti-HBc and HBe Ag or anti-HBe. Importantly, as detailed in Table 2, the anti-HBs levels of patients from group I were rather high, with a median titer of 81 IU/liter (range, 43 to 512).
The two groups were comparable with regard to their HBe Ag status (9 HBe Ag-positive patients in each group), anti-HBe antibody status (6 patients in both groups), and viral load above 6 log copies/ml (10 and 11 patients for groups I and II, respectively). The persistence of an unchanged serological status was also taken into consideration for patient selection, as demonstrated by the median HBs Ag status in follow-ups of 35 and 92 months for patients from groups I and II, respectively. The HBV genotype distributions were heterogeneous in both groups, and two coinfections by different genotypes were observed in group I. Nine patients from group I and eight patients from group II had received treatment for their infections with either interferon or a nucleoside (or nucleotide) analogue. More importantly, two patients from group I and four patients from group II had developed lamivudine resistance, as evidenced by the presence of lamivudine resistance mutations in the polymerase gene.
Sequencing of HBV S region and statistical comparison. A cloning-sequencing approach was used to study the entire S-encoding gene sequences for patients from each group. Two pairs of primers were designed to allow for amplification of the entire S gene and, due to overlapping ORFs, the N terminus of the polymerase-encoding gene. First, a comparison of all clone sequences from each patient was performed; then, all clone sequences derived from a unique patient were compared to a consensus sequence of the same genotype to identify uncommon residues. In order to standardize our comparison, we have arbitrarily chosen to select as reference sequences those used in the NCBI genotyping tool (www.ncbi.nlm.nih.gov/projects/genotyping/) (31).
Full-length sequences generated from all of the clones were analyzed, and the percentages of residue substitutions were compared between the two groups (Fig. 1). The distributions of changes were heterogeneous along the S ORF, and changes were more prevalent for group I than for group II. For simplicity of presentation, the full-length S protein (226 amino acids [aa]) has been hereafter divided into three regions: the N-terminal region (aa 1 to 99), the MHR (aa 100 to 169), which contains the "a" determinant (aa 124 to 147), and the C-terminal region (aa 170 to 226) (34, 35). The most striking accumulation of residue changes was observed in the region encompassing the "a" determinant. Within this region, the numbers of residue changes for group I were two to four times higher than those for group II.
The overall analysis of the S protein sequences for each patient, independent of the serological status, confirmed the heterogeneous distribution of residue changes. Indeed, the highest frequency of residue changes was observed within the "a" determinant (6.25 mutated residues per 100 amino acids). The MHR (3.74 mutated residues per 100 amino acids) and the C-terminal part of the protein (3.58 mutated residues per 100 amino acids) showed less variability, but slightly more than that of the N-terminal region (2.64 mutated residues per 100 amino acids) (Table 3). For statistical analysis, the N-terminal region was taken as the background variability level since it is likely less influenced by immune pressure or antiviral treatment.
Furthermore, a comparison of variability along the S protein indicated a marked difference between patients in groups I and II (Fig. 1; Table 3). When considering the full-length S coding region, the number of mutated residues for group I patients was markedly increased compared to that for group II patients (4.55 versus 1.66 substitutions per 100 amino acids; P = 0.009) (Table 3). Importantly, the amino acid exchange distribution along the S protein indicated that residue changes occurred mostly within the MHR, at 5.71 versus 1.43 mutated residues per 100 amino acids (P = 0.004) for groups I and II, respectively. This difference was even more striking for the "a" determinant (9.52 versus 2.43 mutated residues per 100 amino acids; P = 0.009) (Table 3). These results indicate that the presence of anti-HBs concomitantly with HBs Ag is associated with an accumulation of mutated residues, mostly, but not only, within the antigenic loops. For the S protein C-terminal region, the residue change frequencies between groups I and II were comparable (3.88 versus 3.22 substitutions per 100 amino acids; P was not significant). It should be pointed out that this region overlaps the polymerase ORF and was likely affected by resistance mutations induced by antiviral treatments.
The accumulation of several residue changes for group I patients within the "a" determinant was also much more frequent than that for group II patients (Fig. 2). Indeed, 10 patients (71%) from group I and only 3 patients (25%) from group II carried at least two residue changes in the antigenic loops. The accumulation of several residue changes is likely to change the immunogenicity of the protein.
The percentages of MHR variants infecting individual patients were also different between the two groups (Table 4). Patients from group I were more likely infected by a homogeneous population of clones carrying one or more mutations. Inversely, patients from group II often had a minor population of mutated HBV, with many clones showing no particular mutation. One should point out that for each group I patient, the presence of mutated clones was always associated with the detection of at least one wild-type HBV clone. Interestingly, it was noticed that the same patient could be infected by a mixture of HBV strains with several associations of residue changes in the "a" determinant of the S protein.
The most frequent changes observed in anti-HBs patients were located at positions s145, s126, and s129, but changes were also seen throughout loops 2 to 4 of the MHR. Furthermore, only patients from group I had the G145R, Q129N, I/T126 S/N, D144A, and T123N amino acid changes, which are known to alter immunogenicity.
Analysis of polymerase reading frame. For group I patients, the numbers of residue changes between the S and the polymerase reading frames were compared for the overlapping region encompassing the antigenic loops. The frequency of residue changes was less important for the polymerase than for the S reading frame, at 1.92 versus 3.74 substitutions per 100 amino acids (P = 0.003), respectively, and this difference was even accentuated in the "a" determinant (6.25 versus 3.53 substitutions per 100 amino acids; P = 0.001) overlapping region. This difference in residue substitution between the HBs Ag and polymerase reading frames was not significant for group II patients.
The absence of a difference in the C-terminal portion of the S protein with regard to the number of residue changes between patients in groups I and II is likely explained by the influence of resistance mutations detected in the polymerase reading frame. Indeed, two patients from group I and four patients from group II had been treated with lamivudine and had developed lamivudine-resistant HBV strains.
DISCUSSION
In contrast to what is usually described as a favorable outcome for CHB, the acquisition of anti-HBs antibody is not systematically associated with the loss of HBs Ag. Indeed, the coexistence of HBs Ag and anti-HBs in chronic hepatitis B virus carriers has been described for 10 to 25% of patients by several authors (19, 27). In accordance with these studies, the presence of both HBs Ag and antibody was documented in almost 9% of 864 chronic hepatitis B virus carriers followed in our hospital. The mechanism underlying the presence of both HBs Ag and anti-HBs despite viral replication is unknown, but one reason might be the selection of HBs Ag immune variants. Indeed, an accumulation of residue changes in the MHR of the S protein, the main target of anti-HBs, could lead to escape from recognition by the host immune system. Several studies have described point mutations resulting in amino acid changes in the S protein antigenic loops in vaccines and hepatitis B immune globulin recipients (5, 10, 14, 36). However, HBV escape mutants may also arise naturally in chronic hepatitis B virus carriers due to the sole pressure of the host immune system (13, 19, 22, 25, 26, 28, 39). This hypothesis is reinforced by the results of our study based on a clonal analysis of HBV strains isolated from CHB patients with (n = 14) or without (n = 12) anti-HBs. It should be stressed that one important criterion for selection of our 14 patients in group I was the detection of high anti-HBs titers in several consecutive samples despite detectable viral replication.
In comparing HBs Ag sequences from both groups, it was striking to notice an accumulation of residue changes in strains from Ag/anti-HBs patients (group I) compared to HBs Ag-only carriers (group II). Moreover, the distribution of these amino acid changes was not homogeneous along the protein but, rather, gathered within the MHR, including the highly conformational and cysteine-rich "a" determinant. This antigenic loop region of HBs Ag is described as the main target of the humoral response, and any change of its primary sequence may alter its antigenic structure and would render any anti-HBs humoral defense against this region less effective (29).
Two previous studies using a similar approach, but on a more limited number of patients and clones, have characterized the most prevalent residues affected by such changes (19, 39). Thus, positions 145, 144, 129, 126, and 130 were the most likely to be modified. The description of these residue changes, particularly the G145R change, is remarkable because these substitutions also correspond to common mutations described for HBV vaccine escape variants or for patients who have been therapeutically administered monoclonal anti-HBs (14, 36). One should also point out that these very residues are part of MHR loops 3 and 4 and that other residues within these same loops were also affected by additional changes. In our study, the most frequent residue changes in the MHR were observed principally in patients from group I with both markers (71% versus 22% of patients carried more than one mutation), and they mainly included sG145R, sI/T126N/S, sQ129N, and sD144A. Our results are comparable to those observed by Yamamoto et al. and Kohno et al. but included a larger number of patients and clones per patient. It is important that residue changes observed in Ag/anti-HBs patients are not only at these characteristic positions but also throughout loops 2 to 4 of the antigenic loops, but it is difficult to predict the structural and biochemical effects of these amino acid substitutions. Nevertheless, it is likely that residue changes located at position s146, representing an N-glycosylation site, or s147, the location of a disulfide bridge, could alter the immunological properties of the S protein (19, 38). Altogether, several residue changes within the MHR that are specific to patients carrying both HBs Ag and anti-HBs antibodies need to be further characterized immunologically. Although it seems quite convincing that these changes may alter antibody recognition of the S protein, very little is known about the consequences of such modifications on T-cell epitope recognition. As recently described by Bauer et al. for vaccines, accumulations of changes in CHB patients' T-helper-cell epitopes may also alter immune response efficacies and may facilitate virus persistence (4).
Overall, the accumulation of residue changes in and around the "a" determinant may generate HBs variants with alterations of antigenicity and possibly T-cell epitope structure, leading to a subsequent failure to neutralize HBV (5). Surprisingly, our results are quite similar to those observed by Weinberger et al., who described a high genetic variability of the "a" determinant region for HBs Ag-negative chronic virus carriers (38). In the former study, patients were selected on the basis of anti-HBc reactivity, and a comparison was made according to the detection or lack of detection of HBs Ag. Due to different experimental approaches, i.e., direct sequencing in Weinberger's study and clonal analysis in our work, a comparison of the results is difficult. However, our findings underscore an increased protein variability of the MHR in Ag/anti-HBs-positive patients. The amino acid exchanges found by Weinberger et al. in HBs Ag-negative patients were also found in some clones from Ag/anti-HBs-positive patients from our study. In both cases, a very appealing hypothesis is the selection of immune escape mutants under strong host immune pressure.
The analysis of group I patients revealed the presence of a viral mixture consisting of the wild type and antigenic loop variants, suggesting that the acquisition of anti-HBs is associated with an emergence of immune escape variants persisting for several years in association with wild-type strains. Our study suggests that the presence of HBs Ag variants would arise naturally in chronic carriers and is likely a viral strategy to evade the immune system. In light of this hypothesis, it is possible that the detectable HBs Ag (as measured by the diagnostic assay) is a mixture of the wild type and variants, whereas the detectable anti-HBs antibodies are targeted against the wild-type viruses. In order to eliminate bias due to the serological assay used for anti-HBs determination and to confirm the specificity of the circulating anti-HBs antibodies for wild-type HBs Ag, all patients' anti-HBs titers were controlled using a second anti-HBs assay (Monolisa anti-HBs Plus; Bio-Rad, France), with very comparable results (data not shown).
The selection of immune escape mutants during the natural history of the disease could have several deleterious consequences. First, the accumulation of mutated residues within the MHR may lead to an absence of HBs Ag recognition by commercially available diagnostic tests. This issue has been largely reported by several authors (9, 37). Indeed, many assays are unable to detect G145R mutants, and some of them have dramatically decreased sensitivities for the detection of other mutants. The reasons for such failure depend on the type of antibody (monoclonal or polyclonal) and the targeted epitope used in the assays, with some tests being largely affected by mutations within loops 2 to 4 (6, 9, 10, 16, 18).
Second, the accumulation of HBV carrying MHR mutations in CHB patients with relatively high viral loads raises the problem of transmission of such variants (8, 15). As previously mentioned, MHR variants may not be fully neutralized by vaccine-induced antibodies, and we are still lacking data on their possible transmission to vaccinated subjects. In light of our data, chronic carriers with both Ag and anti-HBs could be considered potential reservoirs of immune escape variants. Further epidemiological studies are needed to scientifically assess the potential threat of such chronic carriers in areas of high endemicity where infant vaccination has been implemented (15).
Third, when screening a patient for the presence of vaccine-induced antibodies, it may also be useful to propose at least the detection of HBs Ag since both markers may be present simultaneously. Testing for anti-HBc antibodies may also be a wise strategy.
We also noted that amino acids located at positions 122 and 160, which determine the d/y and w/r subtypes of HBs Ag, were completely conserved (2). Indeed, amino acid substitutions located at these positions result in changes of the HBs Ag subtype (32). Thus, we may observe over time in the same patient a co-occurrence of HBs Ag particles of a certain subtype and anti-HBs with specificity for another subtypic determinant not borne by the circulating particles. Indeed, for patient 3 (P3) belonging to group I, no amino acid substitutions were evidenced in any of the tested clones, despite the presence of anti-HBs. As recently demonstrated, it is possible that in this particular case the detected antibodies targeted a previously present serotype; however, this hypothesis could only be ruled out by a longitudinal study over time on serial samples from the same patient (24). This mechanism could be another way for HBV variants to persist in spite of immune pressure (19, 28).
Although it was not possible to thoroughly assess the natural history of the disease in this study, the coexistence of HBs Ag and anti-HBs does not seem to affect disease evolution. However, a larger study without selection bias linked to ethnicity, duration of infection, age at infection, and genotype would be needed to specifically address this issue. In light of the recently described selection of a secretion-incompetent variant by Kalinina et al., one may wonder if MHR variants are associated with certain forms of disease (17).
The selection process leading to the emergence of HBs variants could therefore be compared to the well-described selection of HBe-negative variants over the course of chronic hepatitis B virus carriage.
REFERENCES
Asahina, Y., N. Enomoto, Y. Ogura, M. Kurosaki, I. Sakuma, N. Izumi, F. Marumo, and C. Sato. 1996. Sequential changes in full-length genomes of hepatitis B virus accompanying acute exacerbation of chronic hepatitis B. J. Hepatol. 25:787-794.
Ashton-Rickardt, P. G., and K. Murray. 1989. Mutations that change the immunological subtype of hepatitis B virus surface antigen and distinguish between antigenic and immunogenic determination. J. Med. Virol. 29:204-214.
Bahn, A., P. Gerner, U. Martine, F. Bortolotti, and S. Wirth. 1997. Detection of different viral strains of hepatitis B virus in chronically infected children after seroconversion from HBsAg to anti-HBs indicating viral persistence. J. Hepatol. 27:973-978.
Bauer, T., K. Weinberger, and W. Jilg. 2002. Variants of two major T cell epitopes within the hepatitis B surface antigen are not recognized by specific T helper cells of vaccinated individuals. Hepatology 35:455-465.
Carman, W. F. 1997. The clinical significance of surface antigen variants of hepatitis B virus. J. Viral Hepat. 4:11-20.
Carman, W. F., J. Korula, L. Wallace, R. MacPhee, L. Mimms, and R. Decker. 1995. Fulminant reactivation of hepatitis B due to envelope protein mutant that escaped detection by monoclonal HBsAg ELISA. Lancet 345:1406-1407.
Carman, W. F., A. R. Zanetti, P. Karayiannis, J. Waters, G. Manzillo, E. Tanzi, A. J. Zuckerman, and H. C. Thomas. 1990. Vaccine-induced escape mutant of hepatitis B virus. Lancet 336:325-329.
Chakravarty, R., M. Neogi, S. Roychowdhury, and C. K. Panda. 2002. Presence of hepatitis B surface antigen mutant G145R DNA in the peripheral blood leukocytes of the family members of an asymptomatic carrier and evidence of its horizontal transmission. Virus Res. 90:133-141.
Coleman, P. F., Y. C. Chen, and I. K. Mushahwar. 1999. Immunoassay detection of hepatitis B surface antigen mutants. J. Med. Virol. 59:19-24.
Cooreman, M. P., G. Leroux-Roels, and W. P. Paulij. 2001. Vaccine- and hepatitis B immune globulin-induced escape mutations of hepatitis B virus surface antigen. J. Biomed. Sci. 8:237-247.
Ghany, M. G., B. Ayola, F. G. Villamil, R. G. Gish, S. Rojter, J. M. Vierling, and A. S. Lok. 1998. Hepatitis B virus S mutants in liver transplant recipients who were reinfected despite hepatitis B immune globulin prophylaxis. Hepatology 27:213-222.
Günther, S., B. C. Li, S. Miska, D. H. Kruger, H. Meisel, and H. Will. 1995. A novel method for efficient amplification of whole hepatitis B virus genomes permits rapid functional analysis and reveals deletion mutants in immunosuppressed patients. J. Virol. 69:5437-5444.
Hou, J., Z. Wang, J. Cheng, Y. Lin, G. K. Lau, J. Sun, F. Zhou, J. Waters, P. Karayiannis, and K. Luo. 2001. Prevalence of naturally occurring surface gene variants of hepatitis B virus in nonimmunized surface antigen-negative Chinese carriers. Hepatology 34:1027-1034.
Hsu, H. Y., M. H. Chang, Y. H. Ni, and H. L. Chen. 2004. Survey of hepatitis B surface variant infection in children 15 years after a nationwide vaccination programme in Taiwan. Gut 53:1499-1503.
Huang, X., D. Lu, G. Ji, Y. Sun, L. Ma, Z. Chen, L. Zhang, J. Huang, L. Yu, and S. Levicnik-Stezinar. 2004. Hepatitis B virus (HBV) vaccine-induced escape mutants of HBV S gene among children from Qidong area, China. Virus Res. 99:63-68.
Jeantet, D., I. Chemin, B. Mandrand, A. Tran, F. Zoulim, P. Merle, C. Trepo, and A. Kay. 2004. Cloning and expression of surface antigens from occult chronic hepatitis B virus infections and their recognition by commercial detection assays. J. Med. Virol. 73:508-515.
Kalinina, T., A. Riu, L. Fischer, T. Santantonio, H. Will, and M. Sterneck. 2003. Selection of a secretion-incompetent mutant in the serum of a patient with severe hepatitis B. Gastroenterology 125:1077-1084.
Kfoury Baz, E. M., J. Zheng, K. Mazuruk, A. Van Le, and D. L. Peterson. 2001. Characterization of a novel hepatitis B virus mutant: demonstration of mutation-induced hepatitis B virus surface antigen group specific "a" determinant conformation change and its application in diagnostic assays. Transfus. Med. 11:355-362.
Kohno, H., T. Inoue, F. Tsuda, H. Okamoto, and Y. Akahane. 1996. Mutations in the envelope gene of hepatitis B virus variants co-occurring with antibody to surface antigen in sera from patients with chronic hepatitis B. J. Gen. Virol. 77:1825-1831.
Lai, C. L., V. Ratziu, M. F. Yuen, and T. Poynard. 2003. Viral hepatitis B. Lancet 362:2089-2094.
Lavanchy, D. 2004. Hepatitis B virus epidemiology, disease burden, treatment, and current and emerging prevention and control measures. J. Viral Hepat. 11:97-107.
Liu, C. J., J. H. Kao, W. Y. Shau, P. J. Chen, M. Y. Lai, and D. S. Chen. 2002. Naturally occurring hepatitis B surface gene variants in chronic hepatitis B virus infection: correlation with viral serotypes and clinical stages of liver disease. J. Med. Virol. 68:50-59.
Lok, A. S. 2000. Hepatitis B infection: pathogenesis and management. J. Hepatol. 32:89-97.
Margeridon, S., A. Lachaux, C. Trepo, F. Zoulim, and A. Kay. 2005. A quasi-monoclonal anti-HBs response can lead to immune escape of ‘wild-type’ hepatitis B virus. J. Gen. Virol. 86:1687-1693.
Mathet, V. L., M. Feld, L. Espinola, D. O. Sanchez, V. Ruiz, O. Mando, G. Carballal, J. F. Quarleri, F. D'Mello, C. R. Howard, J. R. Oubina, F. Clement, P. Dewint, G. Leroux-Roels, S. Louisirirotchanakul, C. Kanoksinsombat, A. Theamboonlert, P. Puthavatana, C. Wasi, Y. Poovorawan, N. Khan, M. Guarnieri, S. H. Ahn, J. Li, Y. Zhou, G. Bang, K. H. Kim, J. R. Wands, S. Tong, H. Singh, R. Aggarwal, R. L. Singh, S. R. Naik, S. Naik, C. Osiowy, W. N. Chen, C. J. Oon, and M. C. Moh. 2003. Hepatitis B virus S gene mutants in a patient with chronic active hepatitis with circulating Anti-HBs antibodies. J. Med. Virol. 69:18-26.
Mesenas, S. J., W. C. Chow, Y. Zhao, G. K. Lim, C. J. Oon, and H. S. Ng. 2002. Wild-type and ‘a’ epitope variants in chronic hepatitis B virus carriers positive for hepatitis B surface antigen and antibody. J. Gastroenterol. Hepatol. 17:148-152.
Mimms, L. 1995. Hepatitis B virus escape mutants: pushing the envelope of chronic hepatitis B virus infection. Hepatology 21:884-887.
Ogura, Y., M. Kurosaki, Y. Asahina, N. Enomoto, F. Marumo, and C. Sato. 1999. Prevalence and significance of naturally occurring mutations in the surface and polymerase genes of hepatitis B virus. J. Infect. Dis. 180:1444-1451.
Oon, C. J., K. L. Tan, T. Harrison, and A. Zuckerman. 1996. Natural history of hepatitis B surface antigen mutants in children. Lancet 348:1524.
Rehermann, B., and M. Nascimbeni. 2005. Immunology of hepatitis B virus and hepatitis C virus infection. Nat. Rev. Immunol. 5:215-229.
Rozanov, M., U. Plikat, C. Chappey, A. Kochergin, and T. Tatusova. 2004. A web-based genotyping resource for viral sequences. Nucleic Acids Res. 32:W654-W659.
Schaefer, S. 2005. Hepatitis B virus: significance of genotypes. J. Viral Hepat. 12:111-124.
Steward, M. W., C. D. Partidos, F. D'Mello, and C. R. Howard. 1993. Specificity of antibodies reactive with hepatitis B surface antigen following immunization with synthetic peptides. Vaccine 11:1405-1414.
Stirk, H. J., J. M. Thornton, and C. R. Howard. 1992. A topological model for hepatitis B surface antigen. Intervirology 33:148-158.
Thompson, J. D., T. J. Gibson, F. Plewniak, F. Jeanmougin, and D. G. Higgins. 1997. The CLUSTAL_X Windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25:4876-4882.
Tiollais, P., C. Pourcel, and A. Dejean. 1985. The hepatitis B virus. Nature 317:489-495.
Torresi, J., L. Earnest-Silveira, G. Civitico, T. E. Walters, S. R. Lewin, J. Fyfe, S. A. Locarnini, M. Manns, C. Trautwein, T. C. Bock, G. Deliyannis, K. Edgtton, H. Zhuang, T. Sozzi, and D. C. Jackson. 2002. The virological and clinical significance of mutations in the overlapping envelope and polymerase genes of hepatitis B virus. J. Clin. Virol. 25:97-106.
Weber, B. 2005. Genetic variability of the S gene of hepatitis B virus: clinical and diagnostic impact. J. Clin. Virol. 32:102-112.
Weinberger, K. M., T. Bauer, S. Bohm, and W. Jilg. 2000. High genetic variability of the group-specific a-determinant of hepatitis B virus surface antigen (HBsAg) and the corresponding fragment of the viral polymerase in chronic virus carriers lacking detectable HBsAg in serum. J. Gen. Virol. 81:1165-1174.
Yamamoto, K., M. Horikita, F. Tsuda, K. Itoh, Y. Akahane, S. Yotsumoto, H. Okamoto, Y. Miyakawa, and M. Mayumi. 1994. Naturally occurring escape mutants of hepatitis B virus with various mutations in the S gene in carriers seropositive for antibody to hepatitis B surface antigen. J. Virol. 68:2671-2676.(Olivier Lada, Yves Benham)