The Amino Acid Residues at Positions 120 to 123 Are Crucial for the Antigenicity of Hepatitis B Surface Antigen

Plasmids.Some plasmids used in this study were described previously (22, 35) and were provided by the authors of those reports (Table 1). The coding region for HBsAg (nucleotide [nt] 130 to 841 according to HBV genome sequence AF100308) was included. All vectors used the cytomegalovirus immediate-early promoter.

Construction of plasmids encoding mtHBsAg with single mutations within the type a determinant by site-directed mutagenesis.A series of expression vectors of mtHBsAg with amino acid substitutions at positions 120 to 123 was constructed based on pXF3H, a mammalian expression vector with the cytomegalovirus immediate-early promoter (kindly provided by Feng Xing-Hua). The coding region of HBsAg was amplified from a cloned adw2 sequence, as described previously (43, 44). The primers used for PCR, SP1 and SP2, are listed in Table 2. The amplified fragment (nt 160 to 841 according to the HBV genome sequence AF100308) was digested with EcoRI and PstI and inserted into predigested pXF3H. This procedure resulted in a vector expressing HBsAg with an amino-terminal hemagglutinin (HA) tag. The amino acid sequence preceding the S protein is MYP YDV PDY ANS PYP YDV PDY AEF. Mutations were introduced into the wild-type HBsAg sequence by PCR-based mutagenesis using sense primers P120TL, C121SL, K122IL, and T123NL and an antisense primer designated T-primer, as listed in Table 2. PCR-based, site-directed mutagenesis was performed. Briefly, the cloned wild-type HBV S gene was used as a template to generate two DNA fragments using the primer pairs SP1/T-primer and either SP2/P120TL, SP2/C121SL, SP2/K122IL, or SP2/T123NL. The corresponding fragments were combined. Subsequently, the overlapping ends of the fragments were annealed, allowing each strand to serve as a primer for the extension of the complementary strand. The resulting fusion product was amplified further by PCR using the outer primers SP1 and SP2. The mutated sequences were cloned into pXF3H and were verified by sequence analysis.

An additional expression vector for the mtHBsAg with K/R122I was generated on the basis of a cloned HBV genome of the subtype adw2, pHBV991-12-1 (GenBank accession number X51970). The cloned DNA was extracted from Escherichia coli and was subjected to PCR amplification. The region encoding the HBsAg (nt 11 to 1027 according to the numbering of the pHBV991-12-1 sequence) of HBV was amplified using KS13 (nt 11 to 36; 5′-CCT TCC ACC AAA CTC TGC AAG ATC CC-3′) and KS14 (nt 1027 to 1005; 5′-GGA GCA GCA AAG CCC AAA AGA CC-3′) and was cloned into the pCR2.1 vector (41). The correctness of the sequence was verified by DNA sequencing analysis of the cloned fragments. Mutations were introduced by PCR-based, site-directed mutagenesis using specific primers. The wild-type and mutated HBV S genes were subcloned into the EcoRI site of the expression vector pcDNA3 and were tested for the expression of HBsAg by transient transfection in HepG2 cells. In addition, the coding sequence for the HBV pre-S2 region was added to the mutated S sequences to generate pS2-HBsAg122I by using a previously cloned sequence (43). This procedure resulted in a plasmid expressing the HB middle surface antigen with the amino acid substitution K122I within the S domain. The correctness of the sequence and the orientation of the insert in the pcDNA3 vector were verified by sequencing.

Amino acid substitution G145R was introduced into the wild-type HBV S gene by the same technique by using the primers MP-145-1 and MP-145-2 (Table 2).

Transient transfection with the expression plasmids.HepG2 and HeLa cells were maintained in appropriate media, supplemented with 10% fetal calf serum (FCS), in 8-well tissue culture chamber slides at 37°C and 5% CO₂ until the cells reached approximately 60 to 80% confluence. The transfection of cells was performed by using Lipofectamine (Invitrogen) according to a protocol described previously (44). To monitor the transfection efficiency, an expression vector for green fluorescent protein (GFP), pEGFP-N2, was cotransfected with expression vectors of HBsAg. One microgram of the expression vectors of HBsAg and 0.5 μg of pEGFP-N2 were mixed for transfection. Culture supernatants of transfected cells were collected after 72 h for the detection of HBsAg. At the same time, the expression of GFP in transfected cells was judged by fluorescence microscopy. Thirty to 40% of transfected cells regularly expressed GFP. Similar results were generated by using HepG2 and HeLa cells. For detection of HBsAg in enzyme-linked immunosorbent assays (ELISAs), results based on HeLa cells were used, since they usually give higher values due to a better transfection efficiency. The results of immunofluorescence (IF) were obtained by using HepG2 cells.

IF staining of transfected cells.Transfected cells were washed twice with phosphate-buffered saline (PBS) and were fixed with 50% methanol at 4°C for 30 min for IF staining. IF staining of transfected cells with polyclonal goat anti-HB antibodies to plasma-derived HBsAg of ad and ay types (Biotrend, Köln, Germany) (1:80 dilution with PBS) and monoclonal mouse anti-HBsAg type a antibodies (Dako, Hamburg, Germany; Roche, Penzberg, Germany; or our own preparations) (1:100 to 1:200 dilution with PBS) as primary antibodies was performed as described previously (44).

For confocal laser microscopy, transfected cells were fixed with 50% methanol, treated with 0.25% Triton X-100 for 5 min, and washed with PBS. After being stained with primary and secondary antibodies, the cells were incubated for 30 min with a staining solution containing 50 μg of propidium iodide per ml, 300 μg of RNase per ml, 1% Triton X-100, and 10 mM EDTA. After three washes, 30% glycerin was added to cells for confocal laser microscopy.

Preparation of cell lysates for detection of cell-associated HBsAg by Western blotting.To detect cell-associated HBsAg in transfected cells, cell lysates were prepared and subjected to sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis and Western blotting. Cells were transfected with expression vectors and were cultured for 48 h. Cells were washed three times with precooled PBS, detached from the plate with a cell scraper, and collected in Eppendorf tubes by centrifugation at 6,000 rpm. The cell pellets then were resuspended in 100 μl of PBS and lysed by three freeze-thaw cycles. The cell lysates were cleared of cell debris by 2 min of centrifugation in an Eppendorf centrifuge at 12,000 rpm and were stored at −20°C. The protein concentrations in cell lysates were determined by a Bradford bicinchoninic acid assay (Bio-Rad). The lysates were mixed with twofold-concentrated protein sample buffer and were heated for 10 min for SDS-polyacrylamide gel electrophoresis. Fifty micrograms of proteins was loaded on a 12% SDS-polyacrylamide gel. After electrophoresis, proteins were transferred from the SDS-polyacrylamide gel onto a nitrocellulose membrane. HBsAg with the HA tag was detected by polyclonal rabbit anti-HA antibodies at a dilution of 1:300 and horseradish peroxidase (HRP)-labeled goat anti-rabbit antibody at a dilution of 1:5,000. Polyclonal rabbit antibody to HA and HRP-conjugated goat antibody to rabbit immunoglobulin G (IgG) were purchased from Santa Cruz Biotechnology. The detected protein bands were visualized by chemiluminescence.

Determining the reactivity of wild-type HBsAg and mtHBsAg with eight commercial HBsAg ELISAs.Seven commercial HBsAg ELISA kits were purchased. All seven kits use monoclonal anti-HBsAg antibodies to coat the plate and use labeled polyclonal anti-HB antibodies for detection. The detection of wild-type HBsAg and mtHBsAg by the commercial HBsAg ELISA kits was performed according to the manufacturers' instructions. Fifty microliters of culture supernatants or cell lysates was used for the assays. The signal-to-noise (S/N) values were calculated as the optical density (OD) of the sample divided by the OD of the negative control, with S/N ≥ 2.1 as the cutoff. The S/N values of each sample were used to express the reactivity in seven commercial ELISA kits. For each wild-type HBsAg or mtHBsAg, the transfection of cells was repeated four times, and expressed HBsAg in culture supernatants and cells were detected by ELISA. The average values of four test runs were calculated and are given as the reactivity of wild-type HBsAg or mtHBsAg. In addition, the reactivity of wild-type HBsAg and mtHBsAg in supernatants was measured by using the ARCHITECT HBsAg assay kit (Abbott).

Determining the reactivity of wild-type HBsAg and mtHBsAg in ELISAs based on polyclonal anti-HBs and newly generated anti-HB MAbs.Fourteen new mouse anti-HB MAbs were generated by immunization of mice using purified HBV Dane particles of the subtype adr, genotype C. Among them, nine anti-HB MAbs were IgG1, and five were IgG2a. These anti-HB MAbs were purified and verified for their reactivity to HBsAg in commercial anti-HB ELISAs.

The anti-HB MAbs were used to coat microtiter plates at a concentration of 1 μg per ml. After incubation overnight at 4°C, plates were blocked with 5% FCS at 37°C for 2 h. Fifty microliters of each sample was added to wells, and the wells were incubated at 37°C for 1 h. The plates were washed five times with PBS. HRP-labeled polyclonal anti-HB antibodies (Ke Hua Biologicals, Shanghai, China) were dispensed into wells and incubated further at 37°C for 45 min. The plates were washed again five times, 1,3,5-trimethylbenzene was added, and then the plates were incubated for 15 min. The color development was measured at 450 nm. The S/N values were calculated as the OD of the sample divided by the OD of the negative control, with S/N ≥ 2.1 as the cutoff.

Construction of the expression vectors for mtHBsAg with amino acid substitutions at positions 120 to 123 of HBsAg.Previously published work suggested that the amino acid residues at positions 120 to 123 of HBsAg may be crucial for the recognition of anti-HB antibodies (1, 19, 22, 41). However, the natural isolates have multiple amino acid substitutions that may contribute synergistically to the change of HBsAg antigenicity. Thus, it is necessary to construct HBsAg with single-amino-acid substitutions for the assessment of the impact of individual amino acid substitutions on the antigenicity of HBsAg. For this purpose, we generated a series of mtHBsAg with the amino acid substitutions P120T, C121S, K122I, and T123N. The choice of the amino acid substitutions was made by analysis of naturally occurring HBsAg variants that have been found in patients so far (8, 15, 20, 39, 41). To assess the expression of HBsAg with amino acid substitutions, an HA tag was included so that the HBsAg could be detected by Western blotting using a specific antibody to the HA tag. It was found that the wild-type HBsAg and mtHBsAg were expressed at comparable levels after transient transfection of HepG2 cells (Fig. 1).

• Open in new tab
• Download powerpoint

FIG. 1.

Detection of wild-type HBsAg and mtHBsAg expressed in transiently transfected cells by Western blotting. Lysates of transfected cells containing 50 μg of proteins were loaded on a 12% SDS-polyacrylamide gel and were subjected to electrophoresis and Western blotting. The antibody to the HA tag was used for detection of expressed HBsAg. To compare the amounts of protein loaded on the SDS gel, β-actin was detected by using a specific antibody. NC, negative control, untransfected cells; WT, wild type.

Reactivity of mtHBsAg in commercial ELISAs.The wild-type HBsAg and mtHBsAg with amino acid substitutions at positions 120 to 123 were expressed by transient expression in HepG2 cells. Wild-type HBsAg and mtHBsAg in culture supernatants and lysates of transfected cells were subjected to eight commercial diagnostic kits. The detection of HBsAg was performed by using an ARCHITECT HBsAg assay kit (Abbott) and detected 11.2, 8.3, 5.7, 0.27, and 1.8 IU/ml of wild-type HBsAg and mtHBsAg P120T, C121S, K122I, and T123N, respectively. The wild-type HBsAg with the HA tag was clearly detected by different ELISAs, indicating that it was properly assembled and secreted into the culture medium (Fig. 2). Different clones based on pJI expressed HBsAg at levels ranging from 7.3 to 12.1 IU/ml. However, the amino acid substitutions strongly affected the detection of HBsAg (Fig. 2). The mtHBsAg with P120T showed a reactivity like that of wild-type HBsAg in the majority of the commercial assays that have been tested so far (Fig. 2). The reactivity level of the mtHBsAg with P120T in one commercial assay (kit C) was very low. The other three amino acid substitutions, C121S, K122I, and T123N, appeared to impair strongly the detection of HBsAg, despite the normal expression level of these mtHBsAg. The mtHBsAg with C121S was detected by only one of seven commercial assays. All the other assays used did not detect the mtHBsAg with C121S, consistent with the previous finding that the loss of the cysteine residue may influence the antigenicity of HBsAg (2, 28, 29, 38). Both amino acid substitutions K122I and T123N abolished the reactivity of mtHBsAg in all ELISAs included in the study. Similarly, mtHBsAg G145 was not or was weakly recognized by these assays.

• Open in new tab
• Download powerpoint

FIG. 2.

Reactivity of expressed wild-type HBsAg and mtHBsAg in seven commercial ELISAs (kits A to G). The wild-type HBsAg and mtHBsAg were expressed by transient transfection. (A) The reactivity of HBsAg in culture supernatants. (B) The reactivity of HBsAg in cell lysates. The reactivity was expressed as the ratio of sample reactivity to that of the negative control (S/N ratio). The negative control was prepared using untransfected HepG2 cells. WT, wild type.

Reactivity of anti-HB MAbs to mtHBsAg with amino acid substitutions at positions 121 to 123 and comparison with mtHBsAg G145R.To further assess the impact of the amino acid substitutions on HBsAg, the reactivity of mtHBsAg with a panel of new anti-HB MAbs was determined. Among 14 newly generated anti-HB MAbs, four MAbs, TYJ1 to TYJ4, showed high affinity to HBsAg, as determined with serial dilutions in anti-HB ELISAs (data not shown). In particular, the anti-HB IgG2a MAb TYJ3 tested highly positive at the concentrations of 1 and 0.1 ng per ml. TYJ5 was determined to be less reactive and was representative of the other nine anti-HB MAbs prepared in this study. These results indicated that the anti-HB MAbs have great differences in their reactivities to HBsAg. The anti-HB MAbs TYJ1 to TYJ5 were used for the detection of HBsAg of different genotypes and serotypes and three different mtHBsAg with G145R. Three different HBsAg serotypes, adw2, adr, and ayr, were detected by the ELISAs with the anti-HB MAbs without a notable difference (Fig. 3A). Notably, mtHBsAg G145R, on the background of three different serotypes, was detected in ELISAs based on MAbs TYJ1 to TYJ4. The MAb TYJ5, representative of the other nine MAbs (TYJ6 to YYJ14), was not able to recognize mtHBsAg G145R (Fig. 3B).

• Open in new tab
• Download powerpoint

FIG. 3.

Reactivity of expressed wild-type HBsAg (wtHBsAg) and mtHBsAg in ELISAs based on the five newly generated anti-HB MAbs TYJ1 to TYJ5. The wild-type HBsAg and mtHBsAg were expressed by transient transfection. Shown are the reactivities of wild-type HBsAg (A) and mtHBsAg G145R (B) of three HBsAg subtypes, adw, adr, and ayw, with the MAbs. The reactivities of wild-type HBsAg and mtHBsAg P120T, mtHBsAg C121S, mtHBsAg K122I, and mtHBsAg T123N in culture supernatants (C) and cell lysates (D) also are shown. The reactivity was expressed as the ratio of the reactivity of the samples to that of the negative control (S/N ratio). The negative control was prepared using untransfected HepG2 cells. WT, wild type.

The ELISAs based on the MAbs TYJ1 to TYJ5 were used to examine mtHBsAg P120T, C121S, K122I, and T123N. The mtHBsAg P120T was at least detected by some MAbs. Three mtHBsAg, C121S, K/R122I, and T123N, were not sufficiently recognized by any of these MAbs (Fig. 3C). Similar results were obtained with lysates of transfected cells (Fig. 3D). Only T123N in cell lysates was recognized by TYJ4 and TYJ5 to some extent. These findings indicated that anti-HB MAbs TYJ1 to TYJ4 are able to bind to mtHBsAg G145R with a sufficient affinity but still failed to recognize C121S, K122I, and T123N mtHBsAg.

IF staining of mtHBsAg with anti-HB MAbs.We examined whether mtHBsAg could be detected in IF with transfected cells, since only one MAb is needed for the assay (Fig. 4). Two anti-HB MAbs, TYJ1 and TYJ5, from this study and other commercially available anti-HBs were used for IF. Overall, the results were consistent with those generated by ELISAs. Figure 4 shows the IF staining of wild-type HBsAg and mtHBsAg with MAbs TYJ1 and TYJ5. Wild-type HBsAg was stained strongly by both MAbs. The mtHBsAg with P120T was positively stained by both anti-HB MAbs, but the intensity was reduced compared to that with the wild type. The mtHBsAg with C121S stained weakly positively by MAb TYJ1 but not by MAb TYJ5. Both MAbs failed to recognize the mtHBsAg with K122I. The mtHBsAg with T123N was not detected by IF staining using the MAb TYJ1. However, IF staining with TYJ5 showed a weak but clear perinuclear distribution of the mtHBsAg with T123N. Combined with the results generated by ELISAs for HBsAg detection, it could be assumed that the amino acid substitution T123N impaired the secretion of HBsAg, leading to an accumulation of mtHBsAg in cellular fractions.

• Open in new tab
• Download powerpoint

FIG. 4.

IF staining of transfected cells expressing wild-type HBsAg and mtHBsAg with the MAbs TYJ1 (A) and TYJ5 (B). The cells were stained with propidium iodide for confocal laser microscopy. NT, negative control, untransfected cells; WT, wild type.

The amino acid substitution K122I appeared to strongly affect the detectability of HBsAg. To verify the obtained results, we constructed an additional expression vector, pS2-HBsAg122I, for the mtHBsAg with the pre-S2 region based on another expression vector, pcDNA3. Again, the mtHBsAg were not detectable in culture supernatants of transfected cells (data not shown). IF staining with different MAbs failed to show the expression of the mtHBsAg with K122I in transfected cells. The presence of the mtHBsAg with K122I in transfected cells was demonstrated only by using a polyclonal anti-HB antibody (Fig. 5).

• Open in new tab
• Download powerpoint

FIG. 5.

IF staining of cells expressing mtHBsAg K122I in HepG2 cells. NC, negative control, untransfected cells. (A) Wild-type HBsAg, stained with polyclonal anti-HB antibodies; (B) wild-type HBsAg stained with MAb; (C) mtHBsAg K122I stained with polyclonal anti-HB antibodies; (D) mtHBsAg K122I stained with MAb.

The naturally occurring mtHBsAg with amino acid substitutions within MHR 2 of HBsAg were not efficiently recognized by commercial ELISAs.A number of naturally occurring mtHBsAg were found in patients in different clinical settings (Table 1). Few of those mtHBsAg harbored amino acid substitutions at positions 120 and 123. Such amino acid substitutions frequently were accompanied by other amino acid substitutions. To verify the results obtained with the artificially constructed expression vectors for the mtHBsAg, a series of naturally occurring mtHBsAg were expressed by transient transfection, and their reactivity with anti-HB MAbs was investigated. The results are summarized in Table 3. The reactivity of a given mtHBsAg in ELISAs based on the anti-HB MAbs was expressed as the percentage of the reactivity of the wild-type HBsAg with the matched serotype. It could be seen that four MAbs, TYJ1 to TYJ4, were able to recognize the mtHBsAg with the amino acid substitution G145R, although the reactivity was reduced to a level of about 30% of that of the wild-type HBsAg control. Other MAbs showed a lower level of reactivity of less than 20% of that of the wild-type HBsAg control. The mtHBsAg HK188, with L98V and Q101R, was detected by all MAbs tested. In contrast, the mtHBsAg harboring amino acid substitutions within amino acids 120 to 123 were less efficiently recognized by MAbs. The mtHBsAg 1056sp, with P120S/S143L, was shown to be reactive with six MAbs with a level of reactivity that was 30% greater than that of the wild-type HBsAg control. BA2.4, with Y100C/P120T, was significantly less reactive. The mtHBsAg BA3.2 and BA 3.4 showed no or very low reactivity to all 14 MAbs, consistent with the previous results. Other mtHBsAg listed were effectively recognized by at least a few anti-HB MAbs, particularly by TYJ1 to TYJ4. Comparable results were obtained by testing mtHBsAg with the ARCHITECT HBsAg assay kit (data not shown). Taking these results together, the amino acid positions 120 to 123 of HBsAg are essential for HBsAg antigenicity, and the amino acid substitutions in this region may strongly reduce the recognition of HBsAg by anti-HB antibodies. Other mutations studied so far may have an impact on the HBsAg antigenicity, but the impact is less than those in the region of aa 120 to 123. As shown here, this is true for many anti-HB antibodies tested so far.