Partial Delipidation Improves the T-Cell Antigenic
http://www.100md.com
病菌学杂志 2006年第7期
Center for Vaccinology, Department of Clinical Biology, Microbiology and Immunology, Ghent University and Hospital, Ghent, Belgium
ABSTRACT
Hepatitis B virus surface antigen (HBsAg) is a complex macromolecular particle composed of glycoproteins and lipids. The latter, representing 25% of the particle mass, are of host origin and determine the solubility, stability, and, indirectly, B-cell immunogenicity of HBsAg. HBsAg is a T-cell-dependent immunogen that does not elicit a detectable humoral immune response in 5% of HBsAg vaccine recipients and in most subjects suffering from chronic hepatitis B. We investigated the influence of the lipid content on the antigenicity of the particle. Lipids were partially removed from HBsAg by treatment with -D-octyl glucoside and density centrifugation. Sham treatment consisted of density centrifugation of HBsAg only. We compared the in vitro proliferative responses of established T-cell lines and nonfractionated peripheral blood mononuclear cells (PBMC) from HBsAg vaccinees and chronic HBV patients when stimulated with partially delipidated HBsAg, untreated HBsAg, or sham-treated HBsAg. In all experiments, delipidated HBsAg turned out to be 10 to 100 times more antigenic than its untreated or sham-treated counterpart. Remarkably, PBMC from vaccine nonresponders or chronic HBV patients displayed a proliferative response towards delipidated HBsAg, whereas native HBsAg never induced a response. A series of control experiments demonstrated that this enhancement of T-cell antigenicity was HBsAg specific and directly linked to lipid extraction. Nonspecific adjuvant effects of any kind could be ruled out. In vivo evaluation in mice demonstrated that delipidated particles lose most of their B-cell antigenicity. However, when native and delipidated particles were mixed, these mixtures induced equal or slightly superior anti-HBs responses to those induced by the same quantity of native HBsAg alone. In conclusion, our data show that partial delipidation of HBsAg strikingly increases the T-cell antigenicity of this unique viral antigen.
INTRODUCTION
Currently licensed hepatitis B vaccines consist of empty, subviral envelope particles that are commonly known as hepatitis B virus surface antigen (HBsAg). Upon intramuscular injection, these HBsAg particles elicit the production of anti-HBs and induce the expansion of HBsAg-specific T lymphocytes (24). When present at sufficient concentrations (arbitrarily defined as 10 IU/liter), these antibodies convey protection against infection with hepatitis B virus (HBV) (22).
More than 20 years of experience with hepatitis B vaccines and experimental immunization with HBsAg of inbred mouse strains (25) have shown that the immune responses of humans and mice to HBsAg are highly variable. In mice, high-, intermediate-, and nonresponder strains have been defined, and these response patterns are governed by genes located in the major histocompatibility complex (H-2) (23). Evidence is accumulating that the human immune response to HBsAg is also dictated by genes located in the major histocompatibility complex. Poor responses are frequently observed for subjects with HLA haplotypes DR3, DR7, DQ2, and/or DP11 and are seen less frequently for subjects expressing DR1, DR5, DP4, DQ3, and/or DQ5 (7, 10, 17, 21, 24).
HBsAg is a complex macromolecular particle composed of proteins, carbohydrates, and lipids. The envelope proteins and carbohydrates, integral parts of glycoproteins, are of viral origin, whereas the lipid moiety, representing approximately 25 to 30% of the particle mass, is of host origin (26, 33). The complete removal of lipids from HBsAg destabilizes the particle and precipitates the hydrophobic protein moiety. Partial delipidation preserves the particle structure and keeps it in solution but induces minor structural changes that have only been examined at the B-cell level. Some investigators found that lipid extraction did not alter (16) or markedly increased (8, 12, 30) B-cell immunogenicity, whereas others demonstrated that detergent treatment of HBsAg clearly reduced its B-cell antigenicity (9, 11). The effects of partial delipidation on the T-cell antigenicity of HBsAg have never been examined. We have studied this issue because we estimated that changing the interaction of proteins and lipids in HBsAg particles might alter their uptake and processing by antigen-presenting cells (APC) and consequently modify their presentation to and recognition by T lymphocytes. The experiments performed to address this issue are discussed here.
MATERIALS AND METHODS
Subjects. The in vitro T-cell recognition of native and partially delipidated HBsAg was studied using peripheral blood mononuclear cells (PBMC) from six poor/nonresponders (NRs) and two good/high responders (HRs) to hepatitis B vaccines. A vaccinee was considered an NR when the anti-HBs titer measured 1 month after the third of three vaccine doses, given at monthly intervals, did not reach or exceed 10 IU/liter. A vaccinee was considered an HR when the anti-HBs titer measured 1 month after the fourth dose (booster dose given on month 12 in a 0-, 1-, 2-, and 12-month scheme) exceeded 1,000 IU/liter. All six NRs were negative for HBsAg, antibodies to hepatitis B virus core Ag (anti-HBc), and HBV DNA. NRs and HRs were subjects who had participated in clinical vaccine evaluation studies that were performed at our center (18, 19).
PBMC were also obtained from three patients suffering from chronic HBV infection (CC1, CC2, and CC3). Table 1 shows important demographic, histological, and serological data for these subjects. For patients CC1 and CC2, PBMC were obtained during interferon treatment. The study protocol was approved by the Ethical Review Board of the University Hospital of Ghent. All participants gave written informed consent.
Assessment of HB serology. Venous blood samples were collected from each subject for serologic and cellular assays. Anti-HBs was measured in serum with the AUSAB RIA test from Abbott Laboratories (North Chicago, IL), and titers were expressed in IU/liters, using the Hollinger equation (14). The AUSRIA II-125 and CORAB RIA (Abbott Laboratories) tests were used for the detection of HBsAg and anti-HBc, respectively. HBV DNA was measured with the Cobas Amplicor HBV Monitor test from Roche Diagnostics (Mannheim, Germany).
Antigens. The antigens used in the in vitro lymphoproliferation tests were (i) recombinant HBsAg (subtype adw) produced in Saccharomyces cerevisiae (lot DVP23; GlaxoSmithKline, Rixensart, Belgium) (27), (ii) recombinant glycoprotein D2 from herpes simplex virus (gD2, expressed in mammalian cells; GlaxoSmithKline), and (iii) tetanus toxoid (TT; obtained from Statens Seruminstitut, WHO, Copenhagen, Denmark).
Removal of lipids. Partial delipidation of HBsAg was performed as described by Gavilanes et al. (9). In brief, HBsAg particles suspended in 10 mM Tris-HCl-50 mM NaCl, pH 7.0, were incubated with the nonionic, nondenaturing detergent -D-octyl glucoside (OG [C14H28O6]; used at 2% [wt/vol]) (Sigma Aldrich) for 2 h at room temperature. HBsAg incubated with OG (or a control sample not treated with detergent) was layered onto a linear cesium chloride gradient (CsCl density, 1.15 to 1.32 g/ml) and centrifuged for 18 h at 145,000 x g (15°C) in a Beckman SW27 rotor. Fractions of 0.5 ml were collected, beginning at the top of the gradient, and the HBsAg content, measured with an AUSRIA II-125 kit, was determined for each fraction (Fig. 1). HBsAg-positive fractions were pooled and dialyzed against phosphate-buffered saline. After dialysis, the delipidated Ag preparation was concentrated at 4°C using an ultrafiltration device (Centricon 100; Amicon) and called 2% OG. The nominal molecular mass cutoff value of the membrane was 100 kDa. A control preparation, called 0% OG, consisted of sham-delipidated particles and was prepared by running HBsAg through all the steps described above except for delipidation with OG. The 2% OG- and 0% OG-treated preparations were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (Fig. 2), using a PhastSystem setup (Pharmacia LKB Biotechnology, Sweden) with a PhastGel high-density gel (20% homogenous polyacrylamide gel). Gels were stained using the silver nitrate staining method described by Heukeshoven and Dernick (13). In these gels, no fractions with molecular weights lower than that of complete HBsAg were found, indicating that lipid removal did not induce degradation of HBsAg.
Protein quantification. The protein concentrations of the delipidated and sham-delipidated HBsAg preparations were determined by the following two methods: (i) the ESL protein assay, based on a biuret-like reaction (Roche Applied Science, Indianapolis, IN), and (ii) a method based on ninhydrin detection after alkaline hydrolysis (2). Serial dilutions of native HBsAg and a standard bovine serum albumin (BSA) preparation were used as calibrators. The final concentrations of the 0% OG-treated and 2% OG-treated HBsAg stocks were calculated as the means of duplicates with the two different methods (standard deviation [SD], <10%). Neither CsCl nor OG interfered with these methods.
Lymphoproliferation assays. The in vitro cellular immune responses of the vaccinees and chronic HBV patients were measured using an HBsAg-specific lymphoproliferation assay (20). In brief, nonfractionated PBMC were suspended in RPMI 1640 medium supplemented with 25 mM HEPES, 50 U/ml penicillin, 50 μg/ml streptomycin, 2 mM L-glutamine (all from Invitrogen Corporation, Carlsbad, California), 5 x 10–5 M 2-mercaptoethanol (Sigma Chemical Co., St. Louis, MO), and 10% heat-inactivated human AB+ serum (complete medium). PBMC (4 x 105/well) from vaccine recipients or chronic HBV patients were cultured for 6 days (37°C in 5% CO2) in 96-well round-bottomed microtiter plates containing HBsAg or delipidated particles. TT and/or gD2 was used as a positive control antigen. Nonstimulated control cultures (blanks) consisted of PBMC that were kept in culture medium without Ag.
For proliferation assays with T-cell lines, 2 x 104 T cells/well were incubated with complete medium in flat-bottomed plates for 4 days with 1 x 105 autologous, irradiated (2,500 rad; 60Co source) PBMC as APC, in the presence or absence of HBsAg.
All proliferation assays were performed in triplicate, and [3H]thymidine (0.5 μCi/well; Amersham International, Buckinghamshire, United Kingdom) was added 18 h before harvesting. The cultures were harvested using an automated harvesting device and assayed for [3H]thymidine incorporation by liquid scintillation counting in an LKB-Wallac 8100 counter (LKB, Bromma, Sweden). Data are expressed as means of triplicate cultures ± SD, cpm (mean cpm of Ag-stimulated cultures – mean cpm of control cultures), or stimulation indexes (SIs), which were calculated by the following formula: SI = mean experimental cpm (with antigen)/mean control cpm (without antigen). SIs were considered positive when they were 2. The SD of triplicates seldom exceeded 15%.
HBsAg dose and APC dose titration experiments. To examine the antigenic (stimulating) qualities of the different HBsAg preparations, increasing quantities of Ag (from 0.0001 to 9 μg/ml) were added to 105 PBMC. To facilitate comparison of the antigenic quality of native HBsAg with that of 2% OG-treated HBsAg, the concentration required to reach 50% of the maximum proliferation was calculated for each Ag for each subject tested. As an alternative approach, increasing numbers of PBMC as APC (from 2.5 x 104 to 10 x 104) were added to 2 x 104 T cells and 3 μg/ml HBsAg. The number of PBMC required to reach 50% of the maximum proliferation was calculated for each Ag for each subject tested.
Mouse experiments. C57BL/6 and SJL mice, purchased from Charles River (Sulzfeld, Germany), and BALB/c mice, obtained from GlaxoSmithKline, were used for immunization experiments. The experimental groups consisted of female mice (six per group) that were between 5 and 6 weeks of age at the initiation of the experiment. The antigens were suspended in complete Freund's adjuvant or adsorbed to Al(OH)3 and injected intramuscularly (50 μl, in the thigh) on day 0. On day 28, a second dose of Ag (booster) was administered. Animals were bled from the orbital plexus on weeks 0, 2, 4, 6, and 8. Anti-HBs in mouse plasma was measured with the AUSAB RIA test as described for anti-HBs measurements in human serum.
Statistics. The Mann-Whitney U test was used to compare the geometric mean titers (GMTs) in the mouse experiments.
RESULTS
Treatment of HBsAg with -D-octyl glucoside enhances immune recognition by HBsAg-specific T-cell lines. To examine the T-cell antigenicity of delipidated HBsAg in comparison to those of native and sham-delipidated HBsAg, we measured the proliferative responses of HBsAg-specific T-cell lines to these Ag preparations. Increasing doses of treated and untreated HBsAg were added to 2 x 104 HBsAg-specific T cells and 105 irradiated autologous PBMC. Figure 3 shows the proliferative responses of four T-cell lines (HBL-2-TW, HBL-19-TW, HBL-1-VA, and HBL-14-DO) induced by increasing Ag doses (native, 2% OG-treated, or 0% OG-treated HBsAg). The 0% OG-treated preparation served as an internal control for concentration measurements. The proliferative responses induced by the native and the 0% OG-treated preparations were almost identical. The 2% OG-treated HBsAg induced proliferative responses at much lower Ag concentrations, and this was observed in all T-cell lines. To allow an easy comparison of these data, we determined the concentration of each Ag that was required to reach 50% of the maximal proliferation of the cell lines. From the point on the y axis that represents 50% of the maximal proliferation, the dose of HBsAg that is required to reach this 50% proliferative response was obtained by interpolation on the x axis. Figure 3 and the quantitative data derived from it (Table 2) clearly show that delipidation of HBsAg by 2% OG enhances its antigenicity towards selected HBsAg-specific T cells by a factor of 10 to 20. Since the increase in antigenicity occurred in the context of four different HLA class II molecules, it cannot depend on HLA restriction or antigenic fine specificity (6).
In an APC dose titration experiment (results not shown), increasing numbers of autologous, irradiated PBMC were added to 2 x 104 HBsAg-specific T cells and a fixed dose (3 μg/ml) of native or 2% OG-treated HBsAg. The results demonstrate that 2% OG-treated HBsAg requires five times fewer PBMC than native HBsAg to reach 50% of the maximal proliferation. Taken together, these data show that delipidation of HBsAg increases the efficacy of its presentation by APC and/or its T-cell antigenicity.
Treatment of HBsAg with -D-octyl glucoside enhances immune recognition by nonfractionated PBMC from HR and NR hepatitis B vaccinees. The increased T-cell antigenicity of delipidated HBsAg was not only demonstrable with T-cell lines but was even more prominent when nonfractionated PBMC from HRs or NRs to the HBsAg vaccine were examined. When increasing doses of native or 2% OG-treated HBsAg were added to nonfractionated PBMC from HRs, substantially smaller amounts of delipidated HBsAg were required to induce a vigorous response (Fig. 4, left panel). Delipidated HBsAg induced marked proliferative responses in PBMC from an NR (Fig. 4, right panel), whereas native HBsAg was unable to induce any proliferative response in the NR, even at an Ag concentration of 12 μg/ml. Table 3 summarizes the results of experiments performed with PBMC from two HRs and six NRs to the HBsAg vaccine. In HRs, the T-cell antigenicity of delipidated HBsAg was 100 to 200 times higher than that of untreated HBsAg. In NRs, delipidated HBsAg was able to induce a proliferative response, whereas native HBsAg was unable to do so. The amount of delipidated HBsAg required to induce a proliferative response in NRs was much higher than that needed to evoke a comparable or even superior response in HRs. For three individuals (GR1, GR2, and NR5), the experiment was performed twice, and comparable results were obtained.
Treatment of HBsAg with -D-octyl glucoside enhances immune recognition by nonfractionated PBMC from chronic HBV carriers. PBMC from patients suffering from chronic HBV infections were stimulated with native and delipidated HBsAg. In contrast to native HBsAg, which never induced lymphoproliferation in such patients, 2% OG-treated HBsAg was able to do so (Table 4). The amount of delipidated HBsAg needed to induce a proliferative response in PBMC from chronic patients was comparable to that needed to induce proliferation in PBMC from vaccine NRs. These limited data reveal the presence of HBsAg-specific T cells in the circulation of chronic HBV patients.
HBsAg specificity of induced immune responses. To demonstrate that the lymphoproliferation observed upon in vitro stimulation of T-cell lines and PBMC with delipidated Ag was truly an HBsAg-specific response and not some undefined polyclonal or artifactual event, we performed the following experiments. (i) Nonfractionated PBMC from five nonvaccinated individuals were stimulated with 2% OG-treated HBsAg. None of these subjects showed a proliferative response upon stimulation (results not shown). (ii) Five TT-specific T-cell lines were restimulated, in the presence of autologous PBMC, with TT, HBsAg (native, delipidated, or sham treated), or combinations of TT and 2% OG-treated HBsAg preparations (Table 5). None of the HBsAg-only preparations induced a proliferative response (SI < 2), whereas TT clearly did. The 2% OG-treated particles, when added to TT, did not enhance the TT-specific proliferative response, suggesting that the delipidated preparations do not trigger the T-cell receptor nonspecifically or contain factors that promote T-cell proliferation. (iii) We examined whether 2% OG-treated HBsAg required APC for its presentation to T cells. For this purpose, two HBsAg-specific T-cell lines were stimulated with native, delipidated, or sham-treated HBsAg in the presence or absence of irradiated PBMC as APC (Table 6). Only in the presence of APC did stimulation occur, demonstrating that delipidated Ag requires APC to induce T-cell proliferation. (iv) To demonstrate that delipidated HBsAg is presented in a class II-restricted fashion, as is native HBsAg, two HBsAg-specific T-cell lines with known HLA restriction patterns (6) were stimulated with native or 2% OG-treated HBsAg in the presence and absence of monoclonal antibodies towards HLA class II molecules. These assays showed that monoclonal antibodies directed towards HLA class II molecules inhibited the proliferative responses of the lines stimulated with 2% OG-treated HBsAg in the same way as they did for native HBsAg (results not shown). (v) To rule out that the observed effects were due to traces of CsCl or OG in the antigen preparations, increasing amounts of CsCl and OG detergent were added to different T-cell lines and autologous PBMC. In none of the tested settings did the proliferation increase after the addition of CsCl or OG. (vi) We examined whether T-cell lines induced with 2% OG-treated HBsAg are really HBsAg specific. Therefore, PBMC were stimulated with delipidated HBsAg on day 0. On day 17, the growing T cells were restimulated with native (for T-cell line R1) or 2% OG-treated (for T-cell line R2) HBsAg. The Ag specificities of these lines were tested 17 days later (day 34). Both lines R1 and R2 recognized native and delipidated HBsAg but did not proliferate upon stimulation with control antigens TT and gD2 (Table 7). This illustrates the Ag specificities of these lines and suggests that delipidated HBsAg induces an HBsAg-specific response. We concluded that the intrinsic antigenicity of HBsAg was increased. This phenomenon was not a nonspecific adjuvant effect elicited by the delipidation procedure.
Evaluation of the in vivo immunogenicity of 2% OG-treated HBsAg using inbred mouse strains. To examine the in vivo immunogenicity of delipidated HBsAg, we immunized high (BALB/c; H-2d)-, intermediate (C57BL/6; H-2b)-, and low (SJL; H-2s)-responder mice (23) with either native or delipidated HBsAg alone or with a mixture of both (Fig. 5). In BALB/c and C57BL/6 mice, delipidated HBsAg alone did not induce meaningful anti-HBs titers, indicating that the B-cell immunogenicity of HBsAg was largely abolished by the detergent treatment. In BALB/c mice, priming and boosting with 2 μg native HBsAg induced the same GMT titers as immunization with a mixture of 1 μg native HBsAg and 1 μg delipidated HBsAg. In C57BL/6 mice, vaccination with mixtures of native and delipidated HBsAg tended to induce higher anti-HBs titers than vaccination with native HBsAg, but these differences were not significant. In SJL mice, priming and boosting with a mixture of 1 μg native HBsAg and 1 μg delipidated HBsAg induced higher GMT titers than immunization with 2 μg native HBsAg. In this case also, the differences were not significant.
DISCUSSION
The absence of an immune response to HBsAg is unquestionably of pathogenic importance for HBV infections (29). After all, the humoral response to HBsAg is the major protective mechanism against reinfection and against primary infection in vaccinated individuals. Furthermore, the appearance of anti-HBs in the serum of an acutely infected subject is the principal parameter indicating recovery. For this humoral response, however, there is a critical requirement for T cells. There is substantial evidence suggesting that the poor/nonresponsiveness to HBsAg observed in a minority of vaccine recipients and in patients chronically infected with HBV could partially be ascribed to a defective cellular response to this Ag. We speculated that the hampered cellular immune responses seen in these individuals might be caused by the poor T-cell immunogenicity of HBsAg, which could reside in its remarkable composition. Indeed, HBsAg is quite unique as a vaccine Ag because of its high lipid content (33, 35). In the present study, we examined whether delipidation of HBsAg particles alters the antigenicity of the Ag. For this purpose, we treated HBsAg with the detergent -D-octyl glucoside as described by Gavilanes et al. (9). The delipidated HBsAg particles were added to lymphoproliferative assays using PBMC from HRs and NRs to the HBsAg vaccine and from patients with chronic HBV infection (CC patients). The proliferative responses induced with delipidated material were compared to those induced with the native particle. The results presented herein clearly demonstrate that delipidation of HBsAg drastically alters its T-cell antigenicity. The delipidated particle induced better responses than its untreated counterpart, and this was true for HRs and NRs as well as chronically infected patients. This is of particular interest since HBsAg is a notoriously poor T-cell immunogen, inducing only a weak T-cell proliferative response with short-lived memory (29, 36). The mechanism responsible for this remarkable change remains unknown. Based on the present knowledge of antigen processing, antigen presentation, and T-cell recognition and based on the proposed topological models of HBsAg (1, 15, 28, 31, 32, 34), we propose the following hypotheses.
Since the HBsAg polypeptide contains four hydrophobic regions with helical organization that possibly span the lipid bilayer, and since these sequences are endowed with T-cell immunogenicity (3, 5), it is conceivable that partial delipidation facilitates the processing and presentation of these important domains. From this perspective, the expression and biological function of lipase activity in antigen-presenting cells deserve further attention. Since HBsAg-specific T-cell lines can be triggered with 10 to 20 times less delipidated HBsAg than untreated antigen, we assume that delipidated HBsAg is more efficiently (or rapidly) processed and/or that more peptide-class II complexes are exported and exposed at the cell surface of the APC. Most clones recognize peptides in the areas that are considered to be located in hydrophobic -helical transmembrane regions of HBsAg (according to a topological model proposed by Stirk et al. and Howard et al. [15, 34]). Since, after treatment with OG, the content of -helical regions decreased from 52 to 35%, and since most changes occur in the hydrophobic parts of the Ag (9), it is conceivable that some T-cell epitopes become more accessible. It is obvious that the protein moieties of the HBsAg particles and their associated T-cell epitopes could be liberated more efficiently after reducing the lipid content of the particles. We conclude that the increased antigenicity of the delipidated particles in the first series of experiments can be explained by the fact that T-cell epitopes became available at a more rapid pace and possibly at a higher density. We invoke the same mechanism to explain the highly increased antigenicity of delipidated HBsAg towards nonfractionated PBMC from HR vaccinees. However, improved processing and presentation can most probably not explain why PBMC from NR vaccinees and CC patients are stimulated by delipidated HBsAg, whereas they do not react upon exposure to standard or elevated concentrations of untreated HBsAg.
We have previously demonstrated that APC from NR vaccinees are able to take up, process, and adequately present HBsAg to T-cell lines derived from HRs (4, 6). We assume that NRs lack most T cells recognizing the dominant epitopes because their T cells equipped with the appropriate T-cell receptor are systematically eliminated or silenced during thymic maturation or by a postthymic event. Following intense stimulation of the immune system by repetitive vaccination (four or many more doses) with HBsAg, T cells recognizing subdominant epitopes and a minority of T cells recognizing dominant epitopes start to expand and, if present in sufficient numbers, support a humoral anti-HBs response. Subdominant-epitope-restricted T cells can only be detected in vitro when delipidated HBsAg is employed in the assay system. We invoke a similar mechanism to explain the behavior of T cells from CC patients. It has been known for years that CC patients have very weak humoral and cellular immune responses towards HBsAg. The use of delipidated HBsAg in lymphoproliferative assays reveals the presence of very low-affinity HBsAg-specific Th cells in CC patients. Their existence was not unsuspected, since T-cell tolerance or a T-cell nonresponse is clonal and heterogeneous and seldom would be expected to be total. T-cell responses to even delipidated HBsAg are approximately 1,000-fold less efficient in CC patients and vaccine NRs than in vaccine responders. Functionally, there is probably little difference between a zero response and a 1,000-fold-lower Th-cell response in terms of clearing a chronic infection or mediating a vaccine response. This is consistent with the absence of viral clearance and lack of anti-HBs production in CC patients and NRs, respectively. Further studies of the phenotypic and functional characteristics of these T cells, as well as of their immunopathological significance, are needed.
Different control experiments ensured that the observed phenomena are not nonspecific adjuvant effects induced by the delipidation procedure but that the intrinsic antigenicity of HBsAg is increased by lipid extraction.
As demonstrated by others, delipidated HBsAg alone has significantly reduced B-cell antigenicity compared to native HBsAg (9). It was described previously that partial removal of lipids induced conformational changes and a reduction of immune recognition by polyclonal antibodies. However, after delipidation of HBsAg with OG, the particle can still be recognized as an HBsAg in an Abbott AUSRIA assay. This suggests that at least a part of the major "a" determinant remains intact after treatment with the detergent. To bypass the reduced B-cell immunogenicity of the delipidated particles, we mixed delipidated and native HBsAg so that the former contributed to the improved T-cell antigenicity and the latter contributed to the B-cell antigenicity of the vaccine. When vaccination experiments were performed with mixtures of native and delipidated particles, similar (BALB/c mice [high responders to HBsAg; H-2d]) or slightly superior (C57BL/6 mice [intermediate responders to HBsAg; H-2b] and SJL mice [low responders to HBsAg; H-2s]) anti-HBs titers were reached compared to vaccinations with native HBsAg alone.
Further analysis of the T-cell response induced in vitro with delipidated HBsAg can shed new light on the immunopathogenesis of chronic hepatitis B virus infections and on the etiology of vaccine nonresponsiveness. Our studies may be helpful in the search for a more immunogenic or even therapeutic hepatitis B vaccine.
ACKNOWLEDGMENTS
This work was supported by the National Fund for Scientific Research Belgium (grants 7.0013.90 and 3.0024.90), the Special Research Fund of the University of Ghent (OZF 01174591), and the Concerted Research Initiative of the University of Ghent (GOA 12050203) and by financial support from GlaxoSmithKline.
We are grateful to R. Rossau (Innogenetics N.V., Ghent, Belgium) and to B. Vandekerckhove (Blood Transfusion Center of Oost-Vlaanderen, Ghent, Belgium) for performing initial HLA typing. The help of J. Vandekerckhove and L. Van Troys (Department of Biochemistry, University of Ghent) with protein concentration measurements is highly appreciated. We also thank A. Elewaut and his staff (Department of Internal Medicine, Gastroenterology, University Hospital, Ghent, Belgium) for collecting blood from patients chronically infected with hepatitis B virus. We thank GlaxoSmithKline for supplying HBsAg, gD2, and the adjuvant mixtures and Eurocetus (The Netherlands) for their kind gift of recombinant IL-2.
REFERENCES
Berting, A., J. Hahnen, M. Kroger, and W. H. Gerlich. 1995. Computer-aided studies on the spatial structure of the small hepatitis B surface protein. Intervirology 38:8-15.
Brewer, J. M., C. W. Roberts, W. H. Stimson, and J. Alexander. 1995. Accurate determination of adjuvant-associated protein or peptide by ninhydrin assay. Vaccine 13:1441-1444.
Celis, E., and R. W. Karr. 1989. Presentation of an immunodominant T-cell epitope of hepatitis B surface antigen by the HLA-DPw4 molecule. J. Virol. 63:747-752.
Desombere, I., T. Cao, Y. Gijbels, and G. Leroux-Roels. 2005. Non-responsiveness to hepatitis B surface antigen vaccines is not caused by defective antigen presentation or a lack of B7 co-stimulation. Clin. Exp. Immunol. 140:126-137.
Desombere, I., Y. Gijbels, A. Verwulgen, and G. Leroux-Roels. 2000. Characterization of the T cell recognition of hepatitis B surface antigen (HBsAg) by good and poor responders to hepatitis B vaccines. Clin. Exp. Immunol. 122:390-399.
Desombere, I., P. Hauser, R. Rossau, J. Paradijs, and G. Leroux-Roels. 1995. Nonresponders to hepatitis B vaccine can present envelope particles to T lymphocytes. J. Immunol. 154:520-529.
Desombere, I., A. Willems, and G. Leroux-Roels. 1998. Response to hepatitis B vaccine: multiple HLA genes are involved. Tissue Antigens 51:593-604.
Diminsky, D., N. Moav, M. Gorecki, and Y. Barenholz. 1999. Physical, chemical and immunological stability of CHO-derived hepatitis B surface antigen (HBsAg) particles. Vaccine 18:3-17.
Gavilanes, F., J. Gomez-Gutierrez, M. Aracil, J. M. Gonzalez-Ros, J. A. Ferragut, E. Guerrero, and D. L. Peterson. 1990. Hepatitis B surface antigen. Role of lipids in maintaining the structural and antigenic properties of protein components. Biochem. J. 265:857-864.
Godkin, A., M. Davenport, and A. V. Hill. 2005. Molecular analysis of HLA class II associations with hepatitis B virus clearance and vaccine nonresponsiveness. Hepatology 41:1383-1390.
Gomez-Gutierrez, J., I. Rodriguez-Crespo, D. L. Peterson, and F. Gavilanes. 1995. Antigenicity of hepatitis B surface antigen proteins reconstituted with phospholipids. Biochim. Biophys. Acta 1233:205-212.
Goudeau, A., F. Barin, P. Coursaget, P. Sizaret, M. Andre, and P. Maupas. 1978. Effect of deoxycholate on ultrastructure and antigenicity of hepatitis B surface antigen (HBsAg), p. 707. In G. N. Vyas, S. N. Cohen, and R. Schmid (ed.), Viral hepatitis. The Franklin Institute Press, Philadelphia, Pa.
Heukeshoven, J., and R. Dernick. 1988. Improved silver staining procedure for fast staining in PhastSystem Development Unit. I. Staining of sodium dodecyl sulfate gels. Electrophoresis 9:28-32.
Hollinger, B. L., E. Adam, D. Heiberg, and L. J. Melnick. 1982. Responses to hepatitis B vaccine in a young adult population. Franklin Institute Press, Philadelphia, Pa.
Howard, C. R., H. J. Stirk, S. E. Brown, and M. W. Steward. 1988. Towards the development of synthetic hepatitis B vaccines, p. 1094-1101. In A. J. Zuckerman (ed.), Viral hepatitis and liver disease. Liss, New York, N.Y.
Kim, C. Y., and D. M. Bissell. 1971. Stability of the lipid and protein of hepatitis-associated (Australia) antigen. J. Infect. Dis. 123:470-476.
Kruskall, M. S., C. A. Alper, Z. Awdeh, E. J. Yunis, and D. Marcus-Bagley. 1992. The immune response to hepatitis B vaccine in humans: inheritance patterns in families. J. Exp. Med. 175:495-502.
Leroux-Roels, G., I. Desombere, L. Cobbaut, M. A. Petit, P. Desmons, P. Hauser, A. Delem, D. De Grave, and A. Safary. 1997. Hepatitis B vaccine containing surface antigen and selected preS1 and preS2 sequences. 2. Immunogenicity in poor responders to hepatitis B vaccines. Vaccine 15:1732-1736.
Leroux-Roels, G., I. Desombere, G. De Tollenaere, M. A. Petit, P. Desmons, P. Hauser, A. Delem, D. De Grave, and A. Safary. 1997. Hepatitis B vaccine containing surface antigen and selected preS1 and preS2 sequences. 1. Safety and immunogenicity in young, healthy adults. Vaccine 15:1724-1731.
Leroux-Roels, G., E. Van Hecke, W. Michielsen, P. Voet, P. Hauser, and J. Petre. 1994. Correlation between in vivo humoral and in vitro cellular immune responses following immunization with hepatitis B surface antigen (HBsAg) vaccines. Vaccine 12:812-818.
McDermott, A. B., J. N. Zuckerman, C. A. Sabin, S. G. Marsh, and J. A. Madrigal. 1997. Contribution of human leukocyte antigens to the antibody response to hepatitis B vaccination. Tissue Antigens 50:8-14.
McMahon, B. J., D. L. Bruden, K. M. Petersen, L. R. Bulkow, A. J. Parkinson, O. Nainan, M. Khristova, C. Zanis, H. Peters, and H. S. Margolis. 2005. Antibody levels and protection after hepatitis B vaccination: results of a 15-year follow-up. Ann. Intern. Med. 142:333-341.
Milich, D. R., and F. V. Chisari. 1982. Genetic regulation of the immune response to hepatitis B surface antigen (HBsAg). I. H-2 restriction of the murine humoral immune response to the a and d determinants of HBsAg. J. Immunol. 129:320-325.
Milich, D. R., and G. G. Leroux-Roels. 2003. Immunogenetics of the response to HBsAg vaccination. Autoimmun. Rev. 2:248-257.
Milich, D. R., G. G. Leroux-Roels, R. E. Louie, and F. V. Chisari. 1984. Genetic regulation of the immune response to hepatitis B surface antigen (HBsAg). IV. Distinct H-2-linked Ir genes control antibody responses to different HBsAg determinants on the same molecule and map to the I-A and I-C subregions. J. Exp. Med. 159:41-56.
Peterson, D. L. 1981. Isolation and characterization of the major protein and glycoprotein of hepatitis B surface antigen. J. Biol. Chem. 256:6975-6983.
Petre, J., F. Van Wijnendaele, B. De Neys, K. Conrath, O. Van Opstal, P. Hauser, T. Rutgers, T. Cabezon, C. Capiau, N. Harford, et al. 1987. Development of a hepatitis B vaccine from transformed yeast cells. Postgrad. Med. J. 63(Suppl. 2):73-81.
Prange, R., and R. E. Streeck. 1995. Novel transmembrane topology of the hepatitis B virus envelope proteins. EMBO J. 14:247-256.
Rehermann, B., and M. Nascimbeni. 2005. Immunology of hepatitis B virus and hepatitis C virus infection. Nat. Rev. Immunol. 5:215-229.
Skelly, J., C. R. Howard, and A. J. Zuckerman. 1981. Hepatitis B polypeptide vaccine preparation in micelle form. Nature 290:51-54.
Sonveaux, N., K. Conrath, C. Capiau, R. Brasseur, E. Goormaghtigh, and J. M. Ruysschaert. 1994. The topology of the S protein in the yeast-derived hepatitis B surface antigen particles. J. Biol. Chem. 269:25637-25645.
Sonveaux, N., J. M. Ruysschaert, and R. Brasseur. 1995. Proposition of a three-dimensional representation of the constitutive protein of the hepatitis B surface antigen particles. J. Protein Chem. 14:477-486.
Sonveaux, N., D. Thines, and J. M. Ruysschaert. 1995. Characterization of the HBsAg particle lipid membrane. Res. Virol. 146:43-51.
Stirk, H. J., J. M. Thornton, and C. R. Howard. 1992. A topological model for hepatitis B surface antigen. Intervirology 33:148-158.
Van der Meeren, P., W. Van Criekinge, J. Vanderdeelen, and L. Baert. 1994. Phospholipid composition of r-DNA hepatitis B surface antigens. Int. J. Pharm. 106:89-92.
West, D. J., and G. B. Calandra. 1996. Vaccine induced immunologic memory for hepatitis B surface antigen: implications for policy on booster vaccination. Vaccine 14:1019-1027.(Isabelle Desombere, Annic)
ABSTRACT
Hepatitis B virus surface antigen (HBsAg) is a complex macromolecular particle composed of glycoproteins and lipids. The latter, representing 25% of the particle mass, are of host origin and determine the solubility, stability, and, indirectly, B-cell immunogenicity of HBsAg. HBsAg is a T-cell-dependent immunogen that does not elicit a detectable humoral immune response in 5% of HBsAg vaccine recipients and in most subjects suffering from chronic hepatitis B. We investigated the influence of the lipid content on the antigenicity of the particle. Lipids were partially removed from HBsAg by treatment with -D-octyl glucoside and density centrifugation. Sham treatment consisted of density centrifugation of HBsAg only. We compared the in vitro proliferative responses of established T-cell lines and nonfractionated peripheral blood mononuclear cells (PBMC) from HBsAg vaccinees and chronic HBV patients when stimulated with partially delipidated HBsAg, untreated HBsAg, or sham-treated HBsAg. In all experiments, delipidated HBsAg turned out to be 10 to 100 times more antigenic than its untreated or sham-treated counterpart. Remarkably, PBMC from vaccine nonresponders or chronic HBV patients displayed a proliferative response towards delipidated HBsAg, whereas native HBsAg never induced a response. A series of control experiments demonstrated that this enhancement of T-cell antigenicity was HBsAg specific and directly linked to lipid extraction. Nonspecific adjuvant effects of any kind could be ruled out. In vivo evaluation in mice demonstrated that delipidated particles lose most of their B-cell antigenicity. However, when native and delipidated particles were mixed, these mixtures induced equal or slightly superior anti-HBs responses to those induced by the same quantity of native HBsAg alone. In conclusion, our data show that partial delipidation of HBsAg strikingly increases the T-cell antigenicity of this unique viral antigen.
INTRODUCTION
Currently licensed hepatitis B vaccines consist of empty, subviral envelope particles that are commonly known as hepatitis B virus surface antigen (HBsAg). Upon intramuscular injection, these HBsAg particles elicit the production of anti-HBs and induce the expansion of HBsAg-specific T lymphocytes (24). When present at sufficient concentrations (arbitrarily defined as 10 IU/liter), these antibodies convey protection against infection with hepatitis B virus (HBV) (22).
More than 20 years of experience with hepatitis B vaccines and experimental immunization with HBsAg of inbred mouse strains (25) have shown that the immune responses of humans and mice to HBsAg are highly variable. In mice, high-, intermediate-, and nonresponder strains have been defined, and these response patterns are governed by genes located in the major histocompatibility complex (H-2) (23). Evidence is accumulating that the human immune response to HBsAg is also dictated by genes located in the major histocompatibility complex. Poor responses are frequently observed for subjects with HLA haplotypes DR3, DR7, DQ2, and/or DP11 and are seen less frequently for subjects expressing DR1, DR5, DP4, DQ3, and/or DQ5 (7, 10, 17, 21, 24).
HBsAg is a complex macromolecular particle composed of proteins, carbohydrates, and lipids. The envelope proteins and carbohydrates, integral parts of glycoproteins, are of viral origin, whereas the lipid moiety, representing approximately 25 to 30% of the particle mass, is of host origin (26, 33). The complete removal of lipids from HBsAg destabilizes the particle and precipitates the hydrophobic protein moiety. Partial delipidation preserves the particle structure and keeps it in solution but induces minor structural changes that have only been examined at the B-cell level. Some investigators found that lipid extraction did not alter (16) or markedly increased (8, 12, 30) B-cell immunogenicity, whereas others demonstrated that detergent treatment of HBsAg clearly reduced its B-cell antigenicity (9, 11). The effects of partial delipidation on the T-cell antigenicity of HBsAg have never been examined. We have studied this issue because we estimated that changing the interaction of proteins and lipids in HBsAg particles might alter their uptake and processing by antigen-presenting cells (APC) and consequently modify their presentation to and recognition by T lymphocytes. The experiments performed to address this issue are discussed here.
MATERIALS AND METHODS
Subjects. The in vitro T-cell recognition of native and partially delipidated HBsAg was studied using peripheral blood mononuclear cells (PBMC) from six poor/nonresponders (NRs) and two good/high responders (HRs) to hepatitis B vaccines. A vaccinee was considered an NR when the anti-HBs titer measured 1 month after the third of three vaccine doses, given at monthly intervals, did not reach or exceed 10 IU/liter. A vaccinee was considered an HR when the anti-HBs titer measured 1 month after the fourth dose (booster dose given on month 12 in a 0-, 1-, 2-, and 12-month scheme) exceeded 1,000 IU/liter. All six NRs were negative for HBsAg, antibodies to hepatitis B virus core Ag (anti-HBc), and HBV DNA. NRs and HRs were subjects who had participated in clinical vaccine evaluation studies that were performed at our center (18, 19).
PBMC were also obtained from three patients suffering from chronic HBV infection (CC1, CC2, and CC3). Table 1 shows important demographic, histological, and serological data for these subjects. For patients CC1 and CC2, PBMC were obtained during interferon treatment. The study protocol was approved by the Ethical Review Board of the University Hospital of Ghent. All participants gave written informed consent.
Assessment of HB serology. Venous blood samples were collected from each subject for serologic and cellular assays. Anti-HBs was measured in serum with the AUSAB RIA test from Abbott Laboratories (North Chicago, IL), and titers were expressed in IU/liters, using the Hollinger equation (14). The AUSRIA II-125 and CORAB RIA (Abbott Laboratories) tests were used for the detection of HBsAg and anti-HBc, respectively. HBV DNA was measured with the Cobas Amplicor HBV Monitor test from Roche Diagnostics (Mannheim, Germany).
Antigens. The antigens used in the in vitro lymphoproliferation tests were (i) recombinant HBsAg (subtype adw) produced in Saccharomyces cerevisiae (lot DVP23; GlaxoSmithKline, Rixensart, Belgium) (27), (ii) recombinant glycoprotein D2 from herpes simplex virus (gD2, expressed in mammalian cells; GlaxoSmithKline), and (iii) tetanus toxoid (TT; obtained from Statens Seruminstitut, WHO, Copenhagen, Denmark).
Removal of lipids. Partial delipidation of HBsAg was performed as described by Gavilanes et al. (9). In brief, HBsAg particles suspended in 10 mM Tris-HCl-50 mM NaCl, pH 7.0, were incubated with the nonionic, nondenaturing detergent -D-octyl glucoside (OG [C14H28O6]; used at 2% [wt/vol]) (Sigma Aldrich) for 2 h at room temperature. HBsAg incubated with OG (or a control sample not treated with detergent) was layered onto a linear cesium chloride gradient (CsCl density, 1.15 to 1.32 g/ml) and centrifuged for 18 h at 145,000 x g (15°C) in a Beckman SW27 rotor. Fractions of 0.5 ml were collected, beginning at the top of the gradient, and the HBsAg content, measured with an AUSRIA II-125 kit, was determined for each fraction (Fig. 1). HBsAg-positive fractions were pooled and dialyzed against phosphate-buffered saline. After dialysis, the delipidated Ag preparation was concentrated at 4°C using an ultrafiltration device (Centricon 100; Amicon) and called 2% OG. The nominal molecular mass cutoff value of the membrane was 100 kDa. A control preparation, called 0% OG, consisted of sham-delipidated particles and was prepared by running HBsAg through all the steps described above except for delipidation with OG. The 2% OG- and 0% OG-treated preparations were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (Fig. 2), using a PhastSystem setup (Pharmacia LKB Biotechnology, Sweden) with a PhastGel high-density gel (20% homogenous polyacrylamide gel). Gels were stained using the silver nitrate staining method described by Heukeshoven and Dernick (13). In these gels, no fractions with molecular weights lower than that of complete HBsAg were found, indicating that lipid removal did not induce degradation of HBsAg.
Protein quantification. The protein concentrations of the delipidated and sham-delipidated HBsAg preparations were determined by the following two methods: (i) the ESL protein assay, based on a biuret-like reaction (Roche Applied Science, Indianapolis, IN), and (ii) a method based on ninhydrin detection after alkaline hydrolysis (2). Serial dilutions of native HBsAg and a standard bovine serum albumin (BSA) preparation were used as calibrators. The final concentrations of the 0% OG-treated and 2% OG-treated HBsAg stocks were calculated as the means of duplicates with the two different methods (standard deviation [SD], <10%). Neither CsCl nor OG interfered with these methods.
Lymphoproliferation assays. The in vitro cellular immune responses of the vaccinees and chronic HBV patients were measured using an HBsAg-specific lymphoproliferation assay (20). In brief, nonfractionated PBMC were suspended in RPMI 1640 medium supplemented with 25 mM HEPES, 50 U/ml penicillin, 50 μg/ml streptomycin, 2 mM L-glutamine (all from Invitrogen Corporation, Carlsbad, California), 5 x 10–5 M 2-mercaptoethanol (Sigma Chemical Co., St. Louis, MO), and 10% heat-inactivated human AB+ serum (complete medium). PBMC (4 x 105/well) from vaccine recipients or chronic HBV patients were cultured for 6 days (37°C in 5% CO2) in 96-well round-bottomed microtiter plates containing HBsAg or delipidated particles. TT and/or gD2 was used as a positive control antigen. Nonstimulated control cultures (blanks) consisted of PBMC that were kept in culture medium without Ag.
For proliferation assays with T-cell lines, 2 x 104 T cells/well were incubated with complete medium in flat-bottomed plates for 4 days with 1 x 105 autologous, irradiated (2,500 rad; 60Co source) PBMC as APC, in the presence or absence of HBsAg.
All proliferation assays were performed in triplicate, and [3H]thymidine (0.5 μCi/well; Amersham International, Buckinghamshire, United Kingdom) was added 18 h before harvesting. The cultures were harvested using an automated harvesting device and assayed for [3H]thymidine incorporation by liquid scintillation counting in an LKB-Wallac 8100 counter (LKB, Bromma, Sweden). Data are expressed as means of triplicate cultures ± SD, cpm (mean cpm of Ag-stimulated cultures – mean cpm of control cultures), or stimulation indexes (SIs), which were calculated by the following formula: SI = mean experimental cpm (with antigen)/mean control cpm (without antigen). SIs were considered positive when they were 2. The SD of triplicates seldom exceeded 15%.
HBsAg dose and APC dose titration experiments. To examine the antigenic (stimulating) qualities of the different HBsAg preparations, increasing quantities of Ag (from 0.0001 to 9 μg/ml) were added to 105 PBMC. To facilitate comparison of the antigenic quality of native HBsAg with that of 2% OG-treated HBsAg, the concentration required to reach 50% of the maximum proliferation was calculated for each Ag for each subject tested. As an alternative approach, increasing numbers of PBMC as APC (from 2.5 x 104 to 10 x 104) were added to 2 x 104 T cells and 3 μg/ml HBsAg. The number of PBMC required to reach 50% of the maximum proliferation was calculated for each Ag for each subject tested.
Mouse experiments. C57BL/6 and SJL mice, purchased from Charles River (Sulzfeld, Germany), and BALB/c mice, obtained from GlaxoSmithKline, were used for immunization experiments. The experimental groups consisted of female mice (six per group) that were between 5 and 6 weeks of age at the initiation of the experiment. The antigens were suspended in complete Freund's adjuvant or adsorbed to Al(OH)3 and injected intramuscularly (50 μl, in the thigh) on day 0. On day 28, a second dose of Ag (booster) was administered. Animals were bled from the orbital plexus on weeks 0, 2, 4, 6, and 8. Anti-HBs in mouse plasma was measured with the AUSAB RIA test as described for anti-HBs measurements in human serum.
Statistics. The Mann-Whitney U test was used to compare the geometric mean titers (GMTs) in the mouse experiments.
RESULTS
Treatment of HBsAg with -D-octyl glucoside enhances immune recognition by HBsAg-specific T-cell lines. To examine the T-cell antigenicity of delipidated HBsAg in comparison to those of native and sham-delipidated HBsAg, we measured the proliferative responses of HBsAg-specific T-cell lines to these Ag preparations. Increasing doses of treated and untreated HBsAg were added to 2 x 104 HBsAg-specific T cells and 105 irradiated autologous PBMC. Figure 3 shows the proliferative responses of four T-cell lines (HBL-2-TW, HBL-19-TW, HBL-1-VA, and HBL-14-DO) induced by increasing Ag doses (native, 2% OG-treated, or 0% OG-treated HBsAg). The 0% OG-treated preparation served as an internal control for concentration measurements. The proliferative responses induced by the native and the 0% OG-treated preparations were almost identical. The 2% OG-treated HBsAg induced proliferative responses at much lower Ag concentrations, and this was observed in all T-cell lines. To allow an easy comparison of these data, we determined the concentration of each Ag that was required to reach 50% of the maximal proliferation of the cell lines. From the point on the y axis that represents 50% of the maximal proliferation, the dose of HBsAg that is required to reach this 50% proliferative response was obtained by interpolation on the x axis. Figure 3 and the quantitative data derived from it (Table 2) clearly show that delipidation of HBsAg by 2% OG enhances its antigenicity towards selected HBsAg-specific T cells by a factor of 10 to 20. Since the increase in antigenicity occurred in the context of four different HLA class II molecules, it cannot depend on HLA restriction or antigenic fine specificity (6).
In an APC dose titration experiment (results not shown), increasing numbers of autologous, irradiated PBMC were added to 2 x 104 HBsAg-specific T cells and a fixed dose (3 μg/ml) of native or 2% OG-treated HBsAg. The results demonstrate that 2% OG-treated HBsAg requires five times fewer PBMC than native HBsAg to reach 50% of the maximal proliferation. Taken together, these data show that delipidation of HBsAg increases the efficacy of its presentation by APC and/or its T-cell antigenicity.
Treatment of HBsAg with -D-octyl glucoside enhances immune recognition by nonfractionated PBMC from HR and NR hepatitis B vaccinees. The increased T-cell antigenicity of delipidated HBsAg was not only demonstrable with T-cell lines but was even more prominent when nonfractionated PBMC from HRs or NRs to the HBsAg vaccine were examined. When increasing doses of native or 2% OG-treated HBsAg were added to nonfractionated PBMC from HRs, substantially smaller amounts of delipidated HBsAg were required to induce a vigorous response (Fig. 4, left panel). Delipidated HBsAg induced marked proliferative responses in PBMC from an NR (Fig. 4, right panel), whereas native HBsAg was unable to induce any proliferative response in the NR, even at an Ag concentration of 12 μg/ml. Table 3 summarizes the results of experiments performed with PBMC from two HRs and six NRs to the HBsAg vaccine. In HRs, the T-cell antigenicity of delipidated HBsAg was 100 to 200 times higher than that of untreated HBsAg. In NRs, delipidated HBsAg was able to induce a proliferative response, whereas native HBsAg was unable to do so. The amount of delipidated HBsAg required to induce a proliferative response in NRs was much higher than that needed to evoke a comparable or even superior response in HRs. For three individuals (GR1, GR2, and NR5), the experiment was performed twice, and comparable results were obtained.
Treatment of HBsAg with -D-octyl glucoside enhances immune recognition by nonfractionated PBMC from chronic HBV carriers. PBMC from patients suffering from chronic HBV infections were stimulated with native and delipidated HBsAg. In contrast to native HBsAg, which never induced lymphoproliferation in such patients, 2% OG-treated HBsAg was able to do so (Table 4). The amount of delipidated HBsAg needed to induce a proliferative response in PBMC from chronic patients was comparable to that needed to induce proliferation in PBMC from vaccine NRs. These limited data reveal the presence of HBsAg-specific T cells in the circulation of chronic HBV patients.
HBsAg specificity of induced immune responses. To demonstrate that the lymphoproliferation observed upon in vitro stimulation of T-cell lines and PBMC with delipidated Ag was truly an HBsAg-specific response and not some undefined polyclonal or artifactual event, we performed the following experiments. (i) Nonfractionated PBMC from five nonvaccinated individuals were stimulated with 2% OG-treated HBsAg. None of these subjects showed a proliferative response upon stimulation (results not shown). (ii) Five TT-specific T-cell lines were restimulated, in the presence of autologous PBMC, with TT, HBsAg (native, delipidated, or sham treated), or combinations of TT and 2% OG-treated HBsAg preparations (Table 5). None of the HBsAg-only preparations induced a proliferative response (SI < 2), whereas TT clearly did. The 2% OG-treated particles, when added to TT, did not enhance the TT-specific proliferative response, suggesting that the delipidated preparations do not trigger the T-cell receptor nonspecifically or contain factors that promote T-cell proliferation. (iii) We examined whether 2% OG-treated HBsAg required APC for its presentation to T cells. For this purpose, two HBsAg-specific T-cell lines were stimulated with native, delipidated, or sham-treated HBsAg in the presence or absence of irradiated PBMC as APC (Table 6). Only in the presence of APC did stimulation occur, demonstrating that delipidated Ag requires APC to induce T-cell proliferation. (iv) To demonstrate that delipidated HBsAg is presented in a class II-restricted fashion, as is native HBsAg, two HBsAg-specific T-cell lines with known HLA restriction patterns (6) were stimulated with native or 2% OG-treated HBsAg in the presence and absence of monoclonal antibodies towards HLA class II molecules. These assays showed that monoclonal antibodies directed towards HLA class II molecules inhibited the proliferative responses of the lines stimulated with 2% OG-treated HBsAg in the same way as they did for native HBsAg (results not shown). (v) To rule out that the observed effects were due to traces of CsCl or OG in the antigen preparations, increasing amounts of CsCl and OG detergent were added to different T-cell lines and autologous PBMC. In none of the tested settings did the proliferation increase after the addition of CsCl or OG. (vi) We examined whether T-cell lines induced with 2% OG-treated HBsAg are really HBsAg specific. Therefore, PBMC were stimulated with delipidated HBsAg on day 0. On day 17, the growing T cells were restimulated with native (for T-cell line R1) or 2% OG-treated (for T-cell line R2) HBsAg. The Ag specificities of these lines were tested 17 days later (day 34). Both lines R1 and R2 recognized native and delipidated HBsAg but did not proliferate upon stimulation with control antigens TT and gD2 (Table 7). This illustrates the Ag specificities of these lines and suggests that delipidated HBsAg induces an HBsAg-specific response. We concluded that the intrinsic antigenicity of HBsAg was increased. This phenomenon was not a nonspecific adjuvant effect elicited by the delipidation procedure.
Evaluation of the in vivo immunogenicity of 2% OG-treated HBsAg using inbred mouse strains. To examine the in vivo immunogenicity of delipidated HBsAg, we immunized high (BALB/c; H-2d)-, intermediate (C57BL/6; H-2b)-, and low (SJL; H-2s)-responder mice (23) with either native or delipidated HBsAg alone or with a mixture of both (Fig. 5). In BALB/c and C57BL/6 mice, delipidated HBsAg alone did not induce meaningful anti-HBs titers, indicating that the B-cell immunogenicity of HBsAg was largely abolished by the detergent treatment. In BALB/c mice, priming and boosting with 2 μg native HBsAg induced the same GMT titers as immunization with a mixture of 1 μg native HBsAg and 1 μg delipidated HBsAg. In C57BL/6 mice, vaccination with mixtures of native and delipidated HBsAg tended to induce higher anti-HBs titers than vaccination with native HBsAg, but these differences were not significant. In SJL mice, priming and boosting with a mixture of 1 μg native HBsAg and 1 μg delipidated HBsAg induced higher GMT titers than immunization with 2 μg native HBsAg. In this case also, the differences were not significant.
DISCUSSION
The absence of an immune response to HBsAg is unquestionably of pathogenic importance for HBV infections (29). After all, the humoral response to HBsAg is the major protective mechanism against reinfection and against primary infection in vaccinated individuals. Furthermore, the appearance of anti-HBs in the serum of an acutely infected subject is the principal parameter indicating recovery. For this humoral response, however, there is a critical requirement for T cells. There is substantial evidence suggesting that the poor/nonresponsiveness to HBsAg observed in a minority of vaccine recipients and in patients chronically infected with HBV could partially be ascribed to a defective cellular response to this Ag. We speculated that the hampered cellular immune responses seen in these individuals might be caused by the poor T-cell immunogenicity of HBsAg, which could reside in its remarkable composition. Indeed, HBsAg is quite unique as a vaccine Ag because of its high lipid content (33, 35). In the present study, we examined whether delipidation of HBsAg particles alters the antigenicity of the Ag. For this purpose, we treated HBsAg with the detergent -D-octyl glucoside as described by Gavilanes et al. (9). The delipidated HBsAg particles were added to lymphoproliferative assays using PBMC from HRs and NRs to the HBsAg vaccine and from patients with chronic HBV infection (CC patients). The proliferative responses induced with delipidated material were compared to those induced with the native particle. The results presented herein clearly demonstrate that delipidation of HBsAg drastically alters its T-cell antigenicity. The delipidated particle induced better responses than its untreated counterpart, and this was true for HRs and NRs as well as chronically infected patients. This is of particular interest since HBsAg is a notoriously poor T-cell immunogen, inducing only a weak T-cell proliferative response with short-lived memory (29, 36). The mechanism responsible for this remarkable change remains unknown. Based on the present knowledge of antigen processing, antigen presentation, and T-cell recognition and based on the proposed topological models of HBsAg (1, 15, 28, 31, 32, 34), we propose the following hypotheses.
Since the HBsAg polypeptide contains four hydrophobic regions with helical organization that possibly span the lipid bilayer, and since these sequences are endowed with T-cell immunogenicity (3, 5), it is conceivable that partial delipidation facilitates the processing and presentation of these important domains. From this perspective, the expression and biological function of lipase activity in antigen-presenting cells deserve further attention. Since HBsAg-specific T-cell lines can be triggered with 10 to 20 times less delipidated HBsAg than untreated antigen, we assume that delipidated HBsAg is more efficiently (or rapidly) processed and/or that more peptide-class II complexes are exported and exposed at the cell surface of the APC. Most clones recognize peptides in the areas that are considered to be located in hydrophobic -helical transmembrane regions of HBsAg (according to a topological model proposed by Stirk et al. and Howard et al. [15, 34]). Since, after treatment with OG, the content of -helical regions decreased from 52 to 35%, and since most changes occur in the hydrophobic parts of the Ag (9), it is conceivable that some T-cell epitopes become more accessible. It is obvious that the protein moieties of the HBsAg particles and their associated T-cell epitopes could be liberated more efficiently after reducing the lipid content of the particles. We conclude that the increased antigenicity of the delipidated particles in the first series of experiments can be explained by the fact that T-cell epitopes became available at a more rapid pace and possibly at a higher density. We invoke the same mechanism to explain the highly increased antigenicity of delipidated HBsAg towards nonfractionated PBMC from HR vaccinees. However, improved processing and presentation can most probably not explain why PBMC from NR vaccinees and CC patients are stimulated by delipidated HBsAg, whereas they do not react upon exposure to standard or elevated concentrations of untreated HBsAg.
We have previously demonstrated that APC from NR vaccinees are able to take up, process, and adequately present HBsAg to T-cell lines derived from HRs (4, 6). We assume that NRs lack most T cells recognizing the dominant epitopes because their T cells equipped with the appropriate T-cell receptor are systematically eliminated or silenced during thymic maturation or by a postthymic event. Following intense stimulation of the immune system by repetitive vaccination (four or many more doses) with HBsAg, T cells recognizing subdominant epitopes and a minority of T cells recognizing dominant epitopes start to expand and, if present in sufficient numbers, support a humoral anti-HBs response. Subdominant-epitope-restricted T cells can only be detected in vitro when delipidated HBsAg is employed in the assay system. We invoke a similar mechanism to explain the behavior of T cells from CC patients. It has been known for years that CC patients have very weak humoral and cellular immune responses towards HBsAg. The use of delipidated HBsAg in lymphoproliferative assays reveals the presence of very low-affinity HBsAg-specific Th cells in CC patients. Their existence was not unsuspected, since T-cell tolerance or a T-cell nonresponse is clonal and heterogeneous and seldom would be expected to be total. T-cell responses to even delipidated HBsAg are approximately 1,000-fold less efficient in CC patients and vaccine NRs than in vaccine responders. Functionally, there is probably little difference between a zero response and a 1,000-fold-lower Th-cell response in terms of clearing a chronic infection or mediating a vaccine response. This is consistent with the absence of viral clearance and lack of anti-HBs production in CC patients and NRs, respectively. Further studies of the phenotypic and functional characteristics of these T cells, as well as of their immunopathological significance, are needed.
Different control experiments ensured that the observed phenomena are not nonspecific adjuvant effects induced by the delipidation procedure but that the intrinsic antigenicity of HBsAg is increased by lipid extraction.
As demonstrated by others, delipidated HBsAg alone has significantly reduced B-cell antigenicity compared to native HBsAg (9). It was described previously that partial removal of lipids induced conformational changes and a reduction of immune recognition by polyclonal antibodies. However, after delipidation of HBsAg with OG, the particle can still be recognized as an HBsAg in an Abbott AUSRIA assay. This suggests that at least a part of the major "a" determinant remains intact after treatment with the detergent. To bypass the reduced B-cell immunogenicity of the delipidated particles, we mixed delipidated and native HBsAg so that the former contributed to the improved T-cell antigenicity and the latter contributed to the B-cell antigenicity of the vaccine. When vaccination experiments were performed with mixtures of native and delipidated particles, similar (BALB/c mice [high responders to HBsAg; H-2d]) or slightly superior (C57BL/6 mice [intermediate responders to HBsAg; H-2b] and SJL mice [low responders to HBsAg; H-2s]) anti-HBs titers were reached compared to vaccinations with native HBsAg alone.
Further analysis of the T-cell response induced in vitro with delipidated HBsAg can shed new light on the immunopathogenesis of chronic hepatitis B virus infections and on the etiology of vaccine nonresponsiveness. Our studies may be helpful in the search for a more immunogenic or even therapeutic hepatitis B vaccine.
ACKNOWLEDGMENTS
This work was supported by the National Fund for Scientific Research Belgium (grants 7.0013.90 and 3.0024.90), the Special Research Fund of the University of Ghent (OZF 01174591), and the Concerted Research Initiative of the University of Ghent (GOA 12050203) and by financial support from GlaxoSmithKline.
We are grateful to R. Rossau (Innogenetics N.V., Ghent, Belgium) and to B. Vandekerckhove (Blood Transfusion Center of Oost-Vlaanderen, Ghent, Belgium) for performing initial HLA typing. The help of J. Vandekerckhove and L. Van Troys (Department of Biochemistry, University of Ghent) with protein concentration measurements is highly appreciated. We also thank A. Elewaut and his staff (Department of Internal Medicine, Gastroenterology, University Hospital, Ghent, Belgium) for collecting blood from patients chronically infected with hepatitis B virus. We thank GlaxoSmithKline for supplying HBsAg, gD2, and the adjuvant mixtures and Eurocetus (The Netherlands) for their kind gift of recombinant IL-2.
REFERENCES
Berting, A., J. Hahnen, M. Kroger, and W. H. Gerlich. 1995. Computer-aided studies on the spatial structure of the small hepatitis B surface protein. Intervirology 38:8-15.
Brewer, J. M., C. W. Roberts, W. H. Stimson, and J. Alexander. 1995. Accurate determination of adjuvant-associated protein or peptide by ninhydrin assay. Vaccine 13:1441-1444.
Celis, E., and R. W. Karr. 1989. Presentation of an immunodominant T-cell epitope of hepatitis B surface antigen by the HLA-DPw4 molecule. J. Virol. 63:747-752.
Desombere, I., T. Cao, Y. Gijbels, and G. Leroux-Roels. 2005. Non-responsiveness to hepatitis B surface antigen vaccines is not caused by defective antigen presentation or a lack of B7 co-stimulation. Clin. Exp. Immunol. 140:126-137.
Desombere, I., Y. Gijbels, A. Verwulgen, and G. Leroux-Roels. 2000. Characterization of the T cell recognition of hepatitis B surface antigen (HBsAg) by good and poor responders to hepatitis B vaccines. Clin. Exp. Immunol. 122:390-399.
Desombere, I., P. Hauser, R. Rossau, J. Paradijs, and G. Leroux-Roels. 1995. Nonresponders to hepatitis B vaccine can present envelope particles to T lymphocytes. J. Immunol. 154:520-529.
Desombere, I., A. Willems, and G. Leroux-Roels. 1998. Response to hepatitis B vaccine: multiple HLA genes are involved. Tissue Antigens 51:593-604.
Diminsky, D., N. Moav, M. Gorecki, and Y. Barenholz. 1999. Physical, chemical and immunological stability of CHO-derived hepatitis B surface antigen (HBsAg) particles. Vaccine 18:3-17.
Gavilanes, F., J. Gomez-Gutierrez, M. Aracil, J. M. Gonzalez-Ros, J. A. Ferragut, E. Guerrero, and D. L. Peterson. 1990. Hepatitis B surface antigen. Role of lipids in maintaining the structural and antigenic properties of protein components. Biochem. J. 265:857-864.
Godkin, A., M. Davenport, and A. V. Hill. 2005. Molecular analysis of HLA class II associations with hepatitis B virus clearance and vaccine nonresponsiveness. Hepatology 41:1383-1390.
Gomez-Gutierrez, J., I. Rodriguez-Crespo, D. L. Peterson, and F. Gavilanes. 1995. Antigenicity of hepatitis B surface antigen proteins reconstituted with phospholipids. Biochim. Biophys. Acta 1233:205-212.
Goudeau, A., F. Barin, P. Coursaget, P. Sizaret, M. Andre, and P. Maupas. 1978. Effect of deoxycholate on ultrastructure and antigenicity of hepatitis B surface antigen (HBsAg), p. 707. In G. N. Vyas, S. N. Cohen, and R. Schmid (ed.), Viral hepatitis. The Franklin Institute Press, Philadelphia, Pa.
Heukeshoven, J., and R. Dernick. 1988. Improved silver staining procedure for fast staining in PhastSystem Development Unit. I. Staining of sodium dodecyl sulfate gels. Electrophoresis 9:28-32.
Hollinger, B. L., E. Adam, D. Heiberg, and L. J. Melnick. 1982. Responses to hepatitis B vaccine in a young adult population. Franklin Institute Press, Philadelphia, Pa.
Howard, C. R., H. J. Stirk, S. E. Brown, and M. W. Steward. 1988. Towards the development of synthetic hepatitis B vaccines, p. 1094-1101. In A. J. Zuckerman (ed.), Viral hepatitis and liver disease. Liss, New York, N.Y.
Kim, C. Y., and D. M. Bissell. 1971. Stability of the lipid and protein of hepatitis-associated (Australia) antigen. J. Infect. Dis. 123:470-476.
Kruskall, M. S., C. A. Alper, Z. Awdeh, E. J. Yunis, and D. Marcus-Bagley. 1992. The immune response to hepatitis B vaccine in humans: inheritance patterns in families. J. Exp. Med. 175:495-502.
Leroux-Roels, G., I. Desombere, L. Cobbaut, M. A. Petit, P. Desmons, P. Hauser, A. Delem, D. De Grave, and A. Safary. 1997. Hepatitis B vaccine containing surface antigen and selected preS1 and preS2 sequences. 2. Immunogenicity in poor responders to hepatitis B vaccines. Vaccine 15:1732-1736.
Leroux-Roels, G., I. Desombere, G. De Tollenaere, M. A. Petit, P. Desmons, P. Hauser, A. Delem, D. De Grave, and A. Safary. 1997. Hepatitis B vaccine containing surface antigen and selected preS1 and preS2 sequences. 1. Safety and immunogenicity in young, healthy adults. Vaccine 15:1724-1731.
Leroux-Roels, G., E. Van Hecke, W. Michielsen, P. Voet, P. Hauser, and J. Petre. 1994. Correlation between in vivo humoral and in vitro cellular immune responses following immunization with hepatitis B surface antigen (HBsAg) vaccines. Vaccine 12:812-818.
McDermott, A. B., J. N. Zuckerman, C. A. Sabin, S. G. Marsh, and J. A. Madrigal. 1997. Contribution of human leukocyte antigens to the antibody response to hepatitis B vaccination. Tissue Antigens 50:8-14.
McMahon, B. J., D. L. Bruden, K. M. Petersen, L. R. Bulkow, A. J. Parkinson, O. Nainan, M. Khristova, C. Zanis, H. Peters, and H. S. Margolis. 2005. Antibody levels and protection after hepatitis B vaccination: results of a 15-year follow-up. Ann. Intern. Med. 142:333-341.
Milich, D. R., and F. V. Chisari. 1982. Genetic regulation of the immune response to hepatitis B surface antigen (HBsAg). I. H-2 restriction of the murine humoral immune response to the a and d determinants of HBsAg. J. Immunol. 129:320-325.
Milich, D. R., and G. G. Leroux-Roels. 2003. Immunogenetics of the response to HBsAg vaccination. Autoimmun. Rev. 2:248-257.
Milich, D. R., G. G. Leroux-Roels, R. E. Louie, and F. V. Chisari. 1984. Genetic regulation of the immune response to hepatitis B surface antigen (HBsAg). IV. Distinct H-2-linked Ir genes control antibody responses to different HBsAg determinants on the same molecule and map to the I-A and I-C subregions. J. Exp. Med. 159:41-56.
Peterson, D. L. 1981. Isolation and characterization of the major protein and glycoprotein of hepatitis B surface antigen. J. Biol. Chem. 256:6975-6983.
Petre, J., F. Van Wijnendaele, B. De Neys, K. Conrath, O. Van Opstal, P. Hauser, T. Rutgers, T. Cabezon, C. Capiau, N. Harford, et al. 1987. Development of a hepatitis B vaccine from transformed yeast cells. Postgrad. Med. J. 63(Suppl. 2):73-81.
Prange, R., and R. E. Streeck. 1995. Novel transmembrane topology of the hepatitis B virus envelope proteins. EMBO J. 14:247-256.
Rehermann, B., and M. Nascimbeni. 2005. Immunology of hepatitis B virus and hepatitis C virus infection. Nat. Rev. Immunol. 5:215-229.
Skelly, J., C. R. Howard, and A. J. Zuckerman. 1981. Hepatitis B polypeptide vaccine preparation in micelle form. Nature 290:51-54.
Sonveaux, N., K. Conrath, C. Capiau, R. Brasseur, E. Goormaghtigh, and J. M. Ruysschaert. 1994. The topology of the S protein in the yeast-derived hepatitis B surface antigen particles. J. Biol. Chem. 269:25637-25645.
Sonveaux, N., J. M. Ruysschaert, and R. Brasseur. 1995. Proposition of a three-dimensional representation of the constitutive protein of the hepatitis B surface antigen particles. J. Protein Chem. 14:477-486.
Sonveaux, N., D. Thines, and J. M. Ruysschaert. 1995. Characterization of the HBsAg particle lipid membrane. Res. Virol. 146:43-51.
Stirk, H. J., J. M. Thornton, and C. R. Howard. 1992. A topological model for hepatitis B surface antigen. Intervirology 33:148-158.
Van der Meeren, P., W. Van Criekinge, J. Vanderdeelen, and L. Baert. 1994. Phospholipid composition of r-DNA hepatitis B surface antigens. Int. J. Pharm. 106:89-92.
West, D. J., and G. B. Calandra. 1996. Vaccine induced immunologic memory for hepatitis B surface antigen: implications for policy on booster vaccination. Vaccine 14:1019-1027.(Isabelle Desombere, Annic)