Molecular Analysis of HLA Class II Associations With Hepatitis B Virus Clearance and Vaccine Nonresponsiveness

Transcription

1 Molecular Analysis of HLA Class II Associations With Hepatitis B Virus Clearance and Vaccine Nonresponsiveness Andrew Godkin, Miles Davenport, and Adrian V.S. Hill Clearance of acute hepatitis B virus (HBV) infection is associated with a vigorous CD4 T- cell response focusing on the core protein. HLA class II glycoproteins present viral peptides to CD4 T cells and influence the immune responses. HLA-DRB1*1301/2 have been associated with viral clearance, and HLA-DRB1*0301 is associated with nonresponse to vaccination with envelope proteins. Binding affinities of overlapping peptides covering the core and envelope proteins of HBV were measured to HLA glycoproteins encoded by HLA- DRA1*0101,-DRB1*0101 (HLA-DR1), HLA-DRA1*0101,-DRB1*0301 (HLA-DR3), HLA-DRA1*0101,-DRB1*0701 (HLA-DR7) and HLA-DRA1*0101,-DRB1*1301 (HLA- DR13) molecules and compared with published peptide-specific CD4 T-cell responses. There are more high-affinity ligands (IC 50 < 1 mol/l) derived from the core protein than the surface antigen (P <.04 for HLA-DR1/7/13), but there was no increase in the number or the affinity of ligands for HLA-DR13. Clusters of particular core peptides bound to multiple HLA types, explaining the immunodominance of these regions for T-cell responses. Within the envelope protein, the low-affinity ligands (IC 50 < 10 mol/l) are found mainly in the surface antigen, with a marked paucity of ligands for HLA-DR3 (HLA-DR3 vs. non-dr3; P <.05) consistent with the lower vaccination responses for this HLA type. Of all peptides tested, 8 to 10 bound mainly to one HLA type, allowing a substantially greater breadth of response in heterozygotes. In conclusion, these data offer a mechanistic explanation for the dominant response to the HBV core protein during infection and support the direct involvement of the HLA-DRB1 gene in vaccine nonresponsiveness but not altered susceptibility to viral persistence. (HEPATOLOGY 2005;41: ) Hepatitis B virus (HBV) is a small DNA virus that is an important cause of chronic viral infection, with an estimated 250 to 300 million cases worldwide. Although most acutely infected adults clear the virus, approximately 15% develop persistent infection Abbreviations: HBV, hepatitis B virus; sag, surface antigen; HLA, human leukocyte antigen; PBS, phosphate-buffered saline; HPLC, high-pressure liquid chromatography; IC 50, concentration that inhibits 50%. From the Cellular Immunology and Vaccine Development Group, Nuffield Department of Medicine, John Radcliffe Hospital, Oxford, United Kingdom. Received July 21, 2004; accepted March 22, Supported by the Wellcome Trust. A.V.S.H. is a Wellcome Trust Principal Research Fellow. Address reprint requests to: Andrew Godkin, Department of Integrated Medicine, University Hospital of Wales, Cardiff, CF14 4XW, United Kingdom. GodkinAJ@cardiff.ac.uk; fax: (44) Copyright 2005 by the American Association for the Study of Liver Diseases. Published online in Wiley InterScience ( DOI /hep Potential conflict of interest: A.V.S.H is a cofounder of and consultant to Oxon. Pharmaccines Ltd., a company developing therapeutic vaccines for persistent HBV infection. The study sponsors/funding agency had no role in the design or execution of this study nor in the writing of this report. that may lead to complications such as chronic hepatitis, cirrhosis, and hepatocellular carcinoma. The virus contains a compact, well characterized genome of only 3,200 bases encoding structural proteins, including the viral envelope proteins: S protein, M protein (S protein plus the pre-s1 sequence), and the L protein (S protein plus pre-s1 and pre-s2 sequences) and the nucleocapsid core protein. 1 The envelope proteins are found in the surface antigen (sag) of the virus, which is used as the basis for a recombinant vaccine to generate neutralizing antibodies. 2 The pathogenesis of chronic infection is poorly understood. The CD4 T cell response is probably of critical importance in determining the fate of many infections, including the noncytopathic viruses such as HBV (reviewed in Kalams and Walker 3 ). Successful clearance of HBV has been shown to be associated with a vigorous CD4 T cell response to the nucleocapsid protein. 4 The kinetics of this response parallel the CD8 cytotoxic T cell response, which has been demonstrated to both the envelope and nucleocapsid proteins. 5 In comparison, the 1383

2 1384 GODKIN, DAVENPORT, AND HILL HEPATOLOGY, June 2005 CD4 T-cell response to envelope proteins is markedly reduced; the reason for this is unclear. 6 Human leukocyte antigen (HLA) class II molecules are expressed as cell surface glycoproteins that bind and present short peptide epitopes to CD4 T cells, leading to their activation. Each HLA subtype has a particular binding motif that dictates the particular range of peptides that can physically bind in a groove on the surface of the HLA molecule. 7,8 The HLA genes are found on the short arm of chromosome 6 (6p21.3) in a region of strong linkage disequilibrium containing many other immune response genes (reviewed in Trowsdale 9 ). Many infectious diseases show strong associations with particular HLA alleles (reviewed in Hill 10 ). Although differential binding of pathogenic peptides to allele-specific HLA class I glycoproteins has been proposed as a mechanism of disease protection, 11 the mechanism underlying the association of particular HLA class II alleles with infectious diseases is unknown. Whether such associations are caused by particular HLA class II glycoproteins and their specific binding motifs or by another immune response gene in linkage disequilibrium with that HLA class II allele is unknown. Clearance of HBV has also been associated with heterozygosity of HLA class II type, independent of the specific combination of alleles, 12 suggesting that the ability to present an increased range of epitopes to CD4 T cells may be advantageous, and supporting the importance of a broad robust CD4 T-cell response. Although several studies show an association between HLA-DRB1*1301/2 and clearance of HBV, whether this reflects the peptide-binding characteristics of the HLA-DR13 heterodimer is not known. A similar question relates to the influence of HLA class II type on vaccine response to recombinant sag. The vaccine-induced response, which is CD4 T-cell dependent, leads to the production of protective neutralizing antibodies in more than 90% of the recipients. 2 A combination of both HLA and non- HLA genes may influence the magnitude of immune responsiveness to vaccination. 16 Nonresponse to vaccination in humans has been associated with certain HLA types, particularly homozygosity for HLA- DRB1* We used HBV as a model to study the physical interaction of the HLA system with pathogen-derived peptides and assess possible mechanisms underlying HLA class II associations. The very small size of the HBV genome allows a particularly powerful analysis of class II associations with this pathogen. Predicting which sequences will be processed from a protein and presented by the HLA class II glycoprotein dimer to a T cell has proved difficult. The peptides that are eventually presented from the parent protein must be able to bind to the groove in the HLA molecule. We synthesized overlapping peptides covering the entire core and envelope proteins and measured the binding affinities in vitro to purified HLA-DR1, -DR3, -DR7, and -DR13 molecules. HLA-DR13 was chosen as a type associated with viral clearance and -DR3 as a type associated with vaccine nonresponsiveness, and these were compared to the less or nonassociated DR1 and -DR7 types. This analysis allowed the following questions to be addressed: (1) Why is the CD4 T-cell response to HBV focused mainly on the nucleocapsid protein; (2) Do known peptide-specific T-cell responses correlate with the ability of these peptides to bind to particular HLA glycoproteins; and (3) What mechanisms may underlie the documented associations of particular HLA class II variants with disease susceptibility and vaccine nonresponsiveness? Materials and Methods Preparation of HLA Class II Proteins. Epstein-Barr virus (EBV) transformed B-cell lines obtained from the European Collection of Cell Cultures, Salisbury, UK (homozygous for HLA-DRB1*0101, *0301, *0701, and *1302, respectively) were expanded to cells in roller bottles with RPMI 1640 supplemented by 10% fetal calf serum, L-glutamine, and antibiotics. The cells were then lysed with 3% NP-40 in phosphate-buffered saline (PBS) containing leupeptin 1 g/ml, pepstatin 1 g/ml, and 5 mol/l EDTA. The lysate was spun at 100,000g for 90 minutes. The supernatant was passed over a pre-column of sepharose CL 4B followed by an affinity column with cyanogen bromide activated sepharose beads linked to the monoclonal anti-dr L After the supernatant was run down the column, it was washed extensively before elution of the DR molecules with 0.05 mol/l diethylamine (ph 11.5). The DR molecules were immediately neutralized with 1 mol/l TRIS (ph 6.8) and concentrated by ultrafiltration (Centriprep, Amicon, Beverly, MA). The purity was analyzed by 12% SDS-PAGE. Synthesis of Peptides. Peptides were synthesized by standard F-moc chemistry, using the Advanced ChemTech 396 Multiple Peptide Synthesizer, (Advanced ChemTech Europe, Cambridge, UK), following the manufacturer s instructions. After cleavage of the peptide from the resin, the peptide was precipitated in ice-cold diethyl ether. The precipitate was spun down at 3,000 rpm for 5 minutes at 4 C, and resuspended in the ether. This washing step was repeated 3 times. The pellet was then dried and resuspended in distilled water, or 20% to 50% acetic acid if required. The solution was lyophilized overnight.

3 HEPATOLOGY, Vol. 41, No. 6, 2005 GODKIN, DAVENPORT, AND HILL 1385 Table 1. Sequence of Synthesised Peptides Covering the Nucleocapsid Core Protein From HBV Core Number Peptide Sequence Core Number Peptide Sequence c MDIDPYKEFGATVEL c TNMGLKIRQLLWFHIS c YKEFGATVELLSFLPS c KIRQLLWFHISCLTFG c ATVELLSFLPSDFFP c LWFHISCLTFGRETVLE c SFLPSDFFPSVRDLLDT c SCLTFGRETVLEYLVS c SDFFPSVRDLLDTAS c GRETVLEYLVSFGVW c SVRDLLDTASALYREA c EYLVSFGVWIRTPPAY c LDTASALYREALESPE c SFGVWIRTPPAYRPPN c ALYREALESPEHCSPH c RTPPAYRPPNAPILS c LESPEHCSPHHTALR c AYRPPNAPILSTLPET c EHCSPHHTALRQAILC c NAPILSTLPETTVVRRD c CWGELMTLATWVGNNL c TLPETTVVRRRDRGRS c MTLATWVGNNLQDPAS c TVVRRRDRGRSPRRRT c WVGNNLQDPASRDLVV c RDRGRSPRRRTPSPRR c LQDPASRDLVVNYVNT c SPRRRTPSPRRRRSQS c SRDLVVNYVNTNMGLK c PSPRRRRSQSPRRRR c VNYVNTNMGLKIRQLL c RRSQSPRRRRSQSRESQC The purity was assessed to be 70% purity by reversephase high-performance liquid chromatography (HPLC) using a C-18 column (LiChroCART, Merck, BDH Laboratory Supplies, Poole, UK) run with a two-pump Gilson HPLC system (Anachem, Luton, UK). For further use, the peptide was dissolved in dimethylsulfoxide (DMSO), and the concentration measured using a bicinchoninic acid assay. Peptide Binding Assays. The binding assay was carried out using the principle of competition with a promiscuously binding biotinylated CLIP peptide from the invariant chain (96-114) (I*). The test peptide was dissolved in DMSO, and the initial serial 1:10 dilutions 3 (leading to a 1-1:1,000 range) were made in ph 5 buffer (0.02 mol/l 2-N-morpholinoethanesulfonic acid in 0.1 mol/l NaCl and 0.02% Azide). The peptide was incubated at 37 C with I*( g) and the class II protein (0.15 g). The mixture was neutralized with 1 mol/l TRIS (ph 8.0) and transferred to wells precoated with anti-dr antibody. The mixture was incubated for 1 hour, and the wells washed thoroughly with PBS containing 1% TWEEN (PBS-T), then PBS alone. The plate was developed with Avidin-HRP (Extr- Avidin, Sigma, St Louis, MO)/ biotinylated anti-avidin (Sigma)/Avidin-HRP and developed with 100 L o-phenylene diamine (0.4 mg/ml) in phosphate-citrate buffer. The reaction was terminated with 12.5% sulfuric acid, and absorption was measured at 492 nm. The concentration of unlabeled peptide required to inhibit 50% of the binding (IC 50 ) of I* was calculated in mol/l. All of the binding assays were performed in triplicate and the means calculated. We designated higher-affinity peptides as having an IC 50 1 mol/l; lower-affinity peptides include a numerically greater IC 50, for example, 10 mol/l. Statistical Analysis. The number of binding and nonbinding ligands between viral proteins, and to different HLA types, was compared using chi-square tests. Results Synthesis of Peptides. Seventy-seven peptides were synthesized covering the entire core and envelope proteins. Most of the peptides were 15 amino acids in length, overlapping by 5 amino acids (Tables 1 and 2). Each peptide was analyzed by HPLC: if the purity of the peptide was estimated by HPLC to be less than 70%, it was remade. The hydrophobic regions of the S protein proved harder to synthesizes, and the extremely hydrophobic sequence s9 (IIFLFILLLCLIFLLV) was successfully made commercially (Research Genetics, Huntsville, AL). Binding of HBV Core Peptides to HLA Molecules and Correlation to T-Cell Responses. Table 3 shows the results of binding assays with 32 overlapping peptides covering the entire core protein using a range of HLA molecules. For all 4 HLA molecules, several peptides bind with an extremely high affinity. For instance, the number of peptides binding with an IC 50 1 mol/l to HLA- DR1, -DR3, -DR7, and -DR13 is 6/32 (18.8%), 5/32 (15.6%), 10/32 (31%), and 8/32 (25%), respectively. The relative binding of these peptides is compared between HLA molecules in Fig. 1, where it is apparent that clusters of promiscuous higher-affinity binders are found in 4 regions: amino acids 6-33, 47-76, , and ; this figure also shows a close correlation between these promiscuous binding regions and reported immunogenic peptides derived from the core protein and tested in patients who have cleared HBV. 13,19 Hence, the ability of core peptides to bind to the HLA glycoprotein not only correlates extremely well with immunogenicity, but pro-

4 1386 GODKIN, DAVENPORT, AND HILL HEPATOLOGY, June 2005 Table 2. Sequence of Synthesized Peptides Covering the HBV Envelope Including Pre-S1, Pre-S2, and S Protein ps MGQNLSTSNPLGFFP s MENITSGFLGPLLVL ps TSNPLGFFPDHQLDP s PLLVLQAGFFLLTRI ps FFPDHQLDPAFRANT s LLTRILTIPQSLDSW ps LDPAFRANTANPDWD s SLDSWWTSLNFLGGT ps ANTANPDWDFNPNKD s FLGGTTVCLGQNSQS ps DWDFNPNKDTWPDAN s QNSQSPTSNHSPTSC ps NKDTWPDANKVGAGA s SPTSCPPTCPGYRWM ps DANKVGAGAFGLGFT s GYRWMCLRRFIIFLF ps AGAFGLGFTPPHGGL s IIFLFILLLCLIFLLV ps GFTPPHGGLLGWSPQ s CLIFLLVLLDYQGMLPV ps GGLLGWSPQAQGILQ s GMLPVCPLIPGSSTTT ps SPQAQGILQTLPANP s GSTTTSTGPCKTCTT ps ILQTLPANPPPASTN s KTCTTPAQGNSKFPSC ps ANPPPASTNRQSGRQ s SKFPSCCCTKPTDGNC ps STNRQSGRQPTPLSP s TDGNCTCIPIPSSWA ps GRQPTPLSPPLRNTHPQA s PSSWAFGKFLWEWAS ps MQWNSTTFHQTLQD s WASVRFSWLSLLVPFV ps TFHQTLQDPRVRGLY s WLSLLVPFVQWFVGLSPT ps QDPRVRGLYFPAGGS s GLSPTVWLSAIWMMW ps GLYFPAGGSSSGTVN s IWMMWYWGPSLYSIVS ps GGSSSGTVNPVLTTA s PFIPLLPIFFCLWVYI ps TVNPVLTTASPLSSI ps TTASPLSSIFSRIGD ps SSIFSRIGDPALN miscuous HLA binding offers an explanation for these immunodominant regions. Binding of HBV Envelope Peptides to HLA Molecules. The binding affinities of 21 overlapping peptides covering the S protein (sag) to HLA-DR1, -DR3, -DR7, and -DR13 was measured (Table 4). There is a striking paucity of higher-affinity ligands (IC 50 1 mol/l). The number of lower-affinity ligands (IC mol/l) for HLA-DR1, -DR3, -DR7, and -DR13 were 8/21 (38%), 1/21 (4.7%), 7 (33%), and 6/21 (28.6%), respectively. Peptides s1, s2, s3, s7, s8, s16, and s20 all bind to 2 or more of the HLA-DR1, -DR7, and -DR13 molecules yet show no binding to HLA-DR3. The relative absence of binding peptides to HLA-DR3 (HLA-DR3 vs. non-dr3 P.05) fits well with the observation that homozygosity of HLA-DRA1*0101, B1*0301 is associated with nonresponse to the sag vaccine. 17 An N-terminal (amino acids 20-40) peptide of the S protein is recognized more frequently by the CD4 T-cell response, 20,21 and again this correlates with peptides (S1-3) binding to a broad range of HLA glycoproteins (Table 4). Considerable experimental evidence suggests the ability to mount an antibody response to the sag is linked to the MHC class II region. 22 Certain inbred strains of mice have been shown to be high, intermediate, or low responders, with the level of antibody production being linked to the MHC haplotypes. This nonresponsiveness may be overcome by incorporating the pre-s1/s2 proteins into the vaccine. 23 The marked increase in binding of peptides derived from the pre-s1/2 protein suggests a pos- Table 3. Binding Affinities ( mol/l) of Core Peptides to HLA-DR1, -DR3, -DR7, and -DR13 Molecules HLA Glycoproteins HLA-DR1 HLA-DR3 HLA-DR7 HLA-DR13 c c c < c c c c c c c c c c c c c <0.05 c <0.05 <0.5 <0.05 c <0.5 <0.5 c c c c c c c c c c c c c c NOTE. Higher-affinity ligands ( 1 mol/l) are shown in bold.

5 HEPATOLOGY, Vol. 41, No. 6, 2005 GODKIN, DAVENPORT, AND HILL 1387 sible mechanism for these findings (Table 4). The number of lower-affinity ligands (IC mol/l) for HLA- DR1, -DR3, -DR7, and -DR13 were 7/24 (29%), 5/24 (20.8%), 7/24 (29%), and 8/24 (33%), respectively. All alleles also bound at least one peptide with a higher affinity (IC 50 1 mol/l). Vaccination with the extended envelope proteins (M or L protein) may be more efficacious because of the increased number of peptide ligands for HLA class II glycoproteins, and hence increased generation of T-cell epitopes. Comparison of the Number of Ligands Derived From Core and Envelope Proteins. Compared with the core protein, the CD4 T-cell response is far weaker to the sag (reviewed in Chisari and Ferrari 6 ). Table 5 summarizes the number of high-affinity peptide ligands for each HLA type derived from the proteins and identifies a Table 4. Binding Affinities ( mol/l) of Envelope Peptides to HLA-DR1, -DR3, -DR7, and -DR1302 Molecules HLA Glycoproteins HLA-DR1 HLA-DR3 HLA-DR7 HLA-DR13 ps ps ps ps ps ps ps ps ps ps ps ps ps ps ps ps ps ps ps ps ps ps ps ps Fig. 1. Comparison of the relative binding affinities of the core peptides to the 4 HLA types tested. The x-axis shows the core peptides tested ( 1- is equivalent to core peptide 1-15; 6-, to peptide 6-21, as listed in Table 1); the y-axis shows the relative binding affinity ( 1/IC 50 ); hence, the greater the value, the higher the binding affinity of the peptide to the HLA protein. Beneath are shown peptides that have been reported to be broadly recognized by CD4 T cells in most of the patients who have cleared HBV. (Ferrari et al. 19 ; Diepolder et al. 13 ). s s s s s s s s s s s s s s s s s s s s s NOTE. Higher-affinity ligands ( 1 mol/l) are shown in bold.

6 1388 GODKIN, DAVENPORT, AND HILL HEPATOLOGY, June 2005 Table 5. Summary of the Numbers of High-Affinity Ligands (IC 50 <1 mol/l) Derived From the Core and Envelope Proteins for Each HLA Type Tested No. of Peptide Regions With IC 50 <1 mol/l No. of Peptide Regions Tested HLA-DR1 -DR3 -DR7 -DR13 All-DR pre-s1/ sag Core Total Core vs. sag, P NOTE. The peptides derived from the core and pre-s1/2 protein overlap by 9 or 10 amino acids, in contrast to the sag peptides, which overlap mainly by 5 amino acids. To compensate when making a comparison, each consecutive pair of peptides in the core and pre-s proteins is considered a peptide region (i.e., the 32 core peptides contribute 16 peptide regions, the 24 pre-s peptides make 12 peptide regions), and if there are consecutive high-affinity binders from these tested peptides, this is counted as a single high-affinity region for statistical analysis. mechanism for this finding. For all HLA alleles tested, there are far more high-affinity ligands derived from the core protein compared with the envelope proteins (P.04 for HLA-DR1, -DR7, and -DR13). For the 77 peptides tested, between 8 and 10 peptides for each HLA glycoprotein either bound uniquely or to only a single other HLA class II glycoprotein (Tables 3 and 4). This offers a mechanism for the increased frequency of heterozygotes for HLA class II alleles in individuals who have successfully cleared HBV. 12 Thus, heterozygote advantage may simply reflect the wider range of peptides presentable from this small virus by the larger number of class II molecules in heterozygotes, with the resulting broader CD4 T-cell responses leading to more effective pathogen clearance. 24 Discussion This study takes advantage of the small genome of HBV combined with several observations of the influence of HLA class II type on antiviral immune responses to assess possible underlying mechanisms. We used in vitro binding assays to assess the binding capabilities of HBVderived peptides to a series of HLA glycoproteins in an attempt to address the question: does an HLA association reflect the biology of the particular HLA molecule, or is it reflecting the advantages of a particular haplotype of closely linked genes? For a peptide to act as a T-cell epitope, it has to be capable of being processed from the parent protein, bind to the MHC heterodimer, and be recognized by the T-cell receptor. Peptides that bind with a relatively higher affinity are more likely to act as epitopes and perhaps act as immunodominant epitopes. 25 The data shown in this paper strongly support the importance of ligand affinity in determining which peptides act as epitopes (Fig. 1). A comparison of the binding affinities of peptides derived from the core and envelope proteins shows that the former produces significantly more HLA class II ligands (Table 5), and these ligands have a markedly higher affinity for a range of HLA types. Furthermore, the peptide ligands identified correspond closely with recognized CD4 T-cell peptide epitopes. 13,19 This straightforward observation offers an explanation as to why the anti-hbv CD4 T-cell response focuses on the core protein rather than the envelope protein during an acute infection. 6 In several regions of the core protein, the peptides bind to a variety of HLA types. These regions correspond well to those identified as immunodominant for T-cell responses in HLA-disparate subjects (Fig. 1). The converse also applies in regions in which few or no T-cell epitopes are found that do not contain peptide ligands for the tested HLA types. Three regions in the report by Ferrari et al. 19 core 20-34, 70-89, and (corresponding to peptides c4, c13, and c26) have a marked paucity of epitopes. Apart from peptide c4, which binds moderately well to HLA-DR3 and -DR7, no other ligands are identified out of the possible 12 interactions. The core peptide was the most broadly recognized promiscuous epitope (22 of 23 tested subjects). The single patient who failed to mount a CD4 T-cell response to this epitope was homozygous for HLA-DRB1*0301. This allele encodes the only HLA glycoprotein tested that failed to bind the peptide covering this region (C11). It should be pointed out that the pool of peptides covering the S protein (15mers overlapping by 5 amino acids) was not as extensive as the pools covering the pre-s and core proteins (most 15mers overlapping by 10 amino acids). The main contributing factor to this was the marked difficulty in synthesizing the hydrophobic sequences derived from the S protein, as highlighted in the results. The core 9 amino acid binding region of a peptide fits into a groove on the HLA proteins. The groove contains HLAspecific pockets that show preferences for particular side chains of certain amino acids; these pockets correspond-

7 HEPATOLOGY, Vol. 41, No. 6, 2005 GODKIN, DAVENPORT, AND HILL 1389 ing to positions 1, 4, 6/7, and 9 of the core binding peptide Approximately 30% of the binding energy is derived from the interactions of side chains with these pockets. However, we and others have found that even suboptimal peptides, lacking residues from either the N- terminus of the binding core, i.e., P1, P1 P2, or from the C-terminus P9, P8 P9, etc., will still demonstrate measurable binding Hence, although the pool of S protein peptides was slightly more limited, given the dearth of binding peptides, especially to HLA-DR3, we think it unlikely that significant ligands have been missed. Peptides as short as 7 amino acids have been shown to bind to HLA-DR3. 30 Consider the first two S peptides: s1: 1-15, and s2: A theoretical binding core may exist in the overlapping region, for example, 8-16, 9-17, If this were the case, we would have expected to demonstrate some binding of s1 or s2, which covers the majority of this region, bearing in mind the wide range of concentrations of peptides employed in the peptide binding assay. The observed correlation between the binding affinities of peptides and their recognition as T-cell epitopes offers a rationale for exploring the influence of HLA polymorphisms on disease outcome. Several studies highlight the association of HBV clearance with HLA-DRB1*1301 encoding for HLA-DR Diepolder et al. 13 compared proliferative CD4 T-cell responses to core peptides in HLA-DRB1*1301 positive and negative patients. Although HLA-DRB1*1301 patients showed a moderately more vigorous response to the promiscuous regions core and 61-85, they did not recognize core 1-25, in keeping with our observation of the lack of binding peptides from this latter region (Fig. 1). Furthermore, HLA restriction analysis of the T-cell responses did not favor HLA-DRB1*1301 over other alleles. The peptide binding data do not demonstrate superior binding in either number of ligands or affinities to HLA-DR13 (Tables 3 and 5). Thus, in this case the peptide binding data support the hypothesis that HLA-DRB1*1301 may be linked to other important immune response genes and represent a genetic marker for a protective haplotype. This S protein is the main constituent of recombinant vaccines produced in yeast, which have superseded the earlier plasma-derived vaccines. Approximately 2.5% to 5% of vaccine recipients fail to respond, despite multiple inoculations. 32 Apart from predisposing immunocompromising conditions, there is also a genetic susceptibility to nonresponse. 16 Alper et al. 17 described an association between homozygosity of HLA-DRB1*0301 with nonresponse to vaccination in a prospective study. This association between HLA-DRB1*0301 and nonresponse to vaccine has been corroborated by other studies. 32 The HLA-DR3 protein stands out from the other HLA types for its paucity of binding ligands derived from the S protein (Table 4). The S protein, which contains several intra-membrane domains, is far more hydrophobic than the core protein. HLA-DR3 obviously cannot tolerate these hydrophobic residues so well, reflecting a unique binding preference favoring charged or polar residues in two pockets at positions 4 and 6 along the binding groove of the glycoprotein. 7 This situation contrasts with the HLA class I restricted CD8 T-cell responses, which recognize epitopes derived from the envelope protein. Several epitopes have been described for the HLA-A2 molecule, 5 and this would fit with its particular binding motif, which is known to prefer hydrophobic anchors. 7 HLA-DRB1*0701 also has been associated with nonresponse to vaccines. 32,33 Furthermore, there is also a degree of nonresponsiveness with newer vaccines incorporating the pre-s proteins. This finding was strongest when the HLA-DRB1*0701 was part of the extended haplotype DRB1*0701-DQB1*0202, strongly suggesting that this reduced vaccine responsiveness is linked to other immune response genes and that HLA- DRB1*0701 is a marker for the haplotype. The peptide binding data support this observation. Comparing the binding of sag-derived peptides to the different alleles, 7 of 8 ligands for HLA-DR1 also bind to HLA-DR7 with comparable affinities, arguing against such a physical explanation for the reduced response. The S protein contains T- and B-cell epitopes, and the antibodies generated after vaccination are protective. The immunogenicity of the HBV surface antigen vaccine can be augmented by including the pre-s sequences, and studies in mice using overlapping peptides have revealed the C-terminus of the pre-s2 protein to be particularly important for T-cell responses. 34 The region is recognized by multiple mouse MHC class II types. This same region may be important for human vaccine recipients. The pre-s derived peptides ps17 and ps22 bind well to all of the HLA molecules tested (Table 4), and significantly, these peptides are the highest binders from the pre-s2 peptides for all of these alleles. Indeed, the ps22 peptide is the highest binder for HLA-DR3 from all of the envelope peptides. This offers a possible basis for the improved immunogenicity observed in vaccinees receiving newer HBV vaccines that include the pre-s regions: the pre-s peptides clearly bind with a far higher affinity than the protein S peptides. In summary, this analysis of physical binding characteristics of overlapping peptides derived from the HBV envelope and nucleocapsid core proteins to various HLA class II molecules offers a mechanistic explanation for the dominant CD4 T-cell response to the core proteins dur-

8 1390 GODKIN, DAVENPORT, AND HILL HEPATOLOGY, June 2005 ing infection and allows correlation of physical binding data with in vivo responses. The results support the direct involvement of the HLA-DRB1 gene in vaccine non-responsiveness but not altered susceptibility to viral persistence and could facilitate the design of future improved prophylactic and therapeutic vaccines. References 1. Sherlock S. Clinical features of hepatitis. In: Zuckerman A, Thomas H, eds. Viral Hepatitis. Edinburgh: Churchill Livingstone, 1993: Dienstag J, Werner B, Polk P. Hepatitis B vaccine in health care personnel: safety, immunogenicity, and indicators of efficiency. Ann Intern Med 1984;101: Kalams S, Walker B. The critical need for CD4 help in maintaining effective cytotoxic T lymphocyte responses. J Exp Med 1998;188: Ferrari C, Penna A, Bertoletti A, Valli A, Antoni AD, Giuberti T, et al. Cellular immune response to hepatitis B virus-encoded antigens in acute and chronic hepatitis B virus infection. J Immunol 1990;145: Nayersina R, Fowler P, Missale G, Missale G, Cerny A, Schlicht HJ, et al. HLA A2 restricted cytotoxic T lymphocyte responses to multiple hepatitis B surface antigen epitopes during hepatitis B virus infection. J Immunol 1993;10: Chisari F, Ferrari C. Hepatitis B virus pathogenesis. Annu Rev Immunol 1995;13: Rammensee H-G. Chemistry of peptides associated with MHC class I and class II molecules. Curr Opin Immunol 1995;7: Davenport M, Hill A. Peptides associated with MHC class I and II molecules. In: Browning MA, ed. HLA/MHC: Genes, Molecules and Function. Oxford: Bios Scientific Publications, 1996: Trowsdale J. MHC. Res Immunol 1997;148: Hill A. The immunogenetics of human infectious diseases. Annu Rev Immunol 1998;16: Hill A, Elvin J, Willis A, Aidoo M, Allsopp CE, Gotch FM, et al. Molecular analysis of the association of HLA-B53 and resisitance to severe malaria. Nature 1992;360: Thursz M, Thomas H, Greenwood B, Hill A. Heterozygote advantage for HLA class-ii type in hepatitis B virus infection. Nat Genet 1997;17: Diepolder H, Jung M-C, Keller K, Schraut W, Gerlach JT, Gruner N, et al. A vigorous virus-specific CD4 T cell response may contribute to the association of HLA-DR13 with viral clearance in hepatitis B. Clin Exp Immunol 1998;113: Thurz M, Kwiatkowski D, Allsop C, Greenwood B, Thomas H, Hill A. Association between an MHC allele and clearance of hepatitis B virus in the Gambia. N Engl J Med 1995;332: Hohler T, Gerken G, Notghi A, Lubjuhn R, Taheri H, Protzer U, et al. HLA-DRB1*1301 and *1302 protect against chronic hepatitis B. J Hepatol 1997;26: Hohler T, Reuss E, Evers N, Dietrich E, Rittner C, Freitag CM, et al. Differential genetic determination of immune responsiveness to hepatitis B surface antigen and to hepatitis A virus: a vaccination study in twins. Lancet 2002;360: Alper C, Kruskall M, Marcus-Bagley D, Craven DE, Katz AJ, Brink SJ, et al. Genetic prediction of nonresponse to hepatitis B vaccine. N Engl J Med 1989;321: Robbins P, Evans E, Ding A, Warner N, Brodsky F. Monoclonal antibodies that distinguish between class II antigens (HLA-DP, DQ, and DR) in 14 haplotypes. Hum Immunol 1987;18: Ferrari C, Bertoletti A, Penna A, Cavalli A, Valli A, Missale G, et al. Identification of immunodominant T-cell epitopes of the hepatitis B virus nucleocapsid antigen. J Clin Invest 1991;88: Honorati M, Dolzani P, Mariani E, Piacentini A, Lisignoli G, Ferrari C, et al. Epitope specificity of Th0/Th2 CD4 T-lymphocyteclones induced by vaccination with rhbsag vaccine. Gastroenterology 1997;112: Celis E, Ou D, Otvos L. Recognition of hepatitis B surface antigen by human T lymphocytes: proliferative and cytotoxic responses to a major antigenic determinant defined by synthetic peptides. J Immunol 1988; 140: Milich D. Immune response to hepatitis B virus proteins: relevance of the murine model. Semin Liver Dis 1991;11: Milich D, Thornton G, Neurath A, Kent S, Michel M-LTP, Chisari F. Enhanced immunogenicity of the pre-s region of hepatitis B surface antigen. Science 1985;228: Doherty P, Zinkernagel R. A biological role for the majorhistocompatability antigens. Lancet 1975;1: Sercarz E, Lehmann P, Ametani A, Benichou G, Miller A, Moudgil K. Dominance and crypticity of T cell antigenic determinants. Annu Rev Immunol 1993;11: Ghosh P, Amaya M, Mellins E, Wiley D. The structure of an intermediate in class II MHC maturation: CLIP bound to HLA-DR3. Nature 1995; 378: Smith K, Pyrdol J, Gauthier L, Wiley D, Wucherpfennig K. Crystal structure of HLA-DR2 (DRA*0101, DRB1*1501) complexed with a peptide from human myelin basic protein. J Exp Med 1998;188: Stern L, Brown J, Jardetzky T, Gorga JC, Urban RG, Strominger JL, et al. Crystal structure of the human class II MHC protein HLA-DR1 complexed with an influenza virus peptide. Nature 1994;368: Hammer J, Belunis C, Bolin D, Papadopoulos J, Walsky R, Higelin J, et al. High-affinity binding of short peptides to major histocompatibility complex class II molecules by anchor combinations. Proc Natl Acad Sci U S A 1994;91: Geluk A, Van Meijgaarden KE, Janson AA, Drijfhout JW, Meloen RH, De Vries RR, et al. Functional analysis of DR17(DR3)-restricted mycobacterial T cell epitopes reveals DR17-binding motif and enables the design of allele-specific competitor peptides. J Immunol 1992;149: Godkin A, Friede T, Davenport M, Stevanovic S, Willis A, Jewell D, et al. Use of eluted peptide sequence data to identify the binding characteristics of peptides to the insulin-dependant diabetes susceptibility allele HLA- DQ8 (DQ 3.2). Int Immunol 1997;9: Craven D, Awdeh Z, Kunches L, Yunis EJ, Dienstag JL, Werner BG, et al. Nonresponsiveness to hepatitis B vaccine in health care workers. Results of revaccination and genetic typings. Ann Intern Med 1986;105: McDermott A, Zuckerman J, Sabin C, Marsh S, Madrigal J. Contribution of human leukocyte antigens to antibody response to hepatitis B vaccination. Tissue Antigens 1997;50: Milich D, Hughes J, McLachlan A, Langley K, Thornton G, Jones J. Importance of subtype in the immune response to the pre-s(2) region of the hepatitis B surface antigen. J Immunol 1990;9: